Halas Sattasai

Author(s): Hermen Tieken
Publisher: Leiden
Catalog link: https://jainqq.org/explore/006962/1

Page #1
HALA'S SATTASAI STEMMA AND EDITION (GATHAS 1-50), WITH TRANSLATION AND NOTES HERMAN TIEKEN REPORail ghaTavihivAdhAkArtikamayAsamapatityapahAtAnna dinamAnasAmaMtAkatAnakaDaziparikSitasAhanidarSi jhASitAmakArakanimaNamAdehitakAnamA avaMzayAsAradIvAnApanavAsavatikavadhAyA

Page #2
HALA'S SATTASAI STEMMA AND EDITION (GATHAS 1-50), WITH TRANSLATION AND NOTES Leiden 1983

Page #3
On the cover is reproduced the third leaf (recto) of a MS of the Sattasas at present kept in the Shri Hemachandracharya Jain Jnanamandira in Padana in Gujarat, India. For a description of the MS, see A descriptive Catalogue of Manuscripts in the Jaina Bhandars at Pattan, edited by C. D. Dalal, Vol. I, Palm-leaf MSS, Baroda, 1937, p. 178 ff. The leaf reproduced contains a part of the colophon.

Page #4
HALA'S SATTASAI STEMMA AND EDITION (GATHAS 1-50), WITH TRANSLATION AND NOTES PROEFSCHRIFT TER VERKRIJGING VAN DE GRAAD VAN DOCTOR IN DE LETTEREN AAN DE RIJKSUNIVERSITEIT TE UTRECHT, OP GEZAG VAN DE RECTOR MAGNIFICUS : PROF. DR. O. J. DE JONG, VOLGENS BESLUIT VAN HET COLLEGE VAN DECANEN IN HET OPENBAAR TE VERDEDIGEN OP VRIJDAG 4 NOVEMBER 1983 DES NAMIDDAGS TE 2.30 UUR door Herman Joseph Hugo Tieken * geboren op 1 april 1952 te Lijnden

Page #5
Promotor: Prof. Dr. H. W. Bodewitz

Page #6
CONTENTS VI1 Acknowledgements List of abbreviations INTRODUCTION Notes to the Introduction PART I CHAPTER 1 1.1 Previous editions (introductory) 1.2 The first edition by Weber 1.3 The MSS edited by Weber CHAPTER 2: 2.1 The three new South-Indian MSS: Ma, Ti and Tp CHAPTER 3 3.1 The stemma, its scope, method and results CHAPTER 4 4.1 The Third South-Indian recension, Jaina-recension and the Vulgata (introductory) 4.2 The archetype 4.3 The relation between the MSS Ma, Ti and Tp of the Third South-Indian recension 4.4 The relation between the Jaina-recension and the Vulgata 4.5 The relation between the MSS K, V, B, Y and P of the Vulgata 4.6 The contaminated state of the text of MS R of the Jaina-recension 4.7 Diagram showing the relation between the MSS of the Third South-Indian, Jaina and Vulgata-recensions CHAPTER 5 5.1 The eclectic recensions: Sadharanadeva, Telinga and Second Telinga (introductory) 5.2 The Sadharanadeva-recension (S) 5.3 The Telinga-recension (T)

Page #7
VI 5.4 The Second Telinga-recension (W) CHAPTER 6 6.1 Concluding remarks Notes to the stemma. APPENDIX I APPENDIX II APPENDIX III 87 111 119 PART II 143 144 150 151 152 154 CHAPTER 1 1.1 The nature of the text 1.2 The situations 1.3 The audience 1.4 The compiler 1.5 The language of the gathas Notes to Chapter 1 CHAPTER 2 2.1 The method of reconstruction 2.2 The variants 2.3 The presentation Notes to Chapter 2 CHAPTER 3 3.1 Edition of Gathas 1-50, with translations and notes 161 173 176 178 179 267 271 Index of words discussed Index of grammatical points discussed Bibliography Nederlandse samenvatting Curriculum vitae 273 281 283

Page #8
VII Acknowledgements It is to a large extent due to the help and encouragement received from Professor Bodewitz that I have managed to complete my doctoral dissertation on the Sattasaf this year, precisely one hundred years after the last part of the 'first edition' by Weber appeared. It has been a pleasure to finalize my studies under the supervision of the person who also guided my first steps into the study of Sanskrit. I feel deeply indebted to Dr. G. Schokker who introduced me to the study of Prakrit and who has remained ever helpful with suggestions and criticism, I wish to express my gratitude to Professor 0. von Hinuber with whom I discussed earlier versions of most of the Gathas edited here and who kindly allowed me to profit from his extensive knowledge of Middle Indic languages. I feel much obliged to Dr. Th. Damsteegt who read the complete final draught and gave many useful comments. In preparation to the work on this edition I visited several manuscripts libraries in India and Nepal. A subsidy was granted to me by the Netherlands Foundation for the Advancement of Tropical Research (WOTRO) for the expenses involved during the trip. In this connection I should also like to thank Professor K.L. Janert who kindly allowed me to use his extensive library of catalogues of manuscripts collection's, and the authorities of the Government Oriental Manuscripts Library in Madras and of the University Manuscripts Library in Trivandrum for allowing me to photograph several manuscripts from their collections. Finally I wish to thank my wife Ingrid. All problems arising in connection with the interpretation of the text were first discussed with her. These discussions made the work on the Sattasat a real pleasure.

Page #9

Page #10
LIST OF ABBREVIATIONS ABC Abh. AIG AKM AMG. Ara. CDIAL CPD DED Desin. Ed. EWA NA Guj. Janert, An Annotated Bibliography of the Catalogues of Indian Manuscripts Ueber das Saptasatak am des Hala; ein Beitrag zur Kenntniss des Prakrits, by Weber, AKM V 3, 1870 Wackernagel and Debrunner, Altindische Grammatik Abhandlungen fur die Kunde des Morgenlandes Ardha-Magadhi Apabhras a Turner, A Comparative Dictionary of the Indo-Aryan Languages Trenckner, A Critical Pali Dictionary Burrow and Emeneau, A Dravidian Etymological Dictionary Des inamamala Das Saptasatakam des Hala, by Weber, AKM VII 4, 1881 Mayrhofer, Kurzgefasstes Etymologisches Worterbuch des Altindischen Gangadhara('s commentary or his version of the Vulgata) Gujarati Hindi Hemacandra's Grammatik der Prakritsprachen, by Pischel Weber, 'Ueber Bhuvanapala's Commentar zu Hala's Saptasatakam', Indische Studien 16, 1883 Journal Asiatique Jaina Maharastri Maharastrt Marathi Magadhi Middle Indo-Aryan manuscript(s) Monier-Williams, A Sanskrit-English Dictionary Raghavan, New Catalogus Catalogorum New Indo-Aryan e.g. Oguj., Old Gujarati Panjabt Paiyalacchi JA JM. M. Mar. Mg. MIA MS(S) MW NCC NIA . Paiyal.

Page #11
Pkt pr. m. PSM PTSD PW Retr. Prakrit prima manus Sheth, Paias addaman annavo Rhys Davids and Stede, The Pali Text Society's Pali-English Dictionary Bohtlingk, Sanskrit-Worterbuch nebst allen Nachtragen Weber, 'Zum Saptasatakam des Hala', ZDMG 28, 1874 Sindhi Saurasen 1 secunda manus Sinhalese Sanskrit The Prakstaprakasa or the Prakrt Grammar of Vararuchi, by Cowell Zeitschrift der Deutschen Morgen landischen Gesellschaft sec. m. Si. Skt Var ZDMG

Page #12
Corrections and additions p.11, 11.27-32 delete. Insert: On the basis of purely external factors Weber arrived at the conclusion that of these six recensions the Vulgata was the most authentic one available (pp. L-LI). p. 30 after 1. 26 add footnote 119: By definition most retentions are the more difficult readings and are therefore unlikely to be introduced secondarily in a MS for an innovation. This appears nevertheless occasionally to have happened (see, for instance, p. 52), but apparently not systematically or on a large scale. At first the retention could have come to be noted as a variant in the margin of a MS. Subsequently the possibility of choice may have arisen again, which need not always have been settled in favour of the innovation. It should be noted that the source of the secondarily introduced reading need not always have been a second MS. An alternative source could have been formed by the commentaries, which may in certain instances have preserved the original reading and, in explaining it, have turned it into an acceptable variant. These commentaries are also responsible for the introduction occasionally met with of one and the same innovation occurring independently in two different MSS (for a possible instance, see p. 45), especially where it concerns an innovation of the first type, namely a Sanskrit gloss for a typical Prakrit word. p.32, 11.4-7 delete: Because of this...with some other MSS. Insert: These innovations point to the conclusion that the MSS have suffered from contamination, the extent of which remains, however, difficult to gauge. p.83, 11.4-7 delete. Insert: The word is found a second time in a context which is not likely to have been confused with the present one, namely in 741: jai vi hu dinnundalia, etc. (Ma duddalia, Ti dinnandalia, Tp dinnuddalia; Madhavayajvamisra: krtandolika, undaliketi balikakrTCIvisesah; v.1. Tdi(1) lidilia, TC dilandilia, s mandamkillia, w dullia, x vimdilia; cf. dillindilia 'child' in Paiyal. 58 and Desin. V 40).

Page #13

Page #14
INTRODUCTION The present study is preparatory to a critical edition of the complete text of the Sattasai.' The first edition of the Sattasat consists of four successive publications by A. Weber issued in 1870, 1874, 1881 and 1883 respectively, the last one appearing precisely one hundred years ago. Weber used altogether eighteen MSS which he divided into six recensions. As the basis for his edition he selected the so-called Vulgata-recension, which according to him was the most authentic one. However, a closer look at the text of the Vulgata reveals many instances where its readings are obviously secondary compared to those of the other recensions.' Another feature which becomes clear from Weber's edition is the great number of variants available for the text of each individual Gatha. These variants, besides including the usual number of accidental errors, also, and more importantly, concern entirely different words, Padas or even hemistiches. These factors call for a new edition of the text, in the course of which the question of the stemmatical relations between the MSS should be taken up anew. .. The present edition thus begins with the study of the stemmatical relations (Part I). In the edition itself, of which in this first volume only a sample consisting of the first 50 Gathas is given, it has been attempted to establish for each Gatha a text as close to the original as was possible with the MSS available (Part II, Chapter 3). The stemma involves ten of Weber's eighteen MSS, namely those which contain the Prakrit text. The other eight MSS contain only a Sanskrit commentary or a Chaya; thus they need not be considered. Consequently the stemma covers five out of Weber's six recensions, the so-called x-recension being available only through a commentary. To these ten MSS could, however, be added three so far unknown MSS from South-India. The latter three MSS contain a seventh recension of the Sattasal, which, after Weber's Telinga and Second Telinga, will henceforth be referred to as the Third South-Indian recension. One of the results of the stemma is the conclusion that the MSS are basically divided into two branches, a South Indian and a North-Indian one. Furthermore it appears that all available MSS, South Indian as well as North-Indian, ultimately go back to one single North-Indian MS which

Page #15
itself is lost and will henceforth be referred to as the archetype. . The text of the South-Indian branch is in many respects the most authentic one available. As to the order of the Gathas, which differs for each recension, the most original one is found in the North-Indian branch. The stemma should provide the basic arguments in reconstructing the text of the archetype, in respect of the order as well as of the text of the Gathas. This does not mean that the reconstruction has become an automatic process. There appear to be several cases where it is impossible to rely solely on the stemma, for instance, in the case of variation in the forms of the words, which appear to have been considerably influenced by Sanskrit in both the branches of the MSS. The method of reconstruction and several recurrent problems encountered will be discussed in Chapter 2 of Part II. In each case the final test applied to the text is the translation. The Sattasaf has already been translated many times and into many languages, European as well as Indian. Of these only Weber's German translation is critical in that it is verifiable. My translation differs from Weber's in many instances often simply because it is based on a different text. Generally, however, the differences have another reason. : It is to be noted that Weber relied heavily on the traditonal interpretations found in the Sanskrit commentaries which in many MSS accompany the text. For the meanings of words Weber referred in particular to Hemacandra's Prakrit grammar and the same author's Desinamamala (twelfth century A.D.) and to Dhanapala's Paiyalacchi (tenth century A.D.). The latter two works are Prakrit dictionaries. It will appear that often the interpretations given in the Sanskrit commentaries rest on misunderstandings and have been based on faulty divisions of words or compounds, on vague associations, preconceived notions or etymological speculations. Furthermore a considerable number of rules and words included in the above-mentioned grammar and dictionaries appear to be due to their authors' or their predecessors' attempts to account for the derivation and meaning of certain words of, among other texts, the Sattasat. These authors' methods do not differ significantly from those seen in the commentaries. One of the conclusions that may be drawn on the basis of the above is that between the time of the composition of the Gathas of the

Page #16
Sattasai and that of the Sanskrit commentaries, Hemacandra's grammar and the two dictionaries, there must have been a certain period of relative neglect of, or unconcern with, the text; a period, moreover, long enough to lose sight of the original intention of many of the Gathas and of the meaning of many of the words occurring in them. The interpretations of the individual Gathas is often difficult. They contain associations or refer to practices and ideas which are not otherwise known. Many Gathas do not have parallels outside the text itself. A further context is provided by the text as a whole, which seems to have been compiled with the specific aim of amusing well-to-do city-dwellers and courtiers. Practically each Gatha seems to have to be understood in this way, which occasionally may indicate a way out to arrive at a possibly correct interpretation. This broader context, provided by the text as a whole, will be discussed in Chapter 1 of Part II together with a few related problems concerning the compilership of King Hala and the choice of a Prakrit dialect rather than Sanskrit. At this stage I have refrained from making a complete word-index of the Gathas edited. I have restricted myself, instead, to preparing an index of words and an index of grammatical points discussed in the Notes. The numbers of the Gathas refer to Weber's edition of 1881. However, when preceded by an asterisk it concerns a particular Gatha in the part of the text edited here. In the lists in the stemma (Part I) I refer throughout to Weber's above-mentioned edition.

Page #17
Notes to the Introduction The text is known by various names, e.g. Gathasaptasati, Saptasataka, Gathakosa and Sattasai (see Sternbach, 1974: 10). My reason for opting for the last one is in the first place practical: it is the name by which the text is discussed in the most recent book on the anthologygenre, namely the one by Sternbach referred to above. More importantly, it makes immediately clear that the text concerned is written in Prakrit. For a list of translations, see Sternbach (1974: 10-11). The oldest Prakrit grammar is very likely the Prakrtaprakasa by Vararuci. In its original form it dealt exclusively with the so-called Maharastri Prakrit (Nitti-Dolci, 1938: 8 et passim). Its main concern is to define this Prakrit dialect vis-a-vis Sanskrit. Thus, it gives rules concerning sound changes in order to be able to derive Prakrit from Sanskrit words, and lists of Prakrit words, Adesas, which apparently do not have Sanskrit derivations. In the grammar following Vararuci's, written by Hemacandra, there is a considerable increase in rules, often drawn up to account for only one particular word. However, the greatest increase is in the number of Adesas (for which see the fourth Pada). The two dictionaries seem to have been concerned with the same problem, including words which scholars at that time were as yet unable to derive from Sanskrit and as such were characteristic of Prakrit rather than Sanskrit. A case in point is the verb nivvara- mentioned in Hem. IV 92 as an Adesa for Sanskrit duhkham munc- 'to relieve oneself from one's grief'. The verb nivvara- means 'to become calm or satisfied' and is related to Sanskrit nirvrt- in the same way as uvvara- 'to overflow' is to udvrtApparently this meaning was no longer available to Hemac andra. ConsequentTy he relied for the meaning on the various contexts in which the verb occurred. The meaning 'to relieve oneself from one's grief' seems to

Page #18
have been based directly on its occurrence in Gatha 204 of the Sattasai. Another good example of the way Hemacandra worked concerns the way he deals with the word omaliam, discussed in the note on Gatha 32, which probably is merely a corruption for nomaliam (navamalikam) there.

Page #19

Page #20
PARTI

Page #21

Page #22
CHAPTER 1 1.1 Previous editions (introductory) The first edition of the Sattasai is the one by A. Weber (see 1.2). Other noteworthy editions are the so-called Nirnayasagar-edition (1889) and the one by A. N. Upadhye (1970). The former is an edition based on several MSS not known to Weber of the Vulgata-recension. The latter is an edition based on a single MS from Madras of the so-called Second Telinga-recension. Beside these there exist numerous other, mainly Indian, editions, which, however, do not bring forward any new material but are mere, more or less faithful, copies of the Nirnayasagar or Weber's edition." For the present edition only Weber's manuscript material has been used. The material of the two other editions mentioned above has not been taken into account as this would only have resulted in a considerable increase in the number of variants which considering the results of the stemma will play only a marginal role in the final reconstruction of the text. The material could, however, be enlarged by three so far unknown South-Indian MSS. In this chapter I will first give a description (in chronological order) of Weber's four publications which together constitute the 'first edition' (1:2). This is followed by descriptions of the MSS used by Weber (1.3). The latter descriptions have been kept as short as possible. For further information the literature quoted may be consulted, and especially Weber's publications. The order of the presentation of these MSS does not follow the chronology of Weber's successive publications, but is based on their contents; thus, those which contain the Prakrit text only come first, followed by those which contain the Prakrit text with a commentary, and finally those which contain a Sanskrit commentary or Chaya only. Next the three new South-Indian MSS will be described (Ch. 2). 1.2 The first edition by Weber The first edition, by A. Weber, consists of four successive publications, i.e. two text-editions proper (Abh. and Ed.) and two publications con

Page #23
10 taining variant readings and extracts from certain commentaries that were available at the time (Retr. and ISt.). I shall briefly describe each of Weber's publications in their chronological order. Abh. Weber's first publication on the subject, Ueber das Saptasatakam des Hala; ein Beitrag zur Kenntniss des Prakrits, AKM V 3, Leipzig, 1870 (repr. Nendeln, Lichtenstein, 1966) is based on a single MS, referred to as K. Abh. comprises a fragment of the text (1-370; one leaf containing the Gathas 94-102 is missing) with a German translation, extracts from Kulanatha's commentary and notes by Weber. The edition is preceded by a detailed grammar of the Prakrit of the Gathas (pp. 30-70). In an appendix (p. 202ff.) Weber has added in all 67 Gathas of very similar contents, gleaned from Dhanika's commentary on the Dasarupa (nos 1-8), from Mammata's Kavyaprakasa (nos 9-65) and from Visvanatha's Sahityadarpana (nos 66-67). In his edition of 1881 Weber referred to this book with the abbreviation Abh., from 'Abhandlung', which will be used here too." Retr. The above-mentioned edition was followed by an article called 'Zum Saptasatakam des Hala', ZDMG 28, 1874, pp. 345-436. In this article Weber compares several MSS acquired only after he had completed the Abh. The MSS are B. P. 14, D, and S (text and commentary) and A and C (commentary only). The former MSS provided variants to the text established in the Abh. Furthermore, extracts from the available commentaries are given: Gangadhara (A, B and C), Pitambara (P), Sadharanadeva (S) and anonymous (T"). Where necessary Weber adapted his earlier translations and added notes. The material from these MSS is given only for the Gathas they have in common with Abh. Their additional material appears only in the edition of 1881 (see below). In his edition of 1881 Weber referred to this article with the abbreviation Retr., for 'Retractatio'. This abbreviation is used here too. Ed. Weber's third publication on the subject is Das Saptasatakam des Haia,

Page #24
11 AKM VII 4, Leipzig, 1881 (repr. Nendeln, Lichtenstein, 1966). This is the first 'complete' edition of the Sattasas. However, readings, translations, extracts and certain observations of Weber, found in the earlier publications, have not been repeated here. The edition includes material from a number of MSS acquired only after the completion of the Retr., namely of y, Y, E, E, I, X, W and V. In the introduction Weber classifies the available MSS in altogether six recensions: 1) Vulgata: 4 and y (text), K, B and P (text and commentary) and A, C, E,'and (commentary only). 2) Jaina-recension: R (text) (and Bh (text and commentary), edited only in 1883; see below). 3) x-recension: x (commentary only). 4) Sadharanadeva-recension: S (text and commentary). 5) Telinga-recension: T (text) and TC (commentary only). 6) Second Telinga-recension: W (text and commentary) and V (Chaya only). This classification is solely based on features in the outward appearance of the MSS. Each recension appears to differ from the others, in the first place, in respect of the order of the Gathas. In this connection the special status of the Sadharanadeva and Telinga-recensions should be noted: in both the Gathas have been arranged topicwise in groups, called Vrajyas and Paddhatis respectively. Furthermore, each recension is characterized by the absence or inclusion of certain Gathas. In this connection the Second Telinga-recension stands out: it includes even in its longer form only 104 Gathas. Next Weber tries to determine which recension would represent the most original or authentic version of the Sattasai (pp. XLV-LI). In doing so he proceeded from the assumption that the most original recension would be the one which includes the greatest number of Gatas occurring in all the other recensions as well. This appears to be the Vulgata with 430 Gathas that also occur in the other recensions. In this work Weber has therefore edited in the first place the text of the so-called Vulgata, and in particular as it was commented upon by

Page #25
Gangadhara. This text was followed by the Gathas of the other recensions or MSS not found in the Vulgata and by 44 Gathas from several works on Alaskara. These Gathas are the same as the 67 found in Abh. (p. 202ff.), minus 23 which happened already to occur in one or more of the six recensions of the Sattasat. The Gathas have been translated into German from 366 onwards. For the translations of 1-365 one is referred to Abh. and Retr. Variant readings are given below the text. The variants of those MSS, or parts of MSS, already edited in Abh. and Retr. are not given here again.! Beside extracts from Gangadhara's commentary occasionally extracts from the other commentaries are given as well. Where he deemed it necessary Weber has added his own notes and comments. I shall refer to this work as Ed., which stands for Edition, the abbreviation Weber used in the publication that will be described below. ISt. Weber's fourth and last publication concerning the Sattasai is 'Ueber Bhuvanapala's Commentar zu Hala's Saptasatakam', Indische Studien 16, Leipzig, 1883 (repr. Hildesheim, 1973), pp. 1-204. It consists of the edition of the variants of MS Bh, compared to the text of Ed., together with extracts from Bhuvanapala's commentary and, occasionally, Weber's own notes. The text of MS Bh is a specimen of the so-called Jaina-recension. The abbreviation used here for this work, ISt., stands for Indische Studien. 1. 3 The MSS edited by Weber a. MSS containing the Prakrit text only: R, W, Y and T Berlin, Staatsbibliothek, MS or. quart. 740 (pp. 139-258). MS or. quart. 740 contains on pp. 139-258 the text of MS v. In 1876 this MS was compared by Weber with a MS from Bombay lent to him by G. Buhler. The variants are designated with the siglum R. The original was a paper MS of 23 folios containing the complete text (1-710), written in Jaina

Page #26
13 Nagart, of the so-called Jaina-recension. It is dated Samvat 1656 (1598 A.D.). The original MS was badly damaged inside. Lit.: Verzeichniss der Sanskrit- und Prakrit-Handschriften der koniglichen Bibliothek zu Berlin, by A. Weber, Band 2, Abtheilung 1, Berlin, 1886 (ABC 20, 2), no. 1593 b; Ed. pp. XXVIII-XXIX. Present whereabouts are unknown. W is a fairly modern, but undated, paper MS of 37 folios, arranged Europeanwise. The script used is Devanagari. It contains the complete text (1-709) of the so-called Vulgata. It was lent to Weber by G. Buhler and apparently sent back to the latter while he was still in India. Lit.: Ed., p. XXVIII. Berlin, Staatsbibliothek, MS or. quart. 740 (pp. 139-258). the Roman script prepared by Weber in 1875 of MS Wilson 420, Oxford, Bodleian Library. This original is a fairly modern, but undated, MS, written in Devanagari, of 24 folios, arranged Europeanwise, with 22 lines to a page, each line containing 25 Aksaras. It contains the complete text (1-709) of the so-called Vulgata. Lit.: Verzeichniss... (ABC 20, 2), no. 1593 b; Ed., p. XXVIII; Catalogus 'codicum manuscriptorum sanscriticorum postvedicorum quotquot in Bibliotheca Bodleiana adversantur, Vol. I 2, Oxford, 1864 (ABC 238, 1, 2), p. 381a. * Berlin, Staatsbibliothek, MS or. quart. 555. MS or. quart. 555 is a modern copy of 70 folios, arranged Europeanwise, written in the Telugu script, especially made for Weber, of the complete text of the Gathas of MS no. 10283 (in the Telugu script) of the Tanjore Palace Library. Above the lines are written the variants found in the Grantha MSS no. 10285 (for 1-400) and no. 10286 (for 400-700) from the same library. These variants are referred to with the siglum To (Weber's To; I have preferred to in order to avoid a possible confusion with T. To distinguish I from To the former is occasionally labelled ta). I con

Page #27
tains the complete text of the so-called Telinga-recension. For Weber's use a transcript in the Roman script was made of T together with TC (see below) by S. Goldschmidt, who also prepared a complete wordindex of the text: This transcript is kept in the same library, numbered MS or. quart. 706. Lit.: Verzeichniss...(ABC 20, 2), nos 1598 and 1600; Ed., p. XXX-XXXII; .. A classified Index to the Sanskrit MSS in the Palace at Tanjore, by A. C. Burnell, London, 1880 (ABC 306), p. 1740; Retr., p. 347. b. MSS containing the Prakrit text as well as a Sanskrit commentary: Bh, K, B, P, W and S. Bh Berlin, Staatsbibliothek, MS or. quart. 743. MS or. quart. 743, consisting of 122 pages, contains extracts made in 1881-82 by Weber from a MS sent to him by F. Kielhorn from Poona. This original is at present included in the collection of the Bhandarkar Oriental Research Institute in Poona, as no. 245 of 1880-81. This MS, which is undated, consisting of 71 folios, with 17 lines to a page, each line having 46 to 48 Aksaras, contains the text together with a Sanskrit commentary. The script is Jaina Nagart. The text is incomplete (1-600); the order of the Gathas is that of the so-called Jaina-recension. The commentary, styled Chekoktivicaraltla, is by Bhuvanapala. Parts of the text repeated in the commentary are designated with the siglum Bho. Where clear distinction is necessary the text proper is designated Bho. Lit.: Verzeichniss...(ABC 20, 2), no. 1597; ISt., p. 1ff.; Report on the Search for Sanskrit MSS in the Bombay Presidency during the years 188081, Bombay, 1881 (ABC 273, 1), no. 245 (p. 65). Berlin, Staatsbibliothek, MS or. quart. 740 (pp. 1-136). MS or. quart. 740 contains on pp. 1-136 a copy in the Roman script prepared by Weber of an original MS belonging to F. E. Hall. This original, of which the present whereabouts are unknown, consists of 48 folios with the text (1 to 6 lines) in the middle and the Sanskrit commentary (3 to 4 lines) above and below the text. Each line has 55 Aksaras. The script

Page #28
15 is Jaina Nagart. K contains only a fragment, namely of the Gathas 1-370 (folio 17 with 94-102 is missing). The text is a specimen of the so-called Vulgata. The Sanskrit commentary is by Kulanatha. Lit.: Verzeichniss... (ABC 20, 2), 1593 a; Ed., p. XXXIII; Abh., p. 1 ff. B . Present whereabouts are unknown. B is a MS from Bombay. Its form and the script used are not described. It contains a fragment of the text (1-129) with a Sanskrit commentary by Gangadhara (see also below, A, C and E). The text is a specimen of the so-called Vulgata. The MS was lent to Weber by G. Buhler. Weber, who erroneously considered it a copy especially made for him, entered it into the collection of the Staatsbibliothek in Berlin as MS or. quart. 557. Upon being notified of his error Weber sent it back to Buhler in Bombay. Lit.: Ed., p. XXXII; Retr., p. 346. P. Berlin, Staatsbibliothek, MS or. quart. 365. MS or quart. 365 is a collation of 108 pages of a fragment (1-300) of the text of the so-called Vulgata together with extracts of the Sanskrit commentary by Pftambara, prepared in 1872 by R. Pischel of no. 7219, India Office Library, London. This original consists of 63 folios of glazed paper, measuring 9 1/2 by 6 1/8 inches, rather carelessly written in the Kannada script. It is undated but probably goes back to the nineteenth century. The Prakrit text and the Sanskrit commentary are complete up to Gatha 151 (folio 47). From Gatha 229 (folio 59') there is either only the Prakrit text of the Gathas or only the Sanskrit translation. :: Lit.: Catalogue of the Sanskrit and Prakrit MSS in the Library of the India Office II 2, London, 1935 (ABC 164, 2, 1), no. 7219; Verzeichniss ...(ABC 20, 2), no. 1603; A descriptive Catalogue of the oriental MSS and other articles illustrative of the literature, history, statistics and antiquities of the South of India. Collected by the late Colin Mackenzie, by H. H. Wilson, Madras, 1882 (ABC 184), no. 107; Ed., pp. XXXIV-XXXV; Retr., p. 346.

Page #29
Berlin, Staatsbibliothek, MS or. quart. 742. MS or. quart. 742, consisting of 346 pages, is a copy prepared by Weber in 1872 of no. 1750, India Office Library, London. This original consists of 124 folios of grey paper, measuring 12 by 4 3/8 inches, having 11 lines to a page. The script is Devanagart. It is dated Samvat 1672 (1624 A.D.). S contains the complete Prakrit text (1-700) together with the Sanskrit commentary by Sadharanadeva. The text, which is called Muktavalt and in which the Gathas are arranged in Vrajyas, is a specimen of the so-called Sadharanadeva-recension. Lit.: Catalogue of the Sanskrit and Prakrit MSS in the Library of the India Office II 2, London, 1935 (ABC 164, 2, 1), no. 7218; Verzeichniss ...(ABC 20, 2), no. 1596; Ed., pp. XXXVIII-XLI; Retr., p. 347. Berlin, Staatsbibliothek, MS or. quart. 590 (pp.1-49). MS or. quart. 590 (pp. 1-49, pages arranged Europeanwise, 22 lines to a page, each line containing 21 Aksaras) is a modern copy in the Telugu script, especially prepared for Weber, of no. 10287 of the Tanjore Palace Library. This original, which is written in Grantha, contains the text ogether with an anonymous commentary, of the so-called Second Telingarecension. The first eleven Gathas of the original were illegible. For Weber's use of the MS or. quart. 590 a copy in the Roman script was prepared by S. Goldschmidt. The transcript is kept in the same library, numbered Ms or. quart. 707. Lit.: Verzeichniss... (ABC 20, 2), nos 1601 and 1602; Ed., p. XLII; A classified Index to the Sanskrit MSS in the Palace at Tanjore, by A. C. Burnell, London, 1880 (ABC 306), p. 175. c. MSS containing only a Sanskrit commentary or a Sanskrit Chaya: A, C, E, S, T, X, TC and V. Berlin, Staatsbibliothek, MS or. quart. 741. A is a copy of 207 pages in the Roman script prepared by Weber in 1872 of no. 944, India Office Library, London. This original consists of 96

Page #30
17 folios, measuring 12 1/4 by 3 5/8 inches with 8 to 11 lines to a page. The script is Jaina Nagarf. It contains the Sanskrit commentary by Gangadhara Bhatta (cf. B, C and E) on the text of the so-called Vulgata. Lit.: Catalogue of the Sanskrit and Prakrit MSS in the Library of the India Office II 2, London, 1935 (ABC 164, 2, 1), no. 7220; Verzeichniss ... (ABC 20, 2), no. 1594; Ed., p. XXXII; Retr., p. 346 (where the MS is designated with the siglum G(angadhara)). Berlin, Staatsbibliothek, MS or. fol. 1316. C is a modern copy prepared for Weber through R. Griffith, of a MS from Benares. It contains the Sanskrit commentary by Gangadhara (cf. B, A and E) on a fragment (301-600) of the Vulgata. Lit.: Verzeichniss... (ABC 20, 2), no. 1595; Ed., p. XXXIII; Retr., p. 346. E London, India Office Library, Buhler 329. E consists of 75 folios of European paper (watermark 1873), measuring 13 by 4 1/8 inches. The script is Devanagarf. It contains the Sanskrit commentary by Gangadhara Bhatta (cf. B, A and C) on the text of the socalled Vulgata. Lit.: Catalogue of the Sanskrit and Prakrit MSS in the Library of the India Office II 2, London, 1935 (ABC 164, 2, 1), no. 7221; Ed., p. XXXIII. London, India Office Library, Buhler 328. & consists of 32 folios, measuring 10 1/2 by 4 3/4 inches. The script is Devanagari. It probably belongs to the nineteenth century. It contains a fragment of an anonymous commentary on the Vulgata. The commentary is the same as the one found in MS 7 (see below). Lit.: Catalogue of the Sanskrit and Prakrit MSS in the Library of the India Office II 2, London, 1935 (ABC 164, 2, 1), no. 7222; Ed., pp. XXXV-XXXVI. TT London, India Office Library, Buhler 326.

Page #31
18 T is a modern copy in Devanagari on 137 folios of European paper (watermark 1869) bound in book-form. The pages are written on one side only. Each page has 16 to 23 lines. At the end a date is given, viz. Sahvat 1730 (1672 A.D.), which was copied from the original. It contains an anonymous Sanskrit commentary on the Vulgata. It is complete except for 1-20, 62-78 and 415-430. The commentary is the same as the one found in MS & (see above). In the margin of the first 29 folios extracts from Gangadhara's commentary (see B, A, C and E) are found, which in Ed. are designated with the siglum B (see Ed., p. XXXIII). Lit.: Catalogue of the Sanskrit and Prakrit MSS in the Library of the India Office II 2, London, 1935 (ABC 164, 2, 1), no. 7223; Ed., pp. XXXV-XXXVI. London, India Office Library, Buhler 327. x consists of 75 folios. The script is Devanagart. It is probably not older than the nineteenth century. It contains an anonymous commentary on 133-700 of a separate recension, of which it is at the same time the only specimen. Lit.: Catalogue of the Sanskrit and Prakrit MSS in the Library of the India Office II 2, London, 1935 (ABC 164, 2, 1), no. 7224; Ed., pp. XXXVII-XXXVIII. TC Berlin, Staatsbibliothek, MS or. quart. 556. MS or. quart. 556 is a modern copy in the Telugu script, consisting of 74 folios, arranged Europeanwise, with 21 lines to a page and 20 to 23 Aksaras to a line. This copy is based on two originals, namely no. 10283 (written in the Telugu script) and no. 10284 (written in Grantha) belonging to the Tanjore Palace Library. T contains an anonymous commentary on the text of the so-called Telinga-recension. Of MS or. quart. 556 a transcript was made in the Roman script, for which see above, under T. Lit.: Verzeichniss... (ABC 20, 2), nos 1599 and 1600; Ed., pp. XLI-XLII; A classified Index to the Sanskrit MSS in the Palace at Tanjore, by A.

Page #32
C. Burnell, London, 1880 (ABC 306), p. 1740; Retr., p. 347 (where to was designated with the siglum Ty). Berlin, Staatsbibliothek, MS or. quart. 590 (pp. 53-62). MS or. quart. 590 (pp. 53-62; pages arranged Europeanwise, with 22 lines to a page, each line having 21 Aksaras) is a modern copy in the Telugu script prepared for Weber of the Telugu MS no. 10289 of the Tanjore Palace Library. The latter MS is itself a transcript of the Grantha MS no. 10288. V contains the text of a Sanskrit Chaya to the Gathas of the so-called Second Telinga-recension. Of MS or quart. 590 (pp. 53-62) a transcript was made in the Roman script, for which see above, under W. Lit.: Verzeichniss... (ABC 20, 2), nos 1601 and 1602; Ed., p. XLII; A classified Index of the Sanskrit MSS in the Palace at Tanjore, by A. C. Burnell, London, 1880 (ABC 306), p. 174".

Page #33

Page #34
CHAPTER 2 2.1 The three new South Indian MSS: Ma, Ti and Tp The text of these three MSS forms a separate, seventh, recension of the Sattasaf. Like Weber's Telinga and Second Telinga, uhis recension is found only in South-Indian MSS. I have therefore decided to refer to it as the Third South-Indian recension. It is characterized by a distinct order of the Gathas; see Appendix I, in which a concordance is given of the text of the Third South-Indian recension with Weber's recensions or MSS, and Appendix II, in which a concordance is given of Weber's Ed.with the Third South-Indian recension. The numbers in the following descriptions of Ma, Ti and Tp refer to Appendix I. In none of the three MSS the Gathas are individually numbered. The text, though, is divided into Satakas. The order of the Gathas seems random, or, at least, the principle behind the arrangement, if any, has been left unspecified. In Ma and Ti the text is accompanied by the commentary of a certain * Madhavayajvamisra. Specimens will be found in the descriptions of the two MSS. A consistent feature in the commentary is the description of the Nayika and the specification of the poetical figure or figures present in the Gatha, but apparently there is no system in relation to the order of the Gathas. The commentary quotes extensively from Bhoja's sragaraprakasa and Sarasvatikanthabharana. For example, where below in the description of Ma the quotation of the 'beginning' breaks off, the commentary goes on with: uktanca bhojena: nayikascatasras, followed by a quotation from Bhoja's Srngaraprakasa, comprising p. 612, line 19 up to p. 648, line 5 (the examples given by Bhoja are omitted). After that Bhoja is quoted on the sixteen qualities of the Nayika, comprising Srngaraprakasa, p. 777, line 1 up to p. 779, line 6 (again, the examples are omitted). The commentary to this Gatha closes with a remark on a division of Nayikas as given in Bhoja's Sarasvat Tkanthabharana: yacca sarasvatikanthabharane trims adbhedatvakathanam ... In this case the quotations are found after the commentary proper, which otherwise usually closes mentioning the poetical figure. Generally they are found within the commentary. Moreover, the commentary frequently quotes the Kamasutra, for example, in connection with Gatha 12: ukta hi vatsyayan Tye

Page #35
22 dasa tu kamasya sthanani, etc. Madhavayajvamisra is most probably the same person who wrote a commentary on the Setubandha as well as on the Arthasastra; see Handiqui (1976: 89-92). The MSS of these commentaries, like those of his commentary on the Sattasaf, are restricted to, or hail from, Kerala. The Chaya found after the Prakrit text seems to have been added only later. It generally has the same lacunae as the text. Furthermore, where the text is lost completely, no Chaya is found, while in some cases at least the commentary has been preserved. Ma Madras, Government Oriental Manuscripts Library, no. 3378.. Ma is a Devanagarf transcript of an original palmleaf MS written in the Malaya] am script. It consists of two parts of which the pages have been numbered separately: 1-11 and 1-225. The main, second, part contains the Gathas from 11 up to and including 474 of the Third South-Indian recension. The text is accompanied by the Sanskrit commentary of Madhavayajvamisra. Beginning (p. 1): Srt, halas aptasatitatparyatika sanketasthanikrtasya saliksetrasya parinamat pratiksamadhavalibhavena lavanabhayad rudatim svairinti sakht samasvasayati kim ruyasi ona amuhT dhavalaantesu chalicchettesu. harialamandiamuht nadimva cchanavadia jaa. kim rodisi avanat amukhi dhavalayamanesu saliksetresu haritalamanditamukhi nad()iva sanavatika jata, iti. onaamuht avanatamukht, salicchettesu saliksetresu, haritalam manganan yasya taddharitalamandanam, tathabhuta mukham yasyah sa tatha, jaa jata, sanavatika haritalamanditamukht nad Tva jata, yatha nadT purvabhumikan tyaktva bhumikantaram parigrhyan yevavatisthate tadvacchyamavasthan tyaktva jhatiti puspitavastham gatan(!) sanketabhumitvanugunaparinamavatiti phalitarthah, parakTyasvap(/v?)aradh Tra(lacuna)pama

Page #36
23 End (p. 225): vidagdhagosthyam kayoscinmattakasinyos saundaryadina mithassamyaprastave kascid bhujangah kusumavyapadesena vaivarna vyanjayati: gandhena appano malaTe nomalia na cukkihisi(!) anno ko vi haasae masalo parima luggaro. gandhena atmano malatyam navamalika na soksyati anyah ko'pi hatasayah mamsalo parima lodgarah, iti. malaie jatau, navamalika saptala, ca nukkihii(!) na hiyate na hina bhavet, masalo mansalah, parimalasciranubandht gandhah vicitraratabhedopamardasahisnutvan tu dhvanyam, hatasaya ityatra malatya ityanusangah, vesyasuttameyam. In Roman characters the following particulars have been added: Transcribed in 1920-21 from a MS of Narayanan Nambur ipad (sic) of Kutalur Mana, Nareri, Tritala, Malabar Dt. Compared with the original MS. M. A. J., 19-4-1921. In the margin is written: 'ac(?) 4-11-21.' The Gathas have not been numbered separately. The text is divided into Satakas. The end of each Sataka has been indicated after the commentary on its final Gatha. It runs as follows: iti vadikaviculaman imahopadhyayasrfmanmadhavayajvamisraviracitaya halas aptasatTtatparyatTkayam prathama (dvitfyar, etc.) satakan sampurnam. In the margin of the transcript the scribe kept count of the leaves of the original which appears to have consisted of altogether 105 leaves. The leaves themselves, however, do not seem to have been numbered, as would appear from the fact that the counting of the scribe is consecutive also where the order of the leaves has been considerably disturbed. This is the case from 334 onwards. Whereas Gatha 334 is found on p. 157, the Gathas 335-341 are found on pp. 187-190, 342-400 are found on pp.

Page #37
24 161-186 (the commentary on 341 on p. 161.); 400-405 are found on pp. 157-160; 406-474 (end) on p. 190 ff. On p. 158 is found the indication of the end of the fourth Sataka, after 333 Gathas (by counting). That this state of the text is due to the disturbance of the leaves of the original -- which were not numbered -- appears from the fact that precisely opposite the respective breaking-points in the margin a number is found indicating the end of a leaf and the beginning of a new one. At the beginning of this main part a transcript has been added of eleven pages, the original of which must have been very badly damaged, for only part of the text could be recovered. It contains fragments of the text of Gathas 2-10 of the Third South-Indian recension, with parts of a commentary upon them. No part of Gatha 1 of the recension is found. According to a note it was not copied until 1923. No particulars regarding the original are given, but most likely it was the same MS which served as the basis for the main part of the transcript. Possibly copying of the beginning was postponed because of the difficulties encountered in deciphering the text. Lit.: A triennial Catalogue of MSS collected during the Triennium 1916/ 17-1918/19 for the Government Oriental Manuscripts Library, Vol. IV lb, Madras, 1928 (ABC 201, 4, 1b), no. 3378. Ti Trivandrum, University Manuscripts Library, no. 4763. Beginning (p. 2): Srt, halasaptasatftatparyatTka (lacuna)janasankule pi tuha dansanan ti(!) magganti ramne va bhamai dimght pava(!)raide anupinga, cha - sundarayuvanjanasankule pi tava darsanam vimarganti aranya iva (bhramyati omitted) drstir varakya anudvigneti. ramne va aranya iva, varaiae varakya, anupinga anudvigna aganitaparivadabhayetyarthah, utsavas amajadveva(?) kamayamanayor mithodarsanam iti prageva darsitam, parakTyas vadhire(?)pragalbheyam, upamalankarah.

Page #38
25 End (pp. 282-283): daravevirorujaalasu(!) lamliakesasu mauliachisu purusdirtsu kamo piasu samjauho vasai. This text is followed by the Chaya and a long commentary, which on p. 283 is concluded as follows: iti vadikaviculamaNG imahopadhyayasrimanmadhavayajvamisraviracitayas halasaptasatitatparyatTkayam saptaman satakam sampurnan. The title-page contains the following information (p. 1): srth, hala-saptasatt-tatparyatTka padi(:)kavicudamanimahopadhyayasrTmanmadhavayajvamisraviracita. dvitTyasatake ekanavatitamagatham arabhya saptasatakantam. lekhakah: (e)n anantakrsnasarma In Malaya]am the following particulars have been added: Owner of the MS: Ettumanur Kailasapurattu Pisaroti. Copying commissioned by: the same. Script (of the original): Malayalam. Material: palmleaf. Date of completion of the copy: August the 8th of 10 )0 (Kolam Era, probably 1914 A.D.). Signature of the scribe. Date of payment of the scribe (not filled in!) Ti is a Devanagari transcript of 283 pages of an original MS written in the Malayalam script. It contains the text of the Third South-Indian recension from 197 to the end. The text is accompanied by a Sanskrit commentary attributed to Madhavayajvamisra. The Gathas have not been numbered. The text is divided into Satakas. The conclusion of each sataka has been indicated after the commentary to its final Gatha: iti halasaptasatstatparyatikayas dvitTyan (trtTyam, etc.) sataka sampurnam.

Page #39
At the beginning of the MS a group of four Gathas is found which occur again as 374-377 (see also Tp). Gatha 279 is missing (cf. Tp). Of some Gathas the text is missing due to lacunae in the original (577, 595, 596, 606; cf. Tp). Lit.: Alphabetical Index of the Sanskrit MSS in the University Manuscripts Library, Trivandrum, Vol. I, Trivandrum, 1957 (ABC 316), no. 4763; A descriptive Catalogue of Sanskrit MSS in the Curator's Office Library, . Trivandrum, Vol. VI, Trivandrum, 1940 (ABC 312, 6), no. 1439. Tp Trivandrum, University Manuscripts Library, no. 4758. Beginning (1"): ganapaino namo, tam namaha ekkadasanam jo vaanek(!)ellavas aditthena sohai munalakabane (corr. into le)na via(!)dasanuggame (rest of Gatha lost through damage). End (36"): daravevirorujualasu luliakesasu mauliacchTsu purusairtsu kamo piasu sajjauho vasai. SrT ia. Tp is a palmleaf MS of 36 leaves measuring 14 by 2 1/4 inches, with 8 lines to a page and 50 Aksaras to a line. The leaves have been numbered with letters. The script is Malayalam. In this MS the Gathas have not been numbered. It contains the complete text of the Gathas of the socalled Third South-Indian recension (1-699; 272 and 694 are missing). The text is divided into seven Satakas. The end of each Sataka has been indicated as follows: srt, ia halakai(also kavi)viraianam (for ae) sattas atna (for Te) puthamam (vtam, tiam...chatthan; for the seventh Sataka, see above) saan sampunnar, srt.

Page #40
27 While writing the conclusion of the seventh Sataka the scribe was apparently interrupted and appears to have forgotten to complete it. In between 196 and 197 four Gathas are found which occur again as 374377 (see also Ti). Furthermore, for several Gathas lacunae have been noted (577, 595, 596, 606). The MS formerly belonged to srf Narayanan Govindan Namburi, Avl., Kumaramangalattu illam, Ganganasceri. Lit.: Alphabetical Index of the Sanskrit MSS in the University Manuscripts Library, Trivandrum, Vol. I, Trivandrum, 1957 (ABC 316), no. 4758; A descriptive Catalogue of Sanskrit MSS in the Curator's Office Library, Trivandrum, Vol. VI, Trivandrum, 1940 (ABC 312, 6), no. 1437a.

Page #41

Page #42
CHAPTER 3 3.1 The stemma, its scope, method and results The stemma includes only those MSS which contain the Prakrit texts of the Gathas, i.e. those ten MSS edited earlier by Weber described in . 3 (a. and b.), and the three new South-Indian MSS described in 2. As such it covers only six out of the seven now available recensions of the Sattasaf, the so-called x-recension being known through a Sanskrit commentary only. The position in the stemma of the remaining eight MSS described in 1. 3 (c.), which contain only a Sanskrit commentary or a Chaya, can be determined by comparing the order of the Gathas with that found in the first group of MSS. On the other hand, the state of the development of the Prakrit texts on the basis of which the commentaries and Chayas have been written remains in most cases obscure. For instance, on the basis of a gloss nihsarati it is difficult to determine whether the commented text actually read. (original) nii or (secondary) ntsarai (see Gatha 49). In the study of the stemmatical relations between these thirteen MSS an important role has been played by so-called innovations. As innovations should be considered those forms and constructions which can be shown to have been secondarily introduced into the text for certain original forms and constructions (retentions) which apparently were considered * unclear or corrupt, were misunderstood or had a graphical form that might give rise to confusion. Typical innovations are synonyms with an easily recognizable etymology. Examples are ntsarai for nti in 49, viltno for vir(/1)ao in 290 or visarai for pamhusai in 355." It should be noted that many of these innovations, apart from their prakritized forms, cor. respond literally to the glosses found in the commentaries and Chayas which in many MSS accompany the Prakrit text. Some seem to rest on a misunderstanding of the original text, such as martharam tumam vacca in 161 for mamtharam pie vacca and kisa aunnana me sattt (one phrase) in 384 for ktsa a, punnana me satti (two phrases). The innovation alimganaduhiaim for "rahiain in 485 seems due to a confusion of the Jaina Nagart Aksaras -ra- and -du-. Likewise, hoi dana(rahiassa) in 136 seems related to original bhoadana" through a spelling (b) ho(y)adana, with -ifor -(y)a-. On closer examination some innovations have come about

Page #43
30 through corruptions which by a mere coincidence did not disturb the sense. An example is padihasai in 15, for original padihaai, in which the --- seems to have been inserted due to the influence of the following word sahihi. Other innovations have been introduced in order to remove unfamiliar constructions from the text, e.g. in 176 for the genitives vaeriassa and raassa, constructed with bharia 'filled with, the instrumentals vaeriena and raena have been introduced. In all cases the secondarily introduced forms and constructions fit, or were made to fit, the metre, and, what is more important, make sense. As a result, in the transmission of the text innovations have proved to be very stable. This is generally not the case with accidental errors that have arisen through, for instance, transpositions of syllables or omissions. The results are forms and constructions which mostly do not fit the metre and do not make sense, so that at some stage they were emended or replaced by innovations. Innovations are important in two respects. The occurrence of one and the same innovation in two MSS against a retention in the other MSS separates the two former from the latter and indicates that both go back to the same source or that the one derives from the other. As such their function is comparable to that of the so-called separative and conjunctive errors. At the same time they give an indication of the authenticity of the texts of the MSS. A text which contains an innovation against a retention in all the other MSS is less authentic, i.e. farther removed from the supposed archetype, than the latter. The text in, a third MS which contains this same innovation and in addition one of its own, is still less authentic. On the basis of the innovations it can be concluded that the text of the Third South-Indian recension is in many respects the most authentic one available. The Jaina-recension and the Vulgata are characterized by a number of innovations against retentions in the Third South-Indian recension. The text of the Jaina-recension, at least as found in MS Bh, is, in turn, more authentic than that of the Vulgata, which in addition to the innovations referred to above has a number of innovations of its own (4. 1) It can be shown that these three recensions go back to one common source, the archetype, which developed into the Third South-Indian

Page #44
recension, on the one hand, and into the Vulgata, on the other, with the Jaina-recension representing an early offshoot of the Vulgata. The archetype was a North-Indian MS, most probably written in Jaina Nagart. The order of its Gathas must have been very similar to that of the present Jaina-recension and Vulgata (4. 2). The mutual relation between the three available MSS of the Third South-Indian recension can be described fairly accurately (4. 3). They are divided into two branches, represented by Ma and by Ti and Tp respectively, of which the former seems the more conservative one. The Vulgata MSS go back to one common source which at the same time was the source of the present MS Bh of the Jaina-recension (4. 4). Among the Vulgata MSS it is possible to discern two main branches, represented by MSS K and and B respectively. These two branches are found combined in a third, represented by the two South-Indian Vulgata MSS, Y and P (4. 5). MS R of the Jaina-recension probably originally was a MS derived from an immediate predecessor of Bh and subsequently was compared with, and revised with the help of, a Vulgata MS (4. 6). The remaining three recensions, the Sadharanadeva, Telinga and Second Telinga, represent eclectic traditions. Thus the Sadharanadeva-recension combines readings typical of the Jaina-recension with those typical of the Vulgata (5. 2). The Telinga-recension, which generally agrees closely with the Third South-Indian recension, includes readings typical of. the Sadharanadeva-recension, the Vulgata and MS R of the Jaina-recension (5. 3). The Second Telinga recension is markedly different from all other known recensions in that it consists only of approximately 100 Gathas. Its readings agree closely with the other two South-Indian recensions. Furthermore, it has readings in common with some of the other recensions. It should, however, not be ruled out that its compiler drew Gathas from, or compared the text with, quite different sources, namely Alankarasastras. A significantly large number of the Gathas also occur in such works, where they serve as examples. Furthermore, the readings of some Gathas agree more closely to those found in the Alankarasastras than to those found in the other recensions of the Sattasaf (5. 4). However, this does not mean that everything is entirely certain and clear. For instance, while the division of the Vulgata MSS referred to above has been based on the major pattern of the distribution of innova

Page #45
32 tions, it appears that practically each individual Vulgata MS shares at least a few innovations with every other Vulgata MS. Some have furthermore a few readings in common with some of the MSS of the other recensions. Because of this it is doubtful if, for instance, those retentions in v (and B) which separate it from the other Vulgata MSS have been regularly transmitted or have been entered into one of the sources of y after comparison with some other MS. Fortunately these uncertainties only seem to affect the mutual relation of the individual Vulgata MSS and not the position in the stemma of the Vulgata as a whole.

