Madhuvidya

Author(s): S D Laddu, T N Dharmadhikari, Madhvi Kolhatkar, Pratibha Pingle
Publisher: L D Indology Ahmedabad
Catalog link: https://jainqq.org/explore/006525/1

Page #1
MADHU-VIDYA PROF. MADHUKAR ANANT MEHENDALE COLLECTED PAPERS L. D. Series : 125 Editors : S. D. Laddu T. N. Dharmadhikari Madhavi Kolhatkar Pratibha Pingle General Editor: Jitendra B. Shah bhAratIya 88 vidyAmaMki L. D. INSTITUTE OF INDOLOGY AHMEDABAD - 380 009 bAda Jain Education Interational For Private & Personal use only www.jalnelibrary

Page #2
MADHU-VIDYA PROF. MADHUKAR ANANT MEHENDALE COLLECTED PAPERS L. D. Series : 125 General Editor : Jitendra B. Shah Editors : S. D. Laddu T. N. Dharmadhikari Madhavi Kolhatkar Pratibha Pingle L. D. INSTITUTE OF INDOLOGY AHMEDABAD

Page #3
L. D. Series : 125 Madhu-Vidya Editor Dr. Jitendra B. Shah Published by Dr. Jitendra B. Shah Director L. D. Institute of Indology Ahmedabad-380009 (India) First Edition : May, 2001 ISBN 81-85857-07-5 Price : Rs. 560-00 Typesetting Shree Swaminarayan Mudran Mandir 3, Vijay House, Parth Tower, Nava Vadaj, Ahmedabad-380013 (India) Phone : (079) 7432464, 7415750 Shri Kishor Khurjekar Mac Scrip, "Asawari" 10 Aundh, Pune-411 007. Phone : (020) 5883084, 5890925 Printer : Navaprabhat Printery Nr. Old Novelty Cinema, Gheekanta Road, Ahmedabad-380 001. Phone : (079) 5508631, 5509083

Page #4
Preface We were indeed fortunate to have had a personal meeting, some three years ago, with Prof. M. A. Mehendale in the Bhandarkar Oriental Research Institute, Pune. We were of course aware that he is a world-renowned scholar of Indian linguistics, languages, literature, and culture. And even before that meeting, for some years we were acquainted with his erudite contributions. However, on meeting him, besides his very learned bearing, we were also deeply impressed by his humility, courteousness, and no less by his warm and affectionate disposition. Prof. Mehendale's work on the Asokan inscriptions, Avesta, and also on the Mahabharata is too wellknown. His writings are considered authentic and by many as final in his fields of investigation. For long years, his research papers and articles have been published in the journals of high standing both in India and abroad. Reflecting as these writings do serious research and originality, the scholars rate his thought-provoking studies among the valuable contributions to the domain also of Indology in general. Since his articles, paprers etc. are scattered through different published sources, it was decided by Prof. S. D. Laddu and his colleagues at the Bhandarkar and other Oriental Research Institutes in Pune to collect all such writings and publish them in a book form. An editorial board was next instituted. When the collected material was organized and made ready for printing, Dr. Laddu suggested us to publish it as a Volume, a proposal to which we agreed with delight. The format of the book follows the Quail Scripts style. We hope that this Volume will be useful to and welcomed by scholars in the concerned fields as also those who have a general interest in the researches in Indology. Ahmedabad, 11-5-2001 Jitendra B. Shah

Page #5
PRELUDE We have great pleasure in offering to the world of scholars this Volume of Collected Papers of the veteran indologist Professor Madhukar Anant Mehendale, the erstwhile Professor of Sanskrit at the Deccan College Post-Graduate & Research Institute (Pune) and Joint General Editor of the Encyclopaedic Dictionary of Sanskrit on Historical Principles of the same Institute, and currently Editor of the Epilogue of the Mahabharata at the Bhandarkar Oriental Research Insitute (Pune), his contributions in all these positions being exemplary. Prof. Mehendale's reputation as a scholar specialized in the diverse fields of Veda and Avesta, Pali and Prakrit, Historical Linguistics, Inscriptions and Epics has spread far and wide through his sustained and penetrating contributions in these areas. He has also written a few articles in Sanskrit and numerous ones in Marathi showing concern over contemporary linguistic usages. The famous Kalidasan line "na khalu dhimatan kas cid avisayo nama" (Sakuntala IV. 17, line 4) thus easily comes to mind here. Prof. Mehendale has also guided several students working for research degrees of different Universities, who have made mark later in their respective fields. Prof. Mehendale's students and admirers had long cherished the idea of getting his published research papers collected in a single volume and making these (except those written in Marathi and Sanskrit) easily available to interested scholars. A Felicitation Committee for the purpose was formed, with Prof. R. N. Dandekar as President, Profs. A. M. Ghatage and C. G. Kashikar as Vice Presidents, Prof. K. V. Sharina as Advisor, Prof. S. D. Laddu as Secretary, Dr. T. N. Dharmadhikari as Treasurer, and some renowned scholars, mostly local, as its Members. Also, an Editorial Committee was formed comprising Prof. S. D. Laddu, Dr. T. N. Dharmadhikari, Dr. Smt. Pratibha Pingle and Dr. Smt. Madhavi Kolhatkar. An Appeal circulated to a few scholars for financial support in the undertaking met with an enthusiastic response. This, starting from the Editorial Comunittee, was prominently from scholars like : R. N. Dandekar, A. M. Ghatage, C. G. Kashikar, K. V. Sarma, V. N. Misra, M. R. Yardi, A. R. Kulkarni, G. B. Palsule, Saroja Bhate, Shri & Sint. M: G. Dhadphale, A. B. Marathe, Shri & Smt. Vijay Bedekar, Shri & Smt. K. S. Arjunwadkar, M. P. Rege, G. N. Joshi, Ruyintan Peer, H. C. Bhayani, M. A. Dhaky, Jitendra B. Shah, Gerdi Gerschheimer, Ashok Aklujkar, Madhav Deslipande, Uma Chakravarty, A. C. Sarangi, S. S. Bahulkar, W. K. Lele, Yashwantrao Lele,

Page #6
Manisha Pathak, Yashodhara Kar, and from a host others. The Editorial Committee remains sincerely grateful to all these. A ready response in the form of perinission to reproduce Prof. Mehendale's writings was also received from the Publishers and Copyright Managers of Volunes and Journals in which his writings had appeared, and a credit line to this effect to these together will be found given separately under "Acknowledgements". Some of the original papers were in need of type-setting and pre-press work. Also was needed to add the preliminary pages indicating, besides the Preface, Prelude etc., the Life-sketch and Writings of Professor Mehendale. This entire task was very efficiently done at the Mac Script (Aundh, Pune) under the supervision of Shri Kishor Khurjekar. We are very much thankful for his timely services. For the beautiful frontispiece of Professor Mehendale we are indebted to Ms Roberta and Dr. Mrs. Laura Liberale (Torino University, Italy) while on their visit to the Bhandarkar Oriental Research Institute, Pune. With the entire edited material ready, it was our great good fortune to have the veteran scholars from Ahmedabad - Professor M. A. Dhaky (Director Emeritus, Center for Art & Archaeology, A.I.I.S., Gurgaon) and Dr. Jitendra B. Shah (Director of the Sharadaben Chimanbhai Educational Research Centre and of the L. D. Institute of Indology) - offering to undertake the onerous task of photocopying these original offprints of printed papers, and to publish them in one Volume for inclusion in their Research Series. This 'Pattere' (Vidyasauharda) on their part caused great relief to us. In this, we had the free and enthusiastic services of Shri M. D. Bhandare, Director of the American Institute of Indian Studies (West), in securing effective communication with these two. We are extremely grateful to all the three scholars. Finally, the Editors reverentially dedicate the Volume to Professor Mehendale, in the sentiinent : ada ay-guh gaya hufay,' 'tavaiva madhu-kusumam tubhyameva samarpitam.' and pray for his long, healthy and fruitful life. Editorial Committee Pune; Vasanta-panchaini, 29-1-2001

Page #7
ACKNOWLEDGEMENTS The Editors acknowledge with sincere thanks the courtesy shown by the Publishers and Copyright Managers, Indian and foreign, of the Volumes and Journals - such as the Oxford University Press (for the BSOAS), the Sahitya Akademi (New Delhi), the Bharatiya Itihasa Sankalana Samiti, Maharashtra, and others - for granting permission to reproduce the papers of Professor Mehendale collected in the present Volume. Exact titles of these Volumes and Journals, with detailed source references, would be found recorded in the Bibliography (of Papers) that follows. vi

Page #8
Preface... Prelude.. Acknowledgements.. (A) Prof. M. A. Mehendale: A Profile.... Bibliography of Writings of Professor Mehendale.................. I. Books xiv II. Papers ARTICLES TABLE OF CONTENTS (I) Veda (II) (III) (IV) (V) (VI) xv-xxiv Abbreviations: xv List of Papers : xvi-xxiv Life-Sketch & Writings (B) REVIEWS (C) OBITUARY NOTICES Iranian Pali and Prakrit Linguistics Mahabharata Miscellaneous Collected Papers (3-175) (176-213) (214-333) (334-405) (406-530) (531-566) Index to Authors / Editors of Books Reviewed.. vii .iii ..iv-v .............V .xi-xiii ..xiv-xxiv 3-568 569-723 724-735 .736

Page #9

Page #10
LIFE-SKETCH & WRITINGS

Page #11
PROFESSOR MADHUKAR ANANT MEHENDALE : A PROFILE 1. Personalis (i) Present Address : Office Resi. Telephone : Office Resi. : Bhandarkar Oriental Research Institute, Pune-411 004. (India) : B-4, Probodhan Housing Society, Near Income Tax Office, Erandavance, Pune-411 004. (India) : (020) 565 69 32 : (020)543 04 11 : 14-02-1918 : Harsud, Nimar Dist., Madhya Pradesh : 14-12-1941 : Kusum (nee Kusum Kashinath Paralikar) : 1. Col. Pradip 2. Dr. Ashok (ii) Date of Birth Place of Birth (iii) Married Wife Sons 2. Education 1937.B.A.(C1.I) Baroda College, Vadodara (Baroda) 1939.M.A.(cl.I) Wilson College, Mumbai (Bombay) 1943.Ph.D. Deccan College Post-Graduate and Research Institute, Pune (All degress from the University of Bombay.) Scholarships & Prizes : 1939. Bhagwandas Purushottamdas Sanskrit Scholarship, University of Bombay. 1942. V. N. Mandlik Prize, University of Bombay. 1943. Bhagwanlal Indraji Gold Medal & Prize, University of Bombay. 3. Interests Veda & Epics, Pali & Prakrit, Historical Linguistics, Avesta 4. Positions Held 1983 1973-83 Editor, Epilogue of the Mahabharata Bhandarkar Oriental Research Institute, Pune Joint-General Editor, Sanskrit Dictionary Project, Deccan College P.G. & R.I., Pune Professor of Sanskrit (esp. Vedic), Deccan College P. G. & R. I., Pune 1958-78

Page #12
1951-58 Reader, Deccan College P. G. & R. I., Pune 1945-51 Professor, S. B. Garda College, Navasari (Gujarat) 1944-45 Lecturer, Basaveshwar College, Bagalkot (Karnatak) 5. Other Assignments 1952-54 Visiting Lecturer, Gottingen University, Germany 1957-58 Senior Rockefeller Foundation Fellow, Yale University, U.S.A. 6. Special Lectures 1966. Wilson Philological Lectures, University of Bombay 1975. Kaushik Lecture Series (no. 18), S. P- College, Pune 1978. Bhau Vishnu Ashetekar Vedic Lectures, University of Pune 1985. Dr. P. V. Kane Memorial Lecture 1987. Dr. H. R. Divekar Memorial Lecture 1988. Padmavati Pratishthan Lecture Series 1990. Prof. G. K. Bhat Memorial Lecture 1991. Prof. A. G. Mangrulkar Memorial Lectures 1994. Dr. R. S. Walimbe Memorial Lectures 7. Researches Guided (i) For Ph.D., Poona University : 1961. "Linguistic Analysis of Koli - A Dialect of Marathi" (-Smt. Suhasini S. Laddu) 1962. "Indo-Aryan Elements in Cankam Literature" (-Shri S. Vaidyanathan) 1963. "Epic Syntax" (-Smt. K. Meenakshi) 1964. "Language of Yaska's Nirukta" (-Shri Mantrini Prasad) 1967. "Evolution of the Sanskrit Language from Panini to Patanjali (with reference to the Krt of Primary Formations)" (-Shri S. D. Laddu) 1971. "The Maitrayani Samhita - its Ritual and Language" (-Shri M. N. Bopardikar) 1973. "Agastya Legends - A Historical and Cultural Study" (-Shri M. N. Bopardikar) 1974. "Comparative Study of Three Gujarati Dialects : Surti, Charotari and Kathiawadi" (-Smt. Madhu K. Sampat) 1975. "Fables, Parables and Moral Tales in the Mahabharata" (-Shri Palat Raut)

Page #13
(ii) 1978. "The Concept of Vac in the Vedic Literature" (-Smt. Pratibha G. Deshmukh, now Pratibha M. Pingle) 1980. "A Descriptive Study of Tarai Nepali" (-Shri D. c. Gautam) For Ph. D., Ravishankar University, Raipur, M. P. : 1973. "Panini se Patanjali tak Samskrt ka Vikas Taddhita Pratyayom ke Adhar par" (In Hindi) ("Development of the Sanskrit Language from Panini to Patanali, with reference to the Secondary Formations') (-Smt. Manisha Pathak) For M. Phil, Tilak Maharashtra Vidyapeeth, Pune : "Oldenberg yanca akhyana-Siddhanta" (In Marathi) ("The akhyana Theory of Oldenberg") (-Smt. Usha Barve) (iii) 8. Academic Honours 1955. President, 1974. President, "Indian Linguistics" Section, 18th Session (Annamalainagar), All-India Oriental Conference "Vedic" Section, 27th Session (Kurukshetra), All-India Oriental Conference Linguistic Society of India "Sanskrit Literature & Linguistics" Section, Brihanmaharashtra Prachyavidya Parishad, 1st Session, Pune 1979. 1996. President, President, 9. Awards 1976. 1990. 1996. 1997. 1997. 1997. 1998. 1999. Maharashtra State Award for Teachers (University Level) President's Certificate of Honour (Sanskrit) MM. Dr. P. V. Kane Gold Medal for 1992-95 (The Asiatic Society of Bombay) Guru Gangeshvara National Veda-Vedanga Award (Nashik) Pandit Rajaramshastri Natekar Award (Pune) Purushottam Award (Late Mrs. Jayanti Wasudev Trust, Sangli) Prof. N. R. Phatak Memorial Award (Pune) Shreemanta Nanasaheb Peshwe Religious and Spiritual Award (Shree Devadevshwar Samsthan, Pune) 10. Publications Separately listed under : "Bibliography of Writings of Professor Mehendale."

Page #14
I. BOOKS: 1. 2. 3. 4. *5. *6. *7. *8. *9. 10. 11. 1948. 1963. 1965. 1978. 1968. 1976. 1980. 1982. 1990. 1995. 1995. 1996. BIBLIOGRAPHY OF WRITINGS OF PROFESSOR MEHENDALE (As on 14.02.1999) Historical Grammar of Inscriptional Prakrits [Reprint, 1997] (Deccan College P. G. & R. I., Pune) (Jointly edited with Prof. E. Waldschmidt :) Barhut Inscriptions by H. Luders Nirukta Notes: I Nirukta Notes: II Some Aspects of Indo-Aryan Linguistics (University of Bombay, Bombay) Rgveda-Samhitakara and Father Esteller (S. P. Mandali, Pune-411030) [Hindi Translation by M. D. Paradkar] (Yaska, New Delhi, 1993) Questions and Answers in Vedic Literature (University of Poona, Pune-411 007) Rgvedasara [Marathi Trans. of Hymns selected by Acharya Vinoba Bhave] (Vaidik Samshodhan Mandal, Pune-411 037) Game of Dice in Ancient India (Jnanaprabodhini, Pune-411 030) Gathas of Zarathushtra and Yasna Haptanhaiti [Marathi Trans. with Sanskrit parallels] (Vaidik Samshodhan Mandal, Pune-411 037) Reflections on the Mahabharata War (Indian Institute of Advanced Study, Shimla) Cultural Index of the Mahabharata (Edited), Vol. I (Bhandarkar Oriental Research Institute, Pune) 12. (In Press) Prolegomena on the Metre and Textual History of the Rgveda by H. Oldenberg [Eng. Trans. jointly with V. G. Paranjape] (*Written in Marathi) xiii

Page #15
II. PAPERS: ABORI ABSP AORM BDCRI Bh. Vid. BSOAS COI CV FV GJKSV GJV IIJ IJDL IL Indol. Stud. Indol Taur. JABSP JAOS JASB JGJKSV JOIB JUB MV NS PAIOC PICO Proc. SJV SVUOJ VIJ VSM Abbreviations Annals of the Bhandarkar Oriental Research Institute Akhil Bharatiya Sanskrit Parishad Annals of Oriental Research, Madras Bulletin of Deccan College Research Institute Bharatiya Vidya Bulletin of the School of Oriental and African Studies Cama Oriental Institute Commemoration Volume Felicitation Volume Ganganath Jha Kendriya Sanskrit Vidyapeeth Golden Jubilee Volume Indo-Iranian Journal International Journal of Dravidian Linguistics Indian Linguistics Indological Studies Indologica Taurinensia Journal of the Akhil Bharatiya Sanskrit Parishad Journal of the American Oriental Society Journal of the Asiatic Society of Bombay Journal of the Ganganath Jha Kendriya Sanskrit Vidyapeeth Journal of the Oriental Institute, Baroda Journal of the University of Bombay Memorial Volume New Series Proceedings of the All-India Oriental Conference Proceedings of the International Conference of Orientalists Proceedings Silver Jubilee Volume Sri Venkateswara University Oriental Journal Vishveshvaranand Indological Journal Vaidik Samshodhan Mandal xiv

Page #16
List of Papers (Note : The end column refers to pages from the Present Volume.) (A) ARTICLES (1) Veda 1. 1952-53. On Cakran na in the Rgveda X. 12-13 (BDCRI 14.109-18) 2. 157. Vedic Akhkhala : Pali Akkula. (Dr. S. K. Belvalkar FV [Banaras). 12-15) 3. 1960. Yatumavant. (BDCRI 20.375-78) 4. 1961. Satyam eva jayate nanitam. (JAOS 81.405-08) 5. 1962. Two Derivatives in - ya. (BSOAS 25.597-601) 6. 1862. Some Lexicographical Notes on the Upanisads. (IIJ 5.184-86) 7. 1963. Upanisadic Etymologies (Munshi Indological FV/ Bh. Vid. 20-21.40-44 [1960-61] 8. 1965. Vanargu (H.D. Velankar CV [Bombay). 76-81) 9. 1965. A Note on Tena tyaktena bhunjithah. (Appendix to I. Karve, Kinship Organization in India (2nd ed., Bombay]. 376-77) Sure duhita. (BDCRI 25.71-76; also 1969 PICO [Bombay). 3.105-08) 11. 1966. Sanskrit Sakhi. (JASB. NS. 41-42. 80-89) 12. 1968. Madhye lagnam (Baudhayana sSu. 9.3). (ABORI 48-49 [GJV. 193-95) 13. 1971. Sahasah svajah in the Aitareya Brahmana 13.2 (Pancika 3.26). (BSOAS 34.376-79) 14. 1972. Skt. Visipriya. (JGJKSV 27.323-28) Dagh-(Verb), daghna (suffix), and Derivatives : A Study in Historical Lexicography. (IL 33.239-50) 16. 1973. Adhvan : Adhvara. (PAICL [4th Session, Agra). 205-12; also : 1975. AORM (SJV]. 25.481-91) XV

Page #17
97 101 103 109 112 117 137 139 17. 1974. Asusu (VIJ 12.198-201) 18. 1974. Two Vedic Notes. (BSOAS 37.670-71) 19. 1974. Sapta devalokah. (Charudeva Shastri FV [Delhi] 106-11) 20. 1975. Ekavit in the Vedic Literature. (PAIOC [26th Session, Ujjain, 1972]. 149-52; also : Prof. K. A S. Iver FV/Rtam : JABSP (1970-75] 2-6.77-79) 21. 1975. The Abode of Mitra. (A. D. Pusalkar CV (Delhi). Indol. Stud. 3.110-14) 22. 1975. The Rgveda-Samhitakara and Father Esteller. (BDCRI 35.97-116; also : 1976. Presidental Address, Vedic Section, PAIOC [27th Session, Kurukshetra, 1974). 39-60) 23. 1976-77. Two Compounds with Agni (BDCRI 36.115-16) 24. 1977. arnsala. (Ernst Waldschmidt FV/Beitrage zur Indienforschung (Berlin). 315-18) 25. 1978-79. Professor Thieme's Etymology of Skt. asi and its Bearing on the Iron-age in India. (Puratattva 10.79-80) 26. 1979. Prakamodya - its Relevance to the Akhyana Theory. (Ludwik Sternbach FV [ABSP, Lucknow). 143-44) 27. 1980. Atihayat (?). (VIJ 18.19-21) 28. 1980-81. Durativyatha. (Ind. Taur. 8-9.253-55) 29. 1982. Aduvas (?). (JAOS 102.365) 30. 1986. Two Vedic Notes. (Acharya Udayavira Shastri FV/ Rtambhara : Studies in Indology [Ghaziabad). 98-99) 31. 1986. On the Brhaddevata 2.102.03. (IIJ 29.117-20) 32. 1988. On Mitradruh and Mitradroha. (ABORI 69.249-54; also : 1989. Prof. P. V. Bapat FV/Amala Prajna : Aspects of Buddhist Studies (Delhi). 547-51) 33. 1996. Sayana's Interpretation of daurgahe badhyamane in the Rgveda 4.42.8 (Festschrift Paul Thieme/Studien zur Indologie and Iranistik 20.257-64) 143 Bes 145 147 150 153 154 157 161 167 xvi

Page #18
34. 35. (In Press) On anadyamano yad anannam atti [Ch. Up. 4.3.7] (Muni Jambuvijay FV of Nirgrantha, Ahmedabad) Iranian (II) 36. 1959. 37. 1961. 38. 1961. 39. 1966. 40. 1969. 41. 1997. 44. 42. 1980. 43. 1982. 46. 1975. 47. 45. 1987. 1988. 1992. 1995. (III) 48. 1940. 49. 1940-41. 50. 1949. On Vedic Durona. (Dr. Biswanarayan Shastri FV/ Bharatavidya-Saurabham. (Guwahati]. 46-48) Two Sogdian Etymologies (IIJ 3.142-43) Some Avestan Notes. (IIJ. 5.12-14) - Two Avestan Notes. (IIJ 5.61-66) Two Notes on Yasts. (IL 27.70-73) Avestan vawa (n.f.). (COI) [GJV. 220-22; also: 1972. PAIOC [24th Session, Varanasi 1968]. 247f.) On the Significance of the Name Zarathushtra. (PAIOC [26th Session, Ujjain, 1972]. 159-64; also: V. G. Paranjape CV/Some Aspects of Indo-Iranian Literary & Cultural Traditions [Delhi]. 113-17) On Yasna IX.26.(IIJ 22.137-39) Some Remarks on Mihir Yast (Yast X). (GJV [VSM, Poona]. 164-66) Two Notes on Yasna 10. (Dr. Babu Ram Saksena FV/ Rtam [1979-83]. 26]261-63) Some Remarks on Yasna 34. (Festschrift for Henry Hoenigswald [Tubingen]. 261-62) Avestan gared and gerezdi-, (Studien zur Indologie und Iranistik (Reinbek). 16-17 147-48) On Yasna 41.2 (COI [2nd International Congress Proc., Bombay]. 173-76) Pali and Prakrit Takki or Dhakki (BDCRI 1. 189-204) A Further Note on Takki. (BDCRI 2.132-33) Some Phonetic Peculiarities of Early Ceylonese Lena Inscriptions. (Bh. Vid. 10.35-41) xvii 175 178 179 182 188 192 195 200 203 206 209 211 213 216 232 234

Page #19
51. 1952. What was the Place of Issue of The Dhauli and Jaugada 241 Separate Edicts ? (JOIB 1.240-44) 52. 1952-53. On the Prakrit and Sanskrit Names of the Nasik Cave-hill. 246 (BDCRi 14.163-67) 53. 1955-56. Some Remarks on the Language of the Original 251 Buddhist Canon (BDCRI 17.157-71; also : 1955. Presidential Address, Indian Linguistics Section, PAIOC 18th Session, Annamalai.83-94) 54. 1955-56. North Western (and Western) Influence on the Versions 266 of Asoka's Minor Rock Edicts. (BDCRI 17.81-97) 55. 1959. North-Western (and Western) Influence on the Mysore 283 Edicts of Asoka (JASB[NS]. 31-32/Sardha-Satabdi CV, (1956-57] 155-75) 56. 1968. On Gathas 8 and 10 of the Takkariya Jataka. 304 (Prof. M. B. Emeneau FV/Studies in Indian Linguistics. 231-33) 57. 1970. Notes on Asoka's Rock Edicts. (Rtam : JABSP 1.103-08) 307 58. 1970. Notes on Asoka's Seventh and Ninth Rock Edicts. 313 (Umesh Mishra CV [Ganganath Jha Res. Inst., Allahabad). 581-85) 59. 1970. On the Name and Gatha 12 of the Takkariya Jataka. 318 (Proc. of the Seminar in Prakrit Studies [Pune 1969]. 125-30) 60. 1971. The Dictionary of Inscriptional Prakrits - a Desideratum. 324 (Proc. of the Seminar in Prakrit Studies (Bombay). 165-71) 61. (In Press) On the Gatha 56 of Nalinikajataka [526]. (Prof. H. C. Bhayani FV of Nirgrantha, Ahmedabad) 63. 64. (IV) 1957. 1960. Linguistics Trace of an Old Palatal * zh>j in Sanskrit. (IL 17.16-23) Limitations of the Method of Internal Reconstruction. (IL 21.101-03) 336 344 xviii

Page #20
65. 1963. Internal and Comparative Reconstruction : Some 347 Procedural Considerations, (IL 24.41-47) 1965. Sanskrit Balha. (Georgii Kurylowicz FV/Symbolae 354 Linguisticae [Krakow). 199-203) 67. 1968. Two Notes on Internal Reconstruction. (IL 27 359 [Supplement). 53-56) 68. 1970. Evidence for Affricate Pronunciation of the cluster ts 363 in the Maitrayani Samhita. (W. B. Henning MV [London). 299-3021 69. 1970. Three Notes (I) Neutralization of Contrast, II) Analogy, 367 III) Regularity of Phonetic Changes]. (IL-33.85-87) 70. 1972. Notes on Internal Reconstruction and Comparative Method. 371 (JUB [Arts Vol. 41 NS]. No. 77.236-40) 71. 1973. On the Varttika 2 on Panini 6.1.83.(BDCRI 33.249-52) 376 Analysis of Meaning. (Studies in Historical Sanskrit 380 Lexicography (Poona). 76-88) 73. 1980. On TRCA. (IIJ 22.140) 393 74. 1982. On Panini 1.3.41. (ABORI 63.225-26) 394 75. 1998. Prof. Manfred Mayrhofer's New Etymological Dictionary 396 of Sanskrit (ABORI 79.125-36) 76. (In Press) Does Patanjali on Panini 5.2.9 Refer to Chess ? (Prof. G. Cardona FV, ed. by Madhav Deshpande, USA) (V) Mahabharata 77. 1984. Mahabharata Studies I. (ABORI 65.245-50) 78. 1984. Nemesis and Some Mahabharata Episodes. 79. 1986. Draupadi's Question. (JOIB 35.179-94) 80. 1986. The Flora in the Aranyakaparvan of the Mahabharata. (ABORI 67.233-42) 81. 1987. The Fauna in the Aranyakparvan of the Mahabharata. (ABORI 68.327-44) 408 414 421 437 447 xix

Page #21
465 472 484 488 493 82. 1990. Once again Draupadi's Garments. (BDCRI (GMV 50.286-90) 83. 1991. The Custom of Sahagamana : Some Early References. (Proc. of the National Seminar on Living Customs and Their Ancient Indian Sources. [Pune]. 159-71) 84. 1991-92. Damaynti's Svayamvara. (ABORI 72-73.483-86) 85. 1991-92. The Game of Dice in Ancient India. (Professor Katre FV BDCRI 51-52.153056) 86. 1992. Has the Vedic Rajasuya and Relevance for the Epic Game of Dice? (Professor A. M. Ghatage FV/ Vidyavratin (Pune). 61-67). 87. 1995. Is there only one Version of the Game of Dice in the Mahabharata ? (Professor R. K. Sharma FV Modern Evaluations of the Mahabharata (Delhi). 40-47) 88. 1997. Once again Draupadi's Hair. (ABORI 78.159.75) 89. 1998. Urubhanga and the Critical Edition of the Mahabharata. (JASB. NS. 73.91-97) 90. (In Press) Drupadi's Garments, (Prof. H. C. Bhayani FV, ed. by Bharati Modi & P. J. Mistry) 500 509 526 533 (VI) Miscellaneous 91. 1970. Ahinsa and the Spread of Vegetarianism in India. (Humanist Review, 419-26) 92. 1978-79. Medhatithi on Manu IX.256. (SVUOJ 21-22.57-59) 93. 1980. Did Ashoka Ban Cow-slaughter ? (New Quest 39-41) 94. 1982. The Mitani Name Sattiuaza (IIJ 24.295f.) 95. 1982. Galayasti. (IIJ 24.296f.) 96. 1993. The Indo-Aryans, Indo-Iranians and Indo-Europeans. (The Aryan Problem. (Pune). 43-50) 97. 1994. When Kalidasa nods. (Prof. C. G. Kashikar FV/sruti cintamanih (Pune). 87-89) 98. 1996. Some Reflections on the Raghuvamsa. (Dr. G. B. Palsule FV / Amrtamandakini. [Pune). 146-49) 541 543 547 547 549 562 in 565 xx

Page #22
569 571 573 576 599 602 605 (B) REVIEWS 1. 1952-53. Gujarati par Pharasini Asar (in Gujarati) by C. R. Naik [Ahmedabad 1954). (BDCRI 14.239-40) 2. 1952-53. The Etymologies of Yaska by Siddheshwar Varma [Hoshiarpur 1953]. (BDCRI 14.307-09) 3. 1955-56. Prakrta Bhasa by Prabodh Bechardas Pandit (Banaras 1954]. (BDCRI 17.51-52) 4. 1955-56. Beobachtungen uber die Sprache des Buddhistischen Urkanons by Heinrich Luders [Berlin 1954]. (BDCRI 17.54-75) 5. 1955-56. Braja-Bhasa (in Hindi) by Dhirendra Varina [Allahabad 1954). (BDCRI 17.154-56) 6. 1955-56. Buddhist Hybrid Sanskrit Language and Literature by Franklin Edgerton (Banaras 1954]. (BDCRI 17.231-33) 7. 1956. Kanhadade Prabandha by Padmanabha (Vol. 1 : Text). ed. by K. B. Vyas (Jaipur 1953]. (ABORI 37.340-41) 8. 1961. Handbuch des Sanskrit. Mit Texten u. Glossar by Albert Thumb/Richard Hauschild (Darmstadt, Germany]. (ERASMUS 14.731-35) 9. 1965-66. Mitra and Aryaman by Paul Thieme [New Haven 1957]. (BDCRI 26.63-67) 10. 1967. Emphasizing and Connecting Particles in the Thirteen Principal Upanisads by Karl Gustav Hartman [Helsinki 1966]. (Erasmus 19.404-06) 11. 1967-68. Janakiharana by Kumaradasa ed. by S. Paranavitana and C. E. Godakumbura (Shri Lanka Sahitya Mandalaya 1967]. (BDCRI 28.229-31) 12. 1970. Nituka-Mimanisa by Shivanarayana Shastri [Varanasi 1969). (IL 31.58-60) 13. 1970. An Eleventh Century Buddhist Logic of 'exists', with Introduction, Eng. tr. and Notes ed. by A. C. Senape McDermott [Dordrecht, Netherlands 1969). (IL 31.118) 607 612 616 618 620 623 XX

Page #23
14. 1971. 15. 1971. 16. 17. 18. 1972. 19. 1972. 1971. 20. 1973. 22. 1972. 21. 1973. 23. 24. 26. 1973. 1975. 1975. 25. 1981. 1984. Some Aspects of Applied Linguistics by A. M. Ghatage [Kolhapur 1970]. (ABORI 52.241-42) Balts and Aryans by S. K. Chatterji [Simla 1969]. (ABORI 52.242-44) The Pengo Language, Grammar, Text and Vocabulory by T. Burrow and S. Bhattacharya [Calcutta 1960]. (ABORI 52.244-45) A Study of Kamarupi: A Dialect of Assamese by Upendranath Goswami [Gauhati 1970]. (IL 33.88-89) Linguistic Theory and The Study of English: A Selective Outline by Joseph R. Keller [Minneapolis Minn. 1968]. (IL 33.176-77) Praci Jyoti: Digest of Indological Studies Vol. VII ed. by Gopikamohan Bhattacharya [Kurukshetra, 1971]. (IL 33.177-78) Sound Change by D. N. Shankara Bhat (Poona 1972). (IL 34.229-30) Kshetresha Chandra Chattopadhyaya FV. Pt I/JGJKSV 27:3-4, ed. by B. R. Saksena, S. P. Chaturvedi & others. [Allahabad 1972]. (IL 34.316) Kausikasutra-Darilabhasya ed. by H. R. Diwekar, V. P. Limaye, R. N. Dandekar and C. G. Kashikar, [Poona 1972]. (ABORI 54-294-97) Proto-New-Indo-Aryan by Subhadra Kumar Sen [Calcutta 1973]. (IL 36.63-65) A Prakrit Reader by H. S. Ananthanarayana [Mysore 1973]. (ILDL 4.379-80) Indian Lexicography by Klaus Vogel [Wiesbaden 1979]. (ABORI 62.316) Hindi-Gujarati Dhitukosa (A Comparative Study of Hindi-Gujarati Verbal Roots) by Raghuveer Chaudhari [Ahmedabad 1981]. (ABORI 65.288-89) 624 625 627 629 631 632 634 636 637 639 642 644 645

Page #24
4. 647 28. 1984. 651 29. 1984. 656 30. 1985. 660 1985. 661 1985. 662 1985. 663 1985. 665 Dravidian Theories by R. Swaminatha Aiyar [Madras 1975]. (ABORI 65.289-93) Sexual Ethics in the Mahabharata in the Light of Dharmasastra Rulings by Bhakti Datta (London 1979). (ABORI 65.293-97) Mahabharata : 1 Adiparvan tr. into Marathi and ed. by R. S. Walimbe (Pune 1982]. (ABORI 65.298-300) The Mahabharata : The Story and its Significance by S. L. N. Siinha Calcutta 1983]. (ABORI 66.310) Secondary Tales of the Two Great Epics by Rajendra Nanavati (Ahmedabad 1982]. (ABORI 66.311-12) The Brahmanda Purana translated and annotated by G. V. Tagare [Delhi 1983]. (ABORI 66.312-14) The Krsna Cycle in the Puranas by Benjamin PreciadoSolis [Patna 1984). (ABORI 66.314-15) Erich Frauwallner Nachgelassene Werke : Vol. led. by Ernst Steinkellner (Wien 1984]. (ABORI 66.315-16) Poesie Sanskrite Conservee dans les Anthologies et les Inscriptions : Tome III by Ludwik Sternbach [Paris 1985]. (ABORI 67.311-12) Bhrgus - A Study by Jayanti Panda [Delhi 1984]. (ABORI 67.312-13) A History of Classical Sanskrit Poetry : SanskritPali-Prakrit by Siegfried Lienhard (Wiesbadebn 1984). (ABORI 67.313-15) The Mahabharata, its Genesis and Growth by M. R. Yardi [Poona 1986). (ABORI 69.349-55) Kalyanamitraraganam by Eivind Kahrs [Norwegian University Press 1986]. (ABORI 69.355-57) Textkritische Bemerkungen Zur Maitrayani Samhia by Martin Mittwede [Stuttgard 1986]. (ABORI 69.358) Mallanaga Vatsyayana - Das Kamasutra ed. by Klaus Mylius (Leipzig 1987]. (ABORI 69.359) 35. 1986. 666 36. 1986. 666 37. 1986. 667 38. 1988. 670 39. 1988. 676 40. 1988. 679 41. 1988. 680 xxiii

Page #25
42. 1988. 43. 1988. 44. 1988. 45. 1990. 46. 1990. 47. 1990. Siva Dans Le Mahabharata by Jacques Scheuer 681 (Paris 1982]. (ABORI 69.360-61) Prajapati's Rise to Higher Rank by J. Gonda 683 [Leiden 1986). (ABORI 69.361) Encyclopaedia of Puranic Beliefs and Practices Vols I-III 684 by Sadashiv A. Dange [New Delhi 1986, 1987]. (ABORI 69.362-64) Bruderschaft und Wurfelspie (Untersuchungen zur 687 Entwicklungsgeschichte des Vedischen Opfers) by Harry Falk (Freiburg 1986). (ABORI 71.386-92) Elements of Poetry in the Mahabharata by Ram Karan Sharma691 (Delhi 1988] (ABORI 71.392-94) Moral Dilemmas in the Mahabharata ed. by B. K. Matilal 695 [Delhi 1989). (Indian Literature [Non-Dec. 1990) 33.6.81-94) The Astronomical Code of the Rgveda by Subhash Kak 709 [New Delhi 1994]. (ABORI 77.323-25) Johann Otto Ferdinand Kirste : Kleine Schriften ed. by 711 Walter Slaje [Stuttgard 1993]. (ABORI 78.331-32) hermann Oldenberg : Kleine Schriften, Teil 3 ed. by Hanns- 713 Peter Schmidt (Stuttgart 1993). (ABORI 78.332-33) Suresvara's Vartika on sisu and Murtamurta Brahmana ed. 715 by K. P. Jog & Shoun Hino (Delhi 1996]. (ABORI 78.333-37) Hanns Ortel : Kleine Schriften, Teil I & Teil II 719 ed. by H. Hettrich & T. Oberlies (Stuttgart 1994] (ABORI 79.293-95) Paul Thieme : Kleine Schriften 721 (2 Vols. ed. by G. Buddruss (Wiesbaden 1984] & by R. Sohnen Thieme (Stuttgart 1995] respectively (ABORI 79.295-97) 1996. 49. 1997. 50. 1997. 51. 1997. 52. 1998. 53. 1998. 1. 1985. 2. 1985. 3. 1998. (C) OBITUARY NOTICES Sir Ralph Lilley Turner (ABORI 66.359-61) Professor Dr. Ernst Waldschmidt (ABORI 66.364-65) Professor Dr. Sumitra Mangesh Katre (ABORI 79.301-09) 724 727 729 xxiv

Page #26
COLLECTED PAPERS (A) Articles (B) Reviews (C) Obituary Notices

Page #27

Page #28
Reprinted from Deccan Bulletin Vol. XIV, No. 2. (A) ARTICLES (1) Veda ON CAKRAN N.i IN THE RGVEDA X. 95-12-13. By M.A. MENENDALE The Ryveda X,95 comprises the famous dialogue between the king Pururavas and the nymph Urvasi. According to the version of the story as preserved in the Satapatha Brahmana (11.5.1) the dialogue is said to have occurred near a lake in the Kuruksetra where the king had come by chance while he was wandering here and there after his separation from Urvasi. Just at that time Urvasi had also come there with lier female friends and was sporting in the lake in the form of waterbirds. Urvasi recognised Pururavas and let him also recognise her. It is at this point that the dialogue between the two, as given in the present hymn, is supposed to have taken place. Pururavas tries to persuade Urvasi to come and stay with huim again, but she rejects all the approaches made by him. It is only an invisible voice in the end that seems to assure the king of a happy life in the heaven. In order to understand the setting in which the two verses containing the expression under discussion occur, it would be worthwhile to note here in general the contents of the dialogue between the king and the nymph. Pururavas asks Urvasi to stay for a while and have a conversation with him, for, the thoughts that remain unexpressed do not give happiness to anyone. Urvasi replies that this is going to be of no use, as she has left him for ever and has become inaccessible to hin like the wind. Pururavas pleads that he himself is not responsible for Violating the condition which compelled Urvasi to desert him ; it was a trick played by the Gandharvas. Without entering into any discussion on this point, Urvasi only reminds the king that she has kept lier word. Pururavas now complains that not only Urvasi but also her friends had at that time disappeared and why? Urvasi justifies their action by saying that as she was then expecting a child, her friends had to come away to nurse the baby...... The king remembers his enjoyments in the company of Urvasi who had then disappeared like a lightning; now Urvasi has given birth to a child---may she live long ! Urvasi says that the child is nothing but the embodiment of the strength of Pururavas; she had told him of this coming event, but he would not believe her. Pururavas now wants to know if the child longs to see his father, shedding tears when he knows about him (verse 12). Urvasi assures the king that the child does shed tears, and will cry for paternal affection ; she would send the child to Pururavas--but the king would 1. For a detailed account of the story as given in the various versions, cf. GELDNER, Ied. Stud. I.243-295. 2. The condition was that Urvasi should not see the king naked. The Gandharvas had manreuvred to show the king to her in this condition with the help of a lightning. Madhu Vidya/3

Page #29
110 M. A. MEHENDALE not get her back (verse 13). Having lost all hopes, the king threatens to kill himself--to throw himself before the hungry wolves. Urvasi implores hiin to do nothing of the kind ; she advises him not to put too much trust in women, and disappears. As the king makes a final appeal to her to return, a heavenly voice intervenes, assuring the king joy in the heaven. The expression cakran na first occurs in verse 12. Pururavas, having come to know that a son has been born to him, asks Urvasi. kada sunuh pitarar jata icchac cakran nasru vartayad vijanan / The Padapatha renders it as follows--- kada/ sunuh/ pitaram, jatah/ icchat) cakran/ nal asru/ vartayat/ vi sjanan! GELDNER translates---"When will the son, who is born, seek his father? Like & crying (child), he will shed tears, when he knows definitely." Sayana renders it as--" When will the son, born (of you), desire for the father? When, knowing (me as his father), and (na = ca) crying, will he shed tears?" The same expression occurs once again, but in a slightly different context, in the following verse. Urvasi replies to the above question of Pururavas in verse 13 which runs as follows-- prati bravani vartayate asru cakran na krandad adhye sivayai/ The Padapatha renders it as follows - prati/ bravani/ vartayate) asru cakran/ na/ krandat/ as dhye/ sivayai/ GELDFER translates--" I shall be glad to reply to him when he sheds tears. Like a crying (child), he will cry for the kind cares (of the mother).''3 Sayana's interpretation of the line is as follows--" (Oh Pururavas), I tell you by way of reply-(your son) will shed tears; and (na = ca) he will weep, 'shedding tears. when the auspicious occasion, that is longed for, is near at hand." 4 The general purport of the question and the answer is not difficult to follow. The difficulty, however, lies in the interpretation of the expression cakran na which occurs in both the verses and which is no doubt intended to express the object of comparison with which the weeping (v. 12) and the crying child (v. 13) of Pururavas is compared. The earlier attempts to explain the form cakran have been mentioned by OLDENBERGS and rightly rejected by him. OLDENBERG himself would like to derive cakran from V krand and explain it as a reduplicated rootnoun of the type mentioned by WHITNEY in his Sanskrit Granmar 8 1147 c. This 8. GELDNER hiniself disapproves of his earlier suggestion (Kommentar, p. 193) to take nu = not' in this verse (but mui = like' in verse 12) and regard that in a the crying person is the son, in b the father (for his explanation of cakram = father', see f.n.7 helow). 4. adhye adhyate vastuni fivayai sive kalyane samupasthite sati. 5. Noten II. 808-309. Madhu Vidya/4

Page #30
ON CAKRAN N. IN THE RGVEDA X 95.12-13 111 would give us cakrad, like sasyad from Vsyand, and not cakran. But this means only going against the Padapatha and does not involve any emendation in the Samhitapatha ; for, in the Samhita cakrad na would give cakran na.8 GELDNEX. in his new translation of the Rgveda, accepts the derivation of cakran from v krand,? but he would regard it as haplology for cakradan as kanikrat (Rv. 9.63.20) for kanikrodat. As against this explanation it may be observed that in cakradan the accent is on da and hence its loss due to haplology is unlikely ; in kanikradat, however, the accent is on the first syllable and not on da. Sayana already seem to connect cakran with V krand when he explains the text word in verse 12 as krandamanah; that he paraphrases the text word in verse 13 by rudan show that in both the cases he regards the form to be present participle. It is obvious that these attempts to explain cakran are not convincing. Apart from the irregularities involved in connecting the form with V krand, this derivation remains unsatisfactory also from the point of view of the meaning of the verb, The verb V krand, in the Rgveda, is primarily restricted to the cries uttered by the animals like a horse or a bull, or to the shrieking sound made by the birds ; secondarily it is also applied to the thundering or the rumbling sound of the clouds, winds, or waters, or to the sound of the fire and lightning. It is generally believed that'a war, or a war army' is called 'krandas' in the Rgvedas owing to the human shoutings. This is not improbable, but in view of the overwhelming evidence where V krand is applied to the sounds of animals it would be more proper to suppose that krandas is used in the above senses on account of the sound raised by the neighing of the horses. It would certainly not be beside the point to think here also of the war-drum, the beating of which must have added to the great tumult. Thus in the Atharvaveda 5.20.9, a war-drum is called samkrandana. In the same hymn, in verse 7, the verb abhi V krand is used with it, and it is also used in verse 2, where it is compared to a bull (rsabho vasitam iva), or is itself called a bull (vrsa tvam). In the Rgveda, the verb v krand occurs in connection with the 6. In support, OLDENBeuc refers to MACDONELL, Vedic Grammar, 874 a (p. 67). MACDONELL, however, observes--"A final media before a nasal inay become the nasal of its own class. There seems to be no certain instance of this in the Rv.; cakran na (X.95.12.13), however, probably stands for cakrat nd, though the l'p. has cakran na." Thus the sandhi between cakrad na as cakranna is not obligatory, though it is possible. Panini (8.4.45) allows similarly option in such cases in the spoken language. cf. etad murarih or etan murari. 7. Earlier in his Ved. Stud. 1.279 (1880), GELDNER had rejected the derivation from V krand and regarded cakran as the participle of the redupl. Aor. of a rootkr (as vardhantah 4.2.17) As none of the existing three roots kr had similar formation, he proposed to postulate a fourth V kr which would lie at the basis of the words like kiri, karuna, kurara (GELDNER translates here cakrun as unlucky person '). Some years later in his Kommentar (1909) on p. 193 (and Glossar p. 47 under v kr) GELDNER proposes to derive it from v kr, and give it the meaning the pro creator, the father' (he takes the second pada in verse 12 to mean' knowing the facts) he will cry like his father '). 8. 10.88.1 ; 2.12.8 ; 0.25.4. In 10.121.6, by krandasi, heaven and earth are meant. Madhu Vidya/5

Page #31
112 M.A. MEHENDALE deities like Soma and Indra(r) only when they are compared to a horse or a bull. According to the Petersburg Worterbuch the use of this verb as applied to the crying aloud while weeping is instanced in the Rigveda only by the form krandat in our verse 13. As will be shown below, even this instance is explained only on the basis of the object to which the crying child is compared thereby showing that rand is not primarily applied to the cries of human beings. The use of V krandi in the sense to weep or lament' is common only in later literature 10 In such circumstances there is little point in attempting to derive cakran from V krand and translate caleron nasru vartayut as 'He will shed tears like & crying (child)'. (verse 12), or cakran na krandat as he will cry like a crying (child) (verse 13). Moreover, what more sense is gained if one compares a weeping or a crying child with another crying child ? Surely there is nothing gained by saying he smiles like a smiling man'. 'The position is hardly improved when OLDENBERG (Noten II.309) renders the expression in verse 13 as " May be (the child of Urvasi) weep like some one who bitterly wceps (?? for a dead person. like a lamenting wife ? ) as if he could in that way stress his noble longing (for the reunion of his parents)." In order to arrive at a satisfactory explanation of the expression, it is necessary to remember the following three things which seem to be clear (i) that the weeping (verse 12) and the weeping and crying person (verse 13) is one and the seme viz. the child of Purura vas and Urvasi; (ii) that na = iva in both the verses ; (ii) and that cakran represents the object of comparison (upamana) with which once the weeping child (asru Vvart), and once the crying child (V krand) is compared. These facts are sufficiently compelling to lead one to the supposition that there has been some misunderstanding on the part of the author of the Padapatha in the interpretation of cakranna which has further misled the modern interpreters of the verses. I feel that the author of the Padapatha has mistaken the object of comparison cakrarn for cakran. The correct rendering of the Samhita text in the Padapatha therefore should have been cakram na/ and not cakran) wa). The form and accent of cakram are quite regular and the meaning 'wheel' or 'a bird of that name,' undoubtedly gives a better meaning. Also the sandhi between cakram ni would regularly give cakrannall in the recited text, though in the manuscripts it is 9. In the Rv. 10.103.1, Indra, engaged in war, is no doubt called suurkrindana, but in verse 1 he is compared to a bull (ursabho ni bhimil), and in verse ? he is himself called a bull (visan). In 8.100.5 aciktadun sisuwantah sukhayah seerns to refer to the Maruts or the horses of Indra. In 10. 1.2 it is the Fire which is spoken of as fisuh...... kanikradat, and in 10.96.10, aciera. dat primarily goes with the neighing of the horse (9.67.4) with which the charioteer (harivun) is compared. 10. This explains Savana's rendering of the text-words in vers. 13 as cakran tuulann asrini vimuncan / ......krandat......rotsijati! 11. See WACKERNAGEL, I % 283, b (. 333)---final ni is regularly assimilated to the following n. WACKERNACEL further points out that this change of final m ton before dentals and n has led sometimes to mistakes in the Padapatha. So Kv. 4.11.6 yan ni-pisi has been analysed in the Padapatha ay yat ni-pasi instead of yam ni-pasi; in Rv. 4.24.6 avi-venan tam has been anolysed as avi-venan tum instead of civi-uenan tam. OLDENBERG, however, expresses himself against this view, cf. Noten 1.275-276; also cf. SIEG, Die Sagenstoffe des Rgueda, p. 2, f.n. 1. Madhu Vidya/6

Page #32
ON CAKRAN NA IN THE RGVEDA X 95.12-13 customary to show anusvara for such assimilations. That the manuscripts in our case do not show an anusvara is explained by the obvious fact that they were misled by the Padapatha. So far as the recitation is concerned, however, there would be no difference between the text as printed in these two verses es cakranna and at the other four places (Rv. 1.155.6; 4.31.4; 5.36.3; 8.6.38) where the printed text shows cakram na. The first line of verse 12 now means When will the son, born (to me). desire for (his) father? Knowing (about his father) he will let the tears roll like & wheel (set into motion)." To this Urvast replies in the first line of the following verse" I assure you (that) he lets the tear roll--he will cry like a (moving) wheel for the kind thoughts (of the father, i.e. for the paternal affection)." 113 When Pururavas compares the weeping child with a rolling wheel he is no doubt thinking of the round tears rolling down on both sides of the cheek and the rolling on of the two round wheels of a chariot. He may have also used the comparison in view of the long continuous action involved in both the cases11. When on the other hand, Urvasi compares the crying child with a wheel, the point of comparison is certainly the sound (Vkrand) made by the wheel in its forward. movement. The use of the verby/krand with the wheel of a chariot (or of a potter) and its employment as upamana is already found in the Satapatha Brahmana 11.8.1.1. Here it is said tad yatha ha vai idam rathacakram va kaulalacakram va Spratisthitam kranded evan haiveme loka adhruva apratisthita asub EGGELING translates", "Verily, even as this cart-wheel, or a potter's wheel, would creak if not steadied, so, indeed, were these worlds unfirm and unsteadied." The sound nade by the moving wheel has so much struck the early Aryan that he has chosen to name the bird cakravaka1 after it. This bird is noted in later literature for its continuous crying throughout the night.15 From this point of view it would not be unreasonable to suppose that Urvasi perhaps meant this bird when she used the word cakram as upamana for the crying child. The expression yatha......rathacakram....... krandet cited from the Sat.Br. above makes it fairly certain that in our hymn in 12. So far as the long continuous action is concerned, a potter's wheel is also used as the object of comparison. Thus in the Sankhyakarika 67, in order to show that the Yogi, even after attaining the perfect knowledge, does not immediately leave the body but continues to hold it for some time on account of the past impressions, he is compared with a potter's wheel which continues to rotate for some time on account of the past momentum, though the work of preparing the pot on it has been finished. Read samyag jnanadhigamad dharmadinam akaranapraptau! tisthati samskaravasac cakrabhramavad dhrtasarirah || The continuous action of the moving wheel is also suggested by such expressions as cakravrddhi used for the method of taking interest on the interest. 13. The Satapatha-Brahmana part V, p. 126. SBE Vol. 44. (Oxford, 1900). 14. The bird is also called simply cakra. cf. kokas cakras cakravako rathangahvayanamakah Amara. 2.5.23; also cf. Medini (Calcutta edn.) 3.28, and 169.31, Trikandasesa 3.8.351, and Vaijayanti 26.18. Further compare such of its names as cakrahva or cakrasahvaya, and of the sun cakrabandhava (or bandhu friend of the cakra birds') Abhidhanacintamani 2.96, and of the night cakrabhedani ( separater of the cakra birds ') Trikandasesa 1.4.23. 15. The use of a krand is found with these birds in the Kumarasambhava 5.26---parasparakrandini cakravakayoh puro viyukte mithune krpavati. Bull DCIRI-xiv-8 Madhu Vidya/7

Page #33
114 verse 13 we have to interpret cakranna krandat as cakram na krandat. Evidently the same interpretation has also to be adopted for verse 12, as Urvasl in her reply has chosen to use the same expressions as those used by the king in his question. M. A. MEHENDALE Concerning the proposed interpretation for verse 12, viz. cakram na asru vartayat, it may further be pointed out that the use of cakra with the verb vart and its employment as upamana is common throughout the Sanskrit literature,1 In the Rgveda itself we have the following instances where the uses mentioned above. are well illustrated. (1) abhi na a vavrtava cakram na vrttam arvatah niyudbhis carsaninam Rv. 4.31.4. (2) "(Oh Indra) turn yourself towards us, like the wheel set into motion towards the horses, with the yoked horses of the people." anu tva rodasi ubhe cakram na varty etasam | anu suvanasa indavah Rv. 8.6.38. "The two worlds follow you (oh Indra), as the wheel follows Etasa; (to you) follow the pressed Soma juices." (3) caturbhih sakam navatim ca namabhis cakram na vyttam vyatir avivipat Rv. 1.155.6. "He (Visnu) set into motion the pair, with four times ninety names, like a wheel set into motion."17 (4) cakram na vrttam puruhuta vepate mano bhiya me amater id adrivah Rv. 5.86.3. "Oh Indra (lit. invited by many), master of the press-stones, my mind shakes, due to fear of ignorance, like the wheel set into motion." rathyeva rarishyama (5) sakhe sakhayam abhy a vavrtsvaaum na cakram bhyam dasma ramhya Rv. 4.1.8. 16. The close relationship between cakra and Vvart can also be seen from such expressions as cakravartin a sovereign ruler' and cakravarta whirling motion,' ef. cakravarto bhramo bhrantis bhramir ghurnis ca ghurnane / Abhidhanacintamani 1519. The relationship may further also be observed from the fact that a pot prepared on a potter's wheel is called cakravrtia. cf. asuryam va etat patram yat kulalakrtam cakravrttam, Maitr. Sam. 1.8.3.; also ef. Kathaka Sam. 6.3.; the same expression occurs with the negative prefix in the form acakravarta (fem.) in the Ap. Sr. S. 6.3.7 ou which the commentator Rudradatta remarks, varlanan vartah na cakre vartanam yasyah sa tathokta na cakrabhramanena nirmitety arthah! 17. The pair referred to is that of the day and the night. The three hundred and sixty names are the names of the days in a year. 18. Here the point of comparison is also the shaking of the wheel in motion. 19. na cakram also occurs in the Rv. 2.11.20, but there the object of comparison is not the wheel, but the sun. cf. avartayat suryo na cakram bhinad valam indro ungirasvan "As the sun his wheel, (so) he (i.e. Indra) let (the wheel) roll on. Indra, accompanied by Angiras, broke open Vala." GELDNER suggests the possibility of supplying namuceh sirah (Rv. 5.30.7) as the object of avartayat. Sayana suplies vajram, but also gives an alternative rendering where nd not. I would take cakram itself as going with both Surya and Indra. For a reference to Indra letting the sun's wheel roll on for Kutsa, cf. Rv. 5.29.10. Madhu Vidya/8

Page #34
ON CAKRAN N.1 IN THE RGVEDA X 95.12-13 "Oh friend (i.e. Agni) turn yourself towards your friend (Varuna) like the speedy wheel, (turn) towards us like the quick (horses) yoked to the chariot, Oh wonder-worker, like the quick (horses)." (6) satra te dnu kratayo viva cakreva vartuh/ sutra maki asi srutah Rv. 4.30.2. In the above translation asum is taken as an adj. to cakram20 and rathya (nom. dual mas.) as yoked horses'. GELDNER takes asu in the sense of 'horse' and rathya (nom. pl. neut.) as the chariot-wheels.' He translates Oh friend. turn yourself towards the friend (Varuna), like the wheel towards the race-horse, like the chariot-wheels in quick journey for us. Oh Master, in quick journey!" But the placing of na makes this construction improbable. 115 "All people together follow you like the wheels. You are at once famous as the great one." (7) samanam artham caraniyamana cakrain iva navyasy a vavrteva / Rv. 3.61.8. Hastening towards the common goal, come here rolling, Oh New One (lit. Newer, refers to Usas), like the (rolling) wheel." (8) hi vartante rathyeva cakranyamanyam upa tisthanta rayah Rv. 10.117.5. For, (the wealths) roll like the wheels of a chariot. The wealths go to different persons." ()s suryah pary uru varamsy endro vavrtyad rathyeva cakra | Rv. 10.89.2. "This Surya (goes) round the wide regions. May Indra roll him herewards like the wheels of a chariot." In later literature also comparisons with a wheel with the use of the verb vart are often met with. They are especially to be found in the didactic verses like cakravat parivartante hy arthas ca vyasanani ca/2 iti kytva pratiksami bhartrnam udayam punah // Mbh. 4.007, sukhasyanantaram duhkham duhkhasyanantaram sukham | sukhaduhkhe manusyanam cakravat parivartatah // Ind. Spruche 3264,24 20. In that case, however, the use of cakra would be in the masculine. 21. For the use of cakriyeva with Vvart, cf. Rv. 1.185.1; also cf. Rv. 5.30.8. 22. This is according to the Calcutta edn. The critical edn. reads instead-anitya kila martyanam arthasiddhir jayajayau 4.19.3; also cf. Mbh. 18.2860 (Cal. edn.). 23. Also cf. Ind. Spr. 3261. 24. The same simile is implied in the Meghaduta, 109-nicair gacchaty upari ca dasa cakra nemikramena Madhu Vidya/9

Page #35
110 M.A. MEHENDALE The word cakra by itself, without involving any explicit comparison as in the ihove instances, is also often used with Vvart since the earliest times-- pancare cakre parivartamane tasminn a tasthur bhuvanani visva / Rv. 1.164.13.25 "On the wheel (of the year) with five spokes (i.e. the five seasons) and which inoves round and round have stood all the beings." ekacakram vartata ekanemi sahasraksaram pra puru ni pasca / Av. 10.8.7.20 "(It), having one wheel, one rim, and a thousand syllables, rolls forward first, down afterwards." apa junyam bhayir nuda ma cakra avrtsata / Maitr. Sam. 1.2.9.62. "Remove the fear relating to the people ; may the wheels (of the enemies) not turn (towards us)." The same Mantra appears in the Taitt. Br. 3.7.7.1427 slightly differently.-- apa janyam bhayam nuda / apa cakraini vartaya / on which Sayana comments * he uttarahavirdhina cukrini parabalany apavertaya etal losthavad apagamaya/;" and in the Sankh. Sr. S. 5.13,3 (also Kausi. Br. 9.14) the following Mantra is enjoined for repetition hy the Hotr priest while the Soma-carts are being moved forward--apeto janyam bhayam anyajanyan ca vitrahan / 'apacakra avrtsata " Hence (drive forth) the danger which comes from foreign (people) and from others than men, 0 slayer of Vrtra. The wheels (of the foes) have wended away." (Caland's Tr.). It is not necessary to give here inany citations from the later literature illustrating this use. It would be sullicient to note the following few ones-- Having described the wheel of sacrifice, Krsna remarks in the Gita 3.16, evar pravartitam cakran nanuvartayatiha yah / ghayur indriyarimo moghar partha sa jivati //28 Similar expressions recur when it comes to the description of the wheel of law. Thus while describing the condition of the Kuru kingdom under the rule of Blisma it is said in the Mahabharata (1.102.12),20 23. Also Av. 9.14.11. For other references from the Rv. cf. 1.161.11,14 : 4.28.2 ; 7.68.2 ; 10.27.19; the sume idea is expressed slightly differently in Hemicandra's Abhidhanacintamaniatasarpinyah sad ara utsurpinyas ta eva viparitak / evam dvadasabhir arair vivartate kalacakram idam ; in the Mbh. 4.47. 2 evur kalavibhagena kalacakran pravartate /; also cf. Mb). 1.1.38, Ramayana (G. CORRESIO's edn.) 6.73.83. The simile is made quite explicit in the Susrutn (Calcutta edn.) I. p. 19. lines 20-21--90 esa nimesadiyuguparyantah kalas cakravat parivartamanak kalacakram ucyalu ity cke. 20. Also of. Av. 11.6.22. 27. Also Ap. Sr. S. 11.7.2 ; Asval. Sr. S. 4.4.2: Manava Sr. S. 2.2.2.17 28. The same ideu is expressed in the Yajnavalkyasmrti 3.123 (nundasrama edn.). - tasmad annal punar yajnah punar annan punah kratuh / caun etad a nadyantar cakran saiparivariate // 29. Also of the following where cakrann occurs with V var. Mh. 1.60.45 (critical edn.) ; 12.18801, 13.4262 (both Cal. edn.): Bhagavata Purana 9.20.32. ate Madhu Vidya/10

Page #36
ON CAKRAN NA IN THE RGVEDA X 95.12.13 117 sa desah pararastrani pratigrhyabhivardhitah / bhismena vihitam rastre dharmacakram avartata // When the samsara is thought of as a continuously moving wheel, it is said in the Mahabharata 11.162. (Cal. cdn.)--- evar sam saracakrasya parivyttam vidur budhah/ In one of the aphorisms, the relationship between the master and the servants is described in the terms of the spokes of a wheel and its navel--and even with regard to this whcel the use of Vvart is not missed.-- araih sandharyate nabhir nabhau carah pratisthitah / svamisevakayor evamvrtti cakram pravartate !! Ind.Spr. 212, Subhasita ratnabhandagara 3.87, Pancatantra 1.89. The combination of cakra and Vvart seems to have been regarded so fixed that occasionally the same verb is also used with the associates of the wheel like pavi, and as the wheel forms a part of the chariot, it is not uncommon to find Vvart used with ratha itself. visvah pinvathah svasarasya dhena anu vain ekah pavir a vavarta / Rv. 5.62.2. 30 "You (Mitra and Varuna) swell all the cows of the stall (?). Following you rolls hereward the one rim (i.e. the wheel = the sun or the year)." anasvo jato anabhisur ukthyo rathas tricakrah pari vartate rajah / Rv. 4.36.1.31 "The chariot, that has appeared without a horse, which is without a rein, which is praiseworthy, goes round the regions with three wheels." sukho ratha iva vartatam krtya krtyakrtam punah / Av, 5.14.5. "May the witchcraft go back to its maker, like an easy (moving) chariot." tad yathaikapat puruso yann ekatascakro va ratho vartamano bhresar nyeti......! tad yathobhayatahpat puruso yann ubhayatascakro va ratho vartamano na risyati ...... Ait. Br. 5.5.8.32 "Just as a man with one foot while walking, or a chariot with a wheel on one side while moving, fails...... Just as a man with two feet while walking, or a chariot with wheels on both the sides while moving, does not come to harm...." All the instances cited above show a constant use of cakran with Vvart since the earlist times, and also its use as upamana. Hence it would not be going away from the established usage when it is suggested, that in verse 12, the Samhita 30. Also cf. Rv. 10.27.6, (10.156.3 panim for pavim?); 5.81.5. 31. Also cf. Rv. 5.55.1. It may further be noted that suvit (and occasionally trivit) is one of the common epithets of a chariot in the Rgveda. cf. 1.34.9, 12; 47.2,7; 111.1 ; 118.2,3; 188.2,8; 2.40.3; 3.58.3; 4.33.8; 36.2; 44.5; 8.85.8 ; 10.39.1 ; 70.3; 85.20; 107.11. 32. Similar statement is found in the Chand. Up. 4.10 as well, where the relation between cakra and Vvart is made cxplicit-sa yathaikapad vrajan ratho vaikena cakrena vartamuno Tisyati...... / sa yathobhayapad vrajan ratho vobhabhyam cakrabyar vartamanah pratitisthati..........! Madhu Vidya/11

Page #37
118 M. A. MEHENDALE text should be interpreted as cakram na asru vartayat. As already shown above the same interpretation is made more than probable by the parallel use of krand with cakrd in the Satapatha Brahmana. It is, however, not difficult to understand why we find the wrong analysis of the Samhita text just in these two verses when comparisons with a wheel were not uncommon in the Rgveda. For, from among the nine cases cited above, where cakra is used as an object of comparison, the last four cases show the use of iva instead of na, and hence in these cases the possibility of any misunderstanding is altogether excluded. In the 5th instance, na precedes cakram (asum na cakram), and so here also no mistake was likely. In the first four instances where cakram na occurs, it occurs together with vrttam or varti. This fact, as well as the context, make it impossible that in these cases cakram na could ever be mistaken for cakran na. In our present case, however, vrttam is omitted after na; and secondly, the context is that of a weeping child. That is the real ground for the mistake. The use of krandat in verse 13 has misled OLDENBERG to observe (Noten II.309)-"Mir scheint 13deg unverkennbar auf Anschluss an krand- zu deuten redupliziertes Wzlnom, des Typus Wh. 1147". The way in which OLDENBERG has thought in the 20th. century seems to hold a pointer to the way in which the author of the Padapatha, at least some centuries before Christ, must have thought about our verses. He too seemingly felt that in the given context cakran should be some form derived. from krand and hence he analysed cakranna of the Samhita into cakran na, which really should have been cakram na. 83. GELDNER in his Fed. Stud. 1.279 draws, and rightly, a different conclusion-" Ableitung von krand verbietet die nachste Strophe." His view regarding the meaning of cakran as expressed there is, however, unacceptable. Madhu Vidya/12

Page #38
VEDIC AKIKHALA: PALI AKKULA BY DR. M. A. MEHENDALE, DECCAN COLLEGE RESEARCH INSTITUTE, POONA, The vocable akhkhala is taken from the expression akhkhati. kotya which is known to occur only once in the Sanskrit literature, It is found in the third verse of the famous frog-hymn of the Rgveda (VII. 103). It runs as yad im ena usato abhyavursit trsyavatah pravisy agatayam akhkhalikitya pitaram na putro anyo anyam upa vadantam eti // Geldner translates-"Sobald es auf die Gierigen, Durstigen bei Eintritt der Regenzeit geregnet hat, kommt unter lautem Quaken einer auf den Zuruf des anderen zu wie der Sohn zum Vater," Sayana explains the expression as akhlchala iti sabdunukaranan akhkhalasubdam krtva| The dictionary meaning is the same *den Freuderuf akhkhala ausstossen' (Boehtlingk-Roth, also Grassmann), "uttering the exclamation akhkhala' ( Monier-Willia m s). It is thus clear that the expression is an onomatopoetic one referring to the sound of joy made by the frogs. It seems, however, possible to throw some more light on the meaning of the word by suggesting the sound intended to be conveyed by akhkhala. in the frog-hymn two kinds of frogs from the point of view of the sounds uttered by them and their colour--are referred to. They are gomayu uttering sounds similar to that of a cow' and ajamayu 'uttering sounds similar to that of a goat,' or prsni 'the speckled one' and harita 'the green one' (cf. verses 6, 10, also 4). It would, therefore, be interesting to see if the sound referred to by akhkchala can be more closely determined in the light of these descriptions. In the Udana' I. 7 (p. 5) we come across the expression akkulapakkula (or & derivative from it akkulapakkulika). In this account we are told that a certain Yaksa, Aja kalapaka by name, wanted to terrify the Buddha who was sitting in his cetiya. He 1. Ed. by Paul Steinthal, London, 1885. Madhu Vidya/13

Page #39
13 DR. M. A. MEHENDALE (Sec. seems that ther half is just ce has been rights not guessed was therefore made a sound akkulapakkula thrice. On this peculiar expression R. Morris remarks, "Akkulo pa k k ulo are merely instances of an interjectional use of the words (alcula vyakula,) something like the employment of "death and destruction", "fire and fury", to convey the notion of something fearful, in lieu of imitative words." The PTS Dictionary also equates akkula with akula and gives the meaning 'confused, perplexed, agitated, frightened.' The same dictionary derives pakkula or pakula from pa (i.e. pra)+kula 'in great confusion.' But the correct meaning of the word has been already given in the commentary as an onomatopoetic sound uttered by the Yaksa to terrify the Buddha. Cf. tayo vare "akkulo pakkulo' ti bhinsaputukamataya evaru par saddam ukasi. Anukaranasaddo hi ayam. This has been accepted in V. Trenckner's A Criticul Pali Ditionary - Vol. I, which says, "probably a shout to one in order to terrify." It seems that in the compound expression akkula pakkula (or -balkula), the latter half is just an extension of the base akkula in reduplication. Though the base has been rightly explained by the commentator as an onomatopoetic word, he has not guessed whoso sound the Yaksa may have tried to imitate by an utterance like akula. A solution to this problem is, I think, contained in the name of the Yaksa, viz. Ajakalapaka, The commentary referred to above contains (p. 64) two explanations of this name, both of which seem to be fanciful. According to the explanation of the commentator the Yaksa was so called because he liked the gouts brought to him for offering bound in a bundle (aju-kilapaku). Cf. 80 kira yilck ho aje kulipatva bandhunen'i aja-kotthazuna sad lhin laim parichhati no ansiatha. . According to the other explanation offered by some others the Yaksa was so called because he compelled the people who brought him offerings utter crics like a goat (ajal la. paka). Cf. keci pani "ajake viya satte laprti" ti Ajalu.lagako ii. Tassa lira satta baliji u ponetva yadii aju-saddam katra balin upaharanti tadu 80 tussuti. T'usma Ajala-lapako ti vuccati ti. But on the basis of the story itself the correct explanation appears to be that the Yaksa was given the name Ajaka-lapaka* because he himself used to produce sounds similar to those of a goat in order to frighten the people. The second explanation referred to above has missed the point because in it lapaka has been unnecessarily treated as derived from the causal base. But the 1. atha kho Ajakal 7 pako yakkha bhagavato bhayan.... uppid-tukamo (sic.) .bhagavato avidure tikkhattur akkulopakkulo ti akkula pakkulika akisi. 2. JPTS 1886 pp. 94-95. 3. Paramatthadi pani Udinacithakathi of Dhammapilacariya, Ed. by F.L. Woodward, London, 1926. p. 66. 4. Another Yaka has tho nano Ajakilala in one Bharhut inscription (Luders list No. 795). But this has a difforont explanation. Madhu Vidya/14

Page #40
VEDIC ANHKHALA: PALI AKKULA 14 o figure roshten the Bud Sound alkula".es vrddhi of the penultimate u of the root has been taught by Panini in the formation of the agent nouns with the suffix -alcu (cf. 3. 1. 33 and 7.2. 116). It will thus be clear that Ajakalapaka means 'one who utters sounds like a goat' and that the sound alckula attributed to the Yaksa in order to frighten the Buddha is an attempt on the part of the author to figure out the imitation of the bleating of the goat. If this Pali akkula 'bleating of a goat' is connected with Vedic alhkhala, then the latter may also be regarded to represent the cry of that frog," which is supposed to utter cries similar to those of a goat (ajimayu). The loss of aspiration in allula is witnessed in many other instances in Pali ( Geiger 62, rlsa >ilka, etc.). Moreover it may also be noted that for akkula a variant ackhula with aspiration has been recorded. As regards the difference in the vowel in the second syllable (a : u), it is likely that the change occurred as aklila and was confused with akula. Above all it must be remembered that such minor changes are very likely in onomatopoetic expressions. The return to the frog-hymn once again. Gomayu and ajamaya varieties of the frog on the one hand and prsnt and harita varieties on the other are referred to in verses 6 and 10 and in both cases exactly in the same order. This suggests that the speckled variety produced a sound siinilar to that of a cow and the green one produced a sound similar to that of a goat. From the meaning of alchichala suggested above, this seems to be confirmed by verses 3 and 4. In verse 3 we are told that when the rain-water falls, the one frog approaches the other producing the sound a rhakhla. If, as suggested above, the frog producing this sound is ajamiyu, then it would mean that in verse 3 we read 6 In a stanza which occurs in tho Subhisitaratnabhindrigira (Nirnaya. Bigara edn. 1952, p. 207, verse 18) the goat's cries are represented as bobokaran prakurvan. 7. In the Pancaviira Brihmana 12.4.16, the crying of a frog is referred to as al karoti, Otherwiso in tho Voda, the verb use I is vad with or without a pretix like a, pra, u. Cf. V. 7.103.1, 3 etc., 10.166.5 ; AV. 4.15.12,14). In Vagbha a's commentary to his Kivy inufii sana (ch. 2, p. 24, Kavyam. 1 43) whero cortain sounds of animals and objects aro montioned, it is said ravo menu nke u. In modern Indian languages also thoro are different expressions roforring to the crios of a frog : As. !ortorowa, Bg. geigorgengor, Hindi-Urdu ariar Dakhani form as given in J. Shakespeare's Hindustani-English Dictionary is turinturm), Pjb.le.in tein or trein tren Mar. Guj. dari dariv, Kan. vatarala, Tam, vrat vrit, Tel. bekabeka. 8. akichulo bhakkulo is noted in JPTS- 1866, p. 99 (explained as akkheluri khepetu vinisette i ulati parattatiti akkuls, bhakkhituin ulatiti bhakkhulo......). Tho varianty noted in tho toxt of the Udana aro (p. 5. f. n. 1) akkulobakkulo, uklulavakkulika.it. The Gatha 7 referring to this incidont has Oakkulair. In tho Prs Dictionary we are asked to read p kula both for pakkula and bakkulu. 9. Tho text as given by R. Morris (JPTS 1886, p. 94) shows the reading with a in pakkalo (line 3;; but he remarks that we should read pakkulo (line 9). Madhu Vidya/15

Page #41
15 DR. M. A. MEHENDALE [Sec. oachingckled ow, of from the one about ajamayu approaching the gomayu, or presumably the green one approaching the speckled one with a sound of joy. The following verse (4) tells us that now, of the two, the one supports the other in the utterance of cries, and from the context it would seem that now it is the turn of the speckled one to suppout the green one with its utterance of joy. This is happily confirmed by the second half of the verse which expressly tells us "manduko yad abhlvrstah kaniskan prsnih samprakte haritena vacam!"......"wenn der Frosch im Regen hin und her hupft, und der Gefleckte seine Stimme mit dem Grunen vereinigt." ( Geldner). Madhu Vidya/16

Page #42
YATUMAVANT By M. A. MEHENDALE, Poona The word yatuma'vant is attested only in the Rgveda where it occurs four times (1.36.20, 7.1.5, 7.104.23, and 8.60.20). According to the dictionaries yatuma'vant is identical in meaning with yatumant practising witchcraft or sorcery': * Th. BENFEY! (Orient und Occident 1, 1862, p. 387 note 328) rejects Sayana's explanation of this form and proposes to take mavant as an 'organischere Form' of mavan, the latter coming from the verb V man and the suffix -van. GRASSMANN Worterbuch, under stavat) regards -mavat as a double secondary suffix arising out of -man-vant with the loss of n and the consequent lengthening of the preceding vowel a (cf. WACKERNAGEL I S 42). H. H. BENDER (The Suffixes mant and vant in Sanskrit and Avestan, Baltimore, 1910) considers (p. 59) the formation of the word unclear, but looks upon the suggestion of GRASSMANN as plausible, if not convincing. Later on p. 81, f. n. 2 he seems to accept GRASSMANN's suggestion. Louis H. GRAY (Anzeiger fur Indo-germanische Sprach- und Altertumskunde 30, p. 9, 1912) also considers the derivation of yatuma'vant < * yatu-manvant as correct. WACKERNAGEL-DEBRUNNER seem to give two different explanations of this form. In Alt. Gr. II, 2 & 620 they say that yatu-ma'vant is only a Vedic variant for yatu-mant used to obtain a metrically agreeable end of a pada. But in $ 713 b8 they cite it under "Stammerweiterungen" and appear to regard yatu-ma- as an extended base to which is added the suffix -vant. Since they do not give any explanation of the extended base, it is not clear whether they approve of GRASSMANN's suggestion or reject it and have some other explanation in mind (see below). Sayana appears to offer three different explanations of the form. While commenting on RV 1.36.20, where the word occurs for the first time, he takes only -vant as the possessive suffix, and explains the stem yatumaas coming from yatu + V ma? meaning 'activities of demons' (yatavo 1. I am greatly indebted to Prof. W. Wust for sending to me BENFEY's opinion on this expression as well as for giving me the reference to GRAY's opinion referred to below. 2. I do not understand how BENDER (p. 59) says that "Sayana derives yatuma'vant from yatu-ama and the secondary suffix vant." Madhu Vidya/17

Page #43
M. A. MEHENDALE yatanah / tan mimate nirmimate iti raksasavyaparah yatumah / tad esam astiti matup /)s. Later while commenting on kV 7.1.5 Sayana offers two other explanations of the same form. In the first instance he takes yatuma'van as nom. sg. of a base ending in -van (cf. above BENFEY's suggestion) like ya'va from ya'van which occurs in the same line of this verse.* Sayana adds that in that case the absence of the loss of final -n may be looked upon as a Vedic peculiarity (nalopabhavas chandasah). How in this case Sayana interprets yatumavan as himsayah nirgatah is not clear to me.5 In the alternative explanation given here, he seems to take yatuma- as coming from yatu-mat to which the possessive -vant is added as an expletive. Cf. yad va himsayuktah/ paro vatir matvarthiyah purakah. If this interpretation of Sayana's commentary is correct then he was the first to have considered the possibility of explaining yatuma'vant as coming from a double suffix yatu-mat-vant (cf. above GRASSMANN's suggestion). In this paper it is proposed to offer for consideration two explanations of the form. In the first instance we may accept WACKERNAGEL-DEBRUNNER'S suggestion to regard yatu-ma as an extended base and explain it as the noun yatu'- with the secondary derivative -ma meaning "consisting of magic, having the character of magic' (cf. Alt. Gr. II, 2 g 598a where we have instances like druma-, -rgma etc.). Or yatuma- may be regarded as an abridged form of yatumaya-6 To this yatuma- is added the possessive -vant 'one working with something magical or something consisting of witchcraft'. But a better explanation of the form is perhaps to be found in verses 23 and 24 of RV 7.104. The first two quarters of verse 23 read as ma' no rakso abhi nad yatuma'vatam apocchatu mithuna' ya' kimidi'na / GELDNER construes yatuma'vatam and raksah together and translates : "Nicht soll der Zauberer boser Geist uns beikommen, (die Morgenrote) 3. While commenting on RV 7.104.23 Sayana simply renders yatuma'vatam by yatanavatam raksasanam and on 8.60.20 he says yatur yatand pida tadvatam yatudhananam. 4. Ta cam ga"tarati gatuma'san. 5. Is nirgatan a mistake for nirmata (maker, creator)? cf. Sayana's rendering of the text word ya'va as abhiganta. In that case ma'van, like ya'van, can be explained as V ma + primary suffix -van. While commenting on RV 1.36.20 Sayana has already added the prefix nir- before y ma (tan mimate nirmimate). 6. It may be pointed out that WACKERNAGEL-DEBRUNNER (II, 28 598a) consider it likely that chandoma- is an abridged form of chandomaya. Madhu Vidya/18

Page #44
YATUMA'VANT 3 soll das Kimidinpaar wegleuchten". But I think Sayana is justified in separating raksah from yatuma'vatam and construing the latter with mithuna'. The justification for this construction is to be found in the fact that raksas (neuter)? is fairly often used in the sense of raksas (mas.) (GRASSMANN's second meaning Beschadigung, personlich gefasst') and mentioned independently in this hymn (cf. verses 1, 4, 13, 22). Moreover in verse 25 it is distinguished from those who work with yatu' cf. raksobhyo vadham asyatam asanin yatumadbhyah. GELDNER: "Schleudert die Waffe nach den bosen Geistern, den Schleuderstein nach den Zauberern !" It seems, therefore, proper to take raksah in the sense of demon' and separate it from yatuma'vatam in verse 23. Further it is very likely that Sayana was led to his construction of separating yatuma'vatam from raksah and taking it with mithuna by what follows in the immediately following verse (24). The first two quarters of this verse read as--i'ndra jahi pu'mamsan yatudha'nam uta stri'yam mayaya sa'sadanam / GELDNER: "Indra, erschlag den mannlichen Zauberer und seine Frau, die auf ihr Blendwerk pocht!" This line thus refers to a man associated with yatu' and his wife associated with maya'. And if this is the sort of couple that is referred to in the preceding verse then yatuma'vant can be justifiably construed with mithuna' in verse 23. In that case yatuma'vant may be looked upon as an abridged form of yatumaya'vant meaning 'those engaged in magic and deception'. However, it must be remembered that this division of the use of yatu' and maya' is by no means rigid, for we also find men working with maya' and women with yatu'. Similarly it is not necessary that yatuma'vant should always refer to couples. It may as well refer to a single person practising magic and deception as for instance in yatuma'van (nom. sg.) in RV 7.1.5. The expression yatuma'vant has been compared with Vedic silamavant 'stromreich, wasserreich' deriving it from sira' 'stream, water' cf. GRASSMANN, Worterbuch s. v.) and Avestan gaomavant mit Milch versehen, Milch enthaltend'. In both these cases -mavant is supposed to arise from the double suffix -man-vant (cf. BENDER P. 81, f. n. 2 and GRAY, Anzeiger 30, pp. 8-9). But the similarity in these forms may be apparent and the explanation for the one need not be applicable to the other. Thus about silamavant WACKERNAGEL-DEBRUNNER do not seem to agree with 7. In fact according to GRASSMANN'S Worterbuch raksas (neut.) is used in the sense 'Beschadigung'only in two instances, whereas in the majority of cases it is used as identical with raksas (mas.). Of these two instances of GRASSMANN, one is 7-104-23 discussed above, the other is 8-60-20 where both meanings are likely-harm (neut.) or demon (mas.). Cf. ma' no raksa a vesid aghrnivaso ma' yatu'r yatuma'vatam / Madhu Vidya/19

Page #45
4 M. A. MEHENDALE GRASSMANN's explanation referred to above. They deal with it separately ($705 a) from yatuma'vant considering (with a question mark) silama as the name of a shrub. cf. RV. 10.75.8 (si'ndhuh) u'rnavati... silamavati "der an Wolle und Silamastauden (?) reiche Strom." As regards gaomavant it has been already suggested that its formation was influenced by haomavant with which it is found in a majority of cases (cf. BARTHOLOMAE, Altir. Worterbuch s. v.)." Thus there seem to be three possible ways of explaining yatuma vant: (1) Stem yatu- + mavant < double suffix man-vant. This explanation does not offer any ground for the use of the double suffix; (2) Stem yatuma- (from yatu + secondary derivative -ma, or as an abridged form of yatumaya) +vant. This explanation does not offer any special purpose for the stem extension; (3) yatuma'vant an abridged form of yatumaya'vant. Although this explanation suffers from the supposition of an abridgement, it is suggested by the context and gives satisfactory meaning as shown above. 8. Also GELDNER: 'reich an Silamakraut'. 9. The expression antarva'vant 'geraumig' is also cited by WACKERNAGELDEBRUNNER (II, 2 SS 713 b) in this connection. But in SS 701 a (small type on p. 869) they declare antarva'- as not clear. OLDENBERG (Noten), while commenting on RV 1.40.7, expresses his disagreement with PISCHEL (Ved. Stud. 2.214) as regards the meaning of antarva'vant. About its formation he observes "Stammbildung wie bei yatuma'vant," but does not make his point more clear. For the time being I am inclined to explain the expression as antarva + vant (in which antarva is formed with the secondary derivative -va meaning 'having interior, wide space'. [Reprinted from the Bulletin of the De Madhu-Vidya/20 Madhu Vidya/20 itute S. K. De Felicitation Volume. I

Page #46
SATYAM EVA JAYATE NANRTAM M.A.MEHENDALE Deccan College, Poona The above passage is quite well known. It occurs in the Mundaka Upanisad (3.1.6) of the Atharva Veda, and the first part of it has now been inscribed as the motto of the Indian nation. The passage has been mostly taken to mean "Truth alone conquers, not falsehood." In the above interpretation satyam and anrtam are taken to be the subjects, but this does not seem to be correct. Both satyam and anstam have to be regarded as the objects, and a rsi is to be understood as the subject. Taken this way, the sentence would mean "A sage obtains only the Real (i.e., the Brahman), not the unreal." This construction was already seen by Deussen? who translates "Wahrheit ersiegt er (i.e. the ativadin cf. Chand. 7.16), nicht Unwahrheit." This interpretation will be found to be in harmony with the spirit of the Upanisads in general and that of the Mundaka in particular, According to these philosophical texts the highest goal of a sage is to obtain unity with brahman which is the ultimate Reality or satyasya satyam. Whatever is lower than this satya is anrta or unreal, and a sage does not seek after that. About the real and unreal forms of brahman we read in the Maitri 6.3: dve vava brahmano rupe martam camurtam cal atha yan murtam tad asatyam yad amurtam tat satyam / tad brahma taj jyotih. It will be useful to cite here in full the stanza from the Mundaka (3.1.6) in order to be able to understand the context and appreciate the correctness of the above interpretation. satyam eva jayate nanrtam, satyena pantha vitato devayanah/ yenakramanty rsayo hy aptakama, yatra tat satyasya paramam nidhanam/ Here in the last three quarters of the stanza we are told that the heavenly path by which the sages go is laid out by the Real, and the place where the sages reach is the highest abode of the Real. Therefore, when the subject matter is the path taken by the sages to reach the abode of the Real, it would be improper to give a worldly meaning like 'truth alone conquers' to So, for example, Hume, The Thirteen Principal Upanishads. Max Muller (SBE 15), "The true prevails, not the untrue." Radhakrishnan, The Principal Upanisads, "Truth alone conquers, not untruth." Hillebrandt, Aus Brahmanas und Upanisaden. "Die Wahrheit allein siegt, nicht das Unrecht." Sechzig Upanishads des Veda(Istedition, 1987). The same interpretation is also intended by J. Hertel, see below pp.407. Also cf. Br. 2.3.1 dve vava brahmano rupe murtar caivamurtam ca, martyam camstar ca, sthitam ca yac ca,sac ca tyac ca. Madhu Vidya/21

Page #47
406 MEHENDALE: "Satyam Eva Jayate Nanstam" satyam eva jayate. In the context it can only mean "A sage obtains only the Real" because the place where he reaches is the abode of the Real. He does not obtain the unreal or lesser worlds because there are other paths which lead to them which the sage does not take. With the expression satyam eva jayate which means ' A sage obtains only the Real (i.e., brahman),' we may well compare Mundaka 3.2.9 (sa yo ha vai tat paramam brahma veda) brahmaiva bhavati. The word satyam is often used in the Upanisads to designate brahman as the ultimate Reality. In the Chandogya, Uddalaka Aruni taught Svetaketu the doctrine of the essential oneness of the individual and the universal soul. There this highest principle is declared to be satyam : sa ya eso'nima, aitadatmyam idam sarvam, tat satyam, sa atma, tat tvam asi Svetaketo, 6.8.16. Before Svetaketu was initiated in this doctrine he was asked one question by Aruni. While elaborating this question Aruni uses satyam in the sense of basic Rality: ekena mrtpindena sarvam msnmaya vijnatam syat vacarambhanam vikari namadheyam mrttika ity eva satyam / ...... loham ity eva satyam, etc., 6.1. In this Upanisad satyam is expressly said to be the name of brahman: tasya ha va etasya brahmano nama satyam iti, (8.3). In the Mundaka itself, where the passage under discussion occurs, the nature of the brahmavidya is said to be that by which one knows the imperishable Purusa as satya : yenaksaram purusam veda satyam provaca tam tattvato brahmavidyam. 1.2.13. In a few places we find it further stated that this ultimate Reality is covered by a gold vessel (in the form of the sun): hiranmayena patrena satyasyapihitam mukham / tat tvam Pusann apavinu satyadharmaya drstayell, isa. 15; Bp.5.15. In the usual interpretation of styam eva jayate, satyam is taken as the subject. But before we do so it would be well to remember that since satyam is an attribute of brahman it is never employed as subject in the Upanisads. In one passage of the Bshadaranyaka (5.5.1) it may appear at first sight that satyam is used as a subject. But a closer examination will show that that is not the case. We read there as follows in an account of creation :apa evedam agra asuh / ta apah satyam asrjanta / satyam brahma brahma prajapatim prajapatir devan. It may appear that here satyam is said to create brahman. That is, however, not true. In the preceding section (Bt. 5.4) both satyam and brahman are identified and described as first-born : sa yo haitam mahadyaksam prathamajam veda satyn brahmeti ... * We may also compare tasyaisa atma visate brahmadhama Mund.3,2,4; tan....brahmalokan gamayati ....tesar na punaravstrih,Br. 6.2.15;sa enanbrahma gamayari esa devayanah panthaiti, Cha.5.10; also4.15. 'Also, cf., atha namadheyam satyasya satyam iti, Bp.2.3.6.also 2.1.20. "Also cf., tad etad aksaram brahma ... tad eta! satyam, tad amstam ..... Mund.2.2.2. Madhu Vidya/22

Page #48
MEHENDALE: "Satyam Eva Jayate Nanstam" 407 satyam hy eva brahma. It will, therefore, (or atman). tesam evaisu brahmaloko be clear that in the above creation account yesam tapo brahmacaryam yesu satyam satya and brahma can both be first-born if pratisthitam/ ... na yesu jihmam anstan na the two words are understood to be in maya ceti, Prasna 1.15-16; nayamatma apposition. We have, therefore, to pravacanena labhyo na medhaya na translate the passage as: "In the beginning bahuna srutena, Mund. 3.2.3." But since this world was just Water. That Water in our verse the context is the goal reached emitted the Real-Brahma (being) the Real, by a sage, satyam should be taken to mean Brahma, Prajapati, Prajapati, the gods" brahman and not 'true speech.' (Hume). The verb V ji is used both in the sense It is not intended to suggest here that in - 'to win, to obtain'as well as 'to conquer, the Upanisads satyam is employed only in to be victorious' since the earliest times. In the sense of the ultimate Reality, i.e., the usual translation of satyam eva jayate, brahman. It is true that it is also used in the the latter meaning is thought of. But the sense 'truth, true speech.' Thus in the former meaning 'to obtain,' which is advice given by the teacher to his pupil we intended in the suggested interpretation, is read: satyam vada / ... satyan na also to be found in the Upanisads. For pramadita vyam, Taitt. 1.11.1 In the instance, we often come across Chand. 1.2.3 we find tasmat taya expressions like lokar jayati, salokatam ( = vaca) ubhayam vadati satyam jayati. In the Mundaka itself (3.1.10) we canstam ca. A sage has to lead a life of read tam lokam jayate tams ca kaman good conduct and therefore we find where the meaning obviously is 'obtains.' satyam 'true speech' also included in the The above discussion will show that means of obtaining brahman (or atman). the interpretation of the passage under Thus Mund. 3.1.5: satyena labhyas tapasa discussion, viz., 'A sage obtains only the hyesa atma, samyagjnanena Real and not the unreal'is in keeping with brahmacaryena nityam, Svet. 1.15 the Upanisadic usage of the words satyam satyenainam tapasa yo'nupasyati. and ji. Occasionally we also get a mention of While commenting on this passage those means which do not lead to brahman Sankaracarya observes: satyam eva satyavan eva jayate jayati, nanstam ? Also,cf., Chand.6.16; Brhad.5.14.4. nanstavadity arthah / na hi satyanstayoh yas tu vijnanavan bhavati samanaskah sada sucih kevalayoh purusanasritayor jayah sa tu tat padamapnoti yasmad bhuyo na jayatellparajayo va sambhavati / prasiddham loke Katha.3.8. satyavadinanstavady abhibhuyate na For other references giving means to obtain viparyayah / atah siddhamm satyasya brahman, cf., tasmad vidyaya tapasa cintaya copalabhyate brahma, Maitri.4.4. 10 Also,cf., Katha 2.23-24 and Mund.3.2.4. Madhu Vidya/23

Page #49
408 MEHENDALE: "Satyam Eva Jayate Nanstam" balavatsadhanatvam. This will show that have already obtained their desires Sankara also found it difficult, although (aptakamah) and their desireless condition on different grounds, to take satyam by is also emphasized in the Mundaka 3.2.2 itself as subject. But since he accepts : paryaptakamasya krtatmanas tu ihaiva satyam = satyavadi purusah as subject sarve praviliyanti kamah. They live in and renders jayate 'is victorious'his is the forests and finally reach the immortal usual interpretation of the passage Purusa : tapahsraddhe ye hy upavasanty assuring worldly victory to the truthful aranye santa vidvamso bhaiksacaryam man. In the opinion of Sankara this has carantah / suryadvarena te virajah been said in order to praise truthful prayanti yatramstah sa puruso hy behaviour as the most efficacious means. avayayatma, Munda. 1.2.11. If we keep in But it seems unnecessary to single out mind the background of the Mundaka Up. satyam, as 'true speech,' for special praise and the final attainment aimed at by the having included it with austerity, etc., in ascetics, it would be proper for us to the preceding verse among the means to interpret satyam eve jayate as "A sage obtain atman. The Mundaka is specially wins only the Real." intended for ascetics" who wish to attain It is possible to anticipate a few the Imperishable (tad aksaram objections to the above interpretation. In adhigamyate 1.1.5). It is not intended for the first instance it may be argued that we ordinary people of the world who may be may expect an object for the verb Vji if it seeking unreal ends and who need such is used in active voice. If used medially, inducement to be persuaded to follow the i.e., reflexively, no object is necessary. For truth. While the Imperishable is obtained example we find in the Ait. Br. 12.16, by the higher knowledge (para vidya) jayati used with an object, but jayate worldly victories and such other things without it: tathaivaitad yajamano ... would belong to the field of lower jayati svargam lokam, vy asmil loke knowledge (apara vidya). The ascetics jayate. "Similarly the sacrificer wins the heavenly world , he is victorious in this "It appears that the doctrine propounded in the world."Since in our passage jayate is used Mundaka was specially intended for those who in the middle voice it would be proper not shaved their heads and who with controlled minds to anticipate an object and translate 'truth approached the teacher for instruction. Mundaka alone is victorious.' 3.2.10: tesan evaitam brahmavidyam vadeta The objection can be answered in two sirovratam vidhivad yais tu cirnam; also Mund. 1.2.13: tasmai sa vidvan upasannaya samyak ways. First we have to pointout that the uses prasantacittaya samanvitaya / yenaksaram in the middle voice are notalways reflexive. purusam veda satyam provaca tam tattvato In the Mundaka itself we find pasyate used brahmavidyam//Cf. Hertel, Mundaka-Upanisad twice in the active sense : yada pasyah (Leipzig, 1924), p. 19. pasyate rukmavarnam kartaram i sam Madhu Vidya/24

Page #50
MEHENDALE: Satyam Eva Jayate Nanstam 409 purusam brahmayonim,3.1.3; tatas tu tam pasyate niskalam dhyayamanah, 3. 1.8. Even jayate is used unmistakably in the active sense in the Mundaka : tam tam lokam jayate tams ca kaman, 3.1.10. So there should be no objection if in our passage jayate is taken in the active sense with satyam as object and rsi as the subject supplied. Secondly the reason for the use of jayate instead of jayati appears to be in the metre of this Upanisad. From the analysis of the tristubh metre of the Mundaka done by Hertel' it become clear that if of the three parts of a quarter the first one has four syllables and the middle one three, then the latter has never all three short syllables. They are either -----, or ... Therefore when in our instance the line opens with the first division of four syllables (satyameva), we cannot have the middle part with all short syllables. Hence we find the use of jayate( ) instead of jayati( ).Ifthen the use of jayate for jayati is metri causa, there should be no difficulty in understanding the middle form for the active form and take satyam as the object of jayati. Moreover, since this quarter is metrically defective, in the opinion of Hertelone syllable at the end has probably been lost. He suggests to read the quarter as satyam eva jayate, nanrtam sah (pp. 59 and 44). If this is correct, obviously satyam has to be taken as object, and in that case our view will be supported. But we need not rely only on this evidence since it involves emendation. second objection could be that since in the first quarter of the verse jayate is used in the singular number,'* the subject rsih to be supplied has also to be in the singular. But in the third quarter of this verse we find rsayah in the plural. Therefore it would not be proper to assume a subject in the singular number in the first quarter. But such differences in number are not altogether rare. In the Mundaka itself we notice them in the following verses: sa vedaitat paramam brahmadhama yatra visvam nihitam bhati subhram/ upasate purausam ye akamas te sukram etad ativartanti dhirah// 3.2.1 etair upayair yatate yas tu vidvams tasyaina atma visate brahmadhamal samprapyainam rsayo jnanatrptah kstatmano vitaragah prasantah// 3.2.4-5. Thirdly a point may be raised that in the Upanisads we do not come across elsewhere an expression like rsir brahma jayati. This is true. But instead of jayati we find verbs like Vlabh-, V vind-, Vap-, Vas- used in expressions such as satyena labhyah ... atma (Mund 3.1.5); nayam atma pravacanena labhyo (Mund. 3.2.3); tasmad vidyaya ... upalabhyate brahma (Maitri 4.4); brahmacaryena... atmanam anuvindate (Chand. 8.5); tad ya evaitam brahmalokam brahmacaryenanuvin 12 Mundaka Upanisad,p. 28. "I now withdraw my earlier suggestion of 13 This is also true of tam tam lokam jayate, Mund. ragardingjayate as a possible plural form.Cf.Ind. 3.1.10. Linguistics, 17.23. Madhu Vidya/25

Page #51
(Chand. 8.4); brahmapraptah (Katha 6.18); sa tat padam apnoti yasmad bhuyo na jayate (Katha 3.8); atra brahma samasnute (Katha 6.14.Br.4.4.7); amrtam asnute (Isa 11, 14; Prasna 3.11). About the use of jayati itself we may cite the following passages. In the Chand. 2.10.56 while explaining the mystical significance of a sdman it is said that with the twenty-one syllables the knower obtains the sun and with the twentysecond he wins (jayati) what is beyond the sun, viz., the sorrowless heaven (ie. the brahma world). Cf. ekavimsatya adityam apnoti..dvavimsena param adityaj jayati tan nakam tad visokam. Thus here jayati is used for obtaining what is beyond the sun. Sometimes the sun itself is identified with the final goal and to describe the attainment of this the verb Vji is used. Cf. Prasna 1.10: athottarena tapasa brahmacaryena sraddhaya vidyaya atmanam anvisya adityam abhijayante / etad vai prananamayatanam etad amrtam abhayam etat parayanam etasman na punar avartanta iti. Since in this statement coming from an Uoanisad of the Atharva Veda we have a mention of the means for seeking atman which is followed by the expression adityam abhijayante, it would be instructive to read here Mundaka3.1. 15 Cf., Mundaka 1.2.11 where the ascetics are said to reach the immortal Purusa through the gate of the sun (suryadvarena te virajah prayanti yatramrtah, sa puruso by avayaydima) Aslo.cf.. hiranmayena patrena satyasydpihitam mukham, cited above, and sa tejasi sdrye sampannah...sa samabhir unniyate brahmalokam .... Prasna 5.5. 5-6 where we also find mentioned almost the same means like tapas, etc., for the obtaining of atman which is immediately followed by satyam eva jayate. This comparison between adityam abhijayante and satyam jayate should not leave any doubt about the correctness of taking satyam as the object of jayate.16 One thing should be made clear in the end. It has not been assumed in the above discussion that satyam eva jayate cannot at all mean at any place 'truth alone conquers.' If the sentence is used in some different context where this sense in intended it can certainly fulfill that purpose. What has been demonstrated above is the fact that this meaning is unsuitable in the context where it occurs in the Mundaka Upanisad. There we are told about the means for ralizing the self and about the reaching the abode of the ultimate Reality by devayana. In this context the expression should be taken to mean (The sage) wins only the Real, (and) not the unreal." It may also be pointed out that in the Kausi.Up.we find a dialogue between Brahma and the knower of brahman. At the end of this dialogue (1.7) we read sd yd brahmano jitir ya vyastis tam jitim jayati, tam vyastirh vyasnute. It is true that the expression brahmano jitih can mean both 'the conqest made by brahman' or :the winning of brahman.'For the former we may compare tasya ha brahmano vijaye (conquest made by brahman) deva amahiyanta, Kena 3.1, also 4.1. For the latter we may compare apnoti ha adityasya jayam paro hasya adityajayd(winning of aditya) jayo bhavari, Chand.2.10. Madhu Vidya/26

Page #52
NOTES AND COMMUNICATIONS TWO DERIVATIVES IN YA pajasya This word is not very frequent in its occurrence, and its usage is restricted to the Vedic literature. It is obviously derived from pajas- which occurs more often and in usage is restricted to the Rgveda. The meaning assigned to the former is belly, and the latter has been variously interpreted as 'brightness', 'form or appearance, shining form '.1 The connexion between these two words, from the point of view of their meaning, had remained incomprehensible for a long time. A good explanation has been suggested by H. W. Bailey in BSOAS, XII, 2, 1948, 323-26, also in the University of Ceylon Review, xv, 1-2, 1957, 29, and TPS, 1955, 55, n. 1. On the basis of certain cognates from Khotanese and Sogdian, Bailey reconstructs a Middle Iranian paza(h)- meaning generally surface', and when specialized be thinks it means 'face'. Skt. pajas- then could be related to this *paza(h)and its meaning determined as 'surface, face. As noted by Bailey (BSOAS, XIII, 1, 1949, 136), E. Sieg had already assigned the meaning 'surface' to Skt. pdjas in Der Nachtweg der Sonne nach der vedischen Anschauung (1923), pp. 5-6. From pajas-surface' is derived pajasya- belly' as under surface'. Renou in Et, ved. et pan., III, 1957, 68, raised objections to the above view, especially because he felt that pajas- surface, face' would not be suitable when used with Soma. He, therefore, proposed forme', 'masse' as the meaning, which, he thought, would be applicable in all occurrences of pajas. However, in the next volume (tv, 1958, 54) he found that when confronted with tripajasyd-, an epithet of vrgabha-, the word can neither simply mean forme' (Renou) nor surface (Bailey), but that it ought to mean some part of the body." In Et. ved. et pau., VIII, 1961, 62, Renou has once again referred to the difficulty about pajas and given expression to his opinion that the word should mean something more definite than 'form'. He wonders whether pajas is analogous to vareas brilliant form, face'. Now pajasyd was already interpreted as 'belly', a part of the body. But it seems that it is necessary to understand the word somewhat differently, although still referring to a part of the body, when one considers the parallelism which occurs in the Brhadaranyaka Upanisad. There in the opening sections we read about the identifications of the various parts of the body of the sacrificial horse as follows: dyauh prstham antariksam udaram prthivi pajasyam The 1 As early as 1889 Geldner, Ved. Stud., 1, 116, remarked, 'pajas gehort zu den bekannten Wechselbegriffen, welche bald Licht, bald Schnelligkeit und Kraft bedeuten. Much later in 1917 (ZDMG, LXXI, 338) he gave the following meanings for pdjas: (1) Gestalt, Aussehen, Farbe, Erscheinung'; (ii) Lichtgestalt, Lichterscheinung, Farbenglanz'; (iii) pdjas = tejas. Most of these he adopted variously in his translation of the Rgveda. Because tripajasya- occurs by the side of tryudhun- and tryanika-. Also Sat. Br. 10.6.4.1. Madhu Vidya/27

Page #53
598 NOTES AND COMMUNICATIONS having three breachasya This may and AV 9.7.53 krodas sky his back; the atmosphere his stomach; the earth his pajasya'1 (1.1.1), and a little later: dyauh prstham antariksam udaram iyam urah (1.2.3) 'The sky his back: the atmosphere his stomach ; this one (i.e. the earth) his chest'. From this parallelism between the two statements it seems possible to explain pajasya- =uras- chest, breast' and not belly'. The word tripajasyareferred to above (Rg. 3.56.3), can then be rendered as having three breasts '. 2 In TS 5.7.16, VS 25.8, and AV 9.7.5 3 kroda- and pajasyce- are mentioned side by side. This may seem to go against the view that pajasya- means breast ' because, according to the lexicons, kroda- also means 'breast, chest'. But as the commentator Mahidhara on VS 25.8 points out kroda- means the middle part of the breast * (cf. krodah ... vaksomadhyabhagah). That krodais in some way different from breast, as the front surface of the body, is also shown by the Horasastra (1.4) where it is mentioned alongside uras- 5 For pajas- Bailey has suggested the meaning 'face', besides 'surface'. But it seems it would be safe to assume that pajasyd- and pajas-, like asyd- and as-, have the same meaning 'breast'. It is more casy to understand pajas breast' as a specialized meaning from 'surface which Bailey has assumed for *paza(h). This would also satisfactorily explain the Waxi paz, Khowar paz, and Yidgha fiz' breast'. Bailey thought of the meaning 'face' on account of certain contexts in which the Khotanese paysa- and Sogdian p'z (*paz) or Brp'(*frapaz) occur. These refer to acts of salutation and Bailey interpreted them as lying on one's face'. But 'lying on one's breast' would suit in these contexts as well. Actually the phrase 'fall on the face occurs in Zoroastrian Pahlavi where the word rod is used. If we assume that pajas- and *paza(h) mean breast', and not * face', it would mean that the Iranian tradition preserves two kinds of literary references to the act of prostration-one, the eastern, lying on the breast (*paza(h))',& the other, the western, lying on the face (ro8)'.? The meaning breast' assigned to pajas- suits very well in many of the contexts in which it occurs in the Rgveda. It will convey good meaning when used with Agni (3.29.3; 5.1.2; 3.14.1 ; 7.3.4; 1.58.5), with Manyu (10.84.3), and with Usas (3.61.5). It will be found particularly suitable when it is often Sankara explains pajasya = pidasya = pudisanasthana. ?Der Bulle... hat Arei Bauche und drei Euter... Er, der drei Gesichter bat, ..., Geldner. In the footnote he observes, pajasya ist nach AV 4.14.8; Brh. Up. 1.1.1 die untere Bauchseite des Tieros, inguen (zugleich im engeren Sinn dieses Wortes), wahrend udara die Bauchhohle bezeichnet'. For pajasyd- also cf. AV 4.14.8. * In AV 10.9.25, however, the dual forin krodau refers to the two parts of the breast (compared with purodasa). $ This has been already noted by BR. * Bailey refers to Hopkins (BSOS, VI, 2, 1931, 374) who quotes from the epic uruseva pranamuse. ? In a communication datod 5 September 1961, Sir Harold Bailey kindly informs me that Sogdian "paz occurs in a Christian text which is a translation from Syriac. Now the Syriac phrase used in such contexts does moan 'fall on face' (app - face). If our assumption Ir. #paza(1) -- broast is correct then this would mean that the Sogdian translator substituted his inode of prostration lying on the breast' for the Syrian one lying on the face Madhu Vidya/28

Page #54
NOTES AND COMMUNICATIONS 599 used with the epithet prthu 'broad'. Thus Agni is described in 3.15.1 as vi pajasa prthuna sosucano blazing forth with broad chest' (also cf. 7.10.1; 3.2.11 ; 3.3.1 ; 3.5.1 ; 3.27.5; 4.4.1 ; 3.61.2 about the horses of Usas). It will further give excellent meaning when pajasvant-1 is used as an adj. of virain RV 10.77.3: pajasvanto no virah panasyavo (Maruts) like warriors with (broad) chests, fit to be praised'. When used with a chariot prthupajas- may refer to its 'broad front' (RV 4.46.5 ; 8.5.2; 4.48.5 where only pajas- appears). In RV 1.115.5 where rusat pajah is used as opposed to krsnam pajah of Surya what is probably meant is the shining surface and the dark surface as referring to the sun's front and the back. Similarly, as suggested by Bailey (p. 326), pajas- when used in dual about heaven and earth would mean the two (vast) surfaces (cf. 1.121.11; also 1.151.1 although only the singular is used ; in 10.37.8 bphat pajah refers to the sky). When used with reference to Soma (9.68.2-3, 109.21 ; in the plural pajamsi 9.76.1 and 88.5; and in the compound sahasrapajas- 9.13.3, 42.3) it appears that probably it is the image of the horse that has led to the use of pajas- ' breast'. Similarly in RV 2.34.13 the Maruts are imagined in the form of a horse when it is said that by drinking Soma they put on shining and beautiful colour (nimeghamana atyena pajasa suscandram varnam dadhire supesasam). Only I do not understand how the following passage has to be understood : a No vayo mahe tane yahi makhaya pajase (RV 8.46.25) said of Vayu. 2. lokya This derivative is fairly well attested, although mostly in the Brahmana literature. In the Ait. Br. 2.9 (6.9) occurs the following passage in connexion with the cake offering along with the animal offering to Agni and Soma : tasmad ahuh purolasasatram lokyam iti. Sayana renders lokyam = preksaniyam. BR gives various meanings of lokya- of which statthaft, ordentlich; ublich' is assigned to the above occurrence in the Ait. Br. Keith translates the passage as Therefore they say, "The cake offering is the people's sacrificial session >> The above renderings of lokya- seem to be incorrect. The appropriate Avestan pazahvant- used of a dog may mean 'having (fine) breast'. Bailey explains (p. 325) Ossetic Digor fazae, Iron faz' back surface ' from the Middle Iranian *paza(h). Burface'. According to the information suppliecl by Bailey's Digoran friend, however, faza means side' (BS0.4S, XIII, I, 1949, 136). Bailey also quotes Ossetio fuzu, with short vowel, meaning plain' in TPS, 1955, 56. Luders, Varuna, 1, 198, does not agree with Geldner in rendering ni. migh. 88 herabhamnen' but thinks that it means 'in sich hineintraufeln, in sich hineingiessen'. + Sadgurusiaya also explains lokyam drastavyam. He adds purolasena yago yah satratulyo nirikayatam iti brahmavadino vadanti. 5 Other meanings are: (1) as adj., (i) 'Gebiet., freie Stellung gewahrend ', (ii) uber die ganze Welt verbreitet', (iii) die Gewinnung des Himmels bezweckend', (iv) ordentlich, richtig, wirklich ; gewohnlich, tagtaglich '; (2) as noun, freie Stellung! Haug translates, Thence they say: The performance of the Purodass offering is to be attended to Madhu Vidya/29

Page #55
600 NOTES AND COMMUNICATIONS rendering would be winning the world or heaven. In the lines which precede the above passage the Brahmana text explains the identity between the cake and the animal. When one gives a cake offering it is as good as having offered the sap of all animals. The result of the cake offering therefore should be the same as that of the animal offering. Hence the cake offering is declared to be lokya-' world or heaven winning'; cf. sa va esa pasur evalabhyate yat purolasah/ sarvesam va esu pasunam medhena yajate yah purolasena yajate/ tasmad ahuh purolasasatram lokyam iti/. This meaning of lokya- winning (better) world or heaven' has been already mentioned in BR, but it is made applicable only to a few occurrences. However, an examination of the passages listed there reveals that it would be quite suitable for many of the other occurrences. Thus in the Sat. Br. 2.2.3.5 it is said that by the re-establishment of the fire (punaradheya) one not only thrives (in this world) but also attains the better world (lokya). Similarly in the Sat. Br. 10.2.6.7 it is declared that one who lives a hundred years obtains immortality; therefore the life of a hundred years is called lokyn winning heavenly world, i.e. immortality In the By. Up. 1.3.28, -lokya- can only mean world-winning because it is preceded by the word lokajit. Cf. tad haital lokajid eva/ na haivalokyataya abasti ya evam etat sama veda This indeed is world-winning. There is no possibility of his not winning a world who knows this Saman'. Similarly in the Br. Up. 1.5.16 we are told that the world of men can be obtained by a son, the world of fathers by a sacrifice, and the world of gods by knowledge. In the next section (17) it is declared that all the three worlds are included in the word loka. Therefore in this section when it is said that the son who has been instructed is called lokya it means that he procures for the father all the three worlds. The Solagava sacrifice is also called lokya world-winning in the Av. Grhyasutra 4.8.35. In Mbh. 5.4103 Pitamaha tells Yayati that he had obtained heaven by world-winning' (lokya) acts." In the following few passages, however, lokya- appears to have been used in different meanings. Thus in the Sat. Br. 10.5.2.12 16kya- does not mean 'winning the world' but creating the world'. It is said that in sleep the two deities (male and female) enjoy union which leads to the creation of the world: 1 The Trivandrum edition of the Ait. Br. with the commentary of Sadgurusisya gives in the footnote this correct explanation as coming from Bhatta Bhaskara: sarvapunyalokapraptini. mittam. 3 sprhayanty u hasmai tatha pusyati/ lokyam verapi. Eggeling, however, translates and a conspicuous position (is obtained by him)'. 3 yo va satam varsini jivati sa haivaitad amrtam apnoti...lokya satayuta ity evahuh. Also cf. Sat. Br. 9.5.2.16 and 10.3.2.13 (lokyata attainment of (better) world'). so'yam manusyalokah putrenaiva jayyo... karmana pitrloko vidyaya devaloko/... ye vai keca lokas tesam sarvesam loka ity ekata/... tasmat putram anusislam lokyam ahuh/. 5 cf. Stenzler bringt... Welten'. But Oldenberg procures... (open) space'. So also BR. catuspadas tvaya dharmas cilo lokyena (v.1. laukyena) karmana/ aksayas tava loko'yam.../. Also ef. Mbh. 12.1983, 7.696. Eggeling, however, translates makes for heaven'. Madhu Vidya/30

Page #56
NOTES AND COMMUNICATIONS 601 tamad evamvit svapyat/ lokyam ha/ ... ete eva tad devate retah sincatas tasmad retasa idam sarvam sambhavati yad idam kimca. In the Sat. Br: 11.3.3.7 a student is first forbidden to beg alms after the period of his studentship. But again he is permitted to beg alms from a woman in whom he has the greatest confidence' (Eggeling). This is supposed to be lokya 'customary in the world; permissible'.1 In Manu. 2.161 also alokyamay mean what is not usual in the world, not permissible '.' In Mbh. 13.1971 lokya- (v.l. laukya-) as an adj. of tejas- means spread over the world's M. A. MEHENDALE 1 But it can also mean winning world (which is free from death), winning immortality', because as mentioned in 11.3.3.5 the student is supposed to redeem by begging that part of his body which is in death (atha yad ... bhiksate ya evasya mstyau padas tam eva tena parikrinati). Eggeling also translates that makes for heaven'. For the idea of immortality associated with Iolya. also cf. Sat. Br. 10.2.6.7 and Br. Up. 1.3.28 discussed above. yayasyodvijate vaca nalokyam tam udirayet. But the commentator gives the meaning not enabling to win the world or heaven' (svargadilokaprapti pratibandhini). Also Buhler . That will prevent him from gaining heaven'. * This meaning is already noted by BR. Madhu Vidya/31

Page #57
SOME LEXICOGRAPHICAL NOTES ON THE UPANISADS by M. A. MEHENDALE Poona 1. Susi- "Hohlung eines Rohrs" Kath. 19.1 (Bohtlingk-Roth, Skt. Wb.) is to be corrected to susira- in this passage and in KapKS. 29.8 according to Wackernagel-Debrunner, Altind. Gramm. II/2, Nachtrage, p. 939. Bohtlingk (pw.) gives one more reference, the Commentary on ApsS. 7.26.11, but the reference to ChU. 3.13.2-5 should also have been given. On the other hand, devasusi- ChU. 3.13.1 "eine zu den Gottern fuhrende Offnung...." (BR), "Gotterhohlung im Herzen)" (Bohtlingk) has already been recorded. Cf. also parisusiram Sankh., Ar. 11.1 (v. I. parisusiram in the Leiden manuscript). As regards the etymology of this word, Walde-Pokorny I, p. 365 suggest that it is probably to be derived from vsvi- "to swell" with s-extension. Wackernagel-Debrunner, Altind. Gramm. II/2, 8230b (p. 363) derive susila- (Pancar. 3.10.11 cited in Bohtlingk) from sus"to dry". They also suggest, although with a question mark, to equate susila- with susira-, which latter they explain as susi-ra-1 (8686b, p. 857). But how they derive susi- itself is not clear. It may be suggested that susi. is to be derived from svas-:sus. "to blow, etc.". This derivation would be both formally and semantically better. It has to be noted that the word is also written as susi-? which, if the etymology suggested here is correct, should be regarded as the primary form. susi- would then be a secondary form arisen due to dissimilation. As for Amg. jhusira- "lochrig, hohl" (Pischel, Gramm. der Prakrit-Sprachen, $ 211), it seems that the initial s. became ch-, which owing to voicing became jh-. Amg. jhusira- is a more likely development from susira- than from susira-. And not as su-sira- (Grassmann, Monier-Williams, Wackernagel, I, $ 83d, p. 94, II/1 SS 42b, p. 98). 2 Uhlenbeck, Kurzgef. etym. Wb. der ai. Sprache, regards the form susi-as perhaps better than susi-. Madhu Vidya/32

Page #58
SOME LEXICOGRAPHICAL NOTES ON THE UPANISADS 185 2. atighni- BAU. 2.1.19. is explained by BR as follows: "(von han mit ali), adj. vollstandig vernichtend, davon f. oghni, vielleicht mit Erganzung von avastha, ein alles Unangenehme vergessen machender Zustand". This is following Sankara, cf. atisayena duhkham hanti ity atighni, anandasyavastha.... Bohtlingk regards it as a noun and renders it better as die hochste Stufe". The meaning given by Monier-Williams "utter oblivion or profound sleep (obliterating all that is disagreeable in the past, and regarded as the highest condition of bliss)" is not quite suitable for the Upanisadic passage which runs as follows: sa yatha kumaro-va maharajo va mahabrahmaao va 'tighnim anandasya gatva sayitaivam evaisa etac chete. The expression atighnim anandas ya galva obviously means "having reached the excess, or the utmost stage of joy". 8 This meaning of atighni is to be derived from ati Vhan- "to strike beyond (the limit)" which, when used as a mark of excellence in a feat would mean "surpass all others". Regarded this way, the expression would mean "having reached that utmost stage of joy which surpasses all other conditions of joy of lesser degree". We may compare here the use of ati vyadh"to pierce through, beyond" which in the RV. 4.8.8 is used as a mark of excellence: sa vipras carsaninam savasa manusanam ati ksipreva vidhyati "Dieser Redekundige ubertrifft mit Kraft (die Reden aller) Volker, (aller) Menschen wie die schnellende (Bogensehne)" (Geldner). Instead of a bow-string, we should better think of an arrow (isu-). We may also discuss here the word atighnyd- occurring in AV. 11.9.16 (= 11.7.16) which, according to the lexicons, is derived from atighniand hence rendered as "uberwaltigend (?)" BR, "am hochsten stehend" Bohtlingk, "one who is in the condition ati-ghnf" Monier-Williams. The AV. passage runs as follows: sa ksiyati visvasyesano vrsa bhamyam atighnyah "he (= ucchisja- "the remnant of the offering") dwells, ruler of all, an overpowering (?) bull upon the earth" (Whitney). But in this context, as an adjective of bull, the word is better interpreted as "beyond those that may be killed" (ati-*ghnya-). The word would then be almost identical in meaning with aghnya- (or aghnyd-). We may compare this interpretation of ati-ghnya- with ati-martya- "superhuman" Bhag. Pur. 1.1.20 (BR, additions in vol. V), and upari-martya-s "superior to men" RV. 8.19.12. martya- has been treated as a gerundive reformed from : Cf. "ein Ubermass von Wonne geniessend" Deussen, "Hohepunkt der Wonne" Bohtlingk, "the summit of bliss" Hume. * Also Wackernagel-Debrunner, II/2, $ 642y, small type, p. 791. Cr. however Wackernagel-Debrunner, II/2, $ 653, small type, p. 814, where the possibility of upari-martya- being derived from upari-marta- + ya- is also mentioned. Madhu Vidya/33

Page #59
186 M. A. MEHENDALE *mrtya-"one who must die" (Wackernagel-Debrunner, II/2, 8 642a, small type, p. 789). Or alternatively it has also been regarded as a secondary derivative, without any change in meaning, from marta- (WackernagelDebrunner, II/2, $ 651a, small type, p. 807 and $ 652c, p. 813). It has, however, to be admitted that *ghnya- like martya- is not used by itself. Madhu Vidya/34

Page #60
UPANISADIC ETYMOLOGIES M. A. MEHENDALE Deccan College, Poona (1) A hitherto unnoticed etymology of purusa. This occurs as an implied etymology in the Mundaka Up. 2.1.5 where we read puman retah sincati yositayam bahvih prajah purusat samprasutah. Here the wording seems to imply the derivation of prirusa as pu (puman) + ru (retah) + $a (sincati). Although this etymology is based just on sound similarities between parts of the word purusa and three other independent words, it is interesting to note that it reminds one of the etymology of purusa once suggested by Uhlenbeck (Kurz. Etym. Worter.): "Gehort es (=pirusa) vielleicht zur ind. wz. *pers- spruhen u.s.w., welche auch 'semine irrigare' bedeutet haben kann?"2 The Upanisadic etymology also has a striking resemblance with the passage in the Ait. Ar. (II.5.1) referred to by H. W. Bailey in TPS 1960. 84-86 while demonstrating the derivation of purusa from the verbal base par. 'to nourish' attested in Khotanese. In order to show how purusa 'a male' is thought of as 'the nourisher of the child' he cites the Ait. Ar. passage, a part of which runs as: puruse ha va ayam adito garbho bhavati; yad etat retas tad etat sarvebhyo 'ngebhyas tejah sambhutam; atmany evatmanam bibhrati; tad yada striyam sincaty athainaj janayati; tad asya prathamam janma. "This embryo is indeed in the man in the beginning. That which is semen is the strength created from all the limbs. (Thus) one bears in oneself his own self. When one emits that (semen) in a woman, then he begets him. That is his first birth." The well known etymology of purusa in the Upanisads is of course the one which equates it with puricaya, thus suggesting its derivation from pur4 + si.5 sa va ayam purusah sarvasu pursu purisayah Bp 2.5.18; sa etasmaj jivaghanat paratparan purisayam purusam iksate Prasna 5.5. The same etymology is also given in the sata. Br. 13.6.2.1 (where vayu is identified with purusa) and later in the Nirukta 2.3. The difference in the two forms purusa and purisaya (cf. Nir. 1.13) is sought to be made good in the Gopatha Br. 1.39 under the usual concept of the paroksapriyatva of the gods. The Nirukta, of course, also mentions the other possible derivations of purusa viz. pur + sad (purisada) and from pr 'to fill'.8 (purisada) and fro possible derivations of wrukta, of cour Madhu Vidya/35

Page #61
UPANISADIC ETYMOLOGIES A third etymology of the word stated clearly in the Br. 1.4.1, but which does not seem to have found favour with the etymologists, is from purva +us 'to burn'. sa yat purvo 'smat sarvasmat sarvan papmana ausat tasmat purusah. (2) Etymology as evidence for pronunciation. In the case of a few etymologies the varnasamanya between the words derived and the elements from which their derivation is sought is not perfect. If one does not wish to set aside such cases on the ground that slight dissimilarities in the phonetic shape were not cared for, one may argue that this difference, at least in some cases, points to the habits of pronunciation. Thus the etymology of yajna (Page #62
M. A. MEHENDALE many etymologies are based on sound similarities, especially those which occur as explanation of certain identifications. Usually in such etymologies it is the initial letter of the elements from which the noun is sought to be derived that is taken advantage of. Cf. for instance the implied etymology of purusa above. But this by no means is always the case. Thus for example the second letter s of vis in one of the etymologies of purusa or n of the class-sign na in punati in the etymology of yajna are also availed of. What is more striking is the fact that in a few cases the augment a- of the past tense is also used for the purposes of derivation. This perhaps implies that the derivation is based on some event that had once occurred. In the etymology of asva we read: tato 'svah samabhavat, yad asvat Br. 1.2.7.16 The most simple way of giving the etymology in the Upanisads is to state the verb in an inflected form and follow it up by saying that that is how the object has come to have its name. The initial clause giving the verb may begin with yat.17 yad ebhir (= chandobhir) acchadayans tac chandasam chandastvam Ch. 1.4.2;18 sarvam va attiti tad aditer adititvam Br. 1.2.5. In the Upanisads we do not find an etymology being introduced by the word kasmat, as for example we do in the Nirukta, nighantavah kasmat (1.1.). But this later method appears to be foreshadowed in the Upanisads in the use of the word tasmat (or tena) in the latter part of the etymological statements. sa haisa garams tatre ... tad yad gayars tatre tasmad gayatri nama Br. 5.14.4; te yad idam sarvam adadana yanti tasmad aditya iti Br. 3.9.5.19 Instead of tasmat the ablative of the noun derived from the verb underlying the etymology is also used. yad garhapatyat praniyate, pranayanad ahavaniyah pranah Prasna 4.3; vidanad vidyut, vidyaty enam primano... Br. 5.7.1. In one instance where no derivation from a verb is involved, the word niruktam in the sense 'etymology' occurs: tasyaitad eva niruktan hrdy ayam iti tasmad dhrdayam ch. 8.3.3. A large number of Upanisadic etymologies occur due to curious identifications. These etymologies are then made the basis for the various identifications by showing that the meaning derived from the etymology is applicable to the object identified. This is usually expressed in two statements, the first containing the identification and the second the etymological explanation. The former has very often the particle vai, but occasionally also eva or vava, while the latter is overwhelmingly marked by the particle hi (seldom vai or ha). Prano va uktham, prano hidam sarvam utthapayati Br 5.13.1; vag vai gayatri, vag va idam sarvan bhutan gayati ca trayate ca Ch. 3.12.1; iyain vai prusa, iyam hidam sarvam pusyati yad idar kim ca Br 1.4.13.20 Madhu Vidya/37

Page #63
UPANISADIC ETYMOLOGIES In a few instances the etymology is not directly stated as in the above examples. But even so these implied etymologies can be easily noticed. That pati and patni are related to pat is only suggested by the statement sa imam evatmanam dvedha 'patayat; tatah patis ca patni cabhavatam Br 1.4.3. In the case of purusa, as noticed above, the etymology is hinted by putting the grammatically regular form purisaya by its side. In a third type the verb, synonymous with the one intended in the derivation, is mentioned. Thus mune occurs for tra in tasmad enam sarvasmat putro muncati tasmat putro nama Br. 1.5.17.21 But the verb a hu, from which ahuti22 is intended to be derived, is not mentioned in any form. It is to be understood from the expression ehi ehi. cf. ehy ehiti tam ahutayah...yajamanam vahanti Mundaka 1.2.6. The following case is very peculiar in the sense that the entire etymology is left to be inferred. istaphalam evodanah; sa enam yajamanam aharahar brahma gamayati Prasna 4.4. Here obviously udana is traced in the 'unorthodox' etymology to utni, but this is indirectly stated as brahma (ut) gamayati (nayati). The peculiar identifications, to which we have already referred, have sometimes led to what we might call 'unorthodox'23 etymologies. Thus yajus is derived not from yaj but yuj: prano vai yajuh; prane himani sarvani bhutani yujyante Br 5.13.2; samana is derived not from saman but from sama + ni esa hy etad dhutam annam samam nayati Prasna 3.5.24 vidyut is derived from vi da (do) in vidanad vidyut Br 5.7.1. Even the verbal form svapiti is analysed as sva + api + ita (i): svam apito bhavati tasmad enam svapitity acaksate Ch. 6.8.1. In the Ch. 8.3.5 satya is analysed as sat, ti, and yam of which the last element is derived from yam (atha yad yam tenobhe yacchati). We may also mention here sama which is analysed as sd and ama in the Br. 1.3.22:2 Ch. 1.6.1; 1.7.1. But it is also stated to have come from sami+anc in the Br 5.13.3 prano vai sama; prave himani sarvani bhutani samyanci. NOTES 1. This is in keeping with the principle later enunciated by Yaksa (apy aksaravarnasamanyan nirbruyat) in the Nirukta 2.1. The three syllables of the word hrdaya are similarly derived from hr. v da and Vi in the Br 5.3.1. In the Ch. 8.3.3. however, it is analysed as hrdi+ ayam. 2. J. Scheflelowitz (KZ 53.255, 1925) who derives purusa from *par-usa (Lat. pario) remarks that purusa 'bedeutet eingentlich "Erzeuger".... For other explanations see F.B.J. Kuiper, Kirfel Festschrift (Studia Indologica) p. 146 (1955) and M. Mayrhofer, Kurz. ety. Worter, p. 312 (1958). 3. ayam does not seem to refer either to Atman or to the indefinite 'one' as Hume thinks it to be (p. 298 f.n. 2). It is clearly related to garbha which is first supposed to exist in the man in the form of retas. 4. That by pur we have to understand heart' or 'inside of the body' is shown by such expressions purusah.... sada jananam hrdaye samnivistah Katha 6..17, Madhu Vidya/38

Page #64
M. A. MEHENDALE Diva ya esa....purusah.... ya eso 'ntarhrdaya akasas tasmin chete Br. 2.1.17. puritati sete Br 2.1.19, puruso madhya atrani tisthati Katha 4..12, ihaivantahsarire somya sa purusah Prasna 6.2. On the other hand in the Atharvaveda 10.2.28 we read piiran yo brahmano veda yasyah pirusa ucyate. 5. Instead of si we have perhaps pur + vois in a sort of implied etymology: purah sa paksi bhutva purah purusa avisat Br. 2.5.18. 6. prayaty antar ity antarapurusam abhipretya. Nirukta also cites tenedan pur nam purusena sarvam, Taitt. Ar. 10.10.3, Svet. Up. 3.9. 7. This element is also availed of in a different etymology given in the Taitt. Ar. 1.23.3, purvam evaham ihasam iti tat pttrusasya parusatvam (purva + Vas). 8. Cf. Nir. 3.19. Here other etymologies are also given. The sat. Br. 3.9.4.23 presupposes metathesis yanja > yajna. 9. Other etymologies of this type have been noted below under 'unorthodox' ety mologies. 10. The present day pronunciation of jn as dn or dny has been noted by S. K. Chatterji, Ind. Ling. 21.78 (1960). The other 'unorthodox' etymology of yajna! as ya- + Vina appearing in the Ch. Up. 8.5.1 does not give any clue regarding the pronunciation. atha yad yajna ity acaksate brahmacaryam eva tat; brah-1 macaryena hy eva yo jnata tam vindate. 11. On this point again cf. S. K. Chatterji, op. cit. pp. 78-79. 12. The sat. Br. 10.6.2.8-10 gives the same etymology but clearly shows the pro. nunciation to be uktha. 13. Also cf. Sat. Br. 6.1.1.2. S. M. Katre (ABORI 20.277) sees in this etymology the knowledge (unconscious perhaps) of the middle Indic aspiration observed in the MIA treatment of certain OIA clusters with T, e.g. tatra >tattha. In the Ait. Ar. 1.3.14 indra = idandra. 14. This is also how Aupamanyava's view regarding the etymology of nighantu is given in the Nirukta 1..1. te nigantava eva santo nigamanan nighantava ucyante. 15. These remarks are based on the ten principal Upanisads. 16. Also sat. Br. 13.3.1.1. yad asvayat tad asvasyisvatvam. In the Ait. Ar. 2.2.1 we read the following in the etymology of atri: sa yad idan sarvan papmano 'trayata yad idam kinca tasmad atrayah. 17. In the etymology of asva noted above the clause with yat follows and does not precede, 18. Also cf. Ait. Ar. 2.1.4 and 6 where the etymologies of siras, sarira etc. are given in the same way. The etymology of arka given in the Br. 1.2.1 is simi lar, but the verb does not appear in an inflected form. 19. Also cf. the etymology of putra in the Br. 1.5.17, of midra in the Br. 3.9.4, of purusa in the Br. 1.4.1 and yajna in the Ch. 4.16.1 noted above; of adhyardha in the Br 3.9.9 and chandas in the Ait. Ar. 2.1.6. Both tat and tasmat occur in the Br 1.3.22 and tasmat in the Br 3.9.3 where derivation from a verb is not involved. . Also cf. prana == yajus Br 5.13.2, prana = sama Br 5.13.3, 1.6.1 (2, 3) prana = ksatra Br 5.13.4; vac =atri Br 2.2.4; prana = vasul ch. 3.16.1; prana - rudra ch. 3.16.3; prana =aditya ch. 3.16.5; also cf. the explanations of uktha and brahma in the Br 1.6.1, 2, 3 and of udgitha in the Br 1.3.23 and ch. 1.3.6. 21. But vtra is given, for instance, in the Nirukta 2.11, putrah puru trayate or pum narakan tatas trayate. In the Ait. Ar. 2.1.5 where sayam is explained the verb vga occurs in place of vi. 22. In a similar derivation of ahuti in the Ait Br. 1.1.2 the form is first identified with ahuti and then derived from avhu. ahutayo vai namaite yad ahutaya etabhir vai devan yajamano hvayati tad ahutinam ahutitvam. Thus while the Brahmana passage shows consciousness of the difference in the vowel length between ahuti, which can be had from a hu, and ahuti, the Upanisad passage ignores it. However, it may not be justified to infer from this instance that the vowel length was ignored in the pronunciation of certain words. Such etymologies are also known from the Nirukta. Cf. the etymology of anna from anam in the Nirukta 3.9. The other etymology from Vad is also given there. 24. Also cf. 4.4 and the etymology of udana referred to above. In Prasna 4.3 prana is similarly derived from pravni. 25. Also similarly analysed in the Br. 5.5.1 and Ait. Ar. 2.1.5; as sat and tyam in the Kausi. 1.6. 26. Or derived from sama in the place and again in Br 1.6.1, 2, 3. Madhu Vidya/39

Page #65
M. A. MEHENDALE VANARGU Vanargu- has been analysed and interpreted by most Vedic scholarsi as vanar + gu (< Vga or Vgam) meaning moving in the wood, wandering in the forest'. According to an entry (4. gu) in BR Vga appears as gu2 at the end of a compound in adhrigu and vanargu. However, adhrigu has already been supposed to contain -gu-ga-, P. 8. 2. 48. 3 Bergaine's reference RV 1. 145. 4. to be corrected as 1. 145. 5. 4 Wackernagel, Alt. Gr. II 1 $ 42 d & p.99. Madhu Vidya/40

Page #66
Mehendale: Vanargu does not seem to have found followers. It is therefore necessary to state that his interpretation is perfectly justified. Agni is often called a bull' or is compared with a bull, and in this respect he is in many cases described as active in the vana(r). The expression yad ith ghcchanty usatir apisthitam, which occurs in the immediately preceding verse (4), also favours this interpretation. The epithet bull of the forest' appears to be meant to contrast with mrgah apyah the animal of the water'.7 RV 10.4.6 is also addressed to Agni: tanutyajeva taskara vanargu rakanabhir dafabhir abhyadhitam iyam te agne navyasi manisa yuksva ralham na bucayadbhir angaih|| 77 Geldner translates: "Wie zwei im Walde streifende Rauber, die ihr Leben einsetzen, haben (beide Arme) mit zehn Stricken (das Reibholz) festgebunden. Diese neueste Dichtung ist fur dich, Agni ; bespanne gleichsam deinen Wagen mit deinen flammenden Gliedern!" Geldner thus understands ab as a very much abbreviated comparison. The two arms holding the churning stick with ten fingers are compared with two bold robbers, wandering in forest, binding a traveller. Geldner says that while interpreting ab in this way he is only following the traditional explanation available since the days of Yoska. He remarks: "So auch Say. und Durga zu Nir. 8, 14. Er vergleicht die feuerreibenden Arme mit Zwei Raubern', Yaska." Now Sayana indeed interprets the comparison in the way imagnied by Geldner. He, however, does not say anything about the upameyavakya (the two hands holding the aranis) and wants us to follow it from the statement of Yaska which he quotes. Durga also essentially understands the comparison in the same way. It is likely that Yaska also had the same thing in mind and interpreted vanargu 5 As shown by the above translation of Geldner and also by his Glossar. The same interpretation is given by Uhlenbeck's Worterbuch and by Macdonell in Skr. vanara und verwandtes' KZ 34. 292-96 (1897) and in his Practical Sanskrit Dictionary (also Vedic Index S. V. taskara and vanargu). All these scholars have accepted the interpretation of Yaska (see next verse) and Sayana (vanagami). Wackernagel-Debrunner Alt. Gr. II 2, SS287e note (p. 472) also have the same meaning in mind but regarding the form they suggest two possibilities. "In vanar-gu- agre-guscheint eine sinnverwandte Wurzel zu stecken ;... oder war yu- aus Wurzel i- gebildet und nachtraglich auf Wurzel ya- bezogen und zum Vorbild fur gu -zu qa- geworden?" 6 RV 4. 5. 3; 5. 1. 8, 12; 5. 28. 4; 6. 16. 39; 8. 60. 13; 1. 58. 4, 5; 1. 94. 10; 10. 115. 2; 1. 128. 8 6. 6. 5. In some cases where Agni is not directly mentioned as a bull, there is no doubt that this is what is intended, e.g. RV 1. 143. 5; 7.4. 2; 7.7.2; 8.43.8; 10. 79. 2. 7 Cf. here vrsno......garbham......napatam apam RV 5. 41. 10 and ursabho roraviti...... apam upasthe mahiso vavardha RV 10. 8. 1. 8 So also Sarup, Eng. Tr. of the Nirukta, pp. 47-48 and f. n. 1. on p. 48. Geldner mentions the alternative possibility of a passive construction. In that case the two churning sticks, held by the ten fingers, will have been compared with two arrested robbers. 9 We may now add also Skanda-Mahesvara. 10 Thus Yaska does not expressly mention the traveller' bound by the robbers. This has been supplied by commentators, pathika (Sayana), adhuaga (Durga), kascit (SkandaMahesvara). Madhu Vidya/41

Page #67
Velankar Commemoration Volume in the way Durga, Sayana and Geldner have thought him to do (i.e. an epithet meaning vanagaminau to be understood only with reference to the two robbers). But the way Yaska has worded his statement leaves this point doubtful. He says tanutyak tanatyakta" vanarga vanagaminav agnimanthanau bahu taskardbhyam upamimite. This can mean that according to Yaska vanarga vanagaminau i.e. agnimanthanau (because vana stands for the two aranis) and hence the epithet refers only to the upameya arms (baha) and stands for it. This would mean that vanargu (bahu) are compared with taskard. Whether Yaska intended to take vanargi also with taskard or not may be a moot point. But Oldenberg (Noten) certainly does this and understands vanargu as going with both--the robbers and the hands: "vanargis im Vergleich die im Walde ihr Wesen treibenden Rauber, im Hauptsatz die and den Holzern arbeitenden Hande." 78 All the interpreters mentioned above have, however, one thing in common, viz. that they understand vanargu to have come from vanar + Vga or gam and the form as nom. dual. Bergsigne, although as noted above has rejected this explanation (vanagaminau), considers the form to be nom. dual. He accordingly feels that in the above verse the robbers are metaphorically called bulls of the forest'. This, however, does not seem to be the case. If vanargu is taken to be nom. dual going with taskard the comparison appears to be very much abbreviated because, besides not mentioning anything about the upameya, it also omits to mention in the upamanavakya what the robbers hold fast. This difficulty will not arise if one agrees with Bergaigne in understanding vanargu as two forest bulls ' but differs from him by taking the form as acc. dual.13 The first two quarters are therefore to be translated as: "(The two arms) have held fast (the two churning sticks with ten fingers) as do two brigands, risking their life, two forest-bulls with ten ropes." In this way the upamanavakya becomes complete. Moreover, the epithet tanulyaja becomes more relevant in as much as bodily risk is involved rather in binding a forest-bull than in catching hold of a traveller. Of course the dual number in vanargu as well as taskara is occasioned by the two aranists and 11 Why Yaska uses tanutyak (sg.) for tanutyaja (dual) is not clear. His explanation tanutyakta has been taken by Durga and Sarup to be dual of tanutyakta-, by Skanda-Mahesvara to be sg. of tanutyakty. The latter interpretation is to be preferred because Yaska is not likely to use a Vedic dual form in -a; in fact he renders Vedic dual by the classical one as duhanta by duhantau (6. 26), karna by karnau (10. 41), etc. Yaska also occasionally renders a root noun, whether occurring by itself or in a compound, by a derivative in -tr. Hence nid-= (abhi)- -ninditr(10.42), brahma-dvis- brahmana-dvestr (6. 11); similarly tanutyaj. = tanutyaktr-. 12 If the alternative interpretation of the Nirukta passage suggested above is correct then in that case the upameya hands in the form of vanargu will have been mentioned. 13 Wackernagel-Debrunner, Alt. Gr. III SS 121 b x (pp. 218-19) say that a compound with -gu- as the second member has the usual inflexion of u- stems. In the footnote, however, they point out that such compounds occasionally in the Rigveda show strong forms on the analogy of the declension of -go-, e.g. prsni-gav-ah instead of prsni-gav-ah. 14 The idea of stealing is not present. What is intended seems to be that men of this type were employed for catching the forest-bulls. 15 The dual number will cause some difficulty since the two hands are known to hold only one arani. However, in view of the advantages gained by the above interpretation, it seems necessary to set aside this difficulty. The poet apparently has not been exact in this detail, and has been carried away with the usual dual number of the aranis. Or, does this refer to some different way of producing fire? Madhu Vidya/42

Page #68
Mehendale: Vanargu the two arms. Similarly the number ten has a reference to the number of the fingers.16 We read the following in the AV. 4.36.7 : na pisacaih sam saknomi na stenair na vanargubhih/ pisacas tasmin nasyanti yam ahan gramam avisel/ Whitney's translation : "I cannot (bear) with pisacas, nor with thieves, nor with savages (? vanargu); the pisacas disappear from that village which I enter." The hymn in which the verse occurs is supposed to drive away all evil-minded beings, especially the pisicas (cf. sahe pisacan sahasa verse 4, tapano asmi pisacanam vcrse 6, and pisacas tasman nasyanti verses 7-8). In our verse although only the pisacas are mentioned in the second line, there is little doubt that the reciter felt confident of being able to drive away from the village also the stenas and the vanargus mentioned in line 1. These three words seem to stand for three classes of harmful beings. When vanargu appears in this situation there is good reason to believe that it refers to a class different from stena and hence means 'a forest-bull'. But the reciter of the spell may have wished the village to be free from the nuisance not only of the forest bulls but also of all wild animals, 17 and therefore in this instance vanargu appears to stand for all wild animals. Samaveda 6.4.9 runs as hari ta indra smasruny uto te haritau hari/ tar tva stuvanti kavayah parusaso vanargavah|| The reading parusaso is found in the citation given by BR18 which is apparently based on the text of the Aranyakasamhita of the Samaveda published by Goldschmidt (Monatsber. d. Kgl. Preuss. Akad. d. Wiss zu Berlin 1868, 228-248).10. It also occurs in Bloomfield's Concordance which is based on the text of the Aranyasamhita published with the commentary of Sayana and a Bengali translation by Satya Brata Samasrami, Calcutta, 1873.29 But the Bibl. Ind. edition of the Sama 16 Cf. KV 3. 23. 3; 1. 144.5; 1. 95. 2 ; 3. 29. 12-13. 17 AV. 4. 3; 12. 1. 49. Sometimes taskara appears by the side of stena (cf. AV. 19. 50. 5; 19. 47. 6-7 : 4.21.3 and 7), and hence one may be inclined to consider that vanarge(=panagani) refers to him. Thus c. g. Sayana : vanargu sabdas coranama (as listed in the Nighantu 3. 24) ...............vanagamibhis corair api. Bloomfield prowlers in the forest.' But as mentioned above vanargu is better taken to refer to a class different from that of the stenas. 18 qera: in BR's citation to be read as a . Their abbreviation S.V. Naigh. also to be corrected as Sv. Naig. (Naigeya Sukha as callech in Goldschmidt's edition). 19 Also in the Aranyaka gana edited by Narayanasvami Diksita (Aundh, 1942). Aran. yakasainhita is not published in Benfey's edition of the SV (Leipzig, 1848). 20 Sayana's commentary in this edition, however, has purunds, as in the text of the commentary published in the Bibl. Ind. edition, Madhu Vidya/43

Page #69
80 Velankar Commemoration Volume veda (Vol. II. 324? gives the reading purusaso21The same reading also occurs in the Jaiminiya Samhita (2.4.3) edited by Raghu Vira, Lahcre, 1938.22 The reading parusaso, however, is to be favoured because pirusiso would have accent on the first syllable, and secondly perusa- by the side of kavi- hardly adds anything meaningful. Vanargavah in the SV verse shows an accent different from the one (vanurgu-) found in the other three occurrences. This fact has not been noted in the lexicons. Sayana notes it and therefore he does not interpret the word as vanagami, but as vananiyah sambhajaniyah sevaniya gavo yesum te vanargavah. But if the compound was really intended to be a bahuvrihi, since the accent is on the purvapada, it should have been accented on the first syllable on account of prakstisvaraiva (vanar:). Therefore it appears that this is only a case of accent shift and does not amount to different accentuation. The accent shift is probably to be explained as follows: (vanargu + as) *vanargvas > *vanargvas > vanargavas. Sayana does indeed consider here -gu. Page #70
Mehendale: Vanargu (taboo), and therefore had to be learnt in the forest, not in the village." But if the interpretation of vanargu suggested above is the correct one, it does give some ground in support of the original view of Winternitz. The verse refers to persons singing in the forest. This singing is more likely to refer to the ritual one than to the one done at the time of learning the melodies. The Samaveda verse may be translated as: "Oh Indra, tawney are your beard (hair), and tawney your horses. Such as you are, the wise men, dust-coloured, singing in the forest, praise you." 81 A Sanskrit lexicon, it seems, will thus require two entries for this word: (1) vanargu (RV and AV) forest-bull' (vanar + gu < go); (2) vanargu (SV) singing in the forest (vanar + gu< vga to sing "). vanargu 25 But BR: 'Weise und Wilde'. Hence Macdonell and Keith, Vedic Index 2. 241: "In the Samaveda the term is more generally opposed to civilized men (kavayah sages'; vanargavah 'savages'). Madhu Vidya/45

Page #71
376 KINSHIP ORGANIZATION IN INDIA APPENDIX A Note on tena tyaktena bhunjithah M. A. Mehendale This third quarter of the first verse of the Isopanisat has been interpreted in at least two different ways. RADHAKRISHNAN translates it as "Therefore find your enjoyment in renunciation" and adds in his note: "enjoy through tyaga, or renunciation of selfwill. Enjoy all things by renouncing the idea of a personal proprietary relationship to them." He thus partly follows SANKARA who equates tyaktena with tyagena. DEUSSEN" has given a similar interpretation, except that he does not take tena to mean 'therefore, but as referring to jagat. "Wer ihm entsagt, geniesst wahrhaft". DEUSSEN remarks in his introductory note: "In diesem Verzichten auf die vielheitliche Welt liegt der wahre Genuss ........, nicht in dem Trachten nach fremdem Gute." HUME follows him in content: "With this renounced, thou mayest enjoy". RADHAKRISHNAN, in his notes on the passage, refers to a second interpretation : "Sometimes this passage is interpreted as meaning: enjoy what is allotted to you by God (tena). Do not ask for more than what is given". This second interpretation seems to refer to the view of MADIIVA. B. D. BASU, who follows his commentary, renders the line as "Enjoy thou what He hath allotted to thee, and do not beg from any (though he be a king), for wealth." The second interpretation is no doubt more acceptable. The use of the verb bhuj. with instrumental is not found in later Sanskrit. PANINI (2.3.18-27 : also B. LIEBICH, BB 10. 217-220, 1886) does not mention it. But this use is not seldom in the Rgveda" and the Vedic prose. E.g. We find, sasrad hi rah sudanava adit yiz utibhir tayam / purd nunan bubhujmahe "Denn immer wieder I The Principal Upanisads, p. 567. ? Sechzig Upanishad's des Veda, p. 524. 3 The Thirteen Principal Upanishads, p. 362. + The Sacred Books of the Hindus, Vol. I, p. 3. * For the use of bhuj. with instrumental, cf. H. WENZEL, UEBER den Inatrumentalis im Rgveda (1879), p. 83, DELBRUCK, Altindische Syntax (1888), pp. 132-133; SPEIJER, Sanskrit Syntax (1886), 74.9 remark, p. 56; GRASSMANN, Worterbuch zum Roueda and BR, Worterbuch s.v. Bhaj More examples also can be found under these references. Madhu Vidya/46

Page #72
OWNERSHIP OF PROPERTY, SUCCESSION & INHERITANCE 377 haben wir uns eurer Hilfen, ihr gutigen Aditya's, einst und jetzt erfreut" (GELDNER) RV. 8.67.16; d nivarta ni vartaya prinar na Indra ca dehi / jivdbhir bhunajamahai / / "Du Heimfuhrer, fuhre sie heim, gib uns die Kuhe wieder, Indra. Wir woolen uns der Lebenden erfreuen" (GELDNER) RV. 10.19.6; datta nas tasya bhesajam tena vo bhuncjamahai/ / "Give us remedy for that; may we enjoy that (remedy) of yours" (WHITNEY has different translation: "give us the remedy for this; for that would we enjoy you"; also Bloomfield, "grant us the remedy for that: through this (remedy) may we derive benefit from you!)" AV. 6.24.3; ubhayena bhunjate TS 2.5.2.7; urja bhunjate 5.2.8.7, 6.1.3,4; annera brunjate 6.2.5.4; yani tai puuri samvatsarad retamsi jayante ...... na tai tair bhijate 'thu yiyeva dasa masvani jayante yani samvatsarikani tair bhurjate, AB 4.22.5,6. "Whatever seeds are born before the year ...... they do not enjoy them; those that are born in ten months or a year, they enjoy them" (KEITII: "they do not profit by them; . . . . ... by these they profit"). Su yari tam asia nyadadhatu, tenanena manusyu bhunjate, $B 2.2.2.13 "the (fire) which the demons established, that the hu. man beings enjoy." tan mim aratu, tan ma visatu, tena bhuksisiya Ar. G?.s. 1.23.19 'may that protect me, may that enter me, may I enjoy it' (STENZLER: "dadurch moge ich geniessen". In respect of tena tynktene blurjithah it is also possible to think of an instrumental absolute construction: "you may enjoy that when renounced (by him)." In this case tena does not refer to Isa, but to the portion of the jagat given up by him. In the fourth quarter of this verse occurs ma grdhah kasya siid dhanam. Kasya svid is no doubt indefinite 'of any one', and hence is usually taken to refer to any other human being. But in this verse, two entities come up for consideration - isa, 'the master', and the individual addressed. Therefore kasya svid should have particular reference to some one else other than the indivi. dual, viz., the master. Since the jagat should primarily be pervaded by the Lord (isa vasyam), it is his property (dhanam) first. An individual should not long for the Lord's property as long as it has not been renounced by him in the individual's favour. Madhu Vidya/47

Page #73
Sure duhita' - By M. A. MEHENDALE Deccan College, Poona The expression sure duhita occurring in the RV. 1.34.5 (tristham vai sure duhita' (= duhita' a') ruhad ratham) has attracted much attention. Two independent notes have been written on it - one by Bartholomae in Bezz. Beitr. 15. 1-2 (1889) and another by Pischel in Ved. Stud. 3.192-194 (1901). Both Bartholomae and Pischel refer to the views of earlier scholars like Benfey, Grassmann, Ludwig, Bergaigne, Richter and Brugmann. It is not necessary to restate here all that these scholars have said regarding this expression. It would be enough to say that the form sure has been considered either as loc. sg. from sura- or as dat, or gen. sg. from swar- both meaning 'sun'. It may be of some historical interest to know that Bartholomae wanted to treat the expression as one compound word sureduhita' < surazdhuzhita'. In his view the separation of the compound into two words by the author of the Padapatha was a mistake. As far as the form is concerned the more widely acceptod view seems to be to consider sure as gen. sg. of svar- showing an isolated case of external sandhi in -e ( -az < -as) instead of the expected -before a voiced stop. To give a few references, Wackernagel, Alt. Gr. 1, p. XIX and SS 285 bss, p. 338, and 3 $ 160d note p. 314. Renou, Gramm. Ved. SS 137, p. 1031, and W. S. Allen, Sandhi p. 71. This sure (gen. sg.) was naturally construed with duhita' to mean 'the daughter of the sun'. In the first instance we have an exact parallel in suro duhita (7.69.4) where the form and the meaning are not in doubt. Next, we have frequent mention of the daughter of the sun' choosing or mounting the chariot of Asvina. Thus suryasya duhita 4.43.2 and duhita' suryasya 1.116.17, 1.117.13. 1.118.5, 6.63.5, duhita' vivasvatah Khila 1.3.2. The daughter of the sun is also called surya'S who is similarly spoken of as mounting the chariot or being carried by the chariot of Asvina cf. 5.73.5, 4.44.1 (also 4.43.6, 6.63.6, 8.22.1 and Khila 1.2.5). Also cf. the suryasukta 10.85. Paper submitted to the Indology Section of the 26th International Congress of Orientalists held at New Delhi from January 4 to January 10, 1964. In his Histoire de la Langue Sanskrite, Renou had observed: On a meme voulu voir magadhisme dans l'idiome sure duhita "la fille du soleil" (p. 30). On this ree Mayrhofer IF 63.286, fn.3. He regrets the way Renou has made his remark and observes"., dass sure den Sandhi -az -enthalte, der in den ostlichen Dialekten spater verallgemeinert wurde ... lasst sich doch diskutieren." The daughter of the sun is indirectly referred to as vadhu 7.69.3, yosana 8.8.10, yoga 1.119.5. urjani 1.119.2, and surf 1.119.3. The 'three' in the chariot are mentioned in 4.45.1, 8.29.8. Mounting the chariot by surya is mentioned in a simile 1.167.5. The identification of Surya is doubtful. Macdonell (Vedic Myth. p. 51) thought that she was no other than the sun himself conceived as a female. Hillebrandt did not like 2

Page #74
M. A. MEHENDALE Scholars who considered sure to be dat. sg., however, took it to mean 'for the sake of the sun'. cf. Bergaigne "La fille est montee sur votre char pour le soleil.". Those, except Oldenberg for whose opinion see below, who looked upon the form to be loc. sg. thought that the expression ultimately meant the same thing as with gen. sg. 'daughter of the sun'. Sayana got this by simply noting that it was a case of vibhaktivyatyaya i.e. loc. standing for gen. (sure = suryasya). But Pischel wished to render it as 'die, die Tochter bei surya ist's, i.e. taking it to be a round about way of saying the daughter of the sun.' In fact, with regard to Bergaigne's feeling that it seemed impossible to construe sure duhita' as "La fille du soleil,"Pischel asserted "Diese scheinbare Unmoglichkeit ist jedoch das einzig Richtige." (p. 193). . - In spite of this assertion and the apparently compelling evidence cited above in favour of construing sure duhita' together meaning daughter of the sun, Oldenberg (Noten I 36-37) did not favour, and in my opinion rightly, this construction. He rather thought of separating the two and interpreting sure as either loc. or dat. sg. He accepted the suggestion of Pischel to treat sure as loc, but differed from him in rendering it: "euren Wagen bestieg bei Sura die Tochter" (namlich eben des S.). With dat, sg. he translated it as "euren Wagen bestieg um des Suar willen (ihm zu Gefallen? ihn zu fordern?) die Tochter" (namlich eben des S.). I agree with Oldenberg in taking sure as loc. sg., and with him and Bergaigne in separating syntactically sure from duhita'. In view of the fact that the context refers to the mounting of the chariot of Asvina and that elsewhere the daughter of the sun is clearly stated doing that, in the present case a simple duhita' without the qualification suryasya or surah will certainly convey the same meaning. "The daughter mounted the chariot of Asvina" would always mean that the daughter of the sun did it. There can be no ambiguity about it. As regards sure, which I have separated from duhita, I interpret it as standing for loc. abs. sure udite. The full expression thus would be this idea and suggested that she was to be identified with Usas (Vedic Myth. I, 42-44, 61 and 11, 399.) This was also the view of Yaska (12.7), Skandasvamin (cf. his remark on Rv. 1.34.5 rat-(i.e. surya-) prabhavac cosaso duhitruyapadesah / suryasya duhita upah/), and perhaps Hopkins (JAOS 15.270, f.n * 'The red is the dawn which mounts the Asvins' car...'). Although Usas and Asvina are known for their close association the identification of Surya with Usas was contested by Oldenberg (Noten 2.53 on RV. 7.69.4). Geldner (in his translation) and Renou (Et. Ved. et Pan. 3.8 and 91 on RV. 7.75.5a) follow him. 4 La Religion Vedique 2.490. 5 Also Madhava sure surye . . . tasya duhita, and Skandasvamin sasthyarthe saptami. He followed Brugmann's suggestion (IF 12.3) to derive manaav-i fem.) from manau (loc. sg. mas.) meaning "die bei (chez) Manu (seiende). Pischel's view that sure is loc. sg. is in turn referred to by Brugmann IF 13.148-149, f.n.3. 7 La Religion Vedique, 2.490, f.n.2. 8 This may be compared with the use of dative in (ratham)...yamasvina ... surydyai tasthathuh/ 8.22.1 "(den Wagen)... den ihr Asvin, ... fur Surya bestiegen habt..." (Geldner). Madbu Vidya/49

Page #75
SURE DUHITA sure (udite] duhita' a ruhad ratham "when the sun had risen, the daughter mounted the chariot". We may note in the first instance that the expression sura udite occurs quite often in the Rgveda (3.15.2; 7.63.5; 7.65.1; 7.66.4,7,12; 8.1.29; 8.13.13; 8.27.21). Secondly, there are a few cases where a loc. surye has been interpreted as loc. abs. surye udite by Indian commentators and by Geldner. In RV. 1.135.3 we read the following: tavayan bhaga stviyah sdrasmih surye sdca / Here Sayana interprets surye = surye udite. cf. (ayam tava bhagah bhajaniyah surye udite sali lasya rasmibhih saca saha sarasmir bhavati samanadiptir bhavati/yad va taih sahilo bhavati/yatah surye udite sati rasmir bhavaty atah/. Also Madhava: udite surye. Geldner refers to Sayana and has similar interpretation: "Dies ist dein zeitgemasser Anteil, der bei (aufgehender) Sonne bestrahlte". In RV, 4.31.5 where abhaksi surye saca occurs, Madhava again renders it as surye udite. Geldner offers similar translation: "Ich habe meinen Anteil an der (aufgehenden) Sonne bekommen". 10 Sayana, however, has: he Indra tvam ahar surye suryena saca saha abhahsi bhaje. Oldenberg first remarked that whether we should regard su're as loc. or dat. sg. could be decided if we had more exact knowledge of the mythology (Ob nun Dativ oder Lokativ konnte nur konkretere Kenntnis des Mythus, als wir besitzen, entscheiden). But then he also felt that RV 7.69.4 would perhaps show that the event of mounting the chariot of Asvina by Surya occurred at night 'for the sake of the sun', thus indicating his preference for the dative. In RV. 7.69.4 occurs the word pa'ritakmya (for full citation see below) which Oldenberg took to mean 'night'. Now in the first instance the meaning of pa'rilakmy, itself is not certain (See Mayrhofer's Kurz. Wort.). And on the other hand we have the following references to show that when Surya or the daughter of the sun mounts the chariot of Asvina, the horses (or the winged horses called the birds) yoked to their chariot rush to avoid the heat (of the sun), which means that the event did not occur at night but at sun-rise. 1. RV. 4.43.6: sindhur ha vam nasa'ya sincad asvan ghrna' vayo 'rusa'sah pari gman/ tad u su vam ajiram ceti ya'nan yena pali bhavathah surya'yah // "Der Sindhu netzte zusammen mit der Rasa eure Rosse, die roten Vogel entgingen den Gluten. Diese eure rasche Fahrt ward hochberuhmt, durch die ihr die Gatten der Surya wurdet." (Luders, Varuna I. 139). - Geldner in his note on this verse says: "Sie (i.e. the winged horses) entgehen der Hitze, die das Rennen oder die Nahe der Tochter der Sonne verursachte". This is very doubtful. More running will cause more heat; and however fast the horses may run that will not help them to get away from Surya's heat once she had mounted the 9 It is not necessary to consider this as an instance of word haplology by adding stro before duhita: *sure [u'dite suro) duhita > su're duhita'. See the remark on p. 72 about duhita in this context meaning daughter of the sun'. 10 In view of the fact that in these two cases we get surye saca, it is tempting to suggest that in the passage under discussion we understand sure to stand for sure saca. .6: Madhu Vidya/50

Page #76
4 M. A. MEHENDALE chariot. The heat referred to is not that of Sury, but of the sun, and the horses try to avoid it by running away swiftly from him ("Die sengende Glut hat sicherlich ihren Grund in der Nahe der Sonne". (Luders, Varuna I. 139). 2. RV. 5.73. 5: ayad vam surya' ratham tisthad raghusyadam sada / pari vam arusu' vayo ghrna' varanta atapah // "Stets, wenn die Surya euren rasch dahinfahrenden Wagen bestieg, hielten sich eure roten Vogel die Gluten ab, so dass sie nicht versengten." (Luders, Varuna I. 139). The epithet raghusyad becomes significant since the chariot would be quick in getting away from the (sun's) heat. 3. RV. 7.69.4 (referred to by Oldenberg): yuvoh sriyarie pari yosavrnita su'ro duhita' pasitakmyayam / yad devayandam avalhah sacibhih pari ghransam omd'na vam va'yo gal // "The maiden chose your beauty over that of others (pari), the daughter of the sun ........ ..... When you helped the devoted one with (your) strengths, the bird (i.e. the winged horse) of yours escaped with your help the heat (of the sun)."1? (Weil ihr dem Gottergebenen mit euren Kraften halfet, ist euer Vogelgespann durch euren Schutz der Hitze entgangen-Luders, Varuna I. 139 f.n.2. He does not give the translation of the first two quarters). 4. RV. 1.119.2 : svadami ghar mam prati yanty utdya a' vam urja'm ratham asvinaruhal / "I prepare (your) hot drink, your protections approach (me, the worshipper). Urjani (ie. Surya) has mounted your chariot, oh Asvina." The context of mounting the chariot clearly indicates that urjani stands for Surya (cf. Sayana). One wonders why Geldner chooses to interpret it as die personifizierte urj (Starkung),' when in the following verse - yad asvina vahathah surim a' vuram which he translates 'da ihr Asvin die Herrin nach Wunsch fahret'-in suri he sees wordplay with Surya (also Oldenberg). The word gharma clearly indicates that the verse refers to the pravargya which is offered in the forenoon and afternoon (parvahme and aparalme), and not at night. All these references indicate that surya mounted the chariot of Asvina when the sun had appeared. This lends support to the interpretation of sure= sure udite. In the end we will take note of an opinion expressed by Wilhelm Havers in his Untersuchungen zur Kasussyntax der indogermanischen Sprachen.13 In this book he 11 Probably the one offering pravargya, cf. 1.119.2. 12 The translation follows Luders. A different translation is proposed by H. D. Velankar, Indological Studies in Honor of W. Norman Brown, p. 235. He takes vayah as the invigorating food of Alvina. 13 Strassburg, 1911. Madhu Vidya/51

Page #77
SURE DUHITA illustrates mainly the use of the first and the second person pronominal forms me and le as Dativus Sympatheticus. His conclusion regarding the use of sympathetic dative as against the possessive genitive is that Sanskrit has very faithfully preserved the situation of the I.E. stage which was as follows: Sympathetic dative with the pronouns of the first and second person, genitive with all other pronouns, substantives and participles (p. 44). All instances collected by him also show that the use of sympathetic dative is adverbial (ein zum Verbum des Satzes gehorender Dativ' p. 20) and not adnominal. He thus translates dyaur me pita' janita' RV. 1.164.33 der Himmel ist mir Vater und Erzeuger'14 (p. 35, also 8), sakhayas la indra visvaha syama 7.21.9 'dir, O Indra, mogen wir stets Freunde sein' (p. 36), sapatri me para dhama 10.145.2 "blase mir die Nebenfrau hinweg' (p. 30, also p. 19). It is therefore surprising that suddenly in vr'sne sapalni sucaye sabandhu kV. 3.1.10 Havers chooses an adnominal interpretation and translates (p. 36) 'die beiden Frauen des glanzenden Stiers'. The text clearly shows that the dative is adverbial and not used as possessive genitive. Havers should have translated the expression as Geldner, in fact, has done: "Die beiden Verwandten (Himmel und Erde) sind dem reinen Bullen seine geminsamen Frauen"16. The verb Vas- or v bhu- has quite commonly to be supplied in such cases. Since vrsnc sapatni cannot be the proper instance of the use of the dative as possessive genitive, there can be no question of finding possible support for that construction as Havers feels to do (p. 36), in sure duheita' (daughter of the sun). Havers refers to Oldenberg (Noten) and Keith, JRAS 1910. 471, foot note 1. We have already seen that Oldenberg no doubt prefers sure as dative, but he certainly does not interpret it as Havers chooses to do. Keith says that perhaps one may think that surel is dative used as possessive genitive. But he has already remarked that this use of dative to indicate kinship relationship is not illustrated in Whitney's Sanskrit Grammar (pp. 95-96, 1889) and Speyer's Vedische und Sanskrit-Syntax (pp. 13-15, 1896). One may add that this kind of dative is not noted for Vedic or Classical Sanskrit also by Panini (2.3.12 ff. and Liebich, BB 10.214-217, 1886, for Ait, Br, BB 11. 284-288, 1886), Speyer, Sanskrit Syntax pp. 58-67, 1886, Delbruck, Alt. Syntax pp. 140-150, 1888 (earlier also Uber den indogermanischen, speciell den vedischen Dativ KZ 18.81-106, 1869), Hopkins, The Vedic Dative Reconsidered, Trans. and Proc. APA 37. 87-120, 1906, Hopkins, Aspects of the Vedic Dative JAOS 28. 360-406, 1907, Macdonell, Vedic Grammar for Students pp. 310-315, 1916, Renou, Grammaire 14 Geldner apparently ignores the argument of Havers and translates in this and the fol lowing two passages me and te as genitive: "Der Himmel ist mein Vater, der Erzeuger", "Wir wollen allezeit deine Freunde sein, Indra", "blase meine Nebenbuhlerin fort". H. Wenzel, to whom Havers refers, translates 'einem Mannc, vermalt (Ober den Instrumentalis ir Rigveda, Tubingen, 1879, p. 48). Havers' reference p. 47 to be cor rected as p. 48. 16 Keith's reference 1.34.7 to be corrected as 1.34.5. Madhu Vidya/52

Page #78
M. A. MEHENDALE Sanscrite pp. 296-299, 1930, Grammaire de la Langue Vedique pp. 348-350, 1952. Siddheshwar Varma, Syntax of the Dative Case in the Rgveda, (Ganganath) Jha Commemoration Volume pp. 435-456, 1937, and Gonda, The Unity of the Vedic Dative, Lingua 11. 141 - 150, 1962. Since visne is not the correct instance of the dative in the sense of genitive, and since there is no other parallel in Sanskrit for a similar use of the dative, the remark of Havers (p. 45) that "Nur bei Kategorie V (i.e. which concerns instances pertaining to human relationships) ist auch der symp. Dativ sonstiger Nomina ziemlich vertreten und zwar sogar in rein adnominaler Funktion" is not correct. Madhu Vidya/53

Page #79
SANSKRIT SAKHI By M. A. MEHENDALE The etymology of salchi from khya- has been given in the Unadisutras: samane khyah sa codattah (This has been explained as samanam khyayate janair iti sakha) 4.136. Yaska in his Nirukta (7.30) renders sakhayah by samana-khyanah which suggests the same etymology. Grassmann also derives sakhi from sa khya- and compares it with sam khya- 'to belong to some one'. Bohtlingk-Roth, on the other hand, derive sakhi from v sac- 'to be associated with, to be united with'. It is likely that this etymology was suggested by such occurrences in the RV as sakhya saceya 8.48.10, madhvah pitva sacevahi trih sapta sakhyuh pade 8.69.7, na sa sakha yo na dadati sakhye sucabhuve sacamanaya pitvah 10.117.4. But this etymology from v sac- has been already declared to be very doubtful by Uhlenbeck and Walde-Pokorny. Cunyi has attempted to explain away kh in sakhi as due to analogy with plural forms: *sokew a -i-bhih > sakhibhih and from there *sokwo > *saka > sakha. This explanation, apparently, has not found favour with scholars. In the present paper it is proposed to derive sakhi from *sakha with the secondary suffix -. kha means 'hollow, aperture' or 'a hole in the nave of a chariot's wheel. The compound *sa-kha would mean 'who shares with some one else the same hollow of the wheel' or, by the extension of meaning, who shares with some one else the same chariot' i.e. 'driving with some one in the same chariot'. sakhi, derived from *sakha, may be compared with sarathi, derived from saratha. The lack of veddhi in the first syllable of sakhi is not without a parallel. Debrunner has already called attention to a few forms without veddhi, e.g. nioi, sapti (Alt. Gr. II.2$1906, p. 304). It may be mentioned that among the meanings of sakhi given by Boht. -Roth we find 'Genosse, Freund,' and also "Gefahrte, Begleiter'. The word sakhi may thus refer to the charioteer who drives the chariot, or the warrior who fights from it, or the fellow-travellers who go in the same chariot. The deriva 1. Symbolae Rozwadowski, Vol. I, p. 90 (1927). I am indebted to Prof. M. Mayrhofer for this reference. 2. Khe ara iva RV. 8.77.3; khe rathasya 8.91.7. Madhu Vidya/54

Page #80
SANSKRIT SAKHI 81 tive sakhyam would then mean primarily not 'friendship', but the act of journeying together in the same vehicle'. Occurrences of sakhi and sakhyam in the RV fully support the above interpretation and therefore render the newly suggested etymology plausible. This can be seen very clearly, in the first instance, in those passages where sakhi occurs in the context of a chariot or one of its parts. In RV 6.55, Pusan in the first verse is requested to be the charioteer of rta (i.e. of the hymn looked upon as a chariot) so that the seer and Pusan can travel together: ehi...... sam saca. vahai / rather rtasya no bhava. When therefore in-verse 5, the seer calls Pusan as sakha mama, the expression makes better sense when understood as 'my charioteer'rather than 'my friend'. In RV 3.60 we are informed that Rbhus drive with Indra in the same chariot: indrena yatha saratham (verse 4). When therefore in verse 3 we read about them indrasya sakhyam bhavah sam anasuh we would be justified in interpreting indrasya sakhyam as '(the privilege of) driving in the same chariot with Indra'. Also in RV 4.35.7 where it is said sam rbhubhih pibasva ratnadhebhih sakhir ya indra cakrse sukrtya the passage can be understood as drink (soma) with Rbhus, who distribute precious gifts, whom you, oh Indra, have made your driving companions due to their good work.' In RV 10.168.2 waters are described as going with Vata sharing with him the same yoke and the same chariot: tabhih sayuk saratham deva iyate. When in the next verse Vata is called apan sakha it can mean driving with waters in the same chariot'. In RV 4.4.10 we read about Agni: yas tva svasvah suhiranyo agna upayati vasumata rathena / tasya trata bhavasi tasya sakha....... 'Agni, when some one who has good horses, good gold, approaches you with a chariot containing riches, you become his protector, you become his charioteer ....' Similarly it is said about Soma that he drives with Indra (sakha) when he mounts Indra's chariot: a tisthati rathan indrasya sakha (9.96.2). In RV 9.97.6 Soma is asked to go with the gods in the same chariot (devair yahi saratham); now if in the preceding verse we read indur devanam upa sakhyam ayan we have every reason to understand it as 'when Soma approaches the gods to drive with them in the same chariot'. In RV 7.72.2 in the first two quarters of the verse, the Asvins are requested to come in a chariot with the gods (a no devebhir upa yatam arvak 3. It may be noted that in verse 2 of this hymn Pusan is called rathitamam and sakhayam and in verse 3 dhivatodhivatah sakha 'the charioteer of everyone who has a prayer (i.e. prayer looked upon as a chariot). 4. Visthdh, in my opinion, are not the sub-divisions of Vata thought of as his feminine followers (Geldner), but waters. Madhu Vidya/55

Page #81
82 M. A. MEHENDALE sajosasa nasatya rathena). And in the following two quarters of the verse we read yuvor hi nah sakhya pitryani samano bandhur uta tasya vittam. This can refer to the singers' reminding the Asvins of the fact that their parents had driven with them in their chariot (sakhya pitryani), and that since the singers also enjoy the same privilege, this is the common bond between the singers and the gods. A Muni, impelled by the gods, is described as vatasyasvo vayoh sakha (10.136.5). In the context of the horse, sakha seems to mean 'charioteer' ('the horse of Vata, the charioteer of Vayu'). Similarly, the Maruts who bring the healing remedies with them are addressed as yuyain sakhayah saptayah 'you, the charioteers, (you) the horses' (8.20.23). When in RV 10.27:6 we read, ye niniduh sakhayam adhy u no esu pavayo vavrtyuh, this means that those who have ridiculed the warrior (sakhayam), on them may the rims of the chariot roll'. In the RV, if the horses are called sakhaya, what is intended to convey is perhaps not that the horses are friends of each other but that they share the same chariot, i.e. are yoked to the same chariot. Cf. hari vi muca sakhaya 'unyoke the horses who draw the same chariot' (6.40.1), brahmana te brahmayuja yunajmi hari sakhiya sadhamada asu 'I yoke for you with a prayer at a common Soma. session your two horses who are (customarily) yoked by a prayer, (horses) who draw the same chariot, the quick ones' (3.35.4), a ca tvam eta visana vahato hari sckhaya sudhura svanga 'May these strong horses, who draw the same chariot with a good yoke and who have beautiful limbs, carry you here' (3.43.4). The word sakhi is often used in the context of the verb yuj'to yoke' or some derivative from it and these uses also support the new interpretation of sakhi. Visnu is called indrasya yujyah sakha (1.22.19) which means 'riding with Indra in the same chariot, fit to be yoked', i.e. a constant co-traveller with Indra. When, however, sakhi and yujya are used to denote two different persons, one of them may refer to the charioteer and the other to the fighter, both riding the same chariot. RV 2.28.10 reads yo me rajan yujyo va sakha va svapne bhayan bhirave mahyam aha 'who, oh king told me, the timid one, frightful things in dream, whether the fighter 5. Also cf. RV 10.29.8 where many tribes are said to be striving for a drive with Indra (yatante sakhyaya pirvih). This interpretation of sakhya in this verse is suggested by the context where in the same verse Indra is asked to mount the chariot (d sma tathar na prtanasu tistha yar bhadraya sumatya codayase). For the context of the chariot also cf. RV 10.64.7 pra vo vayum rathayujar purandhinu stomaih krudhvar sakhyaya pusanam. "You make with your praise songs Vayu, Purandhi and Pusan yoke the chariot so that all may drive together (sakhyaya)'. Maruts are called sukhayah of Indra in the context of the horses yoked to a chariot vatasya yuktant suyujas cid asvan kavis cid eso ajagann avasyuh / visve te atra marutan salhayah (5.31.10). 6. Also 3.43.1 priya sakhara vi muca. Madhu Vidya/56

Page #82
SANSKRIT SAKHI or the charioteer'. In 7.19.9 the singers request Indra to choose hem for yujya-type of association with him (asman vrnisva yijyaya asmai). When in the preceding verse we read priyasa it te maghavann abhistau naro madema sarane sakhayah, this can mean 'may we, the heroes liked by you, oh bounteous one, take delight in your protection, in your resort, (we) riding in your chariot (sakhayah). Instead of yujya, we find sometimes the root-noun yuj being used. Indra's association with Pusan in the same chariot during their exploits against enemies is expressed as uta gha sa rathttamah sakhya satpatir yujd indro vrtrani jighnate (6.56.2). Indra is described as making a somin (who offers Soma) his associate in the chariot, salchayam krnute yujam (8.62.6)." But one who does not give an offering or press Soma, he does not get this privilege: atra yujam krnute yo havisman nasunvata sakhyamh vasti surah (10.42.4). Indra, as a swan, associating with heavenly waters and moving with them in the same chariot seems to be the reason why they are described as shy (bibhatsu): bibhatsunam sayujam hamsam ahur apam divyanam sakhye carantam (10.124.9)." The idea of common sharing is neatly expressed in dvd suparnd sayuja sakhay samana vrlesam pari sasvajate (1.164.20) Two birds, sharing the same yoke and the same nave (i.e. the same chariot) embrace the same tree',10 The use of the word sakhi in the context of the verbs a vrt and vah also points to the interpretation 'riding in the same chariot'. In RV. 4.1.1, the gods are said to have commissioned Agni as their charioteer (tvam hy agne sadam it samanyavo devaso devam aratim nyerire). And then in verse three, when it is said, sakhe sakhayam abhy a vavrtsvaaum na cakram rathyeva ramhya it can mean 'Oh charioteer, turn towards (Varuna) who will drive with you (sakhayam), as the wheel (rolls) towards the horse, as the two quick horses (rush towards the goal)'." The singer who wishes to ride together (sakhayam) with the Asvina everyday wishes to turn (their chariot) towards himself (a vam nara purubhuja vavrtyam divedive cid asvina sakhiyan (5.49.1). The 7. In 9.66.18 we read about Soma, vrnimahe sakhydya vrnimahe yujyaya. Even vajra is called yujya sakhi (6.21.7). 8. Also cf. 1.129.4, 4.32.6. 9. Yoga, referring to the yoking of horses to a chariot, appears in connection with sakhi in yogeyoge tavastaram vajevaje havamahe / sakhaya indram utaye 'we, riding (with Indra) in the same chariot, invite Indra who is stronger in every yoking, in every race for a prize' (1.30.17). 10. The idea of journeying together is not present here. 11. Or, combined into one simile as done by Sayana, 'as the two speedy horses (turn) the quick wheel'. Rathya in this case is taken to mean 'horse' and not 'wheel'. 83 12. Similarly Yami says in 10.10.1, o cit sakhayam sakhyd vavrtyam 'May I turr hitherwards (Yama), who has come in a chariot, for a common ride'. This common ride seems to be indicative of marriage relation, see below page 85 Madhu Vidya/57

Page #83
84 M. A. MEHENDALE use of the verb a vah. is found in RV 3.4.1: a deva devan yaja. thaya vaksi sakha sakhint sumana yaksy agne 'oh god, bring here (in a chariot) the gods for sacrifice, a charioteer (bringing) those who drive with him; being of good mind, offer them sacrifice, oh Agni'. The meaning is equally clear in RV 8.2.27, where about Indra it is said: eha hari brahmayuja sagma vaksatah sakhayam 'May the horses, yoked with a prayer (to the chariot), the able ones, bring the warrior here.'13 In RV 7.95.4 Sarasvati is described as going over to the sakhis who, with bent knees, approach her: mitajnubhir namasyair iyana Taya yuja cid uttara sakhibhyah.4 The expression mitajnu 'with bent knees' has caused difficulty because it is difficult to imagine some one moving forward with bent knees. But this difficulty disappears when we understand sakhibhyah not as 'friends' but as referring to the singers who go to Sarasvati in the same chariot. Since they drive in a chariot they have their knees bent while approaching Sarasvati. Instead of mitajnu appears abhijnu in 3.39.5. There Indra is called the charioteer (sakha) going with Navagvas, the fighters riding the same chariot (sakhis). The posture of the Navagvas is described as 'with bent knees' (abhijnu): sakha ha yatra sakhibhir navagvair abhijnv a satvabhir ga anugman.15 It has been mentioned above that the term sakhi could be applicable either to the charioteer or to the fighter riding the same chariot. Hence if we get an expression like sakha sakhyuh srnavad vandanani. (3.43.4) it is better interpreted to mean 'may the fighter hear the praises of the charioteer'. That the heroes going to war were often praised by their charioteers is well known. Hence we read in 8.43.14 tvar hy agne agnina .... sakha sakhya samidhyase 'For, you, oh Agni, are kindled by fire .... as a warrior (is roused) by the charioteer'. 16 Incidentally it seems that the place of the charioteer in the chariot was to the right of the warrior. Indra asks Vayu (cf. Luders, Varuna I. 221) to drive his chariot and 13. Also hay all vide also spausus ta Also 13: Also cf. 10.73.4 where Indra is requested to bring Nasatya with him so that they may all ride together, samana turnir upa yisi yajniam nasatya sakhyaya vaksi. Sakhya also appears in relation with viv yu- 'to disjoin, separate', makir na end sakhya vi yausus tava cendra vimadasya ca vseh / ...... asme te santu sakhya sivani (10.23.7). Also 8.86.1. 14. (Sarasvati) is being approached by the worshipful ones with bent knees. She, yoked with riches (i.e. carrying riches with her), is more bountiful 10 (the singers) riding the same chariot.' 15. When the charioteer (Indra) followed the cows with Navagvas as fighters driving with him with bent knees...' The same picture of sitting with bent knees in a chariot is to be seen in the race (1.37.10), in approaching Agni found in a distant place (1.72.5), in Indra's coming to the worshippers (8.92.3), in the worshippers' wanderings on the broad earth (3.59.3), and in Indra's winning the cows with his singers (6.32.3). 16. Cf. sakhya.... stutan 10.50.2 in the context of battles, races or fights with Vrtra. Also cf. 8.92.33 where singers riding with Indra praise him. Also cf Luders' interpretation of RV 10.73.5 (Varuna I 180 f.n. 2 and II 424). Madhu Vidya/58

Page #84
SANSKRIT SAKHI 85 stand on his right. The two would then kill the enemies: asas ca tvam daksivatah sakha me 'dha vrtrani janghanava bhari (8.100.2),17 Sometimes those who went in a chariot to the battle-field abandoned the chariot and fled back. This is what the gods did when they went with Indra to fight against Vrtra. Indra is therefore advised by the Maruts to take them with him in the chariot so that he may win the war (nrtrasya tva svasathad tsamana visve deva ajahur ye sakhayah marudbhir indra sakhyamh te astv athema visvah prtana jayasi 8.96.7).18 In the light of this explanation of sakhi, RV 8.45.37 seems to have the following meaning: ko nu marya amithitah sakha sakhayam abravit jaha, ko asmad isate 'Which warrior (sakha), if he is not provoked (by his charioteer), oh men, has said to the charioteer, 'he has left me.' Who runs away from us?" Surya, or the daughter of the sun, is often described in the Rgveda as going with Asvina in the same chariot. a yad vam sury ratham tisthad raghusyadam sada (5.73.5), tam vam ratham vayam adya huvema.... asvina.... yah suryam vahati (4.44.1). Surya has been identified with Usas by Hillebrandt. This identification gets support from the new interpretation of sakhi because Usas is called the sakhi of Asvina, i.e. she goes with them riding in their chariot (sakhabhud asvinor usdh 4.52.2 and uta sakhasy asvinoh 4.52.3). In 1.119.5 we are told about a yosa who chose Asvina for her husbands and went with them for a common ride (sakhya) in their chariot: yuvor asvina vapuse yuvayujam ratham vani yematur asya sardhyama vam patitvam sakhydya jagmusi yosavrnita jenya yuvam pat: "The two voices have guided, oh Asvina, the chariot yoked by you for your beauty......20 The young woman to be won, who married you (patitvam jagmusi) for a common ride (in your chariot) chose you two for her husbands". This connection between marriage and common ride in a chariot is also reflected in the dialogue between Yama and Yami. In 10.10.2 when Yama rejects the initial entreaty of Yami he says, na te sakha sakhyam vasty etat salaksma yad visurupa bhavati 17. Cf. the word savyasthr- (or savyesthr-), savyastha- (or savyestha-) 'warrior' (lit. standing on the left) in the Sat. Br. 5.4.3.18; also indrah savyasthas candramah sarathih AV 8.8.23 cited in BR. As mentioned in BR, the word savyestha, however, means a charioteer according to Patanjali on P. 8.3.97 and the Skt. lexicons (AK 2.8.60 and H. 760). [Some editions or commentaries give the reading savyesthr-]. 18. But the Maruts also seem to have once abandoned Indra. cf. yad indram ajahatana / ko vah sakhitva ohate (8.7.31) 'that you abandoned Indra. Who (now) is anxious to ride with you?' 19. Luders (Varuna II. p. 370. f.n. 1). however, says that Usas is called the sakha i.e. the friend' of Assvina because the latter appear early in the morning. 20. asya sardhyam is 'quite unclear' according to Geldner. Sayana, asya rathasya yat sardhyam prapyam adityakhyam avadhibhutam laksyam. 21. Sayana identifies yosd with surya. Madhu Vidya/59

Page #85
86 M. A. MEHENDALE "This kind of common ride with you, the rider (sakha i.e. Yama who has come in a chariot) does not desire, (the ride) by which a common origin (lit. sign) may turn into being of different forms (i.e. of different origins, not blood relationship)'. The word sakhi 'who drives in the same chariot' has also been used in the Rgveda with a slight shift in meaning to refer to those who go in the same ship. In 7.88 we read in verse 4 that Varuna placed Vasistha in his ship2la (vasisthan ha varuno navy adhat) in response to the seer's desire expressed in verse 3 to go on a voyage together in the same ship (a yad ruhava varunas ca navam pra yat samudram irayava madhyam). It is with reference to these journeys in the same ship that the seer uses the word sakhya in plural22 when in verse 5 he asks Varuna, kva tyani nau sakhyo babhuvuh and in verse 6 he refers to himself as the sakhi of Varuna. In the story of Bhujyu who was thrown in the sea and was rescued by Asvina we read uta tyam bhujyum asvina sakhayo madhye jahur durevasah samudre (7.68.7) 'And those who went with him in the ship)23 with wicked intention, oh Asvina, threw Bhujyu in the middle of the sea'. When we come across the use of the word sakhi with the verb tr- 'to cross', or its derivative, it is better to interpret it as referring to the one who goes in a ship rather than in a chariot. In 10.53.8 the context makes this quite clear. asmanvati riyate sam rabhadhvam ut tisthata pra tarata sakhayah! atra jahama ye asann asevah sivan vayam ut tarema abhi vajan '(The river) full of stones is flowing; hold fast, stand up, and cross over, you all going in this ship (sakhayah). We will leave those who were unkind (to us). May we cross over (to the other bank) towards the favourable rewards'. The same rnean. ing is also probably intended when in 10.31.1 we read tebhir vayam susakhayo bhaverna taranto visva durita syama 'May we go well with them in the same ship, may we be crossing over all dangers'. 24 In the Rgveda the prayer of the seers is often looked upon as a chariot. Hence the singers who jointly sing the prayer are called sakhayah 'who ride the same vehicle in the form of the ride the comJointly sing the ed upon as a 23. 21a. According to Luders (Varuna I 320-21) this ship is the sun and the sea in which Varuna and Vasistha sail is the heavenly ocean. 22. If 'friendship' was intended, the seer would have probably used sakhyan in singular and Luders (Varuna, I. 315) who interprets sakhyam as friendship does indeed translate the verse as though the singular form was used. It may be noted that Geldner translates the word sakhayah in this verse as 'die Gefahrten'. 24. 3.9.1 sakhayas tva vaurmahe devan martasa utaye apan napatam subhagam sudiditin supraturtim anehusam / 'The men, travelling in ship, chose the kind god for help, the son of waters who shines brightly, who crosses (the waters) easily, and who is faultless'. Madhu Vidya/60

Page #86
SANSKRIT SAKHI 87 prayer'.25 In 1.5.1 we read a tv eta ni sidatendram abhi pra gayata / Sakhayah stomavahasah 'Come here, sit down, (and) sing (a song) with reference to Indra, you who ride together in a praise. song as a vehicle'. Similarly in 6.45.4 we find sakhayo brahma. vahase !rchata praca gayata 'you who ride together, praise and sing (for Indra) who comes in the vehicle in the form of a prayer'. The same picture of a god riding a chariot in the form of a praise-song and leading it appears in 1.173.9, asama yatha susakhaya ena........ asad yatha na indro vandanesthas turo na karma nayamana ultha so that we may be good riders with him .... so that Indra may stand in a praise-song guiding the praises as the clever worker (conducts) his task'.26 The prayer and the sacrifice are also looked upon as a ship and the priests who, as though, sail in it together are called sakhayah. In 10.101.2 we read mandra konudhvani dhiya a tanudhvan mavam aritraparanit krnudhvam / .... prancam yajnam pra nayata sakhayah / 'Make (the songs) agreeable, stretch forth (your) thoughts, fashion a ship which crosses (the river) with rudders.... (Oh priests) who sail together carry forward (your) sacrifice'.27 When sakhi is used in the context of finding a path, or going to or coming from a distant land, it gives better sense if it is interpreted as driving in the same chariot'. In 1.80.6 we read mandana indro andhasah sakhibhyo gatum icchati 'Indra, delighted with Soma, seeks a way out for those who drive with him'.28 In 1.53.7 we find namya ya indra sakhya paravati nibarhayo namucin nama mayinam 'When you, oh Indra, with Nami as your charioteer struck down Namuci, who knows magic, in the distant land'. And in 6.45.1 we have ya anayat paravatah suniti turvasam yadum / indrah sa no yuva sakha 'Indra who brought here Turvasa and Yadu from a distance, leading them well, that Indra is our young charioteer.29 The interpretation of sakhi (sa-khi) suggested in this paper seems to derive support from the way it is placed between words 25. For priests being called salhayah J. Harmata has a different explanation. He observes, "In the light of these data it does not seem an unlikely suggestion that the word sakhi- had been used since very early times, perhaps from the Indo-European period, to designate members of separate social groups. As a result of social development, when the different occupational groups became more and more sharply delimited, this particular use of the word was generally established in old Iranian, more precisely in the language of the Avesta and in old Persian. This development is reflected in the particularized meaning of the Avestan word hacay- 'member of a priestly college'". AO (Hun garica) 5.195 ff (1955). 26. Also cf. 4.25.1; priests, busy with the performance of a sacrifice or engaged in singing a praise-song, are often called sak hayah, cf. 6.16.22, 3.29.9, 10.88.17, 6.26.8. 5.7.1, 8.1.16, 10.61.25, 5.45.6, 1.53.11, 4.17.18, 4.31.3, 7.31.1, 10.71.2. 7, 8, etc. etc. 27. Also cf. suparah sunvatah sakha 1.4.10, 8.32.13. 28. Also cf. sakheva sakiye gatuvittamo bhava 9.104.5. 29. yesam indro yuva schlcha also occurs in 8.45.1, 2, 3. Madhu Vidya/61

Page #87
88 M. A. MEHENDALE like sa-manasah (sharing common thought) and sa-nilah (sharing common nest) in ud budhyadhvar samanasah sakhayah sam agnim indhvam bahavah sanilah (10.101.1).30 Similarly we have a parallelism between sa-jatah and sa-khayah in 10.103.6: imam sajata anu virayadhvam indran sakhayo anu sam rabhadhvam /31. A note on bhratra In Grassmann's Worterbuch bhratta is equated with bhratstva 'Bruderschaft, Verwandtchaft der Bruder'. In all the four cases where the word occurs in the Rgveda, there occurs also the word sakhi (or sakhya). In this context it seems that it woud be better to interpret bhratra, not as 'brotherhood', but as a primary derivative32 from bhr. (in the sense of being carried or borne'): ** RV. 4.10.8: siva nah sakhya santu bhratragne devesu yusme / sa no nabhih sadane sasminn udhan // "May. our driving together, may our being borne together, oh Agni, be auspicious among you gods. That is our relationship in the house, at the same udder". According to Geldner udhan 'udder' is the sacrifice. But Luders (Varuna II 395) points out that udhan refers to the heavenly udder which is the "Urquell des Lebens im hochsten Himmel". The gods and men have their origin there. RV. 4.25.2 cd33: ka indrasya yujyan kah sakhitvam ko bhratran vasti kavaye kauti / 'Who desires for being joined with Indra, who riding (with him) in the same chariot, who being borne (with him), who is fit for the wise god's favour?' RV. 4.23.6: kim ad amatram sakhyam sakhibhyah kada nu te bhratram pru bravama 'What kind of (Soma) cup was it which served for driving together34 for those who rode in it? When shall we proclaim our being borne with you in the chariot)?' RV. 2. 1.9: tvam agne pitaram istibhir naras tvam bhratraya samya tanurucam / tvam putro bhavasi yas te 'vidhat tvam sakha susevah pasy adhrsah // "To you, oh Agni, (come) men with requests as (one 30. It may be noted that the author of the Padapatha does not separate salchayah with 'an avagraha as he does in the case of the other two words. The occurrence of sakhaya with su-dhuri and su-anga qualifying hari may also be cited as giving supporting evidence. & ca tvam eta visana vahato hari sakhaya sudhura svanga (3.43.4). 32. For the lengthened grade of the root vowel see Alt. Gr. 2.2 8 520b, p. 708. 33. Also cf. 4.25.1. 34. Soma cup looked upon as a chariot. Madhu Vidya/62

Page #88
SANSKRIT SAKHI 89 goes) to one's father, to you, with shining body, (they come) with sacrifice for being borne in it (sacrifice = chariot). You become a son (to him) who dedicates himself to you, you become) his well-disposed charioteer (and) protect him from attack.' It is easy to understand that sakhi which originally meant driving in the same chariot, a charioteer, a warrior' came later to mean 'an associate' or 'a friend'. It is likely that in some passages even in the Rgveda sakhi is used in the sense of a companion, a friend! 35 But as shown above, the meaning 'charioteer, warrior' is better suited to the context in a large number of Rgvedic passages. The stem sakhi- is known for its peculiar declension. The first five forms are derived from the strengthened grade (sakhai-). In the nom. sg., however, we have sakha and not sakhai. For this final a, instead of -ai, Wackernagel has already noted a parallel form agna (loc. sg.) for *agnai.36 It has also been pointed out that Avestan has nom. sg. kava (from kavi-). According to Burrow,37 the 2-stems and the u-stems developed three kinds of declension. The oldest is represented by sakhi- with its nom. sg. sakha.38 The second is represented by a few u-stems with the nom. sg. dyaus, gaus. And the third is represented by the usual type with nom. sg. -is, -us. As regards the declension of sakhi, O. Szemerenyi 39 has a different explanation. In his view the nom. sg. sakha is influenced by raja 'king and sasta 'ruler.' The acc. sg. sakhayam shows that the influence must have come from an -n stem and/or an agent noun in -tar. He concludes, "It seems, then, that, on closer scrutiny, sakha reveals itself as an unique deviation due to analogy, not as an archaic IE type." 35. E. g. 10.34.2, 10.95.15. 36. Alt. Gr. I $ 93, p. 106. 37. The Sanskrit Language pp. 180-181. 38. Besides Avestan kava, Burrow also notes aprata RV 8.32.16. He also calls attention to feminine derivatives like agadri (and manavi) and first members in compounds agna-visnu and kavi-sakha. 39. KZ 73, 193-194 (1956). For Kuiper's view in this regard see his Notes on Vedic Noun-Inflerion p. 64. Madhu Vidya/63

Page #89
MADHYE LAGNAM (BAUDHAYANA Sr. S. 9. 3) BY M, A. MEHENDALE Among the various rites which precede the actual day of pressing Soma in a Soma sacrifice occurs Pravargya.. This rite is considered to have been once an independent sacrifice before it became part of a Soma sacrifice. In essence it consists of offerings of hot milk of a cow and a she-goat to Asvina. The milk is first heated in a specially prepared earthen vessel, called Mahavira, which looks like three pots placed on one another. While describing how the Mahavira pot is to be constructed the Baudhayana Sr. Sutra has used the phrase madhye lagnam which has caused some difficulty in understanding it. The text of the Sutra (9.3) reads as follows: pindam karoti makhasya siro'siti/ ... atha trtiyam mrdo'pacchidya trin pindan karoti/ tesam ekam kartre prayacchati gayatro'siti, tenasya budhnam karoti... / atisistayai mrdo'rdham prayacchati traistubho'siti) tenasya madhyam karoti... / sarvam antato mrdan prayacchati jagato'siti / tenasya bilum karoti... / tam pradesamatram prthubudhnam madhye lagnam karoti / The meaning of the passage, except the phrase madhye lagnam at the end, is not difficult to follow. The priest cuts off a third of a big clay ball and divides it into three smaller balls. These he hands over to the artisan one after the other and the latter prepares from them the bottom, the middle part, and the top bowl of the Mahavira. The whole pot has to be one span in height, broad at the bottom and madhye lagnam. In an unpublished monograph on Pravargya, J. A. B. van BUITENEN takes madhye lagnam to mean "joined in the middle' and finds it difficult to understand. He observes : "Baudhayana's language is clear enough; by "half of the remaining clay" we should understand the second clay ball, by "all the remaining clay" the last of the three balls. The picture which thus arises is of three clay balls placed on top of one another, the bottom one flattened out to provide a broad base, the top one hollowed out to provide a receptacle. Curious is Baudhayana's phrasing prthubudhnam madhye lagnam "broad of base and joined in the middle". What middle, since the three elements have two joints? The words read naturally that that part which has a broad base, viz. the central segment, is joined with the top segment, and it is this joint that is described as madhyam". 1. For the latest troatment of Pravar ya, soe H. LUDERA, Varuna II. pp.359-388. 2. Ed, by W. CALAND, P. 268, Madhu Vidya/64

Page #90
194 van BUITENEN thus tries to get out of the difficulty by assuming that essentially the pot is to be looked upon as having only two parts (by excluding the flat bottom). In that case it will have only one joint in the middle, i.e. between the part prepared from the second clay ball and the top receptacle prepared from the third clay ball. According to him, madhye does not refer to the middle part of the vessel but to the joint in the middle. But this is hardly satisfactory. The three parts of the vessel have been clearly distinguished from one another as budhna, madhya, and bila, and these, when placed on one another, should have two joints. None of these two joints could be looked upon as the middle one. Moreover, even if we assume that there are only two parts of the vessel, what purpose is served by stating that they are 'joined in the middle'? The whole difficulty disappears when we understand lagnam not as 'joined but as thin, narrow'. This is suggested by the fact that the word lagnam is intended to contrast with prthu 'broad'. According to the Baudh. Sr. sutra, then, the Mahavira pot is required to be broad at the bottom' (pethubudhnam) and thin or narrow in the middle' (madhye lagnam). The word lagna with the prefix vi is already attested in the epics and the classical literature in the sense slender, thin. To cite from BoHTLINGKROTH we get vilagnamadhya having slender waist' (Mbh. 1. 157. 6), vilagnamadhyah (Mbh, 3. 112.4), vedivilagnamadhya (Mbh. 4.35. *629 (3), Kumara, 1.39) to which APTE's revised dictionary adds vilagnamadhyaya (Vikramo. 4. 37)]. Pali also has vilagga and vilaka slender' which go back to Skt. vilagna. In the opinion of LuDERS Pali vilaka is a case of hyperpalism for the form vilaga occurring in the eastern dialect of the original Buddhist canon. The word lagna in the sense slender, thin' has not been recorded in the available lexicons. It will now be clear that it does occur in this sense in the Baudh. Sr. S. M. A. MEHENDALE That this meaning narrow, thin' of lagnam is what is intended in the Baudh. Sr. S. will become clear also from the other sutra sources where the preparation of the Mahavira, or the Vayu vessels which are similar to it in form, is referred to. In the Ap. Sr. S. occurs the word samnata in 15. 2. 14: mahaviram karot!... pradesamatram urdhvasanum uparistad asecanavantam madhye samnatam vayavyaprakaram "... verfertigt er den Mahavira... Dieser soll eine Spanne hoch sein, einen erhobenen Kopf (Oberteil) und im oberen Teile eine Vertiefung zum Eingiessen der Flussigkeit haben; in der Mitte soll 1. For a kosa with three parts clearly distinguished from each other as budhna 'bottom', udara middle', and bila top', cf. Ch. Up. 3. 15, 1: antariksodarah koso bhumibudino na jiryati......dyaur asyottaram hilam. 2. Beobachtungen uber die Sprache des buddhistischen Urkanons, SS132, pp. 106 Madhu Vidya/65 107,

Page #91
madhye lagnam ( BAUDHAYANA SR. . 9. 3) 195 er eingeknickti sein, wie die Vayugefasse" (CALAND). The Vayu vessels referred to here are earlier described in the Ap. Sr. S. (12. 1. 4) with almost identical words where the commentator says that these vessels resemble a mortar in shape. The Satyasadha Sr. S. 24. 1. 15 simply says that the Mahavira vessel should be vayavyaprakaram. But in the description of the Vayu vessels it uses the word samnata as in the Ap. Sr. S. Instead of sanata, the Manava Sr. S. 2. 3. 1. 13 uses the word sasakta* in the description of the Vayu vessels. When this satra is quoted by the commentator Gopinathabhatta on the Satyasadha Sr. S. 8. 1. 2 he has the reading samlagnamadhyani (instead of samsaktamadhyani as in van GELDER's edition). The commentator adds that this has been explained as samlagnam krsam madhyam yesam tani 5 . those whose middle portion is thin'. The Katyayana Sr. S. 26. 1. 16 uses the phrase madhyasanglhitam in the description of the Mahavira vessel. In the commentary, this is explained as madhyapradese sankucitam ulukhalavat mustigrahanayogyam 'narrow in the mid region, like a mortar, capable of being held in a fist.' 1. in der Mitte gebogen' GARBE, ZDMG 34. 330. 2. ulakhalakstinity rektan Whavati. 3. madhye san natani vayavycini 8. 1. 2. Also cf. Bharadvaja Sr. S. 13. 2. 6 (where the phrase is translated as contracted in the middle' by C. G. KASHIKAR ). 4. sam saklamadhyani translated by J. M. van GELDER as in the middle bent inward'. 6. its bhasyakrtii vyakhyalam says Gopinathabbatla. Madhu Vidya/66

Page #92
NOTES AND COMMUNICATIONS SAHASAH SVAJAH IN THE AITAREYA BRAHMANA, 13.2 (PANCIKA 3.26) At the beginning of the thirteenth adhyaya of the Aitareya Brahmana occurs a story which narrates how the metre Gayatri brought from heaven soma for gods and sages. Having gone to heaven, she frightened the guardians of soma, picked it up in her feet and mouth, and started to return. At this point, one of the guardians of soma, Krsanu by name, shot an arrow at her which cut off the nail of her left foot. From that nail arose a porcupine (salyakah), and the fat (vasam) that flowed became a barrer cow (vasa). The Brahmana text (as read by Keith, following Aufrecht) then goes on to say: atha yah salyo yad anikam asit sa sarpo nirdamey abhavat sahasah svajo1 yani parnani te manthavala yani snavani te gandupada yat tejanam so 'ndhahih so sa tathesur abhavat. Keith (Rgveda Brahmanas (HOS, xxv), Cambridge, Mass., 1920) translates the passage as follows: The socket and the point became a serpent, not biting 3; from its swiftness (came) the viper; the feathers became flying foxes, the sinews earthworms, the shaft the blind snake. Thus became the arrow'. 1 Haug (Bombay, 1863) sahasasvajo, Bibl. Ind. (Calcutta, 1895-1906) and NSP (Bombay, 1925) sahasa svajo; Aufrecht (Bonn, 1879), ASS (Poona, 1896), and TSS (Trivandrum, 1942 ff.) sahasah svajo. Out of the different parts of the arrow mentioned here--salya, anika, tejana, parna, and snavan the first four also occur earlier in Ait. Br. (4.8). The meanings of parna and snavan are not in doubt. The salya socket' (Keith) has also been understood by all the Indian commentators as the metallic arrow-head (e.g. salyah krsnayasanirmito banagre sthapitah Sayana). But about the meanings of anika and tejana, there are differences of opinion, ayana even being guilty of inconsistency. In Ait. Br., 4.8, he interprets anika point (Keith) as 'shaft' (patrayuktad banamulad urdhvavarti bhago mukham) and tejana shaft (Keith) as the sharp point of the metal head (tasya lohasya tiksnam agram). But while commenting on Ait. Br., 13.2, he reverses this position and takes anika to mean the tip of the arrow-head (salyasya yad anikam mukham) and tejana to mean the shaft (lohapatravyatiriktam kastham). Sadgurusisya, whose commentary Sukhaprada has been published in the Trivandrum edition of Ait. Br. (1942), agrees with the latter explanation of Sayana (anikam agrah or barasyasyam and tejanam yastih). But Bhattabhaskara, excerpts from whose commentary have been published in the Trivandrum edition mentioned above, seems to agree somewhat with the former explanation of Sayana. According to him, anika means mukha and together with salya, the metallic arrowhead, it constitutes the danda of the arrow. But he differs from Sayana in that he does not take tejana to mean sharp point' but the part of the arrow which is placed on the bow-string (jyanidhanasthanam). It seems, however, that he refers to Sayana when he says mukhanaisityam ity eke. On TS, 6.2.3, where anika, salya, and tejana occur as names of the parts of an arrow, Bhattabhaskara explains them as anikam mukham (isoh), salyam sariram puccham va, and dharam kastham va. 4 tejanam This is following Sayana (damsanasamarthah sarpah). According to Sayana, it is a non. poisonous water-snake called dundubha. Sadgurusisya also says that it is a non-biting snake but adds that it subsists on earth (mrdbhaksi dambanaksamadantakah). According to Bhattabhaskara, however, nirdamsi nitaram damsanasilah. = According to BR svaja viper' is attested only in the AV and Ait. Br., according to MW also in the Ap. Sr. S. 5 So also Sayana and Sadgurusisya. Bhattabhaskara, however, says that it means mahasarpah (ardhahir seems to be a misprint for andhahir). Madhu Vidya/67 * 4

