Old Bramhi Inscriptions In Udaygiri And Khandagiri

Author(s): Benimadhab Barua
Publisher: University of Calcutta
Catalog link: https://jainqq.org/explore/020516/1

Page #1

। कोबातीर्थमंडन श्री महावीरस्वामिने नमः ।।
।। अनंतलब्धिनिधान श्री गौतमस्वामिने नमः ।।
।। गणधर भगवंत श्री सुधर्मास्वामिने नमः ।।
।। योगनिष्ठ आचार्य श्रीमद् बुद्धिसागरसूरीश्वरेभ्यो नमः ।।
। चारित्रचूडामणि आचार्य श्रीमद् कैलाससागरसूरीश्वरेभ्यो नमः ।।
आचार्य श्री कैलाससागरसूरिज्ञानमंदिर
पुनितप्रेरणा व आशीर्वाद राष्ट्रसंत श्रुतोद्धारक आचार्यदेव श्रीमत् पद्मसागरसूरीश्वरजी म. सा.
जैन मुद्रित ग्रंथ स्केनिंग प्रकल्प
ग्रंथांक :१
जैन आराधना
न
कन्द्र
महावीर
कोबा.
॥
अमर्त
तु विद्या
श्री महावीर जैन आराधना केन्द्र
शहर शाखा
आचार्यश्री कैलाससागरसूरि ज्ञानमंदिर कोबा, गांधीनगर-श्री महावीर जैन आराधना केन्द्र आचार्यश्री कैलाससागरसूरि ज्ञानमंदिर कोबा, गांधीनगर-३८२००७ (गुजरात) (079) 23276252, 23276204 फेक्स : 23276249 Websiet : www.kobatirth.org Email : Kendra@kobatirth.org
आचार्यश्री कैलाससागरसूरि ज्ञानमंदिर शहर शाखा आचार्यश्री कैलाससागरसूरि ज्ञानमंदिर त्रण बंगला, टोलकनगर परिवार डाइनिंग हॉल की गली में पालडी, अहमदाबाद - ३८०००७ (079)26582355
-
-
-

Page #2

OLD BRAHMI INSCRIPTIONS
IN THE UDAYAGIRI AND KHANDAGIRE CAVES
CALCUTTA UNIVERSITY PRESS
1999

Page #3

OLD BRAHMI INSCRIPTIONS

Page #4

,
66
www. kobatirth.org
Acharya Shri Kailasṣagarsuri Gyanmandir
अभिधातृवशात् किंवा द्रव्यशक्तिर्विशिष्यते । अतो मत्सरमुत्सृज्य माध्यस्थमवलम्बताम् ॥ ऋषि-प्रणीते प्रौतिश्चेन्मुक्का चरक-सुश्रुतौ । भेड़ाया: किं न पव्यन्ते तस्माद ग्राह्यं सुभाषितम् ॥
33

Page #5

L
OLD BRAHMI INSCRIPTIONS
IN THE UDAYAGIRI AND KHANDAGIRI CAVES
THE ADVAN
BY
Edited with new readings and critical notes
UNIVERSITY
BENIMADHAB BARUA, M.A., D.LIT. (LOND.)
PROFESSOR IN CHARGE OF THE DEPARTMENT OF PALI, HONORARY LECTURER IN THE DEPARTMENTS OF SANSKRIT AND ANCIENT INDIAN HISTORY AND CULTURE, CALCUTTA UNIVERSITY, FORMERLY A GOVERNMENT OF INDIA RESEARCH SCHOLAR IN PALI
OF CALCUTTA
RDVANCEMENT
OF LEARNING
PUBLISHED BY THE
UNIVERSITY OF CALCUTTA
1929
16942
K

Page #6

93298
PRINTED AND PUBLISHED BY BHUPENDRALAL BANERJER AT THE CALCUTTA UNIVERSITY PRESS, SENATE HOUSE, CALCUTTA
Reg. No. 372B-April, 1929---

Page #7

To All of those who paved the way
for this work
in token of respect and gratitude

Page #8

Serving Jinshasan
093298 gyanmandir@kobatirth.org

Page #9

नन सूचन इस ग्रन्थ के अभ्यास का कार्य पूर्ण होते ही नियत
समयावधि में शीघ्र वापस करने की कृपा करें. जिससे अन्य वाचकगण इसका उपयोग कर सकें.
PREFACE The preparation of a critical edition of fourteen old Brāhmi inscriptions and one table of Brāhmi alphabet in the Udayagiri and Khaņdagiri caves, such as has been attempted in the present volume, is an undertaking, which could not be faced without a feeling of diffidence. So far as I am concerned, it had never been in my contemplation to undertake this difficult work before the task of teaching the Hāthi-Gumphā inscription of Khāravela was thrust upon me by the Boards of Higher Studies in Pāli and Ancient Indian History and Culture in the session 1924-25, during the absence on leave of the colleague in charge of the subject. I proceeded, however, with the task, examining the facsimiles of this bafilling epigraph, reading and discussing its text and interpretation, week after week, in a class of pupils, all of whom readily cooperated with me, and yet the prospect of solving its riddles was far from being bright. After many fruitless attempts I took it up in all earnestness in 1927, devoting myself wholly to it during the Poojah holidays with the firm determination to come to a definite solution. It was only towards the middle of October that the faded lines of the document appeared to me in a somewhat clear light.
I was aware that since Mr. Stirling published the first notice of this inscription in 1827 in a volume of the Asiatic Researches, James Prinsep, General Cunningham and Rājendra Lāla Mitra grappled with it without even succeeding in ascertaining the name of King Khāravela, in whose name it stood. The study of it obtained, no doubt, a right direction and received a fresh impetus when Dr. Bhagawanlal Indraji
(vii)
.

Page #10

viii
PREFACE
published his edition in 1907 in the Acts du Sixieme Congres International des Orientalistes.
I knew quite well that Mr. K. P. Jayaswal stood foremost amongst those who had tried, in recent times, to carry on the work commenced by Indraji, and that he was the scholar and epigraphist to whom the Indologists would ever remain indebted for his untiring energy in leaving no stone unturned to make the contents of Khāravela's inscription known to the world and to emphasize their importance and value. In spite of the fact that he achieved much by way of orientation of the opening and concluding paragraphs of the Hāthi-Gumphā text, I felt that his treatment of the subject left room for a good deal of revision and supplementation.
Making it a point of duty to take best guidance from the publications of previous scholars and epigraphists, from Stirling, Kittoe and Prinsep to Jayaswal and Banerji, I essayed since 1925 with the aid of Locke's plaster-cast in the Calcutta Museum and with the aid of the eye-copies and estampages within my access to probe into secrets of the Hāthi-Gumphā text. And realising that the study of this important text would be incomplete without that of other old Brāhmi inscriptions in the Udayagiri and Khaņdagiri caves, I thought it would be worth while also to carefully examine their readings and renderings published by Prinsep, Cunningham and Indraji, and finally by Mr. R. D. Banerji in the Epigraphia Indica, Vol. XIII. Examining and re-examining these shorter inscriptions as they appeared on original stones, plaster-casts and facsimiles, I was able to detect certain palpable mistakes in previous publications standing badly in need of correction.
This in itself, as I believed, was a sufficient justification for venturing a fresh undertaking. But it seemed to be no less a justification that Khāravela's inscription in the Hathi. Gumphā and the remaining old Brāhmi inscriptions in the Udayagiri and Khaņdagiri caves, were not studied before in

Page #11

PREFACE
is
their inter-connection, as well as in their connection, with the table of Brāhmī alphabet in the first Tattva-Gumphā on the Khaņdagiri bill.
It will be noticed that all the shorter inscriptions have been treated in the present work, together with the table of Brāhmi alphabet, as appendices to the Hāthi-Gumphā inscription, while the Hathi-Gumphā inscription itself has been treated as the main text serving as a complete Khāravela. Carita,-anyhow, a Prakrit panegyric (prasasti) of King Khāravela in ancient Indian epigraphy, composed by some unknown Harisena or Bāņa in an elegant prose diction clearly anticipating the prose style of the Pāli Milinda-Pañha. The scheme of treatment has been conceived in such a manner that it includes Khāravela's inscription as the first text, Khāravela's chief-queen's inscription as the second text, KadampaKudepa's inscription as the third, and so on and so forth till the series closes with the Tattva-Gumphā table. And yet I am afraid that the sense of incompleteness is apt to remain in view of the fact that some four or five caves which have sunk down showing just their covering roofs on the slopes of Udayagiri have not, as yet, thanks to the goodness of the Department of Archäology, been completely exposed to view, affording a chance to the visitors to satisfy themselves as to whether they bear inscriptions or not, and if they bear any inscriptions, what their contents and characters would be. Pressure of work and shortage of funds would be lame excuses, I venture to submit, in this case, for just a few strokes of the pick and shovel are what is required to accomplish the needful work.
Even though some of the inscriptions in some of the buried caves do not find place in this volume, for the reasons stated above, it may be hoped that their contents and characters, when made known, will not have much to add to the information culled from those records which find place in it.
True it is that I have found it expedient to be guided,

Page #12

PREFACE
nay, even misguided, by Mr. R. D. Banerji, to treat the table of Brāhmi alphabet inscribed in scarlet colour, not to say written in red ink, on a dressed portion of the back wall of one of the chambers of the Tattva-Gumphā on the Khaņdagiri hill as of the same date as the Hāthi-Gumphā and other old Brāhmī inscriptions in the Udayagiri caves. On my visit to Udayagiri and Khaņdagiri, the caretaker appointed by the Archæological Department informed me that the late lamented Mr. Panday doubted the contemporaneity of this table with Khāravela's inscription. Mr. Panday might be right, but he did not substantiate his opinion with any definite evidence. Till nothing was found to upset Mr. Banerji's assumption, I thought I should make the best use of it in including the fourteen old Brāhmi inscriptions, as well as the TattvaGumphā table in a complete and coherent scheme of treatment. Whatever be the actual date of this curious table, it is certain that the general forms of the Brāhmi letters which appear in it closely resemble those of the Brāhmi characters in the Hāthi-Gumphā and other old Brāhmi inscriptions. Certainly the importance of this table lies in the fact that it clearly indicates the first step to the engraving of inscriptions with chisel and by means of whetting on pieces of stone or rocky surfaces. It goes at once to show that the engravers used first to make designs of the letters to be engraved in some sort of colour or ink. As the Hatthipāla-Jātaka (Fausböll, Vol. IV, p. 489) goes to prove, inscribing in letters of vermilion upon a wall (jātihingulakena bhittiyā akkharāni likhanam) was a common practice in India. It was not, moreover, unusual, as some of the Central Asian manuscripts attest, to annex a table of alphabet as a key to the reading of the manuscript. Whatever be the actual date of this "table, I find no difficulty in supposing its purpose to be
that of serving as a key to the reading of the whole set of old Brāhmi inscriptions in the Udayagiri and Khaņdagiri caves. I am sure that it will not make any differ.

Page #13

We
PREFACE
xi
ence to the scheme of treatment adopted in the present work, even if the Khandagiri caves, their old Brahmi inscriptions and table of old Brahmi alphabet prove to be posterior to the Udayagiri caves with their old Brahmi records.
The present work comprises two books, the first of which consists of texts and translations of the inscriptions, and the second of notes or dissertations. It is evidently a departure from the usual practice of introducing the texts, translations and word-notes by an introduction dealing appropriately with the points of general importance. In reversing the usual order, my sincere desire is to enable the reader to form his own opinion before reading the author's opinion developed in the notes. It will, nevertheless, be seen that with the text and translation of each of the inscriptions has been attached a short preface or introduction dealing with matters of textual importance and containing a full bibliography.
The reader's attention must also be drawn to the fact that as regards the Hathi-Gumpha inscription, I have preferred the paragraphic arrangement of its text to the usual and mechanical "line arrangement. But the "line" arrangement, too, has not been ignored; it has simply been subordinated to the paragraphic arrangement.
I confess that in building up the dissertations I have had to take the old Brahmi inscriptions at their face value. If King Kharavela had really recorded falsehood in his inscription, there is no means of checking it. But to raise the slightest suspicion as to the veracity of the Hathi-Gumpha inscription is, to my mind, to be over-indulgent to unnecessary scepticism.
Taking the records as they are, or as they can be made out, I have been concerned to discuss, in a threadbare manner, their historical bearings and importance, showing all possible sides and keeping an eye to consistency of the data yielded by them. What I mean by consistency of the data may be best understood in the light of an example. If, as recently

Page #14

www. kobatirth.org
-
xii
PREFACE done by Mr. Jayaswal, Namdarāja in the expression Namdarājativasasata-oghāțita-panādī be taken to signify the Nanda-era, one has got to show that the same Namdarāja in another expression, viz., Namdarāja-nīta-Kālimga-Jināsana, can be taken to convey the same meaning. If it can be shown, I say there is consistency; if not, I say there is no consistency.
As regards the disputed points admitting of two alternative readings or renderings, there are one or two instances (e. g., Goradhagirim ghātāpayitā), where I have indicated my predilection differently in the two books. I humbly request the reader, therefore, not to take anything to be my final opinion without considering the position taken up by me in the book of notes.
While I leave the whole of this work to be judged for what it is worth, I may confidently hope that if the readings proposed by me stand the test of time, they will serve to lay many old ghosts for ever, and to convince the reader once for all (1) that there is no statement in the Hāthi-Gumphā inscription as to the population of Kalinga ; (2) that there is no allusion to Rși Khibīra ; (3) that it contains no statement as to the Greek king Dimita-Demetrios retreating with his troops and transports to abandon Mathurā ; (4) that there is no mention of Āvarāja ; (5) that there is no statement as to Pithuda being ploughed with an ass-plough ; (6) that there is no statement as to the existence of a league of the Tamil powers ; (7) that there is no mention of the Maurya-era ; (8) that there is no allusion to the Nanda-era ; (9) that there is no reference as to the corpus of the Jain canon with its various divisions ; and lastly (10) that there is no such epithet of Khāravela as Bhikhurāja. On the other hand, I shall not be surprised if in the illegible portion of the twelfth year's record (I. 13), the name of Sātakaņi recurs as one of the kings subdued by Khāravela in a second campaign undertaken by him.

Page #15

PREFACE
xiii
Mr. R. D. Banerji says (JBORS, Vol. III, Part IV, p. 487) that he prepared, in 1917, two inked impressions of the Hāthi-Gumphā inscription. One of these two must be the impression reproduced in JBORS, Vol. III, Part IV, p. 472, Plate I. Mr. Jayaswal has, on the other hand, published a plate (JBORS, 1927, Vol. XIII, Parts III-IV) reproducing photographs of select portions of two impressions, which are said to have been prepared by Mr. Madhosarup Vats. And I had the privilege of separately examining the two impressions, one of which resembled the impression reproduced in JBORS, 1917, and the other resembled the extracts from one of the impressions claimed to be of Mr. Vats and reproduced in JBORS, 1927. Although I failed to unravel the mystery which shroud these impressions, I was very happy to find that one of the impressions examined by me clearly showed my reading of ceca-yathi-adhasatikam instead of ca coyathi agasatikam (I. 15), to be quite correct. As for the reading (silā)-thambhāni ca cetiyāni ca kārāpayati (I. 14), I have largely relied upon a small plate published by Mr. Jayaswal in JBORS, 1918. As for the rest of the new readings and interpretations, grounds have been stated in proper places.
A p'ate prepared by patching up select portions of two impressions, such as one published by Mr. Jayaswal in JBORS, Dec., 1927, is bound to be misleading. The sounder procedure in a matter like this is to compare the results obtained from the study of both after carefully studying each of them separately. So far as published estampages go, one has still to place greater reliance upon one published previously by Mr. Jayaswal in JBORS, Vol. III.
I could not but congratulate myself to be able, when the printing of the book of notes was far in progress, to check the errors into which I was at first led by the symmetry of the reading of the text of the inscription of Kbāravela's chiefqueen Hathisāhasa papotasa dhutunā offered by Indraji and

Page #16

xiv
PREFACE
Banerji. There can be little doubt, as the original stone clearly proves, that the correct reading is no other than Hathisā(ī) ha-sampain)ātasa dhutunā.
I cannot, however, claim that finality has been reached or can be reached. If the following pages serve to guide the study of the Hāthi-Gumphā and other old Brāhmi inscriptions along the right path, I shall have sufficient reasons to rejoice.
But I deeply regret an omission on p. 46, namely, the rendering of Arahato nisīdiya samīpe pabhāre (I. 15), "on a slope in the vicinity of the Ārhata (cave)-dwellings." The "Additions and Corrections” will indicate where this is to be inserted.
My grateful thanks are due to Prof. S. Radhakrishnan, President of the Council of Post-Graduate Teaching in Arts, in urging me to prepare a critical edition of the old Brāhmi inscriptions, making it useful to the students and discussing the readings and interpretations of all previous scholars and epigraphists. Mr. Narain Chandra Kundu, Conseil Agrée, Chandernagore, has rendered me a distinct service by translating for me required passages from M. Sylvain Lévi's article in JA. Among my colleagues in the University, Mr. Sailendranath Mitra, Dr. N. P. Chakrabarty, Dr. P. C. Bagchi, Dr. S. K. Chatterji and Dr. H. C. Raychaudhuri have helped me with their valuable suggestions but for which my treatment of the subject would have been, in some places, far different from what it now is. I am indebted also to the Hon'ble Mr. Ramaprasad Mookerjee, Mr. (now Prof.) R. D. Banerji, Dr. Radhakumud Mookerjee, Prof. S. K. Ayengar, Dr. G. N. Banerji, Mr. Cbaru Chandra Roy, and Rai Bahadur Ramaprasad Chanda for the keen interest taken by each of them in this work. I am deeply indebted to Dr. Stella Kramrisch for her note on the reliefs in the Udayagiri and Khaņdagiri caves without which the treatment of the question of chronology of the old Brāhmi inscriptions, their authors and receptacles would

Page #17

PREFACE
XV
have remained incomplete. Among the younger scholars, Mr. Prabodh Chandra Sen, M.A., Mr. Raman Chandra Bhattacharyya, M.A., Mr. Nihar Ranjan Ray, M.A., Mr. Devaprasad Ghosh, M.A., P.R.S., Mr. Sitis Chandra Basu, M.A., and Mr. Mrinal Kumar Ghosh, B.A., have assisted me in various ways, and but for their youthful zeal it is doubtful if I could have made a sustained effort. But I am in no way less indebted to Mr. J. C. Chakravorti, the Assistant Registrar, Mr. A. C. Ghatak, the Press Superintendent, and the Press Staff for their readiness to oblige me while this life-killing work was being seen through the press.
CALCUTTA UNIVERSITY, The 16th December, 1928.
B. M. BARUA

Page #18

Page #19

ALTERNATIVE READINGS AND RENDERINGS
[The asterisk mark indicates preference.]
I. 1-Airena" or Verena. I. 1--caturamta-(rakhaņa)* or caturamta-(luthana), I. 3—Asaka or Asika. I. 4-dampa or dapa. I. 5—(ā)hatapuva* or Arakatapura. I. 9—"having stormed Gorathagiri (the hill-fortress)''**
or “having killed Gorathagiri (the man)'' as a rendering
of “Goradhagirim ghātāpayitā." I. 12-anupa or junasa.
T(i)mira or Tamira. I. 14--kāya-nisidīyam (rājupa)jivakehit or kāya-nisi līyāya
(rāja)- bhatakehi.
rāja-putehi or rāja-sutehi. III.-Airasa* or Verasa. III.--Kadampa* or Kudepa. IV.--Vadukha* or Varikha, X.—“The cave (which is an excavation) of the High Func
tionary Nākiya of Bariyä” or “ The cave (which is an excavation) of Mahāmada, Bariyā and Nākiya."
xvii

Page #20

Page #21

ADDITIONS AND CORRECTIONS
P. 10, Para. 2-Set in the marginal space the Svastika symbol,
the symbol, No. 2, reproduced on p. 141. P. 29, Para. 16-Set at the end the Tree-symbol, the symbol,
No. 4, reproduced on p. 141. P. 32, L. 8, Para. 9—Read" greari" for "fariani." P. 33, L. 3, Para. 15—Read "asfee” for “ asfiy." P. 46, L. 7, Para. 15-Insert after “stone-pillars to be made”
-"on a slope in the vicinity of the Ārhata (Jain) cave-dwellings" as a rendering of Arahata-nisīdīya
samipe pabhāre. P. 47, L. 5—Read " invincible" for " undaunted." P. 47, L. 6—Read " troops and transports " for "carriers of
the realm of royal command" omitting f. n. 4. P. 81, L. 3—Read“ kothão” for “ Kothā°." P. 93, L. 3-Set the Tree-symbol, the symbol No. 1, reproduced
on p. 144, to the left of Kammasa. P. 99, L. 3--Enclose the text between the symbols, Nos. 2 and
3, reproduced on p. 144. P. 155, Foot-note 1-Read "Mālayehi” for “ mālayehi.” P. 161, L. 10, P. 171, L. 32-Read “ pasādānam" for
"pasādáyam”. P. 161, L. 11, P. 171, L. 33-Read "Hathisā(ī)ha-sampanātasa
dhutunā” for “Hathisāhasa papotasa dhītunā.”' P. 211, L. 3, L. 5-Omit"wrongly” and reference to p. 300. P. 211, L. 6-Read “ from another passage in the same Nikāya
(III, pp. 299-300)" instead of " from the Pāli pass
age."
P. 211, L. 14-Add before the sentence--"It Berar and
Konkan were the principalities etc." :-" It cannot certainly be maintained that Prof. Bhandarkar has succeeded in establishing yet a decisive interpretation on the evidence of the first passage relied
xix

Page #22

ADDITIONS AND CORRECTIONS
upon by him (A. III, pp. 76, 78). He has sought to establish that the two words, ratthika and pettanika, represent one item of enumeration in the statement, yadi vā raţthikassa pettanikassa, in the same way as the three words, rājā, khattiya and muddhábhisitta, do in the immediately preceding statement, yadi vā rañño khattiyassa muddhábhisittassa. The falsity of analogy between the two sets of words is clear from the second passage (A. III, pp. 299-300) in which the distinction between the two has been brought out by embodying the series of the three words, rājā, khattiyo, and muddhábhisitto, in one paragraph, and the series of natthiko, pettaniko, senāpatiko, gamagāmiko and pūga-gāmaņika in another. It is impossible to draw any other inference from the enumeration in the second paragraph than that ratthika, pettanika and the rest are intended each to form a separate item. And Buddhaghoşa, too, has precisely treated each of them as a separate item. The argument by analogy is weak because the association of two or more names or designations in a stock enumeration in Pali and Asokan texts (e.g., in R. E. V and R. E. XIII) is generally meant to balance up the expressions and, no less, to indicate the contiguity or similarity of places, persons or
functions, denoted by them.” P. 219, L. 33—Read “ Natya” for “Nātya.” P. 225, F. n. 3—Read "pabhatā” for “ pabbatā." P, 267, L. 22—Read “dūrādayao” for “ dūrādasyao.” P. 292, L. 35~Read "samuthāpitāhi” for “ samthão." P. 293, L. 38, L. 40–Read “ suite" for " suit.”' P. 295, L. 31-Omit (Choța-Hāthi-Gumpha). P. 29.), L. 9-10—Read“ Pāvana-Gumphā” for “Choța-Hālhi
Gumpha."

Page #23

ADDITIONS AND CORRECTIONS
I. 12--Read janasāda-bhavanam for anupa-dabha-vanam, and translate it by " the de
mon's habitat," interpreting janasāda as the same word us the Pali purisa
sādd or porisada which literally means" a man-eater." P. 210-Add : " Vijādharadhidāsa is evidently an expression similar to Vijjahara-nagara.
vāsa (Jaina Jambudvīpa-Prajñapti, Ch. I), which means the Vidyadhara capital (rajadhani).' The Jambudvipa-Prajiapti alludes to 50 Vidyadhara cities in the south, situate on two sides of the Vaitādbya range, extended length wise, dividing South India from the Northern, and to 60 cities in the porth, situate along the spurs of the Himalayan mountain. The cities are described as rich, secure, prosperous and deligntful, where the inbabitants of the towns and districts lived bappily. Their inhabitants who were human
classes of Vidyadharas represented diverse physical types." P. 267--Omit the opinion attributed to Prof. Chatterji and insert : " Prof. S. K. Chatterji
inclines to think that Khāravela is just an Aryanised spelling of a name of Dravidian origin, and that it may be taken to mean 'be of the black lance,' deriving it from the Dravidian karu, meaning 'black' and vel, meaning *a lance,' which both occur in Tamil, and that the Kalinga people inspite of the early inscriptions in the Aryan languages seem to have been non
Aryan speakers down to a late period." P. 312-Read L. 27, if for a; L. 23, is for in and in rows for rows. P. 314-Read L. 8, limbs for imbs; L. 23, female for Säivite.

Page #24

Page #25

www. kobatirth.org
CONTENTS
xvii
PAGES Preface
vii-xvi Alternative readings and renderings Additions and Corrections ...
xviii-xx Book I-Texts and Translations
1-134 No. 1-Inscription of Khāravela
1-48 No. II-Inscription of Khāravela's chief queen ... 49-58 No. III-Inscription of Kadampa-Kudepa
59-64 No. IV-Inscription of Vaďukha-Varikha
65-70 No, V--Inscription of Vadukha-Varikha (?)
71-76 No. VI-Inscription of Cūlakamma...
77-82 No. VII-Inscription of Cūlakamma
83-88 No. VIII-Inscription of Kamma and Khīņā
89-94 No. IX-Inscription of Bhūti
95-100 No. X--Inscription of Nākiya(?)...
101-106 No. XI-Ananta-Gumphā inscription
107-112 No. XII---Ananta-Gumphā inscription
113-118 No. XIII_Tattva-Gumphā I inscription
119-124 No. XIV-Inscription of Kusuma
... 125-130 No. XV-Table of Brāhmi alphabet
131-134 Book II-Notes
135-136 1. The Inscriptions and their Authors
137 2. The relative total of the Caves and Inscriptions 137-140 3. The Symbols ...
141-144 4. The Letter-forms ...
145-153 5. Jayaswal's Five Corrected Readings
154-156 6. The Language ...
157-169 7. The Style
170-174
xxi

Page #26

xxii
CONTENTS
PAGES
! 8. The Contents
9. The Relative Chronological Position 10. The slokas concerning King Aira of Utkala ... 11. The Geographical Allusions 12. Khāravela's Personal History ... 13. Some Points concerning Personal History 14. Kharāvela's Place in History ... 15. The City of Kalinga 16. The Caves, Shrines and Pillars ... 17. Marshall on Chronology of the Caves
18. Stella Kramrisch on Reliefs in the Caves 119. Relative Chronology of the Caves
List of Inscriptional Words Index to Notes
175-180 181-182 183-187 188-231 231-265 266-267 268-287 287-291 292-303 303-307 307-314 314-316 317-321 322-324

Page #27

TEXTS AND TRANSLATIONS

Page #28

Page #29

No. I
INSCRIPTION OF KHĀRAVELA
IN THE HATHIGUMPHĀ

Page #30

ARVAN
WA

Page #31

INSCRIPTION OF KHĀRAVELA
The following inscription, noticed for the first time by Mr. Stirling, is engraved "on the overhanging brow of " the Hathi-Gumphā, better the Bada-Hathi-Gumphả, which is “a natural cavern, very little improved and enlarged by art," and, therefore, "not important from an artistic and architectural point of view." " From the remains which can still be traced in its floor," Dr. Bhagawanlal Indraji is led to think that "the cave had at one time been destroyed, whether by violence or neglect, and had afterwards been repaired and added to.".
“The inscription is carved," says Dr. Indraji, “on the rock which is not perpendicular but bends in. The inscription itself is in seventeen lines occupying a space about eighty-four feet square. The face of the rock does not appear to have been well smooth for the work, but the letters are large and deeply carved. Time and weather have wrought ravages. The first six lines are well preserved. The last four, partly so. The greater part of the intervening space has been much spoilt, portions of it being entirely weather-beaten, while in other portions single letters or groaps of letters can still be made out. The left corner of the inscription, in especial, has been greatly injured, and the initial letters of eight lines in that direction are entirely lost."
Mr. K. P. Jayaswal, who has made the reading, restoration and interpretation of the contents of this important epigraphic record his life's work, writing in 1917, says: “ The rock was roughly dressed on the righthand side. The chisel marks of the dressing are misleading; they tend to produce misreadings. These long and irregular marks left by the original dressing, are not the only pitfalls. Rain-water which trickles down the roof of the cave has cut into the letters and produced & few letter-like marks, Natural decay produced by time has given misleading turns to numerous letters......even hornets like to take liberty with the record of the Emperor Khāravela with perfect impunity and have added a few irregular marks on it....... The inscription is weather-beaten. The first four lines are completely readable. The fifth live has about 13 syllables obliterated by natural decay. Half of the record of the 6th year (L 6) and the entire record of the 7th year (L 7) have disappeared. From the Sth up. to the 15th lines, every line has got large gaps wrought by decay. The 16th and 17th lines are comparatively well-preserved except for the

Page #32

OLD BRAHMI INSCRIPTIONS
loss of about 12 initial syllables. There aro visible signs of a progressive decay."
The text is based upon the facsimile taken by Col. Mackenzie, and published by A. Stirling in 1825 without a transcript in Asiatic Researches, Vol. XV, in connection with his most valuable paper containing “ An Account, Geographical, Statistical and Historical of Orissa Proper-of Cuttack"; Major Kittoe's facsimile read and reproduced by James Prinsep, in 1837, in JASB, old series, Vol. VI, Pl. LVIII, in bis Note on Inscriptions at Udayagiri and Khandagiri in Cuttack in the Lat (Asokan) Character ; Rājendra Lala Mitra's copy of Prinsep's transcript published, in 1880, with a few minor changes in his monumental work, the Antiquities of Orissa, Vol. II, pp. 16 foll. ; Alexander Cunningham's hand-copy prepared partly from Kittoe's facsimile and partly from a photograph of Locke's plaster cast in the Indian Museum, Calcutta, and reproduced, in 1877, in Corpus Inscriptionum Indicarum, old series, Vol. I, Pl. XVIII; Bhagwanlal Indraji's eye-copy and editio princeps, publisbed, in 1885, in Actes du Sixieme International Congres des Orientalistes, Part 111, Sec. II, pp. 152-177, in his bighly suggestive paper The Häthigumpha and Three other Inscriptions in the Udayagiri Caves near Cuttack ; certain corrections proposed by Bühler, in 1895, in his Indian Studies, No. III, pp. 13 foll. and, in 1898, in his monograph - Origin of the Brāhmī Alphabet, pp. 13 foll. ; Lüders' List of Brāhmi Inscriptions, published, in 1910, in EI, Vol. X, Appendix, No. 1345, containing a summary of the contents of the inscription ; certain corrections proposed by J. F. Fleet, as to the 16th lide, in his two short notes published in JRAS, 1910, pp. 242 foll. and 824; R. D. Banerji's inked impression prepared in 1917, and reproduced in JBORS, 1917, Vol. III, Part IV, Pls. I, II and IV; a small plate published by K, P. Jayaswal in JBORS, 1918, Vol. IV, Part IV, p. 364, representing the portion of L 12 and the end of L 15; R. D. Banerji's Note on the Hathi-Gumphā Inscription of Khāravela in JBORS, 1917, Vol. III, Part IV, pp. 486-505; K. P. Jayaswal's masterly edition of the Häthi-Gumpha Inscription, published with a translation and critical notes in JBORS, 1917, Vol. III, Part IV, pp. 425-485; K. P, Jayaswal's revised readings in JBORS, 1918, Vol. IV, Part IV, pp. 364-403; K. P. Jayaswal's corrections in J.BORS, 1927, Vol. XIII, Parts III-IV, pp. 221. 246; certain readings offered by Ramaprasad Chanda, in 1919, in No. 1 of Memoirs of the Archaeological Surrey of India, in his learned paperDates of the rotire Inscriptions on the Stūpas at Sāñchi; certain readings suggested by R. C. Mazumdar in 1A, 1919, p. 190; certain corrections

Page #33

INSCRIPTION OF KHĀRAVELA
proposed by Sylvain Lévi as to the reading of L 11 in JA, 1925, pp. 57-62; certain corrections proposed by Sten Konow as to the reading and interpretation of the inscription in Acta Orientalia, Vol. I, 1923, in his paper-Some Problems raised by the Khāravela Inscription ; Locke's cast preserved, in a broken condition, in the Indian Museum, Calcutta; the plaster cast prepared by H. Panday for the Patna Museum ; and, above all, the plate published by K. P. Jayaswal in JBORS, 1927, Vol. XIII, Parts III-IV, reproducing photographs of judicio usly selected portions of two impressions prepared by Madhosarup Vats.