Page #46
CHAPTER 4 4.1 The Third South-Indian, Jaina and Vulgata-recensions (introductory) The Third South-Indian recension is known through three MSS (Ma, Ti and Tp), the Jaina-recension through two (Bh and R) and the Vulgata through five (K, V, B, Y and P). Each recension differs from the other two in the order of its Gathas and in the inclusion or omission of certain Gathas. In all three, however, the order of the Gathas seems random, showing no internal organization. Each recension is furthermore distinguished from the others by a set of specific readings. From a study of these readings and in particular of the innovations, it appears that the text of the Jaina-recension together with that of the Vulgata is characterized by a certain set of innovations against retentions in the Third South-Indian recension (see under a. below); that the text of MS R of : the Jaina-recension together with that of the Vulgata is characterized by another set of innovations against retentions in the text of the Third South-Indian recension together with that of MS Bh of the Jaina-recension (see under b. below); and, finally, that the text of the Vulgata is characterized by yet another set of innovations against retentions in the two other recensions (see under c. below). .. These facts point to the following -- provisional -- conclusions, namely (1) that the Jaina-recension together with the Vulgata derives from the Third South-Indian recension or its source, (2) that R of the Jaina-recension together with the Vulgata derives from Bh or from its source, and (3) that the Vulgata derives from R or from its source. At this point conclusions 2 and 3 should already be modified. Below it will be shown that R contains a contaminated text, combining readings of a MS closely related to Bh with those of a Vulgata MS (4.6). In this paragraph as well as in the following the Jaina-recension and the Vulgata are considered monolithic wholes, with R either included in the former or in the latter. Conclusions 2 and 3 can consequently be telescoped into one, namely that the Vulgata derives from the Jaina-recension or from its source. In the lists given below minor mistakes or spelling variants found

Page #47
34 in the individual MSS are not always specified. Occasionally a 'retention' had to be reconstructed. Where this is the case first this reconstructed form is given followed by the forms actually found, between brackets. In most instances in the three lists given below the decision as to which is the innovation and which is the retention should be clear. Nevertheless, in a number of cases an explanation seems to be required, which, however, would not infrequently involve, for instance, complete word-studies or the consideration of parallel situations. It will be obvious that a discussion of such type can hardly be taken up in the context of the stemma as it rather belongs to the edition of the text. However, as the present edition contains only the first 50 Gatas no more than a relatively small number of instances could be discussed; in these instances a reference is given to the Gathas concerned. Some instances, though, are taken up later in the stemma itself, in which case a reference is given to the page in question. For the remainder I have had to restrict myself to comments on those cases which could be explained by relatively short remarks. For each list these comments, where available, are brought together in one note. Otherwise I have to refer the reader to the edition of the complete text which is planned to follow. a. Innovations in the Jaina-recension (Bh and R) and the Vulgata (K, V, B, Y and P) against retentions in the Third South Indian recension (Ma, Ti and Tp). Where the Vulgata MSS K, B and P, and Ma and Ti of the Third South-Indian recension, which consist only of fragments, and the Jaina MS Bh, which misses the seventh Sataka, are not available, this has not been expressly mentioned, but should become clear from their absence among the MSS enumerated. I have, on the other hand, included references to the 'eclectic' Telinga (T), Sadharanadeva (S) and Second Telinga-recensions (w), where these are available. For the latter three recensions, see Chapter 5. It is to be noted that T mostly follows the Third South-Indian recension. It should furthermore be noted that the Jaina-recension and the Vulgata differ occasionally in respect of the form the innovation in a

Page #48
35 particular instance takes. The following list, though, is in the first place meant to show that the Jaina-recension and the Vulgata forma distinct group by the mere fact that in some cases they have readings which are demonstrably not original, where the Third South-Indian recension has readings which are demonstrably original. The differences in the forms of the innovations in the Jaina-recension and the Vulgata should belong to the discussion of the relation between these two recensions. .. 15 padihasai Bh, R, K (padiha), , B, Y, P (padiasahi), S against padihaai (cf. T) (see Part II, Gatha 15) 26 suhasutta R, K, B, Y, P, S, pasutta Bh against sahasutta (cf. ) (see Part II, Gatha 26) 54 sari Bh, R, K, 4, B, P, T, sarisa y, s against siri? 105 dajjihisi Bh, R, K, V, B, Y, P, T, S against bhajjihisi 136 hoi daNGarahiassa Bh, R, S, hoi kivinapurisassa K, W, Y, P against bhoadanarahiassa (cf. T: danahoarahiassa) 196 'phuddilliyae Bho, phudilliyai Bho, papphullide R, upphalliai y, uppulidi Y, upphulliai P, daphulliai S against urdaliae (or uddalie) (cf. T: maridulliae) 12 254 oniddhae (onivvae Bho) vasio Bh, unnae vasio R, ulindae vasio K, oladdhie vasio y, volimdae vasio y, gharalimdae sutto T, animae saio s against olambae vasium (cf. W: vasaim; Ma olappie, Ti, Tp olabie) 283 nitthamu Bh, R, K, T, nitthanu , Y, S against niadhamu 316 suhaa Bh, Ry, Y, S against ahava (cf. T and K!) 318 pabbhattha Bh, R, K, , Y, S against pamhuttha (cf. T) (pammut tha Ma, Ti and Tp) 335 jaha na amhe (Bh: amha) musijjamo Bh, R, na jaha amhe musijjamo : K, Y, S, jaha ahme vi na hu musijjamo T against jaha ahme na amus ijjamo (cf. W: na amusajjamo) 360 bharunnayahin R, bharona ahin K, bharonai 4, bharonaahi Y, bharubbharahi T, bharonain S against bharaharani (see Part II, p. 144) *361 amharis Tna Bh, R, amharisThi U, Y, T, assadisthim S against ahmarisena :

Page #49
36 visarai Bh, R, S, vtsario X, Y, visumario T against pahmusai 370 anadharayan tams avalasta Bha, anisamisatasavalata Bh", animisam isantasavalanta R, animisam Tstsi valarta , animisatirichadaravalia y, animisas Tsuvvattanakhalaihta T, animisa valata S against anisa (varnka) valata (see below, p. 50) 384 ktsa au nana me sattt Bh, R, , Y, S against kisa a punnana me satt (cf. T) 386 nesa gahalanghia Bh, R (1 avia), T (1amjhia), na ima gahalanghia U, Y, S against esa gahalambia 410 visakamd (/d)ali vya muddha (vaddhatT) Bh, R, gamanidhua visa kafidali bva (vaththia) , vva(ththia) Y, gamanidhia visalaa vva (vadia) S against gamaNGidhuavisalaa (vaddhartT) (cf. T) 420 paravajjanaccir Thi Bh, R, X, Y, S (degkajja") against paravasana naccirthin (cf. T) 431 taha tassa X, Y, S tassa taha T against tastassa 437 atthakkaulavTs athbhaghai vlocai), Y, S, atthakkagaavisambhagaaa T against atthakkagaavesam (bhana) 460 duttha Bh, R, W, Y, S against dhavala (cf. T)12 466 maacchi Bh, R, W, Y, S against maaccha (cf. T) 483 aigaruena vi tammai (Bh khijjai) hiaan bharena gabbhassa Bh, R, aigaruena vi tammai hiae (s hiaammi) bharena gabbhassa (S gavbhabharena) X, Y, T, S against tamei hiaan ahian bhagga bharena gabbhassa (bhaggam Ma, Ti and Tp) 485 alingan aduhiais Bh, R, W, Y, S against alinganarahiain (cf. T) 487 me hiah tam pi Bh, R, me haan ta pi U, se haan tammi Y, se hia tam pi T, S (5 tammi) against na sahiah tte 514 khijjiavvain Bh, R, V, Y, S, lajjiavvai T against rusiavvai 541 jalolle Bh, R, X, Y, S and Ma! against jalolla (cf. T) 548 taggaamano W, Y, T, S against viahbh amano 553 avanehi Bh, nam nehi R, na nehi W; na nesi y against ninneha 575 marid ira Bh, R, K, M, Y, S against simdi(/a)ra (cf. T: sand ira) 626 muhapundarta R, X, Y, T, S against suhmarntaria 680 pavalaankurayavaNGNGalohillao R, pavalaankuraavannalohitae , paravala ankuravannalohitae Y, pavvalamkur avannalohitae s against pavalaakuralohie niddhe (cf. T) niddhoyadhauras tti kisa hatthe R, dha(t)urae ktsa sahatthe

Page #50
37 V, Y, dhaurae vva kisa hatthes against piddhoadhaurae vi kTsa hatthe (cf. T) 692 ahilijjai pauma lehadalivalaena kamalagovie R, alihijjai pamkaale halalibalaena kamalagovte y, ahilijjai pankaale halatlavalaena kalamagovte Y, ahilijjfai pankaaluhialivalaena kalamao s against ahilijjai kamalullehadalivalaehi kalamagovte (cf. T) 711 volimto Bh, s (otto), volanto R, Y against bolei (cf. T) Innovations in MS R of the Jaina-recension together with the Vulgata (K, V, B, Y and P) against retentions in the Third South-Indian recension together with Bh of the Jaina-recension. 3 majjhaarammi R, K, W, B, Y, P, S against majjhaarao (cf. T) 8 mamdan Thuam R, K, V, B, Y, P, T, S against mamd aNGam bhuam (maidanaabhudam Tp, madanabbhuya Bh (see Part II, Gatha 8) 49 nisarai R, K, V, B, Y, P against nii (cf. T, S and W) 78 pailla R, painna K, W. B, Y, P against pahulla (cf. T) (pa(p) phulla Ti, Tp, pahalli Bh, paala S) (see below, p. 49) 96 parigaa R, W, B, Y, P, T, S, Ma! against pariar as 97 dakkhina R, W, B, Y, P, S against dahina (cf. T) 106 naanesu R, K, W, B, Y, P, T, S and Tp! against vaanesu (vaanenasu Ti) 119 ummatta R, K, W, B, Y, P, S against ummaia 129 anunta R, K, X, Y, P, S suvinto T against asunte! 152 pasaalohilla, aigahiagarua K, X, Y, P, T, S (in R the relevant hemistich is missing) against pasaalohille, aigaruagahia 156 suharasa R, K, X, Y, P against sarahasa (cf. T and s) (saharasa Bh) (see below, p. 49) 163 jivijjai R, K, V, Y, P, T, S against valijjai Tp, Bho (vilajjai Ma, Bhd 165 mamgalan 1 amgale R (se hale), K, V, Y, P, S against lagale mamgalan (mangale mangalam Bh) 170 palottanta R, K, Y, P, T, S against palotta (cf. v) ; '176 vaeriena, "raena R, K, , Y, P, S against vaeriassa, "raassa phukkamto R, V, Y, P, antaa K, S against pumertaa Ma, Tp

Page #51
38 (phumitti Bh) 184 ujjua nahan R, W, Y, anua naha K, s (anua), a paha P against annaman aham (unnamanahan Bh)3 188 pautthamanan R, K, V, Y, P, T, S against maNGapauttham (see Part II, Gatha 31) 290 vilino R, K, U, Y, T, S against vill/r)ao 315 gayavainan R, gaavaian K, , T, S, pautthavaia y against gaapatam 320 vi na taha R, K, X, Y, T, S and Tp! against vi na vi taha 325 anu(raimayan, oraaan) R, K, U, Y, S against ana(ramaan) (cf. T: ara in rame arti) 353 vi na taha R (na taha), K, V, Y, T, S against vi na vi taha 368 thana(k)alasu(ppellana) R, T, S, thanaalamuhay, thanajuala muha y against thanabharu 526 anikkarta R, W, Y, S against anakkamta (cf. T: NGakkamta) 532 khokkha R, W, Y, S against p(h)upp(h)a (cf. T and W) samullihai R, W, y against samalliai (cf. T: samullahai ai, W: savalliai and S s amulliai) 575 anandia R, W, Y, T, S against anardira . C. Innovations in the Vulgata (K, 4, B, Y and P). against retentions in the Third South-Indian and Jaina-recensions. Of the five Vulgata MSS only u and y are complete. Bis only a small fragment and is generally not available below. P contains only the first three satakas and K the first three and more than half of the fourth. Below I have therefore made two extra categories, namely one where K is not available and another where P is not available. 22 litta K, Y, P and Tp (pr. m.), luppa y, B against tuppa (Tp sec. m.) (see Part II, Gatha 22) 61 suhaveti K, W, B, Y, P, T, S and Ma! against suhaati (sam appaht i Bh) 65 anudiahan K, W, B, Y, P against ta diaharh 104 samuccinai K, U, B, Y, P, S against samuccei (cf. T)

Page #52
39 114 goviana K, W, B, Y, P, S against goviae (cf. T)* 160 mamtharam tumarh vacca K, X, Y, P, S against mastharath pie vacca (cf. T)(see Part II, the Note to Gatha 20) 203 vahugunavasto jano tassa K, M, Y, P against vahuguno va anno jaNGo tassa (cf. T and S) 219 na naa K, na gaa V, Y, P, S against hmi gaa Ma, Tp (patta Bh, etta R, gamia T) 220 vedhana K, U, Y, P, S against vedhaa (cf. T) 226 suttassa K, 4, Y, P, s against sutte (cf. T) Innovations in , B, Y and P where K is not available 95 pabbhasio V, B, Y, pakkhasakht P (cf. S: pabbhattho) against pahmusio Ma, Tp (cf. T) (pamhusai Bh, paccusio R) : Innovations in K, , B and y where p is not available 210 L. 240 andha(k)ara K, , adharaa r (cf. S: ardhavera) against arhdhalaal4 4.2. The archetype The three recensions are related in the following way: X Vulgata Jaina Third South- Indian

Page #53
40 The two other possibilities can be excluded: Third South-Indian Vulgata Jaina because the Third South-Indian recension has innovations of its own against retentions in the Jaina-recension together with the Vulgata (see under a. below); and Jaina Third South-Indian Vulgata because the Jaina-recension has innovations of its own against retentions in the Vulgata (see under b. below). The two branches, represented by the Third South-Indian recension, on the one hand, and by the Jaina-recension and the Vulgata, on the other, ultimately go back to one single MS, the archetype X. Its existence is proved by certain errors common to the MSS of the Third South-Indian recension, on the one hand, and MS Bh of the Jaina-recension, on the other (see under c. below). The MS of the archetype was written in a North-Indian script, most likely a kind of Jaina Nagari, as appears from the 'error', found in all available MSS, annavaraha for anavarana in 488." The supposed corruption of an- into ann- points to a source in Jaina Nagart, in which the Aksaras concerned are hardly distinguishable. Furthermore, the MSS of the Third South-Indian recension, which go back more or less directly to the MS of the archetype, have certain orthographical peculiarities which positively point to a source written in Jaina Nagart (see under d. below). Admittedly, this source is probably

Page #54
not the same as the MS of the archetype. This would appear from the innovation paada in Ma and Tp (and T) in Gatha 2 (see Part II) for paua in all the other MSS. Its form shows that it was most probably ad hoc created on the basis of a rule typical of AMg. This seems to suggest that it was introduced in North-India rather than South-India. On the other hand, the word paada is common in other south-Indian Prakrit texts as well, and in South-Indian MSS of Prakrit texts, so that it may equally well have been introduced in these MSS only at a relatively late stage in South-India. As already noted above (4.1) all three recensions have a different Gatha-order. The order of the archetype appears to be most faithfully preserved in the branch represented by the Vulgata, on the one hand, and the Jaina-recension, on the other. In order to make this clear in Appendix III a concordance is given of the Third South-Indian recension (first column), the Vulgata (broken down in hundreds in the following seven columns) and the Jaina-recension (R; last column). It appears that the Third South-Indian recension is based on a selection from a MS the order of the Gathas of which closely resembles that of either the Vulgata or the Jaina-recension, sometimes agreeing more with the Vulgata, Sometimes with the Jaina-recension. It can be observed that the compiler went through this MS from beginning to end picking out individual Gathas or whole strings of them, while rejecting others. The last Gatha selected in this first round is 669 of the Vulgata or, alternatively, 708 of the Jaina-recension (see opposite 425). After that he went through the MS again picking out many of the Gathas rejected first. These were apparently added following the text which resulted from the first round of selection, i.e. after 425. Occasionally, though, some of them were inserted in this first part of the text. While the Vulgata and the Jaina-recension preserve the order of the archetype they do so only in broad outline. For it appears that in each recension large parts of the text of the archetype have come to be displaced. For instance, in the Jaina-recension (R) the sequence 131-187 of the Vulgata is found in the fourth Sataka between 337-391. A comparison with the Third South-Indian recension shows that the text of the Vulgata is original here (see opposite 135-141 and 172-176). Likewise in the Vulgata whole sequences of Gathas have come to be displaced (see

Page #55
opposite 321-2, 333-4, 355-6 and 194-5). These changes are most likely due to the disturbance of the order of the leaves of a MS. Besides, minor changes in the order and position of the Gathas have taken place. Note in this connection the Gathas 188, 189, 190 and 194 of the Vulgata; from their position in the Third South Indian recension as 69, 74, 84 and 70 respectively it appears that the Jaina-recension has preserved their original position here in the first Sataka. On the other hand, the Vulgata has preserved the original order of 332, 333 and 334, which occur as 255, 256 and 257 in the Third South-Indian recension but as 186, 110 (sic) and 287 in the Jaina-recension. It may be noted that the Vulgata has apparently suffered from considerably fewer of these minor changes in the order and position of the Gathas vis-a-vis the archetype than the Jaina-recension. A clear exception is its final Sataka in which the internal order must at some stage have become disturbed considerably (see opposite 10-53). Note in this connection that the first round of selection ended with 669 of the Vulgata, i.e. still rather far removed from the end, and with 708 of the Jaina-recension, i.e. practically at the end. The latter seems more logical." a. Innovations in the Third South-Indian recension (only part 197-474, for which all the three MSS, Ma, Ti and Tp, are available, has been taken into account): 476 viakkhanah (vi aamaih), 669 nisajjai (nimajjai), nisajjasu (cf. T) (pimajjihisi), 724 ukkhippei (cf. T) (utt(n) amghei; ucchappei Bh, utthepei R, utthaggai s). Innovations in the Jaina-recension: 71 na jano piaamakaamanabhamgadukkhakkhamo savvo, 156 hiyayaim, 183 cukkai, 206 taha cci(y)a, 211 hemistiches reversed, 302 va(y) anesu, 385 varis am va, kappam va, 460 sayalagottha, 465 pasanna, 470 atthi Bh, achi R, 505 nayan addhabhallena, 554 tam mama (maha) cc(h)ari(y) an, 715 hatthe.

Page #56
C. Common errors in the MSS of the Third South-Indian recension, Ma, Ti and Tp, and MS Bh of the Jaina-recension: d. 43 49 sartsa Ma, sirTsa Bh (cf. sisira T), for sartra; 167 dharigim Ma, Bh, for dharani; 212 aquhohtaam Ma, Bh, for agahontaan; 292 tanui(d)a (emendation:) Ti, Tp, tanuyaimh Bh (cf. taNGuai and S), for tapuaai; 300 karipim Ti, Bh, for karani; 367 garu(y)ana Ma, Bh, for guruana; 371 bbharata Ma, bbharaftara Bh, for bbhamtara; 379 bhuppaht Te Ma, bhunnamt T(a) Bh, for bhadafitTe; 428 kapphlava Ma, kapalava Bh, for kavvalava; 466 pahia Ma, pasiha Bh, for.pasia; 470 daddham Ma, Bha, for datthum; 487 pavvasa Tp, Bh, for pavasa. Note the omission in both Ma and Bh of kim in 65 peusamattapaddi kim suasi Orthographical peculiarities in the MSS of the Third South-Indian recension, Ma, Ti and Tp, which show that they ultimately derive from a MS written in Jaina NagarT: " 1) -mt-for-nn- (in Jaina MSS -nn-): mamtu for mannu in 260 (Ti, Tp), 683 (Ma, Tp), 920 (T1, Tp), 938 (Ma, Tp) and 941 (Ti, Tp). 2) ru- for -tu- or -tu- for -ru-: 189 ruha (Ma) for tuha (cf. R and K; note Weber's hesitation regarding the reading of Bh; Ist., p. 41), 251 tuvah (Ma) for ravah (sec. m.) (cf. K), 476 rusai (Ti) for tusai, 609 rura (Ma) for tura. 4.3 The relation between Ma, Ti and Tp of the Third SouthIndian recension Of these MSS Tp alone is complete (1-699). Ma contains only Gathas 2-474 and Ti 197-699. This means that only for 197-474 all the three MSS are available. Tp is written in the Malayalam script. Both Ma and Ti are Devanagari transcripts of Malayalam originals. All three MSS have orthographical peculiarities which point to an original in Grantha. Thus, one finds

Page #57
44 -S- for -- in 257 eraso (eram)do) in Ti, and -- for -s- in 70 pa(v)uda (pausa) in Ma, Tp, 444 Tdam (is am) in Ma, 445 pauda in Ti, Tp, 538 pauda in Ma, and 578 pa(v)uda in Ti and Tp. In the preceding paragraph I have shown that they also show traces of having been copied from a MS written in Jaina Nagart. This probably explains the occasional occurrence in Ma and Ti of the Ya-sruti which is otherwise typical of Jaina (Nagart) MSS. E.g. 2 payada Ma, 5 mayulanti Ma, 9 ruyas i Ma and 348 tanuaraya Ti. However, dhaayi in 130 in Ti (dhavai in Tp), which is an innovation in this particular branch of the Third South-Indian recension (see below) for dhuai found in Ma and the North-Indian MSS, sug- . gests that the -y- may also have been inserted in these MSS independently. The three MSS have one common source. This conclusion is based on the following two types of evidence: a. The order of the Gathas. I have shown in 4.2 above that the present recension is based on a selection from a MS which was the source of the Vulgata, on the one hand, and of the Jaina-recension, on the other. It is doubtful if this process of selection could have taken place three times independently, each time with exactly the same result. b. Common errors (only Gathas 197-474, for which all the three MSS are available, are taken into account): 254 olappie Ma, olahbie Ti, Tp (olambae), 341 tte vi taha (tte vi), 354 parunnena ajja (degenan pi ajja), 367 avilambirte (avao), 373 nihai (nihanai), 395 dararuggaa (daruggaa), 407 aumalana (omarana), 409 pahnu(/a)sai, 462 pahnu(/a)sai (panhuai), 496 souna vi (souna), 524 kuviassa paae in Ma, kuviam pasaaeuth Ti, kuvia pasaaeuth Tp * (kuvian ca pasdeun), 572 qahidattho (gahiattho), 643 ujja (ujja pamaro, unmetrical), 924 na atthirammi na (athirammi na), 932 ehi (ei). The relation between Ma, Ti and Tp is as follows. Tp and Ma each have innovations of their own (see under a. and b. below), whereas -Ti has none.

Page #58
45 This excludes the possibility that Tp goes back directly to Ma or vice versa. However, it leaves open the possibility that Tp as well as Ma go go back directly to Ti. Tp, though, shares a number of innovations with Ti (see under c. below) and none with Ma. Ti shares one innovation with Ma, namely as ams ia for asamghia in 707. This form asamsia may, however, have been introduced independently in the two MSS. In this connection it should be noted that in the commentaries which in Ma and Ti accompany the text, asanghia is glossed with asassita. 18 This evidence leads to the conclusion that the first division was between Ma, on the one hand, and Ti and Tp, on the other. a. Tp has the following innovations (only 197-474): 240 pucchahito (cheppahisto), 243 sohai (cf. T) (chajjai), 497 varei mam (cf. T) (vareha nam), 551 unnamia (uttania), 607 siae (sahuli). b Ma has the following innovation: 588 vilakkhena (vilavena). c. Ti and Tp share the following innovations: 130 dhaayi Ti, dhavai Tp (dhuai), 246 halidda (haliddf), 312 gahiuna (ghettuna), 313 nikkudaakola (Tp sec.m. padohar arkola), 325 na aramaan (Ma anaa(!)ramaar) (anaramaam), 656 loluva (lehada), 843 digghohna (d Thu ha), 848 saamdhiae, 849 saadhie (cf. T) (saajjhide), samalam Tp)muham (sama muham), 869 rudiuna (rottunan). The relation between Ti and Tp still needs to be determined. In this connection some significant similarities in the outward appearance of both MSS may be considered. In the first place, in both Gatha 279 (see Appendix I) is missing. Furthermore, Ti and Tp have a number of identical

Page #59
46 lacunae (290 (Ed. 715), 411 (626), 577 (-), 595 (-), 596 (-), 606 (863), 632 (153), 647 (269), 654 (905) and 690 (530)). The extent of the lacunae is, however, not always the same. In Tp the lacunae of 411, 595, 596 and 632 cover the complete texts of the respective Gathas, while in Ti still one or two words have been preserved. The Malayalam original from which Ti was copied consisted of a bundle of palmleaves containing the fragment 197-699. To this bundle was added a stray leaf with the Gathas Ed. 492, 481, 482 and 496 together with Madhavayajvamisra's commentary. In Ti these same four Gathas, with minor variants, are found again as 374-377 (373-376 by counting), embedded be- . tween Ed. 489 and Ed. 488. Ed. 489, however, occurs a second time, namely after Ed. 488. From this it would appear that this group of four Gatha's was inserted here later. In Tp, though containing the complete text (1-699), exactly the same situation is found. Thus, after 196 some space is left blank; then follow Ed. 492, 481, 482 and 496, which are followed by a blank space again, after which the text continues with 197. The same four Gathas occur again in the fourth Sataka: Third S-I recension 371 372 373 374 375 376 377 378 Ma Ed. 487 488 489 (492) 481 Ti/Tp Ed. 487 489 (492) 482 (496 503 482 (496) 488 489 503 The addition of the stray leaf containing Ed. 492-496 before 197 is explicable only in the case of Ti, i.e. its palmleaf original. It then follows that Tp is a collation of Ti, or any of its sources, and some other MS. It remains to determine the extent of the text of this latter MS and its relation to either Ma or Ti, if any. As has been noted above, Tp has the same lacunae as Ti, though

Page #60
47 occasionally the latter has preserved a few words more of the lost text. This would either mean that in preparing Tp the second MS used beside Ti (i.e. its source) also had those same lacunae; or that the second MS, like Ti, was itself a fragment, consisting of precisely the part 1-196 broken off from Ti. The latter assumption, that the second MS con cained only the fragment 1-196, is refuted by the following instances, drawn from Gathas 197-699, in which the scribe of Tp 'corrects' a reading in accordance with Ti: 333 Ti rakkhido, Tp (pr. m.) rakkhio, (sec. m.) rakkhido, Ti sas arkiru Tp (pr. m.) sasamkiu, (sec. m.-) sasank iru, 362 Ti tavijjai, Tp (pr. m.) tanuijjai, (sec. m.) tanuvijjai, 378 Ti anusikkhirie vi, Tp (pr. m.) anus ikkharie vi, (sec. m.) anusikkhirie vi. Probably in these cases the scribe started copying from the second MS and subsequently compared the text with that of Ti. On the basis of this evidence I assume the second to have been a complete MS. The lacunae which Ti and Tp have in common were apparently also found in the second MS used by the scribe of Tp, for otherwise they would almost certainly have been filled in. In any case, the assumption that the scribe of Tp would on purpose have refrained from fiiling in the lacunae found in his first MS is contradicted by his behaviour regarding some smaller omissions. The latter have occasionally been filled in, either immediately or secondarily, i.e. above or below the line. E.g. 23 Ti aha sa cci (the text breaks off and is immediately followed by Madhavayajvamisra's commentary), Tp aha se tti visas imo. This instance possibly shows that the scribe made use of the commentary, which glosses the last word of the lost part of the text, saddahimo, with visvas imah; 460 Ti uva, Tp id., above the line the syllable ha (uvaha) has been added; 522 Ti suraarambho (...)navo, Tp suraarambho uno(!) navo; 688 Ti cadu(...)niunana, Tp id., above the line the word kamma has been added. Occasionally, furthermore, the scribe of Tp made small corrections. E.g. 333 Ti raian, Tp id., corrected into ramiah, 452 Ti manan, Tp mal: deleted)namah, 873 Ti tae, Tp tte, changed into taha (cf. T), 923 Ti bahala, Tp bahala, corrected into bala. It follows that the second MS used by the scribe of Tp was a complete M$. As it apparently had the same lacunae as Ti, it may be concluded that it went back to an immediate source of the present MS Ti.

Page #61
48 The relation between Ma, Ti and Tp may be represented in the following diagram. It should be noted that the position of Tp to the left of Ti is arbitrary; it does not imply that Tp is closer to the source of the three MSS than Ti. 1-699 1-699 1-196 197-699 1-699 Ed. 492-496 197-699 Ma Tr The third South-Indian recension seems to have existed in relative isolation. As will be shown below the Telinga and Second Telinga-recensions have both drawn their Gathas from, among others, this recension, but these seem basically to be instances of one-way traffic only. An exception, however, is Tp, which shows traces of having been contaminated by T. With the latter Tp shares the following innovations: 20 desu suhaa majjha oas a Tp, suhaa maha desu oasa T (de suhaa majjha oas am), 49 lagga (cf. Ma lakka) (lukka), 243 sohai (chajjai), 497 varei mai (cf. , Y, S) (vareha nan), 626 khana (chana), 677 rosanan (rusanan). From the fact that at least one of these innovations concerns a reading

Page #62
49 otherwise typical of the Vulgata, namely 497 varei mam, and from the fact that of the Third South-Indian recension and T, only the latter also otherwise includes readings typical of the Vulgata (see 5.3), it would follow that Tp was influenced by T, and not vice versa. 4.4 The relation between the Jaina-recension and the Vulgata In 4. 1 and 2 I have indicated that the Vulgata derives from the same source as the Jaina-recension. The Jaina-recension, which is represented here by two MSS, Bh and R, is characterized by innovations of its own. This excludes the possibility that the Vulgata derives from any of these two MSS directly. As already said earlier MS R contains a contaminated text. The extent and nature of the contamination can only be determined after the main branches among the five Vulgata MSS have been identified, and will therefore be the subjects of 4.6 only. Below some instances will be given which purport to show that the Vulgata, if indeed not from Bh itself, derives from the same source as Bh, irrespective of the relation of R either to Bh or the Vulgata. a. 78 painna (prak Trna or prat Trna) in K, W, B, Y and P. The word be explained as an emendation (i.e. innovation) for pahalla in Bh (yahalla in Bho). This pahalla is a clerical error for pahulla, for which see T and compare pa(p)phulla in Ti and Tp. The development of pahalla into painna may be reconstructed as follows: pahalla + pa(y)alla + pailla (with -i- for -ya-) + painna (with-on- for -11-). For the intermediate form pa(y)alla, see paala in S, and for pailla, see R. Note in this connection that in Jaina Nagart - 11- and -un- are practically indistinguishable, so that the confusion is easily understood. For an attempt to assign a meaning to pahalla, see Bhuvanapala: pahalladolitam (ISt., p. 51), and Hem. IV 117, where pahalla is mentioned as an Adesa for Sanskrit ghurna-, i.e. as if it is paholla b. 156. Suharasa (sukharasa) in K, y, Y and P (cf. R). The word may be explained as an emendation (i.e.innovation) for saharasa as found in Bh, which seems to be a corruption, through the transposition of -ra- and

Page #63
50 -ha-, of sara(bhasa found in Ma, Tp, T and S and in the commentaries x and G (Patha). In the second hemistich of 237 in), Y and P the following readings are found: u niunasah Tkaralambiakaravalaandoliri, y niunasah Tkaraavalanbiakarabalaamdoliri, P niunas aht(ku)karalavala)valaando. It seems impossible to extract any suitable meaning from these compounds. In the last parts, vPage #64
51 consist of fragments only, namely K of 1-370 (Ed. 1-365), B of 1-129 (Ed. 1-127) and P of 1-300 (Ed. 1-297). All these five MSS seem to go back to one single source. This follows from a number of things, e.g. the innovations they have in common (see 4.1 c.) and those changes in the order of the Gathas they have in common vis-a-vis the archetype and the Jaina-recension (see 4.2). In order to determine the relations between the five MSS the innovations may be considered. At this stage the definition of the term innovation as given on p. 29 above may be broadened so as to include those readings typical of the Vulgata MSS against readings common to Bh (and R) of the Jaina-recension and Ma, Ti and Tp of the Third SouthIndian recension. It appears then that the first division has taken place between K and y, each being characterized by a number of innovations against retentions in the other (see under a. and b. below). It is to be noted that K in this respect seems more conservative than y. These two branches seem to have been combined to form a third, represented by Y and P, which is characterized by those innovations typical of K as well as by those typical of v. Besides, y and P have a number of innovations of their own (see under c. below). MSS Y and P seem to represent a South-Indian branch of the Vulgata, both MSS hailing from South-India. In the case of P this appears directly from the fact that it is written in the Kannada script. Regarding y, which is written in Devanagari, Weber (Ed., p. 2) has shown that it was copied from an original written in a South-Indian script. This follows, among other things, from such orthographical peculiarities as -ms- for -SS-, and -hn- and -hm- for -nh- and -mh- respectively. The short fragment B shares with those retentions typical of the latter (with two exceptions, for which, see under g. below). Furthermore, B and have a few innovations of their own (see under d. below). From these facts it is possible to conclude that the first division took place between K and y. Both traditions came to be combined in a MS, which formed the basis of Y and P (the possibility that y derives directly from the present MS P, or vice versa, can be excluded as each MS has a few innovations of its own). B would go back to an immediate predecessor of y. In this connection it should be mentioned that the possibility that B was copied directly from the present MS may not be

Page #65
52 ruled out, as of W and B only the latter has innovations of its own (e.g. kujjhai for j(n)urai in 38). These conclusions are, however, of a very tentative nature. For instance, the conclusion that y and P go back to one and the same source is contradicted by such instances where y and P differ, the former agreeing specifically with K and the latter with y, and vice versa (see under e. and f. below). It has also been argued that B derives from the same source as v or was copied directly from the present. MS W. On the other hand, in at least two instances B differs from w and agrees specifically with K, Y and P (see under g. below). These instances would seem to show that the Vulgata MSS or their respective sources were considerably influenced by each other. Another problem is that K as well as v occasionally have retentions in common with Ma, Ti and Tp (and T) against innovations in Bh, R, Y and P. E.g. 316 ahava in K, Ma, Ti and Tp against suhaa in v together with Bh, R, Y and P, and 26 sahasutta in v together with Ma and Tp agains suhasutta in k together with Bh, R, Y and P. These instances suggest that in K and u or in their respective sources authentic readings were. secondarily introduced, which, in turn, casts doubt on the status of the retentions in general in these two MSS (see a. and b.). It seems therefore impossible to arrive at an exact definition of the stemmatical relations existing between these five MSS. The diagram on p. 56 is indeed not a stemma; it merely purports to give an idea of the grade of the authenticity of the text of each MS vis-a-vis that in the other four. a. U, B, Y and p have the following innovations against retentions in K (i.e. readings K has in common with Bh, Ma, Ti and Tp). Below, of the other MSS only R is occasionally mentioned, namely where it differs from K. 20 punot, B, Y, P against uno 84 tujjha kae , B, Y, P against tuha kaena 106 dTsai gose v, B, Y, P against gose disai 111 samllavia y, B, paavia y, sallapalaavisam P against ullavia 120 maruehi(), B, Y, P against maruena (Ti, Tp; Gatha missing

Page #66
in Bh and R) 127 vi niuno , B, Y, P against a niuno 153 phalijjamtammi, hiaammi , Y, P against phalijjartassa, hiaassa (hemistich missing in R) 180 samalijjai U, Y, P against samalte 184 mannus amuppaena V, Y, P against mannusamuppannena 187 karihimi , kahimi Y, mi p against kahan padhamuggaa W, Y, P against padhamullaa 206 kuddaliniao X, Y, P against the lihiae K, tie lihirie R (tte liniao Ma, Tp) tahimmia gae tao W, tahimmi gae tao V, tammi dahmi ta etao P tahis cia coriae K, taha ccia coriae R 209 diaharambhe cuovanavalie , Y, diahassar abhe cumbhanavalie P against viadiahasavilakkhalakkhie 213 rahasujjalacunvanam , Y, P against rahasuvvellacumvanan 217 ahilihia , Y, P against alihia 218 padhamuggaa , Y, P against padhamullaa 219 asakkha , Y, P against asajjham 236 ghadiain , Y, P (Sanskrit) against ghadiana (genitive), hiayanan R 263 pattaaphalana sarise , Y, P (Sanskrit:ophalanai sadrse) against pattakalasaride K, pattayaphalasaricche R :..293 viruo vi W, Y, P (Sanskrit virupo'pi) against viruo a b. K, Y and P have the following innovations against retentions in y: 19 hiaatthian K, Y, P, B(!) against hiaacchia 31 sa K, Y, P against se (B) 33 mae (instrumental) K, Y, P against mahan (genitive) (B) paliviuna K, P, oiuna Y, B(!) against palfvium v (+palTvian K) 35 kulavahua nia K, Y, P against kulavahu nia(!) v, B (kulabahu niaa R) 55 samk (h)ini K, Y, P against kamkhiri (B) :56 tena tana K, Y, P against taha tena (B) 123 raena K, Y, P against vaena (B)

Page #67
131 kahin pi K, Y, P (T, S) against tahi pi (cf. tah in va Bh, Ti, Tp, tahin ca R) 205 onaa Y, P, onatta k against oatta Common innovations in y and P: 72 vahula (vahua) 76 dohaggam (doggaccan) 82 imea (unmetrical) 88 airikkammi vi (vairikkammi vi) 90 eanam Y, enam P (ettahe) 102 parirakkhanekkadinna v(cf. R), parirakhkadissa P (rakkhanekka dinna) 112 tti kira Y, tti tira P (tti ia) 122 dupparigejhjham (-) Y, duparigajja (-) P (dupparialla pi) arannan (rannam) 132 kim nu Y (cf. S), kim nam P (kim tha) 137 vinimilia Y, vinimtlia P (samTlia) 138 dajhjhihii Y, dajjahii P (khajjihii) 151 rehai (disai) 177 jaha (-) na taha Y, jaha ma na taha P (...taha na) 181 maulain Y, maulain P (mauain) 201 jaha(m) tan (taha tah) 210 te ahman Y, te appana armpam P (amh an te) 262 sundara Y, P (Sanskrit) (valaa) Common innovations in and B: 6 suhaa (--) suhasai U, suhaa hasai tuha B (with suhaa coming first) 22 luppa (+ tuppa litta) 36 saajjhia a duggaa (saajjhia duggaa a) 41 bharie (cf. R) (maie) 45 aipavas iesu diahesu (cf. I and s) (niccapahiesu diahesu)

Page #68
55 68 aippahaammi (aippahae a) 78 bhasarali , bhasaravali B (unmetrical) (bhamarali) 91 vara (pia) Instances where y and P differ and y agrees with K and P with y: 51 sugandhi Y, k' (cf. 7.) against suamdha P, \ (cf. K?) 178 tujjha virahe'Y, K against tujjha a(j)aso P, (cf. T) 200 genhaha Y, K (cf. s). against mahda P, W (cf. Bh, R) 241 tanut vi Y, tanui vi k against tanus (-) P, U 274 sabbhavaramiain Y, K against sabbhavaneharamiain P, V (cf. Bh; GTti-metre) Instances where y and p differ and y agrees with and p with K: 17 ehai so upaduttho Y, ehii so vi pauttho W, B (cf. Bh) against ehijjai so pauttho P, K (cf. R, Ma, Tp, S) 22 pariunvana Y, (cf. B, Ma, TP, T) against pariunvia P, K (cf. Bh, R, S) 72 chattha maggai Y, V (and all the other MSS) against maggai chattha P, K 76 vimuho Y, V (and all the other MSS) against vimano P, K :: Instances where B differs from and agrees with K, Y and P: 19 hiaacchia w against hiaatthiah B, K, Y, P 33 paltvium v against palTviana B, K, Y, P Note furthermore: 7 vania B, vaniau y against vilaao 46 sunnai K, v?, p against suvvai 42, B, Y

Page #69
56 Diagram showing the relation between K, W, B, Y and P: 4.6 The contaminated state of the text of MS R of the Jaina-recension MS R, together with MS Bh, constitutes the so-called Jaina-recension. Bh is incomplete, missing the last Sataka. The text is accompanied by a Sanskrit commentary, called Chekoktivicaraltla, which is attributed to a certain Bhuvanapala. It may be noted that the text quoted in the commentary (Bho) is occasionally slightly different from the text proper (Bha). MS R is complete, having altogether 710 Gathas (but Gatha 14 occurs a second time, namely as 386). In the first place there is evidence which suggests that R derives from the same source as Bh (the possibility that R derives from Bh itself, or vice versa, can a priori be ruled out as each MS has innovations of its own; for Bh, see, for instance, samappanti in 61, and for R, see 4.1 b.). This appears from the fact that they have in common certain innovations against all the other MSS (see 4.2 b.). It also follows

Page #70
from the changes in the order of the Gathas by which they together are distinguished from the archetype together with the Vulgata (see 4.2).20 Note in this connection also the following peculiar features found in the outward appearance of the two MSS. In both MSS the first half of 482 and the second half of 602 have been combined to form one Gatha, 527 in Bh and 533 in R. The Gatha so formed does not seem to make sense. Note also that Bhuvanapala's commentary on it is lacking. In Bh the last two words of Gatha 187 (Bh 387) are missing: sahasakariyae (hiaath tharatharei). The same lacuna must have occurred in the original of R as would appear from the peculiar reading found: sahasayae gharinThiyayam tharatharei. Above (4.1) it has been shown that Bh, R and the Vulgata are characterized by a number of innovations against retentions in the Third SouthIndian recension. Furthermore it has been shown that R has a number of innovations in common with the Vulgata against retentions in the Third South-Indian recension and Bh. These latter innovations would indicate that what later developed into R, on the one hand, and into the Vulgata, on the other, initially represented one tradition, which must have split off from a MS which was also the source of the present MS Bh. At the same time, though, R has some of those innovations which are typical of (and y and P) (see under a. below) and some of those which are typical of K (and Y and P) (see under b. below) of the Vulgata MSS. a... Innovations in R and W, Y and P against retentions in K: 176 phukkato R, X, Y, P against oastaa K, Ma, Tp (phumitti Bh) cuivato R, X, Y, P against oamtaa K, Ma, Tp (curhbamana Bh) * 183 hiyayao R, W, Y, P against hiahi K etc. 233 naharamka R, W, Y, P (Sanskrit) against dasananka K, Bh, Ti, Tp Innovations in Rand K, Y and P against retentions in y: 147 sasan R, K, Y, P against sase , Bh, Ma, Tp, T '170 palottata R, K, Y, P (T, S) against palotta y , Bh, Ti, Tp 199 payadiyasineha R, paalianeha K, paadiaNGeha Y, padianeha P

Page #71
58 against paadianeha , Bh, Ma, Tp Thus R contains material, as does Bh, which anticipates the Vulgata as a whole (retentions with Bh) and material which represents the last stage in the development of the Jaina-recension (innovations with Bh), on the one hand, and material which represents the first and the last stages in the development of the Vulgata (innovations with all the Vulgata MSS and with y and K, respectively), on the other. This state of affairs is explained by the fact that the text in R is the result of a process of contamination. For traces of this process in R, see, for instance, pattayaphalasari(c)che in 263 which is a combination of pattapphalasaricche of Bh (and Ma, Ti, Tp and k(!)) and pattaaphalana sarise of v, .. Y and P (Sanskrit). Besides, in the order of the Gathas R has a feature typical of Bh next to one typical of the Vulgata. Thus, in R Ed. 14 occurs both as 14, as in the Vulgata (14 K, P, E, G, 13 B, 124, Y), and as 386, as in Bh (382). In the former instance the readings agree with the Vulgata (note chittam), in the latter with Bh (note chikka):21 It should be noted that the occurrence of the Gatha in question as 16 in the Third South-Indian recension indicates that the Vulgata has preserved the more original order here. An explanation of the contaminated state of the text in R should start from a MS closely related to Bh. This would account for the occurrence in R of the retentions it has in common with Bh. It is unlikely that these, as the more difficult forms, would have been entered in a Vulgata MS for the innovations, by definition the easier forms, typical of the latter branch. This assumption dispenses with the necessity to suppose a complete rearrangement of a Vulgata MS in order to make it follow the order of the Gathas of the Jaina-recension. This MS was apparently thoroughly revised by comparing it to one or perhaps more Vulgata MSS. On this point there are basically two possibilities. The first one is that R was compared at least twice, first with a MS like K and next with a MS like y. The second possibility is that was compared only once, namely with a MS belonging to the Vulgata branch represented by Y and P. In the latter case it must have been a MS representing that branch before it became divided into y, on the one hand, and P, on the other, as R does not have any of the innovations common to these two MSS.

Page #72
-- - -. .. ps B g PKR Bh Ma Tp Ti Diagram showing the relation between the MSS of the Third South-Indian, Jaina and Vulgata-recensions

Page #73

Page #74
CHAPTER 5 5.1 The eclectic recensions: Sadharanadeva, Telinga and Second Telinga These three recensions are eclectic in the sense that the text of each of them combines material of at least two already independently existing traditions. In the following paragraphs I shall try to ascertain which traditions were combined in each recension. Among the three recensions the Second Telinga stands out as it consists of only 104 (or 102) Gathas. Weber (Ed., p. XLII) has already suggested that it probably concerns an independent anthology here, which means that the fact that most of its Gathas are also found in the Sattasat would be accidental. However, as most of the Gathas also occur in the Sattas at it seemed interesting to try to find out from which recension or recensions of the Sattas at they originate. In contrast to the compilation of, for instance, the Jaina-recension and the Vulgata, the one in the Sadharanadeva and Telinga appears to have been a very deliberate procedure. Thus, the 700 Gathas of the Sadhara adeva-recension have been arranged topicwise into altogether 60 chapters or Vrajyas. The extent of the Vrajyas varies from 74 Gathas in the Asat T-vrajya to 1 Gatha in the Vesya-vrajya. The principle according to which the Gathas have thus been classified varies too: in some cases it is the type of Nayika, as in the Virahin T-vrajya, in others a specific poetical device, as in the Jati-vrajya. The 700 Gathas of the Telinga-recension are divided into seven satakas. They are further classified into 26 chapters, here called Paddhatis. The extent of the Paddhatis is generally not indicated in the interior of the text, but their names are mentioned, in Prakrit, at the end of each Sataka. In two cases the Paddhati coincides with the Sataka, namely the Viraha-paddhati, which comprises the whole third Sataka, and the Pranayakalaha-paddhati, which comprises the whole sixth Sataka. The other Satakas contain each more than one Paddhati. While the Paddhatis provide a broad division of the Gathas according to erotic situation, poetical device, etc., within the Paddhati the Gathas have been arranged meaningfully as well. For instance, in the

Page #75
62 Viraha-paddhati, the whole third Sataka, several smaller groups can be distinguished. E.g. 201-213 deal with the time immediately before the actual separation. 214-223 describe the traveller; the Gathas have been arranged according to the season: 215-217 spring, 218 summer, 219-220 rainy season and 221-222 winter. Gathas 224-237 describe the Virahini. 238-246 are, again, arranged according to the season: 238-239 spring, 240 summer and 241-246 rainy season. Gathas 297-299 describe what the husband said to his wife on his return. In 300 after the union separation threatens again. A similar arrangement may possibly also be present in the Vrajyas of the Sadharanadeva-recension, especially in the larger ones. 5.2 The Sadharagadeva-recension (S) In Weber's material this recension is represented by only one MS, namely S, dated Samvat 1672 (1616 A.D.), which beside the text. of the Gathas contains the Sanskrit commentary by Sadharanadeva. The MS is written in Devanagari, but Weber (Ed., p. XLI) has suggested that it goes back to an original written in the Bengali script. This would appear from the occasional confusion between -r- and -v-, and from the exclusive occurrence of -v- in favour of -b-. He furthermore suggested that the recension itself should originate from Bengal. In support of this thesis Weber refers to the Bengali Visvanatha (fifteenth century A.D.) who in his Sahityadarpana, verse 565, discusses a type of Kosa in terms applicable to our S-recension, for which, moreover, he mentions as an example a Muktavali, which is the name given to the text in MS S (see Ed., p. XXXVIII). The eclectic nature of S becomes clear from the fact that it has readings otherwise typical of the Jaina-recension (Bh and R) (see under a. below) beside those typical of the Vulgata (with or without R) (see under b. below). Furthermore, it has readings typical of practically every individual Jaina (see under c. below) and Vulgata MS (see under d. below). In order to explain the present state of the text in MS S one has to start from a MS of the Jaina-recension. This would account for the occurrence in S of the retentions, however few, it has in common with Bh (see 4.1 b., Gathas 49, 78 and 156). It is unlikely that these

Page #76
63 would have entered into its text secondarily. Its dependence on a MS of the Jaina-recension becomes furthermore clear from a comparison of the texts of the respective recensions. S has 644 of its 700 Gathas in common with the Jaina-recension (see under e. below). Subsequently S would appear to have been influenced by all the other Jaina and Vulgata MSS individually. It is, however, also possible that some of the latter MSS were, instead, influenced by the text of S. Only in a few cases can the direction of the influence be ascertained. Thus, the reading pat viesattho in 550 in s would indicate that s was influenced by Y, as the reading pat ghare natthi in the latter MS can be interpreted as an 'emendation' (i.e. innovation) for patnaasurae (i.e. pat na asure) in y. It is doubtful, though, if this conclusion may also be applied to all the other cases where S and y agree. paviralo, for pariviralo, in 329 in both S and K would on the other hand show that K was influenced by S. It should be noted that in the Bengali script in which the source of the present MS S was written, the Aksaras -ri- and -vi- are practically identical so that the omission of the syllable -ri- can be more easily explained in s than in K, namely as an instance of haplology. The text of S bears clear traces of this contamination. Consider the following cases:22 354 sunnabhamir ina vahulaiana + sarahasabhamirina vahulaiana K .. (surahasa!), W, Y, Ma, Ti, Tp x sunnovas abhamirTna vahana Bh, R 485 hiaan hiaanihian + hiaan hiaanihitta Bh, R, Ma, Ti, Tp * hiaam hiae nihian W, Y 541 pacamtavanasalilanivahena (GTti-metre) + padatavahanivahena Ma, Ti, Tp x vahasalilanivahena Bh, R, etc. .556 patta niahvapphamsa + patta ni ambaphams am Ma, Ti, Tp x patta n iavapphamsa U, Y, R, onva(p?)phaisa Bh S has furthermore a number of innovations exclusively in common with the Telinga-recension (T). There are reasons to assume that I was influenced by S, and not the other way around. This problem will be discussed in the next paragraph (5.3). It is noteworthy that s includes practically no readings characteristic of the Third South-Indian recension. Exceptions are: 465 pasijjiri

Page #77
and 616 luliakesasu miliacchisu (order). See also 541 pacantavahasalilanivahena and 556 patta niamvapphamsa mentioned above. As some of these instances possibly are retentions it may be that s was, among others, also influenced by a MS of the Jaina-recension the text of which was in certain respects more authentic than that of either Bh or R. Readings in s typical of Bh and R: 9 ma ruasu, 136 hoi danarahiassa, 142 dukkhasokkha, 212 accharian, 223 uvvahai vahu jaha jaha, 224 alam me, 238 vikkei, 314 maulian, 327 vaddhenti S, R, vaftahti Bh, 393 chippai, 400 rusai ccia, 416 ajj(h)a, 432 na niruddho, 444 maulan S, R, Bho, muindalam Bho, 465 vahuna sokkhana S, R, angana vi suhana Bh, 517 visavad (/y) antan, 524 de daun S, je daum Bh, je dar am R, 579 annanafsu, 583 bhahi S, R, hohi Bh, 586 siharammi, 691 gavvio (Bh not available). Readings in s typical of the Vulgata MSS, with or without R: s includes nearly all the innovations found in K, W, B, Y, P and R (against the retentions in Bh, Ma, Ti and Tp) enumerated in 4.1 b. and nearly all the innovations characteristic of the Vulgata enumerated in 4.1 c. To these may be added the following readings which S shares exclusively with and y, where K and P are not available: 409 ekko panhaai thano vto pulaei nahamuhaviddho, 438 atiujjua S, aiujjua , aiajjua Y, 466 de suanu, 474 visao S, vesi(/a)nto , Y, 483 tammai hiaammi (locative) (cf. T), 501 vicchaar, 512 ahaan, 520 jo mam aia(a)rena, 589 tahvira, 595 addams ano, 614 esa guno, 680 vannalohitae (cf. R vannalohillo), 683 ma eam vvia S, ma etti ccia y, ma ettia bia y.

Page #78
65 Readings in s typical of Bh: 212 so tan, 217 jan mittam vasanadesa(k)alammi, 529 paino bhuasiharan. Readings in s typical of R: 196 pari(t)uttha, 586 viala (cf. Bh viulla), 621 bhamai, parimalai, mu(mc)ai. d. Readings in s typical of the Vulgata MSS: S shares a number of the innovations characteristic of U, B, Y and P (against retentions in K) enumerated in 4.5 a.: 127 vi niuno, 153 phalijjantammi, hiaammi, 180 samalijjai, 187 padhamuggaa, 206 kuddaliniai, 213 rahasujjalacunvanaan, 218 padhamuggaa, 219 as akkhan, 236 ghadiais. S goes with K in 20, 84, 106, 111, 187 (a), 209, 217 and 293. Readings in 's typical of K: 185 sans itto, 186 sisirai(), 214 aisochaala s, aisuira K (ciraala), 220 dinnatarala, 228 suhaa jam ca, 238 mammurasachahe, 285 vialu', samattha. Readings in s typical of : 152 anunehi, 360 onaai, phalahte (singular), 409 aviddho S, alitdho , 441 vijjha ham S, ha vijja v.

Page #79
Readings in s typical of Y: 15 padhamuggaa, 307 gananahi(m), 323 laggan, 450 suhaa, 487 na dei sivine vi, 550 pat viesattho S, pas ghare natthi Y, 611 va gotthan. Readings in s typical of P: 105 kaja vi S, karai mia P, 116 dhanummi nisanna S, dunumi nisana P, 155 na anamti. Readings in s typical of .y and P: 122 (duppariallam) nei ar aNGNGah (cf. T), 132 kim nu S, Y, kim nan P. Readings in s typical of y and B: 45 aipavas iesu diahesu (cf. T). The numbers in this section are those of s (see Ed., pp. 433-449). A comparison between the SadharaNGadeva-recension and MS R of the Jainarecension shows that the former includes practically all the Gathas of the latter. Of the 56 Gathas not occurring in R, 11 are found exclusively in S: 50 (Ed. 800), 69 (801), 105 (802), 150 (804), 223 (805), 365 (806), 404 (807), 446 (808), 681 (812), 699 (813) and 700 (814). 10 are found exclusively in the Vulgata (with x, unless indicated): 4, 5, 71, 74, 131, 133, 313, 409 (-x), 508 and 658; 4. in the Vulgata (with x) and T: 447, 586, 573 and 646; 20 in the Vulgata, Ma, Ti and Tp, and T: 51, 98, 107, 172, 207, 247, 265, 275 (also w), 358 (also W), 366, 368, 471 (also W), 482, 507 (also W), 509, 514 (also W), 582, 637, 688 (also W) and 698 (all also in x, except 688; for 582 x is not available); 3 in the Vulgata and Ma, Ti and Tp: 263, 583 and 635 (all also in x; for 583 x is not available); 3 in Ma, Ti and Tp and T: 148, 444 and 501 (also in x and W); 2 Gathas are found exclusively in x: 291 and 512 (?); 1 in T: 411; 1 in Y, P and T: 590; and 1 in Y and x: 667. A comparison of S with the Vulgata yields a practically identical

Page #80
result, s sharing 636 of its 700 Gathas with the Vulgata. Of the 64 Gathas not found in the Vulgata, 11 are exclusively found in s (see above). 13 are found in R: 24, 26, 44, 46, 49, 82, 260 (also Ma, Ti and Tp), 266, 433 (id.), 570 (id.), 627, 688 and 691; 2 in x: 291 and 512 (?); 6 in R and X: 27, 213, 258, 386, 505 and 521; 4 in T (and Ma, Ti and Tp): 148 (Ma, Ti, Tp), 411, 444 (id.) and 501 (id.; also in x and W); 14 in R and T (and Ma, Ti and Tp): 23 (Ma, Ti, Tp), 25 (id.), 45 (id.), 48 (id.), 50 (id.), 190 (id.), 321, 380 (id.), 412, 432 (id.), 497 (id.), 506 (id.), 557 (id.) and 563; 9 in R, X and T (and Ma, Ti and Tp): 28 (Ma, Ti, Tp), 29 (id.), 149 (id.), 224 (id.), 271 (id.; also w), 524 (id.), 626 and 692; 1 in Y and x: 667; 1 in Y, P and T: 590; Gatha 65 occurs in all the MSS minus v (K and P not available), Gatha 169 in all minusy (K and P not available), and Gatha 191 in all minus y and x (K and P not available). Above (4.2) it has been shown that the text of the Third South-Indian recension was compiled through a process of selection from the text of the archetype. Traces of the order of the archetype, as preserved in the Vulgata, on the one hand, and in the Jaina-recension, on the other, can still clearly be seen in the Third South-Indian recension. In the same way in s occasionally sequences of Gathas -- whether interrupted by others or not -- of the Jaina-recension or the Vulgata can be recognized, however, without a clear patter emerging. Though, considering the thorough process of arrangement visible in s this should not be surprising. 5.3 The Telinga-recension (T) In Weber's material the Telinga-recension is represented by several MSS. The Prakrit text is based on a collation of three MSS, one written in the Telugu script (To), the other two in Grantha (Weber's Te but here Tabelled to to avoid confusion with TC). The text is accompanied by an anonymous commentary. In the Telinga-recension the tradition of the Third South-Indian recension has been combined with those of the Sadharanadeva-recension, the Vulgata and MS R of the Jaina-recension. Its dependence on the Third South-Indian recension becomes clear in the first place from the fact that T has many readings otherwise exclusively found in the latter recen

Page #81
68 sion (see under a. below) and in the second place from a comparison of the respective texts. It appears, among other things, that T has 101 Gathas exclusively in common with the Third South-Indian recension (see under b. below). T agrees with the Third South Indian recension in that it, too, has certain forms which point to an original written in Jaina Nagart (see under c. below). T has a relatively large number of readings exclusively in common with S (see under d. below). Besides, it has a few readings in common with the Vulgata and with some individual Vulgata MSS (see under e. below). In this connection it should be noted that T also shares a number of readings with the Vulgata together with S (see under f. below). Finally, I shares a number of readings with R; and with R together with the Vulgata MSS and S (see under g. below). That T was influenced by these respective traditions and not the other way around would appear from the absence in either S, the Vulgata or R of any of those readings typical of T and the Third South-Indian recension. In this connection it may be noted that during my visits of manuscipts libraries in India I did not come across any North-Indian Devanagart MS of the Telinga-recension or, for that matter, of the Third SouthIndian and Second Telinga-recensions, while in Trivandrum I was shown a South-Indian MS of the Vulgata with Gangadhara's commentary. This evidence should, however, not be stretched so far as to mean that there never were MSS of the South-Indian recensions in North-India. Furthermore, the manuscripts libraries contain only the top of the iceberg of the available MSS in India. Above (4.3) it has already been shown that Tp of the Third SouthIndian recension was influenced by T. T has many of the retentions of the Third South-Indian recension against innovations found in the Jaina-recension and the Vulgata, enumerated in 4.1. Besides, it has many readings otherwise exclusively found in the Third South-Indian recension. E.g.