Page #93
NOTES AND COMMUNICATIONS 377 From the above translation it will be clear that Keith treats sahasah as abl. sg. of sahas which he translates as 'swiftness'.. He has in this followed Sayana who explains sahasah svajah as saho vegas tasmat sahaso banavegat svaja ubhayatah sirah sarpo 'bhavat. The only difference is that while Keith renders svaja as viper', Sayana takes it to mean a serpent with mouths at both ends.? This interpretation of sahasah svajah is, however, not correct. For, apart from the point whether sahas can mean vega, it will be observed that in the above account the coming into existence of various creatures is shown to take place from certain concrete objects like the nail, fat, etc. It is therefore very unlikely that one of the creatures, svaja, would be shown as arising from an abstract element like swiftness (sahas). Secondly, for this account of creation the author has chosen & particular kind of sentence pattern in which the source material and the resulting creature both appear in the nominative case, e.g., yad vasam (astavat), sa vasa (abhavat), yah salyah } (asit), sa sarpah (abhavat), yad anikam yani parnani, te manthavalah, yani snavani, te gandupadah, yat tejanam, so 'ndhahih. Hence, if the author had really intended to inform us that the svaja arose from sahas, he would have expressed it as yat sahah, sa svajah. He would not have abandoned his sentence pattern for one instance and given in it the source material in the ablative case. These possible objections to the interpretation of sahasah svajah have been implicitly endorsed by the commentator Bhattabhaskara. He separates sahasah into two words as saha sah and understands the entire line atha yah salyo yad anikam asit sa sarpo nirdamsy abhavat saha sah svajah as meaning that the serpent called nirdamsin arose from salya and the serpent called svaja arose from anika (etad uktam bhavati--yah salyah sa nirdamsi nama sarpo 'bhavat / yad anikam sa svajo nama sarpo 'bhavat). He understands salya (karsnayaso bhagah) and anika (mukham) * together to constitute the danda of the arrow. He says that the two parts of the arrow, the salya and the anika, are mentioned together in the same relative clause in order to indicate the simultaneous coming into existence of the two kinds of serpent from them (ubhayoh sahopadanam tatparinamayoh sarpayoh sahotpattikhyapanartham). He further justifies the use of saha in the principal clause on the ground that it is indicative of the two serpents coming out together from the same danda (tav ekasmad dandat saha jayete / anenabhiprayena sahagrahanam karoti). But this explanation of Bhattabhaskara is not convincing. In the first 6. vehemence' Haug. So also Sadgurusisya and Bhattabhaskara. & cf. the meaning of mukham in the first explanation of Sayana, p. 376, n. 2, above. Madhu Vidya/68

Page #94
378 NOTES AND COMMUNICATIONS instance, his relating anikam (= mukham) to the shaft of the arrow and understanding salya and anika together to constitute the danda is not satisfactory. The salya and the danda are two distinct parts of the arrow. Secondly, the Brahmana passage does not speak about the simultaneous coming into existence of the two kinds of serpent. That is Bhattabhaskara's guess. Actually there is nothing to prohibit us from imagining that all the reptiles mentioned in this section came into existence together from the different parts of the arrow. Obviously then, the Brahmana mentions salya and anika together in one clause simply because they are not really two separate parts of the arrow, anika being only the point of the metal head salya." And if we assume that the two were mentioned together in order to indicate the sahotpatti of the nirdamsi serpent and the svaja, then there was no necessity to use saha again to indicate that they were produced together (saha jayete). Even if we understand Bhattabhaskara as pointing out that the mention together of salya and anika underlines the sameness of time, and the use of saha in the principal clause underlines the sameness of material (ekasmad dandat), even then this latter purpose is already well served by putting together salya and anika in the same clause. There is thus no room for saha in this context. It is therefore better to separate sahasah not into two but into three words as sa ha sah. The expression sa ha sah, thus obtained, can be interpreted in two ways. (1) The first sah may be taken to refer to anukam mentioned in the previous sentence. The sentence thus obtained would read as yad anikam asit sa ha sah svajah 'what was the point, that indeed (became) that svaja'. But in this interpretation we cannot give good reason for mentioning salya and anika together, if they were again to be separated for understanding the rise of two different kinds of serpent from them. Moreover, the sentence, thus obtained, differs from the general sentence pattern in that it shows the emphasizing particle ha and one more sah in the principal clause, which is not the case in the other corresponding sentences. (2) In the second interpretation we do not take sa ha sah svajah as referring to the rise of a serpent (svaja) different from the one mentioned previously (nirdamsi sarpah). We understand it as only giving the name of the serpent which was referred to earlier. The word nirdamsi' not biting' describes 10 the serpent whose name is svaja. The passage beginning with atha yah salyah therefore may be translated as: Now what was the head (of the arrow), what the point, that became the serpent, not biting. That, indeed, (is) that svaja. The construction sa ha sah (svajah), assumed above, has a parallel in tad When in Ait. Br., 4.8, an arrow is said to have three (tripandhi) or four elements (catua. sandhi), anika and balya are counted separately, the third and the fourth elements being tejana and parna. But when an arrow is said to have two elements (dvigandhi), only aalya and tejana are mentioned, anika being then included in the salya. 10 As was seen above, Bhattabhaskara takes nirdambi to be the name of the serpent. This is unlikely. In that case the Brahmana would not have added sarpah before it. The word andhahib blind snake' which occurs at the end of this section is also a description. Obviously that kind of snake did not have a special name. Madhu Vidya/69

Page #95
NOTES AND COMMUNICATIONS 379 dha tat (paran eva) 'that (artvijya), indeed, (is) that (which simply goes away,11 i.e. is futile, does not help the sacrificer)', Ait. Br., 15.2 (pancika 3.46). The repetition of the pronoun tat does not amount to tautology, If sahasah is not to be taken as one word but as two, as suggested by Bhattabhaskara, or three, as suggested in this paper, then we should not get the visarga after the second sa. This latter sa has been interpreted as the demonstrative pronoun and hence it should lose the visarga before any consonant.12 We expect the text in that case to read as sa ha (or saha) sa svajah. Now among the published editions of Ait. Br., that of Aufrecht, the Anandasram edition, and the Trivandrum edition, all read the text with the visarga as sahasah svajah. But Haug's edition of Bombay and the Bibliotheca Indica edition of Calcutta read the text without the visarga as sahasa svajah or sahasasvajah. This, then, seems to have been the original text. A corruption of sa ha sa svajah to sahasah svajah in the oral tradition, especially when the text has been handed down in an unaccented form, is easy to understand. In a continuous recitation sahasasvajah and sahasah svajah are heard alike.13 M. A. MEHENDALE 11 remoto 'Keith, nikrstam Sayana, asubham Sadgurusinya. 11 Pan., 6.1.132. 13 A Sanskrit rendering of this paper will appear in the Pandit Rajeshvar Shastri Dravid felicitation volume. Madhu Vidya/70

Page #96
SKT. VIGIPRIYA M. A. MEHENDALE Deccan College, Poona. This vocable appears in a mantra which is used while drawing one of the five grahas which are prescribed specially for the Vajapeya sacrifice. Its meaning given by BR is : "etwa ohne Backenstucke d. h. ohne Handhaben an den Seiten, von Soma-Gefassen". Bohtlingk probably was not sure of this meaning and hence in his shorter dictionary he did not give it. There he simply says that the word has been variously interpreted by the commentators. Monier Williams only repcats this latter statcmcnt. The mantra in question appears as follows in the Taittiriya Samhita 1.7.12.1-2: yc grahath pancajani nu yesai tisrah paramojah/dalvyah kosahi Sumubjitah/tisuri vesiprijanam isam urjain samagrabhim esi te yonir indraya tval ... (2) Keith translates the above as : "Thic cups of the five folk, Of which three are of highest birth, ( And for which ) the divine cask (1) has been forced out Of these that have no handles The food and strength have I seized; This is thy birthplace; to Indra thee !" Madhu Vidya/71

Page #97
324 M. A. MEIIENDALE Keith's foot-note 1 reads as: "risipriyanam is apparently 'without sipras', and sipra denotes jars' or 'lips', i.e., vessels without mouths or handles, probably." 337 In the Vajasaneyi-Samhita (9.4), the mantra has the following text: gruha urjahutayo vyunto vipraya matim t'sam visipriyanam vo 'ham isam urjam samagrabham/ This mantra, when it appears in the Sat. Br. (5.1.2.8), has been translated by Eggeling as follows: "Ye cups, of strengthening libations, inspiring the sage with thought, I have gathered together the pith and sap of you, the handleless!" "Eggeling and Keith who interpret visipriya to mean 'handleless' have obviously followed the lead given by BR in analysing the word as vi-sipriya. This analysis was already suggested by the Padapatha of the Taitt. Sam. Following it Sayana renders the word while commenting on the Taitt. Samm. as furnished with different kinds of rims'. He interprets sipra (neut. ) as jaw' and thinks that with reference to the cups it is metaphorically used in the sense of rims'. Read: vividhani siprani hanusthaniJani patragrani tair upeta visipriyah. While commenting on the Sat. Br., Sayana refers to the Nirukta view (6.17) according to which sipre (fem. dual) means 'jaws' (hanu) or 'nostrils' (nasike), and says that in the word visipriya it means 'jaw'. His interpretation of the word is the same as above except that instead of 'tair upetam he says here 'tatparyantam' i.e, 'filled upto the 1. One, however, does not understand how this meaning can be reached if sipra means 'jars'. The Taitt. Br. 1.3.9.2 has no comment on this word. 2. The Sat. Br. also has no comment on it. 3. Bhatta Bhaskara also has the same interpretation; visipriy anam vividhahanuparyanlanam. Madhu Vidya/72

Page #98
SKT. VISIPRIYA 325 rims'. But this is not all. He also gives an alternative interpretation which, since it is given first, he seems to prefer. According to this interpretation, sipriya means 'chewing', the action done by the jaws (sipra). Since the Soma stalks are well pressed, the juice extracted from them is fine and needs no chewing at all while drinking. Hence the cups, i.e,, the Soma juices contained in them, are called visipriya "containing well pressed Soma". Read : sisipre hani nasikt va'' iti hi yaskavacanam/ iha hi hanii grhyetel tatra bhavain karma sipriyam vigatam sipriyan yesari tel atra hanuvyaparabhavat sobhanabhisavasarnskyta ity arthah5 / Another way of explaining the word would be to analyse it differently as visi-priya 'lovcd, liked among the people'. This, as an attribute of the Soma cups, suits the context very well since these cups are also called pancajanina 'belonging to the five pcoplc' in the Taitt. Sain, and visvajanina 'bclonging to all the people' in the Maitr. Saia, (1.11.4 ) and the Kith. Sam. ( 14.3). The mantra is used, as incntioned above, while, drawing thc Soma cup of thc Vajapcya, a sacrifice which is charactcrized by ccrtain features like thic chariot racc which must have been very popular among the people. Hencc it is quite understandablc that a Soma cup of the Vajapcya was called visi-priya 'liked among the people'. If visipriya is analyscd as visi-priya, thc loc. sg. form of the first member of the compound is quite under 4. Therc is also another minor difference. Whilc commenting on the Taitt. Saii., he equatcs sipra with palragra; in the Sat. Br. he cquated sipriya with patragra. 5. The explanation of the commentator Mahidhara on the Vaj. San. (9.4) is the same. Read : siprayor hanvoh karma sipriyam hanucalananul vigatan sipriyan yesu grahesu te visipriyah samyag abhisulah supulus cu latra hi hunyor vja putu musti sufijulual. Madhu Vidya/73

Page #99
326 M.A. MEHENDALE standable. As Wackernagel (Alt. Gr. II, I, 8 95c 8, p. 234 ) puts it, "Bei solchem Verhaltnis der Glieder ist Kasusform des Vordergliedes haufig". The adjective priya is again often used with the loc. case. But it is difficult to explain the accent. As a rule, the compound then should have been accented as vistpriya. (cf. Wackernagel, op. cit., 896, p. 238 f.) The shifting of the accent to the first syllable, as seen in the attested form, may be said to have occurred on the analogy of many compounds with prefix vi as the first member showing the accent on the first syllable. We have, of course, then to assume that the accent shift had occurred very early even before the Pada analysis of the Taitt. Samhita. The interpretation of the word visipriya can be attempted also from a different direction. In the corresponding mantras, the Maitr. Sari ( 1.11.4 ) gives the reading visisna and the Kathaka ( 14.3) visisnya. If the reading visisna is taken to bc thc original onc, it is possible to understand it as the description of a Soma cup. The handle of a Soma cup, as something projecting from it, can be looked upon as its sisna (organ or tail ). visisna as applied to a Sona cup, can mcan 'without a handle'. sipra in ovisipra may then be looked upon as a euphimistic substitute for sisna in visisna. #uisiprya would be a further deviativc from #usipra as visisnya of thc Katliaka is from visisna of the Maitr. This avisiprya could be the basis of the attested visipriya." If sipra can be considered as a euphimism or sisna, then it would also be possible to understand better the 6. cf. priyo viksu RV 6.1.6: priyah saryo priyo agna bhavati RV 5.37.5 etc. 7. On thc variation between -ya and -iya cf. All. Gr. 11.2 SS 228 b p. 359. Madhu Vidya/74

Page #100
SKT. VISIPRIYA 327 name of the demon (dasa) Vrsa-sipra (RV. 7.99.4 ). This has been rendered differently as 'die Lippen eines Stieres habend' ( Grassmann ), <Page #101
328 M. A. MAHENDALE sipra in the Padapatha, but it does not make it intelligible. Hence when Sayana renders it as vigatahanu he seems to favour the analysis vi-sisipra. If, as suggested above, it is possible to isolate the root sip then one may agree with Sayana in his analysis of the word and derive sisipra from the reduplicated root sip. But instead of assigning the meaning hanu to osisipra, it would be better to consider it either as a substitute for sisna or connect it with sepa and give it the meaning 'organ'. Vi-sisipra then would mean 'one having a distinct (vi) organ' and, as a name of a demon, it could be compared with the other name of the demon Vrsasipra noticed above. Madhu Vidya/76

Page #102
DAGH- (VERB), -DAGHNA (SUFFIX), AND DERIVATIVES A STUDY IN HISTORICAL LEXICOGRAPHY M. A. MEHENDALE Deccan College, Poona ABSTRACT: Sample entries that could appear in a historical dictionary of Sanskrit have been presented for the root dagh- and 6 related forms and for the taddhita suffix -daghna and 31 derivatives ending in that suffix. I dagh 1.c dagh verb (for IE etymology, Walde-Pokorny 791 ] qeyfa (4), saila (5), get (1, 6?). (FET, EH, E417, 28. ) to go (na ) Nigha. 2. 14; to flow (Faf9.4) Nir. 1. 9; to kill, to protect ( ) Dhatu. 5. 27. Used only with adverbs or prefixes, the basic meaning apparently being : to reach a certain level, to flow at a certain level (see 37TCET) which is modified by the adverbs adhaH, pazcA or pazcAt, and paraH or the prefixes ati, A, pra and prati. (a) with 374' to make something reach below normal level' t u rer.... HIST T'he should not raise the fire) too high...he should not make it reach low 'Katha. 8. 12; Katha 7. 7. (b) with pazcA or pazcAt 'to remain behind ' pazcA sa daghyA yo aghasya dhAtA 'may he, who commits sin, remain behind' RV 1. 123. 5; AT 9475 Le Tout fagio RV 7. 56. 21. (c) with m' to go beyond ' A1 TT-AT TE' do not lag behind, do not go beyond' (addressed to Bhaga) Paippa. 19. 37. 8( see aface). 2. CE atidagh ( ATH) 'to reach beyond, to pass over' F1 azt hud Tut' a' be of help to the praisers, may not good luck reach beyond us' RV 2. 11. 21. This line has been paraphrased by Yaska (Nir, 1. 7) as Althafardetaisaiafagia ar: 'do not give beyond us, abandoning us do not give (to others ). Apparently he derives from 6 6 to which he ascribes the meaning to give'l. Ara at Higitt -AT OR amiga alla! 17 may not a wolf (or) a she-wolf attack you, do not avoid ( us ) and do not go beyond (us) 'RV 1. 183. 4. 1. Sayana also assigns the meaning 'to give and also in 1.183.4. (Grif) to in this passage Madhu Vidya/77

Page #103
240 M. A. MEHENDALE 3. B y adagh (9, (i) 'to flow over, to overtake!? utrafa. 9: TR1 a 37 'Sarasvati...do not go away (from us ), do not overtake us with water 'RV 6. 61. 14; Katha. 17. 18, 30. 3; Katha 46. 6; Maitra. 4.11.2. (ii) to overtake, to attack' (with loc. ) h a rda farz TT when associated with you, do not overtake our own progeny (i. e. do not destroy them)' RV 7. 1. 21; (with acc. ) 19: H kurqat b'do not attack our wish which makes (us) great 'RVI. 178.1; 1 terat fe may not the Gandharva Visvavasu attack (you )'Taitt. Sam. 1. 2. 9. 1; Baudh. Sr. 6. 16. 4. all pradagh ( alarcunt, at: Ved. inf; ma: (?))' to move on (without stopping or returning )' fn9 ar ga is geen a faqAIT #Ha gaudia ' he who takes Visnu-steps is likety to move on going far away; he (therefore ) returns with four (verses). Taitt. Sam. 5. 2. 1. 2; Sarita Q9 ER:191 g atte Taitt. Br. 1. 3. 7. 7. The sat. Br. 13. 1. 3. 4 and 13. 2. 1. 6 give the text as Izvaro vA eSa purAG pradughoryaH purAcIrAhutIrjuhoti. Bo-Ro. point out that the text is here wrong for agat gratiafastarat: 917477 fa' so that they ( 4511: ), wandering without obtaining a stable place, will not keep moving on.' Jaim. Up. Br. 3. 4. 1. 4. 5. pratidaghU pratidagh (dadhnuyAt ) to go against, oppose ' guruM ca yaH pratidadhnuyAt ...gagafa' and one who opposes his teacher... these are minor sins' Vasistha. 1. 23. Schmidt (Nachtrage ), however, feels that afacerea might stand for fact1417. 6. E daghvan used uniquely in 379 TEL 7. 3792 dpascaddaghvan (i)'not remaining behind 'STATE THIS4837 ''for the hero who comes on time, comes often, and who does not remain behind 'RV-6.42.1; SV 1, 352,2. 790 ( also read as 374876a); (Jaim.) 1. 37; Taitt. Br. 3. 7. 10. 6; Ap sr. 14. 29. 2. (ii) 'not falling short of (acc.)' staat er zeri generala wafa Maitra. 3. 9. 4.; Ap. Sr. 7. 28. 2; Manava Sr. 1. 8. 6. 22. AV 19. 55. 5 reads 370 JT Cat 7 which, on the basis of the Maitra. reading, is emended by Whitney-Roth as 32778ealt . Note the use of the gen. Some mss. of the AV, however, read 38 (acc.) for 875161 (gen.). from 2. In all the RV passages cited under this entry Sayana derives with the meaning 'harass' (an) or 'burn'. Madhu Vidya/78

Page #104
A STUDY IN HISTORICAL LEXICOGRAPHY 241 II daghna 8. El daghna taddhita suffix expressive of ( perpendicular) measure. It is used to measure both depth as well as height. Probably it was originally used to indicate the depth of water (cf. 371597) measured by the different parts of the body. Its use was then extended to measure in a similar way the depth of the dug out ground as also the height above the ground. In late classical literature the suffix is found used in a few cases with words which are not body parts of. STAGET, EREHTE, fotog). Yaska (Nir. 1. 9 ) gives the etymology of 76 as oa: gar : 1 graaf ES FA ala i. e. from the verb tren'to flow' (indicating rise in the water level ) or from the verb ac' to be exhausted, reduced' (indicating fall in the water level ). See ziedi azala gazteak #TOT I 3941vai UGT FAT1T. The derivation of a from TL'to flow' may be right since its earliest occurrence in 3717ET expresses depth of water reaching upto mouth. Incidentally, in 371997, TET may not be a suffix but a full word. It may be looked upon either as a verbal noun in -a meaning flowing' or a ppp. in- na meaning "flown.'(For the formation of o see Alt. Gr. II 1, 181 and II 2, 723 ). Panini, however, teaches it only as a secondary suffix (5. 2. 37). He also knew its fem. in -i (-daghni, P. 4. 1. 15 ). Debrunner in Alt. Gr. II 2, 375 observes : however, in pre-classical literature only -daghna is attested. The examples cited by Debrunner are atradaghnas, nabhidaghna, and amsadaghna all from the Sat. Br? But daghni feminines are also found with janu in the Taitt. Ar., Ap. Sulba, with nabhi-and asya- in the Apa. Sr., and with amsa- in the Manava Sr. Not all these examples of course can be looked upon as pre-classicaj. The commentator on the Katy. Sr.:(7.9. 27 ) affirms-daghna, as opposed to-daghni taught by Panini, as Vedic.On nAbhidaghnA he says nAbhiH pramANaM yasyAH sA Cet fa ma 417CERIT :. It is interesting to note that the-daghna fem. forms are, with the exception of the one in the Hiranya. Sr., exclusively restricted to the Sukla YV tradition, while the daghni fem. forms, which alone are known to Papini, occur in the Ktsna YV tradition. It has, however, to be admitted that the occurrences of the fem. forms that have been discovered so far are relatively few. The suffix occurs over whelmingly after words expressive of body parts. The only three attestations that have come to light go far in which is used with words other than those for body parts are all from later litarature. Of these, 375 is doubtful. It is supposed to occur in the Rama, but the line in which it occurs has not been accepted in the critical edition (see s. v.). The other two words, from classical literature, are FRATE (Malatima.) 3. nabhidashna occurs also in the Katy. Sr., the Hirany. Sr., and in the very late work Manasollasa. 4. The case of a n and far is omitted for obvious reasons. Madhu Vidya/79

Page #105
242 M. A. MEHENDALE and foccer (Bhatti. ). In the Vedic literature, the use of ce after a word which is not expressive of a body part is strictly avoided. In the contexts in which its use would seem natural the idea is differently expressed, e. g. we read ATT Tokiula: and not HTCT : Taitti. Sam. 6. 3. 4. 5, Haat aget, and not agen: Katha. Sam. 26.4, Katha Sam. 41. 2, yaits: ganglan kSatriyasya kuryAt, and not udvAhapuruSadaghnaM kSatriyasya kuryAt Sat. Bra. 13. 8. 3. 11. Or, the suffix mAtra is used in the place of dadhna, e.g. atha puruSamAtre juhoti Katha. Sam. 21. 6, Katha Sam. 31. 21, 7: TGTH: a Katha. Sam. 21.4, Katha Sam. 31. 19, yajamAnamAtrI audumbarI bhavati Sadvim. Br. 4. 3. 2. Panini teaches the suffix 2, along with gue and A1, to express measure (THO UHC65H128: 5. 2. 37 ) In the following sutra he teaches the use of an additional suffix aN after puruSa and hastin (puruSahastibhyAM aN ca 38). This has been interpreted to mean that after the words 767 and least, to express measure, one may find the suffix aqu in addition to all the three suffixes of the previous . This leads us to believe that Panini knew four forms each from y#9 and hastin , viz pauruSa (satra 38 ), and puruSadvayasa, puruSadaghna, puruSamAtra (sutra 37), and similarly efe und other forms If this is true we have to admit knowledge on the part of Panini of forms like puruSaghna and hastidadhna in which dughna is affixed to words which are not body parts. However, in view of the fact that there appears to be in Sanskrit literature complete absence of the forms puruSadadhana and hastidaghna (they have been given as examples only since kAzikA) and in view of the fact that in the Vedic texts 77 is affixed only to words expressive of the body parts, it is worth while considering whether the above interpretation of sutra 38 is correct. Perhaps the suira 38 has to be interpreted differently and we have to draw only the last mentioned suffix 197 from the previous and not all the three. This would give us only two forms each: 0167, 69417 and hAstina, hastimAtra. The derivatives with qe are attested only with three case terminations, viz. those of nom., acc., and loc. Of these, the latter two are used both as adjectives and adverbs. The derivatives are usually used singly. A compound of two words is found in the Ea r t: (Manava sr. 4. 4. 8.), and a compound in which can is related to more than one body part is iC451CG TETOISETH (Ragha. Panda, 15. 47). III. Secondary Derivatives formed with the Suffix -daghna Under this head are collected as many formations with 57 as have come to notice so far. The available lexicons list the following : sie-, 372-, 3713997-, re-,575-, 48 - 07-, T69-, -, 127, ge-and ET- (but Madhu Vidya/80

Page #106
A STUDY IN HISTORICAL LEXICOGRAPHY they do not give separate entries for and ). To these, the following have now been added atra, aratni, aSThIvat Asya, upapakSa, kakSa, karNa, afgu-, khajugduma- girikUTa, grIva, zrIvA, cubuka, khutruka, jaghana, nAbhI, nIvi, zIrSa, and f. There could be a few more. The derivatives with occur chiefly in the Vedic literature and there too mostly in the ritual context. Completeness in regard to giving references to their occurrences in the Vedic texts has been sought on the basis of the VVRI Vedic Index. Derivatives with occur less frequently in the post-Vedic literature, but where they do, they do so in non-ritual cantext (except in tier works). 9. sie amsadaghna (i) shoulder-deep (under the ground) topalagAnudrapAmi yanme bhrAtRvyo yaunabhrAtRvyo nicakhAna meM asadapne' 'here I dig up those charms which my enemy, which my non-enemy buried (under the ground), which (as a result lie) shoulder- deep' Mai. Sam. 1.2. 10. (ii) shoulder-high (above the ground) (a) areasdava '(Prajapati) raised (it) shoulder-high' Mai. Sam.6.6; (pia). udgRhvIyAt... aMsadapnam Mana. Sr. 154 12: asadane ( harati ) Vara, Sr. 1. 4. 3.1 sta (fem.) star wafa (the seat) is to be shoulder-high' Sat. Br. 14. 1. 3. 10. aMsadaSnI (fem.) pUrvArdhe chadirante sadagnIzvatasaH sthUNa nihatya... on the eastern side having fixed four shoulder-high posts at the ends of the cover (of the cartshed) Mana. Sr. 2. 2. 2. 28 (The printed text gives side: which is a misprint ). 243 C 10. atradaghna atradaghna anadadhnA (fem.) 'this high ' tasmAdudanA bhavati 'therefore (the seat) is this-high (i. e, navel-high)' Sat. Br. 3. 3. 4. 28. The actual height perhaps was indicated by pointing to the navel (Comm. ' fa fa:). In the Taitti. Sam. 5. 4. 3. 3 the word used to indicate height is iyat ityame juhoti / atho'ya'tyatheya'ti (also Taitti. Ara. 5.9 1. where the Comm. says iyatyAsyadadhne deze ( dhArayitvA ) ... iyati nAbhidadhne deze dhArayitvA iyati jAnudaghne deze dhAravitvA juhoti / ) 11. aratnidaghna aratnidaghna' elbow-high ' zarAvasthamudakaM bhUmerupari nAbhidaghne dhArayitvA tasyopariSTAdaradine svamukhaM kurvannudakasyAdharastAdaratnamAtre mukhaprativimbaM pazyati Sastradipika 1. 1. 6 (139. 8). azma-, 3-, and 12, avadaghna' as high as the horse' at GC: 413TALE: (refers to the piling of the fire altar) Rama. 1. 23. 28 (Gorr) Baroda edn. 1. 13. 23 where the line is differently read: garuDo rukmapakSo vai triguNo'STAdazAtmakaH some mss. read Asyadaghna or azmadaghna. It is likely that was the original one which was corrupted to Even in place of azvadaghna for these three readings -, in that order. Madhu Vidya/81

Page #107
244 M. A. MEHENDALE 13. aSThIvaddana asthivaddaghnd knee-high' aSThIvaddaSvaM zuddhasya (zmazAnaM 5215) (the burial mound) of a Sudra ( should be made) knee-high 'Sat. Br. 13. 8.3, 11. See ATCE, 14. 37165 adaghna 'mouth-deep (water)' NIGER 3429718 a 19 Farat na 'some were seen (like lakes) having mouth-deep (water), having (water) close to armpits, while some others (were seen ) like lakes fit to swim (i. e. with water deep enough for swimming )'RV 10. 71. 7. This is the earliest formation with dana, Adaghna was explained first by yAska as Asyadaghna. Roth ( 1852 ) did not agree. He took it to be from av dagh 'to flow on' and translated it as' anrollend.' But Yaska's interpretation has been upheld since Brugmann (IF 15, 104, 1903-04) who explained it as d-daghna < *uz-daghna. Adaghna was later replaced by Asthadana and mukhadaghna. 15, 31715 asyadaghna (i) 'mouth-deep (water)' 3TIFICER # fag... TTIECT'standing in mouth-deep water... he shall sing' Sama. Bra, 1. 4. 22; 3. 2. 9; 371 REET 3797 Nir. 1. 9 (paraphrase of 3716ER:...). (ii)' mouth-high 'trae (Col) 44 la Maitra. 3. 6. 8; 371RCE:(QUE :) #79: Katha. 23. 4, Katha 36.1; 1, Apa. Sr. 7. 8. 3, 10.10, 5; 37TRICE ( TATT:) Jauh ( shefa) (The Adhvaryu while carrying the fire from the Garhapatya to the Ahavaniya holds it ) mouth-liigh for the third ( of the way ).' Apa. Sr. 5. 14. 8, Hira. Sr. 3. 4; 371697 ARTHOT: ) Eat Apa. Sr. 17. 11. 4, 5; 15. 13. 3; Bhara. Sr 11. 13. 5; 5. 8. 2; Vara. Sr. 2.2.3, 5, 7 (37TRICETTE), 391021 fadece 941... Safa Mana. Sr. 4. 4. 8; 778 agit 'he should not lift ( the fire) beyond mouth-high (level)' Bhara. Sr. 5. 7. 13; 377RCE 197 faraja yangi 127217:' he should pile (the altar) with three thousand (bricks) mouth-high when piling a third time 'Apa. Sr. 16. 13. 11; 311FREA (fem.) 311feff Fafa) free datea Apa. Sulba. 10. 10. 16. 3987 upakaksadaghna (water) deep upto the region near the armpit' yathopakakSadanaM vA kaNThaghnaM vA yato vizramya prasnAyeyustAk tat ' just as after resting at a piace (where water is ) deep up to the region near the armpit or deep up to the neck, whence they would swim, it is like that'Go. Br. 1. 5. 2;Niru. 1.9 ( used only to explain the Samhita word 3979TA:) Durga explains 397497 as kakSasaMnikRSTadezatulyodakAH and Skanda-Manesvara as kakSasamIpaparimANA! . See upapakSadaghna, F1877. 17. 399976 upapaksadaghna '(water) deep upto the region near the armpit'yuthopapakSadanaM dhA kaNThadaghnaM vA yato vizramya prasnAMti 'just as after resting at a place (where water is ) deep upto the region nea_rarmpit or neck - deep, whence they swim' Sat. Br. 12.2. 1. 2, 4 (Comm, 97 g 39921 TATART 39484674) See 398 FHGE. Madhu Vidya/82

Page #108
A STUDY IN HISTORICAL LEXICOGRAPHY 245 18. upasthadaghna upasthadaghna genital high' upasthadaghnaM striyAH ( zmazAnaM kuryAt ) "(the burial mound) of a woman (should be made) genital-high' Sat. Br. 13.8.3.11. See jananadanna, sphigudaghna. 19, rudaghna (i) thigh-deep (water) CERCA: WCCIAN... ...ed: the first svarasaman day (i. e. one of the three days preceding the fa) is (like) thigh- deep (water, as compared to the second and the third svarasaman days which are said to be like knee-deep and ankle-deep water)...... (with samans in the reversed order) the third (svarasaman day) is (like) thigh -deep (water, while the second and the first are knee-deep and ankle-deep)' Sat Br. 12.2 1.3. In the later literature is mostly used with reference to the depth of water. UrudannaM jalam Harivar C 8324 bAri vigAhamAnaM...... rAmA janamUrudaghnam Janakiha 3.34; Urudaghnena payaseottIrya siprAM Kadam. 249.14; depth of the thick dust as fafaga Ravanari. 16.51; UrudaghnI (fem.) kRtvo rudaghnIM ripurakta kulyAM Ravanarj. 14.6. (ii) 'thigh-high'si kazza ( znani gaia) Sat. Br. 13.8.3.11. Hiranya. Pity, 29.5. 20. kakSadana haksadaghna armpit deep water ) ' ( antyaM vA sAma) kakSadane (uda tiSThan prayuJjAnaH ) pazUn ( juSate ) ( or one who employs the last saman i. e. the last verse of the E = 1. 1. 1. 8. while standing) in armpit-deep (water) obtains cattle' Samavidhana Br. 3. 2. 9. See 34, SITE 4 21. kaNThadana hanthadaghna ( i ) neck-deep (water) ya'thopapacadaM vA kaNThadaSnaM vA yato fifa as after resting in a place (where water is) armpit-deep or neckdeep, whence they swim (in deeper water) Sat. Br. 12.2.1.2,4; Gopatha Br. 1.5.2. puthopapada yA kaNThadaM vA yato vibhramyotkrAmanti ( utkrAmanti = they gradually come up in water) Sat. Br. 12.2.1.3; yathopapacadaM yato vibhramyotsnAMti ( utsmAnti they finish swimming, they come out of water) Sat. Br. 12.2.1 5. In later classical literature kaNThadapna is recorded with reference to water depth asasrotasvinInAmajani kumadhAkuNThakaM kaNThadaSnam Haravi; 45.55; kaNTadaSne payakhi Nalabhyu 6.59a; kaNThadaghnamavagAhate Unmattara 69. ( i ); vAriNi kaNThadane Rasaganga. 387.9. vyudastAmAyA mauvAmiha kaNThadanIm (ii) kaNThadanI (fem.) 'neck-high' bemboryugenobhayato Rasasadana 201. See grIvadaghna and grovAdana. 3 22. karnadaghna ear-high ' (a) tha' (Prajapati) raised it (fire) ear-high' Maitra 1.6.6. Apa. Sr. 5.14.8; Hira. Sr. 3.4.18; Vara. Sr. 1.4.3.13,14; fciaga...na Mana. Sr. 1.5.4.13. 23. en hulphadaghna (i) a. ankle-deep (water) 4: (FRA) the third (svarasaman i. e. one of the three days preceding the Madhu Vidya/83

Page #109
246 M, A. MEHENDALE fauca) is (like a place with ) ankle-deep (water) Sat. Br. 12.2.1.3; 2467 gaisaia 'the first FATETHI in reverse order is also (like a place with ) ankle-deep ( water ).' Sat. Br. 12.2.1.3; Gopatha Br. 1.5.2, (i)b. ankle-deep ( ground )' en ala ata una opera Aaala 67''here I dig up those spells which my equal, ( and ) which my unequal buried, ( and ) which ( lie ) ankle-deep ( under ground) 'Maitra. 1.2.10. (ii) ankle-high' 9.29.Cele (Prajaptai) raised it ankle-nigh' Maitra. 1.6.6; sfi n tla 24icely TT TETEHETEara Manava Sr. 1.5.4.12; kulphadadhne harati Varaha Sr. 1.4.3.13. See gulphadana. 24. khajUramadhna kharjiradrumadaghna ' as high as date tree' khajUradrama TE77159, Malatima. 5.14. 25. maza girikutadaghna' summit-high' fafticat: 93: ( demons) Bhatti. 2.30. 26. Tefe gulphadaghna (i)' ankle-deep' 9719 ( 2791 ) ffyrat eget (having dug the place for the ae knee-deep) having filled it (with water?) ankle-deep' Taitt. Ar 1.22.8-9; from later literature we have yetet e Harivam. 8324;#ajzafa UE4767957: Malatima. 3.17; Th... 99: Ragh. Panda. 15.47. (ii) ankle high' a 'he should cut (the tree for yy at a place) ankle-high' Katha. 26.3; Katha 41. 1; Apa. Sr. 7. 2. 6; Bhara. Sr. 7. 1. 15; Hira. Sr. 4. 1. 22. a T E T TE al palaa 'he anoints (the tree) all round with an (at a place which is ) ankle-high or knee-high' Hira. Sr. 4. 1. 20 (Comm. 76: Tafe: 37 : saia la HTAT Fritsarate of safet); sa ga SITR Torce 97'he, turning to the west just from there, holding (the leaf of 3 ) ankle-high' Baudha. Sr. 10. 48; Vaikha Sr. 19. 6; af HTC TEC Tea T817 of Teca greata ed za: 921 "If (the altar is ) knee-high in the front, (it should be ) ankle-high behind, if (it is ) ankle-high in the front, (it should be) level with earth behind' Bauhda. Sr. 17. 30; Baudha. Pits. 1. 15; Agnivesya Gr. 3.8. 1; in late classical literature CZERE Kathasarit. 12. 34.291. See 49.367. 27. grIvadaghna grivadaghna neck-high 'grIvaghnaM cinvIta tRtIya cindhAnaH 'one should pile (the altar) neck-high, piling it a third time. Taiti. Sam. 5. 6. 8. 3; agu feraat 1969Baudha. Sr. 22. 4.; af lagei yalaaaaa 'If (the altar is ) neck-high in front, (it should be) navel-high behind Baudha. Sr. 17. 30; Baudha. Pity 1. 15; Agnivesya Gr. 3. 8. 1; 95 'holding (the Madhu Vidya/84

Page #110
A STUDY IN HISTORICAL LEXICOGRAPHY arka leaf) neck-high Baudha Sr. 10. 48 ( imaM ) jAnudane prathamaM haredatha nAbhidapne'tha zrIcadane Baudha Sr. 20. 17; See kaNThadana, grIvAdana. 28. grIvAdana grtoldaghna neck-deep (water) grIvAdane sneyam | zrI vAdane hi te tAM nidhAya Ayan ' ( the sacrificer ) should bathe in nech-deep ( water ); for they went placing it (ie the dIkSA ) in neck-deep ( water ) ' Kathaka 22. 13. See kaNThadana, grIvadana. 247 , 29. cubukadaghna cubukadaghna 'chin - high' cubukadanaM tRtIyaM cinvAnaizcinvIta one piling (the altar) a third time should pile it chin-high Maitra. 3. 3. 2; Varaha Sr. 2. 2. 2. 1 maitrAvaruNadaNDamAspadaSnaM cubukadanaM yA Ap. Sr. 7.83 10.10.5; mukhena saMmitazdhubukadaghno vA Hira Sr. 7. 155. See chubukadaghna. 30. chubukadana chsabshadaghna chin-high' bubukadaghne tRtIya (juhoti ) ' (he offers ) a third time ( holding the spoon ) chin-high Maitra 3 3 4. See cubukadana. 31. japanadapna jaghanadaghna hip-deep (water) jamanastanadaSnaM payaH Ragha. Papda 15.47ct. sAnadAna) See upasthadapna, sphigudaghna. 32. jAnudanna Janudaghnd (1) a kneedeep (water)' jAnuino dvitIyaH ( svarasAmA ) ' the second ( svarasAman ie one the three days preceding the viSuvat ) is (like a place with) knee-deep (water) 'Sat. Br. 12. 2. 1. 3; Gopa. Br. 1. 5. 2; jAnudano dvitIyaH ( ardhAkRtAmA ) the second ( svarasAman with the sAman in the reversed order ) is ( like a place with ) knee-deep water' Sat. Br. 12. 2. 1. 3; Gopa. Br. 1. 5. 2. antyaM vA jAnudaghna udake tiSThan ( prayuJjAno lakSmIM juSate) 'or (one who employs) the last (verse of the vairUpASTaka = AraNyagAna 1. 1. 1. 8) while standing in knee-deep water (obtains wealth) Sama. Br. 3. 2. 8. In classical literature occurs exclusively with reference to water depth : jAnudaghnaM jalam Harivam C 8324; Kavyami 24.24; Ragh. Pand. 15.47; Sarasvati. 122. 11; Rajatara. 4. 539; Sariga 561; Bhojapra 40 21. ambudhi: jAnudano'yam Saduktika. 230 28.; etc. etc. jAnudaghnI (fem.) jAnudaghnI sarit Kavyal. 6.55. (i)b. 'knee-deep (ground ) ' tenemAM pRthivIM prAvizajjAnudaghna madhastasmAjjAnudaghnaM kheyam ' with that ( wealth Agni) entered this earth down below knee-deep; therefore (the ground) should be dug knee-deep' Katha. 25. 6; Katha 39. 3. devebhyo'pAkrAmatsa pRthivIM prAvizat taM jAnudaghne'nvavindaMstasmAjjAnudaghnaM kheyam Katha 26. 1; Katha. 40.4; Maitra 3. 8. 5: idamahaM tAn yAnudrapAmi yAnme khajAto' yanisajAteo nicakhAna ye' jAnudane' 'here I dig up those spells whom my kinsman whom my nonkinsman buried, which (as a result lie buried) knee-deep (under the ground), Maitra 1. 2. 10 jAnudana triktistaM vA khAtyottara veyarthAn pAMsUna harati 'having dug (the MadhuVidya/85

Page #111
248 M. A. MEHENDALE ground for cAtvAla) knee-deep or three vitasti he carries the earth for uttaravedi' *Apa. Sr. 7. 4, 2; Hiranya. Sr. 4. 1. 10GHT (fem.) mageuraai ar Taitt. Ar. 1. 22. 8; 1. 25. 1; Baudh. Sr. 19. 20. (ii) 'knee-high' FEE a he should cut ( the tree at a place ) kneehigh' Katha. 26. 3; Katha. 41. 1; Apa. Sr. 7. 2. 6; Bhara. Sr. 7. 1. 15; geot gulphaghne jAnudaghne vA paryaNakti 'he anoints (the yUpa) all around with a sruva (at a place) ankle-high or knee-high' Hira. Sr. 4. 1. 20. FICOSI gefa Katha. 21. 6; Katha. 31. 21; Maitra. 3. 3. 4; Sat. Br. 9. 1.1. 11; Vara. Sr. 2. 2. 3. 3; GERTY INTETICE: Mana. Sr. 4. 4. 8; 1 919... FAITET Ria Vai. Sr. 19. 6;Apa. Sr. 5. 17. 11. 4, 5; ( 4 ) FIECE Apa. Sr. 15. 13. 3; Bha. Sr. 11. 13.7; TCE ETYATEJ Aalst epla Apa. Sr. 5. 14. 8; HTC (FR) q4H Bau Sr. 20. 17; Jaca ( gfa Vara. Sr. 1. 4.3. 13; Hira. Sr. 3. 4; Bhara. Sr. 5. 8. 2; Mana. Sr. 4. 1. 5.4. 12; a' ct ( 3 EOTCT) Maitra 1. 6. 6; HTC fra en frana: Taitti. Sam. 5. 6.8 3; Maitra. 3. 3. 2; Kathaka 21. 4; Katha 31. 19. Apa. Sr. 16. 13. 11; Baudha. Sr. 22. 4; Vara. Sr. 2. 2. 2. 21; fa i gitarai CE TEILE FIGC greang Te GG 912 Baudha. Sr. 17.30; Baudha. Pitp 1 15; Hira. Pitp. 16: 7; Agnivesya Gr. 3, 8. 1. FICCE (fem.) WTG area farala geri qala: Ap. Su. 10. 10. See 37 Stare 33. Et nabhidaghna (i)a, navel-deep (water ) IT (39$ faget YFIU HH Tyla:) 977 (998) if he employs the last saman (of the Vairupastaka = 37701a11.1.8 while standing in ) navel-deep ( water, he obtains) corn' Sama. Br. 3.2.9. In later literature is used almost exclusively with reference to depth of water. jale nAbhidadhne Samaradityasam. 4.219; nAbhidadhanamapi HITAHTH: Yatrapra. VI. 21: CTCFFIFI Yajna, Smf. 2.108; etc. (i)b. ' navel-deep ( under the ground )' ata Fanlah t er Hiduaruffaela Ace 'There I dig up those spells ) ...... which (lie) navel -deep ( under the ground ) Maitra. 1.2.10. (ii) navel-high' aige TRACE galeut cela delante Shui he winds (the girdle) round (the q at a place which is) navel-high. He (thereby) puts strength in him ( at a place which is ) navel-high. Therefore people enjoy strengths at navel-high regions' Taitti Sam. 6.3.4.5; Maitra. 3.9.4, Apa. Sr. 7.11.5; 9% ya T4 , H Tula arhet fad now when the Brahman priest sings the HR on the chariot wheel raised navel-high' Sat. Br. 5.1.5.1,2; ' fadi's (alla) Maitra. 9.1. 1.12; 3.3.4; Katha, 2.1.6 Katha 31.21: Sat. Br. 14.3.1.5; Apa. Sr 17. 11. 4, 5; Vara. Sr. 2.2.3.4; Vai. Sr. 19.6; Mana Sr 4.4.8: viagetai . JIE ar la Apa. Sr. 15.13.3; 5.14.8; Bau. Madhu Vidya/86

Page #112
A STUDY IN HISTORICAL LEXICOGRAPHY 249 Sr. 10.4.8; Bhara. Sr. 11.13.6; 374 FIHCET (TE) Eda Bau Sr. 20.17; Bha. Sr. 5. 8. 2; Vara sr. 1. 4. 3. 13; Hira. Sr. 3.4; a' ATTEET ( LETIT Mai. S. 1. 6. 6; Mana. Sr. 1. 5. 4. 12, a i fia faai Faral: Taitti. Sam. 5.6.8.3; Maitra. 3.3.2; Katha. 21.4; Apa. Sr. 16.13.11; Vara. Sr. 2.2.2.21; Datei graf 985 TEE TESTECE TT Bau. Sr. 17.30; Bau. Pitr. 1.15; Agni. Gr. 3,8.1; 311 276) THE 9191 Apa Sr. 10.29.7; 19. 9.10; Bau. Sr. 6.10; HEIG: (comm, SHIN AZ: qara HaiaKaty. Sr. 8.6.1. In later literature we find # FF THE R I... Sastradipik, 1.1.6. ( 139.8 ). TOTE (fem.) HTET (HT) Hala Sat. Br. 3.3.4.28; Katy. Sr. 7.9.27; Piret RTCG1 24H fH Hiranya. Sr. 7.7. 13. In later literature we get nAbhidadhnAM tu tAM bhitti chAdayed viTapavrajaiH Manasollasa 4. 15. 1654. FTATEA (fem.) IRAGET: quia : Apa, Sr. 11. 10. 6; Mayukhamalika on Sastradipika 334.33 (3.8.5 ) Tight FG Tech Faiza Ap. Sulba. 10.10. In later literature we find bAlAnAM krIDArthe nirmitayA nAbhidadhnyA puSkariNyA samA (mImAMsA) bhaviSyati Nyayamalavistara p. 4.20. See FTCG, FRCE. 34.nAbhIdaghna nabhidaghna navel-deep' nAbhIdanapurANapaGkabahalAbhogeva vizvambharA Suktiratnahara 191. 5 (perhaps metri causa for ATGT). See affaqet. 35. pige nividaghna (i) navel-deep (water) ' s a 9A; FITATAT... facetega: (staPE FETHTAI) 'the first svarasaman (i. e. one of of the three days preceding Peeretatis ( like ) navel-deep (water )...... the third (svarasaman with reverse samans ) is (like) navel-deep (water)'. Gopa. Br. 1. 5. 2. (ii) navel-high' corresponding to go of Taitti. Sam 6. 3. 4. 5 ( see sub voce ), the Kanva version has ac. cf. Eggeling SBE 26. 172 f. n. 1. See FT. 36. CE mukhadaghna' mouth-high' yace 2 HOT gefa Sat. Br. 14. 3. 1. 2,3; 9.1. 1. 13; qfare ... H ... C ETTE Bau. Sr. 9. 13; also cf. Bau. Sr. 1o. 14; Vaikha. Sr. 19. 6; geci Huet ( ha pla) Sat. Br. 13. 8.3. 11; F A THETETT 5 Halia Heel they fix the stable ( for the horses ) mouth-high with the Asyattha treos whose roots have grown upwards' Vadhula 3. 76. See 3771, 37FRICE. 37. zIrSadana sirsadaghnd ' head-deep' idameha tAnvalagAnudvapAmi yAnme sejanyo Haar nada 'infaa''here I dig up those spells which my clansman or which my non-clansman buried, (and) which (as a result ) are head-deep (below the ground )' Maj. S. 1. 2. 10. 38. FH16 stanadaghna' breast-deep 'Only in later literature. Fate 527 Hariva. C 8324; TESTTECH674 BU HAMTHATT Since Madhu Vidya/87

Page #113
250 M. A. MEHENDALE water that was only ankle-, knee-, hip-, ( and ) breast-deep has come in contact with women, (but ) not deep water.' Ragha. Pagda. , 15.47 39. ETICE Sphigdaghna' hip-high ' TATITET HTEL..5117CE19EGER EFRIGE GEITA HEY' the measure of the burial mound... (may be) knee-high, thigh-high or hip-high. This is the maximum measure.' Hira. Pi. 29.5. See STEYGE, GETTIGE. COLOPHON I am indebted to the Scriptorium of the Sanskrit Dictionary Department, Poona for information on the later occurrences of the derivatives ending in -daghna. [ Received 1 September 1972 ) Madhu Vidya/88

Page #114
ADHVAN ADHVARA M. A. MEHENDALE Deccan College, Poona ABSTRACT: History of these two words is examined and findings supported from various texts. The word adhvan has been attested since the Rgveda. According to the lexicons it means 'a road, way, orbit'. Its cognate advan, adwan is available in the Avesta. The age of adhvan can therefore be pushed back at least to the Indo-Iranian period. The etymology of the word is in doubt. In the Upadi 4.117(565), it is derived from ad-'to eat' (ader dha ca). In this derivation we have to assume the replacement of d by dh. The relationship between the meanings of ad- and adhvan is explained by the commentators as follows: mamsasonitaviryayamsy attity adhva (vrtti of Svetavanavasin on 4.124), or atti balam ity adhva (Dasapadi vrtti 6.71). The other Upadis suggest the derivation from at-'to be on more constantly' (satatyagamana). In this derivation we have to assume the replacement of t by dh. In the commentary of Devarajayajvan on the Nighaptu 1.3.12 adhvan is derived from the hypothetical root adh-'to go'. The same suggestion is made by Grassmann in his Worterbuch zum Rgveda (s. v. adhvara"). Johansson connects it with Pali *andhati he goes' which verb he presumes on the basis of Pali andhitva (Jataka III. 505) which corresponds to parigantva (III. 506). But scholars are not agreed regarding the existence of Pali andhati. An examination of the passages in the Rgveda (and the Atharvaveda) where the word adhvan occurs indicates that it is used with reference to a watery course, river streams in which the sun moves for his daily course from morning till night. The examples are: (1) ud vam prksaso madhumanto asthur a suryo arubac chukram arpab/ yasma aditya adhvano radanti mitro aryama vafupo sajosah //7.60.4. Madhu Vidya/89

Page #115
206 "Your nourishments, full of honey, have stood up. Sarya has mounted the shining floods (he) for whom the Adityas dig the watery ways delightfully-Mitra, Aryama, (and) Varuna". (2) jagama suro adhvano vimadhyam 10.179.2 (=AV 7.75.2). "The sun has gone to the middle of the watery way". (3) mano na yo, dvanah sadya eti 1.71.9 "who, like the mind, goes instantly (to the end) of the watery way". (4) adhvasya vitato mahan purvas caparas ca yab/AV 13.2.14 his (sun's) watery course is extended long which is castern and westren". (also cf. AV 13.1.36 and 43). The word adhvan is also used with reference to the common course taken by the dawn and the night. This course also lay presumably through. the heavenly waters. samano adhva svasror anantas tam anyanya carato devasiste 1.113.3. "The watery course of the two sisters is the same, endless one, which the one, after the another, as ordained by the gods"." As in the Rgveda, advan (adwan) is used with reference to the path. of the sun and the stars also in the Avesta. (1) kasna ang str@mca dat advan@m Yama 44.3. "Who has made the way for the sun and the stars ?" (2) te...fravazenti durac-urvaes@m adwano urvaesem nasamna yasta 13.58. "Now they (i. e. the stars, the moon, and the sun Yasta 13.57) move forward to the distant turning point of the way, in order to reach the turning point." (3) aat yimo frasusat raoca a upa rapiowam hu paiti adwanom Videvdat 2.10. "Then yama went forward to the light at midday towards the path of the sun". Besides the above passages, adhvan is also used with reference to the heavenly and the earthly waters or streams; it is used also with reference to a vast expanse of water, whether in the heaven or on the earth. (1) ambayo yanty adhvabhir jamayo adhvariyatam/prncantir madhuna payah//1.23.16 (AV 1.4.1). "The mothers move along the watery ways, the sisters of those who wish to go in a vehicle (i. e. the sun and the stars who move in heavenly waters), mixing (their) milk with honey." (2) samano adhva pravatam anusyade 2.13.2. "The streams have the same watery way to go over".7 (3) kim icchanti sarama predam anad dure hyadhva jagurih paracaih 10.108.1 "Wishing for what has Sarama accomplished this indeed the watery way is distant disappearing far away". Madhu Vidya/90

Page #116
way. 207 (4) yad hathur asvina bhujyum astam anarambhane adhvani taugryam astam khila 1.2.1. "When you, oh Asvina, carried Bhujyu" home, the son of Turga who was thrown in the watery expanse having no hold". When we compare this verse with RV 1.116.5 where the same legend is referred to with the words anarambhane tad avirayetham anasthane agrabhane samudre it becomes clear that anarambhane adhvam of the Khila passage corresponds to anarambhane adhvam of the Rgveda passage and that adhvan therefore means a vast watery expanse. 10 In the following verse adhvan is only probably related to a water hamsiv iva patat ho adhvagav iva somam sutam mahisevava gacchathab/ sajosasa usasa suryena ca trir vartir yatam asvina/ 8.35.8. "Like two swans you two fly, like two persons going on adhvan, like two buffaloes you go down to the pressed Soma; sharing (your) delight with Usas and Surya, oh Asvina, you two go on your three fold journey". The expression trir vartir cannot refer to the three pressings, since Aavina are primarily connected with the morning pressing. It therefore seems to refer to the three different kinds of journeys made by them and this is also indicated by the three upamanas hamsau, adhvagau, and mahisa of the first line. Since the first and the third upamana indicate journey through air and on the land, the second upamana may indicate journey on the water. In that case adhvan will refer to watery streams. The above examination shows that adhvan is used with reference to the way which is full of water and this may have been its primary meaning. It is not to be denied that adhvan is used also with reference to the imaginary paths in the mid-region (antariksa) by which the gods come and go (cf. 1.72. 1, 6.10.4, 7.42.2, 1.104.2, 6.50.5, 10.185.2, AV 4.11.2). In fact, it is also occasionally used with reference to land roads (cf. 4.16.2 (AV 20.77.2), 1.42.8, 10.51.6. 10.117.7). But these may be its secondary uses. The word having the primary meaning the road on the land' is pathin. This is made clear by such adjectives used with pathin as arenu (1.35.11, 1.163.6), adhvasma (2.34.5, 9.91.3), a rksara (1.41.4, 10.85.3), avrka (6.4.8), anarvan (8.18.2) which, with the negative prefix, deny such qualities which otherwise are found with pathin.11 The word paripanthin obstructing the way' also shows that pathin primarily means a land road.12 In the passages where adhvan and pathin both occur, it is better to interpret them as waterways and landways.13 If the interpretation of adhvan suggested in this paper is correct, it is possible to explain its derivation as air secondary formation adh-van 'watery, full of water', and adjective turned into substantive. Such formations with the suffix van are known to occur since the Indo-Iranian period (Alt. Gr. 11. Madhu Vidya/91

Page #117
208 2, Sec. 718 ff, p. 900 ff.). The word adh-van will now appear similar to udan-vant 'watery ocean', The above analysis of adh-van assumes a nominal stem *adh 'water'. This stem now will be able to explain satisfactorily the forms ad-bhih (Instrpl.) and ad-bhyah (dat. abl. pl.). So long they were related to ap 'water' and in order to explain them it was necessary to assume either the replacement of p by t (Panini 7.4.48), or dissimilation ab-bhib> ad-bhib (Luders, Varupa I. 83), or hypersanskritisation (Mayrhofer KBW I. 29-30, s. v. adbhih),14 But these attempts become unnecessary once a vocable *adh 'water' is presumed to exist in adh-van. The word adhvara is attested since the Rgveda. It is assigned the meaning 'sacrifice' and this can be exemplified in numerous passages in the Rgveda. The traditional etymology of adhvara is from the root dhvar- with ne negative prefix a- i. e. 'one that does not injure' (dhvaratir himsakarma tatpratisedhash.. Nirukta 1.8).15 This has been disputed by modern scholars. They connect it with adhvan on the assumption of a -r/-n alternating stem and justify it on the basis of the old Indian belief "that there was a way to lead (adhvan-) the sacrificer to heaven" and "the well-known conviction that the sacrifice (adhvara-, which must be of adjectival origin) was the vehicle to convey him and his merits to that destination."16 The view that adhvara originally means 'related to the path, a vehicle' finds full support in the Rgvedic usage of the term. There are clear cases where the word adhvara seems to imply the 'original' meaning 'vehicle'. This can be easily demonstrated in passages where a word meaning 'a charioteer' occurs with adhvara. (1) justo hi duto asi havyavahano 'gne rathif adhvaranam 1.44.2 "for, you are the loved messenger the carrier of oblations, oh Agni, the charioteer of the vehicles."""" (2) vaisvanaram rathyam adhvaranam yajnasya ketum janayanta devah 6.7.2 "Gods have created Vaisvanara, the charioteer of the vehicles, the banner of the sacrifice." (3) sadyo adhvaree rathirath jananta 7.7.4. "They instantly created the charioteer in the vehicle." The meaning adhvara 'vehicle' can also be seen in passages where companion with a horse is involved. The companion is directly expressed or implied.1. (1) so adhvaraya pari piyate kavir atyo na vajasataye canohitab 3.2.7 "He (i. e. Agni) is carried round the vehicle, the wise one, like a horse, for the winning of the prize, the loved one." (2) vaji...... yam devasah havyavaham adadhur adhvaresu 3.29.7. "the prizewinner...... whom the godss (as) the carrier of oblations put to the vehicles "19. Madhu Vidya/92

Page #118
209 The meaning 'vehicle' is certain for adhvara also when it occurs with the verb taks- in the following passage : Pusanvanta sbhavo madayadhvam urdhvagrivano adhvarasm atata 3.54.12 "oh Rohus, being together with Pusan be delighted; with upraised stones they have fashioned the vehicle". 20 in the passages where yajna and adhvara occur together a difference in meaning is indeed implied. Geldner usually tries to bring this out by rendering yajna as opter (or Anbetung) and adhvara as Gottesdienst (or Opfer) etc. 21 Sayana does this by leaving yajna as it is (or equating it with yaga) and explaining adhvara as himsarahita etc. It is better to show this difference by rendering adhvara in its original sense 'vehicle'. (1) agne yar yajnam adhvaram visvatah paribhur asi/sa id devesu gacchati//1.1.4. "Agni, the sacrificer which (as) a vehicle you protect all around, that indeed goes to gods23." ......agnir yajnasyadhvarasya cetati kratva yajnasya cetati 1.128.4 "Agni knows of the sacrifice.of the vehicle; (he) knows of the sacri fice with his insight."24 (3) tvam yajnesv ilate' gne prayaty adhvare 10.21.6 "oh Agni, they worship you in sacrifices when the vehicle has progressed,25" (4) sajosaso yajoam avantu deva urd hvam krpvant adhvarasya ketum 3.8.8 "May the gods, having common delight, protect the sacrifice, may they raise the banner of the vehicle."26 (5) ano niyudohih satimibhir adhvaram sahasranibhir upa yam yajnam 7.92.5 "come to our vehicle with hundreds of mares, to (our) sacri fice with thousands (of mares)". Finally we may compare the compound ratbestha with adhvarestha (10.77.7) to exemplify the meaning vehicle' of adhvara. It is not enough to translate the latter simply as bei der Opferfeir beschaftigt' (Grassmann) or bei der heiligen Handlung zugegen' (Geldner). The meaning 'standing in the vehicle (in the form of the sacrifice)' is very much there. If, as suggested above, adhvan primarily means a water course. a river stream, an expanse of water then it is worthwhile to investigate if traces of the primary meaning of adhvara 'a conveyance to go on watery course, a boat' can be found in the Rgveda. It seems possible to demonstrate this meaning in the following passages : (1) ta no hinvantv adhvaram 1.23.17 "May thosc (heavenly waters) urge on the vehicle." The vehicle to be urged forward by the waters can only be a boat. 27 (2) svidhma yad vanadhitir apasyat suro adhvare pari rodhana goh 1.121.7 "when the axe for cutting the woods, which has accummulated good sticks, is active, the sun (is) in (his) vehicle beyond the obstructions." Since Madhu Vidya/93

Page #119
the sun moves in water, his vehicle must be one to be used on water. The verse apparently refers to the sun's coming out of the stony receptacle in which he dwells at night. 28 (3) a yo mardhanath pitror arabdha ny adhvare dadhire suro arab 10.8.3 "(Agni) who touched the head of the parents, they placed the floods of the sun (light) (viz. the Agni who had gone to the heaven) in the vehicle." Agni when gone to heaven is identified with the sun moving in his vehicle on the waters. (4) urjo napatam adhvare didivamsam upa dyavi 3.27.12 "(Agni) the son of strength, shining in the vehicle, near the heaven". Here also Agni, when near the heaven, is identified with the sun shining in his vehicle. This Agni who is said to be in his adhvara is called aptur "crossing the water" in the preceding verse. 210 In the Nividadhyaya (Khila 5.5.1) occurs the expression rathir adhvaranam. The Ait. Br. 10.2 while commenting on this refers this passage to the sun. If the sun is called the charioteer of the vehicles then, considering the fact that his daily course lics through the water, adhvara must refer to the vehicles on the water. 30 Corresponding to rathiyati from ratha we have adhvariyati from adhvara. It is true that adhvariyati in most cases means 'act as an adhvaryu, perform the sacrificial duties. But in 1.23.16 cited above, it clearly means those going in vehicles, and in the context of water, in boats. Similarly the form adhvaryati, also derived from adhvara, means 'going in a boat in 1.181, lb adhvaryanta yad unninitho apam which is addressed to Asvins. As pointed out by Geldner the line alludes to the legend of Rebha who, as mentioned in 1.116.24, was rescued from the waters by the Asvina. Hence 1.181. lb is to be understood as "when (you i. e. the Asvina) took (him) out of the waters while going in your bout". The primary meaning of adhvara then may be said to be a conveyance to go on water, a boat and then it also means 'a vehicle on land (or air if referring to gods). Secondarily it means sacrifice. Notes 1. 2. 3. Katantra Unadi 6.59: Sarasvatikanthabharana 2.1.283; Prakriyasarvasva 4.126. IF 3. 201 f (1894), 8. 180 ff (1898) Pisani IF 58.254 f. (1942), Mayrhofer KEW 1.32; in Walde-Pokorny' 1. 130 and in Alt. Gr. II. 2. 717b uncertainty is indicated regarding this relationship. Pokorny, Indogerm. Etym. Worterbuch 40-41, records it. under andh, anedh (zero grade ndh) 'hervorstechen, spriessen, bluhen' although there is a difference in the meaning. Madhu Vidya/94

Page #120
211 12. 4. For the sun's daily course through the waters, sce Luders, Varuna 1. 294 ff. Also cf. 1.146.3 (anapavrjya adhvano mimane "measuring out the (watery) courses which cannot be ended." 6. The mothers are the heavenly waters (apo devih) mentioned in 1.23.18. Also cf. 4.58.7 and 9.52.2. Sarama is said to have flown to the ends of the heaven (paridivo antan subhage patanti 10.108.5) Translated against the accent. astam 'thrown' should be accented on the last syllable. But then there does not seem any nccessity for repeating astam home'. 10. Asvina are said to have rescued Bhu from this watery expanse (adhvan or samudra or udamegha (1.116.3) with thc help of a boat or boats floating in the mid region (naubhih ....antariksaprudbhih) 11. In contrast to this, adjectives with adhvan are not commonly used, and where they are, they are colourless ones like uru (8.3.11), dirgha (1.173.11), ananta (1.113.3) etc. The only significant adjective used with adhvan is anarambhana (Khila 1.2.1). Pathin is secondarily used also of river streams on earth (patha var iva dhavaiu 10.145.6) and in heaven (radat patho varunah suryaya 7.87.1). The word devayana occurs more frequently with pathin. In the context of sta, pathin (and pathya) alone : occurs. Cf. 6.16.3, 6.46.13, 10.22.4. Sayana tries to show the distinction either by calling them mahamarga and ksudramarga or as ways on the earth and in the mid-region or as ordinary ways and those on the battlefield. Geldner renders them as die strasse and der weg. Old Persian has abis and Avesta has aiwyo. These seem to be direct descendants of ab-bhih and ab-bhyah. In Sanskrit, these forms were at some stage substituted by ad-bhih and ad-bhyah. This derivation has often been availed of by Sayana to explain adhvara when it occurs by the side of yajsia in the same verse. See below note 23. 16. J. Gonda, Old Indian, p. 168, Leiden. Koln, 1971. Also his earlier arti cle adhvara---and adhvaryu- in the Vishve. Indol. Jour. 3.163-177 (1965). One wonders whether the title of the paper was really intended to be adhvaw and adhvara. All earlier references can be found in Gonda as also in Mayrhofer KEW I. 32. 17. Also cf. 8.11.2 and 4.15.2, and Gonda VIJ 3.173. Also cf. 1.27.1. Also cf. 3.27.8, 4.15.1. According to Gonda, VIJ 3.p.170 the meaning *vehicle' can also be seen in expressions like prasthite adhvare 8.35.23, payaty adhvare 8.13.30, 71.12; 10.21.6, and also in passages like AV 5.12.2 and 18.2.32 (p. 171) where adhvara denotes something which can Madhu Vidya/95 18,

Page #121
212 be removed or transferred'. Further, he sees the same meaning for adhvara in adhvaresya pesah 7.42.1 (p. 172). 20. See Gonda, op. cit., p. 173 and his footnote 52. 21. For adjectival use of adhvara in such cases see. Gonda Vishve. Ind. Journ. III. 174 ff. 22. Cf. above the explanation on adhvara given in the Nirukta. In 4.9.7 and 10.21.6 where, however, adhvara is rendered by him as yaga, he explains yajna as yajnasadham (or yajaniyam) havih. 23. Also cf. 4.9.7. 24. Also cf. 8.10.4. 25. Also cf. 6.10.1 where prayati occurs with yajne. 26, Also cf. 10.1.5. The banner of the sacrifice is Agni (3.11.3) Geldner understands under it the sacrificial post. 27. Gonda, VIJ 3.171 feels that the meaning 'vehicle' of adhvara can also be seen in the other passages where the verb hi-'to set in motion, to urge on' is used with reference to it as in 7.56.12, 10.30.11. 28. The treatment of this verse is not found in Luders, Varuna. 29. The translation follows Luders, Varuna II. 594-595, for the first half. His translation runs as : "Der das Haupt der Eltern erfasst hat, den haben sie beim Opfer niedergesetzt, cine Glut von Sonne." He further remarks: "Das "Haupt" der Eltern, d. h. den Himmel, erfasst Agni, Wern er in den Himmel eingeht." 30. I agree with Gonda (VIJ 3.174 f.n. 58) in not accepting Keith's trans lation of the Ait. Br. passage. 31. E. g. in 2.1.2, 32. If, however, adhvaryati is derived directly from adhvar- (alternating stem of adhvar-) then it would mean 'go on a watery course, on a wide : expanse of water'. 'Is ratharyati formed on the analogy of adhvaryati ? sArAMza prastuta lekha meM 'adhyan' aura 'adhvara' ina do zabdoM kI aitihAsikatA kA vibhinna graMthoM ke AdhAra para vivecana kiyA gayA hai| Rgveda, atharvaveda aura avestA meM 'adhvan' kA makhya artha srota, nadI, jharanA' hai| 'adhvara' Rgveda meM milatA hai jisakA artha hai 'tyaag'| yaha nighaNTa meM bhI milatA hai| isakI utpati* 'adhvara' se mAnI jAtI hai jo ki 'adhvana' kA eka rUpa hai| 'adhvara' kA artha 'vAhana, savArI' bhI hone kA saMketa milatA hai| ataH yaha kahA jA sakatA hai ki 'adhvara' kA mala artha 'jalIya srota ke pAra jAne yogya savArI' rahA hogaa| Madhu Vidya/96

Page #122
ASUSU BY M. A. MEHENDALE The word asush (adj.) occurs only once in the AV 10.10.23. According to the lexicons (BR and MW) it has the same meaning as as (adj.) 'not giving birth, barren'. In his Grammar SS 1147c, Whitney also regards this as a root noun with reduplication (like sasydd. cikit etc.). But he further notes a second possibility: "asus is probably to be understood as a compound, asu-sa," (For the meanings assigned by Whitney, see below his translation and note). The same view is put forth in the Alt Gr. II 2, p. 40. It is regarded as formed either with the reduplication of Suand hence, in mesning, equal to as 'not giving birth (nicht gebarend); or, it is treated as a double compound and assigned the meaning 'giving birth to one who is barren (die Unfruchtbare [a-su-] gebaren machend ?')." AV 10, 10. 23 runs as : sarve garbhad avepanta jdyamanad asusvah/ sasuva hi 14m ahur valeti brahmabhih klptah sa hy dsya bandhuh || Whitney translates the verse as-"All trembled at the embryo, while being born, of her who gives not birth (? asasu); for "the cow hath given birth", they say of her; shaped (m.) by charms. (brahman); for it is her connection." Whitney admits that much. in this verse is not clear and hence doubtful, He notes that the padapatha does not analyse the word asus, but that the accent indicates that it is a compound to be analysed as asu-s 'giving birth to one who does not herself g ve birth'. He considers the connection of klptas with vasa as strange but also says that it probably belongs to garbhas understood. 1. It may be noted that in the BR (also in the abridged edition) the word is recorded as asui (with the final vowel short). 2. Or does the German expression mean 'making the sterile one to produce?' 3. In the paippalada 16.109.3 there is some variation towards the end. brahmana klpia uta bandhur asyah. 4. Also J. Narten agrees with BR in assigning the meaning of asn t asusa barren. In the opinion of Narten the use of asasi and sasiva is meant to create a paradox. See Acta Orientalia Neerlandica p. 128, f.n. 58 (1 71). Madhu Vidya/97

Page #123
VISHVA BANDHU COM. VOL. ( VIJ A:cording to the Kaus. Sutra 65.20, AV 10 10 and 12.4 are recited when a vasa5 is sprinkled with water and given to a Brahmana for sacrifice. According to H.P. Schmidt, KZ 78. 38-39, 1963, this vasi is not a common cow but bas a characteristic by which she is predestined to be given away to a learned Brahmana and then to be sacrificed by him. According to Schmidt this characteristic (Eigenschaft) is stated in 10.10.23. He translates the verse as follows: "All trembled at the embryo as it was born of the (cow), who gives birth to a (cow) who herself does not give birth. For it is said of her: 'the vasa has given birth'; (the embryo) is formed with byoons (durch Gedichte geformt)-it is indeed her origin." According to Schmidt one obtains the following complete sense from the mantra : the va$a to be offered is sterile, her birth is accompanied by a sign (Zeichen) which makes all beings tremble and causes them to adore her (cf verse I names te jayamanayai jataya uta te namah). The embryo is formed by brahmani which, in this context, can only mean that the embryo, even before its birth, is predestined for the carriers of the brahman, i.e. the brahmanah. Perhaps, this interpretation is not likely. The verse does not make any mention of the characteristic (Eigenschaft) which predestines the newly born calf to be given away to a Brahmana. for, this may not be the meaning of brahmabhih klptah-, nor of a sign (Zeichen) which accompanies the birth of the sterile cow and makes the bcings tremble. On the other hand one may feel certain that the verse says four things quite clearly : i. all (onlookers) tremble at the sight of the garbha whilo it is born ; ii. the mother of the garbha is asasa; iii. the reason (hi) of the trembling is given by what the onlookers say about her : "the vasd has given birth'; iv. the garbha is formed (or made complete ?) by brahman. 5. According to Whitney the name vada used for the cow throughout the hymn (AV 10.10) implies non-pregnancy. Madhu Vidya/98

Page #124
XII ] MEHENDALE Asus When one takes into account these points one is inclined to refer this verse to a particular procedure which is enjoined in the sacrifice of a barren cow (vaid), called anibandhya, to Mitra and Varuna after the concluding rite (udayaniya isti) of a Soma sacrifice." When the cow has been immolated it is necessary that the priests. make sure that the cow is, indeed, barren. Her barrenness cannot be taken for granted. Therefore, when the cow is quieted the adhvaryu asks the amity to pull out the omentum and search for an embryo. If no embryo is found then nothing is required to be done, but if an embryo is found then a special procedure for its disposal has to be gone through. 3 It will be clear that the verse in the AV 10. 10. 23 refers to the sight of this garth when a supposed barren cow is sacrificed and, on examination, is found to contain an embryo. This is a surprise and an event which calls for atonement. That is the reason why the onlookers tremble at the sight of the embryo of a cow who was deemed to be barren. They tremble and exclaim "the vasd has given birth." The word asust, applied to the mother cow, is taken to mean 'giving birth to a sterile calf'. This may be justified when the calf later proves to be sterile But how can one say this right at the birth of the embryo ? From what has b:en said above regarding the sacrificial procedure, the word can be interpreted in a different way. As an adjective it can mean 'a barren and a not barren (asus ca sa si cu) and will apply to the sacrificed cow who was deemed to be barren but was, in fact, found to be not barren. Even the old meaning can be understood in the present context in the following way. It will refer to a cow who gives birth to one who, being in au embryonic stage, is not going to beget or give birth to any one. 6. diha yada na kascana rasah parydsisyata tata esa maitravarunt vasi samabhavat tasmad esa na prajayate. Sat.Br. 4.5.1.9. The same 7. The procedure is described in the Sat.Br. 4. 5. 2. 1-18. procedure is also adopted when a pregnant cow (vicitragarbha prajnatagdrbha according to the comm.) is offered to Aditi towards the end of a Rajasuya (Sat. Br. 5.5.2.8). 8. As mentioned in the Sat.Br. 4.5.2.10, the embryo may be one that can be ascertained as male, or as female, or may be indistinguishable. Madhu Vidya/99

Page #125
VISHVA BANDHU COM. VOL. [ VIJ In the procedure referred to above it is enjoined that if the cow is found to contain an embryo then the latter is to be drawn out not from the udara but by tearing as under the thighs (virujya troni). When the embryo is being pulled out the followir g mantra is to be recited ejatu dasamasyo garbho jardyuna saha 'may the embryo, ten months old, come out with the caul', On this the Sat.Br. comments: yadd vai garbhah samrddho phavaty atha dasamasyah/ tam etad apy adakamaryam santam brahmanalva yajusa dasamasyam karoti "when, indeed, an embryo becomes fully developed, then it is ten months old. This (recitation) makes him, even when not ten months old, one of ten months with the help of a brakman, a yajus." Thus with this mantra the garbha is made fully developed (samrddha) although it is really not so.10 The expression brahmabhih" klptah in the verse 10, 10.23 very probably refers to this symbolically making the embryo fully developed although really not so. Or, it may refer to the very conception of the mother cow. Since she was considered as incapable of bearing a calf, the embryo, found in her, is looked. upon as formed, not in the natural way, but by certain brahmans. 4 The end of the verse, sa hy asya bandhuh, means either that. the garbha (referred to by sah), indeed, is her relation (and not a bull, because the conception is believed to have occurred in a miraculous way). Or, it may refer to the Brahmana, called to the mind by the use of the word brahmabhih, who is the chanter of the mantras responsible for the conception of the cow or for making the embryo fully developed and worthy of sacrifice (dasamasya, yajni ya). He is, indeed, her relation. 9. Ultimately the garbha is offered to the Maruts on the cooking fire of the animal (paeuerapana). 10. There is one more mantra (Vaj. S. 8. 29), called brahman or ydjus, with the help of which the priest makes the embryo, which is unfit for sacritice, fit for sacrifice (vajniya vai garbhah tan etdi brdhmanaiva vajusa yajil yah karoti Sat.Br. 4. 5. 2. 10). 11. As noted in f.n. 3, the paippalada version has the singular brahmana MadhuVidva/100

Page #126
670 NOTES AND COMMUNICATIONS TWO VEDIC NOTES (1) kurin? Grassmann, in his Worterbuch zum Rig Veda (1872), has two separate entries kiri subst. (m) and kirin both adj. and subst. (m). To the former he assigns the meaning singer and to the latter ' praising' and 'singer'? He lists three occurrences of the forms of kirin, two as instr. sg. kirina 5.4.10 (= TS 1.4.46.1) and 5.40.8, and one as nom. pl. kirinas 5.52.12.2 But the form kirimu can as well be derived from kiri and for this form a separate stem kirin would not be necessary. But this is not true of kirinas. For its derivation, a sfem kirin would be necessary. The verse where the form in question occurs runs: chandahstabhah kubhanyava utsam a kurino nstuh. The line refers to the dancing of the Maruts near the utsa.4 The author of the Padapatha gives the Samhita word as kirinas, and following him scholars have interpreted the form as nom. pl. of karin and translated it as 'bards' or 'singers'. But it is also possible to analyse the Samhita text kirino nurtuh as kurina u nituh where karina could be instr. sg. and can come from the stem kiri. The line would then mean that the Maruts danced, and the singer of the hymn, to whom they had manifested themselves (te me... asan drsi ... 'they (i.e. the Maruts) appeared to me '5.52.12) danced with them too. In this interpretation we are going against only the Padapatha. But the analysis of kirino as kirina u also gives good meaning and does not compel us to assume a separate stem kurin which would be attested only once. In the RV 1 BR (Worterbuch) has only kiri (aubet.), but the shorter dictionary of Bohtlingk and Mayrhofer's EWA. I. 1956, have both kiri and kirin. According to Bohtlingk the former is a subst., the latter an adj., while according to Mayrhofer both are adj. Mayrhofer (p. 215) thinks that kiri small, hunable, poor' is perhaps a loan from Dravidian. But on p. 565 he refers to Bailey, TPS, 1965, 63, who compares kiri 'weak, poor' with Iranian--Khotanese khiraa 'sad, cpressed '--and suggests the identification of a base ki beside khi with the meaning . depressed'. The etymology of kirin, which M. translates as possibly romping', is a coording to him 'unsicher'. Grassinann (Worterbuch) a.v. kiri makes a suggestion to read kirine in the place of kirini in RV 1.100.9. This would then be another form (dative) to be derived from kirin. But this suggestion is not approved by Ollenberg, Rgveda. Textkr. und exeg. Nolen, 1909, 1, 06, who remarks . Der Instrum. aber ist tadellos. . . !. Geldrer also, as his translation mit dem Armen' (Der Rig. Veda, 1, 1951) indicates, does not accept Grassinann's suggestion. * Wackernagel-Debrunner, Altind. Gr., 111, $ 147, p. 279, note. Geldner, Der Rig Veda, 11, also derives kirina in 5.4.10 and 5.40.8 from kiri and translates it with bloss'. See his note on 5.40.8 where he details the different shades of meaning to be obtainod from the basic moaning.blogs". 4 According to Geldner, Der Rig Veda, I, 58, footnote on 12b, ulaa is perhaps the Soma tub (Somakufe') and at the same time it alludes to the Gotama legend of RV 1.85.11. According to Luders, Varuna, 11, Gottingon, 1959, 384-5, utsa is the inexhaustible source or the container which the Maruts milk, swell, or pour out for the thirsty and around which they dance. 5. Barden Geldner, Der Rig Veda, II, 58; Sunger 'Ludera, Varuna, 11, 386. Geldner thus clearly distinguishes kirin (subst.) from kiri (adj.) possibly assuming a different etymology. On this supposed hapax legomenon see F. B. J. Kuiper, Indian Linguistics, XIX, 1958, 361--2. He considers Skt. krid and *kir to be of foreign origin <*kid. In the first, the cerebral d of the foreign word is kept intact but the initial simple consonant is replaced by a cluster with ; in the second, the foreign d is rondered by . He translates kirin 'frisking, dancing'. This explanation of kirin has, however, certain difficulties of which the author himself shows awareness. Madhu Vidya/101

Page #127
NOTES AND COMMUNICATIONS 671 kiri is usually followed by the particle cid. In the interpretation suggested above it can be said that the particle u has taken the place of cid. This, in all probability, has happened in order to fit the metre, as otherwise with cid, there would be one syllable more (Kurino and kurina cid). (2) mana Grassmann (Worterbuch) has two entries for mana. To mana 1, to be derived from man, he assigns the meanings: 'devotion, affection' (Ergebenheit, Anhanglichkeit '); jealousy, anger' (Eifersucht, Zorn', cf. manyu).? About the second mana he says: perhaps & weight (in gold) '('wol ein Gewicht (an Gold'). He compares with this latter Gk. mna, Lat, mina. In the Rgveda, the second mand is supposed to occur only once, 8.78.2. The verse runs : a no bhara vyanjanam gam asvam abhyanjanam/saca mana hiranyaya. Geldner translates 2c as 'with golden saddle-cloth (?)' (nebst goldener Schabracke(?)'). It does not seem necessary, however, to assume a substantive mand (meaning some kind of weight, or utensil, or ornament, or saddle-cloth) homophonous with mana which means' devotion, thought, etc.'.10 The latter meaning can very well serve for mand occurring in 8.78.2. We can translate the verse as bring to us a bull, (which can serve as) a distinguishing mark, 11 a horse, (which can also serve as) a distinguishing mark, 11 together with a golden (i.e. shining) devotion '. The seer prays to the god to endow him with lustrous devotion. Or we may translate 2c as 'with golden (i.e. full of gold) thought'. In that case the poet wants the god to give him a mind which seeks gold. This would agree with verse 9 of the same hymn in which the poet refers to his kamah hiranyayuk ' a desire which seeks gold '.12 M. A. MEHENDALE 7 BR (Worterbuch) and Mayrhofer (EWA, 4, 1983) also have two ontries for mand. . According to BR it is perhaps to be derived from v ma' to measure'. According to Mayrhofer (EWA, 11, 674) it is a designation of a golden ornament, etymologically perhaps related to Skt. mani. * He (Der Rig Veda, II, footnote on 8.78.2) declares mand to be uncertain, semantically as well as formally. In the next footnote, however, he wonders whether mand can be related to carmamna 'tanner' ('Lohgerber') which ocoure in 8.8.38. He further refers to 8.1.32b where saha ivaca hiranyaya is rendered by him as 'samt goldener Sobabracke! 10 Bohtlingk's shorter dictionary and MW have only one mand with the meanings of mand 1 of Grassmann. 11 This seems to be the meaning of vyanjana and abhyanjana and not what adorne' (WAB ziert ()') and ointment' (Salbol') as Geldner in his translation of this verse understands them to mean. abhyanjana ooours also in 8.3.24 and 10.86.7. In the former it refers to an illustrious horse presented to the singer by his patron. It is very likely that in 10.86.7 also it refers to a distinguished horse which Surya rides to go to her husband. 11 In the first two verses of the bymn, the seer asks Indra to bring him a puroddba, hundreds of cows, a bull, a horse, together with mand. All these are again referred to in verse 9 when he speaks of yavayuh, pavyub, hiranyayud, and asvayub kdmah. Madhu Vidya/102

Page #128
SAPTA DEVALOKAN M. A. MEHENDALE Luders in his Varuna I p. 57 ff. points out that in the Rgurda we often read about the three heavens, the three midregions and the three earths. He then observes (p. 64 f.) that in the late Vedic period the 'seven worlds of gods' (sapta devalokah) take the place of the 'three heavens' of the early Vedic period. In this regard he refers to a passage in the Kaus. Br.20. I which reads as : tad devah samaruhya sarval lokan anu pariplavante devalokam pitslokam jivalokam imam apodakam agnilokam sladhamanam vayulokam aparajitam indralokam adhidivam varunalokam pradivar mflyulokam rocanam brahmano lokam nakan saptamana lokanam. He translates it as : "After mounting this (i.e. the wheel of the gods), the gods move around all the worlds: the world of the gods, the world of the fathers, this world of the living beings, Apodaka, the world of Agni, Rtadhaman, the world of Vayu, Aparajita, the world of Indra, Adhidiv, the world of Varuna, Pradio, the world of Mstyu, Rocana, the world of Brahman, Naka, the seventh of the worlds." Luders rightly observes that in the above cnumeration two different lists of the worlds have been put together. The first comprises the world of the gods, the world of the fathers, and the world of the living beings; the second list gives the seven worlds of the gods beginning with Agniloka and ending with Naka. It is, however, not clear why Luders considers Apodaka, Rtadhaman Aparajita Adhidiv, Pradiv and Rocana as the names of the worlds of Agni, Vayu, Indra, Varuna, Motyu, and Brahman respectively. 1. v. 1. upodatam. 2. v. 1. sattamam. Keith in his translation of the Kau. Br. accepts this reading and translates it as the most real'. He says that thc reading saptamam is "clearly wrong. But Luders considers saptamam as the 'only correct reading, He also rejects Keith's translation of saltamam. In the opinion of Luders sattoma can at the most mean the best' (Varuna I. 65. B a 1). The reading sallanam is accepted in the Kaus. Rr. edited by E. R. Sreekrishna Sarma (Wiesbaden, 1968). Madhu Vidya/103

Page #129
Sapta devalokah 107 He says that Adhidiv is a fitting name for the world of Varuna since he lives in the highest heaven; that Indra's world is called Abaraita because Indra himself is called aparajita invincible':' that Rocana, the world of Brahman, is the oldest designation of the heaven's space of light' (Lichtraum); that Riadhaman as the name of the world of Vayu, appears to be reminiscent of stadhamasi svariyotih occurring in the VS 5. 32;2 and that Apodaka 'waterless' is self-explained as the name of Agni's world. It appears that Luders has been led to the above interpretation of the passage in the Kaus. Br. because in this passage the world of Mrtyu is characterised as pradiva and Luders believes (0.62) that in the AV 18.2.48 Pradiv is mentioned as the name of the heaven in which the fathers live. There is probably nothing wrong in identifying the Mflyuloka of the KB with the third and the highest heaven in which the fathers live, of the AV. But there is nothing in the AV verse to indicate that it purports to give the names of the three heavens. The verse in question reads as: udanvalil dyaut avamd pilumatlli madhyamd tfitya ha pradya ur di vdsyam pitara asate. Luders believes that this verse gives udanvalt. Pilumall, and Pradiv as the names of the lowest, the middle, and the third (which is apparcntly the highest) heaven respectively. But these three worlds can easily be interpreted as only giving the special characteristics or the location of the three heavens, and hence they need not be looked upon as proper names. Thus, according to this verse, the lowest heaven is characterized by water, the middle one by pilu,' and the third by the fathers. The prefix bra in Pradiv is only indicative of the highest location among the 1. It may, however, be noted that in the RV. 3.12.4 and 8.38,2 Agni, together with Indra, is also called aparajita. In the VS 28.2, perhaps, tanunapat agni is called aparajita. In the AV 10.2.33 Brahm a is said to have entered the golden puh which is aparijita. . 2. The connection between VS 5.32 and KB 20.1 is not clear. In the VS, as explained by the commentator, the mantra is addressed to the branch of the U dumbara tree. The word stadhaman occurs once again in the VS 18.38. Here it is Agni who is said to be riadhaman. Apparently the association of Vayu with tta, as given in the Kaus. Br., is based on some different tradition and not on the Vaj. San. 3. Incidentally pratididam is the reading accepted by E. R. Sree Krishna Sarma. 4. The words udanvali and pilumati are accented on the last syllable in Varuna 1.62.4. udanvati is also to be found in the Roth-Whitney edition. 5. Whitney, following the commertator's 'worthless etymological gucss', trans lates pilu as 'stars (?)'. Dut verse 35 of this hymn scenis to suggest that the Madhu Vidya/104

Page #130
Charu Deva Shastri Felicitation Volume three heavens since no such word as uffamd is used regarding the third heaven to contrast with the words avamd and madhyama used of the other two heavens. And what has been achieved by the possessive suffixes vant and mant in udanvart and pilument has been sought to be achieved by the expression ydsyam pitara asate. In other words this expression is as good as pitrmat. If pitymatt would have been actually used, then probably Luders might have considered that, and not Pradiv, as the proper name of the third heaven. If pradio is not the proper name of the highest heaven in the AV, there is little justification in interpreting apodaka and other words of the KB as proper names. It is, therefore, better to understand apodaka without water", ladhaman 'the seat of tta", aparajita as 'invincible', adhidio as 'the heaven above (the preceding three)', pradio as 'the heaven further up', and rocana as 'shining." In the above interpretation rocana is taken as an adjective and not a noun (for a proper noun). Luders has clearly shown (p.66 ff) that in the RV rocand occurs as a noun and refers to the invisible part of the heaven, to the region of light above the firmament (unsichtbarer Himmel, Lichtraum). But he also demonstrates (p. 71 ff) that towards the end of the Vedic period, the word rocand has come to be used as adjective meaning 'shining lustrous' (glanzend, leuchtend). Not only this. He further makes clear (p. 78) that rocand seems to have been used as an adjective already in the Rgveda in three places: 3.5.10; 3.61.5; 10. 189.2. 108 It seems possible to show that rocand occurs as an adjective also in two other places in the Rgueda. In 5.69.4 we read about middle heaven is characterized by svadha (ye......mddhye divdh svadhdya maddyante). Hence pilumati, which also refers to the middle heaven, may be translated as 'full of the drink of the fathers'. However, it should be noted that in the verse 48 it is the highest heaven, and not the middle one, in which the fathers are said to live. 1. If, as seen above, one of the divisions of the world is characterized by water (udane dii) then, perhaps, the reading apodakem, as adjective of Agniloka, is to be preferred to apodakam. It would mean that the Agniloka is close (upa) to that division of the heaven which is characterized by water (udaka). In that case apodakam would be an emendation of one who did not know what to do with upodakam. He thought that apodaka "waterless' would go better with Agniloka. Upodakam is also the reading accepted in the edition of E. R Sreekrishna Sarma. Further, in this edition imam is construed with Agniloka, and not with the preceding jivaloka. 2. Keith also does not look upon these words as proper names. He renders rtadhaman established in moral order', adhidio 'over the sky', pradio the highest sky', rocana 'the welkin'. Madhu Vidya/105

Page #131
Sapla devalokak 109 Mitra and Varuna : ya dhartara rojaso rocandsyotadityt divya pdrthirasya. Luders interprets rotana here as a noun and translatest : "Those you, the heavenly Adityas, who are the holders (Erhalter) of the expanse of the light space (Lichtreichraum) and of the carthly (one)." But in this line there is clearly intended a parallelism between parthiva (rajas) and yocana rajas. In other words, rocand here takes the place of divya and hence, like pdrthiva, it should be understood as an adjective qualifying rajas. The line, therefore, may be translated as : "you, the heavenly Adityas, who are the holders of the shining space (i.c. the invisible heaven) and of the carthly space (i.e. the carth)". Luders usually translates tocand as 'Lichtraum', but since in the present passage he takes both rocand and rajas as nouns he combines the two into a single expression 'Lichtreichraum', which hardly seems to differ from 'Lichtraum'. In the RV 1. 19. 6 we read about the Maruts : ye nakasyddhi rocane divi devasa dsale which Luders translates (p.66); "who live as gods in heaven, in the lightspace of heaven (Die im Lichtraum des Himmels, im Himmel als Gotter wohnen)". He observes on p. 76 that in this line the word ndkasya has been chosen in place of the usual divah because the word divi occurs in the same line. Hence, according to Luders, in the above line nakasya rocand is as good as divdh rocane 'in the lightspace of the heaven'. Understood this way, there occurs an unnecessary repetition of heaven'. It is therefore better to interpret rocane as adjective qualifying divi and understand naka in its usual Rgvedic sense attributed to it by Luders viz. 'the visible heaven, the firmament'. The line may then be translated as : "who, as gods, live in the shining heaven on the firmament."3 According to Luders (pp. 75-76) naka appears, not in its special meaning 'firmament', but in the generalized meaning 'heaven' clearly only in two places in the Rgveda, viz. 1. 164. 50 - 10.90.16 and 1. 19.6. It is true that these references are from the later portions of the Repeda and hence the occurrence of naka in them in its generalized meaning is not surprising. But, as shown above, in 1. 19. 6 naka can very well mean the 'firmament'. If this is true then the clear use of naka 'heaven' in the Rgveda is reduced to one. 1. Varuna 1, p. 66. 2. "die Erhalter der Lichtwelt und der irdischen Welt" Geldner. 3. "Die uber dem Lichte des Firmaments, im Himmel als Gotter wohnen..." Geldner Madhu Vidya/106

Page #132
110 Charu Deva Shastri Felicitation Volume While discussing the Kaus. Br. passage referred to above, Luders calls attention (p. 65) to the fact that naka has been named last in the list of the worlds of the gods which, at least apparently, is thus made also the highest among the sapta devalokah. He says that this is rather peculiar since naka originally designates just the lowest, the visible part of the heaven and not the highest. He tries to reconcile this contradiction by pointing out that in the period of the Brahmanas this distinction between the visible and the invisible parts of the heaven was forgotten. But although naka appears last in the Kaus. Br. enumeration it does not, on that account, mean that it is also to be considered the highest. The author of the Brahmana obviously starts the enu. meration of the invisible worlds of the heaven beginning with Agniloka. And having mentioned the sixth, the Brahmaloka, which is apparently the highest, it mentions in the end naka, the lowest. only to add that it is the saptama, i.e. it makes the total of seven complete. Hence in this passage naka can very well be interpreted as designating only the lowest visible part of the heaven. It can be shown that the word saptama does not point to the highest but to the lowest world in one more passage cited by Luders on p. 65 where the 'seven worlds' include not only the parts of the heaven but also the carth and the midregion. In the Mundaka Up. 1. 2. 3 it is said that one who gives the offering at a wrong time (ahutam) or in a wrong way (avidhina hutam), for him the offering destroys the worlds upto i.e. including the seventh (asaptamams tasya lokan hinasti). Luders says that Sankara is right when under the seven worlds' he understands bhuh, bhuvah, suvah, mahah, janah, tapah, and satyam.' Now when it is said that a sacrificer who offers badly loses the worlds, there can be no question of saying that he loses them all including the highest. What is obviously meant is that he by his wrong act loses all the worlds including 1. P 6.2.48 : tasya purane da! and 49: nanlad asamkhyader maj. 2. Sankara : samyag agnihotrakale 'hulam. 3. Sankara : bhuradayah salyantah sapla lokok. Sankara also gives an alternative explanation of sapta lokah which is far-fetched. Luders, apparently, does not look upon it worth mentioning. According to this explanation sapla lokah may refer to the seven generations, the three preceding and the three following, and the one of the sacrificer himself which are related to him by the rite of the offering to the deceased (pindadanadyanugrahena va sambadhyamanah pitspitamahaprapitamahah putrapaulraprapautrah svarmopakarah sapta lokah ......). The meaning 'generation' assigned to loka by Sankara is rather peculiar. Madhu Vidya/107

Page #133
Sapta devalokah 111 the lowest, i.e. he loses even this world, and this is what the expression a saptaman lokan should mean. Thus, here, carth, which is the first in the enumeration, and lowest in order when looked at from the highest world, is designated as saplama 'seventh', i.e. the one which completes the total of seven. 1. ndyan loko 'sly ayaj nasya kulo 'nyah kuru sattama, Gita 4.31. Madhu Vidya/108

Page #134
EKAVRDH AND EKAVRT IN VEDIC LITERATURE M. A. Mehendale POONA In the Sat. Br. 13.2.1 are mentioned food offerings (annahoma-s) which are offered as a part of the Afvamedha sacrifice. The offerings consist of different materials like ajya, saktu, dhand and laja. They are offered throughout the night. The offerings commence with those offered to prana, apana etc. and then come those which are offered to the cardinal numbers like eka, dvi, tri etc. and go upto data and ekasata. Everytime the numbers are raised successively by one. The formulas, accompanying the offerings, like pranaya svaha, apandya svaha etc. are found in the Vaj. Sam. 22.23-24. While explaining the significance of raising the numbers each time by one. the Sat. Br. (13.2.1.5) observes: ekasmal svaha dvabhyam svaha sataya svahaikasataya svahetyaupurva juhoti | anupurvam evainams tat prinati ekottara juhoti ekavrd vai svargo lokah | ekadhalvainam svargam lokam gamayati "To one hail! to two hail! to a hundred hail ! to a hundred and one hail! He offers in the proper order: in the proper order he thus gratifies them (the gods). He performs oblations successively. increasing by one, for single, indeed, is heaven: singly he thus causes him (the sacrificer) to reach heaven" (Eggeling). The word ekavit is used in the above passage to describe heaven. Eggeling translates it as 'single' which shows that he looked upon the word ekavet to contain the root noun vrt. But since the word occurs while giving justification for raising the numbers each time by one (ekottara juhoti) it appears that it would be better to interpret ekavrt as the nom. sg. of ekavrdh 'rising by one'. In the Vedic literature when the heaven is looked upon as consisting of the three or seven divisions, these are supposed to lie one above the other and the sacrificer rises (a-ruh) gradually up to them. Since the divisions are looked upon as placed singly one above the 1. BR assigns to ekav't the meaning 'einfach'. 2. See "Die Dreiteilung des Himmels" and "Sieben Goetterwelten", Lueders: Varuna I pp. 57ff. Madhu Vidya/109

Page #135
78 RTAM other, the heaven is properly described with the epithet ekavrdh and not ekavrt. Only such interpretation of the word will bring out clearly the relationship between the oblations to the numbers rising by one (eka-uttarah ahutayah) and the heaven also rising by one (eka-vrt Page #136
EKAVRDH AND EKAVRT IN VEDIC LITERATURE sarve saha vyajayantaliadt ekavra asayat samvritam tad deva ittham cettham ca vyatyacarams tan mitravaruna acayatam "The gods, verily, all became pregnant together. They were all born together. That lay alone (i.e. by itself, without procreation) rolled together. The gods passed by and beyond it in this way and that. Mitra and Varuna observed it." In this passage also ekavrt is to be interpreted as containing the root noun -yrt. 1. It is not clear what this tad stands for, It is narrated further that a cow was created from that tad, Madhu Vidya/111

Page #137
THE ABODE OF MITRA M. A. MEHENDALE Having demonstrated (Varuna 1. 9-12) that the abode of Varuna is in waters, Luders observes (pp. 12-13) that it is much more difficult to determine what the abode of Mitra is. The question, he says, can be answered really negatively : it is not water. But he next proceeds to point out that there is only one passage in the Rgueda, viz. 5.3.1 which permits us to draw the conclusion that Mitra has something to do with light. In the verse it is said that Agni is Varuna when born ; he becomes Mitra when kindled (tvam agne varuno jdyase yat, tvar mitro bhavasi yut sdmiddhah). Luders rightly points out that the basis for the first statement is that Agni is born in waters and laruna dwells in waters. He argues that a similar relationship appears to have existed between Mitra and the rising flame which forms the basis for the latter statement of the verse. He admits that at this stage anything more can hardly be said. Luders has, no doubt, very ingeniously thrown out the hint that the flame of the fire might he considered the abode of Mitra. But it appears that while doing so he has, perhaps, misscd the mark only slightly. The two words jd yase and samiddhah in the verse cited above seem significant. Agni is Varuna at the moment of his birth (jdyase) in waters. The birth in waters is implied by his being called Varupa who dwells in waters. Next, he becomes Mitra when he rises in flames (sdmiddhah) and this stage occurs when the spark i hat was produced from water comes into contact with some kind of wood (osadhi, vana, vanaspati). Hence it is more likely that wood, rather than the rising flame of fire, is the abode of Mitra. Agni thus becomes Mitra when it comes into contact with wood ......dass Mitra etwas mit dem Lichte zu tun hat." What Luders very proba. bly means ty Licht' is not 'light' but 'fire'. This becomes clear from his later remark: Mitra hat also wirklich seinen Sitz im Feuer, wie Varuna im Wasser-etwas, was sich uns schon fruher als wahrscheinlich ergab". (P 38). For Agni bcing looked upon as Mitra when born see RV 3. 59.4 and Thieme, Mitra and Anyaman, p. 19. 2 Madhu Vidya/112

Page #138
The Abode of Mitra 111 which is the abode of Mitra. This kind of the rise of fire from waters and wood appears to be the one which happens as a result of the stroke of lightning. This is described in the Nirukia 7.23 as follows: yatra vaidyutah saranam abhihanli yavad anupatto bhavati madhyamadharmaiva lavad bhavaty udakendhanah sariropasamanah upadiyamana evayan sampadyata udakopasamanah sariradiptih "where the lightning (fire) strikes a receiving substance (like wood), as long as it is not received it continues to have the property of the middle (fire), (viz.) being kindled in water and becoming extinguished in a solid substance. Only when it is received, it gets transformed into this one (i.e. the terrestrical fire), (having the property of) becoming extinguished in water and burning in solid bodies." . It is perhaps possible to find some support for the above suggestion, viz. that wood is the abode of Mitra, in a Mantra which, with certain variants, occurs in the Atharvaveda and in the Samhitas of the Yajurveda. In the VS 6.22, for example, it reads like-mdpo mausadhir himsir dhamno-dhamno rajans tdto varuna no munca i yddahur aghnyt iti varuneti sdpamahe idlo varuna no munca l sumitriya na dpa osadhayah santu durmitriydsa tasmai santu yd 'sman duesti yam ca vayam dvismah "Do not injure the waters, nor the plants; from every abode, from there, oh king Varuna, relcasc us. When they said (while taking the oath) "inviolable (are the waters)"; when we take the oath saying "oh Varupa", release us, oh Varuna, from there (i.c. from that oath). May the waters, the plants be goodcontracted (or friendly) to us, bad-contracted (or inimical) they may bc to him who hates us and whom we hate." Notes: (1) The formula mapo mausadhir himsih must have been recited while administering the oath. The person taking the oath is warned not to injure the waters and the plants i.e. not to violate the oaths taken in their presence. This will indicate that oaths were taken not only in the presence of water but also of plants." (2) In the text that follows dhamno-dhamno...munca, Roth suggested to emend the text to damno-damno in spite of the complete This wonderful event is probably meant when Agni's birth in 3. 3. 3. is said. to be citrdm. For similar risc of fire also cf. RV 3. 9. 2. 2. durmitriyds in Weber's edn. looks like a misprint... 3. For a request to the plants to get release from the sbacklc of the oath cl. RV 10. 97. 16. For the ritual use of the formula mdpd etc. scc Sat. Br. 3. 8. 6. 10, Kat $S . 10 3 ff. 4 ZDMG 48, 108 (1896). Madhu Vidya/113

Page #139
112 Indological Studies agreement of the traditional evidence to the contrary, H.P. Sch. midt" accepts this suggestion and renders the passage as : "Von jeder Fessel, o Konig-hiervon lose uns, o Varuna ;" But this emendation is not necessary. The word dahman in the formula obviously refers to the waters which are the abode of Varuna and which are mentioned in the opening portion of the formula. These waters, taken in hand to take the oath, represent the oath and the reciter of the Mantra prays that he be released from every such dhaman i.e. from every oath taken in the presence of this dhamana Since in what precedes this formula both apah and osadhayah are mentioned it is conceivable that the expression dhdmno-dhannah refers to both of them as abodes. Now dpah have already been recognized as the abode of Varuna. Hence it might not be wrong to assume that the osadhayah are the abode of Mitra. Although in the Mantra Varuna alone is mentioned, an appeal to Mitra for release from the bond of oath taken with osadhi can be assumed. (3) In the formula cited above from thc VS 6. 22 we get aghnyi iti vdruneli which are clearly to be treated as the pralikas of the two formulas recited while taking an oath. But since in the VS 20. 18 and in all the other parallel passages we get dpo aghnyt ili udruneti (for the accent differences in respect of aghnya see below) it is clear that in the VS 6. 22 the word dpo has been omitted. J. Narten suggests that the Mantra, of which only the beginning has been cited above, can be reconstructed on the basis of RV 7. 49 and we may assumc that the Mantra read as "apo aghnyt ihd mdm avantu*. But there is no reasonablc ground to connect this pralika with RV 7. 49. If at all we have to imagine how the Mantra might 1. KZ 78. 10 (1963). 2. Also er J. Gonda, The Meaning of the Saoskrit Term dhiman. pp. 78-77 (1987). While recognizing that in this formula dhdinan refers to water astypical representative of Varuna's power and presence, he goes too far in explaining the actual significance of this reference. "The apah and aghnyah mentioned in the formula may....., be regarded as locations or refractions of the god's nature and sa were, it would appear to me, snch displays of his power as the diseases which he sent to punish the wicked. From these dhamani the person praying wistes to be released." Regarding sdrod dhdmani muncalu in AV 7. 83.1 he says "Here also Varuna seems to be requested to "release" i.e. to dismiss", to "eliminate" all dhamani". This is also the view of J. Narten, Acta Orientalia Neerlandica (Proceedings of the Congress of the Dutch Oriental Society held in Lci lon on the occasion of its 60th Anniversary, Sth-9th May, 1970) p. 133 (1971). 4. Op. cit. p. 133. Madhu Vidya/114

Page #140
The Abode of Mitra 113 have continued then, on the basis of the context where it appears, we can think that it ran as dpo aghnyd osadhayo aghnydh* "Waters (are) inviolable, the plants (are) inviolable") (4) In view of the fact that in the opening of the Mantra the oath-taker is warned not to injure (hims) the waters and the plants, aghnyd in the second section of the Mantra means 'invioable, not to be transgressed'. It is neither a name of the waters," nor does it mean 'erlesene Kuhe' as a poetic expression for waters.) (5) There are difficulties about the accent of the word aghnyd. These difficulties have been referred to and discussed by H.P. Schmidt and J. Narten. In some versions, e.g. in the AV 7. 83.2 the text reads ydd dpo dghnya iti varuneti. But in others, c.g. in the TS 1. 3. 11. I, we read yad apo aghniya vdruneli. Since vdruna is voc. and dpah can be voc., Whitney changed the accent of aghnyah in the AV. to aghnyah (as in TS) and treated it as a vocative. H. P. Schinidt agrees with this emendation. But this cmendation is also not necessary. The text as it is can very well be understood as the beginning of a Mantra recited at the time of taking an oath with waters (or with plants) and be a solemn declaration to the effect that (these) waters are inviolable"(and (these) plants are inviolable"). With this declaration he vows not to violate the oath or the contract. As regards the text in the TS where dghniyah is clearly a vocative, we have to agree with J. Narten who looks upon the form of the Mantra in the TS as secondary." Or we may suggest that in this formula the Black YV tradition has the word dghniya with the initial udalta (as, e.g in RV 8. 75. 8, also TS 2.0. 11. 2; MS 1. 11. 6). In that case the words apo aghniyah in the formula need not be treated as vocatives. They can both be nom. .., exactly as dpo azhnydh in the parallel passages, and can be similarly rendered as "the waters (are) inviolable." It may be argued that dghnya with the initial udatta is a peculiarity of the Rgveda. It occurs in the Sarhitas of the Black YV only in so far as the Mantra in question is a repetition from the Rgveda. In that case we may say that the Mantra apo aghnyah etc. 1. Because in the opening of the Mantra both waters and plants are alluded to. 2. H. P. Schmidt, Op.cit p. 10. 3. J Narten, Op. cit p 134. 4. Op. cit. p. 131. She suggests the possibility of magische Gleichsetzung" of aghnydh with diadhayah (p. 132). MadhuVidya/115

Page #141
114 Indological Studies now lost to us except in this pratika was current in two forms, one as dpo dghnyah (both nom.) in the Rgveda tradition and the other as dpo aghrydh (both nom.) in the Atharvaveda tradition. The tradi. tion of the Black YV agrees with the former, that of the white YV with the latter. (5) In the final portion of the Mantra the reciter expresses the wish that the waters and the plants be sumitriya to him and those on his side but durmitriya to his opponent. The two words may. of course, be taken to mean 'friendly' and 'inimical'. But they can also mean 'good-contracted' and 'bad-contracted', i.e. the reciter of the Mantra expresses thc wish that the waters and the plants in whose presence the oaths are taken and, perhaps, the contracts executed, may help him to carry out what he has promised in the oaths and contracts. But let them not do so to his opponent so that he will incur the wrath of Varuna and Mitra.? That the word mitra, from which the two above words are derived, has a reference to a contract becomes clear from the fact that in the comment on the use of this Mantra in the ritual it is said in the Sat. Br. 3. 8. 5. 11 that thereby the sacrificer enters into a contract (mitradheya) with waters and plants. The use of the term sumilriyd and durmitriyd with waters and plants would again indicate a close relationship between osadhi and Mitra, possibly that of the abode and the one who lives in it. With reference to spah, the term sumilriyd is as good as suvarunyd. "Contract' is 'friendly' and 'terrible' in exactly the same way as 'True Spec. ch." P. Thicmc, Mitra and Aryaman, P 52. dthabhiman raya! semitriyd na dpa osadhayah santu durmitriyds tdsmai santu yo smdn duesli ydin ca vay dm duismd iti / ydtra vd alena pracdranty dpas ca hd vd asmdt dvad osadhayaf capakrdmyeva lishanti/ idd u idbhir mitradheyan kuruld totho haindin dhe puna, prdvisanli 'He then addresses the Mantra : sumitriydh etc. Where they practise this (i c. conceal the heart-stake in the ground) the waters, verily, and the plants also, as though going away from him (sacri. ficer) stay away; (by reciting the Mantra Sumitriydh etc.) he enters into a contract with them. Thereby they enter into him again fi.e. return to Lim)." Madhu Vidya/116

Page #142
THE RGVEDA-SAMHITAKARA AND FATHER ESTELLER* M. A. MEHENDALE It is customary in such addresses to take stock of the work done during the period between the past and the present session and it is a good practice. But I feel, this year, for the Vedic Section this is not necessary because the third Volume of Dr. R. N Dandekar's Vedic Bibliography has just appeared. This Volume takes care of all that has been done in the Vedic field until about the middle of 1972. The excellence of this Vedic Bibliography speaks for itself and is there for any one who turns to its pages. For this address I have chosen to confine myself to a limited task, viz, to examine the thesis propounded by Fr. Esteller (E) regarding the nature of the present Samhita text of the Rgveda - a thesis which, I am sure, is known to many of you. I shall state my conclusion first: I find myself in total disagreement with the learned Father. But if I have to criticize his views it does not mean any disrespect for his scholarship. In fact I owe a debt of gratitude to him. He was one of my teachers in Bombay while I was studying for M.A. and it was he who first initiated me into the subject of comparative philology. I am algo aware of the fact that Fr. E has spent more time than I have in the study of both the Veda and the texts which can be called the modern Vedangas -- I mean the texts like Grassmann's Worterbuch, Oldenberg's Prolegomena and Noten, Arnold's Vedic Metre, Bloomfield's Vedic concordance and Rgveda Repetitions, and Bloomfield-Edgerton-Emeneau's Vedic Variants. But in the light of what little experience I have of working in this field I fail to see the correctness of Fr. E's thesis. But let me repeat. I mean no discourtesy to my teacher, to a Vedic scholar, and to a past president of this very section. Gathering my information from Dr. Dandekar's second and third Volumes of the Vedic Bibliography I find that Fr. E has been publishing his views on this subject since 1953. His principal forums have been the platforms of the Vedic Section of the AIOC and his publications have almost * This is a slightly revised version of the Sectional President's Address delivered in the Vedic Section of the 27th Session AIOC, Kurukshetra, 26th-28th December 1974. 1 A few works have since appeared, for instance Peder Kwella's Flussuberschreitung im Rigveda (RV III 33 und Verwandtes), 1973, but these are still only titles for us. Madhu Vidya/117

Page #143
98 M. A. MEHENDALE exclusively appeared in Indian Journals. His thesis is that the traditional text of the Rgveda Samhita is a "palimpsest" i. e. a "written-over" text, because it has been redactorially tampered with by the transmitters whom he chooses to call by the collective name the Samhita-Kara (SK). The question then arises : with all the means that are available to us, is it possible for us to reconstruct the original text as composed by the rsi-kavis ? Fr. E's answer to this question is an emphatic "yes". His guiding principle in this task of text-restoration is that the metre and the so-called archaisms are to be considered paramount and they have a decisive value over the traditionally handed down text. It has been long recognized that some of the metrical discrepancies in the Rgvedic verses regarding the number of syllables required in a pada or the structure of a pada can be removed if the sandhis, and the long vowels and diphthongs are resolved or if we resort to svarabhakti. It is also occasionally necessary to change the vowel length for metrical reasons. But these restorations affect only the pronunciation of the text and not its words, or forms of words, or the word-order. These phonetic restorations referred to above do not call for any addition or suppression of the words. Starting from such phonetic restorations Fr. E proceeds to argue that since the SK did not hesitate to disturb the metre for the sake of his later sandhi rules and other matters of pronunciation, metre should be shown paramount consideration wherever it comes in conflict with the present text and that it must be restored at any cost. Fr. E believes in this unquestioned paramountcy of the metre, i.e. he is convinced that the mantras composed by the rsis were metrically perfect throughout and hence, going beyond and even against Oldenberg, he asserts that "...if mere phonetics, pronunciation and samdhi and the restoration of archaic forms are not sufficient to restore the standard pattern, it must be due to the fact that the SK must have made use of other redactorial devices - the most obvious and likely being naturally a change in the order of words." But it is not just for the change in the word-order that Fr. E blames the SK; he also charges him for haying changed the morphological forms of words, for having suppressed or added words, and lastly for having substituted the words of the original Samhita text with his own and all this sometimes even for extraneous reasons. We are informed by Yaska of the three stages with regard to the composition and the transmission of the Vedic mantras : (1) The first stage was marked by the "seeing" of the mantras (or dharman) by the rsis (saksatkrtadharmara ysayo babhuvuh); (2) the second stage was marked by the handing over of these mantras by the rsis to the later generations who did not or could 2 ABORI 50.16. 9 ABORI 50.37. Madhu Vidya/118

Page #144
THE RGVEDA-SAMHITAKARA AND FATHER ESTELLER 99 not "see" the mantras for themselves (te 'varebhyo 'saksaikytad harmabhya upadesena mantrant sampraduh ); (3) the third and the final stage was marked by the arrangement of the mantras and the composition of the Vedangas (upadesaya glayanto 'vare bilmagrahanaya imam grantham samamnasisuk vedan ca vedangani ca. )." When Fr. E speaks of the redactorial activity by which the original text of the rsis was turned into a palimpsest, he probably has in mind the activity of the second stage and the beginning of the third stage mentioned above. What we should note in this regard is the fact that the persons involved in the first and the second stages are clearly distinguished by a very important characteristic - those to whom the dharman was saksat and those to whom it was not. This would necessarily mean that those who merely received the text from those who had 'seen' it must have looked upon it from the very beginning as very sacred. We have no reason to doubt this. This sacredness of the text was marked by two formal features - the words in the text were looked upon as fixed + (niyatavacoyukti) and their order was also considered fixed (niyatanupurvya)' - and in this respect they markedly differed from any other text. I do not think that the Vedic texts were supposed to have these characteristics only after the third stage noticed above. They were looked upon as revealed " and hence sacred in the second stage already. Fr. E blames the SK for having disturbed the metre of the original text for the sake of his later grammar, But he also says that if the SK could "dodge" the sandhi by changing the word-order he did it and thereby avoided doing any damage to the metre. Changing word-order of the received sacred text meant, according to Fr. E, a minor evil to the SK than reduce the number of syllables in a pada as a result of sandhi. To me the matter seems to be exactly the opposite. Since the SK at numerous other places tolerates a metrically deficient pada, he would have easily tolerated it in a few more cases, but would not have changed the word-order of the " revealed" text. He would do the sandhi if his grammar required it and rest content there. Niruka 1.20. 6 ABORI 51.61. 6 Nirukta 1.15. Ibid. Fr. E himself notes that the SK "lived in a traditional conservative atmosphere" where mantras were looked upon as sacred and religious texts (ABORI 50.20-21). But he fails to draw the necessary conclusion from this fact. Moreover, when Fr. E evaluates the different solutions suggested by him for a given line of the Samhita text, while preferring one to the other solutions, he speaks of "the advantage of preserving every letter of the transmitted text quite in tact" or "of the very weighty advantage of keeping the traditional order" (IL Bagchi Mem. Vol. 65). If Fr. E feels occasionally concerned about the preservation of the text in the 20th century how much more must the SK be in the centuries before Christ. Madhu Vidya/119

Page #145
100 M. A. MEHENDALE To take a hypothetical example given by Fr. E a rsi could say indrasya agneh without doing sandhi and have five syllables. But the SK must do sandhi and thereby reduce the number of syllables to four (indrasyagneh). But if he could avoid the sandhi by reshuffling the order of words as agner indrasya and thus retain five syllables he would do it as a "minor evil". To me the case looks the opposite. The SK in such a case would rather rest content with the sandhi giving him four syllables as a "minor evil", or no evil at all, than change the word-order and commit an unpardonable evil. What we have to remember is that the two characteristics of the received text noted above-viz. fixed words with a fixed order-do not say anything about changes in pronunciation due to sandhi etc. But they clearly point out that no morphological, syntactical or lexical change was to be effected in the mantras. I will now present a few examples to show to you how completely unacceptable are the suggestions of Fr. E regarding the changes he proposes to make in the Samhita text in order to turn the "palimpsest" into the "pure " original text of the rsi-kavis. I admit that I have not read all the articles of Fr. E. I had not enough time for that. But from the eight articles which I read for this purpose I think I am sufficiently well informed about Fr. E's views on this subject. In the examples that I now cite I will, in most cases, state after I have stated Fr. E's ideas, how I interpret the available text without making any change in it. If the solutions suggested by me, or by other scholars working in this field, are not acceptable to you I would admit the continued existence of the problem but not change the text one way cr the other to consider the problem as definitively solved. My examples are restricted to the changes proposed by Fr. E in word-forms or word-order i. e, to cases where he thinks he can go beyond Oldenberg. As was mentioned a little while ago Fr. E blames the SK both for breaking the metre as well as for trying to save it at the expense of the word-order or the word-forms used by the rsi-kavis. RV 2. 1.16.opens as ye stolybhyo... agne ratim upastjanti sura yah. The next half then runs as asman ca tams ca pra hi nesi vasya a brhad vadeama vidathe suvirak. Since the stanza begins with the correlative ye Fr. E expects the next half to begin, "stylistically and idiomatically" as tams ca asman ca.!! But since it does not, Fr. E sees the hand of the SK reshuffling the word-order. The reason ?- to dodge ABORI 50.18. 10 IL (Bagchi Memorial Vol.) 54-75, 1957; Proc. Tr. 22nd AIOC Vol. II, 6-31, 1966; IA (Third Ser.) 2.1. 1-23; and 2.4. 1-20, 1967; JASB 41-42 (1966-67) New Series, 1968; ABORI (Golden Jub. Vol.) 1.16, 1968; ABORI 50. 1-40, 1969; ABORI 51, 59-76, 1970. 11 IA (3rd Ser.) 2.1.13. Madhu Vidya/120

Page #146
THE RGVEDA-SAMHITAKARA AND FATHER ESTELLER 101 the sandhi. The word-order as imagined by Fr. E would have forced the SK to introduce the vowel sandhi and recite it as tamscasmanca which would give him only four syllables instead of the expected five. Hence the SK quietly changed the word-order of the received text! This is absolutely unbelievable. In my opinion, sacrificing a syllable at the altar of his istadevata- sandhi"-would have meant a minor evil, or no evil at all, to the SK than changing the word-order of the received sacred text. The fact must be that he simply retained the word-order as he heard it. The reason for the kavi to begin his pada with asman ( the singers ), and not tan (the patrons ), could be that he wanted to give prominence to the singer-kavi's while requesting Agni to lead them, and their patrons, to a better condition. After all the kavis were the purohitas (lit. placed in front ) and hence to be led first and it was they who were to speak aloud in the vidatha. Although Fr. E has now withdrawn bis suggestion" it may be pointed out that at one time he had proposed to change the refrain byhad vadema vidathe suvirah to brhad vadema vidathesu dhirak, not because there was any metrical irregularity in the pada, or any sandhi to be avoided as a result of this change, but simply because he felt that there existed "a natural connection" between viddtha and dhira (cf. agne yahvasya tava bhagadheyan na pra minanti vidathesu dhirak RV 3.28 4 ), and also because the refrain as reread by him would have "a striking parallel " in gantaro yajnam vidathesu dhirak RV 3.26.6. Fr. E even now wants to change the refrain of the seventh Mandala yuyam pata svastibhih sada) nah in a variety of ways" simply because, to him, "the 'yuyam' in the SK's text hangs completely in the air." I doubt whether any one would agree with Pr. E regarding this ground for the change. He himself must have realized its non-compelling force because in one of the three alternatives proposed by him this very word yzydm has been retained by him. Fr. E attributes to the SK some silly misunderstandings of the text he had received from the sis and says that he made changes in the text due to those misunderstandings. Fr. Ewants to do away with those SK's changes and give us the original text of the rsis". Hymn 10.21 has eight 12 To use Fr. E's words. 13 Cf. RV 4.50.8 : sa it kseti sudhita okasi sve .... yasmin brahma rajani purva eti. 14 IA (3rd Ser.) 2.1.13-14. 15 It may be noted that it is on the basis of this refrain, which Fr. E. once wanted to change, that he now wants to accuse the SK, for having changed suvirasah satahimah badema to madema satahimah suvirah RV 6.4.8. For other fantastic grounds for this change see ABORI Golden Jub. Vol. 5-6. 10 JASB 41-42 (New Ser.) 29, 17 IL Bagchi Vol. 69. Madhu Vidya/121

Page #147
102 M. A. MEHENDALE stanzas, of which seven are addressed directly to Agni, and five of these seven actually contain a voc. agne. But stanza 5 is not directly addressed to Agni. It contains a nom. sg. agnia and two third pers.sgs. vidat and bhuvat. It reads as agnir jato atharvana vidad visvani kavya bhuvad duto vivasvatak... Fr. E feels that the kavi could not have a stanza, in the middle of his sukta, which, like the other stanzas, is not directly addressed to Agni. Therefore he wants to change agnih to agne (voc.) and consequently vidat and bhuvat to vidah and bhuvah ( 2nd per. sg.) assuring us that that was how the kavi had composed his stanza. Fr. E chooses to describe this assumed act of the SK as 'a miserable redactorial mis-correction". Now let us follow the footsteps of Fr. E in order to understand the genesis of SK's 'mis-correction'. According to him the SK felt that agne (voc.) was put in apposition to jatan by the kavi! And since he could not understand how this was possible he changed agne to dgnih, so that it can be "syntactically smoother" with jatah, and then also changed the two verbal forms noticed above. I just cannot believe all this. If the text heard by the SK had really agne (voc.) and two verbal forms in second pers., could he not understand that here we have only to supply tvam? It is better to accept the text as it is and think that the kavi, in the middle of his sakta, changed his style probably to address his colleagues and tell them something about Agni. RV 2.19.2 reads as asya mandano madhuo vajrahasto 'him indro arnovitam vi vrscat/pra yad vayo na sudsarany accha prayamsi ca nadinam cakramanta. Fr. E asserts that in the last quarter, nadinam, coming after ca, should have been nadyah (nom. pl. ) 80 that it could be coordinated with prayamsi due to ca and both could be looked upon as subject of cakramanta. This nadyah (nom. pl.) of the original text was misunderstood by the SK as gen. sg. and was further changed to gen. pl. nadinam. This theory is unacceptable. If the original text really contained nadyah, it could not be misunderstood as it came after prayamsi and the conjunction ca's And even assuming this mistake, why should the SK further be guilty of changing the sg. into the pl.? Here is Fr. E's answer : the SK did it in order to get three syllables as in his orthoepy nad yah gave him only two. Fr. E adds one more reason which, however, is not quite clear to me. He says the SK changed the sg. to pl. "both to fill up HIS pada and possibly to indicate that the RIVERS were concerned some how...when "nadinam" alone would fit into the context AND the plural of "prayamsi". I guess what he means is that nadinam (PL) was used because according to the SK, the water released by killing Vstra belonged to many rivers and also because the form prayamsi was in the plural. 19 IA (3rd series) 2.4.2-3. 10 The particle ca placed after the first, instead of the second word, cf. Macdonell, Ved. Gr. (Students), p. 228. Madhu Vidya/122

Page #148
THE RGVEDA-SAMHITAKARA AND FATHER ESTELLER 103 The possibility of misunderstanding on the part of the SK has first to be set aside. The conclusion forces itself on us that the SK has remained faithful to the text that was received by him. Oldenberg tries to justify the text as it is by suggesting that it is the result of perhaps mixing up of two constructions pra yac cakramanta and pra ca cakramanta. Geldner feels that ca has been put at the improper place. He supplies some verb like 'machte', to be coordinated with vi vrscat"... und machte, dass die Labsale der Flusse (dem Meere) zueilten.'90 Renou wants to supply some word meaning 'strengths' before prayansi (...en sorte que ( les forces ) et les recomforts des riviers pussent marche (vers la mer )." Prof. Velankar takes a clue from arnovytam in b and supplies arnamsi before ca" when the Soma offerings and (the floods ) of the rivers flowed forth (to their respective goals i. e. Indra and the ocean)." It is possible to accept the suggestion to add arnamsi but interpret it as referring to waters in the heaven. The last quarter (arnamsi) prayamsi ca nadinam cakramanta would mean the heavenly waters) and the waters of the rivers flowed forth; or, it may be suggested to add adrayah 'pieces of the rock'," before prayamsi, which are alsodescribed as rushing forward together with the waters freed by Indra in RV 4.19.5 ratha iva pra yayuk sakam adra yak; or, we have to think that the avi has already begun to think of the cows which Indra frees from the cave of Vala and therefore supply gavah. We may note that in the very next stanza (2.19.3) the kavi combines both the Votra and the Vala myth." I shall now give an example where Fr. E has charged the SK for having done word-substitutions due to his misunderstanding of the received text.798 In RV 10.39.14 we read etan vam stomam asvinav akarmataksama bhigavo nd Tatham. At first glance the line seems to convey the sense :" this your praisesong, oh Asvina, we have done, we have fashioned it as Bhrgus (fashion ) the chariot." But since the Bhrgus are not famous for the art of chariot fashioning, but the aebhus are, Fr. E is quick to see here the handiwork of the SK. He asserts that the original line of the rsi must have read as ataksama yhavo na ratham. He admonishes the SK in the following terms: "The flabbergasting connexion with chariot making for the Bhrgus is a pure invention of the SK's over-cleverness." The question that stares us in the face is - if the SK heard the text with Ybhavah, why should he have changed it to bhigavah? The metre, the sandhi, nothing seems to be wrong. Fr. E attempts an answer to this question 20 Fr. E. ridicules these honest attempts of Oldenberg and Geldner by saying "The tortured efforts of GELDNER and OLDENBERG are misspent on the SK's disfigured palimpsest." 21 Luders, Varuna 171, f.n. 9. 22 sa mahina indro arno apam prairayad ahihaccha samudram/ ajanayat suryam vidad sa aktunahnam vayunai sadhat // Also cf. Luders Varuna, 193. 32a IA (3rd Series) 2.4.9-10. Madhu Vidya/123

Page #149
M. A. MEHENDALE when he has dealt with this line again. "The answer seems to be," observes Fr. E, "that the SK misunderstands the text in both cases as involving the sense: 'We, Rbhus, have fashioned this hymn like a chariot'". He continues. "But if it has to be some kind of rsi-kavis who lay claim to the authorship of the hymns in question, it could not naturally be supposed to be the Rbhus, but the Bhrgus, who are repeatedly mentioned in the role of authors of divine praises..." That the SK has misunderstood the text the way Fr.E imagines is out of question. I cannot believe that the SK overlooked the presence of the particle of comparison nd which would have clearly shown to him that rbhavah, if that would have occurred in the original text, was an upamana. 104 As a matter of fact Fr.E himself has drawn our attention to the fact that the SK has retained rbhavah in stanza 12 of this very hymn (10.39) a tona yatam manaso javiyasa ratham yam vam rbhavas cakrur asvina. The SK therefore is quite clear in his mind about the activities of the Rbhus and the Bhrgus and it is fair to assume that he has made no confusion. It is therefore better to attempt an interpretation of the text as it is without changing it for our convenience. I feel that in the line etam vam stomam asvinav akarmataksama bhigavo nd ratham the word ratham is not to be understood literally but as standing for a hymn. Similar considerations apply to RV 4.16.20 eved indraya vrsabhaya vrsne brahmakarma bhigavo na ratham where also Fr. E accuses the SK for 'replacing rbhavah with bhrgavah for similar reasons. To me here also the word ratham stands for a hymn. Fr. E unfortunately, does not end his story about the Bhrgus and the Rbhus here and I am therefore required to dwell a little longer on these lines. According to Fr. E the change proposed by him in the two verses above to read rbhdvah in the place of bhigavah is confirmed by the fact that at least in one place the SK did exactly the opposite, i. e. he replaced bhigavah of the original by his rbhavah - and this, in the words of Fr. E!" to compensate the Rbhus by making them take the place of the Bhrgus in a similar situation." The stanza in question RV 10.80.7 opens with the words agnaye brahma rbhavas tataksah. But since it is the Bhrgus, and not the Rbhus, who are known to have a special connection with Agni, Fr. E is quite confident that in the original text of the rikavi there must have occurred the word bhigavah which the SK altered to rbhavah. Why? You have the Father's answer: In making this change "the SK sees a chance of attributing to the Rbhus exactly what he had taken away from 23 Proc. Tr. 22nd AIOC II 21-22. 24 By mistake printed as Bhrgus. Madhu Vidya/124

Page #150
THE RGVEDA-SAMHITAKARA AND FATHER ESTELLER 105 them" in the two stanzas mentioned above. Thus starting from the requirements of metre Fr. E: has gone far away in detecting the causes for the so-called changes made by the SK. In order to explain this last passage as it stands Geldner proposed that the singers identified themselves with the Rbhus. But Fr. E avers that "Geldner is surely wrong." I would understand the passage the way I did the other two above viz. by interpreting in this context the word brahma as ratham. The singers say : For Agni the Rbhus had fashioned a hymn (like a chariot). Consider one more example where Fr. E detects the hand of the SK making changes in the received text not for the sake of sandhi, nor for metre, nor even due to misunderstanding, but for reasons which one can never believe. 13 RV 1.122.3 reads as mamattu nah parijma vasarha mamattu vato apam vysanvan! sisitam indraparvata yuvam nas tan no visve varivasyantu devah// As regards the third pada Fr. E says that it is metrically perfect. Yet he thinks that the SK has reshuffled the original word-order which Fr. E' sees' as yuvam nah indraparvata sisitam. Here are Fr.E's reasons: (1) "In 3c the style and change of address recommend the transposition..." I do not quite know what Fr. E means by this. Perhaps he seems to suggest that in the first two padas Vayu and Vata are addressed but in the third pada Indra and Parvata are addressed. If the first two padas begin with a verb, the rsi-kavi could not have begun the third pada with a verb when there was a change in the deities addressed. Or, he seems to suggest that in the first two quarters the deites are addressed indirectly while in the third the deities are addressed directly. Hence if the first two quarters begin with a verb the third cannot. (2) Here is one more reason to support the above. If the SK had allowed the text to remain as proposed by Fr. E then the words indraparvata sisitam, recited the sarhita way, might have led those who came after the SK to misurder. stand them as indraparvata asisitam. The SK wanted to save the posterity from this calamity and hence he changed the word-order." (3) And there is one more final reason. In effecting this change the SK tried to imitate not only the first two padas of this stanza by beginning it with a verb ( but then why not begin the fourth one also with varivasyantu ?) but also many other verses in the Rgveda which begin with sisthi and sisite (see Vedic Concordance). I do not think it necessary to take these reasons seriously. I cannot, however, resist the temptation of venturing a guess that if the pada in question 25 ABORI 51.66. 28 For similar reshuffles of word-order to avoid confusion see JASB 41-42 (New Ser.) 29, and for the change of form p. 32. Madhu Vidya/125

Page #151
106 M. A. MEHENDALE had really occurred in our Samhita as Fr. E. desires it to be - that is as yuvan nah indraparvata sisitam - he would have looked upon that as an example of SK's reshuffling of the word-order and would have suggested to read it exactly as it now stands -- that is as sisita m indraparvata yuvam nah - because, he would have then said, there are many parallel lines or in the RV in which sisihi stands at the head of a pada and therefore the kavi must have begun his 3rd line with sisitam. I shall now take up a case of a somewhat different nature. 98 RV 1.25.21 reads as ud uttiman mumugdhi ro vi pasam madhyamam cytal avadhamani jivase. It is strange that in the third quarter we have adhamani (neut. pl. ) when in the first two quarters the rsi-kavi has used pasam (mas, and sg.). We expect to read in c adhamam, and not ad hamani. See, for example, RV 1.24.15 where we read ud uttamam varuna pasam asmad avadhamam vi madhyamam srathaya. Tradition ascribes both the suktas (1.24 and 25 ) to Sunahsepa. Following his practice to ascribe any oddity in the Samhita text to the mishandling of it by the SK, Fr. E hits upon the original text to be ava adhaman nu jivase. But he is unable to pick up metre or grammar as a ground for the SK's change. He therefore imagines that in the oral transmission of the text adhamam nu got altered to adhama nu and thence to adhamani. Normally Fr. E leaves out the question of accent in his discussion. But here he finds it possible to account for the disappearance of the udatta of nu. He wants us to believe that adhamanu was mistaken for adhama + anu and hence was pronounced adhamanu. From this we go over to the final change adhamani. All this is too much for any one to take in. It is impossible to believe that the SK himself did not notice the incongruity between uttamam and madhyamam pasam on the one hand and adhamani on the other. And if he had felt himself free to alter the text he would have changed ad hamani to adhaman long ago. But if he has not done anything of the sort, the conclusion is inescapable that he did not feel himself free to alter the text and hence left to the posterity what he had heard from his predecessory. I would therefore see if I can explain the use of adhamani. Perhaps adhamani refers not so much to the pasa as to the constituent strings or some 21 It is interesting to observe that what are parallels for Fr. E become pseudo-analogies if, as assumed by Fr. E., they are used by the SK 28 IA (3rd series) 2.1.15 ff. 29 The intervening stages imagined by Fr. E. are absurd. They are : First adhamam nu > adhaman nu due to assimilation; then adhamanni > 'adhamanu by hapology (is this really haplology? There is no repeated pronunciation of n, only its implosion is lengthened), and compensatory lengthening: finally -un > -ni due to' the assimilative force of palatals ji in jivase. Madhu Vidya/126

Page #152
THE RGVEDA-SAMHITAKARA AND FATHER ESTELLER 107 such thing which made up the pasa. And if the word for the string, or whatever it was, was neut. a neut. pl. can be understood., Or, it may be argued that in the last quarter the rsi chose to speak in terms of enas instead of pasa and hence the neut. Freedom is sought as much from the pasa as from enas and the same verbs muc and srath are used with them. Cf. kytam cid enah pra mumugdhy asmat RV 1.24.9 and vidva adabdho vi mumoktu pasan 1.24.13; or rajann enamsi sisrathah kytani 1.24.14 and ud uttamam varuna pasam ... srathaya 1.24.15. An example30 of word-substitution by the SK due to his having misunderstood the text is provided, according to Fr. E, by RV 7.8.6 which reads as idam vacah... sam yat stolibhya apaye bhavati dyumad amivacatanam raksoha // Fr, E. observes a problem which is grammatical, not metrical, for Taksoha is mas, and therefore cannot qualify vacah. According to Fr. E the anomally is the creation of the SK who has substituted raksoha for the original jaritre on account of a number of misunderstandinge. In order to understand the how and why of this substitution we have to follow Fr. E.'s detective work. First, with the help of the modern Vedanga texts, Fr. E discovers that RV 2.38.11 has a line somewhat similar to the above: san yat stotibhya apa ye bhavaty urusamsaya savitar jaritre on the basis of which he concludes that the original mantra 7.8.6 must have read as .... dyumdd amivacatanam jaritre. We may ask : why did the SK make the change? Fr.E's reply: since the same line contains synonymous words stotybhyah and jaritre, the SK retained the former and changed the latter. Incidentally, Fr. E does not raise, and therefore does not have to answer, the question : why did the SK not remove jaritre from 2.38. 11 which also contains stolybhyah ? A second question : how did the SK hit upon raksoha when he decided to remove jaritre ? Fr. E's reply : because RV 10.97.6 viprah sa ucyate bhisag raksohamivacatanak contains raksoha and amivacatanah side by side, therefore the SK put raksoha by the side of amivacatanam in 7.8.6. A third question : how is it that the SK brought a neut. vacah and a mas. yaksoha in syntactic relationship ? Fr. E's reply : since in RV 1.129.6 we have ya isavan manma rejati raksoha manma rejati where a manma neut. and raksoha mas. occur close to each other therefore the SK feit no hesitation in relating vacah with raksoha forgetting for a moment that in one case there was subjectobject relationship between raksoha and manma and therefore there was 30 IA (3rd series) 2.1.12 ff; JASB 41-42 (New Ser.) 28. 31 The expression "we, the modern text-critical detectives" is used by Fr. E himself, ABORI 50.15. Madhu Vidya/127

Page #153
108 M. A. MEHENDALE grammatically nothing wrong in having them syntactically related,99 whereas in the other case there was noun-adjective relationship between vacah (neut.) and raksoha (mas.) and hence it would be a grammatical anomally if they were brought in syntactic relationship. Reading this detective story one would come to the conclusion that the SK, as depicted by Fr. E, had very poor knowledge of Sanskrit. But the author of this detective story himself tells us again and again that grammar was the istadevata for the SK and that the SK moved in the circle ot the sistas. If this is true we cannot entertain for a moment the correctness of the reasons given by Fr. E. If the SK retained raksoha, inspite of the apparent difficulty, he must have heard the text that way. We have to see if we can explain it. In the RV 10.162.1 we read brahmanagnik samvidano raksoha badhatam itah. Here, Agni, when joined with bruhman, is called raksoha. It is therefore likely that in our passage also, in which a hymn is said to have come into being for Agni and for whom it is supposed to be doubly strong (?) (idam vacah...agnaye janisista dvibarhah) the singer says "May Agni become raksoha" (lagnih vacasa) raksoha (bhavati )]." Or, we may look upon this as an instance of a slight corruption in the handed down text. Perhaps, the original text contained raksoha (neut. sg.) which would agree with vacah and that in transmission it became raksoha. Even the famous Gayatri mantra tat savitur varenyan bhdrgo devasya dhimahil dhiyo yo nah pracodayat // (RV 3.62.10 ) has not escaped the textual attack of Fr. E34 In his opinion since the stanza opens with tat we must have yat," and not yah, in the relative clause. True, this is what we expect. But if it is not what we expect, there must be some reason for the kavi to use yah. Instead of trying to understand it, Fr. E. blames the oral tradition for altering yan nak to Jo nah in stages, first by haplology yannah > yanah and then by assimilation dhiyo ya nah > dhiyo yo nah. 32 Geldner mentions the possibility of looking upon raksoha as an adj. of manma noting that the former appears in mas, form instead of ncut. Oldenberg, though he prefers to treat raksohd as mas, remarks " "demon-slaying prayer" (Griffith) ist denkbar..." 38 Geldner and Velankar take rak sohd as an adj. of vacah. 34 ABORI 50.31-32. 35 For the impossible and irrelevant supports sought for the tat...yat construction, see JASB 41-42 (New Ser.) 37-38. In one case, RV 1.141.1, Fr. E. extracts yat out of yato and in another text he reads yat where it does not exist (5.82.1). From these he seeks support for his argument. 30 For the other impossible phonetic alterations in the oral transmission imagined by Fr. E one may note na at pathiyah first giving by sandhi and Pali-like assimilation nappathiyah, and further by wrong analysis joining the initial na to the previous word jama and miscorrection to janandh pathiyah of RV 10.14.2 (Cf. IL Bagchi Mcm. Vol. 62; also 70 for other fanciful observations). Again, Fr. E., on metrical grounds wants to read RV 6.40.2 which opens as asya piba, as asya at piba. This, he says, first changed to asyappiba, and then to asya piba, and was finally edited as asya pibal Madhu Vidya/128

Page #154
THE RGVEDA-SAMHITAKARA AND FATHER ESTELLER 109 If Fr. E is correct in his guess that in the original text we had ydi in c then the verb pracodayat will have to be connected with bhargah through it. But this result is undesirable. In almost all the uses of the verb cud, or of its derivatives like coda, codity or codayitri, one notices that it is looked upon as an activity of some deity like Agni, Indra, Asvina, Soma, Usas and some others. Only in a few cases (3.42.8, 8.68.7, 10.120.5, 2.13.9?) does it occur in connection with a priest or a singer. But in not a single case does the verb cud have an abstract notion like bhargah as subject. It is therefore proper to judge that in the Gayatri mantra the seer wanted to relate the activity of stirring up the minds of the poets, to the deity Savits and not to his bhargas and hence he used yak. This conclusion receives support when we find that in the parallel passages it is the deity who is requested to stir (or sharpen) the minds of the poets. In 6.47.10 we read indra mila mahyam jivatum iccha coda ya dhiyam a yaso no dharam,37 Indra is called codayanmali in 8.46.19 and Agni in 5.8.6. Pusan and Vayu are called codilara matinam 5.43.9, and there are many other passages. In addition to the change of yat to yah, Fr. E also wants savituh to be changed to savituh for the sake of rhythm.38 But why should the SK change the long vowel to the short one? For, as noted by Fr. E himself, has he not retained the long i in prasavita in two places 4.53.6 and 7.63.23? The SK had certainly an ear for rhythm, as much as we do, and if in spite of that he has passed on savituh to the posterity, he must have certainly heard the text that way. Dative Singular in -ai Fr. E looks upon his discovery of the archaic dat. sg. in -di in the RV as one of his most significant contributions to the study of this text. As is wellknown the available Samhita knows only the -aya ending. But Fr. E is convinced that that is the result of the SK's changing the archaic -ai to the later - ya ending in the received text. Let us consider a few examples cited by Fr. E. to prove his case, and, to begin with, the one discussed by him as "one of our key-problem texts""). RV 7.88.6 reads as yaapirnityo varuna priyah san tvam agamsi krndvat sakha te, ma ta enasvanto yaksin bhujema yandhi sma viprah stuvate varatham// I shall not say here anything regarding the drastic changes proposed by Fr. E in the third pada but restrict myself only to the consideration of his suggestion to change viprah (nom.) to viprai (dat.) in the last pada. Fr. E gives two reasons for effecting this change. (1) In the first instance he points to the other passages 3? Also cf. codah kuvit tutujyat sataye dhiyah 1.143.6. 38 ABORI 50.31. 39 Only the Padapatha in both places has pra-savita. 10 Proc. Trans. 22nd AIOC Vol. II. 10-14; JASB 41-42 (New Scr.) 35-36. Madhu Vidya/129

Page #155
110 M. A. MEHENDALE in the RV where vipra and the participle stuvat occur in identical case forms, e.g. vipraya stuvate 8.85.5; viprasya va stuvatak 8.19.12 etc. Therefore, he argues, in our passage also we should expect a dat. of vipra by the side of stuvate. (2) Secondly, viprah, as nom. sg., has to be interpreted as referring to Varuna - and this is what Geldner, for example, does" -but, argues Fr.E, although vipra in the RV is used with reference to other deities, chiefly Agni, it is never used with reference to Varuna. Both the arguments are inconclusive. Let me start arguing this way. If the SK heard viprai in the text recited to him and he understood it correctly - Fr. E does not grudge this to the SK - he would have rendered vi prai into vipraya and not viprah. To say that he did not do it because it would have given him one syllable more is not convincing because the SK has allowed metrical irregula. rities to remain elsewhere any way, and, he was not so insensible to meaning as to change a dat. into nom. And, if the number of syllables was his main concern, and not the meaning, why did he not change viprai to vipre (loc. sg. ) as Fr.E himself believes he has done elsewhere? In fact vipre would have been better instead of the "unfitting" viprah. The conclusion, therefore, to be drawn is that the SK heard viprah in the text learnt by him and he faithfully preserved it. In fact we should be grateful to him for having resisted the temptation to change viprah to vipraya in view of the parallel passages mentioned by Fr. E. These passages, certainly, could not have gone unnoticed by the SK. As mentioned above, earlier scholars have understood viprak to refer to Varuna. In support Geldner refers to stanza 4 of this very hymn where Varuga is clearly called viprah (vasistham ha varuno navyadhat... stotaram viprah ... ). But Fr. E wants to get rid of this evidence, inconvenient to him, by blaming the SK for having changed there the original vipram to viprah. And why should the SK do it there? Fr. E says in order to give support to his change in the stanza under discussion and to show that Varuna was called vipra in the RV! But there is one more passage where viprah is clearly used with reference to Varuna. In 6.68.3, which is addressed to Indra and Varupa, we read vajrenan. yah savasa hanti vytrar sisakty anyo vrjanesu viprah. But here too Fr. E wants to change viprah to his dative viprai because the root sac governs dat. (and more frequently accusative). But that is no reason to change viprah to viprai in this verse. It is not necessary that the seer should have clearly expressed the object of the verb. Obviously he has chosen to keep it unexpressed. And if the SK 41 Keith JRAS 1908.1127, on the other hand, looks upon this as an example of the use of nom. As voc. 2 ". ...the SK's locatives aro original datives. ...." Proc, Tr. 22nd AIOC II.25, Madhu Vidya/130

Page #156
THE RGVEDA-SAMHITAKARA AND FATHER ESTELLER 111 wanted to get rid of viprai, why did he not change it to vipram since sac governs also the acc. ? But viprah, the nom. as going with Varuna, is significant. In the first half of the line, in which anyah refers to Indra, he is described as performing the deed ascribed to him with the help of his vajra and savas (indicative of physical strength); in the latter half of the line, in which anyak refers to Varuna, he is described as performing the deed ascribed to him by virtue of his being a vipra ( indicative of the power of inspiration ). And now let us consider what Fr. E has to offer as the final confirmation of the -i dative." RV 1.19.1 reads prati tyan carum adhvaram gopithaya pra huyase. Fr. e agrees with Geldner according to whom gopitha means only "protection"." But if this is true, why does Fr. E want a change?" Are not gods invited to give protection (cf. Grassmann hu)? But Fr. E notes that Sayana interprets gopithaya as somapithaya." And taking his hint from there Fr. E fixes the original text as somapithoi pra huyase with his archaic dative. If the SK had changed it to his classical dative somapithaya it would have given him five syllables and disturbed the metre. He, thus concludes Fr. E, therefore, quietly changed it to gopithaya. All this is nothing but play of imagination. I would not like to join Fr.E in this game, but I am tempted to. I would like to bring to the notice of the SK that if he wanted to change somapithai he could have done it and still avoided the metrical flaw by suppressing pra of the original text : somapithaya huyase would have enabled him to remain closer with the original text and avoid a very bold change to gopithaya pra huyase. And one thing more. If Fr. E really means what he says then he should not have at least expressed his agreement with Geldner in the interpretation of gopitha in this stanza. If the SK changed somapitha to gopitha he must have understood by the latter term 'a drink of milk' and not 'protection'. "m'a drink of min mapitha to cobinterpretation of I do not think it necessary to add more examples. I have given enough to show how untenable Fr. E's conclusions regarding the change in word - forms and word-order are. Now I wish to draw your attention to a theory according to which even the changes which affect pronunciation will have to be looked into carefully. The purpose of such changes was to achieve metrical and rhythmic 15 Pr. Trans. 22nd AIOC II.26.27. 4Grassmann and BR have 1. 8opitha 'protection' and 2. Sopitha 'milk-drink'. The word is similarly treated in Alt. Gr. I 20 and II 719. 45 That an offering of honey mixed with Soma (somyam madhu) is mentioned in stanza 9 does not mean that there must be a reference to Soma in stanza 1. 10 The word actually used by Sayana is somapanaya. Fr. E could have more properly referred to Yaska (10.36) who interprets the word similarly and whom Sayana follows and quotes. Madhu Vidya/131

Page #157
112 M. A. MEHENDALE regularity. But scholars have already started suggesting that at least in some Cases metrical irregularities were introduced into the verses on purpose by the seers. The seers structured the padas in such a way as to reflect symbolically the thought conveyed by them. This idea was expressed by P. Thieme, perhaps some what hesitatingly at first, in a footnote to his article published in Language 31.434 f.n. 4 (1955)". With regard to the metrial defect found in RV 5.59.2 b naur na purna ksar ati vyathir yati" ( the earth ) drips like a fully loaded ship while moving forward rocking"- it has 11 instead of 12 syllables-- he points out that it is not necessary to correct the defect' by pronouncing naur as disyllabic. He first points out that the line is an example of a typically deficient Jagati in which the caesura occurs after the fourth syllable but the line continues as if the caesura had occurred after the fifth." He then says that it is also possible to venture a guess that the choice of structuring the line that way was intended to serve an artistic purpose. " The line first hesitatingly halts at the early caesura and then rushes to the end, skipping on its hurried way, as it were, the fourth (long ) syllable after the caesura, and picturing thus the rocking movement of the ship." Thieme appears now convinced about the correctness of his view expressed above. In an article published recently he takes up this, as well as some other examples, in order to establish his view.point. With regard to the above verse he says firmly that only the monosyllabic pronunciation naus is correct and asks " where else would exist a disyllabic au arisen out of au in the Rgveda?" Another example given by Thime is RV 3.59.2d nainam anho asnoty antito na durat which has a rather longish opening of seven, instead of five, syllables before the caesura. Scholars suggested to get rid of this anomally by dropping, for example, enam (thus Oldenberg, Proleg, 86, also Noten). Thieme looks upon these violent attacks (gewaltsame Angriffe) on the text as absolutely uncalled for. According to him even if one were to forget the word amho'narrowness, distress ''one would get a metrically correct opening. But, he argues, one cannot strike out that word on that account. In fact, by crowding the beginning of the verse with the syllables of the very word anhas the poet has depicted in language the sense of narrowness and distress conveyed by the word.co The examples given by Thieme from the Veda and later literature in his article make a very interesting and refreshing reading. I shall mention here only one more of his examples for the same has been used by Fr. E in support of 47 P. Thieme, Kleine Schriften 702. 48 Oldenberg, Prolegomena 66. 10 Sprachmalerei 'KZ 86. 64-81 (1972). 50 This view was earlier expressed by Thieme in his book Mitra and Aryaman P. 44 f.n. 31a. Madhu Vidya/132

Page #158
THE RGVEDA-SAMHITAKARA AND FATHER ESTELLER 113 his thesis of the archaic dative in the RV." Fr. E suggests to restore rhythm to RV 7.88.3 pra prerkha inkhayavahai subhe kam by word shuffling and diphthongresolution. Thieme points out (pp. 70-71 ) that it is true that the verse does not have a caesura after the fourth or the fifth syllable, but the verse can at once be rendered normal if we decide to locate the caesura in the middle of the word irkha yavanai after the first two syllables. The structure of the pada here symbolises the movement of the ship on a swing (prenkha). The pause in the middle of the word depicts the abrupt stopping of the movement for a moment when the swing has reached its highest point on either side. - Even as it is I doubt very inuch whether any one would agree with Fr.E in accepting the absolute paramountcy of the Vedic metre from which he derives his authority to handle the received text the way he likes. Now the above consideration would force us to take a second look at the very starting point of Fr. E's thesis, viz, the metrical defects, for the removal of which he has taken all these pains. Fr. E's entire attempt is directed towards proving that "the SK's Samhita is a veritable palimpsest'. The dictionary meaning of this somewhat unfamiliar word is -" Writing-material, manuscript, the original writing on which has been effaced to make room for a second." According to Fr. E it means- "a manuscript text that has been written over a pre-existing text by using one and the same writing surface a second time after effacing the original script by white-washing it over in one way or another."* Two considerations emerge from this explanation. First, as emphasized by Fr. E himself by underlining the word 'written', the word 'palimpsest' refers to writing; and, secondly, it indicates that in palimpsests only the writing material is the same, but the texts need not be the same. The second text could be as different from the first as a Ramayana is from a Mahabharata or much worse. The texts to which the word palimpsest is applied are quite different from the ones that are effaced (cf Encycl. Br, and Encycl. Am.). Now first consider that the activity of the SK could certainly not be called a writing activity, Fr. E has described the SK's text as "a real sruti text in the sense that it is originally not a lipi text". And secondl y the nature of the changes which Fr. E alleges the SK has made in the Samhita text concerns orthoepy and beyond that the forms of certain words, suppression or addition of a few words, and the change in the order of the words. These changes do not make the second text totally different from the first. Fr. E has, in fact, asserted that "the SK was not ... producing his redactorial text in a 81 Pr. Trans. 22nd AIOC II. 16. 52 I.A. (3rd Series) 2.1.3. 33 ABORI Golden Jub. Vol. 15. 54 IA (3rd Series) 2.1.6. Madhu Vidya/133

Page #159
114 vaccum, but in circless that were consciously and professionally conservative... hence he could not be an out and out radical iconoclast." He has also given his judgement to the effect that the SK has been moderate in his retouchings with the result that "the overall substantial import of the Rgveda remains practically the same." M. A. MEHENDALE We may leave aside for a while the question whether the word 'palimpsest' applies to the writing material or to the text written on it. But if the SK was conservative by profession and his alleged touches moderate, even if we were to believe everything what Fr. E has said about the changes, is the SK's text qualified to be called a 'palimpsest"? Is this not gross exaggeration calculated to prejudice the reader against the SK even from the start? I may appear oversensitive in my objecting to this word 'palimpsest'. I admit it is difficult for me to accept in academic discussion a word that highly exaggerates the facts. This 'eads me to touch an aspect of Fr. E's writings which I most reluctantly do. I would have preferred to neglect it if it had not extended so profusely over his writings. I have so far practically refrained from giving you samples of the exaggerated expressions used by Fr. E with reference to the activity of the SK. But now consider the following expressions used by him to convey the simple meaning "The SK has changed or altered the rsi-kavi's text":-- "The SK has meddled with the original text; SK's woefully defaced palimpsest; SK's deleterious palimpsest-making; brazen palimpsesting; redactorial spuriousness; redactorial manipulations by which the SK attempted to concoct a few stanzas; the whole text has been ruthlessly tampered with; SK edits his text in that outrageous way; redactorial... trick of the trade; samdhi is SK's hobbyhorse; SK uses the fraudulently" clipped" forms; normal padas now ruined by the SK; any flaw in versification is due to the well-intentioned but deleterious meddling (and ignorance) of the reciters and samhitakara; he adds and reshuffles pseudo-analogically without scruple!; the glaring after-thought represented by the foolish... SK's text is senseless; this is pedantry-panditry in excelsis"; the Vedic Samhita-Karas are a freely editing tribe," But Fr. E generously concedes that there is a "method in "SK's "madness". Yet he deserves to be characterized as "the woebegone redactor" or referred to with such exclamations as "of such stuff is the SK made!" or "The SK is the limit! Poor rsi-kavis." These expressions give an impression that the SK was a wily gangster indulging in fraudulent activities which produced faked and spurious products. 55 IA (3rd Series), 2.1.5. 50 ABORI 50.21. Madhu Vidya/134

Page #160
THE RGVEDA-SAMHITAKARA AND FATHER ESTELLER 115 No wonder Fr. E. speaks in terms of catching the SK "red-handed at his little game." Unfortunately the story does not end here. The Father can use language replete with violence. He speaks in terms of the use of the knife in a surgical room : " ... what kind of a radical redactorial surgeon that SK was who brought his scalpel to bear on the archaic hymns of the ancient Rgvedic bards. Scarcely anything escapes his operating touch"; or to the use of the knife outside the surgical room - but still leaving the unfortunate victim alive : "(the "improvements" of the SK ) ... lamed and maimed the verses, and with them, figuratively, their innocent authors "; or " Add to this the brutally mangled ..." And much to one's regret one notes that Fr. E chooses to speak even in terms which leave the victim altogether dead. The reference may be in terms of a child-sacrifice : "... they (i. e. the SKs ) went so far that for the sake of their new-fangled correctness they ruthlessly sacrificed to their own "modern" fashionable "ista-devatas" ("Siksa" and "vyakarana") the pet child of every self-respecting poet, "chandas" the archaic metre and rhythm of the rsi-kavis..."; "... hence his second rhythmical murder at the analogical Moloch's altar..." (Moloch was the Cannanite god to whom children were sacrificed); or it may be an act of butchery : "This is the way our Samhita palimpsest was butchered innumerable times"; or it may refer to guillotines and murders: "... which has been practically everywhere systematically guillotined (by the Procrustean Redactor ) ..." [Prokroustes, a robber who made victims fit his bed by stretching or lopping them );"... which has throughout been redactorially tampered with in a most cavalierly-procrustean style ..."; "The latter (i. e. the SK) has no scruple to "murder" the original rhythm ..."; "But SK has his own system ... which he clamps upon the original text even if it murders the meter (sic) and rhythm." (Italics in the original, not mine). No wonder Fr. E considers the Samhita text a "dessicated (sic) mummy." I venture to ask : If Fr. E really means what he says, what chance has he of reconstructing the original mantras in their freshness and vigour on the basis of a desiccated mummy, or the material which he looks upon as maimed, butchered or murdered ? Another question: Is it proper to use such strong and unbecoming language in academic writings? True, these child-sacrifices, guillotines and murders are all metaphorica). But is this play of rhetoric, emphasized occasionally by italicizing or printing in bold type, in place ? You will kindly note that all I have said regarding the use of the word 'palimpsest' or of the other exaggerated expressions is not a part of my argument against Fr. E's thesis. It is not that if this was removed from his writings everything will be all right. 87 ABORI Golden Jub. Vol. p. 16; JASB 41-42 (New Ser.) 42. Madhu Vidya/135

Page #161
116 M. A. MEHENDALE Finally, consider one implication of Fr. E's reconstructions. As for him. self he is convinced that his suggestions make the Samhita text "infinitely better 63 because they bring to the available text metrical and rhythmic regularity, archaic orthoepy and grammaticality, better word order and logicality. Now if by some means, more solid than what I have tried to present today, it is proved beyond doubt that the SK has really not altered the text of the kavis, then it would mean that in the opinion of Fr. E the text presented by him is far superior to the one composed by rsi --kavis themselves. I am not sure my arguments will convince Father E. and that he will withdraw all his suggetions. But the picture, happily, is not altogether without its sil. ver lining. On the one hand we note with gratitude that Fr. E does have a few good words for the SK. He calls the Samhita-Karas "master-builders of the Rgvedic tradition " and "towering giants, not only in the religious-cultural aspect but also in the linguistic-literary one." And, on the other, he has stated on occasions that his text-critical observations are only provisional. He has shown Courage to withdraw some of his earlier restorations in favour of the received text. As late as 1969 he has gone to the extent of admitting that even the kavis themselves have already set the example 10 implying thereby, I presume, that the kavis themselves have, in some cases at least, composed metrically irregular padas. I shall therefore overlook for the time being his assertion that "..... the vast majority (of restorations ) have already reached the stage of text - critical definitiveness that shall not be in need of revision.... "1 or his pronouncements, made with supreme confidence : "...there can be not the slightest text-critical doubt that the restored form.... is and can only be the original Rgveda which underlies the SK's Samhita-palimpsest...,' and request Fr. E to give up his role of a "modern text-critical detective" which he has taken up in order " to follow the trail of their (i.e. SKs') innovations, crack their redactorial code, and with growing certainty unmask their well-meant but disfiguring "improvements."69 I appeal to him to be quick in his "readiness for self-correction": and pray to him objectively to evaluate this humble criticism -I shall not say from "a fellow scholar" but--from his one time pupil, a fact which I restate with a feeling of respect for the learned Father. Thank You. 56 IA (3rd Series) 2.1.14. 50 ABORI 50-21. 60 ABORI 50.20. 61 ABORI Golden Jub. Vol. 16. 62 ABORI 50.15. 88 ABORI 50. 39-40. Madhu Vidya/136

Page #162
TWO COMPOUNDS WITH AGNI M. A. MEHENDALE 1. agnihvara The word agnihvara appears in the Maitr. Sam 1.3.35 in one of the mantras used at the time of drawing the dadhigraha. The mantra in which the word occurs runs as : upayamagrhito'si prajapataye tyd jyotismate jyotismantam gihnami ratan de vebhyo daksiya dak sav/dham agnihvarabhyas tyd Ita yubhyo indrajye $phebhyo ...... pithivyai tvaentarik saya rva dive tva ! With a few changes in the order of the words, the above mantra appears in the TS 3.5.8.1, the KS 39.5 and the Kaps 45.6. But the one important difference is that in the mantra of the latter three Samhitas we read devebhyo 'gnijihvebhyah 'to gods whose tongue is Agni' in place of devebhyo ......... agnihvaribhyas of the Maitra-San. MW gives, with a question mark, making a mistake in the fire-ceremonial' as the meaning of agnihvard, apparently deriving hvara from hvar * to deviate or diverge from the right line '. But this meaning is hardly suitable in the context. It is, however, possible to explain hrara as a primary derivative from either hu to call' or hu' to offer' with the suffix -ara (Alt. Gr. II, 28112, pp. 215-216). The word hvara would mean 'inviter' or 'offerer' and the compound agnihvara would mean whose inviter is Agni' or 'whose offering priest is Agni'. This meaning would neatly suit the context. It may be noted that the gods are called agnihotarah in the kV 10.66,8 ( agnihotara rtasapo adr uho 'pa asrjann inu yitraturye ). ind of the indhi offerine ihan part At the end of the M$ 1.3.35 occurs a mantra which is recited at the time of praying after the dadhi offering has been given. The first line of this mantra runs as tisro jihvi sya samidhah prijmano 'gner akinvann usijo amTtyave where jihy: occurs as an adjective of agni. A similar mantra appears in the RV 3.2.9 with yahvi sya in place of jihva sya. It reads as tisro yahvisya samidhah parijmano 'gner apunann usijo amityavah. It is difficult to look upon jihvi of the MS as a corruption of yahve of the RV. The shorter dictionary of Bohtlingk records this jihyd as an adjective of Agni but does not give any meaning. However, Bohtlingk seems to connect jihva with jihva 'tongue'. In the light of the word hvara occurring in the same section of the MS and its explanation suggested above it is possible to look upon jihvi sya Madhu Vidya/137

Page #163
116 M, A. MEHENDALE of the MS as a corruption of juhvasya. The word juhva, like hvard, can be explained as derived either from the reduplicated form of hu to call' or hu to offer'. It can carry the same meaning as hvira. The corruption of juhvasya to jihva sya is easily understandable. 2. agni somabhigau In the Nirukta 14.31 is cited a somewhat obsure stanza the first line of which runs as a yahindra pathibhir ilitibhir yajnim imam no bhagadheyan jusasva. Ja the cryptic comment which is given after it yajiim imin no bhagadhayam has been first repeated with some change as yajram imam no yajnabhagam. This passage thus explains the text word bhagadheyam 'share' as yajnabhagam.share, i. e. offering given in the sacrifice'. This explanation is then immediately followed by the compound word agni somabhagau. As the verse is addressed to Indra and he is requested to come to enjoy the offering given in the sacrifice, agni somabha gau cannot mean offerings given to the deities Angi and Soma, either jointly or separately. Hence the compound has to be taken to mean as further explaining yajnam and yajnabhagam which precede it. This is possibly done because yajnam and yajnabhagam apparently mean the same thing. The commentary, therefore, seems to make further clear that yajnam stands for 'fire' and yajnabhagam for somabhagam Soma offering (given in the sacrifice)'. The compound expression agni somabhagau thus means 'fire and the Soma offering'. The compound in that case should have really been agnisomabha gau. But it appears as agni somabhagau on the analogy of the frequently occurring compound form agni somau. Then onha gaucurrin Madhu Vidya/138

Page #164
arnsala by M. A. MEHENDALE, Poona Among the rules to be observed during the Diksa ceremony we read the following in the Satapatha Brahmana (3.1.2.21): athainam salam prapadayati / sa dhenvai canatuhas ca nasniyat 1 dhenvanaduhau va idam sarvam bibhitah/ ... tad dhaitat sarvasyam iva yo dhenvanaduhayor asniyad antagatir ival ... tasmad dhenvanaduhayor nasniyat/ tad u hovaca yajnavalkyah-asnamy evaham amsalam ced bhavatiti "He (the Adhvaryu) then makes him enter the hall. Let him not eat the flesh) of either the cow or the ox; for the cow and the ox doubtless support everything here on earth. ... Hence, were one to eat (the flesh) of an ox or a cow, there would be, as it were, an eating of everything, or, as it were, a going on to the end (or, to destruction). ... let him therefore not eat (the flesh) of the cow and the ox. Nevertheless Yajnavalkya said, 'I, for one, eat it, provided that it is tender" (Eggeling). Thus Eggeling translates arsala as 'tender'. Similarly Jacobi', Encycl. of Religion and Ethics s.v. Cow (Hindul, remarks: "The Satapatha Brahmana, when prohibiting the eating of the flesh of the cow (iii 1,2,21) adds the interesting statement: Yajnavalkya said: "I, for one, eat it provided that it is tender". The meaning 'tender' has been assigned to arsala also by Hultzsch, Inscriptions of Asoka, p. 127, f.n. 8. According to BR., however, amsala means 'strong'. They derive the word from arsa 'shoulder' and refer to P. 5.2.98 where it is noticed that the suffix -la is added to amsa in the sense 'strong' (vatsamsabhyam kamabale). The Sanskrit lexicons also give the same meaning (balavan, bali) for amsala. In this meaning arsala is used also in the classical literature. Apte's dictionary cites Raghu 3.34 and 16.84, and Dasakumara. 169 (ed. by Godbole and Parab, Bombay 1883) Weber, Ind. Stud. 17.281 (1885), rendered amsala as 'feist', while Oldenberg, Die Weltanschauung der Brahmana-Texte, p. 209, foot-note, used just the word 'gut' to render amsala. Keith (Cambridge History of India I. 137-138) observes as follows in connection with the above passage from the Sat.Br. "... the great sage Yajnavalkya 1 For this and the other references in this paper I am indebted to L. ALSDORF's Bei trage zur Geschichte von Vegetarismus und Rinderverehrung in Indien (1961), pp. 55/56. ? Cf. Amara 2.6.44 balavan marsalo risalah. Hemacandra 448 arsalo bali. On Amara., the commentator says amso balam asyastity arhsalah. 3 On the Dasakumara. passage, the Comm. says ansalarurusah marsalapurusah. Madhu Vidya/139

Page #165
316 M. A. MEHENDALE ate meat of milch cows and oxen, provided that the flesh was amsala, a word of doubtful import, rendered either 'firm' or 'tender' by various authorities." Thus Keith does not wish to commit himself to any particular meaning of the word amsala. In the opinion of Alsdorf, op. cit. p. 56, foot-note 1, the meaning of amsala is unknown. He observes that the different translations of the word, referred to above, are no more than mere guesses. He adds: "dass es 'zart', 'saftig', 'gut', 'feist' 0. dgl. heisse, ist mir ausserst unwahrscheinlich; viel eher durfte es irgend eine rituell relevante Eigenschaft bezeichnen, die das Rindfleisch nach Yajnavalkya auch wahrend der Diksa unbedenklich macht...". One would readily agree with the above observation of Alsdorf. It seems further possible to define the 'rituell relevante Eigenschaft' which, according to Alsdorf, is denoted by amsala, and which, according to Yajnvalkya, renders beef unobjectionable even during the diksa. As will be shown below, the ritually relevant quality is only indirectly, and not directly, denoted by amsala. The word amsala occurs once again in the Satapatha Brahmana (3.8.4.5-6). There in the context of the cutting off of the hind-part (guda) of the sacrificial animal for the principal, the svistakrt, and the upayaja offerings, we read: sa ha tv eva pasum alabheta ya enam medham upanayet/ yadi krsah syad yad udaryasya medasah parisisyeta tad gude nyiset/ ... gudo vai pasuh / medo vai medhah / tad enam medham upanayati / yady u assalo bhavati svayam upeta eva tarhi medham bhavati "He alone, indeed, should slay the animal who may take to it sacrificial essence (medha). If it (i.e. the animal) be lean, let him put into the hind-part what may remain from the fat of the belly (after a portion for the ida offering has been cut off). ... The hind-part, verily, is the animal, the fat, verily, is the sacrificial essence. (When he puts the remainder of the fat from the belly into the hind-part of the lean animal), he carries in this way the sacrificial essence to it (i.e. to the animal). If, however, it (i.e. the animal) be arnsala, then it has itself approached the sacrificial essence". The context in which the word amsala occurs in the above passage leaves no room for doubt with regard to its meaning, amsala is contrasted with kisa 'lean'. Further we are informed that if the animal be lean, it does not itself contain enough sacrificial essence. Hence it is necessary to add some fat from its belly to its hind+ amsala occurs also in the Taitt. Br. 3.4.17.1 in the context of the Purusamedha. An arsala animal is bound for Agni (agnaye 'msalam). But this occurrence does not help in determining the meaning of the word. Sayana explains amsalam = balavantam, and Bhatta Bhaskara explains it as pinarsam. In the Katy. Sr.S. 7.2.24 we read amsalabhojanam va. The option indicated here by the word va has, as explained in the commentary of Karka, a reference to the view held by Yajnavalkya in regard to beef-eating referred to in the Sat.Br. above. Madhu Vidya/140

Page #166
amsala 317 part because fat is identical with sacrificial essence. This is as good as adding sacrificial essence to the animal since guda is identified with pasu. We are further told that if the animal is amsala, and not klsa, no such addition is necessary, for, the animal itself in that case is supposed to be full of sacrificial essence. amsala, therefore, clearly means 'full of, rich in, fat's. Sayana is nearer to the correct interpretation of the word when he explains arsalah as pustangah, although one may not be sure about his derivation of the word from arsa 'shoulder' which he indicates by the use of praviddhamsaho. But Eggeling is not so good when he translates the above line as, "But if it be tender (juicyl, then it has itself obtained the sacrificial essence." amsala thus primarily means 'full of fat', and secondarily denotes, through the Brahmanical identification of medas with medha, 'full of sacrificial essence'. Hence, when Yajnavalkya says that he has no objection to eating beef even during the diksa if it is full of fat (amsala), he does so presumably because in his opinion the fat in the beef makes it full of sacrificial essence. The presence of medha in the beef is the 'ritually relevant quality' which renders it unobjectionable for him. As rightly observed by Alsdorf it is not proper to misunderstand Yajnavalkya and make fun of him by imagining that he was only fond of 'tender' beef. A reference to the above peculiar Vajasaneya teaching regarding beef-eating is found also in the Dharmasutras. In the Vasistha Dharmasutra (14.45f.) we read: dhenvanaduhav apannadantas ca / bhaksyau tu dhenvanaduhau medhyau vajasaneyake vijnayate. Buhler translates the passage as "Not milch-cows, draught-oxen, and animals whose milk teeth have not dropped out'. It is declared in the Vajasaneyaka that the flesh of) milch-cows and oxen is fit for offerings." As already noted by Alsdorf (p. 60), Buhler has inadvertently left out the word bhaksyau in his translation. Alsdorf next refers to J. J. Meyer's view who does not agree with Buhler's translation of medhyau as 'fit for offerings'. In Meyer's view, the word means 'magisch-rituell rein'. In the Ap.Dh.Su. (1.5.17.30f.), where the same topic is dealt with, we find: dhenvanaduhor bhaksyam / medhyam anaduham iti Vajasaneyakam: "Fleisch von) Kuh und Ochs ist essbar. (Elbar ist) opferreines Ochsenfleisch laut Vajasaneyaka". In the opinion of Alsdorf we have to understand the word medhya 5 In the Taitt. Sam. we find the use of the word sthavimat. In 6.4.1 we read jaghanardhad ava dyati ... sthavimato 'va dyati "He cuts off from the hind portions, ... he cuts off from the thick side ..." (Keith). Also cf. Maitr. Sar. 3.10.4 where sthavimat is contrasted with animat. . As for the etymology, perhaps it would be better to extrapolate *arsa 'fat'. 7 ALSDORF, op. cit., p. 56, foot-note 1. * ALSDORF, op. cit., pp. 59-60. i.e. their flesh should not be eaten. Madhu Vidya/141

Page #167
318 M. A. MEHENDALE in the two passages above in a highly significant sense!, viz., as 'intended for sacrifice', i.e. consecrated for sacrifice, or, when freely expressed, 'coming from a sacrificial animal 1. He, therefore, translates the Vas.Dh.Su. passage as, "Verboten sind) Kuh und Ochs und Tiere, deren Milchzahne noch nicht ausgefallen sind. Esbar aber sind Kuh und Ochs, wenn es sich um Opfertiere handelt, nach der Lehre des Vajasaneyaka". One can agree with Alsdorf when he says that the word medhya in the above passages from the Dharmasutras is to be understood as having a highly significant sense; but, in the light of the view held by Yajnavalkya (Vajasaneya), as expressed in the Satapatha Brahmana, it seems that the word medhya has to be interpreted differently. The secondary derivative medhya, with the suffix ya, means, not 'intended for sacrifice', but 'full of sacrificial essence (medha)'. As already mentioned in the Sat.Br. passage cited above, medha and medas are identical. Hence the word medhya can secondarily denote 'full of fat'12, i.e. the same thing as amsala. The Vasistha Dharmasutra 14.45 f. should therefore be translated as: "'The cow and the ox and the animals whose milk teeth have not yet fallen fare forbidden for eating). But it is learnt in the Vajasaneyaka that cow and ox which are full of sacrificial essence can be eaten" 18. The Apastamba Dharmasutra 1.17.30 f. should be translated as: "The flesh) of cow and ox can be eaten. The flesh of an ox, which is full of sacrificial essence (can be eaten) according to Vajasaneyaka." Understood the way suggested above we may draw the following inferences regarding the view held by Yajnavalkya and the tradition established by him with regard to beef-eating. At the time of the Satapatha Brahmana, when beef-eating was not prohibited generally, a restriction was put on its eating by the sacrificer during the diksa. But Yajnavalkya held the view that even during the diksa a sacrificer might eat beef if it was full of fat fi.e. also full of sacrificial essence). Later, in the period of the Dharmasutras, when beef-eating as a general practice was being prohibited, those who followed the teaching of Yajnavalkya held the view that beef might be eaten if it was full of sacrificial essence (by virtue of its being full of fat). 10 'in einem sehr pragnanten Sinne'. 11 "zum Opfer bestimmt", also zum Opfer geweiht oder freier ausgedruckt "von einem Opfertier stammend". 12 It may be noted that the Kosakalpataru Vol. 2.7253 (Deccan College edition, 1966) gives medura, which may mean "full of fat", as one of the meanings of the word medhya (medhyam Sucau medura asrame ca). 13 As noted above, they are full of sacrificial essence, if they are full of fat. 14 And by implication, also of a cow. Madhu Vidya/142

Page #168
Notes and News PROF. THIEME'S ETYMOLOGY OF SKT. asi AND ITS BEARING ON THE IRON-AGE IN INDIA The word asi (m.) is attested in the Sanskrit literature since the Rovecha. Its meaning, there, is knife' used for cutting the slaughtered animals. As 'sword', a weapon used in battles, asi appears since the Atharvaveda. What was this gvedic asi 'knife' made of ? Since the word ayas accurs in the Rgveda and since it is usually taken to mean copper (or bronze) one inference could be that the gvedic asi was made of copper. But this inference would prove wrong if Prof. Thieme's suggestion (Thieme. 1958) regarding the etymology of the word asi is found to be correct.'. We have in Sanskrit correspondences like harita 'yellow': harit hari 'yellow' rohita 'red': rohit, rohi 'red' Similarly in Greek alphito - "barley meal': alphit 'barley meal' (lit. 'the white one') Such correspondences make the following correspondence for Sanskrit a possibility : Skt. asita 'black': *asit, *asi 'black' This means that for Sanskrit we assume an adjective * asi which is not attested. In Greek we have a word asis f. 'river mud'. If we assume an unattested existence of a Sanskrit a djective *asi 'black' it is possible for us to compare Sk. *asi (adj.) 'black': Gk, asis - (Fem.) 'river mud'. Phonetically this correspondence becomes likely only it we assume that the two words Skt. *asi and Gk. asis are derived from IE *i which, as an adjective, would have the meaning 'black'. We have to assume that in Greek the Indo-European adjective 'black' was nominalized to mean 'river mud'. In Latin we have the word 'ensis' m. sword'. Earlier it was not possible to relate Sk. asi (iron) sword, and Lat. ensis (iron) sword' because, although there Madhu Vidya/143

Page #169
was phonetic resemblance between the two, the meaning 'iron sword' could not be attributed to the reconstructed IE word as iron was not known in that period?. But if Latin ensis 'sword'. Jike Greek asis 'river mud', is considered a nominalization of the IE adj. 18 'black'. ie. 'the black one, the black iron sword', then the two words from Sanskrit and Latin can be looked upon as cognates. And not only these two, also Gk. asis 'river mud' can be considered a cognate with them as all three can now be treated as independent nominalizations of the IE adj. r 'black'. If the above explanation of Skt. asi 'knife' as a later development of a nominalization the black one' of a potential adjective *asi 'black' is correct then it implies that at the time (late Rgvedic period) when asi knife' is attested it must have been made of iron. Otherwise the nominalization of the adjective meaning 'black' cannot be explained. This means that on linguistic evidence it is possible for us to say that iron was known in the late gvedic period. All these facts about the etymology of the Skt. noun asi have already been stated by Prof. Thieme in the review referred to above (1964). The purpose of the present paper is only to bring these facts to the notice of the archaeologists and also to make explicit the implication of this etymology for the iron-age in India. Prof. Thiene 1958 Review of 'Dictionnaire etymologique du protoindoeuropeen' Albert Carnoy, Language 34 pp. 510-515. 1964 "The Comparative Method for Reconstruction in Linguistics' Language in Culture and Society (ed.) Dell Hymes, pp. 593-594. M. A. MEHENDALE Prof. Thieme's suggestion seems to have its starting point in W. Schulze's etymological hypothesis regarding Greek asis f. 'river mud' (Kleine Schritten 116 f.). The paper is inaccessible to me. The words for 'iron' in the languages belonging to the IE family are phonetically quite different from one another: Gk. sideros, Lat. ferrum, old Slavic zelezo. old irish zaru. Skt. krsnnayasa. Madhu Vidya/144

Page #170
PRAKAMODYA - ITS RELEVANCE TO THE AKHYANA THEORY M. A. MEHENDALE The word prakamodya is attested only four times in the Vedic literature. Its occurrences in the Vaj. Sam. (30.9) and the Taitt. Br. (3.4.6) are not useful for determining the meaning of the word. It occurs twice in the Sat. Br. (3.2.4.16 and 3.5.3.11) but since the two sentences in which it occurs are identical, we are, in effect, left with only one attestation on the basis of which we have to interpret the word. The Adhvaryu priest repeats the mantra cid asi manasi dhir asi daksinasi' while addressing the cow with which Soma is to be purchased. While commenting on the words dhir asi in the above mantra the Sat. Br. (3.2.4.16) says: dhiya dhiya hy etaya manusya jujyusanti anukteneva prakamodyeneva gathabhir iva tasmad dha dhir asiti "all men seek their sustenance according to their respective knowledge; (some) by the recitation of the mantra-s, (some) by prakamodya, (some) by the singing of the gatha-s. That is why he (i.e. the Adhvaryu) says 'dhir asi". etaya The meanings of the words anukta and gatha in the above passage are clear. They refer respectively to the Vedic mantra-s which the priests learn and recite in the Vedic ritual and the gatha-s which are sung on special occasions. Both these are fixed and are traditionally handed down among the classes which make use of them. The meaning of the word prakamodya alone remains to be determined. Sayana explains the word prakamodya as icchaya laukikabhasanam i.e. everyday conversations done as desired. Bothlingk-Roth give the meaning "Geschwatzigkeit" and Monier-Williams gives 'talking to the heart's content, talkativeness'. Eggeling in his translation of the Sat. Br. renders it as 'readiness of speech'. None of these renderings seems to fit into the context in which the world is used. The context is of earning one's livelihood (jujyusanti) which is done with the recitation of the mantra-s on the one hand and the singing the gatha-s on the other. The Rgveda contains some hymns which are known as the samvada sukta-s (dialogue hymns). It is well known that H. Oldenberg in 1883 put forward his akhyana theory with regard to these hymns. Recently Prof. L. Alsdorf has sought to uphold this theory by furnishing some additional evidence from the Jaina literature.3 Madhu Vidya/145

Page #171
According to the theory of Oldenberg, the dialogue hymns are the remnants of a literary form which once existed and which he calls as akhyana. These akhyana-s contained both verses and prose portions. Whereas the verses were 'fixed', the prose portions were not and were supplied on each occasion by the narrator of the akhyana as desired by him. These prose portions were consequently lost with the result that in the samvada hymns of today we have only the verses and not the prose portions. Thus, according to the theory of Oldenberg. the old akhyana-s consisted of the Vedic sc-s, preserved in the samvada hymns, and the prose portions which were supplied by the narrator. That some of the akhyana-s contained, besides the sc-s and the prose portions, also the gatha-s becomes clear from the following description of the Sunahsepa story in the Ait. Br. 33.6 ; tad etat pararksatagatham saunahsepam akhyanam and of the Trita story in the Nirukta 4.6: tatra brahmetihasamisram rnmisram gathamisram bhavati. It is, therefore, possible to suggest that the word prakamodya which occurs in the Sat. Br. by the side of anukta and gatha and which refers to a means of livelihood, means to be spoken as desired, 'It would refer to the prose passages in the akhyana which the narrator added each time as desired by him. Eggeling seems to have come close to this interpretation of prakamo dya when in foot note 8 to his translation he adds : "Prakamodya, rather either 'fondness for talk'or'effusive speech.' It seems to refer to stoty-tellers (amusing speech)." If this interpretation of the word prakamo dya is correct it will be an additional argument in support of Oldenberg's akhyana theory. Vaj. Sam. 4.19. Cf. for example Sat. Br. 13.1.5.1. ff., 13.4.2.8. They also formed part of the akhyana. The Akhyana Theory Reconsidered, Journal of the Oriental Institute, Baroda, Vol. 13, pp. 195-207, 1963-64. Madhu Vidya/146

Page #172
2. ATIHAYAT (?) By M. A. MEHENDALE Sanskrit Dictionary Deptt., Deccan College, Poona-6 In the Vedic Word-Concordance, Vol. II (Brahmanas), appears an entry atihayat with a question mark. Apparently the form has been looked upon, doubtfully, as a present participle from att ha (jihite). The form in question occurs twice in the Jaimini ya-Brahmana 3. 205 and 3. 208. The Jaim. Br. 3. 205 reads as : kavir girbhih kayyena kavis san somah pavitram aty eti rebhann' ity atihdyan. The Jaim. Br. 3. 208 reads as : supravir astu sa ksayah pra nu yaman sudanavah ye no amho 'ti pi prati ity atihayan. In these two passages, the Brahmana portion has not been correctly read. What appears as atiha yan should really be read as ati hayan - ati ha ayan "they have, indeed, gone beyond, i.e., they have indeed, completed (the performance of)". The Brahmana remark is intended to justify the choice of the mantras which contain the words atyeti 'goes over, or goes beyond' (Jaim. Br. 3. 205) and atipi prati 'they carry over or beyond' (Jaim. Br. 3. 208). The Brahmana passages say that these particular mantras containing the words atyeti and atipi prati are used because "they (the priests) have, indeed, gone beyond, i.e., they have completed (the performance of the ritual pertaining to the pr sthya sadaha)". In the second passage (Jaim. Br. 3. 208) the author of the Brabmana could have said ati ha api paruh, instead of ati ha ayan, because the mantra contains the verbal form ati pi prati (from the root pr) and not at yeti (from the root i). But the author chose to repeat 1. SV 2. 525 ; RV IX. 96. 17 (kuyyena). 2. SV 2.702; RV VII. 66.5. Madhu Vidya/147

Page #173
2 [Vol. XVIII ati ha ayan from 3. 205, apparently because the justification for the use of the mantra was the same, viz., "the priests have gone beyond (the prathya sadaha)". M. A. MERENDALE The mantras in question are used in the Bahispavamanastotra and the Ajya-stotra respectively of the second of the three Chandoma days. In a Dvadasaha rite, lasting for twelve days, the first and the twelfth days are respectively the Prayantya and the Udayaniya days. The remaining ten days are divided into three parts, viz., the prsthya sadaha (lasting six days), the chandomas (lasting three days), and the avivakya (the tenth day)." In the Altareya-Brahmana it is said about the first of the above three parts as prsthyam sajaham upayanti "(the priests) approach the Prsthya sadaha." Similarly while speaking about the last day, the Ait. Br. says ye dasamam ahar agacchanti "(the priests) who arrive at the tenth day." It is interesting to note the use of the forms of the verb meaning 'to go' in these two passages. Since the three Chandoma days occur between the prsthya sadaha and the tenth day, the Jaim. Br. expression ati ha ayan obviously refers to the priests who, after approaching (upayanti) the prsthya sadaha, have gone beyond it in order to perform the rites pertaining to the Chandoma days before finally arriving (agacchanti) at the tenth day. The Jaim. Br. passage contains the form ayan in the imperfect, and not yanti or gacchanti in the present, because what is intended to be conveyed is what the priests have already accomplished. Since they have gone beyond, i.e., completed the ritual pertaining to the prsthya sadaha, the mantras containing the forms at yeti the 'goes beyond' or atipi prati 'they carry beyond' are considered fit (rupasamyddha) for the stomas of the first of the three Chandoma days. The Jaim. Br. remark att ha ayan is no doubt cryptic. But the use of the forms upayanti and agacchantt in the Ait. Br. and of yanti in the Jaim. Br. itself (3. 206, 208) makes it fairly certain that the subject of ayan is the 'priests'. And the context where the cryptic remark occurs makes it almost equally fairly certain that what the 3. Sayana, while introducing the Ait. Br. 24. 3 (p. 631), says: dvadasahe prayani yodayaniyarupe adyante ve ahani tayor madhye dasaratro 'sti tasmime sa trayo bhagah/prathyah sadaha eko bhagah/chandomanamakas trayo 'harvidesa dvitiyo bhogah| datamam ahas trtiyo bhagah | Madhu Vidya/148

Page #174
1980) 1. 2. ATIHAYAT (9) priests have gone beyond, i.e., what they have already completed is the prsthya sadaha. It may, however, bo pointed out that in the Jaim. Br. 3.173 we are told that Prajapati created the three Chandoma days in order to connect the prsthya sadaha with the following four days (viz., the three Chandoma days and the tenth day).* In 3.174 we read Chandamsy evaitac chandoman abhy atinayanty uttarasya tryahasya samparanaya "thus the metres themselves carry (the sacrifice) to the Chandoma days in order (further) to go beyond (the ritual of) the three days." In view of this passage it is possible to understand 'metres as the subject of ati dyan for they are supposed to have gone beyond the prsthya sadaha. It may also be observed in the end that we find in the Brahmanas the use of the preterite forms when there is a reference to what the gods or the sages did in the past. Hence there is also the possibility of the gods' or the 'sages' being the subject of ati ayan. 4. In the Jaim. Br. 3. 174 the progress of the ritual form the prothya sadaha to the seventh day (i.e., the first of the three Chandoma days) is called atikranti. We may also compare atikranto va etarhi yajio bhavati saptame ahan (Jain, Br. 3. 182). Madhu Vidya/149

Page #175
M. A. MEHENDALE DURATIVYATHA xcessive the Upasain the ab In the Kathakasamhita 25.1 we read isum va eta devas samaskurvan yad upasado 'gnim srngam somam salyar visnur tejanam / te 'bruvan yo va ojisthas sa imam visrjatu / ... na prayaja bhavanti nanuyajah / puro va ete yajnasya yat prayajanuyaja yat prayajanuyajan kuryat puro yajnasya kuryad durativyathar syad atiksnam isur kuryat! << What are Upasads that these the gods indeed made into an arrow, (they made) Agni the tip, Soma the socket, Visnu the shaft. They said, "who among you is the strongest, let him release this (arrow)"... There are no Prayaja offerings, nor the Anuyajas. What are the Prayaja and the Anuyaja offerings they are, indeed, the forts of the sacrifice. If he were to offer the Prayaja and the Anuyaja offerings he would erect the forts for the sacrifice. (In that case his act of performing the Upasad) would not inflict excessive pain (on the enemies), (for) he would make the arrow in the form of the Upasads) blunt >>. What is intended to be conveyed in the above passage seems to be that if the sacrificer were to offer the Prayaja and the Anuyaja offerings that would be tantamount to his erecting forts around the sacrifice. Then the arrow, in the form of the Upasad isti, will first strike against these forts, would become blunt, and consequently inflict very little pain (durativyatham) on his enemies. It is, perhaps, possible to understand the passage this way. However, there appears to be one difficulty. If the above sense was intended, the end portion of the passage would have read atiksnam isum kuryad durativyatham syat, and not as the text actually stands. It is therefore tempting to suggest a small emendation and read durativyadham << difficult to pierce through >> in place of durativyatham. The emendation appears appropriate in the context of the shooting of an arrow. If the sacrificer were to offer the Prayaja and the Anuyaja 1. Also Kapisthalakatha, 38.4. Madhu Vidya/150

Page #176
254 M.A. Mehendale would be difficult to pierce thence ineffecti offerings he would be erecting forts around his sacrifice. The Upasadarrow will then find it difficult to pierce through this fort. The sacrificer would also make the arrow blunt, and hence ineffective, by making it first strike against the walls of the fort. Understood this way, the passage yields better sense. In both the interpretations, durativyatham and durativyadham are considered adjectives. The former would qualify a noun like karman <> supplied ?, while the latter would qualify the noun varman <>, which suggests itself from the Aitareya Brahmana passage discussed below. In the Aitareya Brahmana 1.26 (=4.9) we read: devayarma va etad yad prayajas canuyajas caprayajam ananuyaja bhavatisvai sarsitya apratisaraya. This passage is somewhat difficult to understand. Keith translates: << The fore-offerings and the after-offerings are divine armour; (this rite) is without fore-offerings and after-offerings, to sharpen the arrow and to prevent rending >>. Haug's translation is much the same except that he renders apratisaraya as > The above translations are not helpful in understanding the purpose of not performing the fore- and the after-offerings which is expressed with the words samsityai and apratisaraya. Sayana explains the above pasasge in the following way: the Upasad isti should be performed without the Prayaja and the Anuyaja offerings because these offerings are like an armour and in the present case there is no use of an armour. When the arrow in the form of the Upasads is shot against the enemies, due to the sharpness of the arrow all of them would get killed by a single shot and there would be no one left amnog them to strike back. Hence an armour becomes unnecessary. But if the sacrificer were to offer the Prayaja and the Anuyaja offerings it would mean that he was in doubt regarding the sharpness of the arrow and hence, apprehending a counter-attack by the enemy, was providing for protection in the form of the armour. That would not be proper. Hence no Prayaja and Anuyaja offerings are offered in order to demonstrate faith in the sharpness (sarsityai) of the Upasad-arrow and to set aside the apprehension of injury to one's side at the hands of the enemies (apratisaraya)". This is no doubt a good attempt to explain the significance of the two datives samsityai and apratisaraya. But the explanation is, in part, strained. In view of the fact that in the Kathakasarhita the Prayaja and the Anuyaja offerings are likened to puran <>, it would be better to inetrpret the word devavarma in the Ait.Br. passage, not as << divine armour>> (Keith) or > (Haug), but as an enclosure created by the gods, a divine defence >>. The Ait.Br. passage 2. Cf. the way Sayana supplies U' pasadakhyar karma while interpreting the Ait. Br. passage discussed below, 3. Sadgurusisya merely says isval sarasya sarsityai taiksnyaya apratisaraya apratighataya. Madhu Vidya/151

Page #177
Durativyatha 255 the sharpnfered. And once the back with an arrownd the Anuyaja would then mean that the Prayaja and the Anuyaja offerings are not offered because that would be like creating a divine defensive wall around the sacrifice. In that case the Upasad-arrow, when shot, would strike against this wall and would become blunt. To avoid this and to ensure the sharpness (sarsityai) of the arrow, the fore- and the afterofferings are not offered. And once the sharpness of the arrow is ensured, there will be no enemy left to strike back with an arrow (apratisaraya). In the Taittiriya Samhita also (2.6.1.5) the Prayaja and the Anuyaja offerings are said to constitute the varma of the sacrifice and the sacrificer. The fact that in this passage it is further stated that it is on this account that a varatha << a defensive enclosure >> is erected taller in the front and shorter at the back shows that by varma is meant, not <>, but <>. The whole passage in the Taitt. Sam. reads as: yat prayajanuyaja ijyante varmaiva tad yajnaya kriyate varma yajamanaya bhratrvyabhibhutyai/ tasmad varutham purastad varsiyah pascad dhrasiyah. account that a var at the back shows that by The whole passage in Madhu Vidya/152

Page #178
Aduvas (?) In a hymn addressed to Agni in the Rgveda occurs ma tva vayam sahasavann avira ma psavah pari sadama maduvah (7.4.6). The author of the Padapatha has analysed ma duvah as ma aduvah. Geldner, accordingly, translates: "Nicht wollen wir dich, du Machtiger, ohne Sohne, ohne Vieh und ohne Achtung umsitzen." aduvah is thus looked upon as a nom. pl. mas. form coordinated with avirah and apsavah. One arrives at the masculine plural form aduvah either from the base *adu 'Gabenlos' (Grassmann and Alt. Gr. III. 81, 197), or as a haplology form from aduvasah (Alt. Gr. III. 80). Thieme' is against considering aduvah as a haplological masculine plural form. In his opinion aduvah is nom. sg.neut. and the nominal sentence (Nominalsatz) ma duvah contains a most abridged reference to the preceding two requests, viz., not to leave the worshippers without heroes (avirah) and not to leave them without vital force (apsavah). He, accordingly, translates ma duvah as "Nicht (sei] Nichtfreundschaft? [von deiner Seite]" Renou, however, agrees with Geldner in treating a duvah as a mas. plural form; but he agrees with Thieme in relating duvas with the deity. He therefore explains aduvah as 'denues de la faveur (divine).' A simpler solution to the problem posed by aduvah can be found if we go against the Padapatha and interpret the Samhita text maduvahas standing for ma duvah. In that case duvah can be considered as aorist 2nd pers sg. of du 'to burn, to torment, to afflict,' or of * du 'to go afar, to go away from.' The words ma duvah would give expression to the worshipper's request to Agni either not to torment him (by withholding from him (without granting his requests.) For the connection of the verbal form duvah with du 'to torment, etc.' we may compare AV9.4.18 nainam dunvanty agnayahor AV 5.22.2 agnirivabhidunvan; and for the connection of duvah with *du 'to go away' we may compare Agni's going away at a distance and hiding in waters RV, 10.51, especially tasmad bhiya varuna duram ayam (6), or yad dure sann ihabhavah RV.3.9.2. For requests to Agni with the use of ma in this very hymn (RV 7.4) cf. stanza 4 sa ma no atra juhurah sahasvah ("Oh mighty one, that thou mayest not lead us astray."), and stanza 7 ma patho vi duksah ("Do not seek to spoil the ways."). DECCAN COLLEGE M.A. MEHENDALE W. Schubring Festschrift (1951) p. 8 and fn. 1. Incidentally, H. Gunther is not in favor of assigning the meaning 'Vertraulichkeit, 'Freundschaft' to duvas, KZ 69.240 (1951). EVP 13.55, 141 (1964). Madhu Vidya/153

Page #179
atyanUktam TWO VEDIC NOTES M. A. MEHENDALE (1) atyanUktam The following mantra is found in the Taittiriya Brahmana 3.7.11.1: atyAzrAvitam vaSaTkRtam atyanUktaM ca yajJe The meaning of the word atyanaktam is not immediately clear. Caland looks upon it as an adjective,qualifying and translates it as 'uttered too loud' (?) (der alzuweit gesagte Vasatruf)'. Dumount has only followed Caland while translating it as 'the exclamation vasat (a)(if it is) excessively loud"." The above rendering does not seem proper. Caland considers that in the above mantra only two acts, done improperly at the sacrifice, are referred to. But this is not correct. It refers to four such acts, as has been made clear by Sayana by using ufour times in his commentary on the Taittiriya Brahmana. Moreover, Caland has neglected the prefix anu in his translation of atyanUktam. In the mantra atyanUktam comes after vaSaTkRtam In the ritual in the case of certain offerings - comes after aan. Therefore, it may appear at first sight that the prefix anu in atyanUktam is idicative of this anuvaSaTkAra. In that case atyanUktam might mean 'improperly uttered agar. But even this is not acceptable. In the first instance the mantra occurs among those formulas which accompany the expiation offerings offered to atone for the mistakes committed in the afgefare. And there is no offering with gauza prescribed in this isti. Hence the possibility of an arazor being improperly pronounced does not arise and, therefore no expiation is called for. Secondly, anUktam is derived from anu vac and the prefix anu in anuvac or anu-brU does not mean 'after' but 'agreeable to', 'suited for' (3). When the arag instructs the Hotr (a) priest with the words a fra 3 (Satapatha Brahmana 1.3.5.2) what he wants the priest to do is to recite such mantras as are suitable for the occasion. It is only when the priest recites the mantras are suitable for the occasion, that the sacrificial act becomes perfect in form. This has often been told in the Brahmanas as vai yajJasya samRddhaM yad rUpasamRddhaM yat karma kriyamANam Rg abhivadati (Aitareya Brahmana, 1.4.etc. Sayaa te ca (Rcau) karmAnukUlyena rUpeNa samRddhe ) . Whatver mantras the priest recites in response to the fru af whether they accompany the churning of the fire or bringing forward of the Soma, or whether they are Madhu Vidya/154

Page #180
intended to invite the deity to receive the offering, can be called 3. And the prefix arg in argue is indicative of "being suitable for" and not "after". Hence Sea can only mean recitation (of a mantra in response to 3 which is not suited) (for the occasion)". This meaning has already been given by Bhatta Bhaskara while commenting on the Taittiriya Brahmana passage. He explains as asamIcInAnuvacanam improper or unsuited anuvacana Probably the same meaning is intended by Sayana when he glosses atyanUktam (as atikrAntAnuvacanam ). Both Caland and Dumont have neglected this correct explanation given by the commentators. (2) atyudgRhyaH (?) In the Maitrayani Samhita 1.6.6, we read: esi a enfuerit karjadaghnamudagRhNIyAt taM yA uttarasmioka AsaMstA abhisamAvartanta taM karNadanaM nAtyudgRhyo yatkarNadaghnamatyudgRhNIyAd, yajamAno varSiSThaH pazUnAM yajamAnamupariSTAdagnirabhyavadahet / The context is that of carrying the fire from the Garhapatya to the altar of the Ahavaniya at the time of the setting up of the fires (3). At that time the sacrificer should lift the sticks gradually from the ankle-high position to the kneehigh, the navelhigh, the shoulder-high, and finally to the ear-high position. Then, we are told, he should not raise the fire-sticks any further beyond the ears. If he were to do that fire would burn down from above towards the sacrificer. In this connection the Maitrayani Samhita, cited above, reads afgeven (a). The Vedic concordance of the Vishveshvarananda Vedic Research Institute, Vol. I. p. 108 records this from as egg i.e. nom. sg. of the future passive participle (4). But this leads to difficulty, since this nominative form is supposed to qualify (agnim ) which is in the accusative. This requires the text to be emended to sa karNadaghnaH nAtyudgRhyaH It is, therefore, better to consider age as standing for 3, the former word being the gerund and the latter the enclitic particle used for emphasis. In that case we have to supply smurfar after orgy and translate: "him (the fire), which is ear high, not raising beyond indeed, (he puts down on the altar)." For the above interpretation of age and for the supplying of unfar we have the support of the Manava-Srautasutra 1.5.4.13, which reads: fai amegye karNadaghnaM nAtyudgRhya Suma. The interpretation suggested here does not force us to amend the text of the Samhita. Moreover the gerund arge is regular while the potential passive form 3rga: would be irregular. MadhuVidya/155 .....