Page #34

Page #35

INSCRIPTION OF KHĀRAVELA
TEXT
1. Namo ar (i) hamtanam 1 [:] Namo sava-sidhānam [:]* Airena2 mahārājena mahāmeghavāhanena 2(a) Ceta 3-rāja
17
vamsa-vadhanena
9
pasatha-subha- lakhanena caturamta-(rakhana) - guna-upetena Kalimgâdhipatina siri-Khāravelena 10 [11] pamdarasavasani siri-kaḍāra 11 -sariravatā kīḍitā kumārakiḍikā 12 [ ] Tato lekha-rupa 13-ganana-vavahara- 14.vidhisava-vijâvadātena nava-vasāni 16
20
21
vadhamana 23
visāradena yovarajam va sasitam 18.] Sampuna-catuvisati -vaso (so) dān (i) 22 sesayovanâbhivijayo 24 tatiye [12] Kalimgarajavamse 26 purisa-yuge mahārājâbhisecanam 27 pāpunāti 28 [.] +
25
15
Namo arihamtanam ||Namo siḍhānam|| Namo ayariyāṇam||Ņamo uvajhāyāṇam|| Namo loe savva-sāhūṇam||
*Cf. the Jaina formula of namokkāra or nokāra, the Jaina form of invocation, met with on the first page of a Jaina book and quoted by Rajendra Lala Mitra and Bhagawan Lal Indraji :—
8
Cf. Makhadeva-Jataka (Fausböll, No. 9): Atite Videha-ratthe Mithilayam Makhādevo nāma rājā ahosi dhammiko dhammarājā. So caturásīti-vassa-sahassāni kumāra-kīļam tatha oparajjam tathā mahārajjam
katvā.
Cf. Vikrama-Prabandha, verses quoted in the Pattavalis of the Digambaras, edited and translated by Höernle in IA, Vol. XXI, p. 67 :—
Satlari cadusadajutto tiņa kāle Vikkamohavai jammo | Atha-varasa vala-līlā soḍasa-vase hi bhammie dese|} Rasapaṇa-vāsā rajjam kuṇanti micchāvādesa samjutto |

Page #36

OLD BRĀ MĪ INSCRIPTIONS
1. Prinsep and others read arahamtānam. In Banerji's impression there. appears a vowel-mark, the ā-mark, which may be taken to stand for an i-sign, yielding the reading arihamtānam which, in ardha-Māgadhi or Jaina Prakrit, is just & variant of arahamtānam.
2. Prinsep, Mitra and Indraji read Verena, Indraji wrongly suggests that Airena is not to be found in Pāli or Prakrit. See Liders' List of Brähmi Inscriptions, No. 1276– Aira Utavipabhāhi : No. 1280-Cüla Ayira, Ayira Bhuta-rakhita, Ayira-Budharakhita. See Fausboll's JĀtaka, Vol. VI, p. 300 : Ayiro hi dāsassa janinda issaro. See, also, the Old Oriya Text quoted by Jayaswal, where one finds such expressions as “ Ahiro namo rāja," " Airo jitavân bhavet," " Airah Utkalêsvarah." Banerji reads Kharena, which is out of the question. The first letter is far from being kha. Cunningham, Lüders, Jáyaswal and Sten Konow correctly read Airena.
2(a). Jayaswal finally reads māhāmegha.
3. Prinsep, Mitra, Cunningham, Lüders, Banerji and Jayaswal read Ceta. Rama. prasad Chanda justifies this reading by the evidence of the Vessantara-Jätaka (Fausboll, Vol. VI, pp. 514-528), in which the kings of the Ceta royal family are said to have reigned in Kalinga. Sten Konow reads Ceti, which he thinks is absolutely certain. He says that both Mr. Jayaswal and Dr. Thomas, who have examined the stone, agree to accept it. He maintains that the reading Ceti (=Cedi) is quite in keeping with the adjective rājisi-vamsa-kula-vinisito in L 17 whereby Khāravela claimed to have descended' from an ancient family of royal sages. Locke's cast shows, indeed, a deep cutting on the top of the letter ta, which one may take to be an i-mark. But on a closer examina. tion it appears that this cutting is either due to the flowing of water or may be simply & slit in the rock. Jayaswal finally reads Ceti. I have always been tempted to suggest such a reading as [Ka]limga-rāja-vamsa-vadhanena on these three grounds : (1) That there is the faintest trace of a letter after mahāmeghavāhanena, which is no other than ka; (2) that the letter read as ce looks also like a damaged li; and (3) that the letter read as ta or ti is, strictly speaking ga, the upper vertical line standing on the left arm of the angle rather than on its apex, as it should be, had the letter been ta. The appearance of an apper vertical line with an i-mark may just be due to some mysterious erosion around the anusvāra-mark. See Plate in JBORS, 1927, Vol. XIII, Parts III-IV.
4. Prinsep, Mitra and Cunningham wrongly read kā for rä.
5. Prinsep and Cunningham read chadhanena. Nothing is more likely than the mistaking of the Brāhmi letter va for ca or cha.
6. Prinsep and Cunningham read suke, which is meaningless.
7. Prinsep reads caturamtala-thāna; Ounningham, caturamkala-thana, Jayaswal, caturamtala-thuna, correcting his former reading caturamtala-thuna. I maintain that there is no such word as caturamtala, the usual literary expression being caturanta or caturanto, cf. the classical Pāli expression căturanto-vijitāvi. The fifth letter is clearly ra or lu. The supposed u-mark appearing below tha is nothing but a lower exten. sion of the letter kha or erosion on the inscribed surface of the rock, having no organic connection with the main letter. So I propose to read rakhana or luthana. Jayaswal finally reads caturamtaluthita.
8. Prinsep fancifully reads ganenakha te va. Jeyaswal reads gunopahitona, correcting his former reading gunopagatena, which accorded with the reading proposed by Cunning. ham and Indraji. I accept Chanda's guna-upetena, which serves as an instance, where the two words of a compound are juxtaposed without being combiped according

Page #37

INSCRIPTION OF KHĀRAVELA
to the rule of the Sandhi, Jayaswal finally abandons the reading -opahitena in favour of -opapitena.
9. Prinsep reads Kalamgadhapatira, and Cunningham Kalimgadhipati ca, both of which are incorrect.
10. Prinsep reads sikhiravalonam, and Cunningham, sikara-varabelena.
11. Indraji reads kumāra, which is evidently an oversight. Jayaswal in his revised reading, accepts Prinsep's kadāra, in preference to his former reading kadārā.
12. Indraji reads kidakā. Cunningham omits da through mistake. 13. Jayagwal's former reading was rupa.
14. Prinsep reads vavapära and Cunningham, varepāra. Jayaswal adopts vavakāra, giving up his former reading vevahāra. Locke's cast has ve, and I see no harm in reading vevahāra.
15. Canningham reads tijāvadātenn-i, which is fantastic. 16. Prinsep reads vasānā, and Cunningham, vasati, both of which are incorrect.
17. Prinsep reads hovaraja ; Cunningham, yovaraja; Indraji, hovarajam; Jayaswal, correctly yovarajam.
18. Prinsep reads pamsāsiya(sa), separating sa from Sampuna following: Canning. ham, pamsāsiyam. Indraji and Jayaswal read pasāsitam. But the first letter which is somewhat fissured seems to be va rather than pa.
19. Prinsep reads puna, separating it from sam preceding. Cunningham is the first to read sampuna.
20. Prinsep reads cava-vīsati, also visati ; Cunningham, catuvimsati. Jayaswa! reads catuvīsati, correcting his former reading catubisati.
21. Prinsep reads vase; Cunningham, vase(su). Indraji correctly vaso.
22-23. Prinsep reads dānavadhamena, which is accepted by Cunningham and Indraji. Jayaswal reads tadāni-vadhamāna, correcting his former reading ti aana cu dhamena. I am also inclined to read samāna for tadāni.
24. Indraji reads vijaya. Jayaswal in his revised reading, accepts Prinsep's yovenabhivijayo, suggesting that the e-mark in ve is the result of an abrasion. Jayaswal finally reads sesayo Venābhivijayo, taking Venäbhivijayo to mean "a conqueror like Vena, a Vedic personality."
25. Indraji reads vatiye. Others read tative. 26. Indraji reads vamsa ; Jayagwal, vamsa, correcting vamse. 27. Jayaswal finally reads mähā. 28. Cunningham reads pāpenāti. Others read päpunāti.

Page #38

10
OLD BRĀHMÌ INSCRIPTIONS
"2. Abhisitamato? cao padhame 8 vase vāta-vihata-gopura *pākāra-nivesanam patisamkhārayati Kalimga-nagari [ , ]
gabhira 6-sītala-tadāga 8-pāļiyo° ca bamdhāpayati o [] savdyāna" pațisamth[@]panam ca [13] kārayati
panatīsāhi 13 sata-sahasehi [ 1 ] 14 pakatiyo 15 ca 18 ramjayati" [.]
1. Prinsep alone reads 'mate. Others read omato.
2 Prinsep reads va ; Cunningham, cam ; Indraji, correctly ca, though oa is more idiomatio.
3. Prinsep, Cunningham and Indraji read padhama. Jayaswal correctly reads padhame,
4. Prinsep and Cunningham read to pura, which is incorrect,
5. Indraji reads na garim ; Jayaswal, nagari, correcting his former reading nagaram. Both Prinsep and Cunningham read nagari.
6. Prinsep reads Khidira ; Cunningham, khimbira ; Indraji, khibira, suggesting that the roading should be khidhīram ; Jayaswal, 'Khi-biram correcting his former reading khibira. Jayaswal finally reads Khibira, taking it to be the name of a sage. Locke's cast helps us to account for the appearance of ga as khi. It shows that rain-water trickling down from or along the u-mark in pu of sampuna in L 2, has cut into the right arm of ga, lengthening it and making the whole letter look like khi. The second letter appears, at first sight, as bi.
7. Prinsep and Indraji read sitala. Cunningham reads isitāla; Jayaswal, isitala, correcting his former reading sitala.
8. Cunningham alone reads tadiya. . Others read tadāga.
9. Prinsep alone reads pariya. Others read pädiyo. Jayaswal's former reading was pādiyo.
10. Prinsep reads bathupayasi or bathapa yasi ; Cunningham, thāpāpayati ; Indraji, badhāpayati ; Jayaswal, correctly bamdhāpayati.
11. Prinsep reads sapa yāni ; Cunningham, savo yānam ; Indraji, savuyāna ; Jayaswal correctly, sadāyāna.
12. Prinsep and Canningham read patisanthapanam; Indraji and Jayagwal, patisam. thāpanam. Jayaswal finally reads pațio.
13, Prinsep fancifully reads panatisirāsīhi; Cunningham pamnītāsidhi. Indraji is the first to correctly read panatīsāhi.
14. There is a large space between sahasehi and pakatiyo.
15. Prinsep reads pakātiyo. Cunningham and Indraji read pakatiye. Jayaswal reads pakatiyo, correcting his former reading pakātiyo. Jayaswal finally reads Pakatiya.
16. Both Prinsep and Cunningham omit ca.
17. Cunningham reads ijayata ; Indraji, rajayati. Prinsep and Jayaswal correctly road ramjayati.

Page #39

11
INSCRIPTION OF KHĀRAVELA 3. Dutiye 1 ca ? vase : acitayitā * Sātakaņimo pachima - disaip haya?-gajae-nara-radha-bahulamodamdam 10 pathāpayati11 [, ] Ka(limgâ)gatāya? ca 18 senāya 14 vitās[e]ti 15 Asaka 16-nagaram "[.]
1. Prinsep and Canningham read datiye. Indraji reads ditiye ; Jayaswal, dutiye, correcting his former reading ditiye.
2. Jayaswal accepts the reading ca of Prinsep, Cunningham and Indraji, in preference to his former reading cu.
3. Prinsep and Cunningham read väse ; others vase.
4. Cunningham reads acitayata ; Indraji, abhitayitā. Prinsep is the first to correctly read acitayitä.
5. Prinsep reads Satekāri or Sotekāri ; Cunningham, Sotakāni ; Indraji, sātakani ; Jayaswal, sātakanim, correcting his former reading sätakamnim.
6. Prinsep reads pachimā. Others read pachima. 7. Prinsep reads haya or haye ; Cunningham, iha ; Jayaswal, haya. 8. Cunningham alone reads yejam ; others read gaja. 9. Prinsep and Cunningham read babhula : Indrajj and Jayaswal, bahulam. 10. Prinsep reads darim or damdi ; Cunningham, nam te; Indraji, correctly damdam.
11. Prinsep reads pathapayati; Cunningham pathapanati ; Indraji, pathāpayati; Jayaswal, pathāpayati, correcting his former reading pathāpayati.
12. Prinsep reads Kamsabandgatāya ; Cunningham, sabänågatāya ; Indraji, Kusambā. nam khatiyam ; Jayaswal, Kañhabenām gatāya, correcting his former reading Kāsapā. nan khatiyam. Jayaswal finally reads Kanhabemni. I maintain that the reading cannot but be Kalingagatāya. There are three holes between ka and la, and one or two holes between la and ga, which are, after all, responsible for the appearance of a letter like fiha, and for the appearance of gå as be. I concede Kam(sava) ndgatāya as an alternative reading.
13-14. Prinsep reads Qasanāya; Cunningham, disenoya ; Indraji, ca sahāya ; Jayaswal, ca senāya, correcting his former reading ca sahāye.
15. Prinsep reads vātānam ;l Cunningham, vātānamta ; Indraji, (sahāyav)atā patam ; Jayaswal, vitāpati, correcting his earlier reading ditopati. Jayaswal finally reads vitasitam.
16. Prinsep and Cunningham read Saka, ignoring the first letter. Indraji reads Masika. Jayaswal finally reads M(u)sika. That the first letter is a, an not ma or mu, is certain. I am unable to decide whether the name of the place is Asaka or it is Asika. Water has trickled down in a continuous stream from or along the left arm of ti of bamdhapayati in L 3 reaching as far as the top of sa of Asaka or Asika.
17. Prinsep and Cunningham read nagara; Indraji and Jayaswal correctly, nagaram.

Page #40

OLD BRĀHMÍ INSCRIPTIONS 4. Tatiye' puna vase [14] Gamdhava -veda-budho dampa -nata *-gita 6-vādita-saņdasanāhi usava-samāja-kārāpanahi ca kīļāpayati o nagarim'[]
1. Prinsep reads pūsino ; Canningbam, navāye ; Indraji, tatiye ca ; Jayaswal correctly, tatiye.
2, Prinsep, Cunningham and Indraji read Gamdhapa. Jayaswal reads at last Gum. dhava, correcting his former reading Gandhavä.
3. Prinsep and others read dampa. 4. All read nata. 5. Prinsep and Cunningham read gita ; Indraji and Jayaswal correctly, gita..
6. Prinsep reads kidapayati ; Cunningham, kādapayamti; Indraji is the first to correctly read kidāpayati.
7. Prinsep reads nāgari ; Cunningham, nagari ; Indraji nagaram ; Jayaswal correctly, nagarim.

Page #41

INSCRIPTION OF KHARAVELA
13
5. Tathā cavuthe ’ vase : Vijādharâdhivāsaņ* A(raka) ta 5puram Kalimga ?-puva-rājāna(m) (dha)m[e]na o va nitinā va pasāsa(yati) o savata dhamakūte(na) "[, ]bh(i)ta-tasite " ca nikhita 12-chata-(15)-bhimgāre 18 hita 14-ratana 15-sāpateye 16 sava 17. rathika 18-bhojake pāde 19 vamdāpayati 20 [ - ]
1. Prinsep and Canningham read tatha. Indraji reads itha ; Jayagwal, tathā, correcting his former reading itha.
2. Prinsep and Cunningham read riputhe ; Sten Konow reads civuthe ; Indraji, cavuthe; Jayaswal, too, cavuthe, correcting his former reading ciputhe. The appearance of ca as ci may be accounted for by the fact that rain water trickling down from the space between bahulay and pachimam in L 4, has formed a canal reaching the top of ca below.
3. All road oase. 4. Prinsep reads "vāse; Canningham, vasam; Indraji correctly, 'vāsam.
5. Prinsep reads a(ra)hata. Cunningham and others read ahata. In one of the impressions three letters are prominently brought out, the middle one as ha. It seems that the two letters, ra and ka, being somehow connected at the base, appear together as one letter, which is ha. Immediately after Vijādharadhidāsam, anyhow some. where in the sentence, we must expect to get the name of a place. But if the word coming immediately after Vijādharadhivāsam be read as an adjective, as ahata-puram, we get no place-name to stand as the object of the verb pasasayati following:
6. Prinsep reads puve or puba; Cunningham, puva. Indraji and Jayaswal read puram. I think the supposed va is nothing but a fissured ra.
7. Prinsep reads Kalimga or Kalaga. Cunningham and others read Kalimga. Jayaawal finally reads Kālimga.
8. Prinsep reads rājana e satu ; Cunningham, rājān... ; Indraji, rāja.namamsitam; Jayagwal, raja-nivesitam, correcting his former reading rāja-mamamsitam.
9. The reading from dhamena to nitinä is tentative.
10 Prinsep reads "pata dhamatita ; Cunningham, oata dhamatisapà ; Indraji, dhama. kutasa ; Jayaswal, vitadha-makute. The letter na looks, indeed, like sa. Jayaswal finally reads makuta.
11. Prinsep reads (sa)ra...rite ; Cunningham, doubtfully taijatite ; Indraji, pujita ; Jayaswal, (sa)bilamdhite, correcting his former reading sabichidate.
12. Prinsep reads nikhite. Cunningham and Indraji read nikhita. Jayaswal, too, reads nikhita, correcting his former reading nikhite.
13. Prinsep reads (a)bhigare(hi); Cunningham, higāre(hi); Indraji, bhimgārehi ; Jayaswal, correctly bhimgāre.
14-15. Prinsep reads taratuna; Cunningham, taratanam ; Indraji, tiratana(sa); Jayaswal, hita-ratana, correcting his former reading hita-ratana(sa).
16. Prinsep reads sapatena; Cunningham, sāpataye ; Indraji, 'patayo; Jayagwal. sāpateye, correcting his former reading (ratana)sā pateye.
17. Prinsep and others read sava. Jayaswal, too, reads sada, correcting his former reading sava.
18. Prinsep reads rathika or rathike. Cunningham and others read rathika.
19-20. Prinsep reads (bhojake) pā devam dāpayaati : Cunningham, bhojakepā devam dasayati ; Indraji, too, devam dasayated ; Jayaswal correctly reads vam dāpayati.

Page #42

14
OLD BRAHMI INSCRIPTIONS
2
6 Pamcame1 cêdani vase Namdaraja-tivasasata 3-oghațitam Tanasuliya-vātā panāḍim" nagaram pavesayati ['] sata(saha)sehi ca (khanā) pa(yati) 1o [.]
8
1. Prinsep reads paca or pacala; Cunningham, pamcapam. Indraji and Jayaswal correctly read pamcame.
2. Prinsep, Cunningham and Indraji read dāni. Jayaswal reads dāni. One cannot be sure about the second vertical stroke in front of i-mark of ni. It appears that rain. water trickling down from the letter ca of dutiye ca in L 4, has formed a canal, which has cut through the letter ki of kiḍāpayati in L 5, and flowed down as far as and parallel to the i-mark of ni. Dani may be accepted as an alternative reading.
3. Prinsep and Cunningham read tivasasata; Indraji and Jayaswal, tivasasatam. I find no dot-stroke denoting m in front of ta.
4. Prinsep and Cunningham read ughaṭitam; Indraji and Jayaswal correctly, oghatitam. Ramaprasad Chanda rightly argues in favour of the reading tiva sasata-oghāți. tam, instead of tivasasatam oghāṭitam, when he points out (1) that there is no anusvär asign (m) either above or beside the final ta of tivasasata, and (2) that the absence of a suffix after tivasasata is due to the fact that it forms part of a compound word Namdaraja-tivasasata-oghäṭitam, an instance like guna-upetena, where the two words, sata and ogha are not combined according to the rule of the Sandhi. To these arguments of Chanda, I may add (1) that unless Namdaraja-ti vasasata be treated as part of a compound word, it becomes unmeaning, and (2) that Namdaraja-tivasasatam has no meaning at all as a separate word in the present context. Sten Konow, too, reads Namdarāja-tivasasata-oghāțitam.
5. Prinsep reads tamnisaraliya or tanasaraliya; Cunningham, tanamsuliyam; Indraji Tanasuliya; Jayaswal correctly, Tanasuliya.
6. Prinsep reads vaja; Cunningham, văța. Indraji and Jayaswal correctly read vāţă. 7. Prinsep and Cunningham omit m.
8. Prinsep and Cunningham omit m.
9. Prinsep reads pavesa rise; Cunningham, pavesa viso. Jayaswal reads the last letter
as ți.
10. Jayaswal reads the concluding words as so pi vase chadam, completing his former reading so?... vi., and treats them as the opening words of the record of Kharavela's 6th regnal year. I think there can be little doubt about the correctness of the reading proposed above to complete the statement regarding the 5th year.

Page #43

INSCRIPTION OF KHĀRAVELA
15 7. Abhisito ca [chadhe]? vas[e] ° (rāja-sirim) saņdamsayamto 5 sava-kara-vaņa - [16]-anugaha-anekāni? sata-sahasāni visajati pora P-jānapadam [.]
1. Prinsep reads sabhāsari; Cunningham sabhisori ; Jayaswal correctly, abhisito.
2. The gap containing space at least for three letters has altogether been lost sight of by Jayaswal.
3. The letter just beyond the gap appears to be sa.
4. Prinsep reads pāsaca; Cunningham, hāsa ca; Indraji, rājaseyam ; Jayaswal, rājasuyam, correcting his former reading rājaseya. I maintain that the word is neither rājaseyam nor rājasuyam ; it is rājasirim.
5. Prinsep reads sadasa te; Cunningham, samdasamto; Indraji, samdamsanato; Jayagwal, samdamsayamto, correcting his former reading samdemsanato.
6. Prinsep reads pava karavana; Cunningham, sa vakara vana; Indraji, save-karāvanam, which is accepted by R. D, Banerji ; Jayaswal, sava-kara-vanam, correcting his former reading sava-bharāvana[m]. I was inclined at first to read savabharaṇāni. But it is certain that the letter after ra is va.
7. Pringep alone reads anekani. Others read anekāni. 8. Prinsep alone reads sahasani. Others read sahasāni. 9. Jayaswal alone reads poram. Others read pora. I do not find any anusvāra-sign.

Page #44

16
OLD BRĀHMI INSCRIPTIONS
8. Satame' ca vas(e)? (a)si-chata * -dhaja-ra(dha)-rakhitu(ramga) *-sata *-gbațāni savata? saņdasanam & sava!-(mamgalāni) to kārāpayati" [–] sata-sahasehi [.]
1. Prinsep and others read satamam. But the e.mark attached to the final letter of satame is distinct.
2. Prinsep reads tisam ; Cunningham, vesam. I do not find the anusvāra-sign along with the second letter.
3. Prinsep and others read Pasāsato. It is certain that the first letter which is somewhat obscured by fissure is not pa : it is either ma or a, more likely the latter. The third letter is prominently cha.
4. Prinsep reads vajaragharavedham ; Cunningham, vajarighavadhasatima ; Jayaswal, vajira-gharavi-Dhi(si), correcting his former reading jara-ghara khata. Jayaswal finally reads Vajira-ghara-va[m]tigh[u]sitā.
5. Prinsep reads satam; Jayagwal, (Dhi)si ti. The appearance of ra as si may be due to fissure. To me, the reading (Dhi)si is meaningless.
6. Prinsep and Cunningham read gharini. Jayaswal reads gharini, correcting his former reading gharini. None may doubt that the second letter is ta, and not ri.
7. Indraji reads savotu ; Jayaswal, samātu(ka), correcting his former reading savitu. Prinsep and Cunningham correctly read savata.
8. Prinsep reads kahadapana or kahadapamna; Cunningham, kahadapana; Indraji, kula ...... ; Jayaswal, (mâtu)ka-padapuna, correcting his former reading upādapumna.
9. Priusep and Cunningham read nara(pa). Jayaswal reads saku(mārā). Jayaswal finally reads ( ti ?] [ kumāra).
10. Cunningham reads pa ketana. Jayaswal reads the first letter as ma. 11. I have little doubt about the correctness of this reading. 12. The gap must have contained a word denoting a number.

Page #45

INSCRIPTION OF KHĀRAVELA
17
9. Athame' ca’ vase ma(ha)ti-(se)nā(ya)l'a(dhu)ra[m]* a(nupa)to Goradhagirim [17] ghātāpayitā Rājagahānam papid[W]payati? 1,1 (e)tinam ca kamma-padāna '-panādena.losambhīta 11-senavāhane 12 vipamumcitum 18 Madhuram 14 apayāto 15 yeva naridole stories lifgha]ravās[i]nam ca [-] 18 sava-gahapatikānam ca [-]' sa [-] 20 pāna-bhojanam dadāti21 [:] [Kalim]gam (yā)tia palavabhāraos- [1 8]-Kaparukha24-hayagaja-(nara)25-radha%-saha yātia [,] sava28-gharavāsina ma ca sava-rāja-bhatakānam 30 ca sava-gahapatikānamica (sava). Bamhaņānam ca pāna-bhojanam dadāti [,] Arahatānam (samaņānam ca) [—]34 dadāti [–] (sata-sahasehi) 85 [.]
M
1. Prinsep reads thame, omitting the first letter : Cunningham, ye thame. Indraji and Jayaswal correctly read athame.
2. Prinsep omits ca.
3. Prinsep reads manamna-n; Cunningham, munamti mena-va; Jayaswal, mahati. senāya, correcting his former reading manatino dhamani. Jayaswal finally reads mahatā.
4, 5. Cunningham reads...tapabhate; Jayagwal, mahata-bhitti, which is nowhere. The occurrence of such a conjoint consonant as tti in the orthography of Khāravela's inscription is yet a discovery to be made. Mr. Jayaswal seems to have thought of such a reading under the influence of a latent bias that Goradhagiri was an impregnable hill. fortress guarding an entrance to Rajagaha.
6. Prinsep reads de vegiri. Kittoe's facsimile has gedaragiri. Cunningham reads dare sari idha; Jayaswal, Goradhagirin, correcting his former reading Goradhagiri.
7. Prinsep reads rājagabham (or, rājagambha) upapīdapayati; Cunningham, rajagambhu upapīdapayati; Indraji, Rajagaha-napam pīdāpayati; Jayaswal, Rajagaham upapi. dāpayati, correcting his former reading Rājagaha.napam pidāpayati. Sten Konow finds no objection to reading nipa instead of upa. I find that the horizontal base stroke of na is really in line with that of pa of pidäpayati, and that it stands out, at least in Banerji's impression, as the horizontal base line of na. Rājag ahānam in the sense of the people of Rājagaha is an expression of the same kind as Magadhanam signifying the people of Magadha. This reading can well account for the use of (e)tinam following.
8. Prinsep and Canningham read dhatinam ; Indraji reads etinam; Jayaswal, etinā, correcting his former reading etinam. One may read etinä, but etinam gives a better grammatical construction of the sentence. Sten Konow, too, reads etinā. Jayaswal finally reads etinam or etinā.
9. Pringep reads kammupadana; Indraji, kamapadána: Jayagwal, kamma'padana =(kamma + apadana), correcting his former reading kamupadāna. Cunningham correctly reads kamma-padāna. Jayaswal finally reads kammāpadāna.
10. Prinsep reads panādena or panādenā. Others read panādena. Jayagwal finally reads samnådena.
11. Prinsep reads pambata ; Indraji, savata ; Jayaswal, sambīta, correcting his former reading sabata. The second letter is undoubtedly bhi.
12. Prinsep reads sena-vāhayati) or sena-vāhanā(ti); Indraji and Jayaswal correctly read sena-vahane; Jayaswal finally reads "vahano.
13. Prinsep reads ti pamucitu or "ti panimucita; Cunningham, pamacitu ; Indraji, vipamucitu; Jayaswal, vipamumcitum, correcting his former reading vipamumcitam.

Page #46

18
OLD BRAHMI INSCRIPTIONS
14. Prinsep, Cunningham and Indraji read Madhuram. Jayuswul reads Madhuram, correcting his former reading Madhuram. Sten Konow, too, reads Madhuram. It is certain that the reading is Madhuram, and not Madhura.
15. Prinsep reads aparato or apanato; Cunningham, apayato; Indraji and Jayaswal correctly read apayāto. Sten Konow reads upayati, which is doubtful. The o-mark attached to the final letter is very distinct.
16. Cunningham reads navam ranā ba; Indraji, navame ca vase; Jayaswal, yeda narido, correcting his former reading navame ca vase. Here Sten Konow would read Yavana-rājā which is ingenious but uncertain. The fifth letter is distinctly da and not ja. Jayaswal finally reads Yavana-räja..
17. Jayaswal reads nāma; Sten Konow, fancifully Dima(ta). Jayagwal finally reads Dinsite or Dimiti. I cannot offer a better reading than sapa-gho
18. The gap may perhaps be filled in by rāja-bhatakānam ca. 19. The gap may be filled in by Bamhaņānam ca Arahatānam. 20. The gap may be filled in by samaņānam ca.
-21. Prinsep and Cunningham read mora dadāti; Jayaswal reads mo yacati, correct. ing his former reading mo. ya chati.
22. These five letters appear to be above palavabhāra, with which the line closes. Jayaswal finally reads (yachati-vi) palava...
23. Cunningham reads palavama ke; Indraji, pavarako; Jayaswal, palavabhāre, correcting his former reading palavariko.
24. Prinsep and Cunningham read kapam ukha; Indraji reads Kaparukho; Jayaswal, Kalparukhe, correcting his former reading Kaparukho. Jayaswal finally reads Kaparukhe.
25. Prinsep traces three letters after gaja, which he doubtfully reads raluve. Cunningham suggests two letters. Indraji and Jayaswal read radhu immediately after gaia, which seems to be skipping over a fissured short space containing room for two letters, which I propose to read nara.
26. Prinsep reads raluve ; Indraji and Jayaswal correctly read radha.
27. Prinsep and Cunningham read sahāya se; Indraji reads saha-yata; Jayaswal, saha-yamte, correctiug his former reading saha-yate.
28. Prinsep and Cunningham read saca; Indraji reads savam ; Jayaswal, correctly sava.
29. Prinsep reads gharā-vasapa; Cunningham, gharavasaya; Indraji, gharavasadham; Jayaswal, gharadāsa-pa, correcting his former reading gharadasadhan.
30. Cunningham reads anatika-gava yasu(va); Jayagwal, sa-aginathiya, correcting his former reading aginațhiye.
31. Prinsep reads sahanam ca kā(ra); Cunningham, gahanam ca kā(ra); Indraji, (yasavā) gahanam ca kā(ra); Jayaswal, gahanam ca ka(ra).
32. Prinsep reads (kāra)yitum ba imanānam; Cunningham (kara) yitum ba imana. nam; Indraji, (kara)yitum Būmanam; Jayaswal, (kāra)yitum Bamhanānam.
33. Prinsep reads (jatapa para ; Cunningham, (ja) puha sara; Indraji (ta)mhi radhi-säram ; Jayaswal (ja)ti-pamtim parīhāram, correcting his former reading jatharam bhi param. Jayaswal finally reads jātim pariharani.
34. The gap may be filled in by pana-bhojanam.
35. The record of the 8th year may be taken to close with an expression, which was preceded by another denoting a numerical adjective.

Page #47

İNSCRIPTION OF KHĀRAVELA · 10. Na(vame)' ca vase? [/ 9] veduriya: Kali[m]ga-* rāja-nivåsam. Mahāvijaya-pāsādam karayatio atħa-tisgya? satasahasehi [.]
1. Jayaswal reads only the first letter as na. 2. Jayaswal reads giya.
3. Prinsep reads ra-i; Cunningham, venati; Jayaswal, okei, correcting his former road. ing abhaya.
4. Prinsep reads manati or manati ; Cunningham, manati ; Jayagwal, mānehi, correcting his former reading prāci-tațe. Jayaswal finally reads manasti]. The third letter, as it appears in Banerji's estampage, must be read as ga.
5. Prinsep reads pamdarasa or raini rasa ; Cunningham, pamdarasa ; Indraji, nivāsam; Jayaswal, sapnivasam, correcting his former reading nivāsam, which was really correct.
6. Prinsep reads karayati or derayati. Others read kārayati.
7. Prinsep reads thatasaya, omitting the initial letter ; Cunningham, atha hita du; Indraji, atha-tisa ; Jayaswal correctly, atha-tisāya.
8. Prinsep reads sata sarelahi; Cunningham, (du)savasahasehi. Indraji is the first to correctly read sata-sahasehi.

Page #48

20
OLD BRĂHMI INSCRIPTIONS 11. Dasame' ca’ vases (Kalimga-rāja-vamsānam tati(ya). y[u]ga-sagâvasāne (Kallimga* -pu(varājānam ya)sa-sakāram kārā pagati" sata-(sahasehi). [.]
Prinsep and Cunningham read dasāme; Indraji and Jayagwal correctly dasame. 2. Prinsep omits it. 3. Prinsep reads case ; Cunningham, tuse; Indraji and Jayaswal correctly read vase.
4. From Kalinga to Kalimga Prinsep reads dava...rara gavasapa; Cunningham, datibhisara... karatha vasa p-na; Indraji.... Bharadhavasa-pathana; Jayaswal, mahadhi. ta'bhisamayo Bharadhavasa-pathanam, correcting his former reading damdasa......nadasa Bharadhavasa- pathāna. Jayaswal finally reads damda-samdhi-säma-mayoo.
5. Prinsep reads pabayava ; Cunningham, mahajava ; Jayaswal, mahi-jayanam. 6. The reading suggested is tentative. 7. Cunningham reads rācābīyati; Indraji and Jayaswal correctly read kārāpayati.
8. Both Indraji and Jayaswal take the sentence to close with kārāpayati, which seems incorreot.

Page #49

21
INSCRIPTION OF KHÁRAVELA 12. (Ekā)dasame ca vase! [---]-maņi-ratanānio-saha yāti* [,] [1 10] [- ] (Kalim)ga -puvarāja? -nivesita. Pithudaga-dabham Nagalo nekāsayati [,1 (a)nupa-dabhavanam ca terasa-vasa-sata-katam bhidatil Timira-dahasamghātam [.]
1. The opening words of the record of the 11th regnal year are not effaced. Even the first three letters, e, kā and da, can be read, though, to some extent, obscured by fissure.
2. Prinsep reads the letters to be made out in the gap as raharunatanara ; Indraji, uyatānam ca ; Jayaswal, niritaya uyātānam, correcting his former reading ... puna ca. I shall not be astonished if the letters were meant to inscribe Kaparukha-haya-gaja-nararadha.
3. Prinsep reads manetānānā ; Cunningham, soti yo ru-ni; Indraji, manoradhāni; Jayaswal, mani-ratanāni, correcting his former reading manoradhāni. Here I have tentatively accepted Jayaswal's reading. Jayaswal finally reads manio. :
4. For saha yāti Prinsep reads upahi ; Cunningham, upalebhāta; Indraji, upalabhatā ; Jayaswal, upalabhată, correcting his former reading upalabhata.
5. Indraji points out that "the beginning piece of lines 11-17 is lost, and about ten lettere are lost in each."
6. Jayagwal reads mamde ca. He finally reads mamdam. I maintain that the reading is no other than Kalimga.
7. Cunningham reads puve rāja ; Indraji and Jayaswal correctly read puvarāja. 8. Cunningham reads nivesātam ; Indraji and Jayaswal read nivesitam,
9. Prinsep reads pīthu dāga dambha nagare nakasayatta; Cunningham, pithudaga dambha nagalo nakasamyata ; Indraji, pāthudam Gadambha nagale nakāsayati; Jayaswal, Pithudaga-dalbha-nagale nekasayati, correcting his former reading Pithudaga-dabha-nagale nekāsayati. Sylvain Lévi and Sten Konow read pithuda[m] gadabha-na[ m galena käsayati. The reading kept in view by Lüders is pithudam gadabha.namgalena käsayati ("Pithuda, ploughed with an ass plough). Fleet takes pathudaga to be=pämthuddamga, "& market-town for the convenience of travellers," It is certain that there is no anusvāra along with da. Jayaswal finally reads pithuờa-gadabha-namgalena käsayati.
10. Prinsep reads janapade bhāvana ; Canningham, janapuda bhavana ; Indraji, Jayagwal, R. C. Mazumdar and Sten Konow read janapada-bhāvanam. The reading kept in view by Lüders is Jinapāda-bhāvanam ("the meditation on the feet of Jina"). Jayagwal finally reads j(i)nasa dambhāvanam. Read janasao as a variant.
11. Prinsep reads terasa vase satake... ; Cunningham, terasa vase satake bhidasi; Indraji, terasa vasa sata ka...da; Jayagwal, terasa-vasa-sata-Ketubhada(ti); R. C. Mazumdar, terasa-Khasa-satam katabhadata ; Sten Konow, terasa-vasasata-kata bh Cildati. Jayagwal finally reads satikam tu bhidati.
12. Prinsep reads tatemara-dehasapāta ; Cunningham, tamaradehasapāta; Indraji, tamara-dehasamghatam ; Jayaswal, titamara-dehasamghātam, correcting his former read. ing titamara, etc.; Sten Konow, t[a]maradeh[]sa[m]ghātam. Mazumdar reads dehasamghātam.