Page #82
69 2 paada, kaha te, 104 ciae, 127 ummatthaai, halao (cf. W), 154 kunejjasu, 156 bahumagganimmiaii vesavilaana pemma in, 175 golanaikacchakuduga, 177 kalaba, 204 nivvuijjae, addana va padibimban (order), 250 appian, 253 dummuhao, 255 nivvui an, 257 haliasua, 259 nadin, 268 sahavasarala pi thanajuar padian, 284 manan, 291 dimbham anulam(/g)giri, 305 dukkhena, 335 jaggassu (Ma, Tp jaggasu), 354 tuhnikkaparungena pi ajja maNGamsinimuhena, 358 hiaammi, 377 nii, 389 uvvea(m)roirte, 390 manassa, 393 kim tattio, 405 kamalahaena jam viasasi, 409 ekkam passasai thana bfam pulauggamena phuriamuhan, 429 agghia, 441 jammastarai vi saam calanan, 497 ja maa sa maa, -516 piavirahe, 521 pagtajualati, 523 tue, 546 vedha(1)la, 549 maddana, 559. maulantacchai, 573 ajj(h)a, 586 vilola, 589 kauna, 590 sikkhan, 593 savvo, 598 majjham pat kuanda (m), savattt a (hemistiches reversed), 607 dhakkesu, 643 darapikkakavitthagadhalohena, 647 samahiana T (samihian Ma, sumahia Tp), 649 vipphurai, daiena (cf. W), 651 jai (cf. W), 653 dfsataditthisuhao, ullaventasuisuho, 658 asavattie, 660 vahai, 677 pasaanam (T paasanam), 685 muhalo, 689 annam gamam patta, 724 ukkhippei, 747 dhavalia ti kauna, 756 aghadia, 779 so matthae padai. The numbers in this section refer to the Telinga-recension, for which, see .Ed., pp. 450-501. Of the 700 Gathas of the Telinga-recension 28 are exclusively found in this recension (221 also in x) and 14 only here and in W (660 also in x). Of the remaining 658 exactly 101 are otherwise only found in the Third South-Indian recension; some of these also occur in W. Of the reman ing 557, 424 occur in all recensions, i.e. the Vulgata, in which x has been included here, in the Third SouthIndian recension, in R of the Jaina-recension and in S. Of the remaining 133, 35 are found in all except the Third South-Indian recension, 22 in all except R, 16 in all except S and 18 in all except the Vulgata. 3 are found only in the Third South-Indian recension and R, 3 in the Third South-Indian recension and S, 8 only in the Vulgata and R, 4 in the Vulgata and S, 8 in the Vulgata and the Third South-Indian recension, 6 only in the Vulgata, 1 in R (and x), 6 in R and S, 1 in y, P and S, and 1 in y (and W).

Page #83
70 C. There is evidence in T which points to a source written in Jaina Nagari; see mastu for mannu in 683, 938 and 941. Ta which is written in the Telugu script shows numerous traces of having been copied from an original in Grantha. E.g. -i- for -u- in 297 haNGiim ((b)hanium), 298 dijjai (dijjau), bhani im (bhanium), 445 paisa (pausa), 454 ukkhuviin (ukkhivium), 484 valiin (valiun), 685 jamaissa (jamaussa), viinei (viunei), 759 ramei (rameu); --- for --- or -s- for --: 70 pauda (To paudu), 538 id. (To paulo), 578 id., 526 paedam (paes am), 392 kudumbi (kusumbhi), 546 kudumbha (To kusumbha) and 388 pasuria (pazduria); -h- for -bh-: 4 hisini (bhisini), haana (bhaana); and -9- for -bh-: 546 garillTo (bharillto) and 722 garaena (bharaena). Readings T has exclusively in common with S: 146 avah Tri(a), 164 sukha T, sokkhan S, 216 tujjha, 254 sutto T, saio S, 277 te ciram runnah (cf. P), 297 enam, 345 dumesi, 350 saan cia, 363 sahanijjo, 371 kim ta na, 383 khuha (19), 419 lajjaluint vi vahu, 444 pavai Tsi vi T, tsi vi pavai S, bhasala, 487 se hian, 514 kulavahunan, 527 gao ccia, 531 tattha, 542 hiaammi, 546 mettto, 565 vaanehin, 573 ditthinah, 586 uppakkha T, purkham S (-kh-!), 647 neii T, naai S, 745 pasaamanena, 756 bolamte. e. Readings T has exclusively in common with the Vulgata: 451 osarasu, 484 tte valiun, 524 (-) vahuvallaha, 566 pikka.

Page #84
71 Readings T has in common with some individual Vulgata MSS: with K: 235 samaam chia kim vihina, vinimmivia, 265 dese. with Y: 643 alehaladdua. Readings T has in common with (K), 0, Y, (P) (Vulgata) and S: 142 pariaddhana, 205 pasasaki kaa necchai, 278 sovanavant te, 310 mottuna, 316 kin tena sineho, 319 muhahi, 361 ahmaristhi, 365 mammaraan, 380 sappurisa, 388 samudduttarata, 444 maar andapanalohilla, 497 varei mat (cf. Tp). Readings T has in common with S and some of the Vulgata MSS: S and K: 177 (kalaha) ja ma taha na T, S, jam at taha nah K, 354 manas ini, 356 amha. S and Wand B: 45 (s)aiposiesu diahesu. S and Y and P: 122 (dukkhaallam) nei arang an. Readings T has exclusively in common with R: 243 janamtassa vi, 268 hiae ko nama, 291 palaata, 297 paino turiai, 543 edan Ta, eta To, eyath R, 559 likka, 589 avirala, 620 ekko mao T, ekkamao R, 669 ma pahia, sejjae. Readings T has in common with (K), X, Y, (P) (Vulgata), R and S: 90 khamijjanti, 136 horft t, 152 lohillan, aigahiagarua, 153 padihai, 163 jivijjai, 170 palottanta, 188 pautthamanam, 198 jampamana, 290 vilTno, 320 vi na taha, 353 vi na taha (R vi na taha), 368 thanaalasuppellana, 463 kunai kisa, 575 anandiali.

Page #85
72 5.4 The Second Telinga-recension (W) In Weber's material the so-called Second Telinga-recension is represented by two MSS, the one written in Grantha (W), the other in the Telugu script (V). The former contains the text of the Gathas accompanied by a Sanskrit commentary; the latter contains a Sanskrit Chaya only. The beginning of W, comprising the first ten Gathas, was illegible. The recension is markedly different from the other known recensions in that it contains only approximately 100 Gathas, namely 102 in W and 104 in V. The text bears various names. In W is is called PrakrtasrngaraSataka and in V Prakrtasataka. Yet another name seems to have been current. In Upadhye's edition of this recension it bears the title SaptaSatisara.23 In this latter edition the commentary, which probably is the same as the one found in W, is attributed to a certain Vema Bhupala. This Vema Bhupala has been identified with a Reddi king of that name who ruled from 1403 to 1420 at Kondavidu in Andhra (Krishnamachariar, 1970: 479; Sriramamurti, 1972: 68-69). As already said above (5.1). Weber suggested that the Second Telingarecension is possibly an independent anthology, having been compiled independently of the recensions of the. Sattasai. This would, among other things, appear from the fact that the compiler apparently had no intention to create a 'sevenhundred'. The best proof for the assumption that the Second Telinga-recension originated independently of the other known recensions of the Sattasaf, would be to be able to show that there was another source from which the compiler drew the Gathas. In this connection it should be noted that a considerable number of Gathas (17) are also among those Weber collected from the Alamkarasastras (see Abh., p. 202 ff.; collectively labelled 'A' here) (7, 15, 19, 39, 66, 68, 76, 83, 85, 87, 93, 96, 98, 99, 100, 101, 102) and some even exclusively so (68, 93, 99, 101). In some instances the readings agree more closely to the latter. E.g. 875 anaddamana W and A (+Sanskrit; anollamana T), 860 duna W, duni A (+Sanskrit; domia T), 939 nimiliaamh W, milia A (pimiam T), 951 kaccT a W, kattT a A (kattTo T). From whatever source or sources the compiler of the Second Telingarecension drew the Gathas, most of these also occur in the available recensions of the SattasaT. A relatively large number of them otherwise

Page #86
occur exclusively in the Telinga-recension, some of which are found in the Third South-Indian recension as well (see under a. below). The readings generally agree closely with those of the two other south-Indian recensions (see under b. below). Occasionally though it has readings typical of the other recensions (see under c. below). The text of W shows clear traces of contamination, for which consider the following examples: 33 unna vinisasamto as in Ma, Tp and T(?) beside kisa hma para muh Te sayanaddhe as in all the other MSS 237 sahiaasah Tanapasaria as in Ti, Tp and T beside karajuala as in and y 688 vaccasu as in Ma, Ti, Tp and T beside imana as in U, Y. and S. a. In this section the longer form, with 104 Gathas, of the Second Telingarecension is considered (see Ed., pp. 502-508). Of the 104 Gathas 11 are found exclusively in the Second Telinga-recension: 41, 67, 68, 89, 93, 94, 99, 101, 103 and 104. 14 Gathas otherwise occur only in T: 3, 17, 19, 39, 66, 76, 81 (also in x), 82 (also in the South-Indian Vulgata MSS Y and P), 83, 85, 87, 96, 98 and 102. It shares 7 Gathas exclusively with T and Ma, Ti and Tp: 4, 8, 14, 29, 56, 77, 78. The remaining Gathas are found in all the other recensions or MSS. b. Readings common to W and T: 33 petthan, 91 pasaerti, 167 bhana kaha ajja vi dhammia, 149 tassa saasam ghanandhaare tti, 335 bahula, 372 dinno stsammi saha juanehin, 583 dhania(in) uvaguhia mae, 655 navaraaratta, kaaloa, gammiasavvaria, 664 anumannai jaha jaram, 669 atta atta Yamajjai, 938 tanno (Ma, Tp tannai). Readings common to w and the Third South-Indian recension: 33 unna vinisas asto, 190 devaullaan, amha, 237 sahiaasah Tana

Page #87
74 pasaria, 254 olambae vasais, 335 na amusajjamo, 649 paripucchia na jampai. Readings common to W, T and the Third South-Indian recension: 42 dharan, 127 ummattaai, halao, 499 katto khemam katto, mattaovari. (compare khujjamjao with kuvvajjao in Ma), 649 vipphurai, daiena, 651 jai, 688 vaccasu. C. Readings in W not found in the two other south-Indian recensions: 33 ktsa hma par anmuhte sayanaddhe K, 4, B, Y, P, S, 85 nikkaiavena janam gao si k nivvur tanam Vulgata, R, S, 173 raia y 175 goatadaviadakudunga Vulgata, s (Gatha missing in Bh and R), 476 ciasaan (+vidaaman Bh, R, Y; transposition of the Aksaras -da- and -a-, and -S- for --- (Grantha:)), ettha abhavvae...kaha nu kaappa Vulgata, Jaina, s (Ma, Ti, Tp: kaha nu abhavvae...ettha kaavvam; Gatha missing in T), 550 pas vi annatto Y, S, 554 eddahamettan(/g) game , Y, 688 imana V, Y, S (Gatha missing in Bh and R).

Page #88
CHAPTER 6 6.1 Concluding remarks One of the results of the above study is the conclusion that all the available MSS ultimately go back to a single MS. While the text (readings) of this MS appears in many respects to be most faithfully preserved in the Third South-Indian recension, the order of the Gathas is in the Jaina-recension and the Vulgata. The remaining three recensions were compiled by selecting Gathas from the former three and thus suppose their prior existence. This conclusion is of considerable importance for the history of the genre. The Sattasaf is generally considered the oldest anthology (Kosa) of the classical period (Sternbach, 1974: 10). In all later kosas the verses are divided into sections on the basis of categories developed in the Alamkarasastras and in the Nayikabhedaliterature. A similar arrangement is also found in the Sadharanadeva and Telinga-recensions, which, however, have been shown to be relatively late revisions of the text. However, in its original form, as represented by the Jaina-recension and the Vulgata, which is probably older than any of the other known Kosas, the order of the Gathas appears to be random. In these two recensions occasionally two or more Gathas are connected by a similar content, idea or motif. Thus, Gatha *6 and *7 refer to two different trees, the Kurabaka and the asoka, which blossom only after having been embraced and kicked by a young woman respectively. *13 and *14 describe a newly-married wife's clumsiness in the kitchen; and in *41 and 42 a woman who does not want to hurt the feelings of her husband and her neighbour respectively. However, no overall organization on the basis of content is visible. In fact, the Gathas seem to follow each other associatively, in which the occurrence of identical words seems to have played an important role. Thus, it appears that consecutive Gathas or larger sequences of Gathas are linked together through the occurrence of identical words (Tieken, 1978). A similar type of concatenation has also been recognized by Bloomfield (1899: 39) in Vedic literature, while Schubring (1955) has demonstrated its existence in classical Kavya.

Page #89
76 Another conclusion which follows directly from the above study concerns the problem of the authenticity of the tradition which ascribes the Gathas to individual authors. The occurrence of the names is restricted to MSS P (Vulgata) and Bh (Jaina), while a few names are also found in the beginning of MS K (Vulgata). The earliest reference to the Sattasat is probably the one found in the thirteenth verse of the introduction to Bana's Harsacarita (seventh century A.D.) (see Ed., p. XI). In so far as the names found could be identified they belong to persons living before Bana. Thus, a significantly large number of names refers to kings; note those ending in -raja, -adhipa and -sena. Some of these have been identified by Mirashi (1947) as kings of the South-Indian Satavahana-Andhrabhrtya, Vakataka and Rastrakuta-dynasties respectively. The latest datable name appeared to be that of the Rastrakuta king Mana or Mananka (420-455 A.D.). The only positively post-Bana author is Vakpatiraja, if he is indeed the same as the author of the Gauqavaho. Vakpatiraja lived between approximately 715 and 750 A.D. The name Vakpatiraja is, however, restricted to MS Bh, MS P having a different name. On the other hand, there is no evidence for the presence of the names in any stage of the transmission of the Sattasaf earlier than the Vulgata. This appears from the following comparison between the names occurring in the Jaina-recension and the Vulgata. In P the names, occurring in Sanskrit and in the genitive, are available up to the middle of the second Sataka, i.e. only as far as PTtambara's commentary is available (see Ed., pp. LIII-LV). The names are found after the Gathas. In Bh practically each Gatha is provided with a name, in Sanskrit and in the genitive (see ISt., pp. 9-16). They are found before the Gatha (ISt., p. 17). Whereas P and Bh generally agree on the names they differ in the attribution of a particular Gatha to a particular author. It appears that in Bh the names have shifted one -- sometimes more than one -- Gatha forward. This is most probably the result of connecting the name with the Gatha which follows instead of with the one that precedes. That P and not Bh has retained the correct ascriptions becomes clear from those names which were created after some incidence described in the Gatha. E.g. Kalasara, 'Black antilope', found in Bh with 24 (Ed. 26)

Page #90
should belong to 23 (Ed. 25). In 4.2 I have shown that the Jaina-recension and the Vulgata go back to one source. Each is characterized by changes of its own in the order of the Gathas. Below I will consider two instances where the flow of the text was thus interrupted. In Bh the nickname Pavana, 'Wind', is found with 521 (Ed. 498). This name more likely belongs to 450 (Ed. 497), i.e. the Gatha which precedes in the Vulgata, than to 520 (Ed. 627), the Gatha which precedes in the Jaina-recension. It looks as though the names found in Bh were borrowed from a Vulgata MS. It is also possible to maintain that when the order of the archetype, which is preserved in the Vulgata here (see Appendix III, 454, 337 and 412), came to be changed in the Jaina-recension, the name which originally belonged to the preceding Gatha was taken along with the following Gatha to the new position in the text. This latter explanation does, however, not hold in the case of Sthirasaha, 'Strong branch', found in Bh with 99 (Ed. 168). This name should belong to 367 (Ed. 167), the Gatha which precedes in the Vulgata, rather than to 98 (Ed. 96). From Appendix III (117) it appears that in this case it is the Vulgata in which the original order is disturbed, while it is preserved in the Jaina-recension. It follows that the names of the authors found in Bh were indeed borrowed from a MS of the Vulgata. 24 This also means that there is no evidence for the presence of the names in any stage of the transmission of the text prior to that of the Vulgata. This latter conclusion should, however, be treated with caution. In this connection it is to be noted that the names were, or were considered as, part of the commentary. Thus, in P they are available only as far as PTtambara's commentary is; in v and y, which contain only the text, they are completely absent despite their final Gatha which specifically says that the Gathas are accompanied by the names of their authors (see Ed., Gatha 709). Furthermore, they were not considered an essential part of the text, as appears from the fact that in K they are given only for the first few Gathas. Therefore the absence of the names in the Third South-Indian recension, on the one hand, and in the Jainarecension, on the other, cannot be taken as absolute proof that they were absent in the archetype as well. If despite these uncertainties the names were indeed added to the

Page #91
78 text only in the Vulgata, the question arises what may have motivated this addition. Above we have seen that in the Sadharazadeva and Telingarecensions the Gathas are arranged topicwise, as is the case in most other texts of the genre. This revision of the original text most probably took place in an attempt to imitate those other kosas. Only a few of these Sanskrit kosas are original works. Most of them are mere compilations of verses from works of different authors, in which case often with each verse the name of its author is given. The addition of the names to the Gathas of the Sattasat may be explained as the result of an attempt to copy yet another feature of the Sanskrit kosas.

Page #92
79 Notes to the stemma The Gathas aptasatt of Satavahana. With the Commentary of Gangadharabhatta, ed. by Durgaprasad and kasinatha Pandurang Parab, Kavyamala 21, Bombay, 1889. Saptasatisara with BhavadTpika of Vema Bhupala, along with the Chappannaya-Gahao (text and chaya), ed. by A. N. Upadhye, Shivaji University Sanskrit and Prakrit Series Vol. III, Kolhapur, 1970. 3 E.g. Hindt Gathasaptasati, ed. by Narmadesvara Caturvedt, Vidyabhavana Samskrta Granthamala 55, Benares, 1961. E.g. Hindi Gathas aptasatt, ed. by Jagannatha Pathaka, Kashi Sanskrit Series 192, Benares, 1969. 5 Weber's Abh. has been extensively reviewed by G. Garrez in JA., 1872, pp. 197-202. In 1872 Weber published in ZDMG 15, pp. 735-45, a short article called 'Ueber das Saptasatakam des Hala'. It mainly served to announce the discovery of some new MSS, which were edited later in his 'Retractatio' (see below). 6 I use here the siglum T for Weber's Te in order to avoid confusion with T. I use C for Weber's G as he does himself in his edition of 1881. For the first 365 Gathas of the Ed. the variants of the MSS edited earlier in respectively the Abh. and the Retr. have to be gleaned from these two publications themselves. This can be done only by a roundabout way, for the variants given in the Retr. refer back to the text

Page #93
of the Abh., which occasionally differs from that in the Ed. 8 80 Bhuvanapala is apparently otherwise called Tribhuvanapala. Another MS containing the text of the Sattasaf accompanied by the commentary by Tribhuvanapala is no. 76 of the Kaiser Library in Kathmandu, Nepal. The MS, consisting of 230 palmleaf folios, is written in the Old Newari script. It is at the same time the oldest datable MS of the Sattasaf, dated 341 N.S., i.e. 1221 A.D. Tribhuvanapala is the father of a certain Ajada, another commentator on the Sattasat. A commentary by Ajada is found in MS 385 of 1887-1891 in the Bhandarkar Oriental Research Institute in Poona. A short description of the MS is given by Sandesara (1972). The MS was filmed by me in the summer of 1976. It consists of a bundle of altogether 48 folios not forming a consecutive text but being a collection of parts of what once was a complete MS. The text is a specimen of the Jaina-recension. After Gatha 200 the following particulars regarding text and commentary are given: 9 iti mahapi()ditasrtdallaniprthivtpalasutatribhuvanapalasunun trividyena SrTmadajadena krtayah maharajadhirajasratavahanagathakosasaptasatittkaya dvittya Satam samaptar. As far as can be judged from Weber's edition of Bhuvanapala's commentary, which consists mainly of extracts, Ajada's commentary is merely a slightly shortened version of that of his father. A MS containing this commentary is mentioned in the Brhattipanika (published in the Jainagranthavalf; see NCC V, p. 343 b). The Brhattipanika is an old (Safvat 1440, i.e. 1383 A.D.) list of MSS in the possession of several Jaina Libraries of that time in Gujarat (Tripatht, 1977). Kulanatha also wrote a commentary on the Setubandha. For particulars of this commentary, see Handiqui (1976. 96-99 et passim) 10 I have been able to locate five more MSS which may contain the text of

Page #94
the Third South-Indian recension. It should, however, be noted that the catalogues and lists from which the references have been taken with one exception (see 1 below) do not quote the beginnings and ends of the texts. The conclusion regarding the recension to which the texts belong has therefore been solely based on the fact that in each MS the text is accompanied by a commentary which otherwise is exclusively found in connection with the Third South-Indian recension. Note that all the five MSS mentioned below are written in the Malaya]am script. Compare Ma, Ti and Tp which are written or go back to a MS which was written in this script. It seems that the Third South-Indian recension has been preserved solely in Kerala! MS no. 4762 in the University MSS Library, Trivandrum (see Alphabetical Index... (ABC 316), no. 4762, and A descriptive Catalogue...(ABC 321, 6), no. 1438). The MS was formerly owned by Mitran Avl., Cempakasceri 1llam, Kudama lur, Ettumanur. It is a palmleaf MS of 213 leaves of 7 by 1 3/4 inches. The script is Malayalam. From the quotations of the beginning and end given in the last-mentioned catalogue it appears to contain the text' of the Gathas of the Third South Indian recension together with Nadhavayajvamisra's commentary, from 1 to 681. This MS was shown to me while I visited the library in Trivandrum. I was, however, not allowed to film it as it was feared that the MS, which was already very much broken, would deteriorate further. MS no. 294 mentioned in the Lists of MSS collected by the Curator for Publication of Sanskrit MSS, Trivandrum, Vol. I, p. 30 (ABC, UC 32); described as Halas aptasat Ttatparyatika by Madhavayajva. The MS was formerly owned by Govindapisharodi from Kailasapuram (cf. Ti). The material is not specified; the script is Malayalam and it is said to contain 3650 Granthas. 3 MS no. 279 mentioned in the Lists of MSS collected by the Curator for Publication of Sanskrit MSS, Trivandrum, Vol. IV, p. 25 (ABC, UC 32);

Page #95
82 described as Halas aptasatitatparyatika by Madhavayajva. The MS was formerly owned by Mitran Namburi, Kumaranallur. The material is not specified; its script is Malayalam and it is said to contain 4200 Granthas. It lacks part of the beginning. MSS 2 and 3 should at present be in the possession of the University MSS Library in Trivandrum but could not be located. MS no. 4304 mentioned in the Catalogue of the Sanskrit MSS in the Panjab University Library Vol II, Lahore, 1941 (ABC 145); described as Gathasaptasativyakhya tatparyadipika by Madhavayajva. The material is palmleaf, the script Malayalam. It consists of leaves 4 to 106, comprising the fourth and the fifth Sataka. MS no. 4305 mentioned in the same catalogue as '4'; described as Gathasaptasattvyakhya tatparyatTka by Madhavayajva. The material is palmleaf, the script Malayalam. It consists of 79 leaves, said to contain altogether 678 slokas(?). 11 Theoretically it could be assumed that in each case the typical MIA form, e.g. nti, vir(/1)ao and pamhusai, is the innovation, secondarily introduced in order to give the text a more pronounced prakritic appearance. However, the distribution of the sanskritic forms in question, namely nisarai, vilino and visarai respectively, coincides with that of the other categories of innovations mentioned on the same page, which suggests that they are innovations as well. 12 54 Without going any further into the problem of the variation of siri and sari I should like to note that commentator Bhuvanapala (MS Bh) must have read siri (Sanskrit srt) as well (see ISt., pp. 45-6). 196 For udalie or uddali e Madhavayajvamisra in his commentary gives the following explanation:

Page #96
83 dramilair uddalir iti ya krida vyapadisyate soddalika...utkuncitobhayajanukas inaya sajaghanotkamp am rabhasaikaikapadotsaranena nivartaniyatvad uddalikayah purusayitopabhogitvam iti mantavyam. Prakrit umali or uddali seems to be connected with Tamil untal (pronounce undal) 'to play hop-scotch' (for information regarding this meanning, which is not found in any of the dictionaries, I am indebted to my colleague Dr. A. Govindankutty). 431 tamtassa stands for tantasya, the genitive of the past participle of tam-. Misunderstood, it was divided into tah tassa, which triggered the innovations given. 460 In Bh this Gatha is attributed to Dhavala (ISt., p. 161). 487 vabena...na sahian tie, 'her tears did not allow her...' 626 suhmamtaria, '(shade) interspersed (antaria) with small (suhma, Sanskrit suksma) (patches of light)'. 13 96 pariaram in vaivedhapariar am is apparently a third synonym for 'fence! As such pariara, from parikr-, 'to surround', would be typical of MIA. For the type of compound, see the note to Gatha 4 in Part II. 129 asunte, 'without listening'. anunTa is a nonsense-word 184 ann amanaha consists of anyamanas and aham. 315 gaapaiam stands for gataprad Tpam, but was misunderstood. 14 114 Read sirigovide, i.e. Radha. siri (Sanskrit srt) is found in the MSS of the Third South-Indian recension and in MS Bh. 240 amdhalaa consists of andha and -laa, for which suffix see Pischel $ 595. 15 For 488 Weber in his Ed. established the following text: annavarahakuvio vesattanavarahe jaha taha kalena gacchai pasan kuvia kaha tah pasaemi.

Page #97
84 Ein wegen andrer Vergehen Er zurnter wird wohl mit der Zeit irgendwie wieder gut. Wie soll ich aber ihn wieder gewinnen, der daruber zurnt, weil ich ihm feindlich gesinnt sei?! At first sight the text and translation seem in perfect order. No significant variants are available except for pesattana(varakena) in Bho, i.e. Sanskrit presyatva (but Bhuvanapala dvesyatva). In my opinion the text would in fact improve considerably by adopting the word pesattana and emending anna (anya) into ana (ajma). Then, instead of saying that there is an offence consisting of showing one's animosity and that there are other offences, the speaker (a woman) would now distinguish gradation: in offences of essentially the same order, namely between an occasional command (ajna) and treating someone as a servant, the victim being her husband. I should like to propose the following translation: Someone who is angered by being offended with a command always comes round again, even if it takes some time. (But) how will I be able to pacify someone who is angered by being offended because he was treated as a servant? 16 In this connection a few peculiar omissions in the Jaina-recension should be mentioned, in each case Gathas the contents of which may have been offensive to Jainas. Their absence shows the working of censorship. The following Gathas are left out: 120, 121 and 598 which mention a hunter, 175, 669 and 688 which mention a dog (175 also 'killing'), 431 which mentions the felling of a tree, and 405 which mentions a Navakavalia. Note in this connection also the absence of 548 which contains a word which, among others, means 'getting erected' (viabh amano). In the Vulgata this same word was changed into taggaamano. 17 It may be asked if these changes in both the Vulgata and the Jaina-recension, many of which appear to have come about by accident, do not at the same time indicate that the Gathas were originally not numbered.

Page #98
85 In any case they are not in the MSS of the Third South-Indian recension: 18 See also Gatha 76 in which asams ia in K, W, B, Y, P and S beside asaghia in Bh (cf. as adhia in R). 19 The present MS must have been compared with another MS, as appears from the many marginal additions. Occasionally mistakes are corrected: 419 voras avadin (cf. T), v vorasamghadin, 461 ramiavva (cf. Y), ramiavvaa. Other instances concern variants. In a number of cases these variants seem to have been drawn from a MS belonging to the K-version; 'e.g. 214 maida, v2 sunha (K), 289 tti paripucchdeg, tti ia pucch (K); in other cases from a MS like.y: 46 sunnai, v suvvai (v, B), 369 phaggucchana, 2 phaggucchava (Y, T). In some cases the source is not known: 474 nakkhana kaminio surae rehaith vi gananti. 20 The outward appearance of the text in Bh and R is practically the same. The MSS differ, however, on the following points. In R two Gathas of Bh are missing: 282 and 551. Bh 140 is found in R as 109. In R, Bh 382 occurs twice, namely as 14 and as 386. Furthermore, in R, Bh 353 and 354 have accidentally come to be combined into one Gatha: 358. In Bh the following Gatras of R are missing: 2, 101, 144, 304, 315, 414, 416, 548 and 569 (569 in R in the margin). Note that several of these Gathas, namely 101, 144, 304, 416 and 569, are not found in Ma, Ti and Tp either. 21 It is not clear how much importance should be attached to the addition of Ed. 552 (562 W, Y, 563 1, missing in Bh, Ma, Ti and Tp) in R in the margin after (?) 568 (Ed. 550: 560 V, Y, 561 , 562 Bh, 446 Ma, Ti, Tp). It is, for instance, not clear if the addition is to be taken as a correction for an initial oversight on the part of the scribe, or whether We have to do with a marginal addition made by a later reader of the MS. Note that the Gatha in question, like R 568, mentions a buffalo (mahisa).

Page #99
86 22 The present copy of S has numerous marginal additions which indicate that it was compared with one or more other MSS. E.g. 61 samaniain, sa samapiaith (cf. B), 50 jura, s2 juvara (cf. Bh, R), 120 maruena (cf. K), sa marueh it (cf. We B, Y, P), 184 manqusamuppannena (cf. K), s? mannusamuppaena (cf. , Y, P), and 441 a(9) ghissam (cf. Bh, R, Ti, Tp), accissan (cf. W, Y, T). 23 Saptasat Tsara with Bhavad Tpika of Vema Bhupala, along with the Chappannaya-Gahao (text and chaya), ed. by A. N. Upadhye, Shivaji University Sanskrit and Prakrit Series Vol. III, Kolhapur, 1970. The book 'has unfortunately not been available to me. See also Saptasattsara, ed. by D. V. Sitaramaswamy (with Telugu translation), Andhra University, Waltair, 1951, mentioned in Sriramamurti (1972: 68). 24 Apparently only the names were borrowed, as the Jaina-recension as represented by Bh does not show traces of the influence of the Vulgata on its text. 25 As for the two recensions no certain dates are available, theoretically also the reverse may be assumed, namely that they themselves were the first kosas thus arranged and that they set the standard to be followed in all subsequent kosas. It is unlikely, though, that they did indeed excercise such an influence, considering the rarity and geographical limitations of the occurrence of the MSS containing these two recensions.

Page #100
87 APPENDIX I Concordance of the Third South-Indian recension with Weber's Ed. and his MSS. The first column refers to the text of the Third South-Indian recension. The order of the Gathas is basically that of Tp, the only complete MS of this recension, Ma breaking off after 474 and Ti beginning only with 197. Furthermore, the order of the Gathas of the transcript Ma was considerably disturbed as a result of the confusion of the leaves of the original (see above, p. 23). The second column refers to Weber's Ed. The third column to the Jaina-recension as represented by R, the other Jaina MS, viz. Bh, missing the final Sataka. The next four columns refer to the vulgata MSS K, V, Y and P. The fifth Vulgata Ms, B, which contains only a very short fragment (1-129) is subsumed in the second column. MSS K and P are likewise fragments, containing 1-370 and 1-300 respectively. The eighth column refers to the x-recension, the only MS of which, namely x, misses the first and part of the second Sataka. The final three columns refer to the Telinga (T), Sadharanadeva (5) and Second Telinga-recensions (W), respectively.

Page #101
APPENDIX I Concordance between the Third South-Indian recension and Weber's recensions. Third S-I recension 1) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 2) 16 17 18 19 2 2 2 2 2 2 2 20 21 22 23 24 25 26 27 Ed. R(Bh) K 4 1 2 3 893 5 6 7 8 766 9 10 11 236 12 14 61 15 637 16 1 13 23 3 7 ~~3 1 615 9 10 11 186 2 5 56 667 567 co 667 8 8 240 12 12 14 14 65 60 15 15 16 16 13 13 17 17 17 19 19 19 767 616 624 619 680 626 42 45 . 683 627. 9 10 88 11 123 123 YPX TS 9 10 11 239 13 5 8 8 8 9 10 11 240 13 12 12 62 61 17 16 16 14 18 20 15 14 17 19 1 2 3 634 633 690 689 42 43 43 . 693. 692. 5 6 7 PRESELLER 22 9 10 11 239 240 12 14 61 15 16. 13 17. 19 43 1 7 5 501 148 1 7 8 480 643 692 646 173 233 40 238 442 479 455 292 590 209 284 313 427 388 203 26 663 653 611 261 356 389 139 263 367 489 506 13 89 146 30 600 26 487 380 488 276 322 w(v) 659. 565. 410. 1 2 3 4 1) In Tp only; Ti begins only with 197. 2) In R a second time as 386; in Bh missing here; only found as 382.

Page #102
Third S-I recension 28 . 821. 29 1) 30 31 23 365 A I I I I G 33 34 37 38 39 40 41 42 43 44 45 Ed. R(Bh) 46 47 48 49 50 51 2) 52 53 677 628 626 631 605 629 685 630 693 638 63 67 640 646 647 769 745 774 732 499 648 661 681 664 636 783 676 735 651 796 689 641690 692 637 647 697 653 809 . 699 704 701 K Y P XT S 64 49 20 687 686 636 635 615 614 695 694 704 702 64 63 377 646 89 686 377 645 1) Occurs a second time as 411. 2) In Ma after 55. 685 651 650 703 701 657 656 663 662 63 49 648 647 662 20 607 686 651 377 622 654 626 606 90. 170 115 84 369 138 236 423 500 187222229 + 186 335 424 511 222 390 266 671 593 459 392 136 215 408 124 147 94 617 90 23 649 659 658 651 661 660 652 662 661 54 49 53 50 49 57 55 20 20 21 20 174 56 564 613 574 574 87 57 21 21 21 22 21 21 488 175 58 . 270 222. 273. 273. 274. 273. 272. 594. 312 . 100 184 185 437 140 47 50 224 444 320 . - w(v) 14 64 82 32 22 23

Page #103
90 T S W(v) 173 272 322 36 Third S-I Ed. R(BH) K Y P X recension 59 . 24. 24. 24. 25. 24. 24 . 60 25 25 25 26 25 25 61 117 132 120 119 122 120 22 22 2223 22 22 26 26 26 27 26 26 27 27 27 28 27 27 28 28 28 29 28 28 29.29 30 30 30 725 611 31 191 190 192 198 194 32 197 196 198 31 33 31 32 31 32 33 35 33 34 33 33 855 18936 192 191 193 199 35 38 35 36 35 35 36 39 36 37 36 36 38 41 38 39 39 39 40 43 40 41 41 41 39 42 39 40 40 40 41 44 41 42 42 562 617 572 572 828 43 46 43 44 37 37 190 47 193 192 194 50 46 47 46 46 54 50 51 50 50 53 52 53 54 53 53 54 58 54 55 54 54 57 61 57 58 57 57 58 62 59 58 58 5963 58 60 59 59 92 . 60. 64 . 59. 61. 60. 60. . 632. 530 . 205 176 60 649 649 324 529 325 490 368 543 502 589 326 420 653 527 309 354 346 410 394 526 404 527 384 293 305 437 499 : 537 697 10 161 579 342 425 343 160 336 611 235 39 277 62 21 499 656 343 429 209 395 528 . 227 . 344

Page #104
Y P X T S w(v) Third S-I Ed. R(BH) K recension 93 . 65. 69. 62. 94 70 74 69 51 55 51 55 59 55 66 70 63 71 75 70 72 71 66 * 65 . 65 . 71 70 70 52 51 51 56 55 55 66 66 72 73 72 . .522 290 450 614 294 498 657 523 76. 854 101 76 75 75 74 75 74 102 103 296 561 239 353 695 79 588 333 534 602 278 194 647 157 657 589 601 242 104 105 139 16 106 107 82 129 81 84 86 81 132 80 83 85 81 132 80 83 85 108 53 355 109 110 75 79 74 74 78 73 913 81 85 80 127 * 142 132 80 84 79 838782 85 89 84 856 89 91 86 87 92 87 90 94 89 93 97 92 95 99 lac. 96 100 lac. 253 202 256 168 102 169 97 103 lac. 111 87 88 90 112 * '113 114 115 116 117 118 90 88 91 94 96 97 256 170 98 87 88 90 93 96 97 257 171 98 96 94 256 170 97 10 596 623 178 580 320 47 500 265 644 327 334 273 280 445 685 244 345 27 119 120 46 121 122 123 124 125 126 100 104 105 107 108 345 * 861 . 106 116 117 118 120 297 . lac. 105 106 108 109 347 . 102 106 107 109 110 349 . 101 107 108 110 111 349 . 99 105 106 108 109 28 421 487 415 193 348 583 . 361 . 330 454 588 458 389 542 . 8

Page #105
T P X T S W(V) . 284 285 112 113 . 114. 112 . 286 285 115 113 116 114 Third S-I Ed. R(Bh) K recension 127 . 700 . 124. 112 . 128 282 235 285 129 110 123 113 130 111 125 114 131 857 112 126 115 133 118 133 121 121. 124 123 136 126 129 140 130 137 128 141 131 138 135 . 413. 19 669 440 455 206 335 356 9 642 497 579 132 115 121 124 134 135 114 120 123 125 131 130 134 117 123 126 128 134 133 137 128 126 136 130 131 135 136 137 142 659 533 217 277 613 310 32 358 139 132 850 299 140 141 142 143 254 149 143 148 151 152 156 157 254 148 143 149 151 152 157 158 255 150 146 151 153 154 158 159 254 149 144 150 152 153 157 158 128 548 347 - 539 168 607 288 144 145 254 29 608 152.234 157 607 159 . 160 113 165 96 166 482 317 312 34 169 146 251 200 146 352 141 347 147 353 149355 150 356 155 360 156 361 852 310 253 159 364 160 365 162 367 163 368 165 370 167 372 164 369 166 371 147 148 149 150 151 152 153 111 477 90 201 314 160 161 163 164 166 168 165 167 314 161 162 164 165 167 169 166 168 314 162 163 165 166 168 170 167 169 161 162 164 165 167 169 166 168 154 155 156 157 173 175 177 174 176 448 453 511 510 578 491 43 158 159 160 186 390 189 188 190 189 6939 169 . 373. 171 * 171 * 172 * 171 . 178. 426 * 571 .

Page #106
Third S-I Ed. R(Bh) recension K U r P x T S W/v) 58 185 273 61 943 420 38 161 162 163 164 165 166 167 168 169 170 171 1) 172 173 174 2) 175 176 177 178 179 180 400 829 . . . . . . * 171 . . 174 378 176 176 178 177 183 484 215 179 380 181 181 183 182 188 634 216 175 177 177 179 178 184 447 514 176 379 178 178 180 179 480 177 154 180 179 181 180 186 662 180 381 182 182 184 183 189 481 181 382 183 183 185 184 190 301 388 182 383 184 184 186 185 191 347 425 183 384 186 185 187 187 193 172 394 696 388 708 705 138/696 192 108 195 194 196 192 197 552 618 193 127 196 195 197 416 456 195 109199197 199 194 199 513 248 198 147 202 200 202 202 319435 199 148 203 201 203 202 203 340 414 . 201 150 * 205 204 205 204 205 604 544 202 151 206 205 206 205 206 540 204 153 207 208 207 208 281 108 206 157 210 209 210 225 451 210 161 213 214 213 214 179643 207 158 211 210 211 210 211 111 152 444 446 455 455 482 61 562 938 612 223 173 228 226 227 226 227 330 211 162 215 214 215 214 215 122 216 166 220 219 220 219 220 567 299 2 18 168 223 221 222 221 222 401 - 459 219 169 224 222 223 222 223 433 31 : 221 . 171 : 226. 224. 225. 224. 225 . 339. 417. 208 209 210 181 182 183 214 184 185 186 187 188 189 190 191 1) Missing in Tp. 2) In Bh as 140.

Page #107
W(v) 70 - 194 57 199 35 203 523 381 205 254 Third S-I Ed. R(Bh) K W Y P x T S recension 192 * 225 . 175. 230. 228. 229. 228. 229. 463 . 460 . 193 229 179 234 232 233 232 233 251 228 178 233 231 232 231 232 436 250 195 231 181 234 235 234 235 450 503 196 232 182 236 235 236 235 236 492 638 197 235 185 239 238 239 238 239 129 220 198 240 189 244 243 244 243 244 185 243 193 247 246 247 246 246 33 126 200 246 195 250 249 250 249 249198 253 201 247 197 251 250 251 250 250 555 683 202 249 206 253 252 253 252 252 477 462 486 302 204 252 201 255 255 255 255 297 392 203 257 257 258 257 256 405 398 206 914 536 207 819 265 217 267 268 269 268 267 17 674 209 220 271 271 272 271 270 490 * 162 210 271 224 274 274 275 274 273 125 136 211 267 219 270 270 271 270 269 23 676 212 273 226 276 276 276 275 211 187 213 274 227 277 277 278 277 276 156 * 463 276 228 278 279 280 279 277 155 249 215 924 216 460 467 471 471 495 78 597 217 275 229 279 278 279 278 278 629 314 218 280 233 283 283 284 283 282 543 315 219 287 240 290 291 290 289 76 285 220 289 277 291 292 293 292 291 646 624 221 290 278 292 293 294 293 292 654 587 222 291 279 293 294 295 294 293 408 120 223 .292, 280 294 295 296 295 294 135. 53 208 268 214 1) Ti begins here.

Page #108
95 298 297 206 306 230 232 234 315 237 578 Third S-I Ed. R(Bh) K Y P X T S W(V) recension 224 294 . 282 297 . 297 . 298 . 297 . 296 . . 468 225 869 397 226 532 493 542 542 51980 119 227 295 283 298 298 299 298 469 228 302 245 306 306 304 541 316 229 297 285 300 300 301 300 299 409 470 298 242 302 302 302 300 254 651 231 300 244 304 304 304 302 107 219 304 247 308 308 308 306 286 233 312 255 317 316 316 314 174 274 313 256 318 317 317 465 235 316 259 321 320 320 318 364 106 236 849 296 578 572 588 588 580 683 68 238. 239 318 261 323 322 322 320 326 466 240 324 267 327 328 328 326 374 67 241 322 265 326 327 324 239 36 242 325 268 201 329 329 631 243 932 585 244 328 272 334 332 334 330 337 467 246 247 912 532 248 579 573 581 84 594 249 896 505 250 331 275 350 335 341 333 56 326 270 329 330 330 328 295 252 434 618 445 445 444 687 253931 582 254 255 332 286 335 336 335 334 467 472 28 256 333 110 336 337 336 154 452 75 * 257 334 287 337 338 337 1 335 473 258 . 563. 620 . . 573 573 . . . 694. 92. 251

Page #109
Third S-I Ed. R(Bh) recension K V Y P x T S W(V) 525 341 332 342 342 344 342 344 340 342 418 114 262 345 475 516 189 259908 . 260 338 291 261 340 292 910 263 341 293 264 342 294 265 359 314 266 343 295 267 347 300 268 351 305 269 348 301 270 349302 271 354 309 355 310 273 360 315 274 361 316 275 362 317 276 365 320 257 343 344 363 345 351 352 358 359 355 356 364 365 366 370 295 347 352 356 353 354 359 360 365 366 367 370 345 346 364 347 352 356 353 354 359 360 365 366 367 370 343 197 345 . 364 372 346 180 351 355 147 352 210 353 359571 360 365 375 366 187 352 301 272 250 464 262 367 370 418 - 478 277 278 1) 279 280 281 282 283 284 285 286 287 288 289 290 291 367 323 368 327 369 344 371 330 373 332 376 335 523 562 377 336 378 337 706 392 707 393 714 394 715 395 379 . 396. 372 373 374 376 379 382 533 383 384 385 386 372 373 374 376 379 382 533 383 384 385 386 372 332 421 373 531 393 374 468 * 486 376 331 423 378 378 396 381 640 546 576 564 305 382 236 357 383 323 440 384 620 319 385 395 51 386 119 169 17 387 574 191 388 . 406.483. 65 387 388 . 389. 387 1) Missing in Ti and Tp.

Page #110
T SW(V) . 389 61 .21 . 666. 44 689 274 359 70 95 327 51 395 396 390 392 394 396 397 398 401/373 402 130 397 311 371 454 121 267 301 404 89 303 206 207 406 581 268 408 80 413 Third S-1 Ed. R(Bh) KU Y P recension 292 .380. 397. . 390. 388. 293 688 699 697 294 381 398 391 389 295 383 400 393 391 296 385 402 395 393 297 387 404 397 298 388 405 398 299 389406 - 399 300 392 411 402 400 393 412 403 401 302 395 414 405 404 203 152 207 206 207 304 397' 415 407 406 305 400 418 410 409 306 404 . 415 414 307 407 418 417 308 711 504 412 419 423 423 310 311 416 424 427 427 312 417 425 428 428 313 689 634 700 698 314 429429 315 419 426 430 430 420 427 431 317 432 432 318 724 429 319 425 432 436 436 320 431 442 442 321 433 441 444 444 322 565 325 575 575 323 759 564 324 554 598 564 564 325 266 218 . 269 269 270 269 595 361 171 172 265 65 371 37 67 175 243 270 516 422 457 416 419 422 309 902 426 427 611 428 429 430 431 485 520 598 275 268 170 418 321 316 586 421 428 221 432 436 441 443 357269 116 165 578 292 88 609 551 297 637 557 449 498 . 18. 675. 463 54 454 268

Page #111
327 329 330 331 332 334 336 450 Third S-I Ed. R(Bh) K Y P X T S w(v) recension 326 353. 308. 354.358 . 358. . 358 . 626 . 300. 566 435 576 576 464 219 403 328 556 600 566 566 456 697 83 570 440 580 580 46969691 572 443 582 582 471 310 384 571 442 581 581 470 120 693 574 156 584 584 473 181 333 575 204 315 585 585 474 673 33 438 241 449 449 476 167 231 335 439321 450 450 477 108 217 445 447 456 456 547 615 559 337 458 461 461 485 577 322 338 779 659 598 45 339 452 444 463 463 487 519 525 71 340 451 459 462 462 486 625 323 341 454 461 465 465 489 71 100 455 462 466 466 490 2 2 343 457 468 468 492 645 492 344 468 472 472 496 417 493 1) 345 462 469 473 473 497 101 469 476 480 480 504. 694 91 2) 347 463 474 474 498 478: 194 348 472 479 483 507 655 532 349 480 485 485 508 350 482 486 486 510 276 373 351 935 476 483 487 487 178 69 353 691 636 702 700 605 698 17 354 488 488 512 150 181 355 478 486 489489 513 241 356 531 491 541 541 518 298 385 357 .534 . 496. . 544. 544. . 521 . 308 . 430 . 1) Ed. p.198: R 470; see, however, ISt., p. 162. 2) Ed. p.199: R 469; see, however, iSt., p.162. 342 464 461 346 470 483 603 474 475 352 511 477 485

Page #112
99 Third S-I Ed. R(BH) recension K T P x T S W(V) 358 359 360 . 536 . 498. 538 500 541 505 . 546 . 546. 548 548 551 551 . 523. 279 . 374 . 525 427 64 528 244 72 361 539 502 549 549 526 241 63 553 * 490 491 494 496 553 490 491 494 496 543 507 479513 480 514 483 515 485 522 782 663 484 521 487 524 488 528 489 529 492 531 481532 482 533 496 455 503 537 504 538 624 362 363 364 365 366 367 368 369 370 371 1) 372 2) 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 552 495 495 498 498 499 499 500 500 503 504 492 492 493 493 507 508 513 513 514 514 546 530 672 42 535 93 208 536 650 622 537 656 202 539 348 387 675 48 538 342 436 541 309 439 542 551 543 545 338 441 625 344 328 536 555 213 338 556 623 278 557 105 143 414 524 559 336 560 640 561 99 212 562 . 555 563. 584.541 488 - 72 848 515 515 558 31 505 539 720 540 506 541 507 543 508 546 509 548 . 510. 549. 516 516 517 517 518 518 519519 . 520. 520. 1) In Ti and Tp after 377. 2) In Ti and Tp after 371 and, a second time, after 372 (see previous note).

Page #113
100 Third S-I Ed. R(Bh) recension K W Y P T S W(V) 1) 388 389 390 2) 391 392 529 393 394 395 . 523. 523. 526 526 527 527 528 528 529 531 531 535 535 592 592 595 595 598 598 601 600 603 602 604 603 608 607 396 397 398 590 399 400 401 634 402 . 513 . 552. 516 556 517 557 518 519 558 521 560 525 565 582 566 585 577 588 580 585 592 589 593 593 597 717 389 610 517 609 495 612 544 617 591 871 618 619 595 621 602 626 631 627 526 760 574 643 692 839 644 693 . 654. 403 404 405 406 407 408 409 410 3) 411 412 413 414 415 416 417 620 619 619 618 622 621 627 626 495 . 566 . 382 * 615. 569 285 206 570 293 289 571 494 637 572540 290 574 481 525 578 575 519 584 218 587 685 75 590 315 443 592 169 610 594 63 566 595 131 144 629 162 190 676 . 674 474 494 678 345, 442 663 642 400 664 189274 597 379 600 52 567 647 115 84 615 199 441 497 627 54 568 240 92 . 604. 38. 628 629 631 636 637 627 628 630 635 636 653 652 628 654 653 . 664. 663. 1) Missing in Ma. 2) In Bh as 551. 3) In Ma and Tp also as 30.

Page #114
101 Third S-I Ed. recension R(Bh) K Y P x T S W(v) 418 . 419 .645. 695 656 658 655 666 668 654 665 667 . 644. 451. 665 152 264 420 421 670 669 6.71 670 674 673 679678 247 248 599 262 689572 639 407 688 439 247 286 507 37 15 426 248 247 348 111 133 116 132 119 135 117 133 140 373 8 139 460 245 649 501 74 421 421 142 132 233 66 416 420 . 414 417 415 419 418 263 422 660 423 661 424 664 425 669 708 244 * 194 427 428 747 112 429 114 129 430 130 * 143 431 741 432 433 410 512 434 837 435 405 436 408 437 792 673 438 422 583 440 428 442 834 443 429 444 437 549 446 550 568 447 551 571 448 553 586 449 559 606 1) 450 356 311 1) Missing in Ma. 72 409 420 419 387 418 433 423 433 434 439 433 434 439 434 441 434 438 117 164 229 107 445 92 440 440 448 448 559 559 560 560 561 561 563 563 569 569 361 361 439 447 366 449 83 98 450 429 513 451 595 453 459678 198 361/466 289353 357

Page #115
102 P x T S w(V) . 501 . 644. 306. 503 68 162 551 256 41 552 200 205 214 . 691 57 564 407 554 587 376 303. - 5 627 556 459 460 563 585 159 674 406 464 Third S-I Ed. R(Bh) K Y recension 451 466 473 . . 477 477. 452 468 475 479 479 453 497 456 508 509 454 498 527 509 510 455 832 456 615 582 625 624 457 501 535 511 511 458 502 536 512 512 522 561 532 532 524 534 534 461 577 570 587 587 462 583 567 593 593 463 586 578 596 596 843 465 589 584 599 599 594 592 605 604 467 595 706 606 605 468 602 612 611 607 545 617 616 470 614 581 624 623 620 596 630 629 472 263 215 265 266 267 473 279 232 282 282 283 1) 474 281 234 284 284 285 283 236 286 286 287 476 847 477 309 252 313 313 313 478 319 262 324 323 323 479 296 284 299 299 300 480 335 288 338 339 338 481 358 313 362 363 363 482 803 483 * 366 . 322 . . 371 . 371 . 1) After this Gatha Ma breaks off. 466 469 471 93 591 75 571 144 255 596 77 572 632 472 - 509 672 166 211 661 79272 598 666 266 265 69 * 86 282 281 65 284 283 60 636 286 285 251 340 271 311 559 317 321 14 298 257 354 336 431 474 363 370 259 121 148 . 371 . 110 . 141 . 475 679 44

Page #116
103 486 267 884 462 Third S-I Ed. R(BH) K Y P X T S W(V) recension 484 .440. 324. . 451 . 451 . 478 . 303 . 232. 60 485 749 328 260 845 487 370 329 375 375 375 333 422 488 152 3580 153 154 155 154 162 489 154 359 154.156 157 156 164 485 490 883 461 491 598 * 609 608 661 492 493 187 391 190 189191 191 196 402 482 494 386 403 396 394 395 242 370 495 567 436 . 577 577 67 496 817 497 498 830 190 499 823 106 .500 301 305 305 305 303 471 49 502 756 519 334 432 503757 520 433 504 842 505 109 506 86 90 85 87 86 86 192 243 507 859 359 88 93 88 89 89 89 328 509 99 105 lac. 100 100 100 27 665 103 104 105 106 104 453 120 123 125 125 123 662 582 512 133 136 135 138 136 143 228 369 513 143 349 145 145 148 146 154/641 207 38 514 145 351 147 147 148 156 87 608 515 1 84 385 187 186 188 188 194 330 835 230 501 255 824 508 115 510 511 446 516 517 . 196. 145. 198 . 200. 195 . 200 . 123 . 180 .