Page #181
NOTES: 1. Das Srauta-Sutra des Apastamba 3.11.2. Proceedings of The American Philologica Society, 105.35.(11961). asmin yajJe yaddhavirAzrAvitaM samIcInAzrAvaNayuktaM yaccAtyAzrAvitam atiriktA zrAvaNayuktaM yacca vaSaTkRtaM samyagvaSaTkAreNa dattaM yaccAtyanUktam atikrAntAnuvacanayuktaM tat sarvam / But Sayana does not say why the acts done with correct 3 Teiraut and correct quecht would require expiation. We have to suppose that the author had meant that they were improperly performed. Sayanaexplains anuvAcA by anukUlayA vAcA Aitareya Brahmana3.5.p.68. Also of ust , Aitareya Brahmana 3.5,etc. Satapatha Brahmana, 1.3.5.13. Note: The accent marks in this paper have not been shown for the convenience of the Press. Their absence is very much regretted. Madhu Vidya/156

Page #182
BRIEF COMMUNICATION ON THE BRHADDEVATA 2.102-103 While writing on 'Nirukta and Astadhyayi: Their Shared Presuppositions' (IIJ 23, 1-14, 1981), Johannes Bronkhorst has made certain statements which require careful consideration. For the present I wish to confine myself to his following observation. "On many occasions the Nirukta gives several etymologies of one single word in one single meaning. There is reason to believe that all such etymologies were considered simultaneously correct". (pp. 6-7) In support of his statement Bronkhorst first refers to Saunaka's BIhaddevata 2.102--03. The two stanzas are difficult, but it is also difficult to admit Bronkhorst's statement to the effect that "their author considered it possible that a word has simultaneously several etymologies". (p. 7). The stanzas read: yavatam eva dhatunam lingam rudhiga tam bhavet/arthas' capy abhidheyah syat tavadbhir gunavigrahah// dhatupasargavayavagunasabdam dvidhatujam/ bahvekadhatujam vapi padam nirvacyalaksanam// The stanzas are perhaps be rendered as: "As many roots as have their signs present in the conventional form (of a word), and the meaning(s) (of which roots) is meant to be expressed (by those signs in the word), the analysis of the derived form (viz. the word) is to be done with all those (roots). The word, whose analytical formation is to be explained, is a derived form consisting of parts made up of root(s) and prepositions(s), (and) may arise from two roots, many roots, or one root". The author of the above stanzas seems rather to say that a certain word may contain traces of one, two, or more roots (with or without prefixes) and is, accordingly, to be derived from as many roots. He wants to tell one interested in derivations that he need not always feel himself restricted to only one root. The word to be derived may, on occasions, show signs of two or more roots and in such a case as many roots become simultaneously valid for the derivation of that word. It is therefore quite likely that the author of the stanzas has made this suggestion keeping in mind a person like Sakapuni? who perceives in the word agni signs (lingam) of three roots, i-, anj- or dah-, and ni. in the constituents a, &, and ni respectively and accordingly uses the three roots simultaneously for deriving it. The BIhaddevata stanzas thus apply to derivations of words like agni from more than one root. They do not seem to apply, as assumed by Bronkhorst, to the derivations of words like nighantu of which three possible derivations are communicated to us by Yaska in the opening section of the Nirukta. The three Indo-Iranian Journal 29 (1986) 117-120. (c) 1986 by D. Reidel Publishing Company. Madhu Vidya/157

Page #183
118 BRIEF COMMUNICATION derivations are from either gam-, han-, or hr-. The authors of these derivations saw, each at a time, the sign of only one of the above three roots in the word nighantu and not of all of them together. Hence only one derivation, at a time, can be considered correct. The author of the Brhad. stanzas would look upon agni, as derived by Sakapuni, as a bahudhatujam padam. It is extremely unlikely that he would consider nighantu the same way. Next, in support of his view Bronkhorst appeals to Durga. It seems to me that he has perhaps misjudged Durga's commentary on this section (Nir. 1.1). It is true that Durga cites the Brhad, stanza 2.102 and understands it, in my opinion incorrectly, the way Bronkhorst does. That is, Durga too considers that the stanza applies to multiple derivations of words like nighantu. But although, he does this Durga's final conclusion is not that all the roots suggested in the multiple derivations are to be used simultaneously for the derivation of that word and hence the different derivations become simultaneously correct. What Durga says is this: Since all nouns are derived from verbs a given noun will express the action conveyed by the verb from which it is sought to the derived. Now in a case like nighantu, the derivation of which is not clear, one may feel that the word is expressive of different actions and consequently as many roots will be available for derivation. In the word. nighantu, different authors imagined different actions, viz. nigamana, samahanana, and samaharana, and hence three derivations became likely viz. from gam-, han-, or har-. Since there is no way by which one may accept only one of these three roots and refuse the others, one should make a collection of all such roots (as has been done for example, by Yaska), and suggest derivations from them. Perhaps there is some ambiguity in Durga's formulation: yavanto dhatavah svalingam radhigatam darsayanti tavatah sangrhya sa rudhisabdo nirvacyah. This might create an impression that according to Durga the conventional form of the word is to be derived from all the roots together (sangrhya), and not individually from one of them at a time. But the doubt gets cleared up from what Durga says further. He says: The three roots gam-, han-, and har- compete with one another and tell (the etymologist) 'derive this word following me, derive it from me'. It is important to note that each root speaks for itself. All of them do not say together: "derive this word from us". Durga continues: The root gam- feels that the sound gha in the word nighantu is a modification of its ga sound, while the roots han- and harfeel that it is the modification of their ha sound. In such a case if one is true, the other, at the same time, cannot be. If the gh sound has arisen from ga, then it cannot at the same time arise from ha. Earlier also Durga says ahananad eva, na nigamanad ity abhiprayah...evam esa nighantuiabdo gamer vaikopasargad dhantiharatibhyam vd dvyupasargabhyam niruktah. This should leave no doubt about Durga's view. He considers only one root at a time valid for the derivation of nighantu. If there is a competition (ahampurvika) between the three roots, only one of them has a chance to succeed, and not all of them together. Madhu Vidya/158

Page #184
BRIEF COMMUNICATION 119 There is a story in the Mahabharata (3.80.124-127) which illustrates this point. This story of a holy place called Rudrakoti reads in van Buitenen's translation as follows: "(At Rudrakoti) once a crore of seers came diligently and joyfully, desirous of seeing the God. They approached, saying, "I shall be the first to see the bull-bannered God", and "I shall be the first to see him", as the story goes, O king. To prevent these seers of cultivated souls from becoming angry, the Lord of Yoga resorted to his yoga and created a crore of Rudras, one before each of the seers, so that each thought that he had seen him first". The lesson of the story is clear. If the God had not created a crore of Rudras, only one of the crore of seers gathered at Rudrakoti would have been successful in the competition and seen the God first - not all of them simultaneously. Bronkhorst admits that there is no explicit statement in the Nirukta) to the effect that several etymologies of one word in one meaning can be simultaneously correct" (p. 7). Yet he avers that Yaska's procedure seems to indicate the same. It seems to me that Yaska's procedure indicates just the opposite. His use of such expressions as va, api va, yad va while giving alternative etymologies for a given word - and this is also true in the case of the word nighantu - clearly shows that in his opinion these are possible alternatives and it is presumed that a new suggestion cancels the ones made previously. It is not clear why Bronkhorst ascribes the view regarding the simultaneous correctness of different derivations of a word to Yaska because in the case of the word anna, although its derivation from ad- is clear, Yaska has chosen to give an additional derivation from a-nam- (Nir. 3.9). It is true it is not easy to say why Yaska does this. It is possible that he did this under the influence of the Brahmanas and the Upanisads where, occasionally, words of known derivation, have been derived in an 'unorthodox' way, witness, for example, the derivation of yajna in the Sat. Br. 3.9.4.23, Ch. Up. 4.16.1; 8.5.1, or of yajus in the Bt. Up. 5.13.2. Or one may say that Yaska did it because to him the derivation of anna from ad- was not that obvious since not all the roots which end ind yield similar formations, and even in the case of roots like bhid-, nud- etc. alternative forms are available. Be that as it may, a case like this cannot be an argument to in fer that Yaska believed in the simultaneous correctness of alternative derivations. There seems to be only one way of understanding the simultaneous correctness of multiple derivations as of nighantu. We have to suppose that at one time the samamnaya was called nigantu, samahantu and samahartu and that in course of time all these designations assumed one identical form nighantu. Only this way the three derivations become simultaneously correct. Bronkhorst rightly discards this assumption, and yet says that Yaska believed in the simultaneous correctness of the alternative derivations. I must admit that I have not followed Bronkhorst when he says: "But if indeed the different derivations of one word in one meaning Madhu Vidya/159

Page #185
120 BRIEF COMMUNICATION were meant to be simultaneously valid, we must again admit that Yaska's etymologies were not intended as descriptions of the histories of the words concerned" (p. 7). M. A. MEHENDALE Bhandarkar Oriental Research Institute, Pune 411 004 India NOTES As pointed out by the author, this is also the view of J. Gonda, "The Etymologies in the ancient Indian Brahmanas" Lingua S.43. 2 Nir. 7.14. 3 of the two, either aftj is valid or dah- at a time, and not the two together. 4 Gonda says that Durga on Nir. 1.14. has "attempted to show that the derivation of a single word from a plurality of roots was already in vogue in the Brahmanas" (Lingua 5.72). But Durga's expression anekadhatujani there refers to the Brahmanical derivation of the three sounds in the word hrdaya from three roots hr, das, and 1. It is thus a case similar to the derivation of agni referred to above, and not of nighantu. Madhu Vidya/160

Page #186
ON MITRADRUH AND MITRADROHA BY M. A. MEHENDALE Io the Avesta the expression mierom druj- means 'to belio & contract, to break a contract' and the adjectival compound mibro.drug means ono wbo breaks the contract'. Formally, we have an exact parallel mitradruh in the Vedic literature; and this and the other compound formations like mitradroha are met with io the later literature. Wbat do these expressions mean in Sanskrit ? It is generally believed that since in the classical Sanskrit literatura mitra in the sense contract and druh- in the sense to deceive' are not known, these expressions, in later Sanskrit literature, mean only one who injures a friend', injury to a friend'. But does mitradruh, attested in the late Vedic period, also bave the same meaning ? Yes, according to Gonda.? As is well-known the word mitradruh occurs in the Namuci story as narrated in the Taittiriya Brahmapa (1.7.11 Maitr. Sam 4.3.4). Gooda concludes that in that story mitradruh does not mean the one who breaks a contract", but rather one " who injures (the one who is, or considers himself to be ) a mitra". On the other hand, Thieme considers that in the Namuci story the expression mitradruh means contract deceiver.18 According 10 H. P. Schmidt, however, the meaning contract-deceiv. er' for mitradruh remains unestablished for the Brahmanas. "Strictly speaking, " be observes, "the passage does not prove either view : the word mitradruh is spoken by the severed head of Namuci when rolling after Indra. It can be translated either ways and there is no clue in the context that gives us an indication as to wbat the author specifically intended." I mitradruh in the Taitt. Br. is initially accented since, according to Bhatta Bhaskara, it is a vocative. * Vedio Gpd Mitra, p. 98, f. 8. 4. Before him Hillebrandt, who deceives a friend! ZII 3, 5. As mentioned by H. P. Schmidt, Indo-Iranian Mitra Studies: The State of the Central Problem, Etudes Mithraiques, Acta Iranica, 1978, p. 358. Thieme's article is not accessible to me. i. c. the one related to the Namuci story, i. c. as one who deceives a contract' or as "oac who deceives or harms a friend", Op. ost. p. 358. 3? (Annal. BORI) Madhu Vidya/161

Page #187
250 Annals BORI, LXIX (1988) It is truo, the term mitradruh in the Namucl story is not quite unam. biguous. But one is inclined to side with Thieme specifically because of tho context. The story tells us that Namuci's head was cut off by Indra after finding a way out to get rid of the contractual obligation. And if for this act of deceipt - for, Indra certainly violated the spirit, if not the word, of the contract - Namuci's head accuses Indra saying mitradhrug ast, it is more likely that it will accuse Indra of being a contract-deceiver', rather than a fricod-injurer 7. The former is more pertinent to the context than the fatter. It is also true that in this story when there is an occasion to refer to tho agreement entered into by the two parties, the word used for the agreement is sandha, and not mitra. Namuci says to Indra i sandhan te sandadhal (Maitr. Sam. 4.3.4) let me arrange an agreement with you'. Hence, one would be entitled to say that, if Namuci's head wanted to accuse Indra of breaking the agreement it would have said sandhadrhug asi, and not mitradhrig asi. But it is equally true that when, in the same story, they wish to refer to the relationship they wish to establish among themselves as a result of the agreement, they use the word sakhi (sakhaya asava ), and pot mifra. Henco It is possible to counter the above objection by saying that if Namuci's head wanted to accuse Indra of baving injured a friend, it would have said sakhldhrug asi and not mitradhrug asi. In the circumstances, it would be reasonable to look upon mitradruh, 80 close formally to the Avestap mibro.drug, as an inherited technical term signifying one who deceives the contract. It would be worthwhile to examine if we find any traces of the expression mitradruh ( or mitradroha ) as a technical term meaning'one who deceives the contract' (or deceiving the contract') in later Sanskrit literature. For, even according to Thieme, mltradruh in Classical Sanskrit is understood as he who harms a friend'.8 In tho Udyogaparvad of the Mahabharata, Vidura conveys to Krsna bis disapproval of a Degotiated peace with the Kauravas. While giving his reasons be lists the various vices ( dosah ) of Duryodhana and says: kamatma prajnamani ca mitradhruk sarvasarikitah (5. 90. 4). What can mitradruh in this context mean when it is used with reference to Duryodhana? Can it mean one who harms his friends'? But Duryodhana from the Mababbarata is certainly not known for harming his friends. He, no 1 In spite of the contract, Namuci could hardly be considered a friend of ladra.' From the story it seems that not much time passed between the conclusion of the contract and the killing of Namuci. . As reported by H, P. Schmidt, op. cit., D: 358, Madhu Vidya/162

Page #188
MEHENDALB : On mitradrih and mitradroha doubt, wants to injure the Paadavas but they are not bis friends. In fact, a little earlier, when Duryodbana tries to impress on Krsqa that there is no enmity or dispute between him (i. e. Krsaa ) and the Kauravas, Krona retorts by saying that one who hates the Pandavas, also bates him. 10 How can then the Pandavas, whom Duryodhana hates, be described as his friends 711 The expression mitradruh, therefore, in the sense one who harms a friend is not fit for Duryodhana. Whenever there is an occasion to refer to the relationship between the Kauravas and the Pandavas, the word used is bandhavala or svajana," but never a friend'. What Kisoa too wants to establish between the two as a result of his negotiations is not expressed as friendship' but 'good brotherli. ness' (saubhratra ).14 Hence it becomes necessary to look for some other interpretation of mitradruh wben it is used with reference to Duryodhada. Can mitradruh, then mean contract-deceiver'? Yes, it can, for this is what Duryodhana in the Mahabharata really is. : The game of dice was replayed between the Kauravas and the Pandavas with the following undertaking ( vyavasaya )16: the party which loses the game will live, first, for twelve years in the forest. This is to be followed by a year spent incognito; if discovered during this thirteeoth year, another period of forest-life for twelve yoars was to follow. If, howover, the period of living incognito was spent successfully, the party which had lost the game, was to receive its proper sbare of the kingdom.16 The Pandavas lost the gamo and carried out the undertaking. The gograhana incident is reported to have occurred at the end of the thirteenth year.17 Duryodhana, no doubt, was not sure whether the Paodavas had completed the thirteeoth year of life incognito and would have thereforo liked them to go to the forest again for twelve years.18 But as he was not vairan no nusti bhavata govinda na ca vigrahah / 5. 89. 22. 10 yas tan dvesti sa mam dvesti / 5. 89. 28. 11 Actually the Pandavas are referred to as sapatna !cncmy' by Duryodhana and Karna (sapatnavrddhim yat tata manyasa vrddhim atmanah / Mbh. 1. 192. 27.: 13. tasman narha vayri hantuni dhartarastran svabandhavan / Mbh. 6. 23. 37. . 18 svajanani hi kathar hatva / Mbh. 6.23.37; hantum svajanam udyatah 1 Mbh. 6. 23. 45. 5.90. 10; also 5. 72.22. This vyavasaya is referred to as an agreement' (samaya 4. 25. 4; 4. 28.7) or as a' stake' (pana 4. 423). 16 frayodase on niortte, punar ava yathocitan / svarajyam pratipattavyam itarair atha vetaraih // 2. 67. 13. 11 tatas trayodasasyante tasya varsasya bharata / susarmana grhitan tu godhanam tarasa bahu // 4. 30. 3. aniortte tu nirvase yadi bibhatsur agatah / punar dvadasa varsani vano 'vatsyanti pandavah // 4. 42. 3. Madhu Vidya/163

Page #189
252 Annals BORI, LXIX (1988) aure about his calculations be left the decision to Bbisma. Bbisma informed Duryodbana that the Pandavas bad carried out the undertaking exactly as they had promised.18 It was therefore necessary for Duryodhana to invite the Paodavas and return to them their part of the kingdom as agreed to in terms of the game. But Duryodhana was determined not to bonour the contractual undertaking and return to the Pagdavas their share of the kingdom.80 Hence, if Duryodbana has gone back on his word and if he is called mitradruh, the expression is better interpreted to mean one who has broken the contract'. This interpretation of the term mitradruh gets support from a passage that is put in the mouth of Bhlsma towards the end of the Bhismaparvan. Bhisma tries to persuade Duryodhana to give up war and coaclude a treaty with the Pandavas ( sandhis te tata yujyatam / 6. 116. 40.). In this context ho tells Duryodhana i rajyasyardham diyalam pandavanam indraprastham dharmarajo 'nusastu / ma mitradhruk parthivanam jaghanyah...( 6. 116.48). Bhisma asks Duryodhada to save himself from beiag charged to be a mitradruh'. Since, according to Bhisma, this charge could be avoided only by giving back to the Pandavas half of the kiogdom, i. e. by fulfilling the contract entered into before the game of dice, the term mitradruh can only mean its opposite i. e. one who falsifies the contract'. "Fulll the contract, and do not be a mitradruh" that is what Bhlsma tells Duryodhana. Let us consider one more passage where not only Duryodhana, but also his father Dhrtarasfra, is called mitradruh. Sanjaya tells Yudhisthira i "If (Duryodhana ) behaves towards the Pandavas who had not belied the contract (adrug dhesu), as if they had done so drugdhavat), that act of Duryodhana ) is not proper (na sadhu ), that act is not moral(na dharmyam). Dhrtarastra, in that case, along with his son, will become a mitradruh ",11 Here the words adrugdha and drugdhavat cannot mean who had not injured ( a friend )' and 'as though they had injured ( a friend)', but they have the meaning oply in terms of not breaking' or 'breaking (the agreemcot). Hence, here also the term mitradruh can only mean one who has broken the agreement'. 10 sarvain yathavac caritam yad yad ebhih pratisrutam / 4.47.5. Krsna also bad no doubt on tbis point. He says : satye sthitais tac caritam yathavat / pandoh sutaih 5. 1. 11. For the view of Yudhisthira, sce 5. 70. 10. nahan rajyam pradasyami pandavanam pitamaha / 4. 47.15; also cf. 5.90.11. 31 yad yusmakam vartate 'sau na dharmyam adrugdhesu drugdhavat tan na sadhu / mnitradhruk syad dhrtarastrak saputro yusman dvisan sadhuvsttan asudhuh // Mbb. 5. 24. 3. Madhu Vidya/164

Page #190
MEHENDALE I On'mltradrah and mltradroha 233 Obviously it is to the act of breaking the contract by the Kauravas that Arjuna refers to when, before the commencement of the war, he says to Krepa : yady apy ete na pasyanti lobhopahatacetasah kulaksyakstam dosani mitradrohe ca patakam (Mbh. 6. 23. 38). What else could Arjuna refer to by mitradroha? Certainly it could not be to 'harming the friends' becauso the Kauravas never considered the Pandavas their friends. And the possibility of destruction ' in general in the war is already taken care of by the expression kulak saya. As ibe use of the word ca shows, Arjuna is referring to somo additional sinful act wbich the Kauravas would be doing by inviting war. In the light of what has been said above it would be clear that Arjuna .could have referred by mitradroha only to the breach of contract that the Kauravas were committing by starting the war as a result of pot ful6lling the agreement. It is noteworthy that Arjuna is aware of the fact that in a war both sides are responsible for the destruction, and hence the Panlavas, very much like the Kauravas, would be committing the sin involved in killing kinsmen in the war. He therefore pleads with Krsna: kathani na jneyam asmabhih papad asman nivartitum (Mbh. 6. 23. 39). That by papa here he means killing one's kinsmen becomes clear when a litter later he cbides himself by saying aho bata mahat papam kartun vyavasita vayam / yad rajyasukhalobhena hantum svajanam udyata) // (Mbh. 6. 23.45).32 But whereas he feels that in fighting the war the Pandavas would be committing the sin arising from the killing of kinsmen, Duryodhana, and those who side with him, would do not only this but also of mitradroha. Since, according to Arjuna, only the Kauravas are guilty of mitradroha, but not the Panlavas, it becomes clear that by mitradroha he means the breach of the contract the Kauravas had committed in not fulfilling the agreement. It may be noted in this regard that the terms mitradruh and mitradroha do not occur in the earlier parts of the Mahabharata before the game of dice was played. After the unsuccessful attempt on the part of Duryodhana to burn the Pandavas alive, Duryodbana is not said to have tried to commit mitradroha. He, due to his act, is referred to as duratman, 3 durbuddhi, adharmena sthita,as papa2e and with similar expressions, but is not accused of mitradroha, which he could bave been, if the expression had meant anything else. Mbh. 13 Or, as he says earlier, papam evasrayed asman hatvaitan atatayinah 6. 23. 36. -38 1.138.28. 24 # 1515 after 1. 138. 29. 11 1. 144. 7. 96 1. 183.8. Madhu Vidya/165

Page #191
"254 Annals BORI, LXIX (1988) In the end we may pote the use of mitradroha. breaking the agreement' in the Mahabharata outside the Kaurava-Pandava context, A story is narrated in the Mahabharata 14. Sif. which involves Todra, king Marutta, Brbaspati Angirasa, and his brother Samvarta. Bphaspati, on one occasion, promises Indra that he will never officiate as a priest in the sacrifice of a mortal.a? Hence, when King Marutta wants to perform a sacrifice, he, on the advice of Narada, chooses Samvarla as bis priest. When Samvarta asks the king to do something which will sbow his stead fastness to him, the latter takes an oath that he would never leave the king." When Brhaspati, on coming to koow about the sacrifice of Marutta, comes to think about the prosperity that Samvarta would enjoy as a result of his acting as a priest in the sacrifice he becomes jealous of Samvarta. Indra then tries to persuade Marutta, first through Agai, to abandon Samvarta and choose Brbaspati for his priest. Wbco he fails, he makes anothor effort, this time through Gaodbarva Dhstarastra, to persuade Marutta to part company with Samyarta. The king again refuses to oblige. He at that time says that if he were to leave Samvarta and choose some one else as his priest he would incur tbe sin of mitradroha for which there was no expiation ivar caivailad veltha purandaras ca vlssvedeva vasaras cas vinau ca / 1 mitradrohe nisk! tir val yathaiva nastiti lokesu sadalva vadah || 14. 10.5 There was no long-standing friendship between Samvarta and Marutta, In fact, they bad just met each other, Heoce if Marutta were to abandon Samvarta, as desired by Indra and Btbaspati and thereby commit mitradroha this could not be an act of injuring a friend'. The only charge to which be would thereby expose himself was of commiting a breach of contract he had entered into with Samvaria. Hence in this context too mltradroha must be looked upon as a technical term meaning breaking the contract'. All this is not said to assert tbat mitradruh and mirradroha are never used in the epic to mean who injures a friend and injuriag a friend. "30 All that is intended to convey is that there are a few situations in the Mahabharata where one finds the use of these expressions meaning who breaks the contract' (adj.), breaking the contract' (D.). sanasvasihi devesa nuhani martyaya karhioit/ grahisyami sruvam yajne .. Mbh, 14. 5. 24. 18 Sthairyam atra katham te syat sa tvari nihsansayar kuru / Mbb. 14.7.21. 30 yavat tapet sahasranisus tistherams capi parvatah / taval lokun na labheyani tyajeyam samgatam yadi // Mbh. 14. 7. 221. Cl., for such examples, Mbh. 7. 125. 19; 7. 2. 19 (8.26. 33); 12. 166. 25; -12. 167. 20, 22. Madhu Vidya/166

Page #192
Sayana's Interpretation of daurgahe badhyamane in the Rgveda 4.42.8 M. A. MEHENDALE V 4.42 is a hymn of ten stanzas' of which the first six are 'self-praise' of King Trasadasyu (Anukramani), or of Varuna (st. 1-4) and Indra (5-6) GELDNER). Stanzas 8 and 9 together inform us that Indra and Varuna gifted a son, Trasadasyu, to Purukutsani (wife of Purukutsa) as a result of a horse-sacrifice (performed by Purukutsa) and the worship offered to them by his wife. Stanza runs as: asmakam atra pitaras ta asan sapta rsayo daurgahe badhyamane / ta ayajanta trasadasyum asya indram na vrtraturam ardhadevam The stanza refers to a sacrifice for which daurgaha was tied to the post. The seven sages acted as officiating priests at this sacrifice as a result of which they got for the queen (Purukutsani st. 9) a son, Trasadasyu. The stanza does not tell us who offered this sacrifice. It would be a reasonable guess that it was king Purukutsa, the husband of Purukutsani. The stanza also does not tell us to whom the sacrifice was offered, and For a highly informative and critical study of the entire hymn see H.-P. SCHMIDT, "The Place of the Rgveda 4.42 in the Ancient Indian Royal Ritual". In: Ritual, State and History in South Asia, Essays in Honour of J. C. Heesterman, Leiden 1992, pp. 323-349. 2 Stanza 7 is attributed by some to the seer Vamadeva, by others to Varuna; st. 10 is "probably a multi-purpose prayer" (H.-P. SCHMIDT 330). FESTSCHRIFT PAUL THIEME, StII 20 (1996), S. 257-264 Madhu Vidya/167 8

Page #193
258 M. A. MEHENDALE again it would be a reasonable guess that it was offered to Indra and Varuna. The sacrifice was a horse-sacrifice. The next stanza 9 reads as: purukutsani hi vam adasad dhavyebhir indravaruna namobhih / atha rajanam trasadasyum asya vrtrahanam dadathur ardhadevam 11 The stanza gives us some more details of the incident. It tells us that Purukutsani (wife of Purukutsa) honoured Indra and Varuna by offering oblations and obeissance. Apparently this worship was something in addition to the sacrifice referred to in the preceding stanza, for it was only then (arha) that the two gods offered her a son. From the two stanzas it appears that the gift of Trasadasyu to Purukutsa's wife was the result of the joint effort made by the king and his wife - a yajna (ayajania) by the king and a das (adasar) by the queen. This is made explicit by the following Itihasa-stanzas cited by Sayana in his commentary on st. 8: purukursasya mahisi daurgahe bandhanasthite / paryav arajakam drstva rastram putrasya lipsaya // yadrcchaya samayatan saplarsin paryapujayal / te ca pritah punah procur yajendravarunau bhrsam / sa sendravarunav istva trasadasyum ajijanar / itihasam imam janann rsir brute rcay iha // The stanzas tell us that at a particular point (of a sacrifice) when the daurgaha (horse) was already bound (to the stake), there arrived, by chance, the seven sages. (Immediately, the chief queen of Purukutsa, realizing that in her husband (who was probably engaged in the sacrifice referred to above) the kingdom was without a (future) king, and since she desired a son, worshipped the seven sages. They, pleased (with the hon Madhu Vidya/168

Page #194
SAYANA'S INTERPRETATION OF DAURGAHE BADHYA MANE ... 259 our done to them), told her to offer again profusely to Indra and Varuna. The queen did as told and gave birth to Trasadasyu. The Itihasa-stanzas thus make clear what was left vague in the two Rgvedic stanzas. They clearly tell us that what Purukutsani did was something in addition to the horse-sacrifice which was being performed by the king and hence the conclusion that the birth of Trasadasyu was the result of the joint effort done in the same direction. The information supplied by the Itihasa differs from that of the Rgveda in a few respects. In the first instance the Itihasa does not say that the seven sages were already there present at the sacrifice and acted as its officiating priests. The stanzas say that the sages arrived by chance when the sacrifice was in progress. Secondly, the Rgveda does not tell us that Purukutsani worshipped the deities at the instance of the seven sages. Finally, there is a difference in the use of the verbs to express what Purukutsani did: das- (adasat) in the RV, but yaj. (yaja, istva) in the Itihasa. As a consequence it is likely that what Purukutsani did according to the Rgveda was a simple sacrifice requiring some offerings and namas; on the other hand, what Purukutsani was required to do according to the Itihasa was an elaborate sacrifice (yaja ... bhrsam). Sayana, however, has completely misunderstood the Itihasa-stanzas and, consequently, the RV stanzas 4.42.8-9. According to him the word daurgaha occurring in the Itihasa and in the Vedic stanza (8) does not mean a 'horse', but it is a patronymic of Purukutsa 'son of Durgaha (daurgahe durgahasya putre purukutse). Stanza 8, as understood by Sayana, informs us that since king Daurgaha was firmly bound by fetters (badhyamane drdham pasair yasmat) and since the kingdom was without a 'I construe punah with yaja, and not with procuh, because the sages have not said anything before to justify connecting punah with procuh 'spoke again'. On the other hand, a sacrifice was in progress when they arrived, and hence they asked her 'to offer again (punah yaja) to get a son. Madhu Vidya/169

Page #195
260 M. A. MEHENDALE king (asminn arajake dese) the seven sages offered a sacrifice for Purukutsani and by the favour of Indra and Varuna gave her the son Trasadasyu. This sacrifice of Purukutsani to Indra and Varuna was performed by her at the instance of the seven sages (rsiprerita) (St.9). Obviously according to Sayana, Purukutsani's was not a horse-sacrifice but one accomplished with simple offerings and namas. Before Sayana, Venkatamadhava also interpreted daurgaha not as a 'horse', but as referring to Trasadasyu himself, who was born in the family of Durgaha (durgahakulajate mayi). Venkatamadhava secms to be unaware of the Irihasa-stanzas. According to him Daurgaha's (= Trasadasyu's) enemies tried to take him captive (Satrubhih badhyamane), and his commentary implies that during the fight Indra inhabited Trasadasyu and helped him to foil the attempt of his enemies (fathedras ca mam adliyatisthal). The seven sages offered a sacrifice for this victorious Trasadasyu who was still in the condition of being inhabited by Indra (athendradhisthitam mam le a ayajania). Purukulsani, Trasadasyu's mother, also gave gifts to Indra and Varuna with offerings and obeissance (danam kravati ... havirbhih namaskarais ca yukta). According to Venkatamadhava there is no connection between the birth of Trasadasyu and the sacrifice referred to in st. 8 and the worship of Purukutsani referred to in st. 9. The sacrifice (st. 8) in which the seven sages acted as priests seems to have served the purpose of a sort of thanks-giving after Trasadasyu's victory over his enemies. The worship of Purukutsani (st. 9), on the other hand, seems to have been performed during the fight to ensure Trasadasyu's victory. Both the gods Indra and Varuna, Indra by actually inhabiting Trasadasyu, helped him for his rehabilitation on the earth. Venkatamadhava interprets asyah in both the stanzas, not with reference to Purukutsani but with the earth (asyah prihivyah). And not to Purukutsa, Trasadasyu's father. Madhu Vidya/170

Page #196
SAYANA'S INTERPRETATION OF DAURGAHE BADHYA MANE... It is not clear why scholars' have not objected to the commentators' interpretation of daurgaha as something different from a horse. There is nothing in the stanzas, whether Rgvedic or Itihasa, to warrant the interpretation of daurgaha as 'Durgaha's son' (Say.) or as 'one born in the family of Durgaha' (Ven.). Sayana was led to his interpretation of the Itihasa-stanzas probably because he construed the word paryau in the second line with bandhanasthite in the first, and secondly because it was the queen, and not the king, who honoured the seven sages on their arrival and then offered a sacrifice as advised by them. He therefore felt that since the king was bound by fetters he was not free to do anything. But the above interpretation of the stanzas shows that Sayana's construing is wrong and his presumption unnecessary. According to the Itihasa the king's sacrifice was in progress when the sages arrived and he must have been busy with it and therefore there was nothing unusual in Purukutsani's taking the lead to honour the sages. Moreover, as has been already recognized by scholars, the Indian tradition is clearly in favour of interpreting daurgaha as a horse. The Rgveda stanza is quoted in the Sat. Br. 13.5.4.5 among the Gathas which record different types of horse-sacrifice performed by kings in ancient times. There can therefore be no doubt that for the author of the Sat. Br. in this context daurgaha referred to the 'horse' bound for the sacrifice. And he also says in so many words that Purukutsa of the Iksvaku familiy offered a sacrifice with the daurgaha." Next, daurgaha occurs among the synonyms of horse in the Nighan1.14. 261 Since E. SIEG, Sagenstoffe des Rgveda, 1902, 97. He has drawn attention to the fact that the Itihasa-stanzas also find place in the Nitimanjari (on st. 77) of Dya Dviveda. tena ha purukutso daurgaheneja alksvako raja. Madhu Vidya/171

Page #197
262 M. A. MEHENDALE Daurgaha, however, is not likely to be an appellative of horse in spite of Devarajayajvan's (and Madhava's) attempts to derive it either from grh- or from gah-, with an irregular loss of r in the former derivation and an irregular shortening of the vowel a in the latter. The former derivation implies that daurgaha- means the same thing as durgaha- and that durgaha- 'horse' is so called because it is impossible (dur-) to control (grh.) a horse by those who do not know the science of horse-training (?)'; according to the latter derivation durgaha- 'water' is so called because it is difficult (dur-) to fathom (gah-) its bottom, and daurgaha 'horse' is so called because it is born from durgaha- 'water'."' In support of this is cited a Vedic passage apsuyonir va asvah (Sai. Br. 13.2.2.29). Daurgaha is more likely to be the designation of a particular horse (or horses). In the Nighanlu list (1.14) daurgaha occurs between paidva on the one hand and aucsaihsravasa and tarksya on the other." All the four words are secondary derivatives and, as referring to horses, mean '(the horse) belonging to Pedu, Durgaha, Uccaihsravas, and TIksirespectively. Paidva is already known from the Rgveda as a horse gifted to Pedu by the Asvins (1.116.6; 1.117.9; 1.118.9; 1.119.10; 7.71.5; 10.39. 10). Aucsaihsravasa is known from the late Vedic tradition as the horse of Indra (AV 20.128.15; RV Khila 5.14.4). It probably means 'horse belonging to Indra Uccaihsravas (whose fame rises high or is loudly proclaimed)'. On this analogy Daurgaha and Tarksya also could mean "horses be Nirukia (Nighanlu) with the commentary of Pandit Devaraja Yajvan, Calcutta, 1952, Vol. 1 p. 164. This latter derivation is of Madhava who is cited by Devaraja Yajvan. * dugabe upapade grhniteh gahe (? gaher) va ... rephalopah Prpodaraditvai grhnaren / asvahrdayanabhijhair grhirum (? grahitum) asakyarvat durgaha iry ucyale / durgaha eva daurgahah ... / 10 (prsodaradirvat) gaher hrasvarvam /... yad va 'duhkhena ga(? ga hitayyarvat dur. gaham (? durgaham) jalam ucyare' iti madhavah, tarra bhavo daurgahah) "A similar name for a horse, not included in the Nighanlu list (1.14), is Taurvasa (Sar.Br. 13.5.4.16). Madhu Vidya/172

Page #198
SAYANA'S INTERPRETATION OF DAURGAHE BADHYA MANE... longing to Durgaha and Trksi12 respectively. Trksi is known in the Rgveda as the son of Trasadasyu (8.22.7). A king Durgaha is also mentioned in the Rgveda whose grandsons were generous and had acquired fame among the gods (8.65.12). It is difficult to say whether this Durgaha was identical with Purukutsa himself or it was the name of his father, i.e. whether Purukutsa offered the sacrifice with his own horse or with that of his father. As regards the principal points of this incident, that Purukutsa performed a sacrifice with daurgaha, that Purukutsani offered worship to Indra and Varuna, and that as a result of these two acts she got the son Trasadasyu who was destined to be powerful like Indra, there is no conflict in the Indian tradition as represented by the RV stanzas, the Satapatha Brahmana, and the Itihasa stanzas. There is a conflict only between the Vedic tradition and the Itihasa on the one hand and the interpretation of the RV stanzas and the Itihasa by the mediaeval commentators on the other. 263 Finally a word about ardhadeva (8.42.8,9). H.-P. SCHMIDT (329, 330) translates it as 'demi-god" and remarks that a human being, was "thought to be specially endowed with divine powers which entitled him to be called ardhadeva, 'demi-god" (346). This seems to me to slightly miss the mark. Trasadasyu owes his birth to the favour of two gods Indra and Varuna. But he was born inbibed with the qualitiy of only one of them viz. Indra. Like Indra he was the overpowerer or breaker of obstacles (vrtratur-, vrtrahan-). This attribute of Trasadasyu was so striking that it is repeated in both the stanzas 8 and 9 narrating his birth. If Trasadasyu had imbibed at birth the quality also of the other god Varuna, then, in his self-praise, he would have referred to it in one of the two stanzas as 'guardian of rta or vrata', which would have made him comparable to 12 This has been already suggested by FOY (KZ 34.366-367) cited by A. A. MACDONELL, Vedic Mythology p. 149. MadhuVidya/173

Page #199
264 M. A. MEHENDALE Varuna. Trasadasyu was half-god' at birth in this sense (possessed of the quality of one of the two gods) and not because he was partly human and partly divine. Since according to Venkatamadhava Indra had inhabited Trasadasyu, he takes ardhadeva to refer to such Trasadasyu as was inhabited by Indra as his 'owner' (indras ca svami trasadasyus ca). Perhaps now we are in a better position to understand the significance of the word dvita used by Trasadasyu in the very first stanza. The occasion for the self-praise was admittedly his royal consecration. Equipped at birth with the Indra-quality, he was now, after the consecration, going to imbibe and excercise also the Varuna-quality of upholding ria and vrata. Hence, as a Ksatriya, his governance was going to be two-fold (mama dvita rastram ksatriyasya). He was going to be the king doubly - the Indra-way and the Varuna-way. He was thus going to combine in himself what, according to the Avestan tradition, was shared by the two brothers Urvaxsaya and K@r@saspa (Yasna. 9.10.)". 13 H.-P. SCHMIDT, op. cit. p. 342. Madhu Vidya/174

Page #200
On Vedic Durona M.A. MEHENDALE jii. 1. The Vedic word durona is given in its loc. sg. form durne in the Nighantu as a synonym of house (grhanamani 3.4). The Sanskrit dictionaries (Grassmann, Bohuingk-Roth) too assign to it the same general meaning 'house, home, dwelling place'. 2. An examination of the passages where the word occurs in the Rgveda would, however, point out that the primary meaning of the word is a guest house', a dwelling for an atithi'. A start in this direction is given by the following passages: i. atithir durona imam no yajnam upa yahi (5.4.5) where Agni is invited to the sacrifice as a guest to a guest house. ii. atithir duronasat (4.40.5) which expression clearly points out that the place where a guest takes his seat is the durona. yatha sam adhvan cham asad durone'so that there may be weal on the way, weal in the guest house' (10.37.10). 3. Since Agni is the most cherished guest of the Vedic sacrificers his association with durona, a guest house, stands out very prominently in the Rgueda. Thus we read agnih.... nisatto ranvo durone, 'Agni ...., the agreeable one, seated in the guest house' (1.69.2; also 3 ranvo durone); ni durone .... sasada ..... agnih 'Agni has sat down in the guest house' (3.1.18); revato durone... atithih (Agni) a guest in the guest house of a richman' (7.42.4); it is Agni alone who is described as duronayu 'longing for a guest house' (8.60.19). 4. On account of Agni's association with durona, the word has also come to mean 'the place of sacrifice' where fire is kindled and soma is pressed. Hence we read: agne .... samidhyase durone Agni, you are kindled in the place of sacrifice (3.25.5; also 4.2.7 a devayurinadhate durone); samiddho adya manuso durone (Agni is kindled today in the sacrificial place of the man' (10.110.1); 3.18.5; a sacrificer wants to please Agni in his own place of sacrifice (piprisati sva .... durone 4.4.7); also cf. yasi daivansam ..... istaye durone (7.92.3) where Vayu is spoken of as going to the place of sacrifice to receive offering, and durona a deva etu pra no havih (8.19.27) where a sacrificer expresses his wish that his offering may reach the gods in the place of sacrifice. Durona or a place of sacrifice is also a place where soma is pressed: durona a nisitam somasudbhih' (Indra) is made sharp by the pressers of soma in the place of sacrifice' (4.24.8); dasuso durone sutavatah 'in the sacrificial place of the sacrificer who has pressed somal (3.25.4). Sayana too, on occasions, is seen rendering durons as yajnagyha 1.69.2; 4.24.8; 7.92.3; 8.19.27; 10.104.4); Yagagiha 7.70.2; devayajana (8.87.2); devayajanagyha (5.76.4); also of. 1.117.2); Xhavuniya 7.12.1. 5. Besides Agni, there are other deities also who visit the place of sacrifice either with Agni or alone. (i) yad indragni madathah sve durone..... atha somasya pibatam sutagya'since, oh Indra and Agni, you delight in your own place of sacrifice .... drink of this pressed soma' (1.108.7; also 3.25.4).4 Madhu Vidya/175

Page #201
On Vedic Durona 47 (ii) The Asvina also visit the place of sacrifice of a sacrificer: yena gacchathah sukrto duronam.(the chariot on which, you Asvina) go to the place of sacrifice of the pious' (1.117.2, also 1.183.1; 4.13.1); tam patam a gatan durone' (oh Asvina, drink that (soma), come to this place of sacrifice' (3.58.9); the sacrificer directs the Asvina to his sacrificial enclosure as distinct from his house: ime gyha asvinedam duronam'these are our houses, oh Asvina, this the place of sacrifice' (5.76.4); if not soma, it is milk that is ritually made hot which awaits the Asvina (atapi gharmo manuso durone'gharma drink has been heated in the sacrificial place of the man 8.87.2; also Cf. 10.40.13. 6. Since duronaa guest house' has also come to mean 'a sacrificial place', a special house of Agni, the atithi, the word, in a few places, seems to mean 'ones own special place'. Thus in 3.14.3 the dawn and the night are said to step into their respective special houses (usasa ..... a ..... lasthatur durone).5 Similarly in 7.60.5 we are told that the three Adityas, Varuna, Mitra and Aryaman, grow in the special house of truth which is in the highest heaven (ima stasva vavidhur durone 7.60.5). 7. Comments on P. Thiemes paper 'Durona-n.' published in "Amrtadhara" (Dr. R.N. Dandekar Felicitation Volume) 1984, pp. 435-444. Thieme proposes to see in ona of durona, analysed as dur-ona, a stem ona which he considers to be a cognate of Greek eune bedstead'. In that case duronaprimarily would mean '(a house/home] offering a bad bedstead'. Thieme is conscious of the fact that this etymological explanation conflicts with the notion of comfort that is otherwise associated with durona, e.g. in 7.42.4 where a guest is said to rest in a good lap of the durona belonging is a rich person (revato durone syonasir atithih). Thieme's attempt to get over this difficulty by assuming that this way a host is only modestly describing his house to his guest as a place of discomfort' is not likely to succeed. My comments on his paper follow: 1. p. 435: Thieme cites Yaska's explanation of durona (Nir. 4.5): durna iti grhanama durava bhavanti dustarpanand translates it as: "durone'is a name of the house (grhas). It (the house) is difficult to be helped, that is:) difficult to be satisfied". It is not clear why Thieme renders duravah 'as difficult to be helped'. Yaska's dustarpah is only a paraphrase of duravah and hence both the expressions have to be translated alike. This means Yaska's av - = trp-6 Thieme's translation of Durga's remark given in f.n. I on p. 435 is also not accurate. The translation should have started as: "And it is said ......" In the commentary of Durga available to me the words grham upakramya, given by Thieme, do not occur. 2. p. 435: While rejecting the old etymology of ona in durona from the root av 'to satisfy' Thieme says that difficult to be satisfied (by food and drink) does not recommend itself as a designation of a house. That is correct. But it would certainly be an apt designation of a guest house, where guests, notoriously difficult to be satisfied, are accomodated. 3. p. 436: In support of his derivation of duroni given above Thieme cites the paraphrase of this word given in the Satapatha Brahmana6.7.3.11: duronasad visamasad ity etat and translates it as: "duronasat, that is 'taking his seat in an inconvenient / uncomfortable place/house". This does not seem to be correct. Visamasat here means 'taking his seat in different places'. This is shown by the fact that according to the Madhu Vidya/176

Page #202
48 M.A. MEHENDALE Sat. Br. Agni in this passage is said to be sarvesam va esa bhutanam atithih. Since Agni has to go to different persons as their guest he is naturally visamasat'taking his seat in different (and not the same) place'. We may note also that in a different context Agni's epithet sahasrambhara is explained in the Ait. Br. 1.28 as esa ha va asya sahasrambharata yad enam ekam santan bahudha viharanti". "This indeed is his (Agni's) sahasrambharata that, though one, he is carried in different places". 4. pp. 439 440: Thieme suggests a distinction in the use of sva durona as sometimes applied to the place of sojourn of Agni (59) and sometimes to the place of one who offers hospitality (510). But this is not borne out by the evidence. In all the passages cited by him the two words refer to Agni's own place, i.e. the place of sacrifice where he is the guest.? 5. p. 440: Thieme is right when he says that durona is not by origin a "religious" or "poetical term". It is definitely taken, as he says, "from everyday language as used in everyday situations". But this is shown not by Thieme's etymological analysis but by the specific meaning 'guest house' of durona. The word has acquired religious connotation because Agni, as atithi'guest' par excellence, has his place in the sacrificial house. 6. p. 442: Thieme cites RV4.28.3 in which he takes durge durone to refer to the place of refuge in the mountains of the enemies of Indra and Agni. He also takes the phrase kratva ne yatam to refer to the strong will to fight) of Indra and Agni. This does not seem to be the correct interpretation. The expression durge durone as usual refers to the sacrificial place which was specially made inaccessible to those who wanted to disturb the sacrifice. It was this place of sacrifice which the Dasyus attacked as if with a strong will, and where they were destroyed by Indra and Agni. 7. Finally some corrections: (1) Thieme's svadathah in 1.10.8.7 cited twice on p. 437 and p. 440 to be corrected to madathah: (2) 6.41.1 (p. 439) has the word akas and not duma. NOTES 1. The aging Ghosa who lived with her father until she was married seems to have lived separately from the father in his guest house: ghosayai cit pitysade durone. 2. Also cf. 1.70.2; 6.12.1; 7.7.4; 10.104.4. 3. Also cf. 7.7.4; 7.12.1; 7.16.8. 4. For 4.28.3 see below Comments on P. Thieme's paper (6). 5. Thieme is right in interpreting durone in this passage as acc. du and not loc sg. (see Thieme's paper noted below pp. 442-443. 6. Cf. Durga on this passage: avalis tarpanarthah. Madhu Vidya/177

Page #203
Reprint IIJ ID-2 TWO SOGDIAN ETYMOLOGIES (II) Iranian (1) The word for "astrologer" occurs in Buddh. Sogdian as snptsr and in Man. Sogdian as smbtsr (cf. W. B. Henning, Sogdica, p. 21, 22). Benveniste in his edition of the Vessantara Jataka apparently divides the word into two parts, snpt and sr, and suggests to compare the first part of the word with Persian sambat, sambut "vision, fantome" (p. 90.22). It may, however, be suggested that it would be better to connect this word with Skt. samvatsara "astrologer" and regard it as a loan in Sogdian. Skt. samvatsara is attested since the epics. The Sogdian form points to a pronunciation sambatsar with the shortening of the viddhi vowel and the change of my to mb. The fact that both these types of changes are found in the Buddhist Sanskrit and further that the cluster is is preserved in this language would suggest that the speakers of Sogdian took sambatsar as a loan from Buddhist Sanskrit For the shortening of the vyddhi vowell we may compare abhinadita for -nddita "caused to sound", anantariya for anantariya "crime bearing immediate fruition, deadly sin", karaka for karaka "doer, maker", etc. (F. Edgerton, Buddhist Hybrid Sanskrit, Grammar, 3.31). For the change of my to mb we may note that Edgerton in his Dictionary of BHS notes both forms samvara and sambara for samvara "restraint, control". We may also note the following instances of this kind of change from the other north-western Indian and Central Asian material. Thus the shortening of the vrddhi vowel is seen in Khotanese in valmiki for valmiki (Ernst and Manu Leumann, AKM, XX, p. 497; also cf. H. W. Bailey, BSOAS, 10.917 ff., where he gives such instances as nivaysva (-ya < -ika) < naivasika etc.). For the second, we find such cases as sambatsara (also samvatsara) in the North-western Indian Kharosthi inscriptions (cf. Sten Konow, Kharoshthi Inscriptions, CII, Vol. II, part 1, pp. cv and 158, 165 and 62). Similarly we get svayambar as a loan from Skt. svayamvara in Uigur (cf. F. W. K. Muller, Uigurica, II, APAW 1910, p. 21 lines 19 and 7) and svayampar in Tokharian (cf. E. Sieg and W. Siegling, Tocharische Grammatik, p. 517). (2) In Buddhist Sogdian we get 'pks- (apkas-) "side" (Vessantara Jataka 8). At one time it was thought that the word was a borrowing from Skt. paksa.. But this view is now given up since Benveniste (his edition of the Vess. Jat. p. 90) made the good suggestion to derive it from upa-kasa, av. kasa. However, an alternative suggestion may be made to derive the Sogdian form from *apkasa- going back to Old Ir. *api-kasa-, which would correspond to Skt. api-kaksd- "region of the arm-pits" (Rv. 4.40.4, 10.134.7). Deccan College, Poona M. A. Mehendale ! I am greatly indebted to Prof. H. W. Bailey for most of the references in this note. Madhu Vidya/178

Page #204
Some Avestan Notes M.A. Mehendale 1 ta...aparo 'with him as second' In Y. 45.11 occurs yasta daeving aparo masyasca tara masta. Insler' proposes to read a.paro in place of aparo. In his commentary, he interprets a.paro as nom. pl. of *a-par- 'guilty' which he treats as a derivative of par- to judge as guilty'. Accordingly Insler translates the above line as "the person who, in this very way, has opposed the guilty gods and mortals". guilty gods and the person guilty? Accordithhe treats Humbach', on the other hand, translates the passage as "Wer sich mit ihm als zweiter uber Dacvas und Menschen erhaben fuhlt ". If Humbach's interpretation of ta...aparo is correct we have in this construction an exact parallel to the Sanskrit compound of the type atmanadvitiyah 'with self as a second, i.e. accompanied by one more person' recognised by Panini -(6.3.6) 2 abifra- 'incomparable abifra- (adj.) occurs only once in the Gathas (Y. 33. 13). Bartholomae (92) assigns to it the meaning ohne Gleichen, unvergleichlich'. Humbach accepts this meaning and translates the Gathic passage raf@drai vourucasane doisi moi ya va abifra as "Zu Hilfe, o Weithinblickender ! Zeige mir, was eure unvergleichlichen Eigenarten sind". He does not comment on the word. Insler', however, does not accept the above meaning. He renders the passage as "Lord of broad vision, disclose to me for support the safeguards of your rule." Commenting on this word he admits it to be difficult'. Since the attested form is not albifra, he does not look upon the initial a-as alpha privativum but takes it to be a shortened form of the preverb a-. In his opinion the attested form abifra stands for the original *abifra. As regards the form, he derives the stem *a-bifra- from *a-bibhra-, i.e. from the root a-bhar-. For the unexpected change -bhr->-fr-, he cites the parallel AV. jafra- 'deep' = Vedic gabhira. As regards the meaning, he says that the root bar in the sense 'support' occurs in the Gathas. Apparently, according to Insler, the meaning 'safeguard' can be had from 'support'. Madhu Vidya/179

Page #205
SOME AVESTAN NOTES H.P. Schmidt has already pointed out the difficulties in admitting Insler's interpretation, but he accepts Insler's derivation of bifrafrom bhar. He treats a- as alpha privativum. The meaning he assigns to abifra- is 'not to be carried off, inalienable', because in his opinion bhar- has the connotation 'carry off' (cf. Yt. 10.21). Schmidt's interpretation suffers from the fact that bhar- with the preverb apa- can have the connotation 'carry off, but not with the preverb a. That would rather have the opposite meaning to 'carry to or near'. 13 Accepting Bartholomae's suggestion that abifra- means 'incomparable', it is possible to make a suggestion regarding the etymology. While teaching the mode of formation of the ordinals, Panini has used the word purana (tasya purane...5.2.48). Accordingly Sk. dvitiya 'second' literally means "What fills, what completes the number 'two'." Taking a cue from this it is possible to explain -fra- in -bifra- from the root Sk. pra-, Av. par- (ham-pafraiti) 'to fill'. Bifra-, like dvitiya-, then would mean 'what completes the number two, i.e. the second'. Abifra-, like advitiya-, would mean 'without a second, imcomparable' It is likely that bifra-, due to a semantic shift, got the meaning 'double, one to compare with'. The V. 13.44 passage sunahe aevahe asta bifram' would mean 'The one dog has eight doubles or persons to compare with'. 3. mayavant- full of bleats (or bleat-making sheep) The word, in its fem. form mayavaiti-, occurs only once in a younger Avestan text. Bartholomae (1168) assigns to it hesitatingly. the meaning 'wo Begattungen stattfinden. The passage in which the word occurs runs as kata asaum apa.jaso sitibyasca haca gaomaitibyasca vayavaitibyasca haca mayavaitibyasca- (Haboxt Nask 2.16)"... von den Statten, die mit vierfussigen Tieren und mit Vogeln. bevolkert sind und wo deren Begattungen stattfinden (?)" (Bartholomae 1169). Since the dwelling places are here first described as full of cattle (gaomaiti-) and birds (vayavaiti), it would be natural to expect the following word mayavaiti to have a comparable meaning. It is possible to obtain this by interpreting the word mayaas 'bleating sound' from Sk. ma (mimati) 'to bleat'. We may compare with this interpretation Av. anumaya- 'sheep'. The word mayavantwould then mean 'full of bleats (or bleat-making sheep)'. Madhu Vidya/180

Page #206
Notes 1) The Garhas of Zaruthustra, 1975. 2) Die Gathas des Zarathustra. Band I, 1959. 3) op. cit. p. 103. 4) op. cit. p. 53. 5) Form and Meaning of Yasna 33. 1985. 6) Sg. for pl. The explanation ofered here of bifra- should answer Barthoomac's question : 'Was bedeutet *plo- eigentlich ?" (under bi-fra- 965) 7) Accordingly H. Reichelt, Avesta Reuder (p. 252) 'Cohabita tion-supplied, where cohabitations or pairings take place (doubtful)', and M.F. Kanga (Festschrift Prof. P.V. Kane, p. 250) 'full of... enjoyments of love'. 8) Cf. Sk. mayu- 'bleating'. Bibiiography 1) S. Insler: The Gathas of Zarathustra, E. J. Brill, Leiden, 1975. 2) Helmut Humbach: Die Gathas des Zarathustra, Bande I, II, Carl Winter Universitatsverlag, Heidelberg, 1959. 3) Hans-Peter Schmidt: Form and Meaning of Yasna 33, (with contri butions by Wolfgang Lentz and Stanley Insler), American Oriental Society, New Haven, Connecticut, 1985. 4) Hans Reichelt: Avesta Reader, Texts, Notes, Glossary and Index, Karl J Trubner, Strassburg, 1911, Photomechanischer Nachdruck, Walter de Gruyter & Co., Berlin, 1968. 5) Festschrift Prof. P.V. Kane (A Volume of Studies in Indology presented to.....), ed. by S.M. Katre and P.K. Gode, Oriental Book Agency, Poona, 1941. Madhu Vidya/181

Page #207
TWO AVESTAN NOTES* by M. A. MEHENDALE Poona 1. Vibara@want This word is attested only once in the Avesta. In Y. 9.14 we read: tum... ahunam vairim frasravayo vibara@wantam "du hast (das Gebet) AhV. vorgetragen unter Einhaltung der Pausen" (Bartholomae). The meaning assigned by Bartholomae to this word as referring to the mode of recitation by separating the verses or parts of verses with pauses in between seems to be correct. We may compare this use of vivbarwith the similar use of viv/har- "to separate (with a pause or insertions)" found in the Sanskrit ritual texts. While prescribing the hymn for the Ajyasastra of the morning pressing the Aitareya Brahmana (2.35.1) says: pra vo devayagnaya" ity anustubhah prathame pade viharati, tasmat stry aru viharati samasyaty uttare pade tasmat puman uru samasyatl "He separates the first two Padas; therefore a woman separates her thighs. He creates (mistake for 'unites') the last two Padas; therefore a man unites his thighs" (Keith). Cf. Sayana: viharanam prthakkaranam dvayoh padayor madhye viharam vicchedam krtva pathet. According to I am thankful to Dr. H.-P. Schmidt for some valuable references in this paper. 1 Reichelt follows Bartholomae "by observing the pauses" i.e. with pauses between the three verses of the Ahuna Vairya. Lommel renders, "... hast du ... das Ahunavarya-Gebet hergesagt, das in Abschnitte gegliederte (?)..." 3 RV. 3.13.1. 8 Kau.Br. 14.2 uses pade vigrhnati. Cf. with this yas ta uru viharati RV. 10.162.4. For another use of viharati "separates, keeps distinct" cf. Ait. Br. 2.37.1: tad yad Ajyena Pavamanam anusamsati Praugenajyam devarathasyaiva tad antarau rasmi viharati "... in that with the Ajya he follows in recitation the Pavamana, with the Prauga the Ajya (stotra), verily thus he separates the inner reins of the chariot..." (Keith). Kau. Br. 14.4, however, looks upon the recitation of the Ajya and the Prauga as 'intertwining' cf. te etad viharati yatha rathasyantarau rasmi vyatisajed "Thus he transposes the two; it is as if one were to intertwine the outer (? mistake for "inner") reins of a chariot" (Keith). For this use of viharati "intertwine" cf. below. Madhu Vidya/182

Page #208
62 M. A. MEHENDALE the Kau. Br. 14.2 in this mode of recitation breathing in at the time of the pause between the verses is prohibited (tasmad anavanam samkramet). The verb vihar- is used not only in the sense of separating the verses with a pause but also in the sense of separating them and intermingling them with other verses. In the Ait. Br. 6.24.5 we find: pacchah prathamam sad valakhilyanam suktani viharaty, ardharcaso dvitiyam, rksas trtiyam "First he transposes by Padas the six hymns of the Valakhilyas, by half verses the second time, by verses the third time" (Keith). This kind of intertwining of verses is also referred to by viharati and vyatisajati in the Ait. Ar. In 5.1.6 we read nadam va odatinam ity etayaitani vyatisajati padaih padan brhatikaram "He intertwines these hymns with the verse nadam va odatinam (RV. VIII.69.2) joining quarter-verse with quarterverse making them into brhati verses..." (Keith). The same mode of recitation is referred to earlier in the Aranyaka 1.3.5,8 as ta nadena viharati... tristubham canustubham ca viharati. It is interesting to note that in this Aranyaka viharati is also used in the sense of "adding a syllable" in recitation. Thus in 1.3.7 we find nyunaksare prathame pade viharati "He extends the first two verses, which are deficient, by a syllable" (Keith) which refers to the addition of the syllable pu to RV. 10.120.1 and 8.69.2. Similarly with reference to the addition of the syllables pu, ru, and sa to the first, second, and fourth quarters of a stanza we read in the Ait. Ar. 5.1.6 prathamayam ca purusaksarany upadadhati, padesv ekaikam avasane, trtiyavarjam sa khalu viharati "He also inserts in the first stanza the syllables of the word purusa, one in each quarter-verse at the end, save in the case of the third quarterverse" (Keith). The close similarity between the Avestan use of viv/bar- and Skt. vihar- "separate, with pauses or with insertions of other verses" clearly shows that the two are etymologically related and that the latter should go back to viv/bhar-. For the change of bh>h in the Vedic period cf. Wackernagel, Altind. Gramm., 1, SS 217-18. The fact that the Brahmanas have har- for earlier vbhar- is also shown by the following instances. In the Ait. Br. 1.28.35 we notice the use of har- while explaining a form with bhar- in the Vedic citation: sahasrambharah sucijihvo agnir ity esa ha va asya sahasrambharata yad enam ekam santam bahudha viharanti "Bearing a thousand, Agni, of pure tongue' For details cf. As. Sr. Su. 8.2.19-21, Haug, pp. 283-4, fn. 13 and Keith H.O.S., 25, p. 277, fn. 2. For similar use also cf. Ait. Br., 6.26.10; 28.3; 4.2.4 4.5 and Keith, p. 199, fn.3; also cf. Kau.Br. 30.4; 17.2-4. Rv. 2.9.1. Madhu Vidya/183

Page #209
63 TWO AVESTAN NOTES (he says); for this is his character of bearing a thousand, that him being but one they carry apart in many directions" (Keith). Similarly in the Sat. Br. 6.7.4.4 we find vidma te dhdma vibhrta purutraR iti yad idam bahudha vihriyate "We know thy manifold scattered cites" - inasmuch as he (Agni) is here distributed many ways" (Eggeling). These instances. will further point out that the verb vihar- which is often used with reference to the separation and distribution of the sacrificial fire is etymologically earlier vi bhar-, II. Yana 11 The word is attested both in the Gatha and the later Avesta. Bartholomae gives the meaning "Gunst, Gunstgewahrung, Gunstbeweis (seitens der Gottheiten gegenuber den Menschen)". He has been followed by most scholars, who render yana- by "favour". Humbach is probably the first scholar to differ from Bartholomae. He renders yanam vaed- as "einen Weg, ein Mittel finden" (cf. Skt. ydna- "vehicle") and yana- in Y. 28.9 as "Opferumzug" (obviously deriving it from vya- "to go"). He translates the latter passage andis va noit... yanais zaranaema... "Mogen wir Euch nicht durch diese Opferumzuge erzurnen..." W. Lentz in his edition of Yasna 28 (Abh. Akad. Wiss. u. Lit. Mainz, 1954, Nr. 16, pp. 933, 949-50) rejects the view of Humbach especially because he does not understand "... weshalb der Prophet furchten soll, die Gottheit durch Opfer zu erzurnen". Lentz gives the meaning "Wunsch" but does not discuss the etymology of the word. Apparently following Hoffmann and Lentz, Humbach has given up his earlier rendering of yana- in his recent translation of the Gathas.14 Vaj.Sam. 12.19; RV. 10.45.2; also cf. agner dhamani vibhrta purutra/ RV. 10.80.4. "That the use of viv/bhar- had become rare in the Brahmana period can also be seen from the following passage Ait. Br. 1.18.1: tam deva vimethirel sa haibhyo vihrto na prababhuval te hocur deva na vai na ittham vihrto 'lam bhavisyatil hantemam yajnam sambharameti/ "The gods crushed it; it being taken apart was not sufficient for them. The gods said 'It will not be sufficient for us, being taken apart; come let us gather together the sacrifice" (Keith). The Sat. Br., however, once retains vibhriaof the Vedic citation (6.4.4.2). 10 In BB, 10.279 fn. 1 he gives the meaning "Gabe". 11 Also Gershevitch, The Avestan Hymn to Mithra, p. 143 (tr. of Yt. 10.137). Lommel, however, at one place (Yt. 17.26) renders it as "Unterstutzung". Munchener Studien zur Sprachwissenschaft, 2 (1952), p. 18, n. 17. 12 13 Following K. Hoffmann, cf. I.F., 63, p. 100, fn. 1. 14 Die Gathas des Zarathustra, I, p. 78 and II, p. 11 (Heidelberg, 1959). Humbach, however, retains connection with "Gottesdienst" cf. his rendering of yanya- Y. 28.0 "zum Gottesdienst gehorig". But yanya- probably means "favourable to, i.e. willing to fulfill, request". Madhu Vidya/184

Page #210
64 M. A. MEHENDALE He also now rightly derives yana- from Vya- "angehen, bitten" which use is well attested in Sanskrit.15 This particular meaning of vya- "to request, to implore" is so similar to that of Skt. Vyac- that Yaska regards yami (RV. 1.24.11, 8.3.9) = *yacmi 18 That is the reason why in the Nirukta 2.1 he says: athapi varnalopo bhavati tat tva yami iti "Oder geht uberhaupt ein Buchstabe verloren, z.B. in tat tva jami" (Roth). According to the commentator Durga, the letter that has disappeared is c of Vyac- (atra cakaralopah/ yacamityl7 evam etad drastavyam/). Thus in the opinion of Yaska yami in this particular reference is to be derived not from v pa-, but from Vyac. The commentary of Skanda-Mahesvara, 18 however, does not take the above view of yami. In the view of this commentator the varnalopa intended by Yaska is not to be seen in yami but in tatva. Since this form is derived from v tan- it shows the loss of final -n. Then again, according to this author the instance cited by Yaska is not a Vedic quotation but one taken from the daily usage. But this view does not seem to be satisfactory for various reasons. In the first instance if Yaska had intended to give tatra as an illustration of varnalopa he would have used only that word and not tatva yami which looks like a Vedic citation, Secondly this type of loss he has already pointed out in the forms like gatva and gatam (from V gam-). Moreover the new explanation presupposes the reading tatva instead of tat tva for which there does not seem to be any manuscript evidence. With reference to the remarks of Lentz on anais va noit ... yanais zaranaema Y. 28.9 about vexing the god with repeated requests ("Den Dichter befallt nach den verschiedenen dringenden Bitten und Anrufungen plotzlich die Sorge, die man hat, wenn man einen vertrauten 15 W. Schmid, however, has shown that this Vya- "bitten" is to be separated from Vya- "gehen", cf. IF, 62 (1936), p. 219 ff. On p. 225 the author suggests to render Skt. avayana- "Abbitte", although on p. 237 he favours not to connect Av. yanawith Vyas (yami shows the loss of two letters c and d and not one. 18 Published by L. Sarup (1931), pp. 10-11. Madhu Vidya/185

Page #211
TWO AVESTAN NOTES 65 Freund uberfordert", p. 950) Humbach has already compared RV. 2.33.4 ma tva rudra cukrudhama namobhir. We may also add the following passage from the Sat. Br. 2.3.4.4, which is pertinent in this connection: ubhaye ha va idam agre saha asur devas ca manusyas ca/ tad yad dha sma manusyanam na bhavati tad dha sma devan yacanta idam vai no nastida no 'sty itil te tasya eva yacnyayai dvesena devas tirobhuta ned dhinasani ned dvesyo 'saniti tasman nopatistheta/ "Now in the beginning both the gods and men were together here. And whatever did not belong to the men, for that they importuned the gods, saying, 'This is not ours; let it be ours!' Being indignant at this importunity, the gods then disappeared. Hence (it may be argued) one should not approach (the fires), fearing lest he should offend them, lest he should become hateful to them"19 (Eggeling). A discussion of a few passages where yana- occurs may now be added. According to Geldner's edition, which is followed by Bartholomae, we read in Yt. 14.36: yana baraiti astava va taxma| he marayahe parenavo va taxmahe marayahe... Bartholomae suggests to consider yana as acc. pl. mas., which would be an irregular formation, 20 and translates: "Gunstbeweise (der Gotter) tragt davon, wer einen Knochen des starken Vogels oder eine Feder (als Amulet) tragt." Lommel, however, translates differently: "Wer einen Knochen dieses kraftereichen Vogels tragt oder wer eine Feder dieses kraftereichen Vogels tragt..."21 Apparently he reads yo na "the man who..." for yana. He also seems to take astavo and paranavo as acc. sg. The irregularity in Bartholomae's interpretation of yana can be removed if we regard it as acc. pl. fem. from a base yana-. With this we may compare Skt. yac-na- (or yac-nya-), fem. beside yac-na-23 (or yac-nya-) mas. We may then translate the above passage as: "(The amulet) containing the bone of the strong bird or containing the feather of the strong bird brings (him) many requests (i.e. fulfils them)." For a similar use of Vbhar in the sense of the fulfilment of wishes cf. RV. 10 The point that is established, however, is that one should approach the fires. 30 Similarly in Y.43.14 he takes mqora as acc. pl. mas. Perhaps there also we have to take it as a fem. form. 21 Darmesteter (S.B.E., 23, p. 241): "If a man holds a bone of that strong bird, or a feather of that strong bird..." 22 Wackernagel-Debrunner, II, 2, seem to take yaciya- (- yacna-) once as a -na derivative ( 561a and c), and once as a -ya derivative (8 652a). Madhu Vidya/186

Page #212
66 M. A. MEHENDALE 10.95.10 vidyun na ya patanti davidyod bharanti me apya kamyani "who flashed like a flying lightning and brought to me my (apya-7) desires. (i.e. fulfilled them)." In Yt. 10.137 we read: rast@m ahmai naire mainyai mibro maban@m acaraiti, yezi.se yanada bavaiti saNGhamcit anu sastrai sanh@mci anu mainydi. This is translated by Gershevitch as "Straightway Mithra visits the residence of this authoritative man, if as a result of his (the man's) favour (shown to the priest), it (the utterance of Mithra's name) is in accordance with the prescription for recitation, in accordance with the prescription for thinking (= the prescriptions for praying orally and mentally)." But if we understand yana-request and do not interpret the two occurrences of mainyai in two different ways, then we may translate the above passage as: "Straightway Mithra comes to the place (of worship) for this authoritative man, if it (= the utterance of Mithra's name) happens following his ( the man's) request (to the priest). (Then) following the recitation (Mithra comes to the place of worship) for the praise, following the recitation (he comes) for the authoritative man." 23 "Die wie ein fallender Blitz aufblitzte und mir Liebesergusse brachte..." (Geldner). But Oldenberg renders "die mir die kamyani ihrer Heimat, des Wasserreichs, brachte". He also suggests the possibility of taking apya nom. sg. 24 Op. cit., pp. 142-143 and 283-284. One can find here also the earlier interpretations. Madhu Vidya/187

Page #213
Indian Linguistics 27 TWO NOTES ON YASTS by M. A. MEHENDALE, Deccan College, Poona (1) On Hom Yast, verse 1. At the very beginning of the Haoma yast we are informed that Haoma approached Zarathustra at the time of the pressing when the latter asked him the question about his identity. Zarathustra asked : ko nara ahi, yim azam vispahe ayhaus astvato sraestam dadarasa xVahe gayehe xYanvato amasahe ? (Y. 9.1). The last part of Zarathustra's above question-xvahe gayehe x'anvato amasahe-presents difficulty and does not seem to have been satisfactorily explained so far. REICHELT in his Awestisches Elementarbuch $4991 takes x'ahe etc. as genitive of quality and translates the entire clause beginning with yim azam as 'den ich als den schonsten von der ganzen Welt eigenen sonnigen unsterblichen Lebens gesehen habe'. It will be observed in the first instance that Reichelt has not translated the word astvato of the text. Next, he seems to take the difficult portion beginning with r'ahe gayehe as genitive of quality qualifying the word aNGhu 'Welt'. Earlier, BARTHOLOMAE apparently had the same construction in view.3 LOMMEL, on the other hand, takes the clause beginning with rahe gayehe not as going with anhu, but with Haoma himself. The 'shining, immortal, life'is thus attributed not to the world' but is taken to refer to Haoma's superhuman qualities. LOMMEL's translation runs as-"Wer bist du, o Mann, der schonste, den ich im ganzen korperlichen Dasein gesehen habe, (du) von eigenem leuchtendem unsterblichem Leben (?)."5 TARAPOREWALA translates the above passage as follows: "Who O Hero, art thou, whom I see the noblest of all material creation, 1. Also his Avesta Reader, p. 96. 2. Unless he has taken astvant and anhu together to mean Welt'. 3. See his Worterbuch s. v. x'anvant (1865) and Wolff's translation (p. 30) based on it. 4. He translates sraestam (acc.) as though it was nom. 5. Die Yast's des Awesta, p. 188. Madhu Vidya/188

Page #214
TWO NOTES ON YASTS shining with (thine) own eternal life?"6 This translation, which essentially follows that of LOMMEL, suffers from certain drawbacks. In the first instance, avanvato, which is gen. sg., is interpreted as acc. sg. (xvanvantam). Secondly, TARAPOREWALA observes that wahe gayehe, both gen. sg., should have been in the instrumental case. Thirdly, one has to observe that in spite of all these changes 'shining with eternal life' does not give good meaning. One would have rather said shining with eternal lustre'. It is therefore suggested that we should supply sraestam dadarasa at the end of Zarathustra's question, and divide the passage into two parallel clauses-(1) yim az@m vispahe anheus astvato sraest@m dadar@sa, and (2) yim az@m xVahe gayehe xvanvato am@sahe (sraestam dadar@sa). The whole passage may then be translated as follows: "Who art thou, o man, whom I see the best in the whole corporal creation, (whom I see the best) in your own shining and immortal species ?" The translation given above gives good meaning. It brings out the contrast between astvant aqhu = corporal world and x'anvant gaya = lustrous world. It assumes that after Zarathustra started his question to Haoma with the words 'Ko nar@ ahi' taking him to be a human being, he realized that the lustrous person standing before him may not after all be a human being but a superhuman one belonging to the world of immortals. He therefore proceeds to characterize Haoma also as the best among the immortals. In the interpretation given above, gaya has been taken to mean creation, species, world'. This meaning is not given in the dictionary. Corresponding to Sanskrit jiva, gaya has been taken to mean only life'. But if ayhu, corresponding to Sanskrit asu, which originally means 'life' has been interpreted to mean 'material, creation, world' (when combined with astvant), there does not seem to be any strong reason why the same extension of meaning should be denied to gaya, which when combined with zvanvant, may refer to the lustrous world of superhuman beings. (2) On Mithra Yast 5.20 This passage enumerates the discomfitures experienced by those who break the contract, i.e. those who are miero-druj. The horses of such persons refuse to be mounted by them, and even if mounted they 6. Selections from Avesta and Old Persian, Part I, p. 3. 7. Op. cit. p. 17 under xvanvato. Madhu Vidya/189

Page #215
M. A. MEHENDALE do not move from their place. Some of their other discomfitures are described in the following words: apasi vazaite arstis yam anhayeiti avi-mieris frana avanam maeranam ya v@r@zyeiti avi.mioris GERSHEVITCH, The Avestan Hymm to Mithra (p. 83), translates the above passage as follows: 'Back flies the spear which the Antimithra throws, because of the evil spells which the Antimithra performs." Earlier, BARTHOLOMAE had taken frana (loc. sg. of franay meaning Fulle, Menge') to mean at the abundance of (the evil spells)'. In essence, GERSHEVITCH agrees with BARTHOLOMAE in looking upon the recitation of the evil spells as the cause of the return of the spear. He only feels that the number of spells can have nothing to do with the return of the spear, for that was achieved by the fact that they were cast. This enables GERSHEVITCH to dispense with the meaning 'Fulle, Menge'assigned by BARTHOLOMAE to the word. GERSHEVITCH explains frana as an adverbial instrumental of *frana, a derivative of fra that corresponds in form to Latin pronus, and in meaning to Lat. pro and Parth. frh'h because of'.8 LOMMEL also has followed BARTHOLOMAE as can be seen from his translation: "Zuruck fliegt die Lanze, welche der Mithra-feind schleudert, wegen der Menge boser Spruche, welche der Mithra-feind ausubt."9 The above translations are based on the wrong notion regarding the purpose of the evil spells. These spells are definitely not the cause of the return of the spear thrown by the Mithra-enemy. If that were so, he would have certainly stopped reciting them. The spear returns, just as the contract-breaker suffers from other discomfitures. The horse, the spear are of no avail to the Mithra-enemy because of his sin of having broken the contract. The purpose of the spells muttered by him while throwing the spear is to make it more effective. The spells are apparently supposed to reach the body of the enemy together with the weapon. References to the recitetion of the mantras while shooting the weapons to make them more powerful are met with frequently in the literature. cf. mahabanam raksasendrena mantritam, Ram. 6.70.21 8. The Avestan Hymm to Mithra, pp. 177-78. 9. Die Yast's des Awesta, p. 69. Madhu Vidya/190

Page #216
TWO NOTES ON YASTS 73 etc., astram... tvadvadhayabhimantritam, Ram. 2.96.50 etc. cited by Bohtlingk-Roth. If the spells are uttered to make the weapon more effective and yet in the present case the weapon turns back without inflicting any injury, then frana in this context can best be rendered by 'in spite of'. The passage then gives good meaning : "Back flies the spear which the Mitra-enemy throws in spite of the evil spells which the Mitra-enemy puts into action." This suggestion was made by me to GERSHEVITCH when I was reading this yast with him in the summer of 1958. He accepted the suggestion and recorded it in the Addenda, p. 323. He says there, "Such a meaning can be obtained e. g. by taking frana, as a preposition, to mean'in spite of', or by replacing Bth.'s ' because of the abundance of evil spells' by with the abundance......', in the sense of 'with (=despite) all the evil spells." But it is doubtful whether we can imagine for an older Indo-European language such a construction with a preposition governing genitive as is imagined above when one takes frana to mean in spite of'. In Sanskrit one may think of the use of anadstya or the genitive absolute construction to express this meaning. Probably to obviate the difficulty of the use of preposition, GERSHEVITCH thought of the latter explanation mentioned above. But in that case he is required to take 'within the meaning despite which is not likely to meet with approval. If, on the other hand, we assume that avanam maeranam is used as genitive absolute, then the idea of disregard having been conveyed by this constuction itself, frana will remain hanging without any known purpose. I am therefore now inclined to take frana as instr. sg. of a stem in -a, the derivation and meaning being the same as suggested by BARTHOLOMAE. The passage will be accordingly translated as : "Back flies the spear which the Mithra-enemy throws together with the series of the evil spells which the Mithra-enemy sets into action". The spells were thus intended to go along with the spear, but both, the spear and the spells, turn back without causing any injury to the person against whom they were hurled. The same passage frana avanam etc. occurs twice in section 21. We can interpret it in the same way as in section 20. The good throws of the Mithra-enemy, sent along with the evil spells, even if they reach the bodies of the pious, do not harm them. The wind, when it carries off the spear of the Mithra-enemy, carries off also the evil spells. Madhu Vidya/191

Page #217
AVESTAN Vahwa (n., f.) By M. A. Mehendale The word ratwa occurring in the Avesta has been assigned the meaning 'a herd, a flock, a multitude". The word occurs both as a single expression and in compounds. From the contexts in which it appears, it is clear that the meaning of the word cannot be considered doubtful. E.g. in yast 10.28 we read : aat alimai w andi dadaiti gausca vaowa viranaunia yalva asnuto bavaiti "herds of cattle (and multitudes) of heroes he bestows on the house in which he is pleased." In yasna 62.10 we read : upa Owa harsoit gaus vaiwa upa viranan pourutas "Es mogen dir zuteil Werden Herden von Vieh, zuteil eine Fulle von Mannern" (Wolff). In yaxt 8.15, 17 and 19 we get expressions like virayan raram, gavym mowam, and aspayan Owam 'a troop consisting of Jeroes', 'a herd consisting of cattle', and 'a troop oonsisting of horses'. The word voowa is thus used with reference to animals and human beings. When it is used with the expression fsaoni it can mean a group of animals of one kind or of different kinds (horses, cows etc.). In yast 9.9 we have yatha azam fsaoni vunna avabarani ari mazda damabyo "...dass ich den Geschopfen des weisen Herrn fette Herden (rather: herds of animals), verschafte........." (Lommel). Bartholomae in his Worterbuch (1435) does not give any etymology of this word. Louis H. Gray' accepts Jackson's derivation of the word from *"ene (> Skt. V van) 'strive, desire, gain' (Cf. Jackson's Avesta Reader, First Series, which is not successible to me). Gray cites, for comparison, words from the Germanic group like Goth. winja etc. 'meadow'. He adds, "...... the formation of the Avesta word implies that raowa- was THE desirable possession of a pastoral and agricultural folk. 1. Language, 25.378 (1949). Madhu Vidya/192

Page #218
AVESTAN yaowa (n. 1.) 221 From meadow' to 'herd' is not a difficult transition, nor is the reverse". Tarapore wala (Selections from Avesta and Old Persian, n. 20) also suggests the same etymology (w van 'to win, to conquer. to protect, to prosper). The above etymology implies derivation of vafwa from Vran with the primary suffix -tva. But the words derived with the primary suffix -tva in Sanskrit and Avestan are adjectives. "Als Bildungselement fur primare Adjektive ist -tra- nur fur das Induiranische sicher nachgewiesen".! Such adjectives have the meaning of gerundives; e.g. Sk. hantun, Av. jalwa '(deserving) to be killed', Sk. vaktva, Av. vax@twa 'to be spoken'. But since vabwa is a substantive and not a gerundive adjective, the above derivation remains doubtful, It is therefore proposed to derive vriowa from *trantva, a compound formed with Sk. tva 'one, some, several' and comparable in formation with Sk. drandva 'pair, couple'. In Sanskrit, when in a sentence, tva is repeated, it means, when used in the singular, "the one ...... the other', and when used in plural it means *these ...... those, some ...... some'. In such usage there is not always clear opposition implied by the use of tre ......tre. E.g. in RV 10.71.7 we read adaghinasa upakaksasa u tre lirada iva snatunt u toe dadrsre "Some of them (i.e. the sabhayal) appeared like ponds reaching upto mouth, (some like ponds) reaching upto armpit, and others like those fit to bathe in". Thus what are indicated by tve......tre are only different kinds of ponds. If an iterative compound like *fvantra is formed with trai 'some', it oun easily mean a group, a collection, a herd'. Semantically, the derivation of vagwa from *trantra therefore does not seein to encounter any difficulty. The word can refer to a composite herd consisting of 'some' animals of one kind (e.g. cows) and 'some of the other (e.g. horses); or, it can refer to 1. Alt. Gr. II, 28 526, p. 711 Madhu Vidya/193

Page #219
222 K, R, CAMA ORIENTAL INSTITUTE GOLDEN JUBILEE VOL. a herd consisting of only one kind of animals, 'some' of which may be of one type and 'some other (showing the difference in colour etc.). Phonetically, the derivation of taowa from *trant va faces some difficulty which, however, is not unsurmountable. Avestun ow regularly corresponds to Skt. tv (Jackson, Avesta Grammar $94). Medial ow of zagwa is thus well explained. But in the initial position, tv should have shown 0w and not 2, e.g. Skt. tram: Av. Quam. We, therefore, expect to get awabwa and not valua if we wish to derive the word from *tvantra. But this difficulty can be explained in the following way. Double treatment of a comparable cluster in the initial position is seen in Avesta in the case of the cluster dv. It shows both dy and b., e.g. Av. dvarom: Sk. dvaram, Av. bitim: Sk. dvitiyam. Similarly, tv can be looked upon as showing two treatments initially, ow and v. To explain vaswa we have only to assume that the simplification of the initial cluster 6w to v was due to dissimilation (to avoid repetition of ow in two successive syllables). The initial v may have the phonetic value of Avestan v (corresponding to Skt. r). Or it may only be a graphic representation of the voiced spirant w. Occurrences of v for w in the internal position have been already noted by Jackson S 87. He also notes one instance of initial v for w, cf. vaeibya 'with both' for wazibya = uwaeibya : Sk. ubliabliyam. If in catwa initial v is only graphic, then rata = wgowa < *jualwa < *tvantra. Madhu Vidya/194

Page #220
On the Significance of the Name Zarathustra -M.A. Mehendale The significance of the name of the founder of the Zoroastrian religion, which appears as Zarathustra in the Avesta, has been much discussed. The most recent explanation is that of Prof. H.W. Bailey' who observes : "So many attempts have been made to explain this name without a generally accepted solution that it may seem bold to try once more." His explanation will be referred to later in this paper. Bailey feels that his explanation is likely to be more acceptable than the rest. Even so the present writer may be excused for suggesting one more solution to the problem. A general start to the attempts to explain the prophet's name seems to have been given by such Sanskrit proper names as Jarar-karu and Jaral-karna which contain the word jarat as the first member of the compound, and such Iranian proper names like Hitaspa and Yurtaspa which contain the name of an animal (horse) as the second member of the compound. Hence Bartholomac , among others, looked upon zarathustra as a compound formation consisting of two members, *zarant (Sk. jarant) 'old' and usira (Sk. ustra) 'camel', the name of the prophet thus meaning 'one whose camels are old (des Kamele alt sind)'. But scholars felt that a name having this import is not the one which can be looked upon as a suitable name for the prophet. Hence attempts were made to give a different meaning either to the first or the second member of the compound, or to both, but retaining mostly the analysis of the name as a compound form of two members as noted above. Thus, for example, M. Haug in his Gathas Vol. II, pp. 245-246, f.n. 1 (1858) has suggested to equate zarad with hrd 'heart' and ustra with uttara 'higher, excellent', the compound then meaning 'one who has an excellent heart (der ein treffliches Herz hat)', or to connect zarath with Sk. Vir 'to sing', the compound then meaning the excellent singer of praises, most excellent poet (der treffliche Lobsanger)'. Haug prefers the latter explanation as, in his opinion, the singing of praise songs plays an important role in the Gathas and Zarathusbtra himself appears in the Gathas as a poet. In 1862, in his book Essays etc., pp. 252-253 Haug gave up the above explanation. There he calls attention to the fact that the prophet is also known in the Avesta as Zarathustrthiamo which, as a superlative form, must mean 'the greatest or highest Zarathustra'. Haug argues that this denomination can have sense only when it is assumed that there were several Zarathustras and that the prophet was the Madhu Vidya/195

Page #221
114 Indo-Iranian Literary & Cultural Traditions best among them. "The name 'Zarathustra', therefore, must have conveyed in ancient times about the same sense, as the word Destur now-a-days, meaning the spiritual guide and head of a whole district, or even province. The Zarathustrotamo then is to be compared with the Destur-i-Desturin or Chief High priest......" Writing about the significance of the prophet's name, I.J.S. Taraporewala disapproves of the meanings in which ushtra is taken to mean 'camell, and observes : "Far beiter is the suggestion made by some scholars deriving the name from zararha (golden) and ushtra (light), from the root ush "to shine". Thus this designatioa of the Prophet would mean "He of the Golden Light", which is just the appropriate name to be given to one of the Greatest of the Light-Bringers of the world". Bailey, whose article is mentioned above, proposes to establish a verbal base Av, zar, Skt. jar 'to move' and on its strength explain the prophet's name as 'one who can manage camels'. He compares with this name other proper names like yuxraspas 'he who has harnessed horses' and hitaspa 'he who has put horses into a team'. He believes that understood this way Zarathustra can be looked upon as quile a respectable name, for, "This skill among a nomad people deserved the respect its use in a proper name implies." He rejects Bartholomae's explanation, referred to above, in which zarat is taken to mean 'old' for he says "...it is hard to conceive why a mun (or a boy) should be named froin the possession of zarant -- 'old, decrepit' camels. With zarat-moving, driving, controlling' the name has an important meaning befitting the local life". Bailey's explanation thus retains the meaning of the word ustra 'camel', but changes that of the word zarar. The meaning obtained by him does give some respectability to the prophet's name. But even with this new meaning the name appears to be somewhat of a general nature and has nothing specific in it to be looked upon as suitable for the prophet. The explanation attempted in this paper, however, has a direct bearing on an important aspect of the prophet's teaching which must have brought about economic, as well as religious revolution in ancient Iran. It is well known that killing cows and oxen and offering them in sacrifices were practised by Indo-Iranians as acts of religion before the days of Zarathushtra.' It is against these practices that Zarathustra speaks very vehemently in the Gathas. He exhorts people to take good care of the cattle and offer them protection. He says that giving fodder to the cow and offering worship to Ahura Mazdah are the best things for any one to do (Yasna 35.7). It is hardly necessary to cite the different passages from the Gathas in support of this contention. Zarathustra, thus, appears in his new religion in the role of the protector of the ox'. If we remember this as one of the striking features of his new religion, it would be easy for us to recognise the verb tra -'to protect' in the final syllable of the prophet's name. And since we know that the word 'uxsan-(Skt. uksan) 'ox' . is well attested in the Avesta, it should not be difficult for us now to guess that ustra, occurring in the prophet's name, has nothing to do with camels, but is to be Madhu Vidya/196

Page #222
On the Significance of the Name Zarathustra 115 derived from * uxstra meaning 'protector of the ox'. The name Zarathustra would, in that case, mean 'the old protector of the ox'.' There is hardly anything to object to this interpretation from the point of view of the meaning arrived at. Of all the interpretations of the name suggested so far this is the one which would suit the prophet extremely well. There is some controversy on the point whether this name was given to the prophet by his parents in his childhood or whether he adopted it for himself later in his life after he started preaching his new religion. Jackson held the former view and said that although the name, as understood in his time, was unromantic and unpoctic, it was retained by the prophet "as his birth-right even after he became famed as a spiritual and religious teacher. The very fact of his retaining this somewhat prosaic appelative testifies to a strong personality ; Zoroaster remains a man and he is not dubbed a new with a poetic title when later sanctification has thrown a halo of glory about his head."11 On the other hand, Taraporewala, whose view has been referred to above, believes that Zarathustra is the title by which the prophet was known after he had proclaimed his message.1: The interpretation of Zarathustra suggested in the present paper would also lead us to reject the view that this was the prophet's first name. It could come to him only after he had started proclaiming his new message of ox-protection (ustra) and while doing so had become somewhat old (jarat ). It is likely that he did not himself adopt this name but that it was given to him by others--perhaps his revilers as indicated by a reference to his old age. It is, indeed, possible to discover a piece of evidence testifying to the prophet's having become old while carrying on his struggle to save the ox from the tyranny to which it was subjected. This evidence is to be found in the famous Yasna 29, known as the Gatha of the Ox-Soul. The ox complains about the ill-treatment given to him and the cow by the people. He wants to know who created him and for what purpose. If he was endowed with life by the Creator, he had a right to live, and, as an animal helpful to men, a right to seek protection. The wise lord then replied to the ox that the Creator had fashioned him for the milker and the herdsman (and apparently not for the sacrificer). But no one was appointed till then to take care of him. The Good Mind then pointed to Zarathustra and said: "I know but this one : Zarathustra Spitama, the only one who has heard our teachings; He will make known our purpose ---Sweetness of speech shall be given to him." But the ox is not satisfied. He continues to lament that he has been handed over to a powerless man and not to a strong guardian. In Yasna 29.9, the Ox-soul moans : "That I should have to be content with the powerless word of a man without strength for a guardian, I who wisb for a strong master;" What comes after this reads as follows in the original : kada yava hvd anhat, ya hoi dadat zastavat avo which has been rendered as : "Will he ever be, he who shall help him with his bands 2:14 Madhu Vidya/197

Page #223
116 Indo-Iranian Literary & Cultural Traditions In the above translation the two opening words kada yava are taken together to mean 'when, if ever', yava is interpreted as instr. sg. of yay 'Dauer, used as an adverb.16 But this seems to be the only place where kada and yava arc used together. Humbach" asks us to compare kada yava hvo anhat with yada hvo anhat occurring in Yasna 31-16. But this will show that in our passage we should construe only kada with hyo anhat "when will he be ?" The remaining word yava, then, can be interpreted as nom. sg. of yavan, yvan, 'young man, young hero' which is attested in the younger Avesta. The line cited above will thus mean: 'When will that young man be who will give us help with his hands ?" This meaning will suit the context because the Ox-Soul is asking for a powerful man with physical strength as his guardian. If this interpretation is correct, the word yavayoung man' will indicate that the prophet had become old when the Gatha was composed and that he had till then not succeeded in persuading the people to accept the change in the religion as preached by him. Now we are ready to face some other questions before we can admit usira 'protector of the ox' as a compound of uxsan and a form from the verb tra. Such a compound normally should have been uxsathra, and the prophet's name, then, would bave been Zarathuxsathra, 17 First, we will look to the disappearance of * before s in usan for uxsan and in this regard note that a few possible examples of the disappearance of x before s have been recorded. E.g. we get avasata 'he spoke' for *avaxsata<*a-vak-sa-la." Bartholomae'' thinks it possible that x before s bas been lost in the form visano (Acc. pl. of visan) < *vixsand (visanPage #224
On the Significance of the Name Zarathustra of other scholars. One such is that of Burnouf who understood Zarathustra to mean 'having yellow camels'. For other explanations, some of which show "a good deal of fancy", also cf. A.V. Williams Jackson, Zoroaster, the Prophet of Ancient Iran (1901), pp. 12-14 and Appendix I, pp. 147-149. Also cf. John W. Waterhouse, Zoroastrianism (1934). pp. 2-3. 4 The Religion of Zarathushtra, pp.23-24 (1926). Also cf. R.P. Masani, The Religion of the Good Life, Zoroastrianism, London, 1954 (2nd Edn.), p. 36. 5 Bailey cites the form as yuxta.aspa. As a proper name, Bartholomae gives it as yuxtaspa which is adopted in the text above. 6 Tr. Phil. Soc, 1953, p. 41 and f.n. 3. 7 B. Schlerath, Opfergaben, Festgabe Fur Herman Lommel, pp. 129 ff. 8 "Die Pflege und Schonung des Rinds.........steht im Mittelpunkt der zarathustrischen Lehre". Bartholome, Worterbuch, 509. 9 or, 'the protector of old ox', but this is clearly not likely. 10 It seems pertinent here to cite the following passage from J. Duchesne-Guillemin's The Hymns of Zarathustra (Eng. Tr. by Mrs. M. Henning), pp. 5-6: "The society in which Zoroaster lives and preaches is a pastoral society, not yet settled on the land. He teachesas Nyberg has well pointed out-the fertilization of the meadows which makes permanent settlements possible. But these must be defended against the raids of the nomads by force of arins. The nomad is a thief of cattle, which he sacrifices and eats. The good deed is to be summarized shortly as the care and the defence of the cattle, to which is added the duty of extending the area of fertilized meadows at the expense of the nomad. "This emineatly practical and earthbound aspect of Zoroaster's programme is not always recognized. Thus the latest Parsee interpreter of the Gathas refuses, for reasons of piety, to admit that such a trivial thing as cattle-raising could be mentioned in sacred hymns. Therefore, he concludes, it can only be by allegory. In this way some vivid texts are emptied of their sap, and we are supposed to accept Zoroaster as a dreamer or a pure mystic". Thus, for example, Taraporewala in his book, The Divine Songs of Zarathushtra, p. xiii, urges us to interpret the words for 'cattle', 'fodder' and 'pastures' in the Gathas 'in a higher sense'. He takes the word gau to stand for Creation and more especially Humanity, sometimes also Mother Earth", (p. 36). 11 Jackson, op. cit., p. 14. 12 The Religion of Zarathushtra, p. 23. 13 This and the following are translations by Duchesne-Guillemin, op. cit., p. 61. 14 H. Humbach, Die Gathas des Zarathushtra, I. 82 puts it as 'wann jemals wird der zur Stelle sein, der ihm Hand und Hilfe geben wird ?'. 15 Bartholomae, Worterbuch, 1264-65. 16 Op. cit., II. 17. 17 I am not referring here to the difficulty regarding the occurrence of ein zarab for this has been already noticed earlier. Scholars are agreed that zara@ustra is a secondary development out of "zarat.ustra. Haug in his Gathas II. 246, end of the footnote 1, explained the development of r toe due to the following vowel . Bartholomae, Gr. Ir. Phil., I. 1.182 notes also other examples of replacing . Bailey thinks it possible that this is a case of U replacing 8, TPS 1953, 41. 18 Jackson, Avesta Grammar, y 187 (5), p. 58. He also notes the loss of x before i la tuirya for xtuirya. 19 Gr. Ir. Phil., I. 1.149, 8 264, note 2, and Worterbuch, 1472-73 20 Journal of Near Eastern Studies, xxili (1964), p. 38 fur vergl. Sprachf. xxvi (1883); p. 604], the Old zarathustra. According to Gersheivitch, the Median form was *zarat.ustra 21 Wackernagel, Alt. Gr. II. 1 & 26, pp. 64-66 22 Given in the Sakandhvadi gana on the Varttika on p. 6.1.94 117 According to Hubschmann, [Zeitschrift Persian form of the prophet's name was Madhu Vidya/199

Page #225
BRIEF COMMUNICATIONS ON YASNA IX.26* The text of Yasna IX.26 runs as follows: fra te mazda barat paourvanim aiwyanhanam stahrpaesanham mainyutastam, vaNGuhim daenam mazdayasnim. aat ainhe ahi aiwyasto barasnus paiti gairinam drajanhe aiwisaitisca gravasca mgorahe. The first half of the verse presents no difficulty. It informs us that Ahura Mazdah gave Haoma a girdle which was decorated with stars and was fashioned by the spirits (mainyu); he also gave him the good religion of the Mazdah worship. The latter half of the verse contains the words aiwidaiti and grab or grava, which require some comment. Bartholomae (Worterbuch, col. 91) took aiwidatay to mean 'word', comparing the Avestan form with Sanskrit abhidhana 'name, title; word'. But later, following Geldner, he gave up this meaning and preferred 'shelter' (Schirm); see Zum altiran. Worterbuch, p. 176. Geldner himself, in a paper not accessible to me, rendered (with a question mark) the word as 'covering' (Hulle). Lommel (ZII.3, p. 170) believed that aiwidaiti referred to some article of human dress, possibly (allenfalls) of leather, and he imagined it in the present case to be fjerkin, doublet' (Koller). K. Hoffmann (W. B. Henning Memorial Volume, p. 199), on the other hand, thinks that aiwidaiti means 'halter and rein' (aissisaiti- durfte das sein, was man einem Pferd anlegt', also Halfter und Zugel). The word aiwidaiti, as has been long recognised, is to be derived from aiwi + da., Sanskrit abhi +dha- 'to join, to bind; to yoke'. Sanskrit abhihita means 'joined, combined; yoked'. If the same meaning be assumed for the Avestan aiwi + da-, the derivative aiwidaiti could mean 'joining together, closely combining together' and, as related to mqora it could mean 'the recitation of the mantra in which the successive words are joined together, are closely combined together without allowing a . pause between them'. This kind of recitation will resemble the sarhita recitation of the Veda in which the words of the mantra are pronounced in close proximity. The word sarhita is derived from the same root dha- which appears in the form da- in aiwidziti. The difference between the two words lies in the use of the prefix. But we may compare with Av. aiwidaiti also Sanskrit abhinihita which is used for a kind of vowel sandhi in which the initial vowel a of the following word is joined so closely with the final vowel e or o of the preceding word that it gets merged into it. The word abhinihita contains the same prefix abhi which is found in aiwidaiti. On Indo-Iranian Journal 22 (1980) 137-142.0019-7246/80/0222-0137 $00.60. Copyright (c) 1980 by D. Reidel Publishing Co., Dordrecht, Holland, and Boston, U.S.A. Madhu Vidya/200

Page #226
138 BRIEF COMMUNICATIONS account of the additional prefix ni in it, the word abhinihita expresses the idea of 'merging', which is stronger than 'close combination expressed by the terms samhita Oi aiwidditi. The Avestan word gravas was understood by Bartholomae (Worterbuch, col. 529) as acc.pl. of grab 'sentence' (Satz). But, again, following Geldner, he chose to take it as nom.sg. of grava 'stick'. He, however, preferred to translate gravas in our passage as 'support' (Stutze) instead of adopting Geldner's rendering 'Stab'(Z. altir. Wb. p. 176). Lommel (ZII 3. p. 170) accepted the interpretation of Geldner but preferred to look upon gravas as acc.sg. of an -s stem. K. Hoffmann (Henning Volume pp. 197-198) argues for the original' reading *gravusca<*gravasca in place of gravasca. He looks upon the form *grava as acc.pl. of the a stem. As to the meaning he is prepared to accept Geldner's suggestion 'Stab' but thinks that the original meaning of the word was 'handle' (Griff). But, as shown above, if aiwidaiti could mean 'close recitation (of the words of the mantra)', the possibility of gravas referring to something similar in nature, i.e. to some kind of recitation of the mantra, has to be considered. It is well known that the Vedic text is recited in the samhitapatha, as mentioned above, and also in the padapatha, i.e. by separating the padas 'the words' from one another. This separation of the words in the padapatha was apparently known as vigraha. Further, within a pada, analysis of complex forms into its constituents was shown by observing a pause between the constituents. This separation is known as avagraha. Whitney on the Atharva-Pratisakhya 4.78 observes: Vigrhya denotes a word which is altogether independent, and therefore disjoined from others in the pada-text, a nanapada, in distinction from avagrhya, which means 'divisible into its constituents (purvapada and uttarapada), as a compound'. According to the Taittiriya-pratisakhya 22.13, which calls these pauses by a common name virama, the pause between the two padas, viz. the vigraha, is of the duration of two morae (dvimatrah), while the one between the two constituents of a pada, viz. the avagraha, is of the duration of one mora (ekamatrah). It may further be noted that in Sanskrit we have the words pragraha and praghya which are used with reference to the vowels pronounced separately i.e. which are not subject to the rules of euphonic combination. Avestan gravas can be taken as an acc.plur. of a root noun grab (Bartholomae, Worterbuch) or of an a-derivative grava (Hoffmann, Henning Memorial Volume). In both cases it belongs to the root grab., corresponding to Skt grah-/grabh-. Just as the Sanskrit derivatives vigraha, avagraha and pragraha indicate different kinds of 'separation', the Avestan noun can mean 'holding apart, separation (of words and their constituents in recitation), observing pauses (between words and their constituents in recitation)'. Madhu Vidya/201

Page #227
BRIEF COMMUNICATIONS 139 The word drajanhe in this passage has variously been interpreted. Bartholomae (Worterbuch, col. 774) first interpreted it as an infinitive of drag- 'to hold' which meant to hold fast, to preserve (fest zu halten, zu bewahren). Later, following Geldner, he took it to be the dat. sing. of drajah 'stretch or length of time' with the meaning 'for all time, for ever' (fur alle Zeit). Lommel (ZII. 3, p. 170) took it as 2nd pers. sing. middle of drag- 'to hold'. I would agree with Hoffman in interpreting as Bartholomae originally did, i.e. as infinitive of drag. The various translations proposed are: "Und mit diesem umgurtet bist du auf den Hohen der Berge fur alle Zeit Schirm und Stutze des heiligen Worts' (Wolft). 'mit diesem bist du umgurtet auf den Hohen der Berge; du ergreifst (rustest dich mit Koller(?) und Stab des Worts' (Lommel). um zu halten die Zaumungen und Griffe des heiligen Spruches' with the explanation 'Damit wurde dieser Aussage das Bild von Wagenlenker und Pferd zugrunde liegen: Haoma, dem Ahura Mazda cincn Sternengurtel verliehen hat, lenkt den heiligen Spruch wie ein Pferd. 'Griff" durfte dann gegenstandlich etwa 'Griff der Zugel', d. h. der Teil der Zugel, den man ergreift, bedeuten' (Hoffmann). It is, however, to be doubted whether one has to look for a metaphor of a charioteer and a horse in the above expression, to which Hoffmann has been led by abhi dha'to yoke'. It has already been shown above to what meaning we are led by abhi dhato join, to bring in close connection'. It is not probable either, that aiwidaiti and grava both refer to 'Zugel', once in general and a second time to a particular part of it. The latter would have been enough. I would translate the passage as follows: Since then girdled by it you are (seated) on the high summits of the mountains to preserve the close combinations (i.e. recitations without pauses between the words) and the separations (i.e. recitations with pauses between the words and their constituents) of the mantra. Deccan College, Poona M. A. MEHENDALE * Paper read in the Iranian section of the 28th session of the All-India Oriental Conference held at Dharwar, Karnatak, in November 1976. Madhu Vidya/202

Page #228
SOME REMARKS ON MIHIR VAST (YAST X) By M. A. MEHENDALE. (1) radwya ciora hacimno In Yast X. 67 we are told that Mira comes, driving in a chariot, from the eastern continent Ar@zahi to the splendid continent Xanirada. At the end of the verse occur the following lines: radwya Fiora halimno armaNGhaca mazdabata verbrainaca ahurabata These lines are rendered by Gershevitch' as follows: "(Miora comes) equipped with prompt energy, Mazdah-created fortune, and Ahura-created victoriousness," This would mean that Mi@ra, when he comes to the people of the splendid continent X'anirafa, brings with him three things, (i) energy, (ii) fortune, and (iii) victoriousness. But the use of ca only twice, with aranah and var@oraTua, and the close parallelism in the last two lines, where we find the use of very similar attributes mazdabata and ahurabala, will indicate that Mira comes equipped with only two things, fortune and victoriousness, and not three, As regards the reading tiora, Gershevitch notes that it was Geldner's emendation which he himself abandoned in the Addenda to his edition in favour of taxa (Commentary, p. 217). The manuscripts give cixra, rendered by caxia Bartholomae as Tatkraft, Energie', or caxrawheel. Gershevitch accepts the reading ixra, only because it is a lectio difficilior', and the meaning assigned to it by Bartholomae and translates the above lines accordingly. He is, however, required to give somewhat unusual meaning prompt to rawya in order to make it agree with energy'. It seems preferable to accept Geldner's reading taxra and interpret it as standing for a time-cycle. The line rabwya taxra hakimno may be translated as "(Miora comes) associating himself, i.e. according to, the circle of fixed time. " The idea conveyed is that Mira visits the continent X'aniraba at appointed times which are looked upon as moving in a circle, (2) alpi v70ili jala 1. I. Gershevitch: The Avestan Hymn to Mithra, Cambridge, 1959. 2. One ms. has also caerahe. Madhu Vidya/203

Page #229
MEHENDALE M. A.: Somc Remarks on Mihir Yasl (X) 165 In Yast X, 80 we read Owa paiti zi haxadrom daide vahistam vorafrainomca ahurada tam yahmi soire mibro-drujo aipi vioisi jata panrva masyakalho " With you as master it (=the community ) obtains the most excellent succession and Ahura-created victoriousness, (because in it (lit. in which [comin11nity ] ) the many men false to the contract are floored ( lit. lie ), struck at the divinatory trial." (Gershevitch ). Construing yahmi with haxsara and vararayna is natural enough and that is what most of the translators quoted by Gershevitch in luis Commentary (p.230 ) bave done. But Gershevitch docs not follow them because, according to him, in that case "one has to put up with the strange idea that the 'diviuatory trial', is apparently organized by 'victoriousness :'" (p. 229). Gershevitch, as well as the earlier translators, do not bring out the siguificance of nipi in their renderings as they seem to look upon it as a preposition mcaning 'at, in '. Now there seems to be nothing strange 'in saying that the contractdeceivers are floored when they are opposed by the community which, by its policy of non-deceipt, has secured the best companionship of Midra and, consequently, the Alura-created victoriousness. Next, it is possible to render nipi as 'also 'aud translate the above passage as "with you as master (the community of 11011-deceivers ) obtains the best companionship and the Ahuracreated victoriousness in which many men who are contract-breakers lie ( on the ground ), (as) also in the divinatory trial ( those ) struck down ( lie on the ground )." This means that the contract-breakers are floored on both the occasions, many on the battle-field where victory is won by the nou-cleceivers of the contract, and (the individuals ) in the divinatory trial where they lose dite to their falsehood. (3) dvucina pioc hacimna In Yast X.84 wc read yim dvucina pile hacimna bala ustanazasto zbayciti araine, yim driYuscil asotkalso apayato havais datais Madhu Vidya/204

Page #230
166 VSM : Golden Jubilee Volume bada ustanazasto zbayciti avainhe "[Whom ) sometimes she (viz, the cow ) who longs to be milked as she feels (lit, is with the pain of swelling, (invokes for assistauce, with outstretched hands ), sometimes also the pauper who follows the doctrine of Truth but is deprived of his rights; " (Gershevitch ). As regads the earlier translations of the difficult line Gershevitch pronounces the judgenient: "None of the translations suggested for dvucina pioc hacimna inspires confidence." (Commentary p. 230). As regards his own rendering Gershevitch is aware that it is based on "guess-work." In spite of G.'s comments it still seems preferable to follow Bartholomae and translate yim dvacina pibe hocimna......zbayeiti avainhe as 'whom any two coming together for protection call (s) for help.' 'Any two' may refer to an association like that of a warrior and liis charioteer, or that of a warrior and his priest. But the actual call for help inay go out from only one of them and lieuce the use of the singular forms ustanazasti and zbayeiti. The author of the stanza first refers to large associations in which influential persons like vispaiti give the call for help for themselves and for mauy others of whom they are the lead; next, lie passes on to smaller associations of any two personis of whom one may be superior and the other not quite so; in the end he comes to a pauper, who is nobody's head (and nobody's superior ), a lone person, who, all the same, feels entitled for help as he is the follower of the doctrine of Truth. Madhu Vidya/205

Page #231
TWO NOTES ON YASNA 10 M. A. Mehendale Yasna 10.17 reads: vispe haoma upastaomi yaecit azahu daratanho jaininam upadarozahu "Alle Haoma's preise ich,...; auch die, welche, in Gefangenschaft gehalten, sich in der Fesseln der Weiber befinden". Wolff. What is meant by the Haomas held in captivity (or narrowness) with the chains (or in the bundles) by women is not clear. It is generally believed that the jani of 7.10.17 is the same as the mairya jani of 7.10.15. She is a roguish woman who makes use of Haoma for illegitimate purposes. But this can hardly be the correct interpretation since Zarathushtra is not likely to think of praising (upastaomi) such Haomas. Two suggestions are offered here to explain the above passage. (1) It is possible that the word janayo refers to the 'fingers' which hold the Haoma stalks. In the Rgveda, the fingers, in relation to Soma, are often spoken of as the (ten) 'sisters'. In many passages, however, the verbs used with them are expressive of some kind of forward movement, e.g. hinvanti" ; hinvantis; ahesata*; ajantis. But in two passages, the "holding" or the "restraining" of Soma by the 'sisters' is, indeed, referred to, e.g. a grbhnanti" 1. "die in der Enge gehalten sind in den Bundeln der Frauen". Lommel. 2. 9.26.5. 3. 9.65.1. 4. 9.71.5. 5. 9.91.1, 6. 9.1.7. Madhu Vidya/206

Page #232
262 RTAM and yatahl. Hence it is possible to suggest that the word janajo in the rasna? does not refer to real women but refer, figuratively, to fingers. The Haoma stalks held tightly by fingers are apparently spoken of as the ones held in captivity by women. (2) Alternatively, it is possible that the tight holding of the Haomas has a reference to the practice of the tight binding of the Soma stalks. This is referred to in the Satapatha Brahmanas. After the description of the purchase of the Soma stalks for a Soma sacrifice, we read : atha somopanahanasya samutparyantan usnisena vigraihnati ... (18). atha madhye 'nguljakasan karoti ... tam ayativa va enam etat samayacchann apranam iva karotistasyaitad ala eva madhyatdk pranam utssjati(19). "Having gathered up the ends of the Soma-cloth, he (the Adhvaryu) ties them together by means of the head-band... (18). "He then makes a finger-hole in the middle of the knot), ... for, in compressing the cloth), he, as it were, strangles him (Soma and the sacrificer) and renders him breathless; hereby now he emits his breath from inside, ..." Eggeling. In the light of this description it is likely that the word janayo of the Yasna refer to the ends of the cloth. We have to presume that the Avestan word for the end' of the cloth was in the feminine gender, like Sanskrit dasa. If this assumption is correct, the Haomas held tightly by women' would be those which were tied up with the ends of the cloth. It is conceivable that Zarathushtra praises such Haomas. (2) Yasna 10.14 reads : ma me yatha gaus drafsa asito varam acaire frasa frayantu te mado varazyanuhanho jasantu 1. 2. 3. 9.28.4. 10.17. 3.3.2.18-19. 10.17 In the Sat. Br. passage above, however, the word for the end of the cloth is anta (m). Madhu Vidya/207

Page #233
TWO NOTES ON YASNA 10 263 "Nicht sollen sie mir beliebig wie das Stierbanner sich einherbewegen, (wenn) sie (dich) geniessen; stracks vorwarts sollen sie gehen, (die) sich an dir begeistern; mit energischem Schaffensdrang sollen sie sich einstellen". Wolff 1 The above rendering does not seem satisfactory. The one who is drunk is himself likely to falter and not make some one else falter. The word drafso is in the nominative and not accusative. And the connection between the two sentences also is not clear in the above rendering. Prof. H. P. Schmidt has kindly supplied me with a completely different interpretation of the above passage offered by J. Kellens in his book Les noms-racines de l'Avesta, (Wiesbaden, 1974). This interpretation runs as follows : "when you are lying in the strainer", do not stagneto for me like the drop of milk (which, when coagulating does not pass through the strainer) ; let your intoxicating (jets) advance straight, let them come with a vital energy force". This is, no doubt, a better interpretation of the passage. But it seems possible to modify a little Kellen's interpretation. It is difficult to imagine that the poet here thinks of a drop of coagulating milk passing through a strainer as the object of comparison. That would be a very special case. It is more likely that the poet thinks of the drop of milk which does not easily pass through the cow's teat when she, for some reason, holds back the flow of milk. The Haoma juice that does not pass easily through a strainer can be compared with a drop of milk that does not come through the teat. 1. "Nicht soll mir wer davon getrunken hat, wie das Stierbanner nach Belieben (?== schwankend?) sich einherbewegen. Wunderbar mogen herbeikommen deinc Rausche, wirksam mogen sie kommen". Lommel. 2. The book is not accessible to me. 3. asito understood as in asito-gatu. 4. Cf. Skt. vara 'strainer, hair-sieve'. 5. acuire loc, sg. of acara 'non-movement', ma... acaire lit. 'may you not be in immobility": 6. Opc reason could be that she want to feed the calf first, Madhu Vidya/208

Page #234
Some Remarks on Yasna 34 M. A. Mehendale (Bhandarkar Oriental Research Institute, Pune) 1. Yasna 34.2c: It has the expression pairigaeo e xsmavato vahme which has been rendered by Humbach (1959) as 'bei der die Herden ubermittelnden Verherrlichung von euresgleichen'. On the strength of certain parallel passages Humbach believes that pairigacea- stands for something like pairi.da.gaela-or *pairi.data.gaeDa- meaning 'den lebenden Besitz ubermittelnd'. Insler (1975), on the other hand, feels that pairigaeba. means universal'. He compares with this word the Vedic expression parijman- which, according to him, means faround the earth, universal'. He therefore concludes that pairigae0a-, which literally means around the creatures', also means universal'. He translates the above expression as 'in universal glory (p. 55) or praise (p. 221) of your kind'. The context in which the above expression occurs is of a glorification (vahma-), accompanied by praise-songs, performed by an individual, and yet, apparently, many persons are concerned in this act of glorification as is shown by the plural form ahmi by us' in the preceding stanza.' Hence it seems better to interpret pairigacOa- as 'surrounding the living persons in the household', i.e. (a glorification) in which all the members of the household participate 2. Yasna 34.Sab: kat v@ xsaram ka istis Jyao Danai mazda yaea va ahmia, asa vohu mananha Orayoidyai drigum yusmakam. Insler's rendering runs as: "Have ye the mastery, have ye the power, Wise One, for the act to protect your needy dependent - as I indeed am - with truth and with good thinking?" He thus takes the parenthetic phrase yaba va ahmi with reference to drigum yusmakam. He connects asa vohu mananha with Orayoidyai, for he does not approve of treating it as commitative instrumental with vocative mazda. It seems, however, preferable to relate the parenthetic phrase with those words between which the phrase stands, viz. Syaoeanai and asa vohu mananha, and not with those which stand away from it. This also means that one connects asa vohu mananha with syao Danai. What is meant is that Zarathustra wants to know what kind of might, what kind of power Ahura Mazda has for (1) the performer who acts (lit. an act performed) in accordance with truth, and with good mind, as indeed, Zarathushtra himself is (yaoa va ahmi), i.e. a performer of a truthful act done with good mind, and (2) for protecting those who are dependent on Ahura Mazda. 3. Yasna 34.9ab: yoi sp@ntam immaitim ... dus.SyaoBana avazazat. Humbach and Insler, following Bartholomae, treat dus.Iyao Oana as an adjective and render it, respectively, as 'die Ubeltater' and 'those of evil actions' (nominative plural masculine). But it is better to Madhu Vidya/209

Page #235
262 treat it as a noun 'evil action' and render the line as those who have abandoned virtuous piety by evil action'. This interpretation is recommended, on the one hand, by the close proximity of dus. Syao Dana and avazazat, and, on the other, by the fact that dur.Syao Dana is thus contrasted with vanhaus mananho fyao Bana with (good) act stemming from good thinking' of stanza 10. Since stanzas 9 and 10 are related in that the activities of the bad and the good are contrasted, it is possible to imagine dus.xratu-, as opposed to huxratu- (st. 10), as the subject of 9 ab: yoi = dus.xratavo) spontam armaitim ... dus.Syao Dana avazazat. 4. Y. 34.9c: aeibyo mas asa syazdag. Humbach treats asa as nominative plural and translates, von denen sollen sich schnell die Wahrhaftigkeiten entfernen'. Insler, however, treats asa as instrumental singular and, accordingly, translates the passage as 'from them one has certainly retreated in accordance with truth'. It is, however, worthwhile to consider if armaiti can be the subject: (those of evil intentions who have abandoned piety), from them (piety) shall quickly retreat along with truth'. In stanza 10 piety is said to be hieam asahya Genossin der Wahrhaftigkeit, and hence it is reasonable to suppose that when she recedes from the perpetrators of evil actions, she does so along with truth. When evilthinkers abandon piety, piety too, along with truth, leaves them. 5. Yasna 34.13ab: tam advanam ahuri yam moi mraos vanhaus mananho, daena saosyantam ya hu.karata ascit urvaxsat "die Bahn, o Lebensherr, die du mir als die des Guten Gedankens genannt hast, die gutgebahnte auf der die Sinne der Kraftspender mit Wahrhaftigkeit wandeln" (Humbach). Humbach thus treats daen, as the subject and hu.k@r@ta as referring to advan. Insler, however, reads yahu karata and treats karata as nominative singular masculine of karatar- 'extoller, commemorator' and daens as accusative plural feminine to which the relative yahu refers. He translates: "To that, Lord, which Thou hast told me to be the road of good thinking, to the conceptions of those who shall save, along which Thy extroller shall proceed in alliance with truth indeed..." With karatar "extoller' Insler compares carak@rabra-'hymn of praise' (Y. 29.8c). It seems possible to connect these two words etymologically with Vedic kiri- (kirin- ?). Notes Also cf. dama (plural) in stanza 3, and gaeovispd in the same stanza which is nominative plural according to Insler. 2 Humbach prefers to read hahmi wie wenn ich schlafe'. Insler, however, interprets this as referring to Ahura Madza. On the doubtful nature of the stem kirin., cf. M.A. Mehendale, BSOAS 37.1974:670-671. References Humbach, Helmut, 1959. Die Gathas des Zarathustra, Band I, II, (Heidelberg: Carl Winter). Insler, S., 1975. The Gathas of Zarathustra. (Leiden: E.J. Brill). Madhu Vidya/210

Page #236
Avestan garad- and garazdi M. A. MEHENDALE In the Gathas we find garazdi- attested only twice. BARTHOLOMAE (524) assigns to it the meaning taking possession; getting, obtaining (Inbesitznehmen; Bekommen).' He derives g@r@zdi- from garad- (514) to which he assigns the meaning to take a step (gradi)'. BARTHOLOMAE does not give any Sanskrit cognate. In Y.50.9 we read: at hudanaos isayas g@r@zda hyam. HUMBACH assigns the meaning 'favour, grace (Huld)' to g@r@zdi and translates the passage as "furnished with refreshment may I therefore be in the favour of a blessed one". (Mit Labung versehen moge ich mich daher in der Huld eines segensreichen befinden). HUMBACH, however, states in his Kommentar on 50.9 that 'Unklar is g@r@zdi'. INSLER follows BARTHOLOMAE in deriving garazdi- from garad- but assigns to it the meaning 'stride'. He translates: 'then I would, excercising such power, be in the stride' of the blessed one. He feels that the sense 'step, stride', like Latin 'gradus' is better than BARTHOLOMAE's 'Inbesitznehmen'. In Y.51.17 we read: asahya azdyal garazdim which is rendered by HUMBACH as "so that he may obtain the favour of truth" ("damit er die Huld der Wahrheit erreiche), and by INSLER as "in order for him to obtain the stride of truth." The root garad- from which the noun garazdi- is supposed to be derived occurs in three later Avestan passages with the prefix aiwi or avi. To this BARTHOLOMAE assigns the meaning 'to enter upon, to engage oneself in, to begin'. In Vr.17.1 we read: aiwi.garadmahi? yasnahe haptanhatois humataca huxtaca hvarstaca; aiwi.garamahi asam vohu, translated by WOLFF as "We begin the good thoughts and the good words and the good deeds of Yasna Haptanhaiti. We begin the 'Asa is the best possession'" ("Wir heben an die guten In his Commentary on Y.50.9 Insler uses the word 'footing' in place of 'stride'. What INSLER means by his translation is not clear. ? J. KELLENS, Le Verbe Avestique (1984), p. 202 (1.4.1) derives garpomahi from a stem garan- of ?gar- (B. 512) 'to praise'. For garaomahl instead of the expected garanmahi see KELLENS p. 178, b. n. 2. Studien zur Indologie und Iranistik (SUIT) 16 (1992) pp. 147-148. Madhu Vidya/211

Page #237
148 M. A. MEHENDALE ye tch them here for going and backwaro ayese Gedanken und die guten Worte und die guten Tates des Yasna Haptanhatay. Wir heben an das 'Asa ist das beste Gut'"). In Vr.21.1-2 garad- occurs in such contexts as: avi ... asaonamca fravasinam yasnam garadmahi vahmamca / ... avi amasanam spontanam yasnam garaomahi vahmam ca rendered by WOLFF as "For the Fravasis of the thruthful ones we begin prayer and praise ... For the Amasa Spantas we begin prayer and praise" ("Fur ... die Fravasay's der Asaglaubigen heben wir Gebet und Preis an ..."). The third occurrence of aiwi.garad- is in 7.62.11: aiwi.garadmahi apam vanuhinam fraitimca paititimca aibi-jaratimca ayese yesti, rendered by WOLFF as "We begin the forwardgoing and backwardgoing of good waters and (their) praise; I fetch them here for worship" (Wir heben an das Vorwartsgehen und das Zuruckgehen der guten Wasser und ihre) Preisung: zu verehren hole ich (sie her)). It it suggested here to connect Av. garad- with Sk. grdh- (grdhyani) attested since the RV. It means 'to long for, to desire for, to strive after'. If we use this meaning it is possible to obtain such translations of the above later Avestan passages as: "We long for the good thoughts and the good words and the good deeds of Yasna Haptanbaiti" (Vr. 17.1); "We long for the worship of the Fravasis of the truthful ones and for their praise" (Vr.21.1-2); "We long for the forwardflowing and the backward-flowing and the greeting of the good waters" (Y.62.11). Av. g@r@zdi- *(Sk. * grddhi-), would mean "longing, wish'. Accordingly Y.50.9 can be translated as May I be in the longing of the blessed one' (.e. May the blessed one choose me as his priest). Similarly Y.51.17 can be translated as 'in order to obtain the longing of truth' (i.e. longing for truth, or 'what asa longs for viz. complete compliance with the requirements of truth'). os to long touch translation the good worin of the Frava 'It may be pointed out that there are no known Indo-Iranian cognates of Latin gradi. But derivatives of garad- : grdh- are available in Av. garada-, g@redl-? only at the beginning of a compound), as well as in Sk. grdhra-, grdhnu-8/tsa- (?). Madhu Vidya/212

Page #238
ON YASNA 41.2 M.A. MEHENDALE Yasna 41.2 reads : vohu xsa@ram toi mazda ahura apgema vispai yave / huxsa Orastu no na va nairi va xsaeta uboyo anhvo hatam hudastama. J. Narten renders the passage as: "May we, oh wise Lord, reach your good rule for all time. May a good ruler, whether man or woman, rule over us in both the existences, oh Best Worker among those that exist."! Narten treats tu as an expletive particle and construes hyxsaOras with na va nairi va meaning any good ruler, whether man or woman. This interpretation is not likely for the following reasons : 1. In the opening sentence of the above passage the assembled devotees express their wish to obtain the good rule (vohu xsadrom) of Ahura Mazda, It is therefore almost cetain that the word huxsaOras, which immediately follows, refers to Ahura Mazda himself as one of good rule, and not to any human ruler. 2. In the Avesta huxsaora is used specially with reference to Ahura Mazda and the Am@sa Spentas. 3. The words na va nairi va are meant, as in Y.35.6 (yada at uta na va nairi va vaeda haidim), to cover collectively all men and women, and not refer to a single individual, whether man or woman. 4. Hence it follows that the expression na va nairi va is intended to convey collectivity and not indifference regarding gender distinction, i.e. - whether male or female.' In that case probably we expect the use of words nairya and stri (Skt. nar or pums and stri). 5. If the author had intended na va nairi va to be construed with huxsaOras he would have said huxsarasta na va nairi va and would not have used the word na in between them. As the text stands, na va nairi va has to be construed with na. 6. A human ruler, though good, cannot excercise rule in both the kinds of existence of his dependents viz. the one in this world and the other which extends beyond death. Narten is conscious of this difficulty and hence she pleads that a good ruler helps his subjects to lead a righteous life and thereby cross the Cinvant bridge safely after death. This could be a way out, if there was no other. 7. It is hard to imagine that there was a woman ruler governing the country in ancient Iran. MadhuVidya/213

Page #239
These difficulties are easily set aside if we take hyxsaora to refer to Ahura Mazda and interpret na va nairi va as a parenthetical explantion of the immediately preceding word no. Also tu as in Y. 41.3 and 4 may be interpreted as 'thou'. The translation would then run as follows : "May you, whose rule is good, rule over us (all) -- whether (he be) man or woman -- in both the existences." The only difficulty in this interpretation is that the verbal form in the sentence is not, as expected, xsaesa (2nd per. sg. agreeing with tu 'thou'), but xsaeta (3 rd per. sg.). This is in all probability so because of attraction of the immediately preceding na va nairi va or due to the extended force of hursaOras. Two more translations of the Yasna Haptan haiti are now available. Humbach's English translation of our passage substantially agrees with that of Narten. The only difference is that he translates hudastama as 'the most munificent one' and not as 'the best worker'. The French translation of Kellens-Pirart is very different. They treat vohu-xsafra as a Bahuvrihi compound qualifying paq'path'supplied. Their translation runs as: "May we attain the (path) which ensures divine control over you, O Lord Mazda, forever! May a man or a woman having good (ritual) control rule over us in both the existences, O the most generous among those who exist". This translation completely changes the tenor of the passage. Instead of the worshippers attaining the good rule of Ahura Mazda, this translation enables the worshippers to get control of him. The translation depends on the supply of the. word for 'path' which is not easily available from the context, and it introduces the notion of 'ritual' control in the interpretation of xsa@ra'. They take recourse to Y. 33.5 for the word for 'path'. In that case why do they neglect the very word xsara occurring there as object of apano in the expression apana.... xsa@ram vanhus mananho which comes very close to our passage vohu xsa Or@m .... apaema, and choose pado? Kellens-Pirart interpret tu of huxsarastu as a particle and not as a form of the 2nd personal pronoun because they say that, as pointed out by Renou (EGS P. 68 SS 13), the Vedic idiom does not permit the occurrence of two enclitic pronouns (in our case tu and na) following each other. What Renou, however, in fact says is that such a case is not frequent (Le cas de deux pronoms personnels atones n'est pas frequent). There is therefore nothing wrong in interpreting tu occasionally as a personal pronoun when the context requires it. Madhu Vidya/214

Page #240
FOOT NOTES: Der Yasna Haptanhaiti, Wiesbaden, 1986 p. 47 (Mochten wir deine gute Herrschaft,o Weiser Herr,erlangen furalle Zeit, Mochte ein Gutherrschender, sei as Mann oder Frau, uber uns herrschen in beiderlei Dasein, o Bestwirkender von denen, die es gibt). Op. cit., p. 293 (Die Auswirkung der guten Herrschaft eines Menschen auf das geistige Dasein' der von ihm abhangigen besteht wohl darin dass sie es letzteren ermoglicht, ihr Leben als Asahafte zu fuhren und so die Hoffnung haben zu konnen, nach dem Tod sicher uber die Brucke des Bussers zu gelangen). huxsaoras tu no... xsaeta of Y 41.2 and aba tu no .... Yya of 41.3 are so alike that there is no reason to interpret tu differently in these two passage, i.e. as expletive particle in the former and as 'thou' in the latter. It is more economical to seek the explanation of the third personal form of xsaeta somewhere else, as done above in the text. The Gathasof Zarathushtra and the otherold Avestan texts, Parts I&II, Heidelberg, 1991. Les Textes Vieil-Avestique, I, II, III, Wiesbaden 1988, 1990, 1991. "Puissons-nous atteindre le (chemin) qui assure le divine emprise sur toi, Maitre Mazda, pour l'eternite! Qu'un homme ou une femme a la bonne emprise (rituelle) dispose de nous dans les deux existences, ole plus genereux des Existants". For this they refer (Vol. II. 232) to Humbach's view on the point : "It is the magical potency by which the priest makes the god favourable to him" (Die Gathas des Zarathushtra Il. 86). 7. Madhu Vidya/215

Page #241
(III) Pali and Prakrit TAKKI OR DHAKKI By M. A. MEHENDALE Psthvidhara in the introduction to his commentary on Micchakatika classifies the Prakrit dialects used in the play. He has referred to therein to a dialect Dhakkabhasa or Dhakkavibhasa. This is the sakki of Purusottama' and other Prakrit grammarians. According to Pothvidhara this dialect is spoken by Mathura and Dyutakara in the second act of Micchakatika. He also mentions some phonetic characteristics of this dialect. The semi-vowel v is found, according to him, many a time in this dialect, and on account of the presence of Sanskrit words it has both the dental and palatal sibilants, s and s. But Markandeya in his Prakitasarvasva and Purusottama in his Prakitanusasana treat this dialect at some length. A summary of what they have to say about this dialect may be given as follows. They consider the genesis of this dialect to be in the mixture of Sanskrit and Sauraseni. Morphology : The termination -u appears very often ; the termination of the instrumental singular is also -em. Markandeya gives it without anusvara i.e. -e; the terminations for the dative and ablative plurals are -han and -hum also ; the same terminations can be optionally used for genitive plurals ; Markandeya suggests that the same terminations can be used for pronouns and then the penultimate vowel is lengthened. Thus kaham, of whom ? taham, of them, etc.; then, tunga may be employed for tvan. Purusottama alone says that hamam may be used for ahan. Markandeya on the contrary gives three forms of the first personal pronoun ammi, hun, mama. The genitive of this pronoun is given by the same author as mahun or mama. Phonology :-Sanskrit yatha and tatha become jidha and tidhat in Takki according to Purusottama. Markandeya admits of these two forms and in addition gives the two regular ones jaha and taha. This is only to indicate the general character of the dialect. The authors clearly say that the rest is to be learnt from the usage in the writings of sistas. 1 NITTI-DOLCI, Le Prakylanusasana de Purusottama, p. 22. 2 For the acceptance of the reading Takki in favour of Dhakki, cf. JRAS, 1913, pp. 882-3. Cf. also PISCHEL, Gram. d. Pk-Sprachen. 3 Vakarapraya Dhakkavibhasa. Samskytaprayatve dantyatalav yasasakaradvaya yukta ca. . Cf. Pa. idha for Sk, iha. The Pk, forms quoted above bear the same relation to the pronominal bases ya-, ta- as idha bears to i-. On these forms see PISCHEL, 88 103, 266. Madhu Vidya/216

Page #242
190 M. A. MEHENDALE The name of the dialect :-PISCHEL calls it Dhakki and considers that it was spoken in Dhakka country in Eastern Bengal. This view of PISCHEL is criticised and controverted by GRIERSON in JRAS. 1913, p. 882. GRIERSON thinks that the dialect was spoken in the sakka country of the Northern Punjab. But even this is difficult to accept at present in view of the fact that the languages spoken there now share none of its special characteristics. It is indeed very difficult to identify all the different varieties of Prakrit dialects mentioned by the grammarians inasmuch as sufficient literature representing them is not found. Only the three main Prakrit dialects, viz. Maharastri, Sauraseni and Magadhi can be easily distinguished from one another. The subvarieties of these share some of their characteristics and thus lie in the peripheri of these main dialects. The most important characteristic of Takki is the preservation of the distinction of two sibilants out of the original three, Sk. s>s and Sk. s, s > s. This feature may be compared with European Romani sPage #243
TAKKI OR DHAKKI 191 s, s<. Considering the parallelism existing between these we may presume that sakki is a Romani or Gypsy dialect spoken somewhere in India, either in North-West Punjab or in Orissa.8 As admitted by the two Prakrit grammarians the reading of their texts cannot give us a complete idea about the nature of Takki. To gain that end one must look up to the texts making use of this dialect. Hence here is an attempt to study the dialect as it is spoken by the two characters Mathura and Dyutakara in the second act of Micchakatika. A.-PHONOLOGY. (a) Vowels : The simple vowels to be met with are an a, i, u, u. The dipthongs are e and o.. The vowel a comes from Sk. a, cf. aha < atha; jasan yasas. The vowel a comes from Sk. 1, cf. aikasanam Page #244
192 M. A. MEHENDALE k. cf. katthamayi < kasthamayi. kh- cf, khandiavutto Page #245
TAKKI OR DHAKKI 193 In the following examples we get -l- in GODABOLE's edition, whereas STENZLER gives -1-. The reading as found in the latter edition is given in the brackets. judiala (judiara) < dyutakara; dulat (durat) < durat ; ale (ore) < ore. But at many other places even STENZLER gives ale, cf. as corresponding to GODABOLE's edition 87.1 ; 88.5 ; 89.4 ; 90.3; 106.1 ; 102.6 ; 103.1 ; 116.1. -- cf. mandalie < mandlya ; sela < saila- etc. -S. cf. nasika < nasikya. The other two sibilants --- and --- are reduced to the dental one, cf. jasan < yasas ; dosu < dosah ; samavisaman < samavisamam.18 - S- cf. padimasunnu < pratimasunyam ; dasa- < dasa-. It is only at one place that the dental sibilant -s- is changed to the palatal one, cf. pasaru < prasara. STENZLER, however, gives the reading with dental sibilant, cf. pasaru. -h-cf. manahara< manohara.. In the following examples the consonants are lost intervocally. -k- cf. judiara < dyutakara ; sahian < sabhikan. -g- cf. daccha < agaccha. -C- cf. manaharavaanam -h- cf. sukhan > suhan. -th->-h- cf. -pathan > -pahar; atha > cha. -dh- > -N- cf. Tudhira > Tuhira-; adharena > adharena. In one case even initially bh- is changed to h- cf. bhavalu < hodu. But the importance of this singular example is marred by the numerous places where bhodu occurs. Cerebrals : There are indeed few examples where a i Takki cerebral corresponds to a Sanskrit cerebral ; otherwise it corresponds to a Sanskrit dental. --- < Sk. -- kadakkha < kataksa ; -pada- < pata-. -n- salanan < saranam. -n- < -n- cf. nama < nama; niunu < nipunah, etc. tanumajjhe < tanu madhye, ahuna < adhuna. The process of assimilation found in the pre-Prakrit period is carried further in all the Prakrits including Takki. A dental is palatalised or cerebra 13 STENZLER gives samavisamam. Madhu Vidya/220

Page #246
194 M. A. MEHENDALE lised when it comes into contact with a palatal semi-vowel or a cerebral one respectively. Palatalization : dyuta > juda (through *jyuta). mithya > miccha (through *mnichya). - madhye > -majjhe (through * majhye). Cerebralisation : pratima > padima. This tendency to cerebralise a dental after is not widely noticed in Takki. There are many examples where the dental following r is kept in tact. Cl. - pavudo < -pravitah ; -vutto <-vyttah ; vippadiva khalantaa. Sk. pratima > padima ; prasara > pasaru, etc. MEDIAL CONJUNCTS : (a) HOMORGANIC GROUPS : Sk. -cch-> -chh- cf. gacchasi. Sk. -tt-> -tt- cf. -vuttosi < vyttosi. Sk. -kt- > -tt- cf. juttam < yuktai. Sk. -bdh- > -ddh- cf. laddhu < labdhah. (b) HETERORGANIC GROUPS : 1. Consonant groups with a semi-vowel. Sk. -thy-> -cch- cf. miccha < mithya. Sk. -dhy-> jjh- cf. tanumajjhe < tanumadhye. Sk. -kr->-kk- cf. vikkinia < vi + kri-. Sk. -T->-ji- cf. jajjara- -tt- cf. -putta < .putra ; annatta< anytra. Sk. -TI-> -tt- cf. dhutta < dhutta. Sk. -dr-> -dd- cf. ruddo < rudrah. Sk. -td->dd- Daddura < Dardura. Sk. -01- > -pp- vippadiva -kkh- kadakkha < kasaksa. Sk. -sth->-th- kasthamayi < kasthamayi. Sk. -$t- >-th- duttha < dusta; moutthi Page #247
TAKKI OR DHAKKI 195 As in the case of semi-vowels, the sibilants too are dropped in such conjuncts. But here the outgoing sibilant leaves its trace by aspirating the simple consonant in the cluster. 3. Consonant groups with a nasal. Sk. -nd->-nd. mandalie < mandalya. Sk. -iv- < mv. samvahaa < samvahaka. Sk. -ng-> -ng--angaa < angaka. Sk. -mbh->-mbh- -vippalambhaa <-vipralambhaka. Sk. -nt- > .nt- Vasantasena < Vasantasena. In all these cases the conjunct is kept in tact. But in a case where, unlike the examples cited above, a consonant precedes and a nasal follows the latter is assimilattd to the former. Cf. Sk. -gn-> -gg- bhagga -kk- mukke < *muk-na 24 4. Groups of nasals and semi-vowels. Sk. ny- > -^n- cf. -funnu < -sunya ;anysya > annassa. Sk. -In-> -nn- cf. suvanna < suvarna. In both the cases the semi-vowel is dropped, the remaining nasal is uttered with force, and as noted above the dental n is changed to cerebral.n. 5. Groups of two semi-vowels. Sk. -ly->-4. kallavattam < kalyavartam. Sk. -Tv->-09- savvam < sarvam. Groups of sibilant and semi-vowel. Sk. -sy-> -SS- gamissadi -SS- annassa < anyasya. The semi-vowel being weaker of the two is lost. Groups of more than two elements. Sk. -sker-> -kk- nikkamia < nis + V kram-. Sk.-ndr-> nd- Indar < Indram. Sk. mpr->-mp. sampadar < sampratar. 8. There is a single example where the conjunct is retained initially and medially. The word is bodily incorporated from Sanskrit. Cf. pratyabhivadaye. 9. Nasalisation. There are two examples where the conjunct is first simplified and then a nasal is added to it. Cf. jampasi < Sk. jalpasi ; adansami < a-darsayami. B.--MORPHOLOGY. (a) Nouns : 1. Masculine nouns ending in -a. Terminations. Plu. nom. -;-;-. acc. -am inst. -ena. -ehin. 7. Sing. 14 See PISCHEL, SS 566. 15 For manahara in GODABOLE, STENZLER gives mammang Page #248
196 M. A. MEHENDALE Plu. Sing. Sing. Plu. gen. -assa. loc. - e. voc. -a; -a. Examples Sing. nom. dhuttu, niunu, dosu; goho, kalahavida. dhutto, ruddo ; gohe, laddhe. acc. kulavuttan, sahian. inst. kadakkhena, aharena. padehin, vippadivehim. gen. judicrassa. loc. pade. voc. dhatta or dhuttaa, dustha. 2. Neuter nouns ending in-a. Terminations. Plu. nom. -an ;-u; -e. im. acc. an. gen, aha. Examples Sing Plu. nom. sukham, judan; deulu ; bhudain, suvannain. suvannu ; mukke (addhain). acc. deuar, -vaanam, paalan. gen. dasasuvannaha. 3. Feminine nouns ending in -a. Terminations. Sing. Plu. nom. -a. VOC.-. Examples Sing. Plu. nom. padima, nasika. voc. tarumajjhe. 4. Feminine nouns ending in -i. Terminations. Sing. Plu. nom. i. inst. -ie. Examples. Sing. Plu. nom. padavi, katphamayi. inst. -mandalie. 5. Nouns ending in f. Madhu Vidya/223

Page #249
TAKKI OR DHAKKI 197 Sing. Plu. All such nouns are modified to -e base and are then declined like those ending in -Q. Cf. acc. sing. madaran, pidaran. voc. sing. bhatta. 6. Present participles. The present participle in -e has the endings of a noun in -a, cf. nom. sing. akkhanto; voc. sing. kalantaa. The present participle in -i is declined like the corresponding noun. Cf. aloanti. (6) The Pronouns : (1). The first personal pronoun has the following forms : Plu. nom. ahan. amhe. inst. mae (found in STENZLER'S edn, only). gen. mama, mae, me. (2) The second personal pronoun has the following forms :Sing. Plu. nom. tumam, tuhan. inst. tue. gen. tuha, tue. (3) The third personal pron. has the following forms. Sing. nom. so. acc. tarn. gen. tassa. (4) The demonstrative pronouns : etad, idam. i. etad---nom. sing. masc. esu, esa, eso. ii. idar-nom. sing. mac. aam. nom. sing. fem. ian. (5) The interrogative pronoun : kim. nom. sing. masc. ko.. gen. sing. masc. kassa, kassa. nom. sing. neut. kim. (6) Numerals : eka. acc. sing. mas, ekkam. (7) Ordinary pronouns. acc. sing. sauvai. (c) The Verb: (1) Present tense. Terminations. Plu. 1 per. -mi (parasmaipada). -mha. -e (atmanepada). 2 per. -si. 3 per. -di, - . Madhu Vidya/224 Sing.

Page #250
198 Sing. M. A. MEHENDALE 1 per bibkimi, adamsami (par.) pratyabkivadaye (atm.) 2 per. gacchasi, jampa-si, vajja-si. Examples. 3 per. vajja-di, jind-di, dhale-di; tara-i. Plu. The third person singular of the root v as is derived directly from the Sk. Thus asti > atthi. anusare-mha, kile-mha. (2) Imperfect. There is only one example asi (< asit) third sing. (3) Future. First per. sing. atm. -ssarh, cf. pekkhi-ssanh. Sec. per. sing. par. -hast, cf. pavi-hasi. Third per. sing. par. -ssadi, cf. gami-ssadi. (4) Imperative mood. The terminations for the sec. per. sing.-i. nil, cf. paaccha, daccha, genha ii. -hi, cf. de-hi, e-hi. The termination for the sec. per. plu. -ta (par.), cf. passata. The termination for the sec. per. plu. -ha (atm.), cf. ramaha. The termination for the third per. sing. -du, cf. bhodu. (5) Present participle. The distinction between Parasmaipada and Atmanepada is lost in the formation of Present Participles. The termination for their formation is -nta, cf. khalantai, kalantad, akkkanto. The feminine form is made by the addition of the vowel -i in the end. Cf. loanti. (6) Past Passive Participle. These are the assimilated forms of Sanskrit Past Passive Participles. Cf. padittha < pradysta: panalta:< pranasta; avakkanta Page #251
TAKKI OR DHAKKI 199 (d) Indeclinables. i. Conjunction: ca. ii. Interrogation : kahim, tatkim. ili. Interjection : are, ale. iv. Negation : na. v. Emphasis : nam, nama, hu, jjeva. vi. Manner : evvam. vii. Time : sampadan, ta, ahuna, javaccia, tavaccia. viii. Place : ehim, annatta. ix. Conditional : jai, aha. INDEX OF TAKKI WORDS FOUND IN Mfcchakatika.* aan (mas.) 99.6.9, Sk. Pa. ayam. ia (fem.) 85.8, Sk. Pa. iya . annatta 107.1 ; 115.2, Sk. anyatra, Pa. annattha. anna- gen. sg. -ssa 101.6, Sk. anya-, Pa. anna-. aha 101.4, Sk. Pa. atha, S. Pk. adha. aham. nom. sg. first per. pron. 92.5, 6; 101.5, 6, Sk. Pa. aham. Amhe nom. pl. 106.2, Ved. Sk. asme, Cl. Sk. vayam, Pa. mayam, Pk. amha-. mae 100.6 (S. mama), or mama 88.1,2; 115.2 (S. me) gen. sg. first per. pron. Sk. Pa. mama. ahora- inst. sg. -ena 114.9, Sk. Pa. adhara.. avakkanta nom. sg. - 106.3, Sk. apakranta-, Pa. apakkanta-. appana- acc. sg. -an 93.6, Sk, atman-, a. atti- Pk appa-, app@ma-. aikasana- acc. sg. -am 85.7, Sk. atikrsna-, Pa. atikanha-, Pk. ai-kasana-, ai-kanha. Vas- third sg. pre. indic. atthi 115.5, Sk. asti, Pa. Pk. atthi. Cf. natthi 115.2, Sk. nasti, Pa. natthi. Third sg. imperfect asi 106.5, Sk. asit, Pa. asi, Pk. asi, asi. V anusara- first pl. pre-indi. -emha 86.4; 106.5, Sk. V anu-sy., Pa. V anusara-, ahuna 89.2, Sk. Pa. ad huna. ale 84.4 ; 86.1 ; 87.1 ; 88.5 ; 89.4 ; 93.3; 102.6; 103.1 ; 106.1; 116.1, Sk. Pa. are, Mg. ale. Also cf. le 85.5 (S. re) Sk. Pa. re. amhe see under ahami. V gaccha- sec. sg. pre. indi -si 92.1, Sk. Pa. V gacch-; sec. sg. imperative aaccha 116.2, Sk. Pa. av gacch- ; thir. sg. future gamissadi 107.1, Sk. gamisyati, Pa gamissati, Pk. ganchai Absolutive gadua 106.8, Sk. gatva, Pa, gantva, Mg. gadua, gadua. akkhanta- nom. sg. -0, 101.4. False reading for acakkhanto Pischell SS 88 (p. 76) Sk. u-V caks-, Pa. av cikkha-. adamsa- first sing. pre. indi. a-mi 101.6, Sk. av drs., Pa. a- dassa.. aloanta- nom. sg. pre. part. fem. -1 115.1, Sk. a-lokayanta-, Pa. a-lokenta-. Esi STEN * The references are to GODABOLE's edition. The variant readings from ZLER'S editions are given in brackets with S. Madhu Vidya/226

Page #252
200 M. A. MEHENDALE asi see under Vas-. ian see under aan. Inda- acc. sing. -am 85.2, Sk. Indra-, Pa. Inda-. ido 107.1, Sk, itah, Pa. ito, Pk. io. uarod ha- inst. sing. -ena + eva 107.2, Sk. Pa. uparodha-, Pk. uparoha-, ueroha-. ucchadida- nom. sing. -0 113.1, 114.1, Sk. utsaditaekka- acc. sing. -am 85.3, Sk. Pa. eka-, Pk. eka-, ekka-, ega-. evvam 86.5, 87.2, 90.6, 91.3, 101.4, 102.3, 103.7, Sk. Pa, evan, Pk. evam,evvan. etad- nom. sing. masc. esa 107.1, esu 89.7, 100.6, 103.5, eso 85.9 Sk. esah, Pa. eso, Amg, ese, eso, Mg, ese, Ap. eho, ehu. Cf. PISCHEL SS 426 (p. 301). tad- nom. sing. mas. so 106.3, 6. Sk. sah, Pa. so. gen. sing. mas. tassa 106.4, 115.5. Sk. tasya, Pa. tassa. acc. sing. neut. tam 88.6, 92.1, 6, 102.6, 116.1. Sk. tat, Pa. tam. Also cf. tatkim 100.10. Sk. tat-tin. e-hi 87.2, 106.5, Sk. e-hi, Pa. e-hi. Ehim 92.5, Sk. asmin, 61.4 OI-A *esmin. kaffhamayi 86.7, Sk. kasthamayi, Pa. kafphamayi. kadakkha- inst. sing. -ena 115.1, Sk. kataksa-. kadhan 86.7, Sk. Pa. kathan, Pk. kahan, S. Mg. kadham. kalanta. voc. sing. -a 85.7, Sk. kurvat-, Pa. karonta-, Pk. karanta-, karinta-, karentakalahavida- nom. pl. masc. a 106.2 < Sk. *kalahapayita-, Pk. kalahaia < Sk. kalahayita. kallavatta- nom. sing. -u 100.6, Sk. kalyavarta-, acc. sing. -am 99.10. kim- nom. sing. masc. ko 91.3, Sk. kam. Pa. ko, Amg. Mg. ke, gen. sing. masc. kasca (S. kassa) 114.9, Sk. kasya, Pa. Pk. kassa, nom. (acc.) sing. neut. kim 115.5, Sk. Pa, kim. kahim 85.4; 92.1, < Sk. kasmin, Pk. kahi-n. V kila- first plu. pres. indi. -emha 87.2 (S. kilemha), Sk. V krid- Pa. V kila-, Pk. V kidda-, V kila-. . kula acc. sing. -am 85.7, Sk. Pa, kula-. Cf. kula-vutta- acc. sing. -am. 116.1 (S. kula-puttaam) Sk. kula-putra-, Pa. Pk. kula-putta-. V kala- sec. sing. imperative, kulu 93.3, Sk. kuru, Pa. karohi, Pk. karehi. ko see under kim. khandia-vutta- nom. sing. -0101.6-7, 101.9, Sk, khandita-vytta-. V khal- voc. sing. pre. part. khalantaa 85.6, Sk. Vskhal-, Pla. V khal-. gacchasi see under Vgacch- above. ganthu 93.3, Sk. grantha-, something to cover, to bind. gadua see under v gacch- above. gamissadi see under v gacch- above. Vgenha- first plu. pres. indi. -emha 107.2, Sk. V gruna, Pa. V ganhi-, sec. sing, imperative genha 84.5, third sing. ppp. gahida- 88.5 (S. grhida-), Sk. glhita., Pa. gahita-. gosaviaputta 102.3, < Sk. gosvomika-putra. Madhu Vidya/227

Page #253
TAKKI OR DHAKKI 201 goha- nom. sing. -e, 88.4, or -0, 106.3, Pk. goha. On this word see WUST, BSOS 8, 856, THIEME ZDMG 91, 115 ff., KATRE Prakrit goha, Saradesci Commemoration Volume, pp. 173-176. ca 85.2, Sk. Pa. Pk. ca. cittha 84.5 (S. tistha), Sk. tispha, Pa. lifpha. chinalia-puttaa 103.1 (S. chinmalia-putta), Pk. chinnalia- or chinnali-. See THIEME Uber einige Persische Worter im Sanskrit, ZDMG 91, 121. jai 85.2, Sk. Pa. yadi, jajjarapada-ppavuda- nom. sing. - 99.9, Sk. jatjarapatapravita-. Vjampa, sec. sg. pre. indi. -si 115.1. Sk. Vjald-, Pa. Vjappa, V jampa.. infinitive -dum 101.4. jasa- acc. sing. - 85.7, Sk. yasas-, Pa. yasa-. javaccia 106.2, Sk. vavat-+ * cida. On this last see B. SCHWARTZ, Prakrit cia, NIA 2, Sept. 1939, p. 420. ja sec. sg. pre. indi. -si 85.2, Sk, Pa. ya.. Vjina- third sg. pre. indi. -di 101.2, Sk. Vji-, Pa. Vjaya. juttan 101.4, Sk. yuktan, Pa. yuttar. jug- or juda- acc. sing. om 101.5 and 87.2; 102.3; 116.2, Sk. dyuta-, Pa. juta. judiara- nom. sg. -4 86.2 (S. judicru), Sk. dyutakara-, Pa. jutakara-, also cf. judikaru 84.4 (S. judiaru). gen. sg. -ssa 106.4. Also cf. -mandalie 89.7 (S. judialamandalie). ijeva 102.3, Sk. evana 85.3; 87.1 ; 88.2 ; 92.5; 101.2, 4, 6. Sk. Pa. na. nam 89.4, 95.3 ; 102.3. Sk. Pa. nanu. natthi see under atthi above. nama 101.5, Sk. nama, Pa. nama. nasika 106.4, Sk. masikya, Pa, nasika. rpiuna- nom. sg. -4 92.5, Sk. Pa. nipuna. ni v kkama- absolutive -ia 107.1 Sk. nis- V kram-, Pa. ni-vkkamaniu veda- first pl. pres. indi. -(e)mha 106.8, Sk. ni-vid-, Pa. ni-vveda. tam see above under tadtatkir see above under tadtanumajjha- voc. sing. - 114.9 Sk. t'anumadhya-. V tara- third sg. pre, ind. -i 85.3, Sk. Vtr., Pa. V tara-. tassa see above under tad. ta 86.4; 92.6; 106.5; 107.2. Sk. tatah, Pa. tato. More probably Sk. toda > *taa >ta meaning "then, afterwards, at that time or moment." tavaccia 106.3. Sk. tavat + *cida. See javaccia. tumam 89.7 (S. tuham) nom. sing. 2nd per. pron. Sk. tvar, Pa. tvcom, tuvan. gen. sing. tue 100.6; 116.2 (S. tuha); 102.3 (S. mae), tuha 114.6; tuhan 100.4; 101.7, 9; 114.9. Sk. Pla. tava. Daddura. inst. sing. -ena 106.2, Sk. Dardura-. dasa 99.6 (S. dasa-) Sk. dasa, Pa. dasa. -Suvanna- acc. sing. - 88.6; 92.1 ; 99.9 ; 102.6; 115.5. nom. sing. -4 100.6 gen. sing. -dha 84.4 (S. Madhu Vidya/228

Page #254
202 M. A. MEHENDALE dasasuvanaha). Sk. dasa-suvarna-, Pa. dasa-suvanna. V da- 2nd sing. imperative dapaya 97.10 (S. dapaa), Sk. dapaya. V de- 2nd sing. imperative -hi 95.2; 106.1. Sk. V dad-, V paaccha- 2nd sing. imperative peaccha 88.5 ; 89.4, 2; 92.6, 1; 93.2, 4, 6; 101.5 ; 102.6 < Sk. prayaccha. dulpha 103.1. Sk. dusta, Pa. duftha. dulat 84.5 (S. durad) Sk. durat, Pa. dura. deula- acc. sg. -am 86.3, nom. sg. -u 86.2, Sk. deva-kulo. dosa- nom. sg. -4 91.3. Sk. dosa-, Pa. dosa-. vdhala- 3rd sg. pre. indi. -di 99.6; 111.5; Sk. Vdhr-, Pa. V dhara-. dhutta- nom. sig. -4 86.2; 92.5 ; 101.5; or -o 1107.1. Sk. dhurta, Pa. dhutta-, Voc. sg. - 101.5, 6; dhuttijjami 92.6, denom of Sk. dhurta, dhurtiya-mi. padima 86.7. Sk. pratima, Pa. patima, -Sunna- nom. sg. << 86.2 (S. sunu), Sk. -sunya-, Pa. -sunno . padissudia (only STENZLER p. 35. 6), Sk. prati-*Srutya. panatfa 85.8, Sk. pranasta, Pa. panattha. padavi 85.9. Sk. Pa. padavi. padittha- nom. sg. -o 84.5 (S. pradtstosi), Sk. pradista-, Pa. padittha-. pada- loc. sg. -e 85.6. Sk. Pa. pada-. V pala- 2nd sg. pre. indi. -si 85.5, Sk. Pa. V pal-, nom. sg. ppp. papalinu 84.4, 5 [ = Sk. prapalayitah). Cf. Mar. pal-ne; Pk. palai < *palei whence ppp. *pa-linapaviffa- nom. sg. -0 106.6 Sk. pravista-, Pa. pavittha-. pavittha- nom. sg. -u 86.3. See above for Sk. and Pa. forms. pasaru 93.8 (S. pasaru) Sk. prasara, Pa. pasara ; the -u seems to be due to contamination with nom. sg. forms, cf. Sk. prasytah > *pa-satu or * pasadu, *pasau. paala- acc. sg. -an 85.2, Sk. Pa. patala-. patha- nom. sg. - 88.1, 2. pada- nom. sg. -4 86.1, Sk. Pa. pada. The actual reading in GODABOLE'S edn. is padu ; but that seems to be wrong. inst. plu. -ehim 86.3. V pava- fut. sec. sg. -hasi 103.2 (S. pabihasi), Sk. pro-vap-, Pa. V pupuna. pidara- acc. sg. -am 93.2. Sk. pitt.-, Pa. pitara. . pi 101.5. Sk. Pa. api, Pk. vi, avi. Cf. vi 85.3; 101.6 (S. bi). puno 116.2, Sk. punah, Pa. puna (7 before a vowel), prino. pulisa nom. sg. -o 99.9 Sk. purusa-, Pa. purisa, posa- Page #255
TAKKI OR DHAKKI bhaa pali-vevidanga voc. sg d 85.5 (S. bhaapalivebi-), Sk. bkaya-parivepi tangaka-. bhatia voc. sg. - 84.4; 99.8; 100.6; 106.2, 6 (S. bhadda), Sk. bhart bhagga 106.5, Sk. bhagna, P. bhagga. Vbhana- 3rd sg. pre. indi. -di 99.10, Sk. Vbhan, Pa. Vbhana-, 2nd sg. pre indi. si (-i) 116.1 (S. bhanesi). bhuda- nom. sg. neu. -am 116.1. Sk. Pa. bhuta-, nom. pl. neu. -im. Vbho- 3rd sg. imperative -du 86.5; 87.2; 90.6; 91.3; 100.10. Sk. Vbhu-, Pa. Pk. Vho-, also cf. ho-du 114.6 < Sk. bhava-tu. mae. See under aham. manahara-vaana acc. sg. -am 115.1 (S. mammana-). Sk. manohara-vacana-. Mathura- nom. sg. -u 92.5; 101.5. Sk. Mathura-. madara- acc. sg. am 93.4, Sk. matr-, Pa. matara-. miccha 101.5, Sk. mithya, Pa. miccha. mukka- nom. sig. neu. -e 91.5, 7. Sk. mukta-, Pa. mutta. The double -kk shows the OI-A form to be *muk-na. mutthippahara. inst. sg. -ena 106.4, Sk. mustiprahara, Pa. mutthippaharamusida- nom, sig. ppp. -o 113.1, 114.1. Sk. musita-, Pa. musita-. Vrakkha- infinitive -tum 85.3. Sk. Vraks-, Pa. Vrakkha-. radadattaduvvinida- inst. sing. ena 114.9 (S. raidattha-) Sk. rata-dasta-dur vinita-. Vrama- sec. pr. imperative -ha 116.2. Sk. Vram Rudda- nom. sg.0 85.3. Sk. Rudra-, Pa. Rudda-. 203 ruhirapaha- acc. sg. -am 106.5, Sk. rudhirapatha-. laddha- nom. sg. masc. - 84.4 (S. luddhu), -e 88.4. Sk. labdha-, Pa. laddha-. laula- acc. sg. -am 106.8, Sk. Pa. rajakula. le see under ale above. Vvaja 2nd sg. pre. indi. -si 85.2. Sk. Vuraj-. 3rd sg. pre. indi. -di 85.8. 2nd sg. imperative no termination. Vvajja- absolutive -ia 85.3. Sk. Vvarj-. Vasantasenageham 106.6. vi see under pi above. Vvikkina absolutive -ia 93.2, 2, 4, 6. Sk. vi-kri-. vippadiva- nom. sg. - 86.1, vipratipa. Pa vippatipa. inst. pl. -ekish 86.2 (S. vippadibehim) vihava- nom. sg. -u 100.7, -o 115.2. Sk. Pa. vibhava-. Selappadima 87.1. Sk. saila-pratima, Pa. selappatima. sampadam 85.2; 89.4; 92.7, Sk. sampratam. samavisaman 85.6 (S. Samavisamah), Sk. samavisamam, Pa. samavismam. salanamh 85.2, Sk. Saranam, Pa. saranam. Samvahaa 102.6, Sk. Samvahaka. sahia acc. sg. am 85.3, Sk. sabhika-, Pa. sahia-. suvanna- acc. sg. am 91.6; 92.6, Sk. suvarna, Pa. suvanna-, acc. pl. -im 106-7, Madhu Vidya/230

Page #256
204 M. A. MEHENDALE susahiavippalambha- voc. sg. a 84.5. Sk. susabhikavipralambhaka-. suhan 114.6, Sk. Pa. sukhan. sevida- nom. sg. neu. -am 102.4. Sk. Pa. sevita-. so see under tad-above. hu 87.1 ; 88.2 ; 89.7. Sk. khalu, Pa. kho. hodu ser under bho- above. (I am indebted to Professor S. M. KATRE for giving me this interesting topic for investigation and for substantially helping me throughout with books, papers and suggestions.) Madhu Vidya/231

Page #257
A FURTHER NOTE ON TAKKI By M. A. MEHENDALE Dr. S. M. KATRE was kind enough to point out to me a very small Kadavaka from Harivamsapurana which, according to its author, is written in Dhakka-bhasa. The Kadavaka occurs in the 103rd Samdhi of the Purana. The passage is a very small one consisting of barely nine lines and is full of proper names of Yadavas. Hence the passage does not afford to us any great opportunity to study the dialect in which it is written. We find the following vowels in this dialect: a, a, i, i, u, e and o. The consonants in their initial position are generally preserved. In the medial position they are either preserved or lost. There are some cases in which the loss of occlusion in medial aspirates is found. At times the medial surds are sonantised. The laws of palatalisation and cerebralisation are also at work. There is not a single dental n in the whole Kadavaka; everywhere it is cerebralised into . The liquid r is never changed to I. In this respect the present dialect at once differs from the Takki found in Mycchakatika. There is also one more salient aspect in which the present dialect is at variance with the one in Mycchakatika. In the dialect of this Kadavaka all the three sibilants s, s, $, are reduced to the single dental sibilant s. In this respect this dialect is not in conformity with the European and Syrian Romani where we find two sibilants s and f as in the Takki of Mycchakatika. There is nothing very particular to be noted in Morphology. The nom. sg. of mas. nouns in -a ends in -u. The masculine nouns in -i and are brought over to the a type so that in the gen. sg. of all these nouns we get the termination -ssa. In a solitary case we find the termination -him for the gen. sg. instead of -ssa, e.g. Subhanu-kim. Markandeya while describing the characteristics of Takki gives -harh and hush as the optional terminations for the gen. plu. He is silent as regards the terminations for gen. sg. We do not find this termination -hi in Takki of Mycekakatika either. As regards conjugation we get only a few forms of the present indicative, e.g. bhavadi, sakai etc. Below is given a complete index of all the words occurring in the Kadavaka. The references are given to the page number and the line number in the Apabhramsapathavali, edited by M. C. MODI. Anandana-, gen. sg. -ssa, 65.140, Sk. Anandana.. Kisamajjhuddesanteura, gen. sg. -ssa, 65.142, Sk. Kysamadhyoddes antahpura-. Kusumasara-, gen. sg. -ssa, 65.137, Sk. Kusumasara-. Kesava, gen. sg. -ssa, 65.136, Sk. Kesava-. Jaunasida, gen. sg. -ssa, 65.137, Sk. Yamunasrita Madhu Vidya/232

Page #258
A FURTHER NOTE ON TAKKI 133 jaga-, gen. sg. -ssa, 65.139, Sk. jagat. Jara-, gen. sg. -ssa, 65.138, Sk. Jara. Mr. MODI thinks that this is equivalent to Sk. jaratah. I think it is only a proper name. jema jema, 65.135, Sk. yatha yatha. na, oocurs very often, Sk. na. Nanda-, gen. sg. -ssa, 65.140, Sk. Nanda. Nandana-, gen. sg. -SSA, 65.137, Sk. Nandana-. ta-, gen. sg. -SSC, 65.141, Sk. tad.. titti, nom. sg. 65.136, Sk. trptih. tema tema, 65.135, Sk. tatha tatha. Divayana-, gen. sg. -ssa, 65.137, Sk. Dvipayana- or Dvaipayana-. Dunduhi., gen. sg. -ssa, 65.138, Sk. Dundubhi-. Devasena-, gen. sg. -ssa, 65.141, Sk. Devasena-. Paura-, gen. sg. -ssa, 66.143, Sk. Paura-. Paramesata-, nom. sg. -u, 65.135, Sk. Paramesvara-. Bhagad a-, gen. sg. -SSC, 65.139, Sk. Bhagada.. bhava-, pres, third sg. -di, 65.142, Sk. Vbhu-. Bhanu-, gen. sg. -ssa, 65.139, Sk. Bhanu-. Bhoya-, gen. sg. -ssa, 65.141, Sk. Bhoja mahayana-, gen. sg. -ssa, 65.143, Sk. mahajana-. Tisigana-, gen. sg. -ssa, 66.143, Sk. tsigana-. Samba-, gen. sg. -ssa, 66.138, Sk. Samba-. sambhava-, pres. third, sg. -di, 65.136, Sk. sambhu-. Saccai-, gen. sg. -ssa, 65.138, Sk. Satyaki-, sadhaviya-, gen. sg. -ssa, 65.139, Sk. sraddhapitasasimuddasaneuro-, gen. sg. -ssa, 65.142, Sk. sasazimudranupura-. sassuddhasira-, gen. sg. -ssa, 65.143, Sk. svasordhvasiras-. I follow Mr. Modi in the Sanskrit rendering. Chaya gives sasvat- in place of svasa. sahayana-, nom. sg. -u, 65.135, Sk. sabhijana-. Sarana, gen. sg. -ssa, 65.140, Sk. Sarana-. Sara-, gen. sg. -ssa, 65.140, Sk. Sara-. V saha-, pres. third sg. -1, 65.135, for Sk. V kath-. Sini-, gen. sg. -ssa, 65.138, Sk. Sini-. Subhanu-, gen. sg. -him, 65.139, Sk. Subhanu.. sutthuva-, gen sg. -ssa, 65.139, Sk. susthukase, gen. sg. third pers. pron., 65.137, Sk. tasya. sesiyasuyavayanamiyarasa-, gen. sg. -ssa, 65.136, Sk. sesitasrutavacana mptarasaVharisijja-, present third sg. -2, 65.135, Sk. V hys. hu, 65.136, Sk. khalu, Pakhu. Madhu Vidya/233

Page #259
SOME PHONETIC PECULIARITIES OF EARLY CEYLONESE LENA INSCRIPTIONS BY DR. M. A. MEHENDALE A detailed study of the Phonology of the Simhalese inscriptions upto the end of the tenth century has been undertaken by P. B. F. Wijeratne, the first part of which has already appeared in the Bulletin of the School of Oriental and African Studies, Vol. XI, part 4, pp. 823-836 (1946). Below is made only an attempt to point out some of the striking phonetic peculiarities of early Ceylonese inscriptions from the second century B. C. to the second century A. D. It is also attempted there to show some parallelisms between the treatments observed in these inscriptions and those found in the Prakrits of the early Brahmi (and Kharosthi) inscriptions of India, especially from the South India. The following short forms are used in the ensuing discussion. EI EZ L Luders' list of Brahmi inscriptions, appended to EI. 10.1-226. - Epigraphia Indica. Epigraphia Zeilenica. -- I Vowels $1. Treatment of the vowel r: the treatment > a is the predominant treatment in these inscriptions, though the change is noticed in a noun expressive of human relationship (cf. sk. bhraty-> bhatu EZ 1.145.10a). Thus we find that y> a not only in Sk. krtod > katu EZ 1.62, but also in Sk. pity->pita- EZ 1.18.13. This change of > a in a noun expressive of human relationship is very peculiar, for it is nowhere to be found in the Prakrits of the Brahmi or Kharosthi inscriptions. There in all such cases the treatment is a or i. It should, however, be noted that the change r> a even when accompanied by a labial in certain instances and in the 35 Madhu Vidya/234

Page #260
36 word vasabha < Sk. vrabhain Amaravati L 1239 etc., Nagarjunikonda EI. 20.18. B42, 16.C3 etc.) is a peculiarity of South Indian inscriptions e. g. M. A. MEHENDALE Sk. myga -> maka-Amaravati (L 1254). Sk. vyaprta > vapata - Mayidavolu (L 1205) and Kondamudi (L 1328). II Consonants $2. The guttural g: This stop, in certain instances, is changed to the corresponding surd. c. g. Sk. yavagu -> yaku - EZ 1.62.3 Sk. nagara -> nakara EZ 3.116. It is, however, preserved in such instances as ganakaEZ 3.116, and nagara - EZ 3.1222. Similar instances of the change-g->-- are found in the Brahmi inscriptions of Western and Southern India. e. g. Sk. nagara > nakara - Junnar (L 1152). Sk. Sagara -> Sakara- Nasik (L 1123). Sk. bhaga-> bhaka Kanheri ( L 998 ). Sk. Mrgabuddhi-> Makabudhi - Amravati (L 1254). Sh. Nagacandra > Nakacamda Jagayya - peta (L 1202 ). Sk. kosthagara -> kothakara - Nagarjunikonda EI (20.22.F2). etc. etc. SS3. The guttural aspirate kh. Deaspiration of Sanskrit aspirates is one of the chief peculiarities of the Ceylonese inscriptions. It is practically observed in all aspirates. 1. In all other inscriptions of Western and Central India we get usabha-cf. Usabhadata-in Karle ( L 1097, 1099 ), Sailarawadi ( L 1121), and Mathura (L 117). 2. It is further changed to a fricative k (r) in some Kharosthi inscrip. cf. Sk. nagaraka-nak(r)araa and Sk, bhagavat-> bhak( r )a pat()a in the Mathura Lion Capital inscription. Madhu Vidya/235

Page #261
SOME PHONETIC PECULIARITIES 37 Sk. bhiksu -> *Shikkhu bhiku - EZ 1.6.2,3 and EZ 3. 1622. Sk. Dharmaraksita -> Damarakita - EZ 1.18.1. Sk. V khan -> kana - EZ 1.211. As against this general.tendepcy in Ceylonese inscriptions, we have only one instance of similar change from South India. Sk. paksa -> *pakkha -> pakka - Malavalli (L 1195 ). $4. The guttural aspirate gh: This aspirate is also changed to g. e. g. Sk. sangha -> [sa] ga - EZ 1.18.12 and 62.3. But it is preserved in sagha - EZ 1.147. IIIa. This change is also found in an instance from a Kharosthi inscription cf. Sk. ghatika > gad( 7 )iglr)a -- Wardak vase. $5. The palatal c: It is changed to j in, Sk. pracina - > pajina --- EZ 3, 250'. This change is also found in the Brahmi inscriptions from early times but the later tendency there is to change it to y, It may be added that this latter is the only treatment obtained in the Kha rosthi inscriptions. $6. The palatal j: The change of j jh in a solitary instance is very peculiar. cf. Sk. jaya ->jhaya -- EZ 1. 19.6. $7. The cerebral th: Due to the tendency towards deaspiration it becomes l. Sk. jyestha -> * jetha -> jeta - EZ 3. 1542. Sk. artha --> * atha --> ata- EZ 3. 116'. It is interesting to note a similar change, though in a solitary instance, in as early as the Asokan inscriptions. Sk. Kamatha -> kaphata - Pilla'r edicts. ( The aspiration is not lost here but transferred to the neighbouring syllable). Madhu Vidya/236

Page #262
38 M. A. MEHENDALE A similar change is perhaps instanced in, Sk. frestha > seta-Bhattiprolu (L 1337') and Bharaut (cf. Selaka - L 838). SS8. The dental th: It is deaspirated to : Sk. sthavira -> thera-> tera- EZ 1. 18. 1 and 144. 4a. This treatment is perhaps instanced in a placename occurring in a South Indian inscription. Sk. Patha Amdha-patiyagama - Mayidavolu (< 12054). $9. The treatment is only exceptionally noticed in a South Indian inscription. Sk. deyadharma -> deyadama-Amaravati (El 15, 273.53) The dental dh is also deaspirated. Sk. Dharmaraksita-> Damarakita - EZ 1, 18. 1'. Sk. dhr -> dara - EZ 1. 2115. $10. The labial p>v in such instances as, Sk. harjapana- > hahavana-EZ 3. 211". Sk. upaniksepa-> uvanikeva EZ 3. 251. This change is often noticed in the Brahmi and Kharosthi inscriptions. $11. The labial bh, like other aspirates, is deaspirated". Sk. bharya> bariya - EZ 1. 18. 19. Sk. bhakta-> bata-EZ 1.62.3 Sk. vibhajaka -> vibajaka-EZ 1. 211.4. Among Brahmi inscriptions, the following ones from Western, Central, and Southern India give similar instances. Sk. stambha->thaba- Pittalkhora (L 1188), and thaba- Karle (L 1095). 3. Buhler, however, proposes to derive it from Sk. Svaitra -cf. EI 2. 328f. No. 8. 4. It is, however, preserved in Sk, bhratr->bhateu EZ 1. 145.10a. Madhu Vidya/237

Page #263
SOME PHONETIC PECULIARITIES Sk. Sk. Rebhila-> Rebila - Sanci ( L 502), Subhasita (?)-> Subahita-Sanici (L 270). Sk. bhagini-> bak [i]n[i]-Nagarjunikonda EI. 20. 37 Sculpture inscription. SS12. The change m> b is noteworthy. Sk. arama-> araba EZ 3. 116. We may compare with this the change mr> b or mb met with in the following set instances. Sk. Tamraparpi-> Tambapamni Rock edicts of Asoka. Sk. amraamra ambd- Pillar edicts of Asoka. - Sk. amraka-bhrti-> abika-bhati-Junnar (L 1164). Also cf. Aba-at Bhattiprolu ( 13321), and Tambapamni at Nagarjunikonda (EI 20. 22. F') or Tabapana Bodhagaya ( L 946). $13. Sibilants: The sibilant is sometimes preserved. cf. catudisa - EZ 1.18.12, Sivaguta - EZ 1.145.6-7 etc. It is, however, changed to in catudisa - EZ 1.19.4, visiti EZ 1.62.3. As against this, the dental s, though sometimes preserved, is very often changed to . cf. [u] pasika - EZ 1.18.1, faga (< sangha-) EZ 1.19.2F2, and also the gen. sg. term la(sa). This confusion between and has s its parallel in similar confusion between the three sibilants in the Kalsi edicts of Asoka and the Bhattiprolu casket inscriptions. 39 - The sibilant >h(perhaps through s) in Sk. latika -> hatika EZ 1. 62.4. This change is frequently noticed in the case of s in the gen. sg. term. (-sa-ha), cf. Devaha EZ 1.19.3, hulaha EZ 1.20.7, sagaha EZ 1.62.3, maharajaha EZ 1. 61.1 etc. It is interesting to note that a similar tendency is found in some of the Southern Brahmi inscriptions. Sk. Sangha-> Hamgha Amaravati (L 1201', L 1271:etc.) Sk. Sukha -> Hugha (?) Nagarjunikonda EI 20.25 k'. Madhu Vidya/238

Page #264
40 M. A. MEHENDALE Sk. sangha->hagha Nagarjunikonda EI 20.20.C2 and EI 20.17. CI". SS14. Palatalisation: The dentals are usually not palatalised in these inscriptions. Thus whereas Sk. amatya-> amaca in Brahmi and Kharosthi inscriptions, here it becomes ameti instead, cf. EZ 3. 250. The cluster k also becomes k (pagara-EZ 3. 1222 (4th cent. A. D.). Sk. khan-> khana-EZ 1. 2115 (2nd cent. A. D.). Also cf. dini (for dinna) EZ 1. 21. Rock B No. 1 (2nd cent. A. D.). So far as the Brahmi inscriptions are concerned the earliest instance of initial cerebralisation of n is found only in the Kopbal version (cf. Sk. no> no) of Asoka's minor Rock edict (and also perhaps once in the Jaugada separate edict cf. Sk. ni- dhya->ni- jhapa-). Among later inscriptions, we find initial cerebralisation in a few instances from the 1st cent. B. C. to the 2nd cent. A. D. inscriptions. from Western India. cf. Sk. Nanda-> Namdaat Kuda (L 1037) and Sailarwadi (L 1121), Sk. niyukta-> piyutaand Sk. nir-yatita-> pi-yacitaat Nasik.( L. 1127), and Madhu Vidya/239

Page #265
SOME PHONETIC PECULIARITIES 41 Sk. nartaka -> nataka - at Banavasi( L 11868). The sweeping change of n >n in all positions is obtained in the 4th cent. A. D. copper plates of Basim in Central India (cf. El 26. 137 ff). In the Kharosthi inscriptions n- is initially cerebralised from the 1st cent. A. D. (cf. nagara - Taxila silver scroll inscription, nana (< nana) Manikiala inscription etc.. ), and in medial position from the 1st cent. B. C. (cf. bahujana , Sakamuni --Swat relic vase inscription, etc. danamukha-Bimaran vase inscription, pujana - Zeda inscription, tanaya Skarah Dheri image inscription, etc. etc. Madhu Vidya/240

Page #266
Reprinted from JOURNAL OF THE ORIENTAL INSTITUTE, March 1952. WHAT WAS THE PLACE OF ISSUE OF THE DHAULI AND JAUGADA SEPARATE EDICTS? By DR. M. A. MEHENDALE M A., PH.D. Deccan College Post Graduate and Research Institute, Foona. The general belief regarding the issue and the transcription of the edicts. of Asoka is that they were all first written in the eastern or the Magadhan language, as they were issued from the Emperor's capital in the Magadha, and then were translated in the local languages of the different places where they were inscribed. While the veracity of such a general statement may reasonably be doubted, the fact remains that the inscriptions of Asoka show in bold relief three (or two) varieties of the Prakrit dialects-the Western represented by the Girnar version, the north western represented by the Shahbazgarhi version, (though both these labour under strong eastern influence) and the eastern represented by the Dhauli and the Jaugada versions. Hence the theory that. the original drafts of Asoka written in the eastern dialect were altered at Girnar and Shahbazgarhi to suit the requirements of the local dialects seems to be legitimate. Hultzsch observes, while commenting on the word Turamayo which occurs in the Girnar version, "This word proves that the Girnar and the Shahbazgarhi versions are translations from the Magadha dialect, where the (1) of the Greek original is preserved (Tulamaye) at Kalsi." But some of the instances in these western and north-western areas show exceptions to the general set up of the linguistic unity shown by the versions in these regions, and some of them have been explained as due to the eastern influence. Thus, for instance, the presence of a few forms with instead of r, and of nom, sg. mas. ending in-e instead of the regular-o have been explained as due to the influence of the Magadhan language of the original draft. It has, however, not been attempted to see if the exceptions occurring in the eastern versions can be similarly explained as due to outside influence. On a close reading of all the edicts found at Dhauli and Jaugada it will be observed that some of the exceptions which do not conform with the linguistic picture seen in the edicts at these two places occur only in the two separate edicts. The unique feature of these two edicts is that they are found only at these two places and not along with other major edicts at other places. It would, therefore, be reasonable to suppose that the exceptions which occur in the separate edicts do not represent the local forms, but may be due to the influence of a language foreign to the east. It may have been that these two Madhu Vidya/241

Page #267
WHAT WAS THE PLACE OF ISSUE OF THE DHAULI... EDICTS ? 241 edicts were not issued from the capital in the east and hence were not originally drafted in the dialect which for instance, was used in the drafting of the other major edicts at Dhauli and Jaugada; they were perhaps issued from a different place and in a different dialect and were translated into the eastern dialect at the time of their recording on the rocks at Dhauli and Jaugada. Though these translations were more or less accurate, there remained certain forms which appeared in the original draft, but should not have figured in the translations. The following analysis will further show that these exceptional forms show affinity with the north-western language of the Asokan inscriptions, thus pointing towards that area as the most probable source of the issue of these two separate edicts. .. 1. Softening of -- and -C--The D and J (in the following analysis.D and J stand for the other major edicts at Dhauli and Jaugada, while d and j stand for the separate edicts at these two places) versions preserve -k- and -c-. The separate edicts no doubt preserve these surds, but in stray cases they even soften them to -g- and .;-. Sk. loka-> loga- j (logika- cf. MS. I 5), Sk. acala-> ajalad. All other versions show loka-. Now it is true that -k- and -c- are not soften. ed in the north-western versions of Asoka's edicts; but that seems to be due to the strict following of the eastern dialect. What is interesting to observe, however, is that a tendency towards softening can be discerned in other instances in the north-western and the northern regions in the days of Asoka. Thus. Sk, vatika. > .vadika- Kausambi Queen's edict, -vadikya- Topra; Sk. hita-> hidaShahbaz, Mansehra, Kalsi.; Sk. yatra- > -yadra-Mansehra; Sk. lipi. > libiTopra.. The forms loga- and ajala. in j and d, therefore, seem to be due to the borrowings from the north-western dialect and do not reflect the local tendency in the east in the days of Asoka. 2. The next point that deserves our attention is the treatment of cerebral n in the Major Asokan edicts. As is well-known, the D and J versions substitute n by n throughout, and there is no exception to this substitution in the versions of the major edicts at Dhauli and Jaugada. The separate edicts alone, however, though they generally make this substitution as in the eastern dialect, give us four instances where n (obviously as a loan from the original draft) appears. The instances are khana., nijhapetaviye, palaloki ( kena ), and savena. Now as against the n > n treatment noted in the east, the western and north-western versions of Asokan inscriptions preserve n. 3. Equally interesting to note is the treatment of the cluster jn. This cluster is represented by palatal n in the west and north-west, but by dental >> in the east. Now, of the versions at Dhauli and Jaugada, whereas the major Madhu Vidya/242

Page #268
242 M. A. MEHENDALE edicts strictly conform to this treatment, an exception showing the north-western treatment is found only in the separate edict at Dhauli. Sk. pratijna-> patimna. d, Here even i gives the eastern form patimna. . 4. All the versions at Dhauli and Jaugada generally simplify consonant clusters by assimilation. But clusters with y are sporadically preserved only in d and j and not in D and J. Following are the instances : khy : mokhya- d, but-mokhiya- j. ny : ? (anya )- j; vy : samcalitavye j; sy : alasya- j, but alasiya- d. Preservation of clusters is eminently a north-western characteristic, as is shown by the Shah bazgashi version. Yet none of the versions at either Shah. or Manse hra (with one exception) show any instance of the cluster with 3. As in the case of the softening of intervocal surds, this too may have been due to the eastern influence. However, stray instances of preservation of clusters with y are found in the northern and the western versions of Asoka's edicts. khy: mokhya. in the Topra and Mirath versions in the north bhy : ibhya- occurs even in the Mansehra version where Shah. gives ibha-, vy: vyamjana. and katavya- occur in the Girnar version in the west. sy: isya. occurs in the Topra, Mirath and other versions of Pillar edicts, It may be observed in passing that the clusters with ry an with quite often in the Niya Prakrit (cf. Burrow, $ 42). 5. In this connection it is very important to look to the four instances of the preservation of clusters with in the Jaugata version as read by Hultzsch. As we might expect, these instances, do not occur in the separate edicts, but in the versions of the major edicts. Preservation of such clusters is a characteristic of the north-western (and to a certain extent western ) inscriptions of Asoka. In the east such clusters, as a rule, are assimilated. The four instances, however, appear only in the readings of Hultzsch. They are -savatra J IL 4, prativedayamtu J VI 2, drasayitu J IV 3, Piyadrasine J I 3. It is difficult to explain the presence of these four forms in J as the major edicts are believed to have been drafted first in the eastern dialect. However, Buhler, Senart, and Woolner read all these four instances without the cluster, and the estampages given by Hultzsch in CII Vol. I (new edition) show that all the four forms noted above are probably to be read without a cluster. Hultzsch reads here clusters with r not because t, p, or d show the usual curves at the top, but simply because these letters show slight scratches at the top or bottom. 6. The final -a of the various declined forms is regularly shown in the D and J versions. But it is only in the separate edicts that sometimes this final -7 is shown as -a. Though this irregularity in showing the length of the vowel met Madhu Vidya/243

Page #269
WHAT WAS TH" 'LACE OF ISSUE OF THE DHAULI... EDICTS ? 243 can be ascribed to the negligence of the scribe, the point cannot be overlooked, as the instances are quite varied. It is quite known that the Kharosthi inscriptions from the north-west do not show the length of the vowels. If the inseribers of the separate edicts at Dhauli and Jaugada received their copies from the north-west the length of the vowels in the original draft was naturally not marked. Though in the Brahmi versions of these edicts at Dhauli and Jaugada this length, wherever necessary, has been shown in a majority of cases, it escaped the notice of the inscribers in the following instances: (a) Nom. pl. mas. : -anuvigina d, j; ayata d, j; nagalaka j: mahamata d; vataviya d; viyohalaka j; laja-vacanika j. (b) Nom. sg. fem. ---acala j; icha d, j; likhita di solaviya d. (c) Inst. sg. fem. :--anaviliya dj. (d) Nom. sg. mas, :- y base :- pita j. (e) Nom. sg. mas. :-- .an base :-laja d. (f) Inst. sg. neut. :-- -an base :--kammana d; kamana j. 7 In the end we may look to some of the pronominal forms met with in the separate edicts. Though these peculiar forms are not to be found in the north-western dialect, they are noteworthy as almost an unique feature of the separate edicts. The forms in question are :(a) Ist pers. pron. acc. pl. aphe d, apheni j. gen. dat. pl. aphaka, aphakan d. loc. pl. aphesu d, j. (The nom. pl. form maye is also peculiar to d and j versions). (b) 2nd pers. pron. Nom, pl. tuphe, phe d, j. acc. pl. tuphe, tupheni j. inst. pl. tuphehi d, j. gen. pl. tuphaka d, j. loc. pl. luphesu d, j. (of these tuphe occurs also in the Saranatha and Yerragudi versions, tuphakan in the Saranatha version, and tupaka in the Rupanatha version--all being minor rock and pillar edicts). These forms have the base apha. (< * asma.) and tupha- (<* tusma). We may compare with this the Pali, the Prakrit, and the later Prakrit inscriptional bases amha. and tumha.. What interests us for our present discussion is the fact that these forms do not appear in any one of the versions of the major rock edicts including those at Shah. and Mansehra. On the other hand, as against their appearance in the separate edicts may be pointed out that the Madhu Vidya/244

Page #270
244 M. A. MEHENDALE cluster sm is preserved in the north-western dialect of Asoka as sp from which an assimilated form ph is an easy step. Thus we can trace the appearance of these forms to the north-westera influence ( *asma- > aspa. > apha). It is in respect of all the forms detailed in the foregoing discussion that the d and j versions differ from D and J. As all these features are restricted to d and ; and not even once appear in D and J, it would be difficult for us to assume that they represent the sporadic or exceptional tendencies of the eastern dialect. In that case they would have appeared even in a stray instance in D and J. On the contrary the affinity of these features with the north-western dialect suggests that these two edicts were issued by Asoka probably when he was on tour in the north-west in the local language and that they were later translated into the eastern dialect. This woald satisfactorily explain the presence of just a few peculiar forms in d and j. MadhuVidya/245

Page #271
C Reprinted from the Bulletin of the Deccan College', Vol. XIV, No. 3, Dec. 1952 ON THE PRAKRIT AND SANSKRIT NAMES OF THE NASIK CAVE-HILL By M. A. MEHENDALE The name of the hill in which the Buddhist caves at Nasik are excavated appears in some of the Prakrit inscriptions in these caves as tirunhu (5 times)1 or tiramphu (once). The same hill is styled as trirami in two of the inscriptions" written in mixed dialect in some of these caves. The names occur either singly as tiranhu and triraimi, or in composition as tiranhu-pa(v)vata and trirasmi-parvata. That both tiranhu and trirami refer to the Nasik cave-hill is made clear by such references as follows: (this) cave caused to be made on the summit of the mountain Tiranhu and given as a gift to the community of monks; this cave, with a caitya building and cisterns inside it, caused to be made on the mountain Tiramnhu and given to the community of monks in the four quarters; the building of (this) caitya caused to be completed on the mountain Tiranhu"; this cave and these cisterns caused to be made on the Trirasmi hills." The above references further suggest that the two names refer to the identical hill, and in this respect all scholars are agreed. The difficulty only arises in linguistically equating these two names, as the Pkt. tiranhu cannot be regularly derived from Skt. trirami. The difficulty of deriving phPage #272
164 M. A. MEHENDALE The first suggestion for the solution of the problem was made by SENART. He is of the opinion that since the meaning of rasand is partially identical with rasmi, there arose, out of the contamination of these two, a form like *rasni. Hence, according to SENART, tiranhu owes its origin to "trirani. By way of caution, however, he goes on to add, "...local names are subject to dialectic accidents, of which it is often difficult to state the origin and fix the measure."10 Though one may in general agree with SENART as regards the words of caution expressed by him, it is difficult to accepet his suggestion in respect of the derivation of tiranhu. For, apart from the considerations like those of contamination involved in it, the hypothetical form thus arrived at does not enable us to get over the second difficulty referred to above. The form "trirasni may help us to obtain tiranhi (en nh); but this form does not satisfactorily explain the change of the final vowel i>u (*trirasni> tiranhu): On the semantic side, the word rasana' girdle, etc.', has very little propriety in the name of a hill. In a recent article on the subject already referred to above, Dr. KATRE has shown that SENART's suggestion could be bettered by taking the Sanskrit prototype of Prakrit tiranhu as "trirasna from rasna'a girdle." This is no doubt an advance over the suggestion of SENART, for it saves us from the labour of the supposed contamination; but with this explanation also the second difficulty regarding the change in the end vowel remains. The difficulties in offering a satisfactory explanation of tiranhu have remained unsolved because the Sanskrit name of the hill is misleading. Leaving it therefore aside for a while, and keeping our attention on the last syllable of the Prakrit name, remembering at the same time that here we are dealing with the name of a hill, it will be seen that nhu in the given context can come only from Sk. snu. Now snu in the sense of the summit or edge of a mountain' occurs since the earliest times, and has been accepted as the collateral form of the more familiar sanu 'summit of a mountain'. snu has also been given by the Sanskrit lexicons. as an equivalent of sanu.12 Coming to tira, the fact that "the caves are situated in one of the three hills in which the mountain range (at Nasik) ends," or that the cave hill has a pyramidal form,13 suggests that the beginning of the hill-name tira may stand for a word meaning three'. Taken together, we arrive at the conclusion that the original Sanskrit counterpart of the Prakrit tiranku must have been trianu: trisanu, and that the Prakrit name actually owes its origin to a semi-tatsama from trisnu viz. tiramu. Both sanu and snu are surely appropriate as occurring in the name of a hill. 10. EI 8. 64. 11. Indian Linguistics, 14. 145. 1954. 12. Cf. snuh prasthah sanur astriyam Amarakosa 2.8.4; also of. Abhidhanacintamani 1035; Vaijayanti 41.14. 13. Bhagvanlal INDRAJI, op. cit. p. 541. Madhu Vidya/247

Page #273
PRAKRIT AND SANSKRIT NAMES OF NASIK CAVE-HILL 165 It is hardly necessary to add anything here to explain the derivation of Pkt. tiranhu from the semi-tatsama tirasnu. About the semi-tatsama itself in its relation to Skt. trisnu it may be observed that as the difficulty was felt in the pronunciation of the cluster tr at the beginning of the word,tri > tir(a), thus giving rise to tirasnu, which became quite a parallel to trisanu, both from the point of view of the number of the syllables and their metrical value ( -- ~). Normally Skt. tri- is represented as ti- in the MIA languages. But it apepers as tira in the present case as it happens to be a part of the semi-tatsama. Such dissolution of a consonant cluster with rin the first syllable is seen already in such instances of the MIA stage as kriya > kiriya or kiria (Prschel & 131, $ 135); sti > sira, hri > hiri (PISCHEL SS 135, GEIGER 8 8).14 Moreover tri itself appears as tir(a) in some of the numerals of the Modern Indian Languages.15 Thus for instance, for 53 we have tirpan (Hindi, Nepali), tirpanna (Marathi), tirwanja (Lahnda); for 63 we have tirsath (Hindi, Nepali); for 73, tirahattar (Hindi); for 83, tiras? (Hindi, Nepali, Bangali, Panjabi, Lahnda), tiryasi (Marathi); for 93, tiranave (Hindi, Lahnda), tiranabe (Nepali), tiranabbai (Bangali), tiryannav (Marathi); for 43, tirtali (Lahnda), tartali (Panjabi). In addition we may note the use of tiri (Marathi), tiri (Gujarati) in the game of cards.16 Also tir appears as the word for 3 in the Dardic languages. T In this connection it may also be observed that tir occurs for tri in the popular pronunciation of many words beginning with tri. The following examples17 cited from TURNER's Nepali Dictionary have correspondences in many of the modern languages of north India : tirphala (triphala), tirbeni (tribeni), tirbhuvan (tribhuvan), tirsul (trisul). Another similar instance, but not connected with the numeral tri, is Nepali tirsana < trisna garasa, trasyati > tarasai, pramana > paramana, prasanna > parasanna (PiSCUEL SS 132) 15. For illustrations, I am indebted to TURNER, Nepali Dictionary, (London, 1981). 16. Hindi tiyd, Nepali tiya or liyo, Panjabi tia. 17. It is possible to multiply these instances by citing pronunciations like titkal (trikal), lirkon (trikon) etc.; the tendency is witnessed also in other instances like taran (tran), taras (tras), etc. 18. cf. Rv. 8.46.18; 1.117.16, 155.1 ; 6.61.2 ; 8.96.2. Madhu Vidya/248

Page #274
166 M. A. MEHENDALE it occurs in the sense of three raised places above the earth' referring to the three fire altars. The verse runs daivya hotara prathama vidustara rju yaksatah sam rca vapustara / devan yajantav stutna sam anjato nabha prthivya adhi sanusu trisu // "May both the first heavenly Hotrs, the better knowers and those having better forms, offer the sacrifice correctly with the rk; while offering sacrifices to gods at the right time, may they bring them (i.e. the gods) toegther at the navel of the earth, on the three raised summits." Though we do not come across in early literature trisnu or trisanulo as a compound formation and as a name of a mountain, other similar names of mountains are of frequent occurrence since early times. Thus trikakud or trikakubh 'having three summits' occurs as the name of a mountain in old literature.20 varsisthah parvatanam trikakun nama te pita | Av. 4.9.8. " The highest among the mountains, Trikakud by name, is your father, (Oh Ointment)." yatra va Indro Vrtram ahams tasya yad aksy asit tam girin Trikakudam akarot Sat. Br. 3.1.3.12. "When Indra killed Vrtra, that what was his eye, that he turned into mountain Trikakud." (This is observed while discussing the propriety of using the Traikakuda ointment for the eye of the sacrificer). The same name appears as Trikakubh21 in the Kathaka Samhita XXIII.1 cf. Indro vai Vitram ahams tasya caksuh parapatat tat Trikakubham pra visat / With this Trikakud, we may also compare such later names of the mountains as Trikuta, 22 Trisikhara, and Tribrnga occurring in the Epics and the Puranas. 23 Next to the Sanskrit name of the hill, trirasmi. As seen from the Pkt. form, the original name of the hill was trisnu : trisanu. Now it seems fairly certain 19. Its equivalent triprstha, though not connected with a mountain, occurs already in the Rgveda, cf. 7.37.1 ; 9.62.17, 71.7, 75.8, 90.2, 106.11. 20. Trikakuda, As the name of a mountain, occurs in the various Puranas, cf. V.R. Ramachandra DIKSHITAR, The Purana Index, Madras, 1952. 21. trikakubh. having three points' occurs also in the Rv. 1.121.4, but as an adj. of Indra's vajra (*); however, kakubh occurs even in the Rgveda in the sense of the tops of the mountain" ef. avabhinut kukubhah parvatanam (4.19.4). "(Indra), struck down the tops of the mountains,'. For this meaning of kakobh here, see LUDERS, Varuna, p. 90; for references to trikakud, I am indebted to his note on kakubh, kakud, trikakud, kakuha, pp. 89-92. For Trikakud as the name of a mountain ulso ef. Panini 5.4.147 trikakut parvate on which Kasika observes 'na ca sarvas trisikharah parvatas trikakut / kim tarhi samjnaisa parvatavisepasya / 22. According to Arnarakosa, Trikuta and Trikakud are synonyms. cf. trikutas trikakut saman / 2.3.1. 23. For references see DixSHITAR, op. cit., SORENSEN, Index to the Nanes in the Mahabharata, and D. R. PATII, Cultural History from the Vayupurana, Appendix. Madhu Vidya/249

Page #275
PRAKRIT AND SANSKRIT NAMES OF NASIK CAVE-HILL that trirami is just the translation of trisnu: trisanu effected on the following lines. The Sanskrit lexicographers give 'arka' as one of the meanings of sanu." The word arka itself has various meanings, one of which is the ray' which is attested since early times. The Sanskrit name of the hill trirasmi 'having three rays' is therefore not the original name of the hill, but a secondary one obtained by translating sanu with rasmi (on the basis that both have the common meaningarka). Such translations of proper names with the help of synonyms are not uncommon in ancient literature. In fact a similar translation of trisanu itself, but with the use of another word for 'ray' viz. bhanu (instead of rasmi), has been recorded in the case of the name of a king, a descendant of Turvasu. The name of this king appears as Trisanu in the Vayupurana 99-1-2, and in the Brahmandapurana 3.74.1-2. But the name of the same person appears as Tribhanu in the Bhagavatapurana 9-28-16-17, which is evidently a variant of Trisanu due to the translation of sanu by bhanu (in the sense of arka 'ray'). 28 167 It will thus be seen that tiranhu does not come from trirasmi. As shown above, both originate from trisnu: trisinu, the one through the semi-tatsama tirasnu, the other through the translation of sanu by rasmi. 24. cf. Sabdakalpadruma under sanu : arkah/pallavah|iti Jatadharah Also cf. Sabdartha cintamani s.v. 25. turvasos tu suto vahnir vahner gobhanur atmajah | gobhanos tu suto viras trisanu aparajitah || karandhamas trisanos tu.......... 26. turvasos ca suto vahnir vahner bhargo 'tha bhanuman | tribhanus tatsuto 'syapi karandhama udaradhih // The other variants recorded for this name are Traisanu (Harivamsa 1.32.117-118, and also one ms. of the Visnupurana 4.16.2), Trayisanu (Visnupurena 4.16.2, Gita Press, Gorakhpur), Traisani (Agnipurana 276.1, Traisali as noted by WILSON in his translation of the Visnupurana IV.116), Trisari (Matsyapurana 48.1-2, Trisari as noted by WILSON op. cit.), Aisanu (Brahmapurana 18.142-143), and Traisamba (Visnupurana 4.16.2, Calcutta edn.). It is clear that Trayisanu, Traisanu, Aisanu, Traisani, Traisali, Trisari, and Trisari are all corruptions of Trisanu. Traisamba for Trisamba seems to be a translation of Tribhanu as Samba (for Samba) is noted as a particular form of the sun. (See MONIER-WILLIAMS Sanskrit Dictionary, s. v.). For the references to the Purana literature, with regard to these variants, I am intebted to M. M. CHITRAY SASTRI, Pracina Caritra-Kosa, Poona, 1932. Madhu Vidya/250

Page #276
SOME REMARKS ON THE LANGUAGE OF THE ORIGINAL BUDDHIST CANON By M. A. MEHENDALE At the very outset let me express my gratitude to the Executive Committee of the All-India Oriental conference for having elected me to preside over the Indian Linguistics Section of this year. It is a great honour no doubt, but I do feel that it would have been better if my election had waited for some more years. With my limited knowledge and experience I feel some embarrassment in executing the duties of this office which was in previous years held by great digniataries whose names are too well-known to Indologists. Perhaps in electing me as sectional President, the Executive Committee has sought to honour my Purvacaryas through me. Any way it is with this feeling that I shall proceed with the work and discharge my duties as best as I can. But before I proceed I have a sad thing to do. It is with a heavy heart that I refer to the sad demise of Prof. Jules BLOCH who passed away soon after the previous session of the Conference held at Ahmedabad. In his death India has lost a good friend and a great scholar. He not only himself contributed a number of valuable articles and volumes on Indo-Aryan and Dravidian Linguistics, but also trained many a young scholar from this country to do important research in this field. In such Presidential Addresses it is customary to take a review of the work done in the field during the last two years. As this was not done at the previous session I am extending the period of review roughly to four years. With the limitations on time I cannot do any more than just refer to some of the important publications, linking these up wherever necessary, with the previous work done in those particular branches. I earnestly beg to be excused for any omission and request the scholars to bring up these to my notice. (I have to point out that while preparing the review I had to omit references to articles published in * Sectional President's Address delivered at the Indian Linguistics Section of the 18th Session of the All-India Oriental Conference held on 26th-28th December, 1955, at Annamalainagar. Bull. DCRI. xvii-6 Madhu Vidya/251

Page #277
158 M. A. MEHENDALE Research Journals and Commemoration Volumes as that would have taken me too far). Let me start with OIA. With regard to the works of the grammatical nature mention must be made of the second part of the second Volume of DEBRUNNER-WACKERNAGEL'S Altindische Grammatik (1954) dealing with the nominal suffixes. The fact that the volume covers about a thousand pages in the treatment of a subject to which about sixty pages are devoted in WHITNEY'S Grammar will suffice to show the comprehensive nature of this book. Another work is L. RENOU's Grammaire de la Langue Vedique (1952) describing the stage of Sanskrit as represented in the Samhitas. One would realise the importance of this work when one remembers RENOU's noted contributions in the Vedic and the grammatical fields. Mention may also be made of the fact that RENOU has now completed his translation of the Astadhyayi in three volumes, the last of which appeared in 1954. T. BURROW's Sanskrit Language (1955) deals with the subject matter from the historical point of view and takes into account also the evidence of Hittite. J. GONDA has published certain important monographs like Remarques sur la Place du Verbe dans la Phrase Active et Moyenne en Langue Sanscrite (1952), Ancient Sanskrit ojas, Latin *augos, and the Indo-European Nouns in -es/ -08 (1952), and Reftections on the Numerals "one" and "two" in Ancient Indo-European Languages (1953). In the wake of Siddheshwar VARMA's work on Phonetic observa. tions of Indian Grammarians we have now a very good appreciation of ancient Prati akhya and siksa texts in W. S. ALLEN'S Phonetics in Ancient India (1953). M. B. EMENEAU'S Sanskrit Sandhi and Exercises (1952) in a good book for the use of descriptive linguists. Siddheshwar VARMA's book on the Etymologies of Yaska (1953) is a refreshing study discussing which of the etymologies of Yaska may be considered acceptable and which not. Among the works of the type of Dictionaries, reference may be made to SURYA KANTA'S A Grammatical Dictionary of Sanskrit (Vedic) (1953) with an Index to WACKERNAGEL'S first volume of Altindische Grammatik and first 82 pages of MACDONELL'S Vedic Grammar. The fifth part of the Kurzgefasstes Etymologisches Worterbuch des Altindischen (1955, started in 1953) by M. MAYRHOFER, and 9th part of Indogermanisches Etymologisches Worterbuch (1955, started in 1948) by J. POKORNY have recently arrived. J. GONDA in his Sanskrit in Indonesia (1952) ably discusses the history of Sanskrit vocables in the Indonesian languages. Equally important works have appeared in the field of MIA. H. LUEDERS' Beobachtungen uber die Sprache des buddhistischen Urkanons Madhu Vidya/252

Page #278
SOME REMARKS ON LANGUAGE OF ORIGINAL BUDDHIST CANON 169 (1954) has been edited from the Nachlass by E. WALDSCHMIDT. LUEDERS had already given expression to his view that he believed in an original canon composed in an eastern dialect from which the Pali and the Sanskrit versions were translated. This book gives the evidence which led LUEDERS to this belief and his observations on the nature of the eastern language. In this respect F. EDGERTON holds the opposite view--that we cannot speak of an original canon-which he has expressed in the Introduction to his excellent account of the Buddhist Hybrid Sanskrit, Grammar and Dictionary (1953). A summary of the Grammar and an account of the BHS literature can also be found in EDGERTON'S Lectures on Buddhist Hybrid Sanskrit published by the Banaras Hindu University (1954). J. Bloch's book Les Inscriptions D'Asoka (1950) is a very valuable publication since E. HULTZSCH published his work in 1925 and it testifies to the author's grasp on the Indo-Aryan field. The new book includes all the finds except the recently discovered versions of the minor rock edict at Rajula-Nandagiri in the Andhra and at Gujarra in the Vindhya Pradesh. The very extensive field of MIA has been ably covered by S. K. CHATTERJI in his Wilson Philological Lectures delivered at the Bombay University (1954), the publication of which will be eagerly awaited. M. MAYRHOFER'S Handbuch des Pali (1951) is primarily intended for the Indogermanist who wishes to know the nature of a MIA language. Sukumar SEN'S Comparative Grammar of MIA (1951) is now followed up by his Historical Syntax of MIA (1953). No other book on MIA syntax has been written since Jose CANEDO wrote Zur Wort- und Satzstellung in der alt- und mittleindischen Prosa (1937). G. DAVANE'S Ph.D., dissertation, completed under the guidance of S. M. KATRE, on Nominal Composition in MIA has now been published by the Deccan College Research Institute (February 1956). In this field also no work appeared since W. GBABOWSKA wrote on the nominal composition in the Asokan inscriptions (Ro 1927). Very interesting are also a couple of articles by De VREESE on Apabhramsa studies in JAOS 74.1-5; 142-146. Fresh material for the study of Apabhramsa is made available by the edition of Paumacariu of Svayambhu by H. C. BHAYANI (1953). P. B. PANDIT'S three lectures in Hindi on Prakyta Bhasa delivered at Banaras Hindu University were published last year (1954). Scholars are not lagging behind in the NIA field. A book on general Phonetics written with special application to Marathi is Dhvani # Since the Address was read out, the discovery of a version of the ninth rock-ediet of Asoka, near Bombay, has been announced by N. A. GORE in "Times of India" dated 4th January, 1966. Madhu Vidya/253

Page #279
160 M. A. MEHENDALE vicara (1955) by N. G. KALELKAR. It is a good example showing how a subject can be made easy when suitable examples are drawn from the language of those for whom the book is intended. Phonetic and Phonological Study of the Word in Urdu by Masud HUSAIN appeared this year (1955). Similar work on Bhojapuri by B. N. PRASAD and on Aspiration in Oriya by G. B. DHALL are awaiting publication. Though not in a book form I may mention here articles in Hindi on general phonetics like 'Varna-mimarhsa' and 'Varpa-uccarana' by Siddheshwar VARMA published in the Kalpana of Hyderabad. On Hindi and its varieties some important books have appeared. I may mention here Dhirendra VARMA's work on Vraja-bhasa (1954), Baburam SAKSENA'S lectures on Dakkhini Hindi (1952), and U. N. TIWARI'S work on Bhojapuri Bhaza aur Sahitya (1954), and Hindi Bhasa ka Udgama aur Vikasa (1955). A. K. PRIYOLKAR may be congratulated for having brought to light (1954) the hitherto unknown first Marathi Grammar written in Marathi by Marathi scholars. It was composed in about 1824 by KRAMAVANT, PHADAKE and GHAGAWE Shastris. An account of the Gujarati translation of this work by the same Shastris serving as the first Gujarati Grammar is given by K. B. VYAS in the Journal of the Gujarat Research Society, 17.287-299, 1955. PRIYOLKAR has also given us last year an English rendering of the Portuguese Grammatica Marasta, a book written by a missionary and first published in Rome in 1778 describing the dialect spoken round about Bombay. Linguistic Peculiarities of Jnanesvai was published by M. G. PANSE (1953) on the basis of a manuscript which in the opinion of the author is the oldest so far traced (A.D. 1350). A work written in old Gujarati or old western Rajasthani viz. Kanhadade Prabandha has been critically edited by K. B. VYAS (1955). As regards books written in Indian languages on certain aspects of linguistics, I may mention Arthavijnana (1951) in Hindi by Baburam SAKSENA, Sabda-ane Artha (1955) in Gujarati by Bhogilal SANDESARA, Sabda-Udgama va Vikasa (1953) in Marathi by K. P. KULKARNI, Arthavijnana ani Marathi Bhaga in Marathi by S. G. TULPULE and Usha POTDAR (GHATE) (in the Maharastra Sahitya Patrika, 1953), Gujarati par Arabi Pharasiyi Asar (1954) in Gujarati by C. R. NAIK, and Vagvyapara (1955) in Gujarati by H. C. BHAYANI. Though not in an Indian language I may mention here also Lectures in Linguistics by O. L. C. AGUILAR (1954). In the field of Dravidian Linguistics I may refer to R. G. HARSHE'S authorised English translation (1954) of The Grammaticul Structure o Madhu Vidya/254

Page #280
SOME REMARKS ON LANGUAGE OF ORIGINAL BUDDHIST CANON 161 Dravidian Languages by J. BLOCH which very clearly brings out the chief characteristics of this group of languages. N. LAHOVARY'S Substrat Linguistique Mediterraneen, Basque et Dravidien brings out the affini. ties between these two families of languages and seeks to establish that Dravidian was originally a speech of the big mediterranean family of languages. With regard to the studies of the individual languages I may mention R. P. Sethu PILLAI's Words and their Significance: A Study in Tamil Linguistics (1952), C. R. SANKARAN'S Some Problems in Kannada Linguistics (1954), and A. C. SEKHAR's Evolution of Malayalam (1953). Murray FOWLER's phonemic analysis of the Sanskritized literary Tamil spoken in Madras is presented in the article 'The Segmental Phonemes of Sanskritized Tamil, Language, 30.360-367 (1954). T. BURROW and S. BHATTACHARYA in their book on Parji Language (1953) establish that it is an independent language and not a dialect of Gondi. Of great interest is the English version of 17th century Dutch Grammar of Tamil by J. A. B. van BUITENEN and P. C. GANESHSUNDARAM in the Bull. DCRI, Vol. 14. 168-182 (1952). The Jules Bloch Memorial Volume as well as the Chatterji Jubilee Volume of Indian Linguistics (Vol. 14, 1954 and Vol. 16, 1955) contain many important articles on Dravidian linguistics. I may also mention here the article on Dravidian kinship terms by M. B. EMENEAU in Language 29. 339-353 (1953). A. D. TASKAR was recently awarded Ph.D. by the Poona University for his work on 'Intonational Patterns in Indo-Aryan and Dravidian' which he completed under the guidance of C. R.. SANKARAN. A new approach to the typology of phoneme and morpheme distributions as well as a semanto-phonetic study of Dravidian morphemes is being carried on by P. C. GANESHSUNDARAM under the guidance of C. R. SANKARAN. Chaitanya Deva's work on the Tonal Structure of Tambura, which he completed under the guidance of C. R. SANKARAN, awaits publication. C. R. SANKARAN takes his departure from the usual and timehonoured approaches towards the problem of speech-structure by considering the between' (which 'non-temporal-wedge' he calls alphaphoneme) in any consonant vowel configuration as the starting point of his investigations. This departure from the conventional outlook in Phonetics at all levels has led him to a unification of different levels of experience like the articulatory (or auditory-i.e. neuro-physiological) and the physical (the acoustical). Such a study of the 'inter-phenomenon' (or 'the between') to which he was led by consideration of such phenomena as marginal speech-sounds, glides, or more particularly the so-called aytam in old Tamil which has served as a pointer in all his investigations, has taken C. R. SANKARAN far beyond to abstract and symbolic levels of thinking as reflected in many of his latest writings Madhu Vidya/255

Page #281
162 M. A. MEHENDALE published in the Bulletin of Deccan College Research Institute which includes his Presidential Address at the last session. I may add that a paper of C. R. SANKARAN and his collaborators on 'Structure in Speech-The Physical Reality of the Phoneme' is awaiting publication in the Sonderheft of the Fernmelde-technische Zeitschrift of Bonn. II I shall now turn to say some words on a problem which I think is of some importance to the study of MIA languages. I had originally planned to give a critical review together of LUEDERS' Beobachtungen uber die Sprache des buddhistischen Urkanons and EDGERTON'S Buddhist Hybrid Sanskrit Grammar, to both of which I have already referred. This has, however, not been possible. My review of the first part of LUEDERS' Beobachtungen has been already published in the 17th Volume of the Bulletin of the Deccan College Research Institute (1955) pp. 53-75. I therefore planned to give a review of the second part of LUEDERS' book as the main theme of my talk today. Considerations of time again have compelled me to restrict myself only to a few points. In the Zweites Hauptstuck of this book are included LUEDERS' views regarding the phonological and morphological peculiarities of the eastern language of the original Buddhist canon. About the former, we get a detailed discussion of such phenomena as the softening of surds, loss of intervocal consonants, and the treatment of the consonant clusters. About morphology, however, what has survived is only the late Professor's remarks on the flexion of the nominal stems in -a. In order to assess the phonological peculiarities of the original canon, LUEDERS starts from what he considers to be the eastern characteristics of the Asokan inscriptions, and if he finds that these same also. occur in some instances in Pali, he regards them as borrowings from the For the solution of this problem, then, it is of importance to examine the Asokan data very critically and come to certain conclusions. as to what may be called eastern and what non-eastern in the Asokan inscriptions. LUEDERS regards softening of the voiceless stops as an eastern characteristic and for this he gives instances of the change of k>g, t>d. kkh>ggh, and tt or tth>dd or ddh. For the change of k>g he cites (SS87) the following from the Asokan evidence-Sk. loka>loga (cf. hidaloga, palaloga) in the Jaugada separate edict and Sk. adhikrtya> adhigicya in the Calcutta-Bairat inscription. Now in a paper published in the Journal of the Oriental Institute, Baroda, Vol. I. 240-244 (1951-52). I have shown that the two separate Madhu Vidya/256