Page #50

22
OLD BRAHMİ INSCRIPTIONS 13. Bārasamei ca ? vase : [-]* sata.sabasehi vitāsayati 6 Utarā pa (dha)-rājānos [, ] [111] [- ]" Māgadhānam ca vipulam 1 bhayam2 janeto 18 hathi. sam Gamgāya 14 pāyayati 15 [ ] Magadham ca rājānam"? Bahasatimitam pāde vaņdāpayati [, ] Namdarājan(i)tamo Kālimga 2-jinâsanama (Amga-Magad hato Kalim. gam āneti haya-gaja)28-sonavāhana-sahasehi 24 [ , ] AmgaMagadha25-vāsinam28 ca27 pāde 28 (vaņdāpayati) 29 [ , ] [112] [- ]30 vithi-catara-(pa)likhāni gopurāni 81 siharāni" nivesayati 88 [ ] sata-Vāsuki (rata)nam p(e)sayam[ti]34 [ ] abhutamachariyam 35 ca hathi(sa)-pasavam 38 paribaramti 37 [ ] (miga)-haya-hathi 38 u(p)anāmayati 39 ( , ) Paņqa-rājā +0 vi(vidhyabharaṇāni" mutā-maņi-ratanāni 43 āharāpayati 48 idha sata-sahasāļ ni](,). [1 13] [- ]95—sino vasi k(ā)reti [.]
Cf. Mahābhārata, II. 30. 27-29 :
Sa sarvān mleccha-nȚpatin sāgardnūpavāsinaḥ 1 Karamāharayāmāsa ratnāni vividhāni call Candandguri-vastrāni mani-mauktika-kambalami Kāñcanam rajatancaiva vidrumañca mahādhanam # Te kotibata-samkhyena Kaunteyam mahatā tadā 1
Abhyavarşan mahātmānam dhanavarşeņa Pāndavam Cf. Mahābbārata, II. 31, 71-72, 75-76:
Pārdyāmsca Dravidām-scaiva Sahitāmácôndra-Keralaih 1 Andhāmstālavanāmseaiva Kalingānuştrakarnikān# Ātaviñca purīm ramyām Yavanānam puram tathā 1 Dūtair eta vaše cakre karañcinānadāpayat u Tatah sampregayāmāsa ratnāni vividhāni oa Candandguru-kāşthāni divyányábharaṇāni call Vasāmsi ca maharhāni manīmscaiva mahādhanan Nyavarttata tato dhimān Sahadevah pratāpavān ||
1. Prinsep reads Barasama, Cunningham, Barasama; Indraji, Barasamam ; Jayaswal correctly Bārasame.
2. Prinsep roads va ; Cunningham, da; Indraji and Jayaswal correctly read ca. 3. Indraji roads vasam ; Jayaswal correctly vase.
4. The letters are so mach effaced that it is impossible to attempt to make out what they are.

Page #51

www. kobatirth.org
INSCRIPTION OF KHĀRAVELA
23
5. Here I have relied apon Banerji's impression reproduced in JBORS, 1917. For sehi of sahasshi Cunningham reads phahi. Jayaswal finally reads hasa.-keja savasehi.
6. Prinsep reads vitisāyato ; Cunningham, vitisiyitā ; Indraji, pitisāyamto; Jayaswal, vitās ayati, corracting his former reading vitāsayamto. Jayaswal finally reads vitāsayato.
7. Prinsep reads Utiripithi ; Cunningham, Utarapatha. Indraji and Jayaswal read Utarāpatha, which I accept as an alternative reading.
8. Prinsep and others read rājāno. But there cannot be any serious objection to reading it as rājānam, the singular Accusative form.
9. The beginning piece of L 12 is lost for ever. It is likely that such an expression as Amgånarn ca or simply Amga-preceded Magadhānam.
10. Prinsep reads macalāva ; Cunningham, madhānam ; Indraji and Jayaswal read, perhaps correctly, Magadhânam.
11. Prinsep reads vipula ; Cunningham, vipula(ya); Indraji and Jayaswal read vipulam. 12. Prinsep reads leyam. Cunningham is the first to correctly read bhayam. 13. Prinsep and others road janeto. The o-mark seems to be quite clear.
14. Prinsep reeds hithasam gamgasa ; Cunningham, hathasam Gamgāya; Indraji, hathisa Gamgayam ; Jayaswal, hathisu Gamgāya, correcting his former reading h athisu Gamgiya ; Sten Konow, Sugamgāya, suggesting that Sugamgā or Sugamgi is certainly the ancient Maurya palace in Pāțaliputra which is called Sugamgā in the Mudrārākşasa. Jayaswal finally reads hathi Sugamgiya(m). Sugamgaya does not suit the context of the Hathi-Gumpha inscription of Kharavela.
15. Prinsep reads pānayati. Cunningham and Indraji read pāyayati. Jayaswal, too, reads pāyayati, correcting his former reading prāyayati.
16. Prinsep reads ..dha; Cunningham,...ma ; Indraji, Magadham ; Jayaswal, Māgadham, correcting his former reading Magadhā.
17. Prinsep reads rājana. Cunningham is the first to correctly read rājānam.
18. Prinsep reads ibahaga sāsita ; Cunningham, bahasati sita ; Indraji, bahu pațisasita; Jayaswal, Bahasatimitam, correcting his former reading Bahapatimitram.
19. Prinsep reads va dapayata ; Cunningham, pā deva dāpam yati ; Indraji pāde vadāpayati ; Jayaswal päde vamdäpayati.
20. Prinsep reads rāja nita ; Canningham, Namdarājānita ; Indraji, Namdaraja-nita ; Jayaswal, Namdarāja-nitam, correcting his former reading Namdarāja-nitāni.
21. Prinsep rende va a ga ; Cunningham, vāmaga ; Indraji, sa aga ; Jayagwal, Kalimga, correcting his former reading ni Aga.
22. Prinsep reads jinasana ; Canningham, jinasa ; Indraji, Jinasa ; Jayaswal, Jinasamnivasam, which has no meaning. Jayaswal finally reads Jinam.
23. The bracketted reading is tentative. For āneti Canningham seems to read ma ata.
24. For senavāhana-sahasehi Prinsep reads makhana pamda pakhasi ; Canningham, rota na suđiha ; Indraji, gaha ratana-parihārehi ; Jayaswal, gaha-ratanāna-parihārehi, which is an unintelligible jargon.
25. Prinsep reads åe mugadha ; Canningham, mariga Magadha ; Indraji, a Magadhe ; Jayaswal, Amga-Māgadha, correcting his former reading Amga-Magadha.
26. Prinsep reads cajava ; Cunningham, vasasa ; Indraji, vasivu ; Jayaswal, vasum ca. 27-29. For ca pāde vamdāpayati Prinsep reads naghari ; Indraji, nayari ; Jayaswal, neyati. 30, The beginning piece of L 13 is lost.

Page #52

OLD BRAHMĪ INSCRIPTIONS
31. Prinsep reads tajalarala khila barāni; Cunningham, tu jīva ralakhilaye Bāräna(si); Indraji, vijādharu lekhilam barāni; Jayaswal, ta jathara-likhila-barāni, correcting his former reading ta jathara-lekhilāni barāni. Jayaswal finally reads tum for ta.
32. Prinsep reads sihirani ; Cunningham, si hiranā(ni); Indraji, siharāni; Jayaswal, sihirāni, correcting his former reading siharāni. I accept Indraji's reading.
33. Prinsep reads nivaneyati; Cunningham, ni ca iyati; Indraji, nivesayati ; Jayaswal, nivesayati, correcting his former reading nivesayati, which was really correct.
34. Prinsep reads sata vasa sana pariharenam asum ; Cunningham, sata vasadana thari harenam asitam; Indraji, sata-vasu-dāna.pariharena abhutam ; Jayaswal, sata-visikanam parihärena abhutam, correcting his former reading sata-vasu dana, etc. Jayaswal finally reads vesikanam.
35. Prinsep reads (m)asariya ; Cunningham, (m)asāriya ; Indraji, (m)akariyam ; Jayagwal, (nu)achariyam.
36. Prinsep reads hathi-navuna ; Canningham, hathi.naveno ; Indraji, hathi nādana ; Jayaswal, hathi-nāvana.
37. Prinsep reads paripara ara ; Cunningham, pariha; Jayaswal, paripuram upa. Cunningham's reading is nearer the mark,
38. Prinsep reads ranasa yahava; Jayaswal, denha haya-hathi. Jayaswal finally reads savadena for upadenha.
39. Cunningham reads ...na-piva ; Jayaswal, ratanā-manikam, correcting his former reading ratana-janāna. Ratana-mānikam as an adjective of Pamdarājà may be Indian epigraphy but no Indian language.
40. Prinsep reads Padarāja (no); Canningham, maha ri raji (ne); Jayaswal Pamdarāja.
41. Prinsep reads dāvi aneka na ; Cunningham, anekāni; Jayaswal, edāni anekāni. Jayaswal finally reads cedäni for edani.
42. Prinsep reads dato manorata rana; Cunningham, dato mani ratanāni, adding muta to his former reading.
43. Prinsep, Cunningham and Jayagwal read aharāpayati ; Indraji reads āharāpayati.
44. Prinsep reads idha satasa ; Cunningham, idha samte ribha ; Indraji, idham satasa; Jayaswal, idha satasa. Jayaswal finally reads sato for satasa.
45. The beginning piece of L 13 is missing.
46. Prinsep reads sinevasi kadati ; Cunningham, si novasikariti; Indraji, sino oasi. karoti ; Jayaswal, sino vasi karoti, correcting his former reading (vā] sino vasi kareti.

Page #53

INSCRIPTION OF KHARAVELA
. 25 14. Terasame cat vase supavatavijaya’-cakes Kumārīpavate* Arahato* parinivasato bi kāya -nisīdīyāya? (rāja). bhatakehio rājabhātihi' rāja-ñātibio rāja-p(u)tebi" rājamahisihi Khāravela.sirinā satadasa-leņa-satam kārāpitam15 [.] [1 14]
*Cf. Inscription of Udyota keśari iu Lālatêndu-Kesari's cave in Khaņdagiri edited by R. D. Banerji in EI, Vol. XIII, p. 166 :-ŚrīUdyota kesari-vijaya-rājya-samvat 5 sri-Kumāra parvrata-sthāne jirnna vāpi jirnna īsana Udyotita (.] Tasmina thāne catur-vimsati Tirthamkara sthāpita.
1. Prinsep and Cunningham read Terasama va. Indraji omits ca. Jayagwal correctly reads Terasame ca.
2. Prinsep and Canningham read vijaya ; Indraji and Jayagwal, vijayi. Jayaswal finally reads vijayo.
3. Prinsep and Indraji read cako. Jayaswal correctly reads cake, Jayaswal finally reads caka.
4. Prinsep reads arahite ; Jayaswal, too, arahite, correcting his former reading arahato. Cunningham and Indraji correctly read arahato.
5. Prinsep reads papnavisata pi; Cunningham, punapasata hi, also, parinavasamta. Indraji, upanivāse va ; Jayaswal, (ya)pa kh[i]ma-oyasamtāhi correcting his former reading parin[i]väse tähi. Jayaswal finally reads yapa-khina-sapsitehi.
6. Prinsep reads kamra ; Cunningham, kaya ; Indraji, kāyam ; Jayaswal, kāyya, correct. ing bis former reading kāya, which was really correct. Jayaswal finally reads kāya.
7. Prinsep reads nisidināya ; Cunningham, nisidinaya ; Indraji, nisidiyāya ; Jayaswal, nisidiyāya, correcting his former reading nisīdīyāya.
8. Prinsep reads yāpuhavakehi ; Cunningham, yāpujakehi; Indraji, yapajake; Jayaswal, yāpañävakehi, correcting his former reading yäpūjavakehi.
9. Prinsep reads ra. atam ; Cunningham, ra.ladatini ; Jayaswal, rāja-bhitini, correcting his former reading rāja-bhitāni.
10. Prinsep roads cenan devani ; Cunningham, cenam dareni ; Jayaswal, cinavatāni.
11. Prinsep reads sasagutani ; Cunningham nase sitāni, Jayaswal, vosāsitāni, correcting his former reading vusa-satāni. Jayaswal finally reads vasäsitāni.
12. Prinsep roads ujana utasä ; Cunningham, ujani kata uvāsā ; Jayaswal, pūjāni katauvāsā, correcting his former rending pujani sa ba ta. Jayaswal finally reads pujāya. rata-uvāsa.
13. Prinsep reads yarava ladirana ; Cunningham, yeravaladirana ; Jayaswal correctly, Khāravela-sirina,
14. Prinsep reads ji deta dakararikhiti ; Cunningham jivima-kapuri khita; Indraji, ...kāle ; Jayagwal, Jivadeva-siri-kalpam. Jayaswal finally reads jāpa-deha-sirikā.
16. Jayaswal takes the line to close with rakhita. He finally reads parirakhita.
मा. श्री कैलासनगारि सानमार श्री महावीर केन मागवना केन्द्र a. (mind) R. *6*

Page #54

26
OLD BRAHMI INSCRIPTIONS
15.
1 sakata2-samaņa-suvihitānam
ca
sata
disānam (ya) tinam tāpasa- (i)sinam lenam' kārayati [,] Arahata-nisidiya1-samipe pabhare" varâkara1-samuthāpitāhi13 aneka-yojanâbitāhi1 panatisāhi1-sata1-sahasahi" silāhi18 silāthambhani1 ca cetiyani20 ca kīrāpa (yati)" [,] [115] [-]2 patalika28-catare24 ca25 veḍuriya-gabhe thambhe patiṭhāpayati 29 panatariya-sata-sahase(hi) [,] (vedu)riya-nila. vochimnam ceca-yathi-adhasatikam t(i)riyam upădayati[.]
1. The beginning piece of L 15 is missing. Jayaswal traces the remnant of a letter before sakata which he reads as ta. There is, indeed, such a remnant, but precisely of what letter I cannot say. One thing is certain that the sentence embodying the record of the 13th regnal year and running up to the end of I. 14 is in Passive Voice, while the statement filling up L 15 and the first half of L 16 is in Active Voice. This cannot be consistently accounted for unless one assumes (1) that the record of the 13th year end in L 14, and (2) that the missing words in the beginning piece of L 15 constituted the introductory phrase, such as Cudasame ca vase, indicating the particular regnal year to which the record in question refers.
2. Prinsep and Cunningham read sakata. Indraji reads the last two letters as kata; Jayaswal, (tā)su katam, correcting his former reading sukata, which was nearer the mark. Jayaswal finally reads sukati.
3. Prinsep reads same rasavihitinam; Cunningham, samelasa vihitenam; Indraji, samayo suvihitāram; Jayaswal, samana-suvihitanam, correcting his former reading samana.suvihitānam.
4. Prinsep reads suta disanam; Cunningham, suta dasinam; Indraji sava-disānam; Jayaswal correctly, sata-disānam.
5. Prinsep reads jñāta ; Cunningham, tanāpesi; Indraji, yaninam; Jayaswal, ñātānam, correcting his former reading khatiyam. I accept Jayaswal's as an alternative reading. Jayaswal finally reads ñaninam.
6. Prinsep reads yevai-sampa; Cunningham, mapusa isa pu; Indraji, tāpasānam samha; Jayaswal, tapasa-isinam samghä.
7. Prinsep and Cunningham read panam; Indraji reads tanam; Jayaswal, yanam. Jayaswal finally reads tapasi-samghiyanam.
8. Prinsep and others have missed karapayati.
9. Prinsep reads arahasa; Cunningham, cahasa; Indraji is the first to correctly read
arahata.
O
10. Prinsep reads nisidiya; Cunningham, nisidaya; Indraji, nisidiya; Jayaswal nisidiyā.
11. Prinsep and Cunningham read subhare; Indraji and Jayaswal, pabhāre.
12. Prinsep reads vasara; Cunningham, va, bha; Indraji, varakaru; Jayaswal correct. ly, varakara,

Page #55

INSCRIPTION OF KHĀRAVELA
13. Prinsep reada samathaghisipa; Cunningham, sa matha ghisipā; Indraji, samatha. thapatihi; Jayaswal, samuthäpitähi.
14. Prinsep reads anakayā janahi pihi; Cunningham, anake yojanā piti ghi; Indraji, aneka-yojan ähi; Jayagwal correctly, aneka-yojanähitâhi.
15. Prinsep reads the five letters of which there are traces as pa .. ra; Cunningham, as pa.. pipe. Indraji and Jayaswal wisely leave a blank space. Jayaswal finally reads pa, sio.
16, 17. There can hardly be any doubt about the given reading. 18. Prinsep reads silaha; Jayaswal, silāhi, correcting his former reading silahi.
19, 20, 21. Prinsep reads sa papatha dhara si dhasayani nāni; Cunningham, bhagapatha... dhadayana; Jayaswal, Simhapatha-Rañiya Dhusiya nisayāni, correcting his former reading sappatha... Jayaswal finally reads räni Sidhudāya.
22, The beginning piece of L 16 is missing.
23. Prinsep reads patalake; Jayaswal, pataliko. Cunningham and Indraji read pațălake. Jayaswal finally reads ghamtālakto.
24. Prinsep reads catapa; Cunningham, catara; Indraji, Cetake; Jayaswal correctly, catare.
25. Prinsep and Cunningham read ce; others, ca.
26. Prinsep reads veruriya ; Cunningham, teghariya. Indraji is the first to correctly read veduriya.
27. Pringop and Cunningham read gabha, Indraji and Jayaswal correctly, gabhe. 28. Prinsep and Indraji read thabhe; Cunningham and Jayagwal, thambhe.
29. Prinsep reads patipa-yati; Cunningham, patitha payati. Indraji and Jayaswal correctly read patithāpayati.
30. Prinsep reads panatanusa ; Cunningham and Indraji read pamnatariya. Jayaswal reads panatariya, correcting his former reading panamatariya. In Sten Konow's opinion panatariya is the correct reading.
31. Prinsep reads saca. . raja ; Cunningham, sa ca vasa; Indraji, sathi-vasa-sate rāja ; Sten Konow, sacasahite, partly accepting and partly modifying Fleet's reading pamnatariya sacasata ; Jayaswal, sata-sahasehi, correcting his former reading sathi-vasa-sata Räja. Jayaswal has, I believe, placed the reading beyond all dispute.
32. Prinsep reads -riya; Cunningham, ...ya; Indraji, Muriya ; Jayaswal, too, Muriya. Banerji also offers the reading Muriya. Sten Konow thinks that the reading Muriya is certain. Where is the certainty of such a reading ? None need be misled by Banerji's impression. Locke's plaster cast shows that the stone has been peeled off just where some letter was engraved, presenting a socket, which is apt to delude the eye with the appearance of a ma or mu, or to come out in an estampage as a ma or mu. His cast also shows that there is a short space after sahasehi, just enough for engraving one letter. I have carefully examined this space where I find the faint trace of a letter, which is no other than ve. The same has also been made out froin one of the two impressions within my access. See Pl. I in J BORS, 1927, Vol. XIII, Parts III-IV. The letter lost or obscured in the socket is found to be, both in Locke's cast, rather du than ma or mu,
33. Prinsep reads la; Cunningham, kala, Indraji, käle; Fleet, too, kāle, Jayaswal, kalam, correcting his former reading kale or kāle. I doubt very much if the first letter can be read as ka or kā. I say it is not at all ka or ka. Locke's cast clearly shows how

Page #56

28
OLD BRAHMI INSCRIPTIONS
rain-water trickling down along the vertical line of na of ni and flowing to some distance along the right hand side of its horizontal base line, flows down at last cutting deep into the rock and serving to transform the na into a ka-shaped hobgoblin. I shall be the last man to mistake this apparition for kā in which two vertical strokes of equal length go to meet the horizontal base line from two sides at two distant points.
34. Prinsep reads machinam ; Cunningham, ce cinam; Indraji, vochine : Jayaswal, vochimne or vochimnam, correcting his former reading vocchine,
35. Prinsep reads ca coyatha ; Cunningham, cace yatha ; Indraji, ca coyatha ; Jayaswal, ca cayațhi, correcting his former reading ca-cheyathi. The first letter, as it appears in Locke's cast or in Banerji's impressions, in either ce or la, and the second letter, ignoring the vowel mark, is ca or ta. One may correctly read the four letters together as cecayathi or latä-yathi.
36. Prinsep reads agi sati ka : Canningham, ageszti ka ; Indraji, agesati ku; Jayaswal, agasatikam, correcting his former reading Argasi ti kam. Sten Konow, Am gasatikam. tariyam. Jayaswal finally accepts Sten Konow's reading. I frankly confess that it is very difficult to resist the temptation of the reading of the second letter as ga. It is only after a prolonged examination that one will be inclined to read it as dha. I can press these three arguments in favour of reading it as dha : (1) that it is so unlike other forms of ga in the Háthi-Gumpha inscription ; (2) that Locke's cast also shows the dha, and (3) that in one of Vats' impressions it stands out almost as a dha.
37. Prinsep and Cunningham read ka tariyam, which is quite possible. Indraji reads ku tariyam ; Jayaswal, amtariyam, correcting his former reading kam tariyam. Sten Konow accepts Jayaswal's reading satikamtariyam. Jayaswal finally reads turiyam. I find no difficulty in reading tiriyam.
38. Prinsep reads napadayati ; Canningham, napada chati; Indraji, capadayati, Jayagwal, upādāyati, correcting his former reading upādiyati. Sten Konow correctly reads upādayati. I am tempted to read the concluding words of the record of the 14th year also as cetiya-thambhagasatam katariyam upadayati.
AVON

Page #57

I. SCRIPTION OF KHĀRAVELA
29
16. Khema'-rājā sa [,] Vadha-rājā sa [] (Imda)". rājāt sa [] Dhama ®-rājä* pasamto?sunamto 8 anubhavamtoo kalāņāni 10 [116] [-]" guņa 12-visesa 13-kusalo sava pāsamdapājako 14 sava-devậyatana 15.samkāra 18-kārako ?? apatihata. 18 caka 19-vāhana 20-balo 21 caka-dhar(o) 22 guta-cako 28 pavata- 24 cako rāj(i)si 25-vamsa 20-kula 27-vini(s)ito 28 mahāvijayo 29 rājā Khāravela So-siri 31 [-] [117].
* Cf. Manu-Samhitā, VII. 7:
So'gnir bhavati Vāyuś ca so'rkaḥ Somaḥ sa Dharmarāti Sa Kuveraḥ sa Varunaḥ sa Mahendrah prabhāvatah #
1. Prinsop and Cunningham read agama. Indraji is the first to read Khema.
2. Prinsop and Cunninghamn read vatha Indraji reads vadha; Jayaswal, Vadha, correcting his former reading vadha.
3. Prinsep reads rese ; Cunningham, bhi ? Indraji and Jayaswal read bhikhu. The portion of the rock where the two letters were engraved has been peeled off presenting a socket, which deludes the eye with the appearance of two letters resembling bhikhu. I find that bhikhu is but an apparition of Imda.
4. Prinsep reads raja ; Canningbam, -ja. Indraji and Jayaswal correctly read rāja. 5. All have missed sa.
6. Prinsep reads ma. Cunningham and Indraji read nama, Jayaswal correctly reads dhama.
7. Prinsop and Cunningham read pasata. Indraji is the first to correctly read pasamto.
8. Prinsep reads satite ; Cunningham, sanamto; Indraji, sanato, Jayaswal, sunamto, correcting his former reading sunato.
9. Prinsep reads apadhajida ; Vunningham, anubhivato; Indraji, anubhavato; Jayaswal, anubhavamto, correcting his former reading anubhavato.
10. Prinsep reads -lanoni; Cunningham, -Tānani ; Indraji is the first to correctly read kalānāni.
11. The beginning piece of L 17 is missing. 12. Prinsep reada dirono; Cunningham, rutapano; Indraji is the first to read guna.
13. Prinsep and Cunningham read vise, omitting sa; Indraji and Jayaswal, correctly visesa,
14. Prinsep reads sava-påsaida-pājano; Cunningham, sava-pathabhi-pujako; Indraji is the first to reid sava-pāsamda-pūjako.
15. Indraji reads ..tana ; Jayaswal, devāyatana, correoting his former reading ... tina.
16. Prinsep roads kara; Cunningham, makara. Indraji and Jayaswal read samkara. The second lotter might also be treated as a broken kha.
17. Prinsep and Cunningham read kāra; Indraji and Jayaswal correctly, karako.

Page #58

30
OLD BRĀHMI İNSCRIPTIONS
18. Prinsep reads ravato; Cunningham, padahata; Indraji and Jayaswal correctly read apatihata.
19. Prinsep reads laki; Cunningham, cako; Indraji, caki; Jayaswal, caki.
20. Prinsep and Canningham read vähani. Jayaswxl, too, reads vāhani, correcting his former reading vāhanu. Indraji correctly reads vāhana.
21. Prinsep roads bāle; Cunningham, ţhalo. Indraji is the first to read balo.
22. Prinsep reads vākadhara; Canningham, cako dhara ; Jayaswal, caka-dhuro correct. ing his former reading cakadharo. Indraji correctly reads cakadharo.
23. Prinsep alone reads gata cano. Others read guta-cako.
24. Prinsep reads ghavata; Indraji, pasamta; Cunningham, pavata ; Javaswal, too, pavata, correcting his former reading ghisamta.
25. Prinsep reads rājāsa ; Cunningham, rājasa. Indraji ani Jayaswal read räjisi. 26. Cunningham and Indraji read vamsa. Jayaswal reads vasa. 27. Prinsep reads kala. Cunningham is the first to read kula.
28. Prinsep reads dināravato; Canningham, vinigato; Indraji, vinicchito; Jayaswal, vinisrito, correoting his former reading vinicchito.
29. Prinsep roads mahāvijaye ; Cunningham, mahavijayo. Indraji and Jayaswal read mahānijayo.
30. Prinsep reads khāravela; Canningham, kharapela. Indraji is the first to read Khāravela as a name.
31. Prinsep reads samdara; Cunningham, sirino. Indraji is the first to read siri,

Page #59

www. kobatirth.org
TEXT AS IT READS IN THE ORIGINAL १। नमो अरहंतानं [1] नमो सव-सिधानं [1] ऐरेन वेरेन इति वा महाराजेन महामेघ-वाहनेन चेतराजवंसवधनेन पसथ-सुभ-लखनेन चतुरंतरखण-गुण-उपेतेन कलिंगाधिपतिना सिरि-खारवेलेन पंदरस-वसानि सिरिकड़ार-सरीरवता कौड़िता कुमार-कोड़िका [1] ततो लेख-रूप-गणनाववहार-विधि-विसारदेन सवविजावदातन नव-वसानि योवरजं व सासितं [1] संपुण-चतुवी सति वसो सोदानि-वधमान-सेसयोवनाऽभिजयो ततिये कलिंगराजवंसे पुरिस-युगे महाराजाभिसेचनं पापुनाति [1]
२। अभिसितमतो च पधमे वसे वात-विहत-गोपुर-पाकार-निवेसनं पटिसंखारयति कलिंग-नगरि [] गभोर-सोतल-तड़ाग-पाडियो च बंधापयति [] सवयान-पटिसंठापनं च कारयति पनतोसाहि सत-सहसेहि [1] पकतियो च जयति [1]
३। दुतिये च बसे अचितयिता सातकणिं पछिम-दिसं हय-गजनर-रध-बहुलं दंडं पठापयति [,] कलिंगागताय च सेनाय वितामेति असकनगरं ( असिक-नगरमिति वा ) [1]
४। ततिये पुन बसे गंधव-वेद-बुधो दंप-नत-गीत-वादित-संदसनाहि उसव-समाज-कारापनाहि च कोडापयति नगरि [1]
५। तथा चवुथे वसे विजाधराधिवासं भरकतपुरं कलिंग-पुव-राजानं धमेन व नितिना व पसासयति सवत धमकूटेन [,] भीत-तसिते च निखित-छतभिंगारे हित-रतन-सापतये सव-रठिक-भोजके पादे वंदापयति [1]
६। पंचमे च दानि वसे नंदराज-तिवससत-ओघाटितं तनसुलियवाटा पनाडिं नगरं पवेसयति [.] सत-सहसहि च खनापयति [1]
७। अभिसितो च [ छढे व से राज-सिरिं संदंसयंतो सव-कर-वणअनुगह अनेकानि सत-सहसानि विसजति पोर-जानपदं [1]
८। मतमे च बसे असि-छत-धज-रध-रखि-तुरंग-सत-घटानि सक्त संदंसनं सव-मंगलानि कारयति[-] सत-सहमेहि [1]
(31)

Page #60

32
OLD BRĀHMI INSCRIPTIONS
। अठमे च वसे महति-सेनाय मधुरं अनुपतो गोरधगिरिं घातापयिता राजगहानं पपीड़[]पयति[.] एतिनं च कंम-पदान-पनादेन-संभौत-सेन-वाहनेविपमंचितं मधुरं अपयातो येव नरिदो सव-[घरवासिनं च सव-[-] गहपतिनं च [-] स [-] पान-भोजनं ददाति [.] [ कलिं गं याति [,] पलवभार-कपरुख-हय-गज-नर-रध-सह याति [.] सव-घरवासिनं च सव-राजभतकानं च सव-गहपतिकानं च सव-बम्हणानं च पान-भोजनं ददाति [, निगंथानं समणानं च [-] ददाति [-] सत-सहसेहि [1]
१०। नवमे च वसे वेडुरिय-कलि[]ग-राज-निवासं महाविजय-पासादं कारयति अठतिसाय सत सहसेहि [1]
११। दसमे च वसे कलिंग-राजवंसानं ततिय-युग-सगावसाने कलिंगपुव-राजानं यस-सकारं कारापयति सत-सहसेहि [1]
१२। एकादसमे च वसे [-]-मणि-रतनानि सह याति [, [-] कलिंग-पुवराज-निवेसित-पिथडग-दर्भ नगले नेकासयति । अनुपदभवनं च तेरस-वस-सत-कतं भिदति तिमिरदह-संघातं [1]
१३। बारसमे च वसे [...] सत-सहसे हि वितासयति उतरापधगजानो,] [-]मागधानं च विपुलं भयं जनेतो हथीसं गंगाय पाययति[,] मागधानं च राजानं बहसनिमितं पादे वंदापयति[.] नंदराज-नोतं कालिंग-जिनासनं अंग-मगधतो कलिंगं आनेति हय-गज-सेनवाहन-सहसेहि [.] अंग-मगध-वासिनं च पादे वंदापयति [] [-] वोथि-चतर-पलिखानि गोपुरानि सिहरानि निवेसयति [,] सत वासुको रतनं पेसयंति [.] अभूतमछयियं हथीस-पसवं परिहरंति [5] मिग-हय-हथो उपनामयंति [,] पंडराजा विविधाभरणानि मुता-मणि-रतनानि आहरापयति इध सत-सहसानि [] [-]-सिनो वसी कारति [1)
१४। तेरसमे च वसे सुपवत-विजय-चके कुमारी-पवते अरहतो परि निवसतो हि कायनिसौदीयाय राजभतकेहि राज-भातिहि राज-जातिहि राज-पुतेहि राज-महिसिहि खारवेल-सिरिना सतदस-लेण-सतं कारापितं [1]
१५। [-] ? सकत-समण-सुविहितानं च सत-दिसानं यतिनं तापसइसिन लेणं कारयति [] अरहत-निसोदीय-समीपे पभार वराकर-समुथापिताहि अनेक-योजनाहिताहि पनतोसाहि-सत-सहसाहि सिलाहि सिल-थंभानि च

Page #61

INSCRIPTION OF KHĀRAVELA
33
चेतियानि च कारापयति [] [-] पटलिक-चतरे च वेड़ रिय-गर्भ थंभे पतिठापयति पनतरिय-सत-सहसेहि [] [वै]ड़ रिय-नील-वोर्किन चेच-यठि-अधसतिकं तिरियं उपादयति [1]
१६। खेमराजा स [,] वध-राजा स [.] ईद-राजा स [] धम-राजा पसंतो सुनंतो अनुभवंतो कलाणानि [-] गुण-विसेस-कुसलो सव-पासंड-पूजको सव-देवायतन-संकार-कारको अपरिहत-चक-वाहन-बलो चकधरी गुत-चको पवत-चको राजिसि-वस-कुल-विनिसितो महाविजयो राजा खारवेल-सिरि (1)

Page #62

www. kobatirth.org
TEXT AS IT READS IN PALI
१। नमो अरहन्तानं [॥] नमो सब्ब-सिद्धानं [॥] अयिरेन (वीरेण इति वा) महाराजेन महामेघवाहनेन चेतराजवंस-वधनेन पसस्थ-सुभलक्खणेन चतुरन्तरक्खणगुणोपेतन कलिङ्गाधिपतिना सिरि-खारवेलेन पन्नरस-वस्सानि सिरिकलार-सरोरवता कोलिता कुमार-कोलिका [1] ततो लेख-रूप-गणना-वोहारविधि-विसारदेन सब्बविज्जोदातेन नव-वस्सानि योवरज्जं व सासितं [1] संपुरमचतुवोसति-वस्सो सो दानि-वद्धमान-सेसयोब्बनाभिविजयो ततिये कलिङ्गराजवंसे पुरिस-युगे महाराजाभिसे चनं पापुणाति [1]
२। अभिसित्त-मत्तो च पठमे वस्से वात-विहत-गोपुर-पाकार-निवेसनं पटिसंखारयति कलिङ्ग-नगर [.] गभीर-सीतल-तलाक-पालियो च बन्धापयति [.] सब्बु य्यान-पतिसण्ठापनं च कारयति पञ्चतिमेहि सत-सहस्सेहि [.] पकतियो च रञ्जयति [1]
३। दुतिये च वस्से अचिन्तयित्वा सातकणिं पच्छिम-दिसं हय-गज-मररथ-बहुलं दण्डं पट्ठापयति [.] कलिङ्गागताय च सेनाय वित्तासे ति अस्मक-नगरं' ( इसिक-नगर-मिति वा) [1]
४। ततिये पुन वस्से गन्धब्ब-वेद-बुधो दप्प-नच्च-गोत-वादित-सन्दस्मनाहि उस्सव-समज्जा-कारापनाहि च कोलापयति नगरिं [1]
५। तथा चतुस्थे वस्से विज्जाधराधिवासं अरकतपुरं कलिङ्ग-पुब्बराजानं धम्न व नोतिना व पसासयति सब्बत्थ धम्मकूटेन [.] भीत-तसित च निक्वित्त-छत्त-भिङ्गार हित-रतन-सापतेये सव-रहिक-भोजके पादे वन्दापयति [1]
६। पञ्चमे चेदानि वस्मे नन्दराज-तिवस्मसतोग्घाटितं तनसुलिय-वद्या पणालि नगरं पवेसयति [, सत-सहस्सेहि च खनापयति [1]
७। अभिसित्तो च [छठे व स्मे राजसिरिं सन्दस्मयन्तो सब्ब-कर-पण.. अनुग्गह-अनेकानि मत-सहस्सानि विसज्जति पोर-जानपदं [1]
८। सत्तमे च वस्से असि-छत्त-धज-रथ-रक्खि -तुरङ्ग-सत-घटानि सम्बस्थ सन्दस्मनं सव्व-मङ्गलानि कारापयति [-] सत-सहस्मेहि [1]
1. पाठान्तरे-कसपेसागताय......मुसिक-मगरं ।
(34)