Page #117
104 w(v) 840 252 529 945 Third S-I Ed. R(Bh) K Y P x T S recension 518 * 197 . 146. 200. 199. 201 . 196. 201 . 403 . 457 . 519 215 165 219 218 219 218 219 643 298 520 220 170 225 223 224 223 224 329 415 521 226 176 231 229 230 229 230 398 391 522 233 183 237 236 237 236 237 491 639 523 277 230 280 280 281 280 279 349 329 524 305 248 309 309 309 307 284 186 525 526 306 249 310 310 310 308 535 308 527 170 374 172 172 173 172 179 245 612 528 217 167 221 220 221 220 221 25 109 323 266 326 327 326 325 103 140 530 430 441 441 294 115 531 841 532 573 111 583 583 472 101 151 533 633 534 441 326 452 452 479 260 355 535 520 559 530 530 496 641 858 358 898 510 538 868 * 385 539 941 617. 540 915 537 541 224 174 229 227 228 227 228 41 672 542 124 137 127 126 129 127 133 282 282 543 125 138 128 127 130 128 134 158 246 544 126 139 129 128 131 129 135 638 238 545 142 348 144 144 147 145 153 283 130 546 957 547 237 187 241 240 241 240 241 411 461 548 258 210 261 261 262 261 260 476 158 549 259 212 262 262 263 262 261 658 182 550 920 546 551 278 . 231 . 281 . 281, 282 . 281 . 280 . 127 . 137 : 536 537 5 2 59

Page #118
Third S-I recension 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 1) 577 578 579 580 581 582 583 Ed. R(Bh) K 921. 415 423 427 433 900 465 472 546 510 320 263 325 545 509 212 163. 293 281 307 250 838 950 872 891 876 862 890 255 207 903 825 885 864 918 919. 930 923 . 895. 216 296 311 258 1) Lacuna in both Ti and Tp. 426 438 476 5.56 324 105 258 Y P XT S w(v) 426 438 476 556 324 555 555 215 216 215 296 297 296 311 311 259 425 437 500 533 322 532 216 295 309 258 257 547. 473 489 514 504 270 362 204 15 670 141 199 134 661 383 529 226 350 237 651 412 486 432 365 483 40 517 133 464 367 484 163 544 545 576 549 503. 127 29 78

Page #119
106 K P x T S W(V) Third S-I Ed. R(Bh) recension 584 * 909 . . 585 894 586 56 60 587 . . . . . 526 . . 502 56 57 56 56 168 606 588 904 589 590 591 287 287 289 287 287 237 . -520 20 351 641 618 588 - 284 853 948 350 576 592 593 360 94 33 303 205 355 586 355 586 354 475 594 154 413 1) 595 1) 596 597 598 599 600 601 873 927 836 7983 32 34 44 48 45 49 78 32 44 45 80 33 45 46 79 32 44 45 95 79 31 44 45 602 603 604 863 47 51 47 569 232 596 518 - 533 125 553 682 366 201 - 337 151 695 523 177 240 145 390 690 376 431 352 210 157 229 52 605 606 607 608 609 610 611 612 613 614 615 52 77 72 74 73 73 78 82 77 79 78 78 952 84 88 83 85 84 84 91 95 90 92 91 91 98 104 lac. 99 99 98 916 . 122 . 135. 125. 124 . 127. 125. 538 .664 . 580 1) Lacuna in both Ti and Tp.

Page #120
107 w(v) 377 925 623 9 627 399 13 Third S-I Ed. R(BH) K U Y P x T S recension 616 101 . 113. lac.. 103 104 . 102. 506 * 346 617 109 121 110 111 112 110 18 618 865 619 106 119 107 108 109 107 164 179 620 119 134 122 121 124 585 621 131 338 134 133 136 134 524 534 622 556 137 342 139 139 142 140 148 288 234 624 140 * 346 142 142 145 143 151 325 167 625 139 345 141 141 144 142 150 55 593 626 144 350 146 146 149 147 155 597 629 870 628 161 366 162 163 164 163 171 195 264 629 205 155 209 208 209 208 209 630 148 354 150 151 152 151 158 102 214 631 946 635 632 153 3586 155 155 156 155 163 364 633 173 377 175 176 177 176 182 663 581 208 159 212 211 212 211 212 224 339 635 209 160 213 212 213 212 213 165 197 .636 937 213 164 217 216 217 216 217 143 155 638 222 172 227 225 226 225 226 231 347 639 238 237 238 237 238 153 640 238 188 242 242 241 242 241 242 699 641 926 642 250 199 222 253 254 253 253 12 673 643 256 208 259 259 260 259 258 466 644 257 209 260 260 261 260 259363 645 260 211 263 263 264 263 262 118 159 646 917 539 647 269 221 272 272 273 272 271 112 156 648 262 214 265 266 265 264 324 649 315. 258. 320. 319. 319 317. 246. 66. 55 634 605 637 19

Page #121
108 Third S-1 Ed. R(Bh) K Y P x T S W(v) recension 650 .288 . 276 . 290 . 291 . 292 . 291 . 290 * 67654 . 651 136 341 138 138 141 139 147 32 602 652 48 52 48 49 48 48 191 592 18 786 667 204 654 905 521 655 23 23 23 24 23 23 163 200 10 656 299 243 303 303 303 301 657 522 658 246 307 307 307 305 128 138 659 314 257 319 318 318 316 188 653 677 53 256 660 661 662 663 664 665 666 391 667 668 669 670 671 672 317 260 322 321 321 368 576 336 289339 340 339 247 351 901 515 382 399 392 390 98 160 344 296 346 348 348 347 146 604 384 401 394 392 393 619 311 933 591 .. 699 122 111 475 281 514 553 524 524 567 396 616 844 259 391 410 401 399 400 82. 374 333 380 380 563 304 396 420 406 405 * 216 39 897 509 399 417 409 408 218 402 334 381 381 545 772 643 25 413 421 424 424 423 182 362 432 409 443 443 442 568 648. 949 647 569438 579 579 468 85 575 453 460 464 464 488 91 372 . 512 . 551 . . 522 . 522 . . 565 . 636. 554 673 674 675 676 677 375 700 678 679 680 681 682 683

Page #122
109 K r P T S W(V) 470. 470 481 481 484 484 84 558 558 688 689 Third S-I Ed. R(BH) recension 684 .459. 466. 685 470 477 686 473 481 687 763 597 548 822 1) 690 530 691 493 692 831 693 587 579 2) 694 390 407 695 526 449 696 697 743 542 698 529 489 699 .616.590 494. 386. 614. 505 104 103 509680 195 570 448 149 698 97 517 652 548 249 375 212 589 208 378 399 566 307 381 531 540 * 504 540 505 597 400 536 597 398 536 62 539539 626. 625 . 684 508 271 516 648 620 . 666. 153. 696. 1) Lacuna in both Ti and Tp; Gatra has been identified with the help of the commentary. 2) Missing in Tp.

Page #123

Page #124
APPENDIX II 111 Concordance of Weber's Edition (1881) with the third South-Indian recension. Ed Ed 1. Ed 32. 601 33 72 34 - 35 75 36 76 63. 34 64 - Odvor own 66 97 94. - 95 114 96 115 97 118 98 613 99 509 100 120 1 616 11 2 - 12 13 14. 13 15 20 16 72 73 74 75 76 99 608 102 101 - 42 26 43 83 44 602 45 603 85 47 605 48 652 4954 50 86 3 510 4 121 5 122 6 619 7 123 124 9 617 110 129 11 130 132 17 .18 21 .. 78 609 79600 80 106 81 104 1922 55 57 83 59 60 63 107 611 108 506 111 508 110 53 87 54 88 55 96 56 586 57 89 58 90 5991 60 92 61 17 62. - 85 86 87 88 89 26 17 61 18 133 19620 120 511 21 134 22 615 23 135 124 . 542 112 30 67 31. 71 91 612 92 - 93. 113

Page #125
112 Ed Ed Ed Ed 125. 543 26 544 27 105 28 137 29136 130 430 621 159. 150 160 151 61 628 62 152 63 153 64 156 154 194 * 70 95 175 96 517 97 518 98 176 99 177 200 178 228 . 194 29 193 230 . 31 195 32 196 33 522 34 639 35 197 36 14 37 547 38 640 65 138 157 155 67 179 68 303 69 117 160 527 651 623 170 4 5 6 180 629 181 71 240 198 183 140 41 42 73 633 162 75 164 76 165 77 166 43 .44 199 426 625 624 142 545 513 626 514 141 143 630 144 78 47 201 79163 168 634 9 635 210 182 11 187 12 561 13 637 14 - 15 519 16 188 17 528 18 189 19 190 220 520 81 169 46 47 48 49 82 170 171 83 145 515 191 488 632 159 493 49 202 250 642 51 140 52 204 53 116 54 205 571 56 643 57 644 58 548 59 549 260 645 261. - 489 55 146 56 147 57 - 158 . - 8974 190 84 91 - 92 173 193. 174 638 23 186 24 541 25 192 26 521 227 . -

Page #126
113 Ed Ed Ed Ed 330. 250 32 262 .648 63 472 6465 208 325 1211 68 209 69 647 | 58 296 .479 97 229 198 230 199 656 300 231 1 501 2.228 3 658 4 232 524 255 256 257 480 | 662 | 210 39 . 526 340 261 364 .. 65 276 66 483 67 278 68 279 69 280 370 287 171 281 72 - 282 1672 676 1283 177 285 178 286 | 79 291 292 294 82 664 83 295 666 296 212 74.213 41 263 42 264 43 266 44 665 45 125 Xi Shi Dao 217 | 214 77 523 178.551 79 473 280" 218 1474 128. 475.." 590 563 8 . | 9 477 310 149 11 - 12 233 234 659 649 235 117 661 239 478 558 267 | 269 49 | 270 350 1593 51 268 494 297 326 54 271 55 272 56 450 57 . 58 481 241 290 219 | 650 | 220 1221 222 192 223 193 562 194.224 295.227 529 240 242 251 88 298 189 299 390 694 191 671 192 300 93 301 | 94 . 195 302 | 96 673 397,304 59 265 26 27 245 360 273 161 274 162 275 363 .. 28 244 329 ..

Page #127
114 Ed Ed Ed Ed 398 .. 99 675 400 305 432.679 133 321 252 500 .. |1 457 2 458 3 378 379 466 .451 67 68 452 69 346 470 685 71 . 348 686 349 75 350 444 334 335 484 534 39 440 2Nai Nai Yang Men Si 352 184 336 79 480 307 436 438 410 433 11 . | 12 309 | 13 678 14 . 553 16 311 17 312 18 314 315 316 354 355 365 366 375 376 367 370 1368 203 371 51 340 | 381 16 383 7 384, 18 385 386 510 387 11 . | 12 683 | 13 388 | 14 669 ( 15 . 16 389 17 390 | 18'391 | 19 392 520 535 21 393 22459 23 284 460 394 26 695 127 . 28 . 29 698 530 690 131 356 | 32 226 1533 .. 52 339 682 341 86 87 53 as 54 55 317 1342 56 439 440 490 91 57 343 58 319 59 554 441 684 160 216 61 344 345 163 347 164 * 465,556 62 93 691 94 . 95 . 96 377 97 453 198 454 499 . 39. 29 443 430 530 431,320

Page #128
115 Ed Ed Ed Ed 534 * 357 35 - 602 468 636. 40 568. - 69 681 570 329 36 358 331 38 39 640 41 - - - 45 359 362 38 39 540 41 42 43 73 74 75 330 532 332 333 594 461 360 404 43 44 414 416 403 364 44 237 405 248 45 46 47 48 560 557 - 688 16 470 456 699 406 408 49 445 83 462 446 447 51 19 409 52 53 54 53 48 417 448 396 463 693 397 465 620 471 21 410 22 - 55 - 324 56 419 328 398 58 399 411(/30) 660 422 59 449 400 412 466 467 62 81 63 258 64 56 65 322 66 327 567. 495 97 401 98 491 99 - 600 - 601. - 669 . 425

Page #129
116 Ed Ed 704 .. 1.677 | 287 7 288 738 . | 39 740 . 41 431 | 37 | 36 a Nai Nai Xing Dao Zhao Si wa 7933 anuNai Nai Xing Dao Zhao Si waw8w8858 13 . 32 293 502 57 503 89 313 690 91 92 46 33 353 700 127 | 2 . 703 .. 771. . 805 ..

Page #130
117 Ed Ed Ed . 840.525 41 531 874. - 75 - 76 568 42 8 - 9 * 49 43 464 45 670 486 476 908. 259 9584 910 262 11 - 12 247 103 14 206 540 16 614 17 646 18 579 19580 920 550 21 552 380 .49 850 51 52 490 492 575 17 18 192 236 139 . 148 591 100 73 109 131 536 507 55 56 23 28 689 499 505 890 91 92 570 567 - 5 23 582 215 622 641 598 58 59 860 585 126 581 82 161 61 62 569 604 576 63 253 243 498 65 618 692 667 455 583 96 249 97 674 98 537 99 - 900 555 1 663 2 310 3 572 4 589 5 654 6 657 907 . 607 351 37 636 34 442 35 500 36 599 37 434 * 38 564 839. 415 68 69 225 870 627 71 407 72 566 873. 597 38 185 39 - 940 941 . 539

Page #131
118 Ed Ed Ed Ed 952 . 610 53 942. - 43 167 44 - 45 533 946. 631 947. - * 48 592 49 680 950 565 951. - 957. 546 58 .59 - 960. - 54 - 55 956. - -

Page #132
119 APPENDIX III For an explanation of the purpose of this concordance, see p. Third S-I recension Vulgata (Gangadhara-version) 2 3 4 5 6 Jaina-recension (R) 1 7 236 19 616 619 626 45 627 628 631 . . . . . . . 626.

Page #133
120 Third S-I recension 30 . 1 Vulgata (Gangadhara-version) 2 3 4 5 6 . . . . . Jaina-recension 7 (R) . 626. 631 605 629 685 630 693 638 67 63 640 646 647 499 648 661 681 664 636 676 651 689 641 690 692 647 653 637 697 649 651 652 699 704 702 53 . 20 564 613 270 222 24 117 22 * 26 . . . . . . .

Page #134
121 Jaina-recension Vulgata (Gangadhara-version) 2 3 4 5 6 . . . . . 7 Third S-I recension 1 . 27 28 (R) 64 . . 562 190 66 . . . . . . .

Page #135
122 Third S-1 recension 1 Vulgata (Gangadhara-version) 2 3 4 5 6 7 Jaina-recension (R) 75 98 99 72 127 1 253 168 103 106 116 104 105 117 107 108 118 120 297 345 124 282 110 235 123 125 111

Page #136
Third S-I recension 132 33 34 * dii 11 1 1966 19:1 6 1.16199 nn 999 . 9 : ng 35 36 37 38 39 140 42 43 44 45 46 48. 47 150 52 53 55 56 57 58 59 160 61 62 63 -64 165 Vulgata (Gangadhara-version) 1 2 3 4 5 6 7 . 112. 118 121 123 129 128 132 146 141 147 149 150 155 156 159 160 162 163 165 167 164 166 186 169 174 179 175 176. 123 251 310 Jaina-recension (R) 126 133 136 140 141 200 352 347 353 355 356 360 361 253 364 365 367 368 370 372 369 371 390 373 378 380 379

Page #137
124 Third S-1 recension 166 1 Vulgata (Gangadhara-version) 2 3 4 5 6 . 177. . . . . 7 Jaina-recension (R) . 154 68 180 381 382 181 182 383 183 696 384 388 108 127 192 193 195 109 198 199 201 147 148 150 151 153 157 202 204 206 210 207 161 158 446 444 223 211 173 162 166 216 218 219 221 168 169 171 175 179 225 229 228 178 231 181 232 235 182 185 240 189 . . .243 . . . . . 193

Page #138
125 Third S-I recension 1 Vulgata (Gangadhara-version) 2 3 4 5 6 7 Jaina-recension (R) 200 . . . . . . . 246 . 247 195 197 249 206 486 523 201 252 254 203 265 217 220 268 271 224 267 219 226 273 274 227 228 276 460 467 229 275 280 287 289 290 233 240 277 278 279 280 282 291 292 294 532 493 295 283 302 297 298 245 285 242 300 244 304 . 312 . 247 255 . . . . . .

Page #139
126 Third S-I recension 1 7 Jaina-recension (R) Vulgata (Gangadhara-version) 2 3 4 5 6 . . . 313 . . . 316 234 . . . 256 259 578 572 318 261 324 267 265 322 325 268 328 327 272 271 579 573 331 275 326 270 434 618 286 332 333 110 334 563 338 340 341 342 359 343 . 347 . 293 294 314 295 300 . . . . . .

Page #140
127 Third S-I recension 268 . 1 7 Vulgata (Gangadhara-version) 2 3 4 5 6 . . .351 . . . 348 Jaina-recension (R) . 301 305 69 270 349 354 302 309 310 355 360 361 nReng Reng Jiao Si Si Si Si Si 315 316 317 25 320 367 323 368 280 327 344 330 369 371 373 376 332 335 562 523 377 378 336 337 392 393 394 290 191 379 380 395 396 397 688 381 383 385 398 400 402 404 405 406 387 388 389 392 .393. 300 411 412 301 . . . . . .

Page #141
128 Third S-I recension 1 7 302 . Vulgata (Gangadhara-version) 2 3 4 5 6 . . . . 395 . . . 203 397 400 404 407 Jaina-recension (R) . 414 152 415 418 412 416 417 689 418 419 420 421 426 427 428 429 432 425 431 433 565 554 266 353 566 556 570 572 441 325 564 598 218 308 435 600 440 443 442 156 204 241 321 571 574 575 438 . 439. . . . . . .

Page #142
129 Third S-I recension 1 7 Jaina-recension (R) Vulgata (Gangadhara-version) 2 3 4 5 6 . . . 445 . . 450 336 . . . 447 458 659 444 459 461 462 452 451 454 455 457 461 462 469 464 468 469 476 470 463 479 472 474 480 482 475 476 483 691 636 485 477 478 486 491 531 534 496 498 536 500 538 541 505 539 502 543 507 513 514 479 480 483 485 515 522 663

Page #143
130 Third S-I recension 370 1 7 Jaina-recension (R) . . Vulgata (Gangadhara-version) 2 3 4 5 6 . . . . . 484. . 487 488 489 492 481 482 496 521 524 528 529 531 532 533 455 537. 538 503 504 505 539 506 507 540 541 543 546 508 509 548 510 549 513 552 556 516 517 3 557 518 519 521 558 560 565 566 525 582 585 577 588 580 585 590 592 589 593 593 597 389 517 . . . . . . . 610,

Page #144
131 Third S-1 recension 404 1 Vulgata (Gangadhara-version) 2 3 4 5 6 . . . . . Jaina-recension (R) . 7 : 609. 612 617 495 544 591 618 619 621 626 595 602 627 526 574 692 643 644 693 654 695 645 656 658 660 661 664 669 708 244 194 114 130 112 129 143 410 512 405 408 . . . . . . . . 673

Page #145
132 Third S-I recension 438 Vulgata (Gangadhara-version) 2 3 4 5 6 Jaina-recension (R) 1 7 . . . . . 409. . . 39 422 583 430 440 423 428 434 429 437 549 550 568 570 551 586 553 559 356 606 311 473 475 466 468 497 498 456 527 615 582 501 502 522 524 535 536 561 563 570 567 578 577 583 586 589 594 595 584 592 706 602 545 607 614 . 620 . 581 . . . . . . 596

Page #146
133 Third S-I recension 472 1 7 Jaina-recension (R) . Vulgata (Gangadhara-version) 2 3 4 5 6 . . 263. . . 279 281 . . 73 215 232 234 236 252 319 262 284 288 335 358 313 366 322 324 440 328 370 329 152 154 187 391 386 403 436 190 106 230 301 519 520

Page #147
Third S-I recension 506 7 8 9 510 tth dd -- tth tth dd -- 2 8 klu 8 m C/&8sesnum-ttskysk88rky 11 12 13 14 17 19 520 22 23 27 28 29 530 31 32 33 34 35 37 38 539 Vulgata (Gangadhara-version) 1 2 3 4 5 6 7 86. 88 99 103 120 133 143 145 184 196 197 170 215 220 226 233 277 134 217 305 306 323 430 441 573 520 Jaina-recension (R) 90 93 105 115 349 351 385 145 146 165 170 176 183 230 248 249 374 167 266 223 111 326 559

Page #148
135 Jaina-recension Third S-I recension 540 1 Vulgata (Gangadhara-version) 2 3 4 5 6 7 (R) 224 174 124 137 138 125 139 126 142 348 237 258 187 210 259 212 278 231 415 427 423 433 465 472 546 510 320 263 545 509 163 212 293 281 307 250 255 255 207 573 . . . . . . . .

Page #149
136 Third S-I recension 574 1 Vulgata (Gangadhara-version) 2 3 4 5 6 7 Jaina-recension (R) 56 284 237 350 303 205 576

Page #150
Third S-I recension 608 9 610 11 12 18 19 2 2 2 2 2 2 2 8 ---- 620 21. 23 24 27 28 630 32 640 641 1 2 73. 78 8888 84 91 98 Vulgata (Gangadhara-version) 122 101 109 106 119 131 137 140 139 144 161 148 153 173 3 4 5 6 7 205 137 208 209 213 222 234 238 Jaina-recension (R) 77 82 88 95 104 135 113 121 119 134 338 342 346 345 350 366 155 354 358b 377 159 160 164 172 184 188

Page #151
138 Third S-I recension 642 1 7 Jaina-recension . (R) Vulgata (Gangadhara-version) 2 3 4 5 6 * 250 . . . 256 . . 199 44 257 208 209 211 260 269 262 315 221 214 258 276 341 288 136 48 52 299 303 246 314 257 317 260 336 289 382 344 399 296 401 384 699 122 514 553 410 391 374 333 396 420 . . . . 399 . . . . 417

Page #152
139 Third S-I recension 1 Vulgata (Gangadhara-version) 2 3 4 5 6 . . . 375 . . . 7 Jaina-recension (R) 676 . . 334 643 77 413 421 79 432 409 569 438 453 512 460 551 466 459 470 473 477 481 597 548 530 690 91 490 534 493 587 579 390 407 449 526 542 529 489 . . . . . . . . 616 . 590

Page #153

Page #154
PART II

Page #155

Page #156
CHAPTER 1 1.1 The nature of the text The SattasaT is a compilation of 700 (saptasata) Gathas, or arya-verses, in Prakrit. Their content is erotic, depicting, for instance, lovers' attempts to meet, various ways of sexual intercourse, unfaithfulness, lovers' quarrels and their reconciliation. The situations in the majority of the Gathas are set in the country-side, descriptions of city and court-life being conspicuously absent. The people depicted are those inhabiting the country-side such as farmers, hunters, tribals and warleaders charged with the protection of the villages. The Gathas are monologues, mostly spoken by, or describing, these peoples' women. The compilation of the text is traditionally attributed to a certain king, named Hala, to whom I shall return later. This and the fact that the Gathas are composed in a language which, but for its Middle Indic appearance, is basically a form of Sanskrit, and in a style which is highly literary or Kavya-like, suggests that the text was intended for a cultured audience of 'connoisseurs'. This conclusion, which is amply borne out by internal evidence, for which see below, raises the question what interest this audience could have had in a text which basically describes life and in particular love and sex in an environment so completely different from that of their own. Most scholars in fact seem to recognize in the Gathas a kind of romantic attitude on the part of the poets and the audience towards life in the country-side. Thus, according to Weber the text was the product of 'feingebildeter, wo nicht uberfeinter, Dichter aus der Stadt, die ihrerseits das Landleben, wie die Poesie der Natur, die Liebe speciell, verherrlichen' (Ed., p. X). Keith maintains that 'the prevailing tone is gentle and pleasing, describing simple loves set among simple scenes, fostered by the seasons, for even winter brings lovers together, just as a rain-storm drives them to shelter with each other' (Keith, 1928: 223). Compare, finally, Warder according to whom the life in the villages depicted in the Gathas 'was full of hardships, but not without compensating pleasures, especially those of love' (Warder, 1:974: 182). As I will show below these views are widely off the mark. For instance, the people in the villages are almost throughout depicted

Page #157
144 as simpletons and misfits, especially where sexual matters are concerned, while the tone is in fact far from gentle, but rather superior and patronizing. In order to clarify this some of the situations found in the Gathas will be considered more closely below. This should at the same time give a clearer picture of the circumstances, ideals and ambitions of the audience for whom the text was evidently intended. 1.2 The situations A recurrent theme in the situations of the Gathas is the foolishness of the people depicted. Thus, the Pulindas, tribals living in the Vindhyamountains, mistake dark rain-clouds for elephants (116). Frequent references occur to the pamara, a poor farmer." In 238 he barters his cloak for an ox because he is fooled into thinking that his wife's dark breasts, which remind him of heaps of smouldering ashes, will keep him warm during the winter. His stupidity furthermore consists of acquiring an ox when the ploughing-season is still far off. In fact his foolishness is almost proverbial. Compare Gatha 643: kamalan muata mahuara alekkhaladduam pa- darapikkakavitthagandha lohena maro vva chiviuna janihisi.5" Bee, if you abandon the lotus for the kapittha, attracted by the fragrance of their as yet half-ripe fruits, you will soon find out (your mistake) after you have touched them, like a pamara touching a sweetmeat in a painting. The halia or 'ploughman' is equally foolish. Compare Gatha 360: ajjan mohanasudhia daraphudiavetabhara- maa tti mottu palaie halie harani hasian va phalah Thi. When the ploughman ran away from his genteel wife (or mistress), who had fainted from love's extasy, thinking her dead, the cotton plants seemed to laugh with the lips of their bulky fruits slightly opened.

Page #158
The people depicted in the Gathas have in common that they need to work in order to make a living, which leaves them little time and energy for love. Their women suffer accordingly. A case in point is Gatha 324: cikkhillakhuttahalamuha appattamohagasuha 145 kaddhanasudhie paimmi pasutte ghapasamaan pamar savai. The poor farmer's wife who failed to reach her climax as her husband had fallen asleep exhausted from drawing the ploughshare through the thick mud, cursed the rainy-season.9 On this theme several variations occur. For instance, the hunter, having married recently, discovers that he is no longer able to draw his bow, weak from making love with his wife. He has to shave the bow thinner. See, for instance, Gatha 120: ukkhippai mandalina- ruega gehangatahi vahTe sohaggadhaavadaa v- va uaha dhanurufparicho1T. Look at that! From the courtyard of the hunter's house a whirlwind throws up a line of shavings from his bow, which looks like a banner advertising his wife's (conjugal) happiness. ekkallamao divaapiajaassa jaha dhanush As a result the arrows will, of course, lose much of the required force. Furthermore, love has made him weak-hearted. He comes to lack ruthlessness which is necessary in order to kill. Compare Gatha 620: mate taha lalio saanhae padiah vahassa hatthao. 10 The stag wandering alone was fondled by the deer in heat, used as a decoy, in such a way that the hunter, who had a loving wife himself, dropped his bow. As a result the hunter's family is threatened by poverty and starvation.

Page #159
146 In Gatha 951 a trader in skins has to be sent away empty-handed as skins are no longer available in the hunter's house after his marriage. The scars on the chest of the war-leader, contracted in performing his duty of protecting the village, make him an unfit lover as his wife cannot fall asleep on account of their roughness (see Gatha *31). In Gatha 687 the husband's victory in a wrestling-contest indicates his lack of interest in his wife: anard amtena tumam malli na lajjasi naccasi paino pahaena padahas addena dohagge paadijjarte. . Woman, shouldn't you be ashamed instead of dancing, as the drum beaten to celebrate the victory of your husband in a wrestling-contest proclaims you unhappiness. A large number of Gathas concern the itinerant merchant (pahia) and, his wife. They deal invariably with the husband's absence from home, actual or imminent. This is a common theme in Indian erotic poetry. In this text these Gathas provide yet another example of the above-mentioned theme, namely of the impossibility to combine work and love. The travels of the merchant are for him a necessary means of earning a living. He is reluctant to leave his wife (see Gatha *25). During his travels he has to undergo numerous hardships including a ruthlessly hot climate. Furthermore he misses his wife. See, for instance, Gatha 641: agghai chivai cuovai jaakavolasaric thavei hiaammi janiaromasco chi e pahio mahuapuppham. The traveller smells at the madhuka-flower, touches it, kisses it and presses it against his heart, while all the time his hair stands on end because of its resemblance to his wife's cheek. The majority of Gathas concerned depict the fate of his wife and especially her utter unhappiness during his absence from home. Not all details need be treated here. One example (670) should suffice:

Page #160
147 pariosasundarain tais cia una virahe suraesu lahamti jai sokkhai khaugginnai k Tramti. The pleasures of love-making, which gave so much satisfaction, are during his absence turned into stale memories. Other Gathas deal with farmers, notably gahavais and halias. The gahavai or 'lord of the house' seems to be a member of the landed gentry. The halia or 'ploughman' is apparently a small farmer who like a sharecropper partly makes a living by hiring out his labour. Despite the obvious differences in their social positions, they are seen to exchange brides. The gahavai gives his daughter in marriage to the halia, thus placing the girl in a very difficult position as appears from, for instance, Gatha 602: mamdan pi na anai hali- gahavaisua vivajjai anamdano iha hi daddhagamammi avejjae kassa sahamo. The ploughman does not understand (a thing, not) even slowly (or: does not even know what illness is). His genteel wife is dying but who can we tell, for in this wretched village there is no healer for her ills. The gahavai's daughter is complaining in Gatha 164: vaink a ko pulaijjai kena samam va hasijjai kassa kahijjai suham va dukkham va pamar apaure haaggame." Who is there to glance at, who is there for me to talk to about my pleasures and my griefs, who is there to have fun with, in this dreary village full of yokels? After her marriage the halia's daughter, coming from a simple family, encounters problems in her new environment. One thing she lacks is coquetry." In order to attract her husband's attention she behaves rather clumsily. E.g. Gatha 107:

Page #161
148 golaadatthian pec- adhatta uttarium chiuna gahavaisua haliasonha dukkhuttarai paavte. When the farmer's daughter saw her husband, the son of the householder, standing above the sloping bank of the Godavart, she began to clamber upwards along a (steep and) difficult path. 16 The gahavai thus marries beneath his social position, while for the halia the marriage with the gahavai's daughter means a step upwards along the social ladder." However, the gahavai also marries his daughter to the gaman I or war-leader. For the gamant this marriage proves a misalliance as his wife is unable to fulfill his ambitions, which consist, among other things, of having her commit Satt at his funeral. Thus, in Gatha 407 she, however inadvertantly, spoils this ceremony by sweating and thus extinguishing the fire. It should be noted that in the context of lyrical poetry this sweating is in itself quite legitimate: vijjhavijjai jalano anumaranagh analingia- gahavaidhui vitthaas iho vi piaamasuhasijjirang Te. 19 The gahavai's daughter extinguishes the fire, its flames already shooting up high, as she sweated from the pleasure of embracing her dearest one at the occasion of following him in death. In this Gatha the gamant is not specifically mentioned. In the following Gatha (449) he is, again at his funeral which was meant as a spectacle during which all his wives were expected to commit Satt. This time the SatT-ceremony fails because the villagers among whom he is living are unable to appreciate the ceremony, having only eyes for his favourite wife: gaman ino savvasu vi mammaccheesu vi val piasu anumar anagahiaves asu lahai uvar in valai ditthT. Though all the war-leader's wives are beautifully

Page #162
149 attired for the occasion of following their husband in his death, at the most painful moment all eyes are directed towards his favourite. Most of the other situations are common to Indian erotic poetry in general. However, they agree with those discussed above in that they likewise deal basically with the impossibility of lovers to have a satisfying relationship. Before marriage a girl is closely watched by her parents who do not allow her to go out. She peeps through the holes in the fence at the boys passing her house but experiences considerable frustration as they do not respond. If she succeeds in attracting one boy's attention and they decide to meet, they have to do so in the fields (9) or in the dense shrubbery along the river (103). All this, however, provides hardly a romantic experience. The lovers are in fact constantly worrying about the fields which will one day be harvested or about the shrubbery which will in time become less dense. The rainy season is supposedly the ideal time of the year for meeting in secret because then the nights are really dark (249). However, the paths leading to the place of meeting are slippery and dangerous in that season (231, 445). * Married life is likewise represented as full of problems. After the marriage ceremony the girl is taken by her husband to his own house where he lives with his parents, his brothers and their wives in a joint family. This living together is dictated by economic circumstances. The woman is not supposed to entertain extravagant wishes, which would put too great a strain on the resources of her husband's family. This appears to be difficult when she is pregnant and suffers from cravings for rare and expensive things (dohada) (172). The woman is expected to restrain herself in every other respect as well, thus becoming a kulavahu. Sex itself fails to provide real pleasure. For instance, the woman soon discovers that her husband is more quickly exhausted than she is, leaving her unsatisfied (155, 413). When she wants to continue, taking the initiative (viparitasurata), she merely annoys her husband (see *23). A considerable number of Gathas deal with the husband's unfaithfulness which in the first place causes great grief to his wife. She takes revenge by sulking (mana), causing, in turn, considerable frustration

Page #163
150 to her husband, who in the end is forced to throw himself at her feet in an attempt to propitiate her. 1.3 The audience It will be clear that the picture given of the life in the country-side is far from flattering. The text would thus most likely have been intended for people who were not themselves depicted in the Gathas and who, intelligent and living in completely opposite, i.e. better circumstances, could afford to have an interest in the situations described. This audience must have belonged to the cultured and refined leisure-class living in the cities and towns, which brings to mind the so-called Nagaraka as described in Kamasutra I 4. The audience must also have included kings, as :) the ideal warrior, which appears from the descriptions of the gaman1. For, these seem to have been intended for a more succesful member of the warrior-class, namely one who has no problems in having performed the proper rituals and who, furthermore, has established peace, thus having time and energy for things other than waging war, namely love and, as the existence of this text serves to show, enjoying literature. It is evident that the text served in the first place to amuse the audience, which raises the question as to what the humour consisted of. It would seem that it was in any case based on a certain measure of contempt for the people living in the country-side. The main purpose of the text, though, would seem to have been that of providing an outlet for the audience's own frustrations, presumably caused by the painful but inevitable clash of their ideals and ambitions with the realities of life. The text may have acquired this function in that it projects the audience's problems onto a completely different world with different people and different customs. In this way these very problems could become objects for amusement. Whatever the precise nature of the perception of the text may have been, it contains numerous specimen of often crude humour, which is otherwise extremely rare in Indian erotic poetry. An instance of this humour may be found in Gatha 324, quoted above (p. 145), where the farmer's exhaustion which makes him unable to satisfy his wife's desires is attributed, of all things, to his having dragged the plough through thick wet mud. Another example may be

Page #164
151 found in Gatha 360 (see above, p. 144) where the farmer is surrounded by slightly opened cotton-fruits which seem to refer to the state of his mistress from whom he is trying to run away. 1.4 The compiler As already mentioned above the compilation of the SattasaT is attributed to a certain king, called Hala. This Hala is traditionally identified with a satavahana king of that name, who is supposed to have ruled for a brief period in the beginning of the first century A.D.C4 Compare verse 13 of the introduction of Bana's Harsacarita, which seems to contain the earliest reference to the Sattasaf (seventh century A.D.): avinasinam agramyam akarot satavahanah (v.1. salio) visuddhajatibhih kosam ratnair iva subhasitain. Satavahana made an immortal and refined kosa with Subhasitas consisting of pure characterizations, as if it were a treasury filled with the purest pearls. It is commonly assumed that Bana refers to the Sattasas here. This assumption finds support in his use of the word agramya, which with reference to the Sattasaf acquires a double meaning, referring to the subjectmatter of the text, which is village-life (grama), on the one hand, and to the result, which is a kosa for the cultered city-man (agramya), on the other. The ascription of the compilation of the text to a Satavahana is supported by internal evidence. For instance, the Gathas describe precisely the realm of this dynasty, namely the area roughly coinciding with the present state of Maharashtra. Thus, they often refer to the Vindhyamountains (70, 115, 116, 117, etc.) and to the rivers Narmada (549, 760) and Godavart (58, 103, 107).25 The Satavahanas are indeed traditionally represented as patrons of Prakrit literature. It is on the other hand also possible that this tradition was in fact based on their association with the Sattasaf. In this connection it is interesting to note that several of the names of the 'authors' which in some MSS are found with

Page #165
152 the individual Gathas belong to kings of the Satavahana dynasty and their successors, the Vakatakas and subsequently the Rastrakutas. There are reasons to assume that the names were added to the text in a relatively late stage in its transmission (see above,p. 76 ff.). Their presence does, however, show the persistence of the traditional association of the text with Maharashtra kings. 1.5 The language of the Gathas The Sattasai is generally assumed to be the earliest text showing the use of a Middle Indic or Prakrit dialect for purely literary purposes." This particular dialect later became known as Maharastri. Initially Maharastri was closely connected with the erotic-lyrical genre to which, the Sattasai belongs. Thus, it is found in the dramas, in which those characters who otherwise speak SaurasenT Prakrit, used Maharas trt for their verses. These verses concern mainly descriptions of nature and its effects on the characters' moods. The practice of using such Maharastrt verses was begun by Kalidasa. The fact that verses in Maharastrt were absent from the earliest dramas by Asvaghosa suggests that they are indeed innovations in the work of Kalidasa. Subsequently, however, Maharas tri developed into just another literary language beside Sanskrit; it came to be used in epic Kavyas (e.g. Setubandha and Gauqavaho), while the Jains used a mixture of Maharastrt and Ardha-Magadht for their stories. In this connection it may be asked what determined the use of a Prakrit dialect for literary purposes in the first place. In the case of the SattasaT its use is mostly taken for granted, the text having been compiled, if not composed, by a king of the Satavahana dynasty, which for their inscriptions and coins used a Middle Indic dialect rather than Sanskrit. However, the fact that the Satavahanas used a Middle Indic dialect for administrative purposes does not mean that they were ignorant of Sanskrit or that they would not have had the possibility of choice between Sanskrit and Prakrit." Presumably they merely did not consider the use of Sanskrit for administrative purposes." I would therefore like to put forward the following alternative and in many respects speculative explanation. Above it has been argued

Page #166
153 that the Gathas were intended for the amusement of a rich and cultured audience. The humour was at the cost of the simple and somewhat foolish people inhabiting the villages of Maharashtra. The Gathas themselves are mainly monologues spoken by these villagers. In this connection the use of a Prakrit dialect may well have been determined by the intention to give their words a rustic and local flavour. This explanation presupposes pride in the knowledge of Sanskrit on the part of the audience as well as of the authors." The above view seems at least partly to be supported by the way Prakrit dialects were used in the dramas, providing an indication of the character's status being lower than that of one speaking Sanskrit. -

Page #167
154 Notes to Chapter 1 It goes without saying that this observation is based only on the Gathas that can positively be traced back to the MS of the archetype, i.e. those that are found in the Third South-Indian recension, on the one hand, and the Jaina-recension and/or the Vulgata, on the other. Nevertheless there is a small number of Gathas, which can be traced back to the archetype as well, that appear to be exceptional, namely those which mention prostitutes and thus seem to describe life in the city (e.g. *21, 156 and 474). However, their number is so small that they need not affect the argument. Exceptionally a Gatha contains a dialogue, as, for instance, in 384, 676 and 950. This example and the following contain instances of the so-called Bhrantimadalamkara. Other instances of this figure of speech are found in 531, 532 and 640. The word pamara literally means 'having a skin-disease (paman)'. In Gatha 331 the man is indeed scratching his head. Pada bas found in Ti, Tp and T. For alekkhaladduan Ti and Tp have alekkhe ujja (unmetrical); compare alokhkan ujjaya in R and alekkha ujjua in X and S. Perhaps the text should read alekkhe ujjam, in which ujjam represents Sanskrit urjam 'refreshment'. The reading is based on Ma, Ti and Tp. bharaharahi consists of bhara and adhara.

Page #168
155 For ajja 'genteel woman', from Sanskrit arya, see Thieme (1938: 99-100). For the halia being married to a woman of a higher social position, see below. sudhie as found in Ma, Ti, Tp, Bh and R. Note that the rainy season is otherwise associated precisely with conjugal happiness. Yet another point in the Gatha is the stupidity of the woman. Though she curses the rainy season it is, with the harvesting season, the only time of the year when her husband has ample work and consequently enjoys income. 10 The text is based on Ma, Ti, Tp, Bh, R and T. 11 This Gatha is absent in the Third South-Indian recension. 12 . saricchide as found in Ma and Tp. 13 See Fick; 1897: 164. 14 The text is based on Ma and Tp. 15 This applies to the simple villagers in general. Note the complaint in the preceding Gatha: vamk a kopulaijjai...pamar apaure haaggame. 16 The meaning of the Gatha becomes clear only from a comparison with 193:

Page #169
156 golavisamoarac- anuampan iddos am chalena appa urammi se mukko tena vi sa gadham uvaudha. Under the pretext of going down to the Godavart along a rough path she throws herself against his chest, and he embraces her tightly which seemed innocent as he did so (ostensibly) out of pity. 17 The gahavai gave his daughter in marriage to the halia to secure his labour during the ploughing and harvesting seasons, thus sacrificing his daughter's happiness for his need for labourers. 18 The gahavai and gamant, though obviously representing two different life-styles, would seem to be socially equal. In any case the gahavai. will need the services of the gamant to protect his interests and where necessary to enforce his rights. To be certain of the services of the gaman t the gahavai apparently has to establish a relation with him through marriage. 19 alingia as found in Ma, Ti, Tp, T and S. 20 My interpretation of this Gatha differs considerably from Weber's, according to whom it is the eyes of the gaman which even after his death were drawn to his vallabha. On the other hand, the ceremony is a clear instance of potlatch. Note that the women are especially dressed up for the occasion (anumar anagahiavesasu). Therefore ditthi should rather 'refer to the eyes of the spectators. Their special interest in the vallabha lies in the fact that when they want to ask a favour of the gamans they will usually address him through her as his favourite wife

Page #170
157 21 With this assumption I disagree with, for instance, Warder who is of the opinion that the Gathas were in the first place composed by the villagers, and by implication were meant for an audience consisting of villagers. Subsequently the Gathas would have been copied and imitated by poets and scholars from the city (Warder, 1974: 182). In fact, the exclusive preoccupation seen in the Gathas with the villagers' poverty and stupidity and with the awkward situations arising from these factors, make it hardly likely that the Gathas were ever meant for their amusement. It should be noted that Warder on the same page-writes that the Gathas are 'songs of the villagers, in which their joys (sic) and sorrows are recorded'. Compare Kosambi who in the introduction to the Subhasitaratnakosa (p. XLVIII) wrote that 'the new poetry... must have aroused amusement and that tinge of good-natured contempt for the halika, which we find even in the Sattasat. However, in my opinion Kosambi presents only a limited view of what the text is about. 23 See Gatha *3. The word kaivacchala (kavivatsala) would indicate that Hala is a king. 24 The name Hala figures in the Puranas in the lists of the Andhrabhrtya kings, for which see Pargiter (1913: 35-43). The Puranas call the Satavananas Andhrabhrtyas, referring to the time when their realm beside Maharashtra included Andhra as well (see Gairola, 1956: 161-2). The dates of Hala are computed on the basis of the Puranic lists. Two Satavahana coins have been found bearing the name Hala (Rama Rao, 1971: 37). 25 Other rivers mentioned are the Yamuna, the Tapt and the Murala, but in Gathas which are restricted to the North-Indian (Yamuna in 671,

Page #171
158 Tapt in 239) or the South Indian branch of MSS (Murala in 876). 26 Levi, 1902: 111-2. 27 See, for instance, Keith (1928: 30): '... the second century A.D., in which century Maharastrt lyric began its career, made known in the anthology of Hala: Earlier literature in a Middle Indic language is invariably of a religious nature, either Buddhist or Jain. Occasionally these religions produced 'literary' literature as well. See, for instance, the Thera and Ther T-Gathas, from which, however, religious ideals are never completely absent. On the fusion of religious or ascetic tendencies and the erotic in the Thera and TherT-Gathas, see Lienhard (1975). 28 The name Maharastrt turns up only relatively late. It does not occur in the probably oldest description of the dialect in Vararuci's Prakrtaprakasa, but is mentioned only in chapter 12 of that grammar (sutra 32: ses am maharastrivat), which is almost certainly a later addition (Nitti-Dolci, 1938: 19). 29 Keith (1928: 30) writes: 'Of the Prakrits Maharastrt held pre-eminence by its use in the dramas, whence it was introduced by Kalidasa from lyric poetry' (by which he refers to the type of poetry as found in the Sattasaf). The influence of the Sattasas on the works of Kalidasa has been the subject of an extensive study by G. L. Hart III, The Poems of Ancient Tamil. Their Milieu and their Sanskrit Counterparts, Berkeley, 1975. 30 For the Satavahanas exclusively using Prakrit for their official documents, see, among other, Levi (1902: 110) 31 In this connection a legend has been preserved which seems to have an

Page #172
159 historical basis. It is told that the Katantra grammar of Sanskrit was composed for the benefit of a Satavahana king who knew only Prakrit and had difficulties in conversing with his wife who spoke Sanskrit. Levi (1902: 112) argues that this legend could actually refer to king VasisthTputra, who was married to a daughter of the Mahaksatrapa Rudradaman'. His wife composed the only Satavahana inscription in Sanskrit whereas all the other inscriptions of this dynasty are exclusively in Prakrit. 32 In any case, as pointed out by Levi (1902: 112) the language of the Satavahana inscriptions is, but for its Middle Indic phonology, extremely close to Sanskrit: 'Les scribes qui redigeaient en pracrit les panegyriques royaux...n'avaient a faire qu'un effort insignificant pour tourner la louange en sanscrit; ils effleurent le sanscrit de si pres qu'ils semblent s'en garder plutot que d'y pretendre.' 33 In whatever way one looks at Maharastrt Prakrit it is fundamentally a form of Sanskrit, differing from it almost only in pronunciation. Contrast Apabhraca, which represents a thorough breach with Sanskrit, having an entirely different morphology. In the Sattasaf the local, i.e. Maharashtrian, element is mainly present in the vocabulary. Garrez (1872: 203-7) has shown that the Gathas contain several words which are typical of Marathi only. Some of the words collected by him appear much more widely distributed in NIA than he supposed. E.g. bhana-, for the NIA derivatives of which, see Turner, CDIAL 9383. Its use in the Gathas .is, however, indeed foreign to Sanskrit. Garrez was correct on, for instance, dave- 'to show' (CDIAL 6276) and rumda 'large, broad' (10781). To these words may be added saajjhia 'neighbour' (see Gatha *39), Marathi sejf 'female neighbour' and sejar 'neighbourhood' (13077); acchabhalla 'bear' (Gatha 109; see also the note to Gatha *4), Maratht asval, asvil, asol 'id', and Konkant asvel (f.) 'id' (2446); and tuppa 'ghee' (also found in AMg. + M.?), Maraght and Gujarati tup 'id' and Marathi tupat 'oily' (5864) (possibly a loanword from kannada; see DED 2685. Turner, loc. cit., suggests that Prakrit tuppa goes back to RV trpra 'hasty?').

Page #173

Page #174
CHAPTER 2 2. 1 Method of reconstruction In the edition (Gathas 1-50) I have attempted to reconstruct a text as closely to the original as the available MSS allow. The arguments used in the reconstruction are basically derived from the stemma. The MSS, which ultimately go back to one single MS, the archetype, are divided into two branches, consisting of the Third South-Indian recension, on the one hand, and the Jaina and Vulgata-recensions, on the other. Another result of the stemma is the conclusion that the order of the Gathas is most faithfully preserved in the latter two recensions. However, each of the two is characterized by certain transpositions of larger as well as smaller sequences of Gathas of the text of the archetype. This text can only be established again by identifying these sequences and restoring them to their original positions. In doing so the Third South-Indian recension plays an important role as its text was based on a selection from that of the archetype. Traces of the order of the Gathas of the archetype can still be clearly perceived. However, such a reconstruction can only be partly successful. For example, in the Vulgata a group of Gathas somewhere after 433 was apparently displaced and transferred to the sixth Sataka (see the point opposite 321 in Appendix III. Note that the order in the Jaina-recension is considerably disturbed as well). It is impossible, though, to determine whether the part of the text which was displaced began immediately with 434 or only with 435. A suitable principle for dealing with such uncertainties remains to be found. In the part of the text edited here no serious problems in reconstructing the original order of the Gathas occur. The differences between the Jaina and Vulgata-recensions consist only of some transpositions of individual Gathas in the Vulgata. Thus, Gathas *31, *32, *36 and 47, which in the Jaina-recension (R) occur as 31, 32, 36 and 47, are in the Vulgata (G) found as 188, 194, 189 and 190 respectively. Their occurrence in the Third South-Indian recension as 69, 70, 74 and 84 respectively shows that their position in the Vulgata at the end of the second Sataka is the result of some independent development in that recension. A different matter is the question which Gathas belonged to the text

Page #175
162 of the archetype and which represent later additions. Only those which occur in the Third South-Indian recension, on the one hand, and in the Jaina-recension and/or the Vulgata, on the other, can positively be traced back to the archetype. One way of editing the text is by first giving only the Gathas which are clearly authentic, with the remaining Gathas following in appendices arranged according to the position in the stemma of the MS or MSS in which the Gathas occur. The basic text would then probably contain not more than approximately 500 Gathas. The alternative, which I have opted for here, is to take one complete text which is close to that of the archetype as the basis, in this case a combination of the Jaina-recension and the Vulgata. This text will include Gathas of which the authenticity is doubtful. Some of these may possibly go back to the archetype. In the part edited here altogether four Gathas are missing in the Third South-Indian recension. It is almost certain that one of these, viz. *4, originally formed part of the text of the Third South-Indian recension and consequently of that of the archetype (see the note given there). It is equally certain that Gatha *40 is a later addition to the text in the North-Indian branch. This follows from the occurrence of a relatively modern word, bh amagana, which is otherwise found only in Apabhraia. This argument will turn up again in connection with 77 (absent in Ma, Ti, Tp and T (and W)) which contains the Apabhransa verb nada- 'to nag', and with 92 (absent in Ma, Ti and Tp (and W); but found in T) which contains the word phaladia (see Bha phalagaya), a formation, like bhamacana, typical of Apabhraia. The status of the two remaining Gathas, *18 (absent in all the South Indian recensions) and * 37 (absent in Ma, Ti and Tp (and W); but found in T), remains unclear. As already noted earlier, for the text of each individual Gatha often numerous variants are available. The elimination of variants, involved in reconstructing the text of the archetype, is generally settled with reference to the stemma. Thus, the text almost automatically contains that which Ma, Ti and Tp have in common with Bh and R and/or with the Vulgata MSS. Not infrequently two different, but stemmatically equivalent, variants are available. In such cases the basic argument in the elimination is that of the lectio difficilior. Occasionally, however, matters may be more complicated. Certain pro

Page #176
163 blems or ambiguities of the Prakrit text may have been solved identically in the different branches. A case in point is found in * 26, with kunati ccia... janastT in Ma and Tp, and R and y. From the point of view of the stemma this reading should have been adopted in the present text. A closer look, however, shows that probably it is only an 'emendation' introduced independently in the two branches, for kunamta ccia...janamto found in Bh. Another case in point concerns the variation between e and de or te, the personal pronoun of the third person plural and the enclitic forms of the personal pronoun of the second person singular. The variante is usually restricted to only one or two MSS (see the note on Gatha *15). Only once, namely in 683 is it found in practic cally all the MSS. However, this instance differs from the others in that the word e was misunderstood. In ma e mannun, e was combined with ma to form mae, the vocative of maa 'mother'. As such it apparently escaped 'emendation' into clearer de or te. It follows that in all instances e is a retention. A similar situation exists with regard to the variant : i beside vi. The former occurs each time only in one or two MSS, the other MSS having vi (see the note on *17). Only in 149, in jTviasesa i, is it found in practically all the MSS but, again, presumably because it escaped 'emendation' into vi, being misunderstood as the case ending of the instrumental of jTviasesa, i.e. jfviasesai. Moreover, certain changes in the text may have taken place more or less unconsciously in all the branches independently. A case in point is the .corruption of ruvvau (MS Bh) in Gatha * 10 into ruvvasu (Ma, Tp, R, etc.). In all stages of the transmission the text appears to have been influenced in one way or another by Sanskrit. This is apparent in the case of variants consisting of different spellings of one and the same word. Consider the variation between deara and devara as found in, for instance, *28. Whereas Tp and and B have dearo, Ma and R, K, Y, P, T and S have devaro (Bh deuro). Considering their peculiar distribution the variation must have an external cause, which then most likely is Sanskrit. This means that devaro is secondary. In any case there is no positive evidence to assume the opposite tendency here, namely that the text was given a more pronounced prakritic appearance, as, for instance, * by eliding the medial -v-. See in this connection what has been said

Page #177
164 above about the innovations consisting of sanskritic words against retentions consisting of typical Prakrit words (p. 29 and the note given there). A more complicated case is that of the variants chahin (a typical MIA formation) and chaam (sanskritic) found in * 41. The former is restricted to and B only. Following the stemma we would have to adopt chaam found in Ma, Tp, Bh and other MSS. However, there are reasons to assume that chahim is original, chaam having been introduced secondarily as the easier form. In this particular case another external factor, namely the influence of the indigenous grammatical tradition, may have. been involved. Where there are two variants which differ only in spelling it seems justified then to adopt in the text the one which deviates most from Sanskrit, irrespective of the position in the stemma of the MS in which it is found. Thus, in * 29 dena and rao, restricted to T, are adopted instead of tena and ravo found in the other MSS, and in * 30 nikkia found in k rather than nikkiva in the other MSS. The same principle is applied to such instances as bhisinivattammi (only in S; also in Y but in its South-Indian original -v- and -2- are hardly distinguishable; beside opattammi) in *4, paavadiassa (only Bh; beside opadiassa) in *11, and naarahara (only K; beside .ghara) in * 32. Usually the original form of the initial consonant of the second member of a compound is maintained. In these compounds the rules of internal sandhi have been applied. While the above instances are restricted to one or two MSS only this is different in the case of pautthavaid beside pautthapaia. The latter is found only in Ma, Ti and Tp(?). It should be noted, though, that in the Malayalam script 'as in most other South Indian scripts the distinction between v and p is paleographically almost insignificant. The transcription of a word with either v or p can often only be established with reference to the etymology. Other compounds showing this internal sandhi, which can be positively traced back to the MS of the archetype, are, for instance, mahilana (Ma, Bh, R, etc.; beside jana in Tp and B) in *48, Golaada in 107 and vanaava (Ma, Tp, Bh; beside odava in R, K, etc.) in 117. In all these cases we probably have to do with more of less fixed compounds inherited from Sanskrit. In this connection the following two instances may be considered in which an initial consonant is apparently lost, namely

Page #178
165 amdharahian in *32 in K (for gamdharahian) and icchena in * 33 in Tp (for kicchena). In both cases, though, it concerns a free-standing word. In the Note to *15 several other words have been collected which have lost their initial consonant, but those are all either particles, enclitic forms of the personal pronoun or personal pronouns. In this respect the case of icchena, which is an adverb, may be different from that of ahdhao, which is a substantive. Furthermore, amdhao in k may well be due to a clerical error; note the context: suhaa amdhao. As such it has a parallel in the text of Gatha 193, which in K immediately precedes it. MS K reads sa atham uauthu (for sa (9)adham uaudha). In this respect the case of icchena is, again, different. On the basis of the above consideration I have decided to place adhao of k in the critical Apparatus, but to adopt icchena of Tp in the text. In what follows I will discuss several sets of variants, which occur comparatively, frequently, in which the two main branches are diametrically opposed. This is the case with, for instance, the variants didha and dadha, patthi and puttht, and silahanijja and salahanijja, of which the former are found in the South-Indian branch and the latter in the North-Indian one. Occasionally some North-Indian MSS, however, have didha and pattht as well. The first two pairs concern different but equally possible developments of Sanskrit -r-. The third pair concerns a difference in the colouring of the Svarabhakti vowel. If the variation is indeed a purely phonetic matter the South-Indian branch should probably be followed, as I have subsequently decided to do. The South-Indian scribes would in any case have been less influenced by the actual pronunciation in their own language of the words concerned. Another type of variation concerns the particle u(v)a(-ha) 'look' (see, e.g. *4), which occurs as uva(-ha) in Ma, Ti, Tp and T and, where available, w, and as ua(-ha) in Bh, R and the Vulgata MSS. In this connection the question arises whether in uva(-ha) we have to do with an instance of the secondary insertion of the -v- or in ua(-ha) with an instance of its secondary loss. It should be noted that in the MSS of the South Indian recensions one does at first sight not otherwise find evidence of the automatic insertion of a -v- in the combination ua (e.g. uaa (udaka), sua (sruta) and lua (*luta, beside Sanskrit luna)). If the -v- was secondarily inserted here the etymology must have played a role.