Page #282
SOME REMARKS ON LANGUAGE OF ORIGINAL BUDDHIST CANON 163 edicts at Dhauli and Jaugada show certain peculiarities which are never to be found in the versions of the major rock edicts at these two places. Further, as these peculiarities are found in the other non-eastern regions, it is legitimate to conclude that these peculiarities of the separate edicts are not to be considered as eastern but to be due to non-eastern influence. Among these comes the softening of the vioceless stops. The instance loka: loga cited from the separate edict, therefore, cannot be brought forward to establish that softening was an eastern feature. The case would have been certainly otherwise if such an instance were available also from the versions of the major rock edicts at Dhauli and Jaugada. As regards adhigicya it may be pointed out that the find-spot of the Calcutta-Bairat inscription is the head-quarters of a tahasil in the Jaipur State, and an instance of voicing from this inscription, unless corroborated by other evidence, can hardly be considered as showing that particular feature as an eastern characteristic. About the change t>d, LUEDERS cites (894) Sk. hita>hida in the Kalsi, Shah.; and Mans. versions but hita in the Dhauli version. Now it is difficult to know why LUEDERS regards this as an eastern peculiarity when the words noted above show that the change of t>d is witnessed in the northern and north-western versions, but not in the eastern ones. Another instance, Sk. tosa > dosa appearing also in a northern version (Kalsi VI), we may set aside as LUEDERS takes it to be a 'Schreibfehler. But we can certainly take into account the Mansehra form yadra, (Sk. yatra), according to BUEHLER's reading, or ya(d) da, according to HULTZSCH's reading, which also supports the view that the voicing was a non-eastern tendency. But since LUEDERS believes that the change of t>d is an eastern characteristic he has some difficulty in accounting for the presence of t for d in such forms of the separate edict as patipatayeham, patipatayema, etc., from Sk. prativ pad. He observes (p. 81), "Wichtiger ist noch, dass der Redaktor. von Jau. Sep. in seinem Bestreben, die Hochsprache zu gebrauchen, in alle Formen von padipadayati ,zukommen lassen, bewirken" (Pali patipadeti, Sk. pratipadayati) das d falschlich durcht ersetzt hat...." Actually the case seems to have been that since, as suggested above, the separate edict was originally composed in a noneastern dialect it probably contained some other words changing t>d, and these the redactor rightly changed to t. But this misled him in doing the same about patipatayehan etc., because he probably confused the forms of Vpad with those of Vpat. Madhu Vidya/257

Page #283
164 M. A. MEHENDALE As regards the change kkh>ggh, LUEDERS cites ($ 149) Sk. saksyati (future of Vsak): As caghati or caghamti in the 4th PE and caghatha in the separate edicts. The Asokan inscriptions show a base Vcak besides Vsak < Sk. sak. The future base from this stem would be caksya which in the eastern dialect would appear as Vcakkha ACcording to LUEDERS this has further become caggha in the above instance as softening in his opinion is an eastern characteristic. But possibly the Asokan evidence is not clear enough to conclude Vcagghu with the softening of kh>gh to be an eastern form. Its occurrence in the separate edicts makes it a doubtful case in this regard and suggests a possible case of non-eastern influence for softening. But its occurrence in the pillar edict IV, which has an eastern version, is a little: difficult to explain. As I do not regard softening to be an eastern feature I would expect the pillar edict IV to show a base cakkha, and not Vcaggha, if that has to be derived from Vcaksya- (=Vsaksya-). But just on the strength of this one instance I would not like to consider softening to be an eastern tendency. The explanation of the form has perhaps to be sought elsewhere. So far as Pali sagghasi, mentioned by LUEDERS, is concerned, the reading is not absolutely certain and in the context the reading agghasi given by other manuscript suits equally well. And even though sagghasi be the original reading it can be regarded as a non-eastern form arising from eastern sakkhasil. As re.. gards Asokan Voagha I may put forward for your consideration the suggestion that here perhaps we have a base Vcagha of the present tense, and not caggha of the future from Vcak (vsak). This Vcagha indicates that the old IE base from which it is derived had a voiced aspirate in it. It will thus be * kagha (or *kegh) 'to be able, to help' as once suggested by ZUPITZA in Die Germanischen Gutturale (=Shriften Zur Germanischen Philologie. Achtes Heft, Berlin 1896) P. 104, (cited by WALDE-POKORNY 1. p. 333) instead of */kak-kek-?). This IE *kagh seems to have given two bases in old Aryan *Vsagh and Vsak and of these Vsak occurs in the Asokan inscriptions as Vsaka or Vcakx and * sagh as Vcagha. Aryan *Vsagh is perhaps found also in Sk. sagma 'strong or able' which is generally related to the other base Vsak. If this interpretation is accepted Vcagha of the Asokan inscriptions would represent a case of old preservation and not of voicing of intervocal -kh In support of his contention that the change tt> dd is an eastern tendency, LUEDERS cites ($ 152) from Asoka Sk. amravartika > anba. 1. Or it can be explained as coming from Sk. saghnoti 'to take upon one's self, etc.' 2. WACKERNAGEL, Alt. Gr. 1.225 considers the palatal x of Vsak to be secondary. In that case the IE base would be " sagh, Madhu Vidya/258

Page #284
SOME REMARKS ON LANGUAGE OF ORIGINAL BUDDHIST CANON 165 vadikya in the seventh pillar edict at Topra and the Queen's edict at Kosam. Now as noted by LUEDERS himself (p. 118) the eastern treatment of Sk. rt is t and not d, cf. katavirxe etc. in the Dhauli and Jaugada major edicts. In my opinion -vadikya is no exception to this general tendency since the form occurs in the north in the seventh PE of Topra for which there are no eastern parallels. And as regards the Queen's edict, be it noted that Kosam is situated on the left bank of Jamna, about 28 miles west by south of Allahabad, which can hardly be expected to give evidence for an eastern peculiarity unless corroborated by other clearly eastern evidence. The same thing has to be said about the change tth > ddh ($ 152) when LUEDERS cites Sk. astakrogikani > adhakosikyani and Sk. nislisti > nimsidhiya both in the seventh pillar edict at Topra. The normal treatment in the east in similar cases is st or sth > th, cf. setha in the Dhauli major edict, nithuliya in the pillar edicts etc. As noted above the instances which occur only in the seventh pillar edict at Topra should not be used as Asokan evidence in support of a particular feature being eastern. If the same feature is not found in the major edicts at Dhauli and Jaugada or in the other pillar edicts there is every reason for its being treated as non-eastern. In view of the above examination of the Asokan evidence it is difficult for me to agree with LUEDERS when he observes (8 155) -- "Uber. blickt man das gesamte Material, das uns die Asoka- Inschriften und die literarischen Prakrits bieten, so scheint mir deutlich daraus hervor. zugehen, dass die Erweichung des tt, tth zu dd, ddh eine Erscheinung ist, die der Ostsprache angehort." We may now take a couple of instances to show how his hypothesis regarding softening to be an eastern characteristic has led LUEDERS to support or suggest some far-fetched derivations. In $$ 159-165 he gives a very instructive Exkurs uber kasuti, kaddhati, kassati. In Pali we have the verbal base V kaddha, besides kasa (kamsa) and Vkassa which latter go back to Sk. Vkrs or v kars. Now about the origin of V kaddha, WEBER naturally thought to connect it with the Sk. p.p.p. krsta. But LUEDERS objects to this and observes (p. 125), "Die Bestimmung der Heimat von kaddhati ist fur die Entstehung der Formen von Bedeutung. Kaddhati kann sich nicht aus krsta weiter entwickelt haben, wie Weber, allerdings durch die vielfach im Prakrit auftretenden Schreibungen mit tth veranlasst, ZDMG. 28,375 annahm, da der Ubergang von inlautendem trh in ddh auf die ostliche Sprache beschrankt ist. So gewinnt die Zuruckfuhurung von kaddh auf ein vorindisches *krzd, eine Madhu Vidya/259

Page #285
166 M. A. MEHENDALE Nebenform von krs, die Bloomfield JAOS. 41,465 vorgeschlagen hat, an Wahrscheinlichkeit." In the article referred to above BLOOMFIELD notes a 'root-determinative' d in Aryan tongues in such bases as idis-d, pid= pis-d, etc. On this analogy he offers a suggestion, described as 'daring' by himself, that Pali-Pkt. kaddh may go back to kyp-d> *krzd. But this seems impossible because the forms derivable from krzd do not only not appear in Sanskrit or Avesta, but have no parallel in any other Indo-European language. (It is not necessary to discuss GEIGER'S suggestion (SS 130) to consider kaddhati Page #286
SOME REMARKS ON LANGUAGE OF ORIGINAL BUDDHIST CANON 167 But vetana cannot contain the suffix -tana as this suffix is used for the formation of adjectives from adverbs, cf. nutana, sanatana, etc. (WACK.---DEBR. Alt. Gr. II. 2. $ 444). I am, therefore, inclined to equate vetana with Sk. vedana 'wealth, possessions' from Vvid 'to find, to obtain, to give, etc.' which is known since the Rgveda. This vedana seems to have become vetana in some north-western dialect with the devoicing of inter-vocal stop, from where it was taken up again in the OIA and MIA languages. In the course of time vedana also seems to have come to mean 'prize, reward, etc.'. For the semantic change from 'property' to 'reward' we may compare a similar change in the Germanic family from Gothic faihu (Sk. pasu) 'money, property to Eng. fee. Vedana in Pali then may represent the translation of an eastern middleIndic vetana or it may be just the preservation of the old vedana continued in some spoken dialect. As regards the change p > I think LUEDERS (88 99-100) is right in regarding this to be an eastern characteristic. But in his articles on this subject he does not adduce the available Asokan evidence, though small, in support of this view. For instance, among the minor rock edicts only the Sahasram version in the east gives V pava for Sk. pra-vap, while the non-eastern versions of Rup., Brahma., and sidda. give v parpu. That the base v pava had not thouroughly driven out from the east the other base v papa in the Asokan days is shown by the fact that we get such forms as papova in the sixth Araraj pillar edict and papova in the corresponding Topra version. Apart from this, I wish to bring to your notice another explanation of a word dealt with by LUEDERS under Hyperpalismen (8$ 144-147) with regard to the change p > v. I am referring to Pali supana dog ($ 146). LUEDERS considers that the strong form of Sk. svan gave rise to a base suvana in the eastern dialect which was wrongly translated into Pali as supana as the translator was aware of the fact that the eastern dialect changed p > v. Now such a misunderstanding seems unlikely in the case of a word for dog, and if the Pali translator used supana it means that that must have been the regular western correspondent for the eastern suvana. This is quite likely since corresponding to Sk, svan ( *kuon) we have span (nom. sg. spa, acc. sg. span @m) in the Avesta. This span, or a middle-Indic base from it spana, probably survived even in India in some spoken dialect from which we can have the Pali form supana, just as we have suvana from Sk. svan or svara in the eastern dialect. The Gk, word spaka 'hundartig' (besides kuon, kunos) and the Russian word sobaka. 'Hund' show that forms with p were current in other Indo-European languages besides Avesta. Supana is thus an Madhu Vidya/261

Page #287
168 M. A. MEHENDALE RSESE (BSOS 172 self against the Maharastri lof instance not of hyperpalism but of a regular development from an archaic dialect form spana agreeing with the Avesta. I shall now take a couple of cases with regard to LUEDERS' observations on the nominal flexion. In $$ 188-195 LUEDERS has convincingly demonstrated the use of a special abl. sg. in -am in the eastern language of the Buddhist canon. Recently De VREESE (BSOS 17.369-371, 1955) has expressed himself against the use of -am as abl. sg. in old JainaMaharastri (cf. ALSDORF, BSOS 8.329 ff., 1936). But I do not consider that De VREESE has proved his case. The use of the acc. instead of the abl. with certain verbs in the Buddhist Hybrid Skt. does not disprove the argument of ALSDORF (De VREESE had evidently not seen LUEDERS' treatment when he wrote his article), and the interpretations put by De VREESE on the Pali passages are far-fetched. I shall not enter here into the details of my belief that this abl. sg. -am is perhaps not a regular development of -at but is only a graphic representation of the term. -a. I would, on the other hand, add one more instance showing the use of an as abl. sg. to those already cited by LUEDERS. This instance occurs in the Dh. 135 (cited by LUEDERS in SS 140 under a different topic) which runs as yatha dandena gopalo gavo paceti gocaran/ evan jara ca maccu ca ayun pacenti paninam// Max MUELLER-"As a cowherd with his staff drives his cows into the stable, so do Age and Death drive the life of men." LUEDERS rightly objects to gocara being translated as "stable". He corrects it to 'Weide' but continues to regard gocaram as acc. sg. To my mind, however, gocaran cannot be acc.; context shows that it must be abl. In the second half of the stanza we are told about (Old) Age and Death driving back the life of men. Hence in order to understand the simile correctly we must interprete that in the first half of the stanza there is reference to the driving back of the cattle from the gocard and not to it. The first line, therefore, means-"Just as a cowherd with his stick drives away the cows from the pasture......". In $$ 220-225 LUEDERS very ingeniously demonstrates the use of -hi as loc. pl. in the eastern language of the original canon. However, his interpretation of akkhesu in the Suttanipata 659 ($ 221) as a misunderstood loc. pl. does not quite carry conviction. The stanza runs as_ appamatto ayan kali/ yo akkhesu. dhanaparajayo/ sabbassapi sahapi attana/ asam eva mahattaro kali/ vosilgatesu -manam padosayell Madhu Vidya/262

Page #288
SOME REMARKS ON LANGUAGE OF ORIGINAL BUDDHIST CANON 169 According to LUEDERS the stanza means, "Das ist ein geringer Ungluckswurf, wenn einer durch die Wurfel Geld verliert, selbst die ganze Habe samt der eigenen Person. Das vielmehr ist ein grosser Ungluckswurf, wenn einer gegen die Heiligen bose gedanken hegt." LUEDERS considers akkhesu as a misunderstood loc. pl., that is to say in his opinion the original canon had akkhehi as instr. pl., but since this could also be a form of the loc. pl., the Pali translator wrongly gave it as akkhesu. Now in his critical analysis of the second quarter of the first line, LUEDERS has rightly seen that in the original canon we had two separate words dhanam palajaye and that palajaye was a verb form. This has been happily confirmed by the Udanavarga version (8.4) which gives---dhanam parajayet. LUEDERS further argues that the verb form palajaye of the original version was misunderstood by the Pali translator as a nominal form in the nom. sg. and hence he changed dhanam to dhanu and joined the two together in a compound dhanaparajayo. To me it does not seem very likely that the Pali translator could have made a mistake about the verb form if dharum palajaye stood separately in the version before him. The more likely explanation, therefore, seems to be that in the version before the Pali translator dhanampalajaye were read together and hence he mistook it to be a compound formed on the analogy of the more common dhanarjaya (cf. Vopadeva 26.60). But as this was an unusual formation for the Pali translator he changed dhanampalajaye to dhanaparajayo. But whether palajaye is taken as a verb form or a noun form its use with loc. in such constructions as above seems to be quite idiomatic. One may compare here favourably similar uses in some of the NIA languages-'he lost money in horses, in cards, etc.' where one uses loc. and not instr. In the above stanza, therefore, akkhehi could be regarded as loc. pl. and hence rightly rendered by the Pali translator by akkhesu. The Udanavarga translator, however, does not seem to have understood the idiom, and hence he interpreted akkhehi as instr. pl. and rendered it as aksena. The Pali verse is to be translated as "that is an insignificant loss (lit. the unlucky throw, kali) which is loss of money at gambling (lit. dice, akkha).... This indeed is a greater loss when one spoils one's mind against the holy persons." III Before I conclude let me refer to two significant events of the past two years which promise to give a great impetus to the linguistic studies in India-I mean the organising of a series of schools of linguistics at the Deccan College, and the strengthening of the Linguistic Society of India. Thanks to the initiative of the Director and the Council of Management of the Deccan College Research Institute on the one hand, Madhu Vidya/263

Page #289
170 M. A. MEHENDALE and the Rockefeller Foundation on the other a series of three post-graduate Schools of Linguistics was organised in 1954-55. I need not go into the details of the working of these Schools; their success can be seen from the great interest they evoked at each session. Perhaps for the first time in the recent history of University education in India it was possible for the students from all over India to come together and receive instruction from a Faculty constituted by drawing personnel from the Indian as well as foreign Universities. With the growing importance attached to the structural studies of our languages in a free India, the need for imparting scientific instruction in linguistics has been felt, and the holding of short-term schools has partly sought to satisfy this need. The attempt is limited, but it definitely shows steady awakening. However, what is really necessary is the starting of a full-time course in linguistics at all important Universities. This has been often suggested, but for one reason or the other not executed. It is significant, therefore, that our host University here has established a Silver Jubilee Chair in Dravidian Philology from non-recurring grant received from the Union Government, as a beginning in this direction. Let us hope that with the growing country-wide interest evinced in the subject and the proposed establishment of the Linguistic Survey of India by the Central Government, the other Universities will take adequate steps to play their part at an early date. The great task of taking up the descriptive studies of various dialects is ahead of us and this will require a band of field-workers well. grounded in the tool courses of linguistics. The second point to which I would like to refer is with regard to the strengthening of the Linguistic Society of India. I have great pleasure in announcing that the membership of the Society has recently increased from less than fifty to about three hundred. As regards the publication of the official Bulletin of the Society in future it is proposed to bring out the Volumes of Indian Linguistics regularly twice a year towards ultimately converting it into a quartery. Since last year an attempt is being made to hold the meetings of the Linguistic Society annually. In this regard I would request the Universities and Research Institutes to give recognition to this body for the purposes of sending delegates to its meetings and for strengthening its effective membership. Then again it is proposed that at different places where there are ten or more local members, they might form a regional or local circle and hold meetings at least once in two months for reading papers, discussing problems, or even undertaking some organised effort in dialect studies. The reports of these activities of the circles can be published in the Volumes of Indian Linguistics, and even some small monographs can be separately brought out. All such activities will not only help in keeping up the interest of Madhu Vidya/264

Page #290
SOME REMARKS ON LANGUAGE OF ORIGINAL BUDDHIST CANON 171 the members, but also in raising the Society in the eyes of the University authorities to give it due recognition. now conclude with a couple of quotations. The late Dr. SUKTHANKAR said in his sectional President's address at Tirupati 15 years ago, "It is the debt to the rais, which is difficult to discharge and which usually remains unpaid. Let us, however, follow the mandate of the scriptures and let us not forget our debt to the rsis, even if it has been neglected in the past. Let us not forget our debt to the Maharsi Papini, who has made the name of our country resound in the halls of the academies of the world. Let us endeavour by our assiduous and fruitful study to keep bright the fair name of that illustrious Muni of imperishable fame, Panini!" And as recently as 17th October 1955, while requesting Shri Balasaheb KHER to inaugurate the Autumn School of Linguistics, Dr. KATRE repeated, "With your blessings, we are confident that the beginning made here will prove itself fruitful and produce not unworthy successors to the great Acharyas and Rishis who established the science of Linguistics for the first time in India. May these efforts succeed in raising a Panini or a Patanjali in the next few generations to uphold and sustain the position of India as a leading country for the scientific study of languages!" Madhu Vidya/265

Page #291
Sanskrit badhi 20.1 ugnidhro 'bhigyhnati / 'avabadho 'ghasansa iti trtiye / 'avabadha yatudhana iti caturthe (2.2.2) Der Agnidhra bedeckt das Hingestreute (rather dug down, buried') beim zweiten Hinstreuen (rather 'pressing, burying') mit der Formel: ,,Nach unten gedrangt ist das Raksas", beim dritten mit: ,,Nach unten gedrangt ist der Boswillige", beim vierten mit: ,,Nach unten gedrangt sind die Zauberer"8 (Caland). Still in another ritual regarding the preparation of the purodasa, when the husk is separated from the rice taken out for this purpose, the officiating priest puts the husk on the potsherd and pushes it below the black antelope skin. He then presses it down with the mantra 'the evil spirit is dug down, buried' (purodasakapalam tusaih purayitva 'raksasan bhago 'siti ... adhastat krsnajinasyopavapati / nanviksate l'avabaoham raksa' ity avabadhate/' Hiranya. 1. 5. 17). The ritual acts described above should leave no doubt about the meaning of avabadha 'dug down, pressed down, buried'. This follows from the fact that the mantra containing this word accompanies the action of digging and pressing down the charm etc. into the ground. In the RV. 1. 106. 6 (= AV. Paipp. 4. 28. 6) we get nibadha which has the same meaning as avabadha: indram kutso vrtrahanan sacipatinn kate nibalha rsir ahvad utaye / ,,Indra, den Vrtratoter, den Herrn der Kraft, hat der Rsi Kutsa, als er in die Grube hinabgestossen 10 war, zu Hilfe gerufen" (Geldner). Pressed down' or even 'buried' would be better than just 'hinabgestossen'. Baoha without any prefix occurs only once in the Rgveda 1. 181. 7. There it appears to have been used as a substantive, meaning 'dug out * Also Hiranyakesi, with minor variations, 1.6.21. Caland (on Ap. Sr. S. 2.2.2) remarks that these three formulas are not found anywhere else except in the Ap. and the Hiranya. Sr. S. But they are given by the Varaha Sr. S. 1.3.1.42 (with sphyenot. karam apidhatte for abhigyhnati). Commentator: hastenavastabhyoptatusadesam budhale. Ap. Sr. S. mentions only the placing of the husk below the antelope skin (1.20.9), but does not give the following mantra for pressing it down. 10 zvibadha occurs in the general sense of 'overpowering, oppressing' in the Jaim. Br. (Auswahl by W. Caland, pp. 196-197): sa ha Khandikah Kesinam abhibabhuva / sa ha Kesi Khandikena nibadha Uccaihsravasam Kauvayeyam jagama / ,,Khandika nun uberwaltigte den Kesin. Kesin, von Khandika bedrangt, begab sich zu Ucchaihsravasa, der Sohn des Kuvaya". According to PW, additions to Vol. 7 (p. 1779) we get sambalha in the Taitt. Ar. 1.17 which, as mentioned in PW, means sambaddha, drdha according to the commentator. I have not been able to find this word in the reference given. The Vaidikapadanukramakosa of Hoshiarpur also does not give it. Keith, Ait. Ar. p. 189, n. 10), apparently following PW, simply says, ,,It occurs in the Taittiriya Aranyaka" but does not give exact reference. The reference in PW is obviously mistaken and seems to be intended for sambalhatama which occurs in the Ait. Ar. 1.4.1. However, the commentator's explanation referred to by PW remains a moot point. Madhu Vidya/266

Page #292
82 M. A. MEHENDALE in the Jaipur State, Rupnath (ru) in the Madhya Pradesh, Maski (mk) and Kopba] (kpb) in the south Hyderabad State, and Yerragudi (yr) in the Kurnool district of the Andhra State, show the non-eastern influence in their language. Unlike the Mysore versions, these six versions are found at places far away from each other and hence they are not quite identical in their text and language. In the first instance we have to note that the Sahasram version, which is in the east, was translated into the eastern dialect; the Bairat version is more or less in agreement with it and thus it shows the extent of the influence of the Magadhan dialect. The other versions on the other hand are more or less free from the eastern influence in Phonology and show adherence to the non-eastern forms. Since many of the important points as to what may be considered eastern and what non-eastern have been already discussed and the reasons why eastern features are found in the minor rock edict have been stated, these considerations will not be repeated here. The correspondences of the non-eastern forms occurring in the minor rock edict with those in the Niya Prakrit, the later Kharosthi inscriptions and the literary Prakrits have also been pointed out in the paper on the Mysore version referred to above. It will be sufficient in this paper here to show the regional distinctions which occur in the remaining versions of the minor rock edict. However, I have referred to in the foot-notes for the sake of comparison certain forms occurring in the Bhattiprolu inscription in which LUEDERS has observed a north-western characteristic." One thing may be noted regarding the contents of the versions. While most of the versions end with the declaration that the edict was. issued while the king was on tour, and that he had been on tour for 256 nights, only the Mysore and the Yerragudi versions add to it some more matter describing what the king considered to be the dhammaguna. Even with regard to this additional passage the Yerragudt version is more extensive than the Mysore version. Unfortunately the Yerragudi version is not well preserved and hence it has not been fully explained. br-Brahmagiri version of the minor rock edict. sd-Siddapur version of the minor rock edict. MRE-The versions of the Minor Rock Edict. 8. Only the Gavimath and the Palkigundu versions at Kopbal are near to each other and are identical in contents. The Palkigundu version, however, is much damaged. 4. The s version shows only the following few non-eastern features which it has allowed to remain from the original draft. These are-diyadhiya, savachala, pavatave, and palakamamina which are discussed below. 5. On the basis of the distinction made in the sibilants and 8 in the Bhattiprolu inscription, LUEDERS (Philologica Indica, p. 217), observes-Ich bin daher geneigt, den Dialekt von Bhattiprolu diesen Dialekten anzugliedern und in den Leuten, die den Stupa errichteten, Kolonisten aus dem Nordwestern zu sehen." Madhu Vidya/267

Page #293
VERSIONS OF ASOKA'S MINOR ROCK EDICT There is one more point in which the Yerragudi version differs from the others. While all the other versions are written from left to right, the Yerragudi version, though written in Brahmi, shows certain lines being inscribed from right to left as was the custom for the Kharosthi script used in the north-west. While commenting on this state of affairs, BARUA (IHQ 9.114-15, 1953) observes, "The anamoly in the engraving of the inscription might be partly due to the fact that the scribe entrusted with the work was inefficient and careless, and partly due to the fact that he was so much habituated to writing the Kharosthi form of writing from right to left that it was very difficult, nay, impossible for him to shake it off even in writing a Brahmi inscription." But this is a little difficult to admit, because there is at least some consistency in writing almost always the even lines from right to left, and it would be much better to explain the confusion as arising from the fact that it reflected an attempt on the part of the scribe to combine the two ways in which the Brahmi and the Kharosthi scripts were written--the former because he was actually writing the edict in that script, and the latter because the original draft was written in that script. This is a small piece of palaeographic evidence to show that the original draft was issued from the north-west. (For another possible palaeographic evidence, see below p. 94 footnote 58). Two announcements have been made in recent times regarding the discovery of two more versions of the minor rock edict-the one at Rajula-Nandagiri near Pattikonda in the Kurnool district (the find-spot. is only twenty miles from Yerragudi), and the other in a forest near the village Gujarra in the Datia district, Vindhya Pradesh. Both these versions have not been yet published and hence they are not treated in this paper. When published, these versions would be of great help either to support or alter some of the observations made in these two papers on Asoka's minor rock edict. With these introductory remarks I proceed to show the northwestern (or western) characteristics found in the different versions. (1) The vowel ? : In these versions there are not many words with which express human relationship. The only available instance is -pitu- (pitr) occurring in yr version, and it shows the north-western 6. I have already pointed out in my paper on the Mysore versions that the last word in that version, viz. lipikarena is written in the Kharosthi alphabet and that this is one of the facts pointing to the north-western origin of the edict. 7. Indian Archaeology 1953-54. A Review, New Delhi, 1954. 8. "The Times of India", 10th December, 1954. Madhu Vidya/268

Page #294
84 M.A. MEHENDALE treatment ?>u observed in this class of words, and not the corresponding eastern ? > i treatment. cf. pitu and bhratu in S, M, as against piti and bhati in D,J. (2) The vowel e: In a few cases the vowel -e tends to become - in the north-west 10 cf. duv[i] for duve (dve), amni for amne (anya), rajani for rajane (rajanah). Now a similar change may be observed also in kpb in the instance upeti for upete (upetah),11 and in ru in pavatisu12 for pavatesu (parvatesu). (3) The vowel a: With the above tendency may be compared the north-western tendency to change a >i in certain instances.18 For example, ayi (ayam) in S, M, vinikramani M(= vikramanan s). Now with this compare budhi (badham) in rull (but badhan and badha in the other versions). Here we may also consider the forms of the demonstrative pronoun etad, which in some instances in the north-west gives the base eti_15 for the usual eta-. cf. etisa in S, M (etisa in K), but etasa 9. Also cf. the following words in the Bhattipralu inscriptions (references to Philo logica Indica): sapitukasa p. 218, kurapituno and kuramatu p. 220, matugamasa p. 221. 10. For similar tendency in the Niya Pkt., cf. nicinioe, vanti = uparte, T. BURROW, The Language of the Kharosthi Documents from Chinese Turkestan, (Cambridge, 1937) S 1. BURROW also notes that the change was regular in the dialect of Khotan and that similar instances are common in the Kharosthi Dhammapada. In the Bhattipralu inscription too we have kubira for kubera, p. 229 (which, in the form kupira, occurs also in a Bharaut inscription, gee LURDERS' List No. 794). 11. Among other versions, ru has upete, b upayate, br and sd upayite, yr upayite, and mk upagate (see below p. 86). 12. HULTZSCH, p. cxxvi and f.n. 1 takes this as the loc. pl, of the fem. base pavati (parvati) which is unlikely. cf. pavatesu in the s version. Similar change of -e > -i in Pali is noted by LUEDERS, Beobachtungen etc., $ 10, (p. 17) and f.n.3, cf. dani for dane (danan). With pavatisu may be compared Pali hemantagimhisu (hemantagrismesu) cited by LUEDERS. . Among later Kharosthi inscriptions we find sati (sata) and sarvina (sarva-), cf. M. A. MEHENDALE, Historical Grammar of Inscriptional Prakrits, (Poona, 1948), $ 503. Among later Pkts, change a > i is found in Mahar., Amg., and JM; but s and Mg. preserve a, PISCHEL 8 101. 14. pakamasi for pakamasa in ru and kativiya for kataviya in br may be regarded as scribal mistakes. 15. For etina occurring in the Khar. Dhammapada, see BURROW & 1. Pali has also base ti and eti in fem., cf. tiesa, tissaya, tissam, etissaya, GEIGER 8 105. Among later Pkts. M, JM, Amg. and P have bases in -, and 1 in fem., but Mg, and have in -a, PISCHEL SS 424, $ 426. Delhi-Topra pillar edict has kinasu (< kena. svit) but Sarnath has kenapi. For the base ki- in the forms of the interrogative pronoun in Pali see GEIGER $ 111, and in later Prakrits cf. PISCHEL 428. In gen. the form kisa appears also in the eastern Pkt. Magadhi. kinna, instr. in form, has been noted by BURROW $83 for Niya Prakrit. 13. Madhu Vidya/269

Page #295
VERSIONS OF ASOKA'S MINOR ROCK EDICT 85 in d, j, G. Among the versions of the minor rock edict, ru gives the base eti- twice, cf. etina (instr.) and etiya (dat.). Similarly though idam gives forms with the normal base ima-, it also shows non-eastern imi-16 in imisa S and imina G. With this we may compare imina in br, sd, and yr among MRE. The change a >i is also noticed in the non-eastern versions of MRE in the future ending -siti which varies with the eastern ending -sati (-syati). This change may be explained as palatalisation of the vowel.? Thus we have vadhisiti ru, mk, kpb, yr, br, sd, but vadhisati s, b. Now in S we have the future ending -sati in forms like vadhisati etc. where the vowel a is preserved due to eastern influence, but the palatalisation is even then seen in the sibilant , > s. The vocalic change a >i in the future ending may thus be regarded as a north-western characteristic.19 (4) The vowel 2 : In yr we get the non-eastern form garu (guru). with the vowel a (which was also found in the Mysore versions). cf. 10. Pal: 17. Pali has imina (and amina) in mas. and imissa and imissam in fem., GEIGER $108. JM, S, and Mg. have imina. PISCHEL SS 430. BLOCH, op. cit. p. 69 ob serves."Ces formes a i interieur (i.e. forms imissu, imissa and imina), auxquelles il faut joindre dans l'autre demonstratif gen. sg. Sh. M. K, etissa et Rup. instr. etina (mais dans Rup. etiyn atthaya datif, le premier mot doit etre une graphie ou une lecture fautive pour etaya) reposent sans doute sur l'interrogatif, ou le theme lci- du neutre s'est etendu au dela de son domaine primitif." This is, however, unlikely, for the variation a : i shows geographical distinction and should be related to the north-western tendency to change a to i as shown above. For a similar change of a i in the presence of y, cf. muli 'price', esvari 'ownership' and arogi 'health in the Niya Prakrit, of. BURROW 89. For future, BURROW $ 99 gives both the endings -isyati and wisati. Pali has sometimes i for a in the future ending when h appears for & in the Gatha dialect, cf. karihiti for karissati etc. GEIGER $ 150, SS 151, SS 153. Among later Prakrits, Mah., JM and Amg. have similar endings with i vowel, besides the usual endings with the a vowel: -ihisi and -ihiz or -ihi. Hemacandra (4.275) gives -insidi for Saurasen! as well. According to the same grammarian (4.302) the same ending seems to have extended even to Magadhi in bhavissidi (PISCHEL & 520). In ru once we have vadhisata which is to be read as vadhisiti, HULTZSCH, D. 167, f.n. 8. The form vadhisiti is regarded as eastern by TURNER, BSOS 6.532 (1931). This does not seem to be justified in the first instance because the tendency to change a to i is found in the north-west, and secondly because among the versions of the minor rock edict, the one at s (and b) which can be regarded as eastern gives vadhisati and not siti. The form likhiyisami occurring in D XIV 2 does not show the change ayi iyi as TURNER (op. cit., p. 532, and also HULTZSCH, p. xcix) would take it to be, but it is obviously influenced by likhite which immediately precedes it, cf. bahuke ca lilchite likhiyis[ami]. Similarly vadhiyisati of K (IV. 11) is either influenced by vadhite which precedes it, or like vadhiyati, which also occurs in K, it may be due to confusion with the future base vadhiin vadhisati. 18. 19. Madhu Vidya/270

Page #296
M. A. MEHENDALE garu G, S (and galu K), but gulu D, J; (guru occurs in G, S, M also as a borrowing). (5) The semi-vowel y: As in the case of the other non-eastern versions, the semi-vowel y is preserved initially in the forms of the relative pronoun and in the indeclinables cf. yatha ru, yatharaha (yatharha) yr, ya or yarh (yat) ru, b (line 2), kpb, yr, varisa (yadrea) yr. But the eastern s version gives am with the loss of initial y.20 am occurs also once fn b in line 3. What is surprising is to find am in mk and atha in yr. It appears therefore that the original draft contained these forms because instances like e, am and atha were already known in the north-west due to their occurrence in the M version.21 86 As in the Mysore version, we have y developed before i in upayita (upa+ita) in yr., The form upayata in b is a mistake for upayita. An analogical extension of this phenomenon in a compound form when the first member ended in i was noticed in diyadhiya (dvi+ardha) while dealing with the Mysore version. The same form occurs in all the remaining versions, including s and b. Another instance of this nature is hathiyaroha (hasti+aroha) in yr. The optative endings -yu: -vu: The yr version gives the eastern opt. 3rd pl. ending -vu as against the non-eastern -yu given by the Mysore. version, cf. parakamevu and janevu (for pakameyu and janeyu). This will show that the original draft had the ending vu under the eastern 20. This loss in the eastern versions is not regarded by BLOCH (1.c. p. 52) and J. VEKERDI (AO 3.324.1953) as a phonetic phenomenon, because y is not initially lost in the eastern versions in substantives and verb forms like yuta, yati etc. But in my opinion the phenomenon is so regular that perhaps it would be better to assume that the literary dialect has carried to the extreme the tendency which had appeared in the pronunciation of certain indeclinables and forms of the relative pronoun because they were of frequent occurrence in the speech. 21. A few other eastern forms like ava (yavat) are found at S also. 22. Of course it is possible to derive upayata from Vya. Compare, for instance, that the western G version uses Vya or Vniya, where the eastern and under its influence the north-western versions use Vnikhama (or Vnikrama) in the third and the eighth rock edicts (nikhama occurs in d, j also. The form ayaya in G VIII, can as well be derived from Vi, cf. HULTZSCH, p. lxviii, but cf. BLOCH, op. cit., p. 75). In the thirteenth rock edict, however, K and M versions have Vya, and S has vraca (G version is missing). In the same edict, as noun form vinikhamana, occurs in G also (K has vinikhamana, S nikramana, and M vinikramana). But so far as the versions of the MRE are concerned note the use of the verb Vi in the form upeta in kpb and ru. Perhaps upayata may represent an attempt on the part of the scribe to correct the form upayita of the original draft as it was not properly understood. The form upeta noticed above and upagata in mk point to the same inference. 23. Also see below p. 95 where even the base palakama is shown to be eastern, Madhu Vidya/271

Page #297
VERSIONS OF ASOKA'S MINOR ROCK EDICT 87 influence. This was rightly altered in the Mysore version in keeping with the western forms, but allowed to remain in the yr version. (6) Treatment of 1:1: As is to be expected the eastern s version gives the forms with l, and in this respect the b version also agrees with it. cf. udala (udara), cila (cira), etc. As against this, the kpb and the yr versions agree with the non-eastern treatment and show r. cf. udara, cira, etc. The ru version in Central India, however, gives forms with r as well as 1 : satireka,24 aradhe[ta] vaa5 ( a radh-), cira, and savachara, but also satileka, udala, Vpalakama (parav kreem), apalidhiya (aparardhya ?), ahala (ahara), valata (varatan). The mk version, which is further in the south, is allowed to be more free from the eastern influence in this regard. Cf..... [t]ire .., pure, but kalamta (V kar), udula ka. This state of affairs shows that the eastern speech habit to use l was known very much outside its home so that the scribes who worked for the ru and mk versions thought it proper to use some forms with I to show familiarity with the king's speech. Moreover, it is also possible that the original draft contained some of these words with l as a result of the influence of the king's speech.27 While these were completely altered in some versions, it was only partly done in the others. (7) Sibilants: As in the standard western Prakrit, the versions of the minor rock edict have only the dental sibilant in a large majority of cases. But as the north-western dialect shows distinction between the three sibilants we have at least one instance of the palatal s in Saka (sakya), and one instance of the cerebral s in vasa (varsa), both in mk.28 In b, the palatal e is found in an attempt at hyper-north-westernisation in svaga (svarga); exactly as we have saca (satya), sava (sarva), and acarinsa (acaryasya) in the Mysore version. (8) Cerebralisation : The cerebralisation of dentals is particularly an eastern characteristic. But instances of cerebralisation are found in many forms in the north-west also and hence such cases can be witnessed in the versions of the minor rock edict. Thus for t > 1,29 cf. 24. BUEHLER and SENART read satileka, cf. HULTZSCH, p. 166, f.n. 1. 25. TURNER, loc. cit., p. 8, would read aladhetava-; iladhi and Valadha- occur al ready in the K version in the north. Kalanta occurs in K as well. Some Magadhisms showing l for r have been well known in S, M, and G versions also. BLOCH (op. cit., p. 48) regads these cases as scribal errors and considers that they do not represent the actual pronunciation. True, they do not reflect the pronunciation of the west or central dialect but surely they point to the dis tinction maintained in the original draft which came from the north-west. 29. The western dental, however, can be seen in pavatitaviya (pravartitavya) and pakiti (prakrti) in yr. Madhu Vidya/272

Page #298
88 M.A. MEHENDALE kata (kyta) s, ru, kpb, and kataviya (kartavyu) yr; for th>th, cf. atha (artha) s, ru, kpb, yr; for d dh, cf. diyadhiya, adhatiya (ardhatrika), and Vvadha (Vvardh) which occur in all versions; for st>th, cf. thabha (stambha) ru, but thanbhas; for sth > th, cf. thitiki (sthiti) in all versions. Now with regard to st it is to be observed that thambha is the eastern form, as st>th in D,J, K and also M, while thanbha is the west-central form, as st remains or becomes st in G, which latter may have been further assimilated to th in the west-central dialect. With regard to thitika, it may be noted that this form with the cerebral is the eastern one (as sth > sth in the forms of Vstha) and it occurs in the minor rock edict as it was already known in the north-west through the M version.32 It may be added that perhaps in mk we have uthana (utsthana) while uthana is found in S, M also. For vyusta we have vyutha in ru and vyutha:"3 in yr as quite regtilar forms. The s version, on the other hand, gives the eastern form vivutha.34 It appears that vyutha and vivutha came from two different bases derived from vi vas., the former from vi + *ustak and the latter from vi+ *vasta.36 . The treatment of n:n: It is well known that n is eastern and n non-eastern. Now among the versions of the minor rock edict, we have n in kpb, (as also in the Mysore version), but all others give the eastern n: savana (sravana) ru, s, yr,37 porana (*paurana), guna (guna), and prana (prana) in yr. But we have suvana, etc., in kpb. In terminations, all versions have only n, but kph (and the Mysore version) has 30. This form is not known in the north-west so far as the Asokan inscriptions go. Its appearance can be explained only as emanating from the king's speech, The cerebral in this word is clearly due to the influence of the cerebral in udura, cf. khndaki ca udara ca pakamaritu ti. Bhattipralu has 8th >th in yothi (yosthi), but once th in gothi, cf. Phil. Ind. p. 224, 226, 229, and 225; also cf. kunitha (kunistha), p. 220. This is the reading of BARUA; BLOCU reads ryutha Bhattiprolu also has th in yathi (Phil. Ind. p. 221). 34. It has been already shown in the paper dealing with the Mysore versions that we have eastern >> for the non-castern y; also cf, above p. 86. 35. The forms diyachiya etc. above p. 86 would suggest that the north-western forma should be viyutha with the development of y after i. The preservation of the cluster vy shows the western influence, cf. below p. 89. With this cf. later Maharastri vuttha and Jain Mah. pavuttha, PISCHEL SS. 302. vivutha could as well come from vi + *usta and correspond exactly to *viyutha of the north-west (cf. the above f.n.); but as this is made difficult by the absence of the cerebral in vivutha, I am inclined to look upon vivutha as a mistake for vivutha, just as vyitha in the Mysore version has been suggested in the previous paper to be a mistake for vyutha. 37. yr once gives suvana also. 31. 30 36. Madhu Vidya/273

Page #299
VERSIONS OF ASOKA'S MINOR ROCK EDICT 89 vr, 1W, indhs These sometimes n: satirekani ru, but satirekani kpb, adhatiyanis, ru, mk, but adhatiyani kpb.38 The kpb version thus not only preserves n, but also changes n>n (this, however, is not done in the Mysore version) : munisa s, yr, but manusa39 kpb, Vjana s, b, ru, yr, but Vjana kpb; devanan ru etc., but devanam kpb; also cf. dani (idanim) and even initial n in no (no) 40 in kpb. (9) Consonant clusters with y: They are in most cases assimilated as in the north-west and the west. In two cases the cluster is preserved as is also the tendency observed in the west. It is assimilated in Saka (sakya) mk, saka (sakya) mk, aroka (arogya) 41 yr. It is preserved in [ca] kya (sakya) b (line 6) and yugya (?) yr. The eastern influence of dissolution is seen in sakiya:yr, ru, and cakiya kpb, s, b (line 3), acariya (acarya) yr. It may also be found in rdhy > dhiy, cf. aparardhya'i > avaladhiya s, apaladhiya ru, aparadhiya yr. These forms show that the original draft had the eastern forms cakiya, acariya, and avaladhiya. While translating into the western dialect, they were partly westernised by changing caka > saka and I > , but the eastern dissolution was retained. The cluster vy is preserved in the instance vyutha in ru and y: versions as in the west. It is dissolved in ru in the instance vayajana for viyajana42 (vyanjana). In s we have vivutha (viusta or *vivastu see above p. 88 and f.n. 36). In future participles, the non-eastern tendency of assimilation is seen in papotava ru, kpb, adhigatava mk, aradhetava yr (the only instance with -v- in yr; otherwise we have -viy-), and aladhetava b. In the form of this word as also in pavatava43 even the s version adheres to the original in so far as the ending is concerned. The eastern tendency of dissolution is seen in dakhitaviya mk, vataviya mk, 39. eper horizoning to Turner and pakam ven accounts well as in termina According to TURNER (op. cit., Text) kpb gives ? in the terminations also in vasani and mahatena. But in these two cases as well as in adhatiyuni I would read the dental -n- in the endings. Even according to TURNER'S reading kpl gives dental in khudakena and pakamaminena. This is according to TURNER. I would read manusa. There is no place for the upper horizontal stroke of n between the a stroke of the first letter ma and the e stroke of the third letter se in manusehi (line 4). 40. Here also I differ from TURNER and read no both in line 2 and line 4. The estampage and the photograph given by him show slight variation in the symbol which he reads as no. If the left middle horizontal stroke seen on the estampage is due to chance erasure and that the slight upper right hand stroke is an accidental extension of the left hand mark, then the symbol can stand for no. In line 4, we see a symbol which clearly shows that the left hand upper horizontal stroke is a matter of chance occurrence and has no value in reading the symbol as it is separated from the main letter. 41. I intend to show in a separate paper that this derivation is incorrect. 42. BLOCH (op. cit., p. 150, f.n. 13) regards these as scribal errors. Symbolne a separt. 13) intens the symbotroke ikke see a symbbe left hand the Madhu Vidya/274

Page #300
90 M.A. MEHENDALE likhopetavaya (for -taviya) and vivasetaviya ru, and sususitaviya yr. Now the eastern tendency of dissolving the cluster was already known in the north-west in such instances as viyapata or viyapratu (vyaprta), and especially in the endings of the future passive participles, cf. pujetaviya, kataviya etc. (10) Clusters with r: Such clusters are normally assimilated, cf. pakata or pakamta (prakranta), amisa (amiera), etc. The only instances where they are preserved as in the north-west are prakasa ru and prana yr. It may be observed that such clusters are sometimes preserved in G also. (11) The cluster k$: It shows the eastern ks>kh treatment and not the non-eastern ks>ch, cf. khudaka or khudaka (ksudraka) in all versions and dakhitaviya (from v*dyksa or from the future base Vdraksya) mk. Both khudaka (or khudraka) and dakhati occur as loanwords in the S, M versions of the major rock edicts.44 In fact the first major rock edict shows even the base Vdakha (or V drakha?) to be eastern as against v pasa (V pasya-) which is used in G. (12) The cluster jn : In yr we see the eastern change jn>n (and not the non-eastern n > n) in Vanapa (Vajnapa) and natiko (jnati). Now Vanapa occurs already in the S, M versions and the Mysore version which shows that the original draft had Vanapa. The yr version, however, changed it to Vanapa as it changes n>n in almost all instances (cf. above p. 88). In the case of natika, 48 however, it seems that it occurred in the original draft itself as a borrowing from the east; but later, whereas it was correctly replaced by the non-eastern natika in the Mysore version, it was allowed to remain in yr. (13) The cluster tm : The cluster tm > tp as in the west in the instance mahatpa (mahatman) yr. In the corresponding form mahata occurring in s, b, ru, kpb it shows the eastern assimilation om > t, which is already found as a borrowed element in the S, M versions. (14) The cluster sm: The pronominal forms tuphe yr and tupaka (for tuphaka) ru (*tusma) are also north-western in origin since we have sm > sp in the loc. sg. term. in S, which would explain the assimilated form ph.: *tusma > tuspa > tu (p) pha (cf. above th as assimilation from Gst, p. 88). 43. For the probability of the base Vpava being eastern as against non-eastern papa, see below p. 93, f.n. 54. 44. The eastern ko kh treatment seems to have exercised some influence in the north-west so that we see it also in the instance nikhetu (ni ksip) in the Bhattiprolu inscription (Phil. Ind., p. 220). 45. G has Vanapa; the proper north-western v is shown also by Bhattipralu anan ajnam (Phil. Ind. p. 227). 46. K in the north knows natika from the east. Madhu Vidya/275

Page #301
VERSIONS OF ASOKA'S MINOR ROCK EDICT 91 me minor rock edis the presence of his ence of the king (15) Initial h: As in the Mysore version, we have instances of adding h before an initial vowel in such cases as hevan (evam) s, ru, mk, yr, hesa (esa) yr. heta in s is quite understandable. As we have no parallel instances in the north-west or the west, the presence of hevar, hesa in the original draft of the minor rock edict can be attributed only to the influence of the king's speech. (16) As was noted while dealing with the Mysore version in the earlier paper, the main morphological features like the nom. sg. mas, and neut. ending -e7 and the loc. sg. ending -sits found in the minor rock edict are eastern. Their presence was explained to be due to the fact that they were known in the north-west through the versions of the major rock edicts. These features occur in the remaining versions of MRE as well. Some other peculiarities of noun, pronoun, and verb forms may be noted below. (a) Nom. acc. sg. neut. - Though more frequently we find the eastern -e, the western ending -an is found in satirekan and arokan in yr. (b) Dat. sg. -The non-eastern ending -ya is found in etaya athaya ru, kpb, yr; the eastern -ye in etaye athaye s. (c) Acc. pl. m.--The eastern influence in this regard can be seen also in the final portion of the yr version, for which parallel passages in the other versions are not available. Thus we have instances like hathiyarohani, banbhanani, etc., which show the eastern -ani ending. LUEDERS (Philologica Indica, p. 278 ff.) has shown this ending to be of old-Ardhamagadhi. These endings are found in the north-west*) and in a few cases in the west so and hence their occurrence in the original draft is explainable. (d) Among the pronouns, the nom. sg. of the first person is hakan in b, yr which is an eastern form. Its presence in the minor edict is to be attributed to the influence of the king's speech. (cf. hevam etc, above). (e) In the instr. sg., the non-eastern maya is given by yr, and the eastern mumaya by b, kpb and yr give me, which, however, is confused with the gen. sg. form. (f) In the second person, the forms tuphe yr and tupakan (= tuphaka (m)) ru are north-western as explained above, p.90 (cluster sm). 47. Cf. upasake, phale, savane, etc. 48. Cf. Jambudipasi. Cf. yutani, grahathani, etc. Cf. khamdhani, gharastani, pasamdani, pavajitani in Gas given by HULTZSCH, p. lxii; the proper western ending is-e. Cf. yute, athe, etc. Madhu Vidya/276

Page #302
92 M. A. MEHENDALE (g) In the demonstrative, nom. sg. m. is iyan, which is an eastern form along with ayam which is non-eastern. But iyam in the minor It seems rock edict may not be regarded as a borrowing from the east. to occur owing to confusion with the neut. sg. form iyam which as a bor. rowing already occurs in the west. Cf. the following expressions: iyam phule s, br, sd, iyam sagane, br, sd, kpb or iyam savane s, kpb, yr with such expressions as iyam pakame br, sd, yr and iyam athe s, ru, br, kpb. In section G, s gives ese where corresponding ru version gives esa and br, ad, and kpb give iyam. Obviously the original draft had iyari as referring to phale occurring in the preceding section. But this iyam was mistaken to be fem. and rendered as esd in ru, and mistaken to be mas. and rendered as ese in b. I would thus regard ese in b as mas.51 and esa in ru as fem. forms. (h) The non-eastern instrumental imina is given by yr (cf. above p. 85), and the non-eastern dat. imaya is given by ru. Its (i) Among the verb forms, it may be noted that the present part. middle ending mina s, b, ru, yr, or mina kpb is non-eastern.53 occurrence in the s version in palakamaminena is to be regarded as a borrowing from the original draft. Similarly samana in kpb is non-easter, while samta in s is eastern. (17) In the end we may consider the use of certain forms of a word or the use of different words" for expressing an idea which throws 51. Also note that in mk athe is added after iy [am]. It is possible that the translator of the b version had athe in view and not phale and hence he used ese as mas. In any case ese and esa cannot be regarded as neut. as HULTZSCH, p. cxxvii, takes them to be. 52. The same form occurs in G, while the eastern form as shown by D is imaye. 53. BLOCH, op. cit. SS 47 (p. 80) holds a different view. He observes "De toute facon la forme en mina est orientale". That is why he regards the form karamino occurring in S as oriental (SS 34, p. 72). But I have already shown in the paper on the Mysore version why the forms in -mina have to be regarded as non-eastern. LUEDERS (Beobachtungen uber die Sprache des buddhistischen Urkanons, pp. 157-159) rightly treats the present participles in -nt as an eastern characteristic. 54. In the following sections the case of the word idanim has not been considered. The north-western S version (and also the northern K) substitutes in the first. edict aja (adya) of the other versions by idani. Among the versions of the minor rock edict also we find the use of dani in the non-eastern versions viz. ru, mk, yr, and dani in kpb. But the evidence is rather scanty to allow us to look upon it as showing dialectal variation. For other isolated instances we may also compare udara (-la) in all versions except Mysore, where we get mahatpa, and yr, where we have mahadhana; satireka (--) in all versions, but once the Mysore group has adhika; Varadha (--) in all versions, but Vadhigame (adhigatave, adhigacheya) in mk; (for upagata in mk see above p. 86, f.n. 22; for pure in mk and vela in kpb see below P. 94 f.n. 58; for dhamayuta and Madhu Vidya/277

Page #303
VERSIONS OF ASOKA'S MINOR ROCK EDICT light on the dialectal distinctions in the days of Asoka. It was pointed out while dealing with the Mysore version that such a distinction seems to have existed in the use of vagra (i.e. varga) in the north-west and jana in the east (the use is illustrated in the two separate edicts). In the following sections, instances from the Mysore version are also cited as these words were not discussed in the previous paper. 38 (i) We have for Skt. manusya or manusa properly speaking. two forms in the Asokan inscriptions viz, munisa and manusa (or manufa with palatalisation). Of these, munisa occurring in D, J is eastern, and manusa occurring in G, K or manusa in S, M is non-eastern. But munisa occurs as a loan in M also, and this form being known in the north-west it was probably used in the original draft of the minor rock. edict. It was naturally retained in the eastern s version. but was also retained in the Mysore and yr versions. The eastern form, however, was replaced by the non-eastern form manusa in the kpb version which has been read by TURNER as menusu (but which I would read as manusa,58 (see above p. 89, f.n. 39). Now manusa actually occurs also in the western G version in XIII 5 in the form manusanam (gen. pl.). (ii) Mainly two words occur in all the versions of the major rock edicts to express the idea of time, viz. amtala and kala. But their use shows a distinction in meaning: amtala occurs where a long passage of time is intended to be conveyed, and kala occurs where the idea to be conveyed is 'at all times, night and day'. Now this distinction seems to have been peculiar to the east, and similar constructions occur in the major edicts of even the non-eastern regions because they are transla kalamta in mk see below p. 95, f.n. 59). Perhaps papa in all versions, but Vpava in s may be looked upon as showing dialectal distinction. 55. In K we have manusa also. 56. But the point of interest in this section is not the cerebral or dental n, but the vowel variations in the two forms munisa and manusa. For a preference of the vowel sequence (o), i, a in the eastern language, see LUEDERS, Beobachtungen etc., p. 38, where he cites such instances as pulisa, munisa, no mina. Pali and Pkt. on the other hand have manusa. Most of the modern Indian languages show the western form, cf. H. Bg. manus, As, manuh, P. M. manus, G. manus; the i vowel of the eastern munisa can be seen in As. munih, Bg. munis, Or. minisa, Sgl. minisa (miniha); Nep. manisa seems to have arisen from the contamination of the eastern and western forms (for the distribution of the words in the modern Indian languages I am indebted to TURNER'S Nepali Dictionary and op. cit. p. 8, f.n. 2(4) where he considers Nep. manis as contamination of manusa and purisa. 57. TURNER, op. cit., p. 8, f.n. 2 reads here manusanam which is clearly against the palaeographic evidence. The a stroke of ma, so far as I can see, is perfectly visible on the estampage, and is similar to the form of ma in matr[i] line 3, Turamayo line 8, etc., occurring in the same inscription. Bull. DCRI. xvii-2 Madhu Vidya/278

Page #304
9 M.A. MEHENDALE Tha tions of the Magadha version. To illustrate the above usage we may quote the following expressions from the Dhauli version : (a) atikamtamantalan 'in past times'. The expression occurs in the rock edict IV, V, VI, VIII, and it has its parallels in the other versions. (b) atikamtan antalan no huta-puluve savan kalam athakanme va pativedana va (Rock edict VI). "In times past neither the disposal of affairs nor the submission of reports at all times did exist before". (HULTZSCH's translation. Italics mine). This sentence illustrates at the same time the use of both amtala and kala. That this mode of distinction was not properly speaking current in the non-eastern regions can be seen from the versions of the minor rock edict. In these versions, the word kala is used in all non-eastern versions where the eastern s version alone, consistent with the above usage, replaces it by amtula; cf. the following quotation from the E section of the ru version : ya imaya kalaya Jarnbudipasi amisa deva husu...... "The gods who for (all) this time had been unmixed (with men) in Jambudvipa... (HULTZSCH's translation, ital. mine). Correspondingly the Mysore and the yr versions have imina cu kalena ;88 but the s version alone has etena ca antalena, because the idea to be conveyed is 'in past times' and not 'at all times'. (iii) In the versions of the minor rock edict we get two sets of expressions for 'to be zealous' or to have 'zeal in practising morality. These expressions are pakama- (pra Vkram-) and vpalukamu- (para 58. b version is here missing; mk substitutes the expression by using pure just as it has made many other changes in the text of the version. The kpb uses imayam velayam instead. The use of vela in kpb is a little difficult to explain. At the moment I am inclined to explain it as a possible confusion in the reading of the original draft. If this explanation proves to be correct it will show that the edict was issued from the north-west in the Kharosthi script. If the right hand loop of the ka symbol in the Kharosthi script was forgotten or illegibly put in the draft which reached the hands of the translator responsible for the kpb version then the symbol could be easily mistaken for va. Now since the Kharosthi script does not mark the length of the vowels, in the original draft the word kalaya must have stood as kalaya. This being misunderstood as valaya, the scribe seems to have corrected the stem vala to vela and added the final anusvara to the ending -ya which he thought to be loc. sg. He thus arrived at the form velayam. For the present, however, this explanation must be taken for what it is worth. P. MEILE (Misa Devehi Chez Asoka, JA 237, p. 209. 1949), on the other hand, would like to restitute velayan in the ru version on the basis of the same word occurring in the kpb version, as the symbol for ka in kalava in ru is not quite legible. The estampage published by HULTZSCH will show that this suggestion is hardly tenable. Madhu Vidya/279

Page #305
VERSIONS OF ASOKA'S MINOR ROCK EDICT Vkram-).au In fact the main purpose of the minor edict is to encourage all men-those that are highly placed and those that are placed low-to be more and more zealous in the practice of morality. To express this idea all non-eastern versions viz. ru, br, sd, kpb, yro use the verb from Vpakama as in pakamatu, or the past passive participle pakata (prakranta), or the middle pr. part. pakamamina, or the noun pakama (prakrama). But the eastern version s" consistently gives forms of palakama, palakamta, palakamamina, and palakama instead of the above forms. Among the versions of the major edicts, even S, M, G show the eastern parakrama- which, in the light of the above evidence, is to bet attributed clearly to the influence of the eastern palakama as seen in D, J, K. (iv) The word for 'year" used in the versions of the major rock edicts, issued from the east, is vasa (or rasa G, vasa S, M). The context shows that this word was used in official reckonings of regnal years like duvadasavasabhisita (IVth edict) etc.; or it was used when the period of the official tours was to be specified as in pancasu vases (IIIrd edict) or timini vasani (1st separate cdict), or when the passage of many years. was intended to be conveyed, cf. bahini vasa-satani, bahihi vasasatehi (IVth edict). Now when we look at the versions of the minor rock edict we find that in the s version, which is also eastern, the only word for 'year that appears is savachala. Here, however, the context is altogether different. The word does not occur in any of the situations referred to above, but is used while counting the years of the king's relationship with the Buddhist order. Thus it says 25 (adha) tiyani sarachalani am upasuke sumi "(Two and a half) years (and somewhat more have passed) since I am a lay-worshipper", 59. The mk version avoids the use of either of cause of their somewhat unusual meaning. yutena and kalamtam in mk. these two expressions, probably be. Instead we find the use of dhama 60. In yr both BABUA and BLOCH read pak (k) ate (twice), pak (k) amasa, and pak (k) - ame. But in lines 7-8, BARUA reads pakamaminena, which BLOCH would correct to pa (ru) kkaminena. Now the plate given by D. C. SIRCAR (IHQ 7, to face p. 738, 1931) clearly shows that BLOCH's suggestion cannot be accepted. We have, therefore, to retain BARUA's reading which is also in agreement with the regional distinction shown above. A little further in line 9, both the scholars read parak (k) amevi (Bloch -r-). The plate referred to above shows that the letter read as ra or ra is defaced and perhaps testifies to the writer's attempt to cancel whatever symbol he may have inscribed. I, therefore, suggest that in line 9 also we should read pakamevu. 61. The forms cited are from the ru version. 62. The b version is much mutilated; but as it agrees with the s version in general it also may have given forms with palakama. Madhu Vidya/280

Page #306
96 M. A. MEHENDALE and savachale sudhike am "But a year and somewhat more (has) passed) since (I have visited the Samgha and have been more zealous)." It is thus clear that in the eastern dialect vasa was used for mundane purposes, while savachala was used for ecclesiastical purposes. This distinction does not seem to be strictly observed in the non-eastern regions, so that in the context referred to above all the versions except s use vasa in the first occurrence of the word (satirekani adhatiyani vasani ya hakam pasake), but keep savachara in the second occurrence (antireke tu kho sarvachare ya maya samhghe upayite.....). 4) For negation the eastern versions use no while the non-eastern versions have na." Among the versions of the minor rock edict, all have the eastern form, as it has been already borrowed in the G, S., M versions, but only the yr version gives the non-eastern na, and the b and the mk versions have na once. The s version is supposed to give na once in the first line; but in view of the strong evidence for no being eastern, I would regard this a scribal mistake for no. Similarly the particle tu appears to be western, but cu eastern." Among the versions of the minor rock edict, all give the eastern form as a borrowing, except the versions at Mysore and Yerragudi. (vi) In the days of Asoka the word for 'to write' in the northwest was nipisa" which was borrowed from old Persian, cf. such forms as nipista, nipistam, nipesita, and nipesapita occurring only in the S version, edict Nos. IV, V, VI, XIII, and XIV. The corresponding word which occurs in all the other versions is likha. It must be noted that the M version at all places and the S version also in some places give the forms of the verb likha. Now among the versions of the minor rock. edict we get forms of Vlikha only, cf. lekhapeta ru, [likhapa] yathas, likhapetavaya ru, likkita br. The original draft also may have contained Vlikha instead of nipisa as it was known in the north-west as well. (vii) In the use of vocabulary there is one point in which the br version agrees with the west and not with the north-west. Unfortunately the word referred to in this section does not occur in the other versions of the minor rock edict. For 'the writer', the br version uses lipikara which occurs also in G (the K version has lipikala); but the S version gives dipikara. This Persian word dipikara of the orignal draft was thus replaced by the Indo-Aryan lipikara in the br version. Other versions of the major rock edict where this word occurs in G, K, S are defective. It is obvious, however, that the D and J versions must have 63. Translation of HULTZSCH; the lacunae have been filled up from the ru version. The use of the word savachala in the 3 version is, however, clear. 64. Cf. BLOCH, op. cit., p. 82. 65. See HULTZSCH, p. xlii. Madhu Vidya/281

Page #307
VERSIONS OF ASOKA'S MINOR ROCK EDICT contained lipikala, while the M version, though it shows many 'Magadhisms', in this respect must have contained the north-western dipikara. This becomes clear from the fact that the M version always agrees with the S version in giving the ancient Persian word dipi, while in all other versions we have lipi. This happens also when in the same sentence in M we find the use of eastern Vlikha, where S has nipisa, but both S and M give dipi, cf. ayi dhramadipi nipista S, and ayi dhramadipi likhita M (5th edict, section O; also ef. 6th edict, section M, 13th edict, section X, and 14th edict, section A). This shows that while the Old Persian word dipi had completely ousted lipi in the north-west in the days of Asoka, the word nipisa had not succeeded in doing so. 97 Madhu Vidya/282

Page #308
NORTH-WESTERN (AND WESTERN) INFLUENCE ON THE MYSORE EDICTS OF ASOKA By M. A. MEHENDALE In a paper published in the Journal of the Oriental Institute, Baroda, 1, 240-44 (1951-52), I expressed a view that the exceptional forms which occur in the separate edicts of Asoka at Dhauli and Jaugada and which do not occur in the remaining versions of the major rock edicts at these two places suggest that the two separate edicts were issued from a place other than the east. As the exceptional forms in the two separate edicts show affinity with the northwestern dialect of the Asokan inscriptions, I further suggested that the two separate edicts were probably first drafted in the northwestern dialect and then translated into the eastern one. Such an assumption would satisfactorily explain the presence of some northwestern features in the two separate edicts. Such an assumption would further lead to a hypothesis that not all the edicts of Asoka were first drafted in the eastern dialect, as was hitherto believed. This hypothesis is now borne out by the study of the dialect found in the versions of the minor rock edict at Brahmagiri, Siddapura, and Jatinga-Ramesvara in the Mysore State. These records resemble in contents the versions of the minor rock edict found at Rupnath, Sahasram, Bairat, Maski, Kopba! and Yerragudi but they also add to it a second edict' as it were, describing the king's instruction in morality (dhamma). The Mysore versions have another distinct feature viz. that they contain at the end of the edict the name of the scribe, and that whereas the whole edict is written from left to right in the Brahmi script, only the last word lipikarena 'by the writer' is inscribed from right to left in the Kharosthi characters. As the Kharosthi script is other 1. For a description of the places, the text (with plates), and the translation of the versions, see HULTZSCH. Inscriptions of Asoka, CII, 1. xxvi-vii, 175-180 Oxford 1925. The readings and the translations given in this paper are, unless otherwise stated, those adopted by HULTZSCH." The following abbreviations are used in this paper :D-Dhauli version major edicts J - Jaugada d-Dhauli separate edicts K- Kalsi j - Jaugada , S -- Shahbazgashi , br- Brahmagiri minor rock edict M-Mansehra , sd-Siddapur G-Girnar jtr - Jatinga--Ramesvara, 2. The second edict at Jatinga-Ramesvara appears to be longer than the one at Brahmagiri and Siddapura. The Jatinga-Ramesvara version, however, as a whole is in a much damaged condition. The enlarged version of the added second edict is better preserved in the Yerragudi edict. (ed. by B. M. BARUA, IHQ 13. 132-6, 1957, and by J. BLOCH, Les Inscriptions d'Asoka, Paris, 1950), Madhu Vidya/283

Page #309
M. A. MEHENDALE wise used only in the north-western inscriptions of Asoka at Shahblazgashi and Mansehra, a single word in this script in the south tends to show the north-western origin of these inscriptions.3 As will be shown below this supposition based on a small piece of palacographic evidence is amply supported by the linguistic facts preserved in these records. In a recent book, Beobachtungen Uber Die Sprache Des Buddhistischen Urkanons 5 by LUDERS, edited from the Nachlass by E. WALDSCHMIDT, the eidtor makes a reference to the paper referred to above on the Dhauli and Jaugada separate edicts and observes in his Zum Geleit (p. 6, f. n. 1), "M. A. Mehendale hat....einige sprachliche Eigentumlichkeiten, durch die sich die 'Separat-Edikte' von den ubrigen in Dhauli und Jaugada gefundenen Felsen-inschriften unterscheiden, als nordwestlich nachzuweisen versucht. Er vermutet, dass Asoka diese Separat-Edikte auf einer Reise im Nordwesten in der dortigen Sprache erlassen habe, und dass sie vor ihrer Einmeisselung in den Ostdialekt ubersetzt worden seien. Eine Uberprufung dieser rein linguistisch fundierten Annahme auf inhaltliche und historische Wahrscheinlichkeit scheint erwunscht." I am thankful to the editor for these observations, in respect of which I may he permitted to state the following facts taken from the inscriptions which show that Asoka had issued the minor rock edict and the separate edicts while he was away from the capital. (These facts make no reference to the place from where they were issued, nor to the language in which they were originally drafted). (1) In the version of the minor rock edict itself it is expressly stated that the proclamation was issued by the king while he was 3. On the presence of the Kharosthi letters in the Mysore versions, BUHLER (EI 3.135, 1894-95) already observes, "Finally, I have to point out that the Northern or Kharoshtri letters exactly agree with those of the Mansehra and Shahbazgarhi inscriptions .... The use of the two alphabets probably indicates that Pada (now read by HULTZSCH as Capada) was proud of, and wished to exhibit, his accomplishments .... The use of the Northern characters may further be taken to indicate that Pada once served in Northern India, where the Kharoshtri alphabet prevailed." HULTZSCH supports this view when he observes (op. cit. p. xlii), "The three Mysore edicts were drafted by one of this clase (i.e. the class of writers), who wrote at the end of his signature the instrumental case lipikarena in Kharoshthi characters, showing thereby that he had been transferred from North-Western India." As noted above, to me it seems to mean much more than this, viz. that the edicts themselves were first drafted in the north-west. 4. It will be shown in a separate paper that the other versions of the minor rock edict found at Rupnath etc. also support this view. (The paper has now been published in the Bulletin of the Deccan College, Vol. XVII, No. 2, Sept. 1955). 5. Abhandlungen der deutschen Akademie der Wissenschaften zu Berlin, Klasse fur Sprache, Literatur and Kunst, Jahrgang 1952, Nr. 10, Berlin, 1954. Madhu Vidya/284

Page #310
WESTERN INFLUENCE ON THE MYSORE EDICTS OF ASOKA 5 on tour. In the Brahmagiri version we read (line 8) - iyan ca savane savapite vyuthena And this proclamation was issued by me on tour."? 6. Similar statement is found also in the Rupnath, Sahasram, and Yerragudi versions of the minor rock edict. 7. HULTZSCH (op. cit. Corrigenda, p. 259), however, gives up the meaning 'tour' assigned to myustr and expresses his conviction that the word means 'having spent the night (in prayer)'. Accordingly he corrects his translation of thd relevant section (J) of the Sahasram edict as follows -- And this proclamation (was issued) by (me after I had) spent the night (in prayer).' In support of this view HULTZSCH refers to the Baudhayana Dharmasutra IV. 5.30 (also cf. KIELHORN JRAS 1904, 364f., FLEET JRAS 1911. 1106). It is not possible to enter here into a detailed discussion of the much-discussed word vyusta. For our present purposes it is sufficient to note that vivasa occurring in the Rupnath version is clearly connected with viy vas 'to stay away from home, to be on journey' as in prabodhayati mavijan vyusite sokakarsita 'She enlightens me, who am ignorant, and she is harrassed with grief when I am gone on journey.' (Bhag. Pur. 4.28.20) or priyan priyeva vyusitam visarna *as the dejected beloved (longs ta see) her husband who has gone on journey (Bhag. Pur. 6.11.26). Accordingly the section K of the Sahasram version should be translated as 'For two hundred and fifty-six nights I have stayed away from home' and not as HULTZSCH suggests in his Corrigenda - Two hundred and fifty-six nights (had then been) spent (in prayer)." This translation is not possible also because in the minor edict itself (cf. Rupnath, sections B-D) Asoka says "Two and a half years and somewhat more (have passed) since I am openly a sakya. But (I had) not been zealous. But a year and somewhat more (has passed) since I have visited the Sangha and have been very zealous." Thus according to this statement Asoka had been 'very zealous' in his observance of morality for more than one year when he issued the minor rock edict. This statement would contradict with the new translation proposed by HULTZSCH which purports to say that Asoka had spent only 256 nights in prayer which is much less than a year. That Asoka missed some nights of prayer even during that period of more than one ycar when he was very zealous' would be difficult to believe. It is thus clear that vivasa cannot mean spending nights in prayer'. It only shows that while issuing the edict Asoka was on tour. The purpose of mentioning 256 nights may have been just to give proof of the king's great zealousness by showing how long he had been away from home while he was on his religious tour (dhammayata). For the view vyusta =tour cf. F. W. THOMAS IA 37.22 (1908), JA (10 ; 15.51718 (1910), JRAS 1916.117; also cf. K. A. Nilkantha SASTRI, Journal of the Ganganath Jha Research Institute 1.93ff. (1943). The same meaning is assigned to vyasta by J. FILLIOZAT (JA 237. p. 148, 1949) and by J. BLOCH (Les Inscriptions d'Asoka, p. 149, and p. 150, f. n. 14). BLOCH's translation of the relevant passage runs as - Cette proclamation a ete faite (respy. proclamee) apres tournee; deux cent cinquante-six nuits ont ete passee en tournee. The translation of vyustena as 'apres tournee' is not happy, because viv vas means to be on journey' and not to return froin journey'. Probably the expression was so translated because the author believed that the edict was issued by Asoka from his capital when he returned to it during the monsoon period after having spent 256 days outside his capital in pious tour. For a new interpretation of the no, 256, see the article of FILLIOZAT referred to above, Madhu Vidya/285

Page #311
6 M. A. MEHENDALE (2) In the third major rock edict the king records that he has ordered his officers to go on tours every five years to carry out state business and to instruct the people in morality. The details about the instruction in morality to be given to the people as mentioned in this edict viz. obedience to the parents, liberality to friends and relatives, abstention from killing animals, etc., are practically the same as those mentioned in the additional portion of the edict as represented in the Mysore versions. The similarity of this instruction contained in the Mysore versions with the one mentioned above which the officers were expected to give while on tour leads one to believe that the instruction contained in the minor rock edict was also the one which the king himself gave on tour and which he asked to be recorded for the benefit of his officers. (3) In the eighth major rock edict we are told that Asoka took pleasure in undertaking tours of morality since he was anointed ten years. Among the things done on such tours of morality were instructing the people in morality and enquiring with them about morality. Once in such enquiries the king seems to have found out that there was a misunderstanding prevailing among his subjects, especially among those who were away from the capital, as regards the people who could acquire heaven. These people believed that only those that were highly placed, and not others, were able to attain heaven as a fruit of religious merit. It is true the king had stated in his tenth rock edict that "It is indeed difficult either for a lowly person or for a high one to accomplish this (i.e. religious merit) without great zeal (and without) laying aside every (other aim). This is indeed difficult to accomplish for a high (person)." The misunderstanding, therefore, seems to have arisen because in this edict the king has only emphasized how difficult it is to obtain religious merit; and a reference to the highly placed persons alone in the end seems to have left the impression that this difficult task was well-nigh impossible for the lowly persons. In order to remove this serious misunderstanding which he came to know while on religious 8. The translation of the last section (F) given above differs from the one adopted by HULTZSCH. He translates eta tu kho usatena dukaram (Girnar) as But among these (two) it is indeed (more) difficult to accomplish for a high (person). But eta can mean only this' and not among these (two)'. BLOCH (op. cit. p. 119) translates- Mais c'est difficile surtout aux grandes, The Jaugada version which is nearer to the capital has usatena cu dukalatale which means 'This is more difficult for a highly placed'. As the Kalsi version shows, usateneva at Mansehra is to be read as usatena va. BLOCH (op. cit. p. 119, line 20) reads only usaten for the M version. Madhu Vidya/286

Page #312
WESTERN INFLUENCE ON THE MYSORE EDICTS OF ASOKA 7 tour the king possibly immediately issued the minor rock edict which contained the following words -" (H) For this cannot be reached by (a person) of high rank alone, but indeed even a lowly (person) can at liberty attain the great heaven if he is zealous. (I) For the following purpose has this proclamation been issued, [that both the lowly) and those of high rank may be zealous in this manner, and (that even) my borderers may know it....": It is important to note that in the Rupnath version the king asks his officers to go on tour within their districts with the text of the minor rock edict obviously to remove the misunderstanding from the minds of the people in those districts. (cf. BLOCH's translation of the section L, p. 149. For a different interpretation of vayajanena see HULTZSCH, p. 169). (4) A word may now be said about the opening words of the separate edicts at Dhauli and Jaugada which give proof to show that they were issued by the king from outside his capital. (This point was not noted while dealing with the separate edicts in the article referred to above). 9. This shows that the minor rock edict was issued after the major rock edicts a point, as will be shown below, well supported by the linguistic evidence. This, however, goes against the view of HULTZSCH who maintains that the minor rock edicts are the first ones (see pp. xliv, liv, and also l) among the Asokan inscriptions. The arguments given by him (p. xliv), however, are inconclusive. What is stated in the minor rock edict about inscribing the edicts on stones and pillars applies to that edict alone (cf. the words iya ca athe and ina ca afhan at Rup. and Sah. respectively), and it does not apply to the general 'activity of inscribing records on rocks and pillars. The summary of the king's views about dharma is given in the minor rock edict to show to the people how easy and simple it is to acquire morality even for lowly placed persons. and to demonstrate to the officers how they should instruct the people in morality. J. FILLIOZAT (Les Deva D'Asoka, JA 237, p. 232 ff., 1949) contends that the dhammayata referred to in the rock-edict VIII by Asoka is the one that lasted for 256 nights and which is referred to in the minor rock edict. This would mean that the minor rock edict was issued before the major rock edicts. But there is nothing to show that the dharunayata of rock-edict VIII is identical with the vivasa of the minor edict; and further it would be more reasonable to interprete dharmayata (sg) as standing for a type of tours which Asoka contrasts with the viharayata of the former kings. The use of singular does not show that Asoka undertook only one such tour viz. the one referred to in the minor rock edict (FILLIOZAT, op. cit. p. 233, f.n. 1). 10. Line 4 (H) no hiyari sakye mahatpeneva papotave kamant tu khudakena pi Line 5 - pakomi... Rena vipule svage sakye aradhetave (I) etayathaya iyan savane savapite yatha khudaka Line 6 cal mahatpa ca iman pakaneyu ti anta ca me janeyu.... (Brahmagiri version; lacunae in square brackets filled up from the Siddapur version). Madhu Vidya/287

Page #313
M. A. MEHENDALE (a) The first separate edict at Dhauli opens thusLine 1 - (A) Devanar piyasa vacanena Tosaliyar mahamata nagalaviyohalaka vataviya 'At the word of Devanarpriya, the Mahamatras at Tosali, (who are) the judicial officers of the city, have to be told this.' The second separate edict at Dhauli also opens in a similar manner except that it has kumale maharyata ca for mahamata nagalaviyohalaka of the first edict. The first separate edict at Jaugada open as Line 1 - (A) Devanampiye hevam aha (B) Samapaya mahamata nagalaviyohalaka hevam vataviya Devanampriya speaks thus. The Mahamatras as Samapa, (who are) judicial officers of the city, have to be told this.' The second separate edict at Jaugada also opens in the same way as above but it adds the word lajavacanika which means 'The Mahamatras have to be told at the word of the king. The commencement of the Dhauli separate edicts which is so markedly different from that of the other major edicts gives an impression that these edicts were not issued directly by the king from the capital to his officers at Tosali. It rather suggests that the text of these edicts was being conveyed to these officers by some other officers while the king was on tour. If the king had addressed himself directly from the capital to the officers at Tosali, as has been suggested by HULTZSCH (p. 177, f.n. 5), he would have begun his letter, though perhaps less modestly, somewhat like the one found at Calcutta-Bairat in which the king addresses the Sangha 11 As the king was not in the capital while issuing these edicts, he could not send these to the officers at Tosali in the usual manner. Hence in order to give them authority it was found necessary to say expressly at the commencement of the inscriptions that the Mahamatras were being instructed at the instance of the king (Devanampiyasa vacanena).12 These words are not to be found in the first separate edict at Jaugada received by the Mahamatras at Samapa, but the second edict there contains the word lajavacanika 'at the word of the king' which serves the same purpose as the opening words in the Dhauli version. 11. Here the text commences as - Priyadasi laja Magadhe samehar abhivadetunan aha apabad hatar ca phasuvihalatan ca The Magadha king Priyadarsin, having saluted the Sangha, hopes they are well and comfortable.' 12. Similar expression is found also in the so-called Queen's edict (devananpiyasa vacanena savata mahamata vataviya) where the queen seems to register her request, but authority is sought to be given to the edict by saying that the instruction was being conveyed at the instance of the king.".. Madhu Vidya/288

Page #314
WESTERN INFLUENCE ON THE MYSORE EDICTS OF ASOKA 9 All these facts would supply internal evidence offered by the contents of the Asokan inscriptions to show that the two separate edicts and the minor rock edict were issued from outside the capital. It has been already shown that the linguistic evidence given by the separate edicts tends to point out that these edicts were issued from the north-west. In FLEET's opinion13 the Mysore edicts were issued from Suvarnagiri, the head quarters of Asoka's southern province, by Asoka himself. But the opening lines of these edicts clearly go against this view. They show that the officers of the king at Suvarnagiri received the edict from the king from outside which they were now forwarding to the subordinate officers at Isila, Hence they say14Suvarnagirite ayaputasa mahamatanam ca vacanena Isilasi mahamata arogiyan vataviya hevam ca vataviya/Devanampiye anapayati From Suvarnagiri, at the word of the prince (aryaputra) and of the Mahamatras, the Mahamatras at Isila must (better 'may') be wished good health and be told this : Devanampriya commands (as follows)'. As will be seen from the linguistic analysis below, the Mysore edicts reveal certain north-western features which enable us to draw the conclusion that they were issued neither from Suvarnagiri in the south, nor from Magadha in the east but from some place in the north-west. A casual reading of the Mysore versions will easily point out that though the edict was issued originally in the north-western dialect and script, its versions, before being inscribed, were rewritten in a more or less standard western dialect, and in the Brahmi script (witness the use of a single dental sibilant, the absence of the change i>y, and that of the consonant clusters). But though north-western (and western) in phonology, the language of the Mysore edicts shows apparent affinity with the eastern dialect in morphology (witness the nom. sg. mas. and neut. in -e instead of -o and -a (m), and the loc. sg. mas. in -si instead of -e or -mhi). An explanation of this affinity with the eastern dialect will be given below. Here we may first take up the detailed linguistic analysis of these versions to see in what respects they agree with the north-western dialect of the Asokan edicts. The north-western features in the Mysore inscriptions : (1) The vowel 1: In words of relationship, the eastern form favoured -i < -7, while the north-western form favoured 13. JRAS 1909.998 ; 1911.1108. 14. This is according to the Brahmagiri version. The Siddapur version substantially agrees with this, except that it has aha for inapayati. The JatingaRamesvara version is very much defaced. Madhu Vidya/289

Page #315
M. A. MEHENDALE u < -7.40 cf. piti (pit) and bhati (bhradr) in D, J, while pitu nd bhralu in S, M. Among the Mysore edicts jtr gives the northvestern form pitu, but br gives the eastern one piti. The original lraft therefore possibly contained piti owing to the influence of he king's speech. While this was rightly replaced by the non-eastern ritu in jtr, it was allowed to remain in br. In br line 9, HULTZSCH reads pranesu drahyitavyan and translates firmness (of compassion) must be shown towards animals. He lerives (p. cxxviii) drahyitavya from ydth and regards that in this orm the vowel ? > 19.16 According to WOOLNER (Asoka Glossary, Calcutta, 1924) dra here represents dari? as he considers the form o be gerundive of * darhyati (from the same root v dih). In any ase the form can be regarded as north-western, because in S we iave instances showing both the tendencies : Thus we have grahatha (glhastha) showing > ra, and drasana (darsana), drasayitu (darsayitva) showing transposition of 7.18 15. The same tendency is witnessed in the Niya Prakrta, cf. pitu, bhratu, madu, etc. BURROW, The Language of the Kharosthi Documents from Chinese Turkestan (Cambridge, 1937) 8 68, 72, BURROW, however, does not notice this treatment under -T. 85. The later Tor Dherai Kharosthi inscription of about 200 A.D.), however, gives -pitrinan as the language of this inscription is highly Sanskritized. See KONOW, Kharoshthi Inscriptions, (CII, Vol. 2, part 1, Calcutta, 1929), No. 92. Pali has the vowel u as can be seen from the forms of pitar and matar given by GEIGER 891 (for exceptional occurrence of the piti and mati in Pali cf. GEIGER 812, 877). In later Pkts forms with w are more common, though i forms occur in Amg. and JM (PISCHEL 8 55 $ 391). As regards piti in br. it may also be observed that the form had already come to the north through the Kalsi versions of the Asokan edicts. The D version once (IV. 4) gives pitu which may be due to assimilation ci. the form m o}t[il-pitu-sususa. 16. In the Niya Prakta, however, the regular treatment of 1 is 1>1, Ti, but not ta (BURROW $5). With this we can compare the Abokan dridha etc. in S. Both ra and ti treatments are, however, found in the later Kharosthi inscr. cf. MEHENDALE, Historical Grammar of Inscriptional Prakrits (Poona, 1948) $500. 17. Similar transposition of 1 is practically absent in the later Kharosthi documents (BURROW $ 39), but it is found in the Kharosthi Dhammapada. cf. drugati, pravata, etc. BURROW BSOS 8.428. 18. It may, however, be observed that the explanation of the form from V dth is not very satisfactory, as the use of this verb with pranesu would be unusual. Generally in such contexts we find the use of anarambha or analarbha 'abstension from killing? (Rock Edicts 3, 4, 11, Pillar Edict 7). Once we have also sayamo 'gentleness' (better 'self-restraint') (Rock Edict 9), and once pana-dakhina 'boon of life' (Pillar Edict 2). The idea to be conveyed in such expressions is that no violence should be done to the animals, and this is hardly brought out by V drh 'to be firm etc.' with the locative pranesu. BLOCH (p. 150 and f.n. 16) also derives the text word from V druh, and compares with this expression the use of samyama in the rock edict IX noted above. He translates - (Il faut) se contraindre a l'egard des etres vivants. Madhu Vidya/290

Page #316
wesTERN INFLUENCE ON THE MYSORE EDICTS OF ASOKA 11 (2) The vowel 4: In the word guru, we get the vowel u in the eastern form gulu in D, J. The vowel u also occurs in the west and the north-west due to eastern influence, cf. guru in G, S, M. But the form, proper to the west and the north-west, 19 is garu with the vowel a as is shown by the examples in G, S (and galu in K). It is, therefore, worthwhile to note that in the Mysore group we have gatu. (3) The semi-vowel y initially. It is well-known that the initial y of the relative pronouns and the indeclinables is preserved in the western and north-western inscriptions of Asoka, but it is dropped in the eastern inscriptions.20 The Mysore edicts also preserve y in yatha, yatharaham, ya and yam. The presence of an and atha in the Yerragudi version, however, suggests that these eastern forms were present in the original draft, as they were known in the north-west due their occurrence in the M version. While the alteration was carried out in the other non-eastern versions of the minor rock edict, this was not done in yr. In the Mysore version we find a peculiar form upayita (< upa+ ita) which shows the development of the semi-vowel y to break the hiatus. This may be attributed to a north-western tendency to pronounce y before the vowel i which has been noted by BURROW At the moment I am inclined to read dru in the place of dra (druhyitavyam) due to the extension of the wavy line of 1 by a perpendicular stroke below for 4 (cf. pru in bhutapruvam in G 5 and stu in srunaru G 12). If this reading is correct then the form can be derived from v drul to hurt, to seek to harm' which is used also with the locative. But this will give the intended meaning only if we suppose that a word like no 'not' was in mistake omitted in writing - pranesu (no) druhyitavyan violence should not be done to the animals. It appears that as the original draft containing pranesu druhyitavyan gave quite the contrary meaning, the writer at Yerragudi changed the expression to pranesu dayitaviye (IHQ 13. 134, line 18) compassion must be shown to the animals'. SIRCAR's view (IHQ 7.739, f. n. 2) that the plates given by HULTZSCH support the reading dayilaviyar for the Mysore version also does not seem to be correct. 19. The word is not available in the later Kharosthi records. However, Pali has garu (GEIGER $34), Maharastri and Sauraseni also have garu, while Ardhamagadhi and Jaina Maharastri give both garu and guru (PISCHEL $123), 20. MEHENDALE, 831 (2). The same tendency is witnessed in the Niya Praksta, cf. ya (yat), yava (yavat), yatha (yatha) (BURROW $ 85, 892 $ 130). The later Kharosthi inscriptions also preserve y initially (cf. KONOW, Index Verborum). According to J. BLOCH (p. 52) and J. VEKERDI (10 3.324, 1953) this absence of initial y is not a phonetic phenomenon. In the opinion of VEKERDI these forms are better explained as due to contamination with the corresponding forms of the demonstrative pronoun, since in other words like yaso, yati, ctc. no loss of initial y. is to be observed. Madhu Vidya/291

Page #317
12 M. A. MEHENDALE (SS 32) for the Niya Prakrta, cf. such forms as yiyo = iyam, yima ime. I am now inclined to give up the explanation of diyadha or diyadhiya, which occurs in M, K and the versions of the minor rock edict including those at Mysore, as coming directly from duyardha," but regard it as a simple combination of di ( ( di in the compound form di-guna in the Niya Prakrta, which also gives the ordinal biti. See BURROW SS 43, SS 89. Pali also has diguna, though dutiya (GEIGER SS 114, SS 118). 22. The Sahasram version of the minor rock edict, which is in the east, is expected to give the form divadhiyam. But it also gives diyadhiyam as a loan from the north-western dialect of the original draft. PISCHEL'S (SS 230, SS450) explanation of divadha from dvikardha is rejected by LUDERS (op. cit. p. 72, f.n. 2). He regards it as a mixed form from diyaddha (cch (GEIGER SS 57) as also in later Prakrtas except Magadhi which shows ts > sc (PISCHEL SS32). Madhu Vidya/292

Page #318
WESTERN INFLUENCE ON THE MYSORE EDICTS OF ASOKA 13 (5) In Mysore versions we find the retention of the cerebral n as in the west and north-west,25 and not its change to n as in the D, J versions of the Asokan edicts. Thus we have porana, savane, dharmaguna etc. The cerebralisation can be seen in the terminations as well. cf. lipikarena, mahamatanan26 and devanampiya.27 In vasani, however, we find the dental -n- as is also the case in the north-western terminations (HULTZSCH, p. lxxxv) which have n only in Devanapriye. It may also be noted that n in term. occurs in savena in the second Jaugada separate edict and is also possible in palalokikena in the same edict. (6) As in the standard western Prakrta, the Mysore versions have only the single dental sibilant and in this respect they agree also with the eastern dialect of Asokan inscriptions. cf. vasa (varsa), amisa (amisra), v sususa- (V susrusa-) etc. But in the Mysore versions s appears for s in three instances. This fact betrays the writer's incomplete knowledge of the distinction between the two sibilants and his consequent attempt at what may be described as hyper. north-westernisation.28 The instances where s occurs are as follows: (i) In sd sa (ca] m (satya) appears, while the correct form sacar appears in br and jtr. (ii) In jtr one s has been noted in the transcript of HULTZSCH (p. 180, line 19). The jtr version contains some portion more than the corresponding br and sd versions, for which it may be compared with the concluding portion of the Yerragudi version. HULTZSCH reads some of the letters following this palatal sibilant in jtr as sa ...e... [ca] ya ...... A comparison with Yerragudi version (IHQ 13. 134, line 22) will show that this portion in jtr is to be restored to - sa [va] [m] ea (paca) ya (na] .... If the restora 25. Among the later Kharosthi inscr., the earlier ones also have (and n and n in term.), but the later ones show > n, cf. MEHENDALE $ 510C. With regard to this KONOW observes (pp. ciii-iv), "The impression left by this state of affairs is that intervocalic ni and n had the same sound at least over the greater part of the territory, and that the sound was probably a cerebral." One may not agree with this view; it is possible that the n sound had really begun to appear, as can be seen from the Paisaci tendency to change n to n (PISCHEL 8225). So far as the Niya Prakta is concerned n and n are confused, the tendency being to change the cerebral to the dental (BURROW $ 34). In the Dhammapada, however, n appears quite often anatua < anatma etc. cf. BAILEY BSOS 11.499 ff., Glossary. In terminations, the Niya Praksta has the dentals, while the later Kharosthi inscriptions show both x and n. 26. In jtr. we have mahamatana with the dental -^-. 27. In jtr. line 2, devana is rather doubtful. In line 20 we have the 28. The north-western inscriptions of Asoka (MEHENDALE $35), the later Kharosthi inscriptions (MEHENDALE $ 514), and the Niya Praksta (BURROW 833) maintain the distinction between the three sibilants. For the treatment of the sibilants in the Kalsi version see HULTZSCH p. lxxii, Madhu Vidya/293

Page #319
14 M. A. MEHENDALE tion suggested here is correct, the jtr sava29 will stand for sarva. (iii) The third instance also occurs in the additional portion of jtr (line 18) where in the gen. sg. 30 we have sa (sya) in acariyasa. (7) Consonant Clusters: With regard to the clusters we may only note some peculiar clusters like those with y, 1, and m. (a) Clusters with you are normally assimilated in the northwest and the west, but dissolved in the east (except when such clusters are formed with the sibilants.) 32 Accordingly the Mysore versions show assimilation in saka (sakya) in sd, but the clusters ky and hy are preserved in br, cf. sakya and drahyitavya.33 Now the preservation of the clusters with y may have also been a northwestern tendency (though instances are not normally found in the Asokan inscriptions) because such clusters are occasionally preserved in the later Kharosthi inscriptions. With regard to ky itself it may be noted that both the tendencies of the Mysore group are available in the later inscriptions, cf. sakamuni and sakyamuni (MEHENDALE $ 522). In the Niya Praksta we witness again both the tendencies viz, assimilation and preservation of clusters with y (BURROW 41, SS 42). It may be added here that a few instances of the preservation of such clusters are also found in the separate edicts which, as has been suggested, seem to have been issued from the north-west. cf. mokhya in d, but niokhiya in j; perhaps alasya in j, but alasiya in d. (b) The clusters with r are assimilated as in the east. Thus t7 > (1) in ayaputa (aryaputra). But the preservation of 7clusters as in the north-wests4 is witnessed in two instances viz. prakamta (prakranta) and prana in br.35 29. Both sava and cava occur in the Kalsi version (XII 31). 30. It may be noted that the similar wrong use of the sibilant is noted in the gen. sg. tasa and tasa in the Kalsi version. The sya of the gen. sg. is modified as a fricative s(r)a or s(y)a in the later Kharosthi inscriptions (MEHENDALE $ 514). In the Niya Prakta sy > s i.e. 2 (BURROW $ 22, and TURNER JRAS 1927. 232-34. In Khotanese the group sy is preserved, cf. BURROW BSOS 8.431). I am afraid we cannot bring $ of the gen. sg. in relationship with the later Magadhi -assa, as the collective inscriptional evidence goes against it. 31. For the treatment of the cluster vy (and gy) see below. 32. Cf. MEHENDALE SS 43. 33. This portion is more or less defaced in sd and jtr. Still hy is clear in jtr, and possible in sd. 34. Cf. MEHENDALE $ 44; for later Kharosthi material cf. BORROW $ 36, $ 37, and MEHENDALE $ 523. For the more archaic nature of the Niya Prakta than that of the Asokan inscriptions in the matter of preserving clusters, see BURROW, The Dialectical Position of the Niya Prakrit, BSOS 8.422. . 35. Instead of prakamta, BUHLER reads pakamta in line 2. In line 3 even HULTZSCH reads pakarta, and pakama (prakrama) in line 4. Madhu Vidya/294

Page #320
WESTERN INFLUENCE ON THE MYSORE EDICTS OF ASOKA 15 (c) The cluster tm > tp in the Mysore versions under the influence of the standard western Praksta as can be seen from the Girnar instances. Thus we have mahatpa (mahatmanah) in br, sd ; atpa (atman) and catparo (catvarah) in G; but atva 36 and ata (atman) in M, and ata and cature (catvarah) in $. The forms ata and cature in the north-western versions of Asoka are clearly due to the eastern influence.97 That the proper local treatment for the north-west was tv is shown by the Mansehra instance given above and by the later Kharosthi documents. These documents also show that in later times tv > p under the influence of the standard western tp.38 (8) In morphology also the Mysore versions agree with the non-eastern dialect in certain respects as follows: (a) The dat. sg. forms etaya and athaya end in ya and not -ye as in the east. In this case even the north-western versions have the borrowed forms etaye, taye, athaye etc., but G has etaya, taya, athaya, etc. In the Niya Praksta the endings are both -e and -ya. But as BURROW observes the dative is rare except in the infinitives and that it had died out in the popular speech ($ 52, 355). In later inscriptions we have mostly the continuation of the borrowed -e, though -ya and -(y) a are found in two instances (MEHENDALE $ 529). (b) The nom. sg. neut. forms vataviyart and sacam have the non-eastern ending -a (m) and not the -e of the east. The eastern ending is of course seen in the other forms like phale etc. (For the explanation of eastern endings see below). (c) In pronoun forms the inst. sg. of the first personal pronoun maya is north-western. The eastern forms as given in D, J versions are mamaya (and me). In later Kharosthi we have maya 36. Instead of atva, BUHLER reads atma. But cf. HULTZSCH p. 81, f.n. 3. 37. ata also occurs in K, the separate edicts, and the Pillar edicts. 38. In the Niya Praksta we have the treatment tm > tv or p (BURROW $ 44). In the Dhammapada we have tm > tv cf. ajatva < adhyatma, etc., BAILEY, BSOS. 11.498 Glossary. In the later Kharosthi inscriptions we get four forms viz, atmuna-, atvana-, apana-, and ata- (MEHENDALE $ 528). Of these, at mana is clearly a Sanskritised form, and ata- is the continuation of the old borrowing from the east found in the S, M versions. Thus the proper north-western treatment (tv and p) is shown by the remaining two instances. That ata in S, M and the later Kharosthi versions cannot be regarded as due to assimilation of tv > t is shown by the fact that in the later Kharosthi inscriptions (MEHENDALE $524a) and in the Niya Praksta (BURROW $43) tv is either preserved or assimilated to p and not to t, except in the absolutive suffix. The non-eastern treatments leading towards assimilation to p, and the eastern treatment leading towards assimilation to t can also be seen from the later Prakstas (PISCHEL $ 277). Madhu Vidya/295

Page #321
16 M. A. MEHENDALE in the Niya Prakrta (BURROW $78), but me, owing to confusion with the gen., in the inscriptions (MEHENDALE 8 536). The instr. sg. of the demonstrative is imina in the Mysore version which agrees with the form in the Girnar version and the one in Pali (GEIGER $ 108). The eastern form given by the J version is imena. The north-western Asokan forms are not available. (For the other forms of the demonstrative see below). (d) The ending -yu is the non-eastern and -vu the eastern in the optative forms. The Mysore forms janeyu, pakameyu are thus non-eastern. In the Niya Praksta, however, the optative has always the primary endings, thus for 3rd per. pl. -eyanti (BURROW S 100); the same is the case in the later Kharosthi inscriptions (MEHENDALE $ 542). (e) In br once we have sarvacharen, where the anusvara at the end may have been due to its being confused for loc. sg. An anusvara at the end of a loc. sg. form appears in a later Kharosthi inscription, cf. divaser (MEHENDALE 8529a, p. 317). It is interesting to note that a similar anusvara at the end of a loc. pl. form is found in sahasesun in the Dhauli separate edict I, line 4. Or these anusvaras may have been due to the north-western tendency to put an anusvara where it does not properly belong and which has been noted in the Niya Praksta (BURROW $ 47).39 (f) In the Mysore version we have a form of the present participle paka [m] i..na which HULTZSCH proposes to read as pakamamina. If this is correct we may compare with this the Shah. pres. part in -mina, viz. karamina. The other Asokan forms in -mina are sampatipajamina and vipati padaymina which occur in the Dhauli separate edict. Further, in view of the absence of the middle present participles in the D, J versions, we may also attribute the Mysore form samana (from v as) to the north-western influence, though the actually recorded form in the north-western version is samta. It may be noted that samana occurs in Pali, Ardhamagadhi and Jaina Maharastri also. In the literary Prakrts, though the present participles Parasmaipada are available,' a tendency to generalise the middle forms in -mana is noted especially frequently in Ardhamagadhi. In Ardhamagadhi we have also forms with the ending -mina (cf. PISCHEL $ 561, $ 562). In Pali also GEIGER (3191) gives instances of the middle forms from the active bases. In the Niya Praksta, the instances of present participles are rare. But a tendency to use -mana freely has been noted by BURROW & 101. 39. With this we may also compare parkiti (prakti) of the Mysore versions. This reading is adopted by WOOLNER in his Asokan Glossary. HULTZSCH reads the word without the anusvara in all the Mysore versions. Madhu Vidya/296

Page #322
WESTERN INFLUENCE ON THE MYSORE EDICTS OF ASOKA In the later Kharosthi inscriptions, the instances are not available. Among the later Brahmi inscriptions the use of -mana is found in the Nasik in the west and in the Nagarjunikonda in the south (cf. MEHENDALE SS 207, SS 259). There is thus reason to regard the middle participles in mana or mina as originally non-eastern. 17 It has been suggested in f.n. 9 that on the basis of the contents of the inscriptions the minor rock edict seems to have been issued after the major rock edicts. If this suggestion is correct it will show that the affinity of the Mysore version with the east in certain cases can be explained to be due to certain elements borrowed from the east in the north-western versions of Asoka's major rock edicts, and which were thus known in the north-west when the minor rock edict was issued from there. 41 (1) The semi-vowel y is changed to v in dighavusa in the Mysore versions. This is an eastern tendency as the change y> v is seen in the optative terminations in D, J. cf. vasevu, nikhamavu, etc. As against this the non-eastern tendency is to preserve -yu." cf. vaseyu S,M,G, sruneyu S, M, etc. The presence of in v visava (visaya) in S, (but visaya in M and visaya in G) is therefore to be attributed to the eastern influence. The v in dighavusa is thus due to the influence of borrowed words like visava in the north-western versions. It is interesting to note here that among the separate edicts the Dhauli version gives optatives with the eastern -vu (asvasevu etc.) while the Jaugada version gives the same forms with the non-eastern -yu (asvaseyu etc.) (2) The cluster vy is normally dissolved in the eastern versions (vy viy), but assimilated in the north-west (vy vv), and preserved in the west. In respect of this cluster the Mysore versions show a mixture of the eastern and the non-eastern tendencies. Thus the cluster is dissolved, as in the east, in such instances as sususitaviya (susrusitavya), vataviya (vaktavya), and pavatitaviya (pravartitavya). Now though the assimilation of this cluster is noted as the normal north-western tendency above, we find a few 43 40. Pali also has dighayu (GEIGER SS 101), though in Pali instances where y> v are also to be found (GEIGER SS 46). For the preservation of y in ayu and visaya in the later south Indian copper-plate grants cf. PISCHEL SS 253. The change of y> u in a few forms is noted by PISCHEL $254 for Ardhamagadhi Jain Marasti, Apabhrarisa and secondarily for Paisici. 41. The change of y to v is not noted either in the Niya Prakrta or in the later Kharosthi inscriptions. 42. MEHENDALE SS 43 (4)... 43. The cluster gy is also dissolved in arogiya (arogya) in br, sd. Later Kharosthi has the normal assimilation in the case of this word, cf. aroga (KONOW, Nos. 27, 35, etc.) and arogi (BURROW SS9). Madhu Vidya/297

Page #323
18 * M. A. MEHENDALE , instances also of dissolution as in the east in the north-western versions of the major rock edicts. cf. viyapata or viyaprata (vyapnia), pujetaviya (pujayitavya), kataviya (kartavya), etc. Therefore the instances from the Mysore group showing vy > viy are to be attributed to the borrowed eastern elements in the north-west. The preservation of this cluster, as in the west, is found in two instances in the Mysore group, viz. vyutha (vyusta) and drahyitavya. It is possible that this was the standard western tendency in Asokan times.) (3) The borrowing of the eastern trait through north-west is more evident in the treatment of the cluster ks. In the Mysore versions ks is assimilated to the guttural kh as in the east and not to the palatal ch as in (the north-west and) the west. Thus we have in the Mysore version khudaka (ksudraka). Now in the case of this very word it is interesting to note that even the northwestern versions show the eastern form with the guttural. cf. khudaka D, J; khuda or khudaka M, khudraka S; but the Girnar version gives the proper non-eastern form chuda or chudaka. In my opinion this evidence shows that khudaka happened to be in the original draft as by the time the minor rock edict was issued, the major edicts were already inscribed in the north-west and the borrowed eastern form with kh- had gained some currency there. (4) In the Mysore version the cluster jn >n as in the north 44. As mentioned above though instances of the preservation of this cluster are not found in the north-western versions of the Asokan edicts, this may have been a local tendency. cf. divya in the Kalsi version and the instances in the Niya Praksta dadavo and dadavya, BURROW $9, 841. The continuation of the borrowed eastern tendency of dissolution is also seen in the Niya Praksta, cf. viyala (vyala), BURROW $ 42 and word index. It is again worth while to observe that the separate edicts furnish one instance of the preservation, possibly as a north-western characteristic. cf. saricalitavya in j, but samcalitaviya in d. In the later Kharosthi inscriptions, however, we have only the proper north-western tendency of assimilation, cf. MEHENDALE $ 522 (xvi): 45. C1. MEHENDALE $37 (1, iii). According to HULTZSCH (p. lxxxviii) ks remains in the S version. On p. 55, f.n. 5, however, he observes." In order to distinguish this sign (i.e., the one where it corresponds to sk. ks) from the real chh (ie. where it corresponds to skt, ch), transeribe it by ksh, but do not want to imply thereby that it was actually spoken like that." (brackets mine). Later Kharosthi inscriptions show the continuation of the eastern influence as they show kh (MEHENDALE $ 526b), but Niya Prakrta shows ck, while the Kharosthi Dh. has ch (BURROW $ 48). For ks > kk as a borrowing in the Dhammapada cf. DURROW BSOS 8.425. For ks in Gandhari cf. BAILEY, BSOS 11.770-75. Madhu Vidya/298

Page #324
WESTERN INFLUENCE ON THE MYSORE EDICTS OF ASOKA 19 west and the west, and not n as in the east. cf. Mysore natika (jnatika) with nati or natika G,S,M, but nati in D, J. But the cluster in in Vajnapa- has an interesting story. It shows the normal non-eastern treatment (palatal) noted above in G in Vanapa, and the normal eastern treatment (dental) in D, J in yanapa. But among the major edicts the S, M versions show the cerebral Vanapa-, which is nothing but the borrowed n treatment from the east with further north-western cerebralisation of the dental nasal. Now this Vanapa is found also in the Mysore version, which shows that this minor rock edict was issued from the north-west only after this word had been already familiar there due to the versions of the major rock edicts.48 (5) Cerebralisation of dentals in combination with (7) or s This is normally the eastern characteristic, whereas in the west the dentals are preserved (in the north-west cerebralisation is observed due to eastern influence). The Mysore versions show a mixture of both these tendencies, thereby again pointing to the fact that this edict was issued after the major edicts were inscribed at S, M which show sometimes the eastern influence in this regard. Thus we see the dental in pavatitaviya (pravartitavya) and pakiti (prakrti) in the Mysore group. But in kajaviya (kartavya) we have the cerebal of the east. Now it may be noted that kafaviya is found in M as a complete borrowing from the east (cf. the same form in D, J), as katava in S, but katavya in G. Similarly rth >th in the Mysore versions, as in the east, cf. atha (artha). The same word is also found in the major edicts and it is represented as atha D, J, S, athra S, athra S, M, atha G. Thus it will be seen that though the north-western versions show the proper dental treatment, they also show sometimes the cerebral, J 46. Cf. MEHENDALE SS 48 (1); for the same treatment of the cluster in later Kharosthi material cf. MEHENDALE SS 515d, BURROW SS 44. In Pali too the normal treatment in nn, (GEIGER $53), in Paisaci nn, in Ardhamagadhi both un and nn, and in the other Prakrtas nn (PISCHEL SS 276). 47. cf. similar cerebralisation of n in Vprapuna- (pra-Vap-nu) in G, S, V mana (Vman-ya) and ana (anya) in M. 48. The same exceptional treatment in the case of Vajnapa is found in the Niya anati (ajnapti), BURROW SS 44. (For confusion between n and see above f.n. 25). For anata in Khotanese cf. BAILEY, BSOS 11.779. Pali also gives V anapa, but has anna (ajna) 'perfect knowledge' also (GEIGER SS53). A later Kharosthi inscription, however, gives the proper local form ana (ajna) (KONOW Nos. 11,14). 49. In later Kharosthi inscriptions also dentals with are preserved or assimilated to the dentals; only a few instances of cerebralisation are found (MEHENDALE SS 520b, SS 516). The same tendency is found in the Niya Prakta (BURROW SS 37 which also gives katavo). Madhu Vidya/299

Page #325
20 M. A. MEHENDALE borrowed from the east. atha in the Mysore versions is therefore to be attributed to atha or athra in S.50 In the Mysore versions the cerebral is also seen in thitika (sthitika), which is to be attributed to the presence of a similar eastern form twice in the Mansehra version in the north-west. cf. thitika M, thitika D, J, but thitika S (stita G).51 Here may also be noted the treatments of Tdh and st. The cluster rdh > dh in vadhisiti, diyadhiya etc. in the Mysore group in keeping with the north-western and western borrowings from the east. Thus the forms of V vardha > Vvadha are found in S, M, G, as well as in D, J (though the dental is seen in some forms at G).52 The absence of the cerebral in vyutha (vyusta) in bris obviously due to mistake (for a similar mistaken dot in the circle see va in hemeva in line 9 br) unless we attribute it to the presence of the dentals in the west (and the north-west) as against the cerebrals in the east. But in the case of st we find th in the northern version in atha (asta) K, and athami (astami) in two pillar edicts. I am therefore more in favour of regarding th as a mistake for th in vyutha. (6) It has been said above that the Mysore versions agree with the eastern dialect in morphology in certain important respects as nom. sg. in -e and loc, sg. in -si. But even these agreements possibly show that the Mysore versions were issued after these Magadhisms were known in the north-west through the versions of the major rock edicts. 50. In the Niya Prakrta, the cluster tth is preserved or changed to th of doubtful value (BURROW $ 36, 37, 8 49), but in my opinion probably a continuation of the eastern borrowed cerebral th. In the later Kharosthi inscriptions oth is preserved (MEHENDALE 8 520b). In the Dhammapada 7th>th, cf. anatha < anartha, BAILEY, BSOS 11.499, Glossary. 51. In the Niya Praksta also the forms of V stha show st or th, or rarely th of doubtful value. The cerebal is found only in athi (asthi), cf. BURROW $ 49. The same is also true of the later Kharosthi inscriptions, cf. MEHENDALE $ 520 iii, iv, $ 516b. For the treatment in the later Prakstas cf. PISCHEL SS 307, $ 308. 52. In the Niya Prakta, rdh is mostly preserved, though dh and dh are found once each (BURROW $36). Similarly rdh is preserved in the later inscriptions, though dh is found in proper names and dh in vudha < vyddha (MEHEN. DALE $ 520b iv, $ 516d, $ 500c). 53. atha in S, M is doubtful. But the change s# > th in the north-west is shown by the later Kharosthi documents, cf. BURROW 49, and MEHENDALE 519a. The change of this th> dh is found in a Kharosthi inscription (KONOW No. 11). cf. adha (asta). For a similar change in the Asokan inscriptions cf. HULTZSCH, p. 135, f.n. 1, 2. In Khotanese, the group st is preserved, BURROW BSOS 8.431. Madhu Vidya/300

Page #326
WESTERN INFLUENCE ON THE MYSORE EDICTS OF ASOKA 21 (a) The non. sg. mas, in the Mysore version ends in -e as in the east, cf. athe, devanampiye etc. Though the proper northwestern term.is -0, we find some forms in -e already in the S version like jane, vivade etc., and in the M version the -e ending is almost the exclusive one.54 (b) The nom. sg. neut. in the Mysore version ends in -e as in the east, cf. phale etc. The proper north-western term, is -am, but the Magadha forms with -e are quite frequent in those versions, cf, dane, drasane, etc.55 (c) The loc. sg. of -a bases ends in -si as in the east, cf. Isilasi, Jambudipasi. Now the north-western termination is -spi or -e. But the eastern termination is already seen in the north-western versions in such instances as apakaranasi, uthanasi, etc.56 (d) The neuter form se of the base ta occurs in the Mysore versions as in the east. But its use in the north-west is already witnessed in the M version. The nom. pl. mas. of this base is given as se for the Mysore group by HULTZSCH (p. cxxx); here, however, the north-western form is te and eastern se. In the Niya Pitaksta also (BURROW $ 80) te occurs. In view of this evidence I am inclined to regard se in the Mysore group not as nom. pl. mas., but as the nom. sg. neut. used as the beginning of the sentence. (e) The demonstrative base idam gives once the nom. sg. mas. iyam. But this may not be regarded as an eastern mas. form. It is rather to be explained as arising out of confusion with the neuter form iyam because both the nom. sg. mas. (athe) and the 54. For the agreement of the Niya Prakta with the M version in this respect cf. BURROW BSOS 8.420-21, 424. For both -o and -e terminations in the later Kharosthi inscriptions see MEHENDALE $ 529. In the Dhammapada the endiny -e is absent, BURROW BSOS 8.428. 55. In later inscriptions we have sivat hale (KONOW No. 26). Otherwise the normal ending is -am or -a (MEHENDALE $529). For the Niya Prakta ending -a < -am, cf. BURROW $ 52, $ 53. 56. In later Kharosthi material, the term. is usually -mmi (or -mi), and - (see BURROW $ 58, MEHENDALE $529, pp. 316-17). In one case, however, we have si in hasisa written for hasasi (KONOW, No. 31). Perhaps this -st, confused with gen -sa, occurs in the Dhammapada, BURROW BSOS 8.429 57. cf. for instance se in se hevam devanampiye dha (lines 8-9) with se in se ime dhammaguna pavatitaviya (line 10) in the br version. The se in the second instance is dropped in the sd version and just likely in the jtr. version. This second se does not correspond to Skt. te as nom. pl. mas, and agree with ime dhamina guna, but it corresponds to the neuter tad used as a conjunctive. The passage is therefore to be translated as 'that these moral virtues should be practised' and not as HULTZSCH does these same moral virtues must be practised.' Similarly se hevarn devanampiye aha should be translated as 'that Devanampriya speaks thus' and not as HULTZSCH does Moreover, Devanampriya speaks thus,' Madhu Vidya/301

Page #327
22 M. A. MEHENDALE nom. sg. neut. (phale, savane) with which it occurs end in -e. This iyam as neut. occurs in the Mysore versions quite often. It is an eastern form which is found already in the north-western versions together with idam which is the proper form there.58 (7) There remain only two cases in respect of which the Mysore versions agree with the east and for which instances are not available in the north-western versions. Thus the tendency to add h at the beginning of a word with an initial vowel,59 cf. hevan (evam) in the Mysore versions as also in D, J where we get such other words as hida, hedisa, etc., for which in the western and northwestern versions we have evam, idha, edisa, etc. Similarly the nom. sg. of the first personal pronoun in the Mysore versions is hakaneo which agrees with the form in the D, J versions but which disagrees with the western and north-western ahan. In the later Niya Praksta and the Dhammapada also we have ahu (BURROW $78, BAILEY, The Khotan Dhammapada, BSOS 11. 488ff. Index). Therefore hakam is clearly the eastern form with which the later Magadhi hage may be compared. Perhaps the explanation that can be given for these two eastern words hevam and hakar in the Mysore versions and for which parallels are not to be found in the north-western versions is that they are reminiscent of the language of the Maurya king who dictated the edict. This may be especially true about hakam as it refers to king Asoka himself. In the end we may note a point with regard to the separate edicts which goes to show that these two edicts were issued from the north-west. This point was not noticed in the article dealing with the two separate edicts referred to above at the beginning of this paper. The north-western versions no doubt use the word jana as the other versions do. But in the tenth rock edict in section D, jana, which appears in G, is substituted by the word vagra in S, M (and by vaga in K). This use of vagra i.e. varga for jana witnessed in the north-west is again found only in the separate edicts. Thus in the first separate edict, sections AA, at Dhauli we read Ujenite ype is not absent, cf, KOOS 11.791-93. 58. For later Prakta forms cf. PISCHEL SS 429, S 430 and for Pali cf. GEIGER $ 108. 59. In the Niya Prakrta, BURROW $ 28 finds considerable irregularity in the treatment of h, owing to its absence in the native language. But a possible instance of the above type is noticed in hedi = eda (sheep). In later Kharosthi inscriptions such instances are absent, cf. KONOW, Word Index. For the unstability of h in the Gandhari cf. BAILEY BSOS 11.791-93. 60. HULTZSCH (p. lxxviii) follow's PISCHEL. (8 417) and refers to Panini 5.3.71 in explaining thakar from * ahakam. Madhu Vidya/302

Page #328
WESTERN INFLUENCE ON THE MYSORE EDICTS OF ASOKA 23 pi cu kumale etaye va athaye nikhamayisa .... hedisameva vagan *But from Ujjayini also the prince (governor) will send out for the same purpose .... a person (vagam) of the same description.' The corresponding Jaugada version is unfortunately defaced. But in the first separate edict, section L, in Jaugada we read - tata hoti akasma ti tena badhanartika anye ca vage bahuke vedayati 'In this case (an order) cancelling the imprisonment is (obtained) by him accidentally, while many other people (vage) (continue to suffer). Here in the corresponding section K of the Dhauli version, however, we have jane.61 This correspondence between jana and varga which pertains to vocabulary is an important piece of evidence to show that the two separate edicts were issued from the northwest. 61. HULTZSCH has already drawn attention to these substitutions (p. 40, f.n. 2), but without noticing that vagra or vaga is the north-western substitution for the eastern jana, which reappears as a borrowed word in the separate edicts. Madhu Vidya/303

Page #329
ON GATHAS 8 AND 10 OF THE TAKKARIYA JATAKA BY M. A. MEHENDALE Deccan College One of the Bharhut sculptures, now kept in the Indian Museum, Calcutta, bears the inscription kinarajatakam. The sculpture represents a well-dressed man, meant to be a king, seated in a chair, and a man and a woman, intended as a kinnara couple, standing beside him. The sculpture has been variously identified with the Candakinnarajataka (No. 485), the Bhallatiyajataka (No. 504), and the kinnara-cpisode in the Takkariyajataka (No. 481). The last one was first proposed by Hultzsch, and has been rightly accepted by Luders who deals with this subject at some length in his book on Bharhut inscriptions (pp. 134-138).2 The story of the kinnara-episode in the Takkariyajataka is, in short, as follows: A hunter once captured a kinnara couple and brought them to the king of Benares. The king had never seen such beings before. He was told by the hunter that the beings were expert in dancing and singing. The king therefore asked them to show their skill. Being afraid of committing an error, the kinnaras remained silent. The king therefore became angry and said that the beings were in no sense superhuman. They were just ordinary creatures and hence should be killed and served as food. At this, first the female and then the male kinnara uttered some gathas. On hearing them, the king was pleased and set the kinnaras free. Gatha No. 8 nttributed to the kinnari runs as follows: satan sahassam dubbhasitanam kalam pi nagghanti subhasitassa, dubbhasitam sarkamano kileso, tasma tunhi kimpurisa, na balya. Luders translates the gatha as follows: "A hundred thousand of bad speeches do not weigh as much as one piece of good speech. Fearing calamity from bad speech, the kinnaras are silent, not out of stupidity." The third pada of the gatha offers difficulty. Luders says that as it stands in the manuscripts it cannot be right; the explanation of the commentator, who seems to take kileso as verbum finitum is without value." Luders therefore emends the third pada, which then reads as dubbhasita samkamana kilesam. His translation of the gatha given above follows the emended text. In addition, Luders also observes that, "It is to be noted that kilesa apparently has not been used in the Buddhistic sense,''5 1. Complete Bibliography to be found in the book of Luders, referred to below, pp. 134185. One may also look up his List No. 701 in the Appendix to Ep. Ind. Vol. 10. 2. Published by the Government Epigraphist for India as Corpus Inscriptionum Indicarum, Vol. II, Part II (Ootacamund, 1963). 3. Luders, Barh. Inscr. p. 137. 4. dubbhasitar sankamano kilissati kilamati. 5. Luders, ibid., footnote 4. Madhu Vidya/304

Page #330
232 STUDIES IN INDIAN LINGUISTICS The difficulties in interpreting this pada can be overcome with less cffort if we assume that the Pali translator of the original canon in the castern dialect misunderstood the gatha word kilesa- as standing for Pali kilesa-, Skt. klesa.. Actually, it seems to have represented a contraction kilcse = Pali kir' eso? where eso stands for the kinnara standing beside the kinnari, and kil ese has nothing to do with klesa. The last two quarters of the gatha in the eastern dialect of the original canon probably ran as: dubbhasitam (or -dan?) sankamane kilese, tasma tunhi kirpulise na balya, which the Pali translator should havc rendered as dubbhasi'am sarkumano kir' eso, tasma tunhi kimpuriso, na balya, "This one, indeed, (i.e. the kinnara) is afraid of (uttering) bad speech; therefore the kinnara is silent, and not due to stupidity." The kinnari, by uttering this gatha, tried to achieve two purposes. She wanted to prove to the king that she was no ordinary crcature; and further, to plead in behalf of her husband and explain reasonably his silence. She probably thought that this would enable her to secure the release of both of them, and that the king would not demand further proof from her husband to show that he was also a kinnara. But the Pali translator was misled by the word kilese of the original canon and rendered the line mechanically into Pali without worrying for the meaning. Or, he understood it in some such way as "Fearing that calamity occurs due to bad speech (dubbhasitam = dubbhasita kile80 hotiti samkamino), the kinnara has remained silent." Later, some one who construed samkamano kileso, instead of saikamano kimpuriso, and interpreted the former in some such way as done by the commentator, changed kimpuriso (sg.) to kimpurisa (pl.) in order to make it applicable to both the kinnaras who had remained silent. The reading kimpuriso, presumed in the interpretation suggested above, is in fact supported by one manuscript. This interpretation involves only one emendation kimpuriso for kimpurisa; and it spares us from the necessity of having to interpret kilesa (= klesa) in a non-Buddhistic sense. When the king heard the gatha of the kinnari he ordered her release; but he said that the kinnara, who had not yet spoken, should be roasted and served at breakfast. The kinnara therefore thought it was time for him also to speak and recited three gathis, the first (No. 10) of which ran as follows: pajjunnandtha pasavo, pasunatha ayar paja, tvamnatho 'smi maharija, natho 'ham bhariyaya ca, dvinnam aninataram natva mutto gaccheyya pabbatam. The last two quarters of this gatha also are difficult. Luders translates the gatha as: "The cattle depend on the god of rain, these beings on the cattle. On you, oh great king, I depend: on me, my wife depends. One of us when released, could only go into the mountains, after having known that the other 6. "Dem Kanon der buddhistischen Schriften im Pali und im Sanskrit liegt ein Urkanon Zugrunde, der in einem ostlichen Dialekt abgefasst war.," SB4W, 1927, p. 123, cited in Luders, Beobachtungen uber die Sprache des budhistichen Utkanons, p. 8, 1054. 7. "Die Partikel, die im Sk. stets kila lautet, ist im Pali in der gesamten Literatur ikra". Luders, Beobachtungen SS 31, p. 85. 8. About --> -d in the eastern dialect. see Luders, Beobachtungen $ 04 - 98 (esp. 308), pp. 81-88. 8. ck; c' gives kimpurise. Madhu Vidya/305

Page #331
ON GATHAS 8 AND 10 OF THE TAKKARIY JATAKA 283 one is dead." In foot note 6 on p. 137, Luders observes that the last two quarters are not understandable to him. His translation partly follows the explanation of the commentator: amhakam dvinnam antare eko ekam matar hatva sayam maranato multo paccha himavantam gaccheyya jivamana pana mayam annamannam na jahama/tasma sace si imas himavantam pesetukamo pathamar mam maretva paccha pesehitis From this, Luders conjectures that "annataran is perhaps an attempt to rectify annatamam distorted from original annan matar." He also adds, "I have translated accordingly, but I am by no means sure to have found the right meaning." The doubt expressed by Luders regarding the correctness of his interpretation is justificd. His way of understanding the text not only involves a conjecture regarding the distortion of the text and an attempt to rectify it, but also the mcaning arrived at after having done all that is not quite satisfactory. It is therefore proposed to interpret the text in the following way. When the kinnars utters the gatha, the king knows her real nature and orders only her release. The kinnara then realizes that, despite knowing the true nature of the kinnari and receiving an assurance from her that her companion is also a kinnara, the king is not ordering the release of both of them. Now he must break bis silence and by his speech show to the king that he is no ordinary creature. He therefore utters a gatha, and fearing as though that this itself will not be enough to secure his release, pleads that his wife, who has secured her release, depends on him; if the king were also to free him (mutto), realizing (natva) from his speech that the remaining creature (afinataram) is also a kinnara, he will be free to accompany the kinnari to go to the mountain. This interpretation does not do any violence to the text; it supplies what is easily understandable; and the meaning arrived at seems satisfactory. The two quarters may accordingly be translated as: "(Now) having known (me), the other one of the two (also to be a kinnara), when released (by you), I may go to the inountain." 9. gaccheyya has been taken as the form of the 1st pers, sg. Madhu Vidya/306

Page #332
NOTES ON ASOKA'S ROCK EDICTS M. A. Mehendale POONA. (1) ROCK EDICT X The last line of this edict has been divided by Hultzsch into two sections as follows:1 E dukaram tu kho etam chudakena va janena usatena va anatra agena parakramena savam paricajitpa / F eta tu kho usatena dukaram : Hultzsch translates the above sections as follow. : (E) "But it is indeed difficult either for a lowly person or for a high one to accomplish this without great zeal (and without) laying aside every (other aim). (F) But among these two) it is indeed (more) difficult to accomplish for a high (person)." The versions at Shah., Man., Kal. and Yerra. are in substantial agreement with the Girnar version. As understood by Hultzsch, section E does not indicate any difference between the lowly person and the highly placed one regarding the effort they have to make for the attainment of merit. Both are expected to make great effort and leave aside every other aim for its attainment. When this has been once said, there really does not seem to be any point in merely repeating in the following section what has been already said for the highly placed one in the previous section. Hultzsch tries to remedy this situation by taking recourse to the comparative expression dukalatale which occurs in the Jaug. version and translates '(more) difficult to accomplish' although in the Gir, text we have only dukaram without any comparative suffix. A distinction has thus been obtained between the lowly person and the high one. For the former, the accomplishment of merit is difficult, for the latter, it is more difficult.3 Madhu Vidya/307

Page #333
104 RTAM J. Bloch tries to get over the above difficulty and bring out the difference between the low and the high persons by translating the particle kho as 'above all, chiefly'. His translation of the last section runs as "Mais c'est difficile surtout aux grandes" (Les Inscriptipn d'Asoka p. 119). But it is possible to seek the distinction between the low and the high persons without taking any liberty with the text. We should divide the line differently into two sections as follows: E dukaram tu kho etam chudakena va janena usatena va / F anatra agena parakramena savam paricajitpa eta tu kho usatena dukaram / This will mean that accomplishment of merit is difficult for both the classes of people. But the highly placed ones will be required to put in extra effort and abandon every other aim in case they wish to attain merit. Abandoning every other aim seems to go better only with the highly placed persons. The word order in the Dh. and the Jaug. versions is somewhat different. There we have to divide the text as follows: E [dukale ........ anata agena palakamena savam] ca palitijitu khudakena va usatena va / F usatena cu dukalatale / It is now obvious that the division of the text into two sections as done by Hultzsch for the Girnar and the other versions was influenced by the text in the Dh. and the Jaug. versions. According to these versions, both the classes of men are required to make extra effort and abandon every other aim in pursuit of merit. But section 'F' makes a distinction between the two classes by pointing out that this aim is more difficult (dukalatala) for the high persons. It is worth noting that although the Yerra. version in general agrees with the Dh. and the Jaug. versions, in this instance it disagrees with them and agrees with the Girnar and the other versions in placing khudakena va vagena usafenat va before, and not after, amnata agena palakamena sayam palitijitu. This suggests that the Girnar and the other versions faithfully represent the word order of the original text as it was issued from the Madhu Vidya/308

Page #334
NOTES ON ASOKA'S ROCK EDICTS capital. The change in the text was made for some reason in the Dh. and the Jaug. versions and then in order to make the text reasonable dukale was changed to dukalatale. (2) ROCK-EDICT XII Section 'D' of the Girnar version has been read and translated by Hultzsch as follows: tasa tu idam malam ya vaci-guti kimti atpa-pasamda-paja va parapasamda-garaha va no bhave aprakaranamhi lahuka va asa tamhi tamhi prakarane | 105 "But its root is this, viz. guarding (one's) speech, (i.e.) that neither praising one's own sect nor blaming other sects should take place on improper occasions, or (that) it should be moderate in every case."' This gives an impression that Asoka permitted on a moderate scale both the praise of one's own sect and blaming of the other sects. But this is very unlikely. The pious emperor might have permitted only a moderate praise of one's own sect on proper occasions. But since he was anxious for the promotion of all the sects, in no case could he have thought of recommending even moderate criticism of other sects. This becomes quite clear from the next section 'E' which is read and translated by Hultzsch as follows: pujetaya tu eva para-pasamda tena tana prakaranena | "But other sects ought to be duly honoured in every case". Hence it shoud be clearly understood that lahuka va asa applies only to arpa-pasamda-puja and not to para-pasamda-garaha. The last part of the section 'D' therefore should be translated as, "or (that the praise of one's own sect) may be moderate on every occasion." (3) ROCK-EDICT XIV The last section 'E' of the Girnar version of this edict has been read and translated by Hultzsch as follows: tatra ekada asamatam likhitam asa desam va sachaya karanam va [a] locetpa lipikardparddhena va Madhu Vidya/309

Page #335
106 RTAM "In some instances (some) of this may have been written incompletely, either on account of the locality, or because (my) motive was not liked, or by the fault of the writer." As regards alocetpa, Hultzsch derived it from loceti=Skt. rocayatis and took a- as the negative prefix. He, therefore, translated karanam va alocelpa as 'or because (my) motive was not liked'. Bloch also does the same when he translates the words as 'ou faute de considerer le fond (or for want of appreciating the basis)' (p. 134). But this meaning is extremely unlikely. It is difficult to imagine that the king's motive was not liked by them. Hultzsch and those who thought like him, were obviously misled by the reading alocepta in Gir., alocayitu in Kal., and aloceti in Shah., all having an initial a-. This a- was looked upon as the negative prefix. But the Dhauli version has locayitu without initial a- and this could have made it clear that the verb underlying the form alocetpa etc. was not vruc but loc, often used with the prefix a-, 'to consider'. This initial short a- in Girnar is, therefore a, mistake. It should have been a-. As regards the other two versions, it is well known that the vowel length is not marked in the Shah. version, and at Kal. a- is often written as a. Woolner in his Glossary under locayitu had already suggested to read alocayitu 'having regard to'. This has now been confirmed by the Yerragudi version which gives the reading alocayitu. Following 'Woolner, the correct rendering of the words karanam va alocetpa could, therefore, only be 'or having considered (some) reason (for the omission)'. The words desam va sachaya" in this edict have also caused difficulty. Hultzsch took desam to mean 'locality'. As regards, sachaya he equated it with either samksayalo or samkhyaya 'on account of'. Thus, according to him, Asoka's edicts were at times incompletely written on account of the locality concerned. In footnote 6 on p. 26 he mentions as an example of this kind of omission the fact that the Rock-edicts XI-XIII were omitted at Dhauli and Jaugada and the two separate edicts were substituted for them. D.C. Sircar who reads the text of the Yerra. version as desam va samkhayaya (obviously a mistake for samkhaya) translates either as the (particular) place (of a record) was considered (unsuitable for them)' (El 32.15). Apparently he too takes desa to refer to the localities and derives samkhaya from sam khya-. J. Bloch, on the other hand, translates desam va sachaya as 'soit par omission d'un detail'. He apparently follows Senart who derived Kalsi sarkheye from sam ksi- and translated the words as 'in suppressing a Madhu Vidya/310

Page #336
NOTES ON ASOKA'S ROCK EDICTS 107 passage'.11 Bloch, however, notes in footnote 6 that it is tempting to interpret desa as referring to localities (as done by Hultzsch) where the inscriptions are engraved because in fact there do exist local variations in the versions of the edicts. But he notes that in that case we shall have to derive sarkhaya from sam khya-. But the Girnar version gives sachaya as the corresponding form and khy does not normally give ch in MIA. One is, therefore, led to see the verb v ksi in deriving these forms and then it would be impossible to take desa to mean locality. Asoka says that at times his edicts may not have been written completely. This could have happened either because the omission had crept in inadvertently or because it was intentional. In the former case the omission can be attributed to the mistake of the writer (lipikaraparadha). In the latter case, the officer-in-charge must have thought of some good ground for making an omission (karana va alocetpa). But a third factor, not related to the contents of the edicts, also could have been responsible for certain omissions. It was that the place where the edict was to be engraved was not enough for this purpose. It could not contain the entire inscription and hence some omissions were called for. The word desa, therefore, should be taken to refer to the place like the surface of the rock where the inscription was to be engraved and sachaya is to be derived from sam khya.. We, therefore, translate desam va sachaya as 'either having taken into account the smallness of the place (where the edict was to be engraved).' Buhler had long ago correctly explained the word desa. His translation of the Kal. version runs as 'sei es mit Ruecksicht auf den Ort (wo die Inshrift steht)'13 which he further explains as 'weil auf dem steine nicht fuer alles Raum war.' As noted above, Bloch who considered the possibility of interpreting desa as locality rejected it because in that case the form sachaya had to be derived from samkhya, but the ch of the Girnar form could not be explained from the cluster khy. But this difficulty can be overcome by explaining ch in sachaya as an instance of 'hyper-westernism'. It is true that khy does not give ch. But the cluster ks gives ch in the western, and kh in the eastern dialect, cf. chamitave, chuda, and vracha at Girnar, but khamitave, khudaka, and lukha at Dhauli and Jaugada. The translator at Girnar was, therefore, familiar with the fact that a western ch corresponded to the eastern kh in certain words. It is, therefore, quite likely that he in mistake did the same kind of substitution in the eastern form sankhaya, 14 although it was not etymologically justified, and wrote it as sam chaya. Madhu Vidya/311

Page #337
RTAM In the wording of the three possible causes for omissions, the particle va occurs between two words in the case of the first two alternatives in all the versions, as, for example, in Girnar desam va sachaya and karanam va alocetpa. In the case of the third alternative, however, in the Gir., Kal.. and the Yerra. versions only a word precedes va but nothing follows it:14 lipikar paradhena va. This dissimilarity is corrected by the writer of the Shah. version by dissolving the compound and writing one word before and one word after va: lipikarasa va aparadhena. 108 References 1 The text is given following the Girnar version. 2 i.e. absence of demrit. 3 D. C. Sircar (El 32.20-21) also renders dukale of the Yerra. version as "(more) difficult'. His division of the text into two sections is similar to that of Hultzsch. 4 The Yerra. version has in the last section usateneva (instead of usatena) which is not found in any other version. 5 i.e. of the promotion of the essentials of all the sects (sara-vadhi sava-pasamdanam). 6 Bloch's (p. 122) and Sircar's (El 32.26) translations are similar to that of Hultzsch. 7 For tena. 8 See his footnote 7 on p. 26. He looked upon the presence of 1 (in alocetpa) in the western dialect, instead of the expected r, as an instance of pure Magadhism (see his footnote 3 on p. 8). 9 For sachaya of Gir., the other versions have samkheye (K), samkhaya (Sh.), samkhayaya (Yr.) 10 What exactly is intended by this form is not clear to me. 11 As given in Woolner's Glossary p. 138 under samthaye. In the IA 10.272, Senart's translation reads as 'perhaps that a passage has been mutilated.' 12 ZDMG 40. 141-142. 13 His translation of the Shah. version is 'be it on account of the space' El 2.472. It may be noted, however, that Buhler took alocayitu with both disa (for Gir. desa) and kalanam and construed samkheye (for Gir. sachaya) as an adjective of kalanam. As already noted by Hultzsch, this is not natural. Buhler's translation of the whole passage runs as, "But it may be that something has been written here incompletely, be it on account of the space, be it on account of some reason to be specially determined, or through mistake of the writer." (El 2.472), 14 This is how the form must have occurred in the original draft. 15 A closing ti follows va in the Yerra. and also perhaps in the Dh. version. But this ti has nothing to do with the words connected with the three alternatives. Y Madhu Vidya/312

Page #338
NOTES ON ASOKA'S SEVENTH AND NINTH ROCK EDICTS By M. A. MEHENDALE 1. Rock-edict VII The Dhauli version of this edict reads as follows: (A) Devunampiye Piyadasi laja savata ichati sava pasamda vasevu til (B) save hi te sayamam bhavasudhi ca ichanti/ (C) munisa ca ucavucachamda ucavuca laga / (D) te savam ng ekadesam va kachanti/ (E) vipule pi cu dane asa nathi sayame bhavasudhi cazi nice badham/ Hultzsch translates the edict as follows: (A) King Devanampriya Priyadarsin desires (that) all sects may reside everywhere. (B) For all these desire self-control and purity of mind. (C) And men possess various desires (and) various passions, (D) They will fulfil either the whole or (only) a portion of their duties). (E) And even one who (practises) great liberality, (but) does not possess self-control and purity of mind, is very mean. In the above rendering, Hultzsch sees opposition between vipule dane on the one hand and sayame and bhabasudhi on the other. He is therefore required to add the word 'but' in the section E to bring out the contrast. But this interpretation does not satisfactorily bring out the significance of the last two words nice baaham 'very mean.' It is therefore suggested that nathi (=nasti) should be construed with all the three, dane, sayame and bhavasudhi. The * Deccan College, Poona. 1. All versions, except those at Dhauli and Jaugada, add here two more virtues katarata and dadha-bhatita (Girnar version) rendered as 'gratitude and 'firm devotion' by Hultzsch. Madhu Vidya/313

Page #339
582 M. A. MEHENDALE passage may then be translated as: 'And if one does not practise even great liberality, and (also) does not possess self-control and purity of mind, (he) is very mean.' Asoka has said at the beginning of the edict that he desires followers of all sects to practise self-control (sayama) and purity of mind (bhavasudhi). But since this is not possible for all people, it is likely that some practise only a few of the recommended virtues but not the others. It seems that according to Asoka, liberality is a virtue which can be easily practised by all. If one practises only a few virtues and not all, it is bad enough. But if one does not practise even liberality, and also does not have self-control and purity of mind, it is absolutely bad (nice badham). Practising great liberality is thus a beginning of attaining other virtues like self-control and purity of mind. It is only when we understand the edict in this way that the significance of nice badham is clearly brought out. 2. Rock-edict IX (note 1) The ninth rock-edict of Asoka ends as follows in the Girnar version : (K) idam kaca idam sadha idam sadha (read sadhu) iti, imina saka svagam aradhetu iti / (L) ki ca imina katavyataram yatha svagaradhi | This has been translated by Hultzsch as follows: (K) .. This ought to be done; this is meritorious By this (practice) it is possible to attain heaven." (L) And what is more desirable than this, viz, the attainment of heaven. In what precedes the above sections in this edict, Asoka recommends to his people the practice of morality (dhamma-mangala) in place of the other ordinary practices which they were accustomed to follow. He says that many of the latter practices are vulgar and useless. They bear little fruit. But the practice of morality bears much fruit. Even heaven can be obtained be obtained by the practice of morality. Since what is being emphasized all along in the edict is the practice of morality which can ultimately lead to heaven, it is better to translate the last line ki ca imina kataryataramh yatha sun Madhu Vidya/314

Page #340
NOTES ON AsOKA'S SEVENTH AND NINTH ROCK EDICTS 583 garadhi as "And what else than this (viz. the practice of morality) deserves more to be done so that atha) it leads to the attainment of heaven. The pronominal forms idar and imina in section 'K' refer to the practice of morality. It is therefore better to take imina in section 'L' also to refer to the practice of morality than to the attainment of heaven as done by Hultzsch. Moreover, katavyataram can hardly mean 'more desirable." 3. Rock-edict IX (note 2) Towards the end of the ninth rock-edict, from section I onwards, the Kalsi, the Shahbazgarhi, the Mansehra, and the Yerragudi ver. sions differ from the other versions of the edict. It appears that the text in the former four versions is more in keeping with what precedes in this cdict. Here. in this edict, Asoka wants to point out to his subjects the difference between the ordinary ceremonies that are performed on such occasions as marriage, child-birth etc., and the religious practices (dhamma-mangala) which the king recommends to them. The Kalsi version reads as follows: (I) e hi itale magale samsayikye se! (J) siya va tam atham nivateya siya puna no / (K) hidalokike ceva sel (L) syam puna dhammamagale akalikye / (M) harce pi tam atham no niteti (read nivateli) hida athan, palata unartar puna pavasati (read pusavati) / (N) harnce puni tar athari nivateti hida, lato ubhayesain ladhe hoti, hida ca se athe palata ca anamtai puna pasi vali tena dhammamagalena / Hultzsch translates the above sections as follows: (1) For other ceremonies are of doubtful (effect). (1) One may attain his object (by them), but he may not (do so). 2. Only the Dhauli and the Jaugada versions agree with the Gimnar in closing the edict in this way. The other edicts have a different ending for which sec the next note. Both the Dhauli and the Jaugada versions are damaged. 3. "This athain after hida seems to be redundant. In the following section (N) atham is not repeated after hida. Madhu Vidya/315

Page #341
584 M. A. MEHENDALE (K) And they (bear fruit) in this world only. (L) But that practice of morality is not restricted to time. (M) Even if one does not attain attain (by it) his object in this (world), then endless merit is produced in the other (world). (N) But if one attains (by it) his object in this (world), the gain of both (results) arises from it; (viz.) the (desired) object (is attained) in this (world), and endless merit is produced in the other (world) by that practice of morality. In section 'L' in the above text we get the word akalikya (aka likam at Shah., akalike at Man., and akalike at Yerra). Hultzsch has rendered it as 'not restricted to time.' Although this reading and its interpretation give good sense, it seems that in the context in which the word occurs it has probably been mis-spelt. What was intended to be inscribed was not akalika, but alokika, 'not restricted to (this) world. Asoka says that the ordinary mathgala which was practised by the people may or may not bear fruit. And even if it does, the fruit belongs only to this world. But in the case of dhammamamgala, the fruit is not in doubt. It is certain. Either it produces result which has relevance in both the worlds or it produces endless merit valid for the next world. Thus, while the fruit of the ordinary mamgala is restricted to this world alone (hidalokika), the fruit of the dhammamamgala is not restricted to (this or that) world (alokika). The contrast with hidalokika is better brought out by alokika, than by akalika. If the fruit of the religious practice was restricted only to the next world, Asoka would have used the word para-or para-lokika to contrast with hidalokika.* But since the fruit of the religious practice is not restricted to the next world-it may yield fruit even in this world, as made clear in section 'N'-Asoka has chosen to call it alokika. In the context when the location, and not the time, of the result of the action done in this world is emphasized, the desired sense is better obtained by the reading alokika instead of akalika, which involves in writing the transposition of the letters 'k' and 'P. In this very edict we find another instance of the transposition of letters. In place of pasavati in section 'M', we get pavasati. It 4. As he, in fact, does in other edicts. Madhu Vidya/316

Page #342
NOTES ON ASOKA'S SEVENTH AND NINTH ROCK EDICTS is likely that the text issued from the king's capital, on which the four different versions were based, contained these two cases of misspellings, viz. pavasati and akalika. But while the former was corrected in all versions except the the one at Kalsi, the latter was retained in all the four versions, possibly because the word was understood to mean 'not restricted, to time.' It is true that in the case of akalika for alokika, besides the transposition of the letters 'k' and 'l', we also need one more horizontal mark to the left for the vowel 'o'. But such minor errors regarding vowel marks are not uncommon in the Asokan inscriptions. 585 In the end one minor correction to the translation of Hultzsch may be suggested. In the last section (N), Hultzsch takes ubhayesam to refer to the two kinds of results, viz. the attainment of one's object (atham) in this world and endless merit (anamtam puna) in the next. Perhaps it would be better to take ubhayesarh as referring to both the worlds and translate: "But in case one obtains the object in this world, the gain of both (the worlds) arises from it....." Madhu Vidya/317

Page #343
On the Name and Gatha 12 of the Takkarijataka by M. A. Mehendale This Jataka (No. 481) is so called because in the narrative the Purohita of the king has a pupil who is supposed to have the name Takkariya. We find this name clearly mentioned in the prose portion of the narrative. On pp. 245-46 (Vol. IV of Faus boll's edition) we read tada Bodhisatto Takkariyo numa minavo hut va tassa santike sippam uggauhati. "At that time Bodhisatta, having become a young Brahmin by name Takkariya, was learning sciences with him (i. e. with the Purohita)". The author of the prose narrative has apparently given this name to the pupil on the basis of Gatha 1 which runs as aham eva dubbhusitam bhasi bulo bheko aranne ahim avhayano Tukkariye sobbham imam patimi na kir' eva sidhu ativelabhani Luders translates the Gatha as follows: "Ich selbst habe aus Dummheit uble Rede geredet wie der Frosch, der im Walde die Schlange herbeiruft. Takkariya, ich sturze in dies Erdloch. Furwahr, nicht gut ist unzeitiges Reden " In the above translation the form takkariye has been interpreted as vocative singular which is irregular for a stem ending in -a. The commentator has sought to explain it away as a voc. sg. of Takkariya which he looks upon as the pupil's name in the feminine gender (tassa Takkariya ti itthilingam nama). Luders is I. Also cf. atthi deva tass' eva antevasi Takkuriya niunavo nama (IV. 247). 2. That the Gathas alone in the Jatakas are canonical and that the prose is later has now been well established by Luders. Cf. his Bharhut und die Buddhistische Literatur, p. 139; also cf. Alsdorf, The Akhyana Theory Reconsidered, JOI, Baroda, Vol. 13 (1963-64), p. 199. 3. Bharhut etc. p. 96; also cf. Bharhut Inscriptions, CII, Vol. 11, Part II, P. 139. Madhu Vidya/318

Page #344
126 SEMINAR IN PRAKRIT STUDIES justified when he considers this explanation impossible. He is also right when he rejects the view of Hertel,5 according to whom the feminine form suggests that the person addressed to in the Gatha was originally some female being, perhaps the wife of the Purohita. According to Luders the correct explanation has been given by Geiger in his Pali Gr. p. 81. Geiger looks upon this vocative form ending in -e, instead of in -a, as an instance of "Magadhism". This explanation suits well with the theory of Luders who maintains that the Buddhist canon was originally composed in the eastern dialect and was translated later into Pali. - As regards the name Takkariya itself Luders considers it striking' (auffallig). He tries to explain it as related to a place name. He points out that certain inscriptions of the mediaeval times often niention a place Tarkari or Tarkarika, also written as Takkarika. This was a centre reputed for Vedic scholarship. Many Brahmanical families went out from this place to the east and the south: Luders thinks that it would be permissible for us to suppose that Tarkari was a settlement of Brahmins not only in the mediacval period but also in ancient times, centuries before the name began to appear in the inscriptions, and that the people who belonged to that place called themselves with pride Tarkarikas or Takkariyas. This is how the pupil got what appears to be a striking' name. As regards thc vocative form Takkariye, the explanation given by Geiger and accepted by Luders is undoubtedly better than the other two, viz., those given by thc Pali commentator and Ilertcl. As regards the name Takkariya itself, Luders' cxplanation is ingenious and shows his perfect acquaintance with the details of India's past. The explanation should be acceptable, if nothing better than that can be found. But it scems that it is possible to offer another explanation, both of the stem and the form, which comes out of the narrative itself and, therefore, casier to accept. In the first instance we have to note that there is no clear indication in the Gathas themselves to show that Takkariya is the name of the pupil. The Gathas are silent about the name of the 4. Luders has discussed this whole Jataka in his book op Bharhut, referred to above, pp. 94 ff; also cf. CII, Vol. II, Part II, pp. 137 ff. 3. ZDMG 60, 785. Madhu Vidya/319

Page #345
NAME AND GATHA 12 OF THE TAKKARIJATAKA 127 Purohita, and it is very likely that they do not really contain the name of the pupil either. The author of the prose narrative did not understand the Gatha word takkariye, mistook it to be a vocative singular of takkariya and hence interpreted it as the pupil's name. Apparently he did not bother about the difficulty, noticed above, in accepting the form as a vocative singular of an -a stem. But this difficulty, together with the strikingness of the name Takkariya, will disappear if we explain the word in a different way. The Pali form tak kariya can be equated with Sanskrit *tatkarya which may be taken to mean 'to be made or prepared for him ' (tasmai karyam). As the story of this Jataka goes, the wife of the Purohita was in love with another Brahmin who was very much like her husband in appearance. The Purohita did not succeed in persuading his wife to leave her paramour. He, thereforc, carefully devised a plan to kill him by burying him in a pit after he was killed and offered in a sacrifice to please the great spirits. The Purohita was so sure about the success of his plan that he could not keep the secret to himself and, out of jealousy, narrated it to his wife. She, on her part, forewarned her lover of the impending danger, and the latter, to save his life, left the city in good time. The next day when the paramour could not be found, the king ordered that the Purohita himself, who, as men. tioned above, looked like the paramour, be sacrificed and buried in the pit under the city gate. The pupil was then appointed by the king to take the place of the Purohita. When both of them reached the place of sacrifice, the teacher recited Gatha 1, cited above, in which he lamented over his haste in divulging the plan to his wife. His untimely speech had brought about the unfortunate situation in which he himself was to fall in the pit intended for his wife's paramour. He, therefore, rightly described the pit as takkariya < tat-karya which may be interpreted as 'to be made for him'( tasmai karya-). The last two lines of the Gatha have, therefore, to be translated as "I am going to fall in the pit which was to be made for him. Indeed, the speaker of an untimely speech is not doing a right thing". In the above interpretation takkariye has been interpreted as locative singular of takkariya. The form is regular but since the noun it qualifies, sobbhum, is in the accusative, we should have expected takkariyam in place of takkariye. But, in favour of the Madhu Vidya/320

Page #346
128 SEMINAR IN PRAKRIT STUDIES locative, it has to be noted that the use of locative is more common with the verb V pat. One Burmese manuscript actually gives the reading sobbhamhi using the locative. The whole line in this manuscript runs as--fakkariye sobbhamhi ahan pahami. Pahami is probably to be explained as an unreduplicated form of ha .to abandon'. This line makes us suspect that in the original canon composed in the eastern dialect, the third line of the Gatha probably read as-takkaliye sobbhasio mari pahami 'I shall abandon i. e. throw myself in the pit to be made for him'. The Pali rendering of this line seems to have given rise to two kinds of readings (1) takkariye sobbhanhi, ham pahami, and (2) takkarije sobbhamhi man patami. It is the latter which is the basis of the reading given by the Ceylonese manuscripts, viz., takkariye sobblami man patami, interpreted as sobbliam inam patanii. Onc may speculate as one will regarding the way in which the present accusative form appeared in place of the original locative. There seems to be little doubt about the interpretation of takkariye as locative singular to be connected with the following word sobbha. The author of the prose narrative, however, misunderstood it as vocative singular and hence intepreted Takkariya to be the name of the pupil. The title given to this Jataka is thus secondary, based on the wrong interpretation of takkariye. Originally the Jataka may have been named by the first line of the above Gatha. A form *tatkarya in the sense tasmai karyam? perhaps may not be available for the standard dialect of the cducated in those days. 6. For the use of two different locative endings side by side cf. Pischel, Gr. Pkt. Spr. $366 a, p. 251 f. According to Luders, the locative singular term. for the eastern dialect of the original canon was -e. But onc has also to assume, on tho basis of Asokan evidence, the term. -si for the eastern dialect. E. Waldschmidt, in his introductory remarks (p. 10) to Luders' Beobachtungen Uber die Sprache des buddhistischen Urkanons, observes, "Hinzufugen mochte ich, dass Luders' Auseinandersetzungen uber Missdeutungen bei der Ubersetzung von Nominativen bzw. Lokativen aus der Sprache des Urkanons voraussetzen, dass der Lok. Sg. im Urkanon nicht oder nicht durchgehend (ital. minc) den Ausgang auf -si hatte, den die ostlichen Asoka -Inscriften zeigen, sondern den auf -e"This -si is probably to be read as-ssi. 7. takkariya could have also been explained as sarkarila 'caused to be mado for him (tasmai karitam). But there is a difficulty in explaining iya Madhu Vidya/321

Page #347
NAME AND GATHA 12 OF THE TAKKARUATAKA 129 Normally, such a form, if it occurred in the dialect of the educated, would be in the sense tena kuryam or tasya karyam. But it should not be too difficult to assume the existence of such a form in the popular dialects, especially if such an assumption gave a better explanation of the Pali form. Gatha No. 12 in this Jatakao reads as : sabbo loko paraciito acitto sabbo loko cittavasamhi citto paccekacitta puthu sabbasanta kass' idha cittassa vase na vatte W. H. D. Rousc's translation' runs as follows: "Some one there is who cach man foolish finds;1o Each by imagination diffc. rent still; All different, many men and many minds; No universal law is one man's will." It is hardly possible to agree with this translation and Luders scenis to pass over it in silencc. Luders' translation, which is definitely better, reads as follows: "Jederman ist ohne Verstandnis fur den, der die Sache) anders versteht; jederman hat Verstandnis fur den, der sich dem (cigenen) Verstandnis fugt. Alle fron -itu. It is true that the cnuing-ika is often found as -iya, and -iya for -jka has been assumed by Liiders as the normal treatment for the eastern dialect (cf. Beobachtungen ctc. $ $ 89-90 and $ $ 133-138 ). But it is very much to be doubted whether the same treatment, cven as sporadic, can be assumed for -ita. As for the use of the dative with the past participlc, Wackernagel (II. 1 $ 184, p. 199) has noted the use of trad-anvegana prassia started for your search' from the Dajak. 38.1. He also notes the rule of Papini 2.1.36 which sanctions the use of thic dative with the participle rak sita. Ar examples, the commentators give gorakita, asrarak sita which have to be understood as preserved for a cow, preserved for a horse'. This sutra of Papini also includes another participle, hita, but its verbal meaning is not implied. It is taken to mean beneficial, wholesome', thus gohita wholesome for a cow 8. I have discussed Gathas 8 and 10 of this Jataka in Studies in Indian Linguistics presented to Prof. M. B. Emencau, pp. 231-233. 9. The Jataka, ed. by E. B. Cowell, Vol. IV. p. 160. 10. If the reading is paracitte, Rouse would translate : "Everybody is foolish in some other man's opinion". In line 2, Rouse feels, there may be a pun on citto (various): "all the world becomes different through the power of thought". Madhu Vidya/322

Page #348
130 SEMINAR IN PRAKRIT STUDIES Wesen verstehen (die Dinge ) besonders, ein jeder fur sich. Wessen Verstandnis soll ich mich unter diesen Umstanden anschliessen ?" This may be translated into English as : "Everyone is without understanding to him who understands (the things ) in a different way; everyone is with understanding to him, (everyone) who accommodates himself to his (i. e. latter's ) understanding. 11 All beings understand (the things ) separately, everyone for himself. Whose mind, in these circumstances, shall I follow ?" Luders notes in footnote 2 that this Gatha and the commentary on it have been much corrupt. He is certain that in the first line we should read paracittel? corresponding to cittavasamhi of the second line. As regards the last line, Luders suggests that we should read either kass' idha cillassa vasena ratte or kass' idha cittassa rase nu vatte. But it appears that it should be possible to understand the Gatha even as it is and translate it as: "Everyone who follows the mind of some onc clse, (and not his own) is (10 bc considered as) having no mind; (but) everyone who is under the control of his own ) mind, is (to be considered as) having mind. All beings are different, cach one having a mind for himself. (Hence ) nor e, here, should be under the control of anybody else's mind". The Kinnara couple had not sung or danced immediately when the king had ordered them to do so. They spoke only when they found that otherwise they would micct with death. In this Gatha and in the one which precedes it, the Kinnara is justifying their silence. He says that he and his wife were silent for a whilc because men have different dispositions and it is not casy to avoid criticism (G. 11). If thcy had said somcthing, it was quite likely that the king might have thought it to be wrong. In that casc he would have judged them as beings without any intelligence. But since they were really beings possessed of intelligence they wished to follow their own inclination and not act according to the will of some one else. 11. I am not sure whether this translation of the second line comes out of the German. But this I feel is what Luders meant, 12. This reading is found in one manuscript. Madhu Vidya/323

Page #349
THE DICTIONARY OF INSCRIPTIONAL PRAKRITS --- A DESIDERATUM By M. A. Mehendale It is often felt by those working in the MIA field that it is necessary to have a dictionary of inscriptional Prakrits based on both the Brahmi and the Kharosthi inscriptions dating from about 3rd century B.C. to about 4th century A.D.. In order to make the work complete, it may be good to include in this dictionary the vocables found in the early Ceylonese cave inscriptions from about the 2nd century B.C. to the 2nd CenturyA.D. It may be advisable to exclude from the proposed dictionary the material offered by the Asokan inscriptions. And this, for two reasons. In the first instance, good indices to Asokan inscriptions are available, and secondly, these inscriptions are of a nature different from the rest of the Prakrit inscriptions, for they give different versions of the same major or minor rock and pillar edicts. It is true that even the available Asokan indices need a revision in the light of the recent publication of the complete set of the 14 major edicts at Erragudi in the Ep. Indica Vol. 32 and a few other smaller publications of versions of major and minor edicts. Before a dictionary of Inscriptional Prakrits is undertaken, it is necessary as a first step, to reedit many of the Prakrit inscriptions which have not been practically touched almost since the end of the nineteenth century. The only notable exceptions are the group of inscriptions at Sanchi which have been edited by N. G. Majumdar (The Monuments of Sanchi, Vol. I) and those at Bharhut which have been edited by H. Luders (CII Vol. II, part 2, 1968). The work of editing the inscriptions will have to be done carefully. The available texts show that the scribes were not quite particular in showing the anusvara and the vowel length. It would be desirable to restore these signs for the purposes of the dictionary, where such restoration is absolutely certain, indicating, of course, into brackets the forms actually available in the inscriptions. The other peculiarity of these inscriptions is that they show only a single consonant where & geminated stop is intended. It would also be necessary to indicate where such geminated stop is to be read, e.g. nilg) godha in place of niyodha (Skt. nyagrodha) or bha(y)ya for bhaya (Skt.bharya). This practice of indicating geminated stops has been followed by J. Bloch in his edition of the Asokan inscriptions. A large number of Prakrit inscriptions are donative in nature. They are, therefore, short and have a standard form. They mention that a particular object --pillar, or a rail-bar or some such thing is a gift of some Madhu Vidya/324

Page #350
166 one--an individual or a community --coming from such and such a place and following such and such a profession. Thus we have among Bharahut inscriptions s one which reads as Vedisa Capadevaya. Revatimitabhariyaya pathamathabhu danan (Luders List 712) "The first pillar, the gift of Capadeva, the wife of Revatimitta from Vedisa". But such details are not always given. Hence one may find at Bharhut a simple inscription like Vcdisa Phagudevasa danan (Luders 780) "The gift of Phaggudeva from Vedisa", or simpler still Isidatasa danam (Luders 830) "the gift of Isidatta". The word for 'gift' lunam naturally occurs very frequently in these inscriptions and it does not seem necessary to give references to all its occurrences in the dictionary. The few examples given above will also show that these donative inscriptions conto in a large number of personal names, of places and of professions. If the proposed dictionary gives as many identifications of giographical names as possible and attempts a classification of geographical names, e.g. on the basis of their endings like-kata (Karaha. kata, Bhoja-kata), gama (Nava-gama, Sanika-gama), -ghara (Udubara ghara, Kura-ghara) etc., and a similar classification of personal names, it would be of great geographical and sociological interest. Let me illustrate this point by citing the following few lines which indicate what one can say regarding the personal names occurring in the Bharhut inscriptions: "A large number of these names is religious (theophorie). Apparently we are in a period when the worship of old Vedic deities still existed and when the rule of some Grihyasutras recommending to name a person after some nakshatra was in vogue. But the cult of minor deties and spirits like Yakshas, Bhutas and Nagas and of saints seems to have been very popular. Besides, names derived from the Vaishnavite and Saivite deities prove also the existence of these sects in that period. Often the person is called 'protected' (gu(t)tagupta, ra(k)khita=rakshita, palita), or 'given' (da(t)ta=datta) by some deity or star ; or the person is said to have some deity as 'friend' (mi(t)ta=mitra) or "god" (deva), or is said to be the deity's 'servant' (dasa). In the case of such names as may be called Buddhist, however, words as samgha, dha(m)ma dharma, (bu(d)dha = buddha, bodhi and thupa = stupa appear in place of the deity's name. It is surprising that such Buddhist names are relatively few, and that there is no marked difference in naming laymen and clergymen. The non-religious names referring to the appearance of body, mental dispositions, plants or animals are comparatively seldom met with."2 I now give some examples, taken mostly from Bharhut inscriptions, to show how we get from inscriptional Prakrits (A) some new words, or (B) new forms of words already known, and (C) some forms of grammatical interest. 1. A list of the early Brahmi inscriptions has already been published by Luders as an appendix to EI, Vol. 10. It is necessary now to bring this list up to date. 2. Bharhut Inscriptions, pp. 9-4. Madhu Vidya/325

Page #351
167 A New Vocables. 1. a(s)savarika a horseman, cavalier', Sk. afvavara-ka L 728 (A 23). It occurs also at Sanchi (L 381) in the form a(s)savaraka. akranti 'conception', Sk. upakranti L 801 (B 19). The sculpture on which this inscription (bhagavato ukrahti) appears shows the Buddho in the form of a six-tusked elephant coming down to enter the womb of Maya. In works like the Mahavastu and the Lalitavistara the term used for conception is avakranti 'descent' and the verb used is avakram- Pali has okkama) or ukkama as in matu kucchi okkami, Majjihimanikaya III. 20. The Pkt. form to correspond Skt. avakranti would be okkanti. Hence Hultzsch considers the form ukramti of the inscription to be a mistake for okramti. But Luders considers this not quite necessary. He thinks it possible to explain akrashti from Skt. upakranti 'approaching, coming near and compares this with the expression udaram upagatah occurring in the Mahdu. II. 8. 18 and Lalitavi. 55.8. 3. kubhd 'cave' L 954-956 occurs in the Nagarjuni Hill Cave inscriptions of Devanam piya Dasaratha. The word occurs also in the Barabar hill cave inscriptions attributed to Asoka. 4. kotisanithata 'a layer of crores', Skt. kotisamstrta L 731 (B 32). The word occurs in the inscription which is attached to a sculpture depicting Anathapindika's presentation of Jetavana to the Samgha after having bought it for a layer of crores'. The price is referred to in such terms because the coins were spread on the ground. In Pali (Cullavagga), we get santhara (Sk. sashstara) in place of sashthata of the inscription. The Pali ppp. of cantharati is santhata and the Amg. form is aanhthada (Pischel SS 219, p. 157). 5. Na (t) ti announcement', Skt. jnpti.. L 697 (B 64). The word occurs in what is read as a compound sigalanati announcement to the jackals"," The sculpture shows a woman seated on a tree. Three jackals are shown sitting under another tree and the woman. seems to be addressing them. In the foreground a man is shown lying who may be dead since the scene is laid in a susana (smusana), or he may be sleeping, since he is lying on his left side with one arm below his head. The sculpture has not yet been difficult to decide the meaning of lati. should have been written as sigale nati properly identified and hence it is Hultzsch thought that the inscription Skt. srgaalan jnatri 'who has observed 3. In such references, L. refers to Luders' List of Brahmi inscriptions in EI, Vol. X and the number in brackets refers to the numbering adopted in his book on the Bharhut inscriptions. 4. The inscription in full runs as Asada vadhu susane sigalanati. Madhu Vidya/326

Page #352
168 the jackals: Barua-Sinha relate nati to Skt. jnati, Pali nati and translate it as "jackals ... her kinsmen'. Luders prefers to equate nati with Skt. jnapti referring to some announcement made by the woman to the jackals. Perhaps, vinti here stands for vinnatti - Skt. vijnapti "request, appeal'. 6. tana-cakama-pari (repo) The plastering of the fana-walk' L 903 (A 127). The inscription is not attached to any sculpture. The word cakama cumkuma offers no difficulty. It refers to a place, levelled or raised with bricks etc., where monks sat down or walked in meditation. The word lanu presents difficulty. Barua-Sinha read it as vana. Luders' remark on this inscription in his List is 'Yo sense luas been made out and his treatment of the inscription in his manuscript of the book of Bharut inscriptions has been lost. Perhaps, fana stands for thiina which is to be derived from sthuna referring to a cankama used for 'standing' in meditation. If the end part of the inscription is really parirepo it may refer to the plastering of the camkumu, the cost of which was borne by some one and this is recorded in the inscription. 7. tikotika 'having three points', Skt. trikotika, L 765 (B 78). occurs as an adj. of a camkama. The word 8. turam 'music', Skt. turyani, L 743 (B 27). The word does not occur in Puli, but has been noted by Hemachandra for various Prakrits (Cf. 2.68 and Pischel & 284, p. 195). 9. dadani (k) kamo 'strong exertion', L 696 (B 77). This also occurs as the name of a camkama. According to Luders it very probably stands for Pali dadhanikkama referring to the strong exertion displayed by a monk when he was tempted by Mara, 10. dhenachako L 781 (B 76). The inscription is not found with any sculpture. The meaning of this word has not been properly ascertained. Barua-Sinha identify the word with Pali dhonasikhos which, in the Jataka 353.4, appears as a name of a banyan tree. Luders does not agree. He tentatively suggests that dhenachako may have been intended for dhenuchako = Skt. (henulsakah 'a cow-well'i.e. a well which gives milk like a cow. 11. pa(m)ca-nekayika 'a monk who knows the five nikdyas' L 887 (A 57). It occurs as a title of a monk. It also occurs at Sanchi (L 299) in the form pacarekayika. 5. crmnkumam aruyha katipayavare aparaparar cunkami J.V. 182, carilama oritua pannatte arane nisidi Suttan. 1.212. 6. Luders thinks that the Pali word was probably ponasakho - Skt. pravanasakha* with sloping branches ! Madhu Vidya/327

Page #353
169 12. pu(k)kharini 'a pond' L 907. Not in Pali, but occurs in various forins in Pkts, Pischel $ 125, p. 100. $ 802, p. 206. 13. petaki 'a monk who knows the pitakas', L 856 (A 56). 14. bou)dhago(t)thi 'Bauddha Coinmittee' L 234, 351. These inscriptions from Sanchi record the gifts of such committees from the place called Dha(m)mava(!)dhana. 15. mugapha(k)ka '(a jalaka relating to) dumb and paralysed' L 807 (B 59). The word occurs in the title of a Jataka as it appears in the inscription. Pali has the words mugapakkha and mugapakkhika. But phakka appears in the Mahavyutpatti (271, 121) among the words indicating bodily defects (like anha, jatyandha etc.). In the Gathi 33 of the Pali Jataka (No. 588) also pakkha indicates some bodily defect, cf. nihain asandhita pakkho 'I am not pakkha bccause I have no joints ......'. In the opinion of Luders phakka is the correct form which was changed to pakkha in Pali under the intluence of pakkhahata. 16. rajalipikara royal scribe' L 271 occurs in a Sanchi inscription. Pali has lipi, not lipikura, and the only reference given is to Milindapanha 79. As is well known lipikara occurs also in the Asokan inscriptions. 17. rupakamma 'carving' I 845 occurs in a Sanchi inscription. 18. rupakiraka 'sculptor' L 857 (A 55) occurs in a Bharhut inscription. 19. lupada(k)kha 'copyist' I. 921 occurs in the Ramagarh cave inscription. Pali ripadakkha mcans an 'artist'. 20. salt)tupadana 'who has abandoned attachment' L 792 (A 58). This occurs as a monk's title. According to Luders the meaning is not quite certain. Hultzsch thought it to be equal to Skt. Aastropadana 'who is versed in sciences'. Barua-Sinha suggested instead smTtyupasthana 'who is adept in the practice and experiences of mindfulness.' Luders thinks it to be an imperfect spelling for salt)tupadana 'who has abandoned attachment'. He compares safta from srsta with Pali matta (also mattha) from mrsta 'wiped, clean' and Pali sa-upadana 'full of attachment' and an-upadana 'unattached'. 21. Sudikasammadain 'accompanied by (*) a mimic dance' L 743 (B 27). Ho -rnle rightly connected sadika with saftaka which is mentioned as one of the Uparupakas. According to Luders it is possible to have sataka from sattaka and, just as we have nctika by the side of nataka, it is possible to imagine sctika by the side of sataka. The salika of the inscription stands for sutika. The sculpture on which the inscription occurs shows some women playing on musical instruments and four apsaras dancing. This would show MadhuVidya/328

Page #354
170 that originally sattaka was a mimic dance performed by women and from this developed later a reol drama. The word sammadam which occurs as the latter part of the compound is more difficult. It has been suggested to transalate it as 'gladdening, gay, joyous'. Luders offers for consideration the suggestion that originally it meant causing joy together with something else' and later it became a techni. cal term meaning ' accompanied by'. I would rather connect samma la with sarmata and translate the whole inscription sudikasarnmadarii turam devanar as the music of the gods as approved for the sadika', i.e. which is suited for a sidika. 22. so(t)tika weaver', Sk. sautrika L 321. It occurs in a Sanchi inscription. Pali has sutrakira 'spinner'. B.: Vocables having a different form or meuning : 23. ivesani "artisan' L 346. It occurs in a Sanchi inscription. Pali has avesana 'workshop', Skt. dvesana is given by Hemachandra as a word for silpisala. 24. kammanta (in silakarmamta) '(stone) work,' i.e. carving of sculptures in stone' L 687 (A1). With this may be compared selalca(m)ma 'stone work occurring in a Sanchi inscription (L 850). Pali kammanta (working, professjon, house work) does not have this meaning. 25. gharini "house-wife' L 516. It occurs in a Sanchi inscription, Pali has gharani. Mahar. seems to have gharini (Pischel & 885, p. 268). 26. cirati. It occurs as a name of a nun in Sanchi inscriptions (L 888, 624). Pali has kirata. But forms with palatalisation are noted from other Prakrits, Mahar. cilan, cilai, Amg. cilaya, cilaiya. Saura., however, has kirada (Pischel $ 280, p.164). 27. ni(g)gocha 'banyan tree', Skt. nyagrodha L 755 (B 70). Pali has nigrodha. 28. brainana 'Brahmin' L 810 (B 51). Pali has brahmana, Amg. and JM bambhana (also mahana), Pischel $ 250, p. 174. According to Luders bramana is a faulty spelling for bramhana. But it is also possible to read it as hrammana. 29. bhaya i.e. bhayya 'wife', Skt. bharya L 882 (A4). In Bharhut we also get the form bhuriya L 712 (A 34) and L 854 (A 115). Pali has bhariya. The Prakrits give different forms like bhariya, bharia, bhayya and bhajja (Pischel SS 134, p.105, $ 284, p.195). 30. Chanaka 'reciter' L 804 (A 54). Pali has bhanaka. Madhu Vidya/329

Page #355
171 81. bhi(k)khuni and bhi(c)chuni 'nun 'L 764, L 728 (A52, A 24), Pali has bhikkhuni, Amg. bhikku. Buddhist Skt. bhiksuni, Skat, Bhiksuki. According to Luders the forms with -kh- would be eastern and those with -ch- would be western. 32. latuvi quail' L 825 (B 44). Pali has latukika. 33. vald)daki 'carpenter' L 495. Pali has vaddhaki. It occurs in a Sanchi inscription. 34. vo(k)kata 'descended', Sk. vyavakranta L 777(B 18). cokkamti. Pali has 35. su(t)tatikini'nun well-versed in the Sutrantas'. It occurs in three Sanchi inscriptions (L 319, 352, 625). Pali has the mas, form suttantika (which occurs also in Sanchi inscriptions L 625, 797). 26. se(c)cha student'I 704 (B 45). It occurs in the name of a Jataka se(c)chajataka' which has been identified with Dubhiyamakkatajataka (No. 174). Barua-Sinha attempt to derive seccha from Sk. sinc- Luders does not agree and suggests to connect seccha with Pali sekkha (or sekha) 'a monk who has not attained arhatship'. Sanskrit Kosas give saiksa in the meaning prathamakalpika' a beginner, one who has just begun his studies '. According to Luders sekkha is the estern, while seccha is the western form. In Sanchi we have the work selj)jha in the same meaning (L 570). C: A Few verbal Forms. 37. avayesi played (on lute)' L 810 (B 51). This is Aor. 3rd per. sg. of vad. The inscriptional form obviously stands for avayesi. In Pali we get avadesi in yam brahmano avadesi. 38. ke(t)ta having bought' L 731 (B 82). Block derives it from * krayitrii. Skt. has kritvii and Pali kinitva. Perhaps it should be possible to derive ketta also from kritva. We may compare ve(s)sabhu < Visvabhu (B 14) L714, and Anadhape(d)dika < Anathapindika < (B 32) L 731. 39. gahuta mad' L 694 (B 50). It occurs in the title of a Jataka Sujato gahuto, identified with Sujatajataka (No. 352). Cunningham wanted to derive it from go-huta meaning bull-inviter'. But Luders agrees with Hultzsch who equates it with Sk. glhita' caught. seized'. For glhita, Pali has gahita and the Pkts. have gahida or gihida, gahiya, gahia (Pischel $ 56+, P.884). Madhu Vidya/330

Page #356
PALI-ITS HISTORY AND ITS RELATION TO THE ORIGINAL' CANON By M. A. MEHENDALE DALI, the language of the sacred writings of the Buddhists I of Ceylon, represents the earliest recorded stage of the Middle Indo-Aryan. The literature contained in this language is very vast and preserves for us numerous grammatical forms and lexical items to enable us to obtain a clear idea about the language. Although this language belongs to the MIA stage it shows certain phonological and morphological features which cannot be explained on the basis of the classical Sanskrit but have to be traced back to Vedic Sanskrit. Among these features may be counted the occurrence of -!- and -?h- in place of -d-and-dh- exactly as it happens in the Rgveda. Similarly, the instrumental plural termination -chi as in devehi can be explained only as derived from the trisyllabic Vedic form devebhih. Also the Pali absolutive forms in -tvana (pitvana) can be explained only on the basis of the similar ending in the Vedic and not the classical --tva. Such features compel us to conclude that some of the linguistic features of Pali had started to develop already very early when the Vedic Sanskrit--especially the one on which the language of the Rgveda was based--was still a spoken language. Is it possible for us to take further back the history of Pali? Is it possible to demonstrate that some of its features had started to develop even before the period of Vedis Sanskrit, i.e. in the Indo-Iranian period ? Now, among the features which mark off Pali from an eastern dialect which later developed into Magadhi we note the following two: (i) it has only the dental sibilant s, while Magadhi has s; and (ii) it shows final -as developing into -0, while Magadhi has -e (thus we have the opposition devo : deve etc.). It is significant to note that in both these points Pali agrees with Avesta, and hence it seems possible for us to say that the forerunner of Pali shared some of its isoglosses with those of Avesta. If this is true, it would be unnecessary for us to assume the merger on the Indian soil of the two Sanskrit sibilants s and s (we leave out Sanskrit s since for some time it must have been only an allo Madhu Vidya/331

Page #357
260 JOURNAL OF THE UNIVERSITY OF BOMBAY phone of s) into a single sibilant in Pali. In other words, occurrences of Pali s which correspond to Skt. s never passed through the stage of s. Together with Avesta, Pali saw the merger, probably on the Iranian soil, of IE k and s into s. Pali, like Avesta, shows the development of final -as >-0. But in this respect there is a slight difference which we would do well to remember. It is also important to bring to notice one point in this regard which perhaps has not been done before. In Sansktit itself final as > .o in certain conditions. But it also shows in sandhi only the final -a, or -as is either preserved or it changes to -as and -as. Now if we imagine that Pali developed from the spoken form of Sanskrit and if we also imagine that the sandhi rules of grammarians were based on the spoken forms of Sanskrit-and we have to imagine both these things--then it is impossible for us to believe that Pali devo ayati developed from deva ayati, or Pali devo carati developed from devas carati. We must imagine for the proper explanation of the nom.sg. forms of Pali that already in very early times in some spoken dialect of Sanskrit the devo form was generalized so that we had in this dialect expressions like devo carati devo tarati etc. It is from such forms that we can explain Pali nom.sg. -0 ending. In this respect Avesta slightly differs from Pali. Although it is also marked by the isogloss which used the ending -o irrespective of what followed in a continuous utterance, it has preserved as relics the sandhi forms like issavas ca, isavas cit, n@mas te. One more phonetic feature also deserves attention. In Sanskrit no consonant cluster of two stops can occur at the beginning of a word. But a cluster of this type is known to occur in the Girnar version of Asokan inscriptions. True, this cluster does not occur in Pali, but this point is worth mentioning because in many other respects Pali and the Girnar dialect go together. The cluster in question is db which corresponds to Skt. dv. Thus Sk. dvadasa is represented twice in Girnar version as dbadasa, but Pali has barasa. In Avesta we find a similar feature which points to a common isogloss. In Gatha Avesta dv occurs as it is or as dab and in younger Avestan it is represented by tb or b. Pali differs from Magadhi in respect of the sounds r and i. Whereas Pali has both of them, Magadhi has only l. In this case, Sanskrit and probably early Pali, are similar to Avesta for the early Vedic period. All apparently had only r. In later Vedic, and probably just about that time in Pali, 1 forms begin to appear and Madhu Vidya/332

Page #358
PALI-ITS HISTORY AND ITS RELATION then become more frequent. In the early stages of Sanskrit and Pali, both these languages shared with Avesta the isogloss of merging IE r and 1 into r. But later they fall apart and begin to show I forms due to dialect mixture. 261 Language of the Original Buddhist Canon: The Ceylonese tradition asserts that Pali is Magadhi and it represents the language in which Buddha taught his disciples. If this tradition is to be believed the Pali canon would represent the original Buddhist canon. But this Ceylonese claim has been seriously doubted. As was noticed above Pali does not agree with later Magadhi, but it very closely resembles the Girnar dialect i.e. the western dialect of Asokan inscriptions. It is therefore difficult to accept the view that Pali preserves the preachings of Buddha in its original form. Pali is not a very homogeneous dialect. It shows many dialectal features belonging to the eastern dialect. These usually go under the name Magadhisms. The question arises how do we explain these Magadhisms in Pali. Luders, en compering the Pali and the Sanskrit versions of the Buddhist Canon, came to the conclusion that there must have been an original canon in the eastern language of which Pali and Sanskrit writings are tran slations. This original canon however, has been completely lost to us. According to Luders, this language of the original canon agreed in many respects with the Magadhi of the Asokan inscriptions. Although Luders had given expression to this opinion as carly as 1927, he had not come out fully with his complete description of the language of the original canon for quite a long period. It appears that he wanted to publish his findings in his introduction to the edition of the Udanavarga based on the fragments discovered in Turfan. Unfortunately the edition of the Udanavarga was destroyed in the Second World War. But, fortunately, some part of the manuscript related to the introduction of the edition escaped destruction. This has been edited and published in 1954 by Prof. Waldschmidt under the title Beobachtungen uber die Sprache des buddhistischen Urkanons. This publication helps us to get at least some idea of what Luders thought of the eastern language in which he had assumed the original canon was composed. The Magadhisms in Pali occur, according to Luders, because the Pali translator occasionally retained some eastern forms in his translation or because he misunderstood the text of the original canon. Madhu Vidya/333

Page #359
262 JOURNAL OF THE UNIVERSITY OF BOMBAY Some of the linguistic features attributed by Luders to the language of the original canon are as follows: (1) voicing of unvoiced intervocal single and geminated stops, e.g.-k- > -8- t- > -d-, -kkh- > .ggh., -tth-> s -ddh(2) Weakening of intervocal voiced stops to y., c.g.-j- : -Y., -d- > .y.. (8) .P > .V(4) use of 1 (and no r) (5) Abl. sg. of -a stems in -am, (cf. Pali : akatam dukkatam seyyom Dh. 814 instead of akatam dukkata seyyo). (6) Acc. pl. of -a stems in -am, e.g. kanhom dhammam Dh. 87 for kanhe dhamme. (7) Loc. pl. of -a stems in -hi, e.g. virupakkhehi me mettam cullav. V. 6 for virupakkesu. (8) nom. sg. of -& stems in e. But a question is here justified. If the assumed original canon in the eastern language has been completely lost to us, how does Luders formulate his views on the nature of the language of this canon ? The method adopted by Luders can be described in the following way. (i) If Pali, as a western dialect, shows a particular phonetic or morphological feature as its general characteristic, and also shows a few exceptions to the general pattern, then Luders assumes that these exceptions occur due to these being borrowings from the original eastern canon. In that case the phonological or the morphological feature which occurs asexception in Pali is to be looked upon as a regular characteristic of the eastern language. To give an example, if in Pali -k- in the intervocal position happens to correspond generally to Sanskrit -k, but in a few cases Pali -gcorresponds to Skt. -k, then according to Luders, these words with -g- occur in Pali on account of their being borrowings from the castern language. In that case -g- for Skt. -k- or voicing of intervocal sounds to sonants is to be considered as a regular feature of the eastern dialect. Similarly if Pali gives acc. pl. forms of -a Madhu Vidya/334

Page #360
PALI-ITS HISTORY AND ITS RELATION stems generally as jane, and then if in a few cases forms ending in -am have to be interpreted as acc. pl. then Luders considers these as forms retained from the original canon by the Pali translator. In that case, again, am has to be considered as the normal ending of acc. pl. of a stems in the eastern language. If the method is simply stated as above, it is not likely to carry conviction. It must therefore be stated that Luders has tried to justify his conclusions whenever possible by internal and external evidence. This may be described as follows: 268 As an example of internal evidence we may point out that Luders interprets kanham dhammam (vippahaya) of the Dhamma. 87 as acc. pl. because in Pali literature elsewhere these dhammas are spoken of in the pl. (sabbe pi akusala dhamma kanha). This interpretation is further supported by external evidence since in the Udanavarga version in Sanskrit (16.24) we have the pl. forms krsnan dharman (viprahaya). Occasionally supporting evidence is drawn also from the variant readings in the manuscripts. For example, in the Suttanipata 510, we have panhe pucchitum where panhe is the regular Pali acc. pl. form. But in this case the Simhalese manuscripts which usually give the older readings have panham which shows that the original canon had panham as soc. pl. which was changed in the other Pali manuscripts to panhe. It was stated above that Pali has both r and l. Now if Pali has wrong I and nom.sg. in-e, these are attributed to be due to eastern language. In this case the external evidence is given by Asokan inscriptions which clearly show the eastern dialect to be exclusively I and e dialect. Madhu Vidya/335

Page #361
Reprinted from Indian Linguistics, Volume XVII, 1957. IV) Linguistics TRACE OF AN OLD PALATAL *Zh > ; IN SANSKRIT Vu M. A. MEHENDALE, Poona It is well known to the students of the historical phonology of Sanskrit that IE palatal gh survives in Sanskrit as h (< zh). It seems, however, probable to demonstrate its survival also as j ( 2 with loss of aspiration) in an OIA form ujjayati occurring in the Satapatha Brahmana in the description of the Vajapeya sacrifice. One of the peculiarities of this sacrifice is the drawing of the seventeen Suragrahas along with an equal number of the Somagrahas. About the symbolism of this act we read in the Brahmana (5.1.2.10-13): atha saptadasa Somagrahan glhnati/ saptadasa Suragrahan Prajapater va ete andhasi yat Somas ca Sura ca tatah satyan erir jyotih Somo 'nTtam papma tamah Suraite evaitad ubhe andhasi ujjayati sarvam va esa idam ujjayati yo Vajapeyena yajate Prajapatim hy ujjayati sarvam u hy evedam Prajapatih (10). sa yat saptadasa/ Somagrahan grhnati saptadaso vai Prajapatih Prajapatir yajnah sa yavan eva yajno yavanty asya matra tavataivasyai tat satyam sriyan jyotir ujjayati (11). atha yat saptadasa/ Suragrahan glhnati saptadaso vai Prajapatih Prajapatir yajnah sa yavan eva yajno yavaty asya matra tavataivasyai tad anstan papmanam tama ujjayati (12). ta ubhaye catustrimsad grahah sampadyantel trayastrimsad vai devah Prajapatis catustrimsas tat Prajapatim ujjayati (13). EGGELING translates (SBE 41.8-9) the above as follows: "He (the Adhvaryu) then draws seventeen (other) cups of Soma, and (the Neshtri) seventeen cups of Sura (spirituous liquor), for to Pragapati belong these two (saps of) plants, to wit the Soma and the Sura;-and of these two the Soma is truth, prosperity, light; and the Suri untruth, misery, darkness: both these (saps of) plants he thereby wins; for he who offers the Vagapeya wins everything here, since he wins Pragapati, and Pragapati indeed is everything here (10). Now as to why he draws seventeen cups of Soma;-Pragapati is seventeenfold, Pragapati is the sacrifice: as great as the sacrifice is, as great as is its measure, with that much he thus wins its truth, its prosperity, its light (11). And why he draws seventeen cups of Sura;--Pragapati is seventeenfold, Pragapati is the sacrifice: as great as the sacrifice is, as great as is its measure, with that much he thus wins its untruth, its misery, its darkness (12). These two amount to thirty-four cups; for there are thirty-three gods, and Pragapati is the thirty-fourth: he thus wins Pragapati (13)." Madhu Vidya/336

Page #362
TRACE OF AN OLD PALATAL IN SANSKRIT 17 Now in the above extract the use of ujjayati with Prajapati and sarvam idam is quite understandable since this identification is common in the Brahmana literature and according to the Ap. S. S. the very purpose of the Vajapeya sacrifice is the winning of Prajapati (cf. Prajapatim apnoti 18.1.3). What strikes us, however, is its use with both Soma and Sura (ubhe andhasi) without apparent distinction when the former has been identified with truth, prosperity, and light and the latter with untruth, misery, and darkness. EGGELING translates ujjayati in both contexts as 'wins'. But it must be admitted that this reads rather awkward. It is reasonable for the Brahmana passage to tell us that the Adhvaryu seeks to win for the sacrificer truth, prosperity, and light with the help of the Somagrahas, but it is quite contrary to our expectation to hear the same text tell us that the other priest seeks to win for the sacrificer also untruth, misery, and darkness by drawing the Suragrahas. In all our ancient literature, whether philosophical or otherwise, we often find passages to show that gods and men have aspired to gain truth, prosperity, and light and not their opposites. To give only a few instances we may cite the following about truth (satya) : tasya va etasyagnyadheyasya / satyan evopacarah sa yah satyan vadati yathagnim samiddham tam ghrtenabhisinced evam hainan sa uddipayati tasya bhuyo-bhuya eva tejo bhavati svah-svah sreyan bhavaty atha yo 'nstan vadati yathagnim samiddham tam udakenabhisinced evam hainam sa jasayati tasya kaniyah-kaniya eva tejo bhavati svah-svah papiyan bhavati tasmad satyam eva vadet / Sat. Br. 2.2.2.19. EGGELING translates (SBE 12. 312-313) - "Now, attendance on (or, the worship of) that consecrated fire (agnyadheya) means (speaking) the truth. Whosoever speaks the truth, acts as if he sprinkled that lighted fire with ghee; for even so does he enkindle it; and ever the more increases his own vital energy, and day by day does he become better. And whosoever speaks the untruth, acts as if he sprinkled that lighted fire with water; for even so does he enfeeble it: and ever the less becomes his own vital energy, and day by day does he become more wicked. Let him, therefore, speak nothing but the truth."1 Similarly we get passages to show that it was prosperity (sri), and not its opposite, that was considered desirable by gods and men. To quote again from the Satapatha Brahmana 14.1.1.3: ta asata / sriyam gacchema yasah syamannadah syameti tatho eveme satram asate ye satram asate sriyan gacchema yasah syamannadah syameti / EGGELING (SBE. 44.441)-"They. 1. For the desirability of truth as a protective power also cf, the famous instance given by Uddalaka Aruni to svetaketu in Ch. Up. 6.16. For the use of the verb vji with satya cf. satyajit VS 17.83, AV 4.17.2, satyajiti Katy. $.S. 19.5.4. 2. i.e., the gods. Madhu Vidya/337

Page #363
18 M. A. MEHENDALE entered upon the session thinking, 'May we attain excellence! may we become glorious! may we become eaters of food!' And in like manner do these (men) now enter upon the sacrificial session thinking, 'May we attain excellence! may we become glorious! may we become eaters of food!'" It is hardly necessary to give again many instances to show that what applies to truth and prosperity, applies equally to light (jyotis). To give a random instance from the Rgveda we may cite: jiva jyotir asimahi (7.32.26) "living, may we obtain light", and refer to the famous prayer from the Bshad. Up. tamaso ma jyotir gamaya (1.3,28) "Lead me from darkness to light." As against the above passages showing that it were satya, sri, and jyotis which were considered as desirable possessions, we may cite a few others just to show that their opposites were never sought to be 'won', but were deemed as worth 'striking down or driven away' both by gods and men. (1) ansta : avatiratam anitani visva stena Mitravaruna sacethe / RV. 1.152.1. "Alle Ungesetzlichkeiten unterdruecktet ihr; ihr haltet es mit dem Gesetz, Mitra und Varuna!"5 (GELDNER) (2) papman : yatha vai manusya evam deva asan te 'kamayantavart tim papmanam motyum apahatya daivim sarsadan gacchemeti ta etam Caturvimsatiratram apasyan tam aharan tenayajanta tato vai te 'varttim papmanam metyum apahatya daivim samsadam agacchan ya evam vidvansas Caturvimsatiratram asate 'varttim eva papmanam apahatya sriyam gacchanti srir hi manusyasya / daivi samsaj.... (TS, 7.4.2.1-2). KEITH (HOS. 19.600): "As are men, so were the gods in the beginning. They desired, 'Let us strike off the misfortune, the evil of death, and reach the conclave of the gods.' They saw this twenty-four night (rite); they grasped it, and sacrificed with it. Then they struck off the misfortune, the evil of death,6 and reached the conclave of the gods. Those who knowing thus perform the twenty-four night (rite) strike off the misfortune, the evil, and win prosperity, for the conclave of the gods is in the case of man prosperity...." 3. Also cf. RV 2.27.11; 3.34.4; 4.1.14 etc.; VS 8, 52,20.21; AV 8.1.21; 8.2.2; Sat. Br. 14.1.1.33. 4. Rather Suntruths' and 'truth'. 5. Also cf. RV 7.66.13; VS 6.17; etc, 6. Rather 'evil (and) death'. 7. It may also be noted that while taking fire from the Garhapatya, one says: uddhriyamana uddhara papmano ma yad avidvan yac ca vidvarns cakara/ (Sarkh. S.S. 2.6.6). Also cf. RV 1.24.9; 3.7.10; VS 3.45; AV 1.115.1-3; 10.1.10; 3.4; etc. Madhu Vidya/338

Page #364
TRACE OF AN OLD PALATAL IN SANSKRIT 19 (3) tamas: sauryam bahurupam alabhetamum evadityarn svena bha gadheyenopadhavati sa evasmat tamah papmanam apahanti praticy asmai vyucchanti vyucchaty apa tamah papmanam hate / (TS. 2.1.10.3). KEITH (HOS. 18. 144): "he should offer to Surya (a beast) of many forms; verily he has resort to yonder sun with its own share; verily it drives away the darkness, the evil, from him, the dawn shines upon him, he strikes away the darkness, the evil." It will thus be seen that we often come across passages to show that anyta, etc., were not considered fit to be 'won' in the same sense as satya, etc. It should not be argued against the objection raised here to the use of ujjayati with ansta, etc., to say that it is used in the sense 'to conquer, to bring under control'. For apart from the fact that udvji is not used in this sense elsewhere, it is worth noting that we do not come across statements to illustrate the use of Vji with ansta, papman, and tamas. What we get instead is their use with verbs to mean 'to cross over, to go beyond', 'to strike or drive away', 'to burn', or 'to shake away, to abandon'. A few instances may be given here which are taken from the principal Upanisads.9 tarati sokan tarati papmanam Mundaka 3.2.9, nainam papma tarati sarvam papmanam tarati nainan papma tapati sarvar papmanam tapati Brhad. 4.4.23, papmanam apahatya Ait. 3.8.4, Br. 1.3.10,11 (papmanam motyum apahatya athaina motyum atyavahat), hanti papmanam jahati ya evar veda Br. 5.5.3,4; apahatapapma Ch. 8.7; sarire papmano hitva Taitt. 2.5; ya idam sarvam papmano 'trayata Ait. 2.1: asva iva romani vidhuya papam Ch. 8.13; sarvan papmana ausat Br. 1.4.1; yady api bahu iva papan kurute sarvam eva tat sampsaya Br. 5.14.8; papanudan Sve. 6.6; na sa papmano vyavartate Bp. 1.5.2; evam hasya sarve papmanah praduyante Ch. 5.24; tasmai mTditakasayaya tamasas param darsayati Ch. 7.26; tamasah param gamisyati Maitri 6.30; svasti vah paraya tamasah parastat Mundaka 2.2.6; tamah pranudati Maitri 2.2; bhittva tamah Maitri 6.24. Starting then from the fact that anita, papman, and tamas were considered as something 'to be driven away, to be given up' it is possible to suggest a more satisfactory explanation of ujjayati when used with them. In all probability it seems to stand for an older form *ujjhayati (<*uj-zhayati) meaning 'abandons, gives up, etc. The loss of aspiration in this form seems to have occasioned its mingling with ujjayati 'wins, etc. which occurs 8. We may also cite: sasvat putrena pitaro 'tyayan bahulan tamah/ Ait.Br. 7.3. tamas is called ajusta in RV 7.75.1. 9. For other literature one may do well to look up to the Pertersburg Woerterbuch under the respective words. Madhu Vidya/339

Page #365
20 M. A. MEHENDALE so often in this section of the Satapatha Brahmana. This (ud-) *jhayati can be derived from IE root * ghei 'verlassen, fortgehn' which is given by WALDEPOKORNY I 542-43 and POKORNY 5.418-19. *jhayati (1st conj. cf. Avestan participle uzayanto) is obviously akin to Sk. jahati (3rd conj.) 'abandon, etc.' going back to IE * ghe, from which we have the past participle ujjhita (ud + * ghita) with the same prefix as in * ujjhayati.10 With this explanation the text under consideration would mean that the drawing of the seventeen Suragrahas was intended for symbolising the abandoning (and not winning) of ansta, papman, and tamas. When ujjayati occurs in the expression ubhe andhasi ujjayati we have naturally to suppose that here both ujjayati and * ujjhayati have fallen together so that once ujjayati means 'wins' when it refers to Soma, and once it means 'abandons (as coming from * ujjhayati) when it refers to Sura. The passage quoted at the commencement of this article can now be translated, with the necessary changes in EGGELING'S translation, as follows: "He (the Adhvaryu) then draws seventeen (other) cups of Soma, and (the Nesti) seventeen cups of Sura. These two (saps of) plants, to wit the Soma and the Sura, belong to Prajapati; of these two the Soma is truth, prosperity, light; and the Sura is untruth, misery, darkness. Both these very (saps of) plants he thereby (respectively) wins (ujjayati) and abandons (ujjayati <*ujjhayati); for he who offers the Vajapeya wins everything here, since he wins Prajapati and Prajapati is indeed everything here ...... And why he draws seventeen cups of Sura; - Prajapati is seventeen-fold, Prajapati is the sacrifice: as great as the sacrifice is, as great as is its measure, with that much he thus abandons its untruth, its misery, its darkness. These two amount to thirty-four cups; for there are thirty-three gods, and Prajapati is the thirty-fourth; he thus wins Prajapati." It may now be shown that the above interpretation of the passage based on the two-fold derivation of ujjayati is supported by the ritual practice of the Vajapaya and further by a literary usage in the Mundaka Upanisad. First then to the ritual. The method of drawing and disposal of these two sets of cups, viz. those of Soma and of Sura, clearly show that it was intended from the beginning to keep a complete distinction between the two, and that the two were not allowed to co-mingle. The two were purchased separately, had a separate entrance, separate drawing, separate placing, and separate disposal. This procedure is a clear pointer to the fact that there was no question of 'winning the Suragrahas and what it stood for, not also 10. Sk. ujjhati 'leaves, gives up, etc.' occurring from the epics is clearly a new formation from ujjhita as already noted by UHLENBECK (Kurz. Et. W.). His other explanation based on WACKERNAGEL L 164 is not probable. For the above explanation of ujjhita sud-hd through ud-zhita see LEUMANN IF 58.20 ff. (1942). Madhu Vidya/340

Page #366
TRACE OF AN OLD PALATAL IN SANSKRIT 21 of 'conquering' them, but undoubtedly abandoning them, giving them away. To give some details about the procedure it may be mentioned that the seventeen Soma cups are drawn by the Adhvaryu seated in front of the axle of the Soma cart with his face westwards, while the Sura cups are drawn by the Nest (or Pratiprasthaty according to the Ap. S.S.) while sitting behind the axle with his face turned eastward (Sat. Br. 5.12.16). Then there are two separate earthern mounds (khara) erected for depositing the cups, one in front of the axle for the Soma cups and one behind the axle for the Sura cups. The purpose of erecting two mounds is stated as net somagrahams ca suragrahams ca saha sadayama (5.1.2.15) "lest we should deposit together the cups of Soma, and the cups of Sura" (EGGELING). The Adhvaryu and the Nestr do not hold the Soma and the Sura cups beyond the axle nej jyotia ca tamas ca samsrjava (5.1.2.17) "lest we should confound light and darkness" (EGGELING). The Adhvaryu now says samprcau sthah sam ma bhadrena priktam with reference to the Soma cups before. placing them on the mound. The Nestr, however, says, vi prcau stho vi ma papmana priktami with reference to the Sura cups. The significance of the latter mantra is explained in the Sat. Br. 5.1.2.18 as-"Even as one might tear a single reed from a clump of reed-grass, so do they thereby tear him out of all evil: there is not in him so much sin as the point of a grassblade."12 (EGGELING). This passage as well as the following given below should leave no doubt about the interpretation of ujjayati with the Sura cups as coming from ujjhayati. In the following section we are told that the Somagrahas are offered and drunk at the evening pressing (S. B. 5.1.2.19). But about the Suragrahas we read-"And the Neshtri, taking the cups of Sura, steps out by the back door. He walks round by the back of the hall, and placing one (of the cups) in the Vaisya's, or Raganya's, hand, he says, (SB. 5.1.5.28) 'With this I buy him of thee!'13 For the Soma is truth, prosperity, light; and the Sura is untruth, misery, darkness: he thus imbues the Sacrificer with truth, prosperity, and light; and smites the Vaisya with untruth, misery, and darkness." (EGGELING). 11. VS. 9.4. 12. tad yathesikam munjad vivrhed evam enam sarvasmat papmano vivrhatas tasmin na tavac canaino bhavati yavat trnasyagram/ 13. This refers to the taking of the Madhugraha from a Vaisya or a Sudra in exchange of the Suragrahas. The Madhugraha is then given to the Brahman priest (cf. also Katy. S.S. 14.4.15-17). 14. anena ta imam niskrinamiti.... satyam evaitac chriyam jyotir yajamane dadhaty antena papmana tamaad Valiyah vidhyati/. It is worth noting that the Sura is not offered in the Ahavaniya. The Sura cups are taken to the Marjaliya, shaken, and drunk by those who participated in the race. Cf. Ap. S.S. 18.7.2, 4, 8. Madhu Vidya/341

Page #367
M. A. MEHENDALE Now about the Upanisadic passage in favour of the interpretation of ujjayati in the sense 'to win' only when going with satya and not its opposite anrta. In the Mundaka 3.1.6 we read the famous line-satyam eva jayate nanrtam. It has been usual to take satyam here as the subject of jayate15 and accordingly translate the line as 'truth alone conquers, (and) not falsehood.' But there is some difficulty about this interpretation, because in the principal Upanisads satya appears only as something to be described, being often identified with Brahman, Atman, Aditya or Dharma; 16 or as an object of upasand or knowledge; 17 or as a means to obtain the soul.18 About satya as something worth seeing at death we have the well known verse from the Isopanisad 15 (also Br. 5.15.1): hiranmayena patrena satyasyapihitam mukham/ tat tvam pusann apavrou satyadharmaya drstaye//19 But nowhere does satya appear as a subject being associated with any activity as its agent. 20 In the light of these observations it would not be possible to construe satyam as the subject of jayate in the line referred to above. Also the context does not justify it. Just in the preceding verse we are told about truth (and knowledge etc.) being used as a means by an ascetic to obtain the soul (satyena labhyah...... hyesa atma...... yam pasyanti yatayah ksinadosah- Mundaka 3.1.5). An ascetic (yati) is the subject there, satya the means. In our verse, in the second half, we are told that the sages go along the devayana to reach the place which is the highest store of truth (yenakramanty rsayo hy aptakama yatra tat satyasya paramam nidhanam). A sage (rgi) is the subject here, satya apparently the object. In between these two statements, it is not correct to regard satyam as the subject of 22 15. The controversy about the reading jayate or jayati may be left for the time being to a critical examination of the manuscript material. 16. tad etad aksaram brahma sa pranas tad u van manah/ tad etat satyam tad amrtam tad veddhavyam Somya viddhi// Mundaka 2.2.2, tasya ha va etasya brahmano nama satyam iti/ Chand. 8.3, satyam hy eva brahma Br. 5.4; tat satyam sa atma tat tvam asi Svetaketo/ Chand. 6.8.16; tad yat tat satyam asau so adityah/ Br. 5.5.2; yo vai sa dharmah satyam vai tat/ Br. 1.4.14. 17. te devah satyam eva upasate/ Br 5.5.1, also 6.2.15, satyam tv eva vijijnasitavyam/ Chand. 7.16. 18. satyena labhyas tapasa hy esa atma samyajjnanena brahmacaryena nityam/ Mundaka 3.1.5; evam atmani grhyate 'sau satyenainam tapasa yo 'nupasyati/ Sve. 1.15. 19. On which LUEDERS (Varuna, p. 26) observes: "Aber auch die Sonne ist doch nur ein Abglanz der hoechsten, reinen Wahrheit. Erst wenn in der letzten Stunde die Seele den Koerper verlaesst, schaut man rein und unverhuellt die Wahrheit hinter der Decke der Sohne." 20. Perhaps the only exception could be in the cosmogonical account in the Br. Up. 5.5.1 where we read: apa evedam agra asuh/ ta apah satyam asrjanta satyam brahma brahma prajapatim......But usually here satyam and Brahma are considered identical. Hence HUME translates: .....That water emitted the Real - Brahma (being) the Real-; Brahma Prajapati.....". Madhu Vidya/342

Page #368
TRACE OF AN OLD PALATAL IN SANSKRIT 23 jayate. Obviously a yati or a rsi is intended as the subject and satya the object. The line therefore means" (He) wins for himself (hence perhaps the use of Atmanepada) only truth and not untruth (as means and as an end)." It would be interesting to cite here Sankara's comments on this passage, though his ultimate interpretation differs from the one suggested here as he does not take satyam and anytam as objects of jayate. He says, "na hi satyanrtayoh kevalayoh purusanasritayor jayah parajayo va sambhavati/ prasiddham loke satyavadina 'nstavady abhibhuyate na viparyayo 'tah siddham satyasya balavat sadhanatvam/". One cannot object to this interpretation on the ground that in the preceding verse (Mundaka 3.1.5.) and in the present one (3.1.6) yatayah and Tsayah are plural forms while in satyam eva jayate we have supposed yati or rsi in the singular. For, in the present section of the Upanisad we find the use of singular in many other verses. But if the point is still stressed, it has to be pointed out that jayate lends itself being interpreted also as 3rd plural Atm. Vji in the 2nd conjugation attested in Vedic forms like jesi. It is thus clear that the Upanisadic line satyam eva jayate nanrtam gives good evidence to show that in the ancient tradition it was truth alone which was regarded as fit to be won and not untruth and that the use of Vji was suited for satya and not ansta. It would, therefore, be correct to take ujjayati 'wins' only with the Soma cups, and not the Sura cups in the Brahmana passage; in the latter case ujjayati = *ujjhayati 'abandons'. Madhu Vidya/343

Page #369
LIMITATIONS OF THE METHOD OF INTERNAL RECONSTRUCTION* Vu M. A. MEHENDALE, Deccan College, Poona In his article "Sound Change and Linguistic Structure" (Language 22-138-43, 1946) Prof. Henry M. Hoenigswald discusses from the point of view of internal reconstruction various kinds of sound changes and their effects on linguistic structure. In SS 2 of this paper he takes up the case of 'phonemic change without loss of contrast'. To illustrate this point he offers as an example the Germanic sound shift. To quote: "IE voiceless stops were changed in Germanic to spirants (e. g. [t] to [0]), and voiced stops to voiceless ones (e.g. [d] to [t]). But after voiceless consonant, only some kind of [t] occurred in IE. The latter, which was presumably not changed at all at the time of the general shift, was originally most similar to the stop of the old [t]-words; it is new most similar to that of the new[t]words. .... The alternation between th and t in Goth. salbo-b-s 'anointed' and haf-t-s 'restrained' reflects the redistribution of allophones and allows us to reconstruct one original morpheme-to-. See $ 3c and SS 4a." The conclusion in the above quotation about the reconstruction of "one original morpheme-to-" is, however, based upon our knowledge of the pre-history of Germanic derived from the comparative method. It does not arise in the form in which it is presented above (morpheme-to-) out of the application of the method of internal reconstruction demonstrated in g 3c and SS 4a to which we are referred at the end of $ 2. As mentioned in $ 4a, if in a language, of the two given phonemes one occurs more freely and the other is restricted, and further if there is a regular and compulsory paradigmatic alternation between these two phonemes, then it is legitimate to conclude that this alternation reflects a process by which what now appears as a restricted phoneme has undergone a phonetic change into what now appears as a more free phoneme. If we apply this method of internal reconstruction to the above Germanic case we shall have to conclude that the original past passive * I am thankful to Dr. A. M. Ghatage, with whom I had the benefit of discussing Prof. Hoenigswald's paper, for some useful suggestions. Madhu Vidya/344