Page #63

INSCRIPTION OF KHÁRAVELA
35
। अट्ठमे च वस्ने महतिया सेनाय मधुरमनुपत्तो गोरथगिरिं घातापयित्वा राजगहकानं पपोलापयति [,] एतानञ्च कम्म-पदान-पनादेन-सम्भीत-सेनवाहने विप्पमुञ्चितु' मधुरमपयातो येव नरिन्दो सब्बेघरवासिनञ्च सब्ब- [-] गहपतिकानञ्ज [-] स[-] पान-भोजनं ददाति[] [कलिङ्ग याति[, पल्लवभारकप्परुक्ख-हय-गज-नर-रथ-सह याति[] सब्ब-घरवासिनञ्च सब्ब-राजभतकानञ्च सब्ब-गहपतिकानञ्च सब्ब-ब्राह्मणानञ्च पान-भोजनं ददाति,आरहतानं समणानञ्च [-] ददाति [-] सत-सहस्मेहि [1]
१०। नवमे च वस्ने वेलुरिय-कलिङ्ग-राजनिवासं महाविजय-पासादं कारयति अतिसेहि सत-सहस्मेहि [1]
११। दसमे च वस्ने कलिङ्ग-राजवंसानं ततिय-युग-सग्गावसाने कलिङ्गपुब्ब-राजानं यस-सक्कारं-कारापयति सत-सहस्मेहि [1]
१२। एकादसमे च वस्से [...]-मणि-रतनानि सह याति[] [-] कलिङ्ग-पुब्बराज-निवेसित-पिथुदक-दभ नङ्गले नेकासयति [.] अनुपदभवनच तेरस-वस्म-सत-कतं भिन्दति तिमिरदह-संखातं [1]
१३। वारसमे च वस्ने [...] सत-सहस्मेहि वित्तासयति उत्तरापथराजानो[,] [-]-मागधानञ्च विपुलं भयं जनयन्तो हस्थिस्मं गङ्गाय पाययति[,] मागधानञ्च राजानं बेहस्मति-मित्तं ( वेहप्फतिमित्तं इति वा) पादे वन्दापयति [,] नन्दराज नोतं कालिङ्ग-जिनासनं अङ्ग-मगधतो कलिङ्ग आनेति[] य-गज-सेनवाहन-सहस्मेहि[] अङ्ग-मगध-वासिनञ्च पादे वन्दापयति[,] [-] वीथि-चत्तर-पलिखानि गोपुराणि सिखराणि निवेसयति[,] सत-वासुको रतनं पेसयन्ति[,] अन्भुतमच्छरियं हत्थिस्म-पसवं परिहरन्ति [] मिग-हय-हस्थी उपनामयन्ति[] पण्डराजा विविधाभरणानि मुत्ता-मणिरतनानि आहरापयति इध सत-सहस्सानि[] [-] -सिनो वसोकारति [1]
१४। तेरसमे च वस्से सुपवत्त-विजय-चक्के कुमारी-पब्बते अरहतो परिनिवसतो हि काय-निसौदियाय राजभतकेहि राज-भातीहि राज-आतीहि राज-पुत्तेहि राज-महिसोहि खारवेल-सिरिना सत्तदस-लेण-सतं कारापितं [1]
१५। [-]? सक्कत-समण-सुविहितानञ्च सत-दिसानं यतीनं तापसइसोनं लेणं कारयति[,] अरहत-निसौदिय-समीप पबभार वराकर-समुत्थापिताहि

Page #64

36
OLD BRAHMI INSCRIPTIONS अनेक-योजनाहिताहि पञ्चतिसाहि-सतसहस्साहि सिलाहि सिला-थम्भानि च चेतियानि च कारापयति[,] [-] पटलिक-चत्तरे च वेलुरिय-गबभे थम्भे पतिद्वापयति पञ्च-सत्तति-सतसहस्मेहि[,] [वै] लुरिय-नील-वोच्छिवं चेञ्च-यट्टिअवसतिकं तिरियं उप्पादयति [1]
१६। खेमगजा स [,] वडराजा स[] इन्दराजा स [] धम्मराजा पस्मन्ती अनुभवन्तो कल्याणानि [-] गुण-विसेस-कुसलो सब-पासण्ड-पूजको सब्बदेवायतन-संखार-कारको अप्पतिहत-चक्क-वाहन-बलो चक्कधरो गुत्तचक्को पवत्तचक्को राजिसि-वंस-कुल-विनिस्मितो महाविजयो राजा खारवेल-सिरि []

Page #65

www. kobatirth.org
TEXT AS IT READS IN SANSKRIT
१। नमोऽहंद्राः [] नमः सर्वसिद्धेभ्य: [u] आर्येण (वोरण इति वा) महाराजन महामेघवाहनेन 'चेत'-राजवंशवर्धनेन प्रशस्त-शुभ लक्षणन चतुरन्त-रक्षण-गुणोपेतन कलिङ्गाधिपतिना श्री-खारवेलेन पञ्चदश-वर्षाणि श्रो-कड़ार-शरोरवता कोडिताः कुमारकोड़ाः [1] ततो लेख-रूप-गणनाव्यवहार विधि-विशारदेन सर्वविद्यावदातेन नव-वर्षाणि यौवराज्यमेव शासितम् [1] सम्पूर्ण-चतुर्विंशति-वर्षः स इदानों वईमान-शेष-यौवनामिविजयस्ततोये कलिङ्ग-राजवंशे पुरुष-युगे महाराजाभिषेचनं प्राप्नोति [1]
२। अभिषिक्त-मात्रश्च प्रथमे वर्षे वात-विहत-गोपुर-प्राकार निवेशनं प्रतिसंस्कारयति कलिङ्गनगरे [,] गभीर-शीतल-तड़ाग-पालीश्च बन्धयति [] सर्वोद्यान-प्रतिसंस्थापनञ्च कारयति पञ्चविंश-शतसहस्त्रैः [] प्रकृतीच रस्त्रयति [1]
३। हितोये च वर्षे अचिन्तयित्वा शातकर्णि पश्चिमां दिशं हय-गज-नररथ-बहुलं दण्डं प्रस्थापयति [.] कलिङ्गागतया च सेनया वित्रासयति अश्मकनगरम्' (ऋषिक-नगरमिति वा) [1]
४। तृतीये पुनर्वर्षे गन्धर्ववेद-बुधः दर्प-नृत्त-गीत-वादित्र-सन्दर्शनैरुत्सवसमाज-कारणाभिश्च क्रीड़यति नगरौम् [1]
५। तथा चतुर्थे वर्ष विद्याधराधिवासं अर्कतपुरं कलिङ्ग-पूर्वराजानां धर्मेण वै नौत्या वै प्रशासयति सर्वत्र धर्मकूटेन [.] भीत-त्रस्तान् च निक्षिप्त-छत्रभृङ्गारान् हत-रत्न-स्खापतेयान् सर्व-राष्ट्रिक-भोजकान् पादौ वन्दयति [i]
६। पञ्चमे चेदानी वर्षे नन्दराजेन त्रिवर्षशतात् पूर्वमुद्घाटितं तनशूलीयघमनः प्रणालों नगरं प्रवेशयति [5] शतसहस्त्रैश्च खानयति [1]
। अभिषिक्तश्च [षष्ठे वर्षे राजश्रियं सन्दर्शयन् सर्व-करपणाद्यनेकानुअहान् शतसहस्रसंख्यकान् विसृजति पौर-जानपदे [1]
1. पाठान्तर-कृष्णवेणागतया......भूषिक-नगरम् ।
(37)

Page #66

www. kobatirth.org
38
OLD BRAHMI INSCRIPTIONS
८। सप्तमे च वर्षे असिच्छत्र-ध्वज रथ-रक्षि-तुरङ्ग-शत-घटानां सर्वत्र सन्दर्शनं सर्व-मङ्गलानि च कारयति शतसहस्रः [1]
८। अष्टमे च वर्षे महत्या सेनया मथुरामनुप्राप्तः गोरथगिरि घातयित्वा राजग्राहकान् प्रपौड़यति [.] एतेषाञ्च कर्म-प्रदान-प्रणादेन सम्भीतान् सेनावाहनान् विप्रमोक्तं मथुरामपयात एव नरेन्द्रः सर्वग्रहवासिभ्यश्च सर्व [...] गृहपतिभ्यश्च [...] स [-] पान-भोजनं ददाति [.] [कलिङ्गं याति [,] पल्लव भार-कल्पवृक्ष हय गज-नर-रथैः सह याति [.] सर्व ग्रहवासिभ्यश्च सर्व-राज भृत्यकम्यश्च सर्वगृहपतिभ्यश्च सर्व-ब्राह्मणेभ्यश्च पानभोजनं ददाति [.] आईतेभ्यः श्रमणेभ्यश्च [--] ददाति [-] शतसहस्रैः [1]
१०। नवमे च वर्षे वैदूर्य कलिङ्ग-राजनिवासं महाविजय-प्रासादं कारयति अष्टाविंशच्छतसहस्त्रैः [1]
११। दशमे च वर्षे कलिङ्ग-राजवंशानां टतीय-युगसर्गावसाने कलिङ्गपूर्वराजानां यशस्सत्कारं कारयति शतसहस्रः [1]
१२। एकादशे च वर्षे [-] मणि-रत्नैः सह याति [] [-] कलिङ्गपूर्वराज-निवेशित-पृथूदक-दर्भ लागले निष्कासयति [.] अनुपदर्भवनञ्च त्रयोदशवर्ष-शत-कृतं भिनत्ति तिमिरहद-संख्यातम् [1]
१३। हादशे च वर्षे [-] शतसहस्त्रैः वित्रासयति उत्तरापथ-राजान् [] [-] मागधान् च विपुलं भयं जनयती हस्त्यखं गङ्गायां पाययति [, मागधानाञ्च राज्ञां बृहस्पतिमित्रं. पादौ वन्दयति [, नन्दराज-नौतं कालिङ्ग-जिनासनम् अङ्गमगधेभ्यः कलिङ्गमानयति हय-गज-सेनावाहनसहनैः [.] अङ्ग-मगध-वासिनः पादौ वन्दयति [] [-] वीथि चत्वर-पलिघान् गोपुराणि शिखराणि निवेशयति [] शतं वासुकयो रत्नं प्रेषयन्ति [.] अद्भ तानाश्चान् हस्यख-पशून् परिहारयन्ति[,] मृग-हय-हस्तिन: उपनामयन्ति[] पाण्ड्यराजा विविधाभरणानि मुक्ता-मणि-रत्नानि आहारयति इह शतसहस्त्रसंख्यकानि [] [-] सिनः वशोकारयति [i]
१४। त्रयोदशे च वर्षे सुप्रवर्त्त-विजय-चक्रे कुमारो-पर्वते अर्हतः परिनिवसतो हि काय-निषीदों राजभृत्यैः राजधाटभिः राजपुत्रः राजमहिषोभिः खारवेलेन श्रीमता सप्तदश-लयन-शतं कारितम् [1]

Page #67

INSCRIPTION OF KHĀRAVELA
39
१५। [-] सत्कृत-श्रमणेभ्यः सुविहितेभ्यश्च शतस्य दिशां यतिभ्यः तापसर्षिम्यः लयनं कारयति []] अर्हविषीद्याः समीपे प्राग्भारे वराकरसमुत्थापिताभिरनेक-योजनाहिताभिः पञ्चविंशच्छतसहस्राभिः शिलाभिः शिलास्तम्भान् च चैत्यानि च कारयति [.] पाटलिक-चत्वरे च वैदूर्य-गर्भ स्तम्मान् प्रतिष्ठापयति पञ्चसप्तति-शतसहस्रः [] [वैदूर्य-नील-व्यवच्छिन्नं चैत्ययष्टाईशतकं तिर्यक् उत्पादयति [1]
१६। क्षेमराज: स [] वई-राजः स [1] इन्द्रराजः स [5] धर्मराजः पश्यन् शृण्वननुभवन् कल्याणानि [-] गुण विशेष-कुशल: सर्व-पाषण्ड-पूजकः सर्व-देवायतन-संस्कार-कारकः अप्रतिहत-चक्र-वाहन-बल: चक्रधरः गुप्तचक्रः प्रवर्त-चक्र: राजर्षि-वंश-कुल-विनिःसृतो महाविजयो राजा खारवेल-श्रीः [1]

Page #68

INSCRIPTION OF KHĀRAVELA
TRANSLATION
1. Let the head bend low in obeisance to Arhats, the Exalted Ones. Let the head bend low (also) in obeisance to all Siddhas, the Perfect Saints.
By His Lordly' (and) Graceful Majesty 3 Khāravela, the mighty ruler, the Sovran Lord of Kalinga, whose vebicle is Mahāmegha (the greatcloud-like state-elephant), who is the increaser of the Royal House of Ceta, who is possessed of the noble and auspicious bodily marks, who is gifted with the capacity for protecting the earth extending as far as the four seas, were played, for fifteen years, the games befitting the young age of the prince,8 with a body lovely and of fair brown complexion.'! Thereafter,
1. Adoration or salutation as a rendering of namo does not precisely convey its meaning. The Indian namaskāra primarily signifies the bending of the head in token of respect. Here namo is a symbolical expression, into which a whole sentence has been abbreviated.
2. Here Aira has been taken to be a synonym of Ārya or Srämi, Lord' or 'Master.' I am against equating it with Aila, a derivation from Ilā. Adopting the reading Verena, I may offer By His Heroio, etc.' as an alternative rendering
3. The English rendering of Siri is Graceful rather than Gracious.
4, Or, who vies with Mahendra, the mighty king of the gods, whose vehicle is cloud. For the epithet Mahämeghaváhana, cf. the Mahabharata, Sabhāparva, Ch. XIV, Verse 13 : Vakradantah Karugas ca Karabho Meghavāhanah. Jayaswal seems to think that the Megha in the Parānas is nothing but a shortening from Meghavähana. In the Arthasastra, II. 2. 20, the elephants of Kalinga, Aiga, Pracya and Karūsa are said to have been of the best breed: "Kalinganga-gajāḥ śresthah Pracyäś cêti Karūsajäh."
5. The Royal House of Ceta is the same expression as Jayaswal's "the House of King Ceta." Adopting the reading Ceti-tājavamsa, one may translate" of the Cedi rnling race."
6. That is, the marks or features that are of importance to the astrologer, diviner, palmist or physiognomist.
7. Adopting the reading caturamtala.thuna, one may render with Jayaswal-"the support of the whole land"; or, adopting the reading caturama-lathuna, one may render with Sten Konow-"provided with virtues famous to the four ends of the world)."
8. Jayagwal's "princely games" may be preferred for its brevity,
9. With Childers kaļāra or kaļāra is "tawny" or "tan-coloured." Sten Konow suggests that Siri-kadara is the same Prakrit expression as Siri-katāra, which means, according to the Sabdamālā, nāgara and kami. From this he is led to think that " Siri-kadāra is the lover of Sri, i.e., Krona," and that "Kharavela's boyish games are compared with Krana's pranks and sports in Vindāvana." Jayaswal
( 40 )

Page #69

INSCRIPTION OF KHARAVELA
41
for nine years, just the office of a Crown-Prince was administered by (His Royal Highness) who became an expert in (matters relating to) writing, coinage, accounting, administration and procedures, whose self was purified by proficiency in all (Indian) 'polite learning.'2 On completing then (his) twenty-fourth year, he who, as he waxed great, passed the rest of (his) manhood in making conquests, gained the high state implied in the coronation of a great king, in the royal dynasty of Kalinga, in the third generation of two kings."
2. And as soon as he was anointed, in the very first year, (His Majesty) caused the gates, walls and buildings that had been damaged by stormy wind to be repaired in the city of Kalinga, and caused the embankments of the deep and cool tanks to be built up, and (also) caused the work of restoration of all the gardens to be done at the cost of thirty-five hun dred thousand (coins), and (thus) enabled the people (his subjects) to be pleased."
finally accepts Sten Konow's explanation and renders Siri-kadara "the lover of Sri (Vienu)." I think it will not be judicious to commit ourselves so far as that. "Kadaram trna-bahni-vat" is a quotation made by Jayaswal just to the point. Cf. Kadara-Jaimini, Kalara-Janaka, and Kalara-maṭṭuka.
1. See Notes passim.
2. I mean, the whole body of useful sciences and arts. See Notes passim.
3. The expression vadhamana involves the metaphor of the moon waxing day by day. Cf. Kalidasa's Kumara-Sambhava-Kavya, I, 25:
Dine dine sa parivardhamana labdhodaya candramasīva lekha |
4. For maharajabhişecana, cf. yuvarajabhişecana in the Artha-Sastra, II. 36.
5. Jayaswal takes tutiye Kalimga-rajavamse purisa-yuge to mean "in the third dynasty of the Aira line of the kings of Kalinga." According to R. C. Mazumdar, the phrase does not mean any more than "in the third generation of the Kalinga kings, the third generation of the same reigning dynasty." In accordance with the above rendering, the phrase meaus. "in the third generation of the same reigning dynasty of Kalinga, each generation of which was considered as consisting of two kings." Jayaswal's rendering is evidently based upon the authority of a few Sanskrit stanzas quoted by him from an old Oriya manuscript, which according to Prof. Haraprasad Shastri, belongs to the 14th century A.D., These stanzas clearly state that a king named Aira, who was the lord of Utkala, had defeated King Nanda in a battle, and that he was a great friend or enemy of Asoka. There is nothing definite to show that Kharavela belonged to the third dynasty of the Aira line of Kings. For further discussion, see Notes passim.
6. Jayaswal connects pana-tisähi satasahasehi with pakatiyo, taking it to denote the total number of the subjects, the citizens of Kalinga, whom Kharavela sought to please. He cites with confidence, in his support, the authority of the Siddhanta-Kaumudi which
6

Page #70

42
OLD BRAHMI INSCRIPTIONS
3. And in the second year, not (at all) bringing Satakarņi into (bis) thought, (His Najesty) caused a multitudinous troop of horses, elephants, foot-men and chariots to move on to the western quarter, and with the aid of the army from Kalinga, struck terror, into the city of A saka.
4. Again, in the third year, (His Majesty), who was versed in the science of music--the Gandharva-lore, caused the capital (the city of Kalinga) to be amused by exhibiting 'combats,' 2 and dancing, singing and musical performances, and (no less) by arranging festivities and merry gatherings.'3
5. Likewise, in the fourth year, (His Majesty) caused the city of Arakata which was the habitation of Vidyādbaras, the spell-muttering denizens, to be governed in accordance with the principle and usage of the former kings of Kalinga, all over by the highest kind of law, and compelled all the local chiefs '4 and headmen '5 who were robbed of their wealth
ganctions the use of the expression "Satena vatsan pāyayati payah," applying satena to mean "satena parichidya." Supposing satena in this expression is connected with vatsān (calves, an Accusative plural), what does it mean? Does it mean a hundred calves ? I would say, No. Here satena vatsan just means "the calves divided into the groups of one hundred." Applying this meaning of satena vatsan to panatisähi sata-sahasehi pakatiyo in Khåravela's text, one has to say that Khāravela pleased his subjects, dividing them into the groups of thirty-five hundred thousaad persons, a meaning, which hardly suits the context. See what Patañjali has got to say on this point. In commenting on Pāņini's aphorism, II. 3. 18, he says: “ Dvidroñena dhanyaın krīņātiti......dvidroşene kiranyena dhānyam krināti" and 80 as to "pañcakena pašūn, sähasrena aśvān krinātiti." Thas going by the authority of Patañjali, one may explain Khāravela's clause as meaning "he sought to please his subjects by spending thirty-five hundred thousand coins,' if pana-tīsāhi sata-sahasehi be at all connected with pakatiyo. For connecting panatīsāhi satasahasehi with pakatiyo Jayaswal's argument is that there is a large space between kārayati and panatisāhi. I can quite see that spaces in the Håthi-Gumpha inscription have some meaning. But that meaning in not in all places punctuation. The spaces are also intended to keep twe words distinct. Even it is found that spaces are not where they should be, e.g., in L 4, before tatiye puna Dase. • 1. Note that in the Nägik Cave inscription, Cave No. 3, Satukarņi has been described as the lord of Arika and Asaka among other places mentioned by name (A sika-A saka ...... pati). Adopting Jayagwal's reading Kañhapenāgatāya...... Musika-nagaram, one may translate" with the aid of the army which advanced as far as the Kraņavenā river, struck terror into the Mūsika city."
2. For the meaning of dampa or darpa, see Notes passim.
3. For the meaning of samāja, see Notes passim. 4, 6, For the neaping of rathika-bhojalca, see Notes passim.

Page #71

İNSCRIPTION OF KHĀRAVELA.
43
and jewels, whose insignia consisting of umbrellas and golden vases had been cast away, who were frightened and terrified, to bow at (bis) feet.
6. And, then, in the fifth year, (His Majesty) caused the canal opened out by King Nanda, a hundred-and-three years back, to be brought into the capital from the Tanašūli” road, and caused (it) to be dug at the cost of a hundred thousand (coins).
7. And in the sixth year since he was anointed, (His Majesty), while displaying the royal fortune, 3 bestowed an unprecedented) favour on the inhabitants of the towns and districts 4 by remitting all taxes and duties 5 amounting to many hundred thousand (coins).
8. And in the seventh year, (His Majesty) caused a hundred kinds of pom pous parade of swords, umbrellas, flags, chariots, guards and horses, and all ceremonies of victory i to be performed at the cost of some hundred thousand (coins).
9. And in the eighth year, (His Majesty) baving killed 8 Gorathagiri on reaching Matharā with a mighty army, caused a terrible pressure to be brought to bear upon the people of Rājagaha (Rājagļha),10 and, in order to release the troops and vehicles terrified by the uproar of counter-attacks
1. Or, three hundred years.
2. To me, Tanasuli is the same expression as Tanasiva. I am not sure if Tanasult is the same place as Tosali mentioned in Asoka's two Separate Rock Edicta.
3. Adopting tho reading rājasūyam, Jayaswal translater" showing Rājasūya."
4. Jayaswal seems to go too far in explaining Pora.Jānapada as moaning the Paura and the Janapada as two corporate bodies. See Notes passim.
5. Here kara-pana has been taken as an equivalent of the Sanskrit kara-pana or kara. panya. Jayaswal translates "all Government taxes (lit., tad money)."
6. Or, remitted all taxes and duties and did many hundred thousand (other) kinds of farour.
7. Note that in the Sattu-Nipāta, mangala has been used as the opposite of parābhava. The idea of victory attaching to the word mangala is clearly brought out in the conclu. ding verse of the Mangala-Sutta which reads :
Etādisāni katvāna subbattham aparăjitā
Sabbattham sotthim gacchanti tam tesam mangalamuttamuntil 8. Suggesting the reading malata-bhitti Goradharim ghūtāpayitä, Jayaswal translates “having got stormed the Gorathagiri (fortress) of great enclosure." For arguneuts for and against taking Gorathagiri as the name of a hill fortress or as a personal name, see Notes passim.
9. Literally, terribly opp:essed.
10. Adopting the reading Rajagaham upapīdāpayati, Jayaswal translates causes pressure around Rājagrha (lays siege to Rājagrha)."

Page #72

10
44
OLD BRAHMI INSCRIPTIONS
on their part, as he, the lord of men, retreated to Mathura, 1 offered food-and-drink as a means of entertaining all the inbabitants of the place), and all the royal servants, all the householders, all the Brahmins and Arbata (Jain) recluses, returned to Kalinga, marching back with Kalpavşkşa, the Wishing Tree, burdened with foliage, and the troop of the horses, elephants, footmen and chariots, offered (on return) food-anddrink by way of entertaining all the inhabitants, and all the royal servants, and all the householders, and all the Brahmins, and offered food and drink (also) for entertaining the Arbata (Jain) recluses, at the cost of [so many] hundred thousand (coins).
10. And in the ninth year, (His Majesty) caused Mahāvijaya-Pāsāda, the Great-victory Palace, the Beryl® House of the King of Kalinga, to be built at the cost of thirty-eight hundred thousand (coins).
II. And in the tenth year, on the termination of a part of the period of reign of the third generation of two kings of the royal dynasty of Kalinga, (His Majesty) caused homage and honour to be paid to the memory of the former kings of Kalinga at the cost of a hundred thousand (coins). 4
12. And in the eleventh year, (His Majesty) went in procession with jewels, precious stones, etc. [-] caused the grassy jungle of Pșthudaka, founded by the former kings of Kalinga, to be driven into the Langala river, 6 and destroyed the watery jungle of grass called Timira-daha, the Dark-swamp,? which grew up in one hundred and thirteen years. 8
1. Jayaswal translates " on account of this report of the acts of valour (i.e., the capture of Goradhagiri, etc., the king so called to forsake the invested (sambita) division of his army, went away to Mathura indeed." Reading Yavana-rāja Dimata, Sten Konow translates " and through the uproar occasioned by the action the Yavana king Demetrios went off to Mathură in order to relieve his generals who were in trouble."
2. It is mentioned in soine of the Jain books that only the paramount sovereigns we e entitled to perform the ceremony of installing the Wishing Tree.
3. Some of the scholars prefer lapis lazuli as a rendering of vetluriya or raidūrya.
4. Ruadiog mahadhitabhisamayo Bhuradhavasa-pathanam mahi-jayanam...ti kārāpuyati, Jayaswal translates "he having by sacred rites undertaken war, causes... departure for Northern India to conquer the land."
5. Indraji reads the name as Pathula ; Fleet, as Pämthudamga; Syl ain Lévi, as Pithuda (Pihumda); Sten Konow, as Pithumda; and Jayaswal, as Pithuda.
6. Sylvain Lévi and Sten Konow read gadabhanamgalena kāsayati and translate, " caused to be ploughed with the donkey plough," which means "to be destroyed."
7. Or, "abounding in dark swamps."
8. Adopting the reading terasa-casa-sata, Indraji, Jayaswal and Sten Konow translate " thirteen hundred years." Adopting the reading terasa-khasa-satam, R. C. Mazumdar translates "thirteen hundred Khasas."

Page #73

INSCRIPTION OF KHĀRAVELA
45
13. And in the twelfth year, (His Majesty) caused (something to be done] with the aid of (some] hundred thousand, produced consternation among the rulers of Uttarāpatha,' while generating a great fear for the people of [Arga and] Magadha made the elephants and horses to drink at the Ganges, and compelled Bahapatimita, 4 the king of Magadha, to bow at his feet ; caused the honoured seat of the Jina belonging to Kalinga which was taken away by King Nanda to be bronght back from Anga and Magadha to Kalinga by a procession of the horses and elephants and a thousand troops and vehicles, and compelled the inhabitants of Anga and Magadha to bow at his feet; [] caused the streets, courtyards, gate-bars, gates and temples to be set up; one hundred Vāsukis (Dragon Chiefs) sent (him) precious stones, supplied (him) with rare and wonderful elephants, horses and such other animals, and made presents of antelopes, horses and elephants; the king of Pāụdya had procured the various kinds of apparels and ornaments, hundreds and thousands of them, for use here in the city of Kalinga) 9; [-] brought (some ones) into submission.
14. And in the thirteenth year, on the Kumāri hill,' in the well-run realm of victory,8 one hundred and seventeen caves were caused to be made by His Graceful Majesty Khāravela, by his queens, by his sons,
1. Literally, terrifiod.
2. Adopting the reading Utaräpadha-rājānam, one may translate "the king of Uttarā patha."
3. Reading hathi Sugamgaya pāyayati, Jayagwal translates "he effects the crossing of the Ganges on his elephants."
4. See Notes passim. 5. Reading hathi-nāvana for hathisa-pasavam, Jayaswal translates " cargo of elephant.
ships."
6. Note that by idha or iha Asoka had clearly meant Pāțaliputra, his capital. See Rock Edict, V.
7. From the two data supplied in Udyotakeśari's inscription in Lālatendukeśari's cave on the Khaņdagiri hill, the installation of the images of twenty-four Tirthai karas, such as those in the Navamuni cave, and the excavation of a tank, it may easily be inferred, as done by R. D. Banerji, that Kumara-parvata was the ancient name of Khandagiri. If so, by way of contrast, Kumari-pavata nay be regarded as the ancient name of Udayagiri. There is much justification for Banerji to remark that the twin hills, Khaņdagiri and Udayagiri, were known as the Kumāra. Kumāri-pavuta even up to the 10th or 11th century A.D.
8. It seems that supapata-vijaya-cake Kumāri-parate is an expression of the same kind as vijaya-rajye sri-Kumāra-parcata-sthāne. One may also translate : "on the Kumārī hill in the august belt of Suparvata, the noble mountain range,"

Page #74

46
OLD BRAHMI ÎNSCRIPTIONS
by his relatives, by his brothers, by the royal servants for the residing Arhats desiring to rest their bodies."
15. [And in the fourteenth year,] (His Majesty) caused a cave to be built for the honoured 3 recluses of established reputation as well as for the yatis, hermits and sages hailing from a hundred directions ;8 caused the shrines and stone-pillars to be made with stone-slabs quarried out of the best quarries and collected from an extensive area of many leagues; [-] caused the pillars to be set up in a beryl-hall with ornamental floor and ceiling at the cost of seventy-five hundred thousand (coins); and caused half-a-hundred shrine-posts (votive pillars) inlaid with the alternate settings 11 of beryl 1! and emerald (lotuses) 13 to be produced across (alongside).14
16. He who was the lord of security, he who was the lord of bounty, he who was like Inda-rājā, the lord of power, she who was] like Dhamma-rājā, the lord of justice, perceiving, hearing and experiencing the
1. Reading Yāpa-ñāvakehi, Jayaswal translates to the Yăpa professors who are actively engaged in Yūpu und Rhemu practices."
2. For the meaning of kāya-nisidiyāya, see Notes passim. 3. Readiny sukuta, Jayaswal translates "virtuous." 4. Jayaswal translates " well provided for."
5. They are the spiritually advanced Brahmanical ascetics of the fourth stage of effort. Yati is but a synonym of bhikṣu and parivrājuka.
6. They are the Brahmanical religieus of the third stage of effort.
7. They are the high-farned Brahmanical teachers, whether belonging to the second stage of effort or to the third.
8. Reading sava-disanam with Indraji, one may translate" of all quarters."
9. Buddhaghoga explains the term pataliki as meaning "ghanapuppho unnāmayo attharako" "a woollen coveriet thickly woven with flowers." He informs us that a patalika was also known as ūmilaka-patta (Childers, sub ooce Patalika). Jayaswal suggests that pātalikā. aracatoara, meaning "on the lower roofed terrace" is the Sanskrit equivalent of Khāravela's patalika catare.
10. Ceca-yathi=caitya-ganti, cf. chatra-yaşti, Lüders' list, Nos. 925, 927. For cetiya=cecca, see Fausböll's jātika, Vol. V, p. 273 : Ceco'ti Cetiya rājā. Ceca-yathi is but a synonym of cetiya-khambha, which means a caitya-pillar, Lüders' List, Nos. 1210, 1229.
11. The literal rendering of vochimna is "divided alternately into." 12, 13. The vaidūrya and nila varieties of gems are thus enumerated in the Artha-bästra, II. 11.29 : (1) Vaiduryah-utpalavarnaḥ, sirūsapugpakah, udakavarnah, vamiarāgah, sukapatravarnah, pusyarāgah, gomätrakah, gomedakah." (2) "Nilavalīyah-indranilah, kalaya. puspakah, mahānilah, jambavábhah, jimutaprabhah, nandakal, sratanmadhyah, aitavrati, süryakantaś ceti marayaḥ.
14. For justification of the rendering of the concluding clause of the record of the 14th year, see Notes passim.

Page #75

INSCRIPTION OF KHARAVELA
things that are conducive to welfare,' [-] was His Graceful Majesty Kharavela, the great conqueror, who was descended from a family of the dynasty of royal sages, who kept up the realm of royal command, the protector of the realm of royal command, who was strong with undaunted carriers of the realm of royal command, who was the repairer of all abodes of the gods, the honourer of all sects, who became an expert by the possession of special qualities.
3. Literally, by whom is protected.
4. i.e., ministers and officers,
5
i.e., shrines and temples.
1. The five mahākalyānas, according to the Jainas, consist of the Descent, Birth, Initiation, Attainment and Final Release in the life-history of a Jina.
2. Cakka is the anacakka or ājñācakra.
47

Page #76

Page #77

No. II
INSCRIPTION OF KHĀRAVELA'S CHIEF QUEEN
RELATING TO THE VAIKUŅTHAPURA CAVE

Page #78

Page #79

93298
INSCRIPTION OF KHĀRAVELA'S CHIEF QUEEN
The followin, record is connected with the cave which is called Vaikuntha Cave by Mr. James Prinsep and Vaikupthapura Cave by Rājā Rajendra Lāla Mitra. The cave concerned “is in reality," says Mr. R. D. Banerji, “the upper story of a cave with stories and a side-wing, but the local people very often give different names to different parts. It was known as Svargapura sometime ago. In the plan printed with the Puri volume of the Bengal District Gazetteer, the whole group is called Mañcapuri. I have found that the local names of these caves vary with each generation. As one name is forgotten, a new one is immediately invented. The record is incised on the raised space between the second and third doorways in front. This raised space represents a house or verandah with a pointed roof and spires supported little dwarfs who act as brackets. (The record itself) consists of three lines which have suffered much from exposure. The characters of this and the following two inscriptions are slightly later than those of the Hathigumphā inscription of Khāravela."
Regarding this and the following two caves, and regarding this and the following two inscriptions, we have the following authentic account in Rājā Rājendra Lāla Mitra’s Antiquities of Orissa, Vol. II, pp. 14-15:
“Further north-west of the Gopālapura and Muñcapura caves), a little above the level of the surrounding country, we come to a two-storeyed range designed somewhat in the style of the Queen's Palace (Rāņi-nür), but on a smaller scale. The storeys are so arranged as not to rest directly one above the other, but the upper recedes so as to have the top of the lower one open to form a terrace. The upper storey is called Vaikunthapura, and the lower Pātālapura. The lower storey comprises a suite of three rooms, two on a line facing the west, and one on the south, projecting considerably beyond the line of the former, the whole protected by a verandah in front. The eastern rooms are of a trapezoid shape, the extreme length before and behind differing by about I foot, the breadth being 7 feet. Each room has two doors opening into the verandah. The southern room is also a trapezium 10 feet in length on the western side, and 11' 6' on the southern, the breadth being, as before, 7'. It opens into the verandab by a single door. The doorways are framed with pilasters and semi-circular arched
( 51 )
मा. श्री कैलासलागरहुरि गनम दिए श्री महाचीर जैन मागता येत
ST, (ingl ** For Private And Personal Use Only

Page #80

52
OLD BRĀHMT INSCRIPTIONS
bands on the top as in the Queen's Palace, but without any intervening frieze. The two ends of the verandah project forward to the extent of about 5 ft. Its pillars are now very much decayed, and mis-shapen, but originally they were of the Queen's Palace type, square above and below, and octagonal in the centre. The architrave once had a deep frieze formed of basso-relievo figures of men and animals, but it has been very much defaced by the ravages of time.”
"A little beyond the northern projection of the verandah, there is a narrow flight of steps reaching to the upper storey which is an exact counterpart of the lower floor except that on the east side, there is only one room running along the whole length of the verandah, and baving three doors, and the side room has, in the middle of the western wall, a window of a much larger size than any that I have seen on this hill. The dimensions of the eastern room are 22' x 6' 6", that of the southern room 9' X 6', both average, and that of the verandah 24' x 3 ft. The doors measure 3 ft. by 2, and have the usual side pilasters and semi-circular arched bands, but no frieze. The verandah had a range of four detached and two attached pillars, of which the last only are visible. On the architrave on the top of the verandah, there was a frieze of angels, elephants, devotees, the Bo tree, the wheel of law, etc., but they are hardly distinguishable now."
"To the west of the southern room and forming the right wing of the Pätālapura cave, there is a small room running east and west, with two doors and a verandah divided into two intercolumniations by a beavy pillar in the middle. At the corner where the verandah starts from the Pâtālapura range, there is an alto-relievo figure against the side pier, in imitation of the sentinel before the right wing of Queen's Palace. The cave is in a very ruinous condition now, and bears the appropriate name of Yamapura or Hades. The exterior view of a rihāra on the Udayagiri hill in Mr. Fergusson's Illustrations of the Rock-cut Temples of India represents the three caves of Vaikunthapura, Pātālapura and Yamapura."
"On the right hand pier of the centre door of the Vaikuotha range (the upper storey of the Mañcapuri cave in Mr. Banerji's description), there is a Pāli inscription in the Lāt (Asokan) characters, which is very carefully and deeply cut."
The text is based upon Major Kittoe's facsimile read and reproduced by James Prinsep in JAS 3, old series, Vol. VI, PI, LVII, Cave No. 8 called Vaikuntha Cave ; Rājendra Lāla Mitra's copy of Prinsep's transcript

Page #81

INSCRIPTION OF KHARAVELA'S CHIEF QUEEN
53
in his Antiquities of Orissa, Vol. II, p. 15, the Cave called Vaikunthapura; Alexander Cunningham's eye-copy reproduced in Corpus inscriptionum Indicarum, old series, Vol. I, Pl. XVII, Cave No. 9 called Vaikuntha; Bhagawanlal Indraji's eye-copy reproduced in Actes du Sixieme Congres International des Orientalistes, Part III, Sec. II; Lüders' List of Brahmi inscriptions in E I, Vol. X, Appendix, No. 1346, the Cave called Svargapura; Haridas Dutta's inked impression read and reproduced by R. D. Banerji in E I, Vol. XIII, Udayagiri and Khandagiri Cave-inscriptions, Pl. I, No. I, the Cave denoted by the expression "the upper storey of the Mañcapuri Cave"; and, above all, A. E. Caddy's cast preserved in the Indian Museum, Calcutta.