Page #179
166 However, for u(v)a(-ha) no etymon is or ever was available (traditionally it is glossed with pasya(-ta), which is a translation not an etymology). This seems to point to the conclusion that uva(-ha) is original and that the loss of the -v- in the North-Indian branch is the result of a secondary development. However, the case of u(v)a(-ha) does not stand on its own. The same type of variation with exactly the same distribution of the variants can be seen in u(v)a (Sanskrit upa), ru(v)a (rupa), dhu(v)a- 'to shake, to wash' (dhuvati), su(v)a- 'to sleep' (Pali supati, Sanskrit svap(a)ti) and ru(v)a 'to weep' (Vedic ruvati). The instances of u(v)a and ru(v)a need not be specified here. Those of dhu(v)a- are the following: 130 dhuvai T, dhuai Ma(!), Bh, R, K, V, Y, P, S (means here 'to shake' the toes of the husband to wake him up gently. Wrongly interpreted to mean 'to wash'; from there dhaayi in Ti and dhavai in Tp), 280 dhuvai Ma, Ti, Tp, dhuai Bh, R, K, V, Y, T, S, 433 dhuvai Ma, Ti, Tp, S, dhuai Bh, R, , Y, 680 dhuvasi Ma, Tp, dhuasi, Y, T, S (in R the word is missing), 767 dhuvai R, T, S, dhuai Ma, Tp (remarkable reversal:). Note 532 dhuai in R, all the other MSS having dhunai. The instances of su(v)a- are: *33 suvai Ma, Tp, S, suai Bh, R, K, , B, Y, P. 65 suvasi Ma (-vv-). Tp, K, Y, P, S, suasi Bh, R. v. B, 66.suvasi Ma, Tp, Bh, suasi Bhb, R, R, K, , B, Y, P, S, 412 suvaha T1, Tp, T, suaha Bh, R, (suha in Ma, suhaa in Y, and suhaha in S), 717 suvai Ma, Tp, R (-vv-), T, suai Ti(), Bh, S, 804 suvasu S. The instances of ru(v)a- are: *9 ruasi (su) all, 143 ruvasu Ti, Tp, T, ruasu Bh, R, K, 4, y, P, S, 147 ruvamti Ma, Tp, T, ruamti Bh, R, K, W, Y, P, 307 ruvai Ti, Tp, T, ruai Bh, R, K, V, Y, S, 334 ruvasi Ma, Ti, Tp, ruasi Bh, R, , Y, S, 379 ruvatT Na, Ti, Tp, ruafitt Bh, R, W, Y, S, W (ruat T), 400 ruamtam all, 503 ruvant? Na, Ti, Tp, T, ruafitT (ti)-Bh, R,

Page #180
167 W, Y, S, 517 ruvasi Tp, ruasi Ma, Ti(!), Bh, R, W, Y, S, 556 ruvati Ma, Ti, Tp, T, ruamti , Y, S (rovasti Bh, R), 569 ruai Ti, Tp(!), Bh, R, V, Y, T, ruvai S, 596 ruamtammi , Y, 649 ruai Ma, Tp(!), W, Y, W, ruvai R, S, 784 ruai T, S, ruvai R, 837 ruamtie Ma, Ti, Tp(!), ruvat Te T, 848 ruvai Ma, Ti, Tp, ruai T, 885 ruvasu Ti, TP, T, ruvasi W, 895 rua Ti, Tp(!), T, 909 ruvasi Ti, Tp, ruvasu T, 915 ruvasti Ti, TP, T, 941 ruami Ti, Tp(!), T, 960 ruasi W, 963 ruai W. Contrary to u(v)a(-ha) these words have comparatively obvious derivations from Sanskrit. For the -v- in Prakrit the corresponding Sanskrit words have either a -v- or a -p-, which may have provided the basis for the secondary insertion of the -v- in the Prakrit words. For ru(v)a- the situation is different. As already indicated above it most probably goes back to ru-, ruvati (Vedic; Sanskrit has rauti, but also ruvanti). However, in the commentaries and, Chayas on the Sattasat it is almost invariably glossed with rud-, in conformity with Sanskrit, which uses rud- to denote 'to weep' (of human beings) and ru- to denote 'to howl' (mainly of animals). It is significant that precisely for this word the exceptions (rua-) in the South-Indian branch are numerous. A comparison of ru(v)a- with the other words involved suggests a different approach to the problem. It is possible to argue that in the MSS of the South* Indian recensions in ua a -v-, probably suggested by a pronunciation of ua as /uwa/, was inserted unless the etymology of the word prevented it. Thus, we find dhuva-, suva-, etc., but not * uvaa (adaka). For u(v)a(-ha) there simply was no etymology to prevent the insertion of -v-. Whereas the form rua- would be expected throughout on the basis of the traditionally accepted etymology, the occasional occurrence of ruvamust be explained as an instance of a scribal lapse. The problem of the variation between uva(-ha) and ua(-ha), etc. seems still far from definitely solved. Nevertheless some principle has to be decided upon with regard to the text to be edited. I have decided in each instance to adopt the form without the -v- (unless not available, as in, for instance, 909 and 915), thus in the main following the North-Indian , branch. In doing so I adhere in a way to the principle of the lectio difficilior: for most words concerned the form without -v- is the one

Page #181
168 deviating most from Sanskrit. A frequent dilemma occurs in connection with the endings of the nominative and accusative neutre plural, -ai(n), the genitive plural, -anam) etc., and the instrumental plural, -ehim) etc., when found at the end of the hemistich. This position is metrically neutral so that it does not make any difference whether the text reads, for instance, -ai or -ain. On this point, though, there is a significant difference be- . tween Ma, Ti and Tp, on the one hand, and Weber's MSS, on the other. The former, with only a few exceptions, throughout have the endings without Anusvara, while the latter favour those with Anus vara. The actual situation in Weber's MSS is admittedly unclear. K, edited in the Abh., indeed largely favours the endings with Anusvara, but this is to be expected in a MS in which the Anusvara is inserted apparently at random and without regard for the metre (see salak aranan in *3 and land aha in *8). In Retr. all deviations have been noted for B, P, T and S; from K, which for these MSS gives the same picture: generally the endings occur with Anusvara. However, in Ed. Weber refrains from noting this type of variation for the MSS, or parts of MSS, edited there. Moreover, he seems rather careless on this point. Thus, in 354 he reads muhenan , against all the MSS, which have muhena, except possibly Bh, for which in ISt. he does not note a variant (p. 109). This may, however, also be taken as an indication of the fact that he did not note this type of variation of Bh from the text of Ed. either. Before continuing the discussion of these endings in hemistichfinal position a few remarks may be made regarding their occurrence within the Gatha. There the variation between, for instance, -ai and -ain is entirely regulated by the metre. The endings without Anusvara are considerably more frequent than those with Anusvara. The latter are specifically found at the end of the uneven Padas where u-u does not occur. In this position one also finds -ena (relatively frequently) and -es un (once, in 77, a Gatha which probably is a later addition anyhow). -es un is otherwise not found. -enam is otherwise found only once, namely in 525. -ain, -anam and -ehin are otherwise rare too, and a comparatively large number of these instances is found in particular before the enclitics cia, pi and va, -ehin in fact only so. While within the Gatha the occurrence of the short and the long

Page #182
169 endings is entirely regulated by the metre, Weber notes that his MSS are highly irregular in that they add the Anusvara at random. They do so in particular in the case of the plural endings which end in -a and in -i. This agrees with the state in the hemistich-final position where these Same MSS would often have -ain, -anam and -ehin, while -esum (plural, but ending in -u) and -enan (ending in -a, but singular) are conspicuously absent. It would follow that the presence of the Anusvara in -ain, -anan and -ehim in hemistich-final position in Weber's MSS is secondary. However, this is only part of the picture. Ma, Ti and Tp at the end of the hemistich indeed generally have the endings without Anusvara, but exceptions occur, which are, however, usually restricted to one or two of these MSS. See, for instance, naanain (qaanalt) in Ma in *5. The addition of the Anusvara can hardly be explained as secondary, as otherwise, i.e. within the Gatha, it is not found at all. A more likely explanation is that the source of Ma, Ti and Tp occasionally did have -ain, -anan and -ehim, but that most instances have been removed in a process of normalization in favour of the endings without Anus vara. The reason behind the preference for the endings without Anusvara is unclear. It might be a case of extending the endings without Anusvara which prevail within the Gatha to the instances at the end of the hemistich. Note in this connection that in Tp the endings -e and -o, which mostly are to be scanned short, are represented by the characters e and throughout, i.e. also where they are metrically long and also where the distinction does not matter, as at the end of the hemistich. A further illustration of the extent to which the MSS of the Third South-Indian recension were normalized will be given below in connection with the Ya-sruti. Apparently the situation in the source of Ma, Ti and Tp was irregular. It is in any case no longer possible to determine whether in a particular case the archetype read, for instance, -ai or -ain. In the edition I have decided to follow the South-Indian MSS, which means that I have throughout adopted the endings without Anusvara, except where Ma, Ti and/or Tp have an ending with Anus vara. As to these variants the following procedure has been adopted. First those MSS are mentioned of which the existence of a variant is beyond doubt, and then I have added between brackets those MSS of which the reading cannot be definitely ascertained, mostly R, V, Y and Bh, and T

Page #183
170 and S in so far as edited in Weber's Ed. Those MSS which are not specified do not deviate from the text adopted. A further problem concerns the opposition between v and b. Ma, Ti and Tp have both v and b, much in the same way as we are used to from the Sanskrit dictionaries. The same is the case with R, B, Y, P, T and W. For B, P and T, see Retr., p. 349. In Ed., p. 5, Weber notes that R, Y, T and W usually maintain (Sanskrit) b initially. He does not mention anything about b after Anusvara. At least T also has b after Anusvara. The remaining MSS, i.e. Bh, K, and S have only v. In this connection the question arises which group of MSS most faithfully represents the state of the MS of the archetype. As far as Ma, Ti and Tp is concerned, which neatly distinguish between v and b, there is evidence to suggest that the b was introduced only secondarily. A clear instance is bola- 'to pass' for vola- in, for instance, MS Bh. The past participle bolina suggests that it consists of the verb 11-, in which case the first syllable is a prefix. No prefix bo- exists, while vo-, from Sanskrit vyapa- or vyava-, is relatively common. It follows that the b- in bola- is secondary. From the introduction in this word of bfor original v- it appears that its derivation was not known. It is to be noted that in Sanskrit vyap (/v) alf- does not occur with the meaning 'to pass'. It is not clear which other word the scribe had in mind when introducing the spelling bola-. One possibility is that b was introduced almost automatically for v unless the derivation of the word in question prevented it. volaia in Ma in 21 most likely is a survival of this process. Another instance is vaha 'tear' in Ma in 25: Ma, Ti and Tp otherwise have throughout baha. The instance in *25 differs from the others in that the word was wrongly supposed to contain a pun on the words vyadha 'hunter and baspa 'tear'. This pun is recognized by all the available commentaries including the one by Madhavayajvamisra found in Ma. It should be noted that the idea of the pun could only have come up in a text which did not distinguish between v and b! In the South-Indian MSS a dilemma arises. Whether one reads vaha as in Ma or baha as in Tp the pun is lost. It would seem that in Ma the dilemma was solved by leaving the text as it was. It follows that the MS of the archetype most situation has its parallels in the NIA languages. likely only had v. This Only a few of these

Page #184
171 languages distinguish between V and b; most have either v or b. It is a well-known fact that this phenomenon has greatly affected the transmission of Sanskrit texts, in which originally v and b were distinguished (see AIG I,S 161). It is in itself not surprising that it is precisely in the South that apparently a tradition was preserved with regard to writing. v or b in particular words. This almost certainly has to do with the fact that most South-Indian languages do distinguish between v and b. The decision to write in the text v throughout introduces yet another problem in connection with the variants. MS K, edited in the Abh., has only v. For B, P, S and T, edited in the first place in the Retr., the deviations from K in respect of this feature have been carefully noted; at least, Weber does not state explicitly that he makes an exception for the variation between b and v. The text in Ed. has only v. On page 5 Weber remarks that he does not specify the variants, unless the word in question should happen to be mentioned for other reasons. Bh, edited in ISt., seems only to have v; a few instances of b in this MS have been noted. My procedure is the following. Where I have adopted a form with v I first mention those MSS which positively have a variant with b, and subsequently add within brackets those MSS of which the reading cannot be definitely ascertained. As I have shown above (p. 40) the present manuscript tradition goes back to one single MS the text of which was not entirely flawless. Some errors in the MSS can be traced back to this particular Ms. Where possible an emendation has been suggested. In some cases the way to emend the text is pointed out in the MSS themselves. See 292 in which the archetype seems to have read tanuai (cf. Bh tanuai and Ti and Tp tanuai(d)a) which in all the other MSS was corrected into tanuaai. In other cases the emendation could be based on peculiarities of the script of the MS of the archetype, which was Jaina Nagart. In yet a few other cases the emendation could be based on an informed guess, for instance, gammihii for gammihisi of the MSS in 609 (see the note to *10). In a few cases, though, no convincing emendation seems available. The present manuscript tradition seems to have started in a milieu of Jains. This explains the occasional occurrence in the MS of the arche

Page #185
172 type of a word in a form otherwise typical of AMg. An instance is piva, the particle of comparison. In such instances I felt free to adopt in the text what seems the regular M. Prakrit form, namely via (see the note on *1). It appears that the archetype was furthermore characterized by certain orthographical conventions typical of, and otherwise exclusively found in, Jaina MSS, e.g. the Ya-sruti, the use of i and u for e and o respectively, and the use in certain positions of the dental nasal in favour of the retroflex one. Within the transmission of a text these features seem basically interchangeable. That is to say, a Jain copying a non-Jaina MS will add, for instance, the Ya-sruti where appropriate, and, vice versa, a non-Jain copying a Jaina MS will delete the Ya-sruti: 'Because I wish to avoid the impression that the Sattasat is the exclusive property of the Jains I have decided to divest the text of these orthographical features. The Ya-sruti is regularly found in Bh and R. Its occurrence in the MS of the archetype is proved through occasional instances in Ma, which are to be explained as remnants after it had been generally removed from the text. For e and (c) in a closed syllable Jaina MSS commonly write i and u respectively (see Pischel & 109). This spelling seems to show that in this position the phonemic contrast between i and e and between u and o had become less distinct. There are indications that this spelling was also found in the MS of the archetype. In the course of the transmission, however, e and seem to have been restored in most MSS except for Bh and R. In one instance this restoration did not take place because of a misunderstanding of the derivation of the word in question. The instance is found in 782, which reads: durapainpapar imala angacchitt a piva vam- sapallava muddhapupphapamguranan mahena di na mahusirTe. It is almost certain that chittan (found in all available Mss, including Ma, Ti and Tp) stands for chetta (Sanskrit ksetram), and that we should read agan chic/e)ttam. The word chittan, however, was misunderstood as the past participle of chiva- 'to touch', which prevented the restora

Page #186
173 tion of e for i. I should like to suggest to emend the instrumental vammahena (found in all the MSS) into a genitive vammahassa: The nymph of Spring has given her body (amgam), of which the fragrance spreads very far, which is decorated with sprouts and which has tender flowers for its cloak, to the god of Love (vammahassa) as his abode (chi(!e)ttam). This example concerns an instance of the writing of i for e in a closed syllable. In Bh the writing of i for e is also frequently found in an open syllable at the end of a word; e.g. tfi for tTe in *22. The situation in the MS of the archetype is unclear. The instance of jiviasesa i in 149, discussed above (p. 163), for which Ma and Tp have introduced jivias addhae, shows a tendency on the part of the latter MSS to introduce e for i. However, it does not become clear if and to what extent i was written for e in open syllables in the archetype. For initial n- and medial -nn- MS Bh consistently writes n- and -nn- respectively, the latter combination being often written as -rinThis spelling is typical of Jaina MSS. As such, it might have been a feature of the MS of the archetype, too. The spelling seems to be old and to have a basis in some spoken languages (Bloch, 1934: 79). On the other hand, it is possible to discern a development within the Jaina tradition itself, the old palmleaf MSS having n- also initially. It is possible that we have to do here with a locally (Gujarat?) developed convention. The Ya-sruti, i and u for e and o in closed syllables and n- and -nn- for n- and -nn- respectively, are more or less consistent features in Bh and/or R and need not be specifically mentioned in the critical Apparatus. An exception is made for i for cas in tti, which is irregular in Bh. 2.2 The variants In general all available variants have been specified irrespective of their nature. For those of Weber's MSS I make use of his published material only. The text edited here in many cases differs considerably

Page #187
174 from the one in Weber's Ed. Often the readings of his MSS have to be ascertained by inference, i.e. where Weber does not mention any variant on his text his MSS presumably do not deviate from it. Above I have already mentioned several types of variants for which this method does not work, as in the case of the variation between -ai and -aim at the end of the hemistich and of the variation between v and b. On the other hand, a certain number of consistent variants need not be mentioned. Some of these have already been discussed above: the Ya-sruti and the use of i and u in closed syllables for e and respectively, in Bh and R, and the use of n- and -nn-. in favour of n- and -nn- respectively in Bh. A few others will be discussed below, which seem basically determined by the various scripts used. One group of peculiarities concern the South-Indian MSS Ma, Ti, Tp, Y and T. It should be remembered that of these five MSS only Tp is actually written in a South-Indian script, viz. Malayalam. Ma, Ti and y are Nagari transcripts and T is a Roman transcript. For the two other South-Indian MSS, P and W, Weber relied on Roman transcripts as well. In these the peculiarities mentioned below seem to have been silently corrected. Tp throughout writes 1, which is typical of South-Indian MSS (see Luders, 1940: 548). This 1 was apparently also found in the source of Ma and Ti, which confuse 1 with i or u. For 1 in the source of the transcript T, see Ed., p. XXXI, note 2. In most of the South-Indian scripts a dot or circle in the line is used to indicate the doubling of the following consonant. This sign is often confused with that for the Anusvara, a dot or circle written above the line. This confusion is apparent in Ma, Ti, T (Ed., pp. XXXXXXI) and y (Ed., p. 2). In Ma and Ti this 'Anus vara' was subsequently replaced by the homorganic nasal (see below); e.g. agupinga (anuvigga). In Ma and Ti initial i, u and 1 (1) are often interchanged due to the similarity of the characters in the Malayalam script. Ma, Ti, Tp, y and T write ho and hm throughout (in Tp hp and hm are often preceded by the dot or circle indicating doubling). This spelling is typical of South-Indian MSS in general (see Printz, 1921: 24). The spelling represents an interpretation of the Nagar ligatures for ph and mh, in which the nasal is indeed written within and below the h.

Page #188
175 Note in this connection that aspirated nasals are unknown in South-India. Regarding the variants of Bh given in Ist. the following should be observed. Weber notes that in Bh the ligature -jj- is represented in two ways, namely by the usual symbol in Jaina Nagart, which slightly modified is also used for -jjh-, and by a symbol which resembles the ligatures for -sv- (-sb-), -bb- or -mv-. The second ligature Weber transcribed by any of the latter combinations or by -yy-. Unfortunately Weber does not reproduce the ligature in question so that we have to guess what it actually looked like. It may have been the ligature , which, however, in the appendix on paleography in L'Inde Classique (p. 695) is classed among these for -jjh-. Alternatively, it could have been the ligature , which according to Jacobi (Kalpasutra, p. 18, note 1) stands indeed for -jj-. To distinguish this ligature for -jj- from the other one I have transcribed it as -jjh-; thus, kuppijjha in *17 for kuppiyya of ISt., p. 35. Weber (Ed., p. 3) notes that his MSS practically throughout write -ch- for -cch-. In y as well as T -Sch- is found, which occasionally is very similar to -chch-. Likewise his MSS often have -khkh-, -jhjh-, -thth- and -dhoh- for -kkh- etc. The same is the case in MS K, for which see Abh., pp. 26-7. W and y write - thth- throughout. In Ed., p. 3, Weber furthermore notes that his MSS rather randomly add the Anusvara, in particular after the case endings for the plural which end in -a and -i (see also above, p. 169). In his Ed. Weber did not specify such variants unless the word in question should already be mentioned for some other reason. On this point I had to make do with what Weber cared to mention. This type of variation is exactly known for K, through Abh., and for those parts of B, P, T and S edited in his Retr. The majority of the MSS writes -ink-, -nc-, etc. Some, though, use in these combinations the homorganic nasal instead of the Anus vara. This is consistently the case in Ma and Ti, both Devanagari transcripts. One wonders if this feature was introduced secondarily in these MSS at the time of their transcription into Devanagari. It is in any case significant that Tp, an original Malayalam MS, has the Anusvara. -nk-, -nc-, etc. is also found in some of Weber's MSS, but as the exception rather than the rule. For the instances in K, see Abh., p. 20, note 2, and p. 21. Note furthermore tantatatattin in *2 in (but in the same Gatha

Page #189
176 pa ananti) and dant in 23 in T and S. I have the impression, though, that in the edtions following that of MS K in Abh. Weber neglected this type of variation. For instance in 1 K reads samkanta. In Retr. Weber did not give any variants, which would mean that B, P, T and S has samkanta as well, which I have strong reasons to doubt. I follow the practice found in the majority of the MSS, writing in these combinations the Anusvara. The relevant variants from Ma and Ti need not be specified. As far as Weber's MSS are concerned I have mentioned their variants only when they are actually available; thus, samkanta in 1 for K, but no variants for B, P, T and S. 2.3 The presentation The presentation of the text is as follows. Above the text of the Gatha its position in the various MSS is indicated. The relevant MSS are given in the order of the authenticity of their texts, i.e. Ma, Ti, Tp; Bh, R; K, V, B, Y, P; T; S; W, with W, which strictly speaking is probably not a recension of the Sattasat, put at the end. As the order of the Gathas in this edition is basically that of the Jaina-recension and the Vulgata, a different order in the enumeration might be expected. This would, however, result in two systems, one above and another for the Critical Apparatus below the Gatha. For practical purposes it has been decided to use only one system. For completeness' sake also those MSS which contain only a Sanskrit commentary have been given, namely G (Gangadhara; stands for A, C and E), and T, and X, which contains what seems to be a separate recension. G, E, p and x are mentioned in this order after P of the Vulgata. W and V, the latter containing only a Sanskrit Chaya to the Gathas, are treated as a single witness: when a Gatha is illegible or absent in W but its Chaya is found in V, the latter is mentioned. The MSS are broken up recensionwise by semicolons. Below the text the Critical Apparatus is found, preceded by a description of any larger lacunae. The order of the variants follows the text. A variant may concern a single word or a whole part of the text. Two dashes indicate that a following part of the text is referred to. When more than one variant is available these are given recensionwise and when an identical variant occurs in the MSS of two 'consecutive'

Page #190
177 recensions, the variant in question is not repeated. Note in this connection that this meant that the order K, 4, B, Y and P could not always be maintained. The Critical Apparatus is followed by an English translation. The translation in a way had to be a compromise. I have attempted to keep as close as possible to the Prakrit text in order to help the reader to identify the words. On the other hand, in several instances a literal translation fails to convey the intention of the Gatha as a whole, so that recourse had to be taken to certain idiomatic expressions. In several cases even then further comment on the meaning of the Gatha was deemed necessary, which is found in the Notes. I have dispensed with introducing specific symbols which would indicate where the text adopted is doubtful or conjectural. Problems concerning the reconstruction are discussed in the Notes. These same, Notes are furthermore intended to clarify all other problems which may arise in connection with the text. A recurrent theme in the Notes is the discussion of the relation between the text and the indigenous grammatical tradition, with special emphasis on the grammars by Vararuci and Hemacandra, and the relation between the text and traditional lexicography, as exemplified by Dhanapala's Paiyalaccht and Hemacandra's Des Tnamamala. The Gathas of the Sattasaf have often been quoted as illustrations in Sanskrit works on poetics. The earliest work of this type in which Gathas have been quoted is the Dhvanyaloka; e.g. Gatha 669 on p. 20, 110 on p. 119, etc. Another work in which Gathas from the Sattas at occur is Bhoja's Srngaraprakasa. In an appendix (Abhang) to his Abh. (p. 202 ff.) Weber collected the Prakrit Gathas quoted in Dhanika's commentary on the Dasarupa, in Mammata's Kavyaprakasa and in Visvanatha's Sahityadarpana. A number of these Gathas is also found in the Sattasat. Besides in works on poetics, Padas or even smaller parts of Gathas have also been quoted in Hemacandra's Prakrit grammar. In this edition the problem of the testimonia is not taken up. I have made an exception for the quotations collected in Weber's Abhang and those found in Hemacandra's grammar. They are referred to immediately below the Critical Apparatus; only important variants are noted.

Page #191
178 Notes to chapter 2 The -v- in dhuva- and ruva- features in the Prakrit passives' dhuvvata (564) and ruvvau (*10). It is, however, unlikely that the -v- was secondarily inserted in dhuva- and ruva- on the basis of these passives. Traditional scholarship was not so much interested in relations existing between words in Prakrit as in the relation between Prakrit and Sanskrit. It should be noted that in Pkt ru- and rud- have partly coalesced, roissam, roiavva and runna (+*rudna) most likely going back to rud Weber refers to two editions of the Kavyaprakasa, viz. the Calcutta editions of 1829 (Kb) and of 1866 (ka). None of the two was available to me. I refer to the Poona edition by R. D. Karmarkar; of the two editions of the Sahityadarpana, namely the ones of 1828 (Sb) and of 1851 by E. Roer (Sa), only the latter was available to me.

Page #192
CHAPTER 3 3.1 Edition of Gatas 1-50, with translations and notes 11 1 Ma, 2 Tp; 1 eh, R; K, \, B, Y, P, G, 5; T; S; V) I starts only with *21; x only with 133. The first ten Gathas of W were illegible. pasuvaino rosaruna- gahiagghaparik aam via padimas ark amtagor imuhaamd am samjhasalilamjalin namaha. In Ma lacuna for the text; in Tp for pasuvaino...padima (samk anta). pasuvainoh K, vasuvaino P -- sak anta k -- muyacand am Bh -- gehia Tp; gahiagdha K. -- pamkayam piva Bh; okaammia , ommia B, i bia y; kaa via T. Bow before the moon-face of Gaurt which by her anger at Pasupati has become the counterpart of the morning sky; it looks like the cupped hands filled with Sandhya-water, holding a (red) Arghya-lotus. The position of this Gathi at the beginning of the text follows the convention that a literary work should open with a verse praising a god or goddess, or, as here, exhorting the reader to do so. (Cf. Kavyadarsa I 14: sargabandho mahakavyam ucyate tasya laksanam/ astr namaskriya vastunirdeso vapi tanmukham. Note also the presence of a Nandt at the beginning of a play.) These verses are often characterized by particularly intricate descriptions. Here Gaurt is described as being enraged at her husband Pasupati, a situation which is fully in accordance with the general content of the text. It may, incidentally, be noted that the Gatha, in containing the word for 'moon', follows exactly the prescription given for the content of a Nandt (see Levi, 1963: 132). This Gatha, together with the following two, forms an introduction to the text. Gathas *2 and *3 give the impression of having been composed for the occasion and have as a matter of principle to be set apart from

Page #193
180 the other Gathas. The same would seem to be necessary for Gatha *1. Thus, the description of a goddess and her husband seems out of place in a text which is mainly concerned with common people such as farmers and hunters, despite the parallelism in their respective love-lives. A further argument is its form, which is that of a Stotra. It should be noted that most of the other Gathas which have the form of a Stotra, praising either Siva, Visnu or Ganesa, are in all likelihood later additions to the text, as follows from their absence in the Third South-Indian recension. See, for instance, 448 (Siva), 151, 406 and 411 (Visnu), and 403 (Ganesa). In some of these, 403, 406 and 411, erotic connotations are lacking completely. However, Gatha 455, a Stotra praising Siva, like the one under discussion, goes back to the archetype. My translation of the Gatha differs considerably from Weber's. The latter scholar connects the first word of the Gatha, pasuvaino, with sanjhasalilanjalin which he takes directly as the object to namaha, and arrives at the following interpretation: 'Bow before the Sandhya-water in Pasupati's hands' (see 'Retr., p. 347). He may have been led to do so by the position of the particle of comparison, which is here, contrary to common practice, not found after the Upamana but after a compound which describes the Upamana. It should be noted that this is not uncommon in Pkt (see Gatha 93: sivinaap Tena va paniena), nor in Skt (see Speyer SS 430). The -e- in Tp gehia, for gahia, is probably due to the South-Indian pronunciation of -a after certain consonants as -e. This phenomenon has been described for Malaya] am by Godavarman (1935-37) and by Kunhan Raja and Ramachandra Sarma in their edition of the Prakrtaprakasa (p. IX). These authors restrict it to -a occurring after voiced plosives and after C-, y-, r- and 1-. Indeed in the South-Indian MSS of the Sattas at one comes across *7 lehada Tp, 60 deia Ma and hypercorrect 132 davvena Ma and 532 gahnai Ma, Ti, TP, T. But in these same MSS -e for -a or hypercorrect -a for -e is also found after voiceless plosives and after h-. E.g. 189 teha Ma, 412 teio Ma, 642 kendofta Ma and 183 pecchanna Ma (but in the last instance probably under the influence of the following palatal plosive; cf. 373 pacchaito Ti, 457 pacchai Ti); hypercorrect -a in *17 kuppajja Ma, Tp, 4?0 uvakkhio Ti, Tp (written upakkhio), 445 uppakkhimo Ti. After h-: 914 hero Ma (+(b)haro) and 930 hea Ma..Further

Page #194
181 more one comes across *39 e Ma (for (c)a), 93 aviehna Ma and *17 anunajja TP (probably under the influence of following -jj-). In 365 aggapaapallianam Ti and in 578 padipallio Ti, pallio for pellio may be due to the preceding syllable pao. For via MS Bh has piva and MS v ia. This variation needs to be discussed. Corresponding to Skt iva the Pkt of the Sattasaf has commonly (v)va, which Pischel ($ 143) explains as an apocopated form of iva, as (t)ti of iti or pi/vi of api. Besides, but considerably less frequently than the latter, a bisyllabic particle is found, via after vowels, in the SattasaT -- possibly accidentally -- only after -o, but, for instance, in the Setubandha also after - and -e: Gatha 337 parivaho via, 443 kurunaho via, 612 manoraho via, 767 pariamo via; cf. Setubandha V 77, VIII 101, XII 95 (v.1.), XIV 80; after -a in I 7, VI 71, and after -e in I. 47. The same form is also found after Anusvara, but has, as already briefly indicated, several variants there: piva in the Jaina MSS Bh and R as well as in the South-Indian MSS Ma, Ti and Tp which ultimately derive from a Jaina Nagart MS. MS T, which is otherwise closely related to the MSS Ma, Ti and Tp, usually has via. The transcripts of the SouthIndian MSS often read viva, showing a confusion between pand v, characters which are very similar in the South-Indian alphabets. The Vulgata MSS and s usually have iva (or ia), in which case the preceding Anusvara is, replaced by -mm (in the Vulgata MSS) or by -rim (in s). The instances are the following. Note that viva of the South-Indian MSS has been silently corrected into piva: 1 park aan via TP, R, K, B, Y( - bia), P, T, S; piva Bh; -mm ia y, 125 pihin via Ti, Tp, Y, P, T; piva Bh, R; -mm iva K, , -im iva S (Most probably, however, the text should here be emended into niht via, the accusative nihim being ungrammatical) 290 duvvaanam via Ti, Tp, Y (- vio); piva Ma, Bh, R, T; -mm iva K, V, -rim iva S, 331 vand in via T; piva Ma, Ti, Tp, Bh, R; -mm iva K, W, Y, S, 336 os ah am via T; piva Ti, Tp, Bh, R; -mm iva K, W, Y, -im iva S, * 363 das anam via T; piva Bh, R; -mm ia K, -imm iva , Y, -im iva S, 388 lacchin via U, Y, S; piva Ma, Bh, R,

Page #195
182 507 janan piva Ma, Ti, Bh, R; -mm iva W, -m iva Y, -im iva S, 609 duddha piva Ma, Ti, Tp, Bh, R, T; -rim ia , -mm iva Y, -rm iva S, 782 chittan piva Ma, Ti, TP, R, T; -mm iva S. Of these variants iva, which is found exclusively in the Vulgata MSS and S, is almost certainly an innovation. The change of the preceding Anu- . svara into -mm or -rim represents an attempt to preserve the metre. Before iva, which begins with a vowel, the Anusvara may, as in Skt, have changed into -m, in which case the syllable is short. As such the phrase vard imm iva has an exact parallel in Pali mamm iva (see CPD II, p. 331; for alternative mam iva, see Geiger 71). The status of iva in Pali, however, cannot be discussed here. The MS of the archetype seems to have read via as well as piva. The former is not restricted to M. The single prevailing form in s. and Mg. is via. It is, furthermore, undoubtedly identical to pali viya. In all these dialects and languages the particle occurs after vowels and after Anusvara. Geiger ($ 66) suggests that Pali viya, through metathesis, goes back to *yiva, which would be a frozen sandhi-form. He proceeds from this form *yiva in order to be able to account for the medial -y-, the origin of which, however, need not detain us here. I largely agree with Geiger's explanation but should like to add some further remarks concerning the circumstances which may have been responsible for the supposed metathesis. An indication for this may be found by comparing the other most common enclitic particles, namely (t)ti, (v)va, (c)cia and vi/pi. It should be noted that all these begin with a consonant, some after having lost their original vowel. This apparent tendency of enclitics to begin with a consonant may have been responsible for the metathesis in via. The second particle, piva, is otherwise exclusively found in Amg. and JM. texts. It occurs only after Anus vara; after vowels these texts have viva. According to Pischel ($ 337) viva would consist of a kind of proclitic v- plus iva. The existence of such a proclitic v- is, however, doubtful. In the same paragraph Pischel has collected several more instances of this v-, most of which, however, should be explained differ

Page #196
183 ently. E.g. v- in vuccai (Skt ucyate) is from vac-. Alternatively viva may be a frozen sandhi-form. It is difficult, though, to separate viva from the form via of the other dialects. Admittedly, the origin of the medial -v- presents a problem, for which I am unable to offer a satisfactory explanation. One might suppose contamination with iva, found in AMg. texts besides viva. I am in any case hesitant to connect it with Gandhart Dhammapada Pkt viva, about which much is still uncertain; see Brough (1962: 110-2). The conditioning factor for the variant piva beside viva is strikingly similar to that for pi and vi. It has even been suggested by, for instance, Weber (Ed., p. 1), Goldschmidt (Setubandha, index, s.v. via) and Jacobi (Kalpasutra, p. 100) that it is a compound particle made up of pi/vi plus (i)va. This derivation is doubtful. Admittedly, compound enclitic particles are not entirely unknown in Pkt, which has (c)cia, from caiva, but the latter combination is already known in Skt, whereas api plus iva is not. : This derivation has earlier been dismissed by Pischel (5 336) who, instead, suggests that the equation pi : vi may in fact have been responsible for the origin of piva from viva. A logical extension of Pischel's reasoning would be that piva is in reality but a hypercorrect form or a scholarly invention. For, an original initial v- remains vafter Anus vara, and in this connection it does not matter whether it is a proclitic consonant, a glide or a medial -v- transposed through metathesis. Note that in the case of pi/vi the situation is reversed, the P- being the original and the v- its phonologically conditioned variant. The fact that it is well established from the oldest Amg. texts onwards would seem to speak against this view of piva. This may, however, also be taken as an indication that these texts were subjected to rigorous redactional interventions. That this is indeed the case I shall try to demonstrate elsewhere. It remains to explain the occurrence of piva in M. Pkt, in which, at least in the Sattasat, viva does not occur. It should be noted that at the beginning of the present manuscript tradition of the Sattasat stands a MS, or a line of MSS, written in Jaina Nagars, which most likely will have been written by someone well acquainted with AMg. and/or JM. It is not unlikely that at this stage piva was entered into the text for via oc

Page #197
184 curring after Anusvara, either unconsciously or consciously because via disturbed the scribe's notion of correctness. Ag. piva beside viva was, if not invented, at least supported by the parallelism with pi : vi. The case of via after vowels is slightly different. The same scribe may have compared it to AMg. viva. The absence in via of the medial -vneed not have disturbed him, its loss not being uncommon in most of the literary MIA dialects. It should be clear that a definite conclusion regarding the status of piva in M. Pkt needs a closer study of the text-transmissions of, for instance, the Gauqavaho and the Vajjalagga, in which it occurs likewise. In the Setubandha no instances of piva are found, the particle not occurring after Anusvara. Not all the instances of piva in the MSS need automatically go back to the MS of the archetype. This may be especially relevant in the case of Bh and R, which are Jaina MSS. It may also be relevant in connection with the MSS of the Third South-Indian recension. The case of 290, where Ti and Tp read via but Ma piva, should make us aware of the possibility that piva, but also via, may have developed into an independent variant. It may anyhow not be without significance that T, which otherwise agrees so closely with the MSS of the Third South-Indian recension, mostly reads via for piva in the latter. Finally the evidence of the traditional grammars may be considered. It should a priori be noted that not much value should be attached to absence or presence in the grammars of a particular form of the particle, which may merely indicate that the grammarians' observations were restricted to certain texts only or to certain MSS of texts. Thus, Var. IX 16 mentions via, miva (apparently abstracted from -im iva) and mmiva (from -mm iva) but does not mention piva. The latter is given in X 4 for Paisct. Note, however, that according to Pischel ($ 336, note 6) mmiva of IX 16 is a clerical error for piva. Hem. II 182, on the other hand, mentions via and miva as well as piva and viva, but it should be added that in his Sutras not always a clear distinction between M. and AMg. is maintained. In the edition I have throughout adopted via, even where piva can be ascertained for the MS of the archetype. My reason for doing so is twofold: if AMg. piva is indeed a ghost-word the decision is an obyious

Page #198
185 one; and even if it would be linguistically acceptable for AMg., its occurrence in one dialect next to via, which is so much different, is in my opinion very suspect. Its occurrence in the MS of the archetype beside via can be explained through redactional interventions. 21 2 Ma, 3 Tp; 2 R; K, V, B, Y, P, G, 5; 7 T; S; 2 V) Missing in Bh. ama ar pauakavva kamassa tatatattin padhiu soum ca je na ananti kun anti kaha te na lajjamti. In Ma lacuna after kama(ssa). amian R; K, V, B, Y; T; S -- vayada Ma, paada Tp; paian R; paada T; paia S -- pathiin Ma, pathium Tp; v; pathidum T -- sodum P; T -- a K, V, va P; yi. T -- ja Ma -- na jaNGanti Tp; R; K, B, Y, P; T; S -- tattatatti R; tahttatamtti K, tantatatattin V, tattaciitan B, tattattattin Y, tattatatti P; s', tattauttin s2 -- te kaha R, W, Y; te kaham K, B, P; te kaha S -- lajjimti To. Those who do not know how to read and listen to (this) nectarlike Kavya in Prakrit, but are solely concerned with scholarly works (in Sanskrit) on love, .how can they not feel ashamed? The readings adopted for, and my interpretation of, the compound in c differ from Weber's. The final member of the compound should read tattin, which word Des in. V 20 translates with, among others, tatparata 'being devoted'. Indeed it seems to have this meaning as well as that of 'worry, concern' (see Gathas 51, 272, 833 and 866). Turner, CDIAL -5683, derives tatti from Skt tapti 'heat'. For the development of the meaning, note especially Si. tata 'effort, aim' quoted there. The first member of the compound should read tamta (tantra), which has the pregnant meaning 'scholarly work in Sanskrit' here, thus opposing pauak avva 'literary work in Prakrit'. The study of the latter is considered a necessary adjunct here to that of the former. The form paada (praksta) found in Ma, Tp and I seems an independent

Page #199
186 variant turning up in the South-Indian MSS of Pkt texts or in Pkt text's written in South-India. See, for instance, Karpur amanjart I 7, MSS S and U, which have a variant pauca, and MS T, which has yataya, and Candralekha I 11 and Kamsavaho IV 48h. The latter two texts were written by South-Indians, Rudradasa and Rama Panivada respectively. Hence paada could be an innovation in the South Indian MSS of the Sattasat. On the other hand this does not explain how paada could have become prevalent in the South and how the notion that it is the correct form could have arisen there. In this connection it should be noted that paada, from prakrta, is formed after a sound-change, -rta- + -ada, typical of AMg. and JM. (Pischel S 219). Its occurrence in the South-Indian MSS of the Sattasat could then be explained with reference to the Jaina. Nagart MS from which they ultimately derive. The word may have been introduced in this MS by its scribe who very likely was familiar with AMg. and JM. or with the sound-changes accounting for the derivation of these dialects from Skt. Furthermore, from its occurrence precisely in these South Indian MSS of the Sattas as it may have become known in the South as the only correct form for Skt praksta. However, M. Pkt usually has paua. Also the earliest AMg. instances of the word, found in late canonical texts, which, however, are considered in any case, older than the Sattas at, read pagaya or payaya, showing a sound-change similar to that in paua. The word paada therefore most likely seems to be a learned, hypercorrect Alg. form. It may not be without significance that paada, which is thus well known in South India, is not mentioned in any of the grammars (pauda in the pratyadigana in Prakrtasarvasva II 10 is a corruption, through paduga (MSS U, I, O and G), for pahuda (cf. Hem. I 206)). This would indicate that the grammars did not use or know the South Indian MSS of the Sattasaf. After na the verb jana- (Skt jna-) invariably loses its initial j-; see, for instance, 347 na animo. This loss is usually restricted to -j- found within a word. Apparently the phrase na ana-, which occurs frequently, has come to be considered as one word (cf. na ittho in 502). All the MSS, however, except Ma and y, have na jananti here, which nevertheless is to be rejected as the easier reading.

Page #200
187 31 3 Ma, 4 Tp; 2 Bh, 3 R; K, U, B, Y, P, G, 5; 5 T; 8 S; 3 V) sattasaain kaivac- hulena viraiaim chalena kodte majjhaarao salakarana gahana. In Ma lacuna between h(alena)...rana gahana. saisin K -- kaivachalena K, vatthalena , kavatsalena y -- kotiye Ma; kodta U, B, Y; kotte S -- majjhaarammi R; U, B, P, masjhaarammi K, majhjhaarammi Y; majjhaarammi 5 -- salaina T -- viraiana Bh -- salamkarana K -- gahana R; K, U, B, P; T; S; (Bh). From among ten millions of Gathas full of Alankaras, sevenhundred were collected by Hala, the patron of poets. The ending -e in kogte is to be scanned short. In most NagarT MSS this -e, for lack of a separate character, is represented by -e. In some MSS, and practically throughout in B, Y and S, it is represented by -i after -a (-ai) and -u (-ui) and by -a after -T (-Ta). Bh, which usually has -te, sometimes has -Ti; see, for instance, * 22' muhti and tfi. Similarly, - ad and -1o are written as - du and - Tu respectively. -uu for -uo is not found in the Sattasaf. i and u as graphic representations of e and O are also found in closed syllables (Pischel $ 119). The origin of -a in, for instance, kodta, is as yet unclear. It may, however, have arisen from confusion between - i- and -ya- : kodTe + kosti - kogfya + kogfa. The recognition of this ending -a was facilitated by the fact that an ending consisting of a vowel similar to the stem-vowel, as in - Ti, was apparently considered awkward. Thus, while Var. V 22 accepts the ending -a, the commentator Bhamaha, ad Var. V 23, explicitly forbids it after a-stems, where it would give forms like malaa. In the Gathas of the Sattasat the endings -e and -o occur practically only in positions where they have to be scanned short. Long -e and - are found at the end of the hemistich, which is metrically neutral. Furthermore they are found at the end of the uneven Padas where (u) - U does not occur. Otherwise they are rare: see (*22), *43, 226, 228, 322, 492, 513, 593, 599, 739, 789, 797 (ex conj.) and 810. The same is also

Page #201
188 the case with the ablative ending -o as in -ao. In S the endings have been normalized in favour of -a, -i and -u, which are often also found at the end of the hemistich. For a possible derivation of the Dest-word majjhaara (Desin. VI 121: majjhammi majjhaara), see Emeneau (1968: 32-3). 4( 3 Bh, 4 R; K, W, B, Y, P, G, 5; 444 T; 77 S; 97 W) Missing in Ma, Ti and Tp. ua niccalanipphamda nimmalamar agaabhaana- bhisinTvattammi rehai valaa paritthia sarkhasippi wa. uva T; W -- nippada Bh; nippanda T; S; pipanda w -- bhisina K;.. hisint T; visint S; bisint W -- pattammi Bh, R; K, 4, B, 'P, vattami Y; pattammi T; pattami W -- (lacuna)ya Bh; balaa R; B; T; W -- haana T; bhaana missing in W -- bhayanuvaritthia R; parittia k, padithi ya (vt illegible), paritthida B; paditthia W -- samkhasutti Bh, R; K, , B, Y, P; S; samkhicchippi T. Kavyaprakasa, p. 30 (sutti); Sahityadarpana, p. 20 (osutti); Pada a quoted in Hem. II 211. Look! (Standing) absolutely motionless between the lotusleaves, the (female) heron looks like a (white) sippi, a shell, set in (a setting of) spotless emerald: :: In the Gatha the speaker draws a second person's attention to something in his environment, composing as he does a striking comparison. As such the Gatha has exact parallels in, for instance, 62, 63, 64 and 75. The interpretation of the Gatha raises problems which I have been unable to solve so far. According to Weber the speaker draws attention to a quiet place where lovers can meet without being disturbed. This interpretation does not take account of the second line, which refers to something unlikely, namely a shell in a setting of precious stone, viz. emerald. Curiously, the setting is worth more that the thing set. As such the comparison in the second line seems to contain a comment on the situation

Page #202
189 described in the first line. It may, for instance, imply that the balaka spoils the beauty of the place or that it has no right to be there. Apparently the bird is preying, standing absolutely motionless in order not to scare away the fish. It remains unclear to me, though, what precisely the speaker wishes to convey to the person addressed. The particles ua and uaha are used to draw someone's attention to a particular object or person. They are common in the Sattasat (ua in 49, 75, 120, etc. and uaha in *18, 62, 109, etc.) and in the Pkt verses in the Skt dramas (see Pischel in his translation of Hem., pp. 97-8). In the other genres, e.g. the Setubandha and the Gaudavaho, they are absent. Usually ua and uaha are translated with pasya and pasyata respectively (see Hem. II 211 and. Pischel, loc. cit.). These glosses were undoubtedly inspired by their mutual relationship which is similar to that of the imperatives vatta and vattaha. Weber suggests a derivation from the verb uh- (Abh., p. 29, note 4, Ed., p. 4 and word-index, s.v. uh-(2)), which is hardly tenable (see Pischel, loc. cit.). Whereas the North-Indian recensions have ua(-ha) the South-Indian recensions have invariably uva(-ha). For the reason behind the decision to follow the former here, see above, p. 167. 'lotus-plant from Skt Var. II 38 and Hem. I 238 derive bhisin b/v)isini, which seems to consist of bisa 'lotus-stalk' followed by the (possessive?), suffix -int. Skt has bisa, Pali, Pkt and NIA bhisa (for NIA, see Turner, CDIAL 9249). The variant b(/v)isin (here S and W; in 8 T and S) is due to the influence of Skt. According to Mayrhofer (EWA II, p. 433) the existence of bisa beside bhisa would indicate that we have to do here with a loanword from an indigenous Indian language. For the reasons to adopt (bhisint)vattammi, found in S (and y), rather than pattammi found in the other MSS, see above, p. 164. For reha to shine' (Hem. IV 100), from Skt ribh- (rebhati) 'to sing' and 'to shine', see Mayrhofer, EWA III, pp. 73-4. For the meaning of valaa (Skt balaka; originally invariably feminine), 'small white heron', see Thieme (1975: 15-22). The retention of the intervocalic -g- in maragaa 'emerald' in Pkt as well as Apa. (see Hem. IV 349) is unusual. This may have to do with the fact that it concerns a foreign loanword; see Mayrhofer, EWA II, pp. 557-8. A comparison with Greek (o)uapaydos and Tibetan mar-gad, seems

Page #203
190 to show that the voiced -g- in Pkt is more original than the voiceless -k- in Skt marakata. The latter may be hypercorrect. Whether the text reads samkhasutti in accordance with Bh, R, etc. or sankhas ippi with. T and W (for (c)chippi in T, see below), in both cases it concerns a compound of two synonyms for 'shell'. The context suggests that the compound denotes a single entity, which leads to the translation 'a suttt (or sippt), namely a sarkha' (Karmadharaya). As such only samkhasippt makes sense, consisting of saikha, a word for 'shell' which is already long established in Skt, and sippt which is not found in Skt and is thus typical of MIA. For instances in Pali, see PTSD s.v. sippt and sippika, and for Pkt, see Gatha 62 and Setubandha, index s.v. The word sippt is a loanword from Dravidian (see DED 2089). Moreover, the compound sak hasippt seems to have several parallels, namely (vannag)ghaatuppa 'ghee' (see *22), vaivedha(a) 'fence' (see 96, 220, 221, 564) and acchabhalla 'bear' (see 109). Each compound consists of a word commonly known in Skt, ghrta, vrti and rksa, followed by a MIA. synonym, namely tuppa, vedha(a) and bhalla respectively. It would seem that the latter words were still considered new and strange and needed explanations. It is significant that the more original reading, namely saikhasippi, occurs in the two South-Indian eclectic recensions T and W; this points to the conclusion that the Gatha was drawn from a text which was closer to the archetype than that of the North-Indian branch. It is extremely likely that the Gatha originally formed part of the Third South-Indian recension, from which it is now lost. For chippt in T compare *9 chali in Ma, Tp and chana in Ma, Tp and T. ch- in chali may be due to immediately following chettesu. It may, however, not be without significance that in all these cases it concerns an original s: sali and sana, and for sippt, compare Mar. stp or sip. The variants hisini (for bhisint) and haana (bhaana), both found in T, seem to be due to confusion of the Grantha characters for bh and h. 5( 5 Ma, 6 Tp; 4 Bh, 5R; K, , B, Y, P, G, 5; 148 T; 173 S) Missing in W. tava ccia raisamae mahilana vibbh ama viraati

Page #204
191 java na kuvalaasac- chahai maulati naanain. In Ma lacuna from the beginning upto sacchahai. cciam K; via S-- samaye B -- vimbh amas k -- sacchabhai Tp; sachaha in R; K, V, B, Y (ohain), P; T; s-, Sachahaii s -- mayulati Ma; maulenti. w'; enti st -- naanai Tp; nainain R; (, Y). Dasarupa, p. 79. At their love-play women go on coquettishly flashing their eyes until they (, the women, have their orgasm and) close their eyes which resemble the petals of the blue lotus. The derivation of sacchaha 'similar to' (Paiyal. 74: sacchaho samano, and Des Tn. VIII 9: sacchaha...sarise) is uncertain. Those suggested by by Buhler (Paiyal., index, s.v.) and Pischel (Desmn., index, s.v.), from *saksaka, and by Weber (1872: 742), from sa-chaya, are phonetically unsatisfactory. In the language of the Sattas at the original distinction between the verbal e (+ aya) and a-stems seems in the process of becoming blurred, in favour of the a-stems (for this tendency, see Pischel & 491). Most often this is the case before -mti and mta, where the metrical distinction between long e and short a is annulled. Thus, we find here maulanti (from mukulayanti). Otherwise a fore is rare; see, for instance, kahai (kathayati) in 59. Some MSS, and S with great regularity, show a tendency here and there to reintroduce the distinction. See maulenti in v and s, and in 59 kahei in s, which, in fact, spoils the metre. For the reasons to adopt na anain as found in Ma, see above, p. 169. 6( 6 Ma, 7 Tp; 5 Bh, 6 R; K, U, B, Y, P, G, 5; 233 S) Missing in T and W. nohaliam appano kin ean khu tuha hasai suha- na maggase maggase kuravaassa a valiamuhapankaan jaa.

Page #205
192 In Ma lacuna upto (e) am in c. Apparently the variants readings of P have been overlooked by Weber. nohaliyan Bh; dohalia B; noalian sa (nohaliam in Ed., p. 6 probably is a misprint.) -- ayyano K -- kuruvaassa S -- ea K -- kha Ma; tu Bh; vu K, kkhu Y; vvia kkhu S -- hasai tuha suhaa Bh, R; tuha suhaa hasai K, Y, suhaa suhasai U, suhaa hasai tuha B -- viliyamuha Ma; valianana , B, valiana Y -- pamkaja(lacuna) Ma. Pada c quoted in Hem. II 198 (ean khu hasai). Why don't you look for a 'yield of new fruits' for yourself instead of looking for a 'young woman' for the kur abaka-tree? Thus, you know, handsome man, your wife is laughing, her lotus-face turned aside. This Gatha and the following are closely related. The present Gatha refers to the belief that the kur abaka-tree, a red amaranth, blossoms only after having been embraced by a young woman; in * 7 reference is made to the asoka-tree, which needs to be kicked before it blossoms. For these motifs, see, among others, Karpuramanjarf II 43. Furthermore, in the following Gatha a pun is made on the word asoa, meaning 'asokatree' as well as 'griefless'. Likewise in the Gatha under discussion a pun is made, namely on the word nohalia. According to Hem. I 170 this word goes back to Skt navaphalika. In Skt it is restricted to the lexicons, which translate it with 'a girl with whom menstruation has recently begun' and 'a newly-married woman' (PW, s.v.). It has been analysed as a Bahuvrthi-compound, consisting of nava and phala, the latter having the ad hoc meaning 'menses'; see AIG II 2, S 199 dy. The commentaries available on the Gatha give a totally different translation, namely navaphalodgama 'a yield of fresh fruits' or 'a fresh yield of fruits (see Abh., p. 74 for Kulanatha). The word probably is a feminine abstract noun ending in -ika, for which type of noun, see AIG II 2, $210. The situation underlying the first line seems to be the following. The man addressed was in search of a young woman (nohalia), ostensibly for the kurabaka-tree. It is also possible that he looked for one for his own pleasure and used the kurabaka as an excuse. When his wife

Page #206
193 found out she said he had better look for a nohalia for himself. In the latter case we would have to do with the abstract noun, which evidently is used in a metaphorical, or ad hoc, sense here. As such it can have several meanings (note in this connection Omar. navhalt, quoted by Turner, CDIAL 7004, which means 'first fruit of the season' as well as 'prime of youth'). The first sentence may, for instance, contain a reference to the man's old age or even impotence. Hopefully for him the humour is of a milder kind, navaphalika referring to 'children'. The woman says then that her husband does better to invest his energies in her than in the kur abaka-tree. In the second line I have adopted tuha hasai suhaa of Ma, Tp and S (see also E: tava hasati subhaga). The order in Bh, R, K and y, tuha suhaa, with tuha immediately preceding the vocative suhaa, appears more logical, which, however, at the same time serves as an argument to consider it as secondary. Note in this connection eya tu (for khu) hasai tuha suhaa in Bh, which still contains traces of the reading of Ma, Tp and S. In some of the other cases where such a broken-up sentence is found (note also tuha (c)...jaa (d) here) it appears to be due to the particuIar circumstances under which the words are spoken (see the note to *20). The form of the sentence in this Gatha might then be explained by assuming that the messenger, who is laughing about the 'joke' made by the wife, is incapable of formulating a coherent sentence. For valia turned (as ide)' Ma reads vili(y)a. Evidently this is a scribal error or a reading mistake made by the copyist of the transcript Ma. The meaning of vilia 'disappeared' (see 53 paccakkhadifthavilia 'clearly disappeared') hardly makes sense here. 7( 7 Ma, 8 Tp; 6 Bh, 7 R; K, V, B, Y, P, G, 5; 238 T; 40 S) Missing in W. tavijjanti asoeki sahai ko vi kassa vi hi ladahavilaao daiavirahammi paapahar am pahuppato. In Ma lacuna upto paapaharan. tavijjintin K -- asohim Tp; asochin K -- ledaha Tp; land aha K --

Page #207
194 vilayau Bh; vilaa k, viraao V, vania B, vaniau Y, virahao P -- virahasmi k -- paguppan to Ma; pahuspasto Bh; pahuppatto k, pahupanto y, pahupanto P. Even irresistable temptresses let themselves (occasionally) be tormented, (namely) by men who experience no grief when separated from their mistresses. (For) why would someone who is his own master (is the master) suffer to be kicked by someone else? Whereas the usual place of the finite verb is at the end of the sentence, in this Gatha it is found at the beginning. For similar instances, see 562, 692, 693 and 769. According to Gonda (1952: 67) this results in the accentuation of the affirmative nature of the verb. Thus: 'contrary to what one generally believes laqahavilaas can be tormented'. Gatha 817 shows that this indeed happens only rarely. Usually the ladahavilaas themselves are the tormentors. In fact only saints are deaf to their seductive charmes or, as here, people who are completely indifferent to female company: asoehi...daiavirahammi, 'people who experience no grief...when separated from their mistresses'. Gata 817 reads as follows: lagahavilaana loana- jhijjahti mahasatta kadakkhavikkhevajanias aftava cittuvveanas aha hosti. The syntax of the Gatha is misunderstood by Weber. I should like to propose the following translation: Tormented by the side-long glances of the irresistable (ladana) vilaas one (inevitably) becomes thin. (Only) great saints are able to bear the agitation of the mind (such women bring about). In the second line of the Gatha under discussion a pun is made on the word asoa of the first line. Asoka-trees are believed to crave for a kick from a young woman. Only when satisfied in this desire do they bring forth flowers. For this motif, see Karpuramanjart II 43. The

Page #208
asokas of the first line are strange asokas indeed as they do not show the usual reaction to the kick (charms) of the ladahavilaas. In Hem. II 128 the word vilaa is mentioned as a Pkt substitute for Skt vanita. In Paiyal. 12 it is enumerated in a list of synonyms for 'woman', ranging from vahua to niyambint. In both Gatha *7 and 817 the women denoted by ladahavilaa are sexually extremely attractive. Most likely they are prostitutes as they are actively seducing men (see 817). It may be asked if these two connotations, attractive and prostitute, are conveyed by either ladaha or vilaa. The word vilaa is relatively rare. In the Sattasat it occurs three more times and in the Gaugavaho only once (166 varavilaa). It does not occur in the Setubandha. For instances in later texts, see PSM, s.v. The following discussion is restricted to the instances in the Sattasat. 195 In 156 and 890 the compounds vesa- and vesavilaa occur. The meaning 'prostitute' for. vilaa would lead to a redundancy; compare vesavilaa in 890 with Skt vesastri, -yosit, etc., in which the final member denotes simply 'woman'. vesavilaa may be taken as a Karmadharaya; in theory it could also be a Dvandva or Tatpurusa. However, the vesavilaas are no common prostitutes. To the purely perfunctory performance of their profession they have added a special attraction by showing some sort of personal involvement or interest. The clients are fascinated by this as appears from the following two enraptured exclamations. 156 namdamtu suraasarahasa vahumagganimmiaim taghapahari saalaloassa vesavila pa pemmih. Long live the affectionate feelings which the vilaas who are prostitutes know how to show in so many ways, which (alone) are able to completely satisfy the vehement desires for love-making everybody has! For the connotations of the word permai, see the expression premabandhura a prostitute who has, or is able to convince her client she has, a real affection for him', found in Kumarapalapratibodha (ed. Alsdorf), p. 109, verse 76.