Page #370
M. A. MEHENDALE participle morpheme was -00- and not-to- In Gothic e is restricted in its occurrence (it does not occur after voiceless spirants) while t is more free (it can occur after voiceless spirants). The compulsory alternation between restricted and free t in the above morpheme therefore would indicate that there has occurred a previous conditional sound change of e to t in these positions from where it is now excluded in the language. The original reconstructedm orpheme on the basis of this internal evidence. therefore, would be -00- which has become-to- under certain conditons, If we assume that the original morpheme was --to-, how shall we state the conditions about its change? Our reconstruction of the morpheme -00from internal evidence, however, does not agree with the original IE morpheme -to-arrived at on comparative evidence. Hence this seems to be a limitation of the method of internal reconstructon where phonemic change without loss of contrast is involved. This case discussed under $ 2 is partly similar and partly dissimilar with the one discussed under $ 5a. Under the latter Prof. Hoenigswald takes up a case where compulsory alternation arising out of a primary phonetic change is disturbed and made non-compulsory by a secondary sound change and thus prevents internal reconstruction. It may, however, be pointed out that theoretically reconstruction would be possible even in such cases if by some procedure the cases of secondary change can be set aside and the non-compulsory alternation changed to compulsory one! Thus it may happen that some allophones of /x/ merge with /y/ so that /x/ becomes limited and alternates compulsorily with the more free /y/. Further it may happen that subsequently the missing portion of /x/ is filled up by a change in the pronunciation of another phoneme /z/ which is now reassigned to /x/ on the basis of phonetic similarity. This will disturb the nature of the first alternation between /x/ and /y/ and make it non-compulsory. Now this filling up of the missing /x/ could be of various types. If it is of the type illustrated by Prof. Hoenigswald, which is a case of partial filling up, internal reconstruction would be difficult. Also if the phoneme /z/ completely merges with /x/, the previous alternation between /ya/ and /xb/ will be non-compulsory. But if on the other hand /z/ changes to /x/ only in the positions in which /x/ has become /y/, say before /a/, and this affects paradigm, then /z/ becomes restricted and /x/ more free and the alternation between /xa/ and /zb/ would be regular and compulsory. This will permit internal reconstruction of the phoneme /z/ 1. Compare a similar possibility mentioned in SS 6 of recovering the old structure by setting aside foreign vocabulary. Madhu Vidya/345

Page #371
LIMITATIONS OF THE METHOD OF INTERNAL RECONSTRUCTION which has undergone a conditioned sound change. When these cases of /xa/ are recognized as arising out of /za/ and therefore set aside, the alternation. between /ya/ and /xb/ owing to the primary change will also be compulsory and permit internal reconstruction of the phoneme /x/. 3 What has been said above about the reconstruction of the Germanic morpheme -to- applies also to Prof. Hoenigswald's observation on the possibility of reconstructing an IE morpheme dhe/dha on the basis of Latin evidence. After giving a comparative reconstruction on the basis of OscoUmbrian and Latin correspondences he observes: "This is as far as the comparative method itself will take us, on the basis of Osco-Umbrian and Latin alone. It does not tell us, for instance, that the set fff is the reflex of four originally contrasting IE phonemes (*bh-, dh-, *gWh-, and s before r)...... The same facts can also be recovered by internal reconstruc tion on the basis of such alternation as Lat. con-dere 'found' (with d) - feci. I made' (with f), representing an IE morpheme dhe/dha with and without a prefix. "2 But appplying the method of internal reconstruction we do not arrive at this conclusion. As it is the alternation between Latin f- and -d-is regular but not compulsory since instances of -f- are also found. If, however, on same other evidence we are in a position to set aside the instances of -f- as due to borrowing, then the alternation f-"-d- becomes both regular and compulsory. Now in this case f is restricted and d free. Therefore by applying Prof. Hoenigswald's rule of internal reconstruction the reconstructed phoneme will have to be set up as f and not d (much less dh). Our reconstruction will mean that Latin shows a split: pre-Latin "f> f-, -d-, the latter merging with the other d. It cannot reveal that Latin f-Page #372
INTERNAL AND COMPARATIVE RECONSTRUCTION* (Some Procedural Considerations) Vu M. A. MEHENDALE, Deccan College, Poona The procedure for Internal Reconstruction has been outlined by Henry M. Hoenigswald, Language 22.138-43, 1946 (and also in his recent book--Language Change and Linguistic Reconstruction, Chicago 1960). Wallace L. Chafe, while writing on Internal Reconstruction in Seneca, Language 35.477 ff. 1959, points out what he considers to be the shortcomings in the existing theory and gives suggestions for a revised theory, In the first instance Chafe points out that there is an essential methodological similarity between internal and comparative reconstruction in as much as both are based on the comparison of cognate forms. In comparative reconstruction the cognates are taken from different but genetically related languages; in internal reconstruction the comparison is made between cognate allomorphs. Thus we get a correspondence t/d from the German allomorphs Bunt and Bund(e) or Sanskrit allomorphs sarat and sarad(a). We have another set t/t extracted from laut and laut(e) or Sanskrit marut and marut(a). From the point of view of distribution these two sets contrast word final (t, t) because both occur in the environment . We therefore before a vowel (d, t)' assign them to two different phonemes *d and *t. Now we have also a set d/d which never occurs in the above environment in which t/d occurs. We can therefore combine these two sets and assign them to one phoneme *d. This is a good point made by Chafe as it shows how internal reconstruction is similar to comparative reconstruction. But if this was the only method available for reconstruction on internal evidence, it would have a liinitation which does not figure in Hoenigswald's procedure. For, Chafe's method can work only if the language fortunately shows a paradigmatic set Paper submitted for discussion at the Autumn Seminar in Linguistics, M. S. University, Baroda, Oct.-Nov. 1962. 1 Also cf. J. W. MARCHAND, Internal Reconstruction of Phonemic Split, Language 32.245-53 (1956). 2 It is of course presumed that allomorphs are cognates, unless suppletion has clearly taken place. Madhu Vidya/347

Page #373
42 M. A. MEHENDALE t/t, in the above instance, occurring in the same environment as the set t/d. However, it cannot be assumed that such a set in a language would always occur, and if it does not, then there is no set with which the set showing the alternation t/d could be compared. It will remain in complementary distribution with both t/t and d/d which do not occur in the environment of t/d. Chafe's feeling that Hoenigswald has left vague this point of the procedure which helps one to make the choice of the reconstructed phoneme is not correct. Hoenigswald has unfortunately made one statement regarding the reconstruction of Gmc. t, on the basis of the alternation between t and 0, which has led to this impression. It appears that Hoenigswald has been influenced by his knowledge of comparative Indo-European when he lebels his reconstruction as *t. Hoenigswald probably did not feel it necessary to test his statement by applying the procedure of internal reconstruction outlined by him because, in the light of our present knowledge, the reconstruction as such (*t) is the correct one. But if he had done so he would have found that the evidence leads to *0 (and not to *t). That this * is after all to be replaced by *t is shown by comparative evidence and cannot be obtained by internal one. This is not the fault of the procedure but inerely constitutes a limitation of internal reconstruction But for the above statement, Hoenigswald has clearly stated the procedure regarding the choice of the reconstructed phoneme: "Compulsory alternation between restricted /x/ and free /y/ in a paradigm indicates a previous conditioned sound change from /x/ to /y/ in the position from which it is 3 Word final t, as in [unt] (und), would help Hoenigswald but not Chafe as it does not yield any paradigmatic alternation. 4 See A. M. Ghatage, Indian Linguistics 21.88 (1960) and M. A. Mehendale, Limitations of the Method of Internal Reconstruction, Indian Linguistics 21.101-103, 1960. Hoenigswald's defence against Chafe's objection (Phonetic Similarity in Internal Reconstruction, Language, 36.191-92, (1960) is not a very happy one, because it forces him to bring in a statement like "Furthermore, we know-whether from internal reconstruction or in other ways-that a Gothic t may also have another antecedent /*z/, (e.g. in the word which corresponds to E nest)." (p. 191) How one can bring forward this as an argument when one has set oneself the task of internal reconstruction is not clear. His footnote 6"It would have been better to punctuate: one orignal morpheme, "-to"!" is also not very helpful. It is intended to suggest that the reconstruction pertains to "one morpheme' and that the label' to'attached to it is of no significance. All this has come in because one hesitates to admit the limitation of the internal reconstruction. Madhu Vidya/348

Page #374
INTERNAL AND COMPARATIVE RECONSTRUCTION now excluded." (Language, 22.140). When this rule is applied to the German or Sanskrit alternation between t and d noted above it unmistakably points to the reconstruction of *d which is arrived at by Chafe by comparison of similar sets. While there is this similarity between internal and comparative reconstruction, Chafe points out one essential difference between them. For, whereas comparative reconstruction yields inferences about one particular stage of the proto-language, internal reconstruction yields inferences about not one but several different stages in the development of a language. Chafe also shows the way of determining the chronological order of at least some of the phonetic changes which have come about at different periods in the history of a language. This can be done, he argues, wherever we have an automatic (or compulsory) alternation, i. e. one which is predictable in phonological terms applying throughout the language, and a non-automatic (or non-compulsory) alternation, i. e, one which is predictable only in a given grammatical environment. According to Chafe automatic alternations are the results of the most recent sound changes, because the compulsory nature of the alternation has not been disturbed by subsequent sound changes. Therefore the change reconstructed from automatic alternation can be regarded as only one stage removed from the attested stage. On the other hand a change reconstructed from nonautomatic alternation is two stages removed from the attested stage because the automaticity of the earlier stage has been disturbed by a subsequent sound change, Now with regard to this procedure about determining the chronology of sound changes on the basis of the nature of the alternation, compulsory or otherwise, one thing has to be made explicit. What has been said above will apply only in cases where a reconstruction arising out of a compulsory alternation renders another non-compulsory alternation compulsory. We may, however, have a case of a different type where this relationship does not exist. We know that some of the phonetic changes are gradual in their spread, and if the recorded history of a language gives evidence for two different stages, it will show that the change is non-compulsory at the earlier but compulsory at the later stage. In such a case the non-automatic earlier stage is not the result of any disturbance caused by an intervening change but because of the fact that the particular sound change had not covered all cases at a given period, At a later stage, however, these 5 See the example given by Chafe of the reconstruction of Gk. *ti>si as one stage removed from the attested one, while that of Gk, genes-08 >gene-os being two stages removed. Madhu Vidya/349

Page #375
M. A. MEHENDALE exceptions were also covered by the phonetic change and made it automatic. The alternation between s and t in Sanskrit may be cited as an instance, This alternation is automatic in classical Sanskrit vis-a : vit-su, vit-patibut it is not so in Vedic Sanskrit. There it is regular in the paradigm vis-a: vit-su, but not applicable for the whole language, since we get words like -vis-pati, vis-patni (and even vis-pala it that is the correct analysis of the word). If one keeps to the formula, without any other considerations, that a non-automatic alternation reflects a later stage of the language, then it will be seen from the Sanskrit example that that is not always the case. Non-automatic S t is the earlier one, automatic s~t is the later. A reconstruction based on the late Sanskrit automatic alternation s t as *S>t will only find confirmation in the instances like vis-pati found in Vedic language. The Seneca illustration given by Chafe (p. 484, SS 25) of an anomalous form ?oxtne?ta? 'fern' need not necessarily be explained as a result of some recent phonetic change which is responsible for the sequence tn in it. It may as well be the case of an old survival which has until now resisted the change *tn >hn. There is also another way of deciding the chronology of phonetic changes in terms of internal reconstruction when we get a double alternation in a given paradigm. If, among these two alternations, one concerns phonemes which are phonetically closer, then it may be said that the phonetic change which led to this alternation occurred earlier than the one which led to the alternation between phonemes not so close. Thus e. g. in the paradigm of Sanskrit vis we have the alternation s t: vis-a, vit-su and S d: vis-a, vid-bhih. In these two alternations phenemes s, t, and d are involved. Of these, s and tare phonetically closer than s and d. Therefore we conclude that the change *s > occurred before t was affected by another change leading to the alternation s~. Another instance of the above type of alternation is found in s alternating with k and g in Sanskrit dis-a, dik-su, dig-bhih, and as stated above, we suspect that the change s>k occurred before k, in certain environments, was replaced by g as a result of a subsequent change.6 As regards the relative merits of internal and comparative reconstruction Hoenigswald observes, "In principle, internal reconstruction cannot claim to yield phonetic detail any more than the comparative method does". (Lg. 36, p. 192, 1960). However one is inclined to feel that in this respect internal reconstruction stands on a somewhat different level than the comparative one. Since, in a majority of cases, internal reconstruction is likely to 6 By the way, these two alternating types s:t:d and $:k:g will show that these are not cases of compulsory alternations in the strict sense. Madhu Vidya/350

Page #376
INTERNAL AND COMPARATIVE RECONSTRUCTION 45 yield facts which, at a time, are removed only one or two stages from the attested one, it may be regarded to be in a better position to give phonetic details of the reconstructed stage. Therefore it seems reasonable to uphold Chafe's contention that internal reconstruction should precede the comparative method. This in fact will help us to eliminate some of the problems of comparative reconstruction. For example if on the basis of the alternation t~0 in Germanic, the third person sg. present is first internally reconstructed as 0,*is would replace ist'is'. In that case while attempting comparative method, Hoenigswald's set 1 (cf. Principal Step, Lg. 26.358) t/t, extracted from Skt. asti : Gmc ist, will disappear. Instead we will have t/0 which will be identical with his set 3. This will lead to economy, as we have been able to reduce the number of sets and also eliminate the difficulty encountered by Hoenigswald about grouping the set 1 with 2 and 3 or with 5 (Language, 26.360). If it is agreed to attempt internal reconstruction first before proceeding to comparative niethod, then it may be suggested to take one more step in between and apply the information made available by internal reconstruction to some specific items in the data. For instance, we have in Sanskrit an alternation t~t in the past passive participle morpheme-ta, e.g gata : tusta. Internally t is reconstructed as *t in this morpheme since in Skt. t never occurs after s. We now make use of this information and reconstruct t as *t after s also in other items where no alternation between t and t is available, e. g. Skt. asta < *asta 'eight'. This is done because while comparing Skt. with Avesta it helps us to get rid of a set t/t which we would have extracted from cognates Skt. asta: Av. asta. To take another example, it has already been shown that word-final -t can be reconstructed as *-d in German /Bunt/ and Sanskrit /sarat/. We may now proceed further from this step and also reconstruct -t as *-d in some other items where no paradigms are avainable, e.g. German /unt/, and Sanskrit prefix /ut/, indeclinables like /isat/ and /yugapat/, and the abl. ending /-at/ (even when the latter had not shown final /-d/ in external sandhi). It is, of course, not advisable to do this reconstruction if none of the related languages shows final /-d/ in the cognates. A typical problem in comparative reconstruction has been posed by W. S. Allen (TPS 1953, p. 82 ff.) and its solution suggested by A. M. Ghatage (Indian Linguistics 22. 82-85, 1961 and Historical Linguistics and Indo-Aryan Languages, Bombay 1962, pp. 65 ff.). Allen has raised the problem about 7 Chafe observes "In general ... ... ... internally reconstructed features tend to be more recent than those reconstructed by the comparative method" (Lg. 35.495) Madhu Vidya/351

Page #377
M. A. MEHENDALE the reconstruction of Rajasthani by comparing the evidence from Marwari, Mewari and Harauti. The reconstructed Rajasthani as we know had three phonemes bh, b, v occurring both initially and finally. Marwari has preserved all the three phonemes in these positions; Mewari has made one change viz. -bh > -b; Harauti has made two changes-bh > -b and - >-b. The comparison of Marwari and Mewari enables us to recover the older stage without difficulty. But comparison of Marwari with Harauti leads to a difficulty which the procedure available until now cannot solve. Different groupings of partially similar sets lead to an equal number of reconstructed phonemes. The principle that we should accept that grouping which leads to the smallest number of reconstructed phonemes cannot therefore be successfully applied. Allen has phonemicized Harauti data [bh--b, b- -6, 6- -v] as /bh--b, b--b, b--bh/ so that -vis phonemicized as -bh. The comparison of this with the Marwari data bh--bh, b- -, - -v leads to three possible combinations all giving equal number of phonemes. The point therefore is whether we can think of any method which would help us to choose one of the three alternative combinations. Ghatage's suggestion is as follows : "We should admit that reconstruction as valid which leads to the postulation of the least number of changes from the mother language into the daughter languages". This suggestion makes explicit the principle followed by the comparatists for reconstructing the phonemes of the protolanguage. On comparing the sound correspondences of the two related languages, inferences regarding the reconstruction are drawn in such a way that they involve the smallest number of phonetic changes. On comparing a correspondences : h between two languages, although the possibility that they are the reflexes of neither s nor h but a third sound is admitted, it is not entertained because it presupposes more phonetic changes. It seems that the Rajasthani problem can be met also in a different way. The problem has arisen due to a particular kind of phonemicization [-v] as /-bh/, and on a descriptive level it may be possible to do that. When, however, we turn to comparative reconstruction and attempt extraction of sets, a set -:-bh deduced from Marwari and Harauti cognates should always remind the comparatist about the phonetic value of -bh in Harauti which is [-v]. In such cases it is suggested that while preparing the sets it would be better to restore the phonetic value to the phoneme in a particular position. It is true that allophones are not taken into consideration while attempting reconstruction. But in the cases like above an extraction of a set -v :-bh forces us to assume one change for the daughter language whereas the restoration of the allophone -v for -bh (-vi-v) eliminates this necessity. It seems therefore desirable that allophones of a phoneme need not in all cases be neglected while attempting comparative reconstruction. Madhu Vidya/352

Page #378
INTERNAL AND COMPARATIVE RECONSTRUCTION As in the above instance, they should be availed of whenever they lead to economy. Phonemic analysis, which itself is a sort of initial 'reconstruction', regards all allophones as positional variants of a phoneme. This is all right on a synchronic level. But when we come to diachronic linguistics we know that allophones of a phoneme do not always arise because they have developed as positional variants, but because sometimes a gap has occurred somewhere owing to merger. Our aim in historical linguistics is to sort this out and try to understand as far as possible the history of each phoneme. Therefore if we get a correspondence like -V :-bh where -bh is phonetically [-v], this should be taken to yield a set -v/ -v and not -v/ -bh. Once this is done, all unnecessany combinations due to partial similarity with-bh will disappear and economy will be achieved at the very outset. Madhu Vidya/353

Page #379
SYMBOLAE LINGUISTICAE IN HONOREM GEORGII KURYLOWICZ M. A. MEHENDALE Nadbitka Extrait: SANSKRIT BADHA Sanskrit badha-, available only in the Vedic literature, means according to the lexicons "laut, stark, fest. Its accusative badham, used adverbially, 'gewiss, sicherlich etc. is freely available in the later literature. The usual etymology of badhd is given as Page #380
200 M. A. Mehendale we read in the Apastamba Srauta Sutra 11. 11. 8: 'virad asi'ti bahum upavahrtyedam aham tam valagam udvapami'ty udupyoparavanyante 'vabadhate 'gayatrena chandasavabadho valaga' iti ,,Nachdem er mit der Formel: <> 5 (Caland). This occurs with slight variations in the Hiranya. Sr. S. 7. 6. 10 ff. as follows: 'idam aham tam valagam udvapami'ti parsun udvapati / 'idam onam adharam karomity uparavabile' vabadhate / nirasto valaga' iti harati /'avabadho durasyur iti yatra kharam karisyan bhavati tasmin dese nivapati / 'gayatrena chandasavabadho valaga'ity uparavabile 'bhrya nigrhnati ,Here I dig out the charm' thus saying he throws out the earth. Here I press it down' thus saying he presses it down in the uparava hole & The charm is expelled' thus saying he carries away (some part of the dug out earth). "The evil (charm) is pressed down' thus saying he digs down (the earth) on that part where he is going to erect Khara (earthen mound). With the Gayatri metre the charm is pressed down (in the ground) thus saying he presses, it firmly with a shovel in the hole" 7. In a different ritual which concerns the preparation of the Vedi, the officiating priest removes from the place where the Vedi is to be erected pieces of darbha grass and the earth, which is dug out by the wooden sword, and buries (nivap) 'them away at some distance. This he does four times. At the time of each successive nivapana the Agnidhra priest who sits there covers the preceding nivapana with different formulas. This is given in Ap. Sr. S. as follows: 'avabadham raksa' iti dvitiye nivapana * The meaning 'discovered' assigned to avabad ha in the lexicons is clearly wrong. The ritual action consists of two parts --- digging out the earth, which stands for the digging out (udy vap) of the hidden charm, and then pressing down in the ground a part of the earth, which symbolizes the pressing down (avaybadh) of the charm. Instead of avay badh-, we get miyvap- (opposite of udyvap) in the Vaikha. Sr. S. 14. 7, 8. 6 Ap. Sr. S. 11.11.9 and 11 we read: avabadho durasyuh 'Hinabgedruckt ist der Bosartige', purvena purvena mantrena bahum upavahrtyottarenottarona chandasavabadhate nachdem er mit dem je vorhergehenden Spruch den Arm niedergesenkt hat, druckt er mit dem je folgenden Versmasse den Sand nieder' (Caland). Also cf. Baudh. Sr. S. 6.28, line 14; 3.27, lines 21-23 (Caland's edn.). * The commentator explains uparavabile as wparavabilasamipavartisumantapradeso 'vabadhate 'vanato bhutva hastena nipidayati / The place is given as yajamanasyadhaspadan in the Baudh. Sr. S. 6.28deg line 15, the Vaikha. Sr. S. 14.7. ? nigohnati (this is in place of avabadhate of the Ap. Sr. S. cited above)= dardhyar. tham abhrya nitaram grhaati samhantity arthah / Commentator. Madhu Vidya/355

Page #381
Sanskrit badhu 201 agnidhro 'bhigrhnati / avabadho 'ghasamsa' iti trtiye avabadha yatudhana iti caturthe (2.2.2) 'Der Agnidhra bedeckt das Hingestreute (rather 'dug down, buried') beim zweiten Hinstreuen (rather "pressing, burying') mit der Formel: ,,Nach unten gedrangt ist das Raksas", beim dritten mit: "Nach unten gedrangt ist der Boswillige", beim vierten mit: ,,Nach unten gedrangt sind die Zauberer" (Caland). Still in another ritual regarding the preparation of the purodasa, when the husk is separated from the rice taken out for this purpose, the officiating priest puts the husk on the potsherd and pushes it below the black antelope skin. He then presses it down with the mantra 'the evil. spirit is dug down, buried' (purodasakapalam tusaih purayitva 'raksasam bhago 'si'ti... adhastat krsnajinasyopavapati / nanviksate 'avabadham raksa' ity avabadhate/ Hiranya. 1. 5. 17). The ritual acts described above should leave no doubt about the meaning of avabadha 'dug down, pressed down, buried'. This follows from the fact that the mantra containing this word accompanies the action. of digging and pressing down the charm etc. into the ground. In the RV. 1. 106. 6 (= AV. Paipp. 4. 28. 6) we get nibadha which has the same meaning as avabadha: indram kutso vrtrahanam sacipatim kate nibalha rsir ahvad utaye /,,Indra, den Vrtratoter, den Herrn der Kraft, hat der Rsi Kutsa, als er in die Grube hinabgestossen 30 war, zu Hilfe gerufen" Geldner). Pressed down' or even buried' would be better than just hinabgestossen". 10 Badha without any prefix occurs only once in the Rgveda 1. 181. 7. There it appears to have been used as a substantive, meaning 'dug out Also Hiranyakesi, with minor variations, 1.6.21. Caland (on Ap. Sr. S. 2.2.2) remarks that these three formulas are not found anywhere else except in the Ap. and the Hiranya. Sr. S. But they are given by the Varaha Sr. S. 1.3.1.42 (with sphyenotkaram apidhatte for abhigrinati). * Commentator: hastenavastabhyoptatunadebam badhate. Ap. Sr. S. mentions only the placing of the husk below the antelope skin (1.20.9), but does not give the following mantra for pressing it down. 10 nibadha occurs in the general sense of 'overpowering, oppressing' in the Jaim. Br. (Auswahl by W. Caland, pp. 196-197): sa ha Khandikah Kesinam abhibabhuva / sa ha Kei Khandikena nibddha Uccaiheravasam Kauvayeyam jagama ,,Khandika nun uberwaltigte den Kesin. Kesin, von, Khandika bedrangt, begab sich zu Uechaiharavasa, dem Sohn des Kuvaya". According to P W, additions to Vol. 7 (p. 1779) we get sambalha in the Taitt. Ar. 1.17 which, as mentioned in PW, means sambaddha, drdha according to the commentator. I have not been able to find this word in the reference given. The Vaidikapadanukramakosa of Hoshiarpur also does not give it. Keith, Ait. Ar. p. 189, n. 10), apparently following PW, simply says, "It occurs in the Taittiriya Aranyaka but does not give exact reference. The reference in PW is obviously mistaken and seems to be intended for sambalhatama which occurs in the Ait. Ar. 1.4.1. However, the commentator's explanation referred to by PW remains a moot point. Madhu Vidya/356

Page #382
TWO NOTES ON INTERNAL RECONSTRUCTION BY M. A. MEAENDALE, Poona. (1) Reconsideration of Hoenigswald's Rulc of Internal Reconstruction. HOENIGSWALD in his paper on Sound Change and Linguistic Structure observes: "compulsory alternation between restricted / x/ and free /y/ in a paradigm indicates a previous conditional sound change from / x / to /y/ in the position from which it is now excluded." To illustrate his point, Hoenigswald gives an instance from modern Gerinan where voiced stops do not occur in syllable-final position. This gives an alternation/d/~/t/in a paradigm / Bunde/ but / Bunt 1. Since in this position /d/ is restricted in comparison to ltl, the above rule permits us to reconstruct syllable final /t/ as id/ and imagine a conditioned sound change of d to t in that position in the history of modern German. HOENIGSWALD's rule works very well in cases where for a given alternation the pertinent environnents are only two as in the above case. The rule also works satisfactorily in some cases where thc cnvironments concerned are more than two. This can be illustrated with the paradigm of Sanskrit sarad: surad-and farad. bhyon, surat and saral-s. The restriction and freedom of occurrence of the alternating phonemes in the Sanskrit language can be shown in the following way : Before a vowel 2 Before a voiced stop 3 Word final Before an unvoiced fricative X V Since in the environments which are pertinent for this alternation d is more restricted, we reconstruct it as the phoneme of the proto-stage and explain the alternation as a result of the phonetic change d>t in the environments concerned. 1. Readings in linguistics, p. 140. Madhu Vidya/357

Page #383
Sanskrit badha 203 that -dh- is secondary. But the Vedic evidence suggests that badha- is a very old participle. If the authors of the ritual texts did really intend to offer an etymological explanation of it from the root badha-, they did so simply because that was the only root available to them which was phonetically and semantically close to budha-. It is, however, possible to suggest two other etymological explanations. tadha- 'dug out; river-bed' may be derived from IE *bhodh (bhedh) 15 'stechen, bes. in die Erde stechen, graben' (Walde-Pokorny II. 188). In this derivation we will have to assume an old Indic * badh, bah- 'to dig, etc.', which on the analogy of (*gh>)h + ta 16 > dha (cf. sah: sadha) gave rise to badha-. For a similar analogical form one may compare vrudh, ruh: rudha-. The other explanation is to derive badha- (*bhngh-to-. In Avesta we have bazah- 'Tiefe' and in Sanskrit badha- 'dug out'. On the basis of this evidence it is tempting to assume an IE root *//bhengh'to dig' which would explain satisfactorily Skt baoha-, both phonetically and semantically. Poona (India) 1 H. Krahe (Beitrage zur Namenforschung 14.181, 1963) remarks: "Wie einige dieser Worter, so besonders lat. f088a, gall. *bedu., auch nhd. Fhuse-bett zeigen, ist die Wz. *bhedh. geeignet, auch Wasserlaufe zu bezeichnen". 16 H. Krahe, op. cit., p. 183 remarks on the river name Beste: ,,Es kann sich um eine Bildung mit einem t-haltigen Suffix handeln, wobei die Gruppo -d-t- (*bed-4-) zu -st. werden musste...." Madhu Vidya/358

Page #384
TWO NOTES ON INTERNAL RECONSTRUCTION BY M. A. MEAENDALE, Poona. (1) Reconsideration of. Hoenigswald's Rulc of Internal Reconstruction. HOENIGSWALD in his paper on Sound Change and Linguistic Structure observes: "compulsory alternation between restricted / xl and free / y/ in a paradigm indicates a previous conditional sound change from / x / to /y/ in the position from which it is now excluded."" To illustrate his point, Hoenigswald gives an instance from modern German where voiced stops do not occur in syllable-final position. This gives an alternation/d/~/t/ in a paradigm / Bunde/ but / Buntl. Since in this position /d/ is restricted in comparison to It I. the above rule permits us to reconstruct syllable final /t/ as /d/ and imagine a conditioned sound change of al to t in that position in the history of modern German. HOENIGSWALD's rule works very well in cases where for a given alternation the pertinent environments are only two as in the above case. The rule also works satisfactorily in some cases where the cnvironments concerned are more than two. This can be illustrated with the paradigm of Sanskrit Sarad: sarad-a and surad. bhyam, surat and sarat-su. The restriction and freedoin of occurrence of the alternating phonemes in the Sanskrit language can be shown in tlie following way : 2 Before a vowel Before a voiced stop 3 Word final Before an Unvoiced fricative dV X tv x V Since in the environments which are pertinent for this alternation d is more restricted, we reconstruct it as the phoneme of the proto-stage and explain the alternation as a result of the phonetic change d > t in the environments concerned. 1. Readings in Linguistics, p. 140. Madhu Vidya/359

Page #385
54 INDLAN LINGUISTICS It seems, however, that a straightaway application of the above rule without any further consideration may not help in some other cases where the alternating phonemes appear in more than two environments. This is illustrated by the Sanskrit paradigm of marut: marut, marul- and marut-su, but marud-bhih. As in the above case we may chart the distribution as follows : Word final 2 Before a vowel 3 Before an unvoiced fricative Before a voiced stop v v C a x V x v Since the phoneme d is more restricted, we shall have to reconstruct manu and explain the alteration as a result of the phonetic change d > t in the given environments. This will work well for the first and the third environments (word final, and before an unvoiced fricative), but not for the 2nd environment ( before a vowel), because in Sanskrit ad, as well ns a t. can occur in that position. Our reconstruction therefore gocs wrong. To order to get over the difficulty, we have to apply some more considerations. While looking to the distribution of the phonemes concerned we should not take into account all the environments together as done above, but only any two at a time. In the case of the above paradigin the following possibilities present themselves: (1) allernation in the word final position and before a voiced stop; (2) alternation before an unvoiced fricative and a voiced stop: (3) altem. ation before a vowel and a voiced stop. There is no alteration in any other two positions. Now, of the three alternatives stated above, the first two are not helpful because both the alternating phonemes, t and d, appear equally restricted: (1) t can occur in the word final position, but not before a voiced stop; d can occur before a voiced stop, but not in the word final position. The same is true of the second alternative. Therefore on the basis of these, both t and d have equal claims for reconstruction, and any choice between the two will be arbitrary. In the third alternative, however, only t is restricted and d is free: t can occur in one position, before a vowel, while d can occur in both, before a vowel and before a voiced stop. Now applying the rule of HOENIGSWALD, stated above, we can reconstruct t an 1 get rid of the difficulty of having to reconstruct d. The rule therefore should be flated with the following addition : MadhuVidya/360

Page #386
SUPPLIMENT 1968 55 If the environments in which the compulsorily alternating phonemes occur are more than two, it is necessary to consider only any two environments al a time to examine which of the alternating phonemes is more restricted." (2) Compulsory but irregular alternation indicates a merger. Sanskrit paradigm for the word dis show's an alteration $~ ~g in four environments : dis-a, dik and dik-su, dig-bhyam. The environments involed are (1) before a vowel, (2) word final. (3) before an unvoiced fricative, and (4) before a voiced stop. Of the three alternating phonemes, $ is more restricted when any two environments are considered at a tinie and therefore applying the above rule we reconstruct s. But difficulties arise when on the basis of this reconstruction we proceed to make statements about historical phonology. Actually, the following statements can be made: (1) * >k in the word final position; (2) s>k before an unvoiced fricative; (3) 5 > *k>g before a voiced stop. These statements may be correct for this paradigm and some others like those of drs and sprs. But we cannot say that they hold good for the whole Sanskrit language. This is so because we have another word vis which shows a different alternation S ~ ~d, e.g. visa, vit and vit-su, and vid-bhyam in the same environments as mentioned for dis. Thus we See two parallel developments in Sanskrit paradigms: (1) *5> k(g), and (2) s> (d). These are then instances of compulsory but irregular alternation. Compulsory becausc s in a given environment must be replaced by some other phoncme; irregular because the replacing phoncnics are not the same in given environments-sometimes k(and s), sometimes t (and (!). The conditioning factor for the difference in the development is not phonclic but morphological. We suspect, therefore, that the two treatments of $ indicate a merger of two different phoncmes of still older stage. It is thus possible to discern structural traces of two merging phonemes in paradigms. The two proto-phonemes, which subsequently merged into s before a vowel (dis-8 and vis-a), had already developed different allophones in the word final position and before an unvoiced fricative before merger. These allophones subsequently developed as k(g) in one case, and as (d) in the other. It is customary to trace Sk. $ to a single phoneme, palatal k of the I.E. stage. The line of argument followed above, however, indicates that it should be traced to two different phonemes, say ki and K2. One of these gave rise to the k-paradigm and the other to the t-paradigm. When the two phonemes merged, they were s as in disa and visa. In instances where Sk. $ occurs in a non-alternating situation, like satam or dasa, it would be difficult for us to tell from which of the two reconstructed phonemes, ki or k2, they are to be derived. 2. HOENINGSWALD has taken care of this situation by distinguishing between bilaterally automatic and unilaterally automatic alternations (Language Change and Linguistic Reconstruction p. 101). He however, does not give an example from any language, which, as shown obove is offered by the paradigm of sk. marut. Madhu Vidya/361

Page #387
INDIAN LINGUISTICS In the light of the above discussion we may formulate an assumption that an irregular but compulsory alternation, as instanced above, leads us to suspect a merger of two different phonemes of the proto-stage. The difference in the paradigm in the attes!ed stage is thus sought to be explained on the basis of there being two different phonemes at an earlier stage. Madhu Vidya/362

Page #388
EVIDENCE FOR THE AFFRICATE PRONUNCIATION OF THE CLUSTER TS IN THE MAITRAYANI SAMHITA by M. A. MEHENDALE The Maitrayani Samhita 3.4.7 prescribes different forms of the firealtar (agniciti) corresponding to the different desires entertained by the sacrificer. E.g. it lays down that the fire-altar should be of the form of a syena bird if the sacrificer wishes to obtain heaven (syenacitim cinvita svargakamah), etc. In this section we read the following passage: dronacitim cinvitannakamo dronena va annam adyate 'nurupenaivannadyam avarunddhe/ pascaccarur bhavaty anurupatvaya.? This may be translated as: 'One who desires food should construct the fire-altar in the form of a wooden trough, for, verily, food is eaten with a trough. He thereby secures food with a suitable (form of the altar). This (form of the altar) has a caru towards the west for the sake of obtaining the proper form of a trough).' Thus, according to this text, the altar having the form of a wooden trough is to be provided with a caru towards the west. But the usual meaning of caru does not seem applicable in this context. Usually the word means an oblation of rice or barley boiled with butter and milk, but this is hardly appropriate when one is prescribing the form for a fire-altar. The correct meaning of the passage is obtained when we look to the Manava Srauta Sutra, which belongs to the Maitrayani sakha, and to the Sulvasutras. The Manava S.S. 10.3.6.6 reads: dronacit tsaruman esam dasabhago bhavet tsaruh 'Of these, the troughshaped fire-altar has a handle (tsaru). The handle is the tenth part of the whole altar)'.2 Similarly in the Apastamba Sulva sutra3 13.4 ff. we read: dronacitan cinvitannakamo iti vijnayate| dvayani tu khalu dronani caturafrani parimandalani cal ... pascat tsarur bhavaty anurupatyayeti vijnayate sarvasya bhumer dasamam tsaruh). This passage, in the latter half, has an obvious reference to the Maitrayani text. It also adds that the wooden troughs are of two kinds - square and round. The commentator on the 1 The corresponding section (5.4.11) of the Taittiriya Sarn. has dronacitar cinvitannakamo drone vd annam bhriyate sdyony euannam dvarundhe 'He should pile in the form of a wooden trough who desires food; in a wooden trough food is kept; verily he wins food together with its place of birth' (Keith). 2 A Soma vessel (camasa) without a handle (atsaruka) is referred to in the Tandyamahabra. 25.4.4, Ap. Sr.S.12.2.8, Katya. Sr.S. 24.4.42, Latya. Sr.S. 10.12.13. 9 ZDMG, LV, 586, LVI, 369 f Madhu Vidya/363

Page #389
300 HENNING MEMORIAL VOLUME Baudhayana Sulva Sutra (3.217) adds that the square ones are furnished with handles, the round ones with ostha.4 Accordingly, the drona-shaped altar would look like the following figure: N . S The word tsaru in the sense of 'handle of a vessel is well attested in the srauta literature. In the sense of 'hilt' of a sword it is common in the epics and later literature. It is thus clear that what we read in the Maitra. Sam. as caru really stands for tsaru. The passage in question prescribes, as is made clear by the later sutras, that the trough-shaped fire-altar is to be provided with a handle so that it acquires the perfect form of a vessel (anurupatvaya). The normal sandhi between pascat and tsaruh should simply have given pascattsaruh. If instead we find pascaccaruh in the Maitra Sam., this indicates that at some time in the oral transmission of the Samhita text the initial consonant cluster ts of tsaru was pronounced as an affricate. It is very likely that this was a palatal affricate, since its combination with a precedingt resulted in cc. But on this point, i.e. whether the affricate pronunciation was originally palatal or dental, we may not be able to do anything better than speculate. What is certain, is that the text as read today points to an affricate pronunciation of the cluster is in the oral transmission of the Maitrayani Samhita. It is interesting to note that at another place in the Maitra. Samhita (3.8.2) where t and ts come together, the combination has given a different result. The passage runs as follows: yatra puram yudhyeyus tad etabhir juhuyat sara va esa yajnasya tasmad yat kimca practnam agnisomlyat tad upamsu caranti. 'Where they besiege a fortified place, then let him give these offerings. This, verily, is the sara of the sacrifice. Therefore whatever precedes the agnisomiya offering, that is done in a low voice.' The context is that of the three upasad offerings which are given morning and evening on three days preceding the pressing of the Soma in a Soma sacrifice. The word sara here presents a difficulty. It is not recorded in the available lexicons. Schroeder does not note any variant for it. But in another passage in the + Pandit, n.s. I, 626 (tsarumanti caturasrani parimandalany osthavanti). Madhu Vidya/364

Page #390
M. A. MEHENDALE 301 Maitra. Sam. (4.6.4) which has a similar wording, we find tsara in place of sard. There we read: tsara vd esd yajndsya tasmad yat kimca practnam agrayanat tad upamsu caranti. "This, verily, is the tsara of the sacrifice. Therefore, whatever precedes the (drawing of) the Agrayana cup, that is done in a low voice.' The word tsara is also not recorded in the lexicons. But it seems to mean 'action done stealthily' (from the verb tsar-, attested since the Rgveda, meaning 'to sneak, to go stealthily'), and this meaning fits well in the context, for it is said that whatever is recited during that period is done in a low voice. It will thus be clear that sard in the Maitra. Sam. 3.8.2 stands for tsards as in 4.6.4. Obviously, in recitation, one of the two t's was elided (juhuyat tsara > juhuyat sard), so that no sandhi t+ts> cc could take place. In 4.6.4 tsard is preceded by the vowel e of the word vrikte (vag val sarama, vacam evalsam veikte, teard... etc.); hence there was no occasion here for a consonantal sandhi. It may be noted that in Sanskrit the cluster ts, in initial position, is restricted only to the verb tsar- (and its derivatives) and the word tsaru. In most cases ts occurs intervocalically, as in abhitsaranti (RV 8.2.6), tatsara (RV 1.145.4), or is preceded by a word ending in a vowel, yatha tsari (TS 6.4.11.3), gatva tsaran (AV 12.3.13), etc. When in a few cases it is preceded by a consonant, even by t, the whole complex is well preserved in the Rgveda cf. vidat tsaruh (7.50.1,2,3) and latakratus tsarat (8.1.11). In the Maitr. Sam., as was seen above, the treatment differs. The above discussion on the affricate pronunciation of the cluster tr has also some bearing on the question of the pronunciation of the palatal stops of the c class in ancient India. W. S. Allen (Phonetics in Ancient India, 52) is inclined to the view that at the time of the Pratisakhyas and the Siksas the palatals were true palatal plosives and not pre-palatal affricates as is the general pronunciation in modern India. In the Pratisakhyas they are described as articulated at the palate with the middle of the tongue (talau jikvamadhyena Taitti. Prati. 2.36, talavyanam madhyajihvam Atharva Pra. 1.21, also Vaj. Pra. 1.66, 79). The date of the Pratisakhyas has been placed by Siddheshwar Varma between 500-150 B.c. The Rgveda Prati, and the Taitti. Prati. are thought to be earlier than Panini, while the Vajasa. Prati. and the Atharva. Prati. are considered to be later. On the other hand Wackernagel (Aind. Gr. I, SS 119, p. 137) holds the view that the pronunciation of e as an affricate t must be old. As evidence for this pronunciation he gives the Greek spellings of Sanskrit words which take us to about the fourth century B.C. Whitney seems to be undecided on this point. While commenting on the Atharva. Pra. 1.21, he observes: "The ancient Sanskrit e and j can hardly have been so distinctly compound sounds as our ch and j (in church, judge), or they would have been analysed 5 Here Schroeder records the reading sard from one manuscript. 5 Critical Studies in the Phonetic Observations of Indian Grammarians, 21. Madhu Vidya/365

Page #391
302 HENNING MEMORIAL VOLUME and described as such by the phonetists. At the same time, their inability to stand as finals, the euphonic conversion of t and following $ into ch, the Prakritic origin of c and j from ty and dy, etc., are too powerful indications to be overlooked of their close kindred with our sounds, and deviation from strict simplicity of nature'. It is very difficult to decide the age of the Maitrayani sandhi t+ts>cc. Unfortunately we do not have the complete Padapatha of the Maitra. Sam. In view of the fact that the word tsaru, and not caru, is seen in the sutra works cited above, which in their wording clearly refer to the Maitra. Sam., the change t+ts > cc could have occurred only when the recitation of the Sanhita was done independently of these Sutras and with complete disregard for the meaning. It will be useful here to call attention to a fact of the chronology of phonetic changes. We have already seen that the cluster ts- is rare in initial position. In intervocalic position it is well attested in OIA. In MIA it is assimilated to -cch-, e.g. vatsa-> vaccha-. But when the cluster ts arises in composition, it is assimilated to -SS-, e.g. ut-sarp-> ussappa- etc. Now the sandhi of +$ results in cch in OIA, but in MIA it gives ss as above. Cf. ut-srita > OIA ucchrita, MIA ussiya. This will indicate that the MIA development of -ss- antedates the sandhi product -cch- of OIA. It is possible to derive MIA ss from ts, but not from cch. ? Even in such cases we find cch, as in ut-sava > ussava and ucchava. 8 Pischel, Gr. Pkt. Sp., & 327, 327 a. For the development of Marathi s < Pkt. (c)ch, see J. Bloch, La Formation de la Langue Marathe, $ 103. The same treatment in some other NIA languages is noted by him in $ 102. Madhu Vidya/366

Page #392
notes and discussions THREE NOTES M. A. MEHENDALE Deccan College, Poona. (1) Neutralization of contrast This term has been used to designate cases of conditioned merger. If, of the two phonemes of the earlier stage, only one can occur in a given position and not the other, then this is interpreted as a merger of the two phonemes in that particular environment. In this situation the contrast between the phonemes is looked upon as having been neutralized. Penzl in his article on "The Evidence for Phonemic Changes" puts it this way : "thus a suspension of contrast ......, in the terminology of the Prague School a 'neutralization results in this position" (J. Whatmough Felicitation Vol. p. 195). Thus, for example, if in Marathi only s- can occur before -i, and not s-(e. g. t@sa : t@si ) then this conditioned merger of s and sis looked upon as neutralization of these two phonemes before i. It seems, however, that it is not proper to describe this situation as neutralization or suspension of contrast. It is really a case of the restriction on the occurrence of a phoneme in the phonemic pattern of a language. The term neutralization can properly be used in case there is a possibility of both the phonemes occurring in a given position -i. e. the pattern does not stop either the one or the other from occurring there - and yet this does not lead to the native speaker's considering them as different utterances. But if in a given pusition only one phoneme can occur, but not the other, the question of their neutralization does not arise at all. Neutralization is a term better suited for describing what is otherwise known as free variation of phonemes. In such cases there is always a possibility of either of the two phonemes occurring, and if in spite of this the two utterances with the two phonemes are judged to be the same, then we can justifiably say that the phonemic contrast in a given position has been neutralized. We use the term 'neutralization here because the two phonemes, in spite of their use, do not produce different utterances. In Kamarupi, a dialect of Assamese, kh and h are contrasting phonemes, cf, khat 'bed-stead': hat'hand'. But in the final position in the words rakh and rah this contrast is neutralized since both of them mean "juice'. Such examples showing free variation of phonemes may not occur often in languages; but the use of the term neutralization or suspension of contrast is properly applicable to them. Madhu Vidya/367

Page #393
M. A. MEHENDALE The term neutrlization, as is being currently used, is also otherwise misleading. It gives an impression that of the two phonemes, the contrast between which is said to have been neutralized in a given position, any one may occur. For example, if it is said that the contrast between the voiced and the unvoiced stops is neutralized in the word-final position, it implies that any one of the two may occur in that position. But this is not true. Actually only the unvoiced stops seem to occur in that position. Also consider following situation In Sanskrit, examples like maruta: marud-bhih show that the eontrast unvoiced-voiced is neutralized before voiced stops; and when we see examples like sarada: sarat-su we see that the same contrast is neutralized before unvoiced stops. These statements, therefore, in themselves will not be enough. We shall have to add that the terminal phoneme depends on the specified environment: voiced, if the following stop is voiced, unvoiced, if the following stop is unvoiced. (II) Analogy In his Presidential Address delivered at the Annual Meeting of the LSA on 28 December, 1964, Hockett (1965) observes that although linguists are generally agreed that languages change due to Borrowing, Analogy, and Sound Change they are not agreed on the point whether "the attested sound shifts... reflect the workings of a MECHANISM of a linguistic change distinct from and not reducible to analogy and borrowing." Hockett's own position on the above point is that: "There is a mechanism of linguistic change,...called SOUND CHANGE not to be confused with and not reducible to analogy and borrowing."(191) 85 It appears that it is easy to agree with Hockett. On the other hand, it is tempting to point out that although sound change cannot be reduced to analogy (or borrowing), it is possible to look upon ANALOGY itself as a mechanism which is similar to a kind of sound change viz. Assimilation. In assimilation two phonetically dissimilar sounds are partly or fully made similar to each other. In most cases the affecting and the affected sound occur in close proximity (parna panna), but distant assimilation is not unknown (iku>ucchu ). Assimilation across morphemes is also known (mus-na-ti>mus-na-ti, rsi-nam>rsi-nam)... But still the change has occurred within the limits of a 'word' (samanapada). In analogy what happens is that two dissimilar forms are made similar to each other with regard to some phonetic feature or features which, in some cases, tends to bring about uniformity in the employment of a grammatical feature. As is well known, Bloomfield gives this kind of change the form of a proportion: dog: dogs Cow: ? cows (in place of kine) Thus the dissimilarity in the derivation of the plural forms' dogs' and 'kine' has been set aside by analogy and the two forms dogs and cows are made Madhu Vidya/368

Page #394
86 M. A. MEHENDALE similar to each other as far as their plural endings are concerned. then the formations plural' has assimilated 'kine' to it. All analogies, however, are not proportional. In Sanskrit, for example, if *ekadasa has given place to ekadasa on the analogy of dvadasa no such proportion seems to be at work. But this change in the lexical item can also be looked upon as 'assimilation. Two forms *ekadasa and dvadasa which were dissi milar with regard to the vowel length in the syllable preceding -dasa are made similar to each other by analogy. In a way, again, 'dvadasa 'assimilates' *ekadasa to give rise to ekadasa. It is true that, strictly speaking, analogy is not sound change. But this is true to some extent also of assimilation in that assimilation is not a sound change in the sense that a phonemic shift, a split, or a merger is. Both analogy and assimilation affect the phonemic shape of a morpheme but do not seem to affect the phonemic stock of a language. But whereas assimilation occurs within a 'word', analogy works across them. (III) Regularity of Phonetic Changes Bloomfield looks upon sound change as reflecting a change in the speaker's manner of articulation. He therefore is of the opinion that a sound change 'affects a phoneme or a type of phonemes either universally or under certain strictly phonetic conditions, and is neither favoured nor impeded by the semantic character of the forms which happen to contain the phoneme" (1933: 364), This view has been endorsed by Hockett (1965: 190-191) while saying that in a statement of the following type, Parx Dau y (in the environment z) the environment must be a sound or a finite combination of sounds in the parent language. The above view has been challenged by (Robert D. King (Historical Linguistics and Generative Grammar p. 119 ff). He finds that there are certain types of sound change exceptions to which cannot be explained in strictly phonetic environments and hence the hypothesis regarding the regularity of phonetic changes should be modified and stated as follows: Phonological change is regular, but its environment cannot always be stated in strictly phonetic terms. The examples which King gives to refute the Bloomfieldean view of sound change are as follows: (1) The Middle High German e [a] in the word final unaccented position is lost in Standard Yiddish. E. g. tage> teg 'days', erdeerd 'earth'. But in some cases, especially when e is an adjective inflectional ending, it is not lost: di groyse shtot' the big city', dos alte land "the old country. The environments, in the latter examples, are not phonetic but morpho logical (retention of word final -e as an adjective ending). (2) The sequence [kw] from proto-Mohawk to Mohawk undergoes epenthesis: "kwistos> kewistos 'I am cold', But when the kw sequence arises out Madhu Vidya/369

Page #395
THREE NOTES 87 of the coming together of the first person marker and the plural morpheme, no epenthetic e occurs : yakwaks' we several exclusive eat it'. Or, to take a different type, no epenthesis occurs in rakwas 'he picks it up'. King himself observes that it would be possible to explain away the first irregularity by assuming a plus-juncture before the adjective ending and say that the final a disappears in Yiddish except after plus-juncture. But he disfavours any such attempt. In his opinion that would be a gimmick and not a real solution. As for the second case, at least in respect of rakwas, he observes that the underlying form of the cluster kw in that word is / kol. But he says that this does not help us as the rule converting /ko/ to / kw / is older than the rule of epenthesis and hence even in this case kw should have given rise to kew. Now exceptions to the statements of phonetic changes have been observed since the very start of comparative studies. Acceptance of the regularity hypothesis has compelled scholars to look closely at the exceptions and this has led to many good explanations. In the opinion of Verner, who was responsible for one such brilliant explanation, it is the task of the linguists to search for the rule for the irreguarity. The regularity hypothesis has definitely proved its usefulness and hence it is not advisable to discard it or modify it as suggested by King. Rather, it would be better to say that in the Yiddish and Mohawk examples cited above the phonetic conditioning responsible for the exception has been lost to us. It may have been some kind of juncture, a difference in pronunciation, or stress. Some day it may be possible for us to discover it. But no attempt will be made in this direction if the chapter is considered closed by accepting the modified vers. ion of the regularity hypothesis. Verner, for example, could never have explained satisfactorily exceptions to Grimm's Law if Vedic Sanskrit with its accent was not available to him. If he had merely forms from classical Sanskrit, which does not mark accent, Skt. pitar : Gmc, fadar would have continued to remain a probem like the examples cited by King. REFERENCE : Bloomfield, Leonard 1933. Language, New York : Holt. Hockett, Charles F. 1965. Sound change. Language 41. 185ff. [ Received 21 February, 1972). Madhu Vidya/370

Page #396
NOTES ON INTERNAL RECONSTRUCTION AND COMPARATIVE METHOD M. A. MEHENDALE (1) Reconstruction of Skt. s. TF we observe an alternation between two phonemes in a given 1 paradigm, we have reason to believe that the alternation is the result of some phonetic change in the history of that para. digm. In favourable cases it is possible for us to recover the phonetic change and say which of the two phonemes has, in all Kikelihood, undergone the change. To take an example, Skt. has nom.sg. vak, but instr. sg. vaca showing an alternation between k and c. We believe that we get these two sounds in two positions, k in the word final position and c before a vowel, on account of some phonetic change. Now is it possible for us to recover the phonetic change -- i.e. is it possible for us to say whether in the history of this paradigm vac, vaca became vak, vcca or vak, taka became vak, vaca ? Or, to put the question differently, did c>k in the word final position to give us nom. sg. vak or did k > c before a to give us instr, sg. vaca ? Here, the rule of internal reconstruction helps us to determine that it is the c which became k in the wordfinal position giving us vak. These two phonemes in this paradigm, then, are traceable to a single phoneme of the proto stage : cand k here go back to c. The situation becomes a little complicated when a phoneme alternates with two different phonemes in identical positions in two paradigms. To give an example, we have in Sanskrit an alternation between k and j in the paradigm asrk, asrj-a, giving k in the word-final position and j before a. But in the case of another paradigm we see the alternation between t and ja, as in srt, srja. Here, in the same environments as above j alternates with t and not k. A type of this double alternation should lead us to suspect that what we now have in Sanskrit as a single phoneme j is in fact a merger of two different phonemes of the earlier stage. If it was only one phoneme, it should not have led to two different alternations. We therefore conclude that the prototypes of Skt. j were two phonemes jl and j?, whatever the phonetic values of those phonemes in the proto stage may have been. One of these two, Madhu Vidya/371

Page #397
INTERNAL RECONSTRUCTION AND COMPARATIVE METHOD 287 say jd developed allophones in the word final position which were similar to k and ultimately merged with it thus giving us aszk, while ja in the same position developed allophones which were similar to l and hence merged with that phoneme to give us sgt. In the position before the vowel, both of them merged to give a simgle sound j, and hence we have astja and srja. It is customary to denote these two reconstructed sounds as IE W and g(labiovelar and palatal). This kind of non-unique alternation is not restricted to j alone in Sanskrit. The sound $ also shows such double alteration. For example, the sof dis alternates with k, hence dik, disa, but s of vis alternates with t in indentical environments, hence vit, visa. Applying the argument of the above case of twofold alternation of j, we should say that Sanskrit s also represents a merger of two proto-phonemes, whatever might have been the phonetic values of these reconstructed phonemes. We may represent them provisionally as k'i and k'. As in the above case, we would say that k'' developed allophones in the word final position which were similor to k and hence merged with it to give us dik, while k'in the same position developed allophones which were similar to t and hence merged with it to give us vit. Before vowel, however, both of them merged into an identical sound & to give us disa and visa. However, we know that the comparativists have reconstructed only one sound to account for Sanskrit &, and it is represented as k. It is suggested that the comparative method here should be corrected by internal reconstruction and we should assume a twofold origin for Sanskrit s. We may now ask, do we have any traces of the two-fold origin of Sanskrit & in some other language? The answer to the question, it seems, can be in the affirmative. A Sanskrit & corresponds normally to Avestan & (Skt. sapha = Av. safa, Skt. pasu = Av. pasu etc.). But there are a few words where Skt. $ does not correspond to s in. Avesta but to the dental un voiced fricative th' Thus we have Skt. Vsam = Av. tham, Skt. sura =Av. thura (besides Av. sura), Skt. Voi = Av. V this in aiwithyo 'over-sleeping', and Avestan sakhta and thakhta from Vsac for which there is no Skt. Vlac. Normally we would have set aside these few cases as residual forms which cannot be explained. But since we are led by internal reconstruction to a two-fold IE origin for Skt. &, we may well say that these few aberrant cases in Avesta also point to a two-fold origin for Avestan and th' viz. IE kl and k's. We conclude by saying that IE k" became sin Avestan and thus merged Madhu Vidya/372

Page #398
238 with the continuation of IE s, while IE k2 became th, thus merging with Avestan th' which has come from a different origin (e.g. githa, haithya). This may be represented as : JOURNAL OF THE UNIVERSITY OF BOMBAY IE k'1 k'2 Skt. s (alternating with k) s (alternating with t) Avestan (2) Morphological correspondence, a sure indication of genetic relationship. 8 th... Languages are usually said to belong to a single family if they show some similarities. A family of languages implies genetic relationship, i.e. the different languages comprising the family are looked upon as later forms, developed in course of history, of a single, more or less homogeneous, proto language. The similarities or correspondences between languages could be seen in lexical items or in grammar. Usually languages are grouped into families if they show similarities in sound and meaning in a large number of vocabulary items. But scholars have often objected to this procedure. They point out that if Skt. dvs 'sheep' corresponds to Gk. owis, this is looked upon as a sign of genetic relationship. But if Skt. pippali corresponds to Gk. peperi, this is considered as a case of borrowing and hence not indicative of genetic relationship. What is then the difference between the similarities in the Sanskrit and Greek words for sheep and pepepr? The one is no more an item of basic vocabulary than the other. Still, most scholars are agreed that Sk. avis: Gk. owis are cognates, while Sk. pippari: Gk. peperi are not. In the opinion of some linguists like Kretschmer (Einleitung in die Geschichte der grie. chischen Sprache, 1896, p. 18) this is not correct. They feel that this similarity gives us no ground to reconstruct a work like *ovis for IE. It could as well be a loan word from some language outside IE which spread from one language to another in the whole field of IE after the individual languages of this family separated from one another. They believe that the difference between the similarities in the word for sheep' and pepper' are only chronological the former is a case of pre-historical borrowing while the latter is of historical borrowing, the one occurred very early in the history of Sanskrit and Greek after they separated from each other, while the other occurred rather late (Kretschmer o.c. 22, Trubetakoy Gedanken uber das Indogermanenproblem 82). Madhu Vidya/373

Page #399
INTERNAL RECONSTRUCTION AND COMPARATIVE METHOD 239 This difference in the attitude of scholars of looking at words for 'sheep' and 'pepper' has led them to formulate two different hypotheses, viz. the 'assimilation hypothesis and the disintegrattion hypothesis', to explain the similarities that are seen in the languages known to belong to IE family. Those who believe in the assimilation hypothesis believe that the similarities are due to borrowing from one language to another, the languages having thus become assimilated' to one another, while those who believe in the latter feel that the similarities are due to the fact that they had a common origin in the reconstructed IE which in course of time becamc disintegrated into Skt, Gk, Latin etc. Now it is well known that although similarities in words between diffcrent languages are usually taken hold of to formulate the hypothesis of genetic relationship, it is the grammatical similarities which offer more convincing evidence of this kind of relationship. And it has also been pointed out that similarities in grammatical 'irregularities', rather than in grammatical 'regulari. ties', are even stronger in proving genetic relationship. Thieme (Die Hypothese einer einheitlichen indogermanischen Grundsprache which, incidentally, has been described as a 'must' for a student of historical linguistics) uses this very ground in pointing out convincingly that while pippari is a case of borrowing, avi is not. He refers to the morphological peculiarity of the avi declension. It does not form its genitive, as in the case of a large majority of -i stems, with -es, but has the form avyas for which there is in Sanskrit a parallel only in one stem, viz. aris: aryas. Now the Skt: Gk. correspondence for the word for sheep is not restricted to the nom. sg. alone, but shows itself also in the unusual Gen. sg. form Sk. avyas: Gk. awyos. Thus Skt. and Gk. show a correspondence in the peculiar method of its inflexion, and it is this similarity in its unusual grainmatical form that celeraly indicates that it is not a case of a pure lexical borrowing, but that it is a case of inheritance. The Skt. and Gk. forms unmistakably point to a common IE gen. sg. form *dyos. Similar examples of correspondences in grammatical peculiarities pointing to common origin, and not borrowing, can be given from Sanskrit and Avesta. In Skt. the stems ending in -? show a lengthened grade (vrddhi) in the first five forms e.g. data, datarau, dutarah, dataram, datarau. But words ending in and showing human relationship like pit? (but not naptr and svasr) show only guna forins for these cases (except the nom.s.g.), e.g. pitarau, pitarah, Madhu Vidya/374

Page #400
240 " JOURNAL OF THE UNIVERSITY OF BOMBAY pitaram, pitarau. Now in Avesta we have exactly the same situation. The stems ending in s have forms like data, datarem, dataro, but for pits we have forms like pitaram, pitaro. We can take one more example. In Sanskrit and Avesta, the cardinal numbers show gender distinction only for the first four numbers viz. eka, dva, tri and catvar. To take the forms only of the word for 'four' we have in Skt. calvarah (m.), catasrah (f.) and catvari (n.) to which Avesta corresponds for the first two with cathwaro (m.), catanro(f.). Such close resemblances in morphological peculiarities can neither be explained as due to chance or borrowing. They have their basis only in common origin. When an item is borrowed from one language to another, one does not borrow it with all its morphological peculiarities. If today & word like sputnik is borrowed from Russian, no-body borrows with it all the declined forms of that word in Russian. And we have no ground to assume that the linguistic developments which took place in historic or prehistoric times were quite different from, or even opposed to, what we find today. 1. W.D. Whitney : Language and the Study of Language (1867), "So far back as we can trace the history of language, the forces which have been efficient in prodlucing its changes, and the general outlines of their modes of operation, have been the same; and we are justified in concluding, we are even compelled to infer, that they have been the same from the outset." (p. 258): Madhu Vidya/375

Page #401
ON THE VARTTIKA 2 ON PANINI 6.1.83 By M. A. Mehendale In the sutra bhayyapravayye ca cchandasi (6.1.83) Panini lists two Vedic irregular forms viz. bhayya to be feared' (from bhi- ). and pravayya' to be impregnated' (from pra-vi-). The first Varttika on this sutra reads as bhayyadiprakarane hradayya1 upasankhyanam and with this Katyayana adds one more form to the list of irregularities." The meaning of the form hradayyah (pl.) is " existing in, being in, pond' (hrade bhavah) and since Katyayana uses in his Varttika the fem. form it may be gathered that only that was known to him. The example cited by Patanjali for its usage is hradayya apah. The second Varttika on the above sutra runs as av farasya ca and from Patanjali's commentary on it we learn that it accounts for two more forms viz. hradavya and saravya. The meaning of hradavya is the same as of hradayya and the example given in the Mahabhasya is also similar viz. hradavya apah. As for saravya, Patanjali seems to give two different examples - saravya vai tejanam and saravyasya pasun abhighatukah syat. What is the meaning of these two examples? The commentators have not explained them. From their form they seem to be citations from some Brahmana text now lost to us. The first example does not appear to be difficult. Tejanam means 'the shaft of the arrow' and faravya is well attested with the meaning 'arrow' or 'volley of arrows'." Hence the citation seems to identify, as is done usually in the Brahmana texts with the particle vai, the shaft of the arrow with the arrow itself. Or, if tejanam here stands for sugandhitejanam TS 6.2.8.4 etc. (sugandhitejanam MS 3.8.5), which is a kind of grass, then it may be said that in the Mahabhasya passage this grass is identified with the arrow. The basis of identification may be sharpness. Madhu Vidya/376

Page #402
M. A. Mehendale The meaning of the second example does not become immediately clear. We find in the Brahmana texts that when for a certain ritualistic detail two alternatives are available, the Brahmana text shows its preference for the one and rejects the other. In order to ensure that the recommended practice is, in fact, followed the text sometimes goes on to threaten that if the sacrificer follows the practice not approved by the Brahmana, Rudra will be inclined to injure his cattle. But this will not happen if he follows the recommended practice. This is expressed in such words as rudro 'sya pasun abhimanukah syat (if the sacrificer follows the rejected practice) and aghatuko 'sya pasupatih pasun bhavati (if he follows the recommended practice) MS I. 6.4.8 250 In the light of the passages cited above, it is possible to suggest that the Mahabhasya example saravyasya pasun abhighatukah spat is a line from some Brahmana text warning the sacrificer that Rudra would injure his cattle if he followed some practice not approved by the text in question. It is true that the word Rudra does not occur in the text. But this is not strange since the tendency to avoid a direct mention of Rudra is discerned in the Brahmanas.10 But there still remains a difficulty about the interpretation of the word saravyasya. Since in the Brahmana passages cited above we get such expression as asya pasun where asya refers to the sacrificer, Saravyasya pasun may similarly be interpreted and saravya may be taken as an adjective of the sacrificer. In that case saravya cannot have the meaning of saravya (arrow') of the first example. Now in the Nirukta sometimes the secondary derivatives with the suffix -ya or -iya are rendered with sampadin, e. g. yajniyanam = yajnasampadinam (7.27), somyasah somasampadinah (11.19). Similarly Saravya may be interpreted as sarasampadin 'one who procures reed. ' The Mahabhasya passage then would seem to say that Rudra is inclined to injure the cattle of one who procures reed (for some sacrificial purpose). = But it has to be noted that barring one occurrence in the Paipp. 9.10.8 faravyam visam we do not find any other adjectival use of this derivative from sara. What we really obtain is a fem. base saravya used as a noun. And it occurs also in connection with Rudra. Thus we get siva saravya ya tava taya no rudra mrdaya TS 4.5.1.1 Madhu Vidya/377

Page #403
On the Varttika 2 on Papini 6.1.83 251 (with minor variants Katha. 17.11, Katha 27.1., Paipp. 14.2.7, 1.95.2) or rudro va esa yad agnis tasya tisrah saravyah TS 5.5.7.2. saravya even occurs as a personified deity kvasysta para pata saravye brahmasamsite RV 6.75.16. Hence it is possible to suggest two emendations in the Mahabhasya passage and read it as saravyasya ( - saravya asya ) pasun abhighatuka syat '(Rudra's ) arrow (or the Goddess saravya ) is likely to injure his cattle'. The first misreading - saravyasya for saravyasya - is easy to understand and the second -- abhighatukah for abhighatuka - could have occurred under the influence of the occurrence of quite a number of mas. nom. sg. forms of derivatives with the suffix -uka. This suggestion has the advantage that we get the word asya his' to refer to the sacrificer which is found in the sentences of this type. In the end it may be pointed out that in the third Varttika on the sutra Katyayana points out that saravya can be derived also from saru which has the same meaning as sara. In the following two Varttikas he gives two examples to show the use of saru viz. rnjati saruh which occurs in the RV1.172.2 and saruhasta which occurs in the usage of the people (loke ). If derived from saru, baravya becomes a regular formation and it is not necessary to formulate a separate rule for it (as one has to if derived from sara ). It is probable that for this reason the Varttika av sarasya ca and those following it are not treated in the texts like the Kasika and the Siddhantakaumudi. NOTES : 1. Actually one expects hradayyaya upasankhyanam. In the Kasika the Varttika reads as hradayya apa upasankhyanam and in the Siddhantakau. as hradaya upasarkhyanam. 2. The two words listed by Panini arc krt forms while those added by Katyayana in the first two Varttikas on this satra are taddhita forms. 3. The word hradayya is attested only once in the TS 4.5.9.1. The form available is mas, and not fem. Cf. nd no hradayya ya. Apparently it is used as an adj. of Rudra. 4. The form hradi vya (with the accent different from that of hradayd ) is attested once each in the Katha. 17.15 and the Katha 27.5. Here also the form is mas. and not fem. cf. nomo hraddvydya, and is apparently used as an adj. of Rudra. 5. The two examples hradayya apah and hradavya apah given in the Mahabharya look as if they are. from daily usage. It is not clear whether Katyayana gives hradayya and hradawyd as Vedic forms or those current in the bh7$4. 6. Cf. the parts such as salya, anika, tejana, parna and snd van of an arrow mentioned in the Ait. Br 13.2 (also 4.8). 7. E. g. RV 6.75.16, 10.87.13 ; AV 5.18.9 ctc. etc. Madhu Vidya/378

Page #404
252 M. A. Mehendale 8. Occasionally even knowledge of a sacrificial detail is similarly praised : jo va agnihotrd sya daitda devi veda gha tuka enath pafupdtir bhfuaty a gha tuko'sya pafupitih pasin MS 1.8.5. It is noteworthy that in such sentences syat is used when the likelihood of some untoward happening is expre ssed and bhavati when the opposite is the case. 9. The word abhigha tuka does not seem to be otherwise attested. Expressions abhimt nuka and dghaluka are more common. Panini teaches the suffix -uka after certain roots which include han but not man (3.2.154 ). 10. E. g. the Ait, Br. 13.10 (-3.34 ) recommends to say rudriya instead of rudra cf. prajayemahi rudriya prajabhir iti bruyan na rudraty atasyaida na mah parihtiyal. Madhu Vidya/379

Page #405
ANALYSIS OF MEANING M. A. MEHENDALE Deccan College, Poona While studying a somewhat difficult text like the Rgveda one is required to look to a number of available translations and interpretations of single words and passages. While struggling through this mass of literature one discerns a tendency among Vedic philologists, that when they are confronted with difficult passages they are inclined to assume multiplicity either of words (homonymy) or of, meanings (polysemy). But with the advance made in the field of Vedic interpretation it has been observed that in some very important cases these assumptions are not necessary. The glaring example in this regard is that of the Vedic word ttd. Roth, Grassmann, and Geldner all thought it to be both adjective and substantive, and in each case. having more than one meaning.' But Luders has now very convincingly shown that in the Rgveda rid is never used as an adjective, but always a substantive, and that it has only one meaning' truth'. This may be either just simple truth in the usual sense of the word or a magically active cosmic power (Varuna II 405). It is quite true that one should not try to come out with a single meaning of a single word if that leads to strained interpretations. On the other hand it is hardly desirable to complicate the vocabulary of the Rgveda in order to arrive at a simple meaning of the hymns". As in other sciences' economy' may be accepted as a sound methodolo gical principle in philology. Our working assumption should be that there is but one and the same word in the different passages and that it has only one and not multiple meanings. This assumption, of course, should be modified when absolutely necessary. The above principle was, it seems, first formulated by A. Bergaigne as a reaction to, what he calls, the German school (1' ecole allemande ) of Vedic interpretation. This school was then chiefly represented by Roth and Grassmann and later found an important Madhu Vidya/380

Page #406
ANALYSIS OF MEANING follower in Geldner. Bergaigne's principle was to allow the words to have in all their occurrences their ordinary meaning ("a laisser aux mots leur sens ordinaire"). Renou in his Etud es Vediques et Panineennes I p. 3 upholds this principle when he says that it is our duty to try to restore in all passages 'the intial meaning' (" on devra. tenter de restituer partout le sens initial " p. 3) or 'the primary idea' ("l'idee premiere " p. 21 ) of a given word. Thieme in his review of Renou's book points out that the leading French Vedist unfortunately gives up this rewarding principle when, while writing on the word sta, he observes that it demonstrates how futile it is to attempt onesided translation of the essential terms of the RV ("Le mot yta demontre la vanite de toute traduction unilaterale des termes essentiels du RV" p. 22). Thieme comments on this statement to say that "it is vain to entertain any hope of stepping out of the circle of altogether subjective guesswork as long as we do not (as Renou ...... claims is our duty ) try to restore the initial meaning everywhere." Thieme further observes in his review that in so many cases scholars have been at a loss to find out " the initial meaning", the "acception authentique, linguistiquement valable" as distinguished from "values which are just underlying and figurative" (des valeurs...... simplement sous-jacentes et figuratives, Renou p. 11). Thieme suggests the following procedure : "The presumable" notion initiale", the central idea" for which we have to find a name in our language, has to be put to the test as to whether it is recognizable in all the passages of the RV where the word naming this notion occurs, and whether consequently, one single expression can be used when translating this one single word in all the different contexts" (emphasis mine) ?. This is an excellent methodological rule. Of course, it cannot, by its very nature, be applied to words occurring only once, or even twice; but in cases where a given word occurs more than once or twice, and if we are fortunate enough to notice that the contexts in which it occurs are different, the above method deserves to be tried out; not only this, it may be asserted that it is the only proper method to be followed. And if in its application we are able to come out unscathed with only one single word with one single meaning, in spite of the fact that the contexts are different, we are very likely to be nearer the truth. Madhu Vidya/381

Page #407
78 M. A. MEHENDALE Thieme continues (p. 55): "A word is defined by its formal, grammatical features and those traits of usage that are common to all the contexts in which it appears." With regard to the nature of the " penombre" of which Renou speaks (p. 26), Thieme observes (p. 56): "The " shadow " is a matter of the interpretation of the context...... The "shadow " must be taken care of not by the translation, but by a commentary or a commenting paraphrase attached to it." We may add to what Thieme says by pointing out that the context is not only responsible for the "shadow" of the object - here respresented by a given word, it also helps the object to appear in better light and look clear. A good example of the rule: "Let the context speak", was found in one of our recent group discussions. A meaning' dung' was established for sakan in the context gavam gosthad ekavimsatim sakany ahstya juhuyat ( MS 4. 2. 10). The same meaning was taken for granted for saka, in spite of the difference in accent, when it occurred in the MS 3. 9. 3: taksito va esa nagno yan sakam avasyaty anagnam evainam akah and the passage was taken to mean that an animal was here described throwing dung. But the difference in accent raised initial doubt whether both the words were the same or different. On examination of the context it was found that the second citation refers to the chiselled yapa which was looked upon as 'naked'. When the Adhvaryu throws down a saka before fixing the yupa he makes it anagna. A reference to the TS showed further that the saka of the MS was identical with the sakala' a piece of wood, splinter of the TS and thus had nothing to do with 'dung'. It is on this background of Thieme's review that I propose to discuss in the following pages three words: sipra, vanargu, and ekavit. sipra, formally, is a simple word, vanargu and ekavyt are complex - with this difference that in the case of vanar.gu it may be doubted whether the second member is a noun or a verb, while in the case of eka-vyt no such doubt exists. The following discussion shows that it is possible to come out with one meaning for sipra and vanargu but for ekavrt we have to have more than one. (1) spra This word has been variously interpreted. About its form also there are two opinions. Grassmann and BR give only sipra (f.), Madhu Vidya/382

Page #408
ANALYSIS OF MEANING while V. Henry (MSL 9.250), Geldner (on RV 1.101.10), and Debrunner (Alt. Gr. II 2 212 b N p. 329) consider the necessity of extracting a stem sipra n. (cf. sipravant). As regards the meaning of the word, Yaska gave a start by say. ing that it designates some part of the mouth when he said sipre han (jaws) nasike (nostrils) va Nir. 6. 17. Although this meant that Yaska himself was not sure about the meaning, and although his first alternative (hani) was clearly ruled out by the occurrence of sipre by the side of han at least in one Rgvedic passage (a te hani harivah sara sipre rahat somo na parvatasya prsthe 5.36.2), the guess of Yaska found favour with modern scholars for a quite a long time, Geldner (op. cit.) in fact asserts that in any case it designates a part of the mouth which is active while drinking Soma. That is why sipra appears specially in connection with Soma (sipra im Du. Neutr nach 10.96.9, im Pl. Fem.) bezeichnet jedenfalls einen Teil des Gesichts oder Mundes und zwar denjenigen, der bei dem Somatrinken vorzugsweise beteiligt ist. Darum erscheint sipra besonders in Verbindung mit dem Soma von dem Somadurstigen oder Somatrinkenden ). Accordingly, Geldner feels that in some passages the meaning "raised nostrils which greedily take in the Soma smell (Die gehobenen Nustern, die gierig den Somaduft einziehen) would suit well. In other passages the meaning is, opened lips through which the teeth could be seen' and the rows of teeth themselves' especially in connection with hiranyayih and hiranyasipra, harisipra, hirisiprd, and dyaksipra (die geoffneten Lippen, bei denen die Zahne sichtbar werden und die Zahnreihen selbst). In one passage where we get vi syasva sipre, Geldner feels better to interpret it as "the sets of teeth of the mares" ( das Gebiss den Falben ). 79 The lexicons, of course, note these meanings (with one more alternative cheeks' e.g. BR - Backe, as far as the parts of the mouth are concerned.) But they also give the meaning the visors of a helmet (BR- Backenstuck am Helm) as an additional meaning. In the footnote 3 on p. 131 of Vol. I of his translation, Geldner, however, says that the meaning helmet or visor is to be given up since the Indians in ancient times did not know anything of that type (Die Bedeutung Helm' oder 'Visier 'am Helm ist aufzugeben. So etwas kannten die alten Inder nicht). We have to take note of one more meaning which also concerns the mouth or face. Charpentier takes a clue from Geldner's earlier Madhu Vidya/383

Page #409
80 M. A. MEHENDALE translations where in a few passages he translated sipra as 'beard, moustache (Bart, Schnurrbart ), and suggests that the word originally meant only 'hair' and from it developed partly the meaning 'beard-hair, beard, moustache and partly hair on the head, false hair or wig.' (Es scheint mir ganz offenfar, dass nur mit einer ursprunglichen Bedeutung Haar, woraus teils "Barthaar, Bart, Schnurrbart ", teils "Kopfhaar, Haaraufsatz" sich entwickelt hat, durchzukommen ist (p. 31 ). We have thus a choice: lips, jaws, sets of teeth, cheeks, nostrils, beard or moustache as parts of the mouth, and from it somehow the derived meaning helmet or visor. Apparently, scholars do not regard these two as meanings of two different words (1) sipra lips etc. ' ( 2 ) sipra 'helmet'. I for one have the feeling that it is not at all necessary for us to look to some part of the face as the meaning of sipra when it occurs in the context of Soma drinking. It is clear that anyone who drinks Soma has to use his lips or move his jaws etc. and this would hardly require a specific mention. Is it, for example, necessary to say that Indra drinks Soma with open lips? And as for the jaws, or even the sets of teeth, they may have hardly any specific role to play in the drinking of Soma. In fact while commenting on the word visi priya in the TS 1. 17. 12. 2 Sayana observes that since the Soma is well pressed, the jaws have no function to perform while drinking it (vigatam sipriyam (= movement of jaws sipre = hanu; tatra bhavam karma) yesam tel atra hanuvyaparabhavat sobhanabhisavasamskrtali ity arthah). Moreover, it has been suggested that sipra is to be interpreted as a part of the mouth' where Soma drinking is concerned, and as 'helmet' when a war-like activity is the context. But there is hardly a passage in the RV where sipra occurs in connection with Soma drinking but where some war-like attribute or action is not referred to. To give only one example ayam yan puro vibhinatty ojasa mandanah sipry andhasah 8.33.7 ' this is the one who breaks the forts with force, the sipri who gladdens himself with the juice'. The point is, whether, in such contexts, where drinking of Soma is mentioned, it is justified to relate sipra with some part of the body and translate with Geldner "wenn er mit geoffneten Lippen sich am Saft berauscht." The meaning of sipra which refers to some part of the face thus loses any justification. On the other hand, the other meaning Madhu Vidya/384

Page #410
ANALYSIS OF MEANING 81 'helmlet', which Geldner rejects, is the only one which suits best in certain contexts. E.g. it is said of the Maruts : amsesu va istayah patsu khadayo vaksahsu ukma maruto rathe subhah lagnibhrajaso vidyuto gabhastyoh siprah sirsasu vitata hiranyayih Il RV 5.54.11." Since the Maruts are described as carrying different things on different parts of their body, we have here quite clearly to interpret sipra as something put on their heads 13 --and this is more likely to be a helmet, of whatever sort, than the usnisa as conceived by Sayana. Since the meaning 'helmet'has been made certain by the above passage, it is necessary to examine if this meaning alone suits all the occurrences of the word and also its compounds and derivatives. It is found that this is quite possible and hence the word has to be assigned only one meaning. Thus the words siprin, susi pra or sipravan can be well understood with this meaning, especially when we note that in many of the passages where they are used as attributes of Indra, his other war-equipment vajra is mentioned, or some heroic performance is stressed. As regards the fem. form siprini I take it as referring to the sena, perhaps of the Maruts, which may also be looked upon as equipped with the helmets. In the line asmakan siprininam so mapah somapavnam sakhe vajrint sakhinam, I would like to supply siprin (voc.) as referring to Indra after siprininam '(oh helmeted Indra) of our helmeted ( army)'. In the passage vanoti si prabhyas siprinivan, I interpret the adj. siprinivan as referring to Indra who has a helmeted army. This avoids the necessity of looking upon siprinivan as consisting of one pleonastic possessive suffix (Geldner on 1.30.11). Among the dual forms of sipra, it seems to me that in the expression vi syasva sipre (1. 101. 10) it clearly refers to the mares, yoked to Indra's chariot, who wore on the head something for decoration or for protection. The same interpretation is possible in 3.32.1 and 10.96.9. In 5.36.2 a te hangi harivah sara sipre ruhat so mo na parvatasya prsthe it is not necessary to take spre as a dual form. The loc. prsthe allows us also to take spre as loc. sg. of sipra. The line means ' may (the Soma drink ) rise to your jaws, oh heroic one, having horses, on your helmet, as the Soma plant does on the back of the mountain.' Apparently what is implied is the golden colour of Madhu Vidya/385

Page #411
82 M. A. MEHENDALE Soma becoming visible on the jaws and the colour of the helmet also being said to be due to Soma. There is only one passage, RV 8.76.10, which would lead us to believe that sipra probably consisted of two parts: utristhunn ojasa saha pitot sipre avepayah 'Getting up with vehemence, after having drunk (Soma ), you have shaken the two parts of the helmet.' I now find that the same view as above has been expressed long back by V. Henry in Mem. de la Soc. de Ling. 9.249-252 (1896) and Henry says that the same suggestion was earlier made by Bergaigne in the same journal 8.25, note 13 (1894). Bergaigne looks upon sipra as " Les deux pieces d'une sorte de casque". He adds "Indra detache ses deux sapra pour boire le soma (1.101.10; cf. 3.32.1; 8.65. 10; 10.96.9 ), qui monte a ces sipra en meme temps qu'a. ses machoires (5.36.2). Il recoit l'epithete siprin, dont le feminin siprini ( 1.30. 11 ) parait designer une armee d'hommes qui portent les sipra. Indra, avec ses deux sipra, vaut (a lui seul) une pareille armee : 10.105.5." The compounds dyahsipra, hirangasipra, harisipra and hirisipra, can be easily understood with the meaning 'helmet'. The compounds dasasipra, vrsasipra and visisiprd are proper names. The word visipriya probably contains the form sipriya to be derived from sipra. This is a difficult word. sipra, the name of a river, sipra, the name of a lake, sipra mother of pearl and sipra a basket' are of course different words. [sipra is derived from the root *sip' to wag' cf. sepa' tail'in the Alt. Gr. Il 2 687 p. 858. Henry connects it with Lat. caput (MSL 9.250). ). (2) vanargu According to the lexicons, vanargu is attested only four timestwice in the RV and once each in the AV and the SV (vanargu). The complex expression can be analysed easily as vanar-gu and there is little doubt that the first member of the compound is the same as vana-or vanas- and means 'forest'. The latter member, on the other hand, can theoretically represent either the substantive go 'bull', or the verb Iga 'to go' ga 'to sing'. Accordingly there are three possibilities of translating the word :-(1) forest bull; (2) wandering in forest; (3) singing in forest. Madhu Vidya/386

Page #412
ANALYSIS OF MEANING In RV 1.145.5 and 10.4.6 Bergaigne translated the expression as 'bull of the forest' (taureau des bois). 1 RV 1.145.5 is addressed to Agni and runs as sa irit mygo apyo vanargur upa tvacy upamasyam ni dhayi/ vy obravid vayuna martyebhyo 'gnir vidva rtacid dhi salydk // RV 10.4.6 occurs also in a Agni hymn, and the context of the first line is that of the churning of the fire. tanutyajeva taskara vanargu rasanabhir dasabhir abhyadhitam / iyam to agne ndoyasi manisa yuksva rathai na scayadbhir angaih // AV 4.36, where the word occurs in the seventh verse, is intended to drive away all evil beings. There Whitney translates it as 'savages', although with a question mark. In the Nighantu 3.24 vanargu appears, by the side of taskara, among the stenanamani. Whitney obviously has the derivation of -gu- from ga 'to go in mind. The verse runs as - nd pisacaih sam saknomi na stenair na vanargubhih / pisacas tasman nasyanti yam aham gramam avise // SV 6.4.9 is addressed to Indra. Since the verse occurs in the Aranyakasamhita of the SV, I was at one time inclined to interpret vanargu in this verse as' singing in the forest.' The verse runs as hari ta indra smastuny uto te haritau hari/ tam tva stuvanti kavayah parusa so vanargavah // All the three possibilties have thus been exhausted and in fact such that the three possible meanings have been distributed over the three Samhitas. The question arises- is this really necessary ? Is it not possible to get on without assuming the multiplicity of meanings ? Let us examine all the three meanings: (1) singing in the forest' is clearly excluded for the RV and the AV occurrences. It can suit only the SV Passage. (2) 'wander ing in a forest'" does, in fact, appear to suit all the four passages but the following considerations go against it : (1) In the RV 1.145.5 Agni is called apyah mpgah 'water animal and also vanarguss. Ths parallelism shows that corresponding to mrgah, gy-in vanargu is better interpreted as something concrete and 'bull' Madhu Vidya/387

Page #413
84 M. A. MEHENDALE would fit in the context very well.* vanar.gu, thus, expresses in a compound what apyah mygah does in two separate words. (2) In the RV 10.4.6 the adjective vanargu' wandering in forest' used with ta skara would be superfluous." (3) In the AV 4.36.7, since vanargu occurs after pisaca and stend, it should better be interpreted as a substantive referring to a specific class of beings, harmful to villages, and not as an adjective. After pisaca, which refers to a class of non-substantial beings, and stend, which refers to human beings, it would be quite appropriate to assume that vanargu refers to a class of animals, 'forest-bulls'. (4) In the SV 6.4.9 vanargavah, if taken as an adjective, would qualify kavayah. Now it is perhaps possible to think of the kavis going to a forest to praise Indra, may be in conformity with some kind of taboo, but is it on that account necessary to think of them as 'wandering in the forest? On the other hand, if we look to the adjective parusasah occurring by its side and remember that in the RV 6.56.3 we come across the expression paruse gavi and in the RV 5.27.5 parusah uksanah, we will have little hesitation in interpret. ing vanargavah as 'forest bulls'. The kavis who were offering praises to Indra are described as 'rough' or 'dust-coloured 'forest bulls.90 It is thus quite possible to get along satisfactorily with only one meaning of vanargu'forest-bull' and it is not necessary toj assume two or three meanings. In the SV, vanergu is accented differently from the RV and the AV vanargu. But it is extremely doubtful whether it means something different on that account. Sayana's attempt to explain it as a Bahuvribi compound - vananiyah sambhajaniyah sevaniya gavo yesan to vanargavah fails both semantically and formally ( the accent in that case would have been on the first syllable vanargavan ). It is simple to assume some kind of accent shift. The available lexicons have not noted the difference in accent. (3) ekavit The word ekavyt also is restricted to the Vedic literature and occurs there quite a few times. When the word occurs in forms other than the noni. sg., e g. ekavitam (acc. sg.) AV 13.4. 15, ekavrtah (nom. pl. ) AV 13. 4. 12, or ekavrta (instr. sg.) TS 5.2.3.7, the stem ekavyt is not in doubt. But when it occurs as ekavit (nom. sg.) Madhu Vidya/388

Page #414
ANALYSIS OF MEANING . 85 theoretically two possibilities present themselves : (1) nom. sg. of ekavit, and (2) nom. sg. of ekavidh. The available lexicons have accepted only the stem ekavit and assigned it the meaning' being one, simple'. But if we examine the contexts in which the word appears, we notice that it does not quite fit in all of them. We are forced to admit more than one word and more than one meaning. The stem ekavyt with the meaning 'single' is quite in order e. g. in the Sat. Br. 3. 6. 3. 14 where it occurs in connection with the spreading of the sacrificial grass. We read : grhitva prastaram ekavid barhih strnati" having taken the prastara, he spreads the altar-grass in a single layer" (Eggeling). The commentator on the Katy. Sr. S. 8.2.25 and 8.7.12 where the word occurs in a similar context also explains ekavrtekadhatu 'having one layer,' i. e. the grass is spread only once ( ekavaram ). But the same is not true when the word occurs again in the Sat. Br. 13. 2. 1. 5 in a different context. When the food offerings are offered to cardinal numbers like eka, dvi, tri, etc. the numbers are successively raised by one. The word ekavit occurs as an adjective of svarga while attempting to justify this procedure of raising the numbers by one. We read: ekasmai svaha dvabhyam svaha ...... ity anuparvam juhoti/ ...... ekottara juhoti / ekavid vai svargo lokah. Here we cannot accept Eggeling's translation "He performs oblations successively increasing by one, for single indeed, is heaven." This does not bring out the relationship between ekotlaratva of the offerings and ekavrt nature of the heaven. This can be undestood properly only if we interpret ekavit as nom. sg. of ekavidh 'rising by one' (ekena vardhate). The numbers are raised successively by one because the svarga loka also, when thought of as having three or seven layers, successively rises by one. The context thus forces us to admit one more word ekavidh rising by one' Even in the passages where the stem ekavit is quite clear, its analysis, and consequently its meaning, is not always quite certain. Although the difficulty has not been noted by the lexicons, it has been already mentioned in the Alt. Gr. II, 2 p. 43. It is mentioned there that sometimes it is difficult to decide whether the second part of the compound, -vit, is to be understood as the root noun vit from V urt 'to be, to exist' or as the root noun -vr-t from Vur 'to cover.' It is Madhu Vidya/389

Page #415
86 M. A. MEHENDALE noted that there is a difference of opinion in this regard between the authors of BR and Hertel" (Indoir. Qu. u, F. 6. 31 ). When in the AV it is said about the sun sa esa eka ekavid eka eva (13. 4. 12, 20) or ya etam devam ekavitam veda ( 13. 4. 15 ) one can be certain that in these passages ekavyt contains the verb Vurt and it means 'existing alone, single.' But when in the same hymn it is said about the gods sarve asmin deva ekavrto bhavanti ( 13. 4. 21 ) it seems more likely that ekavit contains the verb Vur ("All the gods in him become covered by one "i. e. by the sun ). Whitney 's trying to force the root V urt here and translate the line as "All the gods in him become single " does not recommend itself. Finally the word ekavit seems to contain the verb V vr also in the Maitr. S. 4 2. 13. The passage reads as: deva vai sarve sahantaruanto 'bhavams te sarve saha vyajayantal lad ekavyd asu yat samur. ttam / tad deva ittham cettham ca vyatyacarans tan mitravaruna acayatam "The gods, verily, all became pregnant together. They were all born together. 'That' lay collected together (concealed ) in a single cover. The gods passed by and beyond 'it' in this way and that. Mitra and Varuna (finally) observed it'. [It is not clear what it' stands for. The gods Mitra and Varuna are said to have made gau dvipadi from 'it'). The available lexicons do not list ekavidh. That there are no other forms available of this word like acc., instr. etc. does not in itself go against observing such a vocable in at least one occurrence of ekanei. . We are thus required to list two wordsekavyt and ekavidhand in the former case we are required to assign two meanings (1) existing alone, single, ( 2 ) having a single cover. This example will show that the principle of 'economy' has to be modified if the context requires it. NOTES 1. To cite only GELDNER'S Glossar one notes: "1.adj. recht, wahr. wirklich; wahrhaft, echt, zuverlassig, 2.n. das Rechte, Richtige, Wahre, Reale. (a) rechter Weg. (b) Richtigkeit, Regel: Frasya pathya - pathin der rechte-, richtige, regelmassige Weg; aber auch der Gang des heiligen Werkes, der Ritus. (c) der regulare Gang der Welt, Weltordnung (-lauf), Kosmos; die alles regulierende Zeit. (d) dic Madhu Vidya/390

Page #416
ANALYSIS OF MEANING 87 reale Welt, Realitat. (e) Norm, Gesetz, Recht; stena mit Fug und Recht. (f) Gerechtigkeit, Rechtlichkeit, Redlichkeit. (8) Wahrheit, Aufrichtitgkeit, Wahrheitsliebe. (h) Eid. (i) der rechte Glaube, frommer Sinn, lauterer Gedanke und jede Betatigung des rechten Glaubens, frommer Brauch, rechtschaffner, heiliger Wandel, das Rechte, das rechte Tun, gutes Werk; jedes religiose Gesetz, (Gebot); insbesondere der regelmassige Gottesdienst, die richtige Opferzeit and Opferordnung, Ritus; Gebet und Opfer. (k) stasya sadas - sadana, sadman. (1) die Statte des Glaubens, Opferbaus, Opferstatte, (m) der unsichtbare Teil des Himmels (der spatere satyaloka)." 2. Bergaigne, La Religion Vedique (MM Roth et Grassmann ne craig nent pas, pour simplifier le sens des hymnes, de compliquer souvent le vocabulaire : j'essaie au contraire de retablir la simplicite dans le vocabulaire en admettant la complexite dans les idees (Vol. I pp. iv-v). 3. Etudes sur le Lexique du Rig-Veda I. 4. Op. cit. p. 105. 5. JAOS 77.51 ff. (1957). 6. Op. cit. p. 54. 7. JAOS 77.55. Thieme hastens to add "one single expression" does not mean "one single word". But the different words must center round one single idea ....." He, for example, finds it permissible to translate Sk. gada as 'sickaess, illness, disease'. 8. Geldner in his later translation has completely given up this meaning 9. KZ 46. 26-35 (1914). 10. In the Alt. Gr. the word is given different meanings at different places. Thus "Wange" II 1 p. 121, 294, "Backe" p. 299; "Visier" p. 296, 297 (for this latter, correction "Schnurrbart" occurs in the Nachtrage to II 1 p. 82; "Schnurrbart" II 2 p. 329, 407, 875. 11. To which Mahldhara and Uvata, who have the same explanation to offer as Sayana, add 'well purified' (suputas ca) while com menting on VS 9.4. 12. Also RV 8.7.25 : (of Maruts) vidyuddhasta abhidyavah si prah sirpan hiranyayih/ subhrd vyanjata sriyel 13. But Geldner : "an den Hauptern dehnen sich die goldenen Zahn reihen(?) aus." 14. Twice susipra. 15. MSL 8.482 16. Prof. Velankar Comm. Volume, 80. 17. The available lexicons have, in fact, only this meaning on record. MW, e.g. gives : "moving about in woods, wandering in a forest or wilderness, a savage, a thief-robber Nigh." 18. Agni is often called a bull or is compared with a bull in the RV. 19. For details of interpretation see my paper on vanargu, Velankar Comm. Vol. pp. 77-78. Suffice to say here that I interpret vanargu as acc. dual and translate the line as "(The two arms) have held fast (the two churning sticks with ten fingers) as two brigands, risking Madhu Vidya/391

Page #417
88 M. A. MEHENDALE their life, two forest bulls with ten ropes". This may be compareri with Geldner's translation : "Wie Zwei im Walde streifende Rauber, die ihr Leben einsetzen, haben (beide Arme) mit zehn Stricken (das Reibholz) festgebunden." This translation does not make clear what the robbers hold fast, and Geldner has to supply traveller' as also is done by Sayana (pathika), Durga (adhvaga). and Skanda Mahesvara (kascit). 20. For parusaso, there is a variant reading purusaso. But in that case there is the difficulty about the accent since purusa is accented on the first syllable. Moreover purusasah, by the side of kavayah, has hardly anything significant to convey. 21. Thus in the TS and the AV ekavrt means "einfach" according to BR (hence from V vst), but "einzig umgebend" according to Hertel (hence from Vvr). Madhu Vidya/392

Page #418
140 BRIEF COMMUNICATIONS ON TRCA Mr. Madhav Deshpande in his Brief Communication published in IIJ. 17, pp. 249250 (1975) says that in my discussion on the word trca I have ignored the condition chandasi in the Varttika rci trer uttarapadadilopas chandasi on Pan 6.1.37. But I have not. The condition is too obvious for any one to ignore. The absence of any rule in Panini's grammar to account for the form trea in place of the expected tryred, and the presence of a Varttika by Katyayana to account for it was interpreted by me as indicating a change in the pronunciation of the word in the Vedic texts, the earlier pronunciation being tryrca, the later trcd. My argument was that since Panini heard the pronunciation tryrca in and outside the Vedic texts, it offered no anomaly to him. Katyayana, however, heard the pronunciation tred in the Vedic texts and hence he found it necessary to take note of it. That I have not ignored the condition chandasi in the Varttika will be clear from what I have said on p. 5 of my Nirukta Notes (Series I, 1965): "... since Katyayana restricts the use of trca to Chandas..." Whether we may assume a change in the pronunciation of the Vedic texts. between the time of Panini and Katyayana and whether the piece of evidence based on this assumption is of any value for determining the relative chronology of Panini and Yaska is a different matter. What I wish to bring to the notice of the readers here is the fact that in my argument based on trca I have not ignored the condition. ing factor chandasi in the Varttika. Madhu Vidya/393 M. A. MEHENDALE

Page #419
ON PANINI 1. 3. 41 BY M. A. MEHENDALE Panini 1. 3. 41 reads veh padaviharane, i. e., the root kram with the prefix vi takes Atmanepada terminations if the meaning padaviharana is to be conveyed. What does padaviharana here mean? Patanjali does not comment on this sutra. The authors of the Kasika take the expression to mean 'taking a step, placing a foot-step" (padariksepe ), and give examples like susthu vikramate, sadhu vikramate (he) steps well'. They also observe that this placing of steps is to be understood with reference to the special gait of horses etc. (asradinam gativiseso vikramanam ucyate ) and, by implication, not with reference to the gait of human beings or the deities. It is, therefore, not surprising that we find the use of vi-kram in the Atmanepada with reference to the horses in the sisupalavadha 5.9 ( paryantavartmasu vicakramire mahasvah sailasya) and that Bhattoji Diksita in his Siddhantakaumudi adds the word vaji to the Kasika.example cited above ( sadhu vikramate vaji ). Apparently, nobody has thought of any other animal except horse', although the autbors of the Kasika have said asvadi. Since the authors of the Kasika understand padaviharana to mean padaviksepa, they raise the question : where is the necessity of the present sutra if the root kram, according to the Dhatupatha ( 1.502 ), has no other meaning but padavik sepa? Their reply is that the sutra is necessary because roots have many meanings. The force of their argument seems to be that if the root kram has various meanings, the sutra restricts its use in the Atmanepada only when the root with the prefix vi conveys the meaning placing of a step '; otherwise, as in vikramati sandhih 'the joint gets loosened, it is used in the Parasmaipada. The meaning assigned to padaviharana in the Kasika, however, does not agree with the attested usage. Viharati is well attested in the Vedic literature in the sense disjoin, separate", e.g., tau (i.e. grahau) punar viharatah" (the two priests, viz. the Adhvaryu and the Nestr ) separate again those two (cups, viz. the one filled with soma, and the other filled with sura)" (Sat. Br. 5.1.2. 1 Also S. M. Katre: Dictionary of Panini 2.364, 'placing of footsteps'. 2 Panini himself has used the word viharana in this sense in asyaviharana 'opening of the mouth, separating the jaws' in his sitra: ano do 'nasyaviharane 1.3.20. 29 (Annals BORII Madhu Vidya/394

Page #420
226 Annals BORI, LXIII (1982) 18). Not only this, viharati separates' is also attested specifically with reference to pada i.e. pada.In the Aitareya Brahmana ( 2.35=10.3) we read prathame pade viharati. This refers to the recitation of a rc of which the first two quarters are separated from each other, i.e. they are recited with a pause between them ( Sayana : viharanan prthakkaranam / dyayoh padayor madhye viharam vicchedam kltva pathet). This method of recitation is the opposite of the one in which the two quarters are united, i.e. they are recited without a pause between them ( samasyaty uttare pade Ait. Br. 2.35=10.3 on which Sayana : trtiyacaturthapadayor uttarardhagatayoh samyojangm ). On the basis of the above evidence it is possible to say that the expression padaviharana in the Panini-sutra refers to this peculiar method of recitation. If the Hotr priest recites the stanza by separating its quarters (padas) one would say hota vikramate. If, on the other hand, the word pada in the Paoini-sutra does not refer to the quarter of a stanza, but to a (human ) foot, we may compare the expression tasmat stry uru viharati 'therefore a woman separates her thighs' in the same section of the Ait. Br. referred to above. Even in the RV 10.162.4 we read yas ta uru viharati one who separates your thighs'. Hence, pada. viharana in the Panini-sutra may be taken to refer to the above context, and accordingly an example for the sutra could be strivikramate' a woman separates her feet.' Lastly, if, as our commentators think, padaviharana really refers to the placing of a step' (padaviksepa ), and not to the separation of the padas ( whether in the sense of a quarter of a stanza, or a buman ) foot) as suggested above, then we have the well-known examples of vi-kram in the Atmanepada with reference to the three steps taken by Visnu, e.g. (idam visnur vi cakrame tredha ni dadhe padam ( RV. 1.22.17 ), so'bravid indro yavad evayam visnus trir vikramate ( Ait. Br. 6.15=28. 7), etc. Thus visnur vikramate could be an example of the sutra in question. In view of the evidence given above it is not at all clear why the authors of the Kasika and the later commentators, in the first instance, did not take padaviharana to mean (1) separation of the quarters of a stanza', or (2) * separation of the (human ) feet', and, secondly, even if they took padaviharana to mean placing of a step', why they neglected (3) the famous three steps of Visnu', and chose to restrict the 'placing of the steps' to the horses and the like. 8 paedasabdasya paryayah padasabdah napumsakah (Sadgurusisya on Aitareya Aranyaka 1.3.7). * Kaus. Br. has pade vigrhnati. 6 On which Sayana : trir vikramate pudatrayai praksipati. Madhu Vidya/395

Page #421
PROF. MANFRED MAYRHOFER'S NEW ETYMOLOGICAL DICTIONARY OF SANSKRIT By M. A. MEHENDALE. Prof. Mayrhofer has already carned distinction as an eminent etymo. logist by the publication of his Kurzgefasstes etymologisches Worterbuch des Altindischen (A Concise Etymological Sanskrit Dictionary 1956-1976) in three impressive volumes. Mayrhofer was then conscious of the fact that his dictionary was not an end, but just a beginning in the direction of providing information on etymological questions related to Sanskrit. He had then hoped that some day someone's inspiration will be enkirdled by his book to make a second attempt. It is indeed gratifying to note that instead of waiting for someone else to take up the challenge, Mayrhoter himself bas come forward to embark on a second venture to publish a new etymological dictionary of Sanskrit. As compared with his earlier book, Mayrhofer has made two easily perceptible changes while giving the title of his book. He has dropped the restrictive epithet concise' of his old dictionary and he has chosen to characterize the language dealt with as altindoarisch' (Old Indo-Aryan ) instead of altindisch' (Old Indian ). There is one more change he has made : hc does not give to his new dictionary the alternative English title as he did to his old one. The change may appear small and not worth noting. But it gives as indication of the changes introduced by Mayrhofer inside the dictionary. In his old dictionary Mayrhofer gave, besides German translations of all head words, also their English translations for the convenience of English knowing readers. He has stopped doing this in his new dictionary since he found the practice space consuming." Mayrhofer's old dictionary was completed in 1976. His new dictionary, which is by no means just a new edition of the old one, started appear Etymologisches Worterbuch des Altindoarischen I. Band and II. Band, 1986-1996. Heidelberg, Carl Winter-Universitats Verlag, pp. 837. Vorwort to KEWA Vol. I, pp. IX-X. English translations of Sanskrit words can be of some use to English knowing readers. But their absence does not materially harm the new dictionary. A reader refers to Mayrhofer's dictionary for the etymology of a word, and not for its meaning. The only factor common to both the dictionaries is the author and the languago doalt with, 3 Madhu Vidya/396

Page #422
126 Annals BORI, LXXIX (1998) ing in 1986. In between he published a booklet and a small paper making public his views on an ideal and a practicable etymological dictionary of a large-corpus language. Mayrhofer was engaged on his dictionary work for nearly thirty years and was therefore justified in making knows his views on the forms of etymological dictionaries. His booklet Zur Gestaltung des etymologischen Worterbuches einer ,,Gross-Corpus-Sprache" appeared in 1980 and his paper, "Uberlegungen zu einem neuen etymologischen Worterbuch des Altindoarischen," appeared in 1983.6 Mayrhofer's views expressed in the two publications noted above, especially in the former, are summarized below so that the user of his new dictionary may know what information to expect under each lemma. In Gestaltung Mayrhofer first draws attention to the two aspects of an etymological dictionary: etymology - word history' and ' etymology-origin." In an etymological dictionary of the former type, the author concerns himself with such subjects as : Changes that occur in the meaning of a word, phonetic forms of a word, stylistic and social values that might get associated with a word in the course of its history within that language. In an etymological dictionary of the other type its author has to enquire about the origin of the word, the first combinations of phonetic sequences to which a definitive mcaping was assigned. An ideal etymological dictionary is one which at once takes care of both the above aspects : word history and origin. The realisation of such an ideal, however, Mayrhofer says, depends on the favourable stars of the person concerned ("personlischen Konstellationen"). Mayrhofer, apparently, did not enjoy this good fortune. In vain did he seek, before undertaking his new project, the association of a collaborator of the stature of the late Prof. L. Regou who could look after the "word history" aspect of the * It was ready in 1979 and, in its two abridged versions, was read before learned audi ences on two different occasions. I am obliged to Prof. Mayrhofer for making available to me both these publications. A question was once raised whether, in view of some unscientific excesses committed in the investigation of ctymology.origin', only the other aspect of the dictionary, viz. ctymology-word history' was to be looked upon as scientifically valid. The ques. tion was answered decisively by J. Untermann (Etymologie und Wortgeschichte, 1975. p. 10 ) by pointing out that etymology which in Gk. means instruction on the original meaning of a word' is a justified and necessary form of investigation. ! Gostaltung p. 12. Madhu Vidya/397

Page #423
MEHBNDALE : Mayrhofer's New Etymological Dictionary 127 dictionary, while Mayrhofer would be left to pursue the "origin" aspect. This meant that a "Renou " would describe all the vicissitudes through which a given OIA word has passed in the course of its history within. Sanskrit while Mayrhofer would tell his readers : 1. which OIA words reach back to deinoastrable earlier stages - Indoeuropcan or Indoiranian ; 2. which OTA words are revealed as new lado-Aryan formations, i. e. words not inherited from earlier stages but new forms within Indo-Aryan created from its own linguistic material and with its own rules of grammar; and 3. which OIA words are borrowings from other languages, either i) from MIA languages; or ii) from non-Aryan languages of India; or iii) from languages outside India. If Mayrbofer had succeeded in getting the cooperation he was looking for, he would have presented his new dictionary on the model of Dictionnaire etymologique de la langue Latin of Meillet and Ernout in which the former, an Indoeuropeanist, dealt with the origin-source of Latin words, while the latter, a Latinist, pursued their history within Latin. Since Mayrhofer did pot succeed, he has chosen to deal with only the etymology-origin aspect. This, howerer, does not mean that Mayrhofer did not attempt a philological examination of OIA words before writing his new dictionary. For, as he points out, however well-founded an etymology may be, it will break down if the lexeme, for which it is given, is assigned a meaning which is not borne out by its textual occurrences. A word, though attested, with a wrong meaning is as useless as a ghost word. One of the demands made on an etymologist by J. Untermann is that every linguistic form that is registered as a lemma in a synchronous dictionary of a language must also be entered as an independent lemma in the etymo. logical dictionary of that language. Mayrhofer's view, however, in this regard is that such a practice is not justifiable and is also space consuming. He makes this point clear with the help of an example.16 Both Gk, and Lat. Uberlegungen, D. 148. Etymologie und Wortgeschichte, 1975, p. 15. Mayrhofer tells us that this example is found also in Manu Loumana's Kleine Schri. ften, 1959, p. 187. 10 Madhu Vidya/398

Page #424
128 Annals BORI, LXXIX (1998) nouns : Gk cines. But the same so and Lat. ago have comparable verbal forms: Gk. d go drive, lead', Lat. ago ' drive, lead' which fact permits us to reconstruct their IE origin *ag-drive, lead'. We are therefore justified in giving independent entries Gk. ago and Lat. ago in their respective etymological dictionaries. But the same cannot be said for the comparable ageot nouns: Gk. a ktor leader', Lat. actor driver, actor'. These two words, of course, have a place as independent lemmata in their respective synchronous dictionaries. But they have no such claim in the etymological dictionaries of those two languages because the two agentive forms could arise, independently of each other, in Gk, and Lat. from their respective roots and agentive suffixes. We cannot therefore be sure, as in tbe case of *ag-, that the two forms Gk, aktor and Lat, actor reach back to a common IB reconstruct *eg-tor. Moreover, the requirement that every synchronous lexeme should be cited as an independent lemma in an etymological dictionary is not always practical. It can be fulfilled only in the case of a language like Gothic which has a relatively small vocabulary. Different languages have different kinds of text-traditions and hence there will always be different kinds of etymologi. cal dictionaries. Mayrhofer presents in this connection a four-fold classification of languages 1 1. Informant languages,' i. e, languages that are used as first languages in different communities and the vocabularies of which are always available, at least passively, to a member of that community and, to a large extent, also actively. 2. Corpus languages, i. e. languages which are not witnessed as possessions of speech communities as a whole but which are recorded in their literary works, inscriptions and similar other sources. These languages can be classified into three groups; i) Large-corpus languages: these are attested in writings to such a large exteat that a high percentage of their lexicon which was once available to their speakers, i. e. the informants of that bygone period, is known to us to day. To this group belong languages like Old Indo-Aryan, Greek, Latin. ii) Small-corpus languages; these are available in a small number of 11 Mayrhofer avoids using the metaphorical expression "living laaguages" because its opposite "dead languages" is not suitable for corpus languages like Latin and Sanskrit which are still alive" in special groups like priests, learned men, and poets, and a "dead" language can again come to life as in the case of Hebrew in modoro Israel, while a natural event like " death" is irreversible. Madhu Vidya/399

Page #425
MEHENDALE : Mayrhofer's New Etymological Dictionary 129 texts; but they yield material which is large enough to give us a fairly adequate insight into their structures and basic voabularies. However, we have always to bear in mind that what we possess today is only a fragment of its once large vocabulary. To this group belong languages like Gothic and Old Persian. iii) Rest languages : These cannot in all cases be clearly distinguished from the preceding group. These languages are preserved in small number of inscriptions, glosses, indirect reposts, etc. They are understandable and classifiable if they happen to belong to a language family which is well attes. ted. To this group belong languages like Thracian, Lydian and Phrygian which belong to IB family of languages. There are hardly any problems concerning the form of an etymological dictionary of a real rest'language. The normal type is a complete lexicon along with an index of all references; an edition of its handed down texts followed by the most clear statements on the origin of the demonstrable part of its lexicon. As a model, Mayrhofer names the Lydische Worterbuch of Roberto Gusmani (HeideMerg, 1964). As regards the etymological dictionaries of the small-corpus languages the model presented by S. Feist, Vergleichendes Worterbuch der Gotischen Sprache (3rd edn., Leiden, 1939), has been praised by competent authorities. Instead of enumerating its merits once again Mayrhofer draws attention to its defect, for that would immediately make clear what is and what is not dispensable while writing the dictionary of a large-corpus language. Mayrhofer points out that it is certainly important to know that the Gothic word for . mother' (aithei) is of obscure origin. It is centainly not of Germanic ori. gin (cf. Germ. Mutter); and the Germanic word on its part is of IE origin (Cf. Lat. mater etc.). This negative information Feist could have given just in two words: Germ. Mutter' and, as a representative of IE, Lat. mater. Instead, Feist has used up twelve lines for citing three words from Germanic languages and fifteen words from IE languages other than Germanic. One way at this point cosider the question ; how should one arrange the lexemes in an etymological dictionary? The conventional alphabetical arrangement has been ridiculed by some as a telephone-type.' Mayrhofer does not subscribe to this ridicule for, according to him, the old telephone 13 Mayrhofer informs us that this example was cited earlier by E. Hermann, " Zwei Vorschlage" etc. (IF. 56. 1983, p. 193). Hermann expressed concern that if this practico was not given up, an etymological dictionary would succumb to elephantiasis. 17 (Annals BORI) Madhu Vidya/400

Page #426
130 Annals BORI, LXXIX (1998) type' arrangement has proved over the years as extremely convenient for getting quick information. Those who do not favour the alphabetical arrange. ment propose, instead, three kinds of classificatory groupings: (1) grouping the lexemes according to their origin; (2) grouping them subject-wise ( nach Sachgruppen); and (3) grouping them according to their chronology. For the first type of grouping (origin-wise ), Heinrich Hubschmann's Armenische Etymology (1897 ) is recommended as a model. Mayrhofer admi. res the virtues of this dictionary but points out that it is impossible to imitate fully this model while presenting the lexemes of any other corpus language. The peculiar situation of Old Armenian has enabled the author of its dictionary to classify words according to their origin : Persian, Greek, Syrian and pure Armenian. Even in the case of Armenian, Hubschmann is aware that his well-thought out classification may have to be corrected here and there in view of future rescarch in IE and, especially, in Iranian.' In the case of other languages, particularly in Old Indo-Aryan, such a classification would be impracticable. It would be impractical even to approach Hubschmann's type by adding a supplement in which the lemmata are grouped according to origin. The information to be obtained by the reader from such a supplement viz. that words like pitar, sata, as va are old inherited words is so well known that its mention once again will be waste of space. This is also true of such commonplace information that a late Sanskrit word like bharja (Sk. bhartar), which is of Prakrit origin, is a loan from popular speeches. And one who thinks that he can group the rest of the Sanskrit vocabulary into a Dravidian or a Murda block has no idea of the complexity of the problems involved in arriving at such decisions. As regards the second type of classification ( subject-wise ), Mayrhofer points out that even if an etymologist does not accept it and chooses to present his lemmata alpbabetically, he has to study the subjectwise aspect of bis vocabulary beforehand and take into account synonyms, almost synonyms and meaning areas in the language dealt with by him. The only question is wbether it is advisable to present the dictionary itself subject-wise or present the dictionary first alphabetically and then give the subject wise ordering in a supplement. For the present at any rate Mayrhofer would like to await the application of the subject-wise type first, as an experiment, to a dictionary of a small.corpus language like Gothic. As regards the third type of classification (chronology-wise ), Mayrho. fer has already opted for it. He presents the lemmata in his new dictonary in two broad chronological divisions : 'older language and later language'. In the former division, which he has now completed, he includes words Madhu Vidya/401

Page #427
MEHENDALB : Mayrhofer's New Etymological Dictionary 131 which may be called . Vedic' in the widest sense of the term, and also words which appear for the first time in the works of old gammarians like Panini and Patanjall. Such words which do not appear before the epics and the law books18 have been assigned place in the second division. Mayrbofer points out that such a chronological division is, in a way, justified since almost one hundred percent of the OIA words of the first division (older language ) have le or Indoiranian explanations. Words of foreigo origin or even those which reveal the stamp of MIA phonology are an exception in the first divison. On the other hand, suggestions to derive words of the second division from IE roots are recognized as possibilities only in a few cases. 11 Those, however, who do not subscribe to such chronological divisions and still cling to the slogan Old Indian is Old Indian' adduce in support the evidence of the word pard- which, though it belongs to the later language, is of le origin. They are, however, now faced with the fact that the word pard, is, in fact, attested in a Vedic text (TS. 7.5.1.2) and hence can claim to belong to the older language, 15 One last question of principle dealt with by Mayrhofer is : how much should the author of an etymological dictionary cite from the secondary literature, how many earlier etymological explanations does he have to refer to? Since such explanations are too many, quite obviously one has to make a choice. Aleksander Bruckaer makes this painful task quite simple for him. self. He says that in his Polish dictionary he cites only that explanation which in his opinion is correct, probable, or possible, and passes over the rest in silence. He neither mentions them nor argues against them. Mayrhofer strongly objects to this attitude, not so much because it smacks of self-righteousness and arrogance, but because he is convinced that our linguistic 13 In his Uberlegungen (2. 2. 3, p. 150) Mayrhofer had thought of including words which are first attested in older law:books like those of Manu and Yajoavalkya in the first division. He has since:changed his mind and decided to place all law books in later language' By arranging the lemmata chronologically, which arrrangement incidentally throws light on the two types of the origin of the words, Mayrhofer comes a step closer to satisfying the demand that the lemmata be arranged origin-wise. Mayrhofer's view that it is absolutely necessary to divido OIA lexicon into two chrono. logical parts of an etymological dictionary has some justification no doubt. But in comparing the view of those who wish to combine tho two divisions - old and late into one dictionary with an hypothetical etymological dictionary of Old Greck mixing up in one book also middle and new Greek is going too far. The comparison would be apt only if those who wish to give the etymological dictionary of OIA as one book had also thought of combining with it the etymological dictionaries of MIA and NIA. Madhu Vidya/402

Page #428
132 Annals BORI, LXXIX (1998) science is of dialectic nature. It is therefore not desirable to shut the door completely on all available secondary literature, Finally, as regards the question of the form of a practicable etymologi. cal dictionary, Mayrhofer says that the best way to approach this question is to know first the form of an ideal etymological dictionary of a large-corpus language. This ideal makes the following demands : 1. The author of the dictionary has to transform all independent words of the traditionally handed down texts into etymological lemmata. 2. These lemmata are to be referred to their ultimate constituent ele. ments in accordance with the rules of the grammar of that language ; and, if the word is an inherited one, then in accordance with the rules of the grammar of the older stage of that language ; and, finally, if the word is a loan word, then according to the rules of that source language. 3. The author of an ideal etymological dictionarty has to enquire about not only how his lemmata have arisen, but also where they have gone, i, e., e. g. whether OIA words have survived in later Indo-Aryan languages or whether Latin words in Romance languages. This information is accompanied by the citation of comparable forms from related languages, especia. lly closely related languages. 4. The reader of the dictionary has to be supplied with information on all etymological explanations that have been previously offered. 5. In case the language does not already have a dictionary dealing with word-bistories, or if no such dictionary has been promised for it, informa. tion about the fate of the lemmata within the documented history of the language shall follow the above information on the origin of the words and their comparison with related languages. The ideal etymological dictionary thus will be both etymological and word-historical at the same time. It will be a thesaurus augmented by the information on the rules of derivation, linguistic comparison and history of etymological explanations. That will be a wonderful achievement no doubt, but its execution has to face two formidable burdles. Such a dictionary will run into several volumes, may be fifty, tu which will unbearably tax the As an actual example, Mayrhofer mentions W. von Wartburg's Fransosio os Etymologisches Worterbuch. Madhu Vidya/403

Page #429
MEHENDALE : Mayrhofer's New Etymological Dictionary capacity of any publisher or any buyer. In addition, it presumes that the author of such a dictionary shall have a life of say 250 years' which he will live with full intellectual freshness. So long as these practical and biological Aurdles are not overcome we must look for limitations which will enable us to accomplish a practicable etymological dictionary which is optimal and belongs to this world. 1. The first limitation will be that a practicable etymological dictionary of a large-corpus language will not enter all synchronous lexical items as its independent lemmata. It will be enough to cite word families under each concerned word and herein too selected derivatives as give unexpected evidence for word formations or semantic and phonetic histories. An etymological dictionary of an abundantly attested language cannot be a substitute for its descriptive lexicon and a book teaching its word formation. 133 2. The next limitation pertains to giving information on comparable linguistic forms. In this regard, utmost economy is necessary and possible because such information is repeatedly given in other books. Mayrhofer says that no etymologist can be urged to write a book to help a Robinson, who has with him only this book on his island, to write with its help a doctorate thesis in linguistics. In the case of OIA, information whether a given word has become extinct or whether it survives in NIA languages can be supplied by just referring the reader to R. L. Turner's A Comparative Dictionary of the Indo-Aryan Languages (London, 1966). Next, with regard to information that a word like Skt. matar is of IE origin it is not necessary to document at IE parallels. It will be enough to say: "IE, cf. Lat. mater etc.", and then give a reference to any one of the many books which list all IE "mother" parallels. 3. A special problem that Mayrhofer had to face in his new dictionary was how much-information he might consider adequate to inform his reader that a given lexeme was of Indoiranian origin. In the case of OIA words this information is important. It would have been very convenient for this purpose if he could avail himself of a single book as he could do for Indian by referring the reader to Turner's dictionary. In its absence, Mayrhofer decided to cite a single Iranian parallel to indicate Indoiranian origin of an OIA lexeme and give, in addition, for each such case, the source of the most detailed available secondary literature. 4. As for the information on the earlier etymological explanations of the words dealt with, Mayrhofer has already clearly stated that he is not at all Madhu Vidya/404

Page #430
134 Annals BORI, LXXIX (1998) in favour of withholding such information from the readers. What is required is brevity. In view of the very large number of early explanations one should, as far as possible, only indicate earlier collections where detailed etymological explanations can be easily found. Mayrhofer points out that many modern etymological dictionaries are lucky in having predecessors which contain detailed information on secondary literature. But when he wrote his KEWA he did not enjoy such benefit. The only complete etymological dictionary then available was that of C. C. Uhlenbeck's Kurzgefasstes etymologisches Worterbuch der Altindischen Sprache (Amsterdam, 1898/9), and even this did not make mention of secondary literature. All the other attempts made in this direction remained unfinished. The situation has, since, changed. The one reason that encouraged Mayrhofer to take on himself the difficult task of writing a new dictionary was that he was now in a position to cite for bibliographical references, besides other reference works, his own KEWA. Mayrhofer has thus made it abundantly clear that the etymological dictionary that he is presenting is not an ideal one but a practicable' one which he can think of completing in the foreseeable future. His new dictionary, which he anticipates to run into three volumes, gives, wherever possible, under each lemma the follwing information: 1. Single words or word-families appear at the beginning of each entry.. Not only isolated words are given an independent lemmata, but also derivatives whose formation or semantics is of special interest or whose relationship is not certain appear independently. Hence, although anisatra- and distyaare given under anisa- 'shoulder', amsadhri- and amsald- are recorded as independent lemmata. Similarly, although agratas and agrlyd- are listed under agra tip', agrimd- appears as an independent entry. Primary deri vatives are listed under verbal roots from which they are derived. Thus aja, "djana-, ajira-, ajma-, ajman- are all listed under AJ18 drive, move 17 Of these the first two volumes dealing with older Sanskrit have already: appeared (see above f. n.*. To date, first four fascicules of Vol. III giving words of later Sanskrit. (amhati- to pincha-) have also appeared (1997-1998). 18 Mayrhofer prints verbal roots in capital letters. He gives them in their full grade form as AY (not 'to go'. TAR (not TR) to cross, overcome', BHED (not BHID) to split' etc. He gives the traditional forms, set up by Sanskrit grammarians, as dummy entries with a reference to their corresponding full grade form entries (c. g. I see Ay etc.). The nouos also appear in their full grade froms as bhratar-(not bhratr-), brhant (not brhat-), mahant- (Dot mahat-). In this case, Mayrhofer does not record the weak grade forms as dummy entries. In case of the stems in -iand --, however, he makes an exception and gives giri- (not giray-), guru (not gurav-). Madhu Vidya/405

Page #431
MEHENDALB : Mayrhofer's New Etymological Dictionary 135 forward, throw' and do not occur as different temmata. Even a word like a-yatana 'fixed place' appears not as an independent entry, but under YAT.19 2. In the case of each word, whether it occurs as a head-word or is listed under it, it is stated in which text or text category the word first appears and whether or not it is attested beyond it. Thus arnhars- RV, cidbhuta- RV +, dasa RV +, dasin- listed under dasa, BR +. 3. The next information pertains to the extent to which the OIA words have been inherited by modera Indian languages. As remarked above, in general, a reference to Turner is found sufficient for this purpose. 4. The next information is about the origin of a given lexeme.20 If a word is not a borrowing from any of the non-Sanskritic languages, 41 it is important to know whether, in the first instance, it is of Indoiranian origin. As noted above, in principle, it was enough to cite a sure parallel from a single Iranian language to prove Indoiranian origin. Mayrhofer next informs the reader whether the word further goes back to the IE period. He shows the IE origin by inserting, wherever possible, a reconstruct. There is no better way, according to him, to indicate that the author has been quite clear about the phonological and inflexional processes involved in tracing the lexeme back to its original form, as also about the root structure and word formation of a comparison. As for citiog IE parallels Mayrhofer felt that it was enough to document three or four attestations out of twelve IE languages. In addition, he gives reference to one of the many works where the reader can get full information on comparable forms. 23 Mayrhofer assures his readers that in presenting the individual words of his new dictionary he has kept in mind the duty of an etymologist and 19 It would have been better to record ayatana and similar words as du.amy entries. 40 In the new dictionary too, as in the old one, one comes across labellings like 'etymo. logisch unklar' (arsald- ), nicht befriedigend gedeutet' (akudhry dk ), nicht uberzeugend gedeutet (akapara-) or even nach Form, Bedeutung und Herkunft unsicher (ansadhri-). Such labellings do not mean, as Bartholomae in his criticism of Uhlenbeck's dictionary thought (IF 12. Anz. 23 ), that such words defy explanations right from the start. Rather, they are indicative of a challenge to future research. $1 Mayrhofer includes among these MIA languages also. 9? For this purpose Mayrhofer has consulted a very wide range of secondary literature. This becomes clear from the very large number of bibliographical abbreviations given at the beginning of fascicule 10 of Vol. I (pp. XVI-LVIII) and fascicule 29 of Vol. II (pp. IX-XXIX ). Madhu Vidya/406

Page #432
136 Annals BORI, LXXIX ( 1998) taken into account the entire vocabulary, i. e. he has already studied the semantic areas of words, their synonyms, antonyms, homonyms and home nyms. But since he was dealing with a large-corpus language like OIA, it was not feasible for him to present the dictionary itself into corresponding sections. It is therefore highly encouraging to know that Mayrhofer proposes to publish, after the completion of the third volume, a supplement on the " meaning areas" (Bedenturgsfelder ) found in OIA. Mayrhofer is thus all set to earn a unique distinction. - He will be the first etymologist to bave to bis credit, not one, but two completed etymologi cal dictionaries of a large-corpus language with a supplement on" meaning areas". One can only wish him god speed ! Madhu Vidya/407

Page #433
(V) Mahabharata MAHABHARATA STUDIES 1 BY M. A. MEHENDALE (1) akumaram While writing on akumaram yasah panineh in the ABORI 58-59, pp. 727 ff. (1978), Acharya Shri V. P. Limaye rightly draws attention to the fact that akumaram in the above expression cannot mean upto children' but that kumara must refer to some place-name ('upto Kumara country'). Shri Limaye therefore suggests to emend the text of the Mahabhasya and read akumari in place of akumaram so that the expression could convey that Panini's fame had spread as far as the southern-most tip of India, presectly known as Kanyakumarl. Shri Limaye himself, however, makes it clear that no variant has been recorded in the Mahabhasya editions for akumaram. Now it is perhaps possible to say that the use of the expression akumaram itself was the usual way, at least in the epic times, for conveying the sense of long distances. This becomes clear from the following few passages in the Mahabharata : (1) Duryodhana, on his return from the Rajasuya sacrifice, became very dejected for various reasons. Some of these are listed in the following stanza : pArthAn sumanaso dRSTvA pArthivAMzca vazAnugAn / et ara fes start **ET I 2. 43. 14. Duryodhana began to become pale because, firstly, he saw the Pandavas well-pleased, secondly, the kings, assembled there, were obedient to the Pandavas, and, thirdly, the whole world was well-disposed towards them. In order to convey the very wide extent of the world (lokam), which in the present context would mean the Bharatavarsa, the author has used the express sion akumaram which must mean as far as the Kumara (country)'. It 1 Incidentally it may be noted that in the traditional explanation of this expression the meaning of kumara child' (bala ) is expressly stated very late in the commen. tary Ratnaprakasa (yad balan api sambadhnati) on p. 349 of the Mahabhasya. Pradipa-Vyakhyanani, part IV, ed. by M. S. Narasimhacharya, Pondichery, 1977. (I am thankful to Dr. G. B. Palsule for this information. ) It cannot be stated definitely that this meapipg was intended by Patanjali. Even in the wording of Pradipa (kumarun api yasah praptam ) or of Udd yota ( kumurarupa maryada ), the word kumara could refer to some place-Dame. Madhu Vidya/408

Page #434
Annals BORI, LXV (1984) is most unlikely that akumaram here means down to the children '2 because Duryodhana could have impossibly noticed children in the gathering that had come together for the Rajasaya and hence could say that even children were well-disposed to the Pandavas. Moreover Duryodhana is not particularly likely to be jealous of the Pandavas on seeing even children favour able to them, but certainly on noticing that the people upto the (very distant) land of Kumara were so disposed towards the Pandavas." (2) acaiva tadviditaM pArthivAnAM 246 bhaviSyati AkumAraM ca sUta / nimagno vA samare bhImasena eka: kurUn vA samare vijetA // 8.54.18. Bhimasena tells his charioteer Visoka that that very day the kings would know that either he himself had gone down in the confrontation, or he alone had conquered the Kurus. He says that the kings coming from as far as the Kumara (country)' would come to know that fact. In this context it is hard to believe that Bhimasena declared that the fact would be known by the kings on the battle-field and by the young princes at home that very day. What happened on the battle-field would be known on that very day only to the kings who had gathered there from far and wide. The expression akumaram, therefore, has to be interpreted as referring to some distant place and not to a child." (3) When Bhima points out to Yudhisthira how difficult it is going to be for the Pandavas to remain unknown during the period of the ajnatavasa, about himself he says: mAM cApi rAjAnanti AkumAramimAH prajAH / ajJAtaca pazyAmi meroriva nigUhanam // 3.36.27. In this connection too akumaram has to be understood as indicative of wide geographical area. There is no point in saying that even children knew Bhima and hence it would be difficult for him to remain unknown; it would be more to the point if he were to say that since he was known far and wide, however far he might go from Hastinapura, people would recognize him." The above passages from the Mahabharata make it clear that the word kumara in the expression akumaram in them should be understood to * As translated by van Buitenen. 8 akumaram ia mam va lokah kirtayanty akumaram (8.54. 19) is also to be understood as upto the Kumara (country). P. C. Roy, however, translates 'beginning with the feats of my earliest years'. 4 But P. C. Roy translates including the very children 6 van Buitenen, however, 'down to the kids'. Also cf. 11. 9. 8; 8. 30. 12 (very peculiar is the expression akumarah smaramy aham where akumara- is an adjective). Madhu Vidya/409

Page #435
MEHENDALE: Mahabharata Studies I refer to some place-name far from Hastinapura. About its exact identification we may not be certain now. It may refer to the Kumara-visaya in the east whose king Srepimant was conquered by Bhimasena (Mbh. 2. 27.1) and which lay near Cedi and the Kosala countries"; or it may refer to the Kumara-tirtha (in the south?) referred to in the Brahmandapurana 3. 13. 86". All this is not said to assert that akumaram cannot mean up to the children.' In suitable contexts it can very well do so. In the Mahabharata itself where akumaram is used with reference to nagara or pura it means * upto the children Eg. bhakumAraM naramyAna tarapuraM vai samantataH ArtanAda mahaccakre 9. 1. 16; also 1. 118. 30. 247 (2) A Few Cases of Internal Incongruity in the Sabhaparvan. (i) Enumeration of the Gandharvas in the Indrasabha. In the Sabha parvan, Narada, at the request of Yudhisthira, describes for him the Halls (sabha) of Indra, Yama, Varuna, Kubera, and Brahman (Adhyayas 7-11). When this description is over, Yudhisthira in stanzas 43-48 of Adhyaya 11 recounts the special features of all the five sabhas as described to him by Narada. When he comes to the Indrasabha he says i "Oh sage, you have enumerated for the sabha of Satakratu the gods, the Gandharvas in brief, and the different great seers." When we compare the information contained in this stanza with the contents of the constituted text of Adhyaya 7 which describes the Indrasabha, we find that the gods and the various sages are, indeed, enumerated, but as for the Gandharvas there is only a general reference to them, 10 along with the Apsaris, and they have not been enumerated even briefly. A general reference to the Gandharvas (and the Apsaras) is found in the descriptions of all the sabhas and therefore it cannot be looked upon as a special feature of the Indrasabha. On the other hand, it is for the Kuberasabha that the Apsarlisl and the Gandharvas13 are enumerated in details. This can thus be a specia 7 Among those people who brought tribute to Yudhisthira at the time of the Rajasuya are listed the Kundamanah. For this reading adopted in the critical edition (2. 48. 13) there is a variant Kumaras ca. Since the word occurs with Kasmirah, the Kumaras, mentioned here, could be from the north-west. * kaumAraM ca saraH puSyaM naagbhogaabhirkssitm| kumAratIrthe snAtvA tu tridivaM yAti mAnavaH // 9 satakratusabhayam tu devah samkirtita mune uddesata's ca gandharva vividhai ca maharnayah 2. 11. 47. Devabodha glosses uddesah samkepah. van Buitenen, however, distinctly". 10 tathaivapsaraso rajan gandharvas ca manoramah devarajam satakratum 2. 7. 21. 11 2. 8. 35 (of Yama ), 2. 9. 23 ( of Varuna), 2. 11. 19, 36 ( of Brahman). 12 2. 10. 10-11; also *111, 112. 18 2. 10. 14-17. Stanza 14 mentions two classes of Gandharvas - those that are named 4 Kimnara' and those that are named Nara'. The listing that follows seems to enumerate the Gandharvas of the Kimnara' type. It is likely that this listing concluded somewhere and then followed another list viz. that of the Yaksas (cf. st. 18). Madhu Vidya/410 ramayanti sma nrpate

Page #436
Annals BORI, LXV (1984) feature of the sabha of Kubera, and Yudhisthira, while mentioning the special features of the sabha of Kubera, does make a reference to it.14 Yudhisthira's saying that Narada had briefly enumerated the Gandharvas for the Indrasabha is thus not in harmony with the text that has been admit. ted into the critical edition. 248 But if one looks to the passages marked with the asterisk relegated to the foot-notes one finds that the north-eastern group (except the Nepall version) has, indeed, a brief listing of the Gandharvas visvavasus citrasenah sumanas tarunas tatha *86,15 The southern recension too has a listing of the Gandharvas: visvavasul citrasenah pravatas tumbarus tatha *87,14 Since the listing is not found in all the versions, the editor of the Sabhaparvan has not admitted it into the constituted text; but since the stanza placed in the mouth of Yudhisthira which says that Narada had made a brief listing of the Gandharvas for the Indrasabha is found in all the versions it has been admitted into the critically constituted text, This has led to an internal incongruity which must form the subject of higher criticism. There could be two ways of setting aside the incongruity: (1) First, we may assume that the listing of the Gandharvas for the Indrasabha was not there in the ur-text considered to be at the basis of all the versions and hence a reference to its brief listing in the third quarter of the stanza put in the mouth of Yudhisthira was also not there. When, at a later stage, the listing was added in the southern and the north-eastern (except the Nepali version) recensions, the third quarter of the stanza of Yudhisthira was also suitably modified in order to indicate this listing. Later, only this stanza of Yudhisthira in its modified form, that is as it now appears in the critical edition, got into the north-western recension and the Nepall version due to contamination. Whosoever was responsible for this contamination remained. ignorant of the resulting incongruity. (2) Or, we have to assume that a short listing of the Gandharvas for the Indrasabba did occur in the urtext and, consequently, a reference to it in the third quarter of Yudhisthira's stanza, as we now find it, was also there. Later, this brief listing for the Indrasabha was dropped in the North-western recension and the Nepall version, probably because the enumeration of the Gandharvas (and the Apsaris) was thought fit for the sabha of Kubera, and not of Indra. But it was overlooked that the dropping of this brief listing of the Gandharvas 14 tatha dhanapateh... gandharvapsarasas caiva (kathitah) 2. 11.45. 15 Of course with some variant readings. 18 *87 also contains a list of the Apsaras. 17 How this third quarter looked like in the ur-text before it was modified is hard to say. Madhu Vidya/411

Page #437
MEHENDALE : Mahabharata Studies I 249 from the Indrasabha would lead to an internal incongruity and hence the tbird quarter of 2. 11.47 was allowed to stand as it is.18 (ii) Brhadratha's retirement to the forest. Krsna, while narrating to Yudhisthira about Bihadratha, the father of Jarasandba, has the following to report : Bihadratha went with his two wives to meet the sage Caydakausika who once happened to take his residence on the outskirts of the king's capital. The king satisfied the sage with gifts of choice things. The sage thereupon asked the king to choose a boon. The king, who was feeling despondent due to lack of progeny, said to the sage : " Revered Sir, having left the kingdom I have already started for the penance grove. What do I, an unfortunate person, need a boon for? What is the use of the kingdom 10 me who am without a progeny ? "19 This stapza is incongruous with what we read of the story before. Earlier there is no mention of the king's having left the kingdom (rajyam utssjya ).20 We have been certainly told that the king had become old, 21 but in the constituted text we are not told that he had abdicated the kingdom. Once again we notice that this incongruity disappears if we admit into the constituted text the portion relegated to the footnote ( numbered 178 ) and which is found only in the Southern recension. It reads as : sa bharyabliyam saha tada nirvedam agamad bhrsam / rajyam capi parityajya tapovanam athasrayat / varyamanah prakrtibhir nopabhaktya visam pate/ It is in this portion that we are told of the king's abdication against the wishes of his subjects. We have again with us two possibilities : (1) Either we assume that the account of Bfhadratha's abdication was there in the ur-text and hence also the stanza 2.16.26 which refers to it, and that this account was later dropped in the entire northern recension for some reason.23 The stanza 26, however, could not be dropped because it is related to the following stanza (2.16.27) which begins with etac chrutva, and hence it must be preceded by 18 Only the Kasmiri version has made a clumsy attempt to remove the incongruity by replacing uddesatas tu by vidyadharas tu. The Devanagari group allied to the Kasiniri version has not followed suit apparently because Vidyadharas too are not listed for the Indrasabha. bhagavan rajyam utsrjya prasthitasyo tapovanain/ kin varenalpabhagyasya kiin rajyenaprajasya me // 2. 16. 26 van Buitenen seems to try to get over the incongruity by translating rajyam utsrjya as who will give up his kingdom', but that is hardly possible. tasya yauvanam atyagat 2. 16. 20. 21 Could it be that the account of abdication was dropped because it was considered improper for a king to abdicate without appointing a suitable heir to the kingdom or making some other arrangement for the proper administration ? Cf. how Jarasamdha orders Sahadeva's anointment before he accepts the challenge posed to him by Krsga. 32 [ Annals BORI) Madhu Vidya/412

Page #438
250 Annals BORI, LXV (1984) a stanza (2.16.26) which contains the king's reply. It was therefore allowed to stand without modification in spite of the resulting incongruity. That the ur-text contained an account of the king's abdication as we find it in the Southern recension, but later dropped from the Northern one, is shown by the fact that in the constituted text we have further a stanza in which the sage asks the king to return' (to his kingdom) besides telling him to go'.23 This nivartana must have a reference to the sage's asking the king to retrace his step of abdication and not just to his going back to the city from where he had come out to meet the sage. (2) Or else, we have to suppose that the abdication account was absent in the ur-text and that it was added later in the Southern recension alone with stanza 26 which makes a reference to it. Subsequently only this stanza in its present form got into the Northern recension due to contamination, but not the account of the abdication itself. It will be seen that in both the above examples of internal incongruity the explanation based on later addition', is cumbrous and that in both cases we can give some ground to support the omission theory. Hence we have to choose the explanation based on later omissions. It is true that, generally speaking, when we are confronted with passages that are not found in all the recensions it is safe to assume that here we stand face to face with passages added later in the versions which have them, and not with the passages which are omitted later from the versions which do not have them. Dr. V. S. Sukthankar observes: There is then the question of the "additional" passages, that is, passages found in only one of the rival recensions. There is only one rational way of dealing with these additional passages: they must be carefully segregated from the rest of the text, and examined individually. The onus of proving the originality of these "additional" passages will naturally rest on him who alleges the originality: the documents speak naturally against them, but their evidence is not by any means conclusive "24 (ital. mine). In the case of the two internal incongruities noticed above it has been stated why we have to go against the normal practice and presume that the additional passages are the "original" ones, that is, we have to speak in favour of their later omissions, and not additions. It is at least possible to speak about the omission of certain stanzas in some versions, but it is obvious that we cannot speak of the addition of some stanzas in all the versions. For, that amounts to admitting that they were there in the original text itself. 23 gaccha rajan krtartho' si nivarta manujadhipa 2. 16. 39. 24 ABORI XVI. 91 (1934-35). Madhu Vidya/413

Page #439
NEMESIS AND SOME MAHABHARATA EPISODES* Vu M. A. MEHENDALE The Adiparvan of the Mahabharata (214-225 ) tells us about the burning of the Khandava forest. Once, when Krsna was spending a few days with the Pandavas at Khandavaprastha, he and Arjuna asked for Dharmaraja's permis. sion to go to the river Yamuna to spend there a day. While they were there Agni, in the form of a Brahmana, approached them and expressed his desire to burn the neighbouring Khandava forest. He requested them for help in fulfilling his desire. As Taksaka Naga, the friend of Indra, lived in that forest, the latter had foiled Agni's earlier attempts to burn down the forest. Krsna and Arjuna asked Agni to supply them with the weapons necessary to achieve this end and when that was done, Agni started to burn the forest. Krsna and Arjuna vigilantly guarded the outskirts of the forest and brutally killed all the Danavas and other beings, as well as the beasts and the birds, that tried to flee for their lives from the forest. Those seeking escape were helplessly crying and weeping, but to no avail. Indra, as before, tried to extinguish the fire with the help of rain, but Arjuna had fully covered the forest with his arrows so that pot even a drop of rain could reach the burning forest. Indra started to battle with Arjuna, but was told by the celestial voice that since his friend Taksaka was safe in Kuruksetra he should desist from further fighting. Indra, therefore, left the scene and the forest burned without any hindrance. It was burning for six days. Thousands lost their lives either through the fire or at the hands of Krsna and Arjuna. The denizens of the forest are so listed in the epic : danava raksasa nagas taraksusksavanau kasah / dvipah prabhinnah sardalah simhah kesarinas tatha 11 mrgas ca mahisas caiva satasah paksinas tatha| samudvigna visasrpus tathanya bhutajatayah 1 (1.219. 1-2). Almost all of them died in the holocaust. Only seven remained alive : (1) Naga Taksaka; (2) his son Asvasena; (3) Asura Maya: (4-7) four young birds (Sarngas ). The first, because he was not present in the Khandava, the rest, because they could, or were allowed to, escape. . This is an English version of the authors's Original article in Marathi pub lished in the Navabharat, November 1982. ...35 Madhu Vidya/414

Page #440
274 R. N. Dandekar Felicitation Volume Dr. Mrs. Iravati KARVE was perhaps the first to call attention to this cruel slaughter of the denizens of the Khandava forest. She has tried to explain the slaughter in two ways : (1) The bitds and beasts killed were not really birds and beasts but were human beings with certain birds and beasts for their to. tems (devakas). The forest was burnt down for farming, and it was necessary to see that none of the natives living in the forest survived to claim the ownership of the land. Hence every one had to be killed. (2) In her second explanation, Mrs. KARVE suggests that Krsna and Arjuna did not feel any scruples in killing the residents of the forest since they looked upon them as aliens. The rules implicitly adopted by the Ksatriyas for fairness in war - such as prohibition of killing one who has no weapons, who is running away for life, who is crying for help!-- need be observed only within one's in-group, i. e. while fighting with those who were looked upon as one's own', and not with the aliens. But this explanation, as noted by Mrs. KARYE herself, is not satisfactory since the Nagas at any rate could not be looked upon as aliens by those born in the house of the Kurus. Mrs. Karve, then, asks a question, but fails to come up with a clear answer : Did Krsna and Arjuna feel that the people and the animals living in the forest were fit to be burned down or butchered by powerful weapons? It is time the answer yes', however unpalatable, is explicitely stated. Instead of trying to save living beings from fire, these two heroes did everything to throw them back into it. The author of the Mahabharata is outspoken in his description. He records that Arjuna laughed smugly when he saw pieces of birds, cut down by his arrows, fell into the fire (1.217.11 ). This was an extremely cruel act. It is strange that none of the women who went to the river-side with Krsna and Arjuna objected to what the two were doing. Even Yudhisthira, who permitted Krsna and Arjuna to spend a day at Yamuna, did not ask them on return why they stayed away for six days, and if he knew what had happened, as he must have, ask the two an explanation. The question about the justness of this act has apparently never been raised in our history. One wonders whether the persons responsible for such heinous acts escaped without punishment. We have been tirelessly told in our moral exhortations that 1. In the chapter on Mayasabha in her book Yuganta (in Marathi) pp. 128ff, 2. Mrs. KARVE has not made it clear why she disfavours this explanation. We may think of a few grounds to reject this theory. (1) First, the forest was so large it took six days to burn down. Hence it was not impossible to preserve a part of it and burn the rest for agriculture ; (2) Asvasena, the only Naga who escaped, never came forward to claim his land ; (3) It does not appear that the area surrounding the Khandavaprastha was so thickly populated that it was necessary to burn down a big forest for making land available for agriculture ; (4) Finally, ift the forest was burnt for using it as a farm land why did the author of the Mahabharata not say it so plainly? 3. These are, e. g., made explicit io Mbh. 6. 1. 26 ff. Madhu Vidya/415

Page #441
MEHENDALE: Nemesis and some Mahabharata episodes one has to suffer the fruits of one's bad conduct. Did this law of retribution remain suspended in the case of these two powerful persons? No, it does not seem so. The events in the Mahabharata show that the Pandavas had to suffer for the misdeeds of Arjuna connived at by them they suffered almost the same way, perhaps, even more cruelly. We cannot possibly overlook the connection between the events the one narrated above and those to be narrated below. - Duryodhana fell on the battle-field when he was struck down by Bhima. He was no more able to stand on his feet. From the point of view of the Pandavas the war had ended. Yudhisthira felt that he was now the ruler of the earth. All the warriors on the side of the Pandavas went to Duryodhana's camp and looted it. All of them then decided to spend the night in the camp of the Pandavas but Krsna suggested that he, together with the Pandavas and Satyaki, stay away as they had to perform some auspicious ceremony. The Mahabharata does not tell us what this ceremony was, and how it was performed. It makes only a casual reference to it in the following words: 275 athabravin maharaja vasudevo muhayasah/ asmabhir mangalarthaya vastavyam sibirad bahih // tathety uktva ca te sarve pandavah satyakis tatha vasudevena sahita mangalartham yayur bahih // (9.61.35-36). Although Krsna suggested that the Pandavas should stay away for the performance of an auspicious ceremony, it is quite clear that this was only a pretext. He was aware that Asvatthaman was up to some evil design. He told this clearly to Dhrtarastra when he was deputed by the Pandavas to console Dhrtarastra and Gandharl. He said: apreche tvain kurusrestha ma ca soke manah krthah draunen papo aty abhiprayas tenaha sahasotthitab/ pandavanam vadhe ratrau buddhis tena pradarkita (9. 62.68). When Aivatthaman, Krpa, and Krtavarman meet Duryodhana lying wounded on the battle-field, Asvatthaman, in a fit of anger, vows to kill the Pancalas. At the instance of Duryodhana, Krpa consecrates Aavatthaman as the commander of Duryodhana's forces, which means that from the point of view of the Kauravas the war had not ended. While Aivatthaman anxiously ponders 4. One only does not know why he did not caution the Pandavas against the possible danger to their lives. Madhu Vidya/416

Page #442
276 R. N. Dandelar Felicitation Volume over how to fulfil his vow, the way is shown to him by an owl who kills the crows in their nests at night. Asvatthaman then decides to kill the Pandavas and the Pancalas at night while they are asleep in their camp. Asvatthaman enters the camp of the Pandavas and Kepa and Ketavarman stand guard at the gate of the camp. Asvatthaman is now on his killing-spree, his first victim being Dhrstadyumna. He then starts killing others who are unarmed and not properly prepared for a fight. In the darkness, he finds out his victims by their cries for help. Those that try to escape from him are taken care of at the gate by Krpa and Krtavarman. Their victims too are unarmed and unprepared and are, in fact, seeking for mercy with folded hands. But all this is of no avail. No one escapes them. All the principal warriors on the side of the Pandavas and the Pancalas, including the sons of Draupadi, are slain; no one who slept in the camp that night remained alive to see the light of the day next morning. When Asvatthaman finally leaves the camp it is quiet, exactly as it was when he entered it the previous night - albeit with a difference. The only ones who remain alive on the side of the Pandavas are those who were not in the camp (Ktsna, Satyaki, and the five Pandavas), and, in addition, the charioteer of Dhrstadyumna who, apparently, had a miraculous escape. This charioteer relates to Yudhisthira about thr cruel killings in the camp. One just cannot imagine the condition of the Pandavas on hearing this ghastly news. This incident in the camp runs parallel to the incident in the Khandava forest. It is hardly necessary to comment on them. The similarities between the two stare us in the face. (1) Agni alone entered the Khandava forest and burnt the living beings that came his way; Ktsna and Arjuna stood outside and killed those who tried to escape. Asvatthaman alone entered the camp of the Pandavas and killed those he could lay his hands on ; Krpa and Ketavarman stood out and killed those who tried to escape from Asvatthaman. (2) Agni does not seem to have told Krsna and Arjuna to kill the beings running out to save lives; his only request to the two heroes was that they should foil Indra's attempt to extinguish the fire. Asvatthaman also has not told his companions to kill those who might try to run away from the camp. In fact he boasts that no one will escape him. He may have, however, expected Krpa and Kitavarman to stop any one who tried to enter the camp to help those inside it. Madhu Vidya/417

Page #443
MEHENDALE : Nemesis and some Mahabharuta episodes 277 (3) Only six beings escaped alive from the holocaust - Asvasena, Maya, and the four young ones of Sarnga; and one more, the Naga Taksaka, remained alive as he was not present in the forest - a total of seven.' Only one could escape alive from the holocaust at the camp; and seven others - Krsna, Satyaki, and the five Pandavas - remained alive as they were not present in the camp - a total of eight. (4) The Khandava forest enjoyed the protection of Indra ; Krsna and Arjuna defeated his attempt to extinguish the fire. God Sankara, in the form of a gigantic being, protected the camp of the Pandavas; Asvatthaman himself praised Sankara and obtained entry into the camp. ... (5) Kesna receives from Agni the famous Sudarsana disc, and Arjuna the famous Gandiva bow and two inexhaustible quivers. Sankara himself enters the body of Asvatthaman to confer on him superhuman strength and also presents him an excellent sword. (6) The author of the Mahabharata says that when the fire began to burn the forest it appeared as if the age had come to an end (yuganta 1.216.32) The author says that one had exactly similar feelings when killing took place in the camp (10.8.137). (7) A Celestial Voice which dissuaded Indra from continuing his fight with Arjuna also indicated that the destruction of the Khandava forest was predestined' (distam 1.219.18). God Sankara tells Asvatthaman that the Pancalas have been humbled by Kala and can no longer remain alive (abhibhutas tu kalena 10.7.63 ). These similarities between the incidents related to the forest-fire and the camp-killings are so striking that they lead one to suspect that there is some rela. tion between the two. The author of the Mahabharata, presumably, indicates this relationship by comparing Asvatthaman, about to enter the camp of the Pandavas with the fire burning down a forest of dried up trees (kaksam dipta ivanalah 10.3.28). The Pandavas fought the Kauravas with a view to regaining the kingdom that was theirs and incidentally, fulfil the vows they had taken on various occasions. The joy and the satisfaction which they may have felt at the fall of Duryodhana turned out to be too short-lived. Yudhisthira's sense of total frustration is reflected in his words : " although we conquered the enemies, it is Madhu Vidya/418

Page #444
278 R. N. Dandekar Felicitation Volume we who are defeated " (10.10.9). The author of the epic observes on the catastrophic episode as : " undoubtedly the destined course of events cannot be changed " (10.8.143 ).5 Every reader of the Mahabharata is moyed by the destruction in the camp. That is as it should be. But no one seems to have any tears to shed for those numberless victims from the Khandava forest who were either burnt down or killed by the weapons for no fault at all. Asvatthaman is rightly criticised for his cruelty, but it can be said that he wanted to wreak vengeance for the killing of his father. But the acts of Ktsna and Arjuna can in no way be motivated, let alone justified. But no one apparently has found fault with them, either in the epic or elsewhere in the course of the history. Shall we say that the people, since the time of the Mahabharata, gained in antiquity but not in moral sensibility ? But whether one has asked or not Krsna or Arjuna for explanation of their action, the law of retribution had run its regular course. The Khandava forest was mercilessly burnt down and those responsible for it invited on themselves punishment. Arjuna laughed when he cut down the birds and let them fall into the raging fire; with incomparable grief he heard the killing of those near and dear to him. In burning down the Khandava forest Agni was helped by Krsna. The Mahabharata tells us that this mighty personage of the epic did not escape the working of the law of karman. The third unfortunate incident is narrated in the Mausala Parvan of the epic. It took place outside Dvaraka, near Prabhasa. Thet incident may not be called cruel but it was certainly tragic. Krsna, Balarama and many Yadavas, accompanied by their wives and children, went for pilgrimage to the sea-shore. Under the influence of intoxicant drinks they began to abuse one another. This took a serious turn leading to a massacre. It is significant to note that the mutual recrimination began with a reference to the incident that took place in the camp of the Pandavas described above. Satyaki first reproved Ketavarman for killing the persons who were asleep. And the killing which started at this charge stopped only when all the Yadavas assembled on the sea-shore, including Satyaki, Pradyumna, and 5. asansayan hi kulasya paryayo duratikramah 6. As a matter of fact, this blame should go to Asvatthaman. But one has to remember that Satyaki was under the influence of wine while blaming Krtavarman. Madhu Vidya/419

Page #445
MEHENDALE : Nemesis and some Mahabharata episodes 279 Aniruddha, were slain. The Yadavas were so far intoxricated and furious that while killing one another the father did not spare the son, nor the son the father. Like the animals who died in the Khandava forest, or the heroes who were but. chered in the camp of the Pandavas, the Yadavas too were goaded to this fate by Time (kalaparyayacoditah 16. 4. 29). No one from outside Dvaraka had come to carry out the killings. There was therefore no question of any one trying to flee (16.4.41 ). "Only two Yadavas remained alive - Krsna and Balarama. Balarama met with his end when a Naga (cobra ) left his body and entered the occan. Krsna's life came to an end when a hunter, mistaking him to be a deer, struck him on the sole of his foot with an arrow. The river Yamuna stood witness to the conflagration in the Khandava, the ocean to the destruction of the Yadavas. Krsna had to helplessly suffer the sight of annihilation. It is highly ironical that Krsna, who was responsible for killing the animals of the Khandava, was himself mistaken to be a deer and killed. His last wish to end his life while practising penance remained unfulfilled. The Yadavas had to suffer yet another ignominy, As desired by Krsna, Arjuna left with the remaining Yadavas and the women-folk for Hastinapura. On way, they were attacked by the Abbiras. Arjuna could not protect those in his the charge. The Abhiras kidnapped the Yadava women, and what was worse, some women even lusted and willingly went with them (16.6. 17). The Mahabharata says that the Yadavas were destroyed due to the curse of a sage. It also is on record that Gandhari had cursed Krsna to that effect. But one gets the feeling that Kesna, very much like Arjuna, invited on himself the punishment as an act of retribution. The Yadavas are described as devadandanipiditah (16, 2.5). It this way, one supposes, that one ought to interpret the three harrowing episodes in the Mahabharata. Madhu Vidya/420

Page #446
DRAUPADI'S QUESTION Vu M. A. MEHENDALE $ In the Mahabharata the Kauravas and the Pandavas-or rather Sakupi and Yudhisthira--play the game of dice twice. Towards the end of the first game, Yudhisthira first stakes his brothers, then himself, and finally Draupadi, and loses all the games. In the game of the dice that takes place between Nala and Puskara, the latter suggests to Nala to stake his wife Damayanti,', but he does not oblige. Yudhisthira too could have rejected Sakuni's suggestion to stake Draupadi, but he does not, stakes and loses her, and puts her at the mercy of the Kauravas. After Yudhisthira loses Draupadi's stake, many things happen in the Assembly Hall. And all that takes place ends, according to the Mahabharata version available today, as follows: at a particular instant a jackal suddenly began to howl in the Agnihotra hall of Dhstarasfra ; he was joined by donkeys and some other ferocious birds. Hearing these ill omens Vidura and Grindbari got frightened. They approached Dhstarastra and made him understand the forebodings. Instantly Dhftaristra took Duryodhana to task for having summoned Draupadi to the Assembly Hall, and offered Draupadi a boon of her choice. Draupadi asked, first the freedom from bondage for Yudhisthira, and, when a second boon was offered her, the freedom of the rest of the four Pandavas. Dhstarastra offered her a third boon; but Draupadi politely declined it saying that only two boons were meet for a Ksatriya woman (dvau tu ksatrastriya varau 2.63.35 ). Reading this account one is inclined to believe that all that was happening in the Assembly Hall came to a halt and the Pandavas got their freedom due to the howlings of the jackals and the donkeys, Here are a couple of expressions of this belief : (1) Writing on the Mahabharata in his History of Indian Literature, Vol. I, Prof. M. Winternitz gives a brief account of the main narrative of the Journal of the Oriental Institute, Vol. No. 35, Nos. 3-4, Sept.-Dec. '85 issue, pp. 179-194 Dr. P. V. Kane Memorial Lecture delivered at the Asiatic Society, Bombay, December 16, 1985. This is an English version of the author's original article in Marathi published in the Navabharat, August, 1985. $ Bhandarkar Oriental Research Institute, Poona-4, 1. Mbh. 3.58.3. Madhu Vidya/421

Page #447
180 M. A. MEHENDALE epic. When he comes to the end of the incidents in the Assembly, he writes : " Whilst still further speeches are being exchanged, the loud cry of a jackal and other sounds of ill omen are heard in Dhstarastra's house. Terrified by these, the old king Dhstarastra at last feels himself called upon to intervene." (p. 345). This means that the jackals and the donkeys were responsible for the intervention of Dhstarastra.! (2) Shri Anand Sadhale in his book T 374 1973 16 E' says at one place : "At this stage Fate, in the form of a donkey, came to the rescue of Draupadi.... One does not know how long this humiliation would have gone on, but just at that moment a donkey began to bray which foreboded evil ...... .... Dhrtarastra lost all courage due to those evil signs. He immediately restrained his sons to avoid the evil " 2 (p. 63). The general belief thus is that the sounds uttered by certain animals foreboding calamity were responsible for putting a stop to Draupadi's humiliation and for freeing the Pandavas from bondage. This belief is not of recent origin. It is to be met with in the older poetic epitomes of the Mahabharata, Here are a couple of examples : (1) The Bharatamanjari : Ksemendra wrote this poem in the first half of the eleventh century. Ksemendra's description of the game of dice follows the one in the Mahabharata. When Bhima uttered his terrible vow to break the thigh of Duryodhana the jackals began to howl and flames of fire, covered with smoke, arose (2.435-437). Noticing these ill omens Vidura and Gandhari immediately held consultations with Dhstarastra. Soon Dhstarastra showered his favours on Yudhisthira and the latter started to go home from Hastinapura. Ksemendra does not expressly mention the boons conferred on Draupadi by Dhstar iscra, but his description leaves no one in doubt that in his view it was the ill omens which gave a turn to the events happending in the Assembly Hall. (2) The Balabharata : Amaracandra's Balabharata belongs to the thirteenth century. He too narrates that there were ill omens immediately after Bhima's vow (2.5.61 ). Then Dhstarastra upbraided his eldest son and offered boons to Draupadi. Since the order of the events in the Assembly Hall as reported in the Mahabharata itself shows that the boons were granted by Dhstarastra to Draupadi - 1 Similarly F. Edgerton : "At last evil omens forced Dhstarastra to intervene ". The Sabhaparvan, Poona, 1944, p. xxvi. 2 "azA prasaMgI deva kA gADhavAcyA rUpAne draupadIcyA sAhAyyAsa dhAvUna Ale...hI viTaMbanA ANakhI kuThavara cAlatI na kaLe. paNa tevaDhayAMta akA gADhavAne azubhasUcaka azA AvAjAta rekAvayAsa suruvAta kelI. ... dhRtarASTrAcA dhIra tyA azubha zakunAne akadama khacalA. azubhAceM nirAkaraNa karaNyAsAThI mhaNUna dhRtarASTrAne ApalyA putrAMnA akadama AvarileM." Madhu Vidya/422

Page #448
DRAUPADI'S QUESTION 181 immediately after the occurrence of the ill omens, there is little wonder a similar narration is found in the poetic compositions based on the Mahabharata, Moreover, we find a second allusion in the Mahabharata to these ill omens after the game of dice was over. It is reported this way: After the first game of dice was over, the Kauravas and the Pandavas had a second round, and the defeated Pandavas left Hastinapura for a twelve year's exile. In the capital, Dhrtarastra began to feel nervous. Seeing his plight Samjaya said to him: "In spite of the protests of Bhisma and the others, Duryodhana forced Draupadi to come to the Assembly Hall. (You did not stop him then). You yourself have brought about this calamity. (Why do you feel nervous now ?)" Dhrtarastra replied: "When the moment of disaster arrives, one is off one's guard. That is what happened to me. The children behaved stupidly and dragged, Draupadi to the Hall. Duryodhana and Karca reviled her.' Then there were many bad omens. On the advice of Vidura I offered boons to Draupadi" (2.72.1-25). Thus by repeating the mention of the occurrence of bad omens the Mahdbharata has provided for calling Dhrtarastra himself to witness, should someone doubt that the events in the Sabha took a dramatic turn on account of the cries of jackals and asses. The Pune edition of the Mahabharata has admitted into the critical text the stanzas of bad omens at both the places in the Sabhaparvan. Nevertheless we have to look upon them as interpolations. For, if we do not, we have to assume that the basic question raised by Draupadi regarding her social status remained unresolved to the end, Apparently everybody seems to be convinced about this. Here are a few examples of this conviction: (1) Prof. N. R. Phatak, in the first Volume of the new edition2 of the Chiplunkar Mandali's Marathi translation of the Mahabharata, observes: The significant question which Draupadi had raised at this extremely critical moment could not be answered satisfactorily by any one. Therefore Dhrtarastra managed somehow to get out of the fix by offering boons to Draupadi. " (p. 8) (2) An English translation of the Subhapurvan by J.A.B. van Buitenen has recently been published. He says in his introductory remarks: Yudhisthira staked and lost himself', she asks, before he staked me? If so, he had lost his freedom and, as a slave of the Kauravas, no longer owned her to stake. There is much argument, but it remains inconclusive," (p. 30). In actual fact Duryodhana is not on record to have reviled Draupadi in the Hall, 2 Surekha Prakashan, Bombay, 1967. 3 " yA vikSalaNa ANIbANIcyA veLIM draupadIneM jo eka mArmika prazna vicAralA tyAlA samAdhAnakAraka uttara koNIca devU zakenA, yAmuLe dhRtarASTrAneM hA peca kasAbasA dii| bIlA para devUna soDavilA " 4 The University of Chicago Press, Chicago and London, Vol. 2, 1975. MadhuVidya/423

Page #449
182 M. A. MEHENDALE Prof. Phatak and Prof. Buitenen are thus in no doubt about the inconclusive nature of the debate. Who then gave the turn to the course of events? According to Prof. Phatak this happened some how'. And Prof. Buitenen says "In the end Dhrtar stra rules that the last play was undecisive and that the game as a whole had been neither lost nor won. So Pandavas depart free and still rich men." (p. 30). Prof. Buitenen avoids making any reference to the ill omens and the boons granted by Dhrtarastra and says that Dhrtaristra decided on his own that the game was neither lost nor won. I, for one do not know where in the Sabhaparvan or anywhere else Prof. Buitenen finds basis to make such statements.1 I, on the other hand, feel that we have definite evidence to say that Draupadi's question did not remain unresolved and hence we can assert that the impasse created in the Sabha did not end with the howlings of the jackals and the asses, nor somehow' as Prof. Phatak wants us to believe. We can say that the peculiar situation which developed in the Sabha ended because Draupadi's question was decisively answered. To understand this it is necessary carefully to look at the whole incident the circumstances in which Draupadi's question arose, the persons who were supposed to answer it, the persons who actually did try to answer it, and the answer which finally settled the issue. The narration in the Sabh parvan runs as follows: (1) When Sakuni wins the last game in which Draupadi is staked, Duryodhana assumes that she has become the slave and asks Vidura to bring her over to the Kauravas. When Duryodhana first tells Vidura "Draupadim anayasva" (2. 59. 1) he certainly does not mean "bring Draupadi to the Assembly Hall". He only wants her to be led to the Kaurava apartments to attend to the duties of the menial servants. But, in the opinion of Vidura, Draupadi had not become a slave, hence he does not oblige. (2) Duryodhana then entrusts the job to his messenger Pratikamin. Pratikamin tells Draupadi: "Duryodhana has won you in the game of dice. Hence I have come to take you to Dhitarasira's house to do the household jobs" (2. 60. 4). This message conveyed to Draupadi clearly shows that Duryodhana, to begin with, had no mind to get Draupadi to the Assembly Hall, but wanted her to be taken to the house of the Kauravas. 1. More recently A. Hiltebeitel writing in a paper on "Draupadi's Garments" (II J 22.97, 1980) observes about the fate of this question as follows: "The question remains moot through the entire episode. To the wisest counsellers it is irresolvable, and it drives Yudhisthira to silence, " Madhu Vidya/424

Page #450
DRAUPADI'S QUESTION 183 Draupadi, highly surprised to listen to what Pratika min had said, asks : " What do you say? What prince will stake his wife? Did the king lose his mind while playing, or did he not have anything left to stake?" (2. 60.5) Pratikamin replies: " Wien nothing was left with Yudhisthira, he staked you. First he staked his brothers, then himself, finally you." (2. 60. 6). Even after this clear reply, Draupadi says : "Go, messenger, to the Assembly Hall and ask the gambler ( Yudhisthira): Did you first lose yourself or me?'" (2. 60. 7). It may appear at first sight strange that even when Pratikamin had told Draupadi the order in which Yudhisthira had staked the Pandavas and his wife she asks him to go back to the Hall and ask the same question to Yudhisthira. But it is really not so. Draupadi had known the ordering of the stakes. And although the primary meaning of her question relates to the ordering, what she really demands to know is : If Yudhisthira had lost himself first, did he have the right to stake his wife? If not, how has she become a slave ? This is the first expression of what we generally refer to as Draupadi's Question'. She had first put the question to Yudhisthira--"Whom did you lose first, you or me?" The real import of this question becomes clear from the way in which Pratikamin conveys it to Yudhisthira. (3) Pratikamin returns to the Hall to convey to Yudhisthira the question posed by Draupadi. He tells him: "Draupadi has asked you to answer the following question: "As whose master hast thou lost us? Whom did you lose first, yourselt or me?'" (2. 60. 8). Yudhisthira does not reply. If it was a matter of just the ordering, there was no reason for him to keep quiet. The order of the stakes was clear, and it was known to one and all. The real point at issue was : Did Yudhisthira, who hrd already become a slave of the Kauravas, have the right to stake Draupadi ? When Duryodhana first asked Pratikamin to take Draupadi to the quarters of the Kauravas he had assumed, as probably everyone in the Assembly did, that Draupadi had lost her status as a free woman. But now, for the first time, he realizes that Draupadi does not agree to this position. She disputes the contention that she has become a slave. Duryodhana does not brush aside her viewpoint as irrelevant. He accepts that with Draupadi's question quite a new situation had arisen. He tacitly admits that her question is justified. If Draupadi has any doubt about her status as a slave, it is better, Duryodhana suggests, she came to the Assembly to get her question answered. If it is decided that she has, in fact, lost her freedom she could be sent to do the menial work. Madhu Vidya/425

Page #451
184 M. A. MEHENDALE It is Pratikamin's lot again to take the message to Draupadi." (4) Now we expect the author of the epic to tell us what Draupadi had to say on the message brought by Pratikamin, and also that Pratikamin had to return to the Sabha without Draupadi. But instead of being presented with these details we are confronted with a stanza which says that Duryodhana having observed the faces of the members sitting in the Hall was happy and said something to Pratikamin (2.66.16 tatas tesam mukhan alok ya raja Duryodhanah sutam uvaca hrstah). But we really do not know why Duryodhana should take a look at the faces of the kings sitting there, why he should feel so happy about it, and when did Pratikamin return to the Hall to be addressed by Duryodhana. There is no doubt that there is a gap in the narration, and we have to assume a few events in order to understand the above stanza. What we need to assume would be something as follows: When Pratikamin comes to Draupadi a second time in order to tell her that Yudhisthira does not answer her question and that Duryodhana therefore has asked Draupadi to go to the Assembly, she must have said: "I am not prepared to go to the Assembly. If Yudhisthira does not answer my question, put it to the Kauravas sitting in the Hall, and tell me what they have to say." Draupadi must have shifted the responsibility of answering her question from Yudhisthira to the Kauravas. Pratikamin then must have returned to the Hall to tell the Kauravas that Draupadi is not prepared to come to the Assembly, and expects them to answer her question. But when none of them openes his mouth to give a reply Duryodhana must have felt happy, for he can now ask Draupadi to come to the Hall and, instead of having the messenger go to and fro, get her question answered directly by the Kauravas (2.60.16). Only if we inake a little insertion like this. it is possible for us to understand why Duryodhana feels happy looking at the faces of the Kauravas in the Assembly. (5) When Pratikamin is entrusted with this errand a third time, he hesitates. The Mahabharata tells us that he hesitated because he was afraid of Draupadi's anger (2.60.17). It is likely that he was upbraided by Draupadi when he went to her a second time. Hence instead of doing what Duryodhan wants him to do, he stays in the Hall asking the members, "What shall I tell Draupadi ?" (6) Duryodhana misunderstands Pratikamin's hesitation. He feels that Pratikamin is afraid of Bhima. So he asks Duhsasana to carry out the errand. 1 At this stage there occur two stanzas (2.60.14-15 ) in the Mahabharata which are extremely inconsistent with the narration. We may neglect them for the purpose of this paper. Madhu Vidya/426

Page #452
DRAUPADI'S QUESTION 185 Duhsasana was as if waiting for this opportunity. He immediately rushes to the quarters where Draupadi stays, and instead of simply conveying to her Duryodhana's message, he starts speaking indecently to her. He says: "Draupadi, we have won you. Come to the Hall and, throwing decorum to the winds, have a look at Duryodhana, kurun bhajasva (2. 60. 20)". Draupadi realizes there is no point in arguing with this man, so she hastens to the quarters where the women folk of the Kauravas lived. But Duhsasana restrains her by holding her flowing hair in his hand, and taking no note of her pitiful requests, ruthlessly drags her to the Assembly. Although she has made the issue of the assumption that she is a slave and although Duryodhana has admitted it, Duhsisana insults her by calling her time and again .dasi'. (7) Draupadi is now facing the elders in the Assembly. Duryodhana has got her there so that she may directly put her question to the Assembly, and know her fate. But even before she formally puts the question to the elders, Bhisma assumes she has already done so, and begins his reply : "Draupadi, I am unable to give a decisive answer to your question since the law is subtle. On the one hand, the rules of the game of dice say that one who has been deprived of all his property in the game is not entitled to stake any other property not belonging to him. Since Yudhisthira has lost himself first, Draupadi no longer belongs to him, hence ho cannot stake her. On the other hand, a wife is always dependent on her husband, hence he can stake her. Since I am caught on the horns of the dilemma I cannot decide your case. Moreover, Yudhisthira did not say that there was any deception in the play. (If he did, we could cancel the game and nullify its effect.") (8) Draupadi might have pinned high hopes on Bhisma. But when he disappoints her, she puts her question to the other Kauravas present in the Assembly. When no one opens his mouth, Vikarna, Duryodhana's younger brother, exhorts the assembly members to give their impartial judgement. When his persistent pleas fall on deaf ears, he declares: "Whether you speak out or not, I am now going to say what in my opinion is just. And I declare that Draupadi has not been won (manye na vijitam imam 2.61.24). My opinion is based on the following grounds : (1) Ground No. 1: (i) Four addictions are listed as peculiar to a king, and gambling is one of them'; (ii) When one is under the influence of an addiction, one takes to an unlawful course ; (iii) When Yudhisthira staked Draupadi he had lost control of himself ; | The other three being drinking, women, and hunting. Madhu Vidya/427

Page #453
186 M. A. MBHENDALE (iv) Therefore one may not grant recognition to what Yudhisthira bas done under the influence of his addiction. (2) Ground No. 2: (i) It was not Yudhisthira's own idea to stake Draupadi ; that was Sakuni's suggestion ; (ii) Even if Yudhisthira accepted this suggestion he was not entitled to stake Draupadi since he was not her only husband'; (iii) And even if, for argument's sake, one accepts his right, he could not exercise it since he had already staked himself first and lost; (iv) Hence Draupadi has not become a slave of the Kauravas". Only Vikarna among the Kauravas has pleaded Draupadi's case so cogently. He supports his plea both on general considerations and the particular one--the former being that the society is not bound to recognize the acts of a man who is under the influence of an addiction, and the latter being that a gambler who has lost himself first has no right to stake his wife later. Hence Vikarna's conclusion is that Draupadi is a free woman. The Mahabharata tells us that Karna replied to Vikarna.? But that is not important. What is important is that Duryodhana takes no note of Vikarna's view. (10) Vidura now asks the members in the assembly to reply to Draupadi's question. That was their responsibility, he tells them. However, no one speaks. Draupadi then reminds the members of the traditional law according to which no law abiding (dhrmya ) wife is forced to go to the Assembly to get her question answered (2.62.9). What Draupadi is driving at is that as long as her question--whether she has become a slave or not is not decided she must be treated as a free woman. It was unlawful to force her to the Assembly and make her suffer indignities. It was therefore necessary first to take decision on her social status. She would abide by the ruling given by the Assembly. Even then the members of the Assembly remain silent. 1 As, for example, Arjuna was of Subhadra or Bhima of Hidimba. 2 Hilte beitel thinks that Vikarna's view is based on three reasons, not two as said above. He also feels that Karna in his reply to Vikarna offers a point for point rebuttal" (IIJ 22.98. 1980 ). It is difficult for me to agree to this. To me Karna's reply (2.61.31-36) seems so far removed from Vikarna's argument that I feel that the former must belong to some other version of the Mahabhorata in which Yudhisthira loses his belongings, brothers, himself, and his wife not piecemeal, but in a single game in which he staked his . sarvasva'. 3. At this stage Bhisma once again declares his inability to decide the issue. But that was not necessary. Madhu Vidya/428

Page #454
DRAUPADI'S QUESTION (11) The proceedings thus appear to have come to a deadlock. Draupadi first puts her question to Yudhisthira. When he does not answer, she puts it to the Assembly. Now even the Assembly members do not speak. What is to be done ? At this juncture Duryodhana comes forward and daringly gives a turn to the course of the events. He tells Draupadi: the Assembly members give no reply to you. Let us therefore suppose that you have addressed your question to the rest of the Pandavas-Bhima, Arjuna, Nakula and Sahadeva. If they declare that Yudhisthira was no lord of yours when he staked you, you will be a free woman. If Yudhisthira so feels, let him answer the question even now. All the Kauravas share your grief (sarve hime Kauraveydh sabhayam duhkhantare vartmands tavaiva 2.62.27). 187 All the kings present in the assembly praise Duryodhana and waive their garments to show their approval of what Duryodhana had announced. Then they all turn their necks in the direction of the Pandavas to listen to what they have to say. (12) When the noise made by the kings subsides Bhima begins to speak. He says: "If Yudhisthira had not been our master, we would not have tolerated all this (insult). But Yudhisthira is the master of our meritorious acts, our austerities and lives. If he considers himself won then we too have also been won in dice." (2. 62, 32-33). Bhima's reply is irrelevant. The real issue is whether Yudhisthira, after losing himself, has the right to stake Draupadi, whether in that condition he is any longer her master. This point has not been answered by Bhima. He says that if Yudhisthira considers himself a slave, the rest of the Pandavas are slaves too. But no one has raised the question about the status of the Pandavas, That they were reduced to the state of slaves is well established because when Yudhisthira staked them, he had not lost himself. But Bhima seems to say that it was not even necessary actually to stake the Pandavas; if Yudhisthira for any reason feels himself lost', the Pandavas are lost too. Bhima's reply is totally neglected in the assembly, and it deserved to be, (13) Bhima is followed by Karpa. Actually he has no right to speak now. If at all he wanted to say something he should have done so when Draupadi had put her question to all the members in the Assembly. He does not speak then. Now Duryodhana has put the question specifically to the Pandavas, and it is they who have to answer, not Karpa. But he intervenes and, like Vikarna', argues his case on the basis of certain grounds. He says: 1 Karna carlier refutes Vikarpa's view and now puts forth his own view. Whether the opposition reflected in their names (Karpa and Vikarca) is deliberato or accidental is hard to say. Madhu Vidya/429

Page #455
188 M. A. MEHENDALE (1) There are three categories of persons who are not supposed to own any property (adhana). They are: slaves, pupils, and women.' (2) Accordingly a slave has no wealth of his own which he may use as a stake. But a slave's wife is his property'. (3) Hence Yudhisthira, although a slave of the Kuravas, has Draupadi, his wife, as property which he can stake (2. 63. 1), What Karna is driving at is that although, generally speaking, a slave has no property (and hence cannot participate in a game of dice), if he is married he has his wife as property which he may stake. Hence Yudhisthira is within his rights to stake Draupadi. This he did and lost the game. Hence Draupadi was a slave of the Kauravas. Duryodhana just listens, as everybody else does, to what Karpa has to say. But Duryodhana's decision now no longer rests on what Karna or, for that matter, any one else in the sabha says. He has already declared that the ball was in the court of the Pandavas and that he would abide by their decision. (14) Duryodhana, therefore, completely neglects Karna's reply although the reply was favourable to him. He once more invites the Pandavas to express their opinion. He says: "I shall abide by the decision given by Bhima, Arjuna, or the twins. Let the Pandavas say that Yudhisthira was no longer the lord of Draupadi when he staked her and I shall release Draupadi from the bondage (2, 63, 20; 35 A very critical moment had arrived in Draupadi's life. Her fate-was she to spend the rest of her life as a slave of the Kauravas doing their menial jobs, or was she to lead the life as an independent, honourable woman this question was in the balance. Duryodhana invited the Pandavas to decide the issue, Arjuna did not lose the golden opportunity. (15) No sooner did Duryodhana end his speech, Arjuna began his. He said: "When the game of dice began Yudhisthira was our master. But once he has lost himself, whose master can he be? Kauravas, you take note of this. ( ito raja curvam dsid glahe nah Kuntiputro dharmarajo mahatma | Isas tv ayam kasya pirajitatma taj junidhvam Kuravah sarva eva // 2, 63, 21). Arjuna's reply is quite clear. "Whose master is defeated Yudhisthira ?" Of course of none-not even of Draupadi. I According to Mbh 1. 77. 22, they are a wife, a slave and a son. 2 Before this. Vidura once again (2, 63. 18) expresses his opinion although nobody has asked him to do so. According to him Yudhisthira, once he had lost himself, had no right to stake Draupadi. This he has already told once before (2, 59, 4). Madhu Vidya/430

Page #456
DRAUPADI'S QUESTION 189 This reply of Arjuna resolves the deadlock in the Assembly. One of the Pandavas has clearly declared that when Yudhisthira staked Draupadi he was no longer her lord, (and hence he had no right to stake her). Even if he lost that stake, Draupadi had not become the slave of the Kauravas. After Arjuna's clear reply, it was unnecessary to ask any one else, nor to discuss any further the legality of the matter. As already announced by Duryodhana, the Kauravas accept Arjuna's verdict. Without a monient's delay, Dhitarassfra speaks highly of Draupadi as the lawfully-wedded wife of the Pandavas and his own very special daughter-in-law. Until that moment Draupadi was addressed as a 'dasi' and was humiliated. Against this background the good words spoken by Dhstarastra are particularly striking. Dhstarastra does not stop at merely speaking a few good words to please Draupadi. He offers her a boon. He, as an elderly Kaurava, is perfectly within his rights to do so, and Draupadi, on her part, fully deserves to receive the boon after having been humiliated for some time for no lault of her own. Draupadi asks relief of Yudhisthira from bondage, and when Dhrtarastra offers her another boon, she asks for the relief of the rest of the Pandavas. All the Pandavas had become slaves of the Kauravas and it was necessary to secure their freedom. Draupadi does not ask for her own freedom because that is no longer necessary. After the reply of Arjuna it is accepted that she has not become a slave. If she is not a slave, why ask for her freedom? Hence when Dhstarastra offers Draupadi a third boon she politely declines it saying that only two boons are considered proper for a Ksatriya woman. When one carefully observes this sequence of events which occurred in the Assembly Hall one realizes that the turn which the events took was entirely due to the reply given by Arjuna. If that were not the case, and if the turn was given by the bad omens, we will have to assume that the basic question-whether or not Draupadi had become a slave-remained unresolved. In that situation Dhstarastra's honouring her as a lawfully-wedded wife of the Pandavas and his own special daughter-in-law would be inappropriate. Moreover, in that case it would be necessary for Draupadi to ask her own freedom from slavery, since that question had remained undecided. But since nothing of the sort happens it is clear that nothing in the Assembly had happened due to the bad omens. What happened was due to the decisive reply given by Arjuna, and to the fact that this verdict was accepted by the Kauravas. But the version of the Mahabharata available today makes a meption of the bad omens immediately after Arjuna's reply. It is also said that Vidura and Gandhari, apprehensive of the ill omens, approached Dhstarastra and requested him to intervene. Naturally it has been assumed that Dhstar stra took the matteres in his hand because of the occurrence of the bad omens. Nobody has ever sensed that what happened was due to Arjuna's reply. The importance of Madhu Vidya/431

Page #457
190 M. A. MEHENDALE the stanza put in Arjuna's mouth has been overlooked' and the stanza has been incorrectly interpreted. The latter half of Arjuna's stanza runs as: isas tvayam kasya parajitatma taj janidhvam Kuravah sarva eva (2.63.21 ) This was translated long ago by P. C. Roy as follows: "But having lost himself, let all the Kauravas judge whose master he could be after that ". And more recently (1975) Prof. van Buitenen translates it as: "But whose master is he who has lost himself ? That you should decide, ye Kurus assembled ". (italics mine in both translations ). Both the translators have rendered janidh vam with judge' or decide' which is not correct. It would mean that Arjuna once again passed on the responsibility of answering Draupadi's question to the Kauravas. But Arjuna, a slave, has no right to do that. Duryodhana alone, as the victor in the game of dice, has the right to decide who should answer Draupadi's question. The linguistic usage in the Mahabharata will show that it is wrong to translate janidhvam with judge' or 'decide'. After Draupadi raised the question about her social status, whenever there is an occasion in the Mahabharata asking someone to decide her question, we find the use of the expression prasnam bru, the verb bru being used with or without a prefix. When Bhisma, for instance, says that he is unable to answer the question decisively one way or the other, he says : "tasman na te prasnam imam bravimi (2.60.42). Draupadi requests the members of the Assembly : vibrata me prasnam imam jathavat (2.60.45). While reminding the Assembly-members that it was their responsibility to decide the question Vidura says : na ca vibruta ( ? tha) tam prasnam sabh ya dharmotra pidyate (261.52)? When Duryodhana wants Yudhisthira to decide the issue, he says: prasnam prabruhi kssnam tvam ajitam yadi manyase (2.63.9 )3. Sometimes we find the expression vak yam vi-bru in place of prasnam vi-bru. Vikarna, for instance, says : yajnasenya yad uktam tad vakyam vibruta parthivah (2.61.12); or, occasionally, we find prasnam vi vac" in place of prasnam vi-bru. Bhisma tells Draupadi : na dharmasauksmyat subhage vivaktu saknomi te prasnom imam yathavat ( 2.60.40 ; 2.EUR1.55 ). From the above passages it becomes clear that the epic usage for conveying the meaning to decide the issue, to reply to a question' was prasnam vi )bru.. 1 F. Edgerton (op. cit.) even goes to the extent of saying that none of the Pandavas ventured an opinion when Duryodhana asked them to decide the issue ! 2 Similarly 2.60.42 ; 2.61.50, 54, 55, 56, 66, 67 ; 2.61.14 ; 2.62.27. 3 Also 2.61.19. 4 The root vic-, in place of vac-, or noun derived from it is found in 2.61.12 ; 2.62.16. Madhu Vidya/432

Page #458
DRAUPADI'S QUESTION 191 Hence the renderings in the above English translations of janidhvam as if it was vibruta, are wrong. Arjuna wants the Kauravas to realize (jaaidhvam) that Yudhisthira, in the circumstances, could not be the master of any one ; Arjuna is most certainly not asking them to decide the unsettled issue. Some Maratbi translations of the above passage are as follow: The translation of the Chiplunkar Mandali may be rendered into English as: "Arjuna said: "Kauravas, it is clear that this Dharmraja, son of Kunti, was within his rights to stake us before he lost himself. But once he lost himself, whose master could he be ? You too all know this. Hence it is needless for me to say that Draupadi has not been won." This is not the translation of the stanza, but its paraphrase. However it is not wrong like the English translations. It states correetly what Arjuna wants to convey: "Draupadi has not been won." The other Marathi translation in the one from the Mahabharata edited by Shri C. V. Vaidya? It runs as follows: " First, when he staked, the nobleminded Dharmaraja, the son of Kunti, was our master. When he himself was won (in the game of dice), whose master was he? Ye all Kauravas: You realize this.3" In my opinion this translation is correct. But the translator has failed to recognize its importance and derive the necessary conclusion. In his concluding remarks ( samalocana ) at the end of the volume (p. 25 ) Shri Vaidya observes : "After Yudhisthira staked himself and lost, he became the slave of the Kauravas, Then, at the suggestion of Sakuni he staked Draupadi. Sakuni won the game and hence Draupadi became the slave of Duryodhana". This means that in the opinion of Shri Vaidya whether Draupadi had become the slave or not is not in dispute at all. Although she made an issue of it, Draupadi was, in Shri Vaidya's opinion, a slave. In that case Duryodhana's open invitation to the Pandavas to answer Draupadi's question and Arjuna's reply become moaningless. There is an entry on' Arjuna ' in the index supplied by Shri Vaidya at the end | arjuna mhaNAlA : --he kauravaho, hA kuMtIputra dharmarAja, ApaNa haraNyApUrvI AmhAMlA paNAlA lAvAvayAsa samartha hotA he ughaDa Ahe, paraMtu svatAcA dehaca haravUna basalyAnaMtara maga to koNAcA mAlaka asaNAra? he sAre tumhAMlAhi mAhita Aheca. tevhA draupadI jiMkalI gelI nAhI, he mI ANakhI teM kAya HTTTT? ( g. Yin) 2 Published by Damodar Savalaram and Co., Part 3 ( 1933). 3 arjuna mhaNAlA:-"pUrvI jevhA AmhAlA paNAlA lAvile tevhA kuMtIputra mahAtmA dharmarAja. AmacA svAmI hotA. to svataH ( dyUtAmadhye ) jiMkalA gelyAvara maga koNAcA svAmI ? sarva kauravAMno ! e sala ezi." (p. 356 ) 4 "yudhiSThirAne svatAMsa paNAsa lAvUna dAsa jhAlyAvara zakunIcyA sUcanevaruna draupadIsa paNAsa lAvile va tohi paNa zakunInaM jiMkalyAne draupadI duryodhanAcI dAsI jhAlI." Madhu Vidya/433

Page #459
192 M. A. MEHENDALE of the volume. Under this head we find mention of many other things done by Arjuna, but there is no reference to his reply. As a matter of fact this reply of Arjuna is so important that it should find a place in any entry, big or small, on Arjuna. Just as by hitting the bull's eye at the time of Draupadi's self-choice he won her, similarly at the critical moment in her life he gave the decisive reply and saved her from utter humiliation. What importance attaches to the correct shot of the arrow, the same, or even more, attaches to this straight-forward reply. When Duryodhana invited the Pindavas to give a reply he must have been sure that none of the Pandavas will dare say, "defeated Yudhisthira was no longer the master of Draupadi ". If we look to the way Bhima made his reply, Duryodhana's confidence was perfectly justified. But Duryodhana kept his word although Arjuna's reply was not what he had expected it to be. He did not set aside the verdict given by Arjuna, nor did he object to Dhrtarastra's giving boons to Draupadi. If he had objected, the Pandavas would not have been free from bondage, although Draupadi was admitted to be a free woman, But Duryodhana does nothing of the sort and accepts the result of his bold offer. Someone in the line of the epic transmission did not want this fairness on Duryodhana's part to be observed by the posterity. Perhaps, he also did not want the posterity to know that a woman fought for her right and won. He quietly introduced the incident of the bad omens immediately after the roply of Arjuna and successfully switched off the attention of the listners (and, later, of the readers) from the stanza containing the reply. The interpolator has had no doubt tremendous success in creating the impression that the proceedings in the Assembly after the first game of dice end the way they do because of the ill omens, Our minds are so much obsessed by the fear of bad omens that we never thought that there was anything wrong in assuming that the events in the Assembly took a different turn due to the fear of bad omens. But if we are really to believe this incident of bad omens, one shudders to think of the indignities that would have been heaped on Draupadi in case the jackals and the asses moving around--one fails to know what business these animals had in Dhitarastra's Agnihotrasala had not raised their voice at a very critical moment in her life. She had suffered enough in the Assembly of a famous royal dynasty. She was addressed as a lave' (dasi )', abused as a 'harlot' (bandhaki). She was continuously cragged at her hair causing her unbearable pain. If at this juncture the above-named animals had not risen to the occasion, 1 2, 60. 27, 37. 2 2. 61. 35. 3 The Mahabharata tells us that the Assembly witnessed two very degrading incidents which one would be ashamed to repeat. I do not mention them since in my opinion they are unauthentic. Madhu Vidya/434

Page #460
DRAUPADI'S QUESTION 193 the Pandavas would have remained life-long slaves of the Kauravas and Draupadi would have swept all her life the floors of the Kaurava houses. The Mahabharata would have ended there. But we know that it does not. The events in the Mahabharata continue. Who is responsible for this continuation ? Who has the credit ? We do not give this credit to some eminent ascetic, nor to the virtues of a faithful wife, nor to Sriksna, the incarnation of a god. We choose to give it to a jackal and an ass, who must have been blissfully ignorant of what was going on around them. The Indian tradition has touched a very low level in allowing the stanzas about the bad omens to remain where they are for so long. We are unaware of the fact that in doing so we have tarnished the fair image of an eminent person like Sri Vyasa. If anything were to happen in the Mahabharata just because of bad omens, one may ask--why did the same thing not happen when the Pandavas started on their exile after losing the game of dice a second time? Then too the ill omens occurred, many more than on the previous occasion. The animals and birds like vultures, jackals and crows, of course, shrieked; besides, the lightnings flashed, the earth shook, Rahu unseasonally swallowed the sun, and meteors shattered after making a round of the town the wrong way (2. 71. 25-28). Vidura did draw Dhstar stra's attention to these omens. But Dhstarastra did not get cold feet and ask the Pandavas not to proceed on exile and return to the capital The reason is obvious. Either the ill omens did not occur at all, what appears in the text of the Mahabharata is a figment of some interpolator's imagination, Or, even if they did, Dhstarasta was not so timid as to get panicky and do something he would not have otherwise done. The main purpose of this paper is to make clear that it is wrong to assume that the discussion in the Assembly over Draupadi's question remained inconclusive and that the Pandavas and Draupadi regained their freedom simply due to the accidental occurrence of bad omens. The events in the Assembly end the way they do because Arjuna gave a decisive reply to Draupadi's question and, consequently, Dhstarastra gave two boons to Draupadi. It may now be permitted to refer to a few points incidental to the main topic. The first is: why did Yudhisthira not reply to Draupadi's question. One reason could be that he felt the same difficulty as was felt by Bhisma and hence he did not reply. Or, he was convinced that he had a right to stake Draupadi even after he had lost himself, but he dare not say so openly in order to save Draupadi from the indignities which would have followed. Or, he knew he had no right to stake her but admitting that would have meant that he had done wrong when he staked something on which he had no right. He did not have enough courage to do that. Madhu Vidya/435

Page #461
194 M. A. MEHENDALE The second point is with reference to Vikarna's reply. Before Arjuna gave his decisive reply, five persons - Vidura, Bhisma, Vikarna, Karna and Bhimahad replied or attempted reply Draupadi's question. Of these the replies of Vidura, Bhisma, Karoa, and Bhima were paid no attention to - and this on good or plausible grounds: Vidura's reply went unheeded probably because he was dasiputra and also because he was known for his partiality towards the Pandavas ; of Bhisma, because he could not decide the question one way or the other ; of Karna, because he spoke out of turn; and of Bhima, because the reply was irrelevant. But this cannot be said of Vikarna's reply. He, as a Kaurava, replied when the question was addressed to the Kauravas ; his reply was straight to Draupadi's question ; it was unambiguous, and was based on certain grounds which he detailed in so many words. And yet Duryodhana completely disregarded it! To say that Duryodhana neglected it because Vikarna was younger to him,' or because the reply was not favourable to him is not convincing. It is likely that Vikarna's reply had no effect on the outcome in the assembly because he did not find any Kaurava to support him. But more than this it is not possible to say at this stage. And the third point that needs consideration is the way Bhisma deported himself throughout the Assembly sitting. One cannot but say that his attitude towards Draupadi's question was unbecoming of him; and since Draupadi was not only insulted with abuses, she was also a victim of molestation, his attitude must be judged unpardonable. When asked a question, one may reply to it if one knew the answer. If not, is it necessary to say in so many words: 'I am not in a position to answer the question'? Assuming that there are occasions when this becomes imperative, was it necessary for Bhisma to do so twice ? Moreover, on the second occasion he shifts the responsibility of answering the question to Yudhisthira and feels himself free. Did he not know that Yudhisthira, when asked by Draupadi and by Duryodhana to answer the question, had not done so ? Hence the attitude of Bhisma towards Draupadi's question scems quite improper. If Bhisma on his part felt that the question raised by Draupadi was insolvable. should he then not have given the benefit of doubt to Draupadi ? He was face-to-face with an impudent Duhsasana constantly tugging at the hair of Draupadi. The moment Bhisma saw this, he should have told Duhsasana that he was in the Assembly of civilized Ksatriyas and not in the den of hooligans. He ought to have commanded Duhsasana to leave Draupadi alone. The Grand Sire of the Kauravas did not rise to the occasion and hence his behaviour becomes unpardonable, 1 Vikarna was considered to be one of the four important (pradhana ) Kauravas 1.90.62. Madhu Vidya/436

Page #462
THE FLORA IN THE ARANYAKAPARVAN OF THE MAHABHARATA BY : M. A. MEHENDALE In continuation of "A Cultural Index to the Mahabharata : Tentative Specimen Fascicule" ( ABORI 66. 117-152, 1985) is being published here the information on the plants and trees available in the Aranyakaparvan of the Mahabharaia. It will form, according to the scheme of classification given in the above 'Specimen', entry No. 4. 5. 8: plants and trees'. Tho details given here under each head will naturally be augmented when more information will be available from the rest of the parvans. In the mean time the present entry will serve the purpose of giving the reader an idea of the kind of information he will get on this subject from the Mahabharata. 4. 5.8 Plants and trees ak sa - nut trece, on the banks of the Sarasvati river 3. 174. 23. ajaraka - 1rces, on the Gandhamnadana mountain 3. 155.40. atasi puspa - 'cornlower' (B), for comparing the colour of the skin of the divinc child seen by Markaqdeya 3.186. 86. ambuja - Jotuses', yellow (harita) in colour in the pond on tbe Kailasa mountain 3. 151. 3. arisja -- soapberry trees' (B), growing in the forest 3. 61. 3. arka - trecs, their flowers used for worshipping the five Ganas ( who sprang from Siva's semen) by those who desire wealth and freedom from discases 3. 220. 14. arjuna - trees, growing in the forest 3,61.3; in bloom at the cod of the summer in thc Dvaitavana 3. 25. 17. alabu - ' bottle gourd', from its seeds 60000 sons of king Sagara born 3. 104. 18. asoka -- tree, called tarusrestha 3. 61.97; seen by Damayanti in full bloom in the forest, shining with its chaplets it looked like king of Dramida 3.61.98; pun on the word - yatha visoka gacche yam asokanaga 1 B = J. A. B. van Buiteneo's traoslation. 30 (Annals BORT) Madhu Vidya/437

Page #463
23-4 Annals BORI, LXVII ( 1986) tai huru / salyanama bhara soka mama sokavinasanat 1/ 3.61. 102 (also 3. 61.99); growing on mountain 3.61.38; on mountain Gandhamadana 3. 155. 47; Sveta mountain full of their woods 3. 220.23; found in the liermitage of Rsyasraga 3. 111, 16; asokagrove ( vanika ) ia Laika where Sia was held captive by Ravana 3. 264.41; 3. 263.3; 3. 273. 27. asrantha 'fig trees', growing in the forest 3. 61.3; on the Gandhamadana m. 3. 155.12; was to be embraced by Satyavati's mother to get a son 3. 115. 23. anulaka -'myrobalan trees', growing in the forest 3. 61. 4; on the Gandha. madana mt. 3. 155. 42; its ripe fruit offered by R$yasriga to the couriczaa 3. 111. 12. umra - mango trees,' growing in the forest 3. 61.4; on the Gandhamadana Int. 3. 155.40; in bloom at the end of summer in the Dvaitavana 3. 25. 17; mango groves on the river Narmada in Avanti 3. 87. 2. Amravetasa -mango-canc' (R), growing on the Gandhamadana mt. 3. 155. 41. umratako 'hog.plum', on the Gandhamaldan: mit. 3. 155.40. Trigundu - trecs, growing in the forest 3.61. 3; on the Gandhainadana mt, 3. 155.43; grow on the banks of Sar.svati 3. 174.23; its rips fruit offered by Rsyasriga to thic courtean 3. 111. 12. Indt rara bluc lotus ', found on the Gandham dana mt. 3. 155.4). islka - roed', thrown by Ranna at the crow 3. 266. 67. upala - blue lotus', in the lakes and rivers on the Gandhamadana mt. 3. 155. 49, 67; on the Malyavant mt., ils fragrance smelt by Rama 3. 266. 3; in the lake lampa 3. 264. 1; in a similc ( lotuscs shattered in a fight between the elephants ) 3. 12. 43. udumbara - frces, growing in the forest 3. 61.4; on thc Gandhamadana mt. 3.155-42; was to be cmbraced by Satyavail to get a son 3. 115. 23. osadhi -- plants' (general). when the suo enters the carth and takes the form of a field, the moon (lit. the lord of the plants) produces the plants with heavenly heat turned water and her rays; described as having six flavours ( sadrasah),' sacrificial' (medhyah); they form the food of the living beings 3. 3. 7-8. kanaka - 'thorn', 3. 282. 5: nihutakanlaka- (free from opponents ), a mode of expression 3. 294.11. * B. of the six flowers (a misprint?) Madhu Vidya/438

Page #464
MEHENDALE : The Flora in the Iranyakaparvan of the Mbh. 235 kadamba trees, grow in the forest 3. 61. 4, in the Dvaitavana, in bloom at the end of the summer 3. 25. 17; Sveta mt. full of their groves 3. 220. 24; at the entrance of sage Troabindu's hermitage in the Kamyaka forest 3.249.1; Lohitayani, a nurse (dhatri ) of Skanda, is worshipped on kadamba tree 3. 219. 39. kadall banana tree', a grove (sinda) of them, tall like palm trees, extending over many yojanas on the peaks of the Gandhamadana mt. (kadali skandhan bahutalasamucchrayan) 3. 146. 42, 44; a grove of golden kadalis on the bank of a lake on the Gandhamadana mt. 3. 146. 53; Hanuman sleeps in the midst of banana grove 3. 146. 60; the grove called vana 3. 146. 63, 64; Bhima meets Hanuman in the kadalisanda 3. 149. 4; kadall used for comparisons (falling down of a shaking kalad!) 3. 144. 4, 3. 275. 14; used in a mode of expression (bear fruit and get destroyed) yatha... kadall phalanty abhavaya na bhitaye Imanah 3. 252. 9 (cf. nala and venu). kaplitha trees, growing on the Gandhamadana mt. 3. 155. 42. kamalalotus flowers, growing in the lakes on the Gandhamadana mt. 3. 155. 49, lake Pampa full of them 3. 264. 1; used in comparisons (kamalayataksi 3. 249, 6; kamalek sana 3. 292. 23; kamalapatrakca 3. 275. 3; kamalagarbhahha 3.293. 10; vidhrastaparnakamala padmini 3. 65. 14). karanja -tree, where the mother of the trees lives; people desirous of having progeny worship her on this tree 3. 219. 34. karabha 'vermilion (B), for comparison (karabharunagatra) 3. 268. 25. karavira -- oleander tree', on the Gandhamadana mt. 3. 155. 43; Sveta mt. full of their woods 3. 220, 23; a lake in the Dvaitavana surrounded by them 3,296, 43. karira bamboo shoots", grow on the banks of Sarasvati 3. 174. 23. karusaka -see parazaka. karnikarajasmine' (B), in bloom in the Dvaitavana at the end of the summer 3. 25. 17; on the Gandhamadana mt. 3. 155. 45, looked like the mountain's ear-ornaments 3. 155. 58. kalpavrksa mythical wish-yielding tree, 3. 265. 5. kahlara 'white lilies, in the lakes on the Gandhamidana mt. 3. 155. 49. kaLlcanadruma --? the guise (veza) of a Kirata compared with a kancana druma 3. 40. 2. 3thorns (B). Madhu Vidya/439

Page #465
236 Annals BORI, LXVII (1986) kaliyaka - turmeric trees ', in the regions near Himavat 3. 175. 10. kasmari - trees, in the forest 3. 61. 4; on the Gandhamijana ml. 3. 155. 42. kinisuka - trees, in the forest 3.61.3; on mountain 3.61.38; on the Gandha. madana mt. 3. 155.46; dinavas with their gold ornaments etc, looked like kinsuka trces 3. 103. 11; Valin and Sugriva in fight, covered with blood, coiopared with kinsuka in bloom 3. 264. 32. kutaja --- trees, on the Gandhamadana mnt. 3. 155.45. kunda - jasmine Rower', uscd as a standard of comparison for whiteness 3. 119.4. kumuda - night lotus' (D), appear in rivers and ponds in autuma 3. 179. 13; in the lakes on the Gandham daou mt. 3. 155. 49; on the Malya vant mt., its fragrance spielt by Ruma 3. 266. 3. kurabaka 'red amaranth', on the Gaudhamidana ml., wbeo is bloom look like the arrows of Kama 3. 155. 58. kusa - puialed (sacred) grass, 3.282. S; used for spreading tbe carth (samistura ) when Rama sits for upavasa 3.267.32; Dyumalscoa used a seat (b?si) made of it 3. 279.4; Duryodbaca puis on kusa garments when he sits for praya 3, 239. 17; Draupadi, living in the forest, has an ulturi ya made of it 3. 250. 1; used for Roma's abhiscka 3. 263. 30. kciaka - jasniinc (70) Irccs', a like in thc Dvaitavaca sorrounded by them 3. 296.43; grow on the Gaudhaniadana mt. 3. 155. 44, 45. kokonudu 'red willer-lily llowers in the lakes on the Gandhamadana mt. 3. 155.49. kovidara - trecs, ou the Gandhamadaoa mt. 3. 155.46. ksirin milky trees, on the Gundlamidana mt. 3. 155. 42. k saudra - campaka trocs (B) on the Gandhamadaoa mt. 3. 155.41. kliadira - trees, in the forest 3. 61. H; grow on the banks of the Sarasvati river 3. 174.23; used for making spikes (faniku) inserted in the moals round Lank: 3.268. 3. kharjura - trecs, ia the forest 3. 61. 5; grow on thc Gandhamadapa mt. 3. 155.41. candana -- sandalwood trees', in the forest 3. 61.3. campaka - trees, on the Gandharoidana mt. 3. 155.44. cait j'adruma --'asraliha tree', growing in the cremation grouod, used for comparison 3. 265. 5 ( terrible in appearance though decorated ). japu the China rose, Sveta mt. full of their woods 3. 220.23. Madhu Vidya/440

Page #466
MEHENDALE I The Flora in the ranjakaparvan of the Mbh. 231 jambu - rose apple trees', in the forest 3.61.4; on the Gandhamadana mt. 3. 155. 42. jira --- frees, on the Gandhamadada mt. 3. 155.40. Tamala - trees, on the Gandbamadana mt. 3. 155.46. tarala - thorn apple trees', on the Gandhamadana mi. 3. 155, 46. Tamarasa-day lotus', in the lakes on the Gandhainadada mt. 3. 155. SI. rala - fan-palmin the forest 3. 61. 5; in bloom at the end of the summer in the Dvaita yana 3. 25.17; on the Gandhamadana mt. 3. 155.46; used as a weapon by the monkeys 3.260. 12; 3. 264. 30; 3. 267. 18: for comparison (heads falling on battlefield compared with palm fruit falling from the stems) 3.99.5; as a standard of tallness 3.297.20; occurs in a maxim ( sosayis yani gatrani vyall talagora Jatha ) 3.264. 50; Tolajangha, name of an asura 3.287, 17. rinduka - trecs, in the forest 3.61.3; on the Gandhamadana mot. 3. 155.40; called mahaphala 3. 155.43. tilaka - Trecs, in the hermitage of Rsyasroga 3.111.16; on the Gandha. midada mt. they look like the forehead marks on the long tracks of the forest 3.155.59. tilakanda - joint of sesame', occurs in a simile (cutting off an arm like cutting off a sesa me joint ) 3.263.33(bhujan... nikilas tilakandarat). lunga - trees, (kostharisesa - Nila.) in the regions near Himaval 3. 175. 10. darbha -- grass, spread on the ground while sitting down for prdya 3. 239.16. dadima - 'Pomegranate trees', on the Gandham dana mt. 3. 155. 40. deredaru - pise trees', on the Gandham.dana mt. 3. 155.46; in the regions near Himavat 3. 175. 10; produced from his bones by Agni whilo cptering the carth 3.212. 13. dhanvang - fruil, offered by R$yasriga 10 the courtezao 3. 111. 12. dhava - trees, in the forest 3. 61. 3. nala - reed', occurs in mode of expression ( yatha ... nalo va phalanty abhavaya ) 3.252.9 (cf. kadali and venu). nalina - lotus', a lake with tliesc Bowers on the Vaiduryasikhara mi. 3.87.5. nalini - lotus creeper', in the lake in the Dvaitavana 3. 296.43; * lotus pond' of Kubera. on the Gandhamadana mt. guarded by the roksasas 3. 142. 24; 3. 151. 1; a lotus-pond', called Pamp: 3.264. 1; I significance not known. Madhu Vidya/441