Page #82

Page #83

INSCRIPTION OF KHĀRAVELA'S CHIEF QUEEN
TEXT
8 * *
Arahamta-pasādā (na) m1 Kālimgā(na)m2 (sama)n[a]nam3 leṇam kāritam [‚—]rājino La]lâka (sa)? [1. 1] Hathisa(i) ha-sampa(n)ātasa 8 dhutuna9 Kalimgaca(kavatino)10 (siri-Khārave)lasa " [1.2] aga-mahisi(n)ā 12 ka[r]i-(tam)13 [.] [1.3]
1. The fourth letter is not at all noticed in Kittoe's facsimile. Cunningham and Indraji correctly read pasādānam. Sten Konow thinks that the reading may be pasādāya, a Dative form of pasada in the sense of prasada or "gift." Lüders contemplates a similar reading when he renders the word "in honour of." Banerji reads pasādāyam, treating it as a Locative form of pasada, which is taken to be the equivalent of prasada, "temple or palace." His argument in support of his reading pasādāyam is this: "There is a short vertical stroke attached to the right end of the horizontal base line of the letter," and "there are indications of a similar vertical stroke at the left end." But pasādāyam as a Locative form of pasada in the sense of prasada, "temple or palace" is unexpected in the language of an inscription of this kind, pasada being a neuter stem. Were Banerji's reading at all correct, pasadayam might have been treated as a Sandhi of pasada (=Sk. prasādāt," by the grace of") and ayam, a pronominal adjective qualifying lenam, although here, too, one would expect idam or imam instead of ayam. So far as I am able to ascertain, neither the plaster cast nor the original stone shows any trace of a short vertical stroke attached to the end of the left extension of the horizontal base line. The appearance of such a stroke above the right extension of the horizontal base line may be due to the mysterious work of a hornet. The letter, as we see it, is neither ya, nor pa, nor na. The reading of it as pa (pasādāpam) yields no intelligible meaning. The reading of it as na (Arahamta-pasādānam, "of the Arhata [Jaina] faith") yields an intelligible meaning. The letter with a short vertical stroke appearing above the right extension of the horizontal base line resembles the third letter after Hathisäha in 1. 2 which Prinsep reads no, Cunningham reads nã, and Indraji and Banerji read po. It resembles the third letter of päpunäti in I. 1 (1. 3), and the first letter of nagari in I. 4 (1. 5), see Pl. 1, JBORS, Vol. III, Part IV, p. 472.
2. Prinsep reads Kalimga; Cunningham, Kalimgānam. Indraji and Banerji correctly read Kalimganam.
3. The first letter is represented in Kittoe's fascimile as ya, and the second letter is not at all noticed. Cunningham is the first to correctly read samananam. In Dutta's estampage, the lower half of the letter ma is missing on account of abrasion.
4. Prinsep reads lona, rendering it "excavated." Cunningham and Indraji read lenam. Banerji correctly reads lenam.
(55)

Page #84

56
OLD BRAHMI INSCRIPTIONS
5. Prinsep reads kādatam, mistaking the second letter ri for da. It seems to appear, at first sight, as da. Cunningham is the first to correctly read käritam.
6. Prinsep reads rajino. Bat Kittoe's facsimile has clearly) the a-mark in the first letter.
7. Prinsep reads lasa......Kittoe's facsimile shows the traces of four letters which are clearly set forth in the eye-copies of Cunningham and Indraji. Both Cunningham and Indraji read Lalakasa. In Dutta's estampage, as noted by Banerji, "the 2-mark is more prominent in the second syllable than in the first."
8. Prinsep reads hethisahasampanot asa. Kittoe's facsimile has hethisehasam. Cunningham reads hathisāhanampanātasa; Indraji, Hathisāhasa papotasa, suggesting that the intended name must have been Hathisiha; Banerji, Hathisāhasa papotasa, suggesting that the intended name is either Hathisähasa or Hathisäha. The inscribed - surface of the original stone distinctly shows that the four letters after Hathisāha or Hathis[i]ha are sam-pa-(n)a-ta-sa. It showed the same even a century ago when Kittoe prepared a facsimile of it. It is certain that the third letter may be read as (pā but by no means as (po. When the fourth letter is a clear ta, we need not presume it to be a mysterious transformation from ga, in order to make out sampanāgasa (sarpanagasya) after Hathisāha or Hathis[i]ha. For a similar compound in which a noun is preceded and followed by two adjectives qualifying it, cf. sakata-samaņa-suvihitā in I. 15.
9. Prinsep reads only the first letter as ya. Kittoe's facsimile has ya, una. Cun. ningham reads ca tino. Indraji and Banerji correctly read dhutunā. Jayaswal fails to find out in Dhutină the name Dhuti( = Sk. Dhreti) of King Lālâka's daughter.
10. Cunningham reads only the first letter as ca. Indraji proceeds so far as cakavati. Banerji hits the goal by reading cakavatino.
11. Prinsep reads the last three letters as velasa. Indraji reads the last five letters as Khāravelasa. Banerji reads all the seven letters as siri-Khāravelasa.
12. Prinsep reads mahi pita ; Cunningham, mahi pida. Indraji correctly reads mahi. sinā.
13. Prinsep reads kadā ; Cunningham, kāda; Dutta's estampage bears out Banerji's reading kāritam.

Page #85

www. kobatirth.org
TEXT AS IT READS IN THE ORIGINAL
अरहंत पसादा(नं) कालिंगानं समनानं लेणं कारितं - राजिनो ल[] लाकस हथिस[ ] हसंप-(नातस धुतुना कलिंगचकवतिनो सिरिखारवेलस अग-महिसिना कारितं [1]
TEXT AS IT READS IN PĀLI
अरहन्त-पसादानं कालिङ्गानं समणानं लेणं कारितं [.] राजिनो लालकस्म हथिस[ ] इ-सम्पबत्तस्म धौतुना कलिङ्गचक्कवत्तिनो सिरि-खारवेलम्म अम्गमहिसिना कारितं []
TEXT AS IT READS IN SANSKRIT अर्हत्सु प्रसन्नेभ्यः कालिङ्गेभ्यः श्रमणेभ्यः लयनं कारितं[,-]राज्ञो लालार्कस्य हस्ति[f] संह-सम्पत्रात्मनः दुहिता कलिङ्गचक्रवर्तिनः श्रो-खारवेलस्य अग्रमहिष्था कारितं []
TRANSLATION
The cave has been made for the Kalinga recluses 1 of the Arhata (Jain) faith,2-caused to be made by the chief queen of His Graceful
Majesty Khāravela, the (King) Overlord of Kaliiga,-by the daughter of the high-souled 3 King Lalarka 4 Hastisaha or Hastisimha.
must translate: "The cave of the
1. Treating samanānam as a Genitive case, one Kalinga recluses of the Arhata (Jain) faith."
(57)

Page #86

58
OLD BRAHMI INSCRIPTIONS
2. Adopting the reading Arahamta-pasādāya, one may translate with Sten Konow and Lüders: "In honour of;" and adopting Arahamta-pasadayam, one may translate: "By the grace of the Arhat or Arhats, this (cave)."
#
3. The word sampanata is equated with the Sk. sampannatma, which literally means "self-endowed," "one whose self has prospered by the attainment of desired objects." For the meaning sampannatma, cf. Chandogya Upanisad, VIII. 3: Yam yam antam abhikamo bhavati, yam kamam kamayate so'sya samkalpäd eva samuttisṭhati, tena sampanno mahiyate. Sankara explains sampanna thus: abhipretârthapraptya ca sampanno.
4. L[a]laka or Lalarka means "glorious like the rising sun," apparently as an earlier synonym of Baladitya. For the use of Lalaka as a biruda, cf. Yasalalaka-Tissa occurring in the Mahāvamsa (XXXV. 50) as the name of a king of Ceylon. But it may not be going too far to treat Lalaka. as a local epithet signifying that Hastisäha or Hastisimha was the sun of Lala."

Page #87

No. III
INSCRIPTION OF KADAMPA OR KUDEPA
RELATING TO
THE PATALAPURA CAVE

Page #88

Page #89

INSCRIPTION OF KADAMPA-KUDEPA
The following record is incised on a raised band between the third and fourth doors from the left' in the verandah of the lower storey of the Mañcapuri group of three caves which are designated by Mr Fergusson and Rājā Rājendra Lāla Mitra as Vaikunthapura, Patalapura and Yamapura. There is every reason for believing that the inscription was meant to record the excavation and dedication of the second cave of this interesting group by siri-Kadampa or siri-Kudepa, the Sovran Lord of Kalinga, who may have been either the contemporary or immediate successor of siri-Khāravela.
The text is based upon Major Kittoe's facsimile read and reproduced by James Prinsep in JASB, old series, Vol. VI, Pl. LVII, Cave No. 6 miscalled Manikpura; Rajendra Lala Mitra's copy of Prinsep's transcript in his Antiquities of Orissa, Vol. II, p. 16; Alexander Cunningham's hand-copy reproduced in Corpus Inscriptionum Indicarum, old series, Vol. I, Pl. XVII, Cave No. 7 miscalled Manikpura; Bhagwanlal Indraji's eye-copy reproduced in Actes du Sixieme Congres International des Orientalistes, Part III, Sec. II; Lüders' List of Brahmi inscriptions in E I, Vol. X, Appendix, No. 1347, Cave called Udayagiri Mañcapuri-gumpha; Haridas Dutta's inked impression read and reproduced by R. D. Banerji in EI, Vol. XIII, Udayagiri and Khandagiri Cave Inscriptions, Pl. I, No. II, Cave denoted by the expression Mañcapuri Cave-Lower Story, Front Wall"; and, above all, A. E. Caddy's cast preserved in the Indian Museum, Calcutta.
(61)
(6

Page #90

Page #91

INSCRIPTION OF KADAMPA-KUDEPA
TEXT
Airasa' mahārājasa Kalimgâdhipatino2 mahā(megha). vẫha(nasa)* Kadampa - sirino" lenam" [:]
1. Prinsep, Rajendra Lāla Mitra and Indraji read Verasa, Banerji reads Kharasa, which is impossible Cunningham, Jayaswal and Sten Konow read Airasa. I am strongly inclined to read the royal epithet also as Verisa. The first letter bears the closest possible resemblance to the first letter of the name Vaihidari in the Pabhosā inscriptions of Āsādhasena edited by Führer in EI, Vol. II.
2. Prinsep reads Kalimgadhipatano; Cunningham Kalimgadhipatino. Indraji is the first to correctly read Kalimgadhipatino.
3. Prinsep reads only the first letter as ma. Cunningham reads mahamegharāhana; Indraji, mnhameghavahana; Banerji correctly, mahameghavāhanasa.
4. Prinsep reads Kadepa. Cunningham reads the last two letters as depa. Indraji and Lüders read the name as Vakadepa. Banerji reads Kūdepa. In Dutta's estampage and in the cast one may read the first letter as Ku or kú, but I do not see how the second letter might be read as de. The second letter is unmistakably de or dam.
5. Banerji is perfectly justified in reading sirino as it appears in Dutta's estampage or in the cast. But there is nothing to prevent one reading sirino with Prinsep and others.
6 Prinsep and Rājendra Lāla Mitra wrongly read lonam.
( 63 )

Page #92

www. kobatirth.org
TEXT AS IT READS IN THE ORIGINAL
ऐरस (वेरस इति वा ) महाराजस कलिंगाधिपतिनो महामेघवाहनस कदंपसोरोनो (कूप - सिरिनो इति वा ) लेणं [1]
TEXT AS IT READS IN PALI
अरिस्स (वीरस्म इति वा ) महाराजस्स कलिङ्गाधिपतिनो महामेघवाहनस्स कदम्ब - सिरिनो ( कूदेव - सिरिनो इति वा ) लेणं [ i ]
TEXT AS IT READS IN SANSKRIT
यस्य ( वारसा इति वा ) महाराजस्य कलिङ्गाधिपतेर्महामेघवाहनस्य श्री-कदम्बस्य ( श्री कुदेवस्य इति वा ) लयनं [1]
TRANSLATION
The cave (which is an excavation) of His Lordly Graceful Majesty Kadampa or Kudepa, the Great King, the Sovran Lord of Kalinga, whose vehicle is Mahāmegha (the great cloud-like state-elephant ).
(64)

Page #93

No. IV
INSCRIPTION OF VADUKHA OR VARIKHA
RELATING TO
THE YAMAPURA CAVE

Page #94

Page #95

INSCRIPTION OF VADUKHA-VARIKHA
The following inscription, consisting of one line, 'is incised on the right wall of the verandah of the lower story, to the right of the entrance to the righthand side-chamber of the main wing' of the Mañcapuri group of three caves designated Vaikunthapura, Pātālapura and Yamapura by Mr. Fergusson and Rājā Rajendra Lāla Mitra. There is every reason for believing that the inscription was meant to be a record of the excavation and dedication of the third cave of this group by Prince Vadukha or Varikha.
The text is based upon Major Kittoe's facsimile read and reproduced by James Prinsep in JAS3, old series, Vol, VI, P. LVII, Cave No. 7 miscalled Manikpura cave; Rājendra Lāla Mitra's copy of Prinsęp's transcript in his Antiquities of Orissa, Vol. II, p. 16; Alexander Cunningham's eye-copy reproduced in Corpus Inscriptionum Indicarum, old series, Vol. I, Pl. XVII, Cave No. 8 miscalled Mānikpura Cave; Bhagwanlal Indraji's eye-copy reproduced in Actes du Sixieme Congres International des Orientalistes, Part III, Sec. II; Lüders' List of Brāhmi Inscriptions in EI, Vol. X, Appendix, No, 1348, Cave called Udayagiri Mañcapuri-gumpbā; Hari Das Dutta's inked impression read and reproduced by R. D Banerji in EI, Vol. XIII, Udayagiri and Khandagiri Cave Inscriptions, Pl. I, No. III ; and, above all, A. E. Caddy's cast preserved in the Indian Museum, Calcutta.
( 67 )

Page #96

Page #97

2
INSCRIPTION OF VADUKHA-VARIKHA
TEXT
Kumāro-Vaḍukhasa 2 lenam [.]
3
1 Banerji rightly points out that the a-stroke in the second letter is added to the middle instead of the top. This phenomenon may be observed also in some instances afforded by the Hathi-Gumpha inscription. The o-stroke in the third letter is quite distinct. It is difficult to decide all at once whether the retention of the singular masculine Nominative case-ending in Kumaro, the first word of a compound, is a dialectical peculiarity or a mistake of the mason-engraver. In one of the Barhut Jataka-labels we have Sujatogahuto-Jataka, but in another, Miga-Jātakam.
(69)
Prinsep wrongly reads Vattakasa; Cauningham, Vaddakasa or Vaḍukasa. Indraji is the first to read Vadukhasa. Regarding the second letter, Banerji remarks: "The medial u in Vadukha is very small in size but quite distinct"; and regarding the third letter, he observes that here the kha" has neither a triangle nor a circle at its base." It is not impossible that the intended name is Varikha, the u-stroke being only an accidental chisel mark. It is remarkable how Prinsep and Cunningham have mistaken ri for da in the inscription of Kharavela's chief queen, the Inscription No. II.
Prinsep and Rajendi Lala Mitra wrongly read lonam.

Page #98

TEXT AS IT READS IN THE ORIGINAL
कुमारो वड़खस ( वरिखस इति वा ) लेणं [1]
TEXT AS IT READS IN PALI
कुमार-वलखस्स (वरिक्खस्स इति वा ) लेणं [1]
TEXT AS IT READS IN SANSKRIT
कुमार-वडुखसा ( वरेक्षमा इति वा ) लयनं [1]
TRANSLATION
The cave (which is an excavation) of Prince Vadukhai or Varikha.2
i Vadukha is apparently a pet name of the prince like Tivala in Asoka's "Queen's Edict."
2
Varikha may be equated with Vareksa, and taken to mean "one of excellent look."
(70)

Page #99

No. Y
INSCRIPTION OF VADUKHA OR VARIKHA
IN THE CHOȚA-HATHIGUMPHĀ

Page #100

marTV

Page #101

INSCRIPTION OF VADUKHA-VARIKHA
The following inscription noticed for the first time by Mr. A. E. Caddy, when he was taking casts of the inscriptions of this class for the Indian Museum, Calcutta, “consists,” says Mr. R. D. Banerji, “of a single line, very much mutilated, on the outer face of the tympanum of the arch over the doorway."
The text is based upon Hari Das Datta's inked impression read and reproduced by R.'D. Banerji in EI, Vol. XIII, Udayagiri and Khandagiri Inscriptions, Pl. II, No. IX, the cave called Choța-Hāthigumpbā.
I regret that Caddy's cast of the inscription, which Mr. Banerji saw in the Calcutta Museum, is no longer to be seen there.
( 73 )

Page #102

Page #103

INSCRIPTION OF VADUKHA-VARIKHA
TEXT
? ? kha ? ? ? sa ' loņam [-]
1. Banerji doubtfully reads the first three letters as Agikha. The first letter might be read as Va or Ta, but never as A. The second letter looks like du or ri, even like ta or sa, but never like gi. The letter after kha might be made out as ka or k[u]. The next letter which was probably mã has been split up into two parts, one appearing as dā and the other as na. The letter immediately before sa seems to have been ra. I shall not be surprised if the record was meant to rend Vadukha (Varikha) kumāraga lonam [.]
(75)

Page #104

TEXT AS IT READS IN THE ORIGINAL
?? ? ??? H
U
[1]
var
TEXT AS IT READS IN PALI ?? ??? g adi [1]
TEXT AS IT READS IN SANSKRIT ?? C ??? ero [1]
TRANSLATION
The cave (which is an excavation) of Prince Vadukha or Varikha (?).
(76)

Page #105

No. YI
INSCRIPTION OF CŪLAKAMMA OVER THE DOORWAY OF THE SARPA-GUMPHĀ OR
SNAKE-CAV

Page #106

Page #107

INSCRIPTION OF CULAKAMMA The following “inscription, consisting of one line, is incised,” says Mr. R. D. Banerji, "over the doorway of the Sarpagumphā, which is very close to the Badā-Hāthigumphā (No. 14 of the plan published in the Bengal District Gazetteer, Puri)."
Regarding the cave which is a notable excavation on Udayagiri, Rājä Rājendra Lāla Mitra in his Antiquities of Orissa, Vol. II, p. 30, says:
"Retracing our steps now to the west of the Hathi-Gumphā, the first cave worthy of notice occurs on the most protruding ledge of the fracture between the Udayagiri and the Khandagiri hills. It is remarkable for having in front, on the top of its entrance, a rude carving of the hood of a three-headed cobra, whence its name Ajāgara-Gumpha--ajāgara, a serpent able to swallow (gara) a goat (ajā). Under the hood of the serpent, the cave is cut in the form of a cube of 4 ft. with a door just large enough to admit a man crawling in, and framed in the usual style."
For a similar authentic description of Sarpa-Gumphā, see Mr. Mano Mohan Ganguly's Orissa and Her Remains-Ancient and Mediaval, p. 46, where one reads :
" It is a small cave consisting of a single cell and a verandah ;. the latter having in the front tympanum of the door, the carving of the hood of a three-beaded serpent, and hence the name. The level of the verandah is higher than the ground in front. Drs. Fergusson and Burgess (in their Cave Temples of India, p. 697), have taken the Tiger and the Serpent Caves to be the oldest sculptured caves in the hills. This cave is important for containing an inscription, which, however, does not throw any light on the probable date of its construction."
The text is based upon Major Kittoe's facsimile read and reproduced by James Prinsep in JASB, old series, Vol. VI, Pl. LVII, Cave No. 1 called Snake-Cave ; Rājendra Lāla Mitra's copy of Prinsep's transcript in his Antiquities of Orissa, Vol. II, p. 30; Alexander Cunningham's handcopy reproduced in Corpus Inscriptionum Indicarum, old series, Vol. I, Pl. XVII, Cave No. 2 called Snake-Cave; Lüders' List of Brāhmi Inscriptions in EI, Vol. X, Appendix, No. 1349: Hari Das Dutta's inked impression read and reproduced by R. D. Banerji in EI, Vol. XIII, Udayagiri and Khandagiri Inscriptions, Pl. I, Vol. IV, Cave called Sarpagumphā; and, above all, A. E. Caddy's cast preserved in the Indian Museum, Calcutta.
(79)

Page #108

Page #109

INSCRIPTION OF CŪLAKAMMA
TEXT
Cūlaka[m]masa' Kothājeyā’ ca (pasādo] [ . ]
1. In Dutta's inked impression, the letter a appears to be le, the letter ka to be ko, and the letter ma to be me or mo. These vowel-marks must at once ne dismissed as chisel-marks or fissures. Kittoe need not be blamed for having represented kammasa as kammesa. Cunningham reads Culakammasa. But the ū-mark attached to the first letter is prominent. Sten Konow says, " The first word looks like Cūlakemesa, owing to the abrasions on the inscribed surface. The estampage seems to read Oulakremesa."
2. Prinsep and Rajendra Lala Mitra read kothājayā. Cunningham reads kothājaya ; Läders, kothajeyā; and Banerji, kothajeyā.
3. The occurrence of the Copulative Conjunction ca is indicative of the loss of another word, which seems to be no other than pasado=päsūdo.
(81)
11

Page #110

TEXT AS IT READS IN THE ORIGINAL
2 [] ## often a [UHET] [1]
TEXT AS IT READS IN PĀLI चूलकम्मस्स कोठाजयया च पासादो [1]
TEXT AS IT READS IN SANSKRIT CFÅNT 01877279 [ateic:] [1]
TRANSLATION
The facader and the unconquerable2 cells (which are the meritorious works) of Cũlakamma (Kșudrakarma). S
1. Banerji agrees with Prinsep, Rajendra Lāla Mitra and Cunningham in rendering pasāda (= püsāda) 'temple or palace.' But it seems that a temple or palace is too high an honour to be accorded, in all cases, to any part of a cave-construction. What seems to have been meant by pasāda or pāsāda here is a frontal outer construction consisting of a roofed, pillared and ornamented verandah and having the appearance of a palace. Pasāda or Pāsāda is the same word in sound and meaning as façade.
2. Kothâjeyā is treated as a sandhi joining kotha and ajeyı. Here'unconquerable,' as suggested by Prinsep, rather means 'impregnable, than 'unequalled ' or 'ungurpagsable.' Banerji observes : " Kotha and jeyā have been separated by Dr. Lüders, most probably because the compound kothājeyā) is followed by the conjunction ca, kotha, Sk. Kostha, is still in use in modern vernaculars to denote a brick or stone-built house or chamber, or even a fort. Jeya may have a technical meaning and may denote the verandah or some other part, wbile kotha denotes the main chamber. But it is also possible to take it in another way, in which there is a randhi between this word and ajeyā, unconquerable, qualifying kotha or kotha. The only difficulty is the use of the conjunction." The guggested difficulty can be removed, if it be assumed that the conjunction was followed by a word like pasado or pasado. It is quite possiblo that jeya is a separate technical term denoting some kind of building work. Cf, jeyakah kārā pitah in the inscription No. V from Northern Gujarat, EI, Vol. II. The term kothājeyā may be taken to mean either a cellular shed' or 'chambers with arched entrances.'
3. According to Prinsop, Rajendra Lala Mitra and F. W. Thomas, Culakammar Cudākarma, which is apparently absurd. I agree with Liders in equating the Prakrit form of the name with the Sk. Kșudrakarma.
( 82 )

Page #111

No. VII
INSCRIPTION OF CŪLAKAMMA. IN PĀVANA-GUMPHĀ OTHERWISE CALLED
HARIDĀS-GUMPHĀ.

Page #112

Page #113

INSCRIPTION OF CŪLAKAMMA
The following “record," says Mr. R. D. Banerji, “consists of a single line and is incised over one of the three entrances to the main chamber of the cave from the verandah. The characters belong to the first century B. C. and are distinctly later in form than those of the Mañcapuri inscriptions (meaning Nos. II-IV)."
Of the cave itself which is an excavation on the Udayagiri bill, we bave the following useful information from Rājā Rājendra Lala Mitra's Antiquities of Orissa, Vol. II, p. 30:
“Immediately to the north of the Elephant-Cave (Hāthi-Gumpbā) there is a small excavation which is known under the name of PāvanaGumpbā or the 'Cave of Purification. It is of no importance whatever as a work of art."
The text is based upon Major Kittoe's facsimile read and reproduced by James Prinsep in JASB, old series, Vol. VI, Pl. LVII, Cave No. 5 called Pawan-Cave; lājendra Lāla Mitra's copy of Prinsep's transcript in bis Antiquities of Orissa, Vol. II, p. 30; Alexander Cunningham's band-copy reproduced in Corpus Inscriptionum Indicarum, old series, Vol. I, Pl. XVII, Cave No. 6 called Pawan-Cave; Lüders' List of Brāhmi Inscriptions in EI, Vol. X, Appendix, No. 1353, Cave called Haridāsgumpha; Hari Das Dutta's inked impression read and reproduced by R. D. Banerji in EI, Vol. XIII, Udayagiri and Khandagiri Inscriptions, Pl. II, No. VI, Cave called Haridās Cave; and, above all, A. E. Caddy's cast preserved in the Indian Museum, Calcutta,
(85)

Page #114

Page #115

INSCRIPTION OF CŪLAKAMMA
TEXT
Cūlaka [m]masa è pasāto 2 kothāje(yā) ca + [.]
1. Prinsop and Rajendra Lāla Mitra read Culakamasa. Cunningham reade Culakrammasa ; Banerji Cülakramasa. I cannot conceive of the occurrence of a letter, such as kra, in this series of Jain Cave inscriptions.
2. Prinsep und Rājendra Lala Mitra read paseta. Kittoe's facsimile has pasetā, Cunningham and Banerji read pasāto. The correct reading must be either pasāto or pasātā, pasāto deserving to be treated as a noun, which is the equivalent of pasādo (or pāsādo), and pasātā deserving to be treated as an adjective qualifying kotha.
3. Prinsep and Rajendra Lāla Mitra read Kothâja(yā). Cunningham reads koțhäja. Lüders reads kothājeya; Banerji Kothajeyā.
4. If it be that a word is effaced after the conjunction ca, the reading of the inscription must be this : Cūlaka[m]masa pasāta kothāje(yā) CA (pasādo] [.], the word posāta being an adjective qualifying kotha.
(87)

Page #116

TEXT AS IT READS IN THE ORIGINAL
qan[*]ag qui
TITOLO []
TEXT AS IT READS IN PALI चुलकम्मस्म पासादो कोठाजेयया च [1]
TEXT AS IT READS IN SANSKRIT
क्षुद्रकर्मस्य प्रासादः कोष्ठजेयश्च [1]
TRANSLATION
The facade and the unconquerable cells (which are the meritorious works) of Kșudrakarma,
1. In the cast, as well as in Hari Das Dutta's estampage, the word is pasāto. There is absolutely no ā-mark in the first letter. If pasāto be a phonetic equivalent of pasādo, and the clear estampage of this inscription guarantee the certainty of the reading pasāto, it is to be seen whether pasāto or pasādo is the same word as pāsādo or prūsādah, or it is a word to be equated with a word like pracchadaḥ, meaning a projecting cover, that is, a verandah with an overhanging roof. Sten Konow suggests : "If we should read pasādo (or pasāto), this word is perhaps Sk. prasāda, a gift." To my mind, the use of the word in the sense of a gift is altogether out of the question, Such a word with such a meaning cannot fit in with the context.
(88)

Page #117

No. VIII
INSCRIPTION OF KAMMA AND KHĪNĀ
TO THE LEFT OF THE
DOORWAY OF THE SARPA-GUMPHĀ OR SNAKE-CAVE

Page #118

Page #119

INSCRIPTION OF KAMMA-KHINA
The following "record,” says Mr. R. D. Banerji, “consists of two lines and the characters used in it are about a century later in date than those of the other epigraph in this cave (Sarpagumphā). The characters belong to the first century B.C."
The text is based upon Major Kittoe's facsimile read and reproduced by James Prinsep in JASB, old series, Vol. VI, PI LVII, Cave No. 2 called Snake-cave; Rājendra Lāla Mitra's copy of Prinsep's transcript in his Antiquities of Orissa, Vol. II, p. 30; Alexander Cunningham's hand-copy reproduced in Corpus Inscriptionum Indicarum, old series, Vol. I, Pl. XVII, Cave No. 3 called Snake-Cave; Lüders' List of Brāhmi Inscriptions in EI, Vol. X, Appendix, No. 1350; and Hari Das Dutta's inked impression read and reproduced by R. D. Banerji, in EI, Vol. XIII, Udayagiri and Khandagiri Inscriptions, Pl. II, No. I, Cave called Sarpagumphā.
( 91 )

Page #120

Page #121

INSCRIPTION OF KAMMA-KHĪNĀ
TEXT
Kammasa' (kothā) ca Khi-sl 1]-ặ(ā)ya. ca pasādo“ [.] [1 2].
1. Prinsep wrongly reads Kamesa. Kittoe's facsimile has Kamasa. Cunningham reads Kamase; Banerji correctly, Kammasa.
2 & 3. Prinsep reads rikhinaya; Cunningham, ra...khi Naya(co); Banerji, Halakhinaya. Dutta's estampage yields, at first sight, the reading suggested by Banerji : Halakhinaya= Sk. Slakanāyah (F. W. Thomas), = doubtfully Harakina (Lüders). But it is certain that the letter after khi is na or n[a]. Examining Dutta's estampage closely, I detect that the second lettor is tha.
4. Kittoe's faosimile has ce, and Cunningham's hand.copy co. The o-mark appears also in Datta's inked impression, which is undoubtedly an abrasion. Banerji correctly reads ca.
5. Prinsep, Rajendra Lāla Mitra and Cunningham read pasāde, which is incorrect. Kittoe's facrimile has pasādo. Datta's estampage, too, yields the reading pasādo. Sten Konow remarks: "The plate (reproducing Dutta's estampago) has, however, pāsāde though the a-stroke is indistinct and perhaps erased."
( 93 )

Page #122

TEXT AS IT READS IN THE ORIGINAL कंमस कोठा च खोणाय च पसादो [1]
TEXT AS IT READS IN PALI कम्मरम कोठा च खोणाय च पासादो [1]
TEXT AS IT READS IN SANSKRIT कर्मस्य कोष्ठश्च क्षीणायाश्च प्रासादः [0]
TRANSLATION
The cells (which are excavations) of Karma and the façade (which is a meritorious work) of Khiņā.1
1. Accepting the reading Halakhinaya, one must render the record: "The cells (or chambers which are excavations) of Karma and Slakana.
(94)

Page #123

No. IX
INSCRIPTION OF BHŪTI
IN THE VYAGHRA-GUMPHA OR TIGER-CAVE

Page #124

Page #125

INSCRIPTION OF BHŪTI
The following “record is incised,” says Mr. R. D. Banerji, “on the outer wall of the inner chamber of the Bagh or Tiger cave (No. 15 of the clan published in the Bengal District Gazetteer, Puri). It consists of two lines. The characters used are as old as the inscriptions in the Mañcapuri cave and belong to the second century B.C."
Of the Băgh-Gumphā or Tiger-cave which is an excavation on the Udayagiri, we have the following interesting description in Rājā Rājendra Lāla Mitra's Antiquities of Orissa, Vol. II, p. 31 :
"In size it is similar to the (Alakāpura cave), but its exterior is cut into the shape of a tiger's head. The distended jaws of the animal form the verandah, and the entrance to the cell occupies the place of the gullet. The head is remarkably well-formed, and the chiselling is excellent. By the right of the entrance (we have) a short inscription in the Lāț (Asokan) character."
For a similar description of the cave, see Mr. Mano Mohan Ganguly's Orissa and Her Remains—Ancient and Medieval, pp. 45-46, where one reads : “ The Tiger cave, on the west of the Hāti Gumphā is a small one consisting of a verandah, and a single cell. It is situated higher up the hill than the Ganesa or the Hāti Gumphā. The roof of the front verandah is formed by the upper jaw of the animal. In this cave, the eye, nose and upper jaw of the tiger have been represented ; the two canine teeth on the two sides of the incisors have been shown; the number of incisors is greater than what is noticed in the animals of the feline species. The cell is provided with one door flanked by pilasters resting on raised platforms and surmounted by a semi-circular arch-band."
For a short description of the cave, see Dr. W. W. Hunter's Statistical Account of Bengal, District of Puri and the Orissa Tributary States, p. 73.
The text is based upon Major Kitto e's facsimile read and reproduced by James Prinsep in JAS B, old series, Vol. VI, Pl. LVII, Cave No. 5, called Tiger-cave ; Rājendra Lāla Mitra's copy of Prinsep's transcript in his Antiquities of Orissa, Vol. II, p. 31; Alexander Cunningham's handcopy reproduced in Corpus Inscriptionum Indicarum, old series, Vol. 1, Pl. XVII,
( 97 )
13

Page #126

98
OLD BRĀHMI INSCRIPTIONS
Cave No. 4, called Tiger-cave; Lüders' List of Brāhmi Inscriptions, in EI, Vol. X, Appendix, No. 1351; Hari Das Dutta's inked impression read and reproduced by R. D. Banerji in EI, Vol. XIII, Udayagiri and Khandagiri Inscriptions, Pl. II, No. 7, Cave called Bāgh or Tiger Cave ; and, above all, A. E. Caddy's cast preserved in the Indian Museum, Calcutta.
The inscription is interposed between two symbols, a tree-symbol representing a vľkşa-caitya or woodland-shrine and marking the commencement, and a Svastika marking the close.