Page #209
196 890 sacchadar amanad as ana suviaddhavesavilaa samvaddhiagaruavammahavilas am rami am ko vanniu tarai. Who can describe making love to the skillfull vilaas of the prostitutes' quarters, during which love's joviality, which is already vehement, is further enhanced as they show they do it all spontaneously? The only instance in which the word occurs by itself is 233. It is not clear whether the woman is a prostitute here or not. It is clear, though, that it concerns a woman who at least in her youth was very exciting, her body being compared with the temple of the god of Love. She brought the men she made love with to complete frenzy as appears from the fact that they bit her practically everywhere. thanajahanani arvovari uvvasiananganiva dasananka gaavaana vilaana samulavandha vva dfsamti. The tooth-marks on the breasts, the thighs and the buttocks of old vilaas look like the foundations of a deserted (delapidated) temple of the god of Love. The foregoing instances show that the attractiveness of the vilaas does not depend on the presence of the adjective ladaha. Furthermore, 156 and 890 show that the word vilaa does not by itself mean 'prostitute', though it may be used to refer to such a woman. It would seem that the word means 'woman' but with special reference to her great attractiveness and to the effect she has on men. As an English translation I have therefore adopted 'temptress'. In this connection also the derivation of the word may be considered. I would venture to suggest that it'simply is the feminine of Skt vilaya 'destruction'. In their derivation of pahuppa- 'to be (one's own) master' Weber (Ed., p. 6) as well as Jacobi (1886: 253) proceed from the past participle pahutta, which would go back to Skt prabhuta. Analogically after litta 'smeared' : lippa- 'to be smeared', a passive pahuppa- would have been formed, which has acquired an active meaning. This derivation is

Page #210
197 questionable on two grounds. In the first place no certain instances of hutta from bhuta are known. The latter regularly becomes (b) hua. Secondly, it assumes the existence in Pkt of deponentia of which, however, no certain instances seem to exist either (see the note to *10). Pischel (5 286) suggests instead that pahuppa- is a denominative verb based on the abstract noun prabhutva. Admittedly the sole instance of this noun in the Sattasai, in 91, has pahutta. However, the alternative development of -tv- to -pp- is known dialectically; see Pischel SS 300. In this connection it may be interesting to note that Hemacandra, who in IV 63 mentions pahuppa- as an Adesa of prabhu-, specifically adds that it is used only when the subject or agens (Kartrka) is a prabhu'. However, this derivation creates only a new problem, as no explanation for the origin of the past participle pahutta beside pahuppa- is available. 8( 8 Ma, 9 Tp; 7 Bh, 8 R; K, y, B, Y, P, G, E; 442 T; 487 S) Missing in W. atta taha ramanijjah Tuatilavadisariccha amh am gamassa madanam bhuam sisirena kaam bhisinisamdan. In Ma lacuna from gamassa onwards to the end. amha, Bh -- gomasa Ma -- mamdanaabhudam Tp; mama anabbhuyan Bh, mano an Thuan R; K (haa), V, B, Y, P; mamdan Tbhudam T; S' -- luyatilavada Bh -- sarichan K, sarisan B -- bisinT T; visant S -- mand an K, sandhan Y. Mother-in-law, the beautiful cluster of lotuses that used to be the ornament of our (my parents') village has been treated by the winter in such a way that it has come to look like a harvested sesame-field! According to Weber (Abh., p. 75) the bhisinisamd am and the tilavadi would be places where the woman speaking used to meet her lover, an interpretation apparently also found in the commentaries, which, however, have not been quoted specifically by Weber. It seems unlikely, though,

Page #211
198 that a woman would discuss such matters with her mother-in-law (atta, from Tamil attai; see Turner, CDIAL 222, and Paiyal. 253 atta sasu, but see also below). It is more likely that she complains in guarded terms to her mother-in-law about the behaviour of her husband, and especially about his coldness or indifference towards her, which has physically ruined her 'who used to be the ornament of her parents' village'. The misunderstanding as evinced by the commentaries and Weber may rest on the fact that fields are indeed often referred to as lovers' meetingplaces (see, for instance, Gatha *9). The gloss saht 'friend' for atta in Paiyal. 108 (mauya sahs atta) and in Desin. I 51 (ma i-piuccha-sasusah Tsu atta) may go back to a similar misunderstanding. mand an an bhuam has been reconstructed on the basis of Tp 'mand anabhudam and Bh maid anabbhuyam. A comparison with the phrase unnaehi houna in 83 suggests that bhuam functions here as a kind of emphatic tense marker: 'which was the ornament'. maid an Thuam of R etc. is probably an innovation for the uncertain form in Bh. taha in a has been constructed with sariccham in c, '(it) has been treated in such a way... that it has become like...' (cf. taha...vva in 413). lua is the past participle of Skt lu- 'to cut'. For the short -uin lua, see Jacobi (1898: 569), who explains it from the present stem lunami. Compare dhua from dhu-, dhunoti. For bhisint, 'lotus-plant', see *4. 9( 10 Ma, 11 Tp; 8 Bh, 9 R; K, U, B, Y, P, G, 5; 455 T; 488 s) Missing in W. kin ruasi onaamuhT harialamand i amuht dhavalaatesu salichettesu nadi va sanavadia jaa. kin ruyas i Ma; ma ruasu Bh, R; S -- chalTcchettesu Ma, chalichettesu Tp; cettesum T; salikkhettesu S -- muha Bh; mukht B -- pada Bh; in Tp the final vowel of muht and the first syllable of nadi are illegible -- cchana Ma, chana Tp; cchana T -- savatuya Bh (sanavadiya Bho). Why are you crying with your face bent down now that the .

Page #212
199 rice-fields are turning white? The hemp-field (its surface decorated with yellow flowers) has become like an actress whose face is decorated with yellow paint! The girl is crying because the rice-fields are about to be harvested after which it will be impossible for her to meet her lover there. She is comforted by her friends who point out that the hemp-field, which is in full bloom, could be used instead. For hari(t)ala, 'yellow orpiment', used by actors as paint in dramatic performances, see MW, s.v. Unfortunately no sources are given there. For the variant chali (Ma, Tp) for sali, and for chana (Ma, Tp, T) for sana, see *4. 10( 11 Ma, 12 Tp; 9 Bh, 10 R; K, M, B, Y, P, G, 5; 292 T; 276 S) Missing in W. sahi erisa ccia gaT eana valavalun ma ruvvau tasavaliamuha adan kitamtukudilana pemmana. erisi Tp; R; K, M, B, Y, sa Pi sarisi T; erisi S -- vviya Bh; via S -- gain K, P(?) -- ruvvasu Ma, Tp; R; K (ruccasu), , B (-cc-), rua . v, ruvasu P; ruvvasu T; ruccasu s-, ruchasva st -- tankhalia B -- eana K -- bala Ma, Tp; R; B, Y, cala K, P; bala T -- balukki Tp; valukvi K, valukki , valukki B, valuki P; valukki T. -- bhumtu (for tantu) Bh -- kudilanan k -- pemmanan K, B, P; T; S; (R, W, Y). Pada a is quoted in Hem. II 195 (erisi, v.1. erise, ccia). Friend, stop crying with your moon-face turned as ide: this is the course of these affectionate feelings which 'curl' (indiscriminately around everyone who is available) like the tendrils of the young cucumber-plant. In the case of pronominal adjectives such as erisa the feminine marker, as in Skt, generally is -1. Pischel ($ 245) quotes jadisi,

Page #213
200 tadist, etc. Pali, too, has tadist. Berger (1955: 42) explains the development of -r- of tadrsa into -i- instead of into -a- in tadisa as an instance of assimilation to the final -I in the feminine tadist. The feminine marker -a is extremely rare in these words. Pischel (loc. cit.) quotes only sarisaya (Nayadh.). To this may be added Pali ktdisa in yadisikTdisa jivika (PTSD, p. 553C) and Apa. erisa in erisa joda na jutta (Kumar apalapratibodha, p. 107, verse 66). In the text therefore erisa of Ma and Bh is adopted, the change of the latter into erisi (TP, R, etc.) being much more likely than that of erisi into erisa. We should, however, be aware of the possibility that erisa is a clerical mistake for erisi with -sa from the preceding word sahi. In the Sattasas the word tassa occurs twice more, namely in 148 where it denotes a certain enticing way in which the eyes move (tasavalamt addhataraaloa) and in 692 where it denotes a particular position of the feet (heels down and toes up'). In Paiyal. 269 it has been translated with tiriccha 'oblique(ly), slanting'. Hemacandra (1 26 and II 92) derives it explicitly from tryasra (synonymous with tryasra) 'triangular'. A comparison with caturasra(ka) 'a particular position of the hand, a particular posture in dancing' (from 'having or making four angles') would suggest for tamsa a meaning 'making three angles (on (the rest of) the body)'. It would subsequently have acquired a somewhat less technical sense, like 'slanting' or 'oblique'. valuhk T, 'a kind of cucumber'. Skt has masculine or neutre valurka (MW, s.v.), feminine valunkt being found only in the lexicons. Paiyal. 172 has valunk am and Mar. valuka (quoted in Ed., p. 7). Feminine valuikt is probably ad hoc, the speaker by mentioning this plant referring to a woman. ruvvau is found only in Bh (Bhuvanapala: rudyatam). The corruption of ruvvau into ruvvasu, found in the other MSS, probably rests on a confusion between the grammatical and the logical subject, the latter being a second person. In his Abh.(p. 6) Weber, to whom at the time the variant ruvvau was not available, explains ruvvasu as a deponens. In his editions this explanation turns up several times more. However, a closer look at the instances concerned shows that most if not all can one way or another be eliminated. It follows that as far as the Sattasaf

Page #214
201 is concerned deponentia do not exist in Pkt. It should be noted that Alsdorf (1937: 61-8) after studying those verbs in Apa. earlier identified as deponentia, reached a similar conclusion regarding that language. An obvious case is tirae. This is not, as Weber (Abh., p. 64) assumes, a deponens but a regular passive of the verb tara-, the usage of which is comparable to that of Skt sakyate. gasijjihii in 804, likewise, is a regular passive. Gatha 804 (only found in S) reads as follows: varijjartT navako- ma te amvupisao mui tti ma putti amgane suvasu caido tti muham gasijjihii. If someone tries to stop you, saying '(it is) full moon', you should (obey and not sleep in the courtyard, little girl, lest your face will be swallowed by Rahu) who may think the (real) moon (only) a pisaca in the water (arvupisao). A different case is pijjai in 678. Gatha 678 (only found in w and Y ; and in G and ) reads as follows: pijjai kannanjalihim duddha jalas amilia janar avamilia pi tujjha salavam sa vala raahamsi vva. As far as the first half is concerned the passive pijjai is in perfect order. An anomaly arises only in connection with the second half. One may therefore consider the possibility that the Gatha is a patch-work verse assembled, with little regard for syntax, from two independently existing half-verses. Note in this connection the awkward repetition of the word (Sam)milian. The fact that Hemacandra in IV 10 includes the passive stem pijja- among the Adesas for pa- is no independent support for Weber's (and the traditional) interpretation for pijja- in this Gatha. The inclusion of pijja- in Hemacandra's sutra may in fact rest upon a misunderstanding of an instance like this one. Another instance of the 'deponens' pijja- quoted by Pischel (in his translation of Hem., p. 132), namely pijjanti in Sakuntala (ed. Pischel, p. 29, line 5) is in my opinion a regular passive.

Page #215
202 gammasu in 819 should be eliminated in favour of gamesu which has become available through Ma, Ti and Tp. ruvvasu as a corruption for ruvvau is not unique. Another instance is gammasu in 715 in T, for gammau in the other MSS available. It also explains gammihisi in 609, which against the evidence of all the available MSS should be emended into gammihii. A possible third instance is bhannihisi in : 906, which, alternatively, is a regular passive, meaning 'you will be told'. The case of lagga-, pahuppa-, vacca- and dhakka- (Abh., p. 64) is different yet again. The final -kka in dhakka-, 'to cover', is not from -kya but a MIA verbal suffix (see *14). lagga-, 'to cling to', and vacca'to go', go back to *lagyati and *vrajyati (IV class), rather than being derived from the respective passives. lagga- can also have been based on the past participle lagga (lagna) (the origin of the voiceless cluster -CC-in vacca-, instead of expected -jj-, will be discussed elsewhere). Finally, the instance pahuppa- is, as has already been noted in *8, doubtful, the derivation still being far from certain. 11( 12 Ma, 13 Tp; 10 Bh, 11 R; K, V, B, Y, P, G, 5; 590 T; 322 s) Missing in W. paavadiassa paino didhamannudumiae vi patthim putte samaruhamt ammi haso gharinte nikkamto. paapadiassa Ma (paa), Tp; R; K, M, B, Y, P; T; S -- patthim paino Tp; putthim R; K (oththo), V, B, Y, P; T; putte patthim S -- ruhamt am Ma, Oruhait ainmmi Tp -- dadha Bh, R; K, V (datha), Y, P; S -- mamna Bho, manna Bh"; manu K -- dumiai Y, dummiae P; dumiai S -- ghaso Ma -- harinte Ma; gharinti Bh; gharinta B, Y; S -- nikvasto K. When their son climbed on her husband's back who had fallen at her feet, the wife could not help laughing, even though she was tormented by a fierce anger. For the reasons to adopt paavadiassa found in Bh rather than padiassa in the other MSS, see above, p. 164.

Page #216
203 patthi, 'back', corresponds to Skt prstha. The final -T has been explained as arising from contamination with pitthT (Skt prsti) 'rib' (see Andersen, Pali Glossary, s.v.). Ma, Ti and Tp throughout have patghi, with -a- for -r-. Likewise T, except here (T also often has pattha (+ prstha), for which, see * 35 (also W), 87 and 223). Bh varies: patthi here and in * 35, 216 and 223, but putght in 313 and pittht in 87. The other MSS generally have putthi (note, however, that S has patthi here). The -u- for -r- is due to the influence of the preceding labial. It should be noted that pattht (or putthi) is often found rhyming with pa- as well as with pu-: *11 paino patthin putte, 35 patthin palfvesi, 87 patthipulauggamo, 216 patthie (c), pulaubbhee (d), 313 patthin p(h)usasu, and 676 padara pattht. For drdha Ma, Ti, Tp and T have throughout didha. The other MSS usually have dadha (exceptions: didha in *11 B, 108 K, Y, P and 276 K, S, (x)). For the principle underlying the decision to adopt the forms of the South-Indian recensions here, namely patghi and didha respectively, see above, p. 165. . For didhamannudumiae vi, see also 74. Var. VII 8 and Hem. IV 23 mention dume- (actually they have duma-, but the grammars quote all verbs as a-stems irrespective of their actual conjugation) as an Adesa for Skt du- 'to burn, to afflict'. The verb occurs frequently in M. Pkt and also occasionally in the Panhavagaranaim, a relatively late Amg. text. The most common glosses found in the available commentaries and Chayas are du-, duhkhay-, vyathay-, today- and tapay-. The instances of the word show that the affliction brought about is both mental and physical. For the former meaning, see, for instance, Gathas 792 (dumesi manam) and 937 (dumes i maha hiaam) and Panhavagaranain 2, 17 (hiyayaman adumakain). A clear case of the latter meaning is found in Panhavagaranaii 1,30, where the expression paharadumana occurs in a list of mutilations, such as atthibhasjana, nas abheya and chaviccheyana. Note, furthermore, Setubandha XIII 96 (riusuladumio). In the same text, in V 24 (anuampaduminana) and A 8 (saravian adumianana), it seems to me an 'to distort, to disfigure'. It is also possible that dumia in the latter two cases is in fact a corruption for dusia (Skt dusita). These two sets of meanings refute the derivation, first suggested by Weber (Ed., p. 8) and later elaborated upon by Schwarzschild (1962:

Page #217
204 70), of dume- (Schwarzschild 'se desesperer') from Skt durmanas 'in low spirits'. This derivation assumes a development from 'to afflict mental pain' to 'to afflict physical pain', which is not very likely. Furthermore, formally (dummana + dummia + dumia - dume-) .it rests entirely on the variant dumme - occurring in some of the MSS of the Sattasaf. This dumme - almost certainly is an innovation, probably introduced only. after the link between dume- and durmanas was made. Instead, Turner, CDIAL 6493, suggests to derive dume- from a non-existent * duman 'distress', a noun based on the stem du-. However, by doing so the problem of the derivation of the verb dume- remains unsolved. Beside dume- 'to afflict' Hemacandra, in IV 24, mentions a verb dumameaning 'to whitewash', which would as such be a variant of duma-, for which latter verb, see Gatha 747. duma- 'to whitewash' may, however, well be the same verb as the one discussed above, its meaning having been arrived at ad hoc, in which its resemblance to duma- may also have played a role. The only instance of duma- 'to whitewash' I have been been able to trace back 'is Kalpasutra, p. 42, where it occurs in a standard description, also occurring in Nayadhammakahao 1, 1, 18, of a bed-chamber the outside walls of which are dumiyaghattamatta. In his word-index Jacobi gives for dumiya the gloss dhavalita. It should, however, be noted that dumiya, which, as has been shown above, denotes, among other things, some more or less rough and painful treatment, occurs here side by side with the participles of two other verbs (ghrs - and mrj-), which denote more or less rough treatments as well. 12( 14 Ma, 15 Tp; 11 Bh, 12 R; K, B, P, 13 V, Y, 12 G, 5; 313 T; 472 S; 6 V) saccam janai datthum marau na tuman bhanissam sarisammi janammi jujjae rao maranan pi silahanijjam se. In Ma lacuna between da(tthun) and (sa)risammi. du(thun) Ma; dadhdhun K -- raom k -- marane tuma Ma -- silanagijjan Ma; salahanijjah Bh, R; K, M, B, Y, P; T; S. Dasarupa, p. 29 (salao).

Page #218
205 I must say, she knows what looking is! Passion for a man who is one's equal is right (but not for a man like you). Let her die! I will not tell you (what she asked me to tell you). Even such a dreadful thing as death is to be welcomed by her (in this situation). A messenger addresses her friend's lover. The girl has threatened to die if she does not meet him. For this threat, see also 178 (= 739), 588 and 604 (Pada d in 604 reads marau na bhanissan). However, when the messenger sees that the lover is unworthy of her friend she decides not to intervene. In her opinion the girl is better off dead than alive. saccam in saccam janai datthum is evidently meant cynically here. In the source of Ma the words marau na seem to have been partly or wholly lost, as would appear from the fact that in the accompanying Chaya these two-words have been skipped. mar ane looks like an ad hoc attempt to fill in this lacuna: 'at her death (I will tell you)'. : For silahan ijja, found in Ma and Tp (see also 114 Ma, Ti, Tp and i (s11deg)), beside salo of the other MSS, see Pischel SS 133. For the principle underlying the decision to adopt the form found in the South-Indian recension, see above, p.165. U, B, Y, G, 13 5; 389 T; 139 s) 13( 19 Ma, 20 Tp; 12 Bh, 13 R; K, P, 14 Missing in W. randh anak amman iunie muhamaruan pianto ma jhurasu rattapadalasuandhan dhumai sinf na pajjalai. katmma Ma -- jurasu Ma, Tp; Bho, R; Y, surasu B; jurasu T; S -- padali S -- suamdh1 Ma; suamdhin T -- pivasto Ma; pibato B -- vijjhai Bh; dhumahi P; dhumei T -- pajalai Ma. There is nothing wrong with your cooking. Don't be annoyed. The only reason the fire doesn't start to burn but keeps smoking is that it wants to drink your breath which is as fragrant as the red patala-flower!

Page #219
206 A woman, probably newly married, is teased with her clumsiness in lighting the kitchen-fire. A woman's clums iness in the kitchen is also the subject of the next Gatha and of 732, 733, 734 and 735. jhurasu of Bha, K, V and P is preferred to jurasu of the other MSS. Hemacandra mentions the latter verb in IV 131 as an Adesa for khid- 'to be annoyed' (together with visura-, which explains the variant surasu in B) and in IV 134 as an Adesa for krudh- 'to be angry'. jhura- is mentioned in IV 74 as an Adesa for smr- 'to remember, to think of'; it would seem that Hemacandra thus tried to account for the meaning 'to worry' (see below). Turner derives jura- from jvar- 'to be feverishly hot' (CDIAL 5259) and jhura- from jr- 'to become old' (5409). He suggests that the initial jh- of jhura- is due to the influence of jhi- from ksi'to destroy'. It is doubtful, though, that we have here to do with two different verbs. In the MSS they occur as variants. Admittedly, jhura- is rare. Thus, in *41 jhura- is restricted to K, in 570 to R and in 610 to y. In the other instances of the word, viz. 354, 454, 695 and 822, all available MSS have jura-. Compare the situation in the Setubandha: in XIV 1 jurai (v.1. jhurai), in XI 3 jurai (v.l. kuppai) and in XI 1 cittavia (i.e. cintavia; v.1. juravia). The latter instance seems to undermine Hemacandra's distinction between jura- 'khid-, krudh-' and jhura- 'smr-'. As already said above his classification of jhura- as an Adesa for smr - seems an attempt to account for the meaning 'to worry', which as Setubandha XI 1 shows is equally well conveyed by jura-. In this connection it is not without significance that jura- does not have any NIA derivatives; see Turner, CDIAL 5259. Turner mentions only Prasun zora 'to be angry', which could, however, just as well go back to jhura-. In contrast to jura-, jhura- has survived in most NIA languages; see Turner, CDIAL 5409, who quotes P. jhurna 'to repine', H. jhurna 'to waste away, dry up'. OGuj. jhurai 'to repent' and Guj. jhur vu 'to languish'. I should therefore like to suggest that Pkt jura- is in fact a mere ghost-word. However, the reasons that may have been behind the supposed change of jhura- into jura- remain largely unclear. It may have been due to a simple orthographical problem, the subscribed u having blurred the distinction between the (Jaina) Nagari characters for jh and j.

Page #220
207 In the commentaries on the Sattasaf and the Setubandha there is no evidence of any etymological speculation concerning the word, which might have resulted in a change in form. The usual glosses, khid-, krudh- and kup-, are translations rather than etymologies. For this it would seem that we have to turn to Skt which has jur- 'to be angry, to be annoyed' (only once, namely in Bhattikavya XI 8). This jur- could be a loanword from Pkt, but in the light of what has been said above regarding Pkt jura- it is more likely a loan-translation, showing that jhura- was etymologically connected with jvar- 'to be feverishly hot'; compare jur- with jurni 'glowing fire', which according to the lexicons would also mean 'anger'. suadht in Ma stands for suandhin (cf. T) or for suahdhan, -I, -in and -am in the Malayalam script being hardly distinguishable. padala, 'Stereospermum suaveolens'. For the variant pacalt in S, see also 468 (Y, T), and 621 (Y). Compare Pkt halidd beside Skt haridra in Gathas 58 (Ma, Tp, Bh), 80 (Ma, Tp, Bh, R) and 246 (Ma, Bh, R). haliddt (see Var. V 24 and Hem. I 88) is ascertained for the archetype in these instances. Each time, though, a variant halidda is available, undoubtedly introduced under the influence of Skt haridra. The case of the variation of padala and padalt is different. It is not possible to say with certainty that pacala was secondarily introduced under the influence of Skt for padali as Skt has patalt as well, even though it occurs only relati.vely late in that language. :14( 15 Ma, 16 Tp; 382 Bh, 14 and 386 R; 14 K, P, 12 V, Y, 13 B, G, 14 EUR; 388 T; 203 S) Missing in W. ghar inte mahanasakam- chikkan muhan hasijjai malaggamas imailiena hatthena candavattham gaan paina. ghar inie Ma; na R (14, 386); B, Y; S -- kama Ma -- mailienan K, maliena Y; mailena S -- cchakkam Ma, chittan Tp; chikvan R (386), chittan R (14); K (chimnnam), V, B, Y, P (ci"); cikkan T; chippan S -- hisijjai Y, hasijjayi P. The husband laughed at his wife's face which, touched

Page #221
208 with her hand dirty with soot from working in the kitchen, had come to look like the moon. mailia 'made dirty' is a past participle based on the adjective maila 'dirty', which is from malila (mala+ ila; see Turner, CDIAL 9904). For mas imailia, see also 70; mas imaila in 117. chikka is the past participle of chiva- 'to touch'. For the suffix -kka, see Schwarzschild (1958). Hem. IV 182 mentions chiva- among the Adesas for sprs= 'to touch'. The derivation is uncertain. Turner, CDIAL 5055, suggests that it goes back to chup- (IE skup-) contaminated with ksip-. By way of alternative I should like to suggest that chiva- goes back to Skt ksip- 'to throw, to strike', which became khiva- 'to throw', on the one hand, and chiva'to touch', on the other. The meaning 'to touch' would have developed from that of 'to strike'. For the existence in Pkt of two alternative forms going back to one Skt word, each with a different meaning, compare khana 'moment' beside chana 'festival' (from 'right moment'), which both derive from Skt ksana (see *42). The variant chitta shows the influence of Skt ksipta. 15( 17 Ma, 18 Tp; 13 Bh, 15 R; K, B, P, 174, 16 Y, 15 G, 5; 653 T; 611 S) Missing in W. kin kin de padihaai padhamullaadohalini- sahihi ii pucchiae muddhae e navari daian gaa dittht.' In Ma lacuna from the beginning to (padha )muo. kim ki y -- te Tp; R -- padibhaai Tp; padihasai Bh, R; U, B, Y, padiha k, padiasahi P; padibhaai T; padihasai S -- sahihi R; K, Y, P; sahti S -- ia Tp; Bh, R; K, 4, B, Y, hia P; i T; ia s -- puchiyai Bh, pucchiai R; Y, puchiae K, y, puchiIa B; puchiai S -- muddhai S -- (lacuna)mullua Ma, padhamallaa Tp; padhamu(lacuna)halinie Bho, padhamunnaya Bh; pamthamullaa K, padhamuggaa B, Y, pathamullaa P; padhamullia T; padhamuggaa S -- dohanie R; dohaninta y ; dohalinta S -- nava Ma, navara Tp; Bh, navaram R; namvar am K, navara , B, navaran P; T; navara S -- dimmi Ma; diththt K.

Page #222
209 Pada a quoted in Hem. III 80 (kin ki te padihai). When her friends asked the young wife who was pregnant for the very first time and experienced longings (common to pregnant women): 'whatever do you like?', her eyes fell on her husband and on nothing else! The word dohala (Skt dohada) refers to a woman's cravings for unusual kinds of food during her pregnancy. The woman in this Gatha, who is pregnant for the first time, apparently mistakes the nature of her cravings, thinking they are a desire for her husband. The Skt enclitic pronoun te 'you' is found as de and occasionally after vowels as e. The instances of de and e are the following. 547 kim de (S; te in all the other MSS), 609 muhan de (Bh, , Y, T, S; te Ma, Ti, Tp, R). In these two instances te in the South Indian MSS, which often have -ind - for -it- (677 sandava T, Ma, Tp), may be hypercorrect. Other instances are: 554 tam pi de vahuan (Ma, TP, X, Y, T, S; te W; he Ti), 895 atthi de kajja (Ti, Tp, T), 896 kkhu de (Ti, Tp, T; me Ma), 946 je de eam (Ti, Tp, T), 948 jaha de pia (T; te Bh, R, S). Instances of e are: 683 ma e mannu (Ma, TP, R, T), 734 kaha e (R; de T), 737 vie (R; de s), 908 aha e icchai (Ma, Ti, Tp; de T), 925 kaha nu e 'mano (Ti, Tp; de T). The same development can be seen in the pronouns dena 'by him' (tena) and e 'they' (te). For dena, see *29 (T), 164 (Tp, T), 276 (R), 438 (S), 627 (1), 651 (T) and 905 (T). For e, see 732 chivasti e jala (Ma, R; de Tp; te T, S) and 733 samdhei e jala (R; te T). de for te is the result of the lax pronunciation which goes with the often unemphatic use of the pronoun, which is especially the case with their enclitic forms. This phenomenon has been discussed by Turner (1928). As such the development of te into de has parallels in, for instance, vi (api) and dava (tavat). The instances of e could be due to accidents in the transmission of the text, the original initial consonant, d- or t-, having been deleted unconsciously as it occurred between vowels On the other hand, the loss of the initial consonant could be the result of the same phenomenon discussed above. As in the case of de there are several parallels, namely a (ca), i (vi; see *17)

Page #223
210 ira (kira; see *36), a (ta(t)), una (puna) and u (tu). For e, the personal pronoun of the second person, generally a variant de is available (throughout in T and once in S) except in 683, in which case e was misunderstood (see above, p. 163). This marks de as the innovation. The nature of the innovation is in itself peculiar in that it is another typical Pkt word. For the possible use of de, the personal pronoun of the second person singular, as a kind of particle, see *16. padihaai (TP, T) corresponds to Skt pratibhati. pad ihasai (pratibhasate) found in the other MSS is most likely originally nothing but a clerical error, the -s- having been intruded from the following word sahThi. It was retained as it did in no way disturb the original sense. For ii of the text the available MSS read i(y)a. Var. I 14 and Hem. I 90 (see also I 42) mention this ia as a substitute for Skt iti. Its occurrence would be restricted to the beginning of a word (padadau; Var.) or to the beginning of a sentence (vak yadau; Hem.). As such it differs from the other substitute for iti, namely (t)ti, which functions as an enclitic. Pischel ($ 116) looks upon ia as a direct continuation of Latin ita. However, a comparison with AMg. shows that the situation is somewhat different. In AMg. one finds beside iya a form ii, which goes back directly to Skt iti. In North-Indian, and especially in Jaina MSS, the syllables -i- and -ya- are frequently interchanged, which is due to the fact that they are very similar in (the North-Indian) pronunciation. This explains iya beside ii. It should be noted that Pischel (loc. cit) leaves open another possibility, namely that ii in Amg. stands for original iya. It is very doubtful, though, if a copyist would accept such an awkward form as ii unless it was already there. Compare the case of the ending -(y)a in kodi(y)a, for koqti, which latter form was apparently considered awkward, discussed in * 2. The occurrence in M. Pkt, throughout, of i(y)a would only show that the MSS of most of the earlier M. Pkt texts, go back to North-Indian MSS. This is certainly the case with the MSS of the Sattasaf. For the occurrence of i(y)a in such later texts as the Kams avaho (see I 55) the grammars may be held responsible. In the text I have everywhere adopted ii against the unanimous evidence of the MSS in favour of ia. .

Page #224
211 Vararuci and Hemacandra make a clear distinction between navara(n) 'no other, nothing else' (Var. IX 7 and Hem. II 187) and navari(in) 'immediately after' (Var. IX 8 and Hem. II 188). Hemacandra in his own commentary on II 188 adds, however, that according to 'some' navara (in) and navari (n) are synonyms. In the Setubandha the two meanings are neatly sorted out between the respective words. Note that once, though, in IV 78, navara 'no other' has a variant navari. In the MSS of the Sattasaf the situation is less regular. The words occur nine times in all (I leave out the instances in 515 and 527, where they are found only in Bh and R as variants for nama and taha vi respectively of the other MSS), each time meaning 'no other, nothing else'. The majority of the MSS have navara(in), pavari () occurring as an isolated variant:* 15 navari Y, *34 Bha, 172 V, Y, S, (x), 248 S, 303 Bho, S, 485 R (navari), Ma (navarin), 614 y. Gathas 875 and 953 are found only in T, which has navari. Thus in the Sattasai navara(m) and navarim) are merely variants of the MSS. It should be noted that the occurrence of navari (n) beside navara(m), though each time restricted to only one or two MSS, is too persistent to allow to consider the word as a mere scribal error. What we see is in fact the substitution of navara(m) by navari(n) or vice versa. As far as the Sattasas and the Setubandha are concerned the variation is significantly restricted to the instances where the word means 'no other, nothing else'. A confusion of the two respective contexts is in fact unlikely to have taken place, navari 'immediately after' apparently occurring only in the phrase navari a (for instances, see Setubandha, word-index, s.v.). As a result navari a was taken as one entity; see paiyal. 17: navariya...sahasatti. In the grammars navari is recognized as an independent word, meaning 'immediately after'. From these points of view navara(mn) is more likely the innovation here than zavari(i). The situation may also be approached from an entirely different angle, namely by putting the question whether we have here to do with the substitution by an independently existing synonym or, alternatively, vith a change or emendation of the original word. Evidence bearing on this problem, again, points to the secondary nature of navara(in), the independent existence of a word navara being highly uncertain. Thus,

Page #225
212 in Pkt adverbs ending in a bare -a are virtually non-existent. The supposed instances in Gathas 113, 153 and 256 are in reality first members of a compound (navara should be kept apart from such words as jaha and taha, which belong to a different category. Moreover, in jaha and taha the alternation is not between -a and -am as in navara(m), but between -a and -a). The evidence given above points to the conclusion that originally there existed only one word, navari or navarin, meaning both 'immediately after' and 'no other, nothing else'. The variants pavara and navaram may have been introduced on the basis of etymological speculation. An 'etymology behind this may have been the Skt phrase na param, which was in fact suggested by P. Goldschmidt (as quoted by Pische. S 189) as the possible source for navaram). Goldschmidt derived navari (m) from another such combination, namely na pare. I should like to suggest that navari(in) derives from na and vari(i), the latter an apocopated form of avari which goes back to Skt upari (see Pischel SS 123; for vari, see the word varilla 'overcoat' quoted there). This navarim) may be assumed to have the following meanings: 'no higher, no further, not afterwards' out of which the meanings 'no other, nothing else' as well : as 'immediately after' (from 'and ((c)a) not (long) afterwards') could have developed. The word navara(m) thus formed may, partly or wholly, have replaced navari(m) in the MSS of Pkt texts. On the basis of its occurrence in these texts it was recognized as an independent word by the grammarians. This, in turn, may have had a further 'normalizing' effect on the texts. Thus, we see a highly regular situation in the MSS of the Setubandha. In the MSS of the Sattasaf navari(m) occasionally appears to have escaped normalization. These instances have in the present edition been treated as retentions. padhamullaa 'very first' consists of pachama and -ullaa. The suffix denotes not only possession but also intensity, as here (see Pischel $ 595 and Schwarzschild, 1962: 518). For padhamullaa, see also 187 and 218. For padhamuggaa (prathamodgata) of B, Y and S, see ibidem, MSSV , Y, P and S.

Page #226
213 Y, 16 G, 5; 261 T; 356 S) 16( 18 Ma, 19 Tp; 14 Bh, 16 R; K, W, B, P, 15 Missing in W. amaamaa gaanasehara chikko jehi piaamo raanTmuhatilaa cada de chivasu mamam pi teh is cia karehi. amia K, B, P; T -- gagana K, P -- seara TD -- cchanda Ma -- he TP; B; T -- chivasu Ma; civasu Ta -- chitto Tp; Bh, R; K (chinno), , B, Y, P; S; cikko T -- jehi Bh; K -- pio Bho, pi pio Bho -- manh Ma -- tehi ccia K, teh in vvi P; tehi via S -- karehim K, B, P; T; S; (BH, R, W, Y). 0, moon, consisting of nectar, crest-jewel of the sky, tilaka on the face of the night, will you please touch me with those same beams with which you touched my lover! According to Hem. II 196 the particle de is used in addressing a person and in a friendly request, which would correspond to its occurrence before vocatives (see, for instance, *20) and before imperatives (as here.) respectively (see also Dundas, 1981: 205). It is tempting to connect this de with the enclitic form de of the personal pronoun of the second person, discussed in *15. The latter is used in this text specifically for the genitive-dative. It should be noted, though, that the enclitics te and me (Skt) are basically kinds of stem-forms unmarked for case. This explains their use, occasionally, for other cases than the genitive-dative as well. For the use of these enclitics, see AIG III S 235. .:. For the verb chiva- 'to touch' and the past participle chikka, see*14. , 17 G, E; 263 T; 367 S) 17( 20 Ma, 21 Tp; 15 Bh, 17 R; K, B, Y, P, 18 Missing in W. . ehejja so pauttho ii kassa i phalai mano- ahaan kuppejja so vi anunejja rahana mala piaamammi. ehijja so pauttho Ma, Tp; R, ehai so pauttho Bh; ehijja so pauttho K, P,

Page #227
Saham pi Tp ehii so vi pauttho, B, ehai so upaduttho y; ehijja so pauttho T; kuppajja Ma, Tp; kupijjha Bha; kupejja K anupajja Tp; anigijjha Bha so pi K ia Ma, Tp; Bh, R; K, , B, Y, P; T; S halai Ma kie vi Ma, Tp; kassa vi Bh, K, V, B, Y, P; T; S manomuhaNGa Ma piaamanmT Ma. -- -- -- 214 -- May he come from abroad, may I be angry at him, and may he try to please me. (Perhaps) for some (lucky) person this string of wishes with regard to the dearest one is fulfilled. -- -- The optative ehejja, of which no other instances are known, is peculiar in that it is formed on the future stem ehi- instead of, as is usual, on the present stem e- 'to come' (see Pischel SS 529). It should be noted, though, that if the optative had been formed on the stem e-, the verb itself would have become virtually unrecognizable, completely disappearing in the suffix. The context definitely requires an optative. Therefore the future ehii in ehii so vi pauttho, found in Bh,, B and Y (for ehai in Bh and Y, see Berger, 1955: 79, note 6), for the unusual optative ehejja, is almost certainly secondary. A different problem concerns the spelling ehijja (for ehejja; cf. kuppejja and anunejja). This spelling most likely goes back to the MS of the archetype (see above, p. 172). However, it is not clear why in, for instance, Ma and Tp the -e- in this word was not restored for the i-, as it supposedly was in the other optatives in this Gatha. For kuppajja in Ma and Tp (for kuppejja) and anumajja in Tp (for anunejja), see 1. For ii 'thus', see *15. kassa i is based on R. For the principle underlying the decision to adopt this reading found in R, instead of kassa vi found in the other MSS, see above, p. 163. For other instances of i for vi, see 24 do i (T), 149 jTviasesa i, 255 do i (Bh, R), 304 vamke i (R), 534 mudhahiai (Ma) and 904 do i (T). Te vi in Ma and Tp (Madhavayajvamisra: kasyascid dhanyayas; the Chaya in Ma has kasyapi) seems the innovation, The possibility of a reverse change of kTe into kassa is less likely in the present context.

Page #228
215 The variant halai in Ma for phalai is most probably due to confusion of the Malayalam characters for ph and h which are very similar. 18(16 Bh, 18 R; K, B, Y, P, 19 y, 18 G, 5; 81 s) Missing in Ma, Ti, Tp, T and W. duggaakudumvaal/itthi das iosaramtasalile- kaha nu mae dhoiena sodhavva na uaha runnam va padaena. kudarba R; kudanva v , kudumba B, P, kudduna y -- aftht Bh, R; iththt K, Y, addht , aththt B, ittht P; aththt S -- dhoviena Bh; toiena P -- iti sothacca K -- ors ar ata k -- runa K, B, tuidam P; runa S -- cca K. "How can I, who have been washed, bear the al/i)tthi of a poor family.' Look, it seems as if the cloth cries, water dripping from the seam. The meaning of kudumvaatght or kudumvaittht is unclear to me. The variants seem related through the spelling vayatth. The commentaries gloss aftht or ittht with krsti, which Weber translated with 'pulling'. It is not clear to me, though, what might be meant by this 'pulling of (or by) a poor family'. It should be noted that the commentaries assume that the initial plosive of krsti as the second member of a compound is lost. krsti is then not the only possibility. itth1 may stand for, for instance, drsti 'glance' or visti '(forced) service, drudgery', and at tht for, for instance, patthi 'back'. . The occurrence of the ka-suffix twice in the second line alone, namely in das ja (Skt dasika, from dasa) and in paqaa (Skt pataka (L), from pata) is significant. It is also possible to recognize the suffix in kudumvaal/ittht, which may be divided into kutumbaka and tth1 or al/i)ttht. The meaning remains, however, just as unclear as with the possibilities mentioned above. Besides, it raises another problem, namely what implications, if any, the occurrence three times in one Gatha of the ka-suffix could have here. For dho(v)a- 'to wash' two explanations are available. Pischel $ 482

Page #229
216 suggests it is from dhu(v)a- 'to shake, to wash', which became secondarily inflected according to the I present class. Turner, CDIAL 6886, suggests, instead, that it goes back to *dhauvati, replacing dhavati after dhauta. For uaha 'look', see *4. 19( 21 Ma, 22 Tp; 17 Bh, 19 R; K, U, B, Y, P, G, 5; 498 S) Missing in T and W. kos amvakisalavannaa hiaecchia gharam vac- taNGNGaa unnamiehi kaNGgehi camana dhavalattana pava. kos umbha Ma, Tp; kosambi B, kos arhbo P (BH, R, V, Y) -- kisalaa 'u , B, r (deglaya), P; S -- Carinaya Bha (varnaya Bh); vanga y -- tanaya Bha (tarinaya Bh); omitted in Ma -- ukvaniehih R -- kimhnehi Bh; kannehin K, B, P; S (R, W, Y) -- ehiaaccia Ma, hiaacchia Tp; hiyaichiya Bh, R; hiaatthia K (oththo), B, Y (oththo), P, hiaichian ; hiaatthian S -- vacchamaNGa Tp. Calf, coloured like the leaf-buds of the kosamra-tree! (Before) you go, with your ears pricked up, to the house of one's heart's desire, become a (white) bull! A man is called a 'young calf' when on his way to his mistress's house he is unable to hide his anticipatory excitement. The phrase dhavalattanan pava contains a pun in this connection. It means 'become a bull (dhavala)' and 'become white (dhavala)', i.e. 'try not to show your agitation'. kos ava would correspond to Skt kosamra, a kind of mango-tree. For the red colour of the sprouts (navapallavam) of the mango-tree, see 586. kosumbha of Ma and Tp corresponds to Skt kausumbha, an adjective of kusumbha 'safflower'. The Chaya in Ma has in fact kusumbha! A substantive kausumbha is mentioned only in the lexicons, meaning 'wild safflower' (see PW, s.v.). Then a problem arises in connection with kisala 'sprout, leaf-bud', for it is not the safflower's sprouts or leaf-buds which are red, but its flowers. The division of the .com

Page #230
217 pound into kosumbha 'safflower-coloured' and kisalavannaa 'coloured like a kisala' does not make sense either. Probably the word kosurhbha is the result of contamination of kos anva and kusumbha, which latter plant may have been more familiar from its other instances in the text; see 145, 392 and 546. kisala 'leaf-bud' is rare. The usual form is kisalaa. Skt kisala is found only in the lexicons. The only other instance in Pkt is, according to PSM, found in the Kumar apalacaritra (no place), where it is evidently meant by its author Hemacandra as an exemplification of Sutra I 269 of his own grammar, in which kisala is explained from kisalaa. tannaa 'calf' corresponds to Skt tarnaka, which contains the word taruna (see Mayrhofer, EWA I, p. 485 and Weber, Abh., p. 79, note 2). See Paiyal. 235: tannao vaccho. hiaacchiam in Ma and Tp probably stands for hiaecchian, showing confusion of -e- and -a- before the palatal consonant -ch- (for this phenomenon, see the note to Gatha *1). Hem. IV 225 mentions vacca- as an Adesa for vraj- 'to go'. Its derivation will be discussed elsewhere. Hemacandra mentions vacca- a second time, in IV 192, among the Adesas for karks- 'to desire'. It is not unlikely that this interpretation of vacca- was based on the occurrence of the word in the Gatha under discussion: 'longing (vaccamana) for the house desired in your heart'. 201 53 Ma, 55. Tp; 18 Bh, 20 R; K, B, Y, P, 21 y, 20 G, 5; 140 T; 174 S) Missing in w. aliapasuttaa vinimi- gaind apar i urvana pula liaccha de suhaa majjha oasam iasga na uno ciraissa. Yom alie Ma -- pasutta Ma, Tp; P; T -- 'liacha R; K, olidaccha B -- desu suhaa majjha Tp; de suhaya ahma Bh, de suhaa amha R; desu haamajhu K, de suhaa majhjha B, dehi me suhaa majhjha y, desu haamajha P; suhaa maha desu T; de suhaa majja s -- uvas a Bh; osas am B, oasa P -- pariurbana Ma, pariurbana Tp; pariunvana Y; (BH) -- angari Ma -- po puno , B, na puno Y, na vuno P -- cirayisan P.

Page #231
218 Sahityadarpana, p. 90. (I see,) you pretend to be asleep, keeping your eyes closed. You, handsome man! I promise you -- look! your hair. stands on end as I kiss you on the cheek! -- at a (next) occasion I will not be late again. The woman speaking is probably an abhisarika meeting her lover. He feigns to be asleep in order to make clear to her that she has kept him waiting much too long. The commentaries, and following them Weber, take de as the imperative of the verb de- 'to give' (Skt da-) and link it to oasa, viz. de... :: oas an. de, however, cannot possibly be the imperative of de-, which is dehi or desu (actually found in Tp, K, P and T and in y respectively). Most likely we have to do here with the particle de (de 'suhaa) discussed in *16. The problem then is how to fit in oasa (Gangadhara suggests to supply dehi: he subhaga mamavakasam dehiti sesah; see Ed., p. 10). I should like to suggest to take it, despite the distance (see below), as an accusative of time to na uno ciraissan: 'during (or better: at) a (next) occasion I will not be late again'. For the accusative, which usually denotes the time during which, denoting the time at which as well, see Speyer $ 54. Admittedly a locative would be more regular. It is in fact tempting to suggest that, oasa is an old 'innovation', going back to the MS of the archetype, for oase, for which, among others, the superficial resemblance of de...oas ar to the idiom avakas am da- may be held responsible. Behind all this is the failure to recognize the intentionally interrupted sentences in the Gatha and to appreciate its implications. The sentences are interrupted because the woman is speaking and kissing at the same time. Another clear example where the particular circumstances under which the words are spoken have left their traces on the text itself is found in Gatha 160, which should read: acchodiavatthaddhan- cintesi thaNGabharaa- tapatthie martharam pie vacca siassa majjhassa vi na bhagan.

Page #232
219 The woman addressed has in vain tried to prevent her lover from leaving her. She is now running after him, being in turn followed by a friend who tries to stop her. This friend addresses her, apparently panting and out of breath. This seems to be the most plausible explanation for the chaotic word-order: acchodiavatthaddhant apatthie...pie is a locative absolute construction, interrupted by mamthar an, which is to be taken with vacca. In the second line the finite verb is standing at the beginning of the sentence, while the interrogative particle vi (Skt api) occurs practically at the end! The Gatha should be translated as follows: Now that your lover (pie), held back by you (acchodia), has been able to draw himself free (vyasta) and finally (adhvante) has left (prasthita), go slowly! Are you then not afraid (cintesi...vi na) that your waist, from the burden of your heavy breasts, might break? As in the case of the Gathu under discussion the state of the sentence was not recognized. The text was interpreted quite differently and in the course of the transmission of the text some changes were brought about in it to suit this interpretation. The result of these changes can be seen in Weber's Ed. The commentaries link the word majjha with immediately following basam. It is doubtful if this still makes sense. I have, alternatively, taken it with. gad apar iurvana in the second line. Skt punar, meaning 'on the other hand', becomes una (occasionally uno) losing its initial consonant (for an explanation of this loss, see *15). When meaning 'again' the initial p- is usually retained (see puno in 375), except in the phrase na uro (here and in 941; na una in 818), which apparently was looked upon as one entity, in which case the rules of internal sandhi came to be applied (compare na ana-, for na jana-, discussed in *3). The Gatha is found inscribed on the pedestal of a basrelief, probably from the eleventh century, found near Jabalpur, Madhya Pradesh (see Katare, 1952). The relief itself depicts a man lying on a couch in a kind of bower with a woman kissing his cheek. The inscription has been deciphered -- rather imperfectly as the accompanying ink-impression

Page #233
220 shows -- by Katare as follows: aliapasuttaa vinimtlimai de muha(a)mha uvas an gancapariumba(na) pulai anga na une ciraismam. , 21 G, 5, T; 488 T; . 21( 56 Ma, 57 Tp; 19 Bh, 21 R; K, B, Y, P, 22 175 S; 23 W) # begins here. asamattamano ana ccia volaiahalahalaas vacca gharam se sakouhallassa sa putti citte na laggihisi. ciya Bh; via y; via S -- vaccha T; vauva S; gacha W -- harah S; cari W -- (lacuna se sa)kauhallassa Ma -- bolia Tp; volavity)a Bh, R; K, M, P, bolavia Y, B; T; volavia S; W -- halah aliassa T -- laggihist Ma; ggilarihisi Bha, ligjihisi Bho, laggihasi R; W. Who cares if you haven't finished applying your make-up. Go to his house while he is still eager! Daughter, after his passion is spent you won't stay in his mind anyhow! For vacca-, 'to go', see *19. For kouhalla, 'eagerness', see Paiyal. 156 and Hem. I 117. Hem. IV 162 mentions vola- among the Adesas for gam- 'to go'. The instances in the SattasaT show that it means in particular 'to walk past, to disappear'. Turner, CDIAL 12167, suggests that vola- goes back to vyapacal-. Earlier Weber had suggested a derivation from vyapali- (Abh., p. 32, and, again, in Ed., p. 10). Indeed vola- has much in common with the verb 11-. Beside alliai (898, from a-17-) and samalliai (532) is found ahilenti (266, from abhi-11-), with which may be compared volei (711, 854). Beside volia (232; for vilia, from vi-11-, see 53) is found volina (57, etc.), which corresponds to Skt 1Tna. For the prefix vo-, from Skt vyapa- or vyava-, see Schwarzschild (1965: 350-4). For the variant bola- found in the South Indian MSS for vola-, see above, p. 170 (volaia in Ma is an exception in this MS). For the interpretation of vola(v)ia there seem to be two possibili

Page #234
221 ties. volavia may be the past participle of the causative *vyapalapayati (cf. vilapayati mentioned in CDIAL 11906). This explanation of volavia as a causative must have been current as is shown by the variant bolavia. It is not clear, though, why the -v-, which is otherwise always maintained in such causatives, was deleted in volaia. For that reason I have adopted, volaia of Ma and Tp, which would represent * (vo)layita (compare *apalayayati beside apaltyate, quoted by Turner, CDIAL 447). For halahalaa, 'agitation' or 'passion', see Paiyal. 246 and Desin. VIII 74; see also Gatha. 780. For the possible derivation of Pkt lagga- 'to cling to', see *10. , 22 G, E, T; 649 T; 22( 61 Ma, 62 Tp; 20 Bh, 22 R; K, B, Y, P, 23 619 S) Missing in W. Tarapanaviotthan vannaggh aauppamuhi- aghadianas am as agaanidalam ette pariumvi am bharimo. panamiottham Bh (utthan), R; K (ooththan), , B, Y, P; T; S -- as amhaa K, M, B, Y, as aghaqida P; as andata S -- vadalam Bho, naladam Bha, niladan R; nidaham W; laladam S -- vannaghaa Ma, Tp; R; K (ghia), y (ghia), vannaraghaa , vannacchia P, vannagdhaa B; T; S -- tuppa Ma, litta (sec. m. tuppa) Tp; tuppa R; litta K (lipta), Y, P, luppa V, B; tuppa T; S -- muhti Bh; munie K, B, Y, P; T; S -- tii Bh; tta S. -- pariumban a Ma, Tp; pariumvium Bh, pariumbiya R; pariumv anam , ombana B, ombh anam Y; ombanan T. I still think of how I kissed her while her face was covered with coloured tuppa or ghee: our lips carefully pursed to keep our noses from touching and our foreheads from meeting. During their monthly period women smeared their faces with a kind of coloured ghee (Gangadhara: haridravarmapradhan an ghrtan varnaghstan; & and T: varnapradhanam ghrtam kumkumaharidrasadhitam; Ed. p. 10). probably to mark themselves as 'untouchables'. Contact with them during this time was, as it still is today, prohibited, a taboo which merely

Page #235
222 worked for men as an excitant. When nevertheless engaging in sexual activity, the men had to be careful not to get stained by the coloured ghee, which would betray their transgression (see also 529). This explains the complicated way of kissing described here. For other references to women in their period, see 191, 289, 480, 520, 529 and 530. For the change of -m- into -v- as seen in panavia (Skt pranamita), see Brough (1962: 88-9), who has clearly shown that it is conditioned. by the presence of another nasal in the word. Other instances in the Sattas af are *29 nisavio Ma (nisamio), 83 vane Ma (mane) (vase Tp), 111 vane Ma, Bh (id.) (vase Tp), 154 vane Bh (id.), 332 panavesi Ma, Ti, Tp, T, W (panamesi). See also vammaha (mammaha) in 327 (R) etc. Through its occurrence in Ma and Tp, on the one hand, and in Bh, on the other, the variant vannaghaa(t)uppamuh Te can be traced back to the MS of the archetype. It is, however, to be abandoned in favour of vannagghaa(t)uppamuh Te found in R, V and P, as it gives an unmetrical text with u-u in the third Gana. For the reasons to adopt vanna(9) ghaauppa found in Bh, rather than tuppa found in the other MSS, see above, p. 164. The compound, which is also found in 520, contains two synonyms for 'ghee'. The first is a commonly known Skt word, ghrta, which is also widely distributed in NIA (CDIAL 4501). This is followed by a regional word for 'ghee', tuppa (Pkt, Mar. and Guj.; see CDIAL 5864), which is not known from Skt. It is probably a loanword from Dravidian (see DED 2685; but Turner, loc. cit. suggests, instead, that Kannada tuppa is a loanword from Mar.). The compound occurs itself again in a compound (see also the instance in 520). Theoretically it is therefore possible to translate it as a Dvandva, 'coloured ghee and tuppa(-ghee)'. However, the resemblance of the part ghaa(t)uppa to sakhasippt (discussed in* 4) is too obvious to be ignored. This suggests for vannagghaa(t)uppa the translation 'coloured tuppa, i.e. ghee'. tuppa as occurring in vann agghaa(t)uppamuh Te was traditionally misunderstood as meaning 'smeared with' (Bhuvanapala: upp anh snigdha (ISt., p. 36), Paiyal. 233: makkhiyam tuppan, and Desfn. V 22: siniddha... tuppo); from there litta (Skt lipta) in K, Y and P and luppa (+ tuppa x litta) in v and B. The meaning 'smeared with', however, is, among others, conveyed by tuppalia, tuppavia (529) or tuppia (Vivagasuyam

Page #236
1,2,14; 1,3,13; 1,9,6: nehatuppiyagatta). For tuppalaa (tuppa plus laa; for the suffix, see Pischel $ 595), see 289. For pariumviah Ma, Tp. v. B, y and T have pariunv(/b) anamh. The latter would seem to be the innovation. For the use of the past participle as an action noun, see Renou SS 153 c. Hem. IV 74 includes bhara- among the Adesas for smr- 'to remember'. Pischel 267 suggests that bhara- goes back to this smr-, in the following way: smarati mharaimbharai bharai. See also Mayrhofer, EWA II, p. 482. 23 655 T1, Tp; 21 Bh, 23 R; K, B, Y, P, 24, 23 G, E, p; 163 T; 200 S; 10 V) anasali deftT gose vi on aamuhT 223 taha surae haris aviasiakavola aha se tti piarh saddahimo. F ana Ti; pasaydim R; asasaahim K daft Ti, Tp; dett K, detti P; danti T; S ha (taha) Ti hasiaviasia Y gose aya Bha b gose ya Bh; gos ammi P unaya Bh; onaha Y, ona P -- muhim Ti aha sa cci (breaks off) Ti; aha sa tti Bh, aha si tti R; aha so tti K, aha sa tti, B; sahasatti S pi(y) 8h, R; Y na visasimo Tp, (missing in Ti); saihihmo Y, sahippo P. -- -- -- -- -- When that woman even in the morning still hangs over him (on aamuhT) giving hundreds of instructions and (taha) having cheeks burning with excitement during the love-play, I do not think she is dear (to him). -- Weber seems to have been quite at a loss what to do with the compound on aamuhT. In Abh., p. 81, he changed it, against the metre, into anonaamuhT, 'die...nicht das Antlitz neigt' (from shame). See also Retr., p. 355 and Ed., p. 11. I think, though, that on aamuhT should not be taken as a sign of shame here. It seems, instead, to indicate the position of the woman relative to the man: she is hanging over him or even sits on him (viparttasurata) all the while giving instructions how she wants. to have it. The same situation is understood in Gatha 299.