Page #467
238 .. Annois BORI, LXVII (1986) occurs in a mode of expression (molestation of a woman considered like a jackal jumpiog into a lotus pond ) 3.253. 19. narikela - cocoa-nut trees', on the Gandhamadapa mt. 3. 155.40. nipa - trees, blooming towards the cod of summer in the Dvaitavada 3. 25. 17; on thc Gandbamadana mt. 3. 155. 41; on the banks of the river Sarasvall 3. 179. 14. nivara - 'wild rice', growing along the banks of Sarasvati river 3.179. 14. nyagrodha banyan tree', has cool shade 3.295.15; growing in the forest 3. 61. S; scen by Markandeya while noating on the water at the time of the yuganta 5. 186.81. padma - lotus', a lake fragraot with them 3. 62. 2; fouad in the lakes and rivers on the Gandhamadana ml. 3. 155. 51, 53 (padmasanda), 67; seco by Bbima on tbc Gaudhamadapa mi. 3.150.19, 23, 25 ( golden ); lakes with lotus flowers on the Malyavant mt. 3. 155. 33; Rama smells their fragrance on this mot. 3. 266.3; scals ( mudrah) marked with lotus and lotuscs marked with trisula round at the Piqdaraka tirtha 3. 80. 83-84; a lotus-like birth-mark 3. 66. 5; colour of lotus for comparison 3. 81, 170; 3. 82. 52; 3. 155. 77; 3. 292. 19 (ranra), and also its width ( Ojala ) 3. 292. 19; (scc padmapalasa ); lotuscs arc crushed in fights between clcpbants 3. 12. 48. padmakosa bud of a lotus', for comparing palms of a hand 3. 13. 109. podniopalasa - 'lotus of Icar', for comparing eyes 3. 65. 11; 3. 277.27. padmasougandhika - a special kind of very fragrant goldeo lotus with beryl stoms and of many colours found in the lakes on int. Kailisa 3. 151. 5-6; also found in the lakes in heaven 3.164.46; (sec saugandhika. ). padmini lotus lakc', on mt. Himyant 3. 107.9; on the Gandbamadapa mi. 3.150.23; used in comparison ( vakulam iva padminim) 3.65.14. panasa bread-fruit trec', on the Gandhamadana mt. 3. 155.41. parusako 'kind of fruit, offered by R$yasriga to the courtezan 3. 11L12. palasasanda cluster of palasa tress', form a land mark where the roads in a forest bifurcale 3. 281. 107. palala - Bignonia Irccs', on the Gandbamadana mr. 3. 155. 45. palali --Bigaonia trccs, on the Gandbam dana mt. 3. 155.62. Iv. l. karlak.. Madhu Vidya/442

Page #468
MEHENDALE : The Flora in the Aranyakaparvan of the Mbh. 239 padapa - trees (not named), with blue and radiant colour in the Dvaita forest 3. 296. 41 (see vrksa). paravata - trees, on the Gandhamadana mt. 3. 155. 41. parijata - trees, on the Gandhamadaoa mt. 3. 155. 46; on the Sveta mt. 3. 220.23. pippala - trees, growing around a lake in the Dvaitavana 3. 296.43. pilu - trees, grow on the banks of the river Sarasvat! 3. 174.23. pundarika -- white lotus', appear in rivers in autumn 3.179.13; in the lakes on the Gandhamadana mt. 3. 155. 49; for comparison (of an eye) 3. 264. 49. pumnaga - trees, on a mountain 3. 61.38; on the Gandhamadana mt. 3. 155.. 45. puskara - Jotus flower, at the Sasayana tinto the flowers have the form of a rabbit (?) ( Sasarupapraticchanah) 3.80.120; gold-coloured puskaras in the pood near the residence of Kubera on the Kaildsa mt. 3. 151. 3. puskarini- lotus fake' near the residence of Kubera on the Kailasa mt. (3.151.5; 3. 152. 18.21 ) and those in the heaven (3.164.46 ) full of saugandhika flowers. priyangu - trees, grow along the river Narmada in the Avanti country 3.87.2. priyala - trees, in forset 3. 61. 5; on the Gandhamadana mt. 3. 155. 46; its fruit offered by R$ya fraga to the courtezon 3.111. 12. plaksa - fig tree', the river Sarasvati Nows from it 3.82. 5; they grow along its hanks 3. 174.23; in forest 3.61.4; on the Gandhamadada mi. 3. 155.42; ( Plaksavatarana tirtha on the river Yamuna 3. 129, 13). bakula - trees, on mountain summits 3. 61.38; on the Gandhamadana mt, 3. 155. 44, 46, 62 ( bakulani ). badari - jujube trees', in forset 3. 61, 3; on the Gandham dana mt. 3. 155. 42; on hanks of the river Sarasvall 3.174.23; a large badari free, beyond the Uttara Kurus, near mt. Kailasa, where the asrama of Nara and Narayana is situated 3. 145. 10, 17-19. bibhitaka - trecs, in forest 3. 61. 5; on the Gandhamadana mt. 3. 155. 42; bearing fruit 3.70.6; Kali enters it 3. 70.34; hence becomes inauspi. cious (aprasasta ) 3. 70.36. bilva - 'wood-apple' trees, in forest 3. 61.5; on the Gandhamadana mt. 3. 155. 42; on the banks of the river Sarasvati 3. 174.23. bijapuraka - citron trees', on the Gaadhamidana mt. 3. 155, 40, Madhu Vidya/443

Page #469
240 Annals BORI, LXVII ( 1986 ) bhalla taka - marking-nut trees', on the Gandhamadada mt. 3. 155.42; cashew-out fruit offered by R$yasroga to the courtezan 3.111.12. madhurang 'forest of honcy(-yielding ?) trees', ocar Kiskindba, guarded by Valin and then by Sugriva 3. 266. 26. madhuka 'arrac trccs'in bloom at the end of summer 3. 25. 17. mandara - coral trees', on the Gandbamidana mt. 3. 155. 45. masa - bcaps, raksasas compared with their heaps 3.268.34. mwilja -- grass, 3. 12. 49 (for comparison, munjavaj jarjari bhutah padapah). muru - plant, sparcs made of muru (mouravah pasah)' destroyed by Krsoa and the road to the town Progjyotisa cleared up 3. 13. 26. mula - rools. (Damayanti lived og roots 3. 62.26). mrnall - lotus plant,'. ( for comparison ) 3. 65. 12, 15. moca -- trees, on the Gandhamadana int. 3. 155.41. yarasa - grass, growing in the pond 3. 62. 3. rajiva - blue lolus ', for comparison (cycs ) 3. 277.23. rouhitaka - trecs, growing along tbc bank of Sarasvati 3. 174. 23. lik uca - frecs, on the Gandhamadana ml. 3. 155.41. lodhra - trees, growing in forest 3.61.4. vala b anyan trocs', on the Gandhamidana mot. 3.135.42; a rata tree on the Gayasiras mi., known as Ak sayyakarand, where food givco 10 the mancs never gets cxhausted 3. 85. 8. ranira a sort of canc', grows in large oumbers on Ibc river Narmadi in Avanti 3.87.2. (rksa) - trecs', unnamics, haviog Bowers whicb looked like gold or forcst.fire, which were red or dark like collyrium, aod were like bcryl sound on the Gandhamidana mi. 3. 155. 61; tbe ml. also had shady ficcs (unnamed) blossoming and giving fruit in all tbe scasons 3. 143.3; (scc padapa ). renu -' bamboo', growing in forest 3.61. 3; occurs in a mode of expression (in vitcs its own destruction ) 3.252. (cf. kadali aod nala ); for comparison (sound produced while striking one's arms in a fight compared with the bursting of a bamboo ) 3. 12. 58. vclasa - cane'. growing in river 3. 61. 107; near a lake in the Dvaitavada . According to B. Mauravas aod Pisas are names of tribes. 1 B. lotus sialk' lutus'. Madhu Vidya/444

Page #470
MEHENDALE: The Flora in the Aranyakaparyan of the Mbh. 241 3. 296.43; on the banks of Sarasvati 3. 174. 2.3; A mythical fragrant relasa irce of a thousand branches from whose root flow rivers of honey and water 3. 184.23. vetra - large reed', in forest 3.61. 4. saraparro lolus ', for comparing eyes ( apara) 3. 65.20. sami - trees, on the banks of Saraivali 3. 174. 23; its cultings (familava-) used for the abhiscka of Rama 3.263. 30. Sara - 'roed', Sveta mt. covered with it ( Sarastamba) 3.24. 10; colour or monkeys compared with it (whiteness, Saragaura - ) 3. 268. 27. saka leak trees. growing in forest 3. 61. . sala - trees, growing in forest 3.61.3; on the Gandhamadana mi. 3. 155. 46,62; in bloom at the end of summer in the Dvaitavana 3. 25. 17; used as a weapon during lights by Valin and Sugriva 3. 264.30; by Sugriva 3.271.7-9; by Angada 3.272. 17; by inonkeys 3.260.12; 3. 267. 18; Dyumatsena silling under it 3. 279, 4; used for com parison (tallness ) 3. 36. 24; 3. 263. 25 ( salushundhu- ): 3.297.66. sall - 'rice, vanaras compared with shoots of sali 3. 268. 27; salibhavana rice field', ariny of monkeys compared with it 3. 267. 18. salmali - silk-colon irce', growing in forest 3. 61.3 (sasulmalaih); on the Gandhamidana mt. 3. 155.46; globular growth ( asthili ) round a salmali trec not indicative of its growth 3. 133.9. finisara - sissoo trees' (B), on the Gandhamndana mt. 3. 155.46. Sirise trees, on the hanks of Sarasvall 3. 174. 23; colour of monkeys like its flowers 3. 267. 10; 3.268, 27. santanaka - trees, on the Sveta mt. 3. 220.23. saptaparna --- trees, on the Gandhamadana mt. 3. 155.45. sarja -Irces, in bloom at the end of the summer 3. 25. 17; found in the hermitage of kSyafraga 3. 111. 16; powder (panisu ) made from sarjarasa used as impediment to make the ditch round Lanka dificult to cross 3. 268. 4. sahakara - mango trees', on the Gandhamadana mat. 3. 155. 60 ( look like arrows of the God of Love ). sindhuvara -- trees, around a lake in the Dvaitavana 3. 296.43; full of goldea fowers, on the Gandhamadana mt., they look like the lance (tomara) of the God of Love 3.155. 57. soma - pressed on the bank of the Payosoi river 3. 120.30. 31 Annals BORI) Madhu Vidya/445

Page #471
242 Annals BORI, LIVII (1986) saugandhika -- 'heavenly lotus', of one thousand petals, has heavenly fragrance, lustrous, brought to Pagdavas on the Himavat mt. by a north-casternly breeze 3. 146. 6-8; 3. 152. 2; 3. 150. 18; given by Draupadi to Bhima 3. 153. 13; a forest (yana ) of these flowers on the Gaodhamadapa ml., way to which was shown by Haoumat to Bhima 3. 149.22; the puskarini was in a river 3, 150.27; 3. 152. 22; (a saugandhikarana ( tirtha ?) mcptioned in 3. 82.3). snuha -- 'spurge', along the banks of Sarasvati 3.174.23. haricandlana - yellow sandal', in the regions near the Himavat tot, 3. 175. 10. huritaka - yellow Myrobalan trees', in forest 3.61.; on the Gaodhama. dana mt. 3. 155. 42. MadhuVidya/446

Page #472
THE FAUNA IN THE ARANYAKAPARVAN OF THE MAHABHARATA By M. A. MEHENDALE This paper is in continuation of the one on "The Flora in the Aranyaka. parvan of the Mahabharata" published in the ABORI LXVII (1986), 233-242. As mentioned there these entries can be looked upon as specimens of the "Cultural Index to the Mahabharata ". According to the scheme of classification accepted for the "Cultural Index", the information given here will belong to section 4, 7:"Animals, birds, serpents etc.", The following list does not include words for animals, birds etc. if they occur as names - or as parts of names of persons, like Baka 3. 12. 22 etc., Baka Dalbhya 3. 27. 5, Kunjara 3. 249. 10, Suka 3. 32. 11, Syenajit 3. 190. 73, Brhadasva 3.193. 4, Vrkodara 3. 12. 27 etc.; of mountains, like Rsabha 3. 83. 19, Rsabbakuta 3, 109. 7, Rksavant 3. 58. 20, etc.; of towns, like Nagapura 3. 90. 22, 180. 34, Varanasahvaya 3. 293, 14, etc.; of tirthas, like Rsabha 3. 83. 10, Ekahamsa 3. 81. 16, Nagatirtha 3. 82. 27-29, Dhenukatfrtha 3. 82. 16, Aivatirtha 3, 93. 3. It also does not include words for animals etc, if they occur in epithets like ursadhvaja 3. 81, 62, etc. or vrabhadhvaja (of Siva) 3, 80, 125, etc., suparnaketu (of Krsna) 3. 173. 15; or in expressions showing excellence like narakunjara 3. 266. 15, vanarapuhgava 3. 270. 7, etc., raksasapulguva 3. 264. 43, purusarsabha 3, 262. 7, etc., purusavyaghra 3. 248. 4, etc., kurusardula 3. 83, 97, etc., purusatardula 3. 296. 22, etc., purusasimha 3. 272, 23. ajagara boa', of huge form took hold of Damayanti 3. 60. 20, of Bhima 3. 175. 1, 16, its description of colourful skin, yellow in colour - 12-15, called prdaku 16, bhujuga 19, bhujamga 12, it had four fangs 14; Nahusa cursed to be a boa 3, 178. 45; described as graha 3. 60. 20, 21, 22. See graha.. aja (in ajaidakam) goat', people will milk goats in Kali age (since cows will not be available) 3, 188. 21; ajina-goat-skin' used for wearing by those who live in forest 3. 226. 19, 20; 227, 9. Cf. edaka. 1. Entries on proper names like Valin, Sugriva, Karkotaka etc. will not be found here. They will go under section 1. 5 (names of animals, birds, serpents etc.). Madhu Vidya/447

Page #473
328 ABORI: R. G, Bhandarkar 160th Birth Anniversary Volume anadvah-bull', used for drawing a plough 3. 184. 10. asva- horse', yoked to chariot 3. 255. 24 (vaha), 25; well-bred (ajaneya ) borses of Bhima's chariot 3; 254. 10; bay horses (haryasra ) draw Indra's chariot 3. 274. 12, 13; red horses (sonasva) yoked to the chariot of Sauviraka princes 3. 249. 9; Nala to look after the horses of Rtuparna and train them, he becomes asvadhyaksa 3, 64. 6; king Rtuparna has an asvasala 3. 69.10; also king Bhima has one 3. 71. 6; horses from Sindhu country known for swiftness 3. 69. 1.2 (stanzas 11-12 mention their other characteristics like wide nostrils, broad jaws; they have avartas (curl' B.1 suddhan dasabhir avartaih ); kneel on ground before taking a start and need to be pacified by the driver 3. 69. 18, 19; form one of the constituents of the army 3. 236, 7; form part of a caravan 3. 62. 9; horse set free at asvamedha 3, 105. 9; Satyavan, as a child, made horses of clay and drew paintings 3. 278, 13; asvahrduya or (asva)vidya science of horses' known to Nala 3. 69.27; 70.24, 26, he gave it to Rtuparna 3. 76. 18. Cf. haya., vajin.. asivisa- serpent', jars with serpents kept in the moats (?) around Lanka to make it unassailable 3. 268. 4; an angered serpent used for comparison 3. 222.34; 261. 17; arrows compared with serpents 3. 273. 20; in mode of expression (inviting calamity by offering provocation) 3. 134. 3; 261. 49. Cf. ajagara-, uraga., sarpa.. uksan- bull', offered as alternative food to the hawk by king Usinara 3. 131. 17. Cf. goutsa. u pacakra- duck', produced sweet sounds on the Gandhamadana 3. 155. 76. uraga- serpent', Damayanti sees fierce serpents in the forest 3. 61. 7; swallows Damayanti and gets killed by a hunter 3. 60. 26; five-hooded serpents for comparing the arms of kings (akaravantah suslaksnah) 3.54. 6, arms of Arjuna 3. 79. 19, clenched fist of Bhima 3. 154. 56; those living in lakes snatched by eagles 3. 253. 5; mythical serpents' listed with asura and raksas 3. 105.21 (they cry aloud when earth is dug up); with Gandharva and raksas 3. 157. 20 (Arjuna halted their advance at Khandava); with Kimnaras and raksas 3, 213, 23; mahoragah, mentioned with Kimnaras, different from pannagah (they wait on Siva at Gokarna ) 3, 83. 23; go with Agastya to witness the drinking up of the ocean 3. 102, 20; present at Saugandbika forest 3. 82. 4; mahoragas and uragas mentioned separately ( visit Gandhamadana on parvan days) 1. B-J. 4. B. van Buitone's translation. Madhu Vidya/448

Page #474
MEHENDALE Fauna in the Aranyakaparvan of the MBh. 329 3. 156. 18; Draupadi thought to be daughter of their king (uragaraja) 3. 249. 3 stepping on it is inviting danger 3. 252. 8. Cf. ajagara-, alivia, sarpa.. uluka- owl, they appear, along with other creatures, at the release of raudra astra 3. 170. 45; one called Prakarakarna lived on Himavant 3. 191. 4. ustrika- vessels like camel ustra- camel', as part of a caravan 3. 62. 9; bellies' (7) 3. 16. 7. ksa-bear', infest forests 3. 61. 2, 123; 260. 13; found on mountain 3. 61.37; move in herds (yuthasah) 3. 61. 8; some live at will (yathecchakanivasah) 3.260. 13; chased by Kauravas in Dvaitavana 3. 229. 10; appear at the release of raudra astra 3. 170. 42; bears devoted to Sugriva 3. 266. 6; in the army of Rama 3. 274. 4; black and having marks on faces (mukhapundra) 3. 267. 8; called kalamukha 3. 276. 11; 3. 274. 4; honoured by Rama, they return home 3. 275. 55; their king Jambavant 3. 264. 23: 268. 24; gods and other celestials beget progeny on female bears to help Visnu 3. 260. 7 (rkei), 11 (rksavarastri ). ksavarastri-, ksi-, seo rksa.. riya- 'white-footed antelope', offered at breakfast (pratarasa) 3. 251. 12. rsabha 'bull, appear at the release of raudra astra 3, 170, 43. edaka (in ajaiddakam)owe, in Kali age people will milk them (since cows will not be available) 3. 188. 21. Cf. aja.. aineya-(= ena-) black antelope, offered at breakfast 3. 251, 12. kanka-heron', feasts on flesh and blood on the battle-field 3. 255. 31. kacchapa-tortoise', in the ocean 3. 166. 3; called Akupara lived in the lake Indradyumnasaras 3. 191. 14. Cf. kurma-. kadamba-grey-winged goose' (B.) in the lakes on Gandhamadana 3. 155. 50 161. 5 (kadamba-). kapi monkey, appear as helpers of Rama 3. 267. 1, 51; 268. 31; 270.5, 10, 13: 271, 1; 273.3; Hanuman so called (having a long tail) 3.149.5; Sugriva so called 3. 271. 8; kapisvara-, Valin 3. 264. 20, 25, Sugriva 3. 266. 5; 271. 13; 273. 4; kapikunjara-, Sugriva 3.271. 7; kapitardila, Hanumant 3. 270. 7. Cf. plavaga-, vanara-, sakhamrga-, har.. 1. tilaka Nlakantha) Madhu Vidya/449

Page #475
330 ABORI: R. G. Bhandarkar 150th Birth Anniversary Volume kapila- brown cow', their gift often mentioned in establishing equivalence of punya acquired by visiting holy places 3. 80.76; 81. 38; 82, 8, 29, her hoof-prints along with those of her calf seen even today on a moutain (near Dhenukatirtha ?) 3. 82. 77. Cf. go.. kapota- ! dove', Agni assumes the form of a dove to test king Usinara 3, 130. 19; king weighed with this dove 3. 131. 25; dovo appears bofore the king As Agni 3, 131. 28; kapoti urtti subsisting like a dove on collected grain (alpasangraharupam NI.) 3.246, 4, 5, karabha- young elephant' (camel' NI.)? 3, 264, 45 (the hair of raksasis compared to). harenu- female elephant', sported with male companions in the waters of a lake near the asrama of Dadhica 3.98.15; seen in great number in Dvaitavana 3. 25. 19; on Gandhamadana mt. 3. 155, 77; occurs in a figure of speech 3. 262, 37. karkataki- ' female crab ', in a mode of expression (inviting calamity on one solf) 3.252, 9. kalahansa- goose ', flock the river Ganga on Gandhamadana mt, 3. 155, 85, kaka- crow', Rama threw isi ka at the crow 3. 266, 67, kakola- raven ', feasted on flesh and blood of those dead in war 3. 255, 31, kananakokila- forest cuckoo', produced sweet sounds in Dvaitavana 3. 25. 18, Cf, kokila.. karandava- duck', singing in the Himalayan rivers 3. 39. 18; 175, 9; 150.26 (matta.); in the lakes on Gandhamadana mt. 3. 155, 50; 161. 5; in the lotus lakes near Kubera's residence 3. 151, 6; in the lake Pampa 3. 263, 40, kita- "worm', a cave full of worms 3. 266, 38, kukkuta- 'cock', a red cock given to Skanda by god Fire ornamented his banner 3. 218. 32; of great size, red-crested, held by Kumara 3. 214 24, 23, a sadhana (?) of Kumara 3.215. 10 ( kukku tasya ca sadha. nan cakre mahamunih). kunjara- 'elephant', in forest 3. 61. 623; four-tusked, lotus-coloured shake the lake-waters 3, 155, 77. Cf. gaja., naga., matanga., varana., hastin., !. and the great hermit sang the efficacy of the cook' B. Madhu Vidya/450

Page #476
MEHENDALE : Fauna in the Aranya kaparvan of the MBh. 331 kurara- osprey', river full of their sounds 3: 61, 108; in the lakes on Gandha madana mt. 3. 155, 50, kurari- female osprey', woman in distress compared with themi 3: 128. 4; 170. 56; sound produced by them 3. 60. 19; kurma- * tortoise', in rivers 3. 61. 108, appear at the release of raudra astra 3. 170. 46; a demon (asura ) assumes the form of a tortoise 3. 82. 10. Cf. kacchapa. krsnamrga- black antelope', killed by the Pandavas and offered to Brahmanas 3. 47. 7. kokila- ' cuckoo', on Himavant mountain 3. 107, 8; 175, 7; the asrama of Dadhica full of their potes (pumskokila) 3. 98. 13. Cf, kananakokila-, krostuka- jackal', figures in a simile indicative of its lowness) 3, 262, 28, Cf, krostr.. krostr- jackal', figures in a simile (indicative of lowness ) 3. 248. 17. Cf, kroatuka.. krauzica- curlew', making sounds near the rivers of Himalayan region 3. 39. 18; 61. 108; seen in large numbers in autumn 3. 179. 10. khadyota- fire-fly', in a simile 3. 122. 19. khara- donkey', as part of a caravan 3. 62. 9; yoked to Ravana's chariot (in a dream soen by Avindhya 3. 264. 64. gaja- ' elephant' (see disagaja-), Duryodhana and others trapped them in Dvaitavana 3. 229. 11; many elephants on Himalayas 3.155.13, 35, 63; Subahu's kingdom on Himalayas had many elephants 3. 141. 24; in Jayadratha's following 3. 252. 12; used for defending cities 3. 268. 6; in wars, mounted by warriors (gajaroha) 3.255. 7, (gajaycdhin ) 3.255. 11, ornamented with rings 255. 22, one sent to kill Nakula 3. 55. 20, gajayanavid- one who knew movements of elephants in war 255. 18; his visana tusk' (also cf. disagaja ) and bhuja trunk' mentioned 3. 255, 21, kara trunk 3. 255. 20, for comparing Duryodhana's thigh 3. 11. 28; for comparing gait of woman 3.290. 14, of a man 3. 61, 51 (gajendravikramah); they appear at the release of raudra astra 3. 170. 43. Cf. kunjara-, naga-, matanga-, varana-, hastin.. garuda- appear at the release of raudra astra 3. 170. 44 ; figures in describing sounds of chariots 3. 230, 17, garudi- form taken by Svaha 3. 214. 9; 215. 3. Madhu Vidya/451

Page #477
332 ABORI: R. G. Bhandarkar 150th Birth Anniversary Volume gavaya- gayal', offered for breakfast 3. 251. 12; chased by Kauravas in Dvaitavada 3.229. 10. grdhra- vulturo ', feast on blood and Aesh on battlefield 3. 48. 33 ; 255, 31; go to mount Meru through a hole in the Kraunca mountain created by Kumara 3. 214, 31; appear at the release of raudra astra 3. 170. 44 ; Jatayu so called 3, 258. 2; 262. 41; 263, 3, 5, 15, 19, 20; 266, 45; grdhrapati. 3. 263. 1, 6, 17; uses nails, wings (referred to as bhuja) and beak as weapons 3.263, 4, 5; its feathers attached to arrows, bence called gardhravasas. 3. 34. 83. ago Cow' considered best among the moving ones (? pratisthamananam varah);? food (i, e, source of food ?) 3. 297. 61; considered visible form of gods 3. 196, 4; part of a caravan 3. 62. 9; will yield less milk at the end of Kali-age 3, 186, 37; offered to a guest 3. 279. 6; used as measure to establish equivalence of punya 3. 81, 40, 78, 79; 82. 67, 68; occurs in a subhasita (perishing in mire) 3. 36. 7. Cf. kapila, 'go- bull ';appear at the release of raudra astra 3. 170. 42. Cf. uksan-, reabha., vr 30-, gopuccha- cow-tailed', a kind of monkey 3. 266. 6; 275. 55; = gdan gula- 3. 267. 4. godha- ' a kind of lizard', its skin used for making the arm-guard (baddha godhangulitra.) 3. 18. 3; 38. 16; 143. 1; 262, 19; 267. 17. gomayu- jackal', moving in berds (Puthasah) in forest 3. 61. 8; in great number (bala) invade Rama's asrama 3: 263, 22; described as eaters of the remnants of offerings (vighasasa) in a simile 3. 34.3; feast on blood and flesh on the battle-field 3. 255, 31; 48, 33; uttering sounds on a person's left considered a bad omen 3. 253. 7. Cf. salavika-, 8Tgala.. govrsam 'bull', 3, 31. 25; tied by a nose-rope appears in a simile 3. 31. 25. Cf, uksan-, asabha., go-, vroa, graha- shark', river full of them 3. 61. 108; ocean full of them 3. 101.9, 102. 22. See ajagara., cakora- partridge', uttering sweet sounds in Dvaitavana 3, 25, 18, on Himavant 3. 107. 8 (asitapanga ); 175,7; on Gandhamadana 3. 155. 73. Cf. tittira.. 1. B. the best of the standing, Ni, pratithalipoindim. 2. B.cattlo Madhu Vidya/452

Page #478
MEHENDALE : Fauna ini the Aranyakaparvan of the MBh. 333 cakravaka- a river filled with their sounds 3. 61. 108; in the regions near Himavant 3. 175, 7; in the rivers and lakes on Gandhamadana 3. 150. 26; 155, 50, *camara- . Yak', chowrie (canara ) made from their bushy tails mentioned 3. 240, 42. chaga- goat', Vatapi turned into a goat 3, 94, 8; the sixth bead of Skanda is of a goat 3.217. 12 (Visakha called chagamukha 3. 217.3 or chaga vaktra 3,217,11). jalaku kkuta- water fowl, on Himavant 3. 107.7, the in lakes on Gandha madana 3. 155, 50. jivaka- insect '(B.), Dadhica's asrama full of their sounds 3.98. 13. jivajivaka- pheasant', on Himavant 3. 107, 8, 175, 7; on Gandhamadana 3,155, 74 ( described as raktapitarunah. red, yellow and ruddy'). hase- large fish' ( dolphin B.), found in river 3. 61. 108; in ocean 3. 99. 17. in southern Ocean 3. 266. 44; large fish with elephant's faces (gaja vaktra) appear at the release of raudra astna 3. 170. 45. Cf. timi. jhillikam . cricket', forest full of their sounds 3. 61.1. taraksi- wolf'chased by Duryodhana and others in Dvaitavana 3, 229, 10; killed by Bhima on Gandhamadana 3. 146. 48. tittir;- partridge', for comparison, horses speckled ( kalmasa) like tittiri birds 3. 79. 24. Cf. cakora timi- fabulous-sized fish', found in southern ocean 3. 266. 44; the makara og Pradyumna's banner called sarvalimipramathi 3. 18. 7. Cf. jhasa, timingilatimingila- fabulous-sized fish' (lit. 'swallower of timi' seen by Arjuna in ocean 3. 166. 3. Cf. jhasa-, timi.. timitimingila- seen by Arjuna in ocean 3. 166. 3. damsa - stinging insect', become excited (matta in rainy season 3. 179. 4; region around the asrama of Nara and Narayana free from them 3. 145, 20, one not restrained (aniyata ) meets them on Gandhamadana 3. 142, 27. Cf. masaka.. 1. moor cook'B. 2 'water cockerel B. Madhu Vidya/453

Page #479
334 ABORI: R. G. Bhandarkar 150th Birth Anniversary Volume damstrin- tusked animal', they cannot harm Nala 3. 63. 17. dardura- . frog ', full of pride (dar pita ) jump about in rainy season 3. 179. 8. datyuha- gallinule ', produce sweet sounds in Dvaitavana 3. 25. 18; on Himavant 3. 107. 7. disagaja- quarter elephant', mythical, they scratch the trees on Himavant with their tusks 3. 107. 10; ( see gaja. ). . dvija- bird', many kinds of them in forest 3. 61. 6. Cf. paksin., patrin., vayas.. dvipa- elephant', see mahadvipa-. dvipin- leopard', in forests 3. 61.2, 123. nakra- crocodile', in ocean 3. 101. 9; 266. 44 (southern Ocean); moats round Lanka filled with crocodiles 3. 268. 3. Cf. makara-. inaga- elephant', in the army of Jayadratha 3. 249. 11; buge like mountain summits move at the foot of Himalayas mentioned in a mode of expression) 3. 252.5; measure of strength 3. 260. 13; 297. 68; padmin- a spotted elephant (mahanaga. occurs in a mode of expression) 3. 262. 37; nagas (elephants ? snakes?) will make harsh sounds at the yuganta 3. 188. 81; Markandeya saw them in the belly of the Child 3. 186. 106. Cf. kunjara , gaja-, matanga-, varana., hastin-. anaga- snake', Karkotaka called naga ( speaks to Nala, 3. 63. 4 ff.; 64.1), nagaraja. 3. 70. 32 (Kali burning with the heat of its poison); nagarajan- 3. 63. 3; 75. 17, - mythical, present at Prayaga 3. 83. 67; at Kubera's residence on mt. Kailasa in very large number 3. 140. 11; their world (nagaloka ) referred to 3. 80. 119; 81. 12; situated, below the earth, Samvartaka Fire burns it 3. 186. 62. Cf. ajagara., asivisa., uraga, pannaga., bhogin", sarpa.. nirghatavayasa-, see vayasa.. nyanku- deer', at breakfast 3, 251. 12. paksin- bird '- two kinds sthalaja- and jalaja. 3. 12. 10; jalecara. (referring to & crane) 3. 297. 18; figure in a simile (jagthuh sardulam iva paksi. nah) 3. 268. 18. Cf. dvija-, patrin-, vayas.. patanga- ' moth', attracted by a flame jumps on it 3. 2. 65. 1. 'moor hon. B. Madhu Vidya/454

Page #480
MEHENDALE : Fauna in the Aranyakaparvan of the MBh. 33.5. patrin- bird', not identified but said to be of red, yellow and ruddy colour 3. 155. 73, 74. Cf. dvija-, paksin-, vayas-, padmin-, see naga.. pannaga- serpent', Karkotaka, a pannaga speaks to Nala 3. 63. 7; they appear at the release of raudra asira 3. 170. 42; seen in large number yuthasa!) by, Damayanti in forest 3. 61. 8; - mythical (mentioned along with gods and demons) 3. 170. 7; visit Brahmaksetra (Kuru. ksotra) 3. 81. 4; wait on Siva at Gokarna 3. 83.24; will glorify Rama as long as the earth lasts 3. 275. 48; no pannagas in Kotayuga, Hanumant tells Bhima 3. 147. 12; figure in a simile 3.221. 45. Cf. anaga., ajagara., asivisa-, uraga-, bhogin-, sarpa-. pasu- animal', seven gramya, seven vanya 3. 134. 13. pipilika- 'ant', crawling on an anthill 3. 122. 3. pipiilka- ant., bites Jantu 3. 127. 6. pithasarpa- boa', in a simile (idleness ) 3. 26.21 (NI. ajagara. ). purskokila- male cuckoo', singing in the sorroundings of Himalaya 3. 39. 18; in forests near the asrama of Nara-Narayana 3. 146. 3, 20; excited they move around in rainy season 3. 179. 8; for comparing voice (vani) 3. 112. 7. putrapriya- love-their-youngs (B.), on mountain Himalaya 3. 107. 8. prsata- spotted deer', offered at breakfast 3. 251. 12. plava- ' acquatic bird', in the lakes on Gandhamadana 3. 155. 50. plavaga- ' monkey' 3. 266. 25, plavagarsabha (Angada and others ) 3. 266 27, 29 (Hanumant), plavagasrestha (Sugriva ) 3. 266, 26, plavaga. sattama (Hanumant and others) 3. 266. 52, (Sugriva and others) 3. 273. 13, plavagagadhipa (Sugriva ) 3. 264. 13, 266. 13. Cf. kapi., plavangama-, vanaru-, hari.. plavangama- monkey', attack Kumbhakarna with trees, nails ( karajaih), and other weapons, 3. 271. 2-3; Hanumant and others so called 3. 266. 30; 149. 2. baka- crane', in the lakes on Himalayas 3. 155. 50; living on moss-like plant ( saivala) and fish 3. 297. 11; a certain baka lived in the lake Indra. dyumna 3. 191.9: figures in a figure of speech 3. 33. 7. Madhu Vidya/455

Page #481
336 ABORI: R. G. Bhandarkar 150th Birth Anniversary Volume barhina- peacock', their sounds heard in Dyaitavana 3. 229. 13; Dear Himalayan rivers 3. 39. 18. Cf. barhin-, mayura, sikhin., sikhandin.. barhin- peacock', on Gandhamadana dancing to the tune of the anklets of Apsaras 3, 146, 26; in a simile 3. 255. 11. Cf. barhina., mayura., sikhin., sikhandin.. balaka- crane ', bear the monsoon cloud 3. 84. 11; falls dead when looked at angrily by a Brabmana 3. 197. 3-5. bhasa- kind of vulture', feast on flesh and blood of the dead 3. 255. 31. bhujaga- snake', mythical, in the retinue of Hara 3. 221. 16. Cf. uraga., enaga-, pannaga., bhogin-, sarpa.. bhujanga-, snake', killed by a hunter 3. 60. 27; very poisonous on Himalaya 3. 107. 11. Cf. asivisa-, uraga-, pannaga , bhujaga-, bhogin-, sarpabhurunda- ' wild animal (?)', appear at the release of raudra astra 3. 170. 43. bhrngaraja- large bee', on Himavant mountain 3. 107. 7, 175. 7; uttering sweet sounds on Gandhamadana 3. 155. 76. bhogin- serpent (mythical)', seven headed poison-eyed guard the Sveta moun tain 3. 214. 11; Ravana, their enemy 3. 275. 30, 47. Cf. asivisa-, uraga-, pannaga-, bhujaga-, enaga-, sarpa.. bhramara- bee', humming in forest 3. 296. 41; flowering forests on way to Dvaita vana full of them 3. 229. 13; on mango trees on Gandhamadana 3, 155, 60; figure in comparison (Arjuna's arrows) 3. 167. 22; in a subhasita ( madhuha bhramarair iva) 3. 34. 68. Cf, madhukara.. makara- corcodile ', seen in ocean 3. 166. 3, in the southern Ocean 3, 261, 53; ocean called makaralaya 3. 105. 23, 275, 51 ; on the flag of Pradyumna 3. 18. 2; 19. 11; appear at the release of raudra astra 3. 170. 44 ; figure in similo 3, 254, 19. Cf, nakra.. maksika - fly', one meets them on the Gandhamadana if one is not restrained 3. 142. 27. manduka- frog', figure in a narrative 3. 190. 30 ff. matsya- fish', eat piece of meat in water 3. 2. 39; do not wink when asleep 3. 297. 43; cranes live on fish 3. 297. 11; gold and silvery fish found in Vimalatirtha 3. 80. 103; cursed by Fire to be the food of human beings 3.212. 9-10; fish will be food of men in Kali age 3. 188. 21; a fish requested Manu to save it from the strong fish; this fish later saved Manu from deluge 3. 185. 6 ff. Cf. mina.. Madhu Vidya/456

Page #482
MEHENDALE : Fauna in the Aranyakaparvan of the MBh. 337 madgu- i acquatic bird', in the lakes on the Gandhamadana 3. 155. 50. madhukara- bee', lotus ponds full of their sweet sounds on the Gandha madana 3. 155. 52. Cf. bhra mara.. mayura- peacock', produce sweet sounds from the tops of trees in Dvaitavana 3. 25. 18; also in the forest near Dyumatsena's hermitage 3. 280. 30; on mt. Himavant 3. 107. 8; in the bowers on the lakes on the Gandbamadana 3. 155. 54; 161. 4; yoked (together with swans). to tho vimana 3. 83. 29; Indra's horses compared to peacocks 3. 169. 23. Cf. barhina-, barhin-, sikhin., sikhandin-. .. masaka- mosquito', regions around the hermitage of Nara-Narayana free from them 3. 145. 20; one finds them on the Gandhamadana if one is not restrained 3. 142. 27. Cf. darta-. mahadvipa- big elephant', mentioned as leaders of herds (yuthapa) seen in large number in Dvaitavana 3. 25. 19. mahisa- buffalo', their meat sold in the slaughter-house of Mithila 3. 198. 10; served at breakfast 3. 251. 13; king Usinara willing to give a buffalo to the hawk (Indra) 3. 131. 16; wander freely, without fear of tigers, in the hermitage of Dadhica 3. 98. 14; in forest 3. 61. 123; move in herds (yuthasah) 3. 61. 8; chased by Kauravas in the forest 3. 229. 10; hunted by Pandavas 3. 253. 1; on the Gandbamadana 3. 150.21; water-buffalos on the Gandhamadana 3. 146. 45; overpowered by Bhima 3. 176.4 (measure of strength); monkeys huge like buffalos 3. 267. 11 (measure of size); appear at the release of the raudra astra 3. 170. 42. matanga- elephant', found on mountains 3. 61. 37; many die in wars 3. 255. 29; for comparison of a gait ( especially an elephant in rut) 3. 79. 14, 261. 9; elephant fighting an elephant 3. 12. 56. Cf. kunjara., gaja-, inaga-, varana., hastin-. marjara- 'cat ( wild )', appear at the release of raudra astra 3, 170. 43. mina- fish', moats round Lanka full of them 3. 268. 3; appear at the release of raudra astra 3. 170. 46; figure in a subhasita perish in the absence of water ) 3. 34. 27. Cf. matsya-. mrga- 1) forest animal, game animal (general), 3. 12. 11; 59. 21; 61. 6, 37, 83, 123; 248. 1 (in Kamyaka ), 243. 21 (in Dvaita ); they appear in the dream of Yudhisthira 3. 244. 2 ff.; chased by Kauravas 3. 229. 10, 11; found in the Visakbayupa forest on Yamuna mt. 3. 174. 17; bunted for food 3. 257. 9; 253. 1-4; sacrificial animals hunted for Brahmanas RGB...43 Madhu Vidya/457

Page #483
338 ABORI: R. G. Bhandarkar 150th Birth Anniversary Volume 3. 79. 8; in Dandaka forest 3. 263. 24; many kind of divya' game animals on Sveta mt. 3. 220. 24; raise sounds when struck by rain 3. 179. 7; appear at the relese of raudra astra 3. 170. 42; harsh sounds uttered by them resorting to the direction lit by the sun (adityadipta ), a bad omen indicating insult or attack by enemies 3. 253.3; 2) deer', found in hermitage 3. 61. 61; meat sold in a slaughter-house at Mithila 3. 198. 10; for break-fast 3, 251. 11, 12; many types aincya, preala etc. 3. 251. 12; king Usinara willing to give the hawk (Indra ) a deer 3. 131. 16, not accepted by the hawk 17; Marica took the form of a deer 3. 262. 11, 17; 263. 10 ( mahamiga); Yamadbarma took the form of a deer 3. 298. 13. 20, took away in its antlers the churning sticks of a Brabmana 3. 295. 8, 9 (mahamrga). 10, 14; for comparison (eyes ) 3. 60.29; 264.72; (lion attacking deer) 3. 221. 55; when the deer entered the manusa tirtha they became humans 3. 81. 13. Cf. harina-. mesa- goat', Vatapi turned into a goat whose meat served to Agastya 3. 97. 3. ranku-'a kind of deer', its hair used for making mattresses 3. 225. 9. rajahansa- white-feathered goose' (B.), in the lakes and rivers on the Gandhamadana 3. 155. 66. ruru- a kind of deer', in forests 3. 61. 2; 296. 40; on the Gandhamadana 3. 146. 45; killed by Pandavas for feeding Brahmanas 3. 47.7; offered for breakfast 3. 251. 12. rauhi- ' a kind of female deer', for comparison (rauhiva trasta ) 3. 265. 7. laksmana- heron (B.)', the cries of women whose husbands have fallen on battle-field compared with the autumnal sounds of these birds 3. 169. 21. lohaprstha- ' a species of bird' (red-backed'B.), uttering sweet sounds on the Gandhamadana 3. 155. 76. vadava- mare, 'yoked to a chariot 3. 133. 23; the fire Urdhvabhaj has a maro's mouth 3. 209. 20. vatsa- ' calf', 3. 82. 77. vayas, bird', food for them placed on ground at Vaisvadeva 3. 2. 57. Cf. dvija-, paksin., patrin-. varaha- boar', in forests 3. 61.2, 8 (in herds ); 296. 40; their sounds heard during rainy season 3. 179, 7; wander without fear in the hermitage of Dadhica 3. 98. 14; on mountains 3. 61. 37; on mt. Yamuna 3. 174, 17; in Dvaita vapa, chased by Kauravas 3. 229. 10; hunted by Pandavas Madhu Vidya/458

Page #484
MEHENDALE Fauna in the Aranyakaparvan of the MBh. 3.253. 1; meat sold in the slaughter-house of Mithila 3. 198. 31; offered as breakfast 3. 251. 13; king Usinara offers a boar to the hawk (Indra) 3. 131. 16, 17; appear at the release of the raudra astra 3. 170. 43; daitya Muka takes the form of a boar 3. 40. 8; 163. 17; Visnu takes the form of a boar 3. 100- 19; 294. 28; Visnu in the form of a boar present at Varaha tirtha 3. 81. 15; Markandeya sees boars in the belly of the Child 3. 186. 106. vartika 'quail', of terrible appearance, with one wing, eye, and-foot, vomitting blood, making harsh sounds facing the sun (pratyadityam), a bad omen 3. 176. 42. 339 vajin-horse', yoked to the chariot of Indrajit 3. 273. 25; used in the army 3. 253. 23: 268. 6: Subahu's mountainous kingdom in Himalayas had many horses 3. 141. 24. Cf. atva-, saindhava-, haya-, vanara-monkey', described as vanya 3. 271, 27, and living where it will (yatrecchakanivasz) 3. 260. 13; move in herds (yuthatah) 3. 61.8; many found on way to Gandhamadana 3. 145. 13-14; in the army of Rama 3. 266. 6; 267. 2, 6, 12, 13, 25, 27; 268, 26, 34; 271. 4, 27; 274. 23: 275. 25, 50, 53; fight the raksasas 2. 268. 37, with trees 274. 4, nails and teeth 268, 36; faces (red) like ground vermilion (hingulaka) 3. 267. 11: appearance white like girisa flower 3. 267. 10; 268. 27, or like sara 268. 27, or like autumnal cloud 3. 267. 11; or they look like ricefields or morning sun 3. 268. 27; hairy 3, 268, 36; huge like mountain summits or a buffalo 3. 267. 11; those killed in war brought back to life by Brahman 3. 275, 41-42; honoured by Rama with jewels 3. 275, 53, 54 gods beget them on excellent female monkeys 3.260 7, 11; raksasas, working as spies, assume their form 3, 267. 52; a demon takes this form to attack Krsna 3, 23. 10; Hanumant, Sugriva and others so called 3. 149. 4; 150. 1; 264. 9, 10, 12, 17; 266, 32, 60, 63; vanarendra (Sugriva ) 3, 266, 20; (Gaja and Gavaya) 3. 267. 3; vanarapungava, vanaramukhya (Sugriva and others) 3. 266. 11; 267.23; vanarapati (Valin) 3.264.18; vanararaja (Valin) 3. 263. 41, (Sugriva ) 263. 42; 264. 11, 57; prajnavanara (Angada) 3, 267, 54; vanararsabha 3. 268. 35. Cf. kapi-, plavaga-, takhamrga-, 1hari-. vayasa crow', feast on flesh on blood 3. 255. 31; crying, interpreted as saying 'go, go' (yahi yahi) listed among bad omens 3. 176. 44; their harsh sound heard at the yuganta 3. 188. 81; at the end of Kali age trees will be full of crows 3, 186. 37. Madhu Vidya/459

Page #485
340 ABORI: RG. Bhandarkar 150th Birth-anniversary Vdumo varana- elephant', king Bhima of Vidarbha had a sala for them 3. 71. 6; get excited at the sound of chariot 3.71.7; sport with femalo elephants in the lakes near the asrama of Dadhica 3. 98. 15; on the Gandhamadana mt. 3. 146. 45; 150. 19; overpowered by Bhima in combats 3. 176. 4, Cf. kunjara-, gaja., naga-, matanga-, hastin.. vrka- wolf', in a figure of speech ( simhagostham vrko yatha) 3.251.8. vrsa- bull', setting free a dark one (nila) highly meritorious 3. 82. 85. Cf. uksan-, tsabha-, 'go-, visabha., govrsa-. orsabha- bull', for comparison (two bulls fighting ) 3. 12. 53; its shoulder (vrsabhaskandha) 3. 292.5; setting free a bull (see vrsa- ) 3. 83. 11. Cf. ukpan., tsabha., 'go-, gourga-, ursa.. vyaghra- tiger', infest forests 3. 61, 2, 25; in the caves of Himavant 3. 107. 6; near Himalayas and on the Gandhamadana 3. 155. 13, 63; 146. 39 (some killed by Bhima); one meets them on the Gandbamadana if one is not restrained 3. 142. 27; overpowered by Bhima 3. 176. 4; hermitage of Dadhica filled with their roars 3.98. 18; Sita protected by a tiger (a dream seen by Avindhya) 3. 264. 70; appear at the release of raudra astra 3. 170. 42; Markandeya saw tigers in the belly of the Child 3. 186, 106; for comparison (two tigers fighting each other with nails and fangs ) 3. 12. 54; in a simile (balolkata ) 3.255. 3 (also cf. expressions like purusavyaghra); in a mode of expression (inviting danger to oneself) 3. 134. 3. Cf. sardula-. vyaghravadhu- tigress', in a mode of expression ( krosta vyaghravadhum iva ) 3. 248. 17. Ivyala- beast of prey' (uskavyaghradayah - NT.), in forests 3. 2. 3; 59, 21; in Dvaitavana 3. 243. 21; in Kamyaka forest 3, 253, 2, Ct, svapada-. avyala- ' serpent', in simile (ucchvasantah) 3. 253. 22. wall-female serpent', for comparing Sita's braided hair 3. 265. 25; in a mode of expression (its meaning not clear) $osayisami gatrani vyali talagata yatha) 3. 264. 50. (Cf. tala-, ABORI LXVII. 237, 1986). falcuna- bird', with wings shining like gold seen by Nala (really dice turned into birds ) 2. 58, 11, 15; of various forms and chirping differently on Himavant 3. 107.7; not identified, sitting in the intermediate spaces (vivaresu) on trees on the Gandhamadana, they had plumages like diadema 3. 155. 56. Cf. Sakunta-. Madhu Vidya/460

Page #486
MEHENDALE : Fauna in the Aranyakaparvan of the NBh. 341 sakunta- bird', in a simile 3. 234, 12 ( sakunta iva panjare ); 3. 273. 2 (sakuntav iva). Cf. sakuna.. tatapatra- 'wood-pecker' (M. W.), pheasant' (B.) on the Gandhamadana 3. 155. 73, on Himavant 3. 107. 8. tambara- kind of door', offered at breakfast 3.251. 12. sarabha- has eight foet, kills lions 3. 134. 14; found on the Gandhamadana 3. 155. 35, filled with their roars 64 ; offered as breakfast 3. 251. 12; appear at the release of raudra astra 3. 170. 43. salabha- locust ', in comparison ( appear in large numbers) 3. 167. 23. tasa- rabbit', offered at breakfast 3. 251. 12; at Sasayana tirtha, lotus flowers remain concealed in the form of rabbits ? 3. 80.120; moon has the mark of a rabbit 3. 266. 2. sakhamrga- monkey', in hermitage 3. 61. 61; Rama's allies 3. 276. 11; 266. 61, 63 ( survasakhamTgendra); used as a contemptuous term for ugly woman 3. 251. 3 (sakhamfgastri-). Cf. kapi., plavaga., vanara., hari.. sarika- 'maina', on the Gandhamadana 3. 155. 73. $ardula- tiger', on the Gandhamadana 3.150.21 ; animals, like buffalos, afraid of them 3. 98. 14; called aranyaraj., 3. 61. 19, 30; aranyanypati. 3. 61. 33; mrgendra. 3. 61. 32; figures in comparison 3. 268. 18. Cf. vyaghra-. tardali- tigress', in a mode of expression 3. 262. 28. falavrka- jackal', howling on the left of a person, a bad omen 3. 253. 8; appear at the release of raudra astra 3. 170. 43. Cf. gomayu., siva, $lgalasikhin- peacock', in Bhima's palace dance at the hearing of the chariot sound considering it to be the time for the appearance of the cloud 3. 71, 6, 7; overjoyed peacocks roam about during rainy season 3.179. 8. Cf.barhin., barhina-, mayura, dikhandin-. sikhandin- peacock', move in the company of peahens on the Gandhamadana 3. 155. 72. Cf. barhin., barhinas, mayura-, sikhin-. sikhandini- ' peahen', near the bowers on the Gandhamadana 3. 155. 53; movo there in the company of peacocks 3. 155. 72. fiva- jackal', at the end of the yuga, the crossroads will be sivasulah (will bristle with jackals B.). 3. 186. 36. Cf. gomayu-, falavika-, trgala.. Madhu Vidya/461

Page #487
342 ABORI: R.G. Bhandarkar 150th Birth Anniversary Volume siva- 'female-jackal', their sounds, an ill omen when made standing towards the south 3. 176. 41, or towards the south-west 3. 281. 74. sragala- jackal', entering a lotus pond, an improper act (mode of expression) 3. 253. 19. Cf. gomayu-, siva, salavska. saunda- elephant' (NI.), drunkard' (B.) in a mode of expression (a jar from which water has been drunk by elephants) 3. 253. 5. Syena- hawk', Indra takes the form of a hawk to test king Usinara 3. 130. 19, 28; in a simile (hawks rush at a piece of meat ) 3. 253. 24. svana- dog ', feast on flesh and blood on battle-field 3. 255. 31; mode of expros sion (oblation becomes unworthy when licked by a dog) 3. 275. 13; 253, 19, 20; their mother Sarama 3, 29, 33; svanara. 'a lowly person 3. 252. 3; svadsti. ' a bag made of dog's skin 3. 34, 78; sva paca.' a lowcaste man. 3. 2. 57. svapada- beast of prey', in forests 3. 64. 18. Cf. Ivyala-. satpada- bee', the hermitage of Dadbica humming with their sounds 3. 98. 13; in the lotuses in the ponds on the Gandhamadana 3. 150. 23. sarisrpa- reptile', in forests 3. 2. 3; over-joyed (matta in rainy season 3. 179. 4; one meets them on the Gandhamadana if one is not restrained (uniyata ) 3. 142. 27. sarpa- serpent', (mythical ?) Ravana not to be defeated by them 3, 259. 25; in similes, serpent in a house 3. 222. 1; hissing like a serpent 3. 225, 13. Cf. asivisa., uraga-, 'naga-, pannaga., bhogin., zvyala.. sarasa- 'crane', making sweet sounds on Himavant 3. 107.9; near Himalayan rivers 3. 39. 18; in the lakes and rivers on the Gandhamadana 3. 155. EUR6, 75; their presence indicative of neargess of water 3, 296. 8, 11; yoked ( along with hansa) to the heavenly vimana 3. 246. 31. simha- lion', Dear the hermitage of Dadhica 3. 98. 16; in forest 3. 61, 25; on mountains 3. 61. 37; many on Himalayas 3. 155, 13, 35, 63; in its caves 3. 107. 6; 146. 57; on the Gandhamadana mapy attack Bhima and get killed 3. 146. 39, 46, 48; one meets them on the Gandhamadana if ono is not restrained (aniyata ) 3. 142, 27; having manos 3, 176.4;221. 3; $arabha kills them 3. 134. 14; appear at the release of raudra astra 3. 170. 42; a thousand, yoked to Siva's chariot, move through the air 3. 221. 2-3; Markandeya sees them in the belly of the Child 3. 186. 106; for comparison, man entering his youth 3. 292. 21; gait (vikranta ) 3. 71. 12; chest (uras ) 3. 61. 12; jaws ( daistra) 3. 157. 27; attacking Madhu Vidya/462

Page #488
MEHENDALE : Fauna in the Aranyakaparvan of the MBh. 343 small animals 3. 221. 55; 272. 10; mode of expression not easy to fight them 3. 48. 39; wolf entering lion's den 3. 251, 8; plucking eyelashes and kicking a sleeping lion 3.252. 6; touching his jaws 3. 261. 49; lion's roar 3. 267. 10; 221.51; 230. 5; 274.18; undeserved esteem 3. 133. 20 (siinhikTtah ). Cf. Phari. suparna- sagle', snatch snakes from ponds 3. 253, 5; mythical birds 3. 157. 14, 19; present at Prayaga 3. 83. 67; visit Kubera's residence on mt. Kailasa 3. 140. 11; visit Gandhamadana on parvan days 3. 156. 18. Cf. garuda.. suparni- . female eagle', 3. 215. 4; form taken by Svaha, daughter of Daksa, 3. 214. 9. Cf. garudi.. sukara- pig', mode of expression (high-born woman not likely to be enamoured of a low person) 3.262.37 ( katham ... karenuh sukaram sprset). 8pmara-marsh deer' (B.), move without fear in the hermitage of Dadhica 3. 98. 14; appear at the release of raudra astra 3. 170. 43. saindhava- Sindhu steed', very swift 3. 253. 6. stokaka- (=cataka NI.), move about intoxicated during the rainy season 3, 179. 8. hansa- 'goose', seen in large numbers in autumn 3. 179, 10; lake Pampa full of them 3. 263. 40; on Himavant 3. 107.7; singing in the rivers on the Himalayas 3. 39. 18; 175, 9; Gandhamadana peaks and lotus lakes full of their notes 3. 161, 4, 5; 155. 50; in the lotus pond near Kubera's abode 3.151 6; they go to mt. Meru through a hole in the Kraunca mt, made by Kumara 3. 245. 31; a heavenly vimana to which swang (and peacocks) are yoked 3. 83. 29; 246, 31; 275. 19; swans ornamented with gold (jatarupaparicchada ) caught by Nala, can speak like human beings, bring about marriage of Nala and Damayanti 3. 50. 18, 19, 30-31; 61. 15; 74. 13; for comparison, 3. 288. 16 ( whiteness ), 3. 134, 1 (noise made by debaters ), (sounds ) 3. 112.6, 1. 43. haya- horse', in the army 3.249.11; obtained as tribute by Arjuna from the Gandharvas (speckled like partridge (trttira) and swift as wind) 3. 79. 23-24; yoked to Kubera's chariot called gandharva 3. 158. 23; Nala (Babuka) expert in the knowledge of horses (hayajna3. 69. 23. 33; 73. 6; (hayajvana) 3.76.16; (hayakovida) 3. 70. 17; (hayatattvajna) 3. 69.2, 16; Salihotra, reputed expert in asvavidya (kulatattvavid) 3. 69. 25; Nala (Babuka) expert in controlling horses 3. 69. 23, 34; Daruki ( son of Daruka ) skilled in controlling horses 3. 20.5, knew their Madhu Vidya/463

Page #489
344 ABORI: R. G. Bhandarkar 150th Birth-Anniversary Volume mandala-varieties 3. 20. 8; spurred by pratoda, and controlled by reins 3. 20, 9. Cf. asvrz-, vajin-, saindhava.. 'hari- monkey', Sugriva's followers 3. 266. 17; in Rama's army, swift as wind 3. 268, 23; limbs red like karabha (vermilion 'B.) 3. 268. 25; 267. 7, 21; sala and tala trees and rocks their weapons 3. 267. 18; 269. 4, 5; 270. 8, 15, 16, 23; 271.2; 272. 24; 275.16; hariyuthapa. 3. 271. 6; 274.2; hariyuthapa-yuthapa. 3. 267. 9; harindra. 3. 267, 14; harimahamatra. 3. 267. 17; Hanumant called hari 3. 150.3; 270, 12; Nila so called 3. 271. 25; Sugriva called hari 3.266. 13, harindra 3. 266. 30, haripungava 3. 264. 6. Cf. kapi-, plavaga., vanara., fakha. mrga-. hari- lion', mode of expression (kicking a sleeping lion, a foolish thing to do ) 3, 252. 7. Cf. simha. harina- deer', on the Gandhamadana 3. 146. 28; 150. 20, 25; offered at break. fast 3. 251. 12. Cf. mrga-. harini- female deer', on the Gandhamadana 3. 150. 20; in comparison (ona strayed away from the berd) 3. 61. 23. hastin- elephant', one of the constituents of the army 3. 236. 7, hastyaroha. warrior fighting from elephant-back 3. 255, 22; herd of elephants tramples on the members of a caravan 3. 62. 6; in comparison (a lotus plant harassed by an elephant's trunk) 3. 65. 14. Cf. kunjara-, gaja., Inaga., matanga-, varana., Madhu Vidya/464

Page #490
ONCE AGAIN "DRAUPADI'S GARMENTS" M. A. MEHENDALE The Mahabharata (Mbh.) tells us in the Sabhaparvan, if we follow the text adopted by the Poona Critical Edition (2.61.40-41), that an attempt was made by Duhsasana to snatch away Draupadi's garment. He did not succeed in his attempt to uncover a part of her body because she was covered by garments which appeared on her person one after the other. The text of the Mbh. adopted by the Critical Edition, thus, notes the incident but has no explanation to offer of the miracle of garments (adbhutatamam 2.61.42). A. Hiltebeitel has sought to offer an explanation by resorting to a 'nature mythology' interpretation of the Mbh. scene.' The interpretation suggested is as follows: Draupadi stands for earth; the attempt of Duhsasana to disrobe her is to be understood as the attempt of the sun to lay bare earth in summer; the attempt fails; this happens because garments appear *automatically' on Draupadi's body which signifies that the sun's attempt to lay bare earth in summer fails because the earth is covered in the rainy season with the automatically growing plants. In the opinion of Hiltebeitel we have enough indications in the epic to offer above interpretation: In the first instance Draupadi's name Krsna 'the black one' points to the black earth. Secondly, the suggestion to disrobe Draupadi is made by Karna who, as son of the sun, stands for the latter. Thirdly, Karna is killed by Arjuna, son of Indra who sends down water in the rainy season to cover the earth with plants. Thus, we obtain a chain Sun Earth Indra which is paralleled by KarnaDraupadi - Arjuna. There is one more indication in the epic which, according to Hiltebeitel, further strengthens the above parallelism. Before the great war begins, Krsna attempts to tempt Karna to come over to the side of the Pandavas by telling him that Karna's change of sides will make him the sixth husband of Draupadi (Mbh. 5.138.15). Karna does not accept the suggestion because it is tantamount to the union of the Earth and the Sun, a cosmological impossiblility (p. 103). The constant opposition between the Karma and Draupadi is nothing but the reflection of the opposition between the Earth and the Sun. It is only at the end of the aeon (yuganta) that the one sun begets seven and is able to scorch the earth completely thus making it totally bare. But in the Sabhaparvan the attempt to disrobe Draupadi fails. "In simplest terms, the restoration of Draupadi's sarees shows that, except at the time of pralaya, the Earth's power to restore her garments is inexhaustible". (p. 104). Madhu Vidya/465

Page #491
If we interpret the Sabhaparvan episode in the light of nature-mythology, the unexplained phenomenon of the garments appearing automatically on Draupadi's body is easily explained. It is difficult to agree with the above view of Hiltebeitel. In the first instance, the two assumptions which compel Hilterbeitel to invoke nature mythology are baseless. He assumes that Karna asked Duhsasana to remove from Draupadi's person her saree to disrobe her., This is not correct. What Karna asked Duhsasana to do was to remove only her upper garment (uttariya), witness what the Pandavas did and how they understood the word 'vasas' in Karna's instruction (pandavanam ca vasamsi Draupadyas capyupahara Mbh. 2.61.38).? Hiltebeitel also assumes that the garments appeared automatically on Draupadi's body. Although there is no stanza explaning the miracle in the text accepted in the Critical Edition, we cannot neglect the fact that the tradition has offered two explanations of it. One is that Draupadi prayed Krsna and he supplied her with the necessary garments. The other is that Dharma supplied her the garments. This latter explanation can only mean that Draupadi in the difficult situation she was placed did satyakriya putting at stake her scrupulous observance of dharma' and succeeded in failing Duhsasana's attempt. This explanation, and not the first one, is more likely to be closer to the original epic and hence it is not necessary to assume that garments appeared automatically, without any supernatural intervention, on Draupadi's person. Thus it is unnecessary to bring in nature - mythology to understand the sequence of events in the Mbh. Even otherwise Hiltebeitel's explanation does not stand the test of scrutiny. In the first instance let it be remembered that Karna asked Duhsasana to get the (upper) garment not of Draupadi alone but of the Panadavas as well. The Pandavas obliged him by removing their upper garments themselves and putting them aside. If Duhsasana's unsuccessful attempt is to be understood on the basis of naturemythology, then in this explanation there must be a place also for the Pandavas removing their upper garments and sitting with parts of their bodies bare in the assembly. But in Hiltebeitel's explanation there is nothing to correspond to this part of the incident. If the metaphorical explanation is to be established, it should neatly embrace the whole incident and not just one convenient part of it. Next, as far as the natural event is concerned it is the sun who during the summer months tries to lay bare the earth and does not ask some one else to do it. In the epic narrative it is not Karna (son of the Sun) who tries to snatch the garment of Draupadi but asks Duhsasana to do it. In the nature-mythology imagined by Hiltebeitel there is nothing to correspond to Duhsasana. Hiltebeitel is aware of this discrepancy. He tries to explain it away by pointing out that at the time of pralaya the Sun succeeds in laying bare the Earth and hence the Sun does the act himself. On other occasions, that is during annual summer, the attempt is not successful and hence Madhu Vidya/466

Page #492
in the epic Duhsasana is shown dragging the garment and not Karna. But this escape is not satisfactory. It would have been so, provided in summer the unsuccessful attempt to lay bare the earth was made by some agency other than the sun. Thirdly, Hiltebeitel on the one hand wants us to believe that the sun's attempt does not become successful on account of the earth's intrinsic power to cover herself with plants, but at the same time he also says that the earth needs rain-water supplied by Indra to enable the plants to grow. In the epic narrative, however, Arjuna, son of Indra, is of no help to Draupadi to cover herself with garments. Arjuna kills Karna no doubt, but it has nothing to do with foiling Duhsasana's attempt to carry out Karna's instruction. Fourthly, Hiltebeitel asserts that it is god Indra who supplies rain-water to earth. This is only partially true. In the epics Indra does appear as rain-god. But the older vedic tradition tells us that it is god Sun who sends down rain. That the sun has a hand in the phenomenon of rain is not unknown also in post-Vedic period as can be seen from expression such as navamasadhstan garbhan bhaskarasya gabhastibhih (Ram. 4.28.3) and sahasragunam utsrastum adatte hi rasam ravih (Raghu. 1.18). * If this older and continuing tradition is to be relied upon, then in the epic Karna should have helped Draupadi to get the garments, but that is an impossibility. Hilte beitel at one stage observes : "They did not pick Karna's name out of hat for his role at the disrobing" (p. 102). In his opinion this has happened because in nature mythology it is the sun who tries to disrobe the earth and in the epic Karna is son of the Sun. Hiltebeitel, apparently, has a different view of the role of the epic author. The epic narrator is not a dramatist-cum-producer who has a bunch of actors at his disposal and undertakes to write a stage-play keeping in view certain roles for certain actors. The business of the epic narrator is to narrate the event as it happened in the past. If in the event to be described by him Karna acted in a certain way, he has to say so, and has no choice. The situation in the epic event was not such as if the narrator had many choices and that he picked up one of them. After all who but Karna in the Assembly could issue the particular instruction he did to Duhsasana? There are only three persons - Duryodhana, Sakuni and Karna- who could come up for consideration. In the Sabha, Sakuni's role is restricted to the winning of the dice-game. This done, he remains a passive on-looker. The upper garments of the Pandavas and Draupadi were to be removed because, apparently, the dasas and the dasis had no right to use them. There was no doubt about the Pandavas having become the dasas. The case of Draupadi was yet to be decided. Duryodhana had admitted the question raised by Madhu Vidya/467

Page #493
Draupadi which meant that in his opinion it was not decided whether she had become a dasi or not. Hence, he was most unlikely to issue the instruction. In the whole assembly it was Karna alone who had no doubt that Draupadi had become the slave of the Kauravas. Hence, it is easily understandable that he, and no one else, asked Duhsasana to get the garments. There is nothing here which the author of the Mahabharata has done with some motive. Hiltebeitel attempts to bolster up his explanation by pointing out further that Karma, as suggested to him by Krsna, does not leave the side of the Kauravas and go over to the Pandavas to become the sixth husband of Draupadi. According to Hiltebeitel this happens because the union of sun (represented by Karna) and earth (represented by Draupadi) is an impossibility. But why does one have to imagine a reason for Karna's not leaving the side of the Kauravas? In the case of garments the critical text of the epic has no stanza to explain the incident and hence one can understand Hiltebeitel's imagining some kind of explanation. But in the case of Karna not deserting the Kauravas the Mahabharata does not leave anything to imagination. It tells us quite clearly that Karna does not leave the Kauravas because he had enjoyed his kinghship for thirteen years on account of Duryodhana's support, and that he did not want to betray his friend who had started the conflict with the Pandavas relying mainly on Karma's support (5.139.13, 1518). Why should one not accept this sensible explanation and imagine something which has no basis? The review of Hiltebeitel's interpretation of the garment-episode does not end here. In his opinion, the identification of Krsna Draupadi with black earth is only one facet of her larger identification with praksti constituted of the three gunas: sattva, rajas, and tamas. This identification of Hiltebeitel is based not on the evidence offered by the text of the critical edition but by an interpolated stanza. It is strange that Hiltebeitel discards the evidence of the interpolated passages while treating the first identification of Draupadi with earth, but for the second identification (Draupadi = prakrti), he has no hesitation in resorting to the interpolated passage. One of the interpolated stanzas in the incident of the garments (2.61.553*) says that the garments appearing on Draupadi's body were of different colours and were also viraga (nanaraga-viragani). Actually the above words do not specify the colours of the garments. If they were of different colours one of them could, of course, be red. Hiltebeitel picks it up as it suits his explanation. Next, he requires white colour. This also, like the red, he could have got from the general description nanaraga.But he does not do that because the white colour of the garments is offered to him by the commentator Nilakantha who interprets the word viraga as kevalasveta'pure white'. The third colour, black,required by him for Draupadi's Madhu Vidya/468

Page #494
identification with prakrti is given to him by Draupadi herself who is known to be Krsna. Hiltebeitel, thus, sees before him 'black' Draupadi putting on alternately, 'pure white' and 'red' garments (he of course neglects other equally possible colours of the garments like yellow, blue etc.) which, in his opinion, represent the fundamental 'qualities' of nature, thus enabling him to identify Draupadi with prakrti. "Draupadi's identification with Earth is thus only one facet of her fuller identification with prakrti 'nature'. As the epic's primary embodiment of the Hindu Goddess, she represents all of nature, not just the Earth" (p. 107) For various reason this second identification is as unacceptable as the initial one. In the first instance, the 'garments' in the form of plants which cover the earth in rainy season are strikingly green in colour and they are neither red, nor white. Hence, one cannot set aside this colour and imagine that the garments which, for Hiltebeitel, appeared automatically on Draupadi's body and which therefore stand for plants were red or white in colour. Next, as noticed above, the two colours - white and red required by Hiltebeitel have not been directly stated even in the interpolated text. As a matter of fact the evidence given by the text accepted in the Critical Edition should have settled the point. It clearly states that the garments were only of one colour (tadrupam aparam vastram 2.61.41), i.e. if the original upper garment put on by Draupadi was white, the rest were white, if red red, etc. Hiltebeitel no doubt notes this point (p. 105), yet has no hesitation in drawing upon the evidence of the interpolated passage. In his opinion this passage "makes up for this uncertainty" regarding the colour of the garments. But, in fact, it does not. Even this interpolated text does not specify any colour, much less red and white, but simply says that the garments were of different colours. In these circumstances, it is not proper just to choose the colour which suits the intended interpretation. It is again extremely doubtful if viraga can mean kevalasveta as assumed by Nilakantha. The expression viragavasas can means 'a garment of different colours, variegated'. The interpolated passage means that the garments which appeared to cover Draupadi were of two types, some of different colours, i.e. each one of a different colour, and some variegated, i.e. each one having different colours." - Hiltebeitel approvingly refers to the views of Heino Gehrts who sees connection between the garments of Draupadi and those brought by Uttara, son of Virata, in the Virataparvan. When Uttara is about to march out to get back the cows plundered by the Kauravas his sister Uttara and her friends ask his charioteer Brhannada (= Arjuna) to get for them the (upper) garments (vasamsi) of the Kaurava heroes so that they could make dolls out of them." The garments brought back were of three colours - white of Krpa, Madhu Vidya/469

Page #495
yellow of Karna, and blue of Asvatthaman and Duryodhana. Since the garments in both the episodes are various and colourful, Gehrts has seen connection between the two. Hiltebeitel agrees with Gehrts and asserts that it is possible to find further evidence for connecting the two episodes by pointing to the fact that the stanza in which Brhannada is asked to get the garments we find the use of the word pancalika. The word means both *Draupadi' as well as 'doll.' When BIhannala is asked to get the garments pancalikartham, there is also a hint, argues Hiltebeitel,;'bring the garments for Draupadi.' Now the garments of the Sabhaparvan, since they stand for plants covering the earth, symbolize rebirth. They were given not to Draupadi, although they were meant for 'Pancalika', but to Uttara because they were brought for making dolls. In the opinion of Hiltebeitel this change is highly significant. It suggests that regeneration of the Pandavas, i.e. continuation of their line, was to take place not through Draupadi-since all her five sons were massacred-, but through Uttara. Hence in the Virataparvan, Uttara takes the place of Draupadi. This argument suffers from various drawbacks. It is based on grounds which range from being inadequate to impossible. Before we go into the details, let us once again see what the Mahabharata has actually to tell us. In the Sabhaparvan it tells us that when in the Assembly Duhsasana tried to remove the (upper) garment of Draupadi there appeared other garments of the same colour to cover her. And in the Virataparvan incident we are told that Uttara brought garments of some of the Kaurava warriors which were white, yellow and blue in colour. The relationship between the two garments, as mentioned above, rests on their 'similarity'. But since the colour of the garments is not specified at all in the Sabha, the question of seeking similarity based on colour does not arise at all. And even if we admit for the time being the two colours - red and white - chosen by the Hiltebeitel for the Sabha incident, they do not match the colours of the garments in the Virataparvan. If the garments (standing for plants) symbolize rebirth through Uttara, what kind of rebirth do they indicate when they appear on Draupadi's body? The garments of Draupadi stand for plants, and hence for rebirth, because they are supposed to appear automatically. No such automaticity is seen for the garments of the Kauravas. Yet they symbolize rebirth because they are similar' to those of Draupadi in a very vague way. Admitting, the latter were garments of rebirth, they should have covered Uttara. This definitely does not happen. She at the most touches them while making dolls. But so do her friends who were also apparently involved in the making of the dolls. In spite of all these difficulties, if the garments of the Kauravas are to be looked upon as garments of rebirth, and were handed over to Uttara, and not to Draupadi, for the purpose mentioned above, then to imagine Madhu Vidya/470

Page #496
that the author of the Mahabharata suggested the continuation of the line of the Pandavas through Uttara with garments of the Kauravas is something preposterous. If the author of the Mahabharata did want to convey a suggested meaning by the use of the word pancalikartham it could only be 'bring the garments of the Kaurava heroes for Draupadi, i.e. bring them to avenge the insult done to her by trying to remove her upper garment in the Sabha'. Neither the garments of the first nor of the second episode have anything to do with rebirth or the continuation of the line of the Pandavas. NOTES: 1. Alf Hiltebeitel : Draupadi's Garments, Indo-Iranian Journal, 22.97-112 (1980). 2. I have argued this point at length in a Marathi article published in the Navabharata (Wai, Dt. Satara, Maharashtra) April 1987. Hiltebeitel's translation (p. 98) of the above line" strip the Pandava's and Draupadi's clothes" goes beyond the text. 3. nanaragaviragani vasanany atha vai prabho pradurbhavanti Sataso dharmasya paripalanat 2.61.553.* For Draupadi taking recourse to satyakriya on a different occasion see Mbh.4.14.18. 4. For a detailed and very instructive account, see H. Luders, Varuna, I 308-314, 5. In the Virataparvan (35.23), the garments, in a different context, are similarly described as citrani (variegated -- viraga of the interpolated Sabhaparvan passage) and vividhani (of different colours = nanaraga of the same passage). Mahabharata : Das Geschehen und seine Bedeutung, Bonn, 1975-206-207;224225. This work is not accessible to me. Mbh. 4.35.23. 7. Madhu Vidya/471

Page #497
The Custom of Sahagamana : some early references M. A. Mehendale Writing on the subject of sati about fifty years ago, the late M. M. Prof. Dr. P. V. Kane observed : "The subject is now of academic interest in India, since for over a hundred years (i.e. from 1829) self-immolation of widows has been prohibited by law in British India and has been declared to be a crime... We are now in a position to take a dispassionate view of the practice, to trace its origin and follow its working down to the date of its being declared illegal!". When Dr. Kane wrote these lines India was under the British rule. Now, some forty years after independence, the situation described above unfortunately does not seem to be quite true. The passions aroused by the Rupkanwar incident in Rajasthan a few years ago would not permit us to describe the subject of sati as of mere academic interest. However, it is certainly worth-while to look at the subject academically and try to interpret correctly the proper significance of the available references. After the unfortunate Rajasthan incident allusions were often made to the Vedic mantras and to the Madri-episode in the Mahabharata (Mbh.). It has therefore become imperative to have a fuller account of the Mbh. incident and place it before the public. . But, first, to start with the Vedic literature it may be observed that there is not a single case of self-immolation of a widow reported in the entire Vedic literature. We may further point out that there is no mention of self-immolation in the Avesta either. Hence we are entitled to conclude that the practice of self-immolation was not in vogue in the Indo-Iranian period. It is however necessary to state that there is a mantra in the Rgveda and in the Atharvaveda which together definitely point to the sati custom as a matter of distant past. The mantras in question occur among those which are used at the time of the funeral rites. The mantras have by now become quite famous. The Rgveda mantra runs as follows: udirsva nary abhi jivalokam gatasum etam upa sesa ehi/ hastagrabhasya didhisos tavedam patyur janitvam abhi sam babhutha // Rv. 10.18.8. The situation presented in the mantra is as follows : some one in the family has died (gatasu). Before the body is cremated, his wife lies beside the dead body Madhu Vidya/472

Page #498
(upa sese). Some one holds her hand (hastagrabha) and asks her to get up and come over to the world of the living persons: The holder of the hand apparently is to become her next husband. This mantra occurs also in the Atharvaveda. There it is preceded by another mantra which runs as follows: iyam nari patilokam vrnana ni padyata upa tva martya pretam-/ dharman puranam anupalayanti tasyai prajam dravina ceha dhehi // Av. 18.3.1 This mantra makes the above situation more clear. The woman lying near the dead body does so because she intends to obtain the world of her husband. This she does following an old custom (dharmam puranam). However, some one among those that have gathered round the dead body requests the dead person to let his wife have the progeny and the wealth left behind by him. These two mantras inform us that long before they were composed there was a custom following which the wife of the dead person, if she chose to follow her husband to the next world, lay beside him and was presumably cremated with him. This was the age-old custom (purana dharma) no longer in vogue at the time of the composition of the Vedic mantras. During the period of the mantras, the wife no doubt lay near the dead body of her husband as a relic of the old custom, but was asked to get up and marry again, and to bring to her new husband the progeny and wealth of her dead husband. As an aside it may be pointed out that among certain American Indian tribes the widow lies beside her husband's dead body as a relic of the old custom of selfimmolation but is allowed to withdraw when she is nearly suffocated (as reported by Tawney-Penzer, The Ocean of Story, p. 258). It is not known whether the custom still continues. It was observed a little while ago that there is not even a hint in the Kgveda of a widow committing the act of self-immolation. The later commentators on the Dharmasastras and the Nibandhakaras, however, who favoured this practice included the following mantra from the Rgveda in the sankalpa to be made by the widow at the time of committing sati. They did this to show that the practice of sati was prevalent in the Rgvedic period. The mantra in question partly reads as : ima narir avidhavah supatnih ... a rohantu janayo yonim agre //Kv. 10.18.7. The mantra simply states that the ladies whose husbands are living and who had apparently gone to the cremation ground may return home first. It has thus nothing to do with the widow-burning. The Asvalayana GThyasutra (4.6.11-12) and the Baudhayana Pitsmedhasutra(1.21.14) which prescribe the employment of this mantra have nothing to say about its relationship with sati. The commentator Sayana also does not say Madhu Vidya/473

Page #499
anything of the sort. But the writers who favoured the custom of sati read agne or agnehin place of agre,i.e. they read the quarter as arohantu janayo yonimagne or agneh. The quarter was thus forced to mean 'may the wives mount on the place of Agni'or 'oh Agni, may the wives mount on (your) place' thereby neglecting that if agne were vocative the accent of the end of the mantra would be totally different. It is surprising how the author of the Brahmapurana and of the commentary Apararka say that the Rgveda recommends self-immolation (rgvedavadat sadhvi stri na bhaved atmaghatini - Brahmapurana quoted by Apararka p.111). Apararka specifies that by rgvedavada we have to understand mantras like 'ima narir avidhavah ... Rv. 10. 18.7. It should be by now clear that the Vedic literature does not offer a single instance of any widow having committed sati. Next, it has to be remembered that none of the Grhyasutras like the Asvalayana and the Samkhayana, nor the eariler Smstis like those of Manu and Yajnavalkya prescribe immolation by a widow. The first clear statement that a widow should burn herself with her dead husband is found in the Visnusmrti of about the third century A. D. The statement runs as: mste bhartari brahmacaryam, tadanvarohanm va 25.14 It should be noted that this first statement does not make it obligatory on the part of the widow to follow the dead husband on the funeral pyre, it is mentioned as an alternative of second preference, and not the first. The author of the Visnusmrti would want the widow to lead the life of a celibate as her first preference. Explaining the reasons for this order of preferences a late text, Smrti Candrika of Devannabhatta (13th century A.D.), observes that the life of a celibate (brahmacarya) is to be preferred because it leads to the world of Brahman, whereas self-burning would secure heaven, i.e. a fruit of less value (tatrapi brahmacarye vyavasthanam uttamo dharmah brahmalokapraptikaratvat ... etad dharmantaram (i.e. anvarohanam) api brahmacaryadharmaj jaghanya nikrstaphalavat, V.R. Gharpure's edn p. 254). We may hasten to add that according to Angiras', whose opinion is cited in Apararka, even this second preference alternative was not to be resorted to by a Brahmana woman. If she were to commit sati that would constitute the selfkilling and hence would neither lead herself nor her husband to heaven (ya stri brahmanajatiya mstam patim anuvrajet / sa svargam atmaghatena natmanam na patim nayet)*. This situation however, changes in the 6th - 7th centuries A.D. when Vedavyasa in his smrti recommends sati not only to Brahmana women but gives self-immolation first preference over the life of penances (mrtam bhartaram adaya brahmani vahnim aviset! jivanti ... tapasa sosayed vapuh 2.53). Madhu Vidya/474

Page #500
Although some writers on the Dharmasastra had started recommending selfimmolation to widows as early as the third century A.D., their recommendation had not obtained universal approbation even upto the ninth century. This is clearly seen from the fact that the commentator Medhatithi on Manusmrti vehemently opposes this recommendation. While commenting on Manu 5.157, a stanzas which has nothing to do with sati, Medhatithi goes out of his way to assert that self-killing is as much prohibited for women as for men and a woman who ascends the funeral pyre of her dead husband clearly violates the Vedic injuncation that no one should give up his life before the end of the prescribed life-span (pumvat strinam api pratisiddha atmatyagah... asty eva patim anumarane' pi striyah pratisedhah / kimca 'tasmad u ha na purayusah preyat' iti pratyaksasrutiviruddho' yam.) But the commentators like Vijnanesvara and Aparaditya on the Yajnavalkyamsti who flourished in the 11th and the first half of the 12th century A.D. have very strongly supported the custom of sati. According to them the practice of self-immolation is meant for all women, irrespective of the caste. The only exception they make is those of pregnant women, and those who have young children to look after (ayam ca sarvasam strinam agarbhininam abalapatyanam acandalam sadharano dharmah-- Mita, on Yajn.1.86). These commentators as well as the authors of the Nibandhas like the Dharmasindhu have tried to explain away somehow the statements of the Smrti writers not favourable to them, Occasionally they have exaggerated the statements of earlier texts, or they have cited certain passges as Smrti passages which are not to be found in the extant Smrti texts. For example, we have seen that according to Angiras a Brahmana widow should not commit self-immolation. The author Kasinatha Upadhyaya (late 18th cent.) (Chaukhamba edn. p. 971) of the Dharmasindhu tries to get rid of this exception by interpreting the above statement to mean that a Brahmana woman should not ascend a separate funeral pyre (brahmanya nisedhavacanajatam tat prthakcitiparam/ bhartur mantragnidahottaram anugamanam prthakcitih). As an example of exaggeration we may point to what Aparaka has to say on Yajn. 1.87. In the Ramayana we have a single allusion on sati where Vedavati tells Ravana that her mother entered fire along with the dead body of her husband Kusadhvaja (tato me janani dina tacchariram pitur mama / parisvajya mahabhaga pravista dahanam saha 7.17.13). On the basis of this isolated instance Apararka does not hesitate to make the following sweeping statement : ata eva ramayanadau brahmanyadinam svabhartrsariralinganapurvakam svasariradaham upakhyayate. Having formulated the sentence as ramayanadau ...upakhyayate Apararka seeks to imply that many upakhyanas containing sati-incidents are available in the Ramayana and in other similar works. And by using the would brahmanyadinam he suggests that the women 'committing sati came from all the castes. But this is absolutely not ture. Madhu Vidya/475

Page #501
As to a smetivacana being unjustifiably ascribed to some ancient Smstikara one may cite the example of Kamalakarabhasta (17th Cent.) who in his Nirnayasindhu (trtiya pariccheda, uttarardha, p. 626, Nirnayasagara edn.) cites brahmacaryam cared vapi pravised va hutasanam as said by Manu (tatha ca manuh). But the stanza does not occur in the available Manusmrti. If we now leave the Dharmasastra texts and look to the two epics we find that the Ramayana contains a single instance in its late Uttarakanda of Vedavati's mother. This was alluded to a little while ago. There is no other instance of sati in the Ramayana. None of the wives of king Dasaratha commits sati at his death. It is therefore significant to note that there are few stanzas, considered interpolated in the critical edition in which Kausalya either gives expression of her wish to commit sati idam sariram alingya praveksyami hutasanam 2.60.* 1534) or says that she is not able to ascend the funeral pyre as she craves to see Rama after his return from the forest (nyayyam dharmyam yasasyam ca margam satstrinisevitam / adhigantum na saksyami ramasamdarsanasaya // 2. App. 20.6 ff.; 21. 11. ff.). It is quite clear that these stanzas must have been inserted when the custom of sati had gradually begun to gain ground, When we come to the Mahabharata we notice that the custom of sari was not uncommon among the Yadavas who had established themselves in the west. At the death of Vasudeva his four wives Devaki, Bhadra, Rohini and Madira ascended the funeral pyre (16.8.18.24), and at the death of Krsna five of his wives, viz.,Rukmini, Gandhari, Saibya, Haimavati and Jambavati entered fire independently, i.e. not with Kssna's body (16.8. 71).. In the Visnupurana the number of Krsna's wives who entered fire is raised from five to eight and they are said to have burnt themselves along with Krsna's body (5.38.1-2). There is no other Yadava woman reported in the Mbh. to have committed sati, although Visnupurana again reports Revati to have entered fire with Balarama's body (5.38.3).We can thus see that the custom of sati had gained more ground between the period of the composition of the Mausala parvan of the Mbh.and the composition of the Visnupurana. These instances of sati from among the Yadavas as reported in the Mahabharata are not so commonly known. The incident of sati from the Mbh. that is widely known is of course that of Madri. It has a two-fold significance. In the first instance it is the earliest instance of sati recorded in Sanskrit literature; and, secondly, Madri is the only woman in the Kuru family who is reported to have committed sati. It is therefore necessary to have a close look at it since, as we have seen above, committing sati is not recommended by the authors of the Dharmasastra until about third century A.D. There are two contradictory narrations available in the Mbh.regarding the incidents that took place after the death of Pandu. According to one narration, when Pandu died Kunti offered to ascend the funeral pyre as she was the elder wife (aham jyestha Madhu Vidya/476

Page #502
dharmapatni jyesthamdharmaphalam mama/... mamas madrinivartaya/anvesyamiha bhartaram aham pretavasam gatam/1.116.23-24). But Madri dissuades Kunti from her resolve and herself ascends the funeral pyre (ity uktva tam citagnistham dharmapatni nararsabham/madrarajatmaja turnam anvarohad yasasvini 1.116.31). Afterwards the sages along with Kunti and the five Pandavas go to Hastinapura and hand them over to Bhisma. They then told Bhisma and others that Madri had committed self-immolation (1.117.28).. This narration is the one which is usually known. But there is also a second narration available in the Mbh. which is completely contradictory with the one given above. According to this narration Madri did not commit sati. The entire narration is worth looking into. It is as follows: The sages who arrive at the court of the Kauravas tell that Pandu and Madri had died. They had brought with them the dead bodies of the two and proper funeral rites may be performed on them (tasyas tasya ca yat karyam kriyatam tad anantaram / ime tayoh sarire dve 1.117.29.30). Having heard these words of the sages Dhrtarastra asked Vidura to arrange for a state funeral of Pandu and Madri. We have a somewhat detailed description of this in Adhyaya 118 of the Adiparvan. We are told that Dhstarastra asked Vidura to see that Madri's body is fully covered so that it is not exposed to the sun and wind. When the funeral procession reached the banks of the Ganges Panqu's body was smeared with perfumes and bathed with oils. After washing it they again applied sandal-wood paste mixed with aloe. They then covered his body with white cloth. Pandu then appeared as if he was alive (acchannah sa tu vasobhir jivann iva nararsabhah / susubhe 1.118.20). Pandu's body was then smeared with ghrta and both the bodies were placed on the pyre made of candana and other fragrant wood and were cremated together (ghrtavasiktam rajanam saha madrya svalankrtam/tungapadmakamisrena candanena sugandhina lanyais ca vividhair gandhair analpaih samadahayan (118.21-22). When one reads the above description one immediately realizes that according to this narration Pandu's body was not consigned to flames on the Satasmga mountain where he died but this was done on the banks of Ganges at Hastinapura. One also realizes that Madri did not commit sati. We miss two important details in this second narration. One, we are not told what led to Madri's death. Two, we are not told how the two bodies were kept in tolerably good condition till they were brought to Hastinapura. We are left to imagine that Madri died due to the shock of Pandu's sudden death when he forcibly tried to embrace her (1.116.10,12) and that the sages knew how to preserve the dead bodies in good condition over a period of time. It is very likely that the story teller considered the two details not worth mentioning since they could be easily understood by his audience. Or, and this seems to me more likely, when the attempt was made to bring together the two Madhu Vidya/477

Page #503
contradictory versions these details were omitted in order to make the combination as smooth as possible. The two narrations so contradict each other that both cannot be true at the same time. However, Nilakantha, the famous 17th cent. commentator, has tried to bring about reconciliation between the two accounts by interpreting the word sarira in the second narrative as bone'.This would mean that the sages who came to Hastinapura brought with them not the two bodies of Pandu and Madri, but two bones representing Pandu and Madri. If we understand the passage this way the two accounts become reconciled. Madri did commit sati, and what the sages brought with them were just the two bones. The samskaras which were done on the banks of the Ganges were done on these two bones and not on whole bodies. But is it proper to interpret the word sarira this way to explain away the contradiction? Our answer to this question should be emphatically 'no' because there are many difficulties in accepting Nilakantha's interpretation. A simple objection is that the word sarira occurs not once but thrice in this account. If the author of the Mbh.had really meant 'asthi' by 'sarira', we expect that he would have at least once used the word 'asthi' and made clear what he meant. And the most important objection is that the description of the funeral rites given in the Adhyaya 118 does not read like the one performed on the bone, but performed on the whole body. Dhstarasira's instruction to Vidura to see that Madri's body is not exposed to the sun or to wind would sound ridiculous if it was meant for a mere bone. Similarly, the specific statement that at one stage Pandu looked as though he was alive could not have been made with reference to a bone. For this reason, the late Dr. V.S.Sukthankar, the first General Editor of the critical edition of the Mbh., does not accept Nilakantha's explanation (vol. I, p. LXXXVII). However, according to Dr. Sukthankar it is not possible to decide on the evidence of the manuscripts alone which of the two accounts could be the original one. Both these accounts appear in the identical form in both the northern and the southern recensions of the Mbh. and hence, in accordance with the principles accepted for constituting the critical text, both these accounts had to be accepted for the critical edition. This, however, does not mean that both the accounts are correct; it only means that the attempt to bring together into the text of the Mbh. the two contradictory accounts was made early enough to allow the inflated text percolate into the two recensions. It seems to me that although the manuscripts do not give help in deciding which of the two accounts could be nearer to the original, there are two kinds of other evidence, internal and external, which help us decide that the account according to which Madri did not commit sati is more likely to be nearer to the original. But before I narrate this evidence let me first consider a likely objection to my not accepting Nilakantha's interpretation of sarira as bone. Madhu Vidya/478

Page #504
The objection is based on the fact that in Sanskrit the word sarira does appear in the sense bone'.E.g. in the section of the Aitareya Brahmana which deals with he death of an Agnihotrin, a question is raised as to what should be done to the sacred fire if an Agnihotrin dies while on journey. One option is that the fire should be kept burning until his bones, and in the absence of bones, leaf stalks of a palasa tree (palasavsksasya chinnan vrntan, Sayana) are brought from the place where he was cremated. These bones (or the leaves) are to be then arranged in the form of a body of a man. In this context we read:asariranam ahartoh/yadi sarirani na vidyeran parnasarah sastis trini ca satany ahrtya tesam purusarupakam iva krtva ... Ait. Br. 7.2 (=32.2). Similarly, in the Sat. Br. 12.5.1.13, in a similar context, we hear of the bones of an Agnihotrin being collected and brought to the place of his residence : asthiny etany ahrtya ... purusavidham vidhaya. Here the word asthini is used in the place of barirani of the Ait. Br. and it clearly shows that sarirani means bones. But we cannot justify on this evidence Nilakantha's interpretation of the Mbh. word sarira as 'bone' for the following two reason : (1) In the first instance the Moh. stanza on which Nilakantha comments runs as tasyeman atmajan deham bharyam ca sumhatmanah svarastram grhya gacchamo 1.126.4 (Bombay edn) = Cr.edn. 1.117.1266*). In this stanza, which is not accepted in the Cr.edn., a reference is clearly made to Pandu's body with the word deha'. It is on this word, and not on the word sarira that Nilakantha writes deham dehayor asthini agnihotragnibhih samskaralambhanartham. Now although, as noted above, the word sarira in a definite context is found used in the meaning 'bone'the word deha is never so found. Secondly, in the description of the funeral rites that follows there is no reference to the samskara having been done with the agnihotra fire. Clearly Nilakanatha has interpreted the word deha as 'bone' under the influence of his knowledge of the agnisarskara of an agnihotri Brahmin. (2) Secondly, and this is a more important objection, the word sarira when in the above context is used in the sense of bone'it is always in plural, and never in singular. In the Mbh., in the above narration, the word sarira occurs thrice, but not once is it used in the plural. The plural use in the Ait.Br. indicates that many bones have to be collected to be arranged later in a human form. If any such thing was intended in the Mbh. the sages would have collected a large number of bones of both Pandu and Madri and not one bone each. Dr. Sadashiv A. Dange has tried to answer a part of the above objection to Nilakantha's interpretation. Writing on "Panlu, Madri and their 'Bodies'" he upholds Nilankantha's interpretation of sarira (sg.) as 'bone' and argues that there is no contradiction involved in the two accounts of the Mbh. and that Madri did commit sati. He maintains that sarira is after all used in the sense of bone'even when used in the singular. He says that"...we also have the word sarira(singular) indicating a jar wherein MadhuVidya/479

Page #505
the bones (sarirani) were collected." (p. 3). He cites from the Baudhayana-pirmedhaSutra (3.10) where the jar in which the bones (asthini pl.) of the cremated person are collected is addressed as sariram (sg.). Dr. Dange also refers to the Katyayana - Srauta Sutra 21 3.7-13, and the Sankhayana-Srauta-Sutra (4.15.20), but in both these passages we do not have the use of the word sariram (sg.) in the sense 'bone"". It will therefore be clear that the word sariram occurs only once in the singular and there it refers directly to the jar (kumbha) in which the bones are collected and not to the bones themsleves. The kumbha in which the bones are collected represents the body which has bones. Hence this occurrence of sariram in singular cannot be given as evidence to prove that in Sanskrit sariram in the sg. is used to convey the meaning 'bone'. In order to be able to derive support from the Baudh. Pitr. Sutra we have to assume that when the sages told the Kauravas ime tayoh sarire dve (1.117.30) they gave them two jars in which the bones were collected. If this were really the case the narrator would have most certainly mentioned that the sages collected the bones of Pandu and Madri in two separate jars and brough them to Hastinapura. The narrator was probably quite familiar with this mode of re-cremation. But the narrator does not say anything of the sort. Nor does he say that the sages asked the Kauravas to do the punahsamskara (Baudh. Pitr. Sutra 3.8) of Pandu and Madri. Nilakantha on whom Prof. Dange so much relies does not make any reference to the jar. He simply says deham dehayor asthini, thereby interpreting the word deha straight away as asthini without bringing in the concept of a jar. For him the jar does not figure at all. Otherwise he would have said something like deham dehayoh kumbhayoh avahitani asthini. Next, we have to note that in order to reconcile this 'jar' theory with the detailed account of the antyasamskara in the Mbh. Prof. Dange has to assume that the two small jars were remade into big jars which were later covered with cloth and decorated with ornaments to give them the semblance of actual bodies (p. 5). From this assumption I am glad to note that Prof. Dange shows awareness of the fact that the small jar in which the collected bones are supposed to have been brought could not look like bodies even when covered with cloth and ornaments. Hence the assumption of re-making them into big jars for which, as Prof. Dange notes, there is not even a distant allusion. But, in spite of this, Prof. Dange has no hesitation in giving his judgement on the silence of the text on this importatant point as 'unnecessary' (p. 5). 1 for one do not understand how a jar can be adorned with ornaments and even if this is somehow achieved how it can give the semblance of an actual body. All-in-all, this jar-theory of Prof. Dange, beginning with Nilakantha's changing deham (sg.) of the epic text to dehayoh (dual) and paraphrasing it as asthini and ending with remarking of small jars into big jars, is saddled with too many difficulties to inspire confidence. If the critically admitted text of the Mahabharata had only this single instance of internal contradiction between the two accounts of Pandu and Madri, perhaps, one would have accepted Nilankantha's suggestion to get rid of it as there was no other way out. But since the critical edition presents many instances of such Madhu Vidya/480

Page #506
contradictions I have no hesitation in accepting this as one of them instead of trying to resolve it by impossible assumptions. I have already admitted that in the second narration we do not get information on two points : how Madri died and how the two bodies were kept in tolerably good order until they were brought to Hastinapura. As regards the first point it is easy to assume that Madri died of a serious shock since it was in her arms that Pandu died. Since in the first narration it has been stated that Madri committed sati, the redactor responsible for the combination of the two versions could not retain the stanza in the second narration informing us that Madri died a natural death. Hence I had said in my Marathi article that this part of the narration is lost, and I still favour that view although Prof.Dange has tried to make fun of it. As regards the second point, I would say that either the tapasas living on the Himalayas knew some method of preserving the dead bodies, or that here some element of supernatural is involved since the sages who brought the bodies appear to be siddhas since they are said to have disappeared (ksanenantarhitah sarve .... tatraivantarhitam punah/rsi-siddhaganam drstva vismayam te param yayuh 1.117.3233) after handing over the two bodies to the Kauravas. So much for the objection raised by Prof. Dange and his thesis that in the two accounts of the Mbh. no contradiction is invovled. I now return to my own thesis and present the evidence -- internal and external -- to show that the narration in which Madri is represented as not having committed sati is likely to be the original one. As regards internal evidence I wish to point out that no woman, other than Madri, from the Kuru family is reported to have committed sati. Many Kaurava heroes fell in the battle. But none of their wives or of the families like the Pancalas and the Matsyas committed sati. Hence it is almost certain that Madri did not commit sati. It is true that some of the Yadava women committed sati after the death of Vasudeva and Kisna. This only shows that the custom had stared raising its ugly head in the family of the Yadavas who were located in western India. But on that single evidence we need not assume that the custom of sati had obtained acceptance in other families and in the other parts of India. The external evidence is offered by two independent texts - (1) The Kadambari of Banabhasta (7th century A.D.) and (2) The writing of the Greek Historian Diodorous (1st Century B.C.). While trying to dissuade Mahasveta from her resolve to die following the death of Pundarika, Candrapida tells her that it is futile to give up life at the death of one's father, brother, friend or husband. If on such occassions the breaths do not leave the body themselves, they should not be made to leave (svayam cen na jahati na parityajyah, Nirnayasagaredn.p.364). In order to strengthen his contention Candrapida then cites the examples of Rati, Kunti, Uttara and Duhsala who did not give up their lives at the death Madhu Vidya/481

Page #507
of their respective husbands. If Banabhatta had accepted the sati version of the Madri episode he would not have alluded to Pandu's death at all while citing the instances of women who did not die after their husbands' deaths. Therefore, he must have been familiar with the version according to which Madri died due to grief, and Kunti did not commit sati. Hence, he cites Kunti's example to show that a woman does not abandon life, if death does not come naturally to her as, presumably, it came to Madri, at the death of her husband. Diodorus, the Greek historian of the 1" cent. B. C. informs us that a certain leader of an Indian continent had gone to Iran to fight under the leadership of Eumenes (316 B.C.). When he went to Iran his two wives accompanied him. When the Indian leader lost his life in the battle his two wives vied with each other as to who of the two should commit sati. The dispute between the two was referred to the Greek general for decision. Since the elder of the two was with child, he permitted the younger one to commit sail (Cambridge History of India 1.415). The similarity between this incident and the sati-version reported in the Mbh, is obvious. In both cases the two wives are equally eager to commit sari, they have an argument each one justifying her claim on one ground or the other, and ultimately the younger one commits sati. It is remarkable that in both cases the issue is settled on the ground which is related to the progeny. Of course there is a difference. Whereas in the incident of Iran the elder wife is not allowed to commit sati because she is pregnant, in the Mbh. incident Kunti is dissuaded because she was better able to take care of children without making any distinction between them. On account of this similarity I am inclined to believe that, when this incident which took place on the Iranian soil became known in India, it occurred to some narrator to fabricate a different version of the incidents that took place after Pandu's death and incorporate in it self-immolation by Madri. It should be quite clear that such a thing could have happened only when the custom of committing sati was slowly revived in India. In the post-epic Sanskrit literature, especially Rajatarangini of Kalhana and in the inscriptions dating from the Gupta period, we find references to the custom of sati. It is not necessary to give here a list of them.10 In conclusion it may be said that the custom of sati which was prevalent in the very distant past among the people who came to India had became obsolete at a time very much prior to the Vedic period. Many centuries later the custom raised its head again, apparently first in Western India, with the result that we have stray instances of sati from about the fourth century B. C. When the custom gradually spread in the first few centuries of the Christian era, it was initially restricted to the Ksatriya class. Subsequently it was recommended also for the Brahmanas. After the tenth century A. D. the writers on the Madhu Vidya/482

Page #508
Dharmasastra vigorously wrote in favour of the custom and prescribed it to women of all classes without distinction. Even then the recommendation of these writers in favour of sati never became an obligatory rule in India and the custom was adopted only as an option. When Lord Bentinck prohibited the inhuman custom it was feared that there will be widespread opposition. In fact nothing of the sort happened. Writing about this Dr. Kane observes : "The very fact that there was no disturbance of peace or ebullition of popular feeling or even great protest from the vast Hindu population (except a petition to the Privy Council) against Bentinck's sweeping measure indicates two things : viz. That the burning of widows was a rare occurrence and that people were not very keen on observing the practice nor had they any very deep rooted conviction about its absolute religious necessity."(History of Dharmasastra, II. 1. p. 636). NOTES: 1. History of Dharamasastra, Vol. II, pt. 1, p. 624 (1941). 2. Cf. svargaloke mahiyate said by Angiras. 3. Apparently different from the author of the work now available as Angirahsmrti. 4. This is also the view of Paithinasi, Viraj, and Vyaghrapad, cf. Apararka p. 112. 5. Kamam tu ksapayed deham puspamulaphalaih subhaih na tu namapi grhniyat patyau prete parasya tu // 6. Especially when we are told that it took the sages seventeen days to reach Hastinapura (1.117.27) 7. 'Bharati' (Bulletin of the College of Indology,Banaras Hindu Unviersity, Varanasi) 17 (1987-88) pp. 1-6. 8. Prof. Dange has to rely on the commentator. 9. Actually what I have said in my Marathi article is as follows: "he dona tapazIla mahAbhAratAcI kathA sAMgaNAr2yAne kadhI sAMgitaleca navhate kI, te tapazIla sAMgaNAre zloka OHTOEHT JETTA 154 SI, FUTUONUT IT."7917", urtart, 98cc, q. x) 10. They can be easily found in A. S. Altekar's The Position of Women in Hindu Civilization, 1956, pp. 129-32, and in P. V.Kane's History of Dharmasastra, II, 2. pp. 628-30. Madhu Vidya/483

Page #509
DAMAYANTI'S SVAYAMVARA By M. A. MEHENDALE Writing on the above subject S. Insler! suggests a new interpretation for the Mbl. stanza 3. 54. 26. His interpretation is based on two suggestions both of which are uncalled for. The stanza in question runs as vilajjamana vastrante jagrahayatalocana / skandhadese 'spjac casja srajar para masobhanam || Insler objects to van Buitenen's somewhat free translation of srajam assjat 'threw a garland' because srj doss, not mean to throw but 'to releasc' 11:, however, on that account does not opt for the most natural translation released the garland'. On the other hand he cites two passages from the Albh. 3.264, 33 and Calcutta cdn. 1. 4418 ( = Cricdn. 1. 1131*) in which he gets the expressions malam asajat and srajam samasajat. Insler translates these as hung a garland'.: On the basis of this evidence Insler proposes to cmend ( srajam ) asr jar of the above Mbh. stanza to (srajani) usajat. Is this cmendation called for? Insler admits that flowers can be released' (sti) from the hand. Then what is so impossible in the expression 'release ( sri) the garland from the hands' even if the expression varsa' (shower) is not found with it? Insler himself cites on p. 578 from Pali Jat. I. 8 malari parikkhipitva, but instead of drawing the correct conclusion that to throw a garlaad ' or 'to release a garland' are fairly old and current modes of expression he imagines that the Pali expression malani parikkhipa is influenced by the wrong' Sanskrit expression srajan srj. The two verbs srj and a - or sama-saj express two distinct and successive actions connected with garlarding a person. First comes the act of releasing ( stj) the garland from the hand and then, if necessary, the act of sticking it, adju IJOS 109. 577-580 (1989). Strictly speaking u saj or sa saj also does not mean 'to hang' ( which means 'attach loosely, suspcnd' ), but 'to stick, to fasten'. Madhu Vidya/484

Page #510
484 ABORI: Amslamahotsava Voluinc sting it, putting it in proper place ( saj) % Hence there is nothing wrong in describing that Damayanti released the garland on Nala's shoulders as there is nothing wrong in saying that Kunti stuck the garland on Pandu's shoulder (Cal. edn 1. 4418 cited by Insler ). The author of the Mbi, stanza 3. 54.26 also could have used asajat as found elsewhere, but since that was metrically not possible he chose to say asrjat which was quite suited to the context. But to say that the original asajat was entended to aspjat by some one in the line of transmission just because there was an r sound in the following word ( srajam) is suggesting something which is extremely unlikely. We may ask . if some one in the line of transmission changed asajat to asrjat due to rin srajam, why did he not change avasaktah (Mbh. 1. 37. 3 cited by Insler) to avasts!ah when there was ! ( not just r) to preceda (mrgayam ) as well as to follow ( msia! )? Insler gets his idea to change asrjat of the Mbh. 3. 54.26 to asajal on the basis of the two Mbi. passages, viz, 3. 264.33 and Cal. edn. 1.4418. But he neglects the fact that for the citation from the Cal, edn. we have a variant samaspjat (Cri.cdn. 1,1131") for samusajat. Of the two, Insler picks up the latter simply because that suits his purpose, although th: former is more widely attested. (2) Insicr also fects that the original stanza had vastrantam (acc.), which was later changed to rastranto ( 102.). His search for the motivation of this change leads him to skandhadles (loc.). In accepting this solution Insler assumics too much ignorance on the part of the person allegedly responsible for the change. If this person knew enough Sanskrit to recognize skandhadesc to be loc, sg. he certainly know that vastrantain, if it cxisted in thic text before him, was to be construed with jagrala, and therefore could not . think of changing it to vastrante to bring it in line with skandhadese. Insler interprets rastranta as ni vi end knot by which a woman's lower garinent is secured' and says that Damayanti grasped it to signify her sexual submission to Nala. This mians Damayanti grasped her nivi in the presence of all those that had gathered for th: srayarnvara to indicatc that she would be always ready to loosen it whenever Nala so desired. This is ridiculous. No princess would dream of doing such an act, least of all Damayant) who This is reflected in the use of the two terms avusriut (Insler's asrjat to be so corrected ) and avasuklah used in the stanzas Mbl. 1. 46, 10 and 1. 37. 3 in describing the sanic 'incident (cited by Insler p. 578). Madhu Vidya/485

Page #511
MEHENDALE : Damayanti's Svayanivara 485 is described as vilajjamana.feeling bashful. She is not like Indumati who, on a similar occasion, had almost set aside her bashfulnes (lajja taruky. tya, Raghu. (Nandargikar's edn. ) 6. 80. And what did Indumati after all do after setting aside her bashfulness? She merely fixed her eyes on Aja (drs!ya kumaram pratyagrah! : ) to indicate her choice, and did not do any. thing to indicate her sexual submission to him. Insler wrongly construes vastrante with the following jagraha and therefore wants to chanage vastrante (loc.) to vastrantam (acc.) which would then serve as an accusative complement to the verb jagraja. Vastrante is really to be construed with the preceding vilajjanana. The accusative compleinent for jagraha is srajam (and not the imaginary original reading vastrantam) which is also the accusative complement for asrajat. This is clearly shown by the use of ca (ayatalocana srajani jagraha skandhailesc assjac ca). Damayanti took the garland (apparently from the plate or from the hands of some one in her retinue ) and released it on the shoulders of Nala. Vilajjamana vasarante feelingsly in the end of the garment' means Damayanti expressed her shyness by covering her face a little within the end of her garinent. Drawing forward on the face the cnd of the garment as an act of modesty or as an expression of bashfulness is even now practised in parts of India. Understood this way the whole stanza gets casily construcd without having to do violence to the text and it gives an excellent picture of the conditiou of Damayanti. This picture has been totally distorted by Insler's suggestions, As for the Raghuvanisa stanza VI. 83 cited by Insler, a dillercat intcr. prctation, more plausible than the one admitted so far, is possible. The stanza says : sa (Indumati)... dhatrikarabliyam ... asanjayam asa yathapra # If Damayantidad touched her hivi as imagined by Insler, licr both hands would not be free to hold the garland and hang it on the shoulder of Nala. F. Bopp (Nalus Vaha.bharati Episodium, 1868, p. 31 ) also construes srajain as the accusative complement both of jagriha and asrjat (" ... Vestis.extremo prehendit ... in humneris deposuit ci scritum ...") although he uoderstands vestrante jagraha differently ( grasped in the cad of the garment "). (I am thankful to Dr. A. M. Ghatage for bringing this to my notice). But it is extremely unlikely that Damayanti would first grasp the garland in the end of her garment and thea deposit it on Nala's shoulder. G Cf. su vastrasalnyuddhamukhi lajja ya janasaisadi, Rimu. 6. 102.34. Those who changed the original vostronte to vastrantam as witnessed in the two MSS, K and B4, also must have understood vilajjamani vastrinta jagraha as feeling bashful (Damayanti) licld the end of the garment (to draw it a little forward over her face), Madhu Vidya/486

Page #512
436 ABORI: Amptamahotsava Volume desam kanthe gunam. This has been usually taken to mean that Indumati did not place the garland herself on the neck of Aja but asked her nurse to do it (cf. Mallinatha - dhatryah* ... karabhyani ... asaktam ka rayam asa / na tu svayam asasanja / anaucilyat). But this is very unlikely. What impropriety does Mallinatha see in the princess putting the garland herself around the neck of the prince is not clear. The correct understanding of the whole situation is that Indumati first indicated her choice by favourably look. ing at Aja (6.80). A little later she herself put the garland on his neck. When this was done she asked her nurse only to adjust the garland so that it stuck at the proper place (yat hapradesam 6.83). At this moment Aja felt as if Indumatl had put her arms around his neck (6,84). This description of the condition of Aja would be appropriate only if Indumati, and not her nurse, put the garland. The stanza from Indische Spruche 1780 (688) cited by Insler to prove his point has no place here. One wonders how lajjaksti can mean who feigns modesty.' Even if the compound is supposed to have the clements lajja + aksi it would mean .shame incarnate'. But the compound is really made up of lajja + kstilo and means who has done an act that invites shame' (lajjavaha kstir yasya ). The contrast between nirlajja and lajjaklie in this stanza is not to be understood as one though immodest is feigning modesty' - what sort of modesty could the fellow feign when he was actually dragging the garment and forcibly trying to kiss the beloved ? -- but as one having no feeling of shame although he had done a shameful act.' In the citation, Insler's mErtim to be corrected to murtam. 8 Malliaatha identifics the nurse as Sunando ( dhatrya upamilih sunandayan ). This is unlikely. Earlier Sunanda is specifically said to be the holder of the case in her capacity as a door-keeper or a harem-watcher (ustragrahane niyukta 6.26, vetrablert 6.82; dauvariki 6.59: pratikararuki 6.20, 35; suddhuntarakti 6.45). The nurse therefore must be some one else in Ioduinati's retinue (parivira 6.10 ) and not Sunanda. 8 Ipsler's cumbati in this citation to be corrected to cumbasi. 10 It seems that the only other attestation for bajakrti is in the compound lajjaktlinata. syakah (Skundap. 4.98. 67 ) where, obviously, lajjakrti means a shameful act, an act that brings shame'. Madhu Vidya/487

Page #513
THE GAME OF DICE IN ANCIENT INDIA (Some additions and corrections to the article of H. Lueders on this subject) M.A. MEHENDALE Long ago (1907) H. Lueders wrote a fairly exhaustive article on the game of dice in ancient India (reprinted in Philologica Indica. 1940). The article was noticed by W. Caland in ZDMG 62.123 f. Some of the corrections suggested by Caland, especially the one concerning the meaning of vicinoti, were accepted by Lueders, cf. his Berechtigungen und Zusaetze, Phil. Ind.pp.785-786. Although Lueders in his article did not succeed in clarifying every detail of the game, he certainly did all that was possible for him to do in the circumstances and made many ingenious suggestion of lasting value. It is proposed to make a few additions and suggest some corrections to Lueder's article. 1. In the AV 7.50.5 a gambler tells his opponent : ajaisam tva samlikhitam ajaisam uta samrudham. Lueders observes : "Die Bedeutung von samlikthita und samrudh is voellig unklar"! But in the light of the information given by Lueders (a circle was drawn round the players and the one who lost the game could not come out of the circle unless he paid his dues)?, we may translate the above line as : "I have won you around whom a line has been drawn, I have won you who has been confined." It may be noted that the Paippalada text (19.9.7) has samvstam'enclosed' in place of samrudham which supports this interpretation. If this understanding of the passage is correct, we can say that the custom of drawing circles around the players at the time of the game reaches back to the Vedic times. 2. pravateja - : In RV 10.34.1 the vibhitaka fruit are described as pravatejah. The usual explanation of the word, which Lueders accepts, is 'grown at a windy place' (am windigen Orte geboren, Phil. Ind. p. 122)'. Since 'growing at a windy place' does not look like a significant attribute of a plant or a tree, the explanation given by Yaska (Nir. 9.8) pravaneja 'growing on slopes' is more likely. Yaska apparently derives pravata from pra-vat 'slope'. 3. The technique of the game : This has been very well described by Lueders on the basis of the available evidence. He, however, does not specify whether a player had to separate some minimum number of dice from among those thrown on the adhidevana by his opponent. In the absence of such a condition, it would be easy for any player, with some practice, to separate at once just four pieces, or perhaps even eight, and win the game. Hence, it is essential that a certain minimum Madhu Vidya/488

Page #514
number of pieces must be separated by the player to win the game. This minimum number could be twelve. The assumption is supported by the description of the game played at the time of the Agnyadheya as described by the Baudh $S 2.9. There the sacrificer plays the game with his three sons using 49 pieces of the Vibhitaka fruit. The father and the two elder sons who must win the game take twelve pieces each and win, leaving 13 for the youngest son who loses. If there was no rule about the minimum number of pieces the game at the Agnyadheya could have been played even with 17 or 33 pieces (the father and the two elder sons together removing 12 or 25 pieces respy, leaving 5 or 9 for the youngest). 4. AV 7.50.5 : We read here avim vrko yatha mathad eva mathnami te krtam "I shall crush your krta (throw) as a wolf tears to pieces a sheep". Here a gambler seems to challenge his opponent who has won the game by making a krta and wants to nullify the opponent's success. How do we understand this challenge? According to the mode of play described by Lueders a game comes to an end when the two players in turn have thrown the dice on the ground. In such a game, the loser has no chance to undo what the winner has achieved. Lueders (Phil.Ind.p.154.f.n.2) takes the above AV. line to mean that the speaker wants to prevent his opponent from making the kTta aya by reciting some magic formulae. This can hardly be correct. The verb mathnami expresses drawing apart what actually exists before the speaker and not preventing something from happening. Moreover, understood the way Lueders has done,comparison with a wolftearing to pieces a sheep is out of place. We have, therefore, to say that the loser while issuing the challenge wants to add to the opponent's dice, already thrown on the ground, some more which would also result in a kTta. In such an event, the challenger's krta will have nullified the winner's krta. Such an understanding of the stanza means that it was open for the loser to defeat the winner by throwing dice once more, In this case, the total number of turns the two players would have becomes three and not two as in the unchallenged game. 5. Game in which division by five is decisive : Lueders (Phil. Ind. p. 159) rejects Weber's view that the placing of five aksas on the hand of the king in the rajasuya sacrifice implies that here we have an allusion to a game in which the number of dice is to be divided not by four but by five. According to Lueders the placing of five aksas has nothing to do with the game of dice and that the number of dice, which is five, is detemined by the number of the direction(s) which the king is supposed to make symbolically subservient to him. That the number of dice is determined by the number of directions is correct. But it is not correct to say that here the aksas have nothing to do with the game of dice. Just as in the RV 1.41.9 one who has four dice in his hand arouses fear in the mind of the opponent (because in that kind of game the number of dice was to be divided by four and the holder of Madhu Vidya/489

Page #515
four dice has made krta = the winning throw), similarly in the rajasuya the king who holds five dice in his hand becomes the overpowerer of all (because in this kind of game the number of dice was to be divided by five and five dice meant the winning throw). 6. asphura : In the Mbh. 2.51.3 Sakuni speaks about the game of dice in terms of archery. He says : "Know the glahas to be my bows, dice the arrows, the aksahrdaya my bow-string, and astara my chariot." Lueders had before him the reading asphura of the Bombay edition in place of astara, hesitatingly adopted in the Poona critical edition. Lueders takes asphura to mean adhidevana, i.e. the place where the dice are thrown (Phil. Ind. p. 118). Nilakantha explains the word as aksavinyasapatanadisthana which, according to Lueders, is correct. This interpretation of the Mbh. word asphura cannot be correct. Since Sakuni identifies asphura with his chariot (ratha) and a chariot is a place from where the archer shoots the arrows, asphura must be taken to mean the place near the adhidevana where a gambler sits and from where he throws the dice and not the adhidevana itself. The critical edition reading astara possibly implies that the seat of the gambler was covered with a piece of cloth. That the analogy ofasphura with the chariot was not missed by Lueders is clearly seen from the fact that he translates the word (p.119) as 'Wuerfelplatz'. But "Wuerfelplatz von dem aus der Kaempfer die Pfeileder Wuerfel abschiesst" is quite different from adhidevana which is "eine Vertiefung im Boden, innerhalb dessen die Wuerfel niederfallen mussten" (p.116) 7. dyutamandala: This word occurs in the Mbh. 2.70.600* (=Bom.edn. 2.79.32) and in the Harivamsa 89.49 (= Bom. edn. 2.61.54). At both these places Lueders (Phil. Ind. pp. 113-114) takes it to mean the circle drawn round the gambler which he cannot leave unless he has paid the dues. This interpretation is not likely to be correct. In the Mahabharata the word appears in the context of Draupadi being forcibly taken to the sabha and being dragged around there (gamanam parikarsam ca Krsnaya dyutamandale). If Lueders' interpretation is correct we will have to assume that after Draupadi was taken to the sabha she was made to stand in the circle where earlier Yudhisthira sat while playing the game. But there is nothing in the Mahabharata to warrant such an assumption. Draupadi was brought in the sabha to get the answer to her question directly from the members sitting in the Hall. There was thus, no occasion for making Draupadi stand in the gambler's circle. Hence dyutamandala in this passage must simply mean dyutasabha,i.e.,the place where the game was played. Such an interpretation agrees with another stanza where the dragging of Draupadi is alluded to. It reads as pancalim apakarsadbih sabhamadhye tapasvinim 2.72.12. In the Harivamsa the word dyutamandala is used when it is reported that Balarama killed Rukmin with an astapada in the dyutamandala. In the game of dice Madhu Vidya/490

Page #516
played between Balarama and Rukmin a dispute arose as to who of the two won the game. Enraged Balarama struck Rukmin with the golden dice-board (Hariv. 89.42.43). It is mentioned that Balarama crushed Rukmin on the ground (nispipesa mahitale). It is difficult to believe that all this happened strictly within the space circumscribed by the circle. Since there was no agreement about the outcome of the play both the players were free to come out of the circle and fight out. Hence in this passage too dyutamandala= dyutasabha. 8. Mbh. 3.56.8 (=Bom. edn. 3.59.8): In the Nala episode of the Mbh. we are told that when Puskara challenged Nala for a game of dice Nala, first, did not accept it. But when Puskara persisted in challenging Nala, the latter could bear it no longer and agreed to have a game with him. In this context, we read : na caksame tato raja samahvanam mahamanah / Vaidarbhyan preksamanayah panakalam amanyata // Since Lueders assumed that in ancient India people believed that a gambler could win the game on the strength of the faithfulness of his wife he interpreted the above stanza to mean that Nala agreed to Puskara's request and decided to play while Damayanti was looking on, i.e. in her presence, because he was confident of victory if the game was played in the presence of his faithful wife. Actually Nala lost the game because, according to L'ueders, Kali had entered Nala's body and therefore Damayanti's chastity was of no avail to him (Phil. Ind. pp. 112.113). This interpretation cannot be correct. If Nala thought of defeating Puskara at the game, not on the strength of his own skill, but on the strength of Damayanti's chastity, he would have doubted her character when he lost the game. He at that time did not know that he was possessed by Kali and therefore could not have attributed his defeat to the presence of Kali within him. But the defeated Nala neither expresses surprise at his defeat even when the game was played in the presence of his virtuous wife nor does he cast any reflections on her character. Besides, if we admit Lueders' interpretation we shall have to admit that to Indian mind, Kali was more powerful than a virtuous woman. This can hardly be a correct view. In fact in this episode we are told later that Kali was burning inside the body of Nala due to Damayanti's curse (3.74.18) which demonstrates who, of the two, was considered stronger. The correct interpretation of the stanza can be had on the basis of Panini 2.3.38 sasthi canadare. The stanza means that Nala thought that the time for declaring the pana had come (panakalam amanyata), i.e. he agreed to play the game of dice, in spite of the fact that Damayanti was looking on, i.e. disregarding her presence. On the strength of this epic statement the only conclusion that we may draw is that women in general, and Damayanti in particular, did not favour the husband's playing the game of dice and in spite of this Nala gave in to Puskara's persistent challenge. Madhu Vidya/491

Page #517
NOTES: 3. 1. Phil. Ind. p. 154, n. 1. 2. Phil. Ind.p. 114. Durga on the Nir. 9.8. has a similar explanation, but he also adds pravrtkale during the rainy season.' 4. Maitr. Sam.4.4.6; Taitt. Br. 1.7.10.5; Sat. Br. 5.4.4.6; Katy SS 15.7.5. He could have added that Durga also on the Nir. 5.22 gives conversely asphara as the meaning of devana i.e. adhidevana. The Naradasmsti 17.5 asuddhah kitavo nanyad asrayed dyutamandalam can very well mean: "a gambler who has not cleared himself of the dues should not go to another dyutasabha." Madhu Vidya/492

Page #518
Has the Vedic Rajasuya any Relevance for the Epic Game of Dice?* M.A. Mehendale Narada, who visited Yudhisthira's sabha told him that king Hariscandra, after conquering the entire earth and subduing all the kings (2.11.53, 554) had performed the great sacrifice Rajasuya. Only a king who had performed the Rajasuya got a place in the sabha of Indra (2.11.62). Hence Pandu sent a message to Yudhisthira to perform the Rajasuya (2.11.65-66). Acting on this message, Yudhisthira performed the Rajasuya. For this purpose the Pandavas collected an enormously large amount of wealth by way of tribute from the kings. Soon afterwards they lost everything to the Kauravas in the game of dice. Why must this happen? Prof. van Buitenen in the Introduction to his translation of the Mahabharata' observes that we are entitled to raise the above question. He asks : "Why, when everything has been achieved, must it now be gambled away by the hero in all whose previous life there has not been so much as a hint of a compulsion to gamble, all of whose life has in fact been of exemplary rectitude and prudence? It is this disturbing contradiction in the character of Yudhisthira that demands the question whether this was indeed a contradiction, or whether the events in his life may not have been modeled on a preexisting structure". His conclusion is : "In my opinion there is such a model : the events of The Assembly Hall follow fairly closely the principal moments of the very rajasuya ritual that is central to the book". (p.5) Madhu Vidya/493

Page #519
62 M.A. Mehandala What are the principal moments of Yudhisthira's Rajasuya which van Buitenen thinks "follow fairly closely" those of the Vedic Rajasuya? To start with he argues : "Krsna advises him (i.e. Yudhisthira) that this (i.e. the performance of the Rajasuya) is not a mere matter of personal volition but is dependent on the consent (anumat) of the entire baronage. ... in the Vedic ritual of the rajssuya, too, the ceremony opens with a prayer for consent, which is concretized in an offering to the Goddess of Consent, Anumati" (p. 14). What Krsna really told Yudhisthira was that as long as Jarasandha who had conquered the kings and made them captive was alive it was not possible for Yudhisthira to perform the Rajasuya. It was therefore necessary for him to kill Jarasandha and free the kings (2.13.61-62, 66). One who conquered Jarasandha would surely be the emperor (samraj 2.14.20). Later, when Jarasandha was killed and the kings freed, they asked Krsna what they might do for him. Krsna asked them to offer their assistance to Yudhisthira who wished to perform the Rajasuya. The kings willingly agreed (2.22.94-37). If van Buitenen wants to interpret the readiness of the kings to assist Yudhisthira in his Rajasuya as a "close parallel to the offering to Anumati (Consent, Good Will) in the Vedic ritual, he might as well have interpreted the elimination of Jarasandha, a potential threat to the Rajasuya, as a "close parellel" the accompanying offering to Nirsti (Destruction, Opposition) in the Vedic ritual. Apart from the fact that there is nothing in the Vedic ritual to suggest that the offering to Anumati symbolizes the consent of the barons, Yudhisthira would have been bound to make an offering to Anumati, even after securing the support of the kings, if he was following the Vedic Rajasuya. But the epic has nothing to say about it. The next item in van Buitenen's list of the principal moments" showing close relationship between the epic and the Vedic Rajasuya is the giving of the guest gift to Krsna (2.93.22, 27-31). Since the gift was made only to one person, van Buitenen sees in this "moment" close following of the Vedic rite of bestowing the remnant of the unction water to the heir-apparent of the king who performed the Rajasuya. Seeing these two events as closely related is as far-fetched as one can ever imagine. In the first instance one fails to understand why van Buitenen considers the giving of gift to Krsna as one of the "parting gifts" to the assembled guests. He knows that when this gift was given to Krsna the sacrifice was by no means over (p. 23). How can then this be a parting gift just because it was given after the unction? Madhu Vidya/494

Page #520
Has the Vedic Rajasuya any Relevance for the Epic Game of Dice? Secondly, why does van Buitenen take for granted that the gift was made only to Krsna and to none else? He asserts: "Also no mention is made of any other kings being presented with parting gifts" (p. 22). If this is so, why does he choose to describe the gift given to Krsna as "the first one in a series of parting gifts to the kings who have been Yudhisthira's guests at the Consecration"? (p. 22, italics mine). What he says here is, however, correct because Bhisma has clearly in mind giving gifts to all the assembled kings in accordance with their precedence (kriyatam arhanarh rajnam yatharham; esam ekaikaso rajann arghyam aniyatam 2. 33. 22, 25). There is nothing to prevent us from assuming that this was what Yudhsthira actually did (yatharham pujya nrpatin 2. 42. 38). There is therefore no justification in connecting the guest gift given to Krsna, and after him to the rest of the assembled kings, and the handing over the remaining unction water to the heir-apparent. Thirdly, again on van Buitenen's own admission (p. 23), if suzerainty (samrajya), unlike the king's inherited right to rule his own dominion (rajya), "is not strictly an inheritable office" how could Yudhisthira bestow his samrajya away to Krsna if that was the meaning of his giving argha to Krsna? 63 Finally, it is highly absurd to imagine that when Bhisma asked Yudhisthira to give the argha to the assembled kings he wanted Yudhisthira to show his preferences for receiving the title of samraj after him. The third and the last "principal moment" common to the two Rajasuyas seen by van Buitenen is the challenge issued by Sisupala to Krspa in the Rajasuya of Yudhisthira and the king's chariot drive in the Vedic Rajasuya (p. 23). In this drive the newly anointed king shot arrows at a Ksatriya and then returned to the sadas. According to van Buitenen this drive could symbolize the king's march against a real challenger, a kinsman of the king, who considered himself to be the claiment to the dominion. The epic incident and the king's drive, however, have nothing to do with each other. Sisupala's challenge was not to Yudhisthira who had performed the Rajasuya, but to Krsna. In van Buitenen's own view giving of the argha to Krsna symbolizes the handing over the remnant of the unction water i.e. bestowing the samrajya on Krsna after Yudhisthira. van Buitenen thus makes Sisupala, not a kinsman of the Pandavas, challenge not the emperor but the heir-apparent, and this is not the case in the chariot drive as understood by him. Quite clearly van Buitenen is pleading an extremely weak case or in fact a non-existing case. It is not as if he is totally unaware of it. He says: "To some it might appear that I overstate the connection between incidents in Madhu Vidya/495

Page #521
64 M.A. Mehendale the Vedic rajasuya and in The Book of the Assembly Hall". Why does he do it then? His defence is: "Yet it is not my cause I plead here, but that of the authors of this book". (p. 27). The position now becomes clear. van Buitenen indeed does not very strongly believe what he says. What has happened is that he has noticed an anomaly in the behaviour of Yudhisthira as presented to us by the epic authors. He therefore wants to plead the case of these authors by showing that the anomaly is not real but only apparent. To prove this he feels compelled to find some factors common to the two sacrifices and prove that Yudhisthira played the game not on his own but because he was forced by the rules of the Vedic Rajasuya. van Buitenen blames modern scholars for having failed to notice these common factors. But the modern scholars are not to be blamed. They naturally have not seen what does not exist. In support of his non-existing case van Buitenen next brings forward some lame evidence from the epic itself in order to show that Yudhisthira was bound to play dyuta. In a stanza (2.45.54) addresed by Dhstarastra to Vidura he interprets the word disgam as 'ordained (by the structure of the Vedic Rajasuya)'. According to van Buitenen Dhstarasgra in this stanza as if reminds Vidura, who is not at all enthusiastic about the game (2.45.52), that the game of dice has to take place since it has been ordained (dista). But here the world disa quite clearly does not mean "ordained by the structure of the Rajasuya'; it means 'ordained by fate' which becomes evident from the next stanza (2.45.55) where the word daivavihita occurs. In this stanza Dhstaraspra assures Vidura that he and Bhisma would not allow anything untoward to happen, even if the untoward was ordained by the fate, and that the gods in the heaven would do what was good for them (2.45.55, 53). van Buitenen admits that it was Sakuni who was instrumental for the game of dice (p. 28). How than does he argue that the game had to take place because it was ordained by the Rajasuya? The next hint, according to van Buitenen, given by the epic is contained in Vidura's appeal to Dhstarastra to stop the game of dice when the first ten plays were played out. Vidura appeals to Dhstarasura, argues van Buitenen, and not to Yudhisthira because the latter was bound by the rules of his Rajasuya and was not a free agent (pp. 29-30). This is a very strange argument. Vidura's appeal to Dhstarastra, and not to Yudhisthira who was 'maddened by the game' (dyutamadena mattah 2.60.5), was natural. We cannot forget that it was Dhstaraspra who gave permission to play, and hence it was he who could withdraw it. One of the reasons why Yudhisthira agreed to play was that Dhstarasira had asked him to do that. It was therefore natural if Vidura felt that if Dhstarasira asked, Yudhisthira would stop playing the game. What is extremely difficult to understand, if we follow van Buitenen's line of argument, is that if Yudhisthira played the Madhu Vidya/496

Page #522
Has the Vedic Rajasiya any Relevance for the Epic Game of Dice! 65 game as required by the rules of his Rajasuya, how did Vidura, the most virtuous-minded of all, think of asking Dhstarastra to do what would violate the prescriptions of the Veda? And how could Yudhisthira, if he was not a free agent, stop playing even if Dhstarastra asked him to do that? If one wants to assert that the epic game of dice was played because it formed part of the Vedic Rajasuya, one must first demonstrate that Yudhisthira's Rajasuya was performed mostly following the Vedic prescriptions. But this can absolutely not be done. The only "moments" really common to the two Rajasuyas are that water was used on the abhisecaniya day for unction and that the avabhstha marked the end of the ritual." In the epic there is not even a hint that the Vedic Rajasuya was the cause of Yudhisthira's game. The suggestion for the game came from Sakuni. It required Dhstarastra's consent which Duryodhana could not obtain easily. He had to threaten that he would otherwise commit suicide. Vidura was totally against it. Yudhisthira agreed to play becauce Dhstarastra wanted him to do that and because it was his vow not to refuse if challanged to a game. All these factors do not indicate that the structure of the Vedic Rajasuya was responsible for Yudhisthira's game. In the epic account there is no direct or indirect mention of it. It is possible also to raise the following objections to van Buitenen's contention: 1. If Yudhisthira's game of dice was in accordance with the procedure of the Vedic Rajasuya, the game should have been played at the time of and at the place of the sacrifice, and not several days later and in the Hall specially built for the purpose at Hastinapura. In the epic Dhstarastra wanted the Pandavas to visit the new Hall and have a game. 2. In the epic, Yudhisthira shows no awareness that he had to play the game as a part of the sacrifice performed by him. He did not play it as a matter of course but only because he received an invitation for it from Dhstarastra. 3. Yudhisthira, if he was following the Vedic Rajasuya, should have played the game with a Ksatriya, a Vaisya and a Sudra and not with Sakuni alone as a representative of Duryodhana. 4. On one occasion, after the war began, Sanjaya blamed Dhstarastra and said that if he had dissuaded the two parties from engaging themselves in the game of dice he would not be facing the danger he Madhu Vidya/497

Page #523
66 M.A. Mehendak was (7.62.4). Samjaya could not say this if the dyuta was played as a part of the Rajasuya. Then again Yudhisthira is seen blaming himself for consenting to the game and bringing disaster on the Pandavas and Draupadi (3.85.1, 4). This Yudhisthira could not do if he was bound to play the game as a part of the Rajasuya. van Buitenen feels compelled to make his untenable suggestion because he feels that there is a contradiction in the character of Yudhisthira in agreeing to play the game of dice. By offering his suggestion he wants to show that what looks like a contradiction is an 'apparent' one. But the starting premise itself of van Buitenen is not borne out by the facts. It is true the epic does not tell us that Yudhisthira played the game of dice every day. But the epic certainly tells us that Yudhisthira loved the game of dice and hence it is legitimate to presume that he played it now and then. Sakuni describes Yudhisthira as dyutapriya (2.44.18). Draupadi says Yudhisthira was natiktaprayatna "one who had not made (sustained) effort (to achieve proficiency in the game)" (2.60.43). This means Yudhisthira was an average player. If one were to believe Satyaki Yudhisthira used to play dice with his brothers (5.3.7). We also know that Yudhisthira wished to act as a 'gambling companion' (sabhastara 4.1.20) to Virata during the period of ajnatavasa. He was going to introduce himself to Viraca as a former very dear friend (pranasamah sakha 4.1.22) of Yudhisthira. If a gambler could declare himself to be a close friend of Yudhisthira, the latter must be one who loved gambling enough to feel compelled to accept the challenge issued to him (2.52.15-16). Hence there is no contradiction in the character of Yudhisthira if he agreed to play the game. Hence also it is not necessary to exert oneself to the point of being ridiculous to establish relationship between the Vedic and the epic Rajasuya in order to prove that the contradiction is not real but apparent. The epic war, on archaeological evidence, is supposed to have been fought C. 1200 B.C., ie. at a time much anterior to the formalization of the Rajasuya as represented in the ritual texts. It is therefore improper to seek justification for an earlier event in the texts of a later period. If anything, the epic incidents point to the possibility of a king who performed the Rajasuya being challenged by his kinsman to a game of dice and being deprived of his accumulated wealth. Probably, acceptance of such a challenge was looked upon as brave. Also, perhaps, the challenge could be issued only once. In order to obviate the danger of a king losing his wealth in a real game of dice, the systematizers of the Vedic ritual made the game a part of the ritual in which the performing king must win. Thus one may attempt to explain the game of dice becoming a part of the Vedic ritual on the basis of incidents similar to the one found in the epic, and not vice versa. Madhu Vidya/498

Page #524
is the Vedic Rajasuya any Relevance for the Epic Game of Dice? 67 NOTES AND REFERENCES * Paper read at the Seminar on the Rituals... under the auspices of the Ananthacharya Indological Institute, Bombay, December 28-29, 1991. 1. All references to the Pune Critical Edition of the Mahabharata. 2. 2. 24. 18-2. 27. 28; 2. 28. 54; 2. 29. 1418; 2. 43. 25-35; 2. 45. 15-35; 2. 46. 23-25; 2. 47.1-31; 2. 48. 1.-41. According to Duryodhana only two families did not pay tribute the Pancalas because they were related to the Pandavas, and the Andhaka-Vrsnis because they were the friends of the Pandavas (2. 48. 42). 3. The Mahabharata: 2. The Book of the Assembly Hall, translated and edited by J.A.B. van Buitenen, The University of Chicago Press. Chicago and London, 1975. 4. Buitenen asks: "... is a neighbouring kinsman's envy at Yudhisthira's good fortune sufficient reason for the new king... willingly to gamble it all away?" We may ask : "Why not?* We have a parallel in Nala losing everything away to his brother Puskara in a game of dice. 5. van Buitenen's conclusion that every game of dice must go through twenty turns is absolutely without basis. The anudyuta of the epic concluded only in one play, and there is no knowing how many plays Nala and Puskara had in their dyuta. 6. van Buitenen makes certain statements (p. 30) about this incident which are not of immediate relevance to the subject of this paper, but it has to be pointed out that there is no basis for them. He observes: "In the end Dhstarastra rules that the last play was indecisive and the game as a whole had been neither lost nor won". Where does he find evidence in the epic to make this observation? Similarly, following the general opinion, he states that the debate on the question raised by Draupadi remained inconclusive. The epic, however, tells us that when Duryodhana challenged any one of the Pandavas to answer Draupadi's question, Arjuna did that every neatly (2. 63. 21). Finally, there is nothing in the epic to show that the anudyuta was only a repeat of the twentieth game" and not a new game altogether. 7. J.C. Heestermann: An ancient Indian Royal Consecration, 1957, pp. 114 116; 167-168; Mbh. 2. 33. 1.; 2. 49. 10 fg.; 2. 42. 35. Madhu Vidya/499

Page #525
Is There Only One Version of The Game of Dice In The Mahabharata? M.A. MEHENDALE The Sabhaparvan of the Mahabharata (Mbh.) describes the game of dice played by Sakuni, on behalf of Duryodhana, and Yudhisthira. This version of the game we may call the standard version. But is this the only version of the game available? When one goes through the description in the Sabhaparvan and considers certain other references to the game found elsewhere in the Mbh. one notices certain discrepancies in narration. In order to account for them it becomes necessary to assume that at one time there were current different versions of the game. It is now not possible to have full details of the lost versions, but we can form some idea about them on the basis of the evidence offered by the discrepant account. Their presence in the standard version must be attributed to the anxiety of the narrator to include features of other versions in the standard one that he was narrating. I. Discrepancies in the Sabhaparvan There are five discrepancies in the Sabhaparvan version, two of which have been already noticed by scholars before. 1) When Dhitarastra consulted Vidura about the game of dice, the latter expressed his disapproval. Ignoring it, Dhstarastra asked him to go to the Pandavas to invite Yudhisthira for a game. Distressed, Vidura left for Bhisma's residence (2.45.58). In the next Adhyaya, however, we do not hear about the meeting between Bhisma and Vidura; instead, what has been narrated before is repeated in Adhyayas 46-51 and then we are told that Vidura left for Indraprastha (2.52.1). Obviously the two statements about Vidura's departure belong to two different versions of the game and some one has brought them together'. 2) When Duryodhana sent his messenger to Draupadi a second time to get her to the Sabha to obtain an answer to her question (2.60.10-12), she apparently refused to oblige him. He then returned to the Sabha and repeated Draupadi's question. Since no one replied?, Duryodhana, delighted, asked the messenger to go to Draupadi again to invite her to the Sabha (2.60.16). Between stanzas 12 and 16 which report the above sequence of events there occur two stanzas (2.60.14-15) which are totally incongruous with the narration. They inform us that Yudhisthira, knowing what Duryodhana desired, sent a Madhu Vidya/500

Page #526
messenger acceptable to Draupadi (? draupadyah sammatam dutam prahinot) and she came to the Sabha as she was and stood before her father-in-law. And in spite of the fact that the above stanzas tell us that Draupadi had gone to the Sabha, we are informed in the stanzas that follow (2.60.16 ff.) that Duryodhana asked his messenger to go to Draupadi, and when he hesitated, asked Duhsasana to do the same : In order to explain the incongruency in the narration we have to assume a combination of two versions of the same event -- one in which Duhsasana dragged Draupadi to the Sabha against her wishes, and the other in which Draupadi of her own accord presented herself in the Sabha'. The two stanzas (14 and 15) belonging to this latter version were incorporated in the standard version by some one in the oral transmission". 3) When none of the senior members present in the Sabha answered Draupadi's question, Vikarna did it. He declared that Draupadi had not become a dasi. He supported his contention on four grounds (2.61.19-24): (i) Kings are supposed to be addicts to four pleasures, one of which was dyuta. A person under the influence of these addictions behaved unlawfully (dharmam utsrjya vartate). People therefore did not recognize the acts committed under the influence of addiction. When Yudhisthira was challenged to stake Draupadi he was under the excessive influence of the game; therefore his act could not be recognized; (ii) Draupadi was the common wife of all the Pandavas. Vikarna implied that therefore Yudhisthira alone had no right to stake her; (iii) Yudhisthira staked her after he had staked himself and lost the game; apparently therefore he could not stake her; (iv) finally, it was Sakuni who, desirous of having Draupadi as the stake, mentioned her name. Vikarna implied that Draupadi as stake was not Yudhisthitra's own choice; he was instigated by Sakuni to do that and apparently that was not correct. Karna tried to refute Vikarna's arguments and establish that Draupadi had been lawfully won. In fact, however, he sought to refute only one of Vikarna's four arguments. He pointed out that when Yudhisthira specifically named Draupadi as his stake, the rest of the Pandavas had acquiesced (by remaining silent). Hence Karna implied that Draupadi was the stake of all the Pandavas and not of Yudhisthira alone (this would be in reply to Vikrana's second argument). Karna then said something, as if to refute what Vikarna had said but what Vikarna, in fact, had not said. Karna argued: "Yudhisthira staked all his possession (sarvasva) in the game, and since Draupadi was included in the sarvasva, she was lawfully won by the Kauravas. How can then you say that she was not won?" (yada sabhayah sarvasvam nyastavan pandavagrajah|| abhyantara ca sarvasve draupadi bharatarsabha / evam dharmajitam krsnam manyase na jitam katham II (2.61.31-32). What Karna said implied that Yudhisthira lost all his possessions, including Draupadi, in a single stakes. But this is not what that standard version of the game Madhu Vidya/501

Page #527
tells us. According to it Yudhisthira played twenty games to lose his possessions, brothers, himself, and wife in a certain order. Vikarna had never argued that Yudhisthira staked his sarvasva, and had not mentioned Draupadi by name," therefore she could not be looked upon'as won. It is therefore necessary to assume that what Karna said to Vikarna belongs to a different version of the event and that these stanzas are interpolated in the standard version. 4) In the narration of the Sabhaparvan, stanza 58.30 must be considered interpolated. In that stanza Vaisampayana tells Janamejaya : "Having said this, (Sakuni) who was fond of dice, defeated all those known heroes of the world (i.e. the Pandavas), who stood staked, by throwing the dice, (winning) each one of them separately. (evam uktva matiksas tan glahe sarvan avasthitan/ parajayal lokaviran aksepena prthak pethak) The point of narration where this incongruous stanza occurs is as follows: After Yudhisthira had staked his brothers, one after the other, and finally himself, and played and lost the games immediately after the declaration of each one of the stakes, Sakurti suggested the stake of Draupadi in two stanzas. In the first stanza, he blamed Yudhisthira for having lost himself when he had some 'wealth' left (2.58.29). In the second, Sakuni specifically suggested to Yudhisthira to stake Draupadi and "win yourself back''8 (panasva krsnam pancalim tayatmanam punar jaya (2.58.31). These two stanzas of Sakuni should have occurred one after the other. Instead, between the two (29 and 31) occurs Vaisampayana's stanza noted above (30) which is not only out of the place but is inconsistent with the main narration. Stanza 30 is out of the place because the narrator has already told us that Sakuni had won all the five brothers by playing five games. Now Sakuni had started to suggest to Yudhisthira to stake Draupadi. There is thus no occasion for Vaisampayana to intervene and tell what he has already told before. The stanza is incongruous because it tells a different story about the progress of the game. According to the standard version Yudhisthira and Sakuni played the games immediately after the stake of each one of his brothers and of Yudhisthira himself was announced. According to Vaisarpayana's stanza, however, all the five stakes were deelared first, in a certain order, and when that act was completed the two players played five games one after the other. Stanza 30 of Vaisampayana opens with the words evam uktva. This indicates that in the lost version, before stanza 30, there must have been a stanza in which Sakuni told Yudhisthira something as follows: "Now that you have staked all your brothers and yourself, let us play the games for those five stakes". Only under the assumption of some such course of events Vaisampayana's stanza beginning with 'Having and said thus ...." will be consistent. In the standard version it is obviously interpolated. Once we assume that stanza 2.58.30 is interpolated in the standard version we recognize some more factors which support such an assumption. We notice that MadhuVidya 502

Page #528
the stanza in which Nakula's stake was declared is in the Anustubh while the stanzas in which the stakes of Sahadeva, Arjuna and Bhima were declared are in the Tristubh. We would normally not have laid stress on this difference in the metres because such a change is not uncommon in the Mbh. But the difference between the two stanzas is not merely formal, it also extends to the mode of declaring the stake. We have assumed that in the lost version Yudhisthira first declared only his five stakes without playing the games immediately following the declaration of each stake. In that situation it is possible for Yudhisthira to use such expressions as 'this is my first stake', 'this is my second stake' etc. after the declaration of each stake. In fact Yudhisthira does use such an expression while declaring the first stake of Nakula (nakulo glaha eko me 2.58.11). But in the standard version of the game where each time the game was played immediately after the announcement of the stake, such expressions 'this is my first, this is my second stake, etc.' have no place, and we do not find Yudhisthira saying 'this is my second stake' when he staked Sahadeva and the rest. Hence it appears that stanza 2.58.11 declaring the stake of Nakula also belonged to the same lost version to which stanza 2.58.30 belongs. 5) When Draupadi was dragged to the Assembly hall she said: "The skilled gamblers challenged the inexperienced Yudhisthira to a game of dice and they all together defeated him (ahuya raja kusalaih sabhayam... natikrtaprayatnah... sambhuya sarvais' ca jitah 2.60.43-44). This account does not agree with what had happened before. It was Sakuni who invited Yudhisthira to a game (2.53.1), and when Yudhisthira asked against whom he was supposed to play. Duryodhana announced that he would be responsible for the stakes but the game would be played by Sakuni (2.53.10). The game was then played only between Sakuni and Yudhisthira and none of the gamblers present in the Hall is reported to have taken part in the game at any stage. The stanza in which Draupadi complained that Yudhisthira was defeated, not by Sakuni, but by all the gamblers playing together" must have been interpolated in the standard version from a different account of the game. This assumed version finds its echo at three different places in the Mbh.. First, Vikarna, while stating his case, said that Yudhisthira was challenged by the gamblers when he staked Draupadi (samahutena kitavair asthito draupadipanah 2.61.22). Secondly, Satyaki asked Balarama: "The skilled gamblers challenged the one who was not an expert with dice and defeated him. How could this be called a lawful victory ?" (5.3.6)11. Finally, Yudhisthira himself once said that he was challenged to a game by skilled gamblers and was deprived of his kingdom (3.49.32). It is surprising that Draupadi, Vikarna, Satyaki, and Yudhisthira do not mention Sakuni at all. Hence we have to assume that there was a version of the event in which the gamblers in the Hall, all of them expert, and not Sakuni, defeated Yudhisthira. All the above references really belong to that version."2 Madhu Vidya/503

Page #529
II. Discrepancies in the other Parvans The incongruous elements in the versions of the game noticed in the Udyogaparvan and the Aranyakaparvan are of a more radical nature. They present to us Yudhisthira in a totally different light. 1) The Udyogaparvan : When, before the war, the Pandavas were taking counsel regarding the future course of action, Balarama, at one stage, drew attention to the way how the Pandavas lost their kingdom. He said: "All the principal heroes in the Kuru family and Yudhisthira's friends tried to prevent him from indulging in the game of dice. Although not a skilled player, Yudhisthira challenged Sakuni, a known skilled player, to a game. There were other gamblers in the Hall and Yudhisthira could have challenged them and won. Instead, he challenged Sakuni and was defeated. As the game progressed, the dice always turned out to be unfavourable to Yudhisthira. He was confused and lost miserably. There was no fault of Sakuni (tatraparadhah sakuner na kas cit 5.2.9-11). Balarama's account is at complete variance with the one found in the Sabhaparvan. In the latter, no one is reported to have tried to dissuade Yudhisthira. In fact, the question of dissuasion did not arise at all. Yudhisthira had already told Vidura that he would not himself challenge Sakuni, since he did not wish to play dice; but if he was challenged by Sakuni, he, following his vow, would accept it (2.52.16). There was thus no question of Yudhisthira's challenging Sakuni or any one of the gamblers. According to the Sabhaparvan, the compulsion for playing the game was Sakuni's challenge; he won, because his confidence in his skill was justified. According to Balarama's version, the compulsion for playing the game was Yudhisthira's own urge to play; he lost, because his confidence was. misplaced. 2) The Aranyakaparvan : The information supplied by the account in the Aranyakaparvan changes the image of Yudhisthira even more radically. The information assumes significance because it is supplied by Yudhisthira himself. Once, in the forest, during the course of a conversation, Yudhisthira admitted to Bhima that because he (Yudhisthira) adopted the wrong path they all faced their present difficult condition (mamanayad dhi vyasanam va agat 3.35.1). He continued : 'It was I who took recourse to the dice in order to deprive Duryodhana of his entire kingdom (rajyarn sarastram ?). But the rogue Sakuni played against me for Duryodhana. As the game progressed, I noticed that the dice resulted according to the wishes of Sakuni -- they turned out to be odd or even as he desired. Seeing that it should have been possible for me to control myself (and stop playing). But anger destroys a man's equilibrium" (aham hy aksan anvapadyam jihirsan rajyam sarastram dhTtrarastrasya putrat/ tan ma sathah kitavah pratyadevit suyodhanarthar subalasya putrah // ... aksan hi drstva sakuner yathavat kamanuloman ayujo yujas'ca/ sakyam niyantum abhavisyad atma manyus tu hanti purusasya dhairyam // Madhu Vidya/504

Page #530
3.35.2, 4) It is impossible to reconcile these statements with the account narrated in the Sabhaparvan. There we are told that it was Sakuni who suggested to Duryodhana to take recourse to the safe expedient of dyuta to deprive the Pandavas of their accumulated wealth. Hence Dhtarastra invited the Pandavas for a game dice. But on Yudhisthira's own admission it appears that Dhstarastra may have invited Pandavas just to pay a visit to the newly built Sabha and not for a game of dice. But once in the Sabha, Yudhisthira was seized by the desire to play a game and deprive Duryodhana of his kingdom. It is conceivable that Yudhisthira was superior to Duryodhana in playing the game and hence he challenged him. Unfortunately, however, Sakuni took up the challenge; Yudhisthira could not withdraw, and in the process lost everything to Sakuni. This account in the Aranyakaparvan must belong to a different version. The Udyogaparvan account, noticed above, too holds Yudhisthira responsible for the start of the game, but does not attribute specific motive to him. The Aranyakaparvan account goes beyond it and attributes a selfish motive to Yudhisthira in taking recourse to dice. The Aranyakaparvan account further alludes to a different mode of play. According to Yudhisthira, the dice always resulted as Sakuni wanted them to be. They turned out to be odd if he wanted them odd, or even if that was what he desired. This means that the total number of dice thrown on the playing ground by the two players was divided by two and not by four. It also means that after Yudhisthira threw the dice, when Sakuni followed Yudhisthira he declared in advance how the dice would turn out, odd or even'. The result of each game went in Sakuni's favour. On the other hand, from the various allusions in the Mbh. (5.140.7, 9 etc.; 4.45.24) to the game of dice in the Sabhaparvan it can be assumed that the Sabhaparvan game agreed with the Vedic game of dice in that the total number of dice thrown on the playing ground was divided by four. Apparently Sakuni always succeeded in that game in getting the number exactly divisible by four and hence he won. The different versions of the events in the game of dice may be charted as follows: The Game of Dice in the Muh. -- Different Accounts of certain Events Duryodhana, jealous of Pandavas' wealth suggested the Game of Dice to Dhrtarastra (2.45; 35; 40; 2.51.4) Madhu Vidya/505

Page #531
Duryodhana, jealous of the Mayasabha, asked Dhrtarastra to build a Sabha in Hastinapura Dhstarastra had the Sabha built Dhrtarastra invited the Pandavas to a Game in the Sabha Dhstarastra invited the Pandavas to visit the Sabha Vidura sent to invite the Pandavas Vidura went straight to the Pandavas (2.52.1) Vidura first went to Bhisma (and then to the Pandavas) (2.45.46) The Pandavas in the Hall -- the game of dice Challenge issued not by Yudhisthira Sukani challenged Yudh, and defeated him (2.53.1 ff) Gamblers challenged Yudh, and defeated him (2.60.43-44; 2.61.22; 5.3.6: 3.49.32) Challenge issued by Yudhisthira Yudhisthira challenged Sakuni and lost (5.2.9-11) Yudhisthira challenged Duryodhana, but Sakuni accepted the challenge (3.35; 1-4) Yudh. lost everything in 20 games (2.53.22 ff.) Madhu Vidya/506

Page #532
Yudh. lost everything in a single game (2.61. 31-32) Yudh, staked and played, staked and played five times to lose his brothers and himself. (2.58.11-28) Yudh. first only staked five brothers individually in five stakes and played. five stakes and afterwards played five conscutive games to lose all (2.58.30) Yudhisthira lost Draupadi's stake; Drupadi asked to come to the Sabha Duhsasana dragged Draupadi to the Sabha (2.60.22 ff.) Draupadi herself came to the Sabha (2.60.14-15) Notes: 1. F. Edgerton (Sabhaparvan, Introduction, XXXII) observes: "Again it seems to me that two accounts of the same event are taken into the text side by side". This part of the event has not been specifically narrated but may safely be assumed. Obviously in this version there will be no occasion for Duhsasana to drag Draupadi by her hair. F. Edgerton (Sabhaparvan, Introduction, XXXI), "Clearly we have here parts of two entirely different versions of the story." This was, for example, how Puskara lost everything to Nala in a single stake (3.77.18). Unlike Nala, in the second match between him and Puskara, when he mentioned Damayanti specifically as his stake along with all his wealth (3.77.5). It is likely that in this lost version Draupadi raised the question not only about herself, but also about the four Pandavas. Apparently while staking his sarvasva, Yudhisthira had not named his brothers either. When Duryodhana challenged any one of the Pandavas to reply to Draupadi's question Bhima in his reply said that if Yudhisthira considered himself won, the rest of the Madhu Vidya/507

Page #533
- Pandavas were also won (2.62.33). Significantly he spoke nothing about Draupadi He was as if speaking for himself and his brothers. This reply of Bhima is clearly out of the place in the standard version because nobody had raisedd the question about the Pandavas. It gains significance only in the abovu version assumed by us. The significance of the expression win yourself back" is that Sakuni agreed to stake Yudhistira who had become a dasa, against Yudhisthira's stake of Draupadi; if Yudhisthira won this last game he would be free (and of course Draupadi would not be won). It may be noted that the way Sakuni's victory is announced in the standard version (2.53.25; 2.54.3,7 etc., for the winning of the Pandavas (2.58.13, 15, 21, 25,28) is quite different from the way the same is done in the assumed lost version (2.58.30). Of this mode of play -- one against many -- we know practically nothing. It is however, difficult to understand what Satyaki could have meant when he continued that if the gamblers had approached Yudhisthira while he was playing the game with his brothers at his residence and won, that would have been a lawful victory (5.3.7). In the Nala story too Damayanti gave a different account to the ascetics whom she met in the forest. She told them: "Some mean persons, rogues, who were expert gamblers challenged the king (Nala) to a game of dice and deprived him of his kingdom and wealth" (3.61.78). In the principal narration, however, we are told that Nala's brother Puskara challenged Nala and defeated him. Hence Nala story too apparently was current in two versions. That is, as in the game where the dice are divided by four, it was not necessary that the number be exactly divisible by two and no remainder left. 10. 2. 13. Madhu Vidya/508

Page #534
ONCE AGAIN DRAUPADI'S HAIR By M. A. MEHENDALE In the Sabhaparvan of the Mahabharata (Mbh.) Draupadi is subjected to two humiliations : dragging her by her hair and trying to remove her upper garment. Alf Hiltebeitel (H.) has published two articles concerning these two themes. The first, on Draupadi's Garments, appeared in 1980 and the second, on Draupadi's Hair, in 1981.3 I bave refuted H'.s yiews on the first elsewhere. Here I examine his views on the second, For his second article H. bas used evidence from the Mbh. as well as from the Venisahara of Bhattanarayana, In the present paper I shall consider mainly the Mbh, evidence, referring to the Venisamhara only in a few cases. In his first article on Draupadi's garments H. saw in Kespa ( dark) Draupadi earth, and, covered with white and red garments, also praksti. In the second paper on Draupadi's hair he has chosen to see in Krsna (dark ) Draupadi, wearing dishevelled hair, Goddess Kall, and further in her loose hair the dissolution of the universe (pralaya ). His thrust in the second paper is to present Draupadi wearing her hair loose, unwashed and dishe. velled, since the Sabha episode up to the end of the battle, i. e. for well over thirteen years. It is widely believed in the south, H, informs us, that Draupadi took a vow not to bind up her hair until they were smeared with the blood of either Duryodhana or Duhsasana. The theine is represented in the literary conipositions and stage representations of the epic in the south (pp. 179-185). In the north, the vow related to Draupadi's hair first appears in the Veni sahara (7th cent. A. D.). In this drama, however, it is Bhima, and 1 The original article in Marathi appeared in the Navabharat, Divali Number, 1994, 75-90. 1 IIJ 22. 97-112 ( 1980). 8 Autour de la deesse Hindoue. Ed. M. Biardeau, Paris, 179-213 being Vol. 5 (1981) of Purusartha : Sciences Sociales en asie du Sud, Paris. BDCRI, (Golden Jubilee Volume ) 50, 285-290 (1990). The numbers of pages 287 and 286 to be corrected respectively as 286 and 287 in that article. Madhu Vidya/509

Page #535
160 Annals BORI, LXXVIII (1997) not Draupadi, who takes the vow. In the Mbh. itself we hear of two vows of Bhima: to drink the blood from Duhsasana's chest (2. 61.46 ) and to break the left thigh of Duryodhana (2. 63. 14). None of these has anything to do with the anointing of Draupadi's loose hair with blood and tying them up. H. asserts : "The Mahabharata seems to know more about Drau. padi's hair than it ever makes explicit" (p. 186). The epic events he uses for his discussion are the following: 1. Dragging Draupads to the Sabba According to H. Draupadi's hair were already loose and dishevelled when Duhsasana held them in his hand. His opinion is based on two epithets, urmimant ( 2 (0.22), which he renders flowing', and prakirna (2.60.28 ) dishevelled', used in the Mbh. to describe Draupadi's hair. He justifies the 'flowing' i.e. the loose and dishevelled condition of Draupadi's hair by pointing out that she was at that time rajasvala. Women in period were required to observe certain restrictions on taking care of their hair. For example, according to a statement in the Taitt. Samh. 2. 5. 1 rajasvala women were not expected to comb their hair (ya pralikhate). It is true that according to the Mbh. Draupadl was rajasvala on the day the dice game began. But it does not say the rest of the things imagined, by H. about Draupadi's hair. In the first instance, the epithet armimant, used for her hair, does not mean .flowing', hence loose, but wavy', i, e, curly, Draupadi's hair were dark and curly as seen from the epithets nilakuncitamurdhaja (1.155.42) and vellitagran ( 4.8.1). Next, the Mbh. does not say that Draupadi was already prakirnakes! when Dahsasana beld her by hair. It says that she became so a little later owing to the shakings she recei. ved while being dragged to the Sabha (praklrnake$t patitardhavastra duhsa. sanena vyavadhuyamana 2. 60. 28; and later duhsasanakarotsTstaviprakirna. siroruha 3.12.17). It is also wrong to infer 'the dishevelled condition of her hair on the basis of the Taitt. Samh. passage referred to above. The passage in question mentions such restrictions on rajasvala women as not bathing, not applying ointment to eyes (or oil to the body ), and not combing hair. The last restriction only means that women in period were expected to keep Among the medieval Marathi poets, Muktesvara (17th cent.) shows knowledge of Bhima's vow (Sabhaparvan 16. 72 ). But it is absent both in Sridharasvamin's ( 17th-18th cent.) Pandavapratapa (29.92 tr.) and Moropant's (18th cent.) aryabharata ( 5.91; 6. 29). It would be interesting to find out whether or not the vow figures in the medieval literature of the other parts of northern India. Madhu Vidya/510

Page #536
MEHENDALB : Once Again Draupadi's Hair 161 their hair the way they were before the commencement of the period i, e, as braided before. When in period they were not supposed to untie them for combing or for any other purpose. Conclusion: There is no evidence to show that Draupadl wore hor hair loose and dishevelled before the start of the dice match. Her hair were dishevelled later, though tied up, because of shakings. 2. Starting for the forest Draupadi, when she started for the forest, was rajasvala and mukta. kesi (2.70.9; 2. 71. 18). On this basis H. argues that since some days must have elapsed between the first match of the dice and the start for the forest after the second match, and since Draupadi cannot be rajasvala for a great number of days, the word rajasvala used to describe Draupadi's condition while starting for forest cannot mean observing her period'. It has to be taken to mean being in a gencral state of impurity'. The general state of impurity attaches not only to Draupadi but to the Pandavas as well because they all were entering " into a realm of death symbolized by the gestures of their chaplain Dhaumya who strews darbha grass to the death Goddess Nirfti and sings Saman verses to Yama (2.71.21 )" (p. 189). For his new meaning of rajasvala H. thinks he finds support in the curse-like words uttered by Draupadi. She said : " The Kaurava women, fourteen years hence, would offer water to the dead and then enter the city Hastinapura in a rajasvala state" (2.71.19-20). Since all Kaurava women cannot be rajasvala in its normal sense at the same time, H. argues, the word in this context must mean being in an impure condition' arising out of the death of near relatives. Draupadi's wearing her hair loose at the start of the exile is indicative of this defilement associated with death. All this is not convincing. Draupadi could have started her period the very day the first match of dice was played and if the Mbh. tells us that she was still in her period when the Pandavas started for forest it only means that an interval of not more than two ( or three ) days elapsed between the two 6 While listing the Sabhi incidents that occurred before the Pandavas started for forest H. mentions the question raised by Draupadi and adds that the question was never resolved (p. 188 ). This is an erroneous view about an important event. This view has held its ground for quite a long time. Draupadi's question was squarely answered. by Arjuna (2. 63. 21) and his answer settled that Draupadl had not become a Dasi. For a detailed discussion see the author's paper "Draupadi's question". JOIB 35, 179-194 (1986). 21 [Annals BORT) Madhu Vidya/511

Page #537
162 Annals BORI, LXXVIII (1997) matches of dice. The expression vyadhvagatan (2. 66 24; 2. 67. 1) used with rifercace to the Paodavas on their way to Indraprastha can easily be understood as indicative of the distance they could have traversed in about a day or so. There is therefore no ground to interpret rajasvala used with reference to Draupadi in any other way. And even if H. thinks that he has some. how got over the difficulty presented by the word rajasvala, how is he going to get rid of the other difficulty about Draupadi's garment which is said to be smeared with blood ( sonitaktaikavasana 2. 70. 9; sonitaktardravasana 2. 71. 18 )? This is definitely indicative of her being in period, As for the Kaurava women it is obvious that the word rajasvala cannot mean a woman in period'. But the word can be easily understood otherwise--and H. is aware of it - since it can also mean covered with dust.' The Kaurava women would embrace their husbands fallen on the ground in the battle and hence would be covered with dust' (rajasvalah). This is quite understandable and there is no necessity of bringing in here for the understanding of the word ihe notion of general impurity arising out of death. Similarly the word sonita occurring in the word bandhusonitadigdhangyah (2. 71. 20 ) used to describe the Kaurava women is easily understood as referring to the blood issuing from the wounds of the fallen heroes. It certainly does not refer to the menstrual blood. This is also made clear by the use of the word .bandhu'relative' in the above compound. Now about the word muktakesl used to describe Draupadi when she started for the forest. If the epic informs us that Draupadi wore her bair loose at that time, it means that before that she had her hair tied up. And because she wore her hair loose at that point of time for some reason it does not mean that she kept them so for the rest of the duration of her forest life. Draupadi's purpose in keeping her hair loose at that time is told by Vidura to Dhrtarastra. She did that to foretell that the Kaurava women will have to wear their hair loose at the end of the war (2.71.19 ). As muktakesyah with reference to Kaurava women means that they would wear the hair that way temporarily and afterwards they would tie them up as before, so also muktakest with reference to Draupadi means that she untied her braid only temporarily. Once the purpose for which the hair were untied was served there was no reason why she should keep them upbraided. The behaviour of the other Pandavas also points in the same direction. Whatever they did (e. g. Arjuna released sand from his hand 2.71. 14) while going to the forest was done only for a short while as their actions had specific purpose to serve. When that was achieved they had no reason to continue doing the same for all the twelve years of their forest life. Madhu Vidya/512

Page #538
MEHENDALB i Once Again Draupadt's Hair 163 Also the Purohita of the Pandavas recited the Samans related to Yama not to suggest that be, along with the Pandavas, was entering the realm of death. He did that only to suggest what his counterpart in the service of the Kauravas would have to do at the end of the battle (2. 71. 21-22). Conclusion : At the time of leaving for the forest Draupadi was still rajasvala, i, e, in hier period, and she did untie her hair temporarily on some purpose. 3. Forest life As the Pandavas were entering the Kamyaka forest on the third night of their forest life they were stopped at the entrance by the demon Kirmira. Seeing him Draupadi fainted and was held by her husbands. While referring to her at this point the Mbh. stanza alludes to her hair : " She whose hair were dishevelled and released by the hand of Duhsasana " ( duhsasanakaroiststaviprakirnasiroruha' 3. 12. 17). From this H. concludes: "Draupadi continues to wear her hair dishevelled in the exile" (p. 190). This is not correct. The above expression is not intended to describe Draupadi's hair at that point of time. It is intended to refer to the condition of her hair when, only a few days ago, they were held by Duhsasana in bis hand. A reference to Draupadi's hair here and elsewhere, e. g. in 3. 13. 108, is not to be interpreted as referring to the then prevailing condition of her hair and serving as reminder to the Pandavas of what had happened in the Sabha (p. 190). The Paodavas did not stand in need of such reminders. They would continue to remember the iasult done to their wife although Draupadi had tied her hair up. Besides the reference to Draupadi's hair in the Kirmira incident, H. points to references to Draupadi's tapas in the forest which, according to him, are very likely " directly related to Draupadl's tacit (in the epic ) vow" (p. 190 ), i. e. to her popular vow to wear her hair loose until the death of the Kauravas. It does not seem justifiable to bring into this discussion references to tapas. These references are few and they do not suggest that Draupadi " It is not clear why H, translates this compound as " with her hair widely dishevelled and ruffled by Duhsasapa's hand." Apparently he translates utsrsta as "ruffled ", since elsewhere he renders viprakirna as " dishevelled ". The word utsrsta can only mean released'. The traasposition of viprakirna and utsrsta is apparently due to metrical reasons. Madhu Vidya/513

Page #539
Annals BORI, LXXVIII (1997) practised austerities all through the forest life and consequently wore her hair dishevelled. Draupadl joined her husbands in tapas on two occasions for specific reasons. First, because the mountain Arcika to which they were drawing near was holy ( 3. 125. 13, 19) and, a second time, because nobody who had not practised tapas could mount the mountain Gandhamadana (3. 141. 22-23).8 164 This is all that H. has to bring forward as evidence in support of his contention that Draupadi wore her hair loose and dishevelled during her stay in the forest. His evidence is not conclusive. On the other hand there is evidence to suggest that Draupadi did not practise tapas all along in the forest and that she took good care of her hair. H. has apparently missed this precious little piece of evidence. The evidence is available from an incident which occurred towards the end of the forest life. Jayadratha once saw Draupadi from a distance. He noticed that she had an excellent appearance. She shone with her lustre illuminating the forest ( 3, 248, 9). This descrip. tion does not fit a woman who would be emaciated if she practised continuous tapas. A little later in that incident Kotikasya drew near Draupadi to enquire who she was. He addressed her as sukesi ( voc. )' one whose hair are lovely' (3. 249. 13). This makes clear that Draupadi's hair were in good condition. They were not such as they would be if they were neither washed nor taken care of for eleven years. The expression does not tell us whether the hair were tied up in a braid or they were loose. But one may argue they were not loose. For, if they were in that condition Kotikasya would have almost certainly asked her why Draupadi had not tied them in a braid. Loose hair could be an indication of a woman staying away from her husband. Conclusion: Draupadi took good care of her hair in forest life and had tied them up in a braid. 4. Ajnatavasa According to H. the Ajnatavasa is "rich in its surprising twists and complex symbolism" (p. 191). The twist is observed in the fact that the Pandavas had to perform duties not at all befitting their status. case of Draupadi the twist lies not merely in her having to do a job unsuited to her. The twist is felt more poignantly by H. on account of the condition of her hair. While going to the town of Virata, Draupadi had gathered up her soft hair and concealed them on the right side presumably by a piece of In the case of second tapas, Draupadi is not directly named. However, her tapas can be a reasonable assumption. Madhu Vidya/514

Page #540
MEHENDALE : Once Again Draupadi's Hair cloth. (tatah kesan samutksipya vellitagran aninditan juguha daksine parsve mrdun asitalocana | 4.8.1). Thus, while Draupadi had to keep her beautiful hair concealed, she must take care of Sudespa's hair! One need have no hesitation in admitting the painful twist in the life of Draupadi emphasized by H. But it is necessary to object to his remark 1 "she whose hair is hidden, and, I would argue, stili dishevelled" (p. 191). In support of his assumption H. finds nothing in the epic text, neither critical nor interpolated. All that he can find in support is the artist's representa tions of Draupadi's hair inserted in the critical edition of the Virataparvan (facing pp. 67 and 272) and the Sabhaparvan (p. 305). There Draupadi's hair appear loose, not tied up in a braid. 165 It has to be admitted that the artist's representations in the critical edi. tion are, at times, faulty. The editors of the critical edition must have been conscions of it. Yet, the representations were allowed to appear probably because of the exigencies of time. However, the editors could never have imagined that any one at any time would use those representations, even in a foot-note, as a piece of evidence. In the two specific representations of the Virataparvan mentioned above Draupadi surely wears her hair loose. H. has only noticed that. But those hair cannot be called "dishevelled," not at any rate in the first representation (facing p. 67) and, what is more important, they are not shown gathered up and concealed on the right side as required by the epic text (4.8.1). Draupadi's hair were definitely not dishevelled and unkempt at the commencement of her life incognito. To prove this it is not necessary to point to the two representations noticed above. One would rather point to the two adjectives mrdun and aninditan of her hair found in the passage noticed above (4. 8. 1). They are quite enough to indicate that Draupadi took good care of her hair in the forest. It is likely that Draupadi did not tie up her soft (mrdu) and blameless (anindita) hair in the normal triveni way, as she did in the forest life, but braided them in the ekaveni way because she was going to tell Sudespa that she was living away from her five Gandharva husbands (4. 8. 27). In the southern recension of the Mbh. H. finds a clear proof of the connection between Draupadi's vow to wear her hair dishevelled and her disguise H. has not considered the question why Draupadi had concealed her hair. It is not likely that Draupadi acted that way because that was the way Sairandhris were expec(Continued on the next page) Madhu Vidya/515

Page #541
166 Annals BORI, LXXVIII (1997) as a Sairandhri (p. 192!). That recension informs us that Draupadl as Sairan. dhri assumed the name vratacarini (sairandhri jatisampanna nam naha vratacarini 4.94*). According to H. yratacarini "can hardly be anything but a paradoxical reference to her vow of dishevelment" (p. 192). I beg to differ, I would relate the word vrata in the expression vratacarini not to something which the epic does not state, but to something which it does state. Sairandhri was jatisampanna belonging to a good class (of Sairandbris ).10 Hence she would not do certain things I she would not eat remnants of the food partaken by others and she would not wash anybody's feet. (.4. 8. 29). It is to these norms of life that the word vrata in vratacarini refers. (Also cf. Mbh. 3. 222, 29). Just as vratacarini has nothing to do with the vow of dishevelinent, the word diksitah, used about the Pandavas in a variant in the stanza 2. 68. 1,11 has nothing to do with their occasional practice of tapas in the forest. If the Pandavas, in the variant reading, are said to be vanavasaya diksitah it simply means they were resolved or fully prepared for forest life,' exactly as when the same thing is said, for example, of Yayati (1. 80, 25 ) or Dhsta. tastra (15. 17.3) who resolved to live in forest leaving their respective king. doms. This is shown by the text adopted in the critical edition. According to H., however, the word diksitah occurring in the variant noticed above has connection with what occurred after the twelve years of forest life. At the end of the Ajnatavasa Arjuna told Virata that they had lived unknown in his house as creatures live in the womb (before their birth ) (ajnatavasam usita garbhavasa iva prajah 4. 66. 10). The simile is easily understood. But H. sees in it a hidden meaning. In the diksani ya isi performed at the beginning of a sacrifice, the sacrificer stays in a special hut (diksitavimita). His stay in this hut is looked upon as his living in the (Continued from p. 165 ) ted to have their hair. Damayanti lived for some time as a Sairandhri of noble class ( 3. 62, 26). But it is not said that she had concealed her hair. It would therefore be a reasonable guess that DraupadT concealed her hair so that her identity was not disclosed. It seems her hair were remarkably curly ( 1.155. 42: 2. 60.22: 4. 8. 1). That would have revealed her ideotity. In the case of Damayanti, her ideatity would have been disclosed by the mole between her eyebrows (piplu ). Hence she kept it concea. led ( 3. 66.5) This is what jatisainpanna means and not "belonging to that caste" as rendered by H' H. does not give this reference. Four DevanagarT mss, including that of Nilakantha. one Bengali And one Malayalam manuscript read the first line of the stanza as tatah parujitah partha vanavasaya diksitan. The critical edition has, instead, vang. vasaya cakrus to matim parthah parajitah. 11 Madhu Vidya/516

Page #542
MEHENDALE 1 Once Again Draupadi's Halr 167 womb from which he emerges at the end of the is11.19 According to H. Arjuna's mention of garbhava sa has a bearing on the life of the sacrificer in the hut and that explains the term diksitaused with reference to them, The connections seen by H. are very far-fetched. In the first instance there is no reason to see any ritual significance in the word diksitan. It simpiy means were prepared for.' Secondly, if diksitah had any connec. tion with the garbhava sa of a diksita, then Ajnatavasa should have preceded the vanava sa as the diksani ya isti precedes the sacrifice. H. appears to be aware of this difficulty. However, without mentioning it be quietly relates the expression di ksitah not with vanavasa but with the sacrifice of war' (ranayajna)- a term which does not occur in the epic but in the Venisamhara - which will take place after the Ajnatavasa. But the variant clearly reads vanava saya diksitah and not ranaya dik sitah. H. dwells at some length on the theme of ritual danger, death, and impurity related to rebirth in the diksa rite (pp. 194-196). In this connection he brings in an incident from the Virataparvan (pp. 196-197). Once, Virala hit Yudhisthira's nose with dice and blood began to flow. Sairandhri gathered the blood in a golden vessel and prevented it from falling down, Yudhisthira explained to Virata why Sairandhri did that. He said that if his blood had touched the ground Virata, together with his kingdom, would have perished (4. 64. 8). According to H. this explanation means that the fallen blood would have given rise to destructive forces endangering the king and his kingdom. He gets this idea from an incident in the Devi mahatmya in which demons arose from each drop of blood that fell on the ground from the body of the demon Raktablja. The process was stopped by Kali by drinking up the blood before it could reach the ground. According to H. Draupadi played th: role of Kali and saved the kingdom from destruction. This further means she sayed the womb of the Pandavas the destruction of which would have forced them to emerge from their garbhavasa prematurely and be known before the proper time arrived. The Mbh. gives no occasion to notice such connections. Yudhisthira himself has clarified what he told to Virata, and that should leave no doubt in anybody's mind regarding what Yudbisthira meant. He had earlier told the gatekeeper of Virati that Brozonada would kill any offender who forced out blood from Yudhisthira's body except in war (4. 63. 53-54). If Brhan. 13 Aitareya Brahmana 1. 3. 49 At the most the gatekeeper might have wondered what had Bebannada to do with this, and how could she do what Yudhisthira said. Madhu Vidya/517

Page #543
168 Annals BORI, LXXVIII (1997) nada is able to kill Virata with his ministers and the entire army there is no necessity of imagining the rise of evil forces from the drops of his blood fallen on the ground. And apart from this the parallelism sought to be established between the epic and the Purani incidents does not stand scrutiny, It was not Sairandhri who really stopped the blood from falling on the ground. It was Yudhisthira himself who did it. Sairandhrl came into the picture a little later ( 4.63. 45-47). In the Purana incident Candika fights with Rakta. bija, her enemy. She hits the demon and asks Kali not to let the blood drop down. When this is cione Candika kills Raktabija. In the epic Virata hit Yudhisthir, his friend, in a fit of anger, not his real enemy. Yudhisthira, who is at the receiver's end, and not Virata, the bitter, who asks Sai. randhri to hold the blood in a vessel. When this is done, Yudhisthira does not kill Virata. H. meets with a formidable difficulty with regard to Draupadi's hair in an incident in the Ajnatavasa. When Sairandhri ran to Sudesna's quarters she untied her hair loose (kesan mukiva 4. 15. 36).14 This expression indicates that before Sairandhri loosened her hair they were tied up in some way. This clearly goes against H.'s view that Draupadi's hair were all along loose. He, therefore, takes kesan muktva to mean "Draupadi spreads or shakes loose her already unbound hair." (p. 197). This is impossible. That H. himself is not satisfied with his interpretation is seen from his next statement that "there is no precise and evident consistency in every epic reference to Draupadi's hair" (p. 198). This means that according to H, too kesan muktva means "having loosed her hair" (which were tied up before ). But he would explain it away simply as an example of epic inconsistency, But there is no real inconsistency. The expression in its normal sense is easily understood. When Draupadi was insulted by Kicaka in the presence of her two husbands, she was understandably wild with anger, her eyes red and burning (4,15. 14, 36 ). In this state of mind, no wonder if she suddenly decided to set her hair loose to indicate that she considered her husbands as good as dead. In this connection it is necessary to consider another piece of evidence which H. brings forward in support of his contention of Draupadi's deliberate neglect of her hair. According to him, although there is no consis 11 The Southern recension reads instead keson pramuktan salyamya (4. 338* ). This will not help H. (p. 197). As a variant to kesan in ktva, kesan pramuktan samyanya can only mean that Sairandhri gathered together her hair which she had (just) loosened. Madhu Vidya/518

Page #544
MEHENDALE ! Once Again Draupadi's Hair tency in the epic's reference to Draupadi's hair, there is consistency while referring to her actions. Once, before the death of Kicaka and, again, after his death, Sairandhr. washes her limbs and garments (gatrani vasasi caiva praksalya salilena sa 4. 16. 2; 4. 23. 12). Here, there is consistency in her actions. What strikes H. more in this case is that in both these verses there is no reference to the washing of her hair. Repetition of verses is not unusual in the epic and H.'s conclusion is unwarranted. The word gatrani can include hair as well and no distinct mention of kesan is needed. The Pandavas, while visiting the tirthas, often bathed in holy waters. While referring to such ablutions the epic sometimes uses the word gatrani without specifically mentioning hair ( 3. 93. 5; 3. 109, 20; also Draupadi in another context (aplutang! ) 1. 176. 29). This, however, does not mean that the Pandavas and Draupadi washed only their limbs and not their hair, 169 Kali had figured in the Puranga incident brought in for comparison with an incident in the Virataparvan in which blood issued from Yudhisthira's H. has observed similarity between the roles of Kali and Draupadi in the two incidents. In the death of Kicaka too, which occurred even before the above incident, H. sees a connection between Kall and Draupadi. The person, however, who kills Klcaka is Bhima, not Sairandhri. Hence, H. has to establish some sort of identity between Bhima and Sairandhri in order to associate her with Kali. This he does by approvingly citing M. Biardeau's view that Bhima was the arm of Sairandhri (le bras de la deesse) (p. 198). Next, in this connection H. gets a passage in the epic in which Sairandhri is said to be another form of Death. Kicaka was happy at the prospect of meeting Sairandhri in private, but the fool as he was, he did not realize that Sairandhri was only another form of Death (sairandhri rapinam mudhe mrtyum tam navubuddhavan 4. 21. 19). Again, later, when Kicaka touched Bhima mistaking him for Sairandhri, the epic says that he had touched Death (sayanam sayane tatra mrtyum sutah paramrsat (4.21.42). There is really nothing very striking in such statements. If there was any other woman in place of Sairandhri (and even if she had never worn her hair loose), the narrator might have used similar words. But, H. has the following to contribute to the present scene: "In the epic, Mrtyu is a Goddess, a form evoking Kali and Durga as they break into the Hindu literary tradition. For Death to appear to Kicaka "in the form of a Sairandhri" would seem an unmistakable evocation of Kali, the wild Goddess with the dishevelled hair" (p. 199). Conclusion: There is no evidence to show that Draupadi wore her hair loose in the Ajnatavasa. On the other hand, there is evidence to show that she had tied them up. 22 Annals BORI] Madhu Vidya/519

Page #545
170 Annals BORI, LXXVIII (1997) 5. Attempt at reconciliation B.tore Krsna left for Hastinapura to make a final attempt to avoid war, Draupadi showed her hair to him and asked him not to forget them while negotiating with the Kauravas. Draupadi's hair at that time were delightful to look at, very dark, having the lustre of a cobra, curly at ends, scented with fragrant substances, possessed of all good marks, but first of all, loosely braided (mydusamharam vrjinagram sudartanam sunilam punyagandha dhivasitam sarvalaksanasam pannam mahabhujagavarcasam | kesapaksam15 5. 80. 33-34). Referring to this incident H. says that in these words of Draupadi, the epic comes closest to her popular vow (not to braid her loose hair until they were smeared with Duhsasana's (or Duryodhana's) blood (p. 200). The truth of the matter, however, is that this description of Draupadi's hair is clear evidence to disprove H.'s contention that Draupadi wore her hair loose and did not take care of them for over thirteen years. As for their being loose the epithet mrdusamharam gathered loosely together (in a vent form) goes against H. Hence he disagrees with Nilakantha's correct explanation venirupena samhrtam api mydum. He, on his part, feels satisfied with the rendering softly gathered back' and considers it reasonably neutral' (p. 200, f. n. 71 end). But why should one try to be neutral when the Mbh. epithet is unambiguous ?1 H. has considered the difficulty presented to him by mrdusamharam at least in a footnote. But he takes no note of the other epithets of Draupadi's hair which show that Draupad] did not have them dirty and dishevelled for a long peried. Conclusion Between the Ajnatavasa and the start of the battle Draupadi dit not wear her hair loose. 6. End of the battle After the battle, Yudhisthira was grieved over the great loss of life, including that of Karna. All try to revive his spirits. When Bhima speaks 15 The word paksa added to kesa does not reveal whether the hair are tied up or loose. It signifies only their collection (pasah paksas ca hastas ca kalaparthah kacat pare / Amarakosa 2. 6.98, comm. kalaparthah kesasamuhavacinah ity arthah). 16 A veni may be tied up tight or loose as desired. 17 H. himself renders samhara "binding up" when it occurs in the title of the drama Venisamhara (p. 179). In the context of the word samhara as in mrdusamharam, which is said about Draupadi's hair, the word veni, though not expressed can be easily assumed. Madhu Vidya/520

Page #546
MEHENDALE 1 Once Again Draupadi's Hair 171 he says : distya duryodhanah papo nihatah sanugo yudhi / draupadyah kesapaksasya disiya rvam padavim gatah // (12. 16. 25). H. assures his readers that "This passage establishes beyond any reasonable doubt that Draupadi has worn her hair loose since the dice match" (p. 201 ). How does H. find in the above stanza unambiguous evidence for his contention ? He gets it in two steps: (1) According to H. the second line of the stanza means that "the Paodavas have moved from the depths of defilement to rebirth, from rebirth to revenge and from revenge to coronation" (p. 201 ), just as Draupadi's hair have moved from the state of dishevelment to rebinding and anointing. (2) And the use of the word disiya twice, once with reference to the fall of Duryodhana and again with reference to Yudhisthira's going the way of Draupadi's hair ( as interpreted by H.) shows that the death of Duryodhana is connected with the binding up of Draupadi's hair (p. 201). This is misleading. The stanza does not say that Draupadi's hair were dishevelled and loose before Duryodhana's death. That is H.'s assumption on the basis of which he interprets the stanza. The epic nowhere suggests that Draupadi wore her hair loose when Duhsasana held them and that, after that incident, she kept them loose until the death of Duryodhana. In fact, we have notices to the contrary. The epic, however, does tell us that Duhsasana molested Draupadi's hair. Hence when the stanza says that Draupadi's hair and Yudhisthira have gone the same way, it only means that just as Draupadi's hair bave avenged the insult done to them with the death of Duryodhana and his followers, Yudhisthira too, in their death, has avenged the injustice done to him in being refused his share of kingdom. In the stanza Bhima uses the word distya twice. It may be that when he used it in the first line he admitted that there was an element of luck in his being able to hit the thigh of Duryodhana. By using it again in the second line, Bhima expresses his happiness over the Pagdavas' having been able to achieve a double objective with the death of Duryodhana in the battle. In the first instance Yudhisthira was able to wipe out the injustice and get back the share of the kingdom and, secondly, they were able to avenge the insult done to Draupadi's hair ( draupadyah kesapaksas ya disi ya tvam padavim gatah 12. 16. 25). The main objective of the battle was no doubt to fight against the injustice done to the Pandavas in the denial of their share of kingdom. This is what Kunti's message has empha. sized : "get back the paternal share of kingdom refused to you" (pitryam amsam mahabaho nimagnam punar uddhara 5. 130. 30, 32; 5. 135, 5). This was also the only point made in the story of Vidura told by Kunti. Kuntl's reference to the harsh words spoken to Pancali in the Sabha is only secon, Madhu Vidya/521

Page #547
172 Annals BORI, LXXVIII (1997) dary (5.135. 15).13 If Duryodhana had agreed to return their share of the kingdom to the Paodavas or if he had shown willingness to settle the whole matter by giving the Pa@davas five villages as suggested by Yudhisthira, battle could have been averted. Then the Pandavas would not have got the opportunity to avenge the insuit 10 Draupadi. Hence Bhima was happy (disiya ) that the battle enabled the Pandavas to achieve the double objective. In the first part of his paper H. has tried to establish that Draupadi wore her hair loose and kept them unwashed for a pretty long time. This was only to prepare ground for establishing in the second part of his paper his main thesis viz. that epic Draupadi is the prefiguration of later Kali," the exemplary Goddess of the disbevelled hair" (p. 206). While looking for evidence in this direction, H. notices association of Draupadi with Death and Destruction. As was noticed above, H. points out that according to the epic narrator Sairandhrl meant Death to Kicaka. Next, he points out that at the time of Draupadi's birth an invisible voice prophesied that Krsa will lead the Ksatriyas to destruction ( krsna ksa yam ksatram nini sati 1. 155.44) Draupadi's name Krsoa, which is also a common epithet of Kali is thus directly connected with Destruction, The Kicakavadha incident has been already discussed above. As regards the prophecy it may be noted that the Brahmanas present at the time of Draupadi's birth gave her the name Kssqa specifically on account of her dark complexion (krsnety evabruvan krstam krsnabhut sa hi varnata) 1. 155. 50). The Brahmanas proceeded with their work as though they knew nothing about the prophecy. The name Krsna is intelligible even in the absence of the prophecy. One may therefore consider that that part of the prophecy which connects Draupadl with the destruction of the Ksatriyas did not figure in the original text. Its presence gives rise to an internal incongruity. The text says that hearing the prophecy, the Pancalas were extremely delighted, so much so that the earth could not bear them (na caitan harsasumpurnan iyam sehe vasundhara 1.155. 46). One cannot understand such great joy of the people if the newly born girl was going to be the cause of great destruction. If, however, we omit the lines 1. 155. 44cd and 1. 155. 45cd That in the following two stanzas Kunti considered the insult done to Draupadi even more important than the loss of kingdom is not consistent with what goes before. If that were really the case, Kunti should have mentioned it first. Moreover, the story of Vidura is not at all suited for inciting, a warrior to take up arms because his wife was insulted. This point is irrelevant in the context of the Vidura narrative. Hence the stanzas which refer to Draupadi in Kunti's message are suspect. Madhu Vidya/522

Page #548
MEHENDALE : Once Again Draupadi's Hair 173 and read 1. 155.44 as: tam capi jalam susroni in vaguvacasaririni / surakaryam iyan kale karisyati sumadhyama, it is suitably followed imme. diately by 1.151.46 which speaks of the joy of the Pancalas on hearing the prophecy. Having imagined Draupadi to be the prefiguration of Kali, first on account of her assumed dishevelled hair and, next, on account of her equally assumed relationship with Death and Destruction, H. tries to bring Kali and Draupadi closer through Kalaratri, Kalaratri is mentioned in the epic. It is said that sbe was seen by the soldiers at the time of the noctural massacre in the camp of the Pandavas. H. points out that this Kalaratri of the epic not only receives the epithet Kall dark', but her whole description is such that no other goddess but Kali can be thought of as coming closest to her. H. thus tries to identify Kalaratri, who is mentioned in the epic, with Kali who is not mentioned. But H. has yet to show relationship between Draupadi and Kalaratri before he can relate Draupadi with Kali. Relation. ship between Draupadi and Kalaratri is, however, totally absent in the epic. H. has, therefore, only to rely on certain connections. First, in the description of Kalaratri she receives an epithet fikhandini, and Kalaratri appears in the camp after Asvatthaman has killed the first few beroes one of whom is Sikhaodin, Draupadi's brother. And then in the Veni sanihara, while arranging Draupadi's hair, Bhima calls her veni as the Kalaratri of all the Kauravas (dhartara sirakulakalaratrih ... iyam veni VI, after st. 41 ). One wonders how this can be considered evidence enough to associate epic Draupadi with Kalaratri. Bhima's reference to Draupadi's braid as Kalaratri is from outside the epic. Kalaratri's epithet sikhandini has nothing to do with her mention after Sikhaodin's death. Kalaratri could be present in the camp right from the beginning of the massacre. Only she was noticed by ordinary warriors when Asvatthaman made them his target. Morcover, the epic itself has quite a different story to tell about the identity of this Kaiaratri. The epic identifies her with a krtya an evil spirit' activo on the side of the Kauravas It is said that since the very start of the battle, ordinary fighters in the camp of the Pandavas used to see in a dream Asvatthaman killing sleeping warriors and a krtya carrying them away. When on the night of the real destruction they saw Kularatri, they remembered what they had earlier seen in dreams and identified Kalaratri with the krtya ( 10. 8. 66-69 ). H.'s connecting Draupadi with Kalaratri is horrible. It is said that Kalaratri was all smiles at mass killings in the camp of Pandavas (10.8. 64 Madhu Vidya/523

Page #549
174 Annals BORI, LXXVIII (1997) 65 ). It is unthinkable that Draupadi, if she were the Kalaratri, could do that. Only a Krtya, siding with the Kauravas, could. Draupadi's assumed dishevelled and loose hair were used by H. to see in her the prefiguration of Kali. For this he needed the help of Kalaratri. The induction of Kalaratri the night of Time, the night of dissolution (pralaya ) of the universe' (p. 207), into the discussion has led H. to point out another symbolic significance of Draupadi's hair. In his earlier paper on "Draupadi's Garments" H. had maintained that if Karna himself had tried to disrobe Draupadi and succeeded in his attempt that would bave meant a naimittika pralaya. Draupadi then would stand "depuded like the earth prior to its combustion and ... "bare like the back of the tortoise"" (IIJ 22. 103 ). But since in the epic it was not Karsa, but Duhsasana, who made the attempt to disrobe Draupadi and since he did not succeed in his attempt the epic scene meant that ( except at the naimittika pralaya) the earth has an inexhaustible capacity to restore her garments automatically (IIJ 22. 103 ). In the present paper with regard to Draupadi's untied hair and the dissolution of the universe, H. observes : " The untying of Draupadi's braid represents the potential untying of the universe. For the universe is itself a braid, composed of the three gunas or strands of praksti, nature. Here, bowever, it is not the naimittika pralaya that is evoked, but the praksta pralaya, the very dissolution of nature (prakrti).... Mythologically, this pralaya is the "Night of Time", Kalas atri, with whom Draupadi's hair is directly compared in the Veni samhara .... But by wearing her hair dishevelled for thirteen years, Draupadi also shows that the full dissolution of the universe is at least metaphorically pending should her husbands, with the help of lord Krsna, not regain the sovereignty and restore dharma on earth" (pp. 210-211). I have already refuted elsewhere H.'s views related to Draupadi's garments.10 As for Draupadi's hair, we have seen that there is no justification for seeing her hair untied for thirteen long years. During this period she kept them untied for a short while only on two occasions - first, while starting for the forest, and, again, while going to Sudesna from the Sabha of Virata. There is therefore no question of relating Draupadi with kali, the goddess of dishevelled hair, nor seeing in Draupadi's loose hair Kalaratri, the dissolution of praksti. According to H. Draupadi's loose hair is a fact; 19 See fn. 4 above. Madhu Vidya/524

Page #550
MERENDALE ; Once Again Draupadi's Hair 175 however, the dissolution of praksti, which he thinks the loose hair symbolize, is not a fact but only an imagined possibility. H. has to live with this inconsistency because we know that heavens did not fall when, according to H., Draupadi, resolved to wear her bair loose due to the insult done to them by Duhsasana. This should have happened if untying of hair meant dissolution of the three gunas of prakti. But it did not, and hence what accord. ing to H, is already a fact symbolizes wbat according to him is a possibility ! But then the epic gives no ground for imagining this possibility either. The epic does not inform us that the Kauravas misruled their kingdom during the exile of the Pandavas, that during their rule adharma was ascendant. Hence there is no reason to feel that, if the Pandavas had failed to avenge the insult to Draupadi, i, e. if they had lost the war and not regained sovereignty, the Kauravas would have let loose on earth such terror, such adharma, as would match pralaya. Madhu Vidya/525

Page #551
URUBHANGA AND THE CRITICAL EDITION OF THE MAHABHARATA M. A. MEHENDALE The fighting on the eighteenth day of the epic battle did not end for that day immediately after the fall of Salya but continued for sometime. During the turmoil, Duryodhana felt tired and was a little dismayed. He thought of retiring from the battle. With a mace in his hand he left the battlefield alone and hid himself in a pond.' When the Pandavas learnt about his hideout, they went to the pond and challenged Duryodhana to come out and fight? Duryodhana agreed on one condition, viz. that the Pandavas fight with him singly and not all of them together. The condition was accepted and a duel followed between Bhima and Duryodhana with a mace as their weapon. The duel, with occasional periods of rest, lasted for quite sometime." The fight went on strictly according to the rules, no one showing any inclination to take undue advantage of the adverse situation of the other. On one occasion, for instance, Bhima forced Duryodhana to kneel on the ground and, on another occasion Duryodhana even fell on the ground. On both these occasions, Bhima did not hit Duryodhana while he was unconscious. Only when Duryodhana regained consciousness and was in a position to continue fighting, the duel was resumed. Similarly. When Duryodhana had an upper hand and he succeeded in felling down Bhima and in breaking his armour he did not hit Bhima as long as the latter lay down on the battlefield. By that time, however, Krsna noticed that the duel had reached the stage when Duryodhana had an edge over his opponent. Bhima had lost his armour. Krsna realized that that was the time to throw some hint to Bhima. He plainly told Arjuna that if Bhima continued the fight as he was doing, i.e. fighting strictly according to the rules, there was no chance of his coming out victorious. Bhima must take recourse to some unlawful act. Arjuna too realized the need and he secretly stroked his thigh hinting what Bhima should do. Bhima understood the hint and when he found an opportunity he threw his mace at Duryodhana. Duryodhana, however, succeeded in dodging the hit. When Bhima showed his inclination to hit Duryodhana again, the latter planned to jump up to avoid the hit. Bhima guessed correctly what Duryodhana intended to do and threw his mace exactly at the point where Duryodhana's thighs would be if he really jumped up. This time Bhima did not miss the mark. The mace broke Duryodhana's thighs and he lay on the battle field, his thighs broken? Thus came to an end the duel. Bhima violated an important rule of the Madhu Vidya/526

Page #552
92 M. A. MEHENDALE fight, viz. that one should not strike the opponent below the navel (adho nabhya na hantavyam iti sastrasya niscayah 9.59.6). Now according to the epic narrative, Bhima, on an earlier occasion, had vowed to break the thigh of Duryodhana with his mace. This incident which occurred in the sabha of the Kauravas is wellknown. When Duryodhana bared his left thigh in the sabha, Bhima vowed to break it.9 Duryodhana was also cursed by the sage Maitreya for his misbehaviour. According to this curse Bhima was destined to break the thigh of Duryodhana in the battle, 10 However, on reading the detailed account of the battle one gets the impression that none of the epic heroes-not even Bhima or Duryodhana-had the faintest idea of either the vow or the curse. During the war Bhima never challenged Duryodhana for a duel with a mace. This is what one would expect him to do if he had vowed to break Duryodhana's thighs. Krsna, on one occasion, incited Arjuna to kill Duryodhana with an arrow and end the war, and Arjuna agreed." On another occasion Arjuna himself assured Krsna that he would kill Duryodhana with his sharp arrows if the latter did not run away from the battlefield. Krsna had given his consent to Arjuna's proposal. 12 On both occasions Arjuna, for one reason or the other, could not carry out his resolve to kill Duryodhana with his arrows. But his intention to kill Duryodhana was quite clear. The question then arises, if Krsna and Arjuna had known of either the vow or the curse, how could they think of finishing Duryodhana with arrows?13 Could they think of depriving Bhima of the opportunity to fulfil his vow to break the thigh of Duryodhana with his mace? This only shows that both of them had no knowledge of the vow or the curse. When Duryodhana came out of the pond where he lay hiding and accepted the challenge of the Pandavas, Yudhisthira, on his own, offered three concessions to him. He permitted Duryodhana to choose his weapon. He allowed him. to name the Pandava with whom he would fight. And, finally, he said that if Duryodhana won the duel, the rest of the Pandavas would accept defeat and that would mean the end of the war. No other Pandava would continue the fight with Duryodhana. 14 These concessions were disastrous for the Pandavas. They prove how reckless Yudhisthira was. Once before, at the time of the game of dice, he had gambled away everything in a single stake at the game. Now, by making these concessions, he almost handed over victory to Duryodhana for asking. Duryodhana had only to name the weapon-the mace-and any one of the Pandavas except Bhima for the duel and that would be the end of the matter. Madhu Vidya/527

Page #553
URUBHANGA AND THE CRITICAL EDITION OF THE MAHABHARATA 93 The concessions clearly show that Yudhisthira had no knowledge of either the vow or the curse. If he had, he would have asked Duryodhana to get out of the pond and fight with mace with Bhima so that the latter could fulfil his vow. Krsna took Yudhisthira to task for his thoughtlessness. 15 The Pandavas were, however, very lucky. Duryodhana did not take the easy way to victory. He was too proud for that. He no doubt chose mace as the weapon, but did not challenge any one of the Pandavas since he considered all of them inferior to him for a mace-duel. He, therefore, said: "Let any one of the Pandavas who could think of fighting with me lift up his mace and step forward." 16 Bhima took up the challenge and a duel between the two began." We now come to the two heroes engaged in the duel to see whether they showed any knowledge of the vow (or the curse). A few days before the duel, there was an occasion for Bhima to refer to the yow. When Bhima killed Duhsasana and started to drink the blood from Duhsasana's chest he declared loudly : "Today I have fulfilled my vow related to Duhsasana. I shall also today cut Duryodhana to pieces like a sacrificial animal and give offerings."18 That is surprising. If Bhima had taken the second vow he should have rather declared : "Today I shall also break Duryodhana's thighs with my mace and fulfil my vow related to him." He then should have proceeded to seek Duryodhana out and challenge him for a duel. But this did not happen. in the description of the duel itself we do not find anything that might even distantly suggest that Bhima was looking for an opportunity to hit Duryodhana's thighs. The description rather gives the impression that he was fighting strictly according to the rules. It was only when Arjuna, at the instance of Krsna, gave a hint to him that he threw his mace at Duryodhana's thighs. The vow had nothing to do with it. Bhima thus had no knowledge of the Vow Duryodharia too is never seen conscious of either the vow or the curse. He never mentioned it, nor was he ever worried about it. While fighting his decisive duel with Bhima he is not described taking special care of his thighs. On the other hand, he even once wished to jump up in the air to avoid being hit by the mace of Bhima. 19 This act meant exposing his thighs and Duryodhana would certainly not have wished to do that if he knew of Bhima's vow. The way Bhima had fought the duel must have convinced Duryodhana that he (Bhima) was not going to violate the rule of the duel and hit him below the navel. Unfortunately for him his understanding of the situation was not correct. Bhima had decided to act on the suggestion received by him from Arjuna. When Duryodhana, therefore, did jump in the air, Bhima got his golden opportunity to bring Duryodhana down. The hero with broken Madhu Vidya/528

Page #554
94 M. A. MEHENDALE thighs took this as an act of wilful violation of the rule and not as an inevitable result of a vow (or a curse).20 The only conclusion that can be drawn from the above account is that the two episodes related to the vow and the curse must be interpolations. The interpolations occurred very early in the oral epic tradition giving them time enough to spread to all the versions. And since they are now found in all the versions they could not be set aside and have found place in the critical edition of the epic. The critically edited text is no doubt free from the obvious interpolations which were revealed by the comparison of the versions, but the critically edited text is not free of all interpolations. Dr. V. S. Sukthankar has asked the readers to take note of this fact. He writes that there are many old and new elements in the critically edited text, sometimes even side by side.21 It is now for the future scholars to bring to light the hidden interpolations by a careful reading of the text. The motivation for the interpolation of the vow and the curse is not far to seek Bhima's act of hitting Duryodhana below the navel was inevitable on account of his inferiority to Duryodhana in a mace duel. But the act involved adharma. In order to absolve Bhima from the sin of this adharma an incident was interpolated in the Sabhaparvan in which Duryodhana bared his thigh to show it proudly to Bhima. This enraged Bhima to proclaim his famous vow. Apparently, at some stage, it was felt that the vow was not quite enough to fully absolve Bhima from the adharma. It is true that as a Ksatriya it was his duty to fulfil the vow. But why did he, even in a fit of anger, take the vow which involved adharma? Hence another incident was interpolated in the Aranyakaparvan in which Duryodhana stroked his thigh in the presence of the sage Maitreya. Duryodhana was cursed by the sage for his insolence, 22 and the curse of the sage must prove true. Now there was no question whether Bhima was inferior to Duryodhana or not, or whether he had taken the vow or not. Even if Bhima was superior to Duryodhana, and could bring him down by fighting strictly according to the rules, and even if he had not taken the vow, Duryodhana's thighs had to be hit by his mace so that the words of the sage proved true. It was said above that the account of the war as reported in the epic does not show that any one of the principal characters involved in the war was aware of the vow of Bhima. There are, however, a few exceptions. One exception is found when Krsna suggested to Arjuna that Bhima must take recourse to some adharma to win the duel.23 He then made a casual reference to Bhima's vow, but not to the curse.24 It is quite clear that Krsna would have welcomed Bhima's winning the duel without having to take recourse to adharma. He waited for quite some time to see if that could happen. It was only when he realized that that was not possible that he alerted Arjuna. Madhu Vidya/529

Page #555
URUBHANGA AND THE CRITICAL EDITION OF THE MAHABHARATA 95 But if Bhima were to defeat Duryodhana lawfully, how was he going to fulfil his vow? This shows very clearly that Krsna's reference to the vow is secondary and the stanzas in question are interpolated. . Another exception is found in Krsna's words addressed to Balarama to appease his anger. He told his brother that after all the Pandavas were their friends and relatives. Balarama should also remember that Bhima had vowed to break the thighs and it was the duty of a Ksatriya to fulfil his vow. In addition, there was the curse of Maitreya to the same effect. Finally, what Bhima did was to some extent due to the fact that the Kali era was about to set in.25 These words of Krsna contain a reference to the vow, as well as to the curse. If Bhima's act was the result of the vow and the curse, there was no reason for Krsna to refer either to the relationship between the Yadavas and the Pandavas or to the impending kali age. The reference to the vow and the curse in the words of appeasement of Krsna is therefore interpolated. A reference to the curse, beside one to the vow, in these words of Krsna indicates that the curse episode was interpolated after that of the VOW. Finally, we come to an implied reference to Bhima's vow made by Duryodhana. While criticizing Krsna for his role in the war26 Duryodhana said: "Do I not know what you told Arjuna while asking him to secretly remind Bhima to break the thighs ?"27 This statement contains an implied reference to the vow since it speaks about reminding Bhima. But Duryodhana's remark raises various questions. In the first instance, if Bhima took the vow. is it believable that he needed reminder about it? There was no need, for instance, 10 remind Bhima about his vow with reference to Duhsasana. Secondly, we know that Krsna had said nothing about breaking the thighs. He only said in a general way that Bhima had to take recourse to adharma. The specific nature of the adharma viz. hitting the thigh, was the idea of Arjuna. Thus, Duryodhana's accusation has an element of untruth. Thirdly, if Duryodhana knew that a suggestion had been secretly conveyed to Bhima to hit the thigh, why did he jump up in the air and expose his thighs ? Finally, Krsna was not so naive as to suggest something to Arjuna so vividly that Duryodhana could understand it. It is, therefore, clear that the stanza of Duryodhana is an interpolation which occurred after the addition of the vow incident. In order to understand the spuriousness of the vow incident, it is very instructive to take note of the conversation between Gandhari and Bhima. She asked Bhima how he dared violate the rule. She told him that what had angered her was not the killing of her sons but his striking Duryodhana below the navel.20 In reply, Bhima confessed to Gandhari that whatever he did was out of fear and in self defence. It was impossible for him to defeat Duryodhana otherwise. If he did not take recourse to adharma, he would have died and his brothers would have been deprived of the kingdom. He Madhu Vidya/530