Page #127

INSCRIPTION OF BHŪTI
TEXT
Nagara'-akhadaņsa-[1 1]-sa ? Bhūtinos leņam * [-] [1 2].
1. Prinsep, Rajendra Lala Mitra and Cunningham read Ugara. Lüders and Banerji correctly read Nagara.
2. Prinsep and Rājendra Lāla Mitra wrongly read avedasa, aveda being taken to mean an anti-Vedist, and the Genitive cage-ending sa being regarded as the initial of the donor's name which they read Sasupi. Cunningham reads akhadasa and Banerji, akhadamsa, both agreeing with Prinsep and Rājendra Lāla Mitra in treating the sixth cage-ending sa as the initial of the donor's name which they read Sabhūti (Subhūti). Lüders correctly reads akhadamsasa. Banerji's argument is : " There is plenty of space after the last letter of the first line and so it cannot be said that the possessive case-ending had to be incised in the lower line for want of space. Generally a mason does not mutilate words when there is no dearth of space." Sten Konow's counter-argument is : " The two lines have been kept of the same length, and that is apparently the reason why the termination of akhadamsasa has been written in L 2." I find that, in spite of there being no dearth of space, the mason has written, in one of the Barhut inscriptions, the sixth case-ending in L 2 (Hultzsch's Bharhut Inscriptions in I. A. Vol. XXI, No. 90):
(1) Bhadata-Budharakhitasa Satupadāna(2) 8a dānam thabho.
3. Prinsep and Rājendra Lāla Mitra road Sasuvin ; Cunningham and Banerji, Sabhūtino. Lüders correctly reads Bhūtino. Caddy's cast, which I have carefully examined, clearly bears out Lüders reading.
4. Prinsep and Rajendra Lala Mitra wrongly read lonam.
( 99 )

Page #128

1
TEXT AS IT READS IN THE ORIGINAL
नगर- अखदंसस भूतिनो लेणं [1]
TEXT AS IT READS IN PALI
नगर-अक्खदस्सस्स भूतिनो लेणं [1]
TEXT AS IT READS IN SANSKRIT
नगर-अक्षदर्शसा भूतेः लयनं [1]
TRANSLATION
The cave (which is an excavation) of the Town-judge1 Bhuti.
Nagara-akhadamsa is obviously the same official designation as Nagala. Viyohalaka or Mahāmāta-Nagalaka in Aśoka's First Separate Rock Edict, and Nagarika or Nagarika. Mahamätra in the Arthasästra of Kautilya-Kautalya, Bk. II, Ch. 36 and Bk. IV, Ch. 6.
100)

Page #129

No. X
INSCRIPTION OF NĀKIYA IN THE JAMBEŚVARA CAVE

Page #130

Page #131

INSCRIPTION OF NĀKIYA
The following "record is incised,” says Mr. R. D. Banerji, “over one of the entrances to the inner chamber of the Jambesvara Cave (No. 16 of the plan published in the Bengal District Gazetteer, Puri). The characters of the inscription are of the same age as those used in the Mañcapuri inscriptions." The cave itself is one of the excavations on the Udayagiri bill.
The text is based upon Major Kittoe's facsimile read and reproduced by James Prinsep in J A S B, old series, Vol. VI, Pl. LVII, Cave No. 4, termed a nameless cave ; Alexander Cunningham's hand-copy reproduced in Corpus Inscriptionum Indicarum, old series, Vol. I, Pl. XVII, Cave No, 5, termed a nameless cave; Lüders' List of Brāhmi Inscriptions in EI, Vol. X, Appendix, No. 1352 ; and Hari Das Dutta's inked impression read and reproduced by R. D. Banerji in EI, Vol. XIII, Udayagiri and Khandagiri Inscriptions, Pl. II, No. 8, Cave called Jambeśvara ; and, above all, A. E. Caddy's cast preserved in the Indian Museum, Calcutta.
(108)

Page #132

Page #133

INSCRIPTION OF NĀKIYA
TEXT
Mahāmadasa · Bāriyāya ? Nākiyasa : leņa[m] * [.]
1. Prinsep reads Māpāmadati; Cunningham, Mapāmadāsa; Laders, Mahāmadasa ; Banerji, Mahāmadāsa. Banerji is of opinion that the à in dā is superfluous. I think that the ā-stroke is an appearance rather than a reality.
2. Prinsep reads bākāya (Ya); Cunningham, bāniyāya ; Lüders, bariyāya; Banerji, too, bāriyaya, treating it as a mistake for bhariyāya. I take Bāriya to be the name of a locality, the native place of the donor, or the name of a lady donor.
3. Princep reads (Ya)nākiyasa; Liders, Nākiyasa, Cunningham and Banerji correctly read Nakiyasa. Näki, according to Lüders, or Nakiya, according to Banerji, is the name of the wife of Mahamada. Had it been so, there would have been a feminine Genitive case-ending āya in the declension of Nāki or Näkiya. But we have, instead of Nākiyāya, a masculine Genitive form Nākiyasa. I take Nākiya to be the name of the donor, Bāriya to be his native place, and Mahamada to be his official designation, or all the three of them to be personal names.
4. Prinsep wrongly reads long.
( 105 )
14

Page #134

TEXT AS IT READS IN THE ORIGINAL
महामदस बारियाय नाकियस लेण[] [0]
TEXT AS IT READS IN PALI
महामत्तस्स ( महामदस्स इति वा ) बारियाय नागितस्स लेणं [1]
TEXT AS IT READS IN SANSKRIT
AGIHIABI ( AG(AZ81 efa aı ) arûiai; ziqaar gaaq [1]
TRANSLATION *
2
The cave (which is an excavation) of the High Functionary Nakiya " of Bariyā.
* Another possible rendering is :
"The cave (which is an excavation) of Mahamada, Bariya and Nakiya."
1. Both Lüders and Banerji take Mahamada to be the personal name of a man whose wife Nak or Nakiya dedicated the cave in question. Mahamada seems to me to be the same official designation as the Pāli Mahamatta or the Sk. Mahamatra, which literally means "Great-measure" or "Magnate."
(106
1
2. Nakiya is evidently the same name as the Pali Nagita.
3. I am unable to identify the locality. It must be some place, a village or a town in Kalimga.

Page #135

No. XI
INSCRIPTION ON THE ARCHITRAVE OUTSIDE THE ANANTA-GUMFHĀ

Page #136

Page #137

ANANTA-GUMPHĀ INSCRIPTION
The following “inscription is incised,” says Mr. R. D. Banerji, “on the architrave outside, between the left antæ and the first pillar. The characters of this inscription are certainly later than those used in the inscription in Tatwa Cave, No. 2. No other notice of this inscription has been published except that which has appeared in Mr. Mon Mohan Chakravartti's Notes on the Remains in Dhauli and in the Caves of Udayagiri and Khandagiri.”
The following description of the cave with which the inscription is associated is an extract from Mr. Mano Mohon Ganguly's Orissa and Her Remains--Ancient and Mediæval, pp. 56-60 :
“The Ananta-Gumphā (which is the most important cave on the Khandagiri Hill) consists of an ante-chamber (24'*7') having a covered verandah 26' x 7' in front. The ceiling of the chamber at the rear is curved, the rise of the arch or curve being about a foot; originally there were four doors leading to the room ; these have been reduced to two doors, and one window. The doorways are surmounted by circular arches ending in borizontal bands at the springing. Two fillets of the circular band enclose ornamental figures. The horizontal friezes resemble those noticed in Rāni Nür and Ganesa Cave. The crown of the arches is formed by the interlaced tails of serpents." "A long line of Buddhist rails runs over the arches unlike those in Rāṇi Nür and Ganesa cave broken at intervals by battlemented pyramids, the like of which we notice in the Bharhut sculptures. On the back wall of the ante-chamber are carved the characteristic Bauddha symbols of Swastika, Trisula, etc., and also a standing figure of Buddha (?). The tympana of the arches are carved with representations of various scenes. The scene of Gaja-Lakşmi... is exquisitely beautiful...... The pilasters by the sides of the doorways have striking features characterised by animal capitals and a profusion of sculptures not noticed elsewhere in the locality."
“We agree with Drs. Fergusson and Burgess in detecting similarity between the sculptures of the Ananta-Gumphā and those of the Bharhut Stūpa, built sometime between the 2nd and 3rd cent. B.C.......the probable period of the excavation of the cave cannot be earlier than the third
( 109 )

Page #138

110
OLD BRAHMI INSCRIPTIONS
century B.C., and not second or third century A.D. as fixed by Beglar in the Archæological Survey of India, Vol. XIII (p. 81)."
The text is based upon Hari Das Dutta's inked impression read and reproduced by R. D. Banerji in EI, Vol. XIII, Udayagiri and Khandagiri Inscriptions, Pl. II, No. XI, the cave called Avantagumphā, and A. E. Caddy's cast preserved in the Indian Museum, Calcutta.

Page #139

ANANTA-GUMPHĀ INSCRIPTION
TEXT
[-
]'???samaņ[@]namo lepam [-]
1. The letters representing the opening words of the inscription are irrevocably lost.
2. The stone is peeled off and shows just the sockets of three letters which are read by Banerji as Dohada. I doubt very much the correctness of such a reading. Any one attempting to guess the letters from the outlines of the sockets is apt to glide into a fatal error Anyhow, the outline of the first socket resembles rather that of sa than that of do. Kalimga or Kalimga is the word which may be supposed to have preceded samanānam in an inscription of this kind.
3. samananam in Dutta's estampage.
(111)

Page #140

TEXT AS IT READS IN THE ORIGINAL
[-]??? ĦAT[1]# ami[]
TEXT AS IT READS IN PALI
[—]??? HATI ati [1]
TEXT AS IT READS IN SANSKRIT
[-]??? FUTAI 174 [1]
TRANSLATION
The cave (which is excavated for) the recluses (of Kalinga ?] [-
]
(112)

Page #141

No. XII INSCRIPTION ON THE ROCK
OUTSIDE THE VERANDAH OF
THE ANANTA-GUMPHA
15

Page #142

Page #143

ANANTA-GUMPHĀ INSCRIPTION
The following “record,” says Mr. R. D. Banerji, “is incised on the rock outside the verandah of the Anantagumphā.”
The text is based upon Lüders' List of Brāhmi Inscriptions in El, Vol. X, Appendix, No. 1343; Mon Mohan Chakravartti's Notes on the Remains in Dhauli and in the Caves of Udayagiri and Khandagiri ; and Hari Das Dutta’s inked impression read and reproduced by R. D. Banerji in EI, Vol. XIII, Udayagiri and Khandagiri Inscriptions, Pl. II, No. XII, Cave called Anantagumphā.

Page #144

Page #145

ANANTA-GUMPHA INSCRIPTION
TEXT
[-]" ko(th)āje(yā) *[–] [.]
1. It may be safely assumed that a few letters recording the name of the donor or donors are missing.
2. Mon Mohan Chakrabartti reads Dajacāra. Banerji differing, remarks: "It is really something like a mason's mark. There are three symbols, of which a central one is the Brāhmi letter ja, while the other two may resemble, but are not, letters." It is not too much to say that Banerji's is a cheap explanation, which has the merit of avoiding any difficulty that confronts a man attempting to read an old Brāhmi inscription wherein the letters are effaced and obscured. As far as I can make out, the first letter is ka or ko. The second letter is a fissured tha. The third letter is distinctly je. And the fourth letter is an obscured ya.
3. If the proposed reading koçhājeya be correct, it is easy to understand that it was followed by the conjunction ca and a word like pasādo or pāsādo.

Page #146

TEXT AS IT READS IN THE ORIGINAL
[...] filetfen [-] [1]
TEXT AS IT READS IN PALI
[...] Filzstien [–] [1]
TEXT AS IT READS IN SANSKRIT
[...] 77815fta: [-] [1]
TRANSLATION
The [-]unconquerable cells' (which are the meritorious works of) (-).
1. In commenting on pasāta and kothā occurring in the Inscription No. VII, Banerji observes: "It is interesting to note that here two words denoting almost the same thing, i.e., pasāto (Sk, prāsāda and kotha or kotha, have been used, and that we here find the word pasādo. The Sarpagumphà has only one small chamber but the Haridāsgumphà has a verandah, a large inner chamber with three doorways and one small side chamber on each side. So it is quite possible that the word pasāto refers to the main chamber and the word kotha to the side chambers." I am unable to accept such a suggestion for the simple reason that the word kotha, as plaral form of kotha, denotes all the cells or chambers, and that there is not a word to denote the verandah which is an outer construction.

Page #147

No. XIII
INSCRIPTION IN THE TATTVA-GUMPHĀ No. 1

Page #148

Page #149

TATTVA-GUMPHĀ INSCRIPTION
The following “inscription is written or painted,” says Mr. R. D. Banerji, "on the back wall of the inner chamber of the cave (the Tatwagumpbā, No. 1), and on prolonged examination I found that, in addition to a row of letters which I cannot make out, it was a repetition of the Indian alphabet.”
These observations of Mr. Banerji led me to examine Mr. Hari Das Dutta's estampage closely, and have ultimately enabled me to detect an inscription consisting of one line and a table of Brāhmi alphabet consisting of no less than six rows of letters. The table has been separately presented as the last of the present series (No. XV) of iuscriptions in the caves on the Udayagiri and Khandagiri bills.
The text is based upon Hari Das Dutta's inked impression reproduced by R. D. Banerji in EI, Vol. XIII, Udayagiri and Khandagiri Inscriptions, Pl. II, No. XIII, Cave called Tatwagumphā No. 1.
( 121 )
16

Page #150

Page #151

TATTVA-GUMPHĀ INSCRIPTION
TEXT
......???? ripu (tasa) kaya......???? [-]
(123)

Page #152

www. kobatirth.org
TEXT AS IT READS IN THE ORIGINAL
......????ttgag mal... ???? [1]
TEXT AS IT READS IN PĀLI ......????#Iyan Fill (?)...???? [1]
TEXT AS IT READS IN SANSKRIT ...17118gre (?)—?119 [1]
( 124 )

Page #153

No. XIV
INSCRIPTION OF KUSUMA IN THE TATTVA-GUMPHA No. 2

Page #154

Page #155

INSCRIPTION OF KUSUMA
The following "inscription," says Mr. R. D. Banerji,." is the oldest of the inscriptions in the Khandagiri caves. Most of the caves on the Udaya. giri are ancient, as proved by their inscriptions; but, with the exception of Tatwagumpbā, No. 1, Tatwagumpbā, No. 2 and Anantagumpbā, all other Khandagiri caves appear to be mediæyal, as the inscriptions in them are not earlier than the ninth or tenth century A. D. The record in this cave is incised over one of the entrances to the inner chamber and consists of one line. The cave is No. 1 of the plan of the Gazetteer (the Bengal District Gazetteer, Puri)."
The text is based upon Major Kittoe's facsimile read by James Prinsep in JASB, old series, Vol. VI, p. 1074; Alexander Cunningham's hand-copy reproduced in Corpus Inscriptionum Indicarum, old series, Vol. I, Pl. XVII, Cave No. 1; Lüders' List of Brāhmi Inscriptions in EI, Vol. X, Appendix, No. 1344; Hari Das Dutta's inked impression read and reproduced by R. D. Banerji in EI, Vol. XIII, Udayagiri and Khandagiri Inscriptions, P. II, No. X, Cave called Tatwagumpbā, No. 2.
( 127 )

Page #156

Page #157

INSCRIPTION OF KUSUMA
TEXT
Pādamulikasa' Kusumasa’ leņsā]nio [.]
1. Prinsep and others correctly read Padamulikasa. Kittoe's facsimile has "kasa,
2. Kittoe's facsimile is responsible for Prinsep's reading Kutamāsa instead of Rustmasa. Banerji reads Kusumasa, and remarks that the a-mark in ma is superfluous.
3. Prinsep wrongly reads lonam. Banerji doubtfully reads lena[m]phi, remarking that the last syllable (phi) in this record is superfluous and devoid of any significance. The cast in the Indian Museum clearly shows that the reading is nothing but lenani,
(129)
17

Page #158

TEXT AS IT READS IN THE ORIGINAL
पादमुलिकस कुसुमस लेण [T] नि [1]
TEXT AS IT READS IN PALI
पादमूलिकस्स कुसुमस्स लेणानि [1]
TEXT AS IT READS IN SANSKRIT
पादमूलिकस्य कुसुमसा लयनानि [1]
TRANSLATION
The caves (which are the excavations) of Kusuma of Padamulika.1
1. Padamulika has been rightly explained by Banerji as signifying either the locality or the professional designation of the donor. As a professional designation, it cannot but mean a server of the feet,' that is, a menial. Accepting the second meaning of the word, I may render the record:
"The caves (which are the excavations) of the menial Kusuma."
( 180 )

Page #159

No. XV
TABLE OF BRAHMÍ ALPHABET
IN THE TATTVA-GUMPHĀ No. 1

Page #160

Page #161

TABLE OF BRAHMĨ ALPHABET
"was
The following inscription containing a table of Brahmi alphabet noticed for the first time," says Mr. R. D. Banerji, "by the late Mr. J. D. Beglar in 1882, who published an eye-copy of it with his report (Arch. Surv. Rep., Vol. XIII, p. 82). But unfortunately the eye-copy was printed upside down. Mr. Monmohan Chakrabartti tried to read it from this plate, but apparently did not succeed. The whole inscription is written or painted on the back wall of the inner chamber of the cave, and on prolonged examination I found that, in addition to a row of letters which I cannot make out, it was a repetition of the Indian alphabet. Some young monk had used the back wall of the cell as a (slate or) copy-book and improved his knowledge of the alphabet by writing on it. The characters belong to the first century B. C. or first century A. D."
The table is based upon Hari Das Dutta's inked impression read and reproduced by R. D. Banerji in EI, Vol. XIII, Udayagiri and Khandagiri Inscriptions, Pl. II, No. XIII, Cave called Tatwagumpha No. 1.
"A
(188)

Page #162

Acha
TABLE OF BRÂHMI ALPHABET
ia aj mi u oo
....kha ga gba ? ca cha... .............. .....na ta tha da dha na...........
.............na ta tha da dha na pa pha...sa sa sa ha.......?
........ņa ta tha da dha na pa pha ba bha...ga sa ba... ...............ta tha da dba na pa pha ba...sa sa sa ha.....
..........ta tha........
1. In L1, Banerji reads just one letter, which is gha. 2. In L 3, Banerji does not read the two letters after na. They look like ya and ra,
but they may be pa and pha. 3. In L 6, Banerji reads just the second letter.
(134)

Page #163

NOTES

Page #164

Page #165

NOTES
1. THE INSCRIPTIONS AND THEIR AUTHORS
Of the fifteen old Brāhmi inscriptions in the Udayagiri and Khandagiri Caves, No. I is the Hathi-Gumphā inscription of King Kbāravela; No. II, the Vaikunthapura Cave inscription of Kbäravela's chief queen; No, III, the Pātālapura Cave inscription of King Kadampa-Kudepa; No. IV, the Yamapura Cave inscription of Prince Vadukha-Varikha; No. V, probably the Chota-Häthigumphā inscription of the same prince; No. VI, the Sarpa-Gum phā inscription of Cūlakamma ; No. VII, the Pāvana or Haridas-Gumphā inscription of the same donor; No. VIII, the SarpaGumphä-side Cave inscription of Kamma and Khīņā; No. IX, the Bāgh-Gumpbā inscription of the Town-judge Bhūti; No. X, the Jambesvara Cave inscription of the High-functionary Nākiya of Bāriyā, or it may be, of Mahāmada, Bāriya and Nākiya; Nos. XI and XII, the Ananta-Guimpha inscriptions of some donors ; No. XIII, the inscription of some donor in the Tattva-Gumphā No. 1; No. XIV, the inscription of the menial Kusuma, or it may be, of Kusuma of Padamūla; and No. XV, a table of Brāhmi alphabet in the Tattva-Gumphā No. 1.
2. THE RELATIVE TOTAL OF THE CAVES
AND INSCRIPTIONS
The fifteen inscriptions edited in the present volume, and arranged serially as No. I, No. II, and so forth, are the oldest known Brāhmi inscriptions which have hitherto been discovered as epigraphic records relating to different caves on the Udayagiri and Khandagiri hills in Orissa. The first ten of them belong to the caves on the Udayagiri and the remaining five to those on the Khandagiri hill.
Though, as a general rule, each of these inscriptions is meant to refer to a particular cave, it will be a mistake to presume that there are as many caves as inscriptions. For instance, there are two inscriptions, one of which (No. XI) is incised on the architrave outside the AnantaGumpha, and the other (No. XII) on the rock outside the verandah of the same cave. The concluding words of No. XI, samaņsā]nam leñam,
( 137 ) 18

Page #166

138
OLD BRAHMI INSCRIPTIONS
refer to the cave as a cave (lena), and not in terms of its component parts, while the surviving word of No. XII, koṭhājeya, refers to the cave not as a cave but in terms of its component parts. Here the presumption cannot but be that these two inscriptions, one referring to the cave in term of the whole and the other in terms of the parts, belong as well as refer to one and the same cave.
Secondly, there are two inscriptions in the Tattva-Gumpha No. 1, one of which (No. XIII) is meant to commemorate the name of the excavator of the cave, and the other (No. XV) to serve as a table of Brahmi alphabet. In this particular instance, the two inscriptions. belong to one cave, one referring to the cave as an excavation of some donor and the other serving altogether a different purpose.
Thus it may be shown that the total of the caves falls short of the total of the inscriptions at least by two, Nos. XI and XII being taken as referring to one and the same Ananta-Gumpha, and No. XV being left out of consideration as an alphabetic table.
We maintain that even further reduction of the total of the inscribed caves is possible. For No. XIV refers not to one cave but to caves in the plural number (lenani), which were all the excavations of Kusuma.
Why, it may be argued, if the inscription of Culakamma (No. V) is incised over the doorway of the Sarpa-Gumpha and that of Kamma and Khina (No. VIII) to the left of the doorway of the same cave, should we not take these inscriptions as referring to one and the same cave? Here we have got to distinguish between " belonging to " and "referring to."
same cave.
These two inscriptions belong to the Sarpa-Gumpha in the sense that both are incised over and to the left of the doorway of the But they refer to two caves, each cave in terms of its component parts, koṭhājeyā and pasada. The incising over the doorway of the cave goes to show that the inscription of Culakamma is meant to refer to the Sarpa-Gumpha. The incising to the left of the doorway of this cave may be taken to indicate that the inscription of Kamma and Khina is meant to refer to another cave, which lay on the left side of the Sarpa-Gumpha and would not at once be seen from its entrance. The Sarpa-Gumpha being unicellular and provided with one verandah, it is difficult to think that a second inscription was needed to refer to it in terms of its cell and verandah.
This argument may be substantiated by the study of a parallel case of the Mañcapuri group of three caves and three inscriptions. The inscription of Kharavela's chief queen (No. II) belongs and clearly refers to

Page #167

NOTES
139
the Vaikunthapura cave in the upper storey of the Mañcapuri group. The remaining two inscriptions are both incised in the verandah of the Pātālapura cave in the lower storey, the inscription of King KadampaKudepa (No. III) being incised on a raised band between the middle pillars, and that of Prince Vadukha-Varikha (No. IV) on the right wall of the verandah. Here the inscription of Prince Vadukha-Varikha refers undoubtedly to the Yamapura cave forming the right wing of the Pātālapura. That is to say, two inscriptions belonging to one cave refer indeed to two caves.
The local tradition asserts that there were formerly about 750 caves on the Udayagiri and Khand'agiri hills. This is surely a much exaggerated number, should it be treated as the total of the caves excavated during the reign of King Khāravela and thereabout. For we have a definite statement in the Hāthi-Gumphā inscription (No. 1) regarding the total number of the caves excavated on these two hills during Khāra vela's reign. In accordance with this statement, just 117 caves were constructed (sata. dasa-lena-satam kārā pitum) in the thirteenth year of Khāravela's reign.
We cannot say that all the 1!7 caves were inscribed. If each one of them might be supposed to have an inscription referring to or labelling it, the total of the inscriptions would have been not less than 117. But if all of them were not inscribed, the total of the inscriptions must have been less than 117. Probably all or most of them were not inscribed. Even the famous Rāni Nūr or Queen's Palace is found without an inscription. Though the local tradition ascribes the construction of this cave-architecture to King Lalatendu-Kesari who reigned about the year A.D. 617, the general style of its architecture and the technique, motifs and other details of its basso-relievo are strongly in favour of dating it as old as the Tiger and other inscribed caves wbich may be supposed to have been excavated during the reign of Khāravela or thereabout. Moreover, many of the gurviving caves being natural cavities little improved by human hand, are likely to have been left uninscribed.
In these circumstances, it is not unreasonable to think that the original total of the inscriptions stood far below that of the caves, although the exact ratio of the two totals cannot be determined, nor is it possible to say how many inscriptions have been actually lost due to the ravages of time.
In the Hāthi-Gumphä record of the thirteenth year of Khāravela's reign, we read that His Majesty caused 117 caves to be constructed on the Kumāri hill, obviously a common name for designating the twin-hills of

Page #168

140
OLD BRĀHMI INSCRIPTIONS
Udayagiri and Khandagiri by himself, by his queens, by his sons, by his relatives, by his brothers, and by the royal servants. Among the caves of his queens, we get just one inscribed cave of his chief queen ; among those of the princes, we get just two inscribed caves of Prince Vaduk haVarikha; so on and so forth.
It may be expected that at least those caves which were excavated by the members of Khäravela's royal family were inscribed. We mean that some at least of the inscriptions must have disappeared with the destruction of the ancient caves. And yet the fact would remain that the original total of the inscriptions stood far below that of the caves.
We have wrongly assumed so far as if the inscription of Kbāravela (No. I) were meant to refer exclusively to one cave, namely, the HathiGumpbā on the Udayagiri hill. It is far from being the case. It is not distinctly stated in the record of Khāravela's thirteenth regral year how many out of 117 caves were constructed by His Majesty, how many by his queens, and how many by others. We would think that the majority of these caves were constructed on the strength of the donations made by His Majesty. We shall be doing injustice to him to suggest that the inscription standing in his name was meant to relate exclusively to a cave which is of little importance as a work of art. To suggest this is to leave the acclaimed costly and excellent works of art and architecture accomplished by him altogether unexplained. The so-called Hathi-Gumpbā inscription of Khāravela was, indeed, meant to refer not to one cave but to all the works of art and architecture done by him on the twin-hills of Udayagiri and Khandagiri. As a matter of fact, the Hathi-Gumphā record was meant to refer in a general way to all the 117 caves with their inscriptions and architectural constructions.

Page #169

1
NOTES
3. THE SYMBOLS
The Hathi-Gumpha inscription of Kharavela is enclosed between two pairs of symbols. Two symbols are to be seen on its left side, and two on its right. The first symbol on the left stands over against its second line. The fourth symbol on the right stands over against its seventeenth line. These two symbols were apparently intended to mark respectively the beginning and end of the inscription. The second symbol on the left appears below the first, and stands over against the fourth and fifth lines of the inscription. And the third symbol on the right appears at the end of and between the first and second lines of the inscription.
2
45
3
४
4
141
H
It is not correct to say that the third symbol appears immediately after the name of Kharavela with which the first line terminates.1 Even if this symbol figured just after the name of Kharavela, I do not see how any importance might have been attached to it because of such a position. The symbol, as it now stands, seems to have been set off on the right against the first and second symbols on the left, as an equipoise. In theory, of the four symbols, the first and the fourth were intended to mark the commencement and close of the inscription, and the second and the third to stand, somewhere in the middle, on two sides, enclosing the inscription between them. I believe that this arrangement would not have been departed from, if the right upper corner had not appeared bare as a result of the third symbol having been placed far below the position in which it appears.
What are these four symbols? The first symbol is what the Jains call Vuddha-mangala translated by Dr. Coomarswamy as "Powder-box." The second symbol is the well-known Svastika. The third symbol is what has been labelled as Nandipada or "Bull's foot-mark" in a Buddhist
1. Jayaswal says, "This symbol is inscribed just after the name of Kharavela in the first line." JBORS, 1917, Vol. III, Part IV, p. 428.

Page #170

142
OLD BRAHMI INSCRIPTIONS
inscription on the Padaņa hill near Kanheri.! The fourth symbol is what may be aptly described as Rukh ha-cetiya or Vşkşa-caitya, “ Tree-device" or “ Woodland-shrine."
The first symbol, which looks like a modern crown ? figures, as pointed out by Dr. Bhagawanlal Indraji, in sculptured decorations carved over the doorway arches in some of the Udayagiri caves. It is one of the auspici. ous devices adorning the doorway of a Jain cave at Junagad. It stands over the third figure on the Sanchi gateway. It is found in some of the Buddhist cave inscriptions in Western India, e.g., in the second inscription at Junnar, the first inscription at Karle and the third inscription at Bhaja. It is not uncommon in necklaces (candrahāra). It is still very popular as one of the aşta mangalas or “ eight auspicious symbols " among the Jains in whose modern representation, it appears as "a beautiful bowl with a lid surmounted with three pinnacles." 4
What this first symbol actually signifies none can tell. Nothing is to be inferred as to its meaning from the name Vadiha-mangala offered by the Jains. With regard to its external feature, the Jains may describe it as vaddha or "closed," and the modern scholars as a "shield," or a Trisula, or a tri-ratna device, or a "powder-box." These do not carry us far. But it may be worth while to inquire if any clue might be obtained from the study of some of the devices on the ancient punch-marked coins. On some of these coins, the second, third and fourth symbols of the Hāthi-Gumphi inscription appear around and in association with a device which Sir Alexander Cunningham took to be a representation of Mount Meru, and other numismatists take to be the representation of a mountain. On some of these coins, this device occurs alone, and on some, it figures in association with one or another of the second, third and fourth symbols of the Hāthi-Gumphā inscription. A prototype of this device is met with on the Sohgaura copper-plate where it appears in the middle in association with the third and fourth symbols of the Hathi-Gumphā inscription and with a crescent-mark in its upper curve. So far as this copper-plate and the punch-marked coins are concerned, I have been strongly inclined to explain it as signifying a caravan-camp
1. JBBRAS, XV, p. 320. 2. JBORS, 1917, Vol. III, Part IV, p. 428. 3. Archäological Survey of Western India, separate pamphlet, x, pp. 23, 28, 42. 4. JBORS, 1917, Vol. III, Part IV, p. 429. 5. Ganguly's Orissa and Her Remains, p. 40.

Page #171

NOTES
143
under a canopy. This explanation of mine has been partly based upon a Buddhist description in the Vannupatha Jātaka (Fausbüll's No. 2) of a caravan journey across a desert.
Though there are some points of likeness between this device and the first symbol in the Hathi-Gumphā inscription, the two are not identical. Nor does it appear that the same symbolic device suggested the same idea to all and in all connexions.
As regards the Hāthi-Gumphā inscription, I would, for a proper consideration of their nature, divide the four symbols into two groups, one group consisting of the first and fourth symbols, and the other of the second and third. Taking the first and fourth symbols, one may not be far wrong, I believe, in thinking that these were designed, in this particular connexion, to represent a hilly woodland in which the cave with its inscription was situated. The naming of a hill-cave by a tree adjoining it was an ancient convention, which is evident from the names of certain caves in some of the old inscriptions, e.g., Nigoha-kubhā, "the Banyan cave," in the first Barabar Hill Cave inscription of Asoka, and Imdasāla-guhū, "the Shorea Robusta Cave," in one of the Barhut inscriptions.
Similarly taking the second and third symbols together, it may be said that, whatever may be their origin, as they stand in this inscription, they are intended to convey respectively the idea of svasti (well-being) and that of mangala (victory). The association of these two ideas in Buddhism, and á posteriori in Jainism, is clearly brought out in the concluding verse of the Pāli Margala-Sutta which reads as follows:
Etādisāni katvāna sabbattham aparājitä / Sabbattha sotthim gacchanti, tam tesam mangalamuttamanti
If the Jains and Buddhists associated these two symbols to convey the ideas of svasti and mangala, one need not be surprised if the Vedic religionist associated them to convey his ideas of auspiciousness. I mean to say that Mr. Harit Krishna Deb's suggestion, that these two symbols served, perhaps, as epigraphic devices for representing the mystical formula “OM,” need not be dismisssd as a mere ingenuity.
There are divergent theories regarding the origin of these two symbols, which are, however, too well-known to need repitition. For a general information about the Svastika symbol, the reader is referred to

Page #172

144
OLD BRĀHMĪ INSCRIPTIONS Dr. Bhagavanlal Indraji's note in Actes du Sixieme congres International des Orientalistes, Part III, p. 137. And for the discussion of different theories about the origin and explanations of the form of the Nandipada symbol, he is referred to Dr. Fleet's paper on the Sohgaura copper-plate inscription which he published in JRAS, 1907, pp. 529-32.
Regarding the second of these two symbols, it remains to be seen whether its name Nandipada was suggested in the Buddhist inscription from its external resemblance with the bull's foot-mark or from some other circumstance. There is something to be said for the second alternative. The Cullakālinga-Jātaka (Fausböll's No. 301) says that the tutelary god of Kalinga was represented in the shape of "a white bull," and that of Assaka in the shape of "a black bull.” Dr. Indraji is right to say that the Buddhists, giving it the name Nandipaila, meant to represent it as a “ Bull symbol," and not as a mere “foot-mark of a bull."
One tree-symbol marks the beginning of the inscription of Kamma and Khiņā in the Snake Cave (No. VIII), and another tree symbol that of the inscription of the Town-judge Bhūti in the Tiger Cave (No. IX). These two symbols are just two different varieties of the fourth symbol in the Hāthi-Gumphā inscription. The inscription in the Tiger Cave has, indeed, a second symbol at its end. It is again a Svastika mark carved in a somewhat different form. These three symbols appear as represented in the following table :
7 8
5

Page #173

NOTES
145
4. THE LETTER-FORMS
In the present edition we have so arranged the fifteen old records in a progressive numerical order as to make No. I, the Hathi-Gumpha record of Kharavela, to appear as the peg, and to make the rest, Nos, IIXV, to appear as hanging upon No. I. Looked at from this point of view, Nos. II-XV, are to be taken as subservient to No. I, which is the principal record. In our plan, Kharavela's Hathi-Gumpha record stands first and the Table of Brahmi Alphabet stands last, the series starting from the former and being closed by the latter. Considered in the light of this plan, a special importance has been meant to be attached to the first, as well as to the last number of the series. The special importance of the first is that it furnishes the key to the appreciation of the contents of the records, Nos. II-XIV; and the special importance of the last is that it furnishes the key to the understanding of the letterforms of the preceding records, Nos. I-XIV.
The table is found on the back wall of the inner chamber of the Tattva-Gumphã. No. 1. It presents some six straight rows of Brahmi letters carefully incised one below the other and all below the inscription attached to the cave. The succession of the letters that may yet be read, especially in the right half of the table, indicates that in each row a complete set of the Brahmi alphabet was intended to be shown, and that the letters were meant to be alphabetically arranged, each set beginning with the letter a and ending with the letter ha.
Mr. R. D. Banerji seems to think that the dressed surface of the wall of the cave was used by a young monk as a sort of copy.book for improving his knowledge of the alphabet by writing on it. This is, no doubt, a very plausible explanation for the possibility of the table as it is. It also might be that the mason who was employed to incise the inscription, after having incised the short inscription of one or two lines, found sufficient unfilled space on the dressed surface which he eventually thought of filling up with the rows of Brahmi letters. That the table was the manipulation of a practised hand cannot at all be doubted. And the practising hand of a young monk cannot be expected to have produced such a good specimen.
As bad luck would have it, not a single row of letters in the table is legible to-day from beginning to end, and not a single row can be wholly deciphered. The consequence is that, as regards orthography, we cannot
19

Page #174

146
OLD BRAHMI INSCRIPTIONS
say precisely how many vowel-letters were incised in each row That the vowel-signs were not separately shown and that, at least, these vowelletters, ā, i, ē, 16, ū, e and 0, were incised in each row may be taken for granted. In the texts of the old Brāhmi inscriptions, we have the use of the following vowel-letters and vowel-signs :
(1) . Vowel-letters--a, ā, i, u, e, ai ( ? ), and o; (2) Vowel-signs for ā, i, i, u, ü, e and o
Precisely from what period of time the Brāhmi alphabet was adapted to the needs of Sanskrit language we do not know. If Mr. Jayaswal's "Sunga Inscription of Ayodhya” may be presumed to have been incised actually during the reign of Puşya mitra, or, at least, during the Sunga reign, the second century B. C. is the earliest period to which the date of a Sanskrit record in Brāhmi characters can be pushed. But the Junāgad inscription of Rudradāman I (circa 150 A. D.) is certainly the earliest known lengthy Sanskrit royal record in Brāhmi characters.
In accordance with a statement in the Artha-Sāstra of KautilyaKautalya, the orthography of royal writs (in Sanskrit) comprised some sixty-three letters. The commentator accounts for this total by the summation of twenty-two letters for representing the vowel-sounds and forty-one for representing the consonant-sounds. Thus the dictum in the Artha-Šāstra might be taken to imply that, at the time of its compilation, twenty-two was the conventional maximum of the total of vowel-letters required for orthography of royal writs in Sanskrit). But, in reality, Kautilya-Kantalya thought not so much of the orthography of the royal writs as of the letters, signs and notations of Sanskrit phonology. The notations required for representing nine pluta-svaras were quite out of place in the general orthography of Sanskrit royal writs. Leaving aside the nine pluta-svaras, we get thirteen as the conventional maximum total of Sanskrit vowels. It is impossible to think that the orthography of the Tattva-Gumphā table comprised so many letters as were required to represent all the thirteen Sanskrit vowels. Here the important inquiry is whether this orthography included in it the letters required for representing the two dipthongs, ai and au, or
1. Artha-Šāstra, II, 9. 28 : Akäradayo varnäh trişastih. 2. Artha-Šāstra of Kautalya, edited by Ganapati Sastri, Part I. p. 170.
3. Ganapati Sastri's edition of the Artha-Sāstra, Part I, p. 170 ; Dirghāļ svarūņām tatrástau, pañca hrasvāh, plutá nava.