Page #237
224 aha se tti should be divided into aha sa and tti. Var. VI 24 and Hem. III 87 take aha as a substitute for asau. Like the latter it is used with reference to the masculine: *32 aha sambhaviamaggo, 300 aha...cando, and 908. aha icchai, as well as to the feminine: 613 aha...paai, and 717 aha suai. For the phrase aha sa, see also 57 and 318. In 418 aha so occurs. aha is Apa. in appearance; compare Apa. ehu for esah and eha for esa. For the occurrence of setti (TP, R, Y, P, T) beside sa tti, see also 123 mame tti (Ma, TP, R, T), 360 mae tti (T), 498 mahile tti (Ma, Tp) and 855 na ve tti (Ti, Tp). These instances may be explained by assuming the interference of the corresponding Skt phrases; e.g. seti, mameti, etc. More likely, however, the e for a is due to the palatalizing effect of tt on a preceding a; for another instance of this phenomenon, see Pkt metta beside Skt matra. On Pkt metta, see Mayrhofer (1955: 224) who refers to an article by Schneider (1954: 576-583) in which the latter shows the existence in the Pali canon of a form khettiya beside Skt ksatriya. As such these instances may be compared to 88 se ccia (y) beside sa ccia or to 69 paniggahana ccia (K, S; for gahane ccia), 129 ekkaggama ccia (Ma, Tp, K, M, Y, P, T), 130 saloa ccia (Ti, Tp (pr. m.), W, Y, P, T, S), 290 hiaa ccea (Ma, Ti, Tp, K, , Y, S), 383 hattha ccia . (Ma, Ti, TP, T, S), 612 hiaa ccia (Ma,Ti,Tp,Bh,x,y) and 654 uara ccia (Ma, Ti, Tp). They may also be compared to those instances in which o before the enclitic va is replaced by a; e.g. *29 padaha va (Ma, Tp), 74 mutthia va (y, P, T) and 253 murava vva (Tp). In all these cases it concerns a short vowel of which the particular colouring may already have been less distinct than that of the corresponding long one. This may also explain the replacement of o by the more neutral vowel a in 489 ettia ccea (Ma, Ti, Tp, v), 619 magga ccia (Ma, Ti, Tp) and 767 sea ccia (Ma, TP, R). A problem arises in connection with those instances in which a before vva is replaced by e; e.g.* 32 devae vva (Tp, v), 75 bhatthe wa (Ma), 77 sanusaene va (v), 91 dase vva (Ti, Tp), 120 pade vva (Ti, Tp, T), 183 sake vva (Ma, Tp, T), 217 pancalie vva (T), 417 jae vva (Ti, Tp), 485 cittalihie wa (Ma, Ti, TP, Y, T), 496 male wa (Ma, Ti Tp), 575 mehale wa (Ti) and 654 paohare wa (Ma, Ti, Tp). In the

Page #238
225 following cases the e replaces an o: 253 murave va (Ma) and 714 asarane vva (Ma; asarone vva Ti). Evidently we have to assume intermediate* murava va and *asar ana wa here, of which the former is actually found in Tp (see above). The status of these instances in the MSS is uncertain. The occurrence in 217 of pascalie vva in T (beside vaullaam va in the other MSS; the compound in which the word occurs is a neutre Dvandva, for which see *37) should make us aware of the possibility that the instances of -e vva for -a va, which are mainly restricted to the South-Indian MSS, are the result of an independent development in these MSS. This would explain murave vva in 253 in Ma beside murava va in Tp; the latter can hardly be considered 'an emendation of the former, the most likely emendation being muravo va. On the other hand, instances are also found in y, which is not known to have any South-Indian source. In fact, the occurrence of sanusaene vva in u in 77, a Gatha which was most probably added later to the text in the North-Indian branch, suggests that the development could also have taken place in the North-Indian MSS independently. A difficulty arises when one tries to identify the cause or causes underlying the replacement of a before va by e. This replacement cannot be explained with reference to the effect caused by vva, which would have a labializing effect. It is possible, though, that the e reflects a kind of hypercorrect or exaggeratedly fronted pronunciation of the a in order to prevent any labializing effect caused by following wa. In any case it cannot be explained with reference to the pronunciation found in South India of a after certain consonants as e (see *1). This would not account for the instances in v or for e in, for instance, jae va, in which it occurs after a vowel. A second possibility is, as in the case of setti discussed above, to assume the interference of the corresponding Skt phrases; e.g. maleva, sanus ayeneva, etc. (compare s odaradt via (avatarattva) and rodidi via (roditfva) occurring in Bhasa's Carudatta in 51, 2 and 66, 6 respectively; quoted by Printz, 1921: 9). This, however, does not explain dase va and asar ane wa which correspond to Skt dasa iva and asarana iva respectively. It probably goes too far to assume Skt daseva and asaraneva, representing cases of 'double' sandhi (for this

Page #239
226 phenomenon, see Renou $ 41 b). If indeed in the origin of male wa Skt is involved it would not do either to explain dase va (beside dasa vya) as formed analogically after male vva (beside mala vva) since this analogy does not exist in Skt: dasa iva and maleva. In the edition such instances as se ccia, paniggahana ccia, padaha vva and ettia ccea are considered as retentions. The reason for doing so is that, whether one starts from Pkt or Skt, the cause for emending them into sa ccia, paniggahane ccia, padaho va and ettio ccea respectively presents itself easily. The instance uara ccia in 654 is uncertain as it is restricted to the MSS Ma, Ti and Tp which also have pacchai (for pecchai; see *1) and hiaacchian (for hiaecchian; see *19). The case of such instances as se tti is slightly different, because of Skt seti. In this connection it is to be noted that most of the instances concerned are restricted to the MSS of the South-Indian recensions. An exception is setti which is found in the MSS of the NorthIndian branch as well. At the same time this case differs from the others in that se could somehow be fitted in as the pronoun se 'to him'; e.g. 'we do not believe that she (aha) who... is dear to him (se)'. Admittedly, this involves ignoring tti.This is actually done in Bh, R and y (the latter two having se!) which for pian read pi(y)a. As such se tti could have escaped emendation more easily than the other instances given above. I feel therefore justified in taking these instances as retentions as well. The case of male va etc. constitutes a problem. In the first place it is almost certain that not all instances are original or go back to the MS of the archetype. In the second place the origin of the e beside a remains unclear. My decision to adopt these instances in the text is more or less arbitrary. It is done in order to emphasize the uncertainties concerning the forms, rather than because I believe they are original. An exception is made in the case of pacalie va in 217 (for obvious reasons) and for murave vva in 253. In the latter case murava vva of Tp will be adopted, thus making allowances for the possibility that murave va in Ma is the result of some independent development. For datt in Ti, Tp and T for dett, see *1. The a in dantt in s should probably be explained from the influence of Skt dadatt. Note that MS S makes a strict distinction between a and e-stems (see*5).

Page #240
227 24 58 Ma, 59 Tp; 22 Bh, 24 R; K, B, Y, P, 25, 24 G, E, p; 632 T; 530 S) Missing in W. piaviraho appiadamJTe tuman karijjasi tie namo ahiate. sapa ca garusi do i dukkhai a K, B, Y, P; T; S piya dannah (dasanah) Ma gurusi Tp; Bh; K (im), B, P do vi Ma, Tp; Bh, R; K, W, B, Y, P; S -- duhkhai Tp; dukhkhaim K, dukavaim B, duhkhaim P; dukkhaim T; dukhkhaim S; (Bh, R, V, Y) jTa Ma; jTi Bh; ja w, jTa B; S karijjast Ma, kariissihi Tp; karijjai y; karijacasi S tia Ma; V, Y, tTa B; S (ahi missing) aTe Ma, ahijate Tp; K, B, Y, P, ahija e ; ahijate T; S. -- Being separated from what is dear and seeing what isn't are two causes of great grief. Honour to that noble birth which makes you bring about both! -- -- Weber (Retr., p. 355) suggests that the man much against his liking. left his dear mistress (piaviraho) to show himself in front of his wife whom he does not love (appiadams an am); see also Bhuvanapala as quoted in ISt., p. 36. However, the causative karijjasi shows that the man brings grief not on himself but on someone else, in case his wife. She scolds him for his good upbringing which brought him to her, his heart, though, still being with his mistress. Thus her separation continues and above that he offers to her an unpleasant sight. For the reasons to adopt do i found in T instead of do vi found in the other MSS, see above, p. 163. For other instances of i for vi, see *17. For do (Skt dvau) used with reference to the neutre, see Pischel SS 436; see also Gatha *27. For Pkt Shit (see also Pali abhijati in chalabhijati), beside Skt abhijati, see Var. I 2 and Hem. I 44. Pischel SS 77 and Geiger SS 24 wish to derive ThiaT and abhijati directly from Skt abhijati. Geiger explains the lengthening of the initial syllable from the position of the accent. For the same accent shortening a preceding or following syllable, see also Pischel SS 79 ff. However, for most of the other

Page #241
228 instances of either lengthening or shortening of vowels explained by Pischel and Geiger from the position of the accent, another and better explanation is available. I need not go into details here as several instances will be discussed in the course of this edition (e.g. alia in *27). The lengthening of the initial syllable in ahiat may have its origin in the use of the word in an exaggerated sense, abhijati meaning - "high birth' and ahiat 'very high birth. Compare the -i- in English big which in exaggerations becomes lengthened. A similar explanation may be considered for Pkt samiddhi beside Skt samrddhi 'wealth' and for Pali! anubhava beside Skt anubhava 'power'. For other words of which the forms seem to have been influenced by emotional factors, see diara and viana discussed in *38. In the text ahjate of Bh and R (and Ma?) is preferred above ahijate of the other MSS as it is the form deviating most from Skt. Compare 265 with ahiaim in R, K and and ahijain in Ma, Ti, Bh, Y, T and S. Gatha 905 is found only in Ti, Tp and T which all have ahijat. The case of the past participle ahiaa is strikingly different in that for this word no variant ahijaa is available (see *41 and 266) ; 205 T; 25( 59 Ma, 60 TP; 23 Bh, 25 R; K, B, Y, P, 264, 25 G, 5, 176 S) Missing in W. ekko vi kalasaro kin una vahaulian na dei gantur paahinavalato loanajualan piaamae. kannasaro ta, kahnasaro Tb -- vadamto R; calato , B; valanto P; T; S -- puna Bh -- bahaulian Tp; vahaulia K (pr. m.), oulia K (sec. m.) -- jaalan Ma. For a possible translation, see below. Weber (Abh., pp. 81-2) translates the Gatha as follows: Ein Reh schon, das rechtshin flieht, lasst (den Wanderer) nicht weiter ziehen. Wie viel mehr dies von Thranen umf lorte (Ren-Augenpaar der Liebsten!

Page #242
229 The first line would refer to a specific bad omen, which, however, does not seem to be otherwise known. Admittedly, I have not been able to check srngaratilaka 4, 5 (Gieldemeister), which according to Weber would contain 'eine gewisse Analogie zum Inhalt des Verses' (Retr., p. 356). In the second line a pun would be made on the compound kalasara, which beside 'antilope' also means 'eye' ('having a black -- kala -- pupil -- sara'); and vahaulian meaning 'troubled by a hunter' (vyadha), on the one hand, and 'troubled by tears' (baspa), on the other. For this interpretation Weber relied in the first place on Kulanatha's commentary: yatrasamaye eko'pi krsnamrgah pradaksinas calan vyadhenanudruto 'pyasak un atvena gamanavighnan karoti; vaspakulitan, pakse vyadhakulitan (Abh., p. 82). Apparently the other commentaries concur. The same interpretation is found in Madhavayajvamisra's commentary found in MS Ma. However, a closer look at the vocabulary of especially the first line might suggest quite a different idea. At least two words can be positively identified as technical terms relating to boardgames, such as Indian chess, backgammon or Caturanga. However, in the end there remains for me a problem to combine these terms meaningfully in order to make identification possible. Thus, kalasara (Skt kalas ara) is the common term for a black (kala) piece (sara) in the Indian boardgames. The other term is paahina (pradaksina) in paahinavalamto; it refers to the 'clockwise' movement of certain pieces (Thieme, 1962: 205). In this connection also gam- (in gamtun) and val- (in valaito) are interesting. It is possible that the pair refers to particular but opposite movements. Compare Thieme's (212-3) analysis of the terms gama and cara, occurring in Balabharata II 5, 14: 'moving clockwise' and 'moving counterclockwise', or vice versa. If val- (or perhaps we should read cal- as found in U and B, as paleographically the distinction is almost insignificant) is indeed a technical term for a movement in a specific direction, then the addition of paanina 'clockwise' would be redundant. Alternatively, it is possible to take gam- in the meaning 'to escape', which would in any case much better suit the content of the present Gatha: If only one black piece, (by) making a clockwise movement,

Page #243
230 can prevent (a piece of the other colour) to escape (from the place it occupies), how much more (can prevent me to leave home) the two tear-filled eyes of my beloved. The second line is a pun on kalasara, which stands for Skt kalasara 'black piece in a boardgame' as well as for Skt kalas ara 'eye'. The identification of the game remains uncertain. Backgammon can be ruled out positively. In this game the clockwise movement belongs to the white pieces. Furthermore it needs two pieces to obstruct the movement of another piece. This seems to limit the possibilities to either Indian chess or Catur anga. I would like to leave it to a specialist to decide which of the two may have been intended. For the occurrence of vaha in Ma, which for Skt baspa usually has bana, see above, p. 170. 26( 62 Ma, 63 Tp; 24 Bh, 26 R; K, B, Y, P, 27 W, 26 G, 5, 1; 324 s) Missing in T and W. na kunata ccia maNGNGun sungaiavas apar imusa- nisasu sahasuttadaraviuddhanan naveana jai si jananti. kunanti Ma, Tp; R; Y, kunamto K, U, B, P; S -- a Ma; via ; cvia S maitu Ma, Tp; manah Bh, R; K, W, B, Y, P; S -- pasutta Bh, suhasutta R; K (suhasuh asutta), B, Y, P, sahasultati suhasutta si- ora (dara) Ma -- viddhana Ma, vibuddhanan Tp; viuddhana Bha, vibuddhana Bho, vivuddhana R; K, , B, Y, P; S; (R, W, Y) -- sunnaia R; sunnayia P -- pasa Ma, Tp; Bh, R; K, M, Y, P; S -- parimas ana Bha (musana Bh"), par imhasana R; par imus ana K, o musa B, mans ana Y, mus ana p -- jai vi Ma, Tp -- janamtt Ma, jananti Tp; janantu ccia Bho janamtt R; Y. I would not be so furious if you only realized how much pain is suffered by those who after having fallen asleep together (with their husbands ) wake up a little in the middle of the night to discover that the bed beside them has been vacated.

Page #244
231 As far as the participles kunamta (ccia) and janaoto are concerned the text is based on MS Bh. For my interpretation of the text, see below. For these two participles two sets of variants are available, namely kun anti (ccia) and janamtt (Ma, Tp, R and y), on the one hand, and kunaito (ccia) and janamto, on the other. On the basis of the results of the stemma the former set would have to be adopted. Then the situation is to be understood as follows. For some unspecified reason the woman is angry at her husband. She is warned not to behave in this way or otherwise one night she will find him gone. The speaker is most probably a sympathetic friend. It is not likely that the words are spoken by the husband bimself as he will hardly be aware of, or sympathetic with, the hurt feelings of deserted women as referred to here. Weber in his editions reads kunamto (ccia) and janamto. He analyzes the situation as follows (Abh., p. 82). This time it is woman who is speaking, to her husband. The latter is apparently in the habit of leaving the bed in order to meet his mistress. During this particular night his wife woke up a little and as it were instinctively felt if he was still lying beside her. The husband took this as a sign of mistrust and reacted angrily. To this his wife replied that he would not be angry if he knew how often she had come to grief in this way. In my opinion each set of variants represents a different answer to the problem raised by the reading kunamta ccia...janamto found in Bh, which consequently has been adopted. kunamta in kunamta ccia stands for kunasta, the nominative of the feminine. For another instance of a present participle of the feminine ending in -a, see 136, in which the MS of the archetype read homta (Ti, Tp and Bh; later changed into homtT). The situation is to be understood as follows. A woman is speaking to her husband, who the night before sneaked out of the bed in order to meet his mistress. In the morning he returns home as if nothing has happened, showing no regret or shame, which infuriates his wife. With feigned ignorance the husband asks her the reason for her anger. The text in Bh is unclear in two respects. In the first place kunata in kunamta ccia is ambiguous as it can also be taken as a masculine, e.g. kunamto (for the type of 'sandhi' assumed here, see* 23) This interpretation of kunamta may have been a real alternative, the more so because instances of the participle of the feminine ending in

Page #245
232 - are extremely rare. In the second place the subject of the main clause ('I'), which is different from that of the conditional clause ('you'), is not specifically indicated by, for instance, a personal pronoun to distinguish it from that of the conditional clause. This does not mean that the presence of a personal pronoun in the main clause is obligatory (compare Gatha 216: ga pattiat jai...vahavind bhijjamta), but in this particular case its absence combined with the ambiguity of kunafta is awkward. The two sets of variants would seem to have resulted from attempts to make sense again of the unclear text of Bh. The two sets have in common that the subject of the main clause is the same as that of the conditional clause. In the one case the ambiguity of kunamta (ccia) was resolved in favour of kunato, thus bringing it in agreement with janamto. In the other case the rare feminine gender marker - was replaced by -1, the more common one. Compare the fate of homta, cited above, which was changed into homtt. In order to suit the new context. janahto was changed into janantT. If my interpretation of the text is correct we should prefer mannu 'anger, fury' (Ma, Tp) to the variant manah 'sulking' (all the other MSS), because a sulking woman will never give reasons for her behaviour. Part of mana is precisely the refusal to talk. The spelling mahtu found in Ma and Tp, with -mt- for -nn-, a confusion which is typical of Jaina Nagart, is explained by the fact that the word goes back to the common North-Indian source of Ma and Tp, which, though not necessarily identical to the MS of the archetype, was at least very close to it (see above, pp. 40-1). For the peculiar use of the present participles kupata and janahto found in this Gatha, see Hem. III 180. jai vi janamt 'even if you would know' found in Ma and Tp does not make sense. The position of si, the second person singular of as- 'to be', found in Bh, R, etc. before the word to which it belongs, viz. janamto, is peculiar. It is likely that si functions as a personal pronoun here (see Alsdorf, 1935-37: 326-7). In any case in the conditional clause no finite form of the verb 'to be' seems required. Compare the following Apa. verse quoted in Hem. IV 351:

Page #246
bhalla hua ju maria bahini mahara kamtu lajji(/e)jjamtu vayamsiahu jai bhagga gharu etu. 233 It is good, sister, that my lover was killed. For he would embarrass me in front of my friends if he came home beaten. For the reasons to adopt viuddhana (m) found in Bh rather than vivuddhana(m) found in the other MSS, and vasa found in B rather than pasa in the other MSS, see above, p. 164. sahasutta is found in Ma, Tp and . The remaining MSS have suhasutta, which probably is a mere corruption: gisasu sahasutta. For viuddhanam, with -anamh as found in Tp, see p. 169. sungaia 'vacated' is from sunga 'empty' (Sunya). For the type of formation, compare uttigaia (from uttrna) found in 170 (read: jhamjhavauttinaiagharavivarapalottasaliladharahim as in T). parimusana (masana) is an action noun based on parimu (/a)sa-, from Skt parimrs- 'to touch, to discover'. 27 63 Ma, 64 Tp; 25 Bh, 27 R; K, B, Y, P, 28 , 27 G, E, p; 529 T; 325 S) Missing in W. panaakuviana doNGha vi aliapasuttana maNGaillana giccalaniruddhaNGTs&- sadingaannana ko mallo. -- pana Y ti Bha (vi Bhb) kuvidana Tp; kuviagam K, kuviona y dohna Bhb (dunha Bha); dorahah K, doraha B; doNGa T pasuttana K magaillana Ma; maNGaillagarh K, B, maNGaijjana P; manaillanam T; S; (Bh, R, V, Y) -- bhinnaannana Ma; dinnakannana K (nam), Y. P, disaanana B; dingakannana S. -- -- -- donna Ma, Tp; Dasarupa, p. 184, Sahityadarpana, p. 81. (I wonder) who will be the strongest of these two who, angry though loving, pretend to be asleep, are full of pride and lie still, listening carefully to their

Page #247
234 suppressed breathing? For donha in the MSS of the North-Indian recensions those of the SouthIndian recensions have, here as well as in 162 and 627, donna; see also 878, a Gatha found only in T. According to Pischel SS 436 the genitive ending -nha() is typical of M. (cf. Var. VI 59 and Hem. III 123) as well as of Amg., while - nain) would be typical of s. and Mg. Here they are MSS variants and the question is which was actually found in the MS of the archetype. donna in the South-Indian MSS may be explained as a clerical error for donha. Compare kanna in *25 in To beside kahna in to, or saanna in 624 in Ma. Note in this connection that in Tp hn is invariably preceded by the circle which indicates the doubling of the following consonant. Alternatively, it could be a saurasenism. Note in this connection the occasional retention of the intervocalic -t- as -d- typical of that dialect; e.g. bhudan in *8, kuvidana here and padavie in 107 (but restricted to Tp and T). Another possible Saurasenism in the South-Indian MSS is khu (see *39). On the other hand, it is equally possible that donha in the MSS of the North-Indian recensions represents an instance of the influence of AMG., of which already other possible instances have been discerned (see *1 and *2). It is obvious that these arguments do not lead to any decision as to which form should be adopted in the text. In this connection the derivations of the endings may be discussed. Irrespective of the questior whether the one is a saurasenism and the other an Ardha-Magadhism, it would appear that the ending -nha(n) is the more difficult one and should therefore be adopted. The starting-point of the ending -nnath) may be found in, for instance, caturnam, sannam and possibly also trinam. The derivation of -nha() is much less obvious. Pischel (loc. cit.) suggests that -nha(m) in donha (n) goes back to*dvausnam, itself a contaminated form of *donam and *dosam. As an alternative I should like to suggest to derive -nha() from the feminine, as in catasrnam or tisrnam. It has to be assumed then that these numerals, as a special class of words, have lost their penultimate syllable. *catasnam and "tisnam would have developed into caunha() (the -u- from cau-) and tinha(in) respectively,

Page #248
235 and from there the ending -nha(m) would have been transferred to the other numerals. Regarding the use of the feminine of the number 'two' with reference to the masculine and the neutre as well it should be noted that in Pkt with the numerals above 'one' the gender-distinction appears to have been lost; see Pischel, loc. cit. Note do i dukkhai, with do from Skt dvau, in *24. alia 'false' (cf. Pali alika) corresponds to Skt alska (see Paiyal. 53 and Hem. I 101). Pischel $ 80 and Geiger SS 23 derive ali(k)a directly from Skt alTka. They explain the shortening of the -T- in alia from the position of the accent on the first syllable. Evidence for the workings of such an accent is slender (see *24). Obviously we have here to do with two different suffixes, -ika in ali(k)a and -Tka in altka. manailla 'sulking' consists of manaka and -illa. -ka in manaka may be svarthe. It is also possible that manaka means 'sulking' by itself (AIG II 2, SS 362 d), in which case -illa has merely an emphatic function (compare -ullaa in padhamullaa, discussed in *15). For the meaning of malla, 'the strongest, the winner', compare H. an. 2, 505 (balin) and Trik. 3,3,405 and Med. 1,44 (baltyams) (quoted in PW, s.v.). , 28 G, 5, T; 173 T; 28( 64 Ma, 65 Tp; 26 Bh, 28 R; K, B, Y, P, 29 490 S) Missing in W. . navalaapaharan age romancadarndarat jahin jahin mahai dearo daun tahin tahi dtsai vahue. kuvalaya Bh; navalai K, 4, B, Y, P; T; S -- haran (paharan) B jaha jahim Ma -- devaro Ma; deuro Bh, devaro R; K, Y, P; T; S dandarat K, dastur aimya Y -- vaehue Ma. -- -- Wherever on her body her brother-in-law takes pleasure in beating her with a young creeper, there appears (her defence in the form of) a line of sticks formed by the thrill of her hairs.

Page #249
236 The relation between a woman and her husband's younger brother (devara) is characterized by flirtatious behaviour, in which often the woman herself takes the initiative (see Gatha 456 quoted below). Equally frequently the devara goes too far and the woman has great trouble in keeping him in check (see *38 and 59). In 456 the woman provokes the devara into beating her with a young creeper. Gatha 456 reads as follows (text according to R): dhavai pur ao pase- navalayajuttaa ean su bhamai ditthipahammi samehai. de devara paharasu varain. She runs in front of you, wanders on both sides and stands in your view as you hold a young creeper in your hands. Go ahead, devara, use it on the poor thing! From Gatha 862 it would appear that such beatings were part of the love-play between husband and wife. According to Ti and Tp this Gatha should read: navalaapaharottattha- jam ajja vi juvaijano e tam ki pi haliaputt Te ghare ghare sikkhju mahai. What even now every young woman in every house takes pleasure in learning, that thing the farmer's daughter does only when she is frightened of being beaten with a young creeper. By provoking the devara into beating her with a young creeper she gives him all kinds of ideas. Furthermore, the way she defends herself can only encourage him to go on and take even greater liberties with her. The risen hairs on her body, which look like even so many sticks, are signs of her enjoyment of the situation. For hairs raised on a woman's body from pleasure functioning as a kind of protection, see also Gatha 96. Compare also the Skt expression romancakancuka in Spruche 3274.

Page #250
237 pavalaa is from Skt navalata. The variant navalai is from navalaya, with -i- for -ya-. Pkt pahara 'beating' corresponds to Skt prahara; see Var. I 10 and Hem. I 68. Most likely they represent different formations from the verb prahr-, Skt prahara being formed through Vrddhi and Pkt pahara through Guna. Pkt also has pahara, for which, see *8. maha-, mentioned in Hem. IV 192 among the Adesas for kanks- 'to wish', is from Skt mah- 'to rejoice, to like to'. It is constructed with an infinitive (139 and 862) as well as with an accusative (510). For deara each time a variant devara is available. The latter seems to have been introduced secondarily under the influence of Skt devara (see above, p. 163). Here dearo in Tp, and B, devaro in Ma, R, K, Y, P, T and S, and deuro in Bh, 456 devara in R, deura in Bh, 690 deara in U, Y and S, devara in R, 771 deara in S, devara in R, 872 dearo in Tp (Ma has only'de), devaro in T, and 963 deara in W. Beside deara a form diara occurs, for which, see *38. 29( 65 Ma, 66 Tp; 27 Bh, 29 R; K, B, Y, P, 30 , 29 G, 5, ; 272 T; 368 .S) Missing in W. ajja mae dena vina ahinavamehana rao anuhuasuhai sambhar atte pisavio vajjhapadaha va. tena Ma, Tp; Bh, R; K, 4, B, Y, P; S -- anumhua k; anubhuda T -- sambhavamtie Ma; sahbharanfte K, satmaramtte , sammmar at Te B, sarambhat Te P -- mehana k, meghana B -- ravo Ma, Tp; Bh, R, K, U, B, Y, P; S -- nisamio Tp; Bh, R; K, B, Y, P, nisamio y ; qisamio T; S -- vajja Tp; R; vajhjha K, vajja , Y, P;S -- padahu Bh, padaho R; W, B, Y, P, paduho K; padaho T; S. Today thinking over the pleasures I had without him the thunder of the first rainclouds sounds like the execution-drum. Usually in the absence of the husband thunder produces anxiety about his timely return home. In this Gatha the thunder produces fear of

Page #251
238 punishment, as the comparison of it to the sound of the execution-ut um suggests. Therefore dena vina is to be linked to following anuht&suhai. However, the position of dena vina after ajja mae and before anuhuasuhai is ambiguous, which seems intentional. For the reasons to adopt dena and rao, both found in T, rather than tena and ravo found in the other MSS, see above, p. 164. For other instances of dena, see *15. For sabhara-'to remember and for nisavio beside nisamio, see *22. vajjhapadaha 'execution-drum'; compare Skt vadhyapataha and Pali vajjhapatahabheri. The variant vajja may be due to a misreading of the Jaina Nagart ligature for -jjh- ( 5 ) as -jj- ( 5 ). As Tp has vajja, vajjha in Ma is most likely an emendation. 30( 66 Ma, 67 Tp; 28 Bh, 30 R; K, B, Y, P, 31 W, 30 G, 5, ; 543 s) Missing in T and W. nikkia jaabh Trua gamo gamaninandana duddarhs ana nimyakidaasariccha tujjha kae taha vi tanuai. nikkiva Ma, Tp; Bh, R; U, B, P, nikkava Y; nikkiva S -- jahabh Trua P -- niddass ana Ma -- nibva Ma, nimba Tp; gimva K; (Bh, R, V, Y) -- kidayas ar icha Bh; kidas ariccha K (degicha), , B, Y, P; S -- gamini Ma; Bh, R; K, ggamani v; gamini S -- kujja Ma; tujhjha K -- tave Ma. Afraid of your wife you are cruel, hiding yourself like a worm in a (bitter) neem-fruit. You are a true war-leader! Still the whole village grows thin because of you. Thieme (1937: 117) has suggested that the word gamaNGt (Skt graman occurring in the Sattas at is used in its Vedic sense of 'leader of an army (grama)'. This in a way unexpected sense would explain why the word is marked as 'Dest'; see Paiyal. 104 gamant bhojo daina vady and Desin. II 89: gamapahane gamani ganaudaggamagohagoha ya. The function of the gamani is to protect the village (gama or, as in *33 and 630, a pallt). His heroism and constant desire for battle stand in the way

Page #252
239 of being a good husband or lover (*33) and are a worry to his relatives (630). He is cruel, always from home, preferring a live of hardship above a life of leisure at home. The man addressed is afraid of his wife (jaabh Trua) which makes him cruel (nikkia), too, namely in the eyes of the women he avoids (duddams ana), as he apparently finds the pleasures they promise not worth a domestic quarrel. This has given him the reputation of being like a worm in a bitter neem-fruit (see also 878). With such a gamani-like man about the village would be expected to prosper but instead its women grow thin being unable to meet him. For the reasons to adopt nikkia found only in k instead of pikkiva found in the other MSS, see above, p. 164. The word n andana 'son' in gaman in and ana would specifically denote 'someone belonging to the caste or profession of...'. Note Alsdorf's discussion of putra in the compound mantriputra (1950: 358-360). Beside namdana,, in this text occur (p)utta (haliautta in 84 and gamaniutta in *33) and sua (gahavaisua in 106); for the feminine are found (p)utti, sua and dhua. Y, 198 P, 188 G, 31( 68 Ma, 69 Tp; 29 Bh, 31 R; 191 K, 190 W, 1.92 375, ; 389 T; 326 S) Missing in B, W and X. neurakodivilaggan hiaa manapauttham cihuran daiassa paavadiassa ummo asti ccia kahei. -- aiura Y, nevura P; nepura T -- ciuran Ma; K (ciuran), U, Y, P; S daissa Bho, paino Bh-- paapadiassa Ma, Tp; payalaggassa R; paapadiassa K, M, Y, P; T; S -- manapautthe Ma; padatthamana R; pautthamanah K, M, Y (padatthao), P; T; S -- umoyati Bha, omo asti Bh -- ccaya Bha (cciya Bh); via K, via y ; s. Dasarupa, p. 187 (man apauttham!). By freeing the hair of her husband who had fallen at her feet, from the clasp of her anklet, she indicated that * pride still kept her heart far away.

Page #253
240 For the position of this Gatha here as 31, see above, p. 161. manapauttha which is found in Ma and Tp, on the one hand, and in Bh, on the other, goes back to the MS of the archetype. Bhuvanapala (ISt., p. 40) explains it as a BahuvrThi-compound in which the order of its members is reversed, meaning '(the heart) from which the sulking mood has disappeared' (for instances in Skt of such compounds, see Gonda, 1968: 246). This interpretation seems to have been generally accepted as appears from the variant pautthamana. In my opinion, however, we should accept the compound as it stands and take it as a Tatpurusa, meaning '(the heart) (still) away or distant because of the 'sulking mood' (compare pauttha, literally '(physically) abroad', in the expression ekkaggama ccia pauttho in 129). My interpretation agrees with the fact that the woman is silent using gestures instead of words, which is an indication that her sulking mood is not over at all. Var. II 4 and Hem. I 186 derive Pkt cihura by a special rule from Skt cikura 'hair on the head'. Skt cikura would regularly have become ciura, which is actually found in this and the other instances in which the word occurs as a variant; see, for instance, 273 in which cihura in Ma, Ti, Tp, Bh, R, V, P and S and ciura in Y and T. The latter is obviously an 'emendation' for cihura which does not have a clear derivation from Skt (Skt cikhura is found only in the lexicons): For the reasons to adopt paavadiassa as found in Bh instead of opadiassa found in the other MSS, see above, p. 164. Y, 194.G, 38 5, TT; 32( 69 Ma, 70 Tp; 30 Bh, 32 R; 197 K, 1961, 198 322 T; 420 S) Missing in B, P, W and X. . sa tai sahatthadinnan uvvasi anaaraharade- ajja vi o suhaa gadhar ahi am pi vae vva nomalia vahai. tue Tp; tan R; tui K, Y; tue T -- suhattha K -- dinnaa Ma; dinha K -- ajjha Bh; aja K -- u Bh; re K, y, Y; S -- gamdharaia Bh; andharahian K -- uccasia -- ghara Ma, Tp; Bh, R; W, ghara missing in Y; ghara T; S -- devaa wa Ma; Bh, R; devau vau K, devada va Y; devaa vya T; S -- omaliyah Bh, nomani ah R; (devau va) u malian K; omalian S.

Page #254
241 Pada d quoted in Hem. I 38 (omalayan vahai). Still she wears the navamalika-flower, o handsome man, you gave her personally, though it has lost all fragrance, like a house-god (in a house) in a deserted town. For the present position of the Gatha as 32, see above, p. 161. For tai, the instrumental and locative (once, in 338) of the personal pronoun of the second person singular, practically each time a variant tui is available. Its long form tae has a variant tue. The MS of the archetype. seems in most cases to have had tai and tae, ascertained through its occurrence in Ma, Ti and Tp, on the one hand, and Bh and R, on the other. Admittedly Ti and Tp often have tui.(T always has tue, for which, see below. S and the Vulgata MSS vary.) tai is comparable to the nominative tan, tui to the other nominative tumam and the genitives tuha and tujjha. These three forms, which together are much more numerous than tan, may have suggested the idea that tui was more correct than tai. The instances of tai and tae are the following: * 34 tai ti, TP, R, S, tae , B, tuha Bh, tue K, Y, tu(yavva) P, tue T, tui W; 137 tae Ti, Tp, Bh, R, K, U, tue Y, P, T, S; 194 tai Ma, Bh, W, S, tan R, tue TP, T, tui K, Y; 212 tai Ti, Tp, . Bh, R, K, M, S, tuha Y, ita P, tue T; 215 tai Ti, Tp, Bh, R, K, V, Y, 'tuhi P, tue T, tuha S; 352 tae Bh, R, K, U, Y, tai S, tue T; 437 tai Ma, Ti, TP, U, Y, S, tue T; 478 tae Ma, Bh, R, tue Tp, (Ti lacuna), 4, Y, S; 667 tae , tue Y, T; 700 tai Ma, Bh, K, P, tuha R, tui Y, tue T, S; 811 tai S; 826 tue T; 860 tue W (tumae T); 876 tue Ti, TP, T, W; 912 tae Ma, tue Ti, Tp, T; 919 tai Tp (tui pr. m.), runai Ti, tue T. The occurrence of tai and tae is entirely regulated by the metre. It is possible to explain the final -i in tai as a graphical representation of -e. However, it is uncertain if and how far the writing of i for e was carried out in the MS of the archetype, especially for -e in an open syllable. Alternatively, tai may have been an independent word in Pkt existing side by side with tae. Its form, with -i after short

Page #255
242 a, became only common in Apa. For another Apa. word already occurring in Pkt, see aha discussed in *23. In fact, the occasional occurrence of the variant tae or tue (i.e. tae, tue) for tai would betray attempts to remove this Apa. form from the text. For an explanation of the origin of the variant ahdharahia found in K, see above, p. 165. For the particle o, see Var. IX 4 (o sucanapascattapavikalpesu) and Hem. II 203 (o sucanapascattape). of the verb udvas- Skt has only the causative, i.e. udvasay-, meaning 'to remove, to destroy' (cf. uvvasia 'removed' in Gatha 106). The participle uvvasia, from uvvasa-, seems to me an 'deserted, fallen in disuse' (from there possibly also 'dilapidated'; cf. uvvasia in 233, quoted in the note on *7); compare (late) Skt udvasa 'uninhabited, empty' (RajataranginT), 'solitude' (satruhjaya-Mahatmya; MW). For the reasons to adopt hara found in k instead of ghara found in the other MSS, see above, p. 164. For devaeva (Tp, ) beside devaa va of the other MSS, see *23. nomalia corresponds according to Hem. I 170 to Skt navamalika 'Jasminum sambac'. Var. I 7 has nomallia, i.e. Skt navamallika 'id'. The variant omalian (Bh, K, S) undoubtedly is a corruption for nomaliam, which may be due to preceding o suhaa. As becomes clear from the quotation of part of the last pada of this Gatha in I 38 Hemacandra used a MS of the Sattas at which contained this corruption. His subsequent dealings with the word omalian are illustrative of his method. The context suggests that it denotes flowers offered to a deity, which, moreover, had lost their fragrance. In II 92 Hemacandra ad hoc produced a Skt word avamalya, which would be synonymous with 'nirmalya 'the remains of an offering (of flowers) to a deity' (1 38). , P, 31 G, E, T; 653 s) 33( 70 Ma, 71 Tp; 31 Bh, 33 R; 31 K, B, Y, 32 Missing in T and W. pahar avanamaggavisame gamaniuttassa ure jaa icchena lahai se niddah pallt una se suhah suai. papahara Ma -- vina (vana) R -- maga P; magva S -- kicchena Ma;

Page #256
243 Bh, R; K, W, B, Y, P; S -- mahai Bh (lahai Bh") -- niddhah P -- gomini Ma; gamini Bh; K, Y; S -- outtasya B, P -- vallt P -- una sa K, Y, ua na sa P -- suhai (suh ari suai) P -- suvai Ma, Tp; S. The war-leader's wife sleeps with difficulty, lying against his chest which is rough with scars from wounds of (sword-) strokes. His village, though, sleeps soundly. For pahara 'stroke', see * 28, and for gamant 'war-leader' and the use of the word (p)utta, see *30. For the reasons to adopt icchena found in Tp instead of kicchena found in the other MSS, see above, p. 165. For the reasons to adopt suai instead of suvai, see above, p. 167. 34( 601 Ti, Tp; 32 Bh, 34 R; 32 K, B, Y, 334, 31 P, 32 G, 5, TT; 518 T; 533 S; 95 W) aha sambhaviamaggo enhin hiae ann am suhaa tai ccea navari nivvudho ann am vae loassa. sabbhavia Tp; savbhavi(y)a Bh, R; sambhovia K; sabbhavia T; sabhbhavia S; savbhavia W - tuha Bh; tue K, Y, tae , B, tu(ya?) P; tue T; tui W - - vyea Ma; cceva R; jecca K, jevva Y, (tuya)vva P; vvea S; ccia W -- nivara Ma, qavara Tp; Bho, R; K (Oran), , B, Y (ora), P; T; S; W -- nivvodho Ma -- ihn im S -- iae annan Ma; ann an hi(y)ae Bh, R; hiae ahnan Ta -- second annan missing in Ma -- vai R; K, W, B, Y, P; S. Handsome man, that (old) esteemed course is no longer accomplished by anyone else but you. Nowadays there is one thing in a person's heart and another in his speech. The '(old) esteemed course' refers to speaking the truth. The Gatha is addressed to a man who accidentally addressed his wife with his mistress's name (gotraskhalana). For aha, see * 23; for tai, * 32; and for pavari (Bha), * 15.

Page #257
244 35( 71 Ma, 72 Tp; 33 Bh, 35 R; 33 K, B, Y, P, 34 W, 33 G, 5, 7; 527 'T; 309 S; 36 W) unha vinTsasamto hiaan palTvian anu- alingasi kisa mam parahuttin. saena patthim palTvesi. In Ma lacuna from para (huttin) onwards. uhna vinto Ma; unhai nisas arto Bh, R; K (unhain), W, B, Y, P; T; ungai pts as asto S; unga vinisas arto W -- sa(y) an addhe kisa me par ahuttin Bh" ("hutti Bh"), R; kTsa mae par ammuh Te saanaddhe K (htee), Y (ohta), P (sahan ao ), ktsa maha par ahmuh Ta saan addhe , B; kisa maha par arhk khuhia saanaddhe S; kfsa hma par armuh Te sayanaddhe W -- palTviurh BhD (degiyan Bha); U, palTviuna K, palTviuna B, Y, paliliunam P; palTvia T; palTvium S; paliya W -- putthi R; K (oththo), W, B, Y, P; patthan T; putthin S; petthan W -- palTvesu Y; valavesi To, palavesi To; peleppayasi W. Why do you, sighing hot sighs, embrace me from behind? (First) you burnt my heart (and now) you burn my back with your repentance The word aNGusaa (Skt anusaya) 'regret, repentance' provides a pun, as etymologically it denotes 'lying close'. For hutta 'directed towards' (Hem. II 158), see also 145 and 896 (parahutta), 373 (antohuttan) and 696 (gharahuttamaNGena). The derivation suggested by Weber from bhuta (Retr., p. 410; cf. Hem. II 99: hutta from bhuta) fails to convince, from the point of the meaning as well as from the fact that hutta for bhuta does not actually occur. For pattht 'back' beside Skt prstha, see *11. unhan vinisas ato is based on Ma, Tp and W. unhai in the variant unhai nis as amto is problematic. It would through ellipsis stand for 'hot sighs'. However, such ellipsis is, as far as I know, not very common in Skt, which even tends to the opposite; note such expressions as tapas tap-. Possibly it originates from a corruption, like the one found in Ma, uhna vints as amto, which was subsequently corrected into unhai nis asamto.

Page #258
245 The available variants seem to indicate that the peculiar position in the present text of kisa after, instead of what might be expected before, alimgasi was misunderstood. kisa was subsequently linked to palivesi, as a result of which the word alimgasi lost its proper context, and the accusative mam made no longer any sense. alingas i was replaced by saan addhe, and the accusative mam by a genitive; e.g. me in Bh and R (side by side with accusative parahuttim!). In this new sentence, kisa ...paltvesi, palTvia makes no sense. The word palTvia, which beside in Ma, Tp and I (and W) is also found in Bho, is therefore most likely a retention, even though in the adopted text the absolutive palTvium (Bho, v and S). would also fit well. The second line consists of two independent sentences, namely hiaan palivian, 'my heart is already burnt (by you)', and anus aena patthi palTvesi, '(now) you burn my back with your repentance.' , 193 Y, 199 P, 189 G, 39 g, 36( 73 Ma, 74 Tp; 34 Bh, 36 R; 192 K, 191 ; 346 T; 410 S) Missing in B, W and x. : tujjhangaraasese- sa ira golatuhe na samalt taha kharena somala nhaa jau ukas.ena. tumhangaasesena K, tujhao Y, P -- raga S -- samila Ma -- teha Ma -- Sarena k -- somara K, , Y, saumara P; somara S -- suiram Bh, R; sa kira K, M, Y, P; T (kila); S -- goda K, goda S -- ruhe Ma, titthe Tp; ruhe R; K, lacuna in P; tire T -- jambu Ma, Tp; K (mbu), B, P; T; S; (Bh, R, , Y). They say that that dark and (taha) soft-skinned woman bathed herself in the Godavart at the (same) bathing-place with the biting lotion of the jambu-treebark left over from your bath. For the present position of the Gatha as 36, see above, p. 161. janvukasaa refers to the astringent juice extracted from the bark of the jambu-tree. Because of its astringent nature it must have hurt the woman who is expressly said to have a tender skin (somala). The juice

Page #259
246 is used for a number of things, e.g. as a black dye (see Watts III, pp. 284-6). It is therefore foolish of the woman to apply it to her already dark skin (samall), the more so as for women a light-coloured skin is favoured, which they try to bring about by using a lotion of turmeric (see 58, 80 and 246). The woman is apparently desparately longing to be touched by the man (cf. *16 in which the woman begs to be touched by the same moonbeams that touch her lover). The Gatha has a counterpart in 246 in which a man is drinking the water of the river mixed with bitter turmeric used by his beloved as a bathing-lotion. In this case, too, it probably concerns a bathing-lotion floating on the water after the man's bath. In any case it is unlikely that he left it behind in some jar on the bank of the river for anybody else to use it. It would seem to appear from this Gatha that men tried to enhance the blackness of their skins. Note in this connection the following Apa. verse found in Hem. IV 330, which indeed seems to comment favourably on the contrast of the lightness of the woman's skin and the blackness of that of the man: dholla samala dhan(i)a campavanpT nai suvannareha kasavatgai dingT. The dholla is dark, his wife campaka-coloured. It looks like a streak of gold on a (black) touch-stone! For tuha 'bathing-place' beside Skt tirtha (Pkt tittha), see Mayrhofer, EWA II, p. 507, and Turner, CDIAL 5903. tuha is mentioned in Hem. I 104 and II 72. The variant ruhe in Ma, R and K probably rests on a misreading of the Jaina Nagari symbol for tu as ru (note Weber's hesitation regarding the reading of Bh, which could be tuhe or ruhe; ISt., p. 41). The variant titthe in Tp possibly is an emendation for this ruhe in the source of Tp. See also 58: ruha in R and Ma and tittha in Tp. For the variant teha in Ma, for taha, see *1. Other instances of ira (Ma and Tp; cf. suiram in Bh and R) beside kira are found in 499 tassa (vi?) ira in Ma (cf. Hem. II 186: tassa ira, which probably is a quotation) against tassa kira in Tp, V, Y, .T and W

Page #260
247 and tasseva in R and S, and 773 ko ira in S against ko kira in R. The loss of the initial k-, which is restricted to its occurrence after vowels (contrast kallar kira in *50), may be explained from the less emphatic or tense pronunciation of this particle (see also the note on *15). The variant kira would seem to have been introduced secondarily in the MSS individually or in some earlier stage in the transmission of certain branches. Pkt (k)ira corresponds to Skt kila (Var. IX 5 and Hem. II 186; the latter mentions ira as well). For the meaning 'it is reported or said', see Emeneau (1969: 43-54). . For somala 'soft, tender', corresponding to Skt sukumara, see Var. II 30 and Hem. I 171, 254; see also Turner, CDIAL 13448. The variant somara, which is not taught by the Pkt grammarians, is an innovation based on Skt sukumara. The word gola is glossed with godavart 'the Godavart river' (Paiyal. 132 and Desin. II 104). The relation between this gola and Skt godavarT is as yet unclear; see Luders (1942: 34). W, 34 G, E, T; 304 T; 341 s) 37( 35 Bh, 37 R; 34 K, B, Y, P, 35 Missing in Ma, Ti, Tp and W. tuha virahe ciraaraa rairahasiharadhaena va tissa nivad aitavahamailena muhena chahi ccia na patta. tuva B -- viraye P -- vi(y)alanta Bh, R; nivadaita k -- baha B; T -- mainlena K; sailena s2 -- ravi Bh, R -- dhavena Bho; dhvaena k -- wa P; S -- chai Bh; chayi P; chaah S -- cvia S. As long as you haven't returned, slow traveller, her face spoiled by falling tears, will not regain its natural lustre like the chariot of the sun, a (mountain)peak or a banner, which never enter into the shade. Weber translates rairahas iharadhaena with 'the banner on top of the chariot of the sun'. This translation does not do full justice to the word sihara (Skt sikhara). Instead I should like to suggest to analyze the compound as a Dvandva enumerating three objects which never enter

Page #261
248 the shade, namely the chariot of the sun, a mountainpeak and a banner.' For a Dvandva in the singular, see AIG II 1, $$ 68-70, which, however mentions only instances consisting of two items; but see vayvagnivipram in Manava-Dharmasastra IV 48. Note in rairahas iharadhaa the significant order in the enumeration of the items, from something high in the sky to something comparatively low to the ground, which is a feature of Dvandva-compounds (AIG II 1, $ 71). Another instance in the Sattasaf of this type of Dvandva is probably alihiabhittivaullia in 217. This Gatha reads as follows (vaullian as found in Ti and Tp): tan mittam kaavvam alihiabhittivaul- jam kira vasanammi desaalammi lian va na par ammuham thai. That man is to be made a friend, who is known in times of misfortune, when and where he is needed, not to stand with his face averted, like something drawn, like an (immovable) wall or like a puppet '(? which cannot move away by itself). For Pkt chaht beside chaa (Skt chaya), see Var. II 18, where the intrusion of the -h- is noted, and V 24, where the commentator Bhamaha mentions chaht as showing the substitution of the feminine gender marker -1 for -a (cf. Hem. III 34). Hemacandra in I 129 makes an interesting observation regarding the occurrence of chaht; when it occurs it would always me an 'shadow' or 'shade' and never 'beauty' or 'lustre' (akantau). The few times chaht occurs in the Setubandha it indeed means 'shade' or 'shadow' (see X 37, 47, 48, 54 and 55). In the Sattasaf it occurs four times, namely here, in *41, 49 and 136. Here and in 49 and 136 (for the instance in*41, see below) it means indeed 'shade' or 'shadow'; in the Gatha under discussion, however, only secondarily, by way of a pun. This shows that the distinction made between Pkt chaa and chaht is spurious. It is possible that Hemacandra overlooked this instance or that in the MS of the Sattas at he had before him the text read chaa, as is the case in Bh (chai, i.e. chaya) and s (chaan). The origin of the form chaht beside chaa (chaya) is uncertain. Turner, CDIAL 4027, suggests that it is due to contamination of chaa with a(b)ha or p(r)a(b)ha 'light'. Ghosal (1959: 256- 259) suggests that the.-h- is

Page #262
249 a glide. A fanciful derivation is found in Pischel SS 255. Whatever its origin, the form is in any case also found in some NIA languages. 38( 74 Ma, 75 Tp; 36 Bh, 38 R; 35 K, B, Y, P, 36 V, 35 G, 5, TT; 394 T; 526 S) Missing in W. diarassa asuddhamanas - diaham kahei rama- sa kulavahu piaakuddaliniai nulaggasomitticariai. daar amsa Ma, devaraassa Tp -- addhasuddhamanas sa Ma, asuddhaamanassa Tp; asuddao k, suddao y -- kulabahu R; kulavahua k, kulabahuaY, P- nani Ma; niya Bh; nia K, W, B, Y, P -- kuddi Bha (odda Bh"); kudu K, P, kui B; kuddu S -- lihi(y)ain Bh; K (degamin), B, P; lihiaain T; lihiain S; (R, X, Y) -- somitta Bh; saumitti k -- cariyain Bh; carianis K, cariaim B, P; T; S; (R, W, Y). The good wife takes all day to tell her young brother-in-law, who has improper intentions, about the adventures of Sumitra's son, the devoted brother of Rama, though these were drawn on the walls of their very own house. The term kulavahu denotes a wife who by her behaviour avoids any harm or embarrassment to her husband's family. In this Gatha she tactfully deals with her husband's younger brother, devara, who misbehaves towards her. She does not expose his behaviour to the other members of the household which would only result in a quarrel between him and her husband and finally in a rift in the family (see Gatha 59). Instead she tells him edifying tales about the ideal devara Laksmana. Thus she makes him understand she knows what he is up to. However, these tales were depicted on the walls of their own house, so the devara will already have been quite familiar with them. This and the fact that she takes the whole day for it should indicate to him that she is not at all averse from his company. diara 'brother-in-law', husband's younger brother' goes back to Skt devara (see Var. I 34 and Hem. I 146). Otherwise Pkt has deara (see * 28). It is possible to recognize in this diara beside deara an early instance