Page #556
96 MA MEHENDALE implored Gandhari to forgive him for his misdeed.29 Bhima made no reference to Duryodhana's alleged misdeed in the sabha and his (Bhima's) consequent vow. Bhima's own admission to Gandhari leaves no doubt about the spurious nature of the vow (and also the curse). Notes and References 1. 9.28.18, 24-25, 52;9.29.53,6 1,64;9.30.4. 18. 29.30.17.20.27.29-34. 3 9.56.6,7 4. 9.56.53.61-62 5. 9.56.63-64,67 6. 9.57.4. 7. 9.57.40-45 8. abhyutsmayitva radheyam bhimam adharsayann iva / draupadyan preksamanayah savyam urum adarsayat // 2.63. 12. The stanza has been misunderstood. It is taken to mean that Duryodhana bared his thigh to show it to Draupadi. This is not true. The genitive absolute construction draupadyah preksamanayan is intended to convey Duryodhana's disregard for Draupadi's presence in the Sabha (Panini 2.3.38). A similar construction is found in vaidarbhyan preksamanayan panakalam amanyata (3.56.8) where Nala is described as agreeing to play the game of dice with Puskara in spite of the presence of Damayanti. In such instances the verbs preks-, drs- mis. are used only to indicate the presence of the person concerned. The above stanza (2.63.12) really means that Duryodhana bared his thigh to show it to Karna, which made him smile abhyutsmayitva), and to Bhima as if to assail or challenge him (adharsayann iva). This he did in spite of the presence of Draupadi. 9 263.13-14 10 3.11.32-34 11. 7.77.7, 10.20. 12. 9.23.47;9.26.9.24. 13. For unsuccessful attempts made also by Yudhisthira and Dhrstadyumna to kill Duryodhana see 8.20. 6-31; 8.40. 20-38. 14. 9.31.24-25;52-53. 15. 9.32.1-7 16. 9.31.60 17. 932.15-18. 18. 861.16 Madhu Vidya/531

Page #557
URUBHANGA AND THE CRITICAL EDITION OF THE MAHABHARATA 97 19. 9.57.41 20. 9.63.10. 21. Cr. edn. Vol. 1. Prolegomena, p. Cill. 22 3.11.28-34. 23. 9. 57.4. 8. 17. 24. 9.57. 6-7. 25. 9.59. 11-16, 21. 26. 9.60 26-38. 27. uru bhindhiti bhimasya smrtim mithya prayacchata / kim na vijnatam etan me yad arjunnam avocathah 11 9.60.28. 28.11.13.16-19. 29. 11.14. 1-4, 6. 11. It is well to remember that when Gandhari next asked Bhima about his drinking blood of Duhsasana, Bhima in his reply did refer to his vow. He, however, assured Gandhari that not a drop of blood went beyond his lips and teeth; he only made a pretence of drinking blood (11.14.12-18). How, in that case, Bhima could be said to have fulfilled his vow is a moot point. Madhu Vidya/532

Page #558
(VI) Miscellaneous Ahimsa and the Spread of Vegetarianism in India M. A. MEHENDALE AHIMSA literally means 'non-injury'. Ideally the concept should mean non-injury of any kind, whether physical or mental, to all living beings. But in practice, the concept seems to have made its greatest impact on killing animals for food and for ritual purposes. The Aryans who came to India from outside did not show, at least in ancient times, any aversion to eating meat as an item of food. Even in the early Buddhist texts we find fish and meat included in the five articles of food. Today, Jains are perhaps the strictest among the vegetarians. Among the Hindus, only a small section of the population is vegetarian as a matter of religious practice; a large number among them, however, is vegetarian for economic reasons. It is to be doubted whether climate or any other physical conditions in India. were responsible for the spread of vegetarianism. On the other hand, a clear relationship can be definitely established between ahimsa and vegetarianism. This is seen in the following verse in the Manusmriti (5.48): "It is not possible to obtain meat without causing injury to living beings; killing animals does not lead one to heaven; (but obtaining heaven after death is the cherished goal of all); hence meat has to be avoided" The spread of vegetarianism has not been that easy in India. It had to contend with the ritual partaking of food in sacrifices and other ceremonial occasions which were prescribed by the Vedic texts. The vicissitudes through which it passed are well reflected in the ancient law-books and in many narratives of the Mahabharata. Initially, it seems that even according to the thinkers who were primarily responsible for the spread of Madhu Vidya/533

Page #559
420 M. A. MEHENDALE ahimsad, it was enough if one avoided killing the animal oneself or took care to see that the meat eaten as food was not specially prepared for him. If these conditions were satisfied, meat-eating could not be looked upon as himsa. In the stage that followed, however, certain animals were totally prohibited from being used as food, whether one observed the above conditions or not; and even the permitted animals could not be eaten during certain parts of the year or of the month. The third stage in the spread of vegetarianism was marked by permitting meat-eating only in sacrifices, sraddha ceremonies, and hospitality to guests. Outside these ceremonial occasions, no meat-eating of any sort was permitted. The fact that the ritual meat-eating continued until the third stage is a clear proof of the great influence exercised by the Vedic injunctions on the minds of people. The final stage in this long history was marked by a total and absolute ban on meat-eating, whether done as a part of a ritual or otherwise, and whether the animal was killed by or for oneself or not. This must have been, for those times, a very bold decision indeed. The fifth chapter of the Manusmriti is an important source for the history of vegetarianism in India. It is curious to find the different historical stages reflected in the 'same' text. In this chapter Manu is shown as giving a discourse on the causes for the God of Death having an upper hand in relation to men versed in the Vedic texts. This happens, says Manu, due to the neglect of Vedic studies and other rules of conduct, due to laziness, and due to faults of food. This gives him an occasion for prescribing what may be eaten and what should be avoided. He starts ruling out as food even some vegetables like garlic, leek, and onion and then goes on to enumerate certain animals and birds which were to be regarded as improper for food. Fish of any sort has been condemned outright because one who eats fish is an eater of every kind of meat. Such a detailed enumeration gives the impression that those animals and birds which are not excluded by the law-giver may serve as articles of food. The growing influence of the doctrine of ahimsa and its practical effect seen in the spread of vegetarianism had its impact on the performance of Vedic sacrifices. These latter, in order to be successful, had to be performed according to Vedic injunctions which prescribed, among many other things, killing Madhu Vidya/534

Page #560
AHIMSA AND VEGETARIANISM 421 of animals and offering and partaking of meat portions. The leaders of the society in those days had to declare categorically that ceremonial meat-eating was permitted. Thus it has been declared on the authority of Manu (5.41) that animals are to be killed only for guest-worship, sacrifice, and for honouring gods and manes (pitar). The obvious implication is that they were not to be killed otherwise for being served as food. The scruples of the common man against the killings in sacrifice had to be set at rest by arguing one way or the other. It was pointed out, for instance, that the creator himself had created animals for sacrifices; therefore killing in a sacrifice amounted to non-killing (5.39). It was even maintained that if one ate meat after honouring gods and manes, it did not matter whether the animal for this purpose was killed by oneself or whether the meat was bought or obtained as a gift (5.32). As for the animals themselves, as also for the plants and trees that were cut for sacrifices, it was maintained that they attained a higher status (ucchriti, or uttama gati 5.40, 42) after death and thus, in a sense, were obliged by the sacrificer. It appears that, in spite of all this, there were a few priests who made themselves bold to refuse eating meat in a sacrifice. In order to compel these 'erring priests to do what was sanctioned by the tradition, it was declared that one who refused meat-eating in a sacrifice was born as an animal twenty-one times (5.35). As an aside it may be observed that if one really believed in what one said, then one wonders whether this can be looked upon as a threat. For, a person born as an animal twenty-one times stood a good chance of being caught hold of for a sacrifice and then, as mentioned above, he was assured of a higher status after death! In the final stage, which is reflected in a third group of verses (5.45-55) of the Manusmriti, we hear of the total ban on meat-eating whether ritual or otherwise (nivarteta sarvamamasya bhakshanat). The ban is imposed on the basis of ahimsa. Any meat-eating whatsoever, which involves himsa, cannot lead one to heaven; hence one should abstain from meat-eating. This must have been in those days a very bold stand, for it contradicted the Vedic injunction which prescribed a sacrifice for one who desired for heaven. The old belief that sin does not come to the eater or the seller of meat if he is not himself responsible for killing the animals still lingered on. Hence Manu declares in a MadhuVidya/535

Page #561
M. A. MEHENDALE verse (51) that not only the one who actually kills the animal but also one who consents to killing, who cuts it, who sells or buys it, and who cooks or serves it, all share the responsibility for the killing. The conflict between the spread of ahimsa and the Vedic sacrifice is reflected in many passages of the Mahabharata. The story of the pious hunter (3:198-199) (dharmavyadha) is instructive in this context. The hunter does not approve of meat-eating outside the sacrifices. But he carries on his profession of selling meat because he sees merit in performing one's destined duties although one may not approve of them. His defence of ritual killing is in part the same as mentioned above, viz. that animals killed in sacrifices are liberated from their lower state of life. But he has also a few other interesting observations to add. He says that sacrifices have to be performed since Fires have loved meat from ancient times. In fact the hunter seems to suggest that the very practice of meat-eating in general had originated in ritual meat-eating (3: 199.10). Secondly, he had come to the conclusion after a good deal of thought (bahu samcintya) that total ahimsa was not possible in practice. One may, with effort, be able to reduce its extent (yatnad alpatara bhavet). Even agriculture, which is looked upon as a good profession from ancient times, involves himhsa. The hunter seems to imply that if one does himsa even without one's knowledge, and for nothing, then why not do it in rituals which have the sanction of the Veda and which, for ought we know, may be beneficial. Even for meateating in general, which the hunter himself does not practice, he refers to the law of the big fish swallowing small fish. The hunter, however, seems to forget for a moment that in citing the example of fish he admits that those who live by eating meat are putting themselves on a par with lower animals. The Gokapilya episode which occurs in the Satiparva of the Mahabharata (12: 260) is equally interesting for the history of vegetarianism. Those in favour of ritual meat-eating are here seen as though giving a last-ditch fight against the mounting spread of ahimsa. Sage Kapila begins to doubt the correctness of the Veda which prescribes killing an animal. He asks the sage Syumarasmi if there was anything nobler than ahimsa. Now Syumarasmi enters the very cow which was about to be killed and replies that all persons, even animals and the vegetable world, desire for 422 Madhu Vidya/536

Page #562
AHIMSA AND VEGETARIANISM 423 heaven and heaven cannot be reached without a sacrifice. In fact, everything in this world is created for the sacrifice. The important point to be noticed in this episode is that the justification for a sacrifice is given through the cow herself, and not by a human being, thus adding an edge to the argument. However, complete victory for ahimsu and total vegetarianism is what we find in many other passages of the Mahabharata. Thus in the Anusasanaparva (13: 116) Bhishma advises Yudhishthira not to eat any meat at all. He asserts that meat cannot be had from grass, wood, or stones; to obtain meat, an animal has to be killed. But since ahimsa is the highest truth and the very foundation of dharma, the only conclusion is that eating meat is a sin. In fact, in the Santiparva (12: 257) it has been declared that the pious (dharmatma) Manu had prescribed only ahimsa in all activities. It was only the rogues who introduced himsa in sacrifices which is not what the Vedas really prescribed. The established fact, therefore, was that "ahinsa was superior to all other acts of piety" (ahimsaiva hi sarvebhyo dharmebhyo jyayasi mata). How agonizing the conflict between ahimsu and the Vedic religion must have been can be seen from a very ingenious, but hardly convincing, attempt to give a vegetarian interpretation to the Vedic injuction. It was proposed that "ajena yashtavayam" does not mean "one should offer a sacrifice with a goat", but "one should offer a sacrifice with corn which cannot sprout (a-ja), i.e. a three-year old corn" (Mbh. 14: 94.16). We have offered so far a very concise description of a long drawn struggle in which ahimsa came out victorious. The question may be raised: how did this come about? How is it that an important section of the people, the leaders and thinkers of the time, came to adopt ahimsa and total vegetarianism when this meant the giving up of the ancient and well-established institution of sacrifice? Are there any traces of thought in early times which could later develop in this direction, or did this all come about due to some outside influence? It is customary to trace the spread of ahimsa in Hinduism to the influence of Budhism and Jainism. It is well-known that ahimsa occupies an important place in the teachings of the Budha and Mahavira. It would also not be far from correct if one assumed that the propagation of his dharma by Asoka contributed in no Madhu Vidya/537

Page #563
424 M. A. MEHENDALE small measure to the stopping of the killing of animals either for sacrifice or for food. Asoka was no doubt very tolerant towards other religions. But he did not accept any compromise on the issue of offering animals in sacrifices. In his very first rock edict he proclaimed: "no animal shall be killed here for being offered in a sacrfice". As regards meat-eating as food, he made a very drastic cut in his own kitchen in order to show to the people that example was better than precept. Not that the good qualities of meat, both as tasty food and as a source of quick nourishment were simply lost sight of by those who favoured vegetarianism. They admitted that nothing can take the place of meat which was invaluable in the treatment of those who suffered from injuries, were emaciated, or grown weary by long journey. But, inspite of this, eating meat was objected to by these people on ethical grounds. They said nobody can be considered more mean and more cruel than the one who wished to increase his own flesh by eating that of someone else (Mbh, 13: 117.6-10). An echo of this reasoning is heard in the fifth Pillar Edict of Asoka where he says: "one should not nourish oneself by eating a living being (jivena jivo no pusitaviye)." It may be noted that the words used by Asoka in his edicts are vihimsa and avihimsa and not himsa and ahimsa. Some scholars have observed that the Hindu aversion to killing animals for sacrifice is not to be traced either to Buddhism or to Jainism. They feel, and on quite good grounds, that as far as the Jains are concerned their main concern was to wage a comprehensive struggle against the Brahmanical religion and Brahmanical arrogance on the whole. Their attacks on Brahmanical sacrifice formed only a part of it and hence the main emphasis, at least in the beginning, was in no way on ahimsa. They interpret this fact as suggesting that Jainism-and the same holds good a fortiori also for Buddhism-was not the real source of ahimsa which is supposed to have been later transferred to Brahmanism. They argue that the origin of the aversion to killing animals in sacrifices must be sought in the pre-Aryan, or what we may in a sense call the "original Indian", elements. It has already been shown that such non-Aryan characteristics of Hinduism as phallus worship or the sacredness attached to the pippala tree have to be traced to the pre-Aryan Indus culture. Similarly, ahimsa, or more especially a taboo on killing animals for religious practices, is to be MadhuVidya/538

Page #564
AHIMSA AND VEGETARIANISM 425 looked upon as a pre-Aryan, or a "really" Indian element that has crept into Hinduism, Buddhism and Jainism alike. The above conclusion is a plausible one. However, it has already been suggested that the concept of ahimsa can be explained as a generalisation of the ideas which lie behind the recitation of certain santi mantras in a sacrifice which are recited after an act of injury, actual or imagined, done to the sacrificial animal or the plant. Moreover we find some indications in the Yajurveda, a Vedic text more intimately related to sacrifice than any other, which lead us to feel that ahimsa with its far-reaching influence on animal sacrifice and meat-eating, would have appeared in the post-Vedic Indian thought even in the absence of the pre-Aryan elements assumed in the conclusion referred to above. Although the available Sanskrit lexicons record the word ahimsa in the sense 'non-injury' from the Upanishads onwards, the word in this sense is available in the recensions of the Yajurveda. A yajus formula 'oh plant, protect him, oh axe, do not injure him (oshadhe trayasva enam, svadhite ma enam himsih) is repeated in a sacrifice when the sacrificer is being shaved, or when a branch is being cut from a tree, or when the animal, after having been killed, is being cut open for taking out the limbs for sacrifice. On these occasions, when a razor, an axe, or a knife is used for these purposes, the above formula is repeated in order to ensure that the objects with which these sharp instruments come into contact do not suffer any injury. It was felt that the injury could be avoided if a blade of darbha grass was placed at a point where the razor or the knife came into contact with the man, the (dead) animal, or the tree. First, this blade of grass was addressed as oshadhi, an address significant in itself for bringing into play the healing properties of the plant, and then was appealed to for the protection of the person or the object from the impact of the instrument; next, the sharp instrument itself was commanded not to inflict injury. If one takes into account this concern to avoid injury (ma himsih) reflected in the formulas, it is possible to feel that it could contain the seed of that great feeling which in the course of time developed into the full-fledged doctrine of ahimsa. While this doctrine took root in Buddhism and Jainism, its presence in the Brahmanical thought itself had such an impact on the Vedic religion that it completely pushed the Madhu Vidya/539

Page #565
426 M. A. MEHENDALE animal sacrifice into background and played a substantial role in the spread of vegetarianism. NOTES 1 Cf., for instance, L. Alsdorf, Beitraege Zur Geschichte von Vegetarismus und Rinderverehrung in Indien, pp. 47-54. The present writer has been much benefited by this monograph 2 See H. P. Schmidt, aghnya-, Zeitschrift fur Vergleichende Sprachforschung auf dem Gebiete der Indogermanischen Sprachen, 78.46, n. 1 (1963). Madhu Vidya/540

Page #566
Dr. M.A. MAHENDALE MEDHATITHI ON MANU IX, 256 The Manusmrti IX, 256 runs as dvividhAMstaskarAn vidyAtparadravyApahArakAn / prakAzAMzcAprakAzAMzca cAracakSurmahIpatiH / / This stanza mentions two kinds of thieves' who deprive others of their property. They are referred to as prakasa and aprakasa (or pracchanna) in the Manusmrti (IX, 257). As becomes clear from the following verse, by prakasataskara! (or prakasavancaka)? are meant traders (panyopajivinah) etc., and by aprakasataskaraz (or pracchannavancaka)* are meant thieves, forest-dwellers, etc. (stenagavikadayah). While explaining the verse cited above Medbatitbi makes a remark which does not become clear at first sight. He says: prakAzataskarANAM nAtitaskaravyavahAro yathA loke'nyeSAmaTavIrAtricarANAmAptastaiH sAmAnyopAdAnaM tadvannigrahArtha kriyate / The expression atitaskaravyavaharah is not quite clear. It seems to mean 'dealings or acts (vyavahara) which are beyond i.e., which are different from those of a thief'. The best course to understand the whole sentence would be to start from the end. (i.e.37956127: explained by Medba. tithi as aTavIrAtricaraiH). (prakAzAnAM taskarA iti) sAmAnyopAdAnaM tadvannigrahArtha kriyate. 1. I.e. the daylight 'thieves'. 2. As they are called in Madu IX, 257. 3. I.e, the night 'thieves'. 4. As they are called in Manu IX, 257. 5. The printed editions have *TTEFIT4 or TF1RTEFTTH for which sce below. MadhuVidya/541

Page #567
S.V. UNIVERSITY ORIENTAL JOURNAL [Vol. XXI-XXII This means that the prakAza (thieves) and the aprakAza ( thieves) both are referred to in a similar way (sAmAnyopAdAnaM) as taskara in the stanza (dvividhAMstaskarAnvidyAt) so that the prakAza thieves can be arrested (nigrahArtham) like the amakAza ones (tadvat ). Hence in what precedes this part of the passage one expects to find the reason why both the type of paradravyApahAraka and the aprakAza type of paradravyApahAraka are referred to taskara. 58 The commentator says the reason is that the dealings of the type (ie. traders and others who might indulge in cheating the customers), are not beyond those of a thief, i.e. they are, in effect, similar to those of a thief. The activity (vyavahAra) of this larter type (ic. the forest-dwellers and others who act at night) is of course well known in the world (yathA loke'nyeSAmaTavIrAtricarANAmAsaH). The activity of the former, in so far as it results in depriving others of their wealth, is similar to these. Hence both are called by the common name taskara. That the paradravyApahAraka of the aprakAza type who in IX, 257, are specifically mentioned as stenATavikAdayaH is called taskara in our stanza needs no explanation. What requires explanation is the calling of the deceitful merchants etc. as taskara Medhatithi justifies this by pointing out that the dealings of the latter are not at all different from those of the former. na atitaskaravyavahAra thus in effect means taskarasadRzavyavahAra. The line, with a few additions in the parenthesis for the sake of clarity, may be read as prakAzataskarANAM na atitaskaravyavahAraH / yathA loke anyeSAM (taskarANAM) aTavIrAtricarANAM (vyavahAra) AptaH (tathaiva teSAM prakAzataskarANAM vyavahAraH) / taiH (aTavIrAtricaraiH saha prakAzAnAM 'taskara' iti ) sAmAnyopAdAnaM taMdvanimahArthaM kriyate / In the above discussion, the printed text as given in the editions of V.N. Mandalik (mAkAzastaskarANAm) and Ganganath Jha prakAzastaskarANAm ) has been emended to prakAzaM taskarANAm on the line of the word prakAzaM vaMcakA used in the next stanza ( IX, 257 ). If we, however, retain MadhuVidya / 542

Page #568
1978-19791 MEDHATITHI ON MANU IX, 256 59 the readings in the printed editions, we shall have to interpret. the passage as 9919T: (or 961 9T:) AFFITUTIH (meaning 925737 Ha: FATTOTI 49ETT:) 7 Safa af 549EURlt: 1 7911 19 etc. In the end it is good to examine the translation of Ganganath Jba of the above passage of Medhatithi. His translation runs : Though the action of the 'open' thief does not stand on the same footing as that of the 'concealed' one - such as those who prowl about at night, in forests etc. - yet both have been mentioned together for the purpose of indicating the equality of the punishment to be meted out to them This interpretation, bowever, is not likely. For, apart from the fact that it requires the addition of two crucial words "though..... yet", it is well to remember that in the very next verse (1X, 257) Manu characterizes both the types as vancaka. Thus, according to Manu, the activity of the prakaso type is siinilar to that of the aprakasa type and not different. Hence Medhatithi is not likely to say tbat the action of the one does not stand on the same footing' as that of the other. If the punishment to be meted out to bem is to be equal, their actions have to be on the same footing and not different. Madhu Vidya/543

Page #569
Did Ashoka Ban Cow-slaughter ? A FEW YEARS after he embraced Buddhism, King Ashoka issued what are known as "rescripts on morality" (dhammalipi). He caused them to be inscribed on rocks and pillars throughout his empire for two reasons : he wanted to give them a sort of permanency so that his children and grandchildren could know what he wanted them to do and, secondly, to make his views easily available to the people. Ashoka's dhamma which he tried to propagate through these edicts had nothing specifically Buddhistic in it. His dhamma meant, in effect, practising certain virtues like obedience to parents, proper courtesy to servants, liberality, but above all, abstention from killing animals. This he emphasises time and again in his rescripts. The king admonishes his administrative officers that while touring the districts under their charge, they should give moral instruction to the people. They are to tell the people that "non-injury to animals is meritorious". The king himself had substituted the pleasure tours of the former kings by his religious tours. Earlier, the kings in their pleasure tours used to indulge in hunting and other practices which resulted in injury to animals. King Ashoka, on the other hand, in his religious tours instructed the people in his dhamma and enquired how far they had progressed in the practice of the virtues since he visited them last. In the seventh Pillar Edict Ashoka says that in respect of the promotion of morality among his subjects he has succeeded where his predecessors had failed. The king attributes this singular success to his having taken recourse to two measures, viz. issuing of rescripts on morality and, secondly, persuading the people to give up their old practices in favour of the dhamma recommended by him. The king is on record that he found the method of persuasion much more effective than the issuing of rescripts. It was by persuasion that the king was able to wean away his subjects from causing injury to animals. The bilingual inscription of Ashoka in Greek and Aramaic found at Kandahar in Afghanistan gives the impression that the king was successful in turning his empire into a virtual welfare state for the animals. It says that in the king's dominions hunters had given up hunting and fishermen fishing. The people had followed the king's example and had become vegetarian. There may be some exaggeration in this statement which has been inscribed at a place far away from the capital and some misunderstanding of the text was possible while translating the original edict Madhu Vidya/544

Page #570
into foreign tongues. But the fact remains that it testifies to the growing enthusiasm of the people to put the king's dhamma into practice. The fifth Pillar Edict of Ashoka, issued twenty-six years after his coronation, is devoted primarily to the protection of animals. The king, first, gives a long list of birds and animals which are on no account to be killed. These include parrots, mainas, geese, pigeons, porcupines, bulls set at liberty, also certain kinds of fish, and tortoises. There are certain names of animals in this list like nandimukha and getata which are not intelligible. The list ends with a general statement that all the quadrupeds which are neither useful nor edible are not to be killed. * Ashoka, next, orders through this edict that she-goats, ewes and sows are not to be killed when they are either with young or in milk. Their young ones also were not to be killed if they were less than six months old. He further forbids nourishing of living animals with other living beings. On certain specified days fish were not to be killed and were not to be sold.. He put restrictions on the castration of animals since this involved injury. Thus he enjoined that whereas cocks shall not be caponed at all, bulls, he-goats. rams, boars and whatever other animals are usually castrated shall not be castrated on certain specified days. Similarly, on certain days, horses and bulls shall not be branded. The first Rock Edict of Ashoka, no less than the fifth Pillar Edict, is significant for his concern for the sanctity of animal life. The Rock Edict may be said to have three parts and the pronouncements the king made in them affected to a very large extent the people's religious and social life, and his own personal life. In the first part of the edict the king orders a blanket ban on the slaughter of animals for the performance of the sacrifices : "Here, no animal of any sort whatsoever shall be killed and given as an offering." Although king Ashoka is rightly known for his religious tolerance -- his twelfth Rock Edict is remarkable for his catholic spirit -- he does not hesitate to put a complete stop to the animal sacrifices which must have meant in those days an interference with the religious practices of those who followed the Vedic injunctions. In fact, this is the only pronouncement in the Ashokan inscriptions which may be characterized as specifically Buddhistic. As rightly observed by the late Professor Alsdorf (Beitraege zur Geschichte von Vegetarismus und Rinder-verehrung in Indien, p. 52), it is very difficult to imagine a Hindu king of Ashoka's time ordering, "Here no animal shall be killed for sacrifice". But the attempt of Professor Alsdorf to try to bring this order, at least to some extent, in accord with the king's religious tolerance by interpreting the word 'here' to mean only 'in my capital, Pataliputra', and not 'throughout my empire,' is not convincing. It has rather to be admitted that on this basic issue the king was, clearly, in no mood to compromise. Madhu Vidya/545

Page #571
In the second part of this edict the king asks his subjects not to indulge in certain social entertainments (samaja) which he does not approve, for he sees many objectionable things in such festivities. Ashoka does not elaborate this point, but it will be a reasonable guess to suppose that by prohibiting such festivities the king wanted to put a stop to the suffering caused to animals on such occasions. In the last part of this edict the king announces that he had ordered almost a total ban on killing animals for the royal kitchen. Whereas, formerly, many animals used to be daily slaughtered for preparing curry (supa), now, since the edict had been issued, only two or at most three animals were being killed for this purpose. Since the king has shown honesty in making a public admission of this exception. there is no reason to doubt that the small number of animals mentioned by him (two or three) is correct. It should thus be clear that the Buddhist king shows, both in precept and practice, great regard for the sanctity of animal life. The question arises, does the king in all his efforts to save animal life show any special regard for the cow? Does he specifically ban cow-slaughter? The answer to the question has to be a plain 'No'. Professor Alsdorf (op. cit., p. 59) points out that the cow does not find a place in the fairly long list of animals which are declared inviolable by the king in his Pillar Edict V. The explanation sometimes offered of this omission, namely, that the cow has not been listed among these animals since it was even otherwise not killed in those days, will hardly stand scrutiny. On the other hand, the king's declaration : "I have made inviolable ... all the quadrupeds which are neither useful nor edible" shows that if a cow (or a bull) was edible or useful for therapeutic and other purposes, the king had not made it inviolable. It is possible that owing to the various measures taken by Ashoka--ban on animal slaughter for sacrifices, ban on objectionable social festivities, almost total prohibition of meat-eating in his own household, and his repeated exhortation to abstain from injury to animals -- Cow-slaughter was considerably reduced in his days. But this must have been true of other animals as well. The king in this respect made no distinction between a cow and a goat and had not issued a total ban on cowslaughter leaving other animals to their own fate. M. A. MEHENDALE Madhu Vidya/546

Page #572
THE MITANNI NAME SATTIVAZA In his book Die Arier im Vordern Orient - ein Mythos? M. Mayrhofer mentions (p. 25) a Mitanni royal name Sattiuaza- which, according to him, reflects IndoAryan *sati-vaja. "Kampfguter erlangend'. In support of his view he mentions some Old Indo-Aryan compounds such as vaja-sani-, vaja-sa, vaja-sati (also vajasya sati.), in which we have a combination of vaja- and a derivative of san(i): sa. He also refers to the fact that in Old Indo-Aryan we often come across expressions where waja happens to be the object of san(i). In all the attested compounds cited by Mayrhofer vaja- occurs as the first member, not as the second. He is, therefore, compelled to interpret sati-vdja- as 'Kampfguter erlangend', as though the compound were sanad-vaja-. The attested word vaja-sati-, however, he renders as 'das Erlangen von Gutern'. The difficulty can be avoided if the Mitanni name is considered a reflex of Old Indo-Aryan *satya-vaha "that which carries truth, vehicle of truth, truthful one'. Satyavaha- (Bharadvaja-), as a proper name, is attested in the Mundaka Upanisad 1. 1. 3, Satyavahana in the late Rajatarangini 4. 100. As for the older period, we have a 'Textfigur' in which not satya-, but its near-synonym rta- occurs, which lends itself very well to composition. As an expression for a 'hymn' we have rtasya vahas 'vehicle of truth' in the RV. 8. 6. 2 (Luders, Varuna, p. 442). And the compound itself, asa-vazah, is attested in the Avesta (Y. 10. 14) as an epithet of Haoma. Since, however, sarti- for lartija (cf. nasattija-) constitutes a difficulty. *saptivaha- "one whose draught-animal is a horse' may be preferable. Cf. RV. sapti . . . vahato ratham (8.33.18), saptayo . . . vahantu (1.47.8, cf. 1.85.6), etc. Late parallels would then be the adjectives saptivahana (Kalkipurana 3.8.1) and saptivahin (Paramanandakavya p. 88 verse 4). Deccan College, Poona GALAYASTI In an article 'Lexicalische und grammatische Bemerkungen zu den Asoka-Inschriften' contributed to Ludwig Alsdorf Memorial Volume, Thieme cites (p. 298) the follow. ing stanza from the Sarngadharapaddhati 3.2.9: Madhu Vidya/547 M. A. MCHENDALE

Page #573
calatkamimanominam adatum cittajanmanah / galayastir ivabhati balaveni gunojjvala // In this stanza the braided hair of a young girl are compared to a galayasti. Thieme remarks that the meaning of the word galayasti is not immediately clear. He tentatively suggests fetter, shackle' (Fessel ?), or 'net' (Netz ?) as the meaning. But it seems that galayasti can only mean a fishing rod to which the hook is attached. We have in Pali gala, in Prakrit gala, and in some of the modern Indian languages gal or gal, which mean a hook, a fish-hook'. Tumer, in his etymological dictionary, traces these words to *gada (3971). In the text of the Sarngadharapaddhati edited by P. Peterson (Bombay, 1988), where the above stanza is numbered 3290, occurs a variant reading jalayasti. This would mean a stick which holds the fisherman's net. M. A. MEHENDALE MadhuVidya/548

Page #574
THE INDO-ARYANS, INDO-IRANIANS AND THE INDO-EUROPEANS M. A. MEHENDALE 1. Indo-Aryans The term 'Indo-Aryan' is applied to a specific group of languages spoken primarily in North India and in some countries around India, e.g. Pakistan (Sindhi and Lahnda or Western Panjabi), Nepal (Nepali), Bangladesh (Bengali), and Shri Lanka (Simhalese). The forefathers of the present-day speakers of these languages are supposed to have entered India from Iran along the North-western mountain passes. This event may have happened c. 2000 B. C. The designation 'Indo-Aryan' serves a two-fold purpose in the first instance, it distinguishes this group of languages from Old Iranian together with which it forms the easternmost Indo-Iranian branch of the Indo-European family of languages; and (2) it distinguishes this group of languages from the other three families of languages spoken in India, viz. the Dravidian, the Munda, and the Tibeto-Burman. The term 'Indo-Aryan' has two components: the former component has its justification in the fact that these languages are spoken mainly in a large part of India (before partition) as distinguished from Iran, Afghanistan, Central Asia etc... where other languages of the Indo-European family are spoken. The second component aryan' used in the term 'Indo-Aryan' needs clarification. It must be understood that it has nothing to do with any particular race. The use of the word 'drya' does not mean that the speakers of these languages belong or belonged to any one race called 'Arya'. It is well known that the speakers of these languages today represent a mixture of races. And as for the past, we have no justification to assume that the speakers of the languages at the Indo-Iranian stage belonged to a single race. The word drya has its justification in the designation 'Indo-Aryan' in the fact that the speakers of the languages in the oldest attested stage in India (as well as in Iran) called themselves 'drya'. But this word refers to certain quality or qualities and not to race. The word 'drya' has left no trace in the name of our country where these languages are presently spoken. The country today, as in the past, is called 'Bharata (varsa)'. At one time, however, as witnessed by Patanjali, the word 'arya' appeared in the name Aryavarta which was then applied to northern India. In answer to the Madhu Vidya/549

Page #575
question kah punar aryavartah 'what is Aryavarta', Patanjali on Panini 6.3.109 says: the country bounded by Adarsa' in the west, Kalakavana? in the east, Himavant in the north, and Pariyatra (i.e. Vindhya) in the south is known as Aryavarta. According to Manusmrti 2.21-22 it is the land between the two seas (in the west and east) and between the Himavant and the Vindhya. Today the word arya, through Pali and the Prakrits, survives only in the word aja (m.) aji (f.) meaning 'grand-father' and 'grand-mother' respectively in some of the north Indian languages. The story of the arya is different in Iran. There it continues in the name of the country, Iran, which is derived from airyanam (gen. pl. of airya) = Sk. aryanam 'of the Aryas' with the word for 'country'supplied. 'Iran' thus means '(the country) of the Aryas'. The history of the Indo-Aryan branch of languages in India is conveniently divided into three broad stages--Old, Middle and the New. The oldest attested stage of the Indo-Aryan is found in the Vedas -- particularly the Rgveda. The text of the mantras as composed by rsis has been substantially well preserved except for making some phonetic changes of the type of vowel sandhi. Although the mantras of the Rgveda have been composed by various authors, we do not find in them dialectical variations. The language of the Rgveda gives us an impression of a homogenous language. Later we see successive developments of this language in post-Rgvedic texts like the Brahmanas, the Upanisads and the Sutras. This late Vedic stage approaches closely to the bhasa -- 'spoken dialect' -- described by Panini in his Astadhyayi and which bacame the model for the writers of the Classical Sanskrit. The old Indo-Aryan has more or less maintained without change its linguistic character since the days of Patanjali (2nd cent. B. C.). In the classical period the language does not show the types of changes which are seen earlier from the period of the Rgveda to that of the sutras. This difference between Vedic Sanskrit and Classical Sanskrit is not in phonology. The vowel and the consonantal systems are the same except the disappearance of the intervocal | and lh found in the Rgveda. The difference lies chiefly in grammar and vocabulary. In grammar, many forms like the nom.pl. in-asas, infinitives in -tavai, -dhyai etc. are lost. In vocabulary we see some old items like jalasa (?), ceru (?) disappearing, some others being retained with a difference in meaning, cf. Kratu-, ram-, kup. But more important is the addition of many words which are borrowed from languages, either Middle Indo-Aryan, cf. dohada, mukta, utsuka, aksauhini etc. non-Sanskritic, cf. puja, valaya, kundala etc. The second stage of the development of the Indo-Aryan languages, viz. the Middle Indo-Aryan, is represented by the Pali and the literary Praksts. There is no doubt that certain middle Indo-Aryan tendencies like the disappearance of the vowelr and the emergence of the retroflex sounds had started as early as the Vedic period. Madhu Vidya/550

Page #576
This is witnessed by the presence of such words as (vi) kata (already in the Rv) nata, bhata in the late Vedic literature. The spoken languages of those times must have considerably differed from the languages of the educated and the cultured people. The distinction between the standard, the polished, the hieratic and the substandard forms was expressed by the terms samsksta and praksta applied to them. The term praksta as opposed to samskrta, is to be understood either as 'natural, unsophisticated' form of speech as opposed to samsksta which is 'refined, polished' form of speech; or it can be understood as the grammarians do it : Samskrta language is the praksti 'basis' and praksta is the one 'derived from this prakrti' (cf. Hemacandra 1.1 prakrtih samskrtam/tatra bhavam, tata agatam va prakrtam). It must be remembered that this praklia could not have directly come from the late literary or standard Sanskrit, but from its earlier spoken variety in the Vedic period. This explains why in Pali we have ! and !h in place of d and dh exactly as in the Rgveda but not in the classical Sanskrit, or why we have the instr.pl.endings -ehi (Pali) or ehim (Pkt) < Vedic ebhih, or the nom. pl. ending - aso (Pali), aho (Mg.) < Vedic - asah; the classical Skt. does not have these terminations but has aih for instr.sg. and -as for nom. pl. The period of the middle Indo-Aryan stretches roughly form the 5th cent. B.C. to 1000 A. D. It could be fairly easily assumed that these languages played a distinct role as means of communication in the everyday life of the people since both Buddha and Mahavira favoured the use of these, as against Sanskrit, for teaching their messages, The earliest attestation of this middle Indo-aryan stage is found in Pali, the language of the Buddhist canon and in the inscriptions of Asoka. It is believed that the Buddhist canon was first formulated in the eastern dialect, the dialect of Buddha himself, and that it was later translated into Pali. The assumed eastern canon is nolonger available. The Asokan inscription reflect at least two varieties of the MIA stage, the eastern (with/and nom. sg.in - e) and the western (with r and nom. sg. in - o), and perhaps a north-western (having three sibilants s, S and s and many consonant clusters). The principal languages included in the MIA stage are the Ardhamagadhi, the Sauraseni, the Magadhi, the Paisaci and the Maharastri. The Maharastri does not represent any regional variety of MIA but a stage of development that lies between the literary Prakrits on the one hand and the Apabhramsa on the other. The Apabhramsa of about 1000 A.D. marks the close of the Middle Indo-Aryan period. The new Indo-Aryan or the last stage in the development of the Indo-Aryan languages is said to have commenced in the 11th cent. A. D. and is represented in the various standard and sub-standard regional languages of northern India. These languages apparently grew out of the local Apabhramsas which, although not attested for different regions, must be assumed to precede and be the starting points of the NIA languages. Madhu Vidya/551

Page #577
II. Indo-Iranian The language family designated as 'Indo-Iranian' includes the Old Iranian and the Old Indo-Aryan. It represents that stage in the linguistic development of a branch of the Indo-European which is the source-language for the Avestan and the Old Persian in Iran and the Vedic Sanskrit in India. A certain group of people speaking a certain variety of the Indo-European migrated from the original home of the IE towards the south and east. It is supposed to have settled down for some time to the north of Iran and Afghanistan from where one branch came down to eastern Iran and after some time migrated further south through the passes of the Hindukush to reach the north-western forntiers of India. Those who remained behind in the common home to the north of Iran and Afghanistan came later to western Iran and then even occupied the eastern Iran. It is in common home of the forefathers of the later Iranians and Indians that Indo-Iranian is supposed to have been spoken. The Old Iranian of the settlers of Iran is represented by the Avestan, the language of the holy texts of the Zoroastrians and the Old Persian found in the inscriptions of the Achaemenian rulers of Iran. The date of oldest portions of the Gaths of the Avesta is likely to be c. 1000 B.C., while the old Persian inscriptions date from the 5th century B.C. The Iranian branch, like the Indian branch, can be conveniently divided into three stages -- the old, the middle and the new. The old Iranian is available in two dialects -- the western in the Old Persian inscriptions and the eastern in the Avesta, The middle Iranian is represented by various languages, chief among which is Pahlavi. The new Iranian is represented among others, by modern Persian in Iran, Pusto in Afghanistan and Baluchi in Baluchistan. Some centuries after the fore-fathers of the Vedic Aryans left eastern Iran, the Iranian prophet Zarathushtra introduced certain radical reforms in the traditional religion and gave a new message to the people of Iran. He recognised one god, Ahura Mazda 'the wise Lord', as against the many in the religion inherited by him, and forbade the killing of animals and offerings of flesh together with those of a stimulating drink of haoma (Sk. soma) in religious worship. The linguistic impact of the revolution is seen in the meaning of the daeva (Sk. deva)'demon, evil spirit' and the establishment of a series of parallel pairs of synonyms, one of which was used in connection with the ahuras 'gods' and the other with the daevas demons'- withness the use of us'ear', vayoana'head', zasta "hand', (Sk. hasta) etc. used with reference to Ahuric i.e. divine being and the use of kar@na (Sk.karna)'ear', kamorada 'head (Sk. * ku.murdhan) and 'go','hand' used with reference to daevici.e. demoniac beings. That the forefathers of ancient Indians and the ancient Iranians were at one time one people showing common culture and speaking one language is established beyond doubt due to the close affinity between Vedic Sanskrit and Avestan-an affinity so close as is Madhu Vidya/552

Page #578
found in any other two branches of the Indo-European family of languages. If one glances through the pages of the Altiranisches Worterbuch--the extent of which is far too smaller than of a Sanskrit dictionary--one meets with many vocabulary items which have cognates in Sanskrit and hence prove their common origin-- examples chosen at random: Sk.hiranya, Av.zaranya'gold', Sk.vrksa, Av.var@sa tree; Sk.matsya. Av.masya'fish'; Sk., send, Av. haend 'army'; Sk. yajna, Av. yasna 'worship, sacrifice'; Sk. hotar Av. zaotar 'priest'; Sk. Mitra, Av. miera 'name of a deity', etc. etc. the similarity between the two languages is not restricted to the vocabulary items in large number; it is witnessed even more strongly in grammer -- that is in the manner of root alternation cf. Sk. gam: gacch, Av. gam:jas, in the use of vikaranas to form stems, e.g. Sk. su: sunu, Av. hu: hunu; and in terminations, e.g. Sk. bharami, bharati, Av. barami baraiti. Not only this. Both these branches have made common innovations in the parent Indo-European language, e.g. they have merged IE a, e, o> a, and changed IE m, n > a; both show palatalistion before front vowels as in ca-kara, ja-gama and the introduction of n before gen.pl.term, hence Sk.devanam, Av.daevanam against Gk.hippon, Lat. deum=deorum etc. The similarity between Sanskrit and Avesta is best illustrated by showing how an Avestan line like tom amavantom yazat@m sur@m damohu s@vistem mir@m yazdi zao@rabyo (Yt. 10.6) can be easily rendered into Sanskrit - tam amavantam suram dhamasu favistham mitram yajai hotrebhyah (for hotrebhih) 'that strong, brave, bravest among the creations, Mitra, I worship with oblations". It is very likely that the Indo-Iranians, after they left the original home of the IndoEuropeans, came to Iran and Afhganistan from South Russia. The common view regarding their coming to this new land together and the subsequent migration of a part of them to India has to be a little modified on account of certain linguistic evidence brought to light from the Near East. There, in upper Mesopotamia, Mitanni Kingdom was established by people who spoke a language very similar to the Indo-Iranian or the Old Indo-Aryan. These people are supposed to have come to Mesopotamia between 17411600 B. C. The Cuneiform inscriptions discovered in this area reveal certain terms and names of deities which are unmistakably of the Indo-Iranian branch. Some of the names of the Mitanni princes also show closeness to Indo-Aryan, e.g. suvar-data = Sk. svar-datta or svar-dhata'given or created by the sun'; Indrauta Sk. Indra-uta protectd by Indra'. In a treaty concluded in 1360 B. C. by the Hittite king Suppiluliuma with Mattiwaza of the Mitanni people when the latter was made a king and was given his daughter by the Hittite king, we have the mention of the names of deities as witness to the treaty. In this list of gods we find, by the side of a number of non-Aryan deities, such names as are easily identifiable with Vedic gods Mitra, Varuna, Indra and Nasatya. Then in a Hittite text composed by one Kikkuli on horse-training we find such numerals as aika 'one', tera 'three', panza 'five', satta 'seven', na (va) 'nine', and such words as Madhu Vidya/553

Page #579
asua 'horse' and technical terms like uartana 'circular course (in which a horse moves when under training). The question raised by this unmistakable Aryan evidence in Asia Minor is from where did the people speaking the language akin to the Indo-Iranina reach there. Did they go there sometime during their migration before or after the Indo-Iranian broke into Iranian and Indian, that is did they go there as Ur-Aryans, or as Iranians or as Indians? At one time it was felt that some of the speakers of the Indo-Iranian language separated from the others who later became proto-Indo-Aryans and went to Asia Minor from south Russia along the Caspian sea. Some of them remained there to establish the Mitanni kingdom in south but others went to Iran. This theory is, however, not acceptable because the Aryan remains in Asia Minor are more akin to Indian than to Iranian or to Proto-Indo-Iranian, e.g. we have s and nothin satta, the word for seven'(Sk.sapta: Av. hapta), in nasatya'the name of the deity'(Sk.nasatya, Av.nanhaiya), and the word for 'one'is aika with -k-as in Sk.eka, as against Av.aiva which has -v-. This should force us to admit that those who established the Mitanni kingdom were not Iranians but Indians who may have gone to the Near East from India. Between these two extreme theories, a compromise theory is proposed because the Aryan evidence from Mitanni is not exactly identical with Sanskrit either, but occaisonally shows an older stage of development as witnessed in the word for 'one' aika which shows a diphthong ai as opposed to the monophthong e in Sk. eka. According to the compromise theory the speakers of the Mitanni Aryans went to the near east after the proto - Aryans separated from the IndoIranians but before the forefathers of the Indo-Aryans came to India. T. Burrow chooses to call this stage of separation as proto-Indo-Aryan. Thus we do not imagine the relationship of the Near Eastern Aryan, Old Iranian and Old Indo-Aryan as something like - (1) Proto-Aryans non-Old Indo-Aryan Old Indo-Aryan (in India) Near Eastern Aryan Old Iranian (in Iran) Madhu Vidya/554

Page #580
(2) nor as Proto-Aryans Old Iranian (in Iran) Old Indian (in India) Near Eastern Indian Old Indo-Aryan (3) but as - Proto-Aryans Old Iranian (in Iran) Proto-Indo-Aryan (in Iran) Near-Eastern Indian Old Indo-Aryan Such a view permits us to say that the cluster pt became it in Near-East, but was preserved in India; but the diphthong ai>e in India was preserved in the Near-East. This happened after the branching off of the Near Eastern from Proto-Indo-Aryan. III. Indo-European The Indo-Iranian family togt.her with many other languages or language-families like Greek, Latin, Germanic etc. which were once spread over large parts of Asia and Europe form a family of languages known as the Indo-European family, as distinct from such other families of languages as Dravidian, Munda, Tibeto-Burman (in India) and Semitic, Finno-Ugrian etc. (in Europe and Asia). Similarities between the vocabulary items of Sanskrit and some other languages like Latin or Iralian were long recognised by the missionaries who arrived in India. But they remained more or less as curiosities. The credit for the formulation of the theory of a common origin for these languages goes to Sir Willian Jones who made the now famous pronouncement in 1786 A. D. It was he who for the first time declared that the linguistic similarities between the languages now grouped in the IE family of languages can be Madhu Vidya/555

Page #581
explained only on the assumption of their having sprung from a common origin which, perhaps, no longer exists. Today no one is in doubt that the common source, which was once called IndoGermanic and which is now know as Indo-European, no longer exists, which means that none of the attested language like Hittite, Sanskrit, Greek and others, however old they may be, can be looked upon as the 'mother' language of this family. The languages, or the language families, comprising this large family of languages are: (1) Indo-Aryan, (2) Iranian, (3) Armenian, (4) Albanian, (5) Balto-Slavic, (6) Greek, (7) Latin, (8) Germanic,(9) Celtic, and the two more recently discovered, (10) Hittite and (11) Tokharian. It is likely that the Indo-European common speech had already started showing dialectal variations within its 'home' before its speakers migrated into different lands. This means that not all the characteristics which distinguish these languages from one another developed after their separation from each other. Thus the augment a- in the preterite forms like Sk. abharat, Gk, ehpere, Arm. eber is seen only in these three languages and at one time must have formed a common feature of these three, as distinct from others. Later when the Greek was separated from this group, there occurred another innovation, viz. the change of the Indo-European palatal k'to s in certain languages like Sanskrit, Avestan, Armenian, Albanian and Balto-Slavic, but its continuation as a velar sound k in the rest. This distinction is widely known as a sat@m-centum division based on the words for 'hundred' in Avestan sat@m on the one hand, and Latin centum on the other. On the basis of the comparison made between the different languages comprising the Indo-European family, it has been possible to reconstruct the 'common origin' of these languages. The technique of such a study was first formulated by Franz BOPP in 1816 in his work: Uber das Conjugations-system der Sanskritsprache and was further developed by August SCHLEICHER in 1852 in hiswork: Formenlehre der Kirchenslawischen Sprache. Now it is possible to assert that the 'similarites' between these languages are one-time "identities" and demonstrate what that "identity" could look like. When one compares the sound systems of the different languages of the IE family with the sound system of the reconstructed original language it is observed that Greek and Latin of the centum group have better preserved the vowel system of the original language (it has preserved the original a, e, o vowels while Sanskrit has merged them into a single vowel a) while Sanskrit of the satam group has better preserved the original consonant system (Sanskrit has kept the distinction between voiced unaspirates and aspirates like d and dh, b and bh, while this distinction has been lost in other languages). While it is not likely that the community speaking the original IE language was a racial community, it must certainly have been an economic and a cultural community. The people of this community used a common language, Indo-European, for their Madhu Vidya/556

Page #582
intercourse. As far as the social organisation of this community is concerned it can be said that it was patriarchal in character. The people of this community had some definite ideas about the gods who ruled over the lives of human beings, and about death and the other world. They used some form of worship for these gods and knew the art of composing poetry in which they praised their gods. The question about the original home of this community from where it spread to different parts of Europe and Asia has been discussed for a long time and many theories have been proposed. The two of these which gained some currency were those which placed them in Central Asia or in South Russia and the Kirghiz steppe. The latter theory is mainly based on the assumption that a part of the agricultural terminology which is common to other IE languages is lacking in Indo-Iranian. This means that one section of the community lived a more or less nomadic life which can be only in steppes. Therefore the home of the community must be found at a place where steppe and agricultural land come together. The land when boundaries of Southern Russia and the Kirghiz steppes meet was considered as ideal to fulfil these requirements. P. THIEME who has examined this theory in his Heimat der indogermanischen Gemeinsprache points out(p.25) that P. SCHRADER himself who had advocated this theory had to admit in the first instance that not all, but only a few, IE items related to agriculture were missing in the Indo-Iranians and hence we have to admit that the Indo-Iranians too knew agriculture. It is of course likely that, as compared to the other IE communities, agriculture did not play at that time an important role in their life. Among the items considered missing, it is mentioned that the Indo-Iranian group does not have a cognate for the IE root * sel sei'to sow'. But J. Bloch has demonstrated that it, in fact, is to be seen in Sk. sira (nt.) 'sowing, a plough for sowing sita (f.) furrow' (originally the bedecked'). From the explanations of J. Block it also become clear why there is no correspondance in Sanskrit for Gk. arotron'a plough'. This is because the forefathers of the Indians in their wandering came to know a new type of plough which was equipped with an arrangement of simultaneous sowing. This kind of plough was known in Mesopotamia already in the 3rd millenium B.C. SCHRADER similarly missed the words in Indo-Iranian for IE words meaning 'to grind', 'demestic pig' (according to him Indo-Iranians knew only wild pig.cf. sukara), and 'salt'. It is not necessary here to go into the details of the arguments. It is sufficient to note that these objection have been successfully met and the knowledge of agriculture and pig-breeding can be definitely assumed for all branches of the IE family including the Indo-Iranian. Such people, unlike the nomads, could not have occupied a large territory as, for example, the one stretching from the Baltic sea in the north to the Kirghiz steppe in the south. Madhu Vidya/557

Page #583
It is, therefore, necessary to narrow down this region as far as possible. THIEME, with his now famous 'Lachs (salmon) theory demonstrates how this narrowing down can be achieved. His argument can be briefly stated as follows: Old HG lachs, Anglosaxon leax, old Norse lax etc. point to a common origin IE * laks -/lakso- for salmon fish'. This kind of fish is to be found only in the rivers which flow into the northern oceans (the North sea and the Baltic sea) and their tributaries. There is no salmon in Greece and Italy. Hence we have no correspondences for IE *laks-/*lakso - in Gk.and Latin, and also in south Slavic. The Celts had their own word which was later borrowed as'salmo'in Latin. That the Old HG lachs and the related words are not an innovation in the Germanic and in the north Slavic is demonstrated by the fact that Tokharian, discovered in east Turkestan, has a correspondence in laksi meaning 'fish'. That the word in Tokharian does not mean 'salmon is understandable, since there are no salmon in Central Asia. The Tokharian branch has thus preserved the old IE word but gives it a more general meaning fish'. The presence of laksi in Tokharian clearly shows that the speakers of this branch came from the region where salmon was the fish par excellence. FIE had a word like * laks-/*lakso - its correspondence in Sanskrit would be laksa, but obviously it cannot have the meaning 'salmon'since that kind of fish is unknown to India. Now Sanskrit does have a word laksa which means a very large number. In many languages there are words for things which appear in large number or in great mass and therefore are themselves expressive of large numbers. For example, in Avesta the word baevar meaning ten thousand' is derivable from a word which means a bee' (cf. Gmc. im-pi-'swarm of bees'). In Sanskrit itself the words jaladhi or samudra'ocean'and abja 'lotus'signify large numbers. One of the characteristics of the salmon is that they appear in large numbers. Thus we can argue that Sanskrit has preserved the IE word *lakso- in the form laksa-, not in the meaning 'salmon'but in the meaning one hundred thousand'. A second characteristic of salmon is the red colour of its flesh. An adjective in IE from *lakso/*lakso would be *lakso and would mean 'salmonish,red'. This corresponds exactly with Sanskrit taksa attested since the Atharvaveda, which means red lac tree which originally could have meant 'the red'. In Ossetic, an Iranian dialect, we find a word lasag which means 'trout', a variety of fish, similar to salmon, found in the Caucasian rivers. Allthis evidence points to the north-central Europe as the probable home of the IndoEuropeans and not to the south Russia and the adjoining Kirghiz steppe. On the basis of the evidence given by the IE word for the salmon, it is thus possible to narrow down the limits of the original home of the IE: in the east one does not have to go beyond the beach boundary (from Kalinigrad on the Baltic to Odessa on the Black sea). In the west we need not go upto the river Rhine. This leaves the area of the salmon Madhu Vidya/558

Page #584
rivers Vistula, Oder and Elbe, roughly the area where the eastern IE languages like Baltic (Lithuanian) and Slavic (Polish) and western Indo-European (Germanic) meet, or the area corresponding to Poland and eastern part of Germany. In this region we have trees and animals, words for which can be reconstructed for the IE on the basis of correspondences in the related languages. All these trees and animals are characteristic of this region and there is not a single item in the list which is not compatible with the region. These can be listed as: Trees : birch, beech, aspen, oak, pin, fir. Animals (domestic): dog, fox, sheep, horse, pig, goat. Animals (non-domestic): wolf, bear, fox,stag, hare, mouse, serpent, turtle. Birds : eagle, haron, owl, crane. Acquatic birds : goose, duck, diver. Insects : fly, wasp, bee, louse. D. E. F. THIEME asserts that there is in the above list no animal domestic or otherwise, that we cannot assume for the later stone-, bronze-, and copper-periods in the German low level land. It seems to me that THIEME has very cogently argued his case. Just as on the basis of the linguistic evidence which consists of finding out words which can be reconstructed for the Indo-European and which designate animals and trees which are characteristic of a given region -- in this instance, particularly the words for salmon fish and for beech and fir trees -- he has pointed to the north-central Europe as the home of the Indo-Europeans, he has excluded other countries whose most characteristic plants and animals have no parallers in other languages and hence cannot be considered for being the home of the Indo-European people. All these words must be looked upon as later borrowings in those respective languages which the speakers of these families came to know after they arrived in those lands. He thus excludes India as the home of the IE due to the impossibility of reconstructible words for elephant, tiger, monkey,fig-tree etc; Iran, because there are no IE words for camel, donkey, lion etc; the Mediterranean countries like Greece, Italy, Albania, France, Spain for the absence of IE words for donkey, lion, olive, vine, cypress etc. The case for India can be excluded further on the basis of the available language situation in this country -- the concentration of Indo-Aryan languages in the north and of Dravidian languages in the south with some pockets in the north, -- the geographical knowledge from the Rgveda downwards showing gradual expansion towards east and the south from the north-west, and ancient legends pointing to invasions of the people speaking Sanskritic languages towards south rather than the other way round. Madhu Vidya/559

Page #585
The Rgvedic singers seem to be more or less settled down in the land of the river Sindhu and that of the seven rivers, that is the north-western part and Punjab of the undivided India. They knew the river Ganga, but the river had not then assumed the importance which it did in later Sanskrit literature. They had yet no knowledge of the lotus, the nyagrodha, the tiger and rice, which are all foreign to the north-west of India. The authors of the Atharvaveda hymns show further advance towards east and the south. The tiger now appears as the most feared beast of prey and his skin becomes the sign of king's power. When we come to the period of the Brahmanas we notice that the land known as Brahmavarta or the land of Kurus (Kuruksetra) lying between the rivers Sarasvati and Drsadvati and of the Pancalas, the land between the Ganga and the Yamuna (the Doab) assumes importance. The story of Videgha Mathava as narrated in the Sat. Br. clearly points to the expansion of the Vedic Indians to the east. The fire which came out from the mouth of Videgha Mathava started from the river Sarasvati in the west towards the river Sadanira (modern Gandak?) in the east. The Brahmanas had never crossed the river Sadanira before this incident. But now they settled down' even to the east of the river Sadanira. Similarly the legends of the sage Agastya who first burnt the Raksasas in the north and then destroyed those who had taken shelter in the southern direction, and his crossing the Vindhya mountain and settling down in the south point to the expansion of the Vedic Indians to the South. This gradual expansion of the Vedic people taking them to the newer and newer lands and their getting to know new animals and grains show clearly that they came to India from outside. If the Vedic Indians were the indigenous people there is no reason why they should not mention anything peculiar to the central, eastern and southern parts of India in the hymns of the Rgveda. REFERENCES 1. Adarsa is probably a mistake for adarsa=adarsana or vinasana which is identified with Kuruksetra. Identified with Allahabad in D.C. Sircar's Studies in the Geography of Ancient India, pp. 40, 172, 173, 241 but with Rajmahal Hills in Bihar in N.L. Dey's Geographical Dictionery. 3. "The Proto-Indoryans",JRAS. 1973, 123-140. 4. The information from her upto p. 44 of this article is based on Paul ThiMe's Die Heimat der indogermanischen Gemeinsprasche, Wiesbaden, 1954. 5. BSOS 8 (1936), p. 411 ff. 6. 1.4.1.10 7. Ibid. 1.4.1.14-15 8. Mbh. 13.140.7-13 9. Ibid. 3.102 Madhu Vidya/560

Page #586
REFERENCES : Adarsa is probably a mistake for adarsa = adarsana or vinasana which is identified with Kuruksetra. Identified with Allahabad in D. C. Sincar's Studies in the Geography of Ancient India, pp. 40, 172, 173, 241 but with Rajmahal Hills in Bihar in N. L. Dey's Geographical Dictionary. "The Proto-Indoaryans", JRAS, 1973, 123-140. The information from her = upto p. 44 of this article is based on Paul THIEME's Die Heimat der indogermanischen Gemeinsprache, Wiesbaden. 1954. BSOS, 8 (1936) p. 411 ff. 1.4.1.10 Ibid., 1.4.1.14-15 Mbh. 13.140.7-13 Ibid., 3.102 Madhu Vidya/561

Page #587
face. WHEN KALIDASA NODS When Kalidasa nods, he forgets to describe an important feature of a woman's Kalidasa has, on many occasions, found opportunities to describe the charms of the physical features of women. While doing so he has paid attention to practically every part of her body from hair on the head to the nails on her toes. Some of the physical features like the lower lip, the breasts, or the hips figure more frequently in his descriptions than the others, the neck (kantha, Kumara. 1.42) and loins (jangha. Kumara. 1. 35) figuring probably only once. It is therefore very surprising that Kalidasa on no occasion makes even so much as a passing reference to a woman's nose in any of his works. Kalidasa's references to the limbs of women may broadly be divided into four types: (1) A limb is compared with some object, e.g. a face is said to be like a lotus (Kumara. VIII. 19, 23, 58'), the eye-brows are said to be like waves (Megh. 110; Raghu. XVI.63), the thighs appear like the trunk of a banana tree (Raghu. VI. 35; Mala. III.10), etc.; (2) A particular form or a particular quality of a limb is stressed broadness of the eyes (Sak. III.5; Kumara, 1.46), thinness of the waist (Megh. 88; Mal. III.17), depth of the navel (Vikra. IV.7; Raghu. XVI.63), red colour of the fingers (Kumara. VII.76). etc.; (3) Describing the decoration of particular limb a leaf or a flower is placed on the ear (Kumara. III.62; IV.8), a mark of sandalwood or some other ointment is put on the forehead (Kumara. V.55; VIII. 40), the feet are dyed with the alaktaka juice (Vikra. V1.16; Kumara. V.68), etc.: and (4) Mention of the condition of a limb in certain situations: shoulders drooping due to fatigue (Sak, 1.29), breasts losing their compactness due to grief (Sak. III.9), body bending down a little due to the heaviness of breasts (Megh. 88; Kumara. III.54). Of the different types of references noted above, it is understandable if a reference to the nose is not made in the context of decorating or ornamenting the limbs. A flower cannot be placed on a nose, nor is the nose to be dyed with red colour. Scholars are of the opinion that nose-rings or other similar ornaments for a woman's nose were not common in India before 1000 A. D. But this does not mean that even the existence of the nose which forms a notable feature of the face should have been completely ignored. We find many occasions, especially in the Kumarasambhava, where a reference to Parvati's nose could have easily come up. Madhu Vidya/562 M A. MEHENDALE

Page #588
When the religious rites in connection with Parvati's marriage were in progress, she leaned forward, as instructed by the priest, to bring her face close to the smoke. At that time the smoke which went past her cheek and reached the ear appeared to the poet like a blue lotus adorning it (Kumara. VII.81). In this description Kalidasa has made a reference to Parvati's ears, and cheeks, but not to her nose. He says that the smoke was of 'desirable fragrance' (istagandha), and yet he is not inclined to say that Parvati's nose found full satisfaction due to the agreeableness of the smell, or that, she made a little extra effort with her nose to have some more fragrance. Parvati desired to have Siva for her husband. In order to achieve this objective she began rendering service to Siva when he was practising austerities. On one occasion, when Parvati was on her way to the place of Siva's austerities, the poet found that a certain bee, whose desire was whetted by the sweetness of Parvati's breath, began to hover round her bimba-like lower lip (Kumara. III.56). In this way.. the closest that Kalidasa took his bee to Parvati's face to have the sweet smell was her lower lip,but not a little beyond, to her nose, the very source of the fragrance! When Parvati realized that mere service to Siva was not enough to obtain him as her husband, she started practising severe austerities herself under the open sky. When rainy season arrived, the first drops of rain which fell on her head, in their downward journey, stopped for a while on her eye-lashes; next, they struck against the lower lip, and were shattered when they hit against the elevation of her breasts; from there they stumbled on the folds on her stomach, (as if on the speedbreakers) and finally reached her navel (Kumara. V. 24). Thus ended the long journey of the rain-drops from head to navel with four stations in between. The description is charming no doubt. But no drop took a slightly different course to trickle down from her nose to the lower lip! The bee in the Sakuntala also, like its counterpart in the Kumarasambhava, is inclined to neglect the nose of a woman. A certain bee dared go so close to Sakuntala's face as to touch her eyes, to hum sweetly in her ears, and even to kiss her lower lip (I. 23), but never turned its attention to her nose (to have the sweet smell which must have issued from her breath). When a heroin of Kalidasa, e.g. Malavika (IV.9) or Parvati (V.74), got angry her lips throbbed. In fact in the town where Parvati grew up, all women, when angered, twisted their eye-brows and threatened their lovers with fingers, besides expressing their emotion with throbbing lips (Kumara. VI.45). But neither any of Kalidasa's heroines, nor any other woman figuring in his works, got her nose red due to excitement! On one occasion Kalidasa has gone dangerously close to making a reference to his heroine's nose, but has, in fact, not. The Yaksa of the Meghaduta experienced an imaginary embrace of his wife during which he felt her warm breath. Here was Madhu Vidya/563

Page #589
an opportunity for the poet to say, if he wanted to, that their noses met. But the only thing he says is that their 'limbs met' during the close embrace (angenangam ... usnocchvasam samadhikatarocchvasina duravarti sarnkalpais tair visati ... Megha. 99 (108). The only place where Kalidasa has actually made a direct mention of the nose, not of a female but of a male, is when, while describing Siva practising austerities, he says that Siva concentrated his eyes on the nose (netraih ... laksyikrtaghranam ... Kumara III. 47; cf. Gita : saspreksya nasikagram svam VI. 13). Kalidasa's silence on a woman's (or a man's) nose as contributing to the charm of her appearance is strange in view of the fact that in India attention was paid to nose since the Vedic times. The Dasyus are referred to as 'noseless' (anas RV.V.29.10), apparently in disapproval. In Panini's times the nose having a bent was called avatita, avanata, or avabhrata Panini V. 2.31). A Person having the opposite of it, according to Patanjali, was a tunganasa (on P. I.3.2) 'having a prominent nose' or kalyananasiki or onasika (fem.) (on P. IV. 1.55) 'having an agreeable nose'. In the Mahabharata, Upamanyu describes Asvina deities as sunasau 'having handsome noses' and Yudhisthira is distinguished as carughona (1.180.20) 'having a pleasing nose'. The commentator Nilakantha assigns first place to the nasika 'nose' while giving his idea of the six limbs of Sairandhri (Draupadi) which were 'prominent' (sadunnata - Mbh. IV.8.10: Bom. Ed. IV.9.10). The epic poet does not fail to call attention to the beautiful nose'even of a fallen hero. Indrajit's falling head is said to be sunasa (Mbh. III. 273.23) as also of Aparajita, son of Dhstarastra (Mbh.V1.84.21), and Nila (Mbh. VII.30.25). In the light of this information it is not understandable why Kalidasa has totally ignored a woman's nose in his works. NOTES: 1. 2. These references are only illustrative and not exhaustive. A. S. Altekar, The Position of Women in Hindu Civilization, 1956, pp. 301-303; P.K. Gode, Studies in Indian Cultural History, 1960, Vol.II, P. 153. Madhu Vidya/564

Page #590
Some Reflections on The Raghuvamsa M. A. Mehendale 1 Raghuvamsa 4.80 reads as : tatraksobhyam yasorasim nivesyavaruroha sah: paulastyatulitasyadrer adadhana iva hriyam // Mallinatha interprets the stanza as follows: Raghu established a heap of his fame, which was unshakable, on the Himavant and descended from it. He thereby, as though, put to shame the Kailasa mountain which was moved by Ravana, by not marching on it. The idea is that since the Kailasa mountain was once moved by Ravana it had sustained defeat at his hands. Raghu did not want to invade it again since he thought it below his dignity to defeat someone who was defeated before. By the very act of non-invasion Raghu, as, though, put the Kailasa mountain to shame Kailasam agatvaiva pratini vrtta ity arthahl na hi Surah parena parajitam abhiyujyanta iui bhavahl): Shri M. R. Kale, in his edition of the Raghuvamsa, while elaborating the suggestion of Mallinatha remarks : "Raghu thought it beneath his dignity to advance against Kailasa as the mountain was once overcome by Ravana who was a Brahmana, and therefore was not worthy of his superior prowess as a Ksatriya" (p. 102). This interpretation is extremely unlikely. In support of his interpretation Millinatha should have shown that all the kings who were assailed and defeated by Raghu in this expedition were such as were never defeated before by any one else. But this was not the case. Moreover it is extremely unlikely that Raghu even otherwise would be intending to attack the Kailasa, the seat of Siva. In addition, the above interpretation does not take into account the very significant epithet aksobhya 'unassailable' used by Kalidasa to describe Raghu's yasorasi 'heap of fame'. What the poet really intends to say is that the heap of fame established by Raghu on the 146 11 3744614-Gift in Madhu Vidya/565

Page #591
Himavant was not shakable, while the Kailasa mountain was. This is how he put the mountain to shame. : Raghu. 4.72 reads as sasamsa tulyasattvanam sainyaghose' py asambhramam/guhasayanam sirihanam parivrtyavalokitam // 2. Mallinatha takes avalokitam to be the subject (kart) and asambhramam to be the object in the above sentence. He takes sams- to mean to tell' and hence is required to supply another object 'army' (Safarisa kathayamasa/ 'sainyebby' ity arthal labhyate). Accordingly the stanza is understood by interpreters who follow Mallinatha to convey the following meaning: In spite of the din created by Raghu's army the lions' looking back by turning their necks conveyed to (the army of Raghu) the absence of any agitation of their minds (antahksobhavirahitatvam). This interpretation, though grammatically possible, is extremely unlikely in the context. For, in almost all the stanzas starting from 28, either Raghu himself is the subject or something is told about him. In very few stanzas where this is not the case something is reported either about his soldiers (30, 42, 53, 55, 65, 74), or about the elephants in his army (47, 48. 57, 76), or his horses (56, 67). Hence it is not likely that in an isolated stanza the poet would choose to make the lions the subject and not Raghu. This is avoided if we understand that it is Raghu who is the subject of Sasamsa and that Sams means 'to praise, to admire'. The stanza would then mean that Raghu praised the undisturbed looking back of the lions in spite of the din created by his army. According to the infromation supplied in some of the editions of the text, the interpretation suggested above is, in fact, adopted by some commentators. One of them, Dinakara, for example says: prakaranat sa ity adhyahriyate/sa Raghuh... asambhramarh sadhairyar yatha tatha parivrtyavalokanam sasamsastausit. This interpretation is definitely to be preferred to the one given by Mallinatha for the reason stated above. 3. Two possible variant readings in canto 2. / / amRtamandAkinI // Madhu Vidya/566 147

Page #592
(i) one of these variants is suggested by the commentary of Mallinatha. Its presence is felt in stanza 53. The first line, which is intended by Kalidasa to explain the formation of the word ksatra, reads as : ksatat kila trayata ity udagrah ksatrasya sabdah bhuvanesu rudhah. While commenting on this line Mallinatha starts with the derivation of the root noun ksat from the root ksan '10 injure' (himsayam) and then adds, to explain the compound, "one who protects from destruction is ksatra'' (ksato nasat trayata iu ksatrah). It is thus clear that Mallinatha looks upon the two constitutents of the compound to be ksat + tra and not ksata + tra as implied by the reading of the stanza given above. Mallinatha in his commentary is completely silent about the formation of the word ksata. Hence we must say that he either overlooked the word ksata used by Kalidasa, or had before him a different reading of the stanza, viz. ksatah kila trayata ... That in his commentary as printed today the words ksatad ity adina and ksatat trayate also occur could be explained as an addition made by some later scribe who wanted to bring the commentary in line with the usually accepted text. That this in fact must have been the case is made clear by the fact that the scribe forgot to carry out this correction' a few lines later where we still read ksatas tranam (and not ksatat tranam) akurvatah, which is in conformity with the suggested reading. (ii) Another possibility of a variant reading is felt in stanza 59 of the second canto. The stanza reads as follows : tatheti gam uktavate dilipah... haraye svadeham upanayat ... This is naturally understood as: "To the lion who had said the words 'be it so' Dilipa offered his body". The word go has to be taken to mean 'speech, words'. But in the principal incident that is described in this canto we hear what is happening to Nandini, and in that description we come across, besides of course the word dhenu, the word go to refer to her (29, 48, 61). It is therefore very unlikely that Kalidasa would choose to use the word go in this description in another sence 'speech'. Moreover, there is 148 11 3040ifesti Madhu Vidya/567

Page #593
no necessity to add the words gam uktavate after tatheti. This is shown by 5.59 which reads as tathely upasprsya payah pavitram ... sah ... astramantram jagraha. Here. after talheu some other act viz. sipping the water done by the person who said tatheti is narrated. This is also what we expect in our stanza to take place. Hence it is suggested that the original reading possibly was tatheti gar muktavate dilipah... which would mean "to the lion who having said 'be it so' had released the cow...". The lion, having said tatha proceeded to do something else, viz. releasing the cow. This reading also gives better sense. The lion who had agreed to Dilipa's suggestion by saying 'be it so had immediately acted on it and released (muktavate) the cow. Now it was Dilipa's turn to fulfil his obligation which he also did by instantly offering his body to the lion. 11 3Taifahit 11 149 Madhu Vidya/568

Page #594
(B) REVIEWS Gujarati par Arabi Pharasini Asar : (In Gujarati. The influence of the Arabic and Persian Languages on Gujarati), Part I, by C. R. NAIK, Gujarat Vidya Sabha, Ahmedabad, 1954, pp. 12; 438. Price Rs. 4/8. The modern Indian languages have been greatly influenced, particularly in vocabulary, by the Arabic and Persian languages due to the continuous Islamic rule in this country for several centuries. However, the number of words in the differ. ent Indian languages which could be traced to Perso-Arabic influence varies in different regions owing to various circumstances. In the present work Dr. NAIK makes an attempt to assess the degree of the influence exerted by these forcign languages on Gujarati. The Gujarat Vidya Sabha may be congratulated for having brought out this publication so valuable for the cultural study of Gujarat. In the first three chapters of the book, the author describes the history, language and literature of Iran, and the arrival of foreigners in Gujarat-the topics which supply the background to the main subject. In the fourth chapter he gives his evaluation of the Perso-Arabic element in the language of Gujarat. While doing so he has not merely given the lists of borrowed words in Gujarati. Instead, he has classified these words under suitable heads like words referring to Administration, Law, Revenue, Medicine, Agriculture, etc., and preceded each one of these lists with introductory remarks discussing at times the causes which led to these borrowings. A glance at these lists will enable the reader to appreciate the wide influence exert. ed by the foreigners on the life and culture of the people of Gujarat. In the last chapter we find the treatment of some topics of morphological interest such as the foreign influence on the noun-formation in Gujarati, the use of foreign indeclinables and suffixes. There we also get a discussion of the semantic aspect in so far as the 1. In course of time many of these words became difficult for understanding and attempts were made to compile lexicons of Persian words with Sanskrit equivalents. Efforts in the other direction viz. to avoid the use of foreign words in a language and replace them by old indigenous words were also not unknown. For a description of these Lexicons, cf. Infiltration of Persian words in Indian Languages by Dr. M. R. MAJUMDAR, Journal of Oriental Institute, Baroda, 1.89-92. Madhu Vidya/569

Page #595
240 REVIEWS author cites instances to show how the original meaning of some foreign words was changed in modern Gujarati. Quite interetsing is the collection of some of the Gujarati sayings which are either translations of foreign sayings or which show foreign words in their present form. A few remarks may be made. For a proper and thorough treatment of the subject it was necessary to make a distinction between such foreign words which have completely replaced the words proper to the language and such as are used alongside the indigenous words. It does not again suffice to list together borrowed words of varied frequency in their use. In fact on a proper study it will be observed that in some cases a foreign word is used in a limited context and hence is used less often, whereas the local word finds wider employment. To take only a few instances from amongst those that have been listed together as referring to the parts of the body (pp. 277-278) we may consider such pairs : kadam : pag? ; kamar : ked; khun : lohi ; casm : akh; jaban : jibh; badan : sarir. It is by no means the case that such pairs are restricted only to the words referring to the parts of the body. They can as well be formed with words given in the other lists by the author. Then again in certain cases it may be found that a borrowed word has found inclusion in the language vocabulary not because it expresses a foreign idea, but because the local word has changed its meaning or has come to be used in a narrow sense in the course of its development. Thus, for example, the word for garden bag (p. 243) may have found acceptance in the language, as thic derivative of the inherited word ujani was used in the restricted sense of a picnic. Unless such careful distinction is made in the borrowed words, a casual reader is likely to overestimate the foreign influence on the language and culture of Gujarat. The author's citing of jungh (the printed text gives jung which is not given in the Jodanikosa) by the side of Skt. janu (p. 312) is not correct the former has to be traced to Skt. jangha. On p. 321, the author cites rang (a) as common to all the three languages, viz. Persian, Sanskrit, and Gujarati; but this is in contradiction to the author's earlier observation in footnote 2 on p. 272 where he asserts that rang colour' is a Persian word, and that Sanskrit ranga is altogether a differ. ent word. The absence of a word-index is a great handicap. The utility of works such as the present one would be greatly increased when it is supplied with a good index. As mentioned on the cover page, the present volume is the first part ; but the introduction does not make clear in how many parts the subject is going to be handled, and what we may expect to read in the subsequent parts. M. A. MEHENDALE 2. The second word in each of these pairs has not been cited by the author ; he only gives. the first one as a borrowing. Madhu Vidya/570

Page #596
The Etymologies of Yaska. By SITESITWAR V'ARM., with the assistance of Bum DES SHLISTRI, Vishveshivaranand Indological Series No. 5. Hoshiarpur, 1953. pp. xii. 248. It is now generally recoguises that for a proper understanding of the lyvedla. it is also necessary to take into account the indigenous attempts towards Vedie interpretation. Though the Virukta of Yaska is one of the important contribut. tions to Vedic exegesis which old Indian tradition has preserved for us, it was once customary to brush aside the testimony of Yaska as fanciful and ridiculous. Prof. VARM. has therefore rendered a signal service to the cause of Vedic studies by undertaking it very painstaking and scholarly examination of the etymologies of Yaska and by telling us which of these ctymologies may be regarded as ik'ceptable and which not from the stand-point of comparative philology. The present work is not a mere restatement of Vaska's etymologies, not a hasty decision on them. but a careful cvaluation of Yaska's vicws in the light of Indo-European parallels. All things taken together, the author gives his judgment that the Nirukta may be called a treatise on primitive Vedie etymology. When one reads the etymologies of Yaska one cannot lielp feeling that the while science is reduced to the following a given number of words to be necessarily derived from a given number of verbs on the basis of the similarity of meaning or that of the letters. The bad results which arise out of such mechanisin have been well denionstrated by the author on p. 19 ft. The second chapter of the book, of which these piges for a part, show in the first instance that laska's etvinologies had a sound! scientific basis. He (i.f. Yaska) has evinced a considerable grasp of phonological principles which formed the background of the Vedic language." But the analysis in thus chapter also reveals that * on the whole, the etymologies of Yaska are of primitive type, bearing the stamp of the age in which he lived and, consequently. showing no trace of that comparative historical outlook which crowns the achievements of nodern linguistics." Madhu Vidya/571

Page #597
REVIEWS 308 The third chapter forms the basis of the whole study as it goes on to classify the etymologies of laska under various types such as (i) etymologies based on contamination. (ii) etymologies betraying mechanical nature, (iii) etymologies indicating Yaska' poverty of imagination, etc. This whole classification is based on a critical and Systenatic examination of the data, and it speaks volumes for the scientific outlook and the learning of the author. This chapter shows that roughly 1/6 of the etymologies of Yaska are entirely acceptable, 2/3 are primitive, and 13 obscure. The fourth chapter on - Yaska and the Padakaras " was undertaken at the suggestion of Prof. Vishva Bandhu SHASTRI, and written on the basis of the material collected by Prof. Bluni Dev SHASTRI. We are extremely thankful to Prof. SILASTRI for his brilliant suggestion, for this chapter gives abundant material for further research which will lead us nearer to the correct interpretation of the Rgveda. The chapter first attempts to give the conventions of the Padakara observed in the analysis of words and then a discussion of the analysis made by Yaska in the light of these conventions. Yaska knew Padapatha no doubt, but he was not a blind follower of the views of its author as can be seen from the few occasions on which he disagrees with the Padapatha. The book includes some very useful indices like an index of words etymologiseci by Yaska, an index of the basic vocabulary of Yaska's ctymologies, and an index of comparative linguistic vocabulary. While trying to establish that some of the etymologies of Yaska show that he was far in advance of his times (pp. 4-6), the author gives on p. 5 the instance of kantaka which Yaska derives from v krnt (Nirukta 9.32. This etymology is not given in Rajwade's edition of the Nirukta with Marathi translation). This deriva. tion shows, as it were. that Yaska was aware of the famous law of cerebralisation in OIA. But a little later while deinonstrating how Yaska's etmyologies are primitive (pp. 19-22), the author cites on p. 20 the instance (landa which Yaska <Page #598
30: REVIEWS three etimologies are mentioned in the Nirukta 6.13. One of these, and perhaps the one preferred by Vaska, says that niisatya are so-called as they were born of the pose (nisikuprubhavur babhavatuh). The author notes that according to Wust the word is to be derived from vnas to associate, Indo-Eur. nes to return, Gr. neomai I return home. But . LOMMEL has recently shown in an article contributed to Schubring Comniemoration Volumel (pp. 29-31) that the etymology referring to the birth from nose is based on a myth given in the Brhaddevata and the later Puranas, and that the use of the suffix -tyo is attested since Indo-European times for the formation of words which show origin or descent. In chapter IV, section . are collected together words not analysed by the Padapatha. We may as well noto that damunas has not been analysed by the Padapatha, as it also leaves without analysis vispatni (like vispiti which is recorded on p. 172). M. A. MEHENDALE Madhu Vidya/573

Page #599
REVIEWS 61 Prakrta Bhasa (In Hindi); A series of three lectures delivered under the auspices of the Parsvanatha Vidyasrama at the Banaras Hindu University, by Prabodh Bechardas PANDIT. Banaras, 1954, pp. 58, Price Rs. 1-8-0. The organisers of the Shri Parsvanatha Vidyasrama may be congratulated for having arranged a series of lectures on a topic connected with Jaina religion and Praksta languages as one of their various academic undertakings. They certainly acted properly in securing the known linguist and Prakpta scholar of Gujarat, Dr. P. B. PANDIT, to deliver the first series of the lectures on the middle Indian dialects. The importance of a scientific study of these dialects now needs hardly any emphasis. For, without such a study, as the publisher observes in his introductory remarks, a proper linguistic assessment of the modern IndoAryan languages will not be possible. In his first lecture the lecturer does well to give the historical background of the Prakstas by showing the relationship which Samskrta, as the representative of the Old Indo-Aryan period, has with the other IndoEuropean languages. In the second lecture we get the description of the several middle Indian dialects in their geographical distribution. At one place the lecturer has rightly stressed the point that it is not proper to differentiate between languages if they differ from one another in only one or two phonetic features. The last lecture is devoted to narrate the Madhu Vidya/574

Page #600
52 REVIEWS further development of the middle Indian stage. In this regard the lecturer puts forth two important ideas. In the first place the literature available today does not make it possible to get a clear picture of the further development of the Prakstas in the various geographical areas. If the literature in those days had not developed in isolation from the people, if it had treated of the subjects directly concerning the people, we would have been fortunate today to receive specimens of many Prakrtas which must have been in use in those days in different parts of the country. The second point stressed by the lecturer is that the phonetic changes witnessed in the development of MIA (and for that matter also of OIA) are not absolute innovations. Such changes had already started in the language as a matter of natural course. The contact with the speakers of the other linguistic families in India may have only helped the rapid spread of these changes. It is, therefore, inaccurate to say that the middle Indic phonetic changes (and the morphological changes arising out of them) were due to the ignorance or the laziness of the people who spoke them. Before closing the series, the lecturer has expressed hope that it would be possible for the Parsvanatha Vidyasrama and other similar institutions to take up the publication of good editions of Jaina canonical literature. Such editions should be furnished with an accurate description of the language and a vocabulary which woull ultimately lead to the compilation of a broad-based Praksta grammar. On p. 21, the Shahbazgarhi and the Mansehra versions of the Asokan rock edicts are said to belong to the north-east, which is obviously a mistake due to oversight. On the same page, these versions are twice correctly referred to as belonging to the north-west. The reasons given on pp. 7-8, while discussing as to why the works belonging to the Vedic period were not written down, may not represent the correct state of affairs. The explanation may not be sought in the limited knowledge of the script or in the attempt of the Brahmanas to monopolise the knowledge of the intricate Vedic ritual, but into the fact that written texts easily lead to variations, a thing which had to be scrupulously avoided in order to ensure the efficacy of the sacrifices. In the second lecture, the lecturer has attempted a geographical description of the Prakrtas on the basis of the Asokan inscriptions. This may be done, but only after a critical examination of the data enabling us to fix what really constitutes the eastern or the western, etc. In this respect it may be permitted to refer to an article published in the Journal of the Oriental Institute, Baroda, Vol. I, pp. 240-44 showing that certain exceptional features occurring in the Dhauli and the Jaugada separate edicts are not eastern but north-western. --M. A. MEHENDALE. Madhu Vidya/575

Page #601
Beobachtungen uber die Sprache des buddhistischen Urkanons Von Heinrich LUEDERS. Aus dem Nachlass herausgegeben von Ernst WALDSCHMIDT. Abhandlungen der deutschen Akademie der Wissenschaften zu Berlin. Klasse fur Sprachen, Literatur und Kunst, Jahrgang 1952, Nr. 10, Akademie-Verlag, Berlin, 1954, pp. 196, Price DM 19/ It is always a matter of great pleasure to get to read a book by LUEDERS. It is indeed to be regretted that a part of what LUEDERS wrote could not be published during his life time. The result is that some of his complete works like the edition of the Udanavarga have been altogether destroyed and even his other works which are now being published by his pupils were damaged in part. . The editor's work for the present volume must have been particularly difficult as the work was not finalised by LUEDERS himself. When the Nachlass came to the editor's hands, the pages were not numbered, the book not divided into chapters, and the beginning of the manuscript completely lost. In spite of these difficulties, the editor has done his work exceedingly well and he deserves congratulations and thanks of all lovers of Indological research. He has not simply put the material of the Nachlass into proper shape; he has also made the book more valuable by referring in footnotes to such paragraphs from GEIGER's Pali Litteratur und Sprache as give additional material which was not found in LUEDERS' manuscript. He also gives suitable references to EDGERTON's Buddhist Hybrid Sanskrit, Grammar und Dictionary. In his Zum Geleit the editor elucidates in short the view held by LUEDERS as regards the language of the supposed original canon of Budd'hism. On the basis of the comparison of the Udanavarga verses from the Turfan fragments with their parallels in the Pali version LUEDERS, had come to the conclusion quite early (see references p. 7, footnote 4) that both the Sanskrit and the Pali texts go back to an original which was composed in the eastern dialect. There has been some controversy regarding this view. Recently EDGERTON in his Buddhist Hybrid Sanskrit Grammar has expressed the opinion that we cannot at all speak of an 'original language of Buddhism. Though LUEDERS had made known his opinion early he had not then published the detailed investigation of the problem which we find in the present work. This collected evidence, well substantiated on facts, should go some way to persuade the opponents of LUEDERS' theory to reconsider their opinion, The book has been divided into two major parts to which is added a small Anhang discussing some verses from the Udanavarga. The first Madhu Vidya/576

Page #602
54 REVIEWS part enumerates the various instances from Pali in which it is possible to discern the eastern influence, and the second part furnishes the picture of the phonology and the morphology of the language of the original As I propose to give a detailed review of the second part at some other place, the present account is restricted only to the first part of the book. It is well known that the most striking elements of the eastern language are nom. sg. ending -e and I for r. In the first part LUEDERS has brought together such instances from Pali where he finds the influence of these two eastern characteristics and successfully shows that the passages containing these instances yield good sense only when we suppose that they are based on an original eastern version. The following remarks are therefore offered not in a spirit to controvert the main conclusion of LUEDERS but to elucidate certain points in the material handled by him. 1. 10, p. 17: In Jataka 388,1 we have a word dani in line 1. For this the Simhalese manuscripts read dane and LUEDERS is right in thinking that this is the correct reading. In the translation of the verse, however, LUEDERS renders dani by 'jetzt'. Pali dani for eastern dane shows a change of e>-i probably indicating, as suggested by LUEDERS, short pronunciation of -e. This change is witnessed also in the northwestern inscriptions of Asoka. cf. duvi for duve (dve), ami for amne (anyah), rajani for rajane (rajanah). We also get a similar instance from a version of the minor rock edict, probably under north-western influence. cf. upeti for upete (upetah) in the Kopbal version. For Pali hemantagimhisu for -gimhesu (cited on p. 17, f.n. 3) we have an exact parallel in pavatisu for pavatesu in the Rupnath version of the minor edict.1 2. SS14, pp. 19-20 LUEDERS has shown convincingly that the Mahavastu (II. 238. 17) translator had mistaken anumatte hi of the original eastern version as one word (anumattehi instr. pl.) and misconstrued it as adj. to punnehi. He, therefore, translated it into Sanskrit as anumatraih punyaih. The whole line then read as anumatraih punyaih artho mahyam na vidyati. LUEDERS further observes that the translator while doing so did not pay attention to the fact that in putting anumatraih punyaih (6 syllables) for anumatte hi punnehi (8 syllables) he had shortened the first quarter of the line by two syllables. This is true so far as the written form goes. But perhaps this fact may show that the visarga preceded by the diph 1. For a similar tendency to change ei in the Niya Prakrit and the Kharosthi Dhammapada see BURROW, The Language of the Kharosthi Documents, SS 1. Madhu Vidya/577

Page #603
REVIEWS 55 thong was pronounced as hi (aih as -aihi) exactly as it is done today in single words at least in Maharastra. Hence even though the two words give six syllables in the written form, they were probably pronounced as anumatraihi punyaihi even in a metrical line in order to get eight syllables. This again seems to have been the reason why no sandhi was made between punyaih and artho (=punyairartho), for that would reduce the pada by one syllable. LUEDERS is certainly right when he says that in the Lalitavistara version (261.18) there occurred loss of a syllable when punnehi was translated by punyaih because there sandhi was effected between the visarga and the following a. This loss of a syllable was then made good by the addition of me. cf. anumatram hi me punyair artho Mara na vidyate. 3. $15, pp. 20-21 : Among the cases where the eastern nom. sg. -e was mistaken by the Pali translator as loc. sg., LUEDERS lists kule bhadde which occurs thrice in Jataka 531.50-52. On p. 21 he observes that in verse 51 the interpretation of kule bhadde as loc. sg. seems justified and that this gave occasion to look upon the two words as loc. sg. in form also in verses 50 and 52 where, in the opinion of LUEDERS, they are not justifier and where they have to be interpreted as. nom. sg. It is a little difficult to believe that the interpretation which is regarded as justified for the middle stanza (51) should have influenced the interpretation of the stanza previous to it (50). Hence it may be pointed out that the interpretation of the two words as loc. sg. also seems justified in the first verse 50 (p. 21) which begins with the word yathassu (for yatthassu? FAUSBOELL gives yatth' assu). This word meaning 'where can be more easily connected with kule bhadde as loc. sg. than with nom. sg. (yattha khattiyanam kule bhadde bheri nadati etc.). It is even possible that this yattha coming at the beginning of the first (50) of these three verses is also to be construed with the following two verses (51, 52). Thus when kule bhadde is justified as loc. sg. in the first two verses 50 and 51, it is natural to anticipate that it is justified also in the third verse 52. And this in fact actually seems to be the case. With the supposition of yattha from verse 50, kuld bhadde as loc. sg. give good sense here also. The mistaken nom. singulars in stanza 52 are, therefore, not kule bhadde, but abhirude and -nikunjite occurring in line one of that verse because as loc. sg. they make no sense. They are thus to be interpreted not as participles used adjectively qualifying kule, but as substantives. What is expressed in the other two verses with the use of verb forms nadati, nikunjati, himsati, and uparodati is expressed here with the help of verbal derivatives. abhiruta and nikunjita are not noted in the PTS Dictionary as substantives, but it gives ruta (and ruda) as a noun. In Sanskrit, according to Madhu Vidya/578

Page #604
56 REVIEWS the PW, abhiruta and kujita are attested as nouns since Ramayana. The verse is, therefore, to be translated as-"(where) in the happy house of the Ksatriyas there is the noise of peacocks and kraunea birds and the cries of cuckoos-what is happier than that?"" 4. SS17, pp. 21-22: LUEDERS very ingeneously points out that in the Sutta-Nipata 453 (Theragatha 1229) atthe and dhamme occur as misunderstood loc. singulars. They should have been attho and dhammo (both nom. sg.) respectively. However, his reconstruction (p. 22) of the line from the parallel passage in the Udanavarga 8.14 as found in the old C, manuscript of Berlin is not quite convincing. Actually the line has been read as [sa]tya........rm. ca sam...pra[t]is [th]i.. [m]/ LUEDERS suggests the reconstruction of the line assatyam arthe ca dharme ea santa ahuh pratisthitam/ where satyam and pratisthitam appear as fem. This makes it necessary to supply vacam (going with satyam) from vaca occurring in the first half of the verse (saccam ve amata vaca esa dhammo sanantano). But would it not be more natural to reconstruct the line as satyam arthe ca dharme ca.... pratisthitam (neut.) since the first word, so far as it is legible, does not show the length of the vowel in the second syllable? The Tibetan translation noted by LUEDERS on p. 22, f.n. 2 also points to satyam....pratisthitam. Obviously LUEDERS suggested satyam (vacam) pratisthitam (fem.) because in the ms. F he found vacam, cf. satyam arthe ca dharme ca vacam ahuh pratisthitam as also in ms. x [ca] dharme ca vacam [a]-. But in these manuscripts vacam has nothing to do with satyam; it is to be construed, together with pratisthitam, with ahuh which follows it im mediately. Thus satyam arthe ca dharme ca (iti) pratisthitam vacan (santa) ahuh would give good sense. With vacam ahuh we may compare vacam bhaseyya and bhasati vacam, Sutta-Nipata 451 and 454; or sisyam aha sthitam parave Bharadvajam idam vacah, Ramayana 1.2.20. 5. 21, pp. 24-27 Under cases of the misunderstood loc. sg. LUEDERS notes bhumivaddhano (Jataka 538.51), nom. sg., which in his opinion should have been bhumivaddhane, loc. sg. (The line runs asnihato nuna me putto pathavya bhumivaddhano). LUEDERS starts from Pali vaddheti which, as rightly pointed out by him, is used about 'Aufhaufen eines Breis in einer Schussel'. Therefore, bhumivaddhana, according to him, means 'Erdanhaufung' or 'Grab'. This he connects with nihato which in his opinion is either a wrong translation of nighade or a corruption of nikhato. LUEDERS therefore translates the line as-'sicherlich ist mein Sohn in der Erde in dem Grabe eingegraben'. Madhu Vidya/579

Page #605
REVIEWS 57 This interpretation is not quite convincing. In the first instance, it compels us to presume that nihato is a wrong translation of nighade or a corruption of nikhato. Secondly, on this presumption nihato through nikhato is to be derived from niv khana-" which is not very natural, because it seems difficult to separate nihato from nihantvanas occurring in the first half of the verse (ayan so sarathi eti nihantvana mam' atrajan) and LUEDERS does not seem to object to its derivation from nivhana. Of course it can as well be said that nihantvana is also a corruption of nikhantvana, but this LUEDERS does not do. Thirdly, in this construction nihato and bhumivaddhane (actual text vaddhano) which are brought together stand very much apart. It may, therefore, be suggested that Vvaddha is here used in its usual sense 'to increase'. vaddhana can then mean both increasing or causing to increase (Pa. 3.1.134), in the present instance the latter. bhumivaddhana, lit. 'one who increases the earth', may have been used as an euphimistic expression for a person buried in ground as thereby he may be regarded to cause the earth of the ground to increase. This may have been specially said of a person who was regarded as deaf, dumb, and a fool, and otherwise of no use to the people. The expression pathavivaodhanakakamma which comes at the end of the instructions given by the king to the charioteer can also have similar meaning, the act of increasing the earth (by burying the prince in the ground)'. Thus bhumivaddhano as nom. sg. and adj. to putto seems quite satisfactory in the present verse. 6. $37, pp. 38-39 : $ 37-8 86 are devoted to showing the cases where l in Pali may be traced to the eastern influence. In the beginning, however, LUEDERS examines those instances which seem to be doubtful and where forms with I could be regarded as regular Pali forms. To begin with in $ 37 he takes the case of gilati: girati 'to swallow about which he concludes-"Wie im Sk. konnen aber auch im Pali gilati und girati nebeneinander bestanden haben." The instances examined by LUEDERS in this article, however, seem to show that the two forms were not promiscuously used. In fact they show that the two forms were used to express opposite meanings. Thus 2. Forms of vkhana occur many times in Gathas 3, 4, etc., and also in the prose narration. Hence it may be tempting to connect nihato with nikhana. It is, however, a little difficult to say why exactly here the corruption took place when forms of khana have been well preserved at other places in this Jataka. But even if we regard nihato as a corruption of nikhato or a wrong translation of nighade (cf. nighannasi in G. 8) it is not necessary to take bhumivaddhano as a mistake for -vaddhane. With the meaning of bhumivaddhana here proposed the line can be translated as 'Indeed is my son, the increaser of the earth, buried in the ground'. 3. nihantvana occurs also in Gatha 52 and nihannamano in Gathas, 54, 55. Madhu Vidya/580

Page #606
58 REVIEWS Vgila (also when prefixed with o, ni, pari) was used to express 'to swallow', while Vgira (with the prefix ut) gave the opposite sense 'to vomit, to spit out, etc.'. Therefore the proper Pali word for 'to vomit' would be Vuggira (with r) cited by LUEDERS from the Udana II, 6, etc. (p. 39). When against this we get Vuggila (with 1) 'to vomit out' in Pali (cf. Majjhimanikaya 1,393, Samyuttanikaya IV. 323, and Jataka 436.4), then these may not be regarded as doubtful cases. They can be definitely looked upon as coming in Pali due to eastern influence.* According to the available dictionaries ud Vgil never occurs even in Sanskrit in the sense 'to vomit out, to spit out, eject, etc.', to express which we have ud Vgir. However, udgala in the sense of 'vomiting out' has been noted in the SCHMIDT's Nachtrage as occurring in the late work Yasastilakacampu of Somadeva 1.94 and 3.180-181 (p. 434, line 1). But in both these cases the word occurs with l for the sake of alliteration. Cf. kanthantah-pravilagna-salya-sakal-odgala-skhalat-kuksayah and ghatadasinan hi vadanasaurabham svamitambulodgalang na saubhagyabalat. (This alliteration in end syllables is continued also in the following phrases, cf. prasunavanasamsargan na nisargat, brhadbhanubhavan na svabhavat, etc.). From the text, the word has also gone to the commenttary while explaining the text-word vigirna 1.82. cf. tena vigirna udgirnii udgalita ye phenah. . . . . . Yamena purvan mrtakani bhaksitani pascod udgalah krta ity arthah. The word udgala 'vomiting' is further noted in the Skt. lexicon Vaijayanti 183.126 (udgalo vamathudgarau) only. 7. $ 47, pp. 47-48 : LUEDERS discusses here elamuga in a group of allied words. The citations from the Pali literature make it clear that elamuga means 'an idiot, a fool'.? As LUEDERS points out the word is not attested in Sanskrit literatures and the meanings given to edamuka by the Sanskrit Lexicons do not help us at all in explaining the word. LUEDERS is 4. That seems to be the reason why the Burmese manuscripts show uggiri for uggili in Jataka 436.4, mentioned by LUEDERS on p. 39, f.n. 3. 5. It is also possible that udgala was being mixed up with udgal-'to trickle out, drop out, etc.'That would also explain uggaliadabbhakavala mia (udgalitadarbhakavala mrgah) in the Sakuntala IV. 12. 6. SCHMIDT's equating ridgala with 'Geniessen von Ausgespieenem' with a question mark is obviously wrong. * 7. The meaning 'deaf and dumb' given by Dutoit in translating Jataka 546 (IT) 69 and 20 and seemingly supported by LUEDERS in f.n. 1, p. 48 by referring to Jataka 546 (II) 13 is not supported as in the last instance the reading is not certain. 8. It may, however, be pointed out that anelamuka, as mentioned in the Nachtrage, occurs in the sense 'taubstumm' in the Srikanthacarita 6.10, and probably in the sense 'stumpfsinning' in the Harsacarita (1986) 28.2. 9. To the references given by LUEDERS may be added Rabhasa cited by the commentator on Amara 3.1.38, who also gives the meaning satha for edamuka. Madhu Vidya/581

Page #607
REVIEWS 59 therefore, favourably disposed towards the suggestion made by BOEhrLINGK-ROTH in PW that Skt. edamuka contains eda 'sheep' and therefore it means 'dumb like a sheep'. LUEDERS further observes that the meaning 'idiot' attested in Pali can be easily had from 'dumb like a sheep'. That is true; but being 'dumb' is not in any way a special characteristic of sheep and hence, while explaining the word it would not be sound to proceed from the meaning 'dumb like a sheep'. In fact the word seems to be definitely not known to old Sanskrit literature. Starting, therefore, from the surer meaning 'idiot' attested in Pali literature it may be suggested that the latter part muga of the compound elamuga may not have anything to do with muka 'dumb', but that it may go back to Sk. murkhi 'fool': edamurkha > elamukkha > elamukka > elamuka > elamuga. The word would then literally mean 'fool or simpleton like a sheep'. In that case edamuka would appear as wrong Sanskritisation of Pali clamuga. So far as the form muka < murkhal' is concerned, instances of loss aspiration are witnessed both in Pali and Prakrit, GEIGER 62 (also $ 40) cites ikka < sksa, Takkasila < Taksasila, etc.; PISCHEL S 213 gives sarkala Amg, JM, S< srnkhala, but sankhala M, S, sankhala Mg, etc, and $ 302 sakkuli and samkuli < saskuli; sukka Amg, JM < suska, but sukkhu M, Amg, s, suska Mg, etc. The change k > g occurs in Pali, and it has been noted by LUEDERS as an eastern characteristic in $ 87 and f.n. 1, GEIGER $ 38. He also considers in SS 122-8 132 (see also p. 102, f.n. 2) certain cases of hyper-Palism in which k appears for g. Thus from the formal point of view there is nothing objectionable in deriving ejamuga from eNamurkha.!! Under ela in elamuga PTS Dictionary observes as follows:-"A rather strange use and explanation of elamuga (with reference to a snake "spitting") we find at J III.347 where it is explained as "ela-paggharantena mukhena elamugam", i.e., called elamuga because of the saliva (foam ?) dripping from its mouth, v.l. elamukha." This explanation of the commentator is obviously mistaken. ela does not mean saliva or foam, and secondly this sounds a very strange description of a serpent. In the Gathis the serpent is otherwise described only as kanha, uggateja, 10. LUEDERS does not discuss the loss of aspiration as a regional characteristic. It may be pointed out that in the inscriptions of Asoka the loss of aspiration is noted in the Girnar version in the forms of the root tisth-, cf. Vtista G, but titha or citha in other versions; also cf. sesta (srestha) G, but sretha Shah; Man., setha Kal; gharasta (grhastha) G, but gahatha Man., Kal., grahatha Shah, gihitha Top. The form idha in the Girnar version (and once in the Dhauli) as against hida of the re. maining versions is the case of an old preservation. 11. SS and Mg give mukkha (PISCHEL S 287), and murukkha is noted for the eastern language by Markandeya (SS 139) and Sauraseni ( 131). Madhu Vidya/582

Page #608
60 REVIEWS and asivisa. The commentator was forced to explain away somehow elamugan as he took it to be an adjective of uragam. The instances col. lected by LUEDERS in article $ 47, however, show that elamuga at all places means 'a fool' and that it is used to characterise human beings and not animals. The present occurrence cannot be regarded as a solitary exception to the general use and hence it seems very likely that originally the word stood in the Gatha as a vocative elamuga, an adjective of Brahmana. Addressing the Brahmana as elamuga 'a fool' agrees well with the fact that in the Jataka story the mind of the Brahmana was-worried about the possible cause of his own death or that of his wife when he was himself unknowingly carrying a serpent in his bag. In Gatha 1 he is, therefore, thus described-vibbhantacitto kupitindriyo si, DUTOIT--"Verwirrt ist dir der Geist, angstlich der Sinn" (whence also the commencement of the Jataka vibbhantacitto ti). The Gatha in which elamugam occurs reads as : adaya dandam parisumbha bhastani, pass' elamugan uragam dijivham/ chind' ajja kamkham vicikicchitani, bhujangamam passa, pamunca bhastan// DUTOIT, "Nimm einen Stock und schlage auf den Ranzen; du siehst die Schlange geifernd, doppelzungig. Gib heute auf die angstlichkeit, den Zweifel; sieh auf die Schlange, offne deinen Ranzen !" But as suggested above elamugan should be read as elamuga 'oh fool' referring to the Brahmana. elamuga (voc.) was subjected to the corruption elamugan (acc.) possibly because the metre required the fifth syllable to be long and also because the two words uragan dijivhan following elamuga were acc. sg. LUEDERS himself has put forward ($ 125, p. 103) a somewhat similar suggestion while explaining ajakaran occurring in Jataka 427.2. He suggests that ajakaram may be regarded as a metrical emendation for ajakara- (just as in the present case it has been suggested to treat the reading elamugan as a metrical error for elamuga). But the commentator's explanation interests us for another reason. How did he come to explain the latter part of the compound elamuga with the use of the word mukha? For that speaks the v.l. elamukha which is the same as elamukha. Shall we therefore suppose that the commentator had before him the reading ebamukha and hence he offered a curious explanation of the word with the use of the word mukha? Actually, however, this elamukha in writing must have represented elamukkha (< eda Madhu Vidya/583

Page #609
REVIEWS markhu) (note that the metre requires the fourth syllable to be long) and herein we get further support for our suggestion that elamuga through elamukkha comes from elamurkha. Under elamuga (SS 47) LUEDERS does not treat the above passage. Perhaps he wanted to restrict his remarks in the present article to ela and discuss mga later while dealing with the change k> g (887). This becomes evident from the following remarks of the editor given in f.n. 1 on p. 77-"Auf einem Zettel, der dem Papier und der Schrift nach alter als das eigentliche Manuscript ist, hat Luders unter der Uberschrift,k wird. g' muga und elamuga vermerkt. Fur elamuga hat er dort die Bedeutung, geifernd' gegeben und scheint bei diesem Ansatz an Wurzel muc gedacht zu haben." The meaning .geifernd' given to elamuga by LUEDERS obviously refers to the above passage. But this sudden change in the meaning is not warranted, and as shown above elamuga is, used with reference to the Brahmana and not the serpent. That muga (< murkha) 'a fool' is not merely a supposition but has been actually used in Pali can be shown at least from one passage in the Mugapakkhajataka (VI, 538). That in this Jataka the usual word muga muka 'dumb' is also used is clear from the following Gatha 33 (p. 16): naham asandhita pakkho, na badhiro asotata/ naham ajivhata mugo DUTOIT, "Nicht bin an Gliedern ich ein kruppel, nicht taub, weil kein Gehor ich habe, nicht stumm, weil mir die Zunge fehlt; 61 The word muga occurs also in many other Gathas (4, 5, 38, 58) in this Jataka where it seems to be used in the sense of 'dumb'. The behaviour of prince Temiya as 'dumb' is also assured by the question put to him by the charioteer-kasma pituc ca matuc ca santike na bhani tada (Gatha 32, also 61), DUTOIT, "Weswegen sprachst du fruher nicht bei deinem Vater, deiner Mutter?", as also by the use of the expression vissatthavacano (Gatha 58, 62) 'ganz frei vermag er ja zu sprechen' or 'der Stimme ist er machtig' (DUTOIT) in describing the prince. But the fact that muga murkha 'fool' is also used in this Jataka becomes clear from the last quarter of the Gatha 33 cited above-(naham ajishata mugo), ma man mugam adharayi. DUTOIT renders the line as "(nicht stumm, weil mir die Zunge fehlt); erkenne mich nicht an als stumm". This rendering of muga 'stumm' in the last quarter cannot be correct because there is no. point in asserting twice that the prince, especially when he was speaking Madhu Vidya/584

Page #610
62 REVIEWS the Gatha himself, was not dumb. That he was not dumb has been aiready asserted in the first quarter of the second line, and now the word muga coming again in the second quarter must have been used to deny some other supposed drawback in the prince. That this drawback was nothing but his feigned acting as a fool (hence here muga= murkha) is shown by the very first Gatha in this Jataka where a goddess advised the prince to behave as ma pandicciyan vibhavaya, balamato bhava sabbapaninan/ sabbo tan jano ocinayatu, evan tava attho bhavissati// DUTOIT, "Nicht deine Weisheit lasse sehen, als Tor werde geschatzt bei allen Wesen. Das ganze Volk soll dich verachten; so wird dein Zweck dir in Erfullung gehen." Similarly Gatha 4 tells us that the prince was not only taken to be physically disabled but also as acetasa unvernunftig. Later, when it was discovered that the prince was not really as he pretended to be, he was called pa n n o 'weis'. These references clearly show that the prince for some years not only behaved like deaf, dumb, and crippled but also as a fool.12 That is the reason why having denied crippleness, deafness and dumbness in the first three quarters of the Gatha, he asked the charioteer not to mistake him for a fool (ma man mugan(= murkhan)adharayi) in the last quarter of the Gatha. Now that the meaning of muga = murkha has been ascertained for Gatha 33, it may be examined if in some of the other Gathas also in this Jataka muga means not 'dumb' but 'fool'. In Gatha 4 where muga and acetasa both occur, muga definitely stands for muka 'dumb'. But in Gatha 38 where only muga and pakkha occur, though the prince had shown in all four deficiencies, it is possible that pakkha stands for the three physical disabilities (muga, badhira, and pakkha), and that muga stands for the mental one alone viz. foolishness (cf. amugo mugavannena, apakkho pakkhasammato). In Gatha 58 we read na so mugo na so pakkho, vissatthavacano ca so. Here if we do not regard vissatthavacano as a further elaboration of na so mugo (muko), then that expression alone 12. The prose portion immediately preceding Gatha 1 tells us that the goddess told the prince to act as a fool by taking recourse to the three physical deficiencies of crippleness, deafness, and dumbness apithasappi yeva pithasappi viya hoki. abadhiro badhiro viya hohi, amugo va mugo viya hohi (printed as hoti), imani tini angani adhitthaya panditabhavar ma pakasayi (DUTOIT, "..... So werde, obwohl du kein verwachsener kruppel bist, wie ein verwachsener kruppel; obwohl du nicht stumm bist, sei, als warest du stumm; obwohl nicht taub, stelle dich taub. Indem du diese drei Merkmale zeigst, verrate nicht deine weisheit.") Madhu Vidya/585

Page #611
REVIEWS would declare that the prince could speak clearly. In that case muga (=murkha) and pakkha in the first quarter would be contrasted, as in Gatha 38, as referring to the mental and physical deficiencies,13 In Gatha 5, the meaning of muga is doubtful; but as it stands by the side of badhira 'deaf, it may there have meant 'dumb'. Now as regards the meaning of muga= murkha in relationship with eda, it would be good to start with some of the observations of EDGERTON in BHS Dictionary. Under edaka-muka we read "Here perhaps the literal idea, speechless, is dominant, tho the parallel jadasamena suggests that overtones of dull, stupid must also be present". For edamuka EDGERTON gives the meaning 'stupid (lit. dumb) as a sheep'.14 But as has been shown above the word just literally means 'stupid as a sheep' and not dumb as a sheep', and originally at any rate muka (< murkha) in this compound had nothing to do with muka = dumb. At the close of his remarks under edamuka, EDGERTON observes, "Sheep are proverbially stupid in other countries than India." This remark appears to be almost correct as stories showing the stupidity of sheep are ordinarily not known from Indian literature. However, so far as the proverbial stupidity of sheep is concerned we may do well to refer to a popular nyaya called ajaklpaniya which has been cited in the Kasika under P. 5.3.106 (together with kalcatuliya and andhakavarttakiya). It has been taken there to convey the idea of an unexpected happening, a surprising event (atarkitopanatan citrikaranam ucyate--Kasika). As regards the interpretation of this nyaya we get the following information in the Laukikanyayanjali (A Handful of Popular Maxims, Part I) collected by Colonel G. A. JACOB (2nd edn., Bombay, 1907, p. 1): The maxim of the she-goat and the sword. It is founded on some story of a goat's being suddenly killed by accidental contact with a sword, and is used to illustrate any surprising event happening altogether by chance. It, therefore, belongs to the same class as kakataliya, khalvatabilviya and others of a similar kind. An excellent illustration of its use is found on page 229 of sriharsha's Khandanakhandakhadya :-panau panca vara. takan pidhaya kascit prcchati kati varataka iti/ prstas cajakrpaniyanya. 13. It may further be pointed out that perhaps the title of the jataka muga. pakkha also is intended to show similar contrast between muga (foolishness) and pakkha (physical disabilities) and hence there too muga possibly stands for murkha. 14. To quote LUEDERS again from the end of SS 47 (p. 48), "Die im PW. aus. gesprochene Vermutung, dass edamuka das gewohnliche Wort fur Schaf', eda, enthalt, also stumm wie Schaf' bedeutet, ist daher nicht von der Hand zu weisen. Die fur das Pali gesicherte Bedeutung ,Idiot' wurde sich leicht daraus ableiten lassen." Madhu Vidya/586

Page #612
64 REVIEWS yena braviti panceti/ In a footnote the maxim is thus explained :kanduyanartham stanbhadau sithilabandhakhadge chagi grivan prasiirayati yadrcchaya ca griva chidyate's tathabhuto "jakrpaniyanyayah kakataliyanyayasamah/ Vardhamana puts it differently in his comment on Ganaratnamahodadhi iii.196:-yathajaya bhumin khunantyatmavadhaya kopano darsitas tattulyar vittam kenacid atmavinasaya krtam ajakrpaniyam/" It appears that both the explanations given above are based on different versions of the same story. The first explanation makes the death of the sheep altogether accidental as it is shown to result from the fall of a sword loosely tied to a post or some such thing. The second explanation, on the other hand, allows the room for accident only upto the point that a sword hidden in the ground was accidentally brought to the notice of some one who was probably looking for it and who having found it killed the sheep with it. The expression atmavinasaya seems to suggest the stupidity of the sheep in being responsible for bringing the knife to light. Both these motifs are brought together in a version of the story which is preserved in the Takkariyajataka (481). The occasion at which the story was narrated shows that it was intended to illustrate the foolishness of the sheep. A Brahmana named Takkariya had almost succeeded in bringing about the death of the paramour of his wife. But as he foolishly rushed to reveal the plan to his wife he himself fell a prey to his trick. To this foolish behaviour on the part of the Brahmana a reference is made in the opening line of the Jataka which forms part of the Gatha (1) placed in the mouth of Takkariya. It describes the Brahmana a bala 'a fool'. cf. aham eva dribbhasitan bhasi balo (DUTOIT, "Ich sagt, ich Tor, was ich nicht sagen sollte"). As Takkariya stood there with the threat of death on his head, his pupil began to narrate to hirn how others also had suffered death on different occasions for having acted foolishly or for not having guarded their speech. Among the four stories told by the pupil, the first one describes a merchant's son bringing misfortune to himself by not controlling his tongue and rendering unnecessary advice. In the other three stories death came to those who did some foolish acts. Among these occurs the story of a sheep which was 15. The commentary Padamanjari on Kasika 5.3.106 gives also somewhat similar explanation, but it makes the death of the sheep more accidental as it makes no reference to the scratching of the neck by the sheep, cf. ajaya gacchantyah ky pancnadhahpatata yatha vadhas tatsadream ityarthah/ The Pandit, Vol. XVII.367 (Benares, 1895). Some such explanation was also known to the authors of the Tattvabodhini and the Balamanorama commentaries on the Siddhantakaumudi. Accidental death of a goat is also referred to in the Jataka 18, where the mishap is said to have occurred due to a piece of stone which sprang as a result of a lightning-stroke. Madhu Vidya/587

Page #613
REVIEWS stolen by some thieves and which they wanted to kill and eat the next day. The next day, however, when they went to the bamboo-grove where they had concealed the sheep the night before, they forgot to take a knife with them. They, therefore, in a pious mood released this sheep, which in joy began to jump here and there. Just then it struck against a bamboo tree where a bamboo-cutter had kept his knife. The knife fell down and the thieves having come to know of it killed the sheep with this knife and ate it. Thus it suffered death by its own foolish act. The Gatha (6) referring to this story is slightly different from the above account which is based on the prose narrative. It makes no direct reference to the thieves who are supposed to have made use of the knife to kill the sheep. aja yatha velugumbasmim baddha avekkhipanti asik' ajjhaganchi/ ten' eva tassa galak' avakantam ayam pi attho bahu tadiso va// DUTOIT, "Ne Ziege, die in einem Bambusdickicht war gebunden, stiess beim Hupfen an ein Messer. Mit diesem schnitten sie den Hals ihr ab. Dies bracht' etc." Apparently a similar story but with a clear reference to the digging up of the knife from the ground (cf. the explanation of Vardhamana given above) is referred to in the Mbh 2.59.8 (critical edition) and has been discussed by EDGERTON in JAOS 59.366-68 (1939).10 He has also referred to the ajakrpaniya and the Jataka story (481) mentioned above. The verse and its translation as given by EDGERTON run as ajo hi sastram akhanat kilaikah sastre vipanne padbhir apasya bhumim/ nikrntanam svasya kanthasya ghoram tadvad vairam ma khanih panduputraih// 65 "For it seems that a certain goat dug up a knife, when the knife was lost, knocking away the earth with his feet,-(and so dug up) a cruel cutting of his own throat. In this manner do not you dig up enmity with the sons of Pandu!" The purpose of referring to the story of the sheep is clearly to warn the Kauravas not to do any foolish act which would lead to their own destruction. While mentioning the different versions of the story EDGERTON observes (p. 367), "All versions agree in treating the story as 16. I am indebted to my friend Dr. A. M. GHATAGE for his kind information about this article. 17. The Mbh. verse occurs at 2.66.8 in the Bombay edition where Nilakantha alludes to two versions of the story. They are, however, not very helpful for our purpose. Bull. DCRI. xvii-5. Madhu Vidya/588

Page #614
66 REVIEWS an instance of a mishap brought upon its victim by his own stupid (italics ours) action." The basis of the ajakrpaniya thus seems to be in a story which was intended to show the foolishness of the sheep. Though later on this maxim came to be classed with others like the kakataliya referring to accidental occurrences, originally it alluded only to the foolish act of the sheep and intended to illustrate deaths brought about by one's own stupid acts. This would, therefore, justify our suggestion to take elamuga (voc.) as going with the Brahmana in the Jataka III.402 (p. 347). The Brahmana was called elamuga 'fool like a sheep' because he was likely to bring about his death by his own foolish act of carrying the serpent in his bag. 8. $ 82, pp. 70-72 : LUEDERS is absolutely correct when he say: that the difference in the forms sithila Dhammapada 346 and susthiru Udanavarga 2.6 can be explained only when we presume, that both of them come from a common base in the eastern dialect. I, however, believe that the explanation of the difference in the two forms has to be given in a way different from the one given by LUEDERS. In Dh. 346, sithila (sithila) 'locker' occurs as an adj. to bandhana 'festet. As in the same stanza this bandhana is also described as dalha (drdha) 'fest': etam dalham bandhanam ahu dhira oharinam sithilan duppamuncam / LUEDERS regards the use of sithila =sithila as 'sinnlos'. The ex. planations of sithila given by the commentators are no doubt forced." Moreover LUEDERS points out that sithilan is metrically irregular because the fifth syllable in the Pada is required *o be long (oharinam sithilan.). Now in the parallel passage in the Udanavarga 2.6, we find susthiru dauerhaft which agrees well with the adj. drdha applied to bandhana. etad drdham bandhanam: ahur aryah samantatah susthiram duspramoksam / LUEDERS, therefore, concludes that the original text in the eastern language must have had sutthila (< susthira), written as suthila. This was probably corrupted to sathila which the Pali translator mistook for sithila. The mistake of the Pali translator arose because in Pali we have sathila also corresponding to Skt. sithila The Sanskrit translator of the 18. " -dass eine solche Fessel locker zu nennen sei, da sie Haut und Fleisch nicht zerschneide, kein Blut hervortreibe und die Fesselung nicht empfinden lasse..." Madhu Vidya/589

Page #615
REVIEWS Udanavarga, on the other hand, rendered the eastern form suthila correctly by Skt. susthira. But let us first look at the nature of the bandhana here described. It is the fetter in the form of attachment to worldly pleasures which are detailed in the stanza before (Dh. 345) as manikundalesu puttesu daresu ca yu apekha (Hangen an auserem Schmuck und der Sorge um Weib und Kinder) and which are in a general way referred to as kamasukhan in the present stanza (346) or as kamasya bandhanan in the Tibetan version of the parallel passage in the Udanavarga 2.6 (cited by LUEDERS. on p. 72). In these two verses it is this kamabandhana then which is sought to be contrasted with the other fetters of iron, etc. mentioned in Dh. 345 and which the wise do not regard to be as dalha as the kamabandhana. cf. na tam dalhan bandhanam ahu dhira, yad ayasan, darujam, pabbajan ca; (345) Max MUELLER, "Wise people do not call that a strong fetter which is made of iron, wood or hemp." Then it is said in verse 346 that, etan dalhani bandhanam ahu dhira etc. In this context therefore daha (drdha) does not seem to mean 'fixed or firmly fastened', but only 'strong, hard'. It would not then be 'sinnlos' if a bandhana is called sithila and drdha in the same breath. A fetter can be loose and still so strong that it is difficult to break it. It is from this point of view that the kamasukhabandhana in our stanza is called sithilan, but at the same time duppamunca, or as the Tibetan version puts it sithilam api dusprame ksam where the contrast between sithila and duspramoksa is clearly brought out by the addition of the particle api. That the fetters of love are tender and loose and yet strong as they are difficult to be got rid off is well known from ancient literature. To quote the following verse from the Subhasitaratnabhandagara p. 390.511bandhanani kila santi bahuni premarajju drdhabandhanam uktam/ darubhedanipino 'pi sadanghrir niskriyo bhavati pankajakose// Here it would be futile to say that the bandhana of the lotus flower is tight or immovably fixed on the bee. It is loose and tender also; but at the same time it is strong (drdha) enough for the bee! That drdha and sithila are not incompatible can also be seen from the following passages to which reference is made in the PW. The Taitt. Sam. 3.2.4 gives certain mantras to be repeated when the sacrificer creeps towards the sadas. In the third section (3.2.4.3) occurs a mantra which according to the commentator Madhava is addressed to heaven and earth. The mantra runs as-drdhe sthah sithire samici mar Madhu Vidya/590

Page #616
68 REVIEWS hasas patan/19 "You two who are strong and loose, being united prutect me from misery". The commentator explains it as--he dyavaprthivi upasthanarahitam prati sithile api yuvam upasthataran prati samici anukule satyau drdhe sthah. The same two adjectives drdha and sithira are also used in the Ait. Br. 13.7 with regard to the joints of a human body. Here an explanation is sought to be given for the repetition of the call 'somsavom' twice, once before and once after the dhayyas inserted in the Vaisvadeva sastra. About this it is said-yatha vai purusa evam Vaisvadevam, tasya yathavantaram angany evam suktani, yatha parvany evan dhayyas, tad ubhayato dhayyam paryahvayate, tasmat purusasya paruari sithirani santi drlhani brahmanahitani dhrtani. "The Vaisvadeva Sastra is like a man; just as he has in him limbs, so are the hymns (in the Sastra), just as he has joints, so are the inserted verses; now he repeats the ahava on both sides of the inserted verses; therefore the joints of a man (though) loose, are strong, (because) they are placed on, held together by brahman in the form of ahava)". Sayana, however, explains -tatha' sati dhayya ubhayaparsvaparvani sithirani purvah sithilani santy api pascat prayatnena dharitani drahani sampadyante.20 Having thus seen that dadha and sithila are not necessarily incompatible, and that sithila is justified in view of the fetter being kamabandhana, we may turn to another adjective oharinam21 given to the bandhana which will also support the view that sithila =sithila was really intended. For the use of these two adjectives together, support is given by the reading oharina sisila in the Kh. Dh. and by oharinan sithilan occurring in the Samyutta-Nikaya 1.77 and Jataka 201.2 where the same stanza appears. Thus so far as the Pali and the Kharosthi traditions are concerned this reading is confirmed. The expression oharinan as an adjective of bandhanam has been translated as 'niederziende' which, according to LUEDERS 19. The Ap. Sr. Su. 12.20.4 and the Baudh. Sr. Su. 7.10 give the mantra which has drdhesthah sithire samici. But in the mantra given in the Man. Sr. Su. 2.3.7.2 we find asithire for sithire possibly because it was felt that drdhe and sithire cannot go together. For references see BLOOMFIELD, Vedic Concordance and KEITH's translation of the Taittiriya Samhita, part I, p. 242, footnote 4. 20. In the Ait. Br. 28.8, while justifying the recitation of silpas at the end of the Narasasa sastra of the third pressing, the word drdha seems to be intended in its two senses viz. hard and fixed, so that once it contrasts with mrdu 'soft' and once with sithira 'loose'; cf. athaitan mrdv iva cchandah sithiram yan Narasarsam, athaiso 'ntyo yad acchavakas tad drlhatayai drdhe pratisthasyama iti, "Now this Narasassa hymn is as though soft, (as though) loose. (But) when the Acchavaka comes at the end that is for the sake of hardness and fixity. (They do it thinking) let us establish it on something hard and firm." 21. This is not found in the Tibetan version which has also nothing to corres. pond to dalhan. Madhu Vidya/591 *

Page #617
REVIEWS himself, is 'nicht ganz sicher'. He is, however, prepared to content himself with the explanations of the commentators, viz. that 'die beschriebene Fessel in schlechte Wiedergeburten hinabziehe'. The explanation is farfetched and is offered in view of the metaphorical sense of the bandhana viz. attachment to worldly pleasures. But to me oharinam seems to be a corruption of oharinam falling or slipping down' which suits excellently with the adjective sithila. As the fetter is loose it slips down again and again like a bracelet. With this can be compared favourably the description of the king's golden bracelet in act III.12 of the Sakuntala. As king Dusyanta had become imatiated his bracelet had become loose and he used to push it back. From there, however, it used to slip down to the wrist and the king was required to push it back again and againanabhilulitajya hatanka manibandhanat kanakavalayan srastam srastam maya pratisaryate. Oharinam was slightly changed to oharinam possibly because the metre required the second syllable to be long. Afterwards oharinam was kept on because it was understood in some such manner as given by the commentators. The translator responsible for the Ud. varga probably had before him the uncorrupted form oharinash. But since he rendered suthila of the original by susthira 'fixed', he found it difficult to combine it with oharinam 'slipping down' and hence he was forced to replace it by an altogether different word samantatah. This latter word, from the point of view of meaning, is superfluous and it constitutes a major deviation from the Pali and the Kharosthi versions as they have nothing to correspond to samantatah. muhur Having ascertained that sithilam, corresponding to Skt. sithilam, represents the original reading, let us look at it from the point of view of form. Pali sithila Vedic sithila have been traced back to "arthila. This would give in the eastern middle Indian dialect sathila as in this dialect normally becomes a and the dental following r is cerebralised. LUEDERS is therefore perfectly justified when he considers sathila, which occasionally appears also in Pali, to be an eastern form. Now it has been pointed out that the eastern dialect shows a preference for the vowel sequence ui a in words having more than two syllables (cf. p. 38 where LUEDERS cites pulisa, munisa from Asokan inscriptions). In that case it would not be wrong if we suppose that in the eastern dialect *erthila gave rise also to suthila (uia) by the side of sathila noted above. And in fact the Udanavarga reading susthira (uia) proves that this was really the case. It is thus clear that the original eastern version used suthila (<*rthila 'loose') which was mistaken by the Sanskrit translator of the Udanavarga to be susthira, but which was correctly understood to represent sithila in the r Madhu Vidya/592

Page #618
70 REVIEWS Pali, Kharosthi, and also that Sanskrit version which served as the basis for the Tibetan translation of the Udanavarga." We are thus lead to a conclusion totally different from that of LUEDERS. We see that the mistake was committed by the Udanavarga translator and not by the others. It may be added that perhaps the metre contributed to the mistake for it required the fifth syllable to be long and hence suthilam of the original was in some recension pronounced as sutthilam (to make su metrically long). The explanation here given of sithila: susthira has further the advantage that it does not compel us to assume that the original reading was corrupted to sathila. 9. 84, pp. 74-75 In SS 84-SS86 LUEDERS gives certain very interesting instances of word-plays which can be better appreciated when one presumes that the original text from which the Pali verses were translated had certain words with . Thus for example in the Sutta-Nipata 535, which tells us who, according to the Buddha, can be called ariya (arya), LUEDERS points out that in the original eastern text there must have stood the word aliya (with ) for Pali ariya. This has been suggested because aliya is nearer in sound to abaya 'desire' used in the first line of this verse. This conclusion of LUEDERS may be accepted but we have some other things to suggest with regard to the interpretation of the verse. The first two lines of the verse read chetva asavani alayani vidva so na upeti gabbhaseyyam/ FAUSBOELL. "Whosoever, after having cut off passions and desires, is wise and does not (again) enter the womb... But the use of chetva from Vehida 'to cut off' with asavani is strange. The root meaning of asava (=*asrava) is 'that which flows' and this is not something that can be cut off. It is a thing that is to be destroyed in that it is gradually wasted away. That is the reason why we often get the use of khaya (ksaya) 'extinction' and khina (ksina) with asava. In conformity with this usage one anticipates in the above verse ?? 22. This is important because as mentioned by LUEDERS himself the Tibetan translation often points to a Skt. text which is older than the vulgate of the east Turkestan manuscripts of the Udanavarga. LUEDERS who believes sutthila susthira to be the original reading, finds it therefore difficult to reconcile his view with sithilam of the Tibetan version. "Schwerer wiegt" he observes, "dass die tibetische Version des Udanavarga, die zweifellos eine Ubersetzung des Sanskrittextes ist, ebenfalls die Lesart sithilam kennt." 23. cf. PTS Dictionary under asava, khaya and khina, and also BHS Dictionary under asrava. Madhu Vidya/593

Page #619
REVIEWS 71 a form of the verb khayati (ksayati) 'to destroy, etc.,' with usava, and in fact it is possible to derive chetva from Vksi. As is well known in the western dialect of the inscriptions of Asoka Skt. ks is represented as ch. Hence in that dialect we can get a gerund chetpa from ksayati.24 The Pali correspondent of this western chetpa would be chetva (or chetTana) which occurs in our verse. In chetva asavani 'having destroyed the i savas', therefore, the former is not to be confused with chetva, the gerund from Vchida. Not realising this but feeling at the same time that chetva from Vchida 'to cut off' does not go well with asava, the Burmese scholars made an unsuccessful attempt to mend the matter by transposing the words asavani and alayani and reading the line as chetva alayani asavani.25 Against the proposed derivation of chetva from ksi in chetvu usavani it cannot be argued that sometimes we get the expressions sotari chetva or chinnasota where sota kh and ch in Pali cf. GEIGER $ 56. The eastern counterpart of the western chetpa would be khettu, 25. This seems to be the real ground for the transposition of the words and not, as LUEDERS observes, while in the verse ariya and alaya are brought into relationship by a play on words. cf. "In unserer Strophe wird ariyo mit alayani, den Lusten, die der ariya vernichtet hat, in Verbindung gebracht. Dass es darauf ankommt, haben die birmanischen Gelehrten erkannt, und aus diesem Grunde ist in den birmanischen Handschriften alayani vor asavani gestellt worden." The Mahavastu translator (I11.400.3f.) also seems to have felt the difficulty in connecting chetvit (from Vchid) with asavani and alayani and hence he tried to get over it by substituting hitva for chetva and also adding bandhanani as a gloss to asavani and alayani (p. 75, f.n. 1). cf. hitva alayani asavani bandhanani ativytto neti. It may be added that the regular BHS gerund from Vha 'to abandon' would be jahitva, cf. EDCERTON, Language 13.117 (1937). Madhu Vidya/594

Page #620
REVIEWS of monks-ye te bhikkhu arahanto khinasava... te pi tiriyan Marassa sotan chetva sotthina param gata (M.I. 226).26 Proceeding from the above passage it would be legitimate to conclude that as in sotan chetva so in chinnasota also chinna means 'one who has crossed' and not 'one who has cut off (the stream)'. That this is in fact the case can be gathered from the following occurrences : In Sutta-Nipata 715 we read, yassa ca visata n'atthi chinnasotassa bhikkhuno / FAUSBOELL, "for whom there is no desire, for the Bhikkhu who has cut off the stream (of existence)...." However here chinnasota does not mean 'who has cut off', but 'who has crossed the stream'. This is shown by the fact that in the preceding verse (714) we read, na parang digunan yanti FAUSBOELL, ".... they (i.e. the Samanas) do not go twice to the other shore...." Again in Sutta-Nipata 948 we find, yo 'dha kame accatari samgan loke duraccayan / na so socati najjheti chinnasoto abandhano // FAUSBOELL, "whosoever, has here overcome lust, a tie difficult to do away with in the world, he does not grieve, he does not covet, having cut off the stream, and being without bonds." That here too chinna can only mean 'one who has gone beyond' is shown by the use of the expressions accatari (from ativtara 'to go beyond, to cross') and duraccayam (from dur + accaya 'difficult to cross'). (Note also that in the Sutta-Nipata 945 kamapanka is described as duraccaya where the commentary explains it as duratikkamaniya). This will also be clear from Itivuttaka 5.6 (p. 95) where chetva sotan duraccayam can only mean 'having crossed the stream which is difficult to cross'. In Dhammapada 383 we have, chinda sotan parakkamma kame panuda brahmana / 26. We may compare with this oghatinna and oghatiga, PTS Dictionary s.v. MadhuVidya/595

Page #621
REVIEWS 73 Max MUELLER, "Stop the stream valiantly, drive away the desires, O Brahmana !"27 However, in this instance we have not to think of the stopping or cutting off of the stream but of its crossing. This is shown by the next verse (384) where the idea of 'going beyond', though in a different sense viz. 'to accomplish, to have perfection in, etc.', is similarly present. yada dvayasu dhammesu paragu hoti brahmano/ Max MUELLER, "If Brahmana has reached the other shore in both laws (in restraint and contemplation...."28 Moreover in the same varga of the Dhammapada occurs another verse (414) which leaves no doubt about the way in which chinda sotam in verse 383 has to be interpreted. Verse 414 reads as yo iman palipathan, duggan, sansaram, mohan accaga, tinno para gato, jhayi, anejo akathankathi, anupadaya nibbuto, tam ahan brumi brahmanam / Max MUELLER, "Him I call indeed a Brahmana who has traversed this miry road, the impassable world and its vanity, who has gone through, and-reached the other shore, is thoughtful, guileless, free from doubts, free from attachment, and content". In the end we may refer to Samyutta-Nikaya IV. 291 where chinnasota is used as an adjective of a chariot (ratha) and hence can only mean 'which has crossed the stream'. cf. anigham passa ayantan chinnasotam abandhanan. Here LUEDERS (p. 49) also follows the traditional translation of chinnasota when he renders the line as, 'Sieh den Unerschutterten kommen, der die Flut der begierden abgeschnitten hat, den Fessellosen." Thus it will be clear that in those passages where sota and forms of Vchida are brought togeher the idea to be conveyed is of 'crossing the stream' and not of cutting it off or destroying it'.29 It may then be argu 29. The same line recurs in Samyutta-Nikaya 1.49 which Mrs. Rhys DAVIDS translates as, "Advance with valiant energy, recluse, cut off the stream, scatter desires of sense". (Kindred Sayings, 1.70). 28. It may be added that the idea of crossing the stream also continues in the following verse (385) which, however, is difficult to interprete. Read-yassa puram aparan va, paraparar na vijjati...... tam ahar brumi brahmanan.// Max MUELLER, "He for whom there is neither this nor that shore, nor both, him....I call a Brahmana." 29. Where the idea of drying up of waters is to be conveyed we meet with the use of khina and not chinna with sota. cf. macche va appodake khinasote (SuttaNipata 777), FAUSBOELL, like a fish in a stream nearly dried up, with little water'. The use of ud chid in the sense of 'to break off, to stop, to want' is noted in PW where a passage cited from the Pancatantra II.92 shows its use with kusarit. cf. arthena tu vihinasya purusasyalpamedhasah/ucchidyante kriyah sarva grisme kusarito yatha// But this use of ud chid does not disprove our argument. Madhu Vidya/596

Page #622
74 REVIEWS ed that in chetva asavani also we may understand having cut across, i.e., having crossed the asavas', instead of 'having destroyed the asa vas'. But with a savas the idea is clearly not 'to cross' them but 'to waste them away, to destroy them' as shown by the frequent use of khaya or khina with asava.30 Hence it seems desirable that in our passage chetva is not to be derived from Vchida, but from Vksi as explained above. The attempt of the Burmese scholars to improve matters by transposing asavani and alayani has been described above as 'unsuccessful' because just as chetva from Vchida does not go well with asavani it does not go well with alayani either.31 The new meaning of chetva Page #623
REVIEWS 75 as a gerund. This is quite in keeping with the play on words common in the Sabhiyasutta. Hence with the proposed derivation of chetva from V ksi and vidva from Vuyadh the two lines of our passage mean-"Having wasted away the passions (asava), (and) pierced through the desires (alaya), the wise one does not enter the womb (again).... "33 The book has been well printed and the editor and his collaborators (mentioned on p. 11) deserve our congratulations for having made it practically free from printing mistakes. If in spite of the great care devoted by them a few mistakes have remained, they only show how rare such misprints are. On p. 26, f.n. 1, line 5, vaddhetva appears for vaddhetva; p. 75, 1. 17 duddakhiro appears for duddhakhiro; p. 81, 1. 1 suvo vas uvim for suvo va suvin; p. 110, 1.2 from below nvavaya for uvavaya. On p. 127, line 4 appears the short forms (among the Asokan edicts giving the treatment ny > nn) after Dhau. and Jau. According to the list of the abbreviations $ stands for Sauraseni; but that is impossible here. Ob. viously it should have been Top. which stands for the Delhi-Topra edict of Asoka and which, besides the Dhauli, Jaugada and the Kalsi versions, gives instances of the change ny > nn. -M. A. MEHENDALE. 33. I would rather like to put forward the following suggestion for consideration. The first line of the Gatha is metrically defective. The irregularity can be removed if we suppose that originally vidva (absolutive) going with alayani occurred also in the first line and that chetvasavani were read together. The first line would thus read-chetvasavani ulayani vidva. Later when the significance of vidva (absolutive) in line one was lost sight of and was regarded as superfious in the presence of vidvu ( vidvan) in the second line, it was dropped and the line read as at present. LUEDERS ($ 179, pp. 132-133), however, makes a different suggestion to remove the metrical irregularity. Hlu thinks that in the original text asinava (which occurs in the Asokan Pillar udicts II, III) was used in the place of asava. He, therefore, proposes to reconstruct the first line of the original text as chettu usinavani alayani. Madhu Vidya/598

Page #624
Braja-bhasa (in Hindi), by Dhirendra VARMA. Hindustani Akadami, Allahabad (U. P.), 1954, pp. 162. Price Rs. 6-0-0. The book under review is the Hindi translation of the author's original thesis in French La Langue Braj (Dialecte de Mathura) for which he was awarded the D. Litt. by the Paris University in 1935. Since Jules BLOCH published his memorable work on the Marathi lan. guage, there has been a welcome tendency among Indian scholars themselves to attempt a scientific description of Indian languages. Prof. VARMA thus followed the footsteps of such eminent linguists in the field of New Indo-Aryan Linguistics as Siddheshvar VARMA, S. K. CHATTERJI, and Baburam SAKSENA in writing a valuable treatise on a modern Indian language. The author was fortunate to receive guidance from Prof. BLOCH in successfully achieving the task which he set before himself. The study of the Braj language presented in the work is based on the material drawn from the 16th, 17th, and 18th century Braj literature (which is mentioned on pp. 9-1:1), and the one which the author himself collected during the years 1928-30 from Braj as a spoken language in various villages. Before proceeding to give a detailed account of his investigations, the author begins with a short treatment of certain general topics, such as the geographical area where Braj is spoken, the people who speak it, the Braj literature since the earliest times, and the like. This account serves the useful purpose of introducing the reader to the real problems connected with the linguistic study of Braj. From the linguistic point of view the chapters on phonology, morphology, and syntax are very important. As one goes through these pages one notices how admirably the author has achieved his task of giving an accurate and thorough description of the facts connected with the Braj language. He has also taken care to describe the treatment given in the Braj to the sounds of words borrowed from the Persian and Madhu Vidya/599

Page #625
REVIEWS the English languages. Some of the author's important conclusions are summarised on pp. 127-29. He holds the view that the Braj has not undergone any material change, except in some peculiarities of pronunciation and the use of morphological forms (SS 258), during the last three or four centuries. Today the Braj is being greatly influenced by the Khadi boli. With regard to the other modern Indian languages of North India, the author opines that Kanauji is not different from Braj ($75), that Bundeli can be regarded as a southern form of Braj (SS 261), and that among other languages the Rajasthani is the nearest to the Braj (SS 259). The value of the book has been enhanced by the addition of two appendices which did not appear in the original French Thesis published by the author in 1935. One of these gives the specimens of the Braj as spoken at various places in the interior and on the borders of the Braj territory. The second appendix gives a list of all the Braj words occurring in the book. 155 While discussing the name of the language (pp. 16-17), apparently it is not the intention of the author to give a detailed account of all the references where the word vraja occurs. However, the following remarks may be made with regard to the observations made by the author. Vraja seems to occur either singly or in composition in the sixth and the twelfth major rock edicts of Asoka. The actual forms that we get are vaca or vraca (VI) and vaca-bhumikya (or bhumika) or vraca-bhumika (XII). In both cases vaca has been equated with Skt. vraja by some scholars and the forms have been translated as 'cowpen' and 'inspectors of cowpens' (HULTZSCH, The Inscriptions of Asoka, p. 12, f.n. 6 and p. 22, f.n. 5). In my opinion though this interpretation is possible in the latter case, it is very unlikely in the former, as the hardening of jc is not witnessed otherwise in the east (but vaca occurs in the Dhauli and the Jaugada versions as well). Even the context renders the meaning 'cowpen' improbable in the Rock edict VI. I am, therefore, still inclined to side with BUEHLER and others and take vaca in the sixth edict as coming from varcas. Next it may be noted that the Nighantu (1.10) includes vraja in the list of words meaning 'cloud', and the Nirukta while commenting on this passage (2.21) observes that it may also mean 'mountain'. The Nirukta (6.2) derives vraja from Veraj, and this derivation has been accepted by the author as well (SS 29). BOEHTLINGK-ROTH, (Worterbuch, s.v.), however, derive it from Vvari, which is also the view of WALDEPOKORNY (Vol. I. 290). According to BOEHTLINGK-ROTH, vraja occurs also as the name of an area in the neighbourhood of Agra and Mathura in a certain inscription which has been published in the Zeitschrift fur Madhu Vidya/600

Page #626
156 REVIEWS die Kunde des Morgenlandes 4.146-202.' I have not been able to verify this reference. -M. A. MEHENDALE. I am thankful to my friend Dr. W. THOMAS of Gottingen for the following information kindly supplied by him to me about this inscription: Der Artikel, in dem sich dieses vraja findet, stammt von Chr. Lassen und heisst: "Uber eine alte indische Inschrift der koniglichen Satrapen von Surashtra, worin Kandragupta und sein Enkel Acoka erwahnt werden". Er stellt einen Verbesserungsversuch zu Prinsep's Lesung dieser Inschrift dar, die let. zterer in As. J. of B. VII 334 veroffentlicht hatte...... Die Inschrift ist z.T. auf einer rauhen Oberflache eingehauen. Sie gehort dem Konig Rudradamander Name ist nicht ganz sicher; Prinsep liest in seiner Ausgabe Aridaman und bemerkt in einer Note, dass man auch an Atridaman oder Rudradaman denken konne-und bezieht sich auf den Bau einer Brucke und anderer Schutzmittel gegen den Fluss Palesini. Zeile 11 dieser Inschrift lautet: janopajanasvavirydTjitanam anuraktasarvaprakstinam purvaparakaravantanupavrajanartasurastrasva...... Zu Zeile 11 bemerkt Lassen auf. S. 171 f.: Das Facsimile richtig avarittyanupa, fur vraja aber wie es scheint ativraja........ Zur Erklarung der Namen. Avanti ist Ujjayini, wie bekannt; Anupa aus anvapa "Land langs dem Wasser" ist wohl die Kuste. Guzerat, das Wort ist sonst bekannt genug; vraja ist die Gegend um Mathura.; ativraja ware das Land daruber hinaus;........ Madhu Vidya/601

Page #627
REVIEWS Buddhist Hybrid Sanskrit (Language and Literature): Ten Public Lectures by Franklin EDGERTON, Banaras Hindu University, Banaras, 1954, pp. 88. These are the ten public lectures delivered by EDGERTON in fulfilment of one of the conditions of the Holkar Visiting Professorship at Banaras which he occupied during 1953-54. EDGERTON had already made. himself known as a savant deeply interested in the study of Buddhist Hybrid Sanskrit by bringing his patient research in the subject to a fruitful culmination by publishing in 1953 the Buddhist Hybrid Sanskrit Grammar, Dictionary, and Reader in three volumes (Yale University). The Banaras Hindu University, therefore, could have found no better person than EDGERTON himself to speak on the subject so carefully studied by him. Buddhist Hybrid Sanskrit is the name invented by EDGERTON to describe the language used by the north Indian Buddhists for their religious literature. The purpose of the present series of lectures, as mentioned by the lecturer himself, is not a scholarly and detailed presentation of the subject-that being already achieved in the Yale University publications referred to above-but to offer to a wider public a general survey of the literature and some of the salient features of that language which holds a somewhat peculiar position midway between the standard Sanskrit and a middle Indian dialect. The treatment given to the subject is therefore selective, not exhaustive. Lectures 1 and 7 are important as they give the lecturer's views on the nature of the Buddhist Hybrid Sanskrit and its relationship with the other middle Indic dialects. In, this regard, the lecturer has two. points to make. The one is that BHS is basically a middle Indic dialect, and it is not, as was erroneously assumed by some, a bad or corrupt Sanskrit. It has, however, to be remembered that this applies not to the literary language as we have it preserved today-for in its present form it is highly Sanskritised in many works-but only to the dialect which lies at the bottom of these texts. The next point which the lecturer intends to stress is that this underlying middle Indic dialect is quite an independent one and that it cannot be properly identified with any one of the known middle Indian dialects including Pali. He has arrived at this conclusion on the strength of certain peculiar characteristics of BHS (summarised in lecture 7), which, in his opinion, are not to be met with. Madhu Vidya/602

Page #628
232 REVIEWS in other MIA dialects. He also now gives up his earlier view that BHS has close relations with Ardhamagadhi, the sacred language of the Jains (BSOS 8.501 ff.) A useful survey of BHS literature, which represents both the Mahayana and the Hinayana schools of Buddhism, is given in the lectures 2-6. Quite appropriately the lecturer devotes some more time to such important works as the Mahavastu, the Lalitavistara, and the Saddharmapindarika. Lectures 8-10 present what the lecturer calls 'the normal grammar' of the dialect underlying BHS. For the sake of convenience he omits here the very rare and sporadic forms which he has already dealt with in his big grammar. The lecturer convincingly shows that the study of metre is very essential if we wish to get a correct picture of the phonology (and also of morphology) of BHS especially in respect of the length of vowels and the pronunciation of clusters at the beginning of the words. The metrical evidence clearly shows that at bottom BHS pronunciation was middle Indic, whatever the form in which the words appear in the manuscripts to-day. An instance of how a systematic study can lead to correct interpretations is to be found on p. 86 where the lecturer points out that the tendency observed in BHS to form denominatives like buddhati etc. from the past passive participles shows that the Pali and the Prakrta laggati cannot come from *lagyati (as was supposed by GEIGER $ 136 (2) and PISCHEL 8 488) but from lagnati. But what the lecturer says about the genesis of the loc. sg. of a stems in -esmin etc. may not be the case. He regards, for instance, a loc. sg. form like lokesmim as a blend-form, a cross between the two loc. forms lokasmir and loke. It might, however, be investigated if the starting point of such locatives was not a frequent use of loke and asmin together which resulted into loke(a) smin. In lecture 1 as well as on p. 61, the lecturer refers to his theory that it is not possible to speak of one original language of the Buddhist canon. On his own admission "Many of these canonical works no doubt went back to the earliest times, and were carried everywhere in similar forms; but there is, as we saw in our first lecture, no reason to assume linguistic unity even in the texts as they were thus spread by missionaries; there is no reason to assume any single 'original language of Buddhism' (p.61)." However, the attitude of the Buddha towards the use of various languages as vehicles for religious propaganda, to which the lecturer refers in his first lecture, is not against the hypothesis of there being one language of the original canon. It only shows that the Buddha was definitely against this original canon' itself being imposed on his disciples at all places in India; he was, on the other hand, in favour of the use of the regional Madhu Vidya/603

Page #629
REVIEWS 233 languages for the further propagation of his teachings which originally could only be in one language. In a recent publication Beobachtungen Uber Die Sprache Des Buddhistischen Urkanons (edited by E. WALDSCHMIDT), LUEDERS has demonstrated that the original canon was composed in an easternly dialect. The students of the MIA dialects will eagerly await Prof. EDGERTON'S reactions to this new data presented by LUEDERS. -M. A. MEHENDALE. Madhu Vidya/604

Page #630
340 Annals of the Bhandarkar Oriental Research Institute KANHADADE PRABANDHA BY PADMANABHA (Vol. I Text), critically edited by Prof. K. B. Vyas and published by the Director, Rajasthan Puratattva Mandira, Jaipur, in the Rajasthan Puratan Granthamala, Vol. 11, 1953, pp. 1-8, 1-33, 1-275. Price Rs. 9-8-0 Kanhadade Prabandha is a mediaeval epic poem composed in V. S. 1512. It describes the heroic fight and the great sacrifice of the Chauhan king Kanhadade or Krishna Deva of Jhalor in Rajaathan while struggling against the Muslim invaders of north India. The work is no doubt important from the historical and literary points of view; but it is equally or even more valuable from the linguistic standpoint, for it was composed at an early stage when old Gujarati and old Rajasthan! were not much differentiated from each other. The poem was first published by Shri Navalram Pandya in 1877-78 from a single manuscript, of V. S. 1930 discovered by Buhler. It was published later in 1913 in a better form by shri Derasari with the help of four manuscripts. the oldest among these being of V. S. 1648. Derasari's edition was republished in 1926 without any material change. The present edition, which embodies the editor's patient labour of 14 years, is no doubt superior to those inentioned above. It is based on all the ten available manuscripts in Government and public libraries and in private collections. The editor has been lucky to have obtained a manuscript (designated by him as A which is dated in V. S. 1598 and copied at Jhalor, the place of the poet as well as of the hero of the poem. The importance of this manuscript, dated as it is only 86 years after the composition of the poem, can be easily realized. It was therefore not necessary for the editor to try to enhance its value by surmising without adequate evidence" It is probable that the copyist bad access to the poet's autograph or a direct copy of it." (p. 29). This first volume comprises the text of the poem. The second volume is intended for giving the historical and linguistic background, a translation of the poem, an Index Verborum, and, if possible, critical and explanatory notes. Madhu Vidya/605

Page #631
Reviews 341 On pp. 30-32 the editor gives the principles adopted by him for constituting the text. They will be discussed better by an authority more intimately acquainted with the work of textual criticism than the present reviewer. It may, however, be observed that on the basis of these principles the editor seems to assume that the author was not liable to commit any mistake, or that he could not have used a dialectal form in the place of a norm regar. ded as standard by the editor for a work of V. S. 1500. The question arises as to what should bave been the purpose of the editor,to give a text which is likely to be as near as possible to tbe original as written by the author, or which is as near as possible to the one which should have been written by the author ? It is not possible here to comment in details on the readings adopted in the text. In Appendix II we do not find any note on verse 1 in Khanda I. The very first word gaurinandana, adopted in the text, bas & variant Gorinandana. It will have to be seen whether the vowel o in the first syllable of the variant was used as it was thought to be a near approach to the open pronunciation of the vovel in that syllable, gauri then may be a Sanskritization. It is clear that nut all the forms as used in the poem could be current in actual speech. This is easily seen from the instances of sarasalti (Sk. sarasvati ) and matli ( Sk. mali) in verse 1. The gemination in the final syllable is only due to metrical grounds, and it could not have any relation to the actual pronunciation. This is the reason why the ms. J has mistaken matli for nutti (Sk. muktt) which in the context is quite out of place. We should, however, be grateful to Prof. Vyas for the great labour spent by him in constituting the text and reserve our final remarks until after the publication of the second volume. M. A. Mebendale Madhu Vidya/606

Page #632
ALBERT THUMB/RICHARD HAUSCHILD: Handbuch des Sanskrit. Mit Texten u. Glossar. Eine Einfuhrung in das sprachwissenschaftliche Studium des Altindischen V. A'T'. I. Teil : Grammatik. 1. Einleitung u. Lautlehre. II. Teil : Formenlehre. Dritte, stark umgearbeitete Aufl. v. R' H' (Indogermanische Bibliothek, I. Reihe : Lehr- u. Handbucher). 8o XVI + 347 / XII + 492 p. Carl Winter Universitatsvig., Heidelberg (1958/59). DM 32,-- / 44,-- T's Handbuch des Sanskrit first appeared in 1905 in 2 parts : 1. Grammatik, and 2. Texte und Glossar. When the 1spl was out of print, HERMANN Hirt brought out a 2nd ed. in 1930. This was mostly a reprint of T's work with some corrections in the text, but all additions were given in the Nachtrag. The 2 vol.s under review are the 3rd ed. thoroughly revised by H'. (A 2nd ed. of the 2nd pt also enlarged and thoroughly revised by H' appeared earlier in 1953; cf. the review by C. REGAMEY in Erasmys, X <1957> 214-6.) T' intended to give an introduction to the historical study of Sanskrit since no such book was available for the use of the beginners. There were of course available good descriptive grammars of Sanskrit. But since Sanskrit was studied in the wider circle of comparative philologists, a good and convenient manual giving the position of Sanskrit as a language belonging to the IE. family was called for. The ground for this was already prepared by BRUGMANN'S Grundriss and WACKERNAGEL'S Altindische Grammatik, I. As regards the usefulness of the further publication of the vol.s (esp. III) of the Alt. Gr. in revising T's Handbuch, Hirt observed that "... ich habe nirgends einen Fall gesehen, wo mich Wackernagel veranlasst haotte, von meiner Auffassung abzugehen ..."H'. however, rightly does not agree with Hirt's view and has fully utilised for the 3rd ed. all the available vol.s of the Alt. Gr., DEBRUNNER'S additions to Vol. I, and RENOU'S Introduction gednedrale. Not only this. While trying to include in the body of the text of the 3rd ed. whatever was good and useful in Hirt's Nachtrag, H' found, "... dass die Ausbeute aus diesem 'Nachtrag' nicht allzu erfolgreich war, dass ich vieles als zu weit abgelegen ausscheiden musste und mancherlei mir auch gaonzlich unhaltbar dunkte". Although H'has preserved the inner arrangement of the material presented in T's book, his revisions and additions give a new appearance and value to it. How meaningful the words 'stark umgearbeitete Auflage' are can be realised even by a look at the 2 vol.s. The matter which originally comprised a single vol. of 505 pages has now extended to 2 vol.s of 347 and 492 p. H' has wisely discarded in the new edition the superfluous use of the Nagari script and the Brugmann's method of transliteration which is now no more in use. So far as the text of the Handbuch is concemed, whereas H' has thoroughly revised the Einleitung (e.g. in describing the position of Sanskrit among the Indo-European languages H' has widened the scope in both directions by bringing in Das Mittani-Indische on the one hand and by taking MadhuVidya/607

Page #633
into account the Middle and New Indo-Aryan languages on the other. He also takes note of the substrat - influences) and T's text of the Lautlehre, his additions in the 2nd vol. mostly pertain to the notes and the literature cited. -- H's contribution mainly belongs to the matters of linguistic explanation in respect of which he had planned to give up-to-date information (... das Werk auf den wissenschaftlichen Stand der Gegenwart zu bringen). Here in certain details the experts in the science may not agree with him or find fault with him for not having given the latest view. The book has been already reviewed by F. B.J. KUIPER (Lingua, VIII <1959> 424--41), W. P. SCHMID (IF, 64 <1959> 287-98,65 <1960> 297-302), and W. P. LEHMANN (JAOS, 78 <1958> 212-13), and the present reviewer does not feel competent to judge the merits or otherwise of their criticisms. However, one wishes that H' had taken note of such standard article as that of F. EDGERTON"S "The Indo-Europen Semivowels", Language, 19 <1943> 83-124, and should not have altogether avoided an account of the laryngeal theory which has occupied the attention of Indo-Europeanists for about 50 years. The 2 vol.s should, on the whole, be found very useful not only by the students of Sanskrit but also by those who have made some advance in the Indo-European studies. They give a good historical background of Sanskrit with an equally good Bibliography. Much has been written in the field of Sanskritic and Indo-European studies during the last 50 years, and H's is an admirable attempt to give a compact account of these studies. However, a few remarks, not involving repetitions from earlier reviews and mostly outside the field of linguistic explanation, may be permitted. p. 73: It is extremely doubtful whether we can draw the conclusion H's has drawn from the occurrence of the Asokan inscriptions in Mysore (-ein Beweis dafur, dass zu dieser Zeit arische Sprache und Gesittung dort bereits ein bedeutendes Gewicht besassen). It is also not correct to speak of the Aryan culture as 'superior to that of the Dravidians. p. 104 : Against the derivation of puja < * prnca (cf. KUIPER p. 433-4) also S. M. KATRE, Two Lectures on Linguistics, p. 32--3 (Agra University, Agra 1959). It may be added for the sake of information that the English translation of P. THIEME's views on puja has appeared in the Journal of Oriental Research, Madras, 27 <1960> 1--16. p. 105 : A study of inscriptional Sanskrit is likely to show many Sanskritizations of Persian and modern Indian words e.g. sphurannama for Persian firman, ghranaka for modern Indian ghana, etc. (oil-mill). p. 108 : H' translates ojah samasa-bhuyastvam, etad gadyasya jivitam = "der haoufige Gebrauch von Kompositis verleiht der Prosa Kraft, er ist ihr Lebensnerv". Strictly speaking this does not appear to be correct. The translation should be: "Force or Strength consists in the abundance of compounds; this is the soul of prose." Madhu Vidya/608

Page #634
p. 109 : While speaking about the popular form of the epic Sanskrit H'tries to explain the fact that the Sutas, who did not belong to the Brahmanical circles, did not strictly follow the rules of Sanskrit grammar, because they composed and sang for the princes, the kings, the nobles and the rich, "also fur die weltlichen Staonde."This, however, stands in contradiction to what is said earlier on p. 106--7. There H' regards the use of Sanskrit in dramas by kings, Brahmins, Ksatriyas, etc., to reflect the actual conditions. These male characters in the drama speak classical Sanskrit and not epic Sanskrit. p. 118: In the enumeration of the modern Indian languages, H' mentions such dialects as Bhili, Khandesi among the most important ones. On the other hand he puts Hindustani, Urdu into brackets, which is hardly proper. p. 120 : The linear development of the Indian languages in the form of a family-tree as shown on p. 120 is over - simplified. The correct picture, when worked out somewhat on the lines of the development of the Romance languages given by. ROBERT A. HALL Jr. in Language 26 (1950) 6--27, will certainly look much more complex. p. 136 : About a different view regarding the meaning of upanisad (= connection), see L. RENOU, C. Kunhan Raja Presentation Vol., p. 55 ff., Madras 1946. p. 140 : If the reviewer understands correctly, it is not very complimentary to call the Bhandarkar Institute's edition of the Mahabharata 'Eine Art Kritische Ausgabe'. The principles evolved by V. S. SUKHATNKAR for the critical edition of the epic were upheld by many eminent scholars. H. LUDERS regarded them as deserving the highest possible praise". There were no doubt a few who were critical. But to brush aside on that account the great attempt in the above words is not fair. The Baroda edition of the Ramayan under preparation is referred to by H' simply as 'Eine Neuausgabe'(p. 142), whereas it should correctly be described as "a new critical edition". p. 165 : To name north India as Hindustan and to mention it separately by the side of Bihar and Benares is not good. p. 193 : One does not understand why the Devanagari characters for du and du are mentioned as "besonders zu merken". p. 194: In Devanagari, in the ligatures with na, there is really no loss of the vertical line of na. When the Devanagari ligatures are properly understood many apparent exceptions will disappear. In the ligature dhna, e.g., what is really lost is the vertical line of dha and not of na. p. 197: In the Sonstige Schriftzeichen the a. could have included the use of the numeral three after the long vowel to indicate pluti. II : What is intended to covey when the early Indian grammarians are called "indische Nationalgrammatiker" is not clear. Madhu Vidya/609

Page #635
II 5: When the ablaut in the suffix is intended to be pointed out, it would be better to show it by writing the examples as pi-tar-am, pi-tr-ed, pi-ta instead of pitaram etc. II 12: When the dat. sg. ending -e<* -ei or * -ai we cannot be sure that Skt. vace (for *vake) results form the analogical generalisation as stated on I 293. If the ending in this respect was * -ei, palatalization is what is expected. II 356: Under denominatives, the a. observes that the suffix sometimes appears as -iya- oriya- and gives two meanings for putriyati (or putriyati) "er wunscht einen Sohn" or "er behandelt wie einen Sohn." According to Panini, however, we have two different suffixes to express these two meanings kyac (P. 3, 1.8) for the former and kyan (P.3, 1.11) for the latter. Bibliographical Notes: As observed by Lehmann (p.212) the instructors. will be able to fill the gaps in the bibliography. However, a few notes, not intended to be exhaustive and mostly relating to the Indian publications, may be permitted here. p. 11 The correct title is: Annals of the Bhandarkar Oriental Research Institute. p. 13: Epigraphia Indica is not printed in Calcutta and published from Delhi. p. 13: Indian Linguistics is now published from Calcutta-Poona (Linguistic Society and not Linguistics Society). p.13 In the section y one misses some important Journals like the Journal of the University of Bombay (since 1932), New Indian Antiquary (1938-47), Bulletin of the Deccan College Research Institute (since 1939). p. 28: J. Jolly's work on medicine is translated into English (Indian Medicine) and published with supplementary notes by C. G. KASHIKAR, P. 1951. p. 30-1 M. R. KALE'S Higher Sanskrit Grammar appears twice (sections 8 and 8 a) with slight variations. It is also recently published (1960) from Varanasi. p. 32: V. S. APTE's The Practical Sanskrit-English Dictionary, revised and enlarged (in 3 parts) by P. K. GODE and C. G. KARVE, Poona 1957-59. p. 32 ff: Under Dictionaries now add also ARYENDRA SHARMA'S Beitrage zur vedischen Lexicographie, Munchen 1959, and K. V. ABHYANKAR'S A Dictionary of Sanskrit Grammar, Baroda 1961. p. 35 In the senction 10a where linguistic monographs are brought together one wonders why H. LUDERS: Varuna (now also Vol. II) and P. THIEME : Der Fremdling im Rgveda, both undoubtedly excellent works in themselves, are not mentioned. p. 112 In the Literatur for Pali one misses H. LUDERS: Beobachtungen uber die Sprache des buddhistischen Urkanons (Berlin 1954), although on p. 113 his view is referred to. Madhu Vidya/610

Page #636
p. 122 : Among the probable Dravidian loan words in Sanskrit H'gives toya. This has been recently shown to be a probable loan from Tibeto-Burman (N. M. SEN, Indian Linguistics, 17<1957> 50). p. 126: J. BLOCH Structure grammaticale des langues dravidiennes. An authorised English translation by R. G. HARSHE appeared from the Deccan College, Poona, in 1954. p. 128 : M. WINTERNITZ: A History of Indian Literature, Vol. III, Pt. 1 appeared from Calcutta in 1959. p. 132: Among the editions of the aegveda, the one (with the commentary of Sayana) published by the Vaidika Samsodhana Mandala, Poona, 1933-51(4 vol.s 5th vol. Indices), certainly deserves mention. p. 135: Does nct M.BLOOMFIELD's Hymns of the Atharva-Veda (SBE, 42, Oxford, 1897) merit a mention? p. 142:N. A. GORE: A Bibliography of the Ramayana, Poona 1943. p. 145-6 : Naisadhacarita, English translation with Notes etc., by K. K. HANDIQUI, published by the Deccan College, Poona 1956. p. 148-9: An English translation of the Mudraraksasa by R.S. PANDIT, Bombay 1944. P. 157-8 :In the information about the Dharmasastras a regrettable lapse is P. V. KANE's monumental work History of Dharmasatra, Vol.s 1-5 (Pt 1), Poona 1930-58. p. 159: R. P. KANGLE: A critical edition of the Arthasastra with glossary, Pt. I, Unviersity of Bombay, has recently appeared (1960). p. 160: R. S. PANDIT's English translation of the Rajatarangini, Allahabad 1935. Misprints: It certainly does great cerdit to the a. if the reviewer has to say that the misprints are not many. A few like Devanagari for-nagari (IX), Surasena for Su-(113), Assami for Asami (118), atavi for atavi(112) des uns sonst (unbekannten), Kongs for des (uns...)Konigs (147), duhitaram for duhitaram (207), IE nonce without (245) and once with (246) accent, Im for im (II 5) for -navant- see SS 621 and not s 621 a (II 189), ar for ar (II 191), runadhmi for runadhmi (II 237) are not at all serious. Deccan College, Poona (India) M.A. MEHENDALE Madhu Vidya/611

Page #637
Mitra and Aryaman by Paul Thieme. Transactions of the Connecticut Academy of Arts and Sciences, Vol. 41, pp. 1-96. 1957. Yale University Press, New Haven, Conn., U.S.A. Thieme observes: "The two Vedic gods Mitra and Aryaman--for the Rigvedic poet the two most important figures amongst the Adityas after Varuna-have challenged research again and again." (p. 5). The problem of Mitra and Varuna was discussed by G. Dumezil in his book Mitra Varuna (Paris, 1940 and 1948). He afterwards took up the problem of Aryaman in his book Le troisieme Souverain (Paris, 1949). Before writing the present book, Thieme himself had dealt with Aryaman in his Der Fremdling im Rigveda (Leipzig, 1938) and also in ZDMG 95. 219-221 (1941). As is well known, two very important volumes by Luders on God Varuna have appeared in recent times (Goettingen, 1952, 1959). Dumezil, apparently relying on certain identifications in the Brahmanas, has come to the conclusion that there is a sort of opposition between Mitra and Varuna. Mitra is "le souverain sous son aspect raisonnan, clair, regle, calme, bienveillant, sacerdotal", Varuna, on the other hand, is "le souverain sous son aspect assaillant, sombre, inspire, violent, terrible, guerrier. ....." In fact, to Dumezil, a formula suggests itself : Mitra is brahman, Varuna is the king of the Gandharvas. Thieme does not agree with this view. Dumezil himself admits that a large majority of the texts studied by him 'do not permit distinguishing Mitra and Varuna by clear features'. Thieme therefore rightly asks: "If there is not even a distinction, how should there be an opposition?" (p. 9). In the present book Thieme gives his opinion on the views held by Dumezil, and also asserts his own views, which he had already expressed before, regarding these deities. He firmly believes that God Mitra is the personification of an ethical concept 'contract, God Varuna that of 'true speech', and God Aryaman that of 'hospitality'. Thieme has set forth his arguments so cogently that it is difficult not to agree with him. As for the method, he says : "Everybody, of course also Dume zil, is well within his rights when 'throwing an hypothesis'. The value of such hypothesis, however, depends on the accuracy of the experiments that are meant to prove it. If our experiments are arranged in such a way as to be in contradiction to known facts and involve further assumptions which have no justification apart from making our primary assumption possible, our hypothesis remains a preconceived notion .... A hypothesis has sense only when it restricts our imagination and forces us MadhuVidya/612

Page #638
to try going into a certain direction. It can be disproved when this leads to absurd consequences" (pp. 16 and 17). While one may fully agree with Thieme regarding the method sugge. sted by him and with the results arrived at by him by the application of this method, it seems possible to offer a few suggestions with regard to the interpretation of certain Vedic passages cited by him. It is hoped that these suggestions will lead to a better understanding of these passages. i. On pp. 39 and 42, Thieme cites RV 3.59.1a and b: mitro janan yatayati brutanah mitito dadhara pithivi'm uta dyam and translates, "Contract, when named, makes peoples array (arrange) themselves with regard to each other). Contract has earth and heaven in keeping." While commenting on b Thieme says that it refers to the cosmic function of Mitra and that it is contrasted with his role in human society which is spoken of in a. (Also cf. p. 58. Mitra keeps not only earth but also heaven). One is, however, inclined to feel that in the above passage, b is not intended to contrast with a. On the other hand, it gives a concrete example of the function of Mitra expressed by the term yatayati. Mitra arranges men (a), as he has before arranged earth and heaven (b). The keeping of earth and heaven is looked upon as an arrangement between them. It may be noted that in the RV 5.72. 2b, where the form yatayajjana is used, it is accompanied by dharmana in the instrumental. From this, one is led to suppose that dharmana is to be understood also in our passage. When we do this--mitro janan (dharmana) yatayati-, the use of dudhara in b is easily explained, and further, the interpretation suggested above is confirmed. 2. On p. 44, Thieme cites RV 3.59.2, the last quarter of which-nainam anho asnoty antito ma durat-he translates, "narrowness (anxiety) does not reach him neither from near, nor from afar." From his commentary on this verse it appears that Thieme takes amhas (anxiety) to refer to the possible lawless attacks in peace. It is, however, not clear why the literal meaning of amhas 'narrowness' is considered inadequate by Thieme. The line assures the one who keeps his contractual vow (yas ta aditya siksati vratena) that he will never be in the danger of facing narrowness of space. He will always have wide expanses at his disposal. This becomes quite clear from the use of varimann & prthivydh 'on the width of the earth' in the next verse (RV.3.59.3). 3. On p. 49, Thieme cites RV 3.59.4 : ayam mitro namasyasusevo Taja suksato ajanista vedhih which he translates as, "Contract, (who is) worthy of (sacrificial) reverence, benevolent, a king of good rulership has been born (now) as this leader (?)". Thieme thus connects ayam with vedhah which is at a considerable distance from it. In his commentary on this verse, Thieme says that ayam ajanista should refer to fire because fire is occasionally present Madhu Vidya/613

Page #639
at the time of concluding a contract and because fire is elsewhere identified with mitra. Linguistically, as pointed out by Thieme, ayam must refer to something that is before the eye, and ajanista, as an aorist form, must refer to a happening that has just taken place. But if we accept Thieme's inter pretation, this verse, which occurs in 'the only hymn that is dedicated to Mitra in its entirety' (p. 39), would refer to Agni and not to Mitra. This is not quite happy. Thieme tries to add to the force of his argument by pointing out that the last two quarters of this verse-tasya vaya sumatau yajniyasdpi bhadre saumanase syama-are identical with the RV 3.1.21 cd which are addressed to Agni. But this evidence loses any force it has when one nocices that in the RV 3.59 itself in verse 3 the seer says vayar mitras ya sumatau syama. It is, however, not necessary to interpret this verse as referring to fire in order to justify the use of ayam and ajanista. It can very well refer to the contract that has been just concluded and has been identified with god Contract. This will satisfy all requirements of the linguistic usage, bring abou: a natural connection between ayam and mitrah which s:and close to cach other in the verse, and will not force us to assign a verse in the Mitra hymn to Agni. 4. On pp. 73 f., Thieme shows how in a number of passages in the RV it makes perfect sense when ari is translated by (1) enemy, and (2) guest. On p. 74, however, he says that there are certain passages in which we mighl translate ari by 'host'. In support, he cites RV. 3.43.2 : d yahi purvi'r ati carsani't am arya asisa upa no haribhyam 'Come here, across many peoples! Here, to the blessings of the host, to us with thy horses!' In his Fremdling (p. 27), however, Thieme interprets ari in this verse as 'stranger' and translates arya zsisah by "den Segenswunschen fur den Fremdling". He adds, by way of explanation, that a stranger, when received with blessings, becomes a friend. It should be possible to agree with Thieme in interpreting in the present verse ari as 'stranger', and not as 'hosi', but it would be difficult to agree with his translation of arya asisah "blessings for the stranger' (arye dat.). We better take atyah as gen. sg. (as in the RV. 8.54.7) and translate the expression as 'the blessings of the stranger'. The stranger, in all probability, is the one who has agreed to act as priest at a sacrifice (cf. the very first and other verses in this hymn which speak of sacrifice). It is to these blessings of the priest given to the sacrificer that Indra is invited to come across many peoples. Madhu Vidya/614

Page #640
5. On p. 75, Thieme cites RV 10.27.8 gdvo yavam prayuta atyo aksan. He connects aryah with yavam and translates, "The cows, let loose, ate the stranger's barely". But it would be better to construe aryah with gdvah. The stranger's cows have eaten the native's barley. The calls, to which reference is made in c (hava id aryah), are not those raised by the native, the possessor of the barley (thus, Thieme, Fremdling, p. 12), but those raised by the stranger, the possessor of the cows. He, with his calls, has let loose the cows in the barley field. The owner of the barley field is the suapati, referred to in d, and he wants to know how long he will have to tolerate the nuisances caused by the stranger. 6. A few omissions may be noted in the end. On p. 62, Thieme cites RV 10.89.9. While translating it, he has left out the word durevah. On p. 83, he cites RV 2.1.4. While giving the text, he gives bhavasi in place of satpatih in the third quarter (probably under the influence of RV 5.3.2a). While translating the line, however, he has translated it as if both the words bhavasi and satpatih were present in the text. In d in the same verse he has left out untranslated the word deva. Madhu Vidya/615 M. A. Mehendale

Page #641
PHILOLOGY CARL] G[USTAV] HARTMAN: Emphasizing and Connecting Particles in the 13 Principal Upanishads (Annales Academiae Scientiarum Fannicae; ser. B, tom. 143 2).gr.8deg 180 P. 3 tables. Suomalainen Tiedeakatemia, Helsinki 1966. Fmk 14,50. In the oldest attested stage of Sanskrit, the use of particles is rather limited. They become frequent in certain parts of the Brahmanas and in the Upanishads. Their use in the Upanishads is moreover marked by the abundant use of combinations in which the particles occur. This Upanishadic peculiarity concerning combinations "may be ascribed to the educational character of the texts. The teachers regarded their instructions as important and therefore worthy of being strongly emphasized" (p. 123). H' discusses the use of 13 particles in the 13 principal Upanishads (these are those translated by Hume). The particles dealt with are: api, u, uta, eva, kila, khalu, ca, nu, vava, vai, sma, svid, ha. He has chosen to call the particles "emphasizing", rather than "emphatic", because the former word underlines their function of laying stress upon other words (p. 14). There is a good deal of irregularity in the use of particles in the Upanishads. some paragraphs being practically free from particles while others have a frequent use of them. To the a. (author) this irregularity seems to be as much due to the difference in styles (speculative, descriptive as opposed to narrative) as to the fact that parts of the Upanishads have been written and compiled at different periods and by different authors (p. 165)-- It is often difficult to determine the exact significance of a particle. Notwithstanding this difficulty, the a. has admirably executed his task. The book has 5 chapters, followed by a summary. While the 1" chap. deals with the particles as they occur singly, the 2nd chap. deals with their various combinations; the 3d chap. points out the relation between the particles and the Upanishadic style; the 4th chap. discusses the particles historically and tries to determine the position of the Upanishadic language; the 5th chap. is devoted to the theories of the Upanishadic particles. The a. holds the view that particles are not to be looked upon as normal words but rather as stylistical signs marking different stages in the narrative or underlining important facts in philosophical instructions. They also serve the purpose of punctuation marks like comma, semicolon etc. On p. 23 the a. says that in a few cases yady api is not used concessively (i.e. meaning "even if", "even though"), but as conditional and hence translated simply by "if." By way of illustration he cites Chhandogya 5,24,4. The a.'s judgement Madhu Vidya/616

Page #642
in this case is based on the translation of S. Radhakrishnan, but he does not say why that translation is justified. To the reviewer it seems that in this passage yady api has the meaning "even if, even when." Moreover, Radhakrishnan's translation of this passage seems to contain another slight inaccuracy which is based on Sankara's interpretation of the passage (Hume's translation of the passage is similar to that of Radhakrishnan). The passage runs as : tasmad u haivamvid yady api candalayocchistam prayacchet, atmani haivasya tad vaisvanare hutam syad iti = "Therefore if one who known this should offer the remnant of his food to a Candala, it would be offered in his Universal Self" (RADHAKRISHNAN). In this translation usya is taken to refer to Candalasya and is construed with atmani vaisvanare (cf. SANKARA atmani haivasya candaladehasthe vaisvanare). But in keeping with the style of the Upanishads, asya should be taken to refer to evamvid and construed with hutam,"that (act) of his [i.e. the one who knows] will be an offering in the Universal Self." For a similar use of the demonstrative pronoun going with preceding vidvan one may cite from the same section of the Upanishad, atha ya etad evam vidvan agnihotram juhoti, tasya sarvesu lokesu sarvesu bhutesu sarvesu atmasu hutam bhavati. Instead of tasya, even asya is used in the same section, ... evam hasya sarve papmanah praduyante, ya etad evam vidvan agnihotram juhoti. On p. 30 the a. observes that the use of the particle u becomes more restricted in later literature. "In the classical language it occurs only after atha, na and kim, often quite superfluously." It would have been better if he had mentioned the use of u after yatha in the Nirukta in such expressions as yatho etat or yatho hi nu vaitat (1,14 etc.) which are used to state the view of the purvapaksa before refuting it. Similary on p.40 f, the a. should have noted the uses of kila, either singly or in combination with na and nanu given in the Nirukta 1,5: kileti vidyaprakarse evam kiletilathapi na nanu ity etabhyam samprayujyate 'nuprstelna kilaivam, nanu kilaivam."[The particle) kila [is used) to show excellence of knowledge", as in **[it was) in fact like this". Moreover, [kila) is used in combination with na and nanu when questioned again (regarding the veracity of the statement) "not indeed like this", 'yes, indeed, like this." Deccan College, Poona M.A. MEHENDALE Madhu Vidya/617

Page #643
Janakiharanam by Kumaradasa. Ed. by S. PARANAVITANA and C. E. GODAKUMBURA, published by Shri Lanka Sahitya Mandalaya (Ceylon Academy of Letters), 1967, pp. i-lxxii, 1-401. This is an excellent and a complete critical edition of the famousMaha kavya by the Sinhalese poet Kumaradasa who, according to one tradition, was contemporary of Kalidasa. Apparently this work enjoyed much popularity at one time, since it has been cited in many anthologies and other grammatical and literary works both in India and Ceylon. But in spite of this, the work had remained practically unknown for quite a long time. Even as late as 1947, S. K. DE while writing his History of Sanskrit Literature had to say, "The incomplete and not wholly satisfactory recovery of Kumaradasa's work makes it difficult to make a proper estimate " (p. 187). The Janakiharaya was first noticed by AUFRECHT in his edition of the Unadisutraurti (1859). In 1891. Sthavira Dharmarama published in Ceylon for the first time the text of the poem upto verse 22 of canto XV (the proper extent of the work, as we now know, being 20 cantos). This text, however, was based not on any manuscript of the poem itself-no manuscript having been known to exist in those days-but was reconstructed from Sanne, a wordfor-word Sinhalese version of the poem made in the twelfth century. NANDARGIKAR'S edition of the first ten cantos, based on the above reconstructed text, was the first Indian Devanagari edition of the poem published in 1907. A complete manuscript of the original text of the Janakiharana was discovered by M. Ramakrishna KAVI and S, K. Ramanatha SASTRIAR in Malabar in 1920. A copy of this manuscript is now preserved in the Government Oriental Manuscripts Library, Madras. Subsequently a few more manuscripts of the poem, containing the complete text or a part of it, were also found. The present edition of the complete text in 20 cantos is chiefly based on the Madras manuscript with some help derived from the other manuscripts and the reconstructed version of the Sinhalese Sanne. Apparently the following remark made by S. K. DE in his Hist. of Skt.Lit. (P. 186, fn. 2) about the Madras manuscript does not seem to be justified: "The Madras MS existing in the Govt. Orient, MSS Library contains twenty cantos, but it is a very corrupt transcript of an unknown original, and it is not known-how far it is derived ultimately from the Sinhalese Sanna." On the other hand, as the editors have pointed out, the text now made available clearly shows that almost all the blemishes, like the use of khalu and iva at the beginning of a verse, which the critics had noticed in the Janakiharaya belong not to the original text of the poem but to its redactions. The usefulness of the published edition has beer. considerably increased by one of the editors contributing a number of indexes and some text-critical notes. The information given about the poet (pp. li- xix) is very valuable. It refers to a tradition current in Ceylon in the 15th century which ascribed the authorship of the Madhu Vidya/618

Page #644
Biven. Janakiharana to a Sinhalese king named Kumaradasa. The editor critically examines the available evidence and comes to the concluion that the author of the poem was not a king himself but a member of a Sinhalese royal family. The explanation of Kumaradasa's epithet atisayabhuta given on pp. lxvii-Ixviji is not likely to find favour. According to the editor, the epithet is to be explained as based on old Sinhalese *atisa ( =Sk. adhisa ) +aya (Skt. arya) meaning one who has become heir-apparent or heir-presumptive." On p. xxx, the editor has listed some rare words (or rather, in some cases words with rare meanings ). It would have been better if the meanings of these words had also been given. On p. 282, while commenting on II. 52, the editor translates ajahad ...... dhama as "abandoned its ...... strength". Perhaps, it would be better to render it as "lost ...... (its) lustre''. This rendering will fit in better with the words tamahsthanam samasadya "having encountered the abode of darkness" occurring in this verse. P. 283, VIII. 39: The editor's reason for adopting the reading caksusa (instr.) against the manuscript evidence which gives caksusah (abi.) has not been stated. It is also not clear why he refers to WHITNEY 414 end (p. 155 ). P. 286, XIII 14 : The editor unjustifiably finds fault with Sanne which quite correctly understands jahati as a pl. form agreeing with mrgayositah (also pl.). The editor wrongly looks upon jahati as 3rd per. sg. (the sg. form actually being jahati) and then offers an unlikely translation of the verse. P. 288, XV 62: The editor mentions that the reading syannare which is given by the MSS. can be understood either as syam nare or syal nare. He prefers the foriner; but actually the latter has to be preferred so that syat can be easily construed with astha in (d). It has nothing to do, as supposed by the editor, with manye at (b). The editors should have also accepted the reading -drsi ( loc. sg.), going with nare and vanare, which is given by two manuscripts (the reading accepted by the editors is- dzsi) P. 288, XVI. 45 : The editors have done the right thing in allowing the text in (d) to remain as it is and not amend it to vakyam uce following a possibility suggested by them. . In the end a few misprints, not included in the Errata (p. i), may be pointed out': p. xxx, 1. 5 (from below ) read varllika for vartika; p. xxxi, I. 3, read sarasana for sarasana; p. 223, XX. 57 (d), read prayat for prayat; p. 283, 1. 3 (from lelow), read IX. 39 for IX. 31; p. 282, IV. 2 and p. 284,1. 2 (from below), read Rsya-for Rsyaand Rsya ; p. 282. IV. 20, read, 'while L. and M. have' for ...... has'; p. 286. 1. 15, read mugdhadhiyah for mugha-. At a few places, e g. in III 49 (p. 31), va is printed as ba as in kalahamsasabam. 1 It is not clear why in this account on p. lii Kalidasa has been relerred to specially as the author of the Raghuvamsa and the Meghadura Madhu Vidya/619

Page #645
Pandit SHIVANARAYAN SHASTRI. Nirukta-Mimamsa, Varanasi; Delhi: Indological Book House, 1969. 476 pp. Price Rs. 25. (In Hindi). Reviewed by M. A. MEHENDALE, Deccan College, Poona This is an important and very useful publication on Yaksa's Nirukta after The Etymologies of Yaska by Siddeshwar Varma (1953). As the title of Varma's book indicates, his objective was limited. The present book, on the other hand, is more comprehensive in its nature. It discusses at great length, giving much valuable information at each point, many subjects relating to the text, its author, and its contents. The author has, for example, discussed such questions as the authorship of the Nighantu, the date of Yaska, the nature of etymology, the nature of Vedic deities, and many others which arise in the study of the Nirukta. At the end of the book, the author has contributed three chapters dealing with the state of society as reflected in the Nirukta, the philosophy of Yaska, and Yaska's contribution to grammar and poetics. An enquiry into the nature of words and their analysis are fascinating subjects. It can be said that they attracted the attention of scholars early, and are well reflected in the Brahmana texts. Yaska's Nirukta is no doubt an admirable attempt in this direction. His hypothesis that all nouns without exception are derived from verbs has compelled him to set himself the task of deriving a set number of words from a set number of verbs, without taking. into account such facts that not all the words in the vocabulary of a language are the result of inheritance and that certain words in the course of history change their form so radically that it would be almost impossible to make any guess about their origin without the help obtained from outside the language. He has therefore at times been drawn into making impossible derivations. The author has no doubt given expression to a balanced opinion (p. 221) with regard to Yaska's etymologies. But at times he has expressed his anger in very strong terms for those who have criticised Yaska which cannot be said to be appropriate in a scholarly work like this.' The author believes that Yaska is the author of the Nighantu as well as the Nirukta (p. 29 and pp. 292-93, fn. 3). It is true that not all the arguments put forward by those who believe that Nighantu was composed by some other author before Yaska are convincing. But at the same time it is not easy to convince someone that Yaska was the author of the Nighantu. The beginning of the Nirukta, and especially the second sentence, tam imam samamnayam 1. It is perhaps all to the good that the author has seen only Rajavade's edition of the Nirukta (Yaska 1940) and not his earlier edition of the entire Nirukta with Marathi translation and notes (Yaska 1935). Madhu Vidya/620

Page #646
REVIEWS AND NOTICES 59 nii ghantava il y acaksale does give an impression that Yaska was not the author of the Nighantu. It indicates that the activity of compiling the samamnaya was completed in the past and that people, in the course of time, had come to call the samamnaya by the name nighantus. (Cf. Durga : nitudha hiyan etasmin chabdasamudaye samjnety abhiprayah). On the other hand it would be possible to reconcile Yaska's statement in 7.3 -- yat tie sarvijnanabhutari syut pradhanyasluti tat samamane ---- with the above conclusion by restricting its scope to the daivatakunda of the Nighantu and by not making it applicabie to the first two kandas as well. The author has discussed at some length the question of Yaska's date. It is indeed difficult to agree with him when he considers Yaska to be not only older than Pinini but to have lived in the age preceding the Mahabharala war (p. 79). Although the author has spoken very disparagingly about those who have at times agreed with Western scholars, the reviewer cannot bul recommend to the author to weigh carefully the arguments put forward by P. Thicme (1935, 1958) in support of his view that Yaska is posterior to Panini. On page 10, the author says that the word samamriya etymologically means "traditionally handed down wrillen document" ("parampara prapta lekha"). But how can this be true when we know that the ancient Sanskrit texts were preserved for a long time only in oral tradition ? On p. 65, the author lists certain etymologies for which he feels that Yiiska has relied merely on the similarity of meaning without paying any regard to the similarity of sound. But can we really say this when under this head the author lists such clymologies as putra < purut + Vira, alatina Page #647
60 and says that it indicates that these roots had not gone out of use in the days of Ysaka and hence he could not be considered to be very far removed from the Vedic age. One may point out that the use of the Vedic roots by Yiska for the derivation of the Vedic words (bhara 'battle') is quite normal. His using them further for the derivation of words current in later times (broty. garbha) only exemplifies his dictum athapi naigamebhyo (dhatubhyo) bhasikah (kyto bhasyante) 2.2). REVIEWS AND NOTICES The author's conclusion (pp. 432-434) that in the days of Yaska the vowel was pronounced almost like u is not likely to carry conviction. The examples listed by him for this do not point in that direction. To take only one case, if Yaska derives muhurta muhurta; it only means that according to him. muhurta muhur + Vr. The author often cites the views of other scholars, ancient and modern, while discussing many difficult points in the text. At times, however, he has failed to do so. A few of these omissions may be pointed out. While referring to the view of Audumbarayana (p. 109, 113), it would have been good to look to the interpretation of this passage by Brough (1952). Similarly while discussing the definitions of akhyata (p. 120), the author could have referred to K. Kunjunni Raja's article (1957). Some of the words and passages from the Nirukta discussed by the author have also been dealt with by the Reviewer (Mehendale 1965). On pravahhitam (p. 291, fn. 1) the author inadvertently forgets to mention the view of Skanda-Mahesvara on this word. REFERENCES: Bough, John 1952. Audumbarayana's theory of language. BSOAS 14.73-7. Mchendale, M. A. 1965. Nirukta notes I. Poona: Deccan College. 1968. Some aspects of Indo-Aryan linguistics. Bombay: U. of Bombay. Includes summary of Thieme 1935. 1958 (on pp. 4-7). Renou, Louis 1942. Terminologie grammaticale du Sanskrite. Paris. K. K. Raja 1957. Yaska's definitions of the 'verb' and the 'noun. rescarch. Pp. 86-88. Thieme. Paul 1935. Zur Datierung des Panini. ZDMG 89. *23-*24* (Published 1936). See Mehendale 1968. -1958. Review of Renou 1942. Gottingische Gelehrte Anzeigen 46-9. In German. See Mehendale 1968. Varma, Siddeshwar 1953. The Etymologies of Yaska. Hoshiarpur, Ya-ka 1935. Nirukta. Ed. Marathi tr. and notes. Rajavade, Vishwanath Kashinath, 1940. Nirukta. Ed. Rajavade, Vishwanath Kashinath. With notes, Poona : BORI. Received 15 April 1970] Published 1957). Annals of oriental Madhu Vidya/622

Page #648
118 YAMUNA KACHRU MCDERMOTT, A. C. Senape (ed.). An Eleventh-century Buddhist logic of 'exists': Ratnakiriti's Ksanabhangasiddhih vyatirekatmika. With Introduction, English translation and Notes. Foundations of language supplementary Series, 11. Dordrecht, Netherlands : Reidel, 1969. X + 88 pp. Reviewed by M. A. MEHENDALE, Deccan College, Poona The book under review contains a Romanized transliterated text, which is principally the same as found in Ra!nakitlinibandhavaii (ed. by A. Thakur, Patna, 1957). Minor corrections and deviations from this text have been indicated in the foot notes. In the Introduction, the editor gives information about the author and his philosophy and attempts a comparison of Ratnakirti with other 'flux' philosophers of the west. It would not be possible to understand the text with the help of the editor's translation alone. She has therefore added very useful notes to explain the logical and epistemological problems dealt with in this text. Ratnakirti, the Buddhist philosopher of the early 11th century A.D., flourished at the University of Vikramasila. He was a member of the Yogacara-Vijnanavada school of late Buddhist philosophy. Within this tradition, Ratnakirti belonged to the sub-school of interpretation of Dharmakirti's (7th century A.D.) Pramanavarttika which is "a highly original recasting of the basic tenets of the great Buddhist logician Dignaga (ca. 480 A.D.) into a system of logic and epistemology which became the point of departure for all subsequent developments in Buddhist logic (p. 2, fn. 8). Ratnakirti defends the theory of nonmomentary reality which is expressed in terms "whatever exists is momentary" (yat sat tat ksanikam). The anvaya version of this theory has been established by Ratnakirti in his Ksanabhangasiddhih anvayatmika. What he does in the present book is to establish the contrapositive of it, viz., "whatever is nonmomentary does not exist". Received 22 June 1970 ] Madhu Vidya/623

Page #649
Reviews 241 SOME ASPECTS OF APPLIED LINGUISTICS, by A. M. Ghatage, Shivaji University Extension Lecture Series-1. Shivaji University, Kolhapur, pp. 1-54, Price Rs. 5.00, 1970 This book includes three lectures delivered by Dr. A. M. Ghatage in the Extension Lecture Series organized by the Shivaji University, Kolhapur, in January 1969. The topics covered by the three lectures are : (1) Linguistics-theoretical and applied; (2) Linguistics and Language Competence ; (3) Speech-Teclinology. The lectures are intended to be introductory in nature and are meant for au audience not of expert linguists, but of scholars forming a wider circle of the academic world. Dr. Ghatage's plan seeing to be first to introduce to the audience the subject of linguistics and then explain to them the relationship this science bas to some of the practical problems facing the country today. He, therefore, tells his listeners in his first lecture what linguistics is and what linguistics is not, and then points out to them where precisely linguistics differs from the traditional grammars. In this context he emphasizes that a language has to be looked upon essentially as a system used by a given cominunity for the prirposes of inter-communication. In the second lecture, Dr. Curtage deals with some specific aspects of upplier linguistics like the scope of lunyunye teaching and the number of languages to be tanghat, the medium of instruction, and the use of a foreign language like English in a country like India. He also refers to contrastive linguistics which is com; aratively a recent development in linguistics. Dr. Chatnge points out in his discussion what useful part linguistics can play in tuckling the above problems. Some of the observations made in this lecture-e. "Unless there is * strong motivatiou in the forin of its use ia que form or the other, tbe learning of a language is going to be a mere waste of time and energy and it is likely to be quickly forgotten" (p. 29 )-deserve the notice of those engaged in policy-making. lo bis third lecture, Dr. Ghatage deals with some technical aspects of the study of speech. He describes the functioning of various instruments like the Kymograpb and the Spectograph and the uses to which a language laboratory can be put for teaching a language. He also refers to the use of visible speech for teaching deaf persons and the problems involved in machine translations. Althogh Dr. Ghatage intended to make his lectures not too technical, le could hardly aroid doing this in his third lecture. The reviewer confesses his inability to follow some parts of this lecture. 31 [ Anuals, B. O. R. I. J Madhu Vidya/624

Page #650
Annals of the Bhandarkar Oriental Research Institute On p. 6, the lecturer gives the two words, Germau lass 'allow' and English lass, as examples of same words (meaning thereby words similar in sound) meaning different things. But the words are similar only to the eye, and not to the ear. On p. 8, the lecturer referes to duality by which he probably means the duality of patterning in language as a system; hut as he does not explain it, this point is not likely to be understood by many. On p. 10, Dr. Ghatage says that a language requires for its proper functioning only a limited number of sound units which should be discretely different from each other, easily produced and quickly identified. But is it not true that the sounds of any language are thought to be easy by the native speakers of that language? On p. 36, Dr. Ghatage says that by the use of Kymograph it is possible to record the vibratory actions of different vocal organs like the vocal cords, mouth, lips, nose etc. But shall we say that in the production of speech sounds we have the vibratory action of mouth and nose as well? 242 Dr. A. G. Pawar, the Vice-Chancellor of the University has observed in his Foreword that the aim of founding the Extension Lecture series was to create "genuine interest among the University students for higher studies, acquaint the University Research workers with the recent thoughts and developments in different subjects of science and humanities and establish a constant link between the University and the intelligentsia of the region". It is not to be doubted that these objectives have been fully met. by the lectures under review. M. A. Melendale BALTS AND ARYANS, by Suniti Kumar Chatterji, Indian Institute of Advanced Study Simla, 1968, pp. 1-178 and Plates I-XIX, Price Rs. 20/- (PS2 58., $8.00) In this book Dr. Chatterji gives a lucid account of all aspects of the relationship between the forefathers of the speakers of the Baltic Languages and those of the speakers of the Indo-Aryan languages. He describes what may be called the character of the Balts aud shows how closely their culture and language are related to those of the Vedic Aryans. He elaborates on the beauty of the dainas, the national poetry of the Balts, which is preserved by them over the centuries by oral tradition. These songs are supposed to have preserved glimpses of preChristian life and culture-of a time when joy still walked over the earth'. The Baltic and the Indo-Iranian, together with some others, belong to the eastern group of the Indo-European family of languages. The Baltic group is now represented only by the dialects of Lithuania Madhu Vidya/625

Page #651
Reviews 243 and Latvia in the U.S. S. R. Originally, one more language belonged to this group, old Prussiau, which became extiuct by the 17th ceutury. The Baltic languages, especially the Lithuanian, are supposed to be very conservative in character and hence they are very important for comparative linguistics. Lithuaniau has preserved even pitch accent, which Sanskrit apparently lost in the classical period, - a feature which Dr. Chatterji seems to have inissed to stress in his description of the Baltic languages. A few comparisons will illustrate the close phoneticsemantic resemblance between Sanskrit and Lithuanian : Sk. sunus, Lith. sunus 'son', Sk, devus, Lith. lipunan 'god', Sk. avis, Lith. aris sheep'. In declension, Lithuanian has all the Sanskrit cuses except the ablative. But these resemblances, striking though they are, need not lead any oue to suppose that a Sanskrit scholar can, without special training, read Lithuanian. This will be borne out by looking at the specimeus of olilnud modern Lithuanian and Latvian literature given. by Dr. Chatterji in Ch. XVI. Dr. Chatterji seems to agree with those who are of the opinion that the original line of the Indo-European people lay in the dry steppe laws of Rurasia to the suit of the Ural mountains. It is to be udmitted that the present book is not the place where one may expect to find it detailed discussion of the villcrent theories held on this subject. Nevertheless, it would have been better if the author had referred to P. Thieme' who), on good grounds, holily the view that the original home of the lub-European lily in Central Europe, more specifically in the domnuin of the salmon rivers, i. e. in the region of the rivers Vistula, Oder and Cile, approximately if the point where representatives of eastern (Baltic-Slavic ) und Western (Germanic ) Indo-European languages meet. Ou p. 2.3, Dr. Clintterji ginys that the name Balt may ineun white marsh lands' and then the people connected with these lands'. He wishes to derive the word Balt from the IB root *bhe or *bha and see the continuation of this root in Skt. Vbhris (< hal-8) and V bhan (Page #652
Annals of the Bhandarkin Oriental Research Institute languages of antiquity". But he could have noted at this point the fact that the Indo-Iranian branch had, in fact, in antiquity palatalized IE velars and la bio-velars in similar environments. On p. 118, Dr. Chatterji says: "In Primitive Indo-European a regular pantheon with gols of a nett and precise character appears not to have been fully establislied". One wouilers whether he has used here the word 'nett'under the intluence of the German language. The book has a special interest for Indian readers because, as observed by the author, both the Balts and the Indo-Aryans bare to some extent preserved certain traits of IE religion and culture. Let us therefore join the author in hoping that his book will serve the purpose of bringing together the Baltic and the Indian peoples "through knowledge and understanding, and sympathy and brotherly feeling"." (p. xx). M. A. Mehendale THE PENGO LANGUAGE, GRAMMAR, TEXTS AND VOCABU. LARY, by T. Burrow and S. Bhattacharya, pp. 1-233, Oxford 1970, Price PS 2.60 or 52 s. The authors of this book are alrendy known to Dravidian scholars 48 joint authors of a similar book ou The Purji language of Bastar (1953) und A Comparative Vocabulary of the Ciondi Dialects published in JAS ( Calcutta ) 2.73-229, 1960. The lengo language is spoken nainly in the eastern portion of the Nowrangpur sub-division of the Korapuit district of Orissa. Together witli kui, kuwi, Komdu and Condi, it belongs to the central group of the Druvidian fainily of languages. The language was practically unknown till recently. Although it was first noted in the Cer4u8 of 1891 (Vol. XIII, f. 186 ) no muterial froin it was ever made available for inspection. Iu their book on the l'arji lunguye, the authors refer to Peug loroja of Orissa us the name of a tribe, but there is yet no word from thein about the lengo language. Even iu thc Druvidian Etymologicul Dictionary of 1961, although its authors inake a reference ou p. vi to the discovery of the Penyo language, it was possible for them to use materials from it for their dictionary only in the Supplement published by them in 1968. The present study is based on the materials collected by the authors while tuuring in this region at different periods between 1957 and 1966. It is a fairly complete account of the phonology and morphology of the language. The book iucludes also some texts with translation 1 Another accidental discovery, that of the Manda language, io 1964, bas also been referred to in the Supplement, aod in the book under review, on p. viii, the Authors have promised to publish a description of it in duo course, Madhu Vidya/627

Page #653
Reviews 245 and a vocabulary. The treatment is more historical and comparative than descriptive. Hence, while stating the simple fact that in Pengo both -j- and -2- occur in the intervocalic position, e. g. bajek 'much' and mazi 'son', the authors add : "Such cases are probably due to differences in origin, e. g. -j- may represent earlier -jj-contrasting at some period with -j-" (p. 5). Or aguin, on p. 7, we read : "Pe. his un voiced, as to be expected since it has developed from earlier 8". A distinction is made throughout the description of the language between inherited and loan words. The authors say that they found the language to be remarkably uniform and free from large scale dialectal variations. But they have themselves noticed some features characteristic of the eastern Pengo as opposed to the western Penyo. Moreover they state, on the authority of their principal informant, that there exists a section of Pengos called Mahan Parjas because they use & past tense form mahan for the standard Pengo form macun (p. viii). The eastern Pengo is marked by the use of a single phoueme -- while the western Peago has two phonemes - ; - and -2- It is not clear why the uuthors describe this state of affairs as confusion' (p. vii and 5), instead of calling it'merger' of two phonemes. The authors observe on p. 4 that although in some of the Oriya loan words they heard the open pronunciation ( - ), they have not accepted it in their normal method of transcription. It is to be doubted whether this is the correct procedure tu follow in the description of a lauguuge. On p. 13, the author's note examples of intervocalic - d in l'engo. This being contrary to their earlier statement of the developmeut of intervocalic - (-(---) to - 2-, and since the examples of the intervocalic - d - are numerous, the authors call this an alternative developenent'. This will hardly seem satisfactory to those who still cling to the hypothesis of the regularity of phonetic changes. On p. 48, the authors observe tbat in Pengo tue Dravidian numerals are only the first two. From three onwards, Pengo uses Indo-Aryan numerals. But on p. 50, they record the restricted use of Indo-Aryan numeral for one ek'as & suffix in korek'one score' and of two 'dui' in such expressions as dui kosi 'forty'. The reviewer is not quite familiar with the Dravidian languages and bence it is not possible for him to go into the details of the obser. vations made by the authors regarding Pengo grammar. M. A. Meheadale Madhu Vidya/628

Page #654
reviews and notices GOSWAMI, Upendranath. A Study of Kamarupi: A Dialect of Assamese. Gauhati Department of Historical and Antiquarian Studies, Government of Assam. 1970. Price Rs. 15-00, IX + 312 pages. Reviewed by M. A. MEHENDALE, Deccan College, Poona. In the book we find a detailed study of the Phonology and the Morphology of Kamarupi, a dialect or Western Assam The author notes that "The people of Western Assam and their speech were looked upon with contempt and the term Dhekeri was applied to them." It is to be hoped that the situation has now changed. The treatment of the Phonology and the morphology is mainly descriptive, but we find that the historical and the comparative aspects have also been considered. In the Introduction the author gives some useful information regarding the home of the dialect, the relationship of early Assamese and Kamarupi, some special features of Kamarupi which distinguish it from standard Assamese, and a brief analysis of the vocabulary of Kamarupi. The author says that Kamarupi has a distinct stamp because Kamarup remained as a distinct political entity for a long time and that the means of communication between the cast and the the west Assam were poor. At the end we find an Appendix, which is divided under nine heads, some Specimens of the dialect, a Bibliography, and a Word Index. * On p. 56, the author purports to begin his treatment of 'Initial Vowels'. But his examples, kar, kalah etc., make it clear that what he means by initial vowels is really vowels in the initial syllable. From p. 77 onwards the author gives examples of some' phonological chan. ges of a general character. Some of the examples cited by him are not quite happy. Thus, for example, he cites narayana > naran as a case of haplology (p. 81) which is not correct. Any word contraction is not haplology. Similarly the change noticed in pathasala > patsala (p. 83) is not a case of assimilation, since the retroflex th would have been replaced by th even in the absence of a dental 1. 2. P. 13. The author does not explain the term 'Dhekeri '. On p. 34, the author cites some words which he thinks are similar to Marathi. I am not able to recognise, except in a case or two, anything specially Marathi about these words. MadhuVidya/629

Page #655
REVIEWS and NOTICES 89 sibilant s. This is a general feature of Kamarupi phonoloy. Hence this could be an example of deaspiration (th >t) which was probably due to the influence of the standard colloquial (p. 84). The normal tendency, in the author's opinion, of Kamarupi is aspiration (p. 83 ( 124 ) In many cases the author omits to take into account an intervening phonetic change and hence some of his examples are not immediately in accord with his labellings. Thus his example sari-fa> saitta can fit in his description of the assimilation of plus some consonant only if we assume an intervening stage *sair-fa. Under unvoicing' on p. 87, the change should have been better formu. lated as -nij., .;>*-fic, *.c>s.. Similarly, the change -b >-ph* on the same page should have been shown as -*bh > .*bh >.ph. Some of the examples cited under 'Deaspiration' (p. 86 ) would have been clear if the author had given the intermediate stage like dh> * > r. In effect, however, the change *> shows a phonetic shift. It is necessary to discourage the habit of quoting from authors where such quotations either are out of place or serve no useful purpose. On the page prece. ding Foreword, the author gives quotations (without proper references ) from the works of Bloomfield and Gray. Of these, the former is out of place because in the present book the description of the language is not followed up by an attempt 'to observe a human group' or 'to probe deeper into the ways of the community and understand their historical origin'. On p. 4, in saying that the standard Assa. mese is being used more and more by the people, the author cites the following from I. C. Ward's The Phonetics of English, p 4. : 'should be encouraged since speech is a form of social behaviour.' One fails to understand why the author has found it necessary to give this quotation.' 3. But on p. 86 the author gives some examples of deaspiration also from kamarupi 4. Not intervocal - - > Ph-as indicated by the author, 5. The outhor docs not give reference to the page number of the book, but merely gives the chapter number.. Madhu Vidya/630