Page #175

www. kobatirth.org
NOTES
147
not. The point might be partly decided if we could ascertain that Airena was the openin word of the main portion of the text of Khāra vela's inscription.
The first letter of the opening word, as it appears in the plaster-casts and estampages, is the dipthong ai, and we have agreed with Mr. Jayaswal and others in reading the opening word as Airena, and have differed from Mr. R. D. Banerji who proposed to read Kharena, as well as from Sir Alexander Cunningham who read it as Verena. As we have noted, the name of the king of Utkala who defeated king Nanda in a battle was known to the author of the Sanskrit verses, quoted by Mr. Jayaswal from an Old Osiyā MS. of the 14th or 16th ceutury A. D., to be Aira and Ahira. If Airena be at all a correct reading, one must not forget that it is the only instance where a Brāhmi letter representing the dipthong ai is met with in our old Brāhmi inscriptions.
But our faith in the correctness of the above reading bas, to a large extent, been shaken by the consideration of two facts, one brought to our notice by Sir Alexander Cunningham and the other gleaned by us. Sir Alexander has pointed out that Vera in the sense of Vīra (Heroic) occurs as a royal epithet in some of the coin-legends of Ancient India. And we find, in corroboration of the soundness of Cunningham's argument, that the first letter of the opening word of Khāravela's inscription exactly resembles the first letter of a proper name in the two Pablosa inscriptions of Aşādhasena, which the epigraphists have agreed to read as Vaihidari. It is certain that the first letter of Vaihidari could not be read other than as Vai. If the first letter were meant to be read as Ai, we would have found the lettere with one e-sign attached to it instead of a letter resembling e with two e-signs. But for these two e-signs attached to the first letter, the name would have been read as Vehidari. If this be correct, it goes without saying that the Brāhmi letters to represent the dipthongs, ai and au, were unnecessary for incising our old Brāhmi inscriptions. Further, we should remember that there was no occasion for the scribe or scribes of our old Brāhmi inscriptions to make use of the ai or au sign. These signs were, indeed, needed for incising the Barhut E. Gateway inscription of Dhanabhūti and the Pabhosa inscriptions of Āşādhasena, for inscribing such words as pauteņa and Vaihidari.
Apparently the number of consonants represented in the Tattva-Gumphā table seems to have fallen short of the total, forty-one, contemplated in the dictum of the Artha-Sāstra for the orthography of Sanskrit royal

Page #176

148
OLD BRAHMI INSCRIPTIONS
writs, nay, of Sanskrit language, by as many consonants as eight. For in each row of letters in the table we are not to expect to see more than thirty-three letters intended to represent the following consonants in such order as-ka, kha, ga, gha, na ca, cha, ja, jha, ña ta, tha, da. dha, na ta, tha, da, dha, na pa, pha, ba, bha, ma | ya, ra, la, va sa, şa, sa, ha , that is to say, the five consonantal groups of five each, the two semi-vowels, the two liquids and the four sibilants. It is impossible to conceive that the consonantal signs for representing the Sanskrit ref, ra-phalā,ya-phalā, anusvāra, visarga, as well as for indicating that the pronunciation of da is as in vidala, or that the pronunciation of dha is as in gādha, or that the pronunciation of ta is t as in matsya were meant to be displayed. We have the frequent use of the anusvāra-sign (m) in our old Brāhmi inscriptions. In point of fact, the predominant tendency of the orthography of our old Brāhmi inscriptions is to represent the nasals of all the five consonantal groups by m in all cases of conjoint consonants. There was no occasion for the use of the visarga-sign (h). But this does not mean that the Brāhmi alphabet was wanting in this sign. For we have a clear use of this sign in Mr. Jayaswal's 'Sunga Inscription of Ayodhya.'
There was yet no necessity for having a separate letter for representing the Sanskrit kşa, or for having separate conventional signs for representing the r, ra and ya sounds of conjoint consonants. Looked at from this point of view, it would be wrong to say that the consonantal system covered by the Brähmi alphabet fell short of the orthography of KautilyaKautalya's royal writs in Sanskrit. Even the Brāhmi alphabet used for the inscriptions of Asoka, especially for those at Girnar and Kalsi, was adequate for the entire Sanskrit consonantal system minus the visarga-sign. Anyhow, it is certain that, like the orthography of the Kalsi rock inscriptions of Asoka, the orthography of the Tattva-Gumphā table included three letters representing sa, sa and sa.
So far as the language of our old Brāhmi inscriptions goes, the letters representing sa and şa were of no use. Mr. Jayaswal makes out two words in Khāravela's inscription, one affording a singular instance of the use of the letter sa, and the other that of the use of the letter şa. The two words are respectively rājisi-vamsa-kula-vinisrto (I. 16) and Tramiradeşa-samghātam (I. 12).
Mr. Jayaswal depends for these readings on the estampages taken by Mr. Vats. We can quite see that in Mr. Vats' estampages, at least, in one of them, the letter meant to represent si appears, by some mysterious work of erasion and fissure, as rúa, and that ha of daha appears, for some

Page #177

www.kobatirth.org.
NOTES
149
reason or other, as sa and no less as Asokan ha. But the first letter cannot undisputably be read as de in view of the fact that it presents to view two similar e-signs, attached one to the upper end and the other to the lowerend.
Depending on Mr. Vats' estampages. Mr. Jayaswal makes out a word, Simdhuļāya (I. 14), affording a singular instance of the use of a letter representing the Vedic cerebral ļa which is frequently used in Pāli as a substitute for da. We have to reject his reading on these two grounds: (1) that the Tattva-Gumphä table has no place for such a letter, and (2) that the three letters read by him as dhulāya can be better read, with the aid of the estampages previously published by him, as silähi. As a working hypothesis, it may be taken for granted that, as regards the main letters, what is to be found in the Tattva-Gumphā table may be found in the texts of our old Brāhmi inscriptions and what is not to be found in the Tattva-Gumphā table cannot at all be found in the latter
Prof. Sten Konow reads Calakramesa for Cūlakamasa, If, indeed, the intended reading were Culakramasa (Cūlakarmasa), we might have used it as a remarkable evidence to prove that the scribe or mason-engraver, employed to incise the inscription belonged to the Uttarāpatha, or more particularly, the Gandhāra-region where the prevailing tendency was to spell dharma as lhrama, and darsi as draới.
The letter-forms of our old Brāhmi inscriptions show a stage of development of the Brähmi alphabet, which is posterior to that of the Brābmi inscriptions of Asoka. Dr. Bhagawanlal Ipdraji notices the following as the chief points of difference between the two sets of Brāhmi letter-forms:
“While Asoka's ka has the vertical and horizontal strokes equal in size, here the vertical stroke is longer than the horizontal stroke; ga originally angular now becomes arched; gha, pa, la and ha originally with a rounded base now becomes flat; ma and va with a well-rounded base now becomes triangular; the two base-strokes of ta originally making an angle now tend to be an arch; ikāra-strokes making an angle no longer
1. We are indebted to Dr. Niranjan Prasad Chakrabarty of the Calcutta University for drawing our attention to a number of manuscripts found in E. Tarkestan, each of which has in it a Table of Alphabet, obviously as an aid to the reading of it. Here, too. the table was annexed apart from the consideration that all the letters included in it were actually required for the Ms. concerned.

Page #178

150
OLD BRAHMI INSCRIPTIONS
1.
do so and go up higher. These are all points which prove that the character is later than Asoka's."
Rai Bahadur Ramaprasad Chanda points out the following as the; most notable characteristics of the Hathi-Gumphā alphabet: “(1) A considerable number of letters with thick-headed vertical or serif; (%) ka with the lower part of the vertical prolonged ; (3) invariably rounded ga (4) cha of the butterfly type with two loops ; (5) ta's baving in most cases rounded lower part.” 2 It may also be noticed that in the Hāthi-Gumpha alphabet, the letter a is made up of two side strokes meeting the vertical leavng a wide space between them; the vertical of kha which in its earlier forms had nothing at its base has invariabìy a triangular or circular base; the letter ra which in its earlier forms was of a cork-screw pattern has become a straight vertical; and the letter ha has no longer a short horizontal stroke attached a little below its right vertical.
Applying all these prominent characteristics as a test of chronology of the letter-forms, and observing that the Hāthi-Gumphā inscription shares them, to a considerable extent, with the inscriptions on the Sanchi Gateways, Rai Bahadur Chanda finds indication to be able to say that the Hāthi-Gumphā inscription “is later in date not only than Asoka's edicts and the Besnagar Garuda pillar inscriptions, but also later than the Bharhut torana inscription and the Nânâgbāt inscription of the time of the Andhra king Siri Sātakani I.”
The Rai Bahadur has sought, indeed, to deepen the significance of Dr. Indraji's observations by conceiving as many as seven stages in the evolution of the Brāhmi letter-forms from the Edicts of Asoka to the Sanchi Gateway-inscriptions, the sixth being represented by the HathiGumphā inscription of Khāravela and the fifth by the Besnagar Garuda Pillar inscription of Mahārāja Bhagavata, the Nānäghãt Cave inscription of Nayanikā, the widow of Satakaņi I, and the Barhut E. Gateway inscription of Dhanabhūti, taken in a chronological order.
In the opinion of Prof. Bühler, the characters of the Sanchi South Gateway inscription wherein a king Siri-Sātakani is mentioned were almost identical with those of the Nānāghāt Cave inscription of Nayanikā. As against this opinion, Rai Bahadur Chanda maintains that the characters of the Nānāgbāt Cave inscription of Nayanika and also those of the
T.
1. Actes du Sixieme Congres International des Orientalistes, Part III, p. 140. 2. Memoirs of Archeological Survey of India, No. 1, p. 10.

Page #179

www. kobatirth.org
NOTES
151
Nāsik Cave inscription of king Kanha differ from the Sanchi Gateway inscriptions in these two essential features : "(1) Letters with the socalled serif or thick-headed vertical are quite conspicuous by their absence in these records. (2) In place of ta's with the invariably round lower part of the Sānchi Gateway inscriptions, we have in these earliest Andhra inscriptions and in the coins of king Siri Sāta (rightly identified with Siri Sātakaṇi of Nānāghāt) ta’s with mostly angular lower part."
Sometime ago Mr. R. D. Banerji read, at one of the monthly meetings of the Asiatic Society of Bengal, a carefully written paper embodying the results of his study of the letter-forms of the Nānāghāt Cave inscription of Nayanikā and the Hāthi-Gumpha inscription of Khāravela. His paper, which has been meant for publication in JASB with copious illustrations, showed that he had made a most painstaking analysis of the letter-forms of the two inscriptions. He said, in the course of discussion, that he could see in Khāravela's inscription the engraving of three separate bands. He tried to point out that in all cases ga's and ta's were not arched or rounded but angular, and that, as a matter of fact, there were a number of letter-forms bearing the characteristic features of Asokan Brāhmi, and that, on the other hand, there were certain letter-forms in the Nānāghat Cave inscription that anticipated those of the Kuşāņa inscriptions. Thus he found reasons to differ from Rai Bahadur Chanda who assigned an earlier date to the alphabet of the Nânaghat record. He seemed to think that as, on the ground of anticipation of some of the letter-forms of the Kuşā!a inscriptions, the Nānāghāt record might be dated later, so, on the ground of lingering resemblance with the letter-forms of Asokan Brāhmi, the Hāthi-Gumpha record might be dated earlier. He inclined to explain the occurrence of different types of Brāhmi letters in the Hāthi-Gumphā inscription by the fact that of the scribes employed, one was well-practised in the Eastern variety, one in the Southern variety, and so on.
Mr. Banerji's analysis of the Brāhmi letter-forrns of the two inscriptions leaves nothing to be desired in the way of industry and precision. It convinces us that all the ga's in Khāravela's inscription are not arched ; that the lower parts of all the ti's are not invariably rounded ; that it has a number of letter-forms bearing the characteristic features of the alphabet of the Maurya inscriptions ? ; and that the Nānāghāt Cave inscription shows some letter-forms anticipating those of the Kuşāņa inscriptions.
1. Some of the Nathi-Gumphà letter-forms may be Asokan or Mauryan, but none of the Asokan or Mauryan letter-forms has a serif or thickened top.

Page #180

152
OLD BRĀHMI INSCRIPTIONS
His analysis serves so far as to enable us to think that Rai Bahadur Chanda's statement has erred in facts. But he himself has failed to produce a convincing proof to substantiate his position as to the currency of the Northern, Southern, Eastern and Western varieties of the Brāhmi alphabet precisely at that period of time which had seen the engraving of the inscriptions concerned." In dealing with the language of Khāravela's inscription, he has taken the substitution of dha for tha to be a characteristic of the Southern dialect, the dialect which was current in Kalinga and Andhradeśa. Is it a right assumption that because the inscription of Khāravela and that of Nayanikā are found engraved in Southern countries their language must have been a Southern dialect ? Can we seriously take, on the same ground, the letter-forms of the Nānāghāt Cave inscription as representative of the Southern variety of the Brāhmi alphabet ? We cannot. We find that for the engraving of the most southern copies of his Minor Rock Edicts, King Asoka employed a scribe, named Pada, who hailed from a country, say the Gandhāra region, where Kharoşthi was the script in common use. The same remark holds true of the scribe or mason-engraver employed to inscribe Dhanabhūti's record on the Barhut E. Gateway.
We are prepared to entertain Mr. Banerji's hypothesis in so far as it leads us to believe that the scribes employed to incise Khāravela's inscription represented different localities. In examining the letter-forms of the Barhut inscriptions, we found some clear data compelling us to think on similar lines,-to feel that certain carvings and inscriptions were the work. manship of a particular group of artists, and others those of other groups. We were constrained to think that Rai Bahadur Chanda's view of the development of the Brāhmi alphabet was defective in so far as it did not consider what effects might have been produced if different groups of artistscribes hailing from different parts of India were employed to do work on the same piece of stone, either at the same time or at different times. The Rai Babadur failed to notice that, so far as the Barhut Inner Railing was concerned, there were two carvings 2 with a number of inscriptions, in which the letter cha was of the butterfly type with two loops, the letter pa had a
1. We mean the period anterior to the date of composition of the Lalita-Vistara which mentions various local alphabets.
2. Cunningham's Stāpa of Bharhut, PI, XIV

Page #181

NOTES
153
fat base, and the letters tended to show thickened tops or serifs. Although we felt that these were the workmanship of one group of artists who represented a particular locality, we found no means of ascertaining what locality of India was actually represented by them.
We bave now, thanks to the continued efforts and successes of the explorers and epigraphists, a good number of inscriptions, the letter forms of which show a close resemblance in many respects, to wit: (1) our old Brāhmi inscriptions in the Udayagiri and Khandagiri caves ; (2) Mr. Jackson's second Barābar inscription recording the name of the rock on which it is engraved as Goradhagiri ; (3) the Sanchi Gateway inscriptions incised during the reign of one King Siri-Sātakaội ; (4) two Pabhosa inscriptions of Āşā lhasena referring to Bahasatimita; (5) Gotamimitrā's inscription referring to King Vişnumitra 1 ; and (6) Mr. Jayaswal's “Sunga Inscription of Ayodhya," ? referring to Puşyamitra. Curiously enough, while these epigraphs, especially the inscriptions of Khāravela and Aşādbasena, all referring to Bahasatimita, show all the signs of later development, later than the Mauryan, Yasamitā's Brick Tablet inscription referring to Brhāsvātimita is found in Brāhmi characters which are decidedly Mauryan
If Brhāsvätimita of Yasamita's Brick-tablet was, as Dr. Vogel takes him to be, the same person as Bahasatimita of Āsādhasena's inscriptions, or the same person as Bahasatimita of Khāravela's Hathi-Gumphā record, we get at last a fact, which altogether upsets Rai Bahadur Chanda's chronological order of the Brābmi inscriptions. And we are inclined also to think alike with Mr. Jayaswal who says: “The Sungas both in inscription and on coins used a script whose letter-formas are of a different style, and more advanced than the letter-forms of the later Mauryas. In the Besnagar inscription of the later Sunga Bhāgabhadra, we have the Maurya forms. There being thus more than one style of writing in vogue at one and the same time, the chronology of the records of the second century B.C. and thereabout is not deducible merely from letter-forms."'4
1. Indian Historical Quarterly, Vol. II, No. 3, p. 56. 2. JBORS, Vol. XIII, Parts III-IV, p. 247. 3. JRAS, 1912, p. 120. 4. JBORS, Vol. XIII, Parts III-IV, p. 249.
20

Page #182

154
OLD BRĀHMI INSCRIPTIONS
5. JAYASWAL'S FIVE CORRECTED READINGS
The third instalment of corrections published by Mr. Jayaswal in JBORS, Vol. XIV, Part I, pp. 150-1, goes to show that he keeps yet an open mind as to the reading of the text of Kbāravela's inscription. The fact that he has so far revolved like a weather-cock at every gust of wind is indicative of pothing but his indecision. We shall consider below five of the corrected readings which he has offered and see whether and how far they bear scrutiny.
First, he proposes to read Māhāmeghavāhana, instead of Mahāmeghavāhana, as a common patronymical epithet of King Khāravela and King Kadampa-Kudepa in the old Brāhmi inscriptions Nos. I and III. He seems to think that the patronymic Māhāmeghavāhana implies that Mahämeghavāhana wag the personal name of Kbāravela's grandfather. Even if his reading be correct, we do not see how his conclusion is correct. There are numerous instances in the Brāhmi inscriptions where for a we have the use of ā, e.g., Brhāsvāti for Brhaspati in Yasamata's Brick-tablet inscription. Even supposing that in Māhāmeghavāhana, a has not been represented by ā, where is the evidence to prove that Khāravela's grandfather was succeeded by his father and that his father was succeeded by him? Seeing that, whether as Māhāmeghavāhana or as Mahāmeghavāhana, it occurs as a common epithet of two kings of the same Kalinga royal family, we may say without much fear of contradiction that Mahāmeghavāhana was a common epithet of King Kbāravela and King Kadampa-Kudepa, as well as that of their predecessors. But here his reading itself is open to dispute. As bad luck would have it, exactly that portion of the inscribed surface of the rock is fissured which contains the letters inscribing the epithet. So far as No. I is concerned, the choice lies not only between ma and mā, but also between mă and mo. As regards No. III, the ā-stroke is not where it should be. A similar apparition can be noticed in the inscription of Kusuma (No. XIV) inducing one to read Kusumasa as Kusumāsa. We do not see much use making fuss over ma and mä.
Secondly, he proposes to read in 1. 9 sa-b(i)ta-senavāhano vipamumcitum Madhuram apayāto, correcting samb(i)ta his previous correction to sa-b(i)ta, and to translate sa-bita-senavāhano: “ with his army and transport having lost their morale.” He thinks that bita is the same word as vīta, which has a technical military sigpificance,

Page #183

NOTES
155
as may be proved on the authority of the Amarakoşa where (in Bk. Kșatriya, 43) we read : vitam tu asāram hastyasvam. Here vita is explained as a substantive. But Mr. Jayaswal interprets bīta as an adjective qualifying sena-vāhana, which is impossible on the authority of the Amarakosa. Another objection is that after panādena there must be a verbal form to denote the effect of the action implied, wbile Mr. Jayaswal's reading and rendering serves to remove the verbal form altogether. Moreover, one can be sure more about ne as the final syllable than about no. We do not deny the possibility of the reading sabita as it appears in the estampages. What we deny is the possibility of Mr. Jayaswal's interpretation of it. Even sabita must be taken in the same sense as sam hita. For the appearance bha as ba, one may compare gabhira in 1. 2. For sabita = sambhita, one may compare samkāra = samkhāra in I. 16.
Thirdly, he proposes to read in I. 2 mamờam Avarāja-nivesitaPithudam gadabha-namgalena kāsayati, correcting his previous correction mamdam · avarāja-nivesitam pithuda, and takes it to mean “He (Khāravela) razes to the ground (lit. ploughs down) with an ass-plough Pothumla the market town (maņņa), founded by (the) Āva king." Iu support of his reading Āva-rāja, he argues : " There was an ancient Tamil race Āya (Cam. H.I., p. 596) and a kingdom called Āva-mukta, to the north of Kānchi, in the time of Samudragupta (Allahabad inscription), and a people spelt as Avarni and Aruarni below or by the Kistna as described by Ptolemy." He also puts forward as an argument: “ The Bhāgavata records an Āra dynasty (xii, 2-29) which the Vişnu (iv. 24. 13) equates with Andhra. This agrees with Ptolemy. The Āvas were in the Andhra country at the time." We have looked in vain into the Vişnu-Purāņa (iv. 24. 13) for the equation of any dynasty by the name of Āra with the Andhra. The letter read as ā in Mr. Jayaswal's Avarāja cannot be read as ā in view of the fact that in it the two left strokes do
1. A statement in the Nasik Cave inscription No. 10 (EI, Vol. VIII) goes indeed to prove that panadena may be used even without a verbal expression after it to denote the effect produced by the action implied in it. But it is difficult to understand that in the Hathi-Gumphā statement the verbal expression vipamumcitum (to release, to relieve) governs madhurā instead of senavahanas. The statement in the Nasik Cave inscription reads: "Bhatáraka amñatiya ca gatosmim varsaratum målayehi rudham Utamabhädram moca yitum ca Malayā pranăden' eva apayātā,” which Senart translates : " And by order of the lord I went to release the chief of the Uttamabhadras, who had been besieged for the rainy season by the Mālayas, and those Mālayas fled at the mere roar (of my approaching) as it were.

Page #184

156
OLD BRAHMI INSCRIPTIONS
not meet the vertical leaving a space between them. We doubt the correctness of his reading mamdam. Mr. R. D. Banerji's estampage clearly shows that the first letter of Mr. Jayaswal's Avarāja is pu. We make bold to dispute namgalena kasayati because the letter after namgale or nagale is ne and is intended to be connected with kasayati.
Fourthly, he proposes to read in I. 12 Tramira-desa-samghātam, correcting his previous correction tamara-deha-samghatam, and takes it to mean "the combination (or league) of the Tramira (Tamil) countries." His reading Tramira has come very near to our tamira or timira. Mr. Jayaswal does not at all consider that here the word samghata may also be interpreted in the sense of samkhata meaning "called". Regarding his reading desa, our first objection is that in the estampage of Mr. Vats, the letter da shows similar e-strokes on both ends. Our second objection is that the use of the cerebral sibilant ga is unexpected. Mr. R. D. Banerji's estampage tends to produce the appearance of the first letter as ti, and the estampage of Mr. Vats that of tra. We cannot think that here Tramira occurs as a spelling for the Tamil (Tamila or the Sk. Dravida), due to the fact that la or da is in all cases represented by da, and nowhere by ra.
Fifthly, he proposes to read in I. 16 vinisṛto, correcting his previous reading vinis(i) to. He has evidently relied too much on the estampage of Mr. Vats, forgetting that in taking an impression, a man, however careful and cautious he may be, is apt to exercise his own discretion in determining the outlines of the letters. If we were the men engaged for the purpose, we would have found ourselves quite at liberty to mark out the letter as s(i). Anyhow, it is certain that the letter was never meant to be sr, cf. the letter by in Bṛhasväti on Yasamita's Brick-tablet (JRAS, 1912, p. 120).

Page #185

NOTES
157
6. THE LANGUAGE
Judged by the sound-system and syntax, the language of the old Brahmi inscriptions is very nearly Pali, the language of the Buddhist Tipitaka preserved in Ceylon, Siam and Burma. Leaving the spelling and pronunciation of a few words out of consideration, we can say that their language is Pali, and nothing but Pali. The exceptional cases of spelling and pronunciation are important as enabling the reader to detect the under-current of a dialect having affinity, in respect of its phonetics, with Ardha-Magadhi, the language of the extant Jaina Agama.
Broadly speaking, the differences between the language of the old Brahmi inscriptions and Ardha-Magadhi are the differences which exist between Pali and the Jain Prakrit and have been specified by Prof. Jacobi.1 For example, in the Jain Prakrit, in final syllables, as well as in the middle of words, o is frequently represented by e, while in the language of the old Brahmi inscriptions and in Pali, o is nowhere represented by e. At the same time, the language of the old Brahmi inscriptions shows a complete agreement with both Pāli and Ardha-Magadhi in its tendency not to replace the Sanskrit r-sound by the l-sound.
The language of the old Brahmi inscriptions differs from ArdhaMagadhi and agrees with Pali also in its main tendency not to indiscriminately cerebralize the dental nasal.
Now let us examine the exceptional cases of spelling and pronunciation and see what results we obtain and hoy we can account for them. Let us, first of all, examine the invocation formula as it occurs in Kharavela's inscription. This was obviously a rendering from a current Jain formula, which probably survives in the later full-fledged formula: Namo arihamtānam, namo siḍhāņam, etc. In the inscriptional formula, the cerebral nasals are replaced by the dental, but the spelling arihamta is yet retained, while the commoner spelling, as evidenced by the inscriptions of Kharavela and Khāravela's chief queen, is arahamta. Both of these two spellings are met with in Ardha-Magadhi, the spelling arihamta being met with more frequently than arahamta, while only one spelling, arahanta, is met with in Pali. We also notice that, in the second clause, the cerebral
1. Preface to the Ayaramga-Sutta, edited by H. Jacobi, Part 1-Text, pp. VIII-XIV. Read also P. V. Bapat's interesting paper "The Relation between Pali and ArdhaMagadhi," Indian Historical Quarterly, Vol. IV, No. 1.

Page #186

158
OLD BRÄHMİ INSCRIPTIONS
sonant aspirate dha has been replaced by the dental dha, and we have, for the Sk. sarva, sava ( = savva), precisely as in ardha-Māgadhi, instead of saba ( = sabba) as in Pali.1
The language of the old Brāhmi inscriptions shows an agreement with Ardha-Māgadbi in its tendency to retain the Sk. d-sound instead of substituting for it the Vedic cerebral sound ? as in Pāli : kaļāra (I. 1)=Ardha-Māgadhi kačāra=Sk. kaờāra= Pāli kaļára ; kūdikā, kīđitā (I. 1)=Ārdha-Māgadhi kīờiyā, kiddā, kādıyā=Sk. krida, krīditā = Pāli kīļikā, kāļitā; kīļāpayati (I. 4)= Ārdha-Māgadhi kīļāvayai =Sk. krīdayati = Pâli kīļāpayati ; pīļāpayati (I. 9) - Ardha-Māgadhi pīdāvayai =Sk. pīdayati= Pāli pīļāpayali ; tadāga (I. 2)=Ārdha-Māgadhi tađāga= Sk. tadāga=Pāli taļāka ; veduriya (I. 10, I. 15)=Ardha-Māgadhi veduriya =Sk. vaidūrya=Pāli veļuriya.
Further, as to the exceptional cases of spelling and pronunciation, the language of the old Brāhmi inscriptions differs from Pāli and agrees with Ardha-Māgadhi in its greater tendency to replace the surd by the sonant of a consonantal group : tadāga (I. 2)=Pali taļāka= Ardha-Māgadbi tadāga; padhame (I. 2)=Pāli pathame, pathame=Ardha-Māgadhi padhame; opadha (I. 13)=Paliopatha=Ārdha-Māgadhi ”paha (a change from padha); nadha, Goradha (I. 3, I. 8, I. 9)=Pāli radha, Goratha=Ārdha-Māgadhi raha, Goraha (changes from radha, Goradha); Pithuduga (1. 12)=Päli Puthudaku=Ārdha-Māgadhi Pihudaga, Pihudaga ; samghāta (I. 12)=Pali sum khūta, samkhyāta=Ārdha-Māgadhi saņkhātu ? ; Madhuram (I. 9)=Pâli Madhuram=Sk. Muthurām=Ārdha-Māgadhi Mahurum (a change from Madhuram).
The language under discussion differs from Pāli and agrees with Ārdba-Māgadbi also in its tendency to form a gerund by adding tā (= ĀrdhaMāgadhi ttā) instead of tvā to the verb ; acıtayita (1.3)=Ārdha-Māgadhi acimtayittā=Pāli acintayitvā ; ghātāpayitā (I. 9) =Ārdha-Magadhi ghātā. vayittā=Pāli ghātāpayitvā; in its tendency to change p into v in the middle of a word : kara-vaņa-anugaha (I. 7)=Pāli karu-paņa-anugaha ; in its tendency to frequently use pana (= Ārd ha- Māgad bi panna, Pāli panata, papa, pañña) for pañca : palatĩ8@li (I. 2)=Pali pañcatiz sahi ; panatariya (1. 15) = Pāli pusca-sattali ; in its tendency to represent the r of
1. Cf. also pura (1. 5, 1, 12)= Ārdha- Magadhī puva=Pāli pubba i savata (I. 8)=ardhaMavadhi savuvattha = Pali sabbattha ; youana (1.1) = Ārdha-Màgadhi yovvana = Pali yobbana.
2. Here samyhata may be equated also with Sk. and Päli samghäto.

Page #187

NOTES
159
a conjoint consonant by ņ: dampa (1. 4)= Pali dappa=Sk. darpa 1; samdaņsayamto (I. 7) = Pāli samdassayamto=Sk. sandarśayan ; akhadamsa (IX) - Pāli akkhadassa =Sk. akşadarśa ? ; in its tendency to dispense with tsound in the middle of a word ; cavuthe (I. 5)= Ardha-Māgadhi cautthe = Pāli catuithe ; and in its tendency to represent kl by h : siharāni (1.13)= Ardha-Māgadhi siharāni = Pāli sikharāni.
If Mr. R. D. Banerji's reading dhut[u]nā be correct, we find in it another point of agreement between the language of the old Brāhmi inscriptions and Ardha. Māgadhi : dhutunā (II)=Ardba-Māgadhi dhutuņā= Păli dhītunā.
In pāờiyo (I. 2) = Sk. pālīḥ= Pāli and ardha- Māgadhi pāliyo, one has an exceptional instance, where l is represented by d= Pāli ?. In veduriya (1, 10, I. 15)=Sk. vaidūrya, one has a case, in which d is represented by d=!. In raņa (I. 7)=Sk. paņa and in dapa (if it is a correct reading)=Sk. drava, we have a case, in which p and v are interchanged. In palikhāni (I. 13)=Sk. parighān and in samghālam (I. 12)=Sk. samkhyālam, ove has a case, in which ich and gh are interchanged. In lenin=Sk. layanam, one has a case, in which the dental nasal is represented by the cere. bral. In Madhuram (I. 9)=Sk. Mathurām, one has a case, in which th is represented by dh. But these are exceptions which are met with equally in Pali. ! Regarding the use of two spellings of pāli as pāli and pāli, of Lāļa as Lāla and Lāla, and of lena as lena and ļena in Pāli, Dr. E. Müller's observations are worth quoting: “ It is very difficult to give exact rules for the use of l as the manuscripts are even less consistent in this respect than with regard to the dental and cerebral n."3
As regards palikhāni =Sk. parighān, the usual Pāli spelling is palighāni or palighāni. The palikha spelling occurs in one of the Jātaka verses (Fausbölls Jataka, Vol. VI, P. 276):–
Esikā parikhāyo ca palikham aggalāni ca. The commentator suggests paligha as an alternative reading : palikhan ti paligham, ayam era vā pātho. “Here palikha means paligha. This paligha may also be the intended reading."
1. If dapa be the correct reading, the word must be equated with Pāli and Ardha. Māgadhi dava, and Sk. drava.
2. The only exception in Pāli is lomahamsa=lomaharga, 3. E, Müller's Pali Grammar, p. 27.

Page #188

160
OLD BRAHMI INSCRIPTIONS
As for samghata=Sk. samkhyāta, we find a parallel instance in Pāli Maghadeva (Culla-Niddesa, p. 80) = Makhadeva, although here both the forms may be regarded as phonetic changes from Mahadeva, which is the usual Pali spelling.
No hypothesis regarding the character of the language should be built on the basis of the readings etinam (I. 9) and samkarakarako (I. 16). For it seems that etinam is but the engraver's mistake 1 for etisānam = Pāli etesanam, and samkara for samkhara, ef. patisamkharayati (I. 2).
One need not be surprised if the intended reading in Kharavela's inscription (I. 9) was bramhananam or brahmaṇānam, which is the same in pronunciation as the Pāli brāhmaṇānam. So far as the present inscribed surface of the rock goes, one has to read bamhananaṁ or bahmaṇānam.
Apart from the question of chronology, the following Brahmi inscriptions may be so arranged as to indicate a march of the official language of ancient India from a stage of old Magadhi towards Sanskrit through a Pali stage reached in the language of Ananda's Sanchi Gateway inscription of Satakarni's time and in that of the old Brahmi inscriptions:
I. Piprahwa Vase Inscription
II. Asoka's Rummindei Pillar Inscription
IV.
Iyam salila-nidhane Budhasa Bhagavate Sakiyanam sukitibhatinam sabhaginikanam saputadalanam.
III. Yasamata's Brick-tablet Inscription2
Devanampiyena Piyadasina lajina visativasdbhisitena atana agaca...silavigaḍabhīcā kālāpita silathabhe ca usapapite.
Jivaputaye rājabharyāye Bṛhāsvātimitadhitu-Yasamatāye kāritam.
Gotamimitra's Pillar Inscription3
...mitrasa-putrasa-raño-Visnumitrasadhitu-Īdragibhadaye dhātiye Gotamiye Mitraye danam thambho.
1. Etinam may be a genuine genitive plural form from etisa.
2. JRAS, 1912, the inscription edited by J. Ph. Vogel.
3. IHQ, Vol. II, No. 3, p. 442, the inscription edited by N. C. Majumdar.