Page #263
250 of the development found in Bengali, of -e- in the neighbourhood of an hiatus changing into -i-. As an example Bloch (1934: 44) quotes siult from Skt sephalika. Turner, on the other hand, suggests that the special development found in diara goes back to the use of the word expressing a certain degree of intimacy (CDIAL 6546). The pair diara-deara has an exact parallel in viana (64) beside veana (e.g. * 26) (both from vedana), of which the former could then be taken to denote a 'terrible pain' against the latter which would simply denote 'pain'. 39( 75 Ma, 76 Tp; 37 Bh, 39 R; 36 K, B, Y, P, 37 V, 85 PP, 36 G, 5, ; 404 T; 572 S) In P this Gatha occurs as 36 and a second time after 85, itself numbered 85. Missing in W. . caccaragharint piadan- asaT saajjhia dug sana a tarunt pautthavaia a gaa a na hu khandi an silan.. cattara Bh; K, W, B, cattaa y, po, a P -- dassana Ma; dasana P, dasana po -- e Ma; vi S -- pautthapaja Ma, Tp; pautthadaiya Bha (vaiya Bh); odaia K, Y, P, o taia PD, paia , B; T -- e Ma; a missing in P, pB -- saamdhia Tp; saajjia R; K, V, B, Y (onjia), P, sahadhya P; saaccia T; saajhjhia S -- duggaa e Ma; a duggaa , B -- na kkhu Ma, na khu Tp; na ya Bh, puna R; nu hu B. -- khandia B, kharidi am missing in P -- silai Ma. A beautiful housewife living on a busy road, a young woman whose husband is abroad, an unchaste neighbour who is poor, (who believes that) propriety is not lost? Hem. II 12 equates caccara with Skt catvara 'crossroad' (i.e. where four roads meet). The etymology of caccara is as yet unclear. Schwarzschild (1956: 112) suggests that caccara derives from cattara (+catvara) which was contaminated with the verb car-, as in sacara 'road'. I should like to suggest that Pkt (and Pali) caccara is from*carcara 'a place where there is a constant coming and going', from there 'crossroad' (as a place where traffic accumulates). Note that in the Sattasat caccara each time indeed in particular denotes a place where a lot of people

Page #264
251 pass (see *47). catvara is merely a convenient Skt gloss. Its Pkt derivative, cattara, has in some MSS been entered into the text as the more familiar formation, for original caccara. For the reasons to adopt pautthavai a (Bho, s) instead of paia, see above, p. 164. Paiyal. 236 mentions the word saajjha, of which saajjhia is an extended formation, as a synonym of samosia 'living together' (sam-a-vas-). Des in. VIII 10 translates saijjha (for sayajjha) with padivesia 'neighbour', which as a gloss is also found in most of the commentaries on the Sattasat; see, for instance, Kulanatha, quoted in Abh., p. 86, and padivesa in 335 in W in the Pkt text. In this connection it should be noted that Desfn. VIII 49 translates samosia (see above) with pacivesi. This meaning 'neighbour' returns in the derivatives of the word in Mar.; e.g. sejt 'neighbour'. Turner, CDIAL 13077, derives saajjha from * sagrhya which would mean 'living side by side' (cf. Vedic saksit living together or side by side'). . The variant saajjia may be due to a misreading of the Jaina Nagari ligature for -jjh- as -jj-; see the note to * 29. saandhia in Tp is the Pkt form of sagandhika 'relative' with which Madhavayajvamisra translates saajjhia. See also T, quoted in Retr., p. 428: (ta?) sagandhika prativesika, and. V, quoted in Ed., p. 124: sam) gandha. For the variant saandhia, for saajjhia, see also 335 (Tp dhia), 848 (Ti, Tp) and 849 (Ti, Tp, I). In Pkt. the Skt particle khalu was first shortened into khu, which is found after Anusvara (see, for instance, *6). After vowels its pronunciation in M. Pkt was further weakened, according to the rules of internal sandhi, into hu. This development of khalu into khu and hu should probably be explained from the unstressed or unemphatic nature of the particle (see *15). Note in this connection that khalu is never found in the emphatic position at the beginning of the sentence. On the other hand, it is often found in fixed idioms, such as na hu (here, 48, 227, 279, 478, 666), ma hu (521, 607, 909; in 909 kin nu khu in Ti and Tp) and taha vi hu (229, 275 (?), 470, 647, 741), of which the first is known from Skt, whereas the latter two are not. For taha vi hu a variant taha vi a tathapi ca) is found occasionally in the MSS of the South-Indian recensions (275 Ma, Ti, Tp, 470 T,

Page #265
252 647 Tp; in the last instance taha vi a khu in Ma). Otherwise hu occurs freely (203, 212, 230, 242, 265, 401, 425, 498, 775, 889, 891, 896, 903, 926). For the meaning of (k)hu, see Emeneau (1969: 240-68). For hu in the Jaina and the Vulgata MSS, Ma, Ti and Tp have throughout a variant khu (note one typical reversal, in 478, with hu in Ma, Ti and Tp, and khu in v and ku(?) in R and y). khu is also occasionally found in T and, where available, in W. S often has kkhu (against the metre). Khu in these MS would seem to be secondary. As such it may be an indication of the influence of Saurasent on these MSS, which dialect has either khu or kkhu but never hu (see Pischel & 94) (for other possible traces of this influence on the South Indian MSS. see *27). 40( 38 Bh, 40 R; 37 K, B, 38 V, V, P, 37 G, 405, IT; 560 S) Missing in Ma, Ti, Tp, T and W. . talurabh amadanakhudi- daravuddavuddanivvud- akesaro girisate purena damahuaro hirai kalavo. khuddiya Bha, bhamanas ankhudia R; bh amaulakhudia K, M, B, Y, P ! (okhadia); S -- naie K, Y; naya , nata B, naf P; nata S. -- daravuddapun ibudda Bho, vudduvuddun ibudda Bh", "buddabuddan ibbudda R; cuduvudun ivudu K, vuddaghudanighuda W, dakhuduchuunichuda B, ovuduvudunivvudu Y, budubuduni budu P; ovuddhavuddhanivuddha S -- h Trae S -- kalambo K, B, P; kaavo S (Bh, R, W, Y). The Kadamba ( -flower) is dragged along the stream of the mountain-river, its filament broken by the violent movement of the whirling water; the bee (sitting on it) first sank a little, then sank and finally went under completely. For talura 'whirlpool', see Paiyal 129: talaro avatto, and Desin. V 21 (Hemacandra's own commentary): talaro avartasceti kecit. For bhamacana 'tossing around', see Hem. III 151, IV 30 and 161, Pischel S 554 and, finally, Turner, CDIAL 9648. Bloch (1934: 224) connects the suffix -ad- with the nominal suffix -- which is unknown in

Page #266
253 Pkt but characteristic of Apa. (see Alsdorf 1937: 18-9). The readings found in R (bhamanasan(khudia)) and the Vulgata MSS (bhamaula(khudia)) seem to represent respective attempts to remove this Apa. formation, which is foreign to Pkt proper, from the text. A similar phenomenon can be seen in 92. The fact that bhamadana is a relatively modern formation and the fact that the Gatha was not present in the South-Indian recensions, suggest that the Gatha was not included in the archetype but was added to the text only in a later stage. A similar conclusion on similar grounds can be reached regarding 77 and 92 (see also above, p. 162). In Hem. IV 116 khuda- is enumerated among the Adesas for tud- 'to break'; for the participle khudia, see Hem. I 53. In the SattasaT otherwise only ukkhudia 'broken open' occurs (331). For khuda-, see also Turner, CDIAL 3892. For vudda- 'to sink' (or 'sunk' the past participle being identical to the stem), see Hem. IV 101. The verb is amply attested in NIA (see Turner, CDIAL 9272). vudda- does not occur in the Setubandha or the Gaud avaho. The verb is, however, found in Apa. (see Hem. IV 415 and 423); nir- in nivvudda-, seems here to intensify, as in Skt nirmajj kalava "Nauclea cadamba' corresponds to Skt kadamba (Paiyal. 255, Var: II 12 and Hem. I 30, 222). As Tamil katampu shows, the - 1 - in Pkt is more original than the -d- in Skt (see Mayrhofer, EWA I, p. 144). , ; 41( 76 Ma, 77 Tp; 39 Bh, 41 R; 38 K, B, P, 39 X, Y, 38 G, 41 384 T; 293 S) Missing in W. ahiaamanino dugniavandhavana jhurai gaassa chahim paissa rakkhaitT gharinT vihavena ertana. In Ma lacuna from ahiuaman(ino) to (ghari)nT. ahiai Tp; ihiiyaa Bha ahiyai Bho, ahiyani R; ahiai , B -- dugaassa P -- chaan Tp; 'Bh, R; K, Y, P, chaahi U, chaam T; S -- payassa Bh; piassa K, Y, P -- rakhkhanty K -- bandhavana Tp; B, P, vandhavanah K; bamdhavana T; (Bh, R, X, Y) -- kuppai Tp; jurai Bh, R; V, Y, kujhjhai B, surai P; jurai T; jurai S -- vibhavena Tp -- ertana Ma; istanan Bha (ento Bh), R; pattana K, Y (degana), P, ettana , B;

Page #267
ehtagam T; ettaga S. 254 Not wishing to rob her husband who was poor but had high notions of his birth, of this delusion, the house-wife was angry at her own relatives when they arrived with pomp and splendour. I suggest to take chaht in chahim...rakkhamt in the sense of 'delusion' (i.e. about his own greatness). Compare Skt chaya 'hallucination' in Susruta-Samhita 1.114, 13, 15 (PW: 'Schattenbild, Hallucination"). For the meaning given to rakkha-, compare the expression papaan...rakkhanto in Gatha 122, "protecting", 1.e. "not wishing to hurt" the affection (his first wife has for him)'. chahi is restricted to and B, the other MSS having chaam. There are reasons to assume that chahim is actually a retention.. It is in any case difficult to see why, if the text originally had chaam, the latter was replaced by the more difficult form chahim (no clear explanation of the form chaht beside chaa is available; see *37). There was no necessity to do so from the point of the meaning of the word either. chahi occurred side by side with chaa. If chahT is indeed original here, one has to explain what in this particular instance could have been behind its change into chaa. The fact that this change took place independently in the South as well as the North-Indian branch points to an external factor. In this connection the traditional interpretations of the word should be considered. (Unfortunately in Madhavayajvamisra's commentary in MS Ma nothing specific is found in relation to chahim.) Bhuvanapala (ISt., p. 42) apparently takes the word in the sense of 'complexion': vailaks (y)ena ma malinimanam Systu iti nijabahdhavebhyo as@yati, in order to prevent (the husband) from acquiring a dark complexion from embarrassment she shows anger to her own relatives'. In S (Retr., p. 360) and in E and m (Ed., p. 15) the word is glossed with mahatmyam; in G (Retr.. p. 360) with mahattvam. 'Greatness' can here only denote the image. which the husband thinks he projects. As such the meaning of these glosses comes very near to that of 'delusion' suggested above. The gloss kirti 'fame' for chaa found in DesTn. III 34 may well have been based

Page #268
255 on these, taken literally. On the other hand, these interpretations all seem to proceed from the meaning 'beauty' or 'lustre'. In *37 I have referred to Hem. I 129 where it is said that where the word chaht occurs it invariably means 'shade' or 'shadow'. Undoubtedly this restriction was initially based on the observation of the occurrence of the word in the texts. It is not unlikely, though, that it subsequently acquired the character of a prescriptive rule. This may have been responsible for the change of chaht into chaa in the Gatha under discussion. In any case the rule did not prevent the substitution of chaa as the easier form, for chaht, the more difficult one. For jhura- 'to be angry at' and the variant jura-, see *13. The variant pattana(m) for entana rests on a misreading of the Nagart character e- for pa-. Regarding the variant entana found in Ma it should be noted that in this MS a vowel at the end of the Gatha is occasionally lengthened; e.g. piaamammt in * 17. , ; 421 78 Ma, 79 Tp; 40 Bh, 42 R; 39 K, B, 40 V, Y, P, 39 G, 42 449 S) Missing in T and W. sahtne vi piaame duggaapautthavai am patte vi chane na madio appa saajjhiam santhavamtte. sahine Ma -- pi P -- pate Ma -- cchane Ma; khane S -- duggai R; dugaa P; ciraara S -- pautthapaiam Ma, Tp -- saacchiaam Ma; sajjhian R; saajjian K, V, B, Y, P; saajhjhia S. Though her dearest one is at her disposal (and) the festivalday has arrived she has not adorned herself, (ostensibly) in an attempt to cheer up her neighbour who is poor and whose husband is abroad. The woman would not adorn herself out of sympathy with her neighbour, who is poor, whose husband is still abroad and who consequently cannot afford and is not allowed to adorn herself. However, so much sympathy between neighbours is suspect. A comparison, with, for instance, 848 suggests that the woman may in fact have an affair with her neighbour's

Page #269
256 husband and does not adorn herself because he has not yet returned home. The Gatha reads as follows: ruai ruamtte mae piavirahe mami saaj- ohidine ganai jhijjai aham va jhiae neho ccia apuvvo. She cries as I cry, she counts off the days and grows thin like I do. Aunt, the affection of the the neighbour (for me) during my dear one's absence is unprecedented. chana 'festival' (see Paiyal. 248 and Hem. II 20) goes back to Skt ksana '(right) moment', which otherwise became khana, meaning 'moment'; see *14 and Turner, CDIAL 3642. For the reasons to adopt pautthavaiam as found in Bh, R, etc. in- . stead of paiah found in Ma and Tp, see above, p. 164. For saajjhia 'neighbour', see *39. 5, it; 43( 77 Ma, 78 Tp; 41 Bh, 43 R; 40 K, B, 41 W, Y, P, 40 G, 43 305 T; 437 s) Missing in W. tujjha vasai tti hiaa tuha virahe kisiaii imehi dittho tuman ti acchti ti tte angai vi piai.. tujha K, tujhjha , B, P; S -- vasahi Bh, R; vasihi S -- diththo K -- acch Thi Ma; achThin Bh, achihim R; Y, achTin K, acchTim B, P; T; S; (v) -- kisiaati tti Tp; kisiaati R; kisiani tik, kisiai ti , B, Y, P; kasiai ti T; kisiai ti S -- tena Bh; tfa S -- asgain K -- puai Ma; piain K, B, P; T; S; (Bh, R, W, Y). Her heart is dear (to you) because it is your house, her eyes because through these she looks at you, (and) her limbs (I suppose) because they have become thin during your absence. The variant kisiaaiti (R, Tp) is due to the spelling kisiayah ti for

Page #270
257 kisiain ti. kisiaati is interpreted as krsayante (R). In Tp the quotative particle which was thus lost is added again: kisiaati tti (cf. & and : krs Tbhavanttti). , ; 44( 79 Ma, 80 Tp; 42 Bh, 44 R; 41 K, B, 42 U, Y, P, 41 G, 44 557 S) Missing in T and W. sabbhavanehamaie anahiae una hiaan ratte rajjijjai tti juttam in a ja dijjai tam jano hasai. sabbhavangeha Ma; sabhbhava k -- bharie R; W, B, mahie P-- rattijjai Ma, raccijjai R; rattijjai y -- annahiae y -- in Ma lacuna for hasai. You may be full of true love (but, mind you) falling in love with a loving man, that is (a) good (combination). But at this people laugh, namely when the heart is given to a man who is completely heartless. Kulanatha (Abh., p. 87) and Bhuvanapala (ISt., p. 42) translate the word maie with maye. maia would be an extended formation: * mayika. It may be asked, though, if maia does not go back to madita (as in ummaia, from unmadita, in 498 etc.) 'drunk' or 'full with' (from saturated!; cf. matta in 66). For the translation this hardly makes any difference. nehamaia occurs also in 450. Weber, following the commentaries, takes sabbhavanehamaie as a locative singular masculine, linking it to following ratte. Alternatively, it could be taken as a vocative of the feminine, in the same way as vahuvihavilas ar asie in 477 (not so interpreted by Weber and the commentaries). This interpretation may in fact be preferred. Note in this connection the absence of a connective particle between omaie and ratte. Furthermore, maie as a locative linked to ratte, seems to disturb the clear parallelism between ratte rajjijjai, on the one hand, and anahiae una hiaas dijjai, on the other. For the variant raccijjai in R, see Hem. IV 422 (23) (Apa.) and Turner, CDIAL. 10583.

Page #271
258 ana- in anahiae is a double a-privans, an- + a (cf. CPD I, p. 142"), meaning '(completely) without' (cf. Pali anabhava 'non-existence, annihilation'). See in this text also anahontaa (212), anarama am (so read; 325), anarasia (555) and ana(horana) (774). See also Hem. II 190 and Pischel in his translation of Hem., p. 92, where earlier litarature on ana- is quoted. 45( 25 Ma, 26 Tp; 43 Bh, 45 R; 42 K, B, 43 X, Y, P, 42 G, 45 I, T; 30 T; 600 S; 26 W) arambhamtassa dhuan ta maraNGam anarasbhe lacchi maranai va hoi purisassa vi hoi laccht una na hoi. ar arbhatassa Ma -- dhura Ma, Tp; dhuvan Bho; dhur an T; dhar an W -- lacht -- ca R -- purusassa Tp; R -- maranam ar an ar abhe Ma -- hovi P -- lacht K. A man who shows initiative is certain of either wealth or death. Even when he lacks all initiative (that same) death is there but there will be no wealth. dhuan of the North-Indian branch is preferred to dhuram of the SouthIndian branch. dhur am would go back to Skt dhura 'yoke, burden'. It is doubtful, though, if dhura provides a suitable object to arambha- 'to undertake, to take the initiative' or if arambha- is a suitable verb to dhura (cf. Pali vahanto...dhuram; PTSD, p. 342"). Originally dhura may well be nothing more than a corruption for dhuam. Note the frequent occurrence of -r- in this and the next line. For purisa 'man', see Var. I 23 and Hem. I 111. Ma, Ti, Tp and W have in all instances purusa (purisa here in Ma is the sole exception in this MS), a form which is also occasionally found in some of the other MSS, as here in R, in 196 in R and S, and in 380 in 8. purusa is most likely an innovation in these MSS introduced under the influence of Skt.

Page #272
259 46( 82 Ma, 83 Tp; 44 Bh, 46 R; 43 K, B, pa, 44 , 37 Y, P, 43 G, 46 5, T; 579 T; 342 S) In P as 37 and as 43 (pa), in the latter case with Chaya but without commentary. Missing in W. virahanalo sahijjai ekkaggamapavaso asavandhena vallahajanassa mae maraNGam visesei. saijjai y -- asibamdhena Ma, Tp; P, po; T; (Bh, R, W, Y) -- dullaha R; vallabha K, vallaa P; pa -- ikkagjama Bh -- u mai Bha, mami Bho. The fire of separation from a beloved person is endured by the band of hope. (But) a man's absence from home while he is still in the same village, mother, is worse than death. Garrez (1872: 213) compares the first line with Sakuntala, verse 91 (ed. Bohtlingk): garuan pi virahadukkham asabandho sahabedi. For as avamdha earlier Weber (Abh., p. 88) referred to Meghaduta 11 (ed. Hultzsch). For the present interpretation of ekkaggamapavasa, see Garrez, loc. cit. Compare ekkaggama ccia pauttho in 129. 47( 83 Ma, 84 Tp; 45 Bh, 47 R; 193 K, 192 , 194 Y, 200 P, 190 G, 425 T; 343. S: 21 W) in k the commentary is also found between 43 (*46) and 44 (+48), numbered '44'(!). Missing in B, E, and x. ajjah cea pauttho racchamuhadeulacac- ajja cia sunnai jaai carai amh am ca hiaai. ajja ccea Tp; ajja ceva R; ajja ccea K; T; ajja cia W -- a cia Ma, ajja ccia Tp; ajja cia K, ajam cia P; ajja cchia T; ajja cvia S -- sunnaaim R; K (suttanaain), W, Y, P (sunama in); S (sungain); W -- jayam in R; jaain K, P; T; S (Bh, X, Y, W) -- rattha k -- devaulacaccarai Ma, Tp; caccaradeulaimh Bha, deulacaccarain Bho, ocaccarain R; X, Y, cattarai K, P; devulacaccarai T; cattaradeuldim s - - ahma ca Ma, Tp; amh an va Bh -- hiaain K, P; T; S; (Bh, R, W, Y) --

Page #273
260 racchamuhan devaulla cattarai amha ca gihaai sungai W. Today he went abroad and today everything has become empty: the streets, the temples, the crossroads and our hearts. For the present position of the Gatha as 47, see above, p. 161. For the phrase ajja cea pauttho, see also 58. deula (see also 109) is from devakula, with loss of the syllable -va-; see Hem. I 172. Compare raula from rajakula, quoted by Bhamaha, ad Var. IV 1. For deula, see also Turner, CDIAL 6524. For caccara, see *39. , 44 G, 47 , ; 48( 602 Ti, Tp; 46 Bh, 48 R; 44 K, B, Y, P, 45 125 S) Missing in T and W. akkhudai pia hiae ditthe sarisammi gune ann an mahildan am ramastassa sarisammi gune aTs arte. na kkhudai Tp; akhudai Bho, avadai R; akkhadai K (akhkho), \, B, Y, P (akhkho); S -- mahilajana Tp; B -- maratassa Ti -- diththe K - - sarimmi Ti -- sarimmasammi Ti; sarise va S. -- afsate Ti; ad Tsamte K, Y, P, aisante y. A man's dear mistress inevitably comes back to his mind when enjoying other women, namely when he sees a virtue equal to hers as well as when he doesn't! akkhudai has been based on Bh and on Ti and Tp (for Tp na kkhudai, see below). akkhuda- seems a variant of ahuda- the past participle of which, ahudia, is mentioned in Desfn. I 69 among the synonyms for padia 'fallen'. A NIA derivative is S. akhuranu 'to trip' (see Turner, CDIAL 1028: from *aksutati, or 1036: * akhutati, Pkt ahuda- 'to fall', *akhotayati, Pkt ahoda- 'to strike, to beat' (from 'to cause to fall')). The variant akkhadai shows the influence of askhalati; cf. the gloss askhalati given by PTtambara and Gangadhara (Retr., p. 360). na kkhudai

Page #274
261 found in Tp seems due to the connection made with khuda- 'to break' (see *40). The Chaya in Ti reads na trusiti, and Madhavayajvamisra has khuhai (khudai?) trutati vicchidyate (passive!), smaryata iti yavat. The emphasis in the first line is on the verb, which is found at the head of the sentence (see 8). ats amte, from a + disante, shows a rule of internal sandhi (see also 224 and 388; cf. aittha in 414, 526 and 748). This seems to have to do with the relatively frequent occurrence of this combination, as a result of which it has come to be treated as one word. Compare such phrases as na ittha (502, v.1.) or na anasti (*2), for which a similar explanation should be given.. , 45 G, 48 5, T; 49( 603 Ti, Tp; 47 Bh, 49 R; 45 K, B, Y, P, 46 533 T; 682 S) Missing in W. naipurasacchahe jov- aniattasu a rat- vanammi diahesu niccapahiesu su putti kim daddhamanena. ; sachahe K, B, P, satthahet; sariche T; sachahe S -- jauvvanammi Ti; jovvaemmi K -- divasesu piccapahiesu R; niccapahiesu diahesu K (diahamsu), Y (dihaesu), P (pahiesa), aipavasiesu dianesu , B; saiposiesu diahesu T; aipavasiesu diahesu S -- aniattasu vi R; K, Y,.P; T;. anivuttasu vi S -- rasu V , rattin T -- puttali T -- dhaddha Tp; d'addha Bh, R; daththa K, dagha v', dattha ?, didhdha B, ddaththa Y, lacuna P. What use is (this) cursed sulking, little girl, when youth (passes quickly) like the stream of a river, when the days are constantly on the move and the nights do not return: For sacchahe 'similar to', see 5. In MIA the initial d- of the Skt verbs dah- 'to burn' and dam) - 'to bite' becomes - Pali invariably has dahati, Skt daha occurs as daha as well as as daha. The past participle invariably occurs as daddha. Thus either the initial plosive or the medial cluster is retroflex, but

Page #275
262 not both at the same time (on the medial -ddh-, see below). Likewise . Pali has dasati, sandasa and damsa, and dattha and datha. In the Sattasas one finds dakka 'bitten' in 237 (Ti, Tp, Bh, R, K,Y; dakka in W, P, T, S, W) and 532 (Ma, Ti, Tp, Bh, R, W, S; dakka in Y, T, 54, W), dattha in 636 (Ma, TP, W, Y, T; cakka in R (Bh not available)), dadha 'fang' in 402 ( U, Y, S) and in 761 (Bh, R, T, S), and dasana 'tooth' in 522 (Ti, TP, BH, K, T, S, nahara in R, W, Y, P) (note that Var. II 35 explicitly mentions dasana!). In 105 dajjhihisi is found (but bhamjihisi in Ma and Tp), in 321 Jahai (dahai in T), in 373 qajjhai, in 401 dajjhasi (su), in 758 dahiuna, in 193 qana (Ma, Tp, Bh, V, Y; daha R, K, S), in 726 dahana (R; dahana s), and in 799 dahana (R). The past participle of daha- 'to burn' occurs frequently, in several cases, as here, meaning 'cursed'. Both daddha and daddha are found, which are variants. Thus, the MSS of the South-Indian recensions, Ma, Ti, Tp and T, where available, have qaddha throughout. The other MSS have daddha as well as daddha: *49 gaddha Ti, Tp, P, T, S, daddha Bh, R, K, W, B, Y, 134 daddha Bh, K, , P, daddha R, Y, 149 daddha Tp (lacuna in Ma), Bh, K, V, S, daddha R, Y, P, 229 daddha Ma, Tp, Bh, k, , (utthe, but uprobably for da-), daddha R, 502 daddha Ma, Ti, Tp, V, T, daddha Bh, R, S, 511 daddha , daddha S (dutthu Bhi, ucha R), 558 daddha U, S, daddha R, Y, 602 daddha Ma, Ti, Tp. V, T, S, 710 daddha S, daddha Y, 714 daddha Ma, Bh, T, S, daddha Tp(!), R, V, W, maddha Ti(!), 781 daddha R, daddha S. This state of affairs partly coincides with the observation made by Hemacandra in I 217: - would be optional in dasana and dasta, and in dagdha and daha, but obligatory in dahati and dasati. Note that Var. II 35 mentions only dasana. In the Gatha under discussion the two branches are diametrically opposed, Ti and Tp having daddha and Bh and R daddha. It is almost certain that the latter, here as well as in all its other instances, is secondary, introduced in order to remove caddha from the text which with its many retroflexes must have appeared a linguistic monstrosity. A different question concerns the origin of the initial d- in MIA

Page #276
263 Jaha- and casa-. H. Smith, quoted by Bloch (1934: 58) suggests that the - represents a case of anticipatory assimilation to the medial -ddhand - th- in the respective past participles. On the other hand, we may have to do here with early instances of the variation occasionally seen in NIA between an initial d- and 4-. See H. dtthna/dtthna (CDIAL 6518) and P. dambhna/ Jambhna (CDIAL 6121). As far as NIA is concerned an explanation for the variation is probably that at the beginning of a word the phonemic distinction between a d- and a d- is less clear. The variation may also reflect a difference in the speech-situation, varying between quiet and un involved and emphatic and involved. Such a distinction seems interesting when applied to MIA daha- and dasa-, which both denote more or less violent or painful actions. A second problem is the origin of the medial -ddh- in caddha. In fact the problem of its origin is largely ignored (not discussed by, for instance, Geiger, in SS 42, 3, and Pischel, in SS 222). AIG I S 236 suggests that Pali daddha is from *dazgha, i.e. as if the -h in dahbelongs to the old palatal series. However, *dazdha should then have become *dadha (cf. 1Tdha from *lizdha). In fact, beside daddha a form *daddha seems to have existed. Thus, side by side with darhna 'to burn' and dadha 'to blaze' or 'fire' H. has dadhna 'to burn' and dadhana 'to blaze furiously'. P. dambhna (da) 'to burn' (see above) supposes *daddha, which after laddha : lambha, has developed a stem dambha-. In my opinion the existence of daddha beside *daddha should be compared to Pali kathati 'to boil' beside Skt kvath- and Pkt kutthasu 'boil!' (for Pali kathati, see von Hinuber, 1979). The fact that it concerns two semantically closely related words can hardly be accidental. Probably the cerebralization of the medial dental(s), which appears to be optional, is contextual, signifying, for instance, the speaker's fear or horror of the actions denoted by the verbs. 50( 84 Ma, 85 Tp; 48 Bh, 50 R; 46 K, B, Y, P, 47 U, 46 G, 49 }, ; 160 T; 336 S) Missing in W. kallan kira kharahiao taha vaddha bhaavai pise pavas ihii pio tti suvvai janammi jaha se kallam cia na hoi.

Page #277
264 kallin S -- khala T -- hiyau Bh -- pavas ihai Bh, pavisahai R; pavasai K, pavasei Y, pavasehai p -- piu tti Bh -- Sungai K, u', sunai P -- manammi Ma -- vaddaha Ma; vaththa K, V, Y, vatha P; vadhdha S -- jahi Bha (jaha Bho) -- mey - kalan Ma; kallaa P; kallin S -- cia missing in P. Pada a quoted in Hem. II 186. People say that at daybreak a lover will leave cold-heartedly. Venerable Night, grow so long that for him daybreak will never come! For kalla 'dawn, next morning (or 'yesterday'; cf. Paiyal. 211: hijjo (hyas) kallam), from Skt kalya or kalya, see Turner, CDIAL 3104. For kira beside Skt kila, see *36.

Page #278

Page #279

Page #280
Index of words discussed in the notes aisamta akkhuda ana anusaa. atta alia asoa aha 44 uva(ha) uvvasia 48 35 8 27 7 23 Thi(j) T 24 i (v.l. vi) 17 ia (v.1. of via) 1 ia (v.1. of ii) 15 ii (v.1. ia) 15 icchena (v.1. kicchega) ira (v.1. kira) 36 iva (v.1. of via) 1 ua(ha) 4 -utta 33 4 32 48 e (pronoun) ehejja 0 32 omalia kalamva kalla 50 kicchena 33 kira 36 kisala 19 kudumvaa(/i)tthi (?) kuravaa 8 kulavahu 38 kouhalla 21 kosamva 19 17 32 40 15 18 267 33

Page #281
268 khu 39 khuda- 40 gamant 30 gola 36 ghaa(t)uppa 22 caccara 39 cattara 39 cihura 31 chana 42 chaht 37, 41 chikka 14 chiva- 14 jura- (v.1. of jhura-) 13 jhura- (v.1. jura-) 13 daddha 49 na ana- 2 na uno 20 -nahdana 30 navari(n) (v.l. navara(ih)) 15 navara() (v.l. of navari(m)) pavalaa 28 . na hu 39 . nivvudda- 40 nisavia (v.1. nisamia) 29 nomalia 32 nohalia 6 tamsa 10 tai 32 tae 32 taNGNGaa 19 tatt 2 taha vi hu 39 talura 40 tui 32 tue 32 tuppa 22 15

Page #282
tuha 36 daddha dadhal 11 dasia 18 diara 38 digha. 11 dume 11 49 de (pronoun) de (particle) deara deula dega devara donga (m) dogha (m) dohala dho (v)a patthT 11 padaal 18 28 47 15, 29 28 putth purisa 27 15 27. 18 15 padhamullaa 15 panavia (v.1. pagamia) 22 parahutta 35 parimusaga 26 16. pahara 28 pahuppa- 7 paada (v.1. of paua) padala (11) 13 piva (v.1. of via) 1 11 45 purusa 45 bola (v.1. of vola-) bhamadana 40 bhara- 22 bhisinT 4 'maia 44 2 21 269

Page #283
mailia majjhaara mar agaa malla maha reha lagga ladahavila lua 8 vacca 14 27 28 4 maNGailla 27 ma hu 39 miva (v.1. of via) 1 rakkha 41 4 21 19 vajjhapadaha vannagghaa(t)uppa valaa 4 vola- 21 volaia 21 sarkhasippT sambhara sacchaha 5 saajjia 39 sal hagijja sippi 4 silahagijja sungaia somala hu 39 -hutta 2 valumki 10 via (v.1. fa, iva, piva, miva, viva) viaga 38 vilaa 7 viva (v.1. of via) 1 vudda- 40 vesavila 7 26 36 7 35 29 4 29 22 12 121 270

Page #284
271 Index of grammatical points discussed -a -accia + -eccia Accusative Anusvara Apabhras a : -a tti + -etti -a va * -e va bhua in the South Indian MSS ending of the instrumental and genitive of feminine I-stems, e.g. kodta feminine gender-marker, e.g. erisa e.g. kunamta e.g. seccia denoting the time at which, e.g. oasah added after the endings of the plural ending in -i and -a, e.g. -aim, -ehin and -ana p. 168 ff. early Apa. forms in Pkt, e.g. aha 23 e.g. tai 32 Apa. features in Gathas added later in the North-Indian branch, e.g. bhamadana 40, p. 162 e.g. se tti 23 e.g. devae va 23, 32 functioning as a tense-marker, e.g. mand anan bhuam position of the particle of comparison non-existent in Pkt consisting of more than two members, in the neutre singular, e.g. rairahasiharadhaa in the South Indian MSS e.g. paniggahana ccia development in Pkt to a-stems position at the beginning of the sentence 7 retention in a foreign loan-word, e.g. maragaa position of ii, in contrast to that of (t)ti, which is an enclitic possible emphatic function, e.g. manailla (cf. -ullaa) Comparison Deponentia Dvandva -eccia + -a ccia e-stems Finite verb -9 ii (Skt iti) -illa

Page #285
272 -jjh-/-jj Karmadharaya -kka Lengthening numbers confusion, e.g. vajjha/vajja e.g. saajjhia/saajjia bilingual', e.g. sankhasippi e.g. (vannag) ghaa(t)uppa verbal suffix of initial syllable, e.g. ahi(j)at in the neighbourhood of other nasals above 'one', loss of gender distinction e.g. sea ccia e.g. padaha va e.g. murave va functioning as an action noun in conditional clauses internal sandhi in compounds influence of Skt on the spelling influence on the MSS broken-up functioning as a personal pronoun 23 -occia + -a ccia -Ova + -a vva -Ova + -evva Past participle Present participle Sandhi Sanskrit Saurasen 1 Sentence 23 22 26 p. 164 p. 163 27, 39 20 si <6 taha...sariccha -ullaa -V emphatic, e.g. padhamullaa secondary insertion or deletion in -ua- (or -uva-) p. 165 ff.

Page #286
273 Bibliography AKM Abbreviations Abhandlungen fur die Kunde des Morgenlandes BB Bezzenberger's Beitrage zur Kunde der Indogermanischen Sprachen BSOAS * Bulletin of the School of Oriental and African Studies IHQ The Indian Historical Quarterly IJJ Indo-Iranian Journal Indian Linguistics ISt Indische Studien JA Journal.Asiatique JAOS Journal of the American Oriental Society JOIB Journal of the Oriental Institute of Baroda KZ (Kuhn's) Zeitschrift fur vergleichende Sprachforschung STII Studien zur Indologie und Iranistik ZDMG Zeitschrift der Deutschen Morgenlandischen Gesellschaft Alsdorf, L., 1935-37: 'The Vasudevahindi, a Specimen of Archaic Jaina Maharas tri', BSOAS VIII, pp. 319-333 (Kleine Schriften, pp. 56-70) Alsdorf, L., 1937: Apabhraia-Studien, AKM XXII 2, Leipzig Alsdorf, L., 1950: "Pancatantra-Miszellen', ZDMG 100, pp. 356-361 (Kleine Schriften, pp. 586-91) Alsdorf, L., 1974: Kleine Schriften, ed. by A. Wezler, Glasenapp Stiftung Band 10, Wiesbaden Andersen, D., 1901: A Pali Glossary, repr. Delhi, 1979 Berger, H., 1955: Zwei Probleme der mittelindischen Lautlehre, Munchen The Bhattikavyam of Bhatti with the Commentary (Jayamangala) of Jaya mangala, ed. by V. N. S. Joshi and K. V. Joshi, Bombay, 1900 Bloch, J., 1934: L'Indo-aryen. Du Veda aux temps modernes, Paris Bloomfield, M., 1899: The Atharvaveda, Encyclopedia of Indo-Aryan Research II 1 B, Strassburg Brough, J., 1962: The Gandhari Dharmapada, London Burrow, T. and Emeneau, M. B., 1961: A Dravidian Etymological Dictionary, repr. of the 1st ed., Oxford, 1966 Candralekha: Rudradasa's Candralekha, ed. by A. N. Upadhye, Bharatiya Vidya Series No. 6, 2nd ed., Bombay, 1967 The Dasarupa, or Hindu Canon of Dramaturgy, by Dhananjaya; with the

Page #287
274 Exposition of Dhanika, the Avaloka, ed. by F. E. Hall, Bibliotheca Indica, New Series Nos 12, 24 and 82, Calcutta, 1865 The Desinamamala of Hemacandra, ed. by R. Pischel, Bombay Sanskrit Series No XVII, 2nd ed., Bombay, 1938 The Dhvanyaloka of Anandavardhanacharya, with the Commentary of Abhinavaguptacharya, ed. by Pandit Durgaprasad and K. P. Parab, Kavyamala 25, Bombay, 1891 Dundas, P., 1981: 'Some linguistic Notes on the SattasaT', JOIB XXX, pp. 205-13 Emeneau, M. B., 1968: 'Prakrit etymological Notes', IL 29, pp. 30-33 Emeneau, M. B., 1969: 'Sanskrit Syntactic Particles kila, khalu, nunam', IIJ XI, pp. 241-68 Fick, R., 1897: Die sociale Gliederung im nordostlichen Indien zu Buddha's Zeit. Mit besonderer Berucksichtigung der Kastenfrage. Vornehmlich auf Grund der Jataka dargestellt, Kiel, repr., Graz, 1974 Gairola, Ch. K., 1956: 'Die Anfange der Satavahana-Dynastie in Indien', ZDMG 106, pp. 155-165 Garrez, G., 1872: Review of Ueber das Saptasatakam des Hala. Ein Beitrag zur Kenntniss des Prakrits, ed. by A. Weber, AKM V 3, Leipzig, 1870, in JA, pp. 197-220 The Gaudavaho, a Prakrit historical Poem by Vakpati, 1st ed. by Shankar Pandurang Pandit, 2nd ed. by Narayan Bapuji Utgikar, Bombay Sanskrit and Prakrit Series No. XXXIV, Poona, 1927 Geiger, W., 1916: Pali Literatur und Sprache, Grundriss der Indo Arischen Philologie und Altertumskunde, 1. Band, 7. Heft, Stassburg Ghosal, S. N., 1959: 'The Euphonic Glide H in Prakrit', JOIB IX, pp. 256-259 Godavarman, L., 1935: 'The Change of a to e in the Indo-Aryan Loan Words of Malayalam', BSOAS VIII, pp. 559-562 Gonda, J., 1952: Remarques sur la Place du Verbe dans la Phrase active et moyenne en Langue Sanscrite, Utrecht. Gonda, J., 1968: 'Abbreviated and inverted nominal Compounds in Sanskrit', in Pratidanam, Indian, Iranian and Indo-European Studies Presented to F. B. J. Kuiper on His Sixtieth Birthday, ed. by J. C. Heesterman et al., Janua linguarum. Studia Memoriae Nicolai van Wijk dedicata Series Major 34, the Hague, pp. 221-246 (Selected Studies 111

Page #288
275 43-68) Gonda, J., 1975: Selected Studies, in five volumes, Leiden Handiqui, K. K., 1976: Pravarasena's Setubandha, translated into English with extracts from unpublished commentaries, critical notes and an introduction, Prakrit Text Series No. 20, Ahmedabad von Hinuber, 0., 1979: 'Pali kathati: ein Beitrag zur Uberlieferungs geschichte des Theravada-Kanons', IIJ 21, pp. 21-26 Jacobi, H., 1886: 'Unregelmassige Passiva im Prakrit', KZ 28, pp. 249 256 (Kleine Schriften I, pp. 110-117) Jacobi, H., 1898: 'Der Akzent im Mittelindischen', KZ 35, pp. 563-578 (Kleine Schriften I, pp. 73-88) Jacobi, H., 1970: Kleine Schriften, ed. by B. Kolver, in two volumes, Glasenapp-Stiftung Band 4, 1-2, Wiesbaden Janert, K. L., 1965: An Annotated Bibliography of the Catalogues of Indian Manuscripts, Part I, Verzeichniss der orientalischen Hand schriften in Deutschland, Supplementband 1, Wiesbaden : The Kalpasutra of Bhadrabahu, ed. by H. Jacobi, AKM VII 1, Leipzig, 1879, repr. Nendeln, 1966 Kas avaho: Rama Panivada's Kais avaho, ed. by A. N. Upadhye, 1940 2nd issue Delhi, 1966 Karpuramanjart: Rajasekhara's Karpuramanjart, ed. by S. Konow and Ch. R. Lanman, Harvard Oriental Series Vol. IV, 1901, repr. Delhi, 1963 Katare, s. L., 1952: 'An inscribed Sculpture inspired by Hala's Gatha saptasati', IHQ XXVIII, pp. 379-385 Kavyadarsa: Dangin's Poetik (kavjadarsa), ed. by O. Bohtlingk, Leipzig, 1890 Kavyaprakasa of Mammata, with the Balabodhint of the late V. R. jhalakikar, ed. by R. D. Kamarkar, 6th ed., Poona, 1950 Keith, A. B., 1928: A History of Sanskrit Literature, Oxford Krishnamachariar, M., 1970: History of Classical Sanskrit Literature, 1st repr., Delhi Der Kumarapalapratibodha. Ein Beitrag zur Kenntniss des Apabhr ansa und der Erzahlungs-Literatur der Jainas, ed. by L. Alsdorf, Alt- und Neu-Indische Studien 2, Hamburg, 1928 Lienhard, S., 1975: 'Sur la Structure poetique des Therather Tgatha', JA CCLXIII, pp. 375-396 Levi, s., 1902: 'Sur quelques Termes employes dans les Inscriptions des Ksatrapas', JA XIX, pp. 94-125

Page #289
276 Levi, S., 1963: Le Theatre Indien, 2nd ed., Paris Luders, H., 1940: Philologica Indica, Gottingen Luders, H., 1942: 'Von indischen Tieren', ZDMG 96, pp. 23-81 (Kleine Schriften, pp. 490-548) Luders, H., 1973: Kleine Schriften, ed. by 0. von Hinuber, Glasen app Stiftung Band 7. Wiesbaden Manava Dharma-sastra, the Code of Manu, ed. by J. Jolly, London, 1887 Mayrhofer, M., 1955: 'Das Problem indo-germanischer Altertumlichkeiten im Mittelindischen', in Studia Indologica. Festschrift fur W. Kirfel, ed. by 0. Spies, Bonn, pp. 219-241 Mayrhofer, M., 1956-1980: Kurzgefasstes etymologisches Wortenbuch des Altindischen, in four volumes, Heidelberg Meghaduta: Kalidasa's Meghaduta, ed. by E. Hultzsch, Prize Publications Fund Vol. III, London, 1911. Mirashi, V. V., 1947: 'the Date of the Gathasaptasatt', IHQ XXIII, pp. 300-310 Monier-Williams, M., 1970: A Sanskrit-English Dictionary, New Edition, Oxford Nayadhammakahao, in Anga Suttani, ed. by Muni Nathamala, Vol. III, pp. 1-391, Ladnun, 1973 Nitti-Dolci, L., 1938: Les Grammairiens Prakrits, Paris The PaiyalachchT Namamala, a Prakrit Kosha, by Dhanapala, ed. by G. Buhler, BB 4, 1878, pp. 70-166 Panhavagaranain, in Anga Suttani, ed. by Muni Nathamala, Vol. III, pp. 635-713, Ladnun, 1973 Pargiter, F. E., 1913: The Purana Text of the Dynasties of the Kali Age, Oxford Pischel, R., 1900: Grammatik der Prakrit-Sprachen, Grundriss der Indo Arischen Philologie und Altertumskunde, 1. Band, 8. Heft, Strassburg, repr. Hildesheim, 1973 The Prakrta-Prakasa or the Prakrt Grammar of Vararuchi, with the Com mentary (Manorama) of Bhamaha, ed. by E. B. Cowell, 3rd ed. Calcutta, 1962 Prakrtaprakasa of Vararuci with the Commentary of Ramapanivada, ed. by C. Kunhan Raja and K. Ramachandra Sarma, The Adyar Library Series No. 54, Madras, 1946

Page #290
277 Prakrtasarvasva: Markandeya's Prakrta-Sarvasva, ed. by K. Ch. Acharya, Prakrit Text Society No. XI, Pkt Grammar Series No. 1, Ahmedabad, 1968 Printz, W., 1921: Bhasa's Prakrit, Frankfurt Raghavan, V., 1968-... : New Catalogus Catalogorum. An alphabetical Register of Sanskrit and allied Works and Authors, Madras Rama Rao, M., 1971: Studies in the early History of Andhradesa, Madras Renou, L., 1930: Grammaire Sanscrite, Paris, 2nd ed., 1968 Renou, L. and Filliozat, J., 1947-1953: L'Inde Classique, in two volumes, Paris-Hanoi Rhys Davids, T. W. and William Stede, 1921-1925: The Pali Text Society's Pali-English Dictionary, repr., London, 1972 The Sahityadarpana or Mirror of Composition, A Treatise on Literary Criticism, by Visvanatha Kaviraja, ed. by E. Roer, Bibliotheca Indica Vol. X, Calcutta, 1851 Sakuntala: Kalidasa's Sakuntala, ed. by 0. Bohtlingk, Bonn, 1846 Sakuntala: Kalidasa's Sakuntala. The Bengal1 Recension, ed. by R. Pischel, * Kiel, 1877 Sandesara, B. J., 1972: 'Ajada's Commentary on the Gathas aptasatt of Hala', JOIB 22, pp. 45-47 Ueber das Saptasatakam des Hala. Ein Beitrag zur Kenntniss des Pracit, ed. by A. Weber, AKM V 3, Leipzig, 1870, repr. Nendeln, 1966 Das Saptasatakam des Hala, ed. by A. Weber, AKM VII 4, Leipzig, 1881, repr. Nendeln, 1966 Saptasatisara with Bhavad Tpika of Vema Bhupala, along with the Chappannaya-Gahao (text and Chaya), ed. by A. N. Upadhye, Shivaji University Sanskrit and Prakrit Series Vol. III, Kolhapur, 1970 Schneider, U., 1954: 'Acht Etymologien aus dem Agganna-Sutta', in Asiatica. Festschrift Friedrich Weller, ed. by J. Schubert and U. Schneider, Leipzig, pp. 575-583 Schubring, W., 1955: 'Jinasena, Mallinatha, Kalidasa', ZDMG 105, pp. 331-337 (Kleine Schriften, pp. 314-320) Schubring, W., 1977: Kleine Schriften, ed. by K. Bruhn, Glasenapp Stiftung Band 13, Wiesbaden Schwarzschild, L. A., 1956: 'Some Forms of the Absolutive in Middle Indo-Aryan', JAOS 76, pp. 111-115

Page #291
278 Schwarzschild, L. A., 1958: 'Prakrit Thakka, "Tired", IL 19, Sir Ralph Turner Jubilee Volume I, pp. 311-318 Schwarzschild, L. A., 1962: 'Ghummira, Gholira "agite, branlant", JA CCL, pp. 65-75 Schwarzschild, L. A., 1962: "First", "Second" and "Third" in Middle Indo-Aryan', JAOS 82, pp. 517-522 Setubandha: Ravanavaha oder Setubandha, ed. by S. and P. Goldschmidt, in two volumes, Strassburg-London, 1880 and 1884 Speijer, J. S., 1886: Sanskrit Syntax, Leiden, repr. Delhi, 1973 Spruche: Bohtlingk, 0., 1863-1864: Indische Spruche. Sanskrit und Deutsch, in two volumes, S Petersburg Sriramamurti, P., 1972: Contributions of Andhra to Sanskrit Literature, Andhra University Series No. 105, Waltair Srigaraprakasa: Maharaja Bhojaraja's Sringara Prakasha, ed. by G. R. Josyer, in four volumes, Mysore, 1955-n.d Sternbach, L., 1974: Subhasita, gnomic and didactic Literature, part of Vol. IV of A History of Indian Literature, ed. by J. Gonda, Wiesbaden The Subhasitaratnakosa compiled by Vidyakara, ed. by D. D. Kosambi and V. V. Gokhale, Harvard Oriental Series 42, Massachusetts, 1957 Thieme, P., 1937: 'Uber einige persische Worter im Sanskrit', ZDMG 91, pp. 88-146 (Kleine Schriften I, pp. 278-336) Thieme, P., 1938: Der Fremdling im Rigveda. Eine Studie uber die Bedeutung der Worte ari, arya, aryaman und arya, AKM XXIII 2, Leipzig, repr. Nendeln, 1966. Thieme, P., 1962: 'Chess and Backgammon (Tric-trac) in Sanskrit Literature', in Indological Studies in Honor of W. Norman Brown, ed. by E. Bender, Amarican Oriental Series Vol. 47, New Haven, pp. 204-216 (Kleine Schriften II, pp. 413-425) Thieme, P., 1971: Kleine Schriften, ed. by G. Buddruss, in two volumes, Glasenapp-Stiftung Band 5, 1-2, Wiesbaden Thieme, P., 1975: 'Kranich und Reiher im Sanskrit', StII 1, pp. 3-36 Tieken, H., 1978: 'A formal Type of Arrangement in the Vulgata of the Gathasaptasatt of Hala', Stil 4, pp. 111-130 Trenckner, V., et al., 1924-: A Critical Pali Dictionary, Copenhagen. Tripatht, Ch., 1977: 'Katalogisierungstradition der Jainas', ZDMG, Supplement III 2, XIX. Deutschen Orientalistentag, Vortrage,

Page #292
279 pp. 992-1001 Turner, R. L., 1928: 'Romani les and Sanskrit tasya', BSOAS V, pp. 43-51 (Collected Papers, pp. 310-318) Turner, R. L., 1966: A Comparative Dictionary of the Indo-Aryan Languages, London Turner, R. L., 1975: Collected Papers, ed. by J. Brough, London Vajjalagga: Jayavallabha's Vajjalaggam with the Sanskrit Commentary of Ratnadeva, ed. by M. V. Patwardhan, Prakrit Text Society Series No. 14, Ahmedabad, 1969 Vivagasuyam: in Arga-Suttani, ed. by Muni Nathamala, Vol. III, pp. 715-813, Ladnun, 1973 Wackernagel, J. and, Debrunner, A., 1957-: Altindische Grammatik, Gottingen Warder, A. K., 1974: Indian Kavya Literature II: Origins and Formation of the classical Kavya, Delhi Watts, G., 1972: A Dictionary of the Economic Products of India, in six volumes, 2nd repr., Delhi Weber, A., 1872: 'Ueber das Saptasatakam' ZDMG 15, pp. 735-745. Weber, A., 1883: 'Ueber Bhuvanapala's Commentar zu Hala's Saptasatakam', ISt. XVI, Leipzig, pp. 1-204, repr. Hildesheim, 1973.

Page #293

Page #294
281 Nederlandse samenvatting Het voorliggende werk is een voorstudie tot een kritische tekstuitgave van Hala's Sattasat, een anthologie, daterend uit het begin van onze jaartelling, van 700 Prakrit verzen erotisch van inhoud. De eerste editie van deze tekst, van A. Weber, dateert uit het einde van de vorige eeuw. Weber beschikte over 18 handschriften (HSS). De tekst bleek overgeleverd in zes verschillende recensies. Voor de huidige tekstuitgave wordt gebruik gemaakt van Webers materiaal, uitgebreid met drie tot nu toe onuitgegeven HSS uit Zuid-India. Deze laatste HSS bevatten de tekst van een zevende. recensie. In deel I van dit boek wordt verslag gedaan van een onderzoek naar de onderlinge afhankelijkheid van de verschillende HSS (Stemma). De belangrijkste conclusie is dat de HSS zijn onderverdeeld in een Zuidindische en een Noord indische tak. Beide gaan uiteindelijk terug op een enkel Noordindisch HS, dat echter zelf verloren is gegaan. In de : tekstuitgave wordt getracht de tekst van dit ene origineel te reconstrueren. Een andere conclusie van dit onderzoek is dat de tekst van de afzonderlijke verzen van het verloren origineel in vele opzichten het best bewaard is gebleven in de Zuidindische tak, maar de volgorde van de verzen in de Noordindische tak. A Deel II bevat een reconstructie van de tekst van de eerste 50 verzen van het bovengenoemde origineel. Hieraan vooraf gaan twee korte hoofdstukken. Het eerste hoofdstuk handelt over de vraag voor welk publiek de tekst was geschreven en/of verzameld. Gezien de superieure en patroniserende kijk die blijkt uit de verzen, op het leven op het platteland en op dat van de 'werkende klasse' moet worden aangenomen dat de tekst zich richtte tot een publiek van rijke stedelingen en hovelingen. Verder wordt gespeculeerd over de vraag waarom de dichter of dichters een Middelindisch dialect gebruikten in plaats van het 'hoof se' Sanskrit. Een mogelijk antwoord hierop is dat men zo poogde de woorden van de bewoners van het platteland, in wier mond de verzen zijn gelegd, een boers tintje te geven. Het tweede hoofdstuk handelt over de gevolgde tekstkritische methode. Sommige principes vloeien voort uit de conclusies van het stemma. In een aantal gevallen echter werkt het stemma niet, met name daar waar

Page #295
282 het variaties betreft in de spelling van de woorden. Op dit punt is de invloed van het Sanskrit in elk stadium van de handschriftoverlevering aantoonbaar. Dit leidt tot het principe dat in geval er twee spellingsvarianten beschikbaar zijn, die waarvan de vorm het meest afwijkt van het Sanskrit, in de tekst moet worden opgenomen, ongeacht de positie in het stemma van het HS waarin deze wordt gevonden. In het derde hoofdstuk volgt de eigenlijke tekstuitgave (verzen 1-50), met vertaling en noten. De vertalingen wijken in vele gevallen af van die van Weber. De verschillen worden voor een groot deel bepaald door een strikt volgen van de Prakrit tekst zelf, terwijl Weber zich juist vaak baseerde op de interpretaties zoals gevonden in de mee-overgeleverde commentaren. Deze commentaren blijken echter daartoe te onbetrouwbaar, zoals meer dan eens in dit proefschrift in de noten bij de vertalingen wordt aangetoond. In deze zelfde noten wordt verder ingegaan op, onder andere, problemen betreffende de reconstructie van de tekst, de oorsprong van de belangrijkste varianten en de betekenis van de woorden.

Page #296
283 Curriculum vitae Geboren op 1 april 1952 te Lijnden, behaalde de auteur van dit proefschrift in 1970 het einddiploma Gymnasium Alpha aan het Bonaventura Lyceum te Leiden. Vanaf 1970 studeerde hij Indo-iraanse talen en culturen te Leiden. In 1973 behaalde hij het kandidaatsexamen; in 1975 het doctoraalexamen met als hoofdvak Sanskrit en Middelindisch en als bijvak Tamil. Sinds 1 januari 1976 is hij werkzaam als wetenschappelijk medewerker voor het Sanskrit en het Middelindisch aan de Rijksuniversiteit te Leiden. Naast zijn onderwijs en onderzoek houdt hij zich bezig met het vertalen van Tamil -, Sanskrit- and Prakritliteratuur in het Nederlands.

Page #297

Page #298
on kamITanAmadhAraNapatinApaTIyamAnA tApadAyamanipAtitAnipatomeraThA dhamadhAmagaralaMyarovaniyatiIANi dvAragadAnamasaMpadAdhikamAsakhaNDA