Page #656
KELLER, Joseph R. Linguistic Theory and the Study of English: A Selective Outline. Minneapolis, Minn. : Burgess Publishing Company, 1968 91 pp. $ 2.75. Received by M. A. MEHENDALE, Deccan College, Poona. The purpose of this outline is stated to be, first, to point out that there is a basic coherence in the development of linguistics, from Grimm even to Chomsky and, next, to clarify "the cultural lags within linguistics." It is pointed out that linguistics in its very early stage was prescriptive; from this, after the discovery of Sanskrit, it passed through Comparative Philology and structuralism to transformational hypothesis. The impact of these developments in linguistics on the study of old and middle English is not to be seen in the available descriptions. Tne author therefore indicates how this can be achieved. From the late Middle English to the Modern English period, however, the phonological and morphological changes are not so extreme and what one sees in this period is a large-scale borrowing of foreign words into the English vocabulary. The author therefore turns to the problems of style and usage. With regard to the former he refers to the Boas-Sapir-Whorf hypothesis regarding the relationship of world view and language structure; and as regards the latter he mentions the two extreme attitudes of doctrinaire permissiveness and doctrinaire prescriptiveness and rightly observes that''neither extreme is valid ". As an outline, the book will be found very useful. It goes to the credit of the author to have emphasized that although the term structuralism was not invented till 1920, Grimm, while thinking about phonetic changes, thought in terms of changes in the habits of articulation which led to the restructuring of the sound systems. He also aptly points out that the practices of the neogrammarians and the structuralists do not throw each other out. While pointing out that it would be arrogant to call modern Western European Linguistics as the only science of language the author refers to the descriptive thoroughness of Panini and his predecessors who are said to shave analyzed Sanskrit in the fourth century (p. 11 ). The author here does not ay whether B. C. or A. D. But that he means the fourth century A. D. becomes clear from his later statement: 'the Hindu grammarians who described Sanskrit in the first centuries of our era" (p. 85). Since no one has ever thought of bringing down Panini, much less his predecessors, to a date after the beginning of the Christian era, these remarks by the author are very hard to follow. On pp. 22-23, the author gives dates of the Germanic sound shift. He says that IE bh, dh, gh became Gmc. voiced stops b, d, g, ca. 1000 B. C., and that IE voiced stops b, d, g, became p, t, k, ca. 100-500 A. D. But if the change IE bh > Gmc. b occurred before the change IE b > Gmc. p, the question arises why the Gmc, b, arising out of IE bh, did not become p along with IEb which became p in Gmc. Obviously the change bh > b must be supposed to follow, and not Madhu Vidya/631

Page #657
REVIEWS AND NOTICES 177 precede, the change b > p. If the dates are given following E. Prokosch : A comparative Germanic Grammar p. 53 (Keller refers to Prokosch on p. 22 ), then it is worth while to note that Prokosch accepts an intermediate stage bo z through which bh dh gh passed before becoming bd g, and the date ca. 1000 B. C. given by hi.n may have been intended for the first change bh dh gh > 50 % and not for their ultimate appearance as bd g. It is rather awkward to see (Bernard) Bloch's name printed as Block (p. 25). (Received 27 April 1972), BHATTACHARYA, Gopikamohan. (ed.). Praci- Jyoti. Digest of Indological Studies Kurukshetra, Hariyana : Kurukshetra University, Annual, Vol. VII (1971). I, 281 pp. Subscription : Rs. 30; 50 s.; $ 8 per year. Revicwed by M. A. MEHENDALE, Deccan College, Poona. During the last two decades, the number of Journals devoted to Indological siudics has steadily grown. There have also appeared during this period a number of felicitation or memorial volumes. While this increase in publications offers frequent opportunities to scholars to publish their research, it also makes it difficult for them to keep pace with the tide of new arrivaly. The Institute of Indic Studies of the Kurukshetra University therefore started in 1963 the publication of the Digest of Indological Studies --- Praci-Jyoti in order to enable the scholars to find in one place information about the articles appearing in different volumes. This was indeed a welcome objective. Unfortunately, the Institute, after having published, the first five volumes and part 1 of Vol. VI between 1963-68, had to stop the publication, and this in spite of the fact that the Digest "had won universal acclaim all over the world." It is therefore gratifying to see that the Institute has once again found it possible to start the publication of the Digest beginning Vol. VII for the year 1971. The present volume contains the same different sections as in the previous volumes. It also publishes topics on which work is being carried on in different Universities for research degrees. In the end is given an (incomplete) list of Research Institutes and reviews of a few books. A look at the Digest will convince any one of the great amount of labour that has gone into the preparation of the volume. The publication of such a useful reference work becomes possible only as a result of team work. It is pleasant to note that the Editor has found a willing body of workers to help him in preparing the material for the Digest, Madhu Vidya/632

Page #658
178 M. A. MEHENDALE In the first volume the Editor had said that it was impossible for him to include everything that had appeared in the different Journals and that he was therefore forced to make a "judicious selection of articles". On going through the matter presented in the volume under review one has to say that this principle of 'judicious selection was not applied rigorously. Again, although one observes that the abstracts of papers published in this volume are on an average shorter than those in Vol. I (where they often extended to the length of a full page, and occasionally even beyond that), it must be said that they could have been still shorter. If both these principles-careful selection-and abstracts of minimum length-are strictly adhered to, it would be possible for the Institute to bring out future volumes expeditiously and at a considerable less cost, without sacrificing its utility. [ Received 29 April 1972] Madhu Vidya/633

Page #659
reviews and notices SHANKARA BHAT, D. N. Sound Change, Poona, Bhasha Prakashan, 1972. 98 pp. Rs. 12; $3 Reviewed by M. A. MEHENDALE, Deccan College, Poona Ever since the beginning of modern linguistic studies, sound change has, quite understandably, engaged the attention of scholars. More than any other aspect of linguistic change, studies in sound change have produced some very important and interesting results. Owing to a general decline in the interest in historical linguistics in the past few decades the theory of sound change did not receive the attention it did in the 19th and the early years of the 20th century. However, in recent years there have been some refreshingly new approaches to the study of sound change, Dr. Bhat's being one of them. The author, in the small compass of less than a hundred pages, tries to examine the bases of language change, the characteristics of sound change, the effects produced by sound change on language structure, and the methods employed to recover the sound changes. He has critically examined the well known hypotheses regarding the gradualness and regularity of sound change in the light of the evidence collected by him in his dialect studies in South India. He has offered some good material for a reassessment of these basic notions regarding sound change. But it may be doubted whether it would be good to throw away completely a hypothesis like that of gradualness and regularity which has proved its worth over a long period. The most important contribution made by Dr. Bhat in this book is to draw attention to the fact that in the history of a given language no novel contrasts are brought about. What was until recently believed to be a new contrast, Dr. Bhat points out, is in effect only a shift of contrast (4-29). Thus, when in Indo-Iranian velar k and palatal c come to be contrasted, this is really a shift of an earlier contrast between vowels a and e. But it does not seem correct to state this as: "Sound changes are incapable of introducing new sets of contrasts into a language" (p. 82). It would be more pertinent to say: Sound changes are incapable of adding to the number of contrasts already existing in a language. Equally interesting is Dr. Bhat's remark that "the basis for connecting together various written records of a language belonging to different periods of time is nothing but genetic hypothesis itself..." (p. 20). Dr. Bhat has not elaborated on this point, but he seems to see a similarity of approach in comparing two synchronic stages of related languages and two diachronic stages of suppo sedly the same language. Both approaches lead us to the establishment of historical relationship. Comparison of OIA karna and MIA kanna would be on par with the comparison of Skt. soma and Avestan haoma if the two sets of corres. pondences are considered as cognates in the sense that they are etymologically related. That in one case the established etymology, karna, happens to be identical with one of the two items compared makes it no different from the other where the suggested etyinology *saonta is different from both the compared items. Madhu Vidya/634

Page #660
230 A. M. MEHENDALE Even in the latter type of comparison, occasionally the suggested etymologies do turn out to be identical with one of the compared items, as with IE esli >Sk. asti, Gk esti. In essence Dr. Bhat's suggestion comes very close to that of Hockett who, while pointing out the reflexive nature of genetic relationship observes : "Consider the limiting case in which we compare a language with itself, Recurring correspondences are then indentities, every form is cognate with itself and from the beginning there is no unexplained residue." ( 1965 : p. 189). But there is a contradiction involved in Dr. Bhat's exposition. He looks upon the comparison of two dialects on the basis of genetic hypothesis as an internal criterion for observing sound change, and even though he considers comparison of two diachronic records of a language as a case of genetic hypothesis, he looks upon it as external evidence for language change (p. 20). A few other comments may be offered : (1) On p. 31 Dr. Bhat gives three English words knight, knob, and knowledge as examples where childern must have failed to internalize the distinction between k and its absence in the initial pre-nasal environment. But whereas in the case of knight there is a comparable word night where one can imagine a child to have failed to note the distinction between the initial kn- and n-, there is no such comparable item for knowledge. As for knob, although a word nob exists in English it is hardly likely to come a child's way. (2) The fact (and not "the assumption") that a lost phonological contrast cannot be recovered through following sound changes is a limtation and not the main basis of comparative method (pp. 35-6 ). (3) The diagram (p. 61) showing the split of k'to k and c is so drawn as to indicate that the velar part of l' has developed into k and the labial part of it has developed into c. But this is not correct. It would have been better for the author to indicate the split of a simple velar k and not that of a labiovelar ko (4) The Sanskrit stem is not suhrt (p. 64 ) but suhrd. (5) The alternation - and ju in Sanskrit cannot be called non-automatic. It is automatic, but non-unique. . (6) The appearence of u in wives is not in the intervocalic position (although it appears intervocal in spelling) (wives was once wi:v@z.--ARK.) (7) A few comments regarding examples or glosses: (i) p.53: Pali mottia is incorrect; (ii) p. 69: Hindi parni 'bride' is doubtful; (iii) p.87: German Bund'pocket' is incorrect; (iv) p.57: palatalization of a velar before a palatal vowel e as in *kedi > cedu is not a good example of partial assimilation. A better example would be Skt.* ud-matta > un-matta. A word of caution may be given to the reader, Dr. Bhat seems to be specially fond of the phrase a number of.' He has used it no less than five times in a single page of preface. It is likely that he does not mean it literally every time he uses it. REFERENCE Hockett, Charles F. 1965. Sound change. Lg. 41. (Received 30 March 1973] Madhu Vidya/635

Page #661
SAKSENA, B. R., CHATURVEDI, S. P. and others, (ed). Kshetresa Chandra Chattopadhyaya Felicitation Volume, Part I, being Vol. XXVII parts 3-4 of the Journal of the Ganganatha Jha Kendriya Sanskrit Vidyapeetha, Allahabad, 1972, pp. 373. Reviewed by M. A. MEHENDALE, Deccan College, Poona. Prof. K. C. CHATTOPADHYAYA is known to scholars as a diligent researcher in the field of Veda and Sanskrit grammar. On account of his long association with Allahabad University and the fact that he was "one of the most brilliant students of Dr. Ganganatha Jha" it is but natural that the Kendriya Sanskrit Vidyapeetha at Allahabad has thought of designating one of the Volumes of its Journal as a felicitation volume to Prof. CHATTOPADHYAYA. The Volume contains many interesting articles. It is difficult to make a detailed reference to all of them in a short notice. L. ALSDORF convincingly points to one more hymn RV 5.78 as a "legend-spell" (Legendenzauber). Although he considers this hymn as a 'composite' one, he asserts that "I am far from maining that every sukta of the legend spell type must be a secondary combination of a charm and an existing old legend prefixed to it later." (p. 73). P. N. BANERJEE in his article on Calcutta-Tokyo Language Axis gives examples to substantiate his view that "the Bengali and Japanese languages are strikingly akin to each other. Morphologically, syntactically, lexically, and idiomatically there are striking parallelisms between these two oriental languages. ..." (p. 35). Ramasvarup CHATURVEDI offers a brief grammatical analysis of the language of Suradasa. S. P. CHATURVEDI gives the text of the letters in poetic Sanskrit exchanged between Pandit RAMACHANDRA and an Englishman Lancelot WILKINSON. S. K. CHATTERJI refers to India-Central Asia contacts from pre-historic times. "We may however be justified in assessing that the Hindu.... was face to face with the Altaic Turk long before either of them came to India and the lands of the Middle East...." (p. 129). Dalsukh MALAVANIA makes it perfectly clear that the word vari in the phrase sabbavarivarito used with reference to Mahavira in the Pali text cannot mean "water" but "sin". H. K. MIRZA suggests to render Sanskrit amnaya used to translate Pahlavi patvand (Pazand paevand) as 'descendant, lineage, family, succession' (pp. 330-331). Among other articles mention may be made of Vedic statya and Panini 3.1.123 by M. D. BALASUBRAHMANYAM, On the Sanskrit Literary Genres Paryayabandha and Samghata by H. C. BHAYANI, An Etymological Note on the word s'man by B. B. CHAUBEY, Atharva Sarnhita and its Forms by H. R. DIVEKAR, Laksana, "Grammatical Rule" by Sergin Al-GEORGE, The Adjectives of Early and Middle Oriya by S. HOTA, Fresh Light on Panini's Sutra 1.2.32 by A, N. JANI, On the Galitapradipa of Laksmidharasuri by K. P. JOG, A Critical Study of Ch. V of Epistle I of Manuscihr Gosn-Jaman by M. F. KANGA. Madhu Vidya/636

Page #662
REVIEWS H.R. DIWEKAR, V. P. LIMAYE, R. N. DANDEKAR, C. G. KASHIKAR and V. V. BHIDE, Ed., Kausikasutra Darilabhasya, Post-Graduate and Research Department of Tilak Maharashtra Vidyapitha, Poona, 1972, pp. xvi, 36, 136 + 136, 59, Price Rs. 50, PS 3, $ 8. Among the ancilliary texts belonging to the different Vedas, the Kausikasutra (KS), belonging to the Atharvaveda, occupies a peculiar position. It is neither a Srauta, nor a Grhya-sutra but "a mixture of two distinct kinds of Sutras, Atharva-sutras and Grhya-sutras" (Bloomfield JAOS 14.xxi). The text of the KS itself is not easily intelligible and hence one is often required to take help from the commentary of Darila and the Paddhati of Kesava. Until recently only extracts from these two texts, published by Bloomfield in JAOS 14, along with the text of the Kausikasutra, were available to scholars. It was therefore necessary to publish the entire text of the commentary of Darila. This task has now been accomplished by a group of eminent and devoted Sanskrit scholars. Unfortunately they were compelled to base their text on a single manuscript, a microfilm of which was made available to them by the authorities of the University Library at Tubingen (W. Germany). Three other manuscripts of the commentary are known to exist; but these could not be traced by the editors in spite of their great efforts to procure them. The only manuscript, on the other hand, on which the present text is based, is corrupt beyond imagination. The diff. culties of the editors in this situation therefore can be imagined. They must have been required indeed to struggle very hard to obtain a fairly intelligible text out of the corrupt manuscript. A look at the original text, which also has been reproduced in the editon by the off-set process, will convince any onc about the truth of the following statement of the editors: "The editors had literally to wrestle with many passages for hours together--not unoften, in several sittings--before they could restore them to an intelligible form " (p. xiii). One really admires the patience, the tenacity and the ingenuity expended by the editors in bringing out this excellent edition. It is indeed difficult to express adequately our gratitude to the editors for this devoted work. The edition first gives the text of the KS upto the end of Kandika 48. This is followed by the commentary of Darila---both as it appears in the original and as read by the editors. In the end are given some very useful and informative Appendices. In the Appendix where citations from the accented texts like the Atharvaveda are given it would have been better to give those citations with accent. In App. A (p. 12) we have a note ( 28.6) on sarvayajnanain cendrabhaktitvat. In this note we are referred to the Madhu Vidya/637

Page #663
BOOK REVIEWS 175 Nirukta 7.10. But this section of the Nirukta does not list sacrifice in general among the shares (hhakti) of Indra. In the note 53.3 on p. 31 of App. A, it would have been better to add that the Nighantu (1.12 ) itself does not list rajah among the synonyms of water. In the note 61-8 given in App. A. p. 24 one could have also referred to H. Luders : Das Wurfelspiel im alten Indien, Phil. Indica p. 106 ff. On p. 61 lines 10-11 of the commentary occurs the word talpah as qualifying rajaputrah. Here the editors might have referred to the Sat. Br. 13.1.6.2 where we get the word talpya qualifying raja putra (saram var ralpya rajaputra asapalah). The commentator explains the word talpya as talpasadhavas talpyah sayyagarah talpena saman ratrau ye sadhu raksanti. On p. 107, line 13 Darila explains the sutra word akarsa as akarsah lohakaranam | angarakarsanarthani kutakah. One may note in passing that the commentary Tattvabodhini explains akarsa as akrsyate' nena khaladigatam dhanyani ity akarsah (while commenting on the example akarsasvah given under P. 5.4-97) On p. 123 of the text, in line l occurs the expression amusyah putrasya as two different words. It is also given as two words in the Padapatha of AV 10.5.36. In App. B (p. 49 ) the editors offer the following comment : "But according to P. 5.1.133 amusyahputrasya is a compound-word." This statement, however, does not seem to be correct. P. 5.1.133 (dvundvamanojnadibhyas ca) only tells us that the suffix aka ( vun ) may occur after a dvandva compound and the words listed in the manojnadi gana. Hence we can have forms like gaupalapasupalika or manojnaka etc. Now the expression amus yapautra (but not amusyahputra) occurs as one word in the manojnadi gana and hence, according to the sutra in question, we can have a form like amusyaputraka. But the Sutra itself does not say anything about the formation of amusyaputra, much less of amus yahputra. It would have been therefore better to state simply that amusyaputra as a compound from occurs in the manojnadi gana (P. 5.1.133) On p. 123 line 3 we read idam ahum ak sabrahmanayanaputrasya veccikaputrasya pranapanay apakyntami. On this, in App. B (p. 49 ), the editors have the following comment : " It is better to read aksasya brahmana. yanaputras ya." But this may not be justified. The sutra ( 44.31 ) on which Darila is commenting runs as idam aham-amusyayanasyamusyah putrasya... Thus the Sutra does not give any scope to name the individual, against whom the black magic is to be practised, by his personal name. The Sutra wants him to be referred to only by way of his father and mother. The title of the text as given by Darija himself is Kausikabhasya. The same could have been retained without change. Madhu Vidya/638

Page #664
reviews and notices Indian linguistics 36. 63-5 (1975) SEN, Subhadra Kumar. Proto-New Indo-Aryan. Calcutta: Eastern Publications, 1973. viii; 182 pp. Rs. 25. Reviewed by M. A. MEHENDALE, Deccan College, Poona The present book, a doctoral thesis of Calcutta University, is written. to describe linguistically a stage in the development of Indo-Aryan which the author chooses to call Proto-New Indo-Aryan. The title of the work is somewhat misleading. One picks up the book with the thought that one would find in it the description of a linguistic stage which is arrived at by reconstruction on the basis of the oldest recorded stages of some of the New Indo-Aryan languages. But this is not what the author does. He informs us in his Preface that his work is the result of his study of the language recorded in the Dohakosas, and the allied Pahudadoha, Savayadhamadoha, verses quoted by Hemacandra in his Siddha-HemaSabdanusasana ch. 8 and some Jaina works like the Kumarapalapratibodha, etc. He considers that the linguistic stage which he describes represents Avahattha i.e. the later phase of Apabhramsa. But because this phase is very close to the oldest form of New Indo-Aryan, the author avoids calling his work a description of Avahattha, and calls it a description of Proto-New Indo-Aryan. He does it also because the term Avahattha is not used by major Prakrit grammarians. What, in essence, the author wishes to tell us is about the transitional stage between late Middle IndoAryan and early New Indo-Aryan. "This process of transition is neither uniform nor simultaneous in the different linguistic areas, Uniformity in the linguistic structure therefore is not to be expected. But a rough picture can be drawn out and the following pages are an attempt in that direction." (pp. 1-8). According to the author's estimate 'Avahattha is about 70 per cent Proto-NIA. Of the remaining 30 per cent 15 per cent form the Apabhrarsa strain and the other 15 per cent are its own peculiarities." (p. 18) Once we realize exactly what we are going to get from the author, we find in his work a perfectly readable account of the phonology, morphology, syntax, and vocabulary of this 'Proto-New Indo-Aryan'. To make his account more intelligible the author has prefixed to his description a brief sketch of the earlier stages of Indo-Aryan. All this forms part I of the book. Part II is supplementary to the first. The author gives in it specimens, with notes, of the texts on which his thesis is based. One wishes he had not called these specimens Proto-New Indo-Aryan Madhu Vidya/639

Page #665
64 M. A. MEHENDALE texts since as he has rightly observed : "No Proto-NIA text marked as such is available to us. It has to be reconstructed from the Avahattha (and Laukika) texts known as yet." (p. 18) A concise vocabulary follows these specimens of texts. Some comments follow : (1) On p. 3, while describing how Vedic and Classical Sanskrit differ in vocabulary, the author cites examples of only vocabulary loss. He could have as well given examples of new arrivals in Classical Sanskrit, especially by way of borrowing. (2) On p. 7 the author expresses his opinion that there exists "a perfect parallelism" (emphasis added) between 'Proto-NIA' and Buddhist Hybrid Sanskrit because the literary products of the close of the MIA period show some features of late Apabhraisa although in point of time they are closer to New Indo-Aryan. But such mixtures are likely to be found in any transitional literature. Moreover, as the author himself points out on the same page, the BHS is "to some extent artificial" while the proto-NIA is "hased on colloquial". (3) On p. 16 the author improperly includes items of lexicon like dola 'eye', bani "horse', etc. under morphology. (4) On p. 22 the author notes the retention of geminated -5s- in parassu and suanassu as an Apabhraisa feature as opposed to the occur. rence of a single -s- in tasu which is an Avahattha feature. But he also says that the retention of double consonants is due to metrical necessity. If this is true, we cannot point to it as a retention of an old feature. (5) In chapter 9 on Avahattha morphology one misses certain features. For example, one does not find on p. 68 the instr. sg. termination -ina (cf. sasanalina < svasanala 'hot breath' p. 146 verse 4); the very first item airattie in vocabulary does not figure under genitive. The two forms caijja and bhamijja which occur in verse 1 on p. 146 are not treated under the passive (pp. 89-91). (6) Under Avahattha syntax (p. 103) the author mentions the use of the ablative-genitive and gives as an example guruhu pasac 'by (=from) the grace of the guru'. This is not the correct example. The use of *by = from' applies to pasaa and not to guru. If at all, the author could have used it as an example of ablative-instrumental. (7) The second line of stanza 3 cited on p. 146 runs as -- navari mayarku vi taha tavai jaha dinayaru khayakali. The author's translation of khayakali 'at the time of mutilation (i.e. in the waning phase)' is not correct. The word khayakali goes with dinayaru and not mayanku. of the Prakrit grammarian Purusottama. * Prof. P. L. Vaidya's rendering (in his edition of the eighth adhyaya of Siddha-HemaSabdanusasana, p. 683) at the close (of the day)' is equally unhappy. Madhu Vidya/640

Page #666
65 REVIEWS AND NOTICES It means '(as the sun burns) at the time of (the world's) destruction'. (8) The vocabulary at the end leaves much to be desired. For one thing it is not complete. And the words included have not received a uniform treatment. In some cases only the Sanskrit equivalents are given; in others, these are followed by meaning. At times the Sanskrit equivalent is followed by formal identification (aksaye loc. sg.), in others only the form is identified (loc. sg. of abhra) without actually giving the Skt. equivalent. With all this, however, the book is undoubtedly a welcome addition to the available treatments of Indo-Aryan. [ Received 6 May 1975) Madhu Vidya/641

Page #667
Iniernational Journal of Dravidian Linguistics REVIEWS A PRAKRIT READER (A Linguistic Introduction Based on Selections from Hala's Sattasai] H. S. Ananthanarayana : Central Institute of Languages, Mysore, 1973, pp. x; 99, Rs. 6/ M. A. Mebandale Poona As the sub-title of the book states this is a modest attempt to introduce the reader to a dialect of Middle Indo-Aryan as reflected in, the Sattasai attributed to Hala. The author, first, gives a clear picture of the phonology and the morphology of the dialect as seen from the verses. He treats phonology both descriptively and historically, but makes no historical statements with reference to morphology. The description of the dialect is followed by a selection of some hundred verses and a useful Glossary. In the text of the verses, detailed notes on individual words are given to help the reader to understand the text. But it is not clear why the author has chosen not to translate the verses into English. On the whole the Reader will be found very useful by the students of Prakrit for whom it is intended. A few comments may follow : 1) p. 6: The remark "The plosives occur medially as geminates and with a nasal" is misleading. It is likely to give the impression that the plosives, otherwise, do not occur medially. How then would one account for the medial plosives in words like kodaa, kapurisa etc.? 2) p. 10: The author could have pointed out specifically that as geminates, in the medial position only the semi-vowels 7 and v occur, but not r. 3. p. 10: The author looks upon my in cumvai as an instance of v with a nasal; but he considers mh in gimha as an example of a Madhu Vidya/642

Page #668
ijd! Vol. IV No. 2 380 Mehandale cluster. He should have explained the reasons for the two different treatments. 4) In historical phonology one misses a statement to the effect that intervocalic stops generally tend to disappear as in paa, pai etc., and in descriptive phonology a statement to say that two vowels may occur side by side without coalescence. 5) In morphology one expects to have a clear statement about the loss of the dative. 6) pp. 28-29 : The author notes ejja as an optative marker and then, among the terminations, he gives zero term for 1st sg. and for 3rd sg. and pl. However, in the paradigm there, three forms are shown with a final short vowel -a (hapejja) for which no statement is made. 7) Certain words are not taken care of in the notes which occur after the text of the verses. Thus the words pamkaa, in verse 1, tanti in verse 2, dunia in verse 3, tog- (tog-gaa > ivad-gata) in verse 5 are not explained. The last itern does not occur in the Glossary either. In verse 5, the use of the particle ma has not been explained. 8) The expression kamassa taitatamtim kunamti (p. 35) is a difficult one. The commentator explains tattatamtim as tattyacintam, but it is difficult to relate tamti with cinta. It is, however, true that the word tamti occurs twice in the Sattasai (1.51, 3.73) where it means cinta "anxiety, thought'. Dr. Ananthanarayana suggests to derive tanti from tantri (Glossary under tattatamrim) but does not indicate how it can mean 'secrets of love'. One possibility could be to look upon tamti to have the same meaning as tantra. The expression then means "those who formulate the doctrine about the real state of love'. 9) The word pesio in verse 101 (p. 67) has not been explained. In the Glossary it is derived from presita 'sent'. The derivation is possible, but not quite appropriate in the context. The explanation prave sita 'made to enter' would be more to the point. The first line of the verse means 'pride which was made to enter my heart by my friends having some how found on opening'. As a matter of information it may be added that the Sattasai has been published with Marathi translation and notes by S. A. Jogalekar (1956) and with English translation by Radhagovinda Basak (1971.) Madhu Vidya/643

Page #669
316 Annals BORI, LXII (1981) INDIAN LEXICOGRAPHY; by Claus Vogel (In the series A History of Indian Literature, ed. by Jan Gonda, Vol. V, Fasc. 4), Otto Harrassowitz, Wiesbaden, 1979, pp. 303-401. This is a very useful account of the Indian Lexicons, both general and special, as also bilingual and multilingual, and the commentaries on them. The whole account is neatly presented in a small compass of less than huadred pages. The author gives a brief introduction and then narrates the characteristic features of Indian lexica: the principles followed by the authors of the lexicons in organizing the lexical matter and the method of structuring individual items in this organization. This is followed by systematic information about the individual authors and their works. In supplying this information to the reader, the author has, apparently, brought together all available material on the subject. In the opening statement the author says that Indian lexicographic work started with the compilation of word-lists (nighantu) giving "rare, unexplained, vague, or otherwise difficult terms culled from sacred writings." The well-known specimen of this kind, however, contains, besides the words of the above description, lists of synonyms and also a section called the daivatakanda. Madhu Vidya/644 M. A. Mehendale

Page #670
Reviews 289 HINDI-GUJARATI DHATUKOSA (A Comparative Study of Hindi Gujarati Verbal Roots): by Raghuveer Chaudhari. L. D. Institute of Indology, Ahmedabad 380009, 1982, pp. 230. Price Rs. 45/= This study of Hindi-Gujarati verbal roots is a welcome addition to the linguistic studies in the field of Indo-Aryan linguistics. The study was undertaken by Dr. Chaudhari for the Ph.D. degree of the Gujarat University. The work constitutes, in fact, a historical and comparative study of the Hindi verbal roots. It is historical in the sense that author tries to determine the sources of Hindi roots and it is comparative in the sense that in every respect be tries to ascertain the relationship of Hindi roots with those of Gujarati. In the Introduction, called Section 1, the author Darrates the work done in India and abroad in the field of the investigation of roots. The narration is informative and therefore useful. The author next lists the Hindi roots with their meanings, indicating their etymology and listing a Gujarati comparable root wherever available. For the preparation of this list the author got his material from the standard Hindi lexicons and some dialectal studies of Hindi. He also includes some obsolete Hindi roots. The total number of roots comes to 4270. But for his further investigation the author rightly excludes from this list such roots which show differences only in phonetic forms and thus obtains the revised number 2981. This list of Hindi roots, together with notes on them, forms Section 2 of the book though not specifically so described by the author. The third and the last section consists of two parts. In part 1. the author gives the classification of the roots such as tadbhava, desaja, onomatopoetic, tatsama, ardha-tatsama, and foreign roots. It may be remarked that it is necessary to consider the desaja roots also as those coming from foreiga sources, and, as the author rightly emphasizes, it is important to identify, as far as possible, the source which is vaguely labelled as desi. In part 2, the author gives his conclusions based on statistical account. The entire study has been carefully carried out and neatly presented (misprints are to be regretted). There could be some occasions of disagreement with the author, for example, one may not agree with his listing of **64 (729) in addition to ( 781), or his mentioning of Guj. TH 'to like, to prefer' under Hindi TH 'to lose' (1073). Incidentally, don't we have Guj. Tre in this sense ? The author's use of zabdakozIya sAmagrI-saMrakSaNa-zAstra for lexicostatistics (p. Madhu Vidya/645

Page #671
Annals BORI, LXV (1984) 205, also a p. 3) may not meet with approval. Lexicostatistics would properly refer to the statistical study of a lexicon for drawing historical inferences or for any other practical purpose. One doubts whether this sense is conveyed by sAmagrI -saMrakSaNazAstra. 290 It would be easy to find fault with the author's use of abbreviations. These have to be, as far as possible, easily recognizable and are meant to be only to save space. The use of a for way, (which one may suspect as standing for fra) for get, and many others of similar nature are not happy choices. Madhu Vidya/646 M. A. Mehendale

Page #672
Reviews DRAVIDIAN THEORIES: by R. Swaminatha Aiyar, The Madras Law Journal Office, Madras 600 004, 1975, pp. xlvii, 574, Price : Rs. 30/ 289 The contents of this book formed originally the subject of a series of lectures delivered by Shri Aiyar over sixty years ago to a selected audience. These were subsequently serially published in 1922-23 in The Tamilian's Friend, the Journal of The Tamil Education Society, Madras. Shri Aiyar thereafter continued his research in the subject and wrote a number of articles and delivered several lectures. In the light of this new material Shri Aiyar revised his original lectures with a view to publishing them in a book form.1 But before the book could be published the author, unfortunately, passed away. Realizing, however, the importance of the work the officers of the Madras Law Journal decided to publish Shri Aiyar's book although it meant a good deal of labour. Some difficulties arose even after the printing was started, but they were got over and the book finally appeared to the delight of all interested in understanding the relationship of the Indo-Aryan and the Dravidian Languages. There was a time, says the author, when it was believed that the Dravidian languages had sprung from Sanskrit. This theory was rightly given up when Bishop Caldwell's monumental Comparative Grammar of Dravidian Languages revealed a grammatical structure of these languages quite different from that of Sanskrit. But, the author complains, now some scholars have gone to the other extreme and have formulated a theory according to which the Dravidian languages exercised a profound influence on the grammatical structures of the Indo-Aryan languages including Sanskrit. This theory implies that the Dravidian languages were fully developed before the advent of the Sanskrit speakers into India and that they were spoken all over the north. The author believes that in his research he has tried to avoid the two extremes and has struck the middle path; but, in reality, he too has taken up a position which, although not identical with the erstwhile theory, is very much close to it. "The aim of the present book" the author states "is to secure a reconsideration of the current theory of Dravidian languages on the ground that the theory is based on a misapprehension of the real facts of the situation, and is supported by assumptions some of which can be easily disproved" (p. 1 P. N. Appuswami in his A Note on the Book, p. iv, does not give the date of this revision. Madhu Vidya/647

Page #673
Annals BORI, LXV (1984) 538). The large number of material collected and investigated by the author have led him to infer that "a very considerable majority, if not all, of the Dravidian grammatical forms have arisen from suffixed elements borrowed from Sanskrit and the Prakrts during the last twenty or twenty-five centuries, and that the basic portion of the Dravidian vocabulary not traceable directly or indirectly to Sanskrit or other Aryan sources is not very large." (pp. 4-5) 290 There is no doubt that the author has collected a very large number of details, both regarding vocabulary and grammar, on the basis of which he attempts to base the above inferences. There is also no doubt that the author. shows a remarkable control of both the Indo-Aryan and the Dravidian languages. Naturally a scholar who has similar qualifications will be in a better position to assess the value of the author's observations. However, a few comments are made here regarding some of the inferences of the author. As to the grammar, the author infers that most of the tense and modal suffixes used in Dravidian are of Aryan origin. This very important subject the author has dealt with fairly extensively with copious illustrations in chapters VII-XI. As mentioned above this portion will have to be reviewed by a competent scholar. However, it may be pointed out that the following statement made by the author while analysing the Malayalam future form alikkum will destroy does not carry conviction. He says: "kuPage #674
Reviews 291 Survey of India have said about the periphrastic future constructions and the past active participles. He wants to take these formations back to the IndoIranian period by adducing Avestan evidence (pp. 173-175). While doing so he has relied heavily on Kanga's Avesta Grammar ( $471, 8563 ). Without claiming any finality on the subject it may be pointed out that Jackson in his Avesta Grammar, although he notcs periphrastic verbal phrases ( $8 722721), does not give periphrastic future. Jackson also does not give any examples of past active participle. All the Avestan examples given by Kanga as those of past active participle have been treated differently by Bartholomae in his Worterbuch. The author has at various places discussed the vocabulary items. Since the book under review was writteu, we have now good etymological dictionaries and vocabulary studies which were not available to him. In the study presented by the author certain methodological drawbacks are encountered with which do not inspire confidence in his etymological inferences. 1) In considering sound-meaning resemblances it is better to take up items of polysyllabic structure than those of monosyllabic ones. In the latter case chance resemblances cannot be ruled out. 2) In historical studies it is necessary to state when a given item is attested. An item like Skt. mrga-: Ta. madeer' can be considered at all if it is shown that the latter is attested only since the post-Apabhramsa stage. 3) Wrong generalisations have led the author to unwarranted inferences. He states that the final short vowels of Sanskrit words of two or more syllable are quiescent (p. 17). But not realising that this is not true if the final voweld comes after a consonant cluster he expects the Tamil form of the Sanskrit word ustra- to become ollu- (p. 18). What actually occurs, however, is otta- which seems to be quite in order under the circumstances. 4) The author's etymological speculations are occasionally vitiated by inconsistency. For the derivation of ta, aravu and aravam.serpent' from Skt. * sarpaka- he imagines the intervention of a glide vowel (*sarapaka ), while for the derivation of Ta. pap from *prasarpa he apparently assumes assimilation and also certain other things (p. 21). The result is that the etymologies suggested by the author appear, on the whole, far-fetched. $ Bartholomae ( 887) treats pata ... ahmi, nipata ahmi, cited by Kanga, quite diffc rently. Cf. anvar stavas-tompa ( 140 ), stara-wat (1606), vib@ra -wat (1448, this one having a Vedic parallel vibhft-van). xsvipta-vat (562) is etymologically doubtful. The author has wrongly read Kanga's yanavat as sanavat (B. *yana-vant 1286). Madhu Vidya/649

Page #675
292 Annals BORI, LXV (1984) There is one item which we may take up for investigation. Op p. 20 the author observes that in Vedic Sanskrit vac meantmouth' and it is from this word that Ta. vay mouth' is to be derived. In support of his contention he cites a passage from the Srautapadarthanir vacanam p. 170. Since in the ritual, the author argues, the sacrificer's wife washes the mouth of the sacrificial animal to the accompaniment of the mantra vak ta apyayatam, the word vac in this mantra must mean mouth'. The mantra in question occurs in the Vaj. Su:n. 6.15 and it becomes clear that the Adhvaryu and the sacrificer repeat the mantra manas ta apyayatam, vak ta apyayaram etc. while pouring water on the animal's head, mouth etc. in order to make the mind, the speech etc. 'swell' in the head, mouth and other limbs of the animal. Similarly if the sacrificer's wife cleanses the animal's mouth while saying vacam le sundhami etc. ( Vaj. Sam. 6.14 ) it does not, on that account, mean that vac means mouth, but it only shows, as the commentator says, that the mouth etc. are supposed to be the abodes of the vital airs (pranayatanani) According to the Sat. Br. 3.8.2.6 the purpose of reciting the mantra is to put the vital airs and to revive them (tat pranan dadhati tat samirayati) in the dead animal's mouth, nostrils, and other organs which are considered to be the abodes of the vital airs. There is thus no qusetion of vac meaning' mouth' in the Veda. A few sentences in the book like. It is these particles and letters which are referred to as tense and mood signs in the tittle of this book' (p.5) are unintelligible. On p. v we are told that some material is missing in the manuscript of the book without indicating where the gap occurs. Probably this remark is to be understood in the light of the Editor's foot note on p. 539. M. A. Mebendale Madhu Vidya/650

Page #676
Reviews 293 SEXUAL ETHICS IN THE MAHABHARATA IN THE LIGHT OF DHARMASASTRA RULINGS: by Bhakti Datta, Asia Publications, London, 1979, pp. 124 + Annexe I-XVII (pp. 11 ) In this book the author deals with the question of the evolution of sexual ethics in the light of four stories selected, at the instance of Prof. P. Thieme, from the Adiparvan of the Mahabharata. The four stories are: (1) Bphaspati and Mamata, (2) Yayati and Sarmistha, (3) Sveta ketu and (4) Dirghalamas. The two particular points investigated by the author on the basis of these stories are: (1) whether the stories and their modific &tions indioate any pattern of evolution of the dharma concept, and (2) whether the modifications reveal any attempt to bridge the differences, if any, between the practices observed in the stories and the rulings of the Dharmasastras. For this purpose the author investigates the different stories on the following lines: the author first gives (1) the summary of the story, which is followed by (2) the translation of the relevant stanzas from the Maha. bharata. The author then offers ( 3 ) criticism of the text of the stanzas, and finally ( 4 ) the story is considered in the light of the Dharmasastra rulings. The author's conclusions are given at the end. They are as follows : 1. In the Mahabharata " legends with conflicting sexual ethics appear side by side without any attempt at explanation". (p. 110) (2) These legends reflect "changing scenes of social values indicating time-related evolution in sex morality" (p. 104). This evolution centres round the following four main characteristics : a) "Promiscuous situation when women were free outside the period of their riu. b) Privileged intimacy between brother-in-law and sister-in-law. c) niyoga'... d) niyoga' slowly giving way to restricted and chaste conduct of wife". (p. 111). 3. The interpolations in the stories reveal "a strong current of brahmanification emerging within the society." (p. 115) It is possible to agree with much that has been said by the author. However, it seems that the above conclusions stand in need of slight modi. fications. In the first instance, although it is true that these legends with conflicting morals stand side by side in the Mahabharata without any attempt at Madhu Vidya/651

Page #677
Annals BORI, LXV (1984) explanation, we have to observe that within each story "an attempt at explanation" is to be seen in the changes that are introduced into the legend by way of interpolations or later versions. 294 Secondly, it is not possible to say that all the stories in their interpolations or later modifications reveal a current of brahmanification. This would have been true if the characters in all the stories had belonged to the non-Brahmana class and then the changes introduced had shown an attempt to modify the behaviour of the characters so as to make it conform to the prescriptions of the Dharmasastras. But this is by no means the case. Some other observations seem necessary. 1. While dealing with the story of Brhaspati and Mamata the author has not noticed one deviation between the Adiparvan version (1. 98. 6 ff.) and the Santiparvan version (12. 328. 44 ff.). While in the former it is said. that the sage, while still in the womb of the mother, had studied the Vedas and the ancillary texts, in the latter he appears to have studied these sometime after the birth. It is possible that the author overlooked this deviation since it did not pertain to any aspect of sexual ethics. 2. The author has already stated (p. 18) that in the third version of the story of Mamata and Brhaspati in the Mbh., Utathya, Mamath's husband, is made to disappear in order to legitimize Brhaspati's behaviour towards Mamata. But there is one more thing to note. This version makes no mention of Brhaspati's having come into physical contact with Mamata. 3. On p. 29, the author observes: "After a thousand years Yayati got rid of his desire...". But this has no basis in the text (Mbh. 1. 70. 44) and, in fact, it goes against what is stated there (arrpta eva kamanam). 4. On p. 49, the author states that in the Mbh. 1. 70. 42 we are told that Yayati could transfer his decreptitude because of his own power of austerities, but that in the Mbh. 1. 78. 38, 40-41 king Yayati is shown to be quite dependent on the brahmin Sukra for this purpose. This modification, in the opinion of the author, is an example of the rising brahmanic supremacy. In fact there seems to be no modification. In the version of the story as it appears in the Mbh. 1. 70 we are told that the king's fulfilment of kama had remained deficient due to the curse of Usanas, but the details of the curse are not given. But we have no reason to assume that they were different from those narrated in the next version in the Mbh. 1. 77-78. There, the reason for the curse is said to be the king's flirtation with Sarmistha. Sukra makes it possible for Yayati to transfer his old age to anyone of Madhu Vidya/652

Page #678
Reviews 293 his sons by inteptly thinking of the sage ( mam anudhyaya cu bhavena 1. 78. 40). This cannot be said to contradict the statement of the earlier version where it is stated that the king transferred his old age by resorting to the strength of asceticism (tapoviryasamasrayat 1. 70.42). The author has wrongly assumed that this tapovirya belonged to Yayati. But that could not be the case. The king bad, no doubt, performed many sacrifices but that was not tapas. The king's resorting to the tapovirya in the earlier version is the same thing as his intently thinking of the sage' in the later version. Hence this does not seem to be an example of the growing ascendancy of Brahmanism. (Yayati practised penance only later in life 1. 70.694). 5. It appears that the author has misunderstood the significance of the maryada on the behaviour of women laid down by the sage Dirghatamas (Mbh. 1. Appendix 1. 56). His restriction' has to be understood in the light of the story told there. Pradvesi, the wife of Dirgbatamas, refuses to accept any money won by her husband and at the same time declares her intention no longer to maintain her husband as she did before. This means that she either wanted to take to another man or remain without a man. The purpose of the maryada proclaimed by Dirghatamas was only to prevent Pradvesi from taking recourse to any of these two alternatives. There is no mention that the maryada was intended to forbid any kind of niyoga. Hence the conclusion drawn by the author (p. 114) in this regard seems to be unwarranted. 6. On p. 116 the author states that Pandu appointed brahmins to raise issues on his wives. But this is not correct. It is true that in the Mbh. 1.111. 32, 36, Pandu says that he would like his wife to give birth to sons from some one either equal to him (a ksatriya ) or one superior to him (a brahmana ), but in fact, he permits Kunti to invite the gods (1.113.39 ) as suggested by her. 7. The author makes a mention of the shady profession' (pp. 116-117; also p. 97) carried on by Pradvesl, the wife of Dirghatamas. But it is not clear where the author got any evidence to cast such aspersions on Pradvesl's character. In the text cited by the author there is a passage which makes reference to the manual labour ( framenarta ) wbich Pradvesi had to do to support her husband and children. But nothing more than that. If the wife had to do srama to maintain her husband it does not mean that she led an immoral life. 8. Another example of reading too much into the text is offered when the author says about Dirghatamas that he " learot the practice of cattle and continuted to do so in public" (p. 110). Now it is gross exaggeration to 1 . Madhu Vidya/653

Page #679
296 Annals BORI, LXV (1984) use such an expression on the basis of the words vitathamaryada or bhinnamaryada vsed about the sage in the Mbh. 1.98. 1038. Violating the socially accepted norms of sexual behaviour does not mean that such acts were carried on 'openly' (p. 89), or 'in public'or, public cohabitation' (p. 99) was practised. The author has rightly decided to translate the relevant Sanskrit passages for being used in the text and not rely on the translations of other scholars (p. 5). The author's translations are, on the whole, accurate. But occa. sionally we meet with lapses. One may neglect a few cases of omissions,.c.g. the word kami of st. 1.98.12 is not translated on p. 12, or the word praksipya ( st. 1.98.18 ) is omitted in the translation on p. 861; one may also neglect slight inaccuracies like the translation holy' of dhl man (1.98.6 on p. 11 ) or "charioteer' (in its usually accepted sepse ) of maharatha ( st. 1.111,35 on p. 82). But one cannot say the same thing when the author translates avindat of st. 12.328.44 as appeared there before' (p. 14) instead of obtained' or even married'. yajato dirghasatrair me sapac cosanaso munen (st. 1.70. 38) is rendered as " through a curse of the hermit Usanas (at a time) while (I was ) sacrificing in long session'" (p. 31). What is really meant is " (of me) who was sacrificing with long sacrificial sessions and by the curse of the sage Usanas ". The same is true of the author's translation of the Mbh. 1.77.10 given on p. 34. asokavanikabhya se sarmis:ham prapya visthitah can impossibly mean having come near Sarmistha by the * Asoka-wood' (the king) emptied his bowels'. The author should have at least thought about the propriety of the translation when thereby the king is required to do an act near the wood in the presence of Sarmistha which he should have done in private. The line simply means .( the king ) having found Sarmistha near the Asoka grove stood there'. On p. 41 while translating mam anijnatum arhasi ( Annexe X, line 237) the author adds into the brackets the words (to go away). But the contexto shows that what we have to add is (to enter fire )'. The author uses (p. 71 ) the expression for your sake' to translate tvarkple of the Mbh 1.113.304. 1 Or, alternatively the word samavastjan is overlooked. 9 Thus the reasons given by Yayati for his kama having remained deficient are two (cf the use of ca), and not one as the author's translation makes out. avam ca sahitau devi pravisava hutasanam Rama. Uttarakanda. App. 1, No. 8, line 235. + Avatkrte 'han prthusroni gaccheyam putrinam gatim. Madhu Vidya/654

Page #680
Reviews 297 This translation is not accurate if.for your sake' means in your interest'. In the context, tvatkrte is better rendered when this is done by you (i.e. when you have given birth to sons)' or on your account '. The book is no doubt a welcome contribution to the understanding of the development of the notion of dharma so far as it relates to sexual ethics. As such it deserves a careful study. It is hoped that the author will continue to work in this field and study a larger number of the epic narratives. M. A. Mehendale Madhu Vidya/655

Page #681
Reviews 297 MAHABHARATA - 1 Adiparvan translated into Marathi and edited by R. S. Walimbe, published by the Vidarbha Marathwada Book Company, Pune, 1982, pp. 778, Price Rs. 120/ Quite a few Marathi translations of the great Indian epic are already available, and of these two have recently appeared as reprint editions : 1. the Chiplunkar Mandali's translation ( 1967-1972 ) and, 2. the one by Shri Bhalba Kelkar (1982). The publisher of the book under review has justified the appearance of one more Marathi translation on the plea that the language used for the earlier translations is rather literary, used by the Pandits in their writings, and that those early translations tend to be uniutelligible, when they are too literal or they are loose, because they strive to convey only the purport of the original text. The publisher therefore set himself the task of providing a new translation to the reader which is faithful in the extreme to the original text and yet is written in a style that will make it readable. He no doubt found in Prof. Walimbe the person most suitable to execute this task. As he points out Prof. Walimbe is known for his scholarship in Marathi as well as Sans. krit. He has already to his credit Marathi translations of some well-known classical Sanskrit texts. The translator on his part informs the reader in the preface that while translating the epic he has tried to avoid the two dangers noticed above, viz. unintelligibility and laxity. His principal aim, he says, is to lay bare in his trauslation what the author of the epic wished to convey to his audience. He assures the reader that while executiog this task he has taken utmost care to see that he does not say anything that would go against the original text. With these assuarances in mind when the reader goes through the translation he is no doubt satisfied with the readability and the intelligibility of the translation. But when one compares it with the original text one feels that the traoslator had at many places, perhaps, better. done otherwise. In a brief review like this it is not possible to list all such places. Only a few examples can be given by way of illustration. To start with very simple things, it is not clear why the words like ayatanani (1.12), sarvesam ca mahiksitam (1.13), Kurukseire (3.1), vagbhir Igbhih (3. 56), aprameyam ( 88.8) are left out in the translation. It is also not clear why the translator takes such small liberties with the text 1 The references in this review are to the Chitrasala press edition followed by the translator. Madhu Vidya/656

Page #682
298 Annals BORI, LXV (1984) when he renders the word sprhaniya- (88.10 ) as atulaniya, or janmana ( 89. 2) as due to birth in a noble family' (when what is meant is, as shown by the context, vayasa). As examples of wrong construeing may be cited the following: (1) The translation of 89.1 presupposes the text to be surasiddharsilokat prabhramtitah atah alpapunyah. But actually prabhramsitah and alpapunyah are to be separated and the latter is to be taken together with prapatami. (2) The last two quarters of 88. 7 belong together and form one sentence. But the translator makes two different sentences of them, as if the author of the stanza wanted to compare Yayati in point of lustre and appearance also with the sun, as he compares him with Indra and Agai in the first two quarters of the stanza. (3) ID 175. 2 the sentence really runs as idam akhyanan sarvesu lokesu puranan paricaksate. The translation, however, runs as "this story, banded down by tradition since very old times, is famous in the three worlds". (4) In the next stanza ( 175. 3) the construction is parthivah gadhiti visrutah loke. But the translation runs as "The king was named Gadhi. He was famous in the world". These are trifling little things, no doubt. But could they not have been easily avoided ? More regrettable is the translator's slip of the pen when while translating the very first line of the epic he says that Lomaharsana arrived at the Naimisa forest. In fact it is Ugrasravas, the son of Lomaharsapa, who arrived there. The translation of this line suffers from other defects too. The text of the original says that Ugrasravas who came to the Naimisa forest was well-versed in the Puranas (pauranikah). But the translator, relying entirely on the fanciful explanation of the word given by Nilakantha, credits the visitor also with the knowledge of the secret doctrine of the Uparisads. The translator renders the text word sautih as sutaputra' and yet he retains the word sauti in his translation. It is well known that This first line occurs again in the Mahabharata ( 4. 1). The translator has rendered the line there correctly. It is difficult to understand why the same line has been rendered differently at two places not far removed from each other. The translator informs the reader in the preface that he has made full use of the commentary of Nilakaptha. He is convinced that Nilakaatha has made a careful attempt to reveal the mind of the author of the Maha. bharata. The translator has therefore found it fit to follow him to a very great extent While no one will deny the advisability of consulting a commentator, Madhu Vidya/657

Page #683
Revlews 299 one wonders whether it was necessary to follow him to the extent the translator has done. This can best be exemplified by pointing to the translation of the twelve stanzas (3. 57-68 ) in praise of Asyinau. The stanzas have been composed clearly in imitation of a Vedic hymn, but this bearing is completely lost in the translation which is based on Nilakaotha's explanation. The well known reference to Asvinau with words like nasatya and dasra (3. 58 ) lose all their Vedic background when they are rendered as those in whom unreal and perishable things are totally absent 2; this is also true of the epithet citrabhanu (3. 57 ) rendered as who appear in the form of the diverse universe'. The allusion to the Vedic myth according to which the Asvinau saved a vartika from the mouth of a wolf is completely missed when vartika (3. 59 ) is rendered as 'jiva in the form of a female bird '4. The stanza 3.66 is not even translated fully.5 All in all one forms the impression that the translation of this piece is not that strict and faithful as it should have been. There is one advantage for a translator when he undertakes to render a Sanskrit text in an Indian language. He can conveniently retain in his rendering such words from the original text as kulapati or satra. There is no objection to the translator's doing this, but since one cannot always be sure what ideas such words would evoke in the mind of the general reader, they are better explained in the notes. Moreover one has to be on one's guard. While translating the expression prastavayan kathah (1.6) the translator says . while doing the prastavana of stories' which will be taken to mean while introducing the stories'. What is actually meant in the text is while initiating the conversational talks'. The notes meant for nearly the first half of the translation are given at the end; the rest appear in the form of foot-notes. Most of the notes are on proper names. While these in themselves are useful, it would have been better if the translator had extended the scope of his notes. His last note is on 125.31 relating to the cremation of Pandu and Madri's ascending the funeral pyre. He points out that the account given here is not in harmony with what is said a little later in 126.4. There it is said that the dead bodies asatyam rajjubhujangadi dasram upaksayadharmakan ghatadi tadubhaya bhavar pau--Nilakantha. 3 vicitraprapancakarena bhasamanau-Nila. vartikam ( suparnasarapam ) jivapaksinim-Nila. 5 The word nasatyaut, therefore, which occurs twice in this stanza, is not translated. The translator describes this as Madris jumping into the funcral pyre, which does not exactly reflect what is found in the text. The original simply says anvarohat which can impossibly mean jumped'. Madhu Vidya/658

Page #684
300 Annals BORI, LXV (1984) were taken to Hastinapura. To resolve the contradiction, Nllakagtha interprets the word deha as asthini (bones) and the translator finds this interpretation quite satisfactory. He, however, admits that this whole account is very confusing, Instead of agreeing with the commentator, the translator could have referred to Dr. V. S. Sukthankar's remarks on this passage in his Prolegomena (p. LXXXVII). The translator has taken for his basis the Chitrasala press edition of 19287. He does not state the reasons for his choice. One therefore does not know why he did not give preference to the critical edition published by the Bhandarkar Oriental Research Institute. One reason cau be that the trapslator chose that text which is widely known in Maharashtra. But since the B.O.R.I. edition offers the reader the critically constituted text it would have been better if the translator had accepted the critical edition as his basis and popularised that text among the general readers. A golden opportunity, one feels, bas been regrettably lost. M. A. Mebendale Madhu Vidya1659

Page #685
310 Annals BORI, LXVI (1983) THE MAHABHARATA - THE STORY AND ITS SIGNIFICANCE : by S. L. N. Simba, Birla Academy of Art and Culture, 108-109, Southern Avenue, Calcutta 700 029, 1983, pp. xii, 182. Price : Rs. 25/-, $ 5. As indicated by the sub-title, the book parrates, first, the story of the Mahabharata (pp. 8-73) and, next, trics to make clear the significance of the epic for the present day. For this purpose, the author describes briefly but pointedly the chief characteristics of the principal figures in the epic, narrates some important fables and moral tales, and then informs the reader what the Mahabharata is as a Dharmasastra and a Yogasastra. This is a very readable account. It is well known that the Mahabharata contains mucb matter that is not essential to the main narrative. Yet there is no doubt that this is considered by many to be the inost important aspect of the epic. The author shares this view and is not prepared to accept that the didactic portions in the epic are interpolations. He puts it the other way round i "the story is subscrvient to the didactic theme" (p. 79). As regards the homogeneity of the book the author is quite emphatic : "the Mahabharata is not a lose collection of various themes, but a book of remarkable unity of conception and treatment" (p. 80). It would be difficult to agree with this view. While writing on the Mahabharata as a dharmasarstra, the author rightly cmphasizes that it is necessary to consider the term dharma in its widest sense. While writing on danadharma the author says that the Mahabharata wants to stress that " Giving is not charity, but an act of duty. Disinterestedness is its main quality" (p. 137). Although the author has great reverence for the Mahabharata, he does not fail to point out that as far as stridharma is concerned, the prescriptions of the epic are one-sided, making it appear that "women exist only for the benefit of men, whom they should serve with ihe utmost devotion and loyalty" (p. 136). The author on various occasions, especially while delineating the principal characters of the Mahabharata, bas sought comparisons with the Ramayana. This forms an interesting feature of the book. In order to enable the general reader to appreciate these comparisons the autbor gives at tbe end of the book the suminary of the Ramayana (pp. 163-182 ). The author has made it clear in the Preface (p.ix) that for the purpose of writing the book he has almost wholly relied on English and Kannada translations of the epics. M. A. Mehendale Madhu Vidya/660

Page #686
Reviews 311 SECONDARY TALES OF THE TWO GREAT EPICS : by Rajendra 1. Nanavati, L. D. Institute of Indology, Ahmedabad - 380009, pp. xii, 195. Price : Rs. 50/ Shri Nanavati discusses fully in this book some 'secondary' tales in the two epics - the Ramayana and the Mahabharata. By the word secondary' he meaas'any parrative that does not form part of the original story' (p. 15). A little later (p. 27) he wishes to define the term as that tale which is complete in itself, independent of the original tale'. This has been elaborated towards the end of the book (p. 163 ) as that tale or episode may be called secondary which is not original, which is not connected in any way with the principal characters, or more correctly, with the central cvent of the epic, or again, shows stylistic differences or creates internal contradictions or unnecessary duplications etc.' The author has restricted himself in this book to the tales in the Balakanda and the Uttarakanda of the Ramayana and those in the Adiparvan and the Sabhaparvan of the Mahabharata. In the author's opinion these tales represent nearly all significant varieties of tales' (p. 27) in the two epics. The author has taken great pains to make a full study of all the implications of the tales. He seems to adopt in his treatment a historical, a functional, and where pertinent, a comparative approach. His view is that such a study "must be related to its (i, e, the tale's ) mythical, ritualistic, traditional, linguistic aspects so as to decide whether it yields any real bistorical data or has ritualistic iconisation, or symbolises some philosophical concept or metaphorises some actual event" (p. 165 ). As a result of a full-scale study of the tales on these lipss his conclusion is : "Cumulatively speaking the epic-stories are more often than not personications or concretisations of some Abstract Vedic ideas. Philosophical concepts or sacrificial rites are often put in the concrete form of a story" (p. 166). It will not be possible to enter into a detailed discussion of the tales and say whether or not the author has succeeded in proving his poiat of view. One may only note that the author occasionally makes statements which may not carry conviction. On p. 26 the author points out what he thinks to be the basic similarity in the motif-pattern of the Ramayana, the Mahabharata, and the Nala story. According to him, jealousy of the antagonist is the pivotal emotion in all the three tales. But the jealousy of Kaikeyl, or for that matter of Kali, for Rama, and Nala is not the same as that of Duryodhana for the Paodavas. Again he says that the jealousy makes the antagonist use some trick against the protagonist. But neither Kaikeyi, uor Kali, por Sakuni. Madhu Vidya/661

Page #687
312 Annals BORI, LXVI ( 1985) made use of a trick (Sakuni won the game by his great skill and we have Yudbisthira's testimony to the effect that Sakuni did not employ trick na manyate tam nikTtim mahatma 2. 60.420). The author's statedient " Duryodbana asks Vidura to bring Draupadl in the assembly. hall to receive commands from her new masters" (p. 153) does not agree with the text of the epic. When Duryodhana asks Vidura draupadim anayasva (2. 59. 1) he does not mean get her to the assembly-hall)'. This is made clear when Pratikamin conveys Duryodhana's message to Draupadl with the words sa prapadya ivam dhitara strasya vesma nayami tram karmane yajnaseni (2. 60.4). It is only when Draupadl raises the question of her social status that Duryodhana asks her to come to the Assembly to get the answer to her question (2. 60. 10). It would have been to the author's great advantage if he had utilized the treatments by H. Luders of the Rsyasaga episode and of the concept of ria. A list of abbreviations would have been useful. The book is happily free from glaring misprints. An occasional slip like "The word R$ya, spelt morc correctly as Rsya" (p. 41 ) leaves the reader guessibg. The author's work is admirable. The L. D. Institute of Indology may justifiably feel happy about its publication. M. A. Mehepdale THE BRAHMANDA PURANA : trapslated and annotated by Ganesh Vasudeo Tagare, Motilal Banarsidass, Delhi, Varanasi, Patna, 1983, Part I, pp. Ixxxii, 1-389; Part II, pp. xiii, 395-719; Part III, pp. xv, 721-953; Part IV, pp. xv, 957-1223, Part V, pp. xxiii, 1225-1399. The present publications in five volumes are in continuation of the laudable epterprise undertaken by Messrs Motilal Baparsidass to publish in a series English translations of Sanskrit texts related to Ancient Indian Tradition and Mythology. They have already published the translations of siva, Linga, Bhagavata, Garuda, Narada and Kurma Pusanas comprising the first 21 volumes in this series. As said by the Editor, the series has been started Madhu Vidya/662

Page #688
314 Annals BORI, LXVI (1985) chariot'. But the word saptarasmiratha) seems to mean whose chariot has i, e. is controlled by seven reins.' (5) Onp. xlvi the translator cites a stanza drstani vai etc. in which occurs the word mahatman. This he readers as great soul'. But this word is not likely to be a poun. It should have been reodered as high-souled' (adj.). The General Editor of the Series, Prof. J. L. Shastri, in the Prefaces to the five Volumes briefly indicates the contents of the respective volumes. M. A. Mehendale THE KRSNA CYCLE IN THE PURANAS : by Benjamin Preciado Solis, Motilal Banarsidass, Delhi, Varanasi, Patna, 1984, pp. xii, 151, and plates 75. The sub-title of the work Themes and Motifs in a Heroic Saga' is quite modest. The work, in fact, is much more comprehensive. It deals with "the stories of Krsna, their origins and formation, the themes and motifs that constitute them, and their representation both in art and literature, specially in the Puranas up to the tenth century" (p. 124). The author has done full justice to what he has aimed to do. In Chapter 1 he discusses the Vedic Antecedents of Krsna and comes to the conclusion that although the main traits of the Vedic deities Visnu and Narayana are reflected in the later Krsoa stories, the name Kropa in the Rgveda cannot be said to refer to the hero and the god of the Puranas. In Chapter 2, the author examines the next phase of the literary cvideace beginning with the Chandogya Upanisad (3.17.6 ) and ending with the Mahabhasya of Patanjali. He also draws on the epigraphic evidence like the Besanagar column and the Ghosundi stone inscriptions. It was during this period that Krsna began to acquire the form in which he is widoly known. Towards the end of the chapter the author establishes for this period the chronology of the development of Krsaa as a god and of the legends connected with him. . Io the following two chapters (3, 4) the autbor has inade an admirable survey of some characteristic themes and motifs in the Ktsa stories as they occur in the Harivamsa, the Visnu, the Bhagavala and the Devi Bhagavata Purapas, the Ghata Jataka and the Balacarita. He also compares these motifs with their parallels in the stories of Indra in India and of Heracles in Greece. Madhu Vidya/663

Page #689
Reviews 315 In the last chaptar (5) the author studies the representation of Kroga legends in the sculptures up to the tenth century. He has, for this purpose, included also the representations found in the Lara Jongrang temple at Prambanam in Java. This iconographic study opens with the representations of Krspa and Balarama on the coins of Agathokles discovered at AiKhanum on the Oxus river in Afghanistan in 1970. In his detailed study of the sculptures the author rejects some of the earlier wrong identifications, e. g. of an amorous couple io a sculpture at Pabarpur, Bangladesh (pl. 40), with either Radba and Krsna or with Rukmini and Krsaa; or be suggests some new identifications, e. g. of the representation of the Pralambasuravadba in the central scene on the pillar at Mandor, now at the Jodhpur Museum in Rajasthan. The author identifies the female figure in this scene (pl. 8) with Radha which, if accepted, would be her earliest known representation io sculpture (4th century A. D.) The author's appraisal of the evidence before him is generally strict. He scems to have slightly relaxed his rigour when on p. 36, column 2 bottom, he is willing to take the tradition of a certain Krsma, a hero of the Vrsnis of the Yadava tribc as far back as the Vedic period since the Yadavas (actually Yadus or Yadva ) and the Bhojas (?) are mcationed in the Rgveda. The author shows acquaintance with the wide range of literature connected directly or indirectly with the various details of the subject of his research. The author says that he undertook this study at the suggestion of Prof. Basham. Readers will agree that Prof. Basham's suggestion was directed to the right person. There is an unfortunate slip on p. 125, column 1, line 10, where we find killing of Klsna' instead of killing of Kamsa'. M. A. Mehendale Madhu Vidya/664

Page #690
316 Annals BORI, LXVI (1985) ERICH FRAUWALLNER NACHGELASSENE WERKE, Vol. I Aufsatze, Beitrage, Skizzen: edited by Ernst Steinkellner, publi shed by Die Osterreichischen Akademie der Wissenschaften, Wien, 1984. After the death of Prof. Frauwallner on July 5, 1974, his literary Nachlass was first handed over to the editor and next it was gifted to the Institut fur Tibetologie und Buddhismuskunde of the University of Vienna. This Nachlass, the editor informs us, consists mainly of four parts: 1. Materialsammlungen, 2. Arbeitstexte, 3. Ubersetzungen, and 4. Darstellungen. It is not the intension of the editor to publish Frauwallner's entire Nachlass. What will be published would be chiefly those texts which Frauwallner himself had finally intended to publish. This will be done in three volumes. Volume I, which is already published, contains the following titles: I Uber den geschichtlichen Wert der alten ceylonesischen (pp. 7-33), 11 Der ursprungliche Anfang der Vaisesika-Sutren chroniken (35-41), III Der Navyanyayah (43-36), IV Der Navyanyayah, ein Artikel fur das "Worter buch der Philosophie (57-61), V Geschichte der indischen Philosophie, IV Band, Nachgelassene Skizzen (63-133): J. Erkenntnistheorie und Logik der klassi schen Zeit (66-87), 2. Tantrayuktayah (88-92), 3. Sprachtheorie (93-119), 4. Mimamsa (120-122), 5. [Sprachtheorie des Vrttikara ]1 (123-124), 6. Kumarila (125-126), 7. Dharmakirti (127-133). Each of the titles is preceded by the Editor's prefatory remarks which contain useful details. In the text proper the Editor has not made any changes except completing the incomplete references and partly modifying the references to literature since Frauwallner had often used editions not easily accessible today. The present volume will be followed by one containing "Texte indischer Philosophie in Ubersetzung" to be edited by Gerhard Oberhammer. and a third one containing select "Vorarbeiten und unvollendete Skizzen " meant for the first part of the third volume of the "Geschichte der Indischen Philosophie" which should have dealt with the rise of the Hinayana system. Apparently the third volume will be edited by the editor of the present volume. 1 The title is supplied by the Editor. Madhu Vidya/665 M. A. Mehendale

Page #691
POESIE SANSKRITE CONSERVEE DANS LES ANTHOLOGIES ET LES INSCRIPTIONS, Tome III : by Ludwik Sternbach, Publications De L'Institut De Civilisation ladienne, Serie in-8", Fascicule 51, Paris, 1985. pp. xlii, 415. The list of the authors in this volume giving stanzas ascribed to them in the anthologies and the inscriptions begins with Yampyaku (1138) ( Papaka 788) and ends with Hexidhanesora (1903). This volume also notes a few additions (1904-1919) to the first two volumes and has an Errata to Vol. II. The first two volumes were published in 1980 and 1982*. The present volume has been prepared on the samne lines as those followed in the first two. As has been noted by the author in his Introduction (p. xv) published in Tome I, the purpose of these volumes is " d'enregistrer les vers des pocies, soit inconnus, soit connus, sous la forme qui est la leur, dans les anthologies et les inscriptions." This purpose has been admirably achieved in this as well as the other two volumes. All of them bear testimony to the indefatigable industry and great devotion of the author. There is no doubt that the three volumes will be highly useful as reference works. M. A. Mehendale BHRGUS - A STUDY : by Jayanti Panda, B. R. Publishing Corpora tion, Delhi 110052, 1984, pp. xii, 200. Price: Rs. 120/ The Bhrgus or the Bhiirgavas form an important family of the ancient secrs known since the Vedic times. Their influence extends to the following periods and is reflected in the cpics and the Puranas. The authoress has made a bold attempt to cover this entire period in the first four chapters of her book in the last chapter - Conclusion -- she calls attention to certain important aspects of the Bhargavas such as their relationship with the deitics, * These have been noticed by Shri S. V. Sohoni in the ABORI LXV (1894), pp. 323 327, Madhu Vidya/666

Page #692
their initial close relationship with the Ksatriyas later turned into hostiliy, their mastery over peculiar sciences like the sanjivani vidya, as also on archery. "Thus the Bhargavas are pictured as sages, domineering, arrogant, unbending, and revengeful in nature. Simultaneously they are shown to be omniscient, omnipotent and supermen. ... Besides these qualities they are well known as the promulgator of Dharma and Nitisastra... The tendency of deification of Bhrgu which originates in the Vedas reaches its final stage in the present versions of the Puranas." (p. 181) " The book contains some unfortunate lapses which could have been easily avoided. On p. 4 we find bhrgave mentioned as the plural form. On p. 53, Finding the demons unarmed" is followed by the word pragrhitayudha, which means just the opposite. On p. 78 the name of the lake is given as Brahmasara instead of Brahmasaras (Mbh. 13. 96. 7). It is difficult to understand what the authoress means when she says: "It appears that the descendants of Bhargava Cyavana unanimously take Ksatriya wives. " (p. 178). Equally unhappy is the sentence... and Narayana incarnates to restore the fallen dharma" (p. 181). A HISTORY OF CLASSICAL SANSKRIT POETRY, SANSKRITPALI-PRAKRIT: by Siegfried Lienhard, published by Otto Harrassowitz, Wiesbaden, 1984, pp. 307. The present volume makes fascicle I of Vol. III of A History of Indian Literature' edited by Jan Gonda. M. A. Mehendale "The poetry presented here is that poetry which is partly epic, partly lyrical. It is either written. entirely in verse, entirely in prose or alternates between verse and prose. Formally it falls into two main categories: poetry of the major form (mahakavya ), which is almost always sub-divided into sections, and poetry of the minor form (laghukavya ), whose important relative is single stanza and short poems" (p. 2). In his concluding remarks (p. 273), the author notes that he has excluded from his treatment classical 1 The expressions that actually occur in the Puranic texts are myastabastrah, tyaktaia stra. 49 [Annals BQRI] Madhu Vidya/667

Page #693
Sanskrit drama, inscriptions (prasasti) and letters of spiritual instruction ( lekha ). These two stateraents together give an accurate description of the material presented in the volume. The author believes that classical poetry did not begin with epic works of considerable length, as has been usually assumed, but with laghukayya, above all with muktaka, the one-stanza poem. It is from here that the classical poetry blossoms into the mahakavyas composed in the sargas, from wbere it proceeds to the mahakavyas in prose, ultimatley to end in the mixed variety known as Campu. Accordingly the author presents bis description and analysis of the material in the following order : Chapter III Poetry of the Minor Form, Chapter IV Poetry of the Major Form, Chapter V Poetry of the Major Form - Prose, and Chapter VI Poetry of the Major Form - Campu. His first two chapters are intended to offer a proper background to the create ment of the subject. To his prefatory note (p. v) Lienhard observes that the General Editor of the scheme had told him that the volume was meant to be for both laymao and scholar alike. On going through a major portion of the work one can confidently say that the author has done full justice to the General Editor's request. While writing on the beginning of the Kavya tradition (p. 53 ), Lienhard lists four arguments which are meant to demonstrate that the Rama. yana "can hardly be said to represent poetry of the major form in the narrow sease". Of these, only the first argument - the Ramayana does not consist of sargas but of kandas - seems to be relevant. It is not clear how his second argument that many parts of the Ramayana have been interpolated can have relevance for the point to be proved. His third argument that we know early kavyas that antcdate Valmiki's work will only show that the Ramayana may got be called the adikavya, but this certainly cannot be used to disprove its being a mahakavya. His last argument that the form in which classical poetry arose was not mahaka vya but laghukavya will only show that such laghuka vyas antedate mahakavyas but will not, on that account, disqualify the Ramayana being called a mahakayva. The author has profusely illustrated his work with citations from the different works dealt with by him and these citations have been translated into English. While these translations are, on the whole, good, occasoinally one comes across certain lapses. The word sikha ( Raghuv. VI. 67 ) is better. rendered as filame' tbao as 'rays' (p. 36); similarly himadhaman (Sisup. IV, 20) is not place of cold' (p. 36), but whose rays are cool'. On Madhu Vidya/668

Page #694
p. 141 ne quotes, apparently approvingly, W. N. Brown's traoslation of the Mahimaastava 24. The word nama occurring in the third line is surely an cmpbatic particle and it does not mean 'name'. Rather unfortuante is the author's assumption (p.40 ) that the Sanskrit word suhrd ( Meghaduta II. 16) is vowel-ending (suhrda. ) and not consonant-coding. When be cites the Meghaduta stanza he reads it as suhrdah, which is metrically defective, instead of as suhrd val. On p. 9 Liephard cites Nitisataka 60. While translating it he omits line 2, and while commenting on its composition says that it "cods in true praise of these genuinely noble beings who are above fraud and deceipt" (ital. mine ), but the stanza does not have anything of this sort. The book is happily free from misprints. The few that occur (like the one noted above suhrdah for suh!d vah, or rambha' for 'rarambha in Nitisataka 60 op p. 9, or gadas ya for gadyasya p. 231, f. n. 5) can be easily corrected. The use of kapalaka "the group" for kalapaka (twice on p. 66 and once on p. 67 ) needs to be pointed out. The writing of an author's pame as Jayamaghava instead of Jayamadhava (on pp. 24, 25, 26, 28, also Index p. 280) has arisen due to misrcading of the Devadagari dh as gh. These are, however, minor points. The author has taken considerable paios to go through a larger number of primary texts and an equally large number of items of secondary literature and has produced an admirable account that is critical, informative and pleasing to read. M. A. Mehendale Madhu Vidya/669

Page #695
Reviews THE MAHABHARATA, ITS GENESIS AND GROWTH-A Statistical Study by M. R. Yardi, Bhandarkar Oriental Research Insti tute, Poona, India, 1986, Pp. viii. 2, XIV, 235, Price Rs. 110/-. 349 The Bhandarkar Oriental Research Institute undertook to publish the Critical Edition of the Mababharata with the objective of making available to scholars the oldest form of the text which could be reached on the basis of the available manuscripts (Prolegomena to Adiparvan p. LXXXVI). The result thus obtained was never claimed to be the Ur-Mababharata, not even the form in which the epic was narrated by Sabti in the Naimisa forest. It was admitted that the critical edition contained interpolations which must have been added to the original text, if there ever was one such, at a time which the manuscripts cannot reach. Hence in order to rid the Ur-text of these accretions an approach different from the one based on the manuscripts was necessary. Shri Yardi believes that a statistical approach would enable us not only to identify the original Bbarata and the interpolations but also determine the chronology of the additions. (Preface, p. vii). Shri Yardi's statistical study is based on the form of the Anustubh, the most predominant metre of the epic. He has excluded from his analysis stanzas longer than the Anustubh and prose passages (as also Adhyayas having Jess than ten stanzas). The Anustubh, as is well known, has four quarters, each having eight syllables. In a quarter the fifth syllable is required to be short, the sixth long and the seventh alternately long and short. For the purposes of his statistical study, Shri Yardi has taken a line, i. e. two conse cutive quarters forming half the sloka, as a unit. This means that in his unit of 16 syllables, those syllables occurring at places 5, 13, 15 will be short, at 6, 7, 14 long. The remaining syllables can be either short or long. In a line, then, since in 10 out of 16 places the poet is free to use short or long syllables, this can give rise to many stylistic variations. Shri Yardi has based his study on these variations. Shri Yardi started his statistical study of the critically constituted text some twelve years ago. The results of his preliminary investigations appeared, beginning 1978, in different journals. All these have been reprinted as appendices to the present work which contains full results of his study. In a brief Introduction (pp. I-XIV) the author tells us that as a result of his analysis he has been able to identify five different styles which he characterizes as A, Alpha, B. C, and Beta. Accordingly he concludes that there were four revisions of the original text composed by Vyasa or Vaisampayana. The four revisions were made by Suta Lomaharsapa, his son Sauti Ugrasravas, Madhu Vidya/670

Page #696
350 Annals BORI, LXIX ( 1988 ) the author of the Harivamsa, and the author of the Parvasangraha-parvan, in that order. In the first five chapters the author gives an account of the contents of the Ur-Mahabharata ( or should it be Ur-Bharata ?) and the additions made to this original text in four successive stages. This is followed by five very important and informative chapters on * The Kauravas and the Pandavas', . Vasudeva-Krsaa and Rudra-Siva', * History of the Bharata War', Dating the Epics', and the Date of the Bharata War'. These chapters have an independent importance of their own and merit a separate review at some later stage. The present reviewer cannot say anything about the operations carried out by the author in his statistical analysis of the epic. He will take for granted the correctness of the five different homogeneous styles identified by the author. But he looks askance at the conclusions arrived at by the author on the basis of these identifications. It has been believed since long (cf. e. g. C. V. Vaidya- The Mahabharata : A Criticism, pp. 1-8, 1904 ), that the Mahabharata itself mentions three different titles of the epic -- Jaya, Bharata, and Mahabbarata, and that these were composed respectively by Vyasa, Vaisampayapa and Sauti Ugrasravas. Shri Yardi rightly dismisses (p. VI) the view that Jaya' is mentioned as the name of the pic in the lide jayo nametihaso 'yam ( 1. 56. 19 ) and agrees with Dr. V. S. Sukthankar in looking upon Jaya not as a specific name of the epic but a generic name applicable to different works of this type. (Adiparvan, p. 989). It has also been believed, again erroneously and since long, that the Mahabbarata tells us that the Bharata composed by Vaisampayana had an extent of 24000 slokas (1. 1. 61 ) and that the extent of one hundred thousand slokas, as given in the colophon of the Adiparvan ( fatasahasryam samhitayam), was reached by the epic when it was expanded with the addition of the upakhyanas by Ugrasravas. The epic itself says nothing of the sort. In the first instance it must be noted that the epic nowhere makes a clear-cut distinction between the two titles, Bharata and Mahabharata. It calls, e. g., the composition of Vyasa Bharata in 1. 1. 17, 56, 58, 199, 201, 1 The extent of this Jaya was supposed to be eight thousand eisht hundred stanzas ( astou slokasahasrani astau slokasatani ca 1. App. I. I f. n. 1, lime 15, p. 884 - Bombay edo, 1. 1. 81 ). ? So also Shri Yardi, Preface p. i. Also 1. 1. 27; 1. 56. 13. Madhu Vidya/671

Page #697
Reviews 351 206, 208; 53. 51; 2. 242; it calls the same text Mohabharata in 1. 1. 10; 53. 32, 35; 56. 1, 30-32. Secondly, it makes no distinction between the story as narrated by Vyasa and Vaisampayana. The stanza 1. 1. 61 -- caturvimsatisahasrin cakre bharatasamhitam / upakhyanair vina javad bharatam procyate budhaih 1l - has been erroneously interpreted to mean that Vaisam. payana originally composed the text of 24000 stanzas and that it was then free from upakhyanas. That is not the case. The stanza only means that the sage Vyasa, after having composed the entire Bharatasambita ( which also contained the upakhyanas ), himself made (cakre ) a shorter version of it, having 24000 stanzas, omitting the upakhyanas. Some * learned mea' call this abridged versionBharata', which, by implication, means that these learned men' call the upabridged Bharatasambila Maha. bharata'. It is quite clearly in this sease that the Asvalayana Gphyasutra 3. 4. 4 refers to two texts -- the Bharata and the Mahabharata. But, as seen above, the Mahabharata itself makes no such distinction. The epic is on record to point out the identity of the two titles by pointiog out that the composition, besides being bharata, is mahat and hence is known as maha. bharata (mahattvat bharatatvacs ca mahabharatam ucyate 1. 1. 209 ). In fact if the epic statement is to be believed Vyasa's text contained not only this upakhyanas, but also khilas like Harivamsa? and Bhavisyat, but that the Suta parrated only the eighteen parvans in the Naimisa forest ( 1. 2.69, 233; 1. 2.71 ). If then the epic recognizes only one title (maha)bharata for the composition of Vyasa and two reciters Vaisampayana and Suta Ugrasravas, it is difficult to see how Shri Yardi asserts that the epic gives evidence for Su a Lomaharsaga also as one of its authors. It is true that the epic informs us that Suta Lomaharsana told the story of Astika to his son UgraSravas ( 1. 13. 7).' But this gives no ground to assume that Ugra sravas The Suta also knows the shorter and the longer versions (1. 1. 23, 25 ) made by Vyasa himself (1. 1. 49). 5 The critical edition accepts hesitatingly the reading bharavatvat. Dr. Sukthankar defends this reading against Wioternitz who favours bharatatont. But he admits that the manuscript evidence is almost evenly balanced ( A BORI 11. 179--180 ). For other etymologies of mahabharata cf. 1. 56. 31, 93. 46, and App. I. 2.3-4. The term Harivamsa in these references quite clearly does not refer to the whole of Harivamsa, but to on: of its three parvans. It is curious that on both occasions where the khilas are named Visnu parvan is not mentioned. It is once mentioned in some Dorthen mss. 1. 2. 177*. Although there are clear statemeats to the effect that Vaisampayana Darrated the story to Japaniejaya at the instance of Vyasa (1. 1. 18, 57-58; 1. 54. 21-23), there are indications that Vyasa himself narrated the story (1. 1. 59-60; 1. 53. 31-33), 0 Ugrasravas had also heard from his father, who was versed in the Puranas (1. 5. 1). the narration of the Bhrguvamsa 1. 5. 5. Madhu Vidya/672

Page #698
Annals BORI, LXIX (1988) learnt from his father the epic as expanded by him by making additions to Vaisampayana's original text. There is no statement in the epic to this effect. According to the epic, Suta Lomaharsana does not figure at all in the line of transmission of the text. 352 That Vaisampayana and Sara Ugrasravas (Sauti) are among the authors of the epic is admitted at all hands. Now what has the statistical analysis to say about their share of composition? It wants us to believe that Vaisampayana, the author of the Ur-Bharata, composed only 13 of the extant 18 parvans and, what is still more striking, he had no hand in the composition of the two very important parvans of the epic the Sabba and the Aranyaka ! How can Janamejaya understand the epic story if the Incidents of the Sabhaparvan are omitted from the narration? It is not known how Shri Yardi in his account of the Ur-Mahabharata can include the following lines if the original text did not have parvans 2-4: "... who (i. e. Asura Maya) in gratitude built for the Pandavas a celestial palace. Duryodhana, on beholding that palace, became jealous and desired to possess it. And so deceiving Yudhisthira in a dice game with the help of his maternal uncle Sakuni, he contrived to send the Pandavas to the forest to spend a period of twelve years in exile and one more year in disguise." (p. 5). This much about the omission of the two entire parvans in the ori ginal composition. If one goes through the contents of the parvans ascribed to Vaisampayana one may come across some serious omissions. The one such is to be found in the Dronaparvan. Thus the original Bharata is supposed to have contained Adbyayas 11-25, 50-51, but it did not contain 26-49. This means that the original epic had nothing to say about the killing of Abhimanyu. This would be a very glaring omission and it is hardly believable that none of those who heard the narration asked Veisampayana about the death of Abhimanyu on the battle-field. Even Suta, who according to Shri Yardi was the first to make additions did not think it necessary to say a word about the killing of Abhimanyu, and it was left to his son, Sauti Ugradravas, to make good the omission in his third redaction. Since family discord (bheda) and the destruction of kingdom (rajyavinata) form the kernel of the epic story (1. 54. 19, 22-24; 55. 4-5), and since the brief narration of the incidents in 1-55 stops at the death of Duryodhana, it is likely that the Ur-Bharata ended there and that Vaisampayana had no hand in the composition of the parvans 11-18. But whereas the statistical analysis points to Vaisampayana as the author of some of the adhyayas of Santi- (12) and Anusasana-parvans (13), the contents of which Madhu Vidya/673

Page #699
Reviews are not even remotely connected with the epic story it does not assign to him any share in the parvans 11, 14-18 the contents of which have some relationship with the epic story! According to the line of transmission revealed by the statistical analysis, Suta Lomaharsana, whom the epic does not recognize as the reciter, occupies the second place. He is supposed to have made good practically all the deficiencies in the text of his predecessor and gave it a form of 18 parvans. His son Ugrasravas, whom the epic recognizes as the reciter, is then left with the task of only making some additions to the text of his father. This he did to parvans 1-7 and 12-14, but none at all to 8-11 and 15-18. 353 What strikes one very pominently is that according to the statistical analysis the Karpaparvan was left absolutely untouched by all the four redactors who followed Vaisampayana. At least the statistical study has not been able to reveal any such additions. All these are not realistic conclusions which would inspire confidence in the method. On the other hand it has to be observed that the analysis does give some realistic results like the Bhargava legends turning out to be later additions. After Shri Yardi subjected the constituted text to statistical analysis and identified five different styles, one thought he would carry out the same operations on the longer passages considered, on manuscript evidence, to be interpolations by editors of the critical text. He should have thereby demonstrated that none of these interpolated passages was composed in a style identical with any of the five styles used for the constituted text and thus even on the evidence of style they proved to be interpolations. If this does not turn out to be the case it would mean that imitation of style wes possible and hence identification of authors on the basis of statistical analysis alone is not possible. Moreover, there does not seem to be any reason why an epic poet should go on monotonously composing in a single style when so many different alternatives were available to him. The statistical analysis reveals that the upakhyanas were written in three styles, alpha, B and in C (p. VIII). Of these, alpha style has been assigned to Sata and B to his son, Sauti. If Sauti had bis father's text before him, would he not, for a change, somewhere pick up the style of his father? And, as was mentioned above, if Suta Lomaharsana had no hand in the transmission of the text, we have to say that Sata Ugrasravas used two styles for the upakhyanas. And further, 10 Except the last adhyaya (154) which narrates the death and cremation of Bhisma. Madhu Vidya/674

Page #700
354 Amals BORI, LXIX ( 1988) if the adhyayas from the Santi- and the Anusasana-parvans are not likely to have been composed by Vaisa mpayana, it would mean that Sauti composed ip A style as well. All this has not been said to suggest that the constituted text was composed just by two authors - Vaisampayana and Ugrasravas It is not disputed at all that the original composition underwent various revisions at different periods and different places. The epic itself gives perroission to do that. It characterizes the Bharata-Samhita (1. 1.61), not only as equal in extent 'to the four Vedas '(vedais caturbhih samitam)", but calls the samhita itself a. Veda'wbose author was Krsna Dvaipayana (1.1.205, 1:56. 17). The epic advises the reciters to expand this. Veda' of Krsna with additions of itihasas and puranas (itihasapuranabhyam vedam samupabrmhayet 1, 1. 204).12 To do this the epic reciter bad to be well-versed in composition of the type of itihasas and puranas. If that were not the case the Veda' (of Krsna ) is afraid that this reciter will pass over me, will skip me (i. e, will abridge me ) ( bibhely alpasrutad vedo mam ayan pratarisyati 1. 1. 204 ). Accordingly the epic has, in the course of its long history, experienced many * expansions' (samupabrmhana ). It is hardly possible to limit these just 10 four and demarcate very neatly their individual extent, in addition 10 ascribing them to definite authors. One word at the end regarding an assumption made by the author. which a layman, like the reviewer, not familiar with the statistical studies, is not likely to understand. When the user of the Anustubh metre was free to use short or long syllables at certain places, why does the author say: * The stylistic variations in the Anusubh slokas could arise from the natural propensity of the authors to make use, unconsciously of course, of more or less long syllables, where they are free to do so". But why would they not use, if they have the freedom, short syllables, or both short and loog in some what equal proportion? Further, does the sentence mean that if the authors had a natural propensity to use more or less short syllables stylistic variations would not have arisen and hence ihe statistical study would not have been possible ? One would of course understand that this would be the case if the authors had used long and short in somewbat equal proportion. But not if they had used short. M.A. Mehendale 11.1. 19. Nilakantha - samitam iti pathe tulyam ity arthah. V. S. Sukthankar JBBRAS (NS) 4. 158-161. Also cf Mbh. 18.5.43 and App. I. 5. 16. 19 The dictum does not apply to the four Vedas, as is generally assumed, but, in the con text, only to the Veda of Krsna (Dvaipayana ). samupabrnihayet also does not meanstirengthen' (Winteroitz, Hist, lod. Lit. I. p. 505), much less interpret (T. G. Maiokar, The Upabrmbana and the Rgveda laterprotation ). Madhu Vidya/675

Page #701
Reviews KALYANAMITRARAGANAM: Essays in honour of Nils Simonsson! edited by Eivind Kahrs, Norwegian University Press, 1986, Pp. xiii, 312. 355 This volume of research papers, instead of a cow', is presented to a guru by his pupils, colleagues and friends on the occasion of his 65th birthday. It is indeed a great tribute paid to the genial character of the teacher when his pupils choose to describe him Kalyanamitra good friend'. There are in all eighteen contributions in the Volume. They deal with such variety of subjects as would make it difficult for the present reviewer to take note of all of them. In the article on "Arhatschaft und Selbstmord ", the authors P.- A. Berglie and C. Suneson-come to the following conclusion: "Suicide has at all times occupied a certain place in the Indian religious practice. While self-killing has been an ideal way to end life for the Jainas and has exercised at certain times not an inconsiderable attraction also for the Hindus, this way of ending life remained in Buddhism reserved only for the monks" (p. 37). In the article on "The Crystal and the Hibiscus Flower", Erik af Edholm discusses the crystal-and-flower simile, illustrating error or false. knowledge, which is widely used especially by Yoga-Samkhya and Advaitavedanta authors. In the article "The Cots are crying" Gunilla GrenEklund outlines the attitude of Indian Thinkers to the metaphorical use of language. Eivind Kahrs makes a sustained attempt to interpret a discourse on bhava offered by Durga in his commentary on Yaska's Nirukta." (p. 115). Before examining Durga's discourse he discusses two well-known passages from the introductory portion of the Nirukta 1.1. and 1.2 (bhavapradhanam... vrajya paktir iti). It appears the author favours the interpretation offered in the two commentaries on the Nir. according to which the Nirukta passage (1.1) tad yatra-ubhe bhavapradhane bhavatah makes a complete sentence and hence a danda is to be put after it. In this sentence yatra is taken to mean vakye: (tad yatra (i. e. vakye) ubhe (i. e. namakhyate), (tatra te) bhavapradhane bhavatah ). But certain objections can immediately be raised against this interpretation. In the first instance, if this interpretation is to be accepted Yaska's wording, as suggested by Durga's commentary, would have been (tatra te) akhyatapradhane bhavatah (cf.. vakye hy akhyatam pradhanam in Durga's comm. ), and not bhavapradhane bhavatah; secondly, the expression tad yatra ubhe bhavatah implies kutracit ubhe (=namakhyate) na bhavatah which renders the interpretation yatra vakye impossible; and finally, there is little possibility of Yaska shifting his attention to a sentence' Madhu Vidya/676

Page #702
356 Annals BORI, LXIX (1988) when he is busy defining nama and akhyata. The interpretation of this passage very much depends on what immediately follows, viz. purvaparl. bhutam bhavam etc. which is clearly intended to explain the distinction between vrajati and vrajya, pacati and pakti. Hence tad yatra ubhe can only meanpow in forms where both ( i. e. a verb and a verbal ) noun ...' and not pow in a sentence where both ( i. e. verb and noun in general )... '. On the basis of what has been said above it is difficult to agree with the author when he says "Whether the interpretations offered by Durga and Skanda-Mahesvara are in keeping with the intentions of Yaska or not, we shall probably never be able to make out" (p. 121). One can, and say, they are not '1 In the article " A Kafir on the Kafir Life Cycle," Knut Kristiansen notes the birth, marriage, illness and death customs in pre-Islamic Kafiristan ( now Nuristan ) on the basis of a Urdu text written by a Kafir whose name was Azar. In the article on " The' water-Miracle' in Tibet " Per kavaerne calls attention to one of the later legends, known as 'water-miracle', associated with the Iranian god Mitbra. This legend is not known in Iran or Iodia, but is known in Tibet (p. 160 ). In a very interesting and well-reasoned article "R$yasraga: Ursprung und Hintergrund " Georg von Simson states that the motif of drought is pot secondary in the legend and that the legend could have, right from the beginning, the two "Frauengestalten " viz. the hetaera and the daughter of the king for two different functions -- one as seducer, and the other as marraige-partner (p. 207 ). The author of this article discusses at some length the Rgvedic Suryasukta (X, 85 ) and raises the question whether it is possible to discern pature-mythology as the background of the RSyasaga legend ip addition to its interpretation as a raio-myth (p. 214). The autbor seems to favour such a view but has to assume for tbis purpose far-fetched identifications such as Roma pada ( Haarfuss ) = tree (which has hair for feet, or wbicb has hair on foot', where hair = roots p. 215), or Saota = earth, one whose burning ( as the result of the preceding summer-heat is extinguished ( by the onset of the rainy season) (die, deren Brand (nach der vorsommerlichen Hitze durch die einsetzende Regenzeit) geloscht ist. p. 216 ). The fact that to make this adjective-turned-noun (santa ) applicable to earth requires putting into brackets so many words shows that the identification is not easily obtained. And even after this strenuous effort what one gets is not the complete identification of Santa with the earth but Only a temporary one i. e. the earth in rainy season. 1 Tocidentally vinasyati in the Nirukta passage 1. 2. (sad bhavavikara bhavanti...) is better rendered disappears' and not is destroyed' (p. 117). Madhu Vidya/677

Page #703
Reviews Writing on "Uses of the Curse in Rama Literature" William L. Smith points out that almost all the curses have some specific purpose : "they answer questions, resolve contradictions and establish moral causality" (pp. 261-262). He then goes on to classify the curses into a number of types. He notes that when a curse is given under a certain condition then the vernacular expression used for it is sapa vimocana. To convey the same meaning we have also such expressions as ussapa, uhsapa or usapa in some of the modern Indian languages. In the article on "Jang u asti: War and Peace in Iran " Bo Utas makes an important observation: "..in Zoroastrianism peace' (or at least * concord") is something morally good and desirable in itself, while 'war' (or discord') is something morally bad and worthy of blame." M. A. Mehendale 357 Madhu Vidya/678

Page #704
358 Annals BORI, LXIX (1988) TEXTKRITISCHE BEMERKUNGEN ZUR MAITRAYANI SAMAITA : Martin Mittwede, Franz Steiner Verlag Wiesbaned GmbH., Stuttgart, 1986, pp. 207. This is a very useful study of the text-critical suggestions already made by different scholars towards the improvement or correction of the text of the Maitrayani Samhita during the course of tbe last hundred years. Although the pumber of suggestions presented in this book is quite large, the author is aware of ihe fact that there could be some he has missed aod which await the attention of a researcher. The MS. is available in two editions - the one by Leopold von Schroder (Leipzig, 1881-86) and the other by S. D. Satavalekar ( Auddh; 1941-42), The author notes that the latter edition gives at many places the text correctly. He also notes that it has some better readings from the accent point of view. All the same he feels that the usefulness of Satavalekar's edition from the text-critical point of view is very limited, and be gives good reasons for holding this view. Wbile evaluating the text-critical suggestions made by scholars, the author's main task has been to make the position of the correct suggestions stronger by pointing to the parallel passages, otherwise to lay bare the wrong suggestions. Where such clear decisions were not possible, the author has chosen only to report the facts without arriving at any conclusion.. The "pure forms" "(Reinformen ) which come to light as a result of the author's thorough study pertain to - (1) improvements regarding interpunctuation, (2) corrections of obvious printing errors, (3) corrections on the basis of the material provided by the manuscripts, (4) emendation of the text against the manuscript evidence, (5) reconstruction -a) of the authentic text (not of a siogle word, but of a sborter or longer passage wrongly handed down), and (b) of the original' composition (Urfassuog) which must not correspond with any handed down text. While giving information under each entry the author says he has taken care to give enough to make a particular correction intelligible to the reader and also enable him, if he so desires, to ascertain it by referring to parallel passages. This practice is higbly commendable. M. A. Mehendalo Madhu Vidya/679

Page #705
Reylews 359 MALLANAGA VATSYAYANA - DAS KAMASUTRA : translated into German by Klaus Mylius, Verlag Philipp Reclam jun., Leipzig, 1987, pp. 196. The Kamasutra of Vatsyayada has been translated more than once since the first translation by R. W. Burton and F. F. Arbuthnot appeared in 1883 (cf. the very useful intormation given by the author at the end of the book under the heading "Weiterfuhrende Litteratur" pp. 194-196). However, the present translation need not be looked upon as superfluous. The aim of the translator has been to acquaint the reader...with what really is to be found in the Kamasutra without interpreting into the work something from the modern view point" (p. 16 ). It may be said that the author has been largely successful in achieving his objective. In the " Einleitung " (pp. 5-16 ) the author briefly deals with, among other topics, the question of the date and the place of Vatsyayana. In the author's view v. belonged presumably to the second half of the 3rd cent. A. D., and that he probably hailed from Western India. The explanatory notes (pp. 169-185 ) and the glossary ( 186-192 ) should go a long way to make the translation easily intelligible to the reader. Althougb the translation, at some places, cannot be called inaccurate, yet it may not stand the test of "bochstmogliche Genauigkeit" (p. 16) aimed at by the author. This may be illustrated with only a few examples. In the translation of 1. 1.3 (ebenfalls )' which is meant for the textword ca, and of 1. 2. 1 ( zum Streben ) which corresponds to the text-wor seveta (trivargam seveta (zum Streben ) nach den drei Gutern 'Deed not have been put into the brackets. In the translation of 1. 1. 4 ( auch )' is unnecessary. In 1.1. 3 again avabodhaka has been translated as .die...erkanat haben'. It should have been who have taught' (cf. R. Schmidt .die... zur Erkenntnis gebracht haben). In 1. 2. 38. werden Topfgerichte nicht etwa nicht zubereitet 'translates nahi ... sthalyo nadhisriyante. What the translator means could have been achieved by remaining close to the text " it is not that the cooking-vessels are not put on fire", M. A. Mehendalo Madhu Vidya/680

Page #706
Reviews SIVA DANS LE MAHABHARATA by Jacques Scheuer, Presses Universitaires de France, Paris, 1982, pp. 376, Price 180 Francs. Although much has been written on the Mahabharata and equally much on Siva, no work of any importance has appeared on the role of Siva in the epic. The present work is intended to fill in this gap, but with one limitation. The author of the work has no intention of collecting and presenting everything on Siva in the epic, for although this might serve the purpose of being a very useful index, it will not in itself enable the reader to establish any kind of coherence in the traits, attributes etc. of Siva. What the author there fore attempts to do is to define for the readers "the place which Siva occupies in the fabric of the narrative which should permit not simply to describe the personage, but to understand the place which he occupies in the Hindu mythical world contemporaneous with the epic" (p. 15). In the Introduction the author reviews the opinions of scholars on Siva Oldenberg, for example, maintained that although Siva is most Indian of the (Indian) gods (dieser indischeste der Gotter), he occupies, in the Mbh. a place that is rather marginal. That such an important deity as Siva appears in the epic only at scattered places and does not ever display his fondness for killing is, according to Oldenberg, perhaps the greatest gap in the epic. The author intends to demonstrate in this work that the fondness for killing (sa folie meurtriere) is very much there in the battle on the Kuruksetra and is limitlessly displayed in the massacre that took place at night in the camp of the Pandavas. The author is willing to concede that if Siva was to be found in the Mbh. only in some references to the cult of linga and in the litany of his thousand names, his place in the epic would be secondary. But the author believes that Siva's intervention in the Mbh. is less rare and less marginal than what appears at first sight. 359 In the principal part of the work (chapters I-IX) the author tries to depict Siva's intervention in the following parvans of the Mbh. : Adi. Sabha, Aranyaka, Udyoga, Droga and Sauptika, The author has based this study on the text of the Mbh. as is available today, i. e. the one which in its essentials was established (according to Winternitz and others) in the 4th century A.D. He has not tried to distinguish in this text the older and the later strata, a nucleus and the successive accretions around it, an authentic text and the suspected interpolations. 1 Das Mahabharata: Seine Entstehung... Gottingen, 1922, Madhu Vidya/681

Page #707
360 Annals BORI, LXIX (1988) What does the author's study based on this uncritical' text show to him? It shows that Siva's place in the epic is far from being negligible, either from the point of view of volume or its significance. The intervention of Siva in the epic theme, from the birth of the heroes upto the final massacre, is continuous, coherent and well-integrated (p. 345 ). M. A, Mehendale Madhu Vidya/682

Page #708
Reviews 361 PRAJAPATI'S RISE TO HIGHER RANK : by J. Gonda, E. J. Brill, Leiden, 1986, Pp. X, 208. Writing on Prajapati on an earlier occasion ( History of Religions 22 (1982) p. 129 ff.), Prof. Gonda had already expressed his doubts about the correctness of the view, widely held, that this deity is essentially a product of ritualistic and theological speculation and consequently comparatively young'. In that article Prof. Gonda had demonstrated that already in the early period of Indian religious history Prajapati must have been a popular deity meeting various wishes of a common man. In the present book be proposes to deal with the widening of the god's horizon and interests; the increasing number of bis relations with various other deities; his engagement in more complicated sacrificial rites; and his rise to higher rank and power." (p. 1). There is no doubt that Prof. Gonda has admirably achieved his objective. In Part One of the Book on " Prajapati's position and Gradual Rise in the Veda " he has collected and studied all the relevant texts in order to bring to light the various aspects of the deity and his association with other Vedic gods. of all the different activities ascribed to the deity by far the most important is the creative one. Whenever new rites were added to the ritual or the existing ones altered, the ritualists sought to give authority to the additions and alterations by ascribing them to Praja pati himself (p. 42). Io Part Two on " Prajapati in the Srauta Ritual" Prof. Gonda examines in details Prof. Oldenberg's view, which he does not share, that Prajapati is a young and vague, broadly defined, deity who has no access to the main and fundamental rites of the Soma ceremonies. In order to substantiato his own view Prof, Gonda makes a detailed study of "the mantras and other passages from the descriptions of the frauta rites in which Prajapati's name occurs or mention is made of his activity" (p. 119 ). Prof. Gonda cautions bis readers not to see in his book chronological survey of Prajapati's career. "What seems to emerge from the preceding disquition is that whereas the beginning of Praja pati's career can be satisfactorily reconstructed and the outcome of the various associations and identifications of this god with other deities and powers is an unquestionablo historical fact, the intervening period cannot be adequately and definitely described in a few words as a strictly historical account" (p. 195 ). It is somewhat surprising that while dealing with (p. 266 ) tbe honeywhip ( madhukasa) of the Asvina and the madhusukta (AV. 9. 1) Prof. Gonda makes no reference to H. Luders : Varuna II p. 370 ff. M. A. Mehendale Madhu Vidya/683

Page #709
Reviews 361 ENCYCLOPAEDIA OF PURANIC BELIEFS AND PRACTICES: by Sadashiv A. Dange, Three Volumes, Vol. I (A-C), Vol. II (DG. ), Vol. III (H-N.). Navrang, New Delhi, 1986, 1987. Pp. 370, 371-746, 747-1069, Price : Rs. 325/- each. Here we have, as the title immediately indicates, an extremely useful tool for further research. Prof. Dange has excerpted the relevant material from the nineteen Mahapuranas (the well-known eighteen plus the Sivapura a which is supposed to be an upapuraoa ) and published it for the benefit of research scholars as well as general readers. The interest in knowing the ancient fojian beliefs and practices is ever-growing and when all the volumes are published they will no doubt go a long way in satisfying this curiosity. The author and his colleagues, whose names appear in the Preface, havo undertaken a task which must have required a good deal of industry. . The author's emphasis wbile collectiog the material has been sociomythological'. He specifically tells us that personal names appear in the Iodex "only when they have some belief attached to them, or when there is some custom or practice associated with them. Mere dynastic or personal details are discarded" (Preface p. vii ). After having thus told the readers what to expect in the lodex, Prof. Dange tells us that the entire material bas been arranged subject-wise'. This gives an impression that the entries have been first classified under different subjects like social, political, economic, religious etc. and then under these heads the entries have been arranged in an alphabetical order. But this is not the case. All the entries in the Index appear in their alphabetical order. Wbat the author apparently meant by subject-wise' arrangement is that the entry-heads are chosen not on the basis of the priacipal characters but on the basis of the socio-mythological' content of the entries. Thus, for example, the second entry-head in Vol. I is not Bharadvaja, but Adoption. Even a cursory look at the entries would be enough to give the reader an idea of the wealth of information supplied in these Volumes. It is not possible here to go into the details of even some selected entries for their evaluation that could be the subject of an independent article. However, on reading the first four entries one gets the impression that much more care should have been devoted to check the accuracy of the statements made. This becomes in particular necessary because the Volumes are likely to be used by non-Sanskritists and such readers cannot be expected to look up Madhu Vidya/684

Page #710
362 Annals BORI, LXIX (1988) the original Sanskrit text and satisfy themselves about the veracity of the information they happen to come across in the Volumes. In the first entry on Acchoda: we are told: "Acchoda was mind-born daughter of the Manes (pitrs). Once she saw the king Vasu Uparicara going with a nymph named Adrika;... and followed them. As the manes had no body or form. she did not comprehend Vasu Uparicara : and thought him to be the pitr." The last sentence is pot intelligble since there can be no connection between pit!'s being body-less and Acchoda's not comprehending Vasu. The fact, however, is that since the manes had no bodily form, Acchoda, their mindborn daughter, did not know her fathers, and hence, when she saw Vasu she took him to be her father. The text is quite clear, and offers no difficulty for interpretation ) amurtatrat pitrn svan sa na jananti sucismita tam Vasuni pitaran mene ... Skanda P. 2. (9), 7. 16). Further we read: " The other manes cursed both of them ..." But who are the 'other' mapes ? The Puraoa text makes no scuh reservation and simply says: "The mapes cursed the two ( viz. Vasu apd Acchoda)" (tau tatah pitarah sepuh st. 17). In the last sentence of the entry we are told that as a result of the curse Accboda would be born of Adrika, a nymph, in her first form'. This also is not intelligible and becomes clear only when we know, from the text, that what is really meant is in her fish-form "(...matsyadehayam adrikapam jantsyase st. 25). The second entry is on Adoption. Here we are told that a Brahmaga (Bbaradvaja ) could be adopted by a Ksatriya (Bharata ). At the end of the entry we are told that Bharadvaja was given away by his Brahmaga father. But according to the story as narrated in the Matsya P. 49. 1777., to which the author refers, Bharadvaja was not given away by his father (Bphaspati) to Bharata - that was done by the Maruts. As far as one can make out Bphaspati, his father, did not know, or care to know, what happened to the child ( Bharadvaja ) tbat was deserted soon after its birth ( 49.25). Since Bharadvaja was not given by bis father to Bharata, there is further no question of his being given away on the condition that he will be the son of both' (i.e. the natural father and the one who adopted him). It is true the author does not specifically say that Bharadvaja was given away on this 1 It is not clear why the author does not give reference also to the Brabmanda P. 2. 10. 53-72 for his entry on Acchoda. I e., apparebly 'manes'. 1 In this account we will beglect the expression "followed them" for which there is nothing corresponding in the text. 4 first form' typing error for fisb-form'? Madhu Vidya/685

Page #711
Reviews 363 condition, but since he says that Bharadvaja became a dvyamusyayana, and mentions this condition while explaining the term, the reader is likely to believe that this condition was stipulated by Bshaspati while giving Bharadvaja to Bbarata. The text, as mentioned above, has nothing of the sort. Incidentally, dvyamusyayana does not mean, as stated by the author, "one belonging to two fathers", but "one belonging to two gotras". According to the commentator on the Trikandasesa ( 3. 1. 1) amusyayana means one who is sadyamsodbhava i. e. saikulajata according to the Sabdakalpadruma. Note the words vamsa and kula in these explanations ). In the pext entry on " Adultery" first we are told the punishments prescribed or the expiations recommended for this sin. At the end we read: "Having union with a woman elder than oneself is also said to be adultery" and for this statement we are referred to Brahmavaip. 1. 30. 44 ( yah sevate mahamudho gurvinim ca svakaminim). Several factors come here for comment: (1) gurvini does not mean a woman elder than oneself', but a pregnant woman'; (2) the stanza does not note this as a case of adultery; (3) it cannot be a case of adultery if svakamini' means 'one's wife'. In spite of these ipaccuracies, there is no doubt that the Index will be of immense help to students of lodian Culture. M, A. Mehendale Madhu Vidya/686

Page #712
Revlews 387 BRUDERSCHAFT UND WURFELSPIEL (UNTERSUCHUNGEN ZUR ENTWICKLUNGSGESCHICHTE DES VEDISCHEN OPFERS) : by Harry Falk; Hedwig Falk, Freiburg, 1986. Pp. 216. In this book Falk gives the results of his investigations on the relationship between the Vratyas and the game of dice. The book has five chapters : (1) Bruderschaften in Indien, (2) Das Wurfelspiel im Veda, (3) Das rituelle Wurfelspiel, (4) Das profane Spiel, (5) Schluss. Falk has studied extensively the relevant literature on the subject. He cites frequently from original texts and he does not cite a Sanskrit passage without translating it into German. This is of great advantage. According to Falk Vratyas were not converts in the religious sense but were a product of a social reform ( Objekte einer Gesellschaftsreform'p. 55). In his concluding chapter (3), summarising the results, Falk observes 1 " If we now consider the separate results in their totality, then the technical side at least of the game of dice seems explained. The three times fifty nuts corresponded to the members of the Young Mens' Associations who appeared together not only as students of the Veda but, in the period of intermission of study, as receivers of sacrificial animals they had to fulfil an important task in the religious life of the Aryan community. In the circle of these young men the one was procured with the help of 150 nuts who, as dog', must kill the sacrificial animal. "At the time of dividing the glaha it was important that there was no remainder left over. This conception is inseparably connected with the way an animal should be cut up : A perfect cutting up leaving no remainder permitted the rising again of the sacrificial) victim. Just as the horse in the Asvamedha could, as a whole, continue to live in the heaven, so also the nuts, after separation, were again collected together. "In the original form of the game there was no opponent. The companions stepped up before the nuts one after the other. The one who retained one piece as remainder became the first, the leader of the group', Kali, Rudra, Pasupati, the 'master of the sacrificial animal.'" (p. 188) In order to be able to pronounce judgement on the conclusions arrived at by Falk at various stages in his study it is necessary that one has first obtained a clear understanding of Falk's reasoning. Since the reviewer does not claim to have done this, what is done here is to offer some observations on details some of which may turn out to be corrections. driving rod' is 1. p. 24: It is not clear why Falk says that pratoda- Madhu Vidya/687

Page #713
388 Annals BORI, LXXI (1990) not attested outside PB and KsS. (der sonst nicht in der Sanskrit-Literatur bekannt ist ). 2. p. 26: According to Falk Yajnikadeva on the KSS 15.3.32 gives the information that the quivers received by Palagala are each filled with three arrows (mit je drei Pfeilen gefullte Lederkocher). This is not correct. What the commentator tells us is that the Palagala receives three quivers filled with arrows (tisra isupurnas carmatunyah). The number three' thus restricts the number of quivers and not that of arrows. The remark of Yajnikadeva has therefore nothing to do with tissdhanya- of the BaydhsS 18. 24. The meaning of that word ' a bow and three arrows' can be obtained from the MS 4. 5.9 (cited by Falk pp. 26-27) and the explanation provided by the commentators on the TS 1. 8. 19. 1 (tissbhir isubhir yuktam dhanuh). Just as there could be a unit consisting of a bow and three arrows (in a quiver ), there could be others where the quivers contained more or less than three arrows. When, for example, in the Mahabharata war different heroes shoot at Abhimanyu different numbers of arrows the reason could be that their individual quivers had that number of arrows (7.36. 15 ff.). 3. p. 26, f. n. 49: According to Falk, ba navan VS 16.9 refers to the quiver. This is possible, but in the present context it seems to refer to Rudra. In the first half of the stanza the god is requested to loosen the string from the two ends of his bow and to throw away the arrows held by him in his hand (yas ca te lid sta isavah ). The second half of the stanza is intended to convey that the god conceded the request. It is therefore better to translate visal yo banavani uta as the one with the arrows (in his hand, viz. the god Rudra ) is (now) without the arrows'. 4. p. 27: It is not clear why Falk translates punyajanma as 'some one born (anew) at a sacrifice' (jemand bei einem Ritual (neu) geboren wurde ) when it simply means 'one whose origin is holy. Also there is here no question of something being produced at a ritual. The bow and the arrows are supposed to have come out of Makha's left and righ thands. Makha also means sacrifice'. Because the bow and the arrows together (isudhanyasa. muhah ) are yajnajanma therefore they are considered panyajanma. 5. p. 51, f. n. 135: The Baudhos 18.24 has the subject in the plural ( te ye ... upetya (read upetah?)... nopeyuh), but Falk renders it in the singular (Wer ... betrieben hat, moge ... nicht tun ). 6. p. 52: The Nidanasutra 6. 1 runs as ye tu khalu janmana kanistha ye ho vayaseti / ye janmanety ahus tan kanistha ity acaksate. It does not become immediately clear what distinction Falk wants to make when he ren. Madhu Vidya/688

Page #714
Reviews 389 ders the first line as " (There are ), however, such that are younger by (the sequence of the) birth, and on the other hand (such) that are younger) due to age" ((es gibt ) aber ( solche ), die nach der Reihenfolge ) der Geburt die Jungeren sind, (und) andererseits (solche), die durch das Lebensalter (die Jungeren sind)". It would have been better to say such as are younger by birth (i. e., others than the eldest son in every following generation or according to the law of primogeniture). Or vayas may mean vigour, virility, strength. The two kinds of kanisthah, then, are those that are younger by birth and those (though older by birth ) are lower in virility. This interpretation is favoured by the fourth class of the Vratyas (Falk p. 51 ) who are described as jyesthah samani camed hrah (PB, 17. 4. 1). 7. p. 53, f. n. 142: Falk does not give the exact reference from the Mahabharata where he says Sikhandin is described as one lacking the sex organ (dessen fehlendes Geschlechtsteil ). What we know from the Mbh. (5. 189-193) is that Sikha qcin was born a woman, later turned into a man, and was married to the princess of the Dasarna country. 8. pp. 55-56 and f. n. 150 : The difficult passage from the Baudhss 18. 26 does not seem to have been correctly rendered. It really means i "Gandharvayana Valeya Agnivesya asked Aupoditi Gaupalayana Vaiya. ghrapadya : Whatever you have done (viz. taking hold of those who were creeping for Bahispavamana for purifying yourselves ), who has so acted (before )?'. (Instead of giving a reply, Aupoditi thinks :) with this question ( second iti) he has indeed touched the secret observance'. (Mistaking his silence for ignorance) Gandharvayana cursed Aupoditi and the sons of the Kurubrahmins :) We have cursed you as ignorant ones...'" 9. p. 57, f. n. 155: For PB 24. 18 te ha va aniryacya ... didiksuh Falk proposes an alterpative translation in the footnote. This is to be preferred to the one adopted in the text because yac-requires two accusative complements, * 10. pp 78-79: While dealing with irina as a salty depression in the ground (Senke mit Salzerde ) caused by the evaporation of salty river-water, Falk could have also noted the passage in the Mahabharata (13. 139.25) where it is mentioned that the land from where the ocean receded became irina. 11. p. 82: JB 3. 236 does not speak of the cows' who had fallen into the Indus (die in den Indus gefallenen Kuhe), but of the bulls' (te (not tah) sindhum pravisan). Hence also the bulls, not the cows, became salt (tal lavanam abhavat). Further, the JB statement tasmad u yo lavanena panan carati gava eva bhavanti does not mean 'one who always sells (some Madhu Vidya/689

Page #715
Annals BORI, LXXI (1990) thing) for salt, (his) cows become something' (dessen Kuhe werden etwas), but "Hence if one gives salt in exchange for something, (these what he gives, viz, salt) are bulls ". 390 12. p. 83: For Gedeihen' we cannot have bhuyamsa by the side of pusti and posa, but bhayastva. 13 p. 84 and f. n. 260: Falk looks upon tenasmiml loke dhrto hiranyam abhyavarohati as one sentence and translates: thereby he descends in this world secure on the gold (-plate) (damit steigt er in dieser Welt sicher auf die Gold (-Platte) ab). But actually we have here two sentences tina... dhrtah and hiranyam abhyavarohati thereby he becomes firm in this world. (Then) he descends on the gold (plate)'. This construction is made quite clear by the KS 14.8 and the Sat. Br. 5. 2. 1, where for the second act there is a separate section (20) which reads atha hiranyam abhyavarohati. 14. p. 87, f. n. 270: Medhatithi not on Manu 7. 3. but 8. 3. 15. p. 91 From Draupadi's stanza Mbh. 2. 62. 9 one cannot say that a woman stepping into the Sabha could be looked upon as a prostitute (Eine Frau in der Sabha konnte also als Hure angesehen werden. Dies meint auch Draupadi im Mbh...). What Draupadi wants to emphasize is that there is a long-standing custom according to which virtuous women are not forced to go to the Sabha to seek justice. But if she was (as is the case with Draupadi), that in itself will not mean that she was a prostitute. Karpa, no doubt, calls Draupadi a bandhaki (2. 61. 35), but that is for a different reason. That has nothing to do with her presence in the Sabha. The grounds for which Drau. padi did not wish to go to the Sabha are different (2. 60, 25, 29). 16. p. 99 According to the JB 3. 72 a gambler shares with an eunuch and a prostitute one third of the grief not of this world alone, as Falk states (ein Drittel der Sorge dieser Welt), but one-third of the total grief of all the three worlds together. [Falk's upavesayan to be corrected to nyavesayan ]. 17. p. 100: Falk says that while dealing with the ritual game of dice the texts avoid the use of the word kitava and, instead, use the derivatives of the root div-. In support he refers to the ApDS 2. 25. 13 where the word used for gamblers is not kitavah but divitarah. However there is nothing in the context of the ApDS to show that preparations are afoot, not for a profane, but a ritual game. 18. pa 102: Falk looks upon vaibhitaka as a patronym of the vibhitaka tree and considers Vibhitaka to be the name of the father of the tree, who, according to him, is Rudra. But vaibhi taka- may simply have svarthe -a Juffix, Madhu Vidya/690

Page #716
Revicws 391 19. pp. 104-105: Mbb. 5. 152. 24 reads naranam pahcapancasad esa pattir vidhi yate / sena mukham ca tisras ta gulma ity abhisamjnitah. Falk takes this to mean 250 men make a patti and three pattis a sena mukha or a gulma. He obviously follows van Buitenen's translation (Vol. 3, p. 469). This interpretation (pancapanicasat=250) is theoretically possible but it should not be adopted in preference to the conventional one (55) unless the context very clearly demands it. For, theoreticaly pancapancasat can mean even 45 ( 50 less by 5 ) or 10 ( 50 divided by 5)! ( This additional information was given to mo by Pandit Vamanshastri Bhagvat). But no one normally adopts these thoretically possible meanings. The stanza has been already correctly understood by BR. It means 55 men make a patti and a senamukha; three pattis are called gulma. In Mbh. 1. 2. 15 we find a different computation. Here patti is looked upon not as identical with senamukha, but forms a part of it (3 pattis = senamukha). 20. pp. 108-109: Falk translates yo aksesu tanuvast (AV 7. 114. 1 ) as who has the body of the gambler) under his control' (der den Leib (des Spielers ) in seiner Gewalt hat). Falk does not translate the word aks esu. The passage seems to mean who among the dice controls (their bodies' i: e., the ugra and the babhru, to whom namas is made, controls the pieces of the dice. This may be the senanih, or the raja of the dice referred to in RV 10. 34.12. Falk's aksesu to be read as aksesu ). 21. p. 112: In the stanza Mbh. 2. 51.3 Falk takes the word astara to mean a piece of cloth used in the game (Spieltuch ) which in his opinion was used to cover the ground of the adhidevana and on which the dice were thrown. But since the player's astara is equated with the chariot (ratham) of the warrior, it must refer to something spread on the place where the player himself sits while throwing the dice on the adhidevana. (Falk's vidhi to be corrected to viddhi ). 22. p. 127 : Falk takes the expression vi cinoti occurring in the Vedic passages to mean 'separate into fours the dice held by the player in his fist (gldha). (vi-cinoti bezeichnet das Abtrennen von Viereineiten (sic) vom glaha). This cannot be the case. Making units of four dice each of those held in the hand requires no skill. The skill of the svaghnin lies in picking up in his grip exactly that number of dice which, when divided by four, will leave no remainder. Hence vi-cinoti must refer to this act of separating the exact number of dice by the player with his glaha. 23. pp. 128-129 : It would have been possible to accept Falk's suggestion to read aksan iva svaghni ni minoti ta ni (AV 4, 16.5) in place of Madhu Vidya/691 coti ta mi (AV 4. 16. 5in e

Page #717
Annals BORI, LXXI (1990) the transmitted tani if this had not involved the change of accent and an unnecessary repetition of ni. It is therefore better to admit the difficulty regarding the gender (nimisah f. tani nt., as done by Luders (Phil. Ind. 164). or one may admit nimis (nt.) by the side of nimis (f.). 392 24. p. 175 Falk thinks that the AV 5. 18. 2 aksadrugdho rajany dh papa at maparajitah speaks of a king who has lost himself in the game of dice (der sich selbst verloren hat). Such a king, according to Falk, may eat the cow of a king. But such a king, if he has gambled himself away, will have neither the authority nor the money to take away or buy the cow of a Brahmin. The stanza refers to a king who, due to his own fault, has lost much in the game of dice (papah). Such a king may take away the cow of a Brahmin and enjoy her milk. 25. p. 185 Falk translates sabham eti kitavah prechamano jesyamitt tanva sasujanah (RV 10. 34. 6) "The gambler goes to the Sabha asking himself shall I win?" (Zur Sabha geht der Spieler, indem er sica fragt: , Werde ich gewinnen?). This is not correct. A challenge issued by the gambler to his opponents is here implied by the verb prach. We have to translate "The gambler goes to the Sabha challenging with the words I am going to win, feeling proud of himself (?)". Or, "The gambler goes to the Sabha challenging ( who wants to play with me?'). With the thought (irl) 'I am going to win' he feels proud of himself (?)" The book has unfortunately many misprints. It would be purposeless to list only a few. Madhu Vidya/692 M. A. Mchendale

Page #718
ELEMENTS OF POETRY IN THE MAHABHARATA : by Ram Karan Sharma ; Motilal Banarsidass, Delhi, 1983. Pp. 179, Price : Rs. 125/ In this book the author presents a classification of the vairous figures of speech collected from the critically edited text of the three parvans of the Mahabharata (Mbh): Adi, Vana (but 'Aranyaka' according to the cr, edn. ) and Bhisma. As the author states the book is "not a rhetorical discussion of the soul of the poetry of the Mahabharata. It rather aims at enumerat. ing the symbolic, alliterative, paronomastic, or repetitive linguisic features that beautify the body of the Mahabharata." (p. 1) Yet the author chooses to use the words elements of poetry in the title of the book, and not just poetic embellishments', apparently because he believes what Jayadeva (Candraloka 1.8) has to say about poetry : "One who regards poetry as word ard n:eaning without alamkara, why does he not regard fire as without beat?" (p. 8). : The author has presented the account of the "poetic expressions of the corpus" (p. 1) as follows: Chapters 1-3 classify the arthalamkaras, Chapter 9 details the poetic idioms, and chapter 10 classifies the sabdalarkaras. In chapter 11 we have a useful discussion of repetition as a technique of oral poetry. The book evinces the author's deep study of the alamkarasastra and his familiartiy with the classical Sanskrit literature. His detailed presentation of the figures of speech reveals the amazing variety of the objects used in the epic as upamanas. Students of Sanskrit poetry would eagerly await a similar treatment at the hands of the author for the remaining parvans of the Mbh. On p. 7 the author says that Vyasa taught Bharata to his five pupils. Strictly speaking, in this context, he should have said: 'to his son and four pupils (1.57. 74-75 and 1. 1. 63). In a passage cited on p. 12 from the Citramimamsa (p. 6) the author renders the word bhumika as costume' ( sailusi samprapta citrabhumika. bhedan). The word is better rendered as 'rolc' or character' (sthitibhedan Comm. Sudha, although Tativaloka has vesaparigrahah). MadhuVidya/693

Page #719
The author renders the figure of specch Svabhavokti as description of nature' (p. 12). But on the same page he also observes that for this alamkara, the cpic poct's imagination was confined to forest life and battle scenes. Hence, and also because of the definition of Svabhavokti cited by the author from the Kuvalayananda on p. 146, the name of the alamkara is better rendered as description of the natural state or condition (jatyadisthasya svabhavasya varnanani ). In the chapter on idioms the author could have mentioned atmanam urdhvaretasani kr (6. 115. 13 ) which is used to signify to practise celibacy ( adyaprabhisti me da sa brahmacaryam bhavis yati 1. 94. 88). On pp. 167-168 the author approvingly cites from an article : "Oral Poets of South India : The Todas" by Emeneau in which he says that according to the Mbh. I, I the text has undergone three successive recitations each of which was of a different length. In this connection it is better to note, in the first instance, that the Mbh. I. 1 reports the composition of the Bharata by Vyasa (1. 1. 52 ) and its two (not three ) recitations : one by Vaisampayana (1.1.18, 57-58 ) and the other by Suta Ugrasravas (1. 1. 23). Next, it is equally important to note that the first recitation, since it was carried out in the presence of Vyasa himself, could not have very materially diffcred in length from the original composition. The case was quite different in the case of the second recitation, In the last paragraph of the book the author observes that oral litera. ture like the Mbh., when compared with written literaturc, is devoid of "artificial refinement in syntax or embellishments" (p. 175). But as far as the embellishments are concerned this observation is contradicted by the wealth of the ala mkaras gathered by the author and by his statement on p. 13: "there is hardly a page in our corpus that does not have at least half a dozen striking examples of the figures of sound". The book under review is the second edition of the text first published in 1964 by the University of California. This proves its wide populaity which it rightly deseves. The author says that the second edition is just a reprint of the first edition.except that the indices have been appended in order to facilitate its referential readability." (Author's Preface to the second edition). Actually, we have only one index, and not indices. M. A. Mehendale Madhu Vidya/694

Page #720
Diverse Treatment M. A. MEHENDALE Moral Dilemmas in the Mahabharata. Edited by Bimal Krishna Matilal. Shimla: Indian Institute of Advanced Study and Motilal Banarsidass. Delhi. 1989, pp. xiv + 157, Rs. 95. THE HE book contains thirteen contributions on the subject of Moral Dilemmas in the Mahabharata. The authors are B. K. Matilal, T.S. Rukmani, S.P. Dubey, K. Kunjunni Raja, Y. Krishnan, A.N. Jani, Amiya Dev, 8. G. Kantawala, P. D. Santina, S. Paul Kashyap, M. M. Agrawal, E.R. Sreekrishna Sarma and S. M. Kulkarni. They have picked up different incidents from the Mahabharata like Arjuna's laying down the arms, or Draupadi's raising the question about the validity of the last play in the game of dice which in their opinion constitute. moral dilemmas. The treatment is diverse and unfortunately only a few of the contributions have a direct bearing on the subject. Some contributors, Matilal, for instance, also make clear the situation in which a character could be said to face a moral dilemma. The contributions were presented for discussion at a colloquium organised by the Indian Institute of Advanced Study at Shimla, apparently in the year 1988. The date is nowhere stated. One has to guess it. In the opinion of the Editor, the topic Review Madhu Vidya/695

Page #721
82 | Indian Literature set for discussion is "rather an outstanding one, not a bit less daring and intriguing" (p.xi), and as stated by him, the credit for choosing the topic goes to Margaret Chatterji of IIAS. Matilal, in his contribution, which was probably meant as a key-note address, rightly observes that the dilemmas presented in the Mahabharata are in some sense universal and therefore a discussion on them would be quite relevant even in present times. Matilal describes the situation in which it could be said that a given character in the epic is faced with a moral dilemma as follows: This happens when the character in question "is committed to two or more moral obligations, but circumstances are such that an obligation to do x cannot be fulfilled without violating an obligation to do y" (p.6). But it does not seem correct when he goes on to say that the agent's choice in such cases is either irrational, i.e. arbitrary, or is based on grounds other than moral (p.6). It would be equally incorrect to say that the agent's weakness of the will plays an important role in his decision-making procedure. One would rather say that the choice was governed by what appeared to the agent to be a superior moral law of the two. It is in this way, for example, that Bhisma made his choice when he was faced with the dilemma--should he stick to his vow of celibacy or should he do what Satyavati asked him to do, (1.97. 8-26); or, Dusyanta made his choice in the drama of Kalidasa when he was faced with the dilemma-should he accept Sakuntala's word and be guilty of accepting somebody else's wife, or should he rely on his own memory and be guilty of abandoning his own wife? (Sakuntala 5.29 : daratyagi bhavamy' aho parastri-sparsapusulah). Matilal mentions two typical cases of moral dilemma, both faced by Arjuna, one just before the war and other in the midst of the war. Arjuna's alternatives in the first case were : (i) to fight the war and kill the elders; (ii) not to fight the war and save the elders. He chose for himself the latter, but Krsna persuaded him to choose the former and yet do justice to both by fighting the war without any self-interest. Similarly in the second case Arjuna was faced with the alternatives: (i) to keep his word Nov.-Dec. 1990 Madhu Vidya/696

Page #722
Diverse Treatment / 83 and kill Yudhisthira; (ii) not to keep his word and save the life of his elder brother. Even here Krsna asked him to choose the latter and yet do justice to both. Matilal seems to have missed this point. On p. 17 Matilal observes: "But the acknowledgement of possible flexibility does not mean that the fixity and universality of ethical laws will be entirely negotiable." To prove his point Matilal refers to the story of Rama and says that he "idealized" dharma (p.18, also p.14). It is not clear to which incident in the Ramayana Matilai refers. Perhaps, he has in mind Rama's going to the forest as asked for by Kaikeyi, and this in spite of the entreaties of Bharata, and the advice tendered to him by the sages Jabali and Vasistha. If this surmise is correct, it will not prove Matilal's point that Rama "idealized" dharma. For, Rama did accept the kingdom after his return from the forest. This is not a matter of course as many seem to assume. A very ideal Rama could have said: "When Kaikeyi asked the boon her intention clearly was that I should never ascend the throne. If Bharata does not wish to be a king, that is a matter for him. I for one am out of the question." But Rama does not say this and rightly so. He accepts the kingdom in the interest of the subjects of Ayodhya. It therefore seems that the ethical laws were looked upon as negotiable as long as they were not violated for self-aggrandisement. Besides Matilal's contribution, there are very few other papers which have a direct bearing on the subject of the colloquium. Among these few may be mentioned S.P. Dubey's paper on "The Concept of Moral Dilemma: Its Applicability in the Context of the Mahabhurata." Dubey maintains that there are three characters in the Mahabharata who do not experience moral dilemmas: Duryodhana, Karya and Krsna. Whereas this statement seems to be true of Duryodhana, one cannot say the same about the other two. It could be said that Krsa faced the dilemma when, during the war, he started to attack Bhisma on two occasions (6.55. 86-92; 102.37-58). His dilemma could be: Should he take up the weapon and break his vow not to do No. 140 Madhu Vidya/697

Page #723
84 | Indian Literature so, or should he kill Bhisma and safeguard the interest of the Pandavas? Similarly Karna could be said to have faced a dilemma when Indra asked for his armour and kundalas: Should he keep his vow and part with the armour or should he not do so in the interest of Duryodhana (3.285.1)? It has, however, to be observed that Karna compromised his vow when he asked for Indra's sakii in return (3.294. 17.23). Dubey chooses to pick up the Amba-episode to point out a moral dilemma for Bhisma but he does not state it correctly. The moral dilemma before Bhisma was not whether he should marry her or send her back to Salva as Dubey seems to make out. The dilemma, if at all, before Bhisma arose when Parasurama, his teacher, asked him to marry Ambi or be ready for a fight. Dubey's narration of the episode is not in keeping with the critically edited text of the Mahabharata. Dubey, at one stage, observes: "Amba insists on marrying Bhisma himself as it was he who abducted her..." (p.41). But Amba never does this. Not only this, there are many other statements in Dubey's article for which there is no justification in the epic. On p. 39 he says that Karna was jealous of Bhisma, and that he decided to stay away from the battle while Bhisma led the army because of this jealousy. This is not the whole truth. Karna stayed away from the battle because he was underrated by Bhisma (5.165.27). According to Bhisma, Karna was not an atirathi, not even a rathi, but just an ardharathi (5.165.5-6). Karna stayed away also because Bhisma agreed to lead the army of Duryodhana on condition that either Karna fights first or Bhisma fights first (5.153.24). Karna was jealous of Arjuna no doubt, but that was not because of Draupadi, but because Arjuna was a match for him and, on occasions, had proved himself superior to him. It is incorrect to say that Karna was not allowed to participate in the swayamvara of Draupadi (p.39), because he was a sutapurra. The passage in which Draupadi is supposed to have said nilam varayumi sutam occurs only in one Nepali, one Kasmiri allied to Devanagari, and partly in the Devanigari version (including that of Nilkantha). It is totally absent in the Kasmiri, the oldest Nov.-Dec. 1990 Madhu Vidya/698

Page #724
Diverse Treatment / 85 Nepali and the Bengali version of the north and in the whole of the southern recension. It is therefore rightly looked upon as an interpolation in the critical edition (1.178.1827'). On the other hand we have a clear statement in the Mahabharata, which is found in all versions, viz., that Karna, like many other kings, did try to string the bow but failed to do so (1.179.4). It is again wrong to state that Duryodhana sent his messenger to Draupadi straight away to bring her to the assembly hall (p.42). Dubey blames Bhisma and others for not restraining Duryodhana at this stage. But it is well to remember that Duryodhana in the first instance sent his messenger to take Draupadi, who in his view had become a dasi, to the house of the Kauravas to do the menial work (2.59.1). How does Dubey expect Bhisma to intervene at this stage? It was only when Draupadi raised the question about the validity of the last game in which she was staked that Duryodhana wanted her to come to the assembly and get the answer directly from those present there (1.60.16). Finally, it may be noted that Bhisma was able to extend his life for fifty-eight days not on account of the boon (vara) received by him from his father (cf. Dubcy's f.n.66) but because he could do so with the help of yoga (6.114.112; 13.154.2-6). What Bhisma received from his father could at best be described as a blessing (asirvada). Bhisma turned it into a boon (vara) by making the usirvada come true. That Bhisma extended his life with the help of yoga is mentioned also by Sri Ramanujicarya -- he makes no reference to vara at all while commenting on the brahmasutra (4.2.19). Writing on "A Note on the Moral Dilemmas in the Mahabharata", Kunjunni Raja obscrves that Dhrtarasgra was always faced with moral dilemmas (p.51). But this cannot be correct in view of the definition of the moral dilemma given by the author himself. A moral dilemma has to be a conflict, not between right and wrong but between two 'rights' (p.49). But Dhstarastra's alternatives are between what is right and what is wrong viz., his unmindful affection for his son (p.51). Kunjunni Raja does not give any foot-notes and consequently no text references even No, 140 Madhu Vidya/699

Page #725
86 / Indian Literature for such important statements as those made by Markandeya (p.50), and Bhisma (p.51)'. This practice has led him to assume, as is widely done, that the famous stanza janami dharmam etc. (p.51) occurs in the Mahabhirata. But this is not true. The stanza occurs in the Pandavagita (57) of unknown authorship and there it is put in the mouth, not of Dhrtarasira but, of Duryodhana. However a stanza occurs in the Mahobharata (2.57.8) the last two lines of which are somewhat similar to those of stanza janami dharmam etc. The last two lines of the Pandavagita stanza run as kenapi devena hrdi sthitena vatha niyukto 'smi tathi carimi. The Mahabharata lines run as : teninusispah pravanad frumbho yatha niyukto 'smi tatha valami. Arjuna's dilemma---should he fight the war and be responsible for killing his kinsmen or should he lay down the arms and spare their lives-has been referred to by many contributors participating in the colloquium. M.M. Agrawal devotes an entire paper to this subject: "Arjuna's Moral Predicament". In Agrawal's opinion Arjuna's hesitation reflects partly a conflict of sentiments and partly of prudence, nevertheless it is a moral dilemma. In the light of the famous Gita stanza nasso mohah smrtir labdha ... (18.73). Agrawal states the net result of the Gira-teaching to be as follows: "He (i.e. Arjuna) is free from moha, and relinquishing all thoughts of personal gain he is now in a position to act from the motive of duty alone" (p.140). But it would be well to remember that this is true only in the limited context of the Gita. It is not true in the larger Mahabharata context because Arjuna's subsequent behaviour during and after the war does not bear this out. Also Arjuna does not really say what Agrawal makes him say. He simply says that he is now free from the mola regarding his duty (he was dharmasammudhacetah [2.7]a before the Gita was told to him), he has regained the memory of his duty which he had lost and is therefore willing to do what Kssna has asked him to do. These are about all the papers (a total of four out of thirteen) that have a direct bearing on the subject of the moral dilemma. The rest of the papers have only a somewhat loose connection Nov.-Dec. 1990 Madhu Vidya/700

Page #726
Diverse Treatment / 87 with it or no connection at all. Writing on "Moral Dilemmas in the Mahabharata", T.S. Rukmani starts with the assumption that there is such a thing as a standard moral behaviour and then says that all acts which deviate from this behaviour give risc to dilemmas. But this statement is vety different from what Matilal and many other participants have to say on what constitutes a moral dilemma. Hence if Duryodhana lists the misdeeds of the Pandavas and of Krsoa (pp.21-22) these cannot be illustrations of moral dilemmas. On account of her very different notion of what a moral dilemma is, Rukmani is the sole contributor who feels: "The dilemma is never brought to the forefront as a 'dilemma' and that the question of moral dilemmas in the Mahubharata is difficult to understand" (p.32).3 Rukmani objects to the game of dice because it has been condemned by Manu (9.221 ff). But how does she expect the characters of the Mahabharata to be guided by the rules of Manu who came much later? Rukmani says that Manu permitted gambling, but not betting (p.27) and in support refers to Manu (9.223). But Manu says nothing of the sort in this stanza. What Manu says is apranibhir yat kriyate tal loke dyutam ucyale pranibhih kriyate yas tu sa vijneyah samahvayah in which he makes distinction between the gaine to be played with inanimate objects (dyuta) and one to be played with animate objects (samahvaya). Gambling with betting is known in India since the Vedic period. On p. 32, Rukmani says that the story of Kausika occurs in the Virataparvan but does not say where exactly in the Virafaparvan. Actually it is to be found in the Aranyakaparvan (Adhyaya 197). A.N. Jani has contributed a paper on "The Socio-Moral Implications of Draupadi's marriage to Five Husbands". This out-of-the-way marriage is not a case of a moral dilemma. For whom could it be a dilemma-for Drupada, Yudhisthira, or No. 140 Madhu Vidya/701

Page #727
88 | Indian Literature Draupadi? It is really a question of finding justification for an act which went against the rules of marriage sanctioned by tradition (1.187.26-27). Jani sees the solution of the problem in an 'ethnological fact' (p.72). For this Jani simply relies on the statement of Yudhisthira in which he says that in suggesting a polyandrous marriage he was merely going the way his ancestors did (1.187.28). But we are entitled to ask: Who were these ancestors of Yudhithira who entered into a marriage of this type? How were Jatila, Varksi or even Saibya related to the Pandavas? It is going too far when Jani says: "Polyandry was in vogue in their (i.e. of the Pandavas) family" (p.73) because Kunti had three sons, one each from Dharma, Vayu and Indra. As we know this had happened apparently because Kunti was acting under constraints of the mantra received by her-each, mantra could be used only once, and as a corollary the same deity could not be invited a second time (1.104.3). Would Jani say that Kunti was the 'wife' of the three deities whom she invited? Yudhisthira, no doubt, is called 'Dharmaputra', but is Kunti ever referred to as 'Dharmapatni', or for that matter Vayupatni or Indrapatni? If not, how can we say that Kunti had entered into a polyandrous marriage? And granting that to be the case, why does Yudhisthira then not justify the marriage on the ground that his mother had done the same? Jani looks down upon polyandry as an uncivilized custom (p.73). This means that whether a community is civilized or not depends on its marriage customs. Will the author then be prepared to accept that Indians, as against the communities who practise monogamy, were uncivilized because they, until recently, were polygamous? The author at one stage suggests, as has been suggested by other scholars before him, that since Pandavas had lived on the Himalayas for some years and that since some Aryan tribes. who came to India via the Himalayas practised polyandry, Pandavas too chose that form of marriage. But there is no reason why the Pandavas should not have followed the customs of their own family and adopted alien ones. It is hardly necessary Nov.-Dec. 1990 Madhu Vidya/702

Page #728
Diverse Treatment / 89 to discuss here C.V. Vaidya's view, which Jani approvingly cites (p.73), that Pandavas belonged to a family other than that of the Kauravas. Matilal puts very mildly his reservations on this suggestion when he says, "The evidence to support it seems to be insufficient" (p.xi). While writing on "Marriage and Family in the Mahabharata: Some Aspects", S.G. Kantawala deals with the subject of niyoga. The custom has been looked upon differently at different times, but one would find it hard to agree with Kantawala when he says that Satyavati's suggestion to Bhisma to procreate sons on Vicitravirya's widows presented a dharma-dilemma for him (p.90). Kantawala seems to think that Bhisma's expression dharmed apetem (1.97.23) was made with reference to niyoga. That is not the case. About the practice of niyoga, Bhisma agrees with Satyavati that it was paro dharmah (1.97.13; earlier Satyavati: dharmam kartum iharhasi. 10). It is with regard to Satyavati's insistence that Bhisma should break his vow to remain a celibate, if not as dharma, then as apaddharma, that he considers the suggestion to be something that is 'far away from dharma'. It is difficult to find anything of a moral dilemma in E.R. Sreekrishna Sarma's paper on Arjunvisadayoga. In his paper Sarma does not discuss the dejection felt by Arjuna at the commencement of the war, as one would be inclined to believe from the title of the paper, but the one he felt at the sight of Duhsala in the Sindhu country when he was out on a victorious expedition for Yudhisthira's Asvamedha. Duhsala came up to Arjuna and begged him to protect her child whom she held in her arms. Sarma describes at length, much of which is the result of his imagination, the remorse Arjuna must have felt at her sight. Arjuna, of course, must have remembered that it was he who was responsible for killing Jayadratha, Duhsala's husband, in the war. He therefore blamed Duryodhana whose greed for the kingdom had forced war on the Pandavas which they had to fight according to the dharma of a ksatriya (14.77.39-40). For the rest of what Sarma says there is no evidence in the epic. No. 140 Madhu Vidya/703

Page #729
90 Indian Literature There is a statement in Sarma's paper which is difficult to accept. He makes Arjuna say: "Should I not feel the same sense of shame and insult to dharma now, as ought to have been felt by Duryodhana and his associates on that occasion (i.e. when Draupadi's modesty was outraged in the Sabha)?" (pp. 14647). Does Sarma consider the shameful acts of the Kauravas in the Sabha and Arjuna's killing Jayadratha to fulfil his vow which reduced Duhsala to her pitiable plight to be on a par with each other? As a point of detail one may note that it is not true, as Sarma makes it out, that Gandhiri wished to have a daughter (p.145). This occurs in an interpolated passage (1. App.1. No.63). The question raised by Draupadi regarding the validity of the last play in the game of dice in which she was staked has been mentioned by more than one contributor as an instance of a moral dilemma. S.M. Kulkarni has an entire paper (pp.150156) devoted to it. All those who speak about it agree that the question remained unsolved to the end. In fact, Matilal goes to the extent of saying that Draupadi's dilemma is not only unresolved but also unresolvable" (p.x). . It has to be observed that the problem posed by Draupadi's question has not been correctly followed. In the first instance, let it be remembered that it is not a moral dilemma, but is one which has legal implication (scc also Matilal, p. xi, 2.), especially the one related to the rules of the game of dice. According to a remark made by Sakuni we learn that a gambler could stake himself only when nothing else was left with him to stake. In the eyes of Sakuni, Yudhisthira had committed adharma since he staked himself when a certain item of his property, viz. his wife, had remained unstaked (2.58.29). In order to tempt Yudhisthira to stake Draupadi even after he had staked himself hc offered Yudhisthira freedom from bondage if after staking Draupadi he won the last game. He told Yudhisthira : "You stake Draupadi on your part, and I will stake you. If you win this game, you would not only not lose Draupadi, but you will win back your own freedom (2.58.31)". Sakuni could do this since Yudhisthira Nov.-Dec. 1990 Madhu Vidya/704

Page #730
Diverse Treatment / 91 had become the disa of the Kauravas. Unfortunately for Yudhisthira he lost the last game too. But now Draupadi raises a question regarding the legality of the last game. Her question boils down to this: Can a dasa, who is not supposed to own any dhana (2.63.1), play a game of dice? However, raising this question does not make Draupadi 'a social rebel' or 'a nonconformist' (p. 2); nor was she 'standing up for the rights and autonomy of the entire womanhood of that time' (p. 3), as Matilal observes. It is no use making Draupadi what she is not. Matilal also feels that Draupadi's question raises problem of a general nature like: Can wives be regarded as chattels? Can they be gambled away? (p. xi). This also is not true. As he himself observes, her question is a specific one and it has only legal implication. With regard to the conduct of the game of dice, Matilal observes: "If Sakuni cheated Yudhisthira in the game of dice and Yudhisthira did not claim that he had been cheated, even when this was openly known to him, would Sakuni be morally. reprehensible? I believe he could be but he would not be legally condemned" (p. xi). I do not wish to argue whether or not Sakuni was morally reprehensible or deserved to be legally condemned. But it has to be pointed out that the presumption from which these considerations follow, viz., that Sakuni cheated Yudhisthira and that this was known to the latter, is baseless. In the first instance Sakuni was an acknowledged expert in the game, whereas Yudhisthira was admittedly not. In a match between these two, where is the necessity for Sakuni to resort to fraudulent play? Next, both parties, before the start of the game, had expressly agreed not to resort to cheating (2.53.2-5) and there is no reason to believe that any one of them violated the agreement. Bhisma has openly declared that Yudhishira has at no stage complained of fraudulent play on the part of Sakuni (2.60.42). When Yudhisthira lost the first game he felt that he lost it because the number of vibhitaka fruit with which they were playing was small. He therefore suggests to use large number fruit (2.54.1). This did not help Yudhisthira, and he No. 140 Madhu Vidya/705

Page #731
92 / Indian Literature went on losing. If a gambler suspects fraudulent play he had the right to object to the winner's declaration jitam (cf. Harivamsa, 89-38). Yudhisthira docs nothing of the sort, on the contrary he declares that he was defeated (2.60.41). All contributors who have referred to the subject of Draupadi's problem have declared that the problem remained unresolved till the end: Matilal (p. 2), T.S. Rukmani (p. 31), S.P. Dubey (p. 43), S.M, Kulkarni (p. 155); Kunjunni Raja is of the opinion that Draupadi should not have raised the question because that would prove Yudhisthira to be a liar (p. 51). Apparently according to him too the question remained unanswered. As was mentioned above Matilal has gone a step further and has declared it to be 'unresolvable' (p. x). This again is not true. What the Mahabharata tells us in this regard is as follows: Duryodhana challenged the Paodavas that if any one of them said that Yudhisthira was not the master of Draupadi when he staked her he (Duryodhana) would accept the verdict and free Draupadi (2.63.20). Arjuna accepted the challenge and boldly asked the Kauravas that when Yudhisthira had lost himself whose master could he be (2.63.21), thereby implying that he could not be the master of Draupadi, Arjuna's reply had settled the question of Draupadi once for all. The Kauravas accept the verdict and Dhstar/stra steps forward to confer boons on Draupadi. The intervening stanzas about bad omens (2.63.22-26) are clearly an interpolation and should have really no place in the epic narrative. The ugly situation was saved by Arjuna's reply and not hy 'a miracle' as Matilal thinks (p. 3)". Y. Krishnan's contribution on "The Meaning of the Purusarthas", Amiya Dev's on "La Guerre de Kuruksetra n'aura pas lieu: Udyoga Reconsidered'', Peter Della Santina's on "Conceptions of Dharma in the Sramanical and Brahmnical Traditions: Buddhism and the Mahabhurata", and S. Paul Kashyap's on "Reflections, on the Concept of Action in the Gita" have no relevance to the subject of moral dilemma. They may be good papers in themselves (Matilal describes some of them as 'scholarly' (p. xiii), 'illuminating' (p. ix), raising 'some interesting Nov.-Dec. 1990 Madhu Vidya/706

Page #732
Diverse Treatment | 93 points' (p. xii), but they have very little to do with the topic of moral dilemmas. The editor does not seem to have given his attention to the details of the editorial work. He has no doubt written an excellent Introduction, summarizing the contents of the papers and offering his own comments on them. But there are certain other functions which he should have carried out. He, as Editor, should have, for example, asked those contributors who have not supplied references to give them in the footnotes to substantiate the statements made by them in the text. There is a paper in which the author does, indeed, give references but they are of no avail since the references are only to the parvans without further details. Where should a reader look for verification of a statement if the author simply says that it occurs in the Santiparvan (without giving the Adhyaya and the sloka number)? Further the Editor could have brought to the notice of the writers certain inconsistencies, e.g. the very first sentence of the first paper says that the Mahabharata describes itself as a fastra of dharma, artha, kama and moksa (p. 20) whereas the stanza cited in support of this statement in f.n.1 on p. 32 mentions only three subjects viz. artha, dharma and kama. The Editor could have also told a contributor that the critical edition of the Mahabharata has not been edited by Sukthankar and Karmarkar (p. 48) but that the General Editors of the critical edition were V.S. Sukthankar, S.K. Belvalkar and P.L. Vaidya in that order. The book is happily free from serious printing errors (except 'unusually' for 'usually' p. 49). But the same, unfortunately, cannot be said of the Sanskrit stanzas cited in it. References 1. Which can now be given as 3.26.10-15 and 2.62.15. 2. Also cf. mohakalilam of Arjuna mentioned by Krsna 2.52. 3. Other contributors have a different view, e.g. Matilal p.5,7; S.P. No. 140 Madhu Vidya/707

Page #733
94 Indian Literature Dubey p.35, A.N. Jani p.69. 4. That a husband could stake his wife is made clear by Puskara's suggestion to Nala to stake Damayanti (3.58.3). That Nala does not oblige him is a different matter. 5. For a detailed discussion of the subject, see M.A. Mehendale: Draupadi's Question, Journal of the Oriental Institute, Baroda, Vol. 35.179-194.1986. 6. To say, as Amiya Dev does, that "He (i.e. Krsna) comes to sell peace, he goes back buying war, and without regret" (p. 84) is, in this reviewer's opinion, totally misinterpreting Krsoa's peace mission. The author says Dhrtarastra's sight was temporarily restored (p.85). But this is reported only in the Southern recension and in some contaminated Devanagari manuscripts, cf. 5. 129: 495-496. It is not correct to say that Duryodhana won't yield even a 'needlepick' of earth (p. 88). What he would not yield is a needle-prick of earth (5.125.26). 7. Gita does not promise both heaven and kingdom to Ksatriyas as is said by the contributor on p. 113. It is a case of either-or (Gita 2.37), and not both. Nov.-Dec. 1990 8058 Madhu Vidya/708 12

Page #734
THE ASTRONOMICAL CODE OF THE RGVEDA : by Subhash Kak; Aditya Prakashan, New Delhi; 1994; Pp. xi, 144; Price : Rs. 175/ While reading an essay is a popular magazine the author of this book suddenly got the idea that the identical size of the sun and the moon when viewed from the earth had something to do with the structure of the Rgveda. In the year 1997 he discovered that the number of hymns in the Magdalas of the Rgveda encoded certain facts about the passage of the sun and the moon. His later investigations with the help of a computer convinced him that this correspondence was not a coincidence but was deliberately achieved. (p. x). The book deals with a variety of subjects related to Vedic astronomy. After taking a review of the context in which Vedic studies were carried out in the nineteenth century, the author deals with the chronology of the Vedic texts and astronomy of the fire aliars.. Next, he deals with the proper subject of the book viz. the architecture of the Rgveda' and the Rigvedic code. In the end he also analyzes the text of the Atharvaveda and the Bhagavadgita to point out that they also reveal a kaowledge of the code. The author has based his calculations on the number of hymns and their internal groupings in each Mandala. But as has been long recognized the number of hymns in each Mandala as found in the text today cannot be the same in the original collection of the ten Mandalas. The present collection violates at several places the principle of arranging the hymns. in the descending number of stanzas. To restore the arrangement to its proper order it is necessary to split the longer hymns into shorter ones. To give a single cxample, in the present arrangement of the Rgveda Samhita the number of hymns in Magdala 3 is 62. The last hymn consits of 18 stanzas, whereas the one immediately preceding it has 7 stanzas. This violates the principle of arrangement referred to above. To restore the proper order it is necessary to split the last hymn of eigbteen stanzas into six trcas which are addressed to different deities. When this is done the number of hymns in the third Mandala becomes 67, instead of 62 as at present. If the author's conclu. sions are to be accepted it would mean that the redactor who gave the final shape to the Samhita deliberately combined the six frcos nientioned above into a single hymd so that the iotal number of hyains in Mandala three becomes 62 as is required for the code discovered by the author. This nust have been a very bold and purposeful act of the redactor and it is difficult to imagine that such a significant step left no trace bebind it. The author himself is aware of the fact that the astronomical code discovered' by him as the basis for the i The author does not give any details about this magazine, Madhu Vidya/709

Page #735
organization of the Rgveda was "forgotten very early as there is no explicit mention of it even in the earliest indexes" (p. 109). He therefore feels safe to assume that if a text is organized according to the numbers of the code then that constitutis evidence supporting a date that is pre-Buddhistic" (p. 19). He then proceeds to demonstrate (chapter 7) that the organization of the texts of the Atharvaveda and the Bhagavadgita reflects the code numbers. But it should be quite cicar that the Bhagavadglta in its present form cao impossibly be a part of the origiaal' Mahabharata. There are some inaccurate and misleading statements in the book. On p. 14 the author says: "Rasabha which literally means the twin asses are defi. ned in the Nigha otu 1. 15 as Asvinau ..." But what the Nighantu here does is simply list the different animals that are yoked to the vehicles of different deities, c. g. hari of Indra, rohitah of Agni, etc. In this list are mentioned rasabhau, i.e. two asses which serve as draught-animals for Asvinau. One may therefore say that the asses are characteristic animals of Asvina. Similarly the Nighanlu at the same place does not define Aja (goat) as sun. It only -mentions goats (ajah pl., oot sg.) as the characteristic draught-animals of Pasan. The identification of Pusan with sun and of the Asvind with Gemini cannot be taken for granted for the Rgveda. M. A. Mebendale : Cf. H. Oldenberg : Prolegomena, p. 198; for a complete discussion sec pp. 191-202. 3 One should read Rasabba or Rasabbau. The dual number does not convey twin asses, but two asses. This, however, is not correct. The characteristic animals of Asvina are winged horses (Cr. Luders, Varuna. p. 86 and his footnote 3 on p. 89. Although Luders does not say it here, it is likely that rusabha, in this context, means screaming, and not ass (cf. Luders. Philologica indica, p. 754). Madhu Vidya/710

Page #736
REVIEWS JOHANN OTTO FERDINAND KIRSTE KLEINE SCHRIFTEN Edited by Walter Slaje, Franzsteiner Verlag, Stuttgart; 1993; Pp. XIII + 374; Price : DM 98/ The present volume No. 33 in The Glasenapp-Stiftung series brings together Kirste's ( 1851-1920) Kleine Schriften only in the field of Indology and a few of his reviews. Kirste's other writings pertain to Iranian and Slavic studies. Kirste studied at varions universities. His chosen subjects were : Classi. cal Philology, Sapskrit, and Comparative Grammar of Indoeuropean languages. He was graduated in 1876 in Comparative Linguistics at Vienna. After carrying on bis studies further in Sanskrit and in Old Iranian and Middle Persian at Paris, Old Slovenian and Serbian at Belgrade, Kirste returned to Vienna where he worked in close collaboration with G. Bubler. He was then introduced by Bubler to the studies of Indian manuscripts as a result of which he described ( 1889-1892) a number of manuscripts, related to law books, which were made available to him by Buhler from the libraries of Elphinstone College, Bombay, and Deccan College, Pune. Stand Kinrants for toited for In 1891 an independent teaching post was established for Oriental Philology in the University of Graz (Austria) and Kirste became its first occupant ( April 1, 1892). Among the three Austrian aspirants for the post, Kirste, Hultzsch and Winternitz, Kirste was considered to be best suited for the post as he could teach, besides Sanskrit, Semitic and Iranian languages. Kirste fully justified these expectations. As far as Indology is concerned, Kirste's research was centered mainly around Phonetics ( Pratisakhyas ), Grammar and Lexicography of the Jainas (Hemacandra ), Grhyasutras and the Mahabharata. The editor of the present volume has given a complete bibliography of Kirste's writings which is extremely useful. His writings are divided into three groups : I Independent Works, II Short papers (1. ladology, 2 Iranistik, 3 Slavic ), and III Reviews (Indology aad Iranistik). Among his short papers, only the papers listed in the section Indology have been published. From among his numerous reviews, 1 The above account is taken from the editor's preface to the Volume, Madhu Vidya/711

Page #737
332 Annals BORI, Lxxvm (1997) only three, viz. the ones on H. Oldenberg's The Grihya-Sutras. Pt. 2 (Oxford, 1892), E. Felber's Die indische Musik der vedischen uad der klassischen Zeit. Nach den Platten des Phonogramm-Archives des kais. Akademie. Mit Texten und Ubers. v. B. Geiger (Wien, 1912), and J. Hertel, Das Pancatantra ... ( Leipzig, Berlin 1914 ) appear in the Volume. Perhaps it would bave been good also to include Kirste's reviews of such books as M. Winternitz, Das altindische Hochzeitsrituell (Wien, 1892), H. Luders, Die Vyasa-Siksa... (Gottingen, 1894), W. Caland, Die altindischen Todten- und Bestattungsgebrauche ... (Argsterdam, 1896 ) and some others. At the end the editor has listed obituary notices on J. Kriste and appreciations. The volume contains three Indexes, all selective : (1) Names and Subjects, (2) Words, and (3) Text-passages, M. A. Mehendale Kirste laments the fact that the plates No. 445 and 465 were not ideatified for him although: he had sought help from India. He therefore expressed the hope that an Indian who knew the texts recorded on those plates would some day come to Vienna and identify them. It is not know whether Kirste's hope has been fulfilled Madhu Vidya/712

Page #738
Reviews 333 HERMANN OLDENBERG : KLEINE SCHRIFTEN, Teil 3, edited by Hanns-Peter Schmidt, Franz Steiner Verlag, Stuttgart ; 1993; Pp. VIII + 1571-2103 + (Register) 17. TI The first two parts of H. Oldenberg's Kleine Schriften appeared in 1967 ( 2nd edition 1987). They were edited by K. L. Janert. The two parts bring together only papers of H. Oldenberg to the exclusion of his monogra. phs and reviews. In addition, however, Janert gives at the begioning of Part 1, a complete Bibliography of Oldenberg's writings indicating simultaneously which of those writings find place in the two parts. Oldenberg's writings fall into the following five sections : 1. Kleine Schriften; 2. Monographien ; 3 Einzelbeitrage zu Sammelwerken; 4. Textausgaben und Ubersetzungen : 5. Rezensionen. The present third part in this series of Oldenberg's Kleine Schriften is edited by H. P. Schmidt. It forms a complement to the first two parts in the sense that it brings together under two heads three of Oldenberg's Monographs (Vedaforschung, Indien und Religionswissenschaft, and Zur Geschichte der altindischen Prosa ), one Excursus from his book on Buddha, and some selected reviews. Besides, in the first section Schmidt also publishes Oldenberg's replies to Mommsen's enquiries on the oldest criminal law. Schmidt states in his Vorwort his reasons for bringing together the above writings as follows: (1) The monograph on the history of old Indian prose is closely related to two of his papers on narratives containing prose and poetry. (2) The monographs on Vedaforschung and Indien und die Religions. wissenschaft describe and justify Oldenberg's methodological standpoint. (3) The excursus "Uber das geographische Verhaltnis der vedischen und buddhistischen Kultur" does not appear in the subsequent editions of Oldenberg's book on Buddha, and it deserves to be saved from oblivion. (4) Oldenberg's replies to Mommsen's queries is not mentioned in Janert's Bibliography appearing in Part 1. (5) Several of Oldenberg's reviews, since they make positive contributions to the subject, are as good as independent papers, Schmidt makes good one more deficiency in Janert's Bibliography by publisbing Oldenberg's review of Pischel-Geldner's "Vedische Studien, 1. Heft". In his Vorwort Schmidt mentions two entries conveyed to him by Janert which should have figured in the Bibliography. One need have no hesitation in agreeing with Schmidt for publishing the third part of Oldenberg's writings for the reasons stated by him. Scholars Madhu Vidya/713

Page #739
Annals BORI, LXXVIII (1997) all over the world would be thankful to him, as indeed they were earlier to Janert, for making Oldenberg's writings easily accessible. 334 Madhu Vidya/714 M. A. Melendale

Page #740
Reviews SURESVARA'S VARTIKA ON SISU AND MORTAMORTA BRAH. MANA ed. by K. P. Jog & Shoun Hino (Advaita Tradition Series, Vol. 7); pub, by Motilal Banarsidass Publishers Private Limited, Delhi, 1996; Pp. XX + 119; Price: Rs. 150/-. In this volume K. P. Jog and Shoun Hino have edited, translated and annotated Suresvara's Brhadaranyakopanisadbhasyavartika' on Sisu (BrUp. 2.2) and Murtamarta Brahmana (Br. Up. 2. 3). The two authors have already earned the gratitude of scholars in general and of students of Indian philosophy in particular by similarly editing, translating and annotating several other Vartikas on the Brhadaranyakopanisadbhasya (Advaita Tradition Series Volumes 1-6). 333 The importance of the Varttikas on the Brhadaranyakopanisadbhasya for understanding the thought of Sankara, the Guru, and of Suresvara, the pupil, is recognized at all hands. As observed by Hajime Nakamura in his Foreword: Suresvara has underlined every small detail in the varied arguments in the Bhasya on the Upanisad and clarified the same with characteristic skill... The Vartika of Suresvara on the Brhadaranyakopanisad is truly his magnum opus and needed to be translated in full". It is therefore gratifying to note that Jog and Hino have undertaken to execute the task of translating with annotations the Varttikas of Suresvara and supply the long-feit need. The authors observe: "There is inherent difficulty in rendering into very simple English structure the slightly (and comparatively) truncated or complex Sanskrit structure; we have yet tried at a number of places to simplify the same by avoiding as much literal English rendering as in the earlier parts of our series." (Preface). Even after admitting this difficulty of translation, one feels, occasionally, while going through the translation that the authors should have devoted a little more care and aimed at more accuracy in giving their renderings. A few examples are given below. 1. The first line of stanza 2 on the Sisu Br. reads as-aeutqfaqftiyaa tadvyAkhyA cAdhunocyate agearea, angaisa. This is translated as: "(Also) it was stated This is the secret doctrine of the Upanisad"." It is not clear why the authors repeat the text word Upanisad'. It confuses the reader, since the word has been already rendered as secret doctrine'. The line is better rendered as "(What) was stated as its secret teaching (viz. satyasya satyam), its exposition is now given here". The authors rightly point out that the words tasyopanisat in the Varttika refer to BrUp. 2. 1. 20 where it is stated i This is how the word is spelt in the text. It would have been better to spell it as varttika. Madhu Vidya/715

Page #741
Annals BORI, LXXVIII (1997) tasyopaniSat satyasya satyamiti which is there immediately followed by the words prANA vai satyaM teSAmeSa satyam. 2. Stanza 23 of the same Brahmana reads as: 334 upAsate'kSaNi prANaM rudrAdyAH sapta devatAH / akSINA iti tA jJAtvA nAnnazcayamupAznute || This has been rendered as: "The seven gods Rudra and others wait on Prapa (residing) in the eye (etc.). Having known them as non-decreasing, a worshipper does not experience decrease in (lit. destruction of) food." Comments: (i) It is not known why the authors add (etc.) after the eye. It is warranted neither by the text of the Varttika nor by the BrUp. 2. 2. 2 on which the stanza is based. The Upanisad passage does not mention any other sense organ like ear, nose etc. besides the eye (aksan). Moreover the Varttika 25 also emphatically states that the group of gods (Rudra and others) is only in the eye (akSagveva yata: pUrva vyAkhyAto devatAgaNa:) (Unfortunately the authors' rendering of st. 25 is also not very happy.) (ii) Upasate has been rendered as wait on which means attend on, But this meaning is not intended here. Upasate of the Varttika stands for upatisthante of the Br. Up. 2. 2. 2 where it is further paraphrased as anvayatta 'connected with'. Obviously upasare in the Varttika is used in its literal meaning to sit or be near. The most surprising thing is that the authors in a foot-note state offer worship to to be the literal meaning of upasate ! (iii) The last quarter of the stanza nannaksayam upainute is rendered as does not experience decrease in (lit. destruction of) food. In the first instance one does not know what the authors gain by adding (lit. destructtion of) into the brackets, especially when they in their Preface say that they have tried to avoid literal English rendering. Secondly, having rendered aksinah, with refence to the deities as non-decreasing' they say in a footnote Or, imperishable. This is not necessary and, perhaps, also wrong in the context. For, although imperishableness may be true of gods, it is not true of food. What is true of gods has to be true of food as well in the Varttika since the non-decrease of food results from the knowledge of the non-decreasing nature of gods. 3. The next stanza 24 reads as: mUrtina pratyAhitaM prANamupayaH prANasaMjJakAH / rudrAcA madhyamaM yasmAt satataM paryupAsate / Madhu Vidya/716

Page #742
Reviews 335 This has been rendered as : "(siace) those seers called Prapa', the seven gods Rudra and others ever particularly wait upon that Praga which is supported in the head ". Observations : (i) The authors have unnecessarily combined stanzas 23 and 24 into one sentence. The two stanzas contain two statements, complete in themselves. (ii) The authors apparently render the text word yasmat as since'. In that case it is not clear why they have put it into brackets. On the other band, the words sapla devatah of stanza 23 are not repeated in stanza 24. Therefore the words the seven gods' should bave been put into brackets. (iii) The authors surprisingly omit translating the word madhyama which occurs in the stanza and instead choose to inform the readers in footnote 6 that the word Prana in this stanza refers to madhyama prana. In this footnote they should have better referred to Br. Up. 2. 2. 1 : 7 ara fertutsa : TIT: (iv) What applies to the word upasate of stanze 23 ( see observation ii above ) equally applies to paryupasate of this stanza. (v) Murdhni pratyahitam does not mean which is supported in the head', but which is placed in the head'. This has reference to Br Up. 2.2.3 where we find the word nihitam (gafast: ... afar Tait fafca favegfafai 970 a 7 fara aromaatik ). This makes clear what is meant by murdhni pratyahitam. (vi) The authors have completely misunderstood the meaning of st. 24. The stanza says that the seers called Pranass sit around the Praga placed in the head' since, as told in the preceding stanza 23, the seven gods, Rudra and others, sit around the Prapa ( madhyama ) in the eye. The stanza is therefore to be rendered as : "The seers called the Pragas ( sit around ) the Prana placed in the head, since (the seven gods ), Rudra and others, constantly sit around the middle ( Prana). . Rather Pranas. (pl.). 9 Cf. By Up. 2. 2. 3: ST191 444 M ACR I This is explained by Sankara as ? prANAH parispandAtmakAH, ta eva ca RSayaH / * Explained by sai kara as : 91on: 1191 21977 AF: A ag weat: * Explained by saikara as : ASHTA: 9107: Erica q: spuit FITATI Madhu Vidya/717

Page #743
336 Annals BORI, LXXVIII (1997) 4. Stanze 15 in the Murtamurta Brahmapa (p. 22) reads as : yaniSedhamakhenedaM neti netIti bhaNyate / avidyAmadhiM kRtvA dve rUpe brahmaNastvime / / This has been rendered as : " These are the two forms of the Brahman which are denied in these words, viz. neti, neti, which purport to negate (all duality), which keep in view ignorance in its full extent." The rendering of avidyam avadhim kstva given above is unintelligible. It is difficult to understand why the authors have allowed it to be that way when in their annotation they state correctly what is meant by avidyam avadhin kriva. They say: "It purports to say that one can mention the two forms only so long as ignorance persists and, on removal of ignorance, one cannot talk of any form of the Brahman". In the light of this explanation a better way to render the stanza seems to be "what is said by way of negation with the words 'not this, not this' (neti neti) refers to these two forms of Brahman (viz. murta and amurta) which exist only as long as the ignorance (of the true nature of Brahman) persists." It is hoped that instances of the above type are few in this as well as in the other volumes published earlier in this series. M. A. Mebendale Madhu Vidya/718

Page #744
Reviews HANNS OERTEL: Kleine Shriften, Teil I and Teil II, edited by Heinrich Hettrich and Thomas Oberlies; Franz Steiner Verlag, Stuttgart; 1994; Pp. XV+ 1669. 293 These two solid volumes containing selected papers, reviews, and monographs of H. Oertei have appeared in the famous Glasenapp-Stiftung series as Band 32. H. Oertel (1868-1952) was born at Geithain (Sachsen). He studied Sanskrit with W. D. Whitney at Yale University in the United States. He taught Linguistics and Comparative Philology in different capacities at the same University from 1891 to 1917. In 1914 he went to Germany and then did not return to the States, probabby due to the difficulties of the first World War. He settled down at Basel in 1920. He taught there Indian Philosophy, Religion, and Literature at the University. In 1922 he succeeded K. F. Geldner at Marburg. Finally, in 1925, he took charge, as successor to W. Geiger, of the teaching post of Indian and Iranian Philology at Munchen. He retired in 1935. While Oertel was in Munchen three dissertations were completed under his guidance: F. J. Meier's Der Archaismus in der Sprache des Bhagawata-Purana, V. Trapp's Die ersten funf Ahnikas des Mahabhasyam (transleted into German and explained), and B. K. Ghosh's collection of the Fragments of Lost Brahmanas (all three are published). In the second World War Oertel suffered an irreparable loss when during the b mbarding of Munchen (1944) his entire library and his valuable card collection (on which he started work when he was 19) were completely destroyed. Undeterred, he took up teaching again after the end of the war and gave lectures until he was eighty. Besides his teacher Whitney, scholars who decisively impressed Oertel were J. Wackernagel and B. Delbruck. Hence it is understandable that Oertel's main interest lay in the field of Vedic prose - especially the prose of the Jaimini ya Brahmana and Sanskrit Grammar- especially syntax. Oertel has also published his observations on words like sunrta, causal of II (tapayate), vadaba, vagura and others. In his younger years, comparative and general linguistics engaged the 1 The information given here about H. Oertel is taken from the Editor's Vorwort to Leil 1. Madhu Vidya/719

Page #745
24 Annals BORI, LXXIX (1998) attention of Oertel. His Lectures on the Study of Language gave in those days occasion for lively discussion. Besides doing teaching work, Oertel acted as co-editor of JAOS from vol. 29 (1939) to vol. 34 (1915), and as co-editor of KZ (Zeitschirft fur Verglechende Sprachforschung) from vol. 54 (1927) until his death. W. Kierfel in his obituary note on Oertel justifiably describes him as "one of the last Indologists of the old school." He also observes that "Oertel's name will be for ever associated with Vedic research" (ZDMG 102. 12, 16). The Editors of the present volumes give at the beginning of Part I a complete Bibliography of the writings of Oertel and at the same time indicate, by giving page numbers to the Volumes, the writings which have been included in the Kleine Schriften. The Bibliography is divided into the following six sections: I. Articles, II. Reviews, III. Monographs, IV. (Short) Notices, V. Miscellaneous. The last section VI, which actually does not form part of Oertel's writings, gives five entries on appreciative writings and obituary noties on H. Oertel. At the end one finds a very useful Register (pp. 16011664). A look at the articles and reviews selected by the Editors shows that they are such as are directly related to Indology. Oertel's article "A practical proposal for preliminary work on a new Sanskrit dictionary" which appeared in Woolner Commemoration Volume (1:40) pp. 177-182 has not been reproduced. The article is quite interesting and its reproduction, even after the start of the publication of a new Sanskrit dictionary at the Deccan College, Pune, would have been worthwhile. It is also not clear why the editors omitted Oertel's reviews like those on S. Levi's La Doctrine du Sacrifice dans les Brahmanas (AJP 20 (1899). 444-447), P. E. Dumont's L Avamedha (OLZ 1928. Col, 995-996), and such others. From among Oertel's 14 monographs, no less than ten have been. included in the two Volumes. None of his short notices or his appreciative writings or obituaries (listed section V) finds place in the Kleine Schriften. Perhaps, inclusion of a few from the latter would have been useful. M. A. Mehendale Madhu Vidya/720

Page #746
295 PAUL THIEME KLEINE SCHRIFTEN (Glasenapp-Stiftung Band 5) Ed. by Georg Buddru's; pub, by Franz Steiner Verlag Wiesbaden; 1st ed. 1971, 2nd ed. 1984 (unaltered, but with an added Supple. ment 1984 to Bibliography: p. xvi); Pp. xvi +813; PAUL THIEME KLEINE SCHRIFTEN II (Glasenapp Stiftung Band 5 II) Ed, by Renate Sohnen-Thieme; pub. by Franz Steiner Verlag Stuttgart; 1995; Fp. ix + 815-2178. Reviaws Paul Thieme's contributions, as pointed out by G. Buddruss in his Vorwort to the first of the above two Volumes, pertain to two main spheres: Veda and the Sanskrit grammatical tradition. He has emphasized three characteristics of Thieme's Vedic writings: richness in brilliant ideas, uniforms clarity, and strictness in methodical execution. As to Thime's writings in the sphere of grammar, Buddruss points out that Thieme's interpretation of Indian grammarians and his valuation of their procedure culminate in attem. pts o arrive at a comparative and historical understanding of the development of grammatical scholasticism in India. Thieme's writings included in these Volumes are such as have not appear. d in book form or as longer dissertations. A few of his total writings have been omitted and the selection made by the editor has met with the approval of Thieme. The editor's wish, expressed in the first Volume, that Thieme should be able to write more articles after the publication of that Volume, has been amply fulfilled as can be seen from the publication of the second Volume of his Kleine Schriften within twentyfive years of the appearance of the first one. The first Volume gives a Bibliography of Thieme's writings up to about 1970, which at the same time serves as the List of Contents of that Volume. It has two following divisions: 1. Monographs (which are not included in the volume). 2. Articles: (A) Veda exegesis and Word studies (only one of the listed articles is omitted, but it appears in the second volume noticed below), (B) Miscellaneous contributions to Cultural and Religious History (a few of those listed are omitted), (C) Saskrit Grammar (a couple of those listed are omitted), (D) Various Articles (none of the five listed are included). (E). Appreciations (four, not included), (F) Reviews (of the 55 reviews listed, a dozen are included; all these have appeared after 1951 ). Two of these monographs: (1) Der Fremdling Im Bgveda (with Exkurs: Ari "Fremder"), and (2) Mitra and Aryaman are now published together in Paul Thieme: Opera Ma ora Band I edited by W. Knobl and N. Kobayaski. Hozokan Publi. shing Co., Kyoto, Japan, 1995. The editors hope to publish if not all, at least some of Thieme's remaining monographs in future Volumes of Opera-Maiora, Madhu Vidya/721

Page #747
296 Annals BORI, LXXIX ( 1998) At the end we have two Indices, one, of the words and the other, of the passages discussed. The second of Kleine Schriften Volume contains Thieme's articles and reviews published since 1970 and up to 1990 They appeared after those contained in the first collection noticed above. In addition, there appear in the second volume : (1) three articles which from the point of view of the time of their publication belong to the first collection. (They are : (i) " Sanskrit sindhu- Sindhu- and Old Iranian hindu- Hindu-", 1970; (ii) Merkwurdige indische Worte", 1942; and (ii) " The Comparative Method for Reconstruction in Linguistics ", 1954, and (2) an unpublished lecture delivered by Thieme in the Deutschen Orientalistentag in Erlangen in 1977 on "Stand und Aufgaben der Rigveda- Philologie". 'H.wever, two unpublished lectures of Thieme which he delivered in Japan in Novembe 1988 at the time of his being honoured with the Kyoto-Preis" could not be included in the present Volume. Like wise, a few of Thieme's articles and reviews which he wrote after 1990 (some in presss and some already published' ) do not find place in this Volume. The editor informs us that they have been reserved for a supplementary Volume which will also include Thieme's article on Ancient Indian Theatre, contributed to the Fernostliches Theatre 1966, but which is now out of print. The Table of Contents of this Volume serves also the purpose of being a Bibliograpliy of Thieme's writings of the period 1970-1990. It has the Following seven divisions : (A) Veda exegesis and Word-studies (only one of the listed articles not included in the Volume); (B) General and Indo-European Linguistics ( all the listed articles included); (C) Miscella. neous contributions to Cultural and Religious History (all the listed articles included); (D) Sanskrit Grammar ( both the listed articles included ); (E) Various Articles (two of the three listed articles included ); (F) Appreciations (the only listed item is not included); and (G) Reviews (both the listed reviews included ). At the end we have Addenda and Corrigenda, (wo Lists (i) of words dealt with and (ii) of passages dealt with, and finally a combined Index of # For this reason they are marked in the List of Contents of this Volume as Nachtrag zu Bd. I. 3 These are : his article (title not given) which is to appear in Georg Buddruss Felicita tion Volume, and "Reflections on the Vocabulary of Zarathushtra's Gathas." Review of M. Mayrbofer's Etymologisches Worterbuch des Altindoarischen in BSOAS LYII, 1994, and his essay: "Zur Fruhgeschichte des Scbacb," 1994, Madhu Vidya/722

Page #748
Reviews 297 Names and Subjects. As regards the Addenda, the Editor has a special remark to make : Thieme had no doubt taken cogoizance of wide ranging corrections and suggestions made during the period 1970-1990 and bad partly noted them down. However, during the last two years bis eye-sight was so much impaired that he was unable to execute bis original plan of systematically going through his articles. So also the suggestions and improvements conveyed to the Editor in recent times have not been taken into account for this Volume since they could not be discussed in sufficient details with Thieme. M. A. Mebendale MadhuVidya/723

Page #749
(C) OBITUARY NOTICES SIR RALPH LILLEY TURNER b. 5-10-18881 [ d. 22-4-1983 Sir Ralph Lilley Turner, Honorary Member of the Bhandarkar Oriental Research Institute (1948) passed away on April 22, 1983. He was 94. Sir Ralph Turner was born on October 5, 1888. Educated in the Perse Grammar School and Christ's College, Cambridge, Sir Ralph joined the Indian Educational Service and in 1913 was appointed Lecturer of Sanskrit at Queen's College in Benares. In the very following year (1914) he was invited by the University of Bombay to deliver the Wilson Philological Lectu res. Soon afterwards the First World War broke out and Sir Ralph served for four years (1915-1919) in the Queen Alexandra's Third Gurkha Rifles. For Sir Ralph this did not mean a complete break in his academic career for it was during this period that he learned the language of his comrades-in-arms and collected materials for his first major achievement the famous Nepali Dictionary. After the war Sir Ralph found himself again at Benares, this time as Professor of Linguistics at the Benares Hindu University. Two years later the scene of his academic activities shifted from India to the United Kingdom. In 1922 he was called to be the first full-time Professor of Sanskrit at the School of Oriental (and, since 1938, African) Studies, University of London. Later, in 1937 he became the School's Director. He held this post till his retirement from the School in 1957. The great expansion of the School under his Directorship was the result of Sir Ralph's far-sightedness and the untiring efforts with which he pursued his objectives. The littleknown story of the battle on the home-front which Sir Ralph fought against the "official apathy or lack of fore-sight" is vividly narrated by J. C. Wright and C. D. Cowan in their obituary published in the BSOAS 47 (1984), 540-548. Sir Ralph's research career in which he handled a number of problems 1 In a letter dated 20-12-1985 Miss Diana Matias, Editorial Secretary, BSOAS, informs me that Prof. Turner was invited a second time in 1922 to deliver Wilson Philological Lectures, but he could not deliver them because he was that year appointed to the Chair of Sanskrit in SOS', London, Madhu Vidya/724

Page #750
360 Atinats BORI, LXVI (1985) related to Indo-Aryan linguistics extends over seventy years. His first publi. shed paper, however, was " Agaiost the stress accent in Latin " (1912), and the last << Implosive d- and y-orr-'. (1982). The list of Sir Ralph's published papers and Addresses is quite large. His " Collected Papers (1912-1973 )" appeared in 1975. Sir Ralph's Wilson Philological Lectures (apparently unpublisbed ) have been referred to above. He was invited by the University of Poopa in 1958 to deliver the first P. D. Gupe Memorial Lectures which were published in 1960. The subject of his lectures was "Some Problems of Sound Change in Indo-Aryan". In these lectures Sir Ralph dealt with some special conditions affecting the sound change, the disturbances caused by ana. logy, and the cases presented by loanwords. Sir Ralph's eminence as a comparatist depends on his two monumental Dictionaries. Sir Ralph's acquaintance with the New Indo-Aryan languages began with Gujarati which he learnt even as he was a student of Christ's College, Cambridge. During the First World War he became acquainted with Nepali. His subsequent writings suggest that he was also familiar with Sindhi, Hindi, and Romani, the language of the Gypsies. His ever-widening interest in the New Indo-Aryan languages enabled him to give to the world of scholars A Comparative and Etymological Dictionary of the Nepali language in 1931. During the years wben Sir Ralph was entrusted with the onerous duties of the Directorsbip of the School, he had very little time to study the materials he had collected over the years for his other major work, A Comparative Dictionary of the Indo-Aryan Languages. But at the age of 70 when he was free from these duties he returned to his academic pursuit with great zeal. The entire publication of the Dictionary was completed in a remarkably short span of four years (1962-1966). In this Dictionary are brought together about 1,40,000 words from the Indo-Aryan languages spokeu in five countries - India, Pakistan, Bangladesh, Nepal, and Sri Lanka. These words have been arranged under some 15000 Sanskrit head-words, attested or reconstructed, which suggest their etymologies. Already in 1920, in the Inaugural Address delivered by Sir Ralph when appointed Professor of Linguistics at the Benares Hindu University, he had visualized the possiblity of the scientific studies of Indian languages 2 In the Obituary mentioned above Sir Ralph's last piece of research work is said to be his re.copying the slip meant to be used in the Addenda as his Dictionary entry No. 6672, Madhu Vidya/725

Page #751
Obltuary Notices 361 done by Indians themselves. This work, already started by a few pupils of Sir Ralph, is now being done more extensively at some of the University Departments of Linguistics. The establishment of these Departments is the direct result of the deliberations carried on and the resolutions passed at the Conference of Linguists and Educationists held at Deccan College, Pune, from 26th to 29th May 1953 under the General Presidentship of Sir Ralph Turper. In recognition of the services rendered by Prof. Turner to the cause of the advancement of knowledge, knighthood was conferred on him in 1950. He also received many other academic distinctions from the Universities in the U. K., India, and Nepal. In addition, he was twice invested by the two successive kings of Nepal with the Order of Gorkha Daksina Babu First class. The task of preparing an Index to any volume is laborious, all the more so when it comes to the preparation of an Index to a dictionary. But this was accomplished with great devotion by Mrs. Turner. An Index of all the words cited from the different languages in the Nepali Dictionary was prepared by her and published along with the Dictionary in 1931. The Index to the Comparative Dictionary, prepared on similar lines, was also compiled by Mrs. Turner and published subsequently as a supplementary volume (1969). Not only this. Mrs. Turner also collaborated with her husband in the preparation and publication (1971) of a second supplementary volume designated as Phonetic Analysis, which, in fact, turns out to be a regrouping of the materials presented in the Etymological Dictionary, with a view to helping any one interested in examining the histories of some 1500 sounds or sound.groups from the earliest to the latest phases of Indo-Aryan. "No account of the works of Sir Ralph would be adequate", observes Prof. Brough in his Foreword to the " Collected Papers ", " if it did not ioclude a tribute to his wife, Dorothy Rivers Turner, who, until her death in 1972, gave him such constant aid and support, not least in compiling the lodexes ..." M. A. Mebendale 1 Mention of bis Presidential Address is missing in the list of the "Writings of Sir Ralph (Books and Articles )" (upto 1957) published in BSOAS Vol. 20 (1957) published in honour of Sir Ralph Turner. 46 (Annals BORI) Madhu Vidya/726

Page #752
Obituary Notices 365 PROFESSOR DR. ERNST WALDSCHMIDT b. 15-7-1897 1 [ d. 25-2-1985 Prof. E. Waldschmidt, Honorary Member of the Bhandarkar Oriental Research Institute (1978), breathed his last on February 25, 1985. He was 87. In his death Indological studies in general and Buddhist studies in particular have lost a diligent and painstaking scholar, Prof. Waldschmidt was born on July 15, 1897 at Lunen (Westfalen). He began his University education in 1919 under Prof. Paul Deusssen and Prof. Emil Sieg. He obtained the Ph. D. degree in 1924 at Berlin by writing a dissertation on Das Beichtformular der buddhistischen Noonen'. He served for some time in the Museum fur Volkerkunde in Berlin wbere, as the Curator, he was in charge of the Indian colletions. In these collections were stored, among other things, the Turfan manuscripts discovered in Central Asia. Here Prof. Waldschmidt got his first opportunity to study the materials which later were destined to be the mainstay of his and bis pupils' brilliant research work. In 1930 Prof. Waldschmidt was appointed Lecturer in Indologie at the University of Berlin. He left this appointment in 1936 when he was called upon to be the successor to Prof. B. Sieg at the University in Gottingen. He remained at this post for nearly thirty years until in 1965 he retired as Professor Emeritus. It was during this period that Prof. Waldschmidt and his pupils (to name only a few, Dr. H. Hartel, Prof. D. Schlingloff, Dr. Mrs. V. StacheRosen, Prof. H. Bechert) made significant contributions to the study of Buddhism. With single-mioded devotion they worked on the Turfag manuscripts, mostly fragmentary in nature, and published them along with parallel versions in other languages, translations and explanations. Whatever work was done by Prof. Luders and Prof. Waldschmidt on these manuscripts before 1939 was misplaced during the second world war. Most of this could be recovered and Prof. Waldschmidt aad his pupils made a fresh bid to work on the Turfan fragments. Among Prof. Waldschmidt's major publications - besides, of course, a number of papersi - may be mentioned Das Mahaparinirvanasutra (1950-51), Das Mahavadanasutra ( 1956 ), Das Catusparisatsutra ( 1952-62 ), Faksimile 1 A collection of his writings appeared in 1967 under the title on his 70th firthday. Von Ceylon bis Turfan Madhu Vidya/727

Page #753
366 Annals BORI, LXVI (1985) Wiedergaben von Sapskrithandschriften aus der Berliner Turfaofundea (1963); Miniatures of Musical Inspiration in the collection of the Berlin Museum of Indian Art, Parts I and II; from the Nachlass of Prof. Luders, Beobachtungen uber die Sprache des buddhistischen Urkanons (1954), and Bharhut Inscriptions ( 1963). As early as 1953 Prof. Waldschmidt had planned to publish a lexicon based on the Buddhist Sanskrit texts published by him and by his pupils. The work had progressed slowly and the actual publication, edited by Georg von Simson and Heinz Bechert, started in 1973 wheu part 1 appeared, which was followed by part 2 in 1976. Prof. Waldschmidt received many honours during his life time. He was elected President of the International Congress of Orientalists held in Munchen in 1957, A Felicitation Volume - Beitrage zur Indienforschung - Ernst Waldschmidt zum 80. Geburtstag gewidmet - was presented to him in 1977 on the occasion of his 80th birthday. In 1957 Prof. Waldschmidt donated his house, and also his library, in the Hainbundstrasse to the Uoiversity in Gottingen for housing the Indologisches Seminar which till then was located in the Prinzenstrasse, Prof. Waldschmidt had to serve in the army in both the world wars, When his 85th birthday was celebrated with great honour on 15. 7. 1982 he told bis audience that in both the wars on many occasions he came very near to serious danger to his life but luckily survied them all to be able to live long enough to carry out his destined work. Wbat he said on that occasion, while concluding his address, was quite characteristic of him. He then said: "I would have very much liked to come close to the ideal of a scholar that I have in mind, the one characterized by an endeavour for truth and, as far as possible, by an objectivity in the service of research, such one having admiration for the achievements of the predecessors, being conscious of his responsibility and keeping himself very much in the background. I do not know how far I succeeded. It is my wish at any rate that the typical objective researcher does not die out. I feel the urge to speak this out and may God belp me towards the fulfilment of my wish." M. A. Mebendale Madhu Vidya/728

Page #754
OBITUARY NOTICE PROFESSOR Dr. SUMITRA MANGESH KATRE b. 11-04-1906 ] [d. 21-10-1998 On the 26th October 1998 I first heard the sad news of the passing away of Dr. Katre. He was 92. He died at the residence of his elder daughter at San Jos:, Calif., U. S. A. In Dr. Katre's death the world of scholars has lost a versatile personality who combined in himself high Sans krit scholarship, fruitful planning, and administrative skill.. I met Dr. Katre first in Bombay in 1939 when I was a student for M. A. The University of Bombay had then prescribed for the M. A. exami. nation the Jasaharacariu of Puspidanta. Since the text was in Apabhramsa it was not easy to understand it without some guidance. No one at that time was free to teach the text. At the suggestion of Prof. H D. Velankar I approached Dr. Katre to request him to read the text w th us. He prom. ptly agreed and our classes began almost immediately. Little did I dream that the contact I then had with Dr. Katre was destined to develop later into my being first his pupil and then his colleague in the Deccan College. When the old defunct under-graduate Deccan College was revived in 1939 as Deccan College Post-Graduate and Research Institute, Dr. Katre joined it as Professor of Indo-European Philology. I too joined the Institute the same year to do my Ph. D. under bis guidance on the subject, "Histori. cal Grammar of Inscriptional Prakrits". E. D. Kulkarni was his another Ph. D. student who worked on Epic Variants, Dr. Katre had his primary and secondary school education in Mangalore. Even as a high school student, he had studied Sanskrit Grammar the traditional way. However, when he went to Madras for his graduation he chose Mathematics as his subject. After his receiving B. A. degree in that subject in 1928 he was advised to enrol himself in the Trinity College, Cambridge, for higher qualifications in Mathematics. By a queer course of incidents he could not produce in time the necessary documents for getting admission to the College. With the failure in one direction is linked the story of his magnificent success in the other. Instead of returning to India empty. handed, Dr. Katre sought admission to the School of Oriental Studies in London on the strength of his knowledge of Sanskrit Grammar. After gett ing through the qualifying examinations in record time he started working for his Ph. D. ungr the gidance of Dr, William Stede on the subject "Early Madhu Vidya/729

Page #755
304 Annals BORI, LXXIX ( 1998) Buddhist Ballads and their Relation to the Older Upanisads". During his stay in London he regularly attended Prof. Turner's classes in Indo-European and Indo-Aryan, which proved extremely helpful to him in his later career. During his tenure in London he had an opportunity to spend a summer sem. ester in Germany. He completed his dissertation for Ph. D. in 1931. At the time when Dr. Katre return:d to ladia with initiation in IndoEuropean and lado-Aryan linguistics and in Prakrits, there were no openings for tcaching and doing research in Linguistics in any of the academic institutes associated with the Indian Universities. The University of Calcutta was perhaps the only exception. Dr. Katre, therefore, had to start teaching Prakrits first in the Nowrosji Wadia College and then in the S. P. College, both in Pune. The chances of finding an opportunity to make use of his special talents were then bleak. The revival of the old Deccan College in 1939 bence proved to be a very crucial event in his career. His selection as the first Professor of Indo-European Philology in the newly revived Institute gave him ample opportunities for planning and execution of his brilliant and bold ideas, His close association during his stay in Pune with Prof, P. K. Gode and Dr. V. S. Sukthankar, both of the Bhandarkar Oriental Research Iostitute, proved very valuable for him. In collaboration with Prof. Gode he revived the de. funct journal, Indian Antiquary, in the form of the New Indian Antiquary ( 1938-39 ) and also launched the publication of a new journal, viz. the Oriental Literary Diges! ( 1937). These publications made it possible for Dr. Katre to come into close association with the scholars of his time and their published research. Dr. Sukthankar, the then General Editor of the Mahabharata, introduced Dr. Katre into yet another area of fruitful research, viz. Texual Criticism, Some of the early Ph. D. dissertations completed under the guidance of Dr. Katre were related to the Critical Edition of the Mahabharata. Dr. Katre bimself later published his Introduction to Indian Textual Criticism ( 1941 ). When Dr: Katre was appointed as Professor in tbe Deccan College Research Tostitute (1939), there was provision only for three Professors and five Readers. The other two Professors who joined the Institute along with Dr. Katra were Dr. V. M. Apte ( Vedic Sanskrit ) and Dr. H. D. Sankalia (Anci. ent Indian History). The arrangement which was then decided upod was that one of the three Professors was to function also as the Director of the Institute. Owing to some reasons none of the three Professors named above was nominated to the post of the Director, and hence a Committee of Direction was appointed to carry out the Director's functions for one year. In the next year (1940) Dr. I, J, S. Taraporwala was appointed as the Director, Madhu Vidya/730

Page #756
Obituary Notice 305 According to the original provision there were to be three Professors in the Institute, and since Dr. Taraporwala and Dr. Katre bad specialization in the same subject (Indo-European Philology), a situation arose in which Dr. Katre would have to leave the Deccan College. But the experts who then guided the destiny of the Deccan College had realized the value of Dr. Katre for the all-round development of the Institute and hence, as a special case, they arranged for the provision of a fourth Professor. As it happened Dr. Taraporwala resigned his post only two years later (1942) and, in his place, Dr. Katre was appointed the next Director. He thus bappens to be the youngest Director of the Institute and also the one who occupied-that post for the longest duration (1942-1971). His name has justifiably been identified with the Deccan College, so much so that the Government Bungalow No. I on the campus in which Dr. Katre lived from 1950'-1971 is even now known as Dr. Katre's Bungalow although some four or five persons have lived there after 1971, Dr. Katre's successful career in the Deccan College has various facets. The most promnient among these pertain to 1 ) the planning of the Dictionary of Sanskrit on Historical Principles, 2) the impetus to the study and teaching of Linguistics in all parts of India, and 3) the various publications of the Deccan College. Dr. Katre, it appears, had in mind the organizing of the great Historical Sanskrit Dictionary even as he joined the Deccan College in 1939. His plan was no doubt very bold and needed the cooperation of many scholars. This meant availability of large funds which were not immediately at his disposal. The annual budget of the Deccan College in those days was only Rs. 75,000, just enough to maintain the staff and to carry on the routine administration. But Dr. Katre was not willing to wait till he could gather around him sufficient number of scholars and was assured of enough financial support. Robust optimism was the mainstay of his character. He made a very inodest start almost immediately after he took charge as Director of the Institute in 1942. He initiated a limited project of Dictionary of Inscripti. onal Sanskrit. He appointed two assistants for this purpose (the present writer being one of them ) and got started the work of collecting material from published Sanskrit inscriptions. His idea was highly practical. Even if Dr. Katre had failed to make a start for his ambitious plan of the great dictionary, he could have certainly completed this small dictionary of During the second world war, the Deccan College was temporarily shifted to another Flace in the city where it remained from 1940 to 1950, 39 (Annals BORI) Madhu Vidya/731

Page #757
Annals BORI, LXXIX (1998) Sanskrit inscriptions which would have offered the information not available in the dictionaries then in use. However, Dr. Katre made good progrers in his attempts to secure academic and financial assistance for the large dictionary, and in 1948 he was able to lay the foundation of his Dictionary Project. Many developments took place in the coming years end Dr. Katre finally succeedel in establishing an autonomous Sanskrit Dictionary Department in the Deccan College. The earlier Sanskrit Dictionaries were based on not more than about 500 works which number, considering the wide range of Sanskrit literature, was rather small and hence a fresh attempt based on a larger number of texts was called far. Speaking of the proposed dictionary in Bombay in 1949 Prof. Renou observed: "A scheme for a Sanskrit Thesaurus on the lines undertaken by the Deccan College Research Institute has long been considered by many Orientalists as absolutely indispensable." The learned Prefessor further went on to observe : " Hence it becomes necessary that the task be undertaken by your country. Only among you can be found in adequate numbers practised philologisrs able to direct the work and, above all, the greater multitude of humble collaborators, the local pandits, themselves good Sanskrit scholars and ready to dedicate themselves to the common task.? 306 Dr. Katre's proposed dictionary was to embrace nearly 1500 Sanskrit texts besides the material collected from Sanskrit inscriptions. A selection of these texts appeared in the form of a Minimum Programme which was drawn up with the assistance of the late Prof. L. Renou of Paris. This was the first and the most essencial step to be taken. On its basis the work of extracting vocables with their exact references and citations was done by the editorial staff and their assistants employed in the Dictionary Department. When Dr. Katre retired from his post in the Deccan College this spade work, which lasted for nearly twentyfive years, was to some extent completed. Dr. Katre himself contributed his mite by publishing his Dictionary of Panini in three parts (1968-69). Much and more important work, however remained to be done. The whole responsibility of editing the Dictionary in the real sense of the term was shouldered by Dr. A. M. Ghatage, the next General Editor of the Dictionary. He and his calleagues succeeded in giving Earlier in 1940, H. Ortel had thought of the same idea in his "A Practical Proposal for Preliminary Work on a New Sanrkrit Dictionary" (Woolner Commemoration Volume, Lahore, pp. 177-82), where he aiso records in a footnote similar earlier suggestions from W. Wust (19-9) and Vanamali Vedantatirtha (1930). His complete English translation of Panini's Astadhyayi was to appear much later (1989). Madhu Vidya/732

Page #758
Obituary Notice 307 the final shape to the Dictionary. So that ultimatedy it started publication in 1978. The work is still in progress. Dr. Katre's another area of interest pertained to the all-round development of linguistic studies in India. In 1939 a lone Chair for Indo. European Philology was established in the Deccan College. With Dr. Katre's occupancy of that Chair the research carried out by him and his students was given the name ". Katre School of Linguistics" by the renowned scholar Dr. Siddheshvar Varma of the University of Jammu. But Dr. Katre was not satisfied by the progress he had made for he realized that what was then being done in the Deccan College went only in one direction viz, historical linguistics. The science of linguistics had many other equally important aspects. Dr. Katre desired all-round development of the science. To fulfil his plans be took the first step in organizing in the Deccan College in 1953 a Conference of Linguists and Educationists in India. As a result of the deli. berations of this Conference wbich were published in the form of a Report, Dr. Katre was able to secure a handsome grant from the Rockefeller Fountion of the United States. With this help, which lasted for five years, Dr. Katre was able to organize Winter Seminars and Summer Schools of Lingui. stics at Pune and at different Universities in India from 1955 to 1960. Rocke. feller Foundation's help also enabled Dr. Katre to seek cooperation of lingui. sts from America and Great Britain to participate in the teaching progiamme at those Schools and Seminars. Some of the Indian scholars who attended these Schools and Seminars were awarded Junior and Senior Fellowships by the Rockefeller Foundation to enable them to proceed to the United States for a year or two for advanced study in linguistics. As a result of this total activity, Departments of Linguistics were established at many Universities in India and they were staffed by the products of the linguistic activity intiated by Dr. Katre. In 1964 the Deccan College was recognized by the University Grants Commission as a Centre of Advanced Study in Linguistica, Dr. Katre's third principal activity pertains to the Deccan Collega publications. From the very first year of its coming into existence the Deccan College Research Institute started publishing its own annual Bulletin which provided the means to its staff and students to publish their research. The Bulletin was mainly looked after by Dr. Katre. Much later after he became tbc Director of the Institute, he initiated many new Series like the Deccan College Monograph Series ( 1946 ), Deccan College Dissertation Series (1946), Deccan College Hand-Book Series (1951), and Building Centenary Series (1964). Besides, as a supplement to the work that was being dune in the Sanskrit Dictionary Department he started a series called Sources of Indo. Madhu Vidya/733

Page #759
308 Annals RORI, LXXIX (1998) Aryan Lexicography (1947), in which many unpublished Sanskrit Kosas were published. Yet another publication of the Dictionary Department was Vak (1951) which published articles mainly of lexicographical interest. lodia as a linguistic area where languages and dialects of not less than four language families are spoken, offers an excellent field for language descriptions. Dr. Katre, therefore, desired that a permanent department of Linguistic Survey of India be established by the Central Goverment on the lines of the Archacological Survey of India. Dr. Katre himself had in the initial stages of his career published Konkani Phonetics (1935) and Forma. tion of Konkani (1942, 2nd edn. 1965). Althongh Dr. Katre did not succeed in his efforts in this direction during his stewardship of the Deccan College, a sort of mini-linguistic survey of the dialects spoken in Maharashtra was carried out and published by Dr. A. M. Ghatage and his studenes with the support of the Government of Maharashtra, Dr. Katre wanted to establish also a printing press on the campus of the College mainly for the purpose of getting the publications of the Deccan College printed to his satisfaction. Even this dream of his remained unful. filled. In retrospect, Dr. Katre's academic and administrative career can be looked upon as highly successful and satisfying for him. He was able to put into practice most of his major ideas. Any one who has seen the Deccan College of 1939 with a small staff of seven members active only ia a part of its old building, its hostels almost empty, its Library with a modest collec. tion of books, an institute baving no Museums of its own (Dr. Sankalia's first few estampages of cave inscriptions were then displayed on two newspaper stands in his room), and compares it with the present Deccan College with a large staff functioning in its spaced out buildings, not on one but two campuses, having an excellent, well-maintained Library which can boast of a very rich collection of books and an excellent periodical section, Boys' and an added Ladies' Hostel with all its rooms occupied, two MusiumsArchaeology and Maratha History -- and a Phonetics Laboratory, will realize what Dr. Katre has been able to achieve within a span of a little over thirty years of his association with the Deccan College. Extremely gentle in his behaviour, Dr. Katre was politeness incarnate. He spoke little and was soft-spoken. His personality was impressive and was blessed with very good health. He enjoyed playing tennis, Hospitable by Dature he treated alike all those who visited him. He had full cooperation of Mrs. Katre in this regard. He himself also enjoyed different dishes. His Madhu Vidya/734

Page #760
Obituary Notice 309 bebaviour with his colleagues was graced by respect for them. Any one ex. periencing some difficulty or the other had easy access to him, and Dr. Katre would do all that he could to remedy the situation. Not only this. If any one had any differences of opinion with him, academic or otherwise, he felt ao hesitation in approaching him and giving expression to his vicws. He felt no fear of being disrespected for his boldness. All those who got an opportunity of working in the Institute at the time when he was at the helm of affairs must be looked upon as really fortunate, I offer Dr. Katre my respectful homage. M. A. Mehendale Madhu Vidya/735

Page #761
Index to Authors / Editors of Books Reviewed Aiyar R. Swaminatha Ananthanarayana, H. S. Bhat, D.N.S. Bhattacharya, Gopikamohan (ed.) Bhattacharya, S. Burrow, T. Chatterji, S. K. Chaturvedi, Raghuveer Dandekar, R. N. (ed.) Jange, Sadashiv A. Datta, Bhakti Diwekar, H. R. (ed.) Edgerton, Franklin Falk, Harry Ghatage, A. M. Godakumbura, C. E. (ed.) Gonda, Jan Goswami, Upendranath Hartman, Karl Gustav Hauschild, R. Hino, Shoun (ed.) Jog, K. P. (ed.) Kahrs, Eivind Kak, Subhash Kashikar, C. G. (ed.) Keller, Joseph R. Lienhard, Siegfried Limaye, V. P. Luders, H. McDermott, A. C. Senape (ed.) Matilal, B. K. (ed.) Mittwede, Martin Mylius, Klaus (ed.) Naik C. R. Nanavati, Rajendra Panda, Smt. Jayanti Pandit, P. B. Paranavitana, S. (ed.) Presiado - Solis, Benjamin Saksena, B. R. (ed.). Sen, Subhadra Kumar Shastri, Shivanarayana Sheuer, Jacques Schmidt, Hanns - Peter (ed.) Simha S. L. N. Slaje. Walter (ed.) Steinkellner, Ernst (ed.) Sternbach, Ludwik Tagare, G: V. Thieme, Paul Thumb, Albert Varma, Dhirendra Varma, Siddheshwar Vogel, Klaus Vyas, K. B. (ed.) Walimle, R. S. (ed.) Yardi, M. R. Madhu Vidya/736

Page #762
Our Latest Publications (2000-2001) Some Topics in the Development of OIA, MIA, NIA - Dr. H. C. Bhayani (1998) 75.00 Anantanaha Jina Cariyam - Ed. Pt. Rupendrakumar Pagariya (1998) 400.00 Alankaradappana - Ed. Dr. H. C. Bhayani (1999) 50.00 Astaka Prakarana - Dr. K. K. Dixit (1999) 75.00 Siri Candappahasami Cariyam - Ed. Pt. Rupendrakumar Pagariya (1999) 250.00 Tattvartha Sutra - (Translated into English by Dr. K. K. Dixit) 300.00 Tarka-Tarangini - Vasant G. Parikh 270.00 Sanmati Tarka - Pandit Sukhlal Sanghavi, Pandit Bechardas Doshi 225.00 SAMBODHI 100.00 The Journal of the L. D. Institute of Indology : (Back Vol. 1-21) Per Vol. Current Vol. 23 (2000), 24 (2001) 150.00 Our Forthcoming Publications DaoN. mo. da. desAI saMpAdita prAcIna kRtio Sastravarta samuccaya Saptapadaarthi Jain Edu le www.jaimellibre