Page #189

NOTES
161
:
V. Dhanabhūti's Barbut E. Gateway Inscription
Sugaram raje raño Gigiputasa Visalevasa pantrar Gotiputasa
Agarajasi putena lāchipulena Dhanabhūtina kūrilan toranām xilākamnamto on upamna.
VI. Ananda's Sanchi Gateway Inscription
Rūño siri-Sātakaạisa ūvesanisa Vasithiputasa Anamlasa dā
nam. VII. Khāravela's Chief-queen's Inscription
Arahamta-pasādáyar Kūlingāna samanānań lenai kāri
tam-rājino Lāli lasa Ilathisīhasa papatasa dhitunā Kalimga-cakavatino siri-khārarelasa agama hisinā kāritam.
VIII. Āşādhasona's Pabhosa Inscription, No. II--
Adhichatrāyā rāño songkāyanaputrasa Vamsa pālasa putrasya
raño Tevaniputrasya Bhāgavatasya putreņa Vaihidarīputrena Āsādhasenena lāritam.
IX. Sunga Inscription in Ayodhya 1 –
Kosaládhipena dvirasvamedhayājinaḥ Senāpateh Pusyamitrasya şaşthena Kaušikāputreņa Dhana...dharmarājñaḥa pituḥ Phalgudlevasya ketanam kāritan.
We maintain that, from the point of view of antiquity of the Pāli language, the foregoing setting of the Brāhmi inscriptions is of little importance. For going back to earlier times, we find that, upon the whole, the sound system and grammatical forms of the language of the Girnar version of Asoka's Rock Edicts are strikingly similar to those of Pāli. It appears that, for the adaptation of wordings of Asoka's edicts to the local dialects, arrangements were made in Asoka's Imperial Secretariat to have the drafts prepared by the drafters who were supposed to have been conversant with the local dialects of different centres, for which the edicts were intended. In theory, the Girnar copies of the fourteen rock edicts were meant to be drafted by the drafter familiar with the local dialect of Girnar. But in practice, in some instances, the drafters by a mutual
1. JBORS, Vol. XIII, Parts III-IV, p. 247, the inscription edited by K. P. Jayaswal. 2. Jayaswal reads dharmarājñih.
21

Page #190

162
OLD BRAHMI INSCRIPTIONS
arrangement worked for one another or, for some reasons or other, the drafter employed to prepare drafts for one place was called upon to do the work for another place, and, in some instances, the draft prepared for one place was despatched by mistake to another place. For example, by the sareness of opening words, Devānampiyasa vacanena, we may ascertain that the Dhauli copies of Asoka's Separate Rock Edicts and the copy of Queen's Edict were drafted by one and the same drafter. We think there is no better way than this of explaining the irregularities of spellings and grammatical forms in the Girnar copies of Asoka's fourteen Rock Edicts.
Among distinctive characteristics of the sound system, we notice that the Girnar language invariably retains the st-sound instead of assimilating it into th(=ith), and the st-sound instead of assimilating it into th(=tth): asti, nāsli, samstuta, stita, tistamti, anusasti, nistānāya ; invariably retains the r-sound in the word sarvata or sarvatra and optionally in the word purva, while in sava, rv is assimilated into v(=vv); and shows a greater tendency to optionally retain the ra-stop: prānā, priya, sarvatra, brāhmana, sramaņa, parākrama. The last characteristic, viz., the retention of r-stop has its vestiges in such Pāli words as yatra, tatra, atra, aññatra, amutra, bhadra, brāhmaṇa, gadrabha. In Pāli, we have sm and mh as two alternative forms, while in the Girnar language, we have the use of only one form, namely, mh: vijitamhi, a pakaraṇamhi. The retention of the ya-stop is another point of agreement between the two languages : Girnar-katanya, vyamjanato, divyāni; Pāli-Sakya, vākya, vyamjanalo, samkhyāta.
This may suffice to convince the reader of the fact that the På language was modelled on a western form of the Indian Prakritic dialect as typified by the language of the Girnar version of Asoka's rock edicts. And if the language of our old Brāhmi inscriptions is found to be very nearly Pāli, we cannot, for that reason, be justified in thinking either that it was representative of a dialect, which was prevalent in the kingdom of Kalinga, or that the Pāli language was based upon the Kalinga dialect presupposed by it.1 Apart from other arguments advanced by other scholars against Prof. Oldenberg's view, we find that, unlike Asoka's
1. Vinaya-Pitaka edited by Oldenberg, Vol. I, Introduction, p. liv.
2. Nalinaksha Dutt's Early History of the Spread of Buddhism and the Buddhist Schools, pp. 262-4 ; Sunitikumar Chatterji's The Origin and Development of the Bengali Language, Vol. I, pp. 56 ff.

Page #191

NOTES
163
edicts, our old Brahmi inscriptions bear no proof of adaptation to local dialects. In many essential points, the language of our old Brāhmı inscriptions bears likeness to that of the Girnar version of Asoka's rock edicts. We can account for this likeness without resorting to Prof. Oldenberg's hypothesis. The likeness might be simply due to the fact that our old Brāhmi inscriptions were drafted by a Jain recluse who came to live in the Udayagiri and K bandagiri caves from Gujrāt, or that the Jain recluse who composed our old Brāhmi inscriptions was won over from the Buddhist faith and utilised his knowledge of the Pāli language. To render an adequate explanation for the Pāli diction, as well as for the exceptional cases of spelling and pronunciation, we see no better way than to presume that the old Brāhmi inscriptions, as we now have them, were a rendering in a kind of Girnar language or of Pali from an original draft prepared by a Jain recluse in an eastern dialect presupposed by ArdhaMāgadhi or Jain Prakrit, and that this rendering was done by another Jain recluse in the course of rewriting it, the Jain recluse who was either born and brought up in the Girnar region or won over from the Buddhist faith having an opportunity of being conversant with a dialect similar to the Girnar language or with Pāli itself.
Whether or no the spoken dialect of the people of Kalinga was originally a Dravidian form of speech is a disputed question. But it seems certain that the language of our old Brāhmi inscriptions was not the spoken dialect of the people of Kalinga. The Udayagiri and Khandagiri caves where these inscriptions were engraved are situated in the heart of the Puri District of Orissa. The spoken dialect of the inhabitants of this District is now known as Oriya. The spoken dialect of the people of Utkala. varşa, the country of Utkala, was known to the outsiders, say to the cultured people in Magadha, as an unintelligible and uncouth jargon as might be inferred from the expression Ukkalāvassa-bhaññā which has been used as a term of contempt in one of the passages of the Pāli Tipitaka. Buddhaghosa, the Pāli commentator, has altogether missed the significance of this expression. He takes it to mean “two men of Utkala, Vassa and Bhañña by name."2 This meaning is out of the question in the Pāli passage where, while speaking of the philosophers propounding
1. Aiguttara-Nikaya, Part II, p. 31.
2. Manoratha-Pūraại, Siamese edition, Part II, p. 377: Ukkalati Ukkala-janapada. väsino. Vassa-Bhaññati Vasso ca Bhañño ca dve janā.

Page #192

164
OLD BRAHMI INSCRIPTIONS
views contrary to his own, the Buddha has referred to them as Ukkalavassa-bhañña, that is, as persons speaking in terms of the unintelligible and uncouth jargon of the country of Utkala.
Prof.
Our old Brahmi inscriptions' is a conventional language, which tended to remain clear of Magadhisms, the elements of eastern dialects.1 S. K. Chatterji rightly observes: "The Aryan language...came in the wake of the North Indian religions, Brahmanism, Buddhism and Jainism, and was used by royal patrons of these religions among Dravidians and other non-Aryans, merely as a sort of religious language, in documents of a religious, and often of an administrative character."2
3
We have sought to maintain that the Pali language, too, tended to remain clear of Magadhisms. It is most astonishing that Prof. Bapat freely admits attakare, parakūre, sukhe, dukkhe, jīva-sattame, and the like to be the genuine Pali forms, forgetting that these Ardha-Magadhisms are clearly associated in the earliest Pali texts, notably the SamaññaphalaSutta of the Digha-Nikaya, Vol. I, with the doctrines of such recluse teachers as Pūraṇa-Kassapa, Pakudha-Kaccayana and Makkhali-Gosāla, and that, as such, these are meant to have been put within inverted commas as a means of being kept distinct from the Pali forms, attakaro, parakaro, sukham, dukkham, and jiva-sattamo.1
It is interesting to observe how scholars after scholars have erred on the wrong side in subsuming that all that is in Pali is Pāli. The spelling of Raigiri as Isigili, met with in the title and body of the Pali Isigili-sutta of the Majjhima Nikaya, is cited by Prof. Lüders as a notable instance of lingering old Magadhism in Pali, and no less as an evidence in support of his theory about the rendering in the existing Pali canon of old Magadhi texts in a western dialect. But what can be more misleading than this? As we have sought to show elsewhere, the usual Pâli name of the hill would have been Isigiri, and the Isigili form had
1. The word palikhani which occurs in Kharavela's inscription as an equivalent of the Pāli palikhani or palighani, and the Sk. parighan, might be cited as an exception. But the alternative Pali spelling of paligha as paligha, met within the Vaseṭṭha-Sutta, SuttaNipata, indicates that the case is not that of replacement of the ra-sound by the la-sound but that of an inter-consonantal change effected through la or da.
2. The Origin and Development of the Bengali Language, Vol. I, p. 63.
3. Indian Historical Quarterly, Vol. IV, No. 1, p. 23.
4. See Barua's paper-The Ajivikas-in the Calcutta University Journal of the Department of Letters, Vol. II, pp. 46-51.
5. Barua's paper-Jinalogy and Buddhalogy-Calcutta Review, Oct., 1924, p. 60.

Page #193

NOTES
165
to be adopted for a very special reason, meaning it to be put within inverted commas. The reason for adoption or retention of the spelling Isigili has been stated in the Sutta itself :
“ Bhutapubbam, Bhilkhave, pañca-Pacceka-buddha-satāni imasmim Isigilismim pabbate ciranivūsino ahesum. Te imam pabbatam pavisantā dissanti, pavitthā na dissanti. Tam enam manussā disvā evam āhamsu:
Ayam pabbato ime isi gilatiti Isigili,' Isigili tveva samañña uda
pādi.”1
"Formerly, O Bhikkhus, some five hundred Eyotistic Buddhas (hermit teachers) came to live for ever i.e., to cast off their bodies) in the dark caverns of) this Isigili mountain. They could be seen entering (the caverns of) this mountain, and once they entered into the mountain, they could no longer be seen. Observing this strange happening, the people said : This mountain swallows these sages,' and hence arose the name of the hill-Isigili, 'the swallower of sages.'”
The explanation offered by the Buddha enables us to understand that the real name of the mountain which was Rşigiri or Isigiri, “ the Mountain abode of the Hermits," was locally pronounced as Isigili, and acquired a new association of ideas in the fanciful etymological derivation “Isigilatiti Isigili," “the Hermit-swallower because it swallows the hermits." and that this new association could not be embodied without coining such a longer name in Pāli as I sigili-giri=Rşigili-giri, “the Hermit-swallowermountain."
As we have seen, the substitution of the Dental sonant aspirate for the Dental surd aspirate in such words as padha, radda, Goradha, padhama, and Madhurā is a characteristic of the dialect presupposed by the language of our old Brābmi inscriptions, that is to say, of the language of the supposed first draft. Among the later Prakrit dialects, the Sauraseni, the vernacular of the people of Śūrasena or Mathurā region, has alone been characterised by this kind of phonetic change in Vararuci's PrākṣtaPrakāśa, in the aphorism (XII. 3) : " The Dental consonants la and tha which are not the initials of any words and are not conjoined with other consonants change respectively into da and dha."
One need not be astonished if the first draft of Khāravela's inscription was prepared in the dialect of the Mathurā region by a Jain recluse who
1, Majjhima - Nikāya, Vol. III, Part I, p. 68.

Page #194

166
OLD BRĀHMI INSCRIPTIONS
was familiar with it. The two inscriptions discovered and published by Mr. Jackson in the Barābar Hills offer us an instance of the substitution of dha for tha.? In the earlier inscription, the letter-forms of which are similar to those of Asokan inscriptions, the name of the particular bill to which it is attached is recorded as Gorathaqiri, while in the later inscription, the letter-forms of which are strikingly similar to those of Khāravela's inscription, the paine of the sau.e hill has been recorded as Goradhagiri.?
Mr. R. D. Banerji offers this explanation for the use of the spelling Goradhagiri in the second inscription : “ The substitution of dha for tha shows that the second record was incised by an inhabitant of Southern India. It is quite possible that this record was incised by an inbabitant of Kalinga, probably one of the men who had accompanied Khāravela in bis first campaign against the king of Magadha,”3
We find it difficult to agree with Mr. Banerji in thinking this kind of substitution was a peculiarity of the spoken dialect of the people of Kalinga in view of the fact that in none of Asoka's inscriptions in Kalinga we notice it. If it be true that the second record was actually incised by one of the men who had accompanied King Khāravela in his campaign against the people of Rājagaha, the fact may be explained differently. The author of the record might be a man who was brought up in Mathurā or the Mathurā region. Our explanation is more plau on the ground that, according to the Hāthi-Gumphā record, Mathurā was used by King Khāravela as the military base in his campaign against the King of Anga-Magadha, and no less in his campaign against the kings of Uttarāpatha. 4
We may, on another ground, maintain that the presupposition of the language of the supposed first draft of Khāravela's inscription was not the spoken dialect of the people of Kalinga. In Khāravela's inscription, we have cavuthe (1. 5) for the Pāli catutthe. We find that cāvudasa occurs in all the versions of Asoka's fifth Pillar Edict as a common spelling for cātuddasa. The dialects of all the versions of Asoka's Pillar Edicts point, beyond any doubt, to the existence of a widely prevalent lingua franca, or language of the cultured laity as Professor Rhys Davids would like to
1, JBORS, Vol. I, Part II, pp. 159-71. 2. The point is ably discussed by R. D. Banerji in JBORS, Vol. III, Part IV, p. 500. 3. JBORS, Vol. III, Part IV, pp. 500-501. 4. See for fuller discussion passim, under Geographical Allusions.

Page #195

NOTES
167
call it, in the third century B. C., in the Middle Country extending as far, say, as Kaušām bi and Mathură. It is apparent that the dialects of these Pillar Edicts are full of Māgadbisms. And if we go by the verdict of Vararuci, the predominant tendency of the Māgadhi dialects was Sauraseni, that is to say, of the dialect of the people of Mathurā.
Thus we are led to assume that the Udayagiri-Kbandagiri caves in Orissa, Mathura and Ujjeni-Girinagara were the three important centres of Jainism during the reign of Khāravela, and that the language of Kbāravela's inscription is, so far as its grammatical forms go, the same as the dialect of the Girnar version of Asoka's Rock Edicts, and, so far as its sound system goes, a combination of the Girnur and Mathura dialects.
The substitution of dha for tha cannot be said to be a peculiarity of the Pāli language on the ground that in Pāli, too, we have Madhurā as a spelling for Mathurā, just in the same way that we have in it Isigili as a spelling for Isigiri, and Makhādeva and Maghādeva as two spellings for Mahā. deva. All these were locally current proper names retained in Pāli, and meant to be put within inverted commas, that is to say, to be kept distinct from the standard Páli spellings. The spelling of the name MādhavaVideha as Māthava-Videgha occurring in the Satapatha-Brāhmana (I. 4. 1) is an apt parallel in Sanskrit literature. This spelling shows that the personal name Mādhava-Videha was locally pronounced, most probably by the inhabitants of Videha or Mithilā region, as Māthava-Videgha.
We mean to say that the spelling Madhurā was not due to a Pāli rendering of the sk. Mathurā,--that, in other words, Madhură was a janapada-nirutti or deši-nāma, that is to say, a locally current proper name, which found recognition in Pali. In accordance with a significant statement made by the Buddha in the Aranavibhanga-Sutta of the MajjhimaNikāya (Vol. III, Part III), one locally current proper name, if it signifies an object for which it is intended, is as good as the other, and there is no sound reason for regarding one of them as more correct than the other. The importance of his statement lies also in the fact that it contains a reasonable explanation for the recognition of a particular form of the proper name not from any intrinsic superior value of its own but on account
1. Vararuci's Prákyta-Prakāśa, XI.2 : Prakrtih Sauraseni.
2. See Actes du Sisieme Congres International des Orientalistes, Part III. p. 140, where Bhagawanlal Indraji remarks: "The whole inscription is in prose. Its language is Prakrit, different from the Lāt (Pillar) inscriptions of Asoka, but resembling the old Mahārāsýra Prakrit of the Western India cave inscriptions." This characterisation is too general to need comments.

Page #196

168
OLD BRAHMI INSCRIPTIONS
of local adaptation. We quote below his statement in ertenso to enable the reader to judge for himself what it is and what it implies :
Janapada-niruttim nábhiniveseyya, samaññam nátidhāveyyati—iti kho pana etam vuttam. Kiñc'etam pațicca vuttam? Kathañ ca, Bhikkhave, janapada niruttiyū ca abhiniveso hoti samaññāya ca atisāro ? Idha, Bhi. kkhave, tad ev'l ekaccesu janapadesu 'Päti'ti sañjānanti, Pattan'ti, Vitthan' ti, Sarāvan'ti, 'Dhāropan'ti, 'Ponan'ti, 'Pisilan'ti sanjananti. Iti yatha yathā nam tesu tesu janapadesu saījānanti tathā tathā thāmasā parāmassa abhinivissa voharati? : "Idam cvu saccam, mogham aññan" ti. Evam kho, Bhikkhave, janapada-niruttiyā ca abhiniveso hoti samññāya ca atisāro.3 Kathañ ca, Bhikkhave, janapada-niruttiyā ca anabhiniveso hoti sam. aññāya cu anatisūro ? Idha. Bhikkhavo, tad cv'ekaccesu janapadesu "Pūtí'ti sañjānanti, Pattan'li, "Vitthan'ti, Sarāvan'ti, Dhāropan'ti, Ponan'ti, 'Pisilan'ti sañjūnanti. Iti yathi yathā nam tesu tesu janapa. desu sañjānanti “Idam kira'me āyasmanto sandhāya voharanti" ti tathā tatha voharati aparāmasam.4 Evam kho, Bhikkhave, janapada-niruttiyā anabhiniveso hoti samaññāya ca anatisaro. Janapada-niruttim náb hiniveseyya, samaññam natidhāveyyati-iti yam tam vuttam idam etam pațicca vuttam.
" The local form of a proper name is not to be dogmatically adhered to the local designation is not to be pressed too far. Such is the rule as stated. For what reason is the rule so stated ? And how is it, О Bhi. kkhus, that a man becomes inclined to dogmatically adhere to the local form of a proper name, and to press the local designation too far? Here, 0 Bhikkhus, it so happens that in some locality a utensil is known by the name of Pāti, in some by the name of Patta, in some by the name of Vittha, in some by the name of Sarūva, in some by the name of Dhāropa, in some by the name of Poņa, and in some by the name of Pisila. The inhabitant of a particular locality having strongly embraced and dogmatically adhered to a particular form of the proper name whereby the utensil
1. Buddhaghoşa's Papasca-Südani, Siamese edition, Part IIT, p. 471 : Tad evan tam yeva bhajanan.
2. Ibid, p. 471 : Abhinivissa voharatiti 'Pattan'ti-sanjānana-janapadam gantva "Pattam aharatha dhovatha'' ti sutvā "Andha-puthujjano, nayidam 'Pattan' 'Pati' nama eså, eram vadāht" ti abhinivissa voharati.
3. Ibid, p. 471 : Atisaro'ti abhivādanan (a misprint for atiahāvanam).
4. Ibid, pp. 471-2: Tathā tathā voharati aparamasanti " Amhākam janapade bhajanan Pate'ti' puccati, ime pana nam 'Pattan'ti vadanti. Tato patthāya janapada. voharam muncetvā pattam 'Pattan' t'eva aparūmasanto voharati.

Page #197

169
is known in this or that locality, boastfully says: This is the only correct form, and the others are incorrect.' Thus it is, O Bhikkhus, that a man becomes inclined to dogmatically adhere to the local form of a proper name, and to press the local designation too far. How is it, O Bhikkhus, that a man does not become inclined to dogmatically adhere to the local form of a proper name, and to press the local designation too far? Here, O Bhikkhus, it so happens that a utensil is known by different proper names in different localities, in some by the name of Pati, in some by the name of Patta, and so on and so forth. The inhabitant of one locality, when he is in another locality, realising that the gentlemen of the second locality conventionally use this form of the proper name to designate this object,' wisely use that particular form whereby the object is known in that particular locality without any local attachment (that is, abandoning the form whereby the object is known in his own locality). Thus it is, O Bhikkhus, that a man does not become inclined to dogmatically adhere to the local form of a proper name, and to press the local designation too far. It is for this reason that the rule is so stated as: The local form of a proper name is not to be dogmatically adhered to, the local designation is not to be pressed too far." "1
22
NOTES
1. Here we have refrained from introducing a discussion as to the origin and antiquity of Pali language, reserving it for a separate monograph. But it is our decided opinion that a conclusion about the antiquity of Pali drawn from the similarity observed between it and the language of Kharavela's inscription is bound to be a dogmatism and dangerous presumption. Truman Michelson's paper on Magachisms in the Language of the Girnar, Shahbazgarhi and Mansehra Inscriptions' (American Journal of Philology, 1909, pp. 234 ff., and JAOS, 1909, I, pp. 77 ff.) will receive full attention in the proposed monograph.
4

Page #198

170
OLD BRAHMI INSCRIPTIONS
7. THE STYLE
Our old Brāhmi inscriptions are all written in prose, a rhythmic prose abounding in alliterations, elegant expressions, and balanced sentences, clauses and phrases. In reading these inscriptions, especially those which stand in the names of King Khāravela and his chief-queen, one cannot but be tempted to make out verses in them. We venture to say that all attempts made in this direction are destined to end in failure. Their diction is metrical prose without revealing the actual process of versification. In reading out the inscription of Khāravela one is apt to feel as though one were chanting verses in marked cadences, the invocation formula in a variety of Kumāra-Lalita metre, and the main text in a kind of Simhavikridita,
Scan the invocation formula as carefully as you may, you cannot confidently class it as a metrical composition in any of the known metres, and yet your inclination will all along be to trace in it a process of versification in the Kumāra-Lalita metre. The fact is that the desire to produce a complete rhythmic effect has led the composer to balance up the groups of sounds in successive sentences, clauses or phrases, inducing a tendency to versification within a prose construction. So far as its greater rhythmic effect goes, the inscriptional formula || Namo ari hamtānam || Namo savasidhānam || shows an improvenient upon an earlier Jain formula || Ņamo arihamtūnam || Namo sidhanam || which seems to linger in the later fullfledged Jain formala || Namo ari hamtānam | Namo sidhānam || Namo āyari. yāṇam || Namo uvajhāyānam || Ņamo loe savra-sāhūņam || quoted on p. 8 (ante). Comparing the inscriptional formula with the later Jain formula, we detect that after the monotonous repetition of four similar clauses, the latter is closed with a clause || Ņamo loe savva-sāhūnam || constructed so as to maintain the rhythm of the whole, and that, in this final clause, one has a two-syllabic word in addition to those contained in the second clause of the former. In the formula consisting of two clauses, the insertion of one two-syllabic word suffices for the purpose, while in the formula consisting of five clauses, the insertion of an additional word is needed.
1. Inscriptional formula
|| Namo arihamtānam || Ņamo sava-sidhānam ||
11. Later Jain Formula || Ņamo arihamtāṇam || Ņamo siờhāņam || Namo ārjariyānam ||
Il Ņamo uvajhāyānam || Ņamo loe savva-sāhūņam ||

Page #199

NOTES
171
A similar Buddhist formula of invocation consisting of three clauses can be traced in the Petakôpadesa which is one of the extra canonical works presupposed by Buddhaghosa's Atthasalini, and probably also by the Milindapanha.
III. Buddhist Formula in the Peṭakôpadesa
Namo sammāsambuddhanam || Paramatthadassinam || || Siladiguna-paramippattānam ||
In the Buddhist formula, the first two clauses joined together read like a line of a verse, while the third clause shows that the composition is yet rhythmic prose. The real character of the composition will be manifest if we supply the omissions and read it as follows:
Namo sammasambuddhanam || [Namo] paramatthadassinam || || [Namo] siladigna-paramippattanam |
Coming to the main text of Kharavela's inscription, we find that the effect of rhythm is heightened by a mathematical progression of the volumes of sound, and that the main statement commences from the point where the climax is reached. In such a text as this the verbs are bound to be sparingly used, and a rhyming process is bound to play its part as will be evident from the following quotation :
I. (1)|| Airena mahārājena || mahāmeghavahanena || Ceta-rajavamsavadhanena pasatha-subha-lakhanena || caturamta-rakhana-guna-upetena || Kalimgddhipatina siri-Khararelena || pamdarasa-vasāni || siri-kaḍāra || sarī. ravata kiḍita kumāra-kiḍikā ||
The rhyming process plays its part also in a sentence consisting of several clauses, each with a separate verb :
1. (2)-pākāra-nivesanam patisamkharayati || tadagapadiyo ca bamdhāpayati || savāyāna-patisamṭhāpanam ca kārayati || pakatiyo ca ramjayati ||
In the text of the inscription of Kharavela's chief queen, the same verb is repeated for the sake of rhythm and emphasis :
II- Arahamta-pasadayam Kalimganam samaṇānam leņam kāri. tamrajino Lalákasa Hathisihasa papotasa dhituna Kalimga-cakavatino siri-Kharavelasa aga-mahisinā kāritam ||
The rhythm is sought to be maintained even in such a short inscription as that of Kamma and Khina :
VII- Kammasa kotha ca | Khīņāya ca pasado ||

Page #200

172
OLD BRAHMI INSCRIPTIONS
For the sake of rhythm the words are left as they are without being joined according to the rules of Sandhi, e.g., guna-upetena (I. 1), and tivasa. sata-oghāțitam (I. 6).1
So far as these peculiarities are concerned, our old Brāhmi inscriptions clearly anticipate the Pali prose style of the Milindapañha, another of the extra-canonical works presupposed by Buddhaghoşa's commentaries, which, even according to the tradition embodied in it, was not composed within the first five centuries of Buddha's demise. And so far as these peculiarities are concerned, there is nothing save a few long-drawn conpounds within the four corners of the Pāli Tipitaka to anticipate the prose style of our old Brāhmi inscriptions. The Milinda descriptions of the city of Sāgala and the earthquake signalising the memorable character of Vessantara's charity, quoted below, will, we believe, show how close is the resemblance of its prose style with that of Khāravela's inscription, in spite of the fact that the prose style of the latter has not, as yet, attained the maturity of the prose style of the former :
I. Description of Sāgala--
Atthi Yonakānam nānāputabhedanam Sāgalam nāma nagaram nadi. pabbatı-sobhitam ramaniya-bhūmippadesabhāgam ārām-uyyānopavanatalāka-pokkharani-sampannam nadi-pabbata-vana-rāmaneyyakam, etc.
II. Description of the Earthquake
Hetthā mahāvātā sascalanti, sanikam sanikam sakim sakim ākulākulā vāyanti, onamanti unnamanti vinamanti, sinappattā pādapā papatanti, gumbagumbam vaiāhakā sandhāvanti,...rudanti yakkhā appesakkhā, hasanti yakkhā mahesakkhā kampamānāya mahāpathaviyā.
Our old Brābmi inscriptions' is not the prose style of the Pali Tipitaka, nor that of earlier portions of the Jaina Āgama, nor that of the Vedas, Brāhmaṇas, older Upanişads, Kalpa-Sūtras, Niruktas and Prātisākhyas. So far as their prose style goes, they stand out, in point of time, as a notable landmark in the literary history of India. Just as with reference to the accidental unconscious beginning and maturer development of the style of Kāvya poetry Asvaghoşa's Buddha-Carita stands midway between
1. Cf. Vasa-abhisita or vara-abhisita (Asoka's R. E. XIII, Sabbazgarbi; P. E, V Delbi-Topră, Delhi-Mirath; P. E. VI, Delhi-Toprā).
For Private And Personal Use Only ----------

Page #201

NOTES
173
the Prologue of the Pali Nālaka-Sutta as found in the Sutta-Nipāta? on one side and Kālidāsa's Kumāra-Sambhava on the other, so with reference to the accidental unconscivus beginning and maturer development of the rhythmical prose style, our old Brāhmi inscriptions represent a link of transition between some of Pāli set formulas of Buddhist precepts and the Milinda's descriptions of the city of Sāyala and the earthquake. The Pāli formulas representing an accidental unconscious beginning of the prose style of Khāravela's inscription are being quoted below :
I. Pāli Formulas of Buddhist Precepts(a) Nacca-gīta-vādita-visīka-dassanā-veramaņi
(Vinaya Mahāvagga) Nacca-gita-vādita-visūka-dassanā-veramaņi-sikkhāpadam
(Khuddaka-Patha) Nacca-gita-vādita-visuka-dassanā-veramani-sikkhāpadam samādi
yāmi
(Khuddaka-Patha-Comy.) (b) mālā-gandha-vilepana-dhāraņa-mandana-vibhūsana-tthānā veramaņi
(Vinaya Mahāvagga] mālā-gandha-vilepana-dhāraņa-mandana-vibhūsana-tthānā-vcramani-sikkhāpadam
(Khuddaka-Patha) mālā gandha-vilepana-dhāra na-mandana-vibhūsana-tthūnā veramaņi-sikkhāpadam samādiyāmi
(Khuddaka-Pātha-Comy.]
1. We mean that the Sutta-Nipāta contains one of the two versions of the discourse in I ndian languages, the other version being found in the Mahāvasta, III, pp. 386-87. If Dhammananda Kosambi's identification be correct, as we believe it is, this is the very discourse which was recommended for study by King Asok a in his Bhabru Edict under the title "Moneya-Sute." The verses of the Prologue appear to be a supergrowth and later addition. As to the relation between the Pāli Prologue and the Buddhacarita, the follow. ing quotations will suffice :(4) Pāli PrologueDadallamānam siriya anomarannam
dassesi puttam A sitaphayassa Sakya. Disvä kumāram sikhim iva pajjalanttam
tārāsamam da nabhasigamam pisuddham. (b) Buddha-Carita, III, 23 :Dretvà ca tam rajasutam striyastā jājvalyamānam vapugā óriyā ca.

Page #202

174
OLD BRAHMI INSCRIPTIONS
II. Kbāravela's Inscription
I. 4-Dampa-nata-gita-vādita-samdasanāhi kidāpayati nagarim. I. 1-Sava-kara-vana.anugaha-anekāni-sata-sahasāni visajati pora-jāna
padam. Striking, indeed, is the resemblance in respect of the prose style and letter-forms which exists between the Barhut inscriptions attached as labels to two companion suenes depicting one of the Buddha-legends presupposed by those in the Lalita-Vistarat and the Hāthi-Gumphā inscription of King Khāravela, although the latter shows a maturer development :
I.
Barhut Inscriptions
(a) Utaram disa tini savata-nisisāni.
Dakhinam disa cha-kāmávacara-sahasāni. (6) Sāờikarsammadam turam devānam.
II. Hāthi-Gumphā Inscription
(a) I. 3-Sātakanim pachima-disam haya-gaja-nara-radha-bahulam
damdam. (6) I. 12-Anupadabhavanam ca terasa-vesa-sata-katam bhidati
tamiradaha-samghātam.
- 1. Cunningham's Stūpa of Bharhut, Pl. XIV, S, Gate, Prasenajit Pillar, Middle and Lower Bas-Reliefs. The point is discussed by the author in the Buddhist India, a Buddhist Quarterly edited by Barua and Dharmacharya, Vol. I, No. 3.

Page #203

NOTES
8. THE CONTENTS
175
The contents of the Hathi-Gumpha inscription entitle it to a unique position. Among the remaining inscriptions, No. XV contains, as we saw, a table of Brahmi alphabet and the rest are short epigraphs to label the caves to which they are attached, or to which they are intended to refer, with records commemorating the names of their excavators or builders, that is to say, of their donors. In two instances (Nos. II and XIII), the records expressly mention the class of persons for whom the caves are built, the donee to whom the caves are meant to be dedicated. In so far as these records commemorate the names of the donors and the donee, they remind us of the Barabar Hill Cave inscriptions of King Asoka and of the Nagarjunt Hill Cave inscriptions of King Dasaratha. And in so far as these records stand without the expression dana or dina signifying the act of dedication, and emphasize the excavation or building work in the verbal expression karita, they remind us of the Rummindei Pillar inscription of King Aśoka and the Barhut E. Gateway inscription of King Dhanabhūti. But in the majority of the cases the inscriptions are so worded as to indicate that the caves are intended to bear the names of their excavators, to be known as the caves of so and so, of King KadampaKudepa, of Prince Vadukha- Varikha and the like.
The reader will entirely misjudge to differentiate the bulk of the remaining inscriptions from Nos. II and XIV on the ground that they commemorate the names of the excavators or builders of the caves but not those of the persons for whom the caves are excavated or built. We would say that the mention of the persons for whom the caves are excavated or built is unnecessary in these inscriptions. The fact that these caves are all meant for the residence of the recluses in Kalinga (Kaliṁgānam samaṇānam) is implied in each one of these inscriptions. For reading with reference to the context supplied in the Hathi-Gumpha record of Kharavela's thirteenth regnal year, we find that all of the 117 caves are meant to serve as kaya-nisīdī or resting places of the Arhats who happen to dwell on the Kumari hill (Kumari pavate Arahato parinivasato hi kayanisīdīyāya).
It is not only for finding out the context of the remaining inscriptions and for understanding the specific purpose of the 117 caves excavated during the thirteenth year of Kharavela's reign that one is to turn one's attention to the Hathi-Gumpha inscription. One has to read the last-mentioned

Page #204

176
OLD BRĀHMI INSCRIPTIONS
inscription as well for a general record of all the architectural construotions done during Kbāravela's reign and under his auspices. This is again not all. This inscription presents a systematic record of all the notable achievements of Khāravela, including all the works of art andil architecture done under his auspices. And this, too, will be but an imperfect and inaccurate characterisation of the contents of this inscription.
The Hatbi-Gumpha inscription appears to be a systematic record of K bāravela's personal history and successive achievements. The recorded facts are presented in the garb of an autobiographical sketch. The concluding paragraph is so designed as to make the record appear as closed with the name of King Khāravela-siri, that is, to create the impression that the record is written and signed by the king with his own hand. The invocation formula with which the record begins is meant to indicate the nature of the religious faith of the king.
The opening statement contains a succinct account of how King Khāravela spent the first twenty-four years of his life, the first fifteen years as a young prince and the remaining nine years as a crown-prince, previous to bis installation in the throne of Kalinga, which took place after he had completed his twenty-fourth year. Here certain relevant details are given as to the physical feat ires and other kingly qualities of Khāravela, the name of the royal family of which he was the scion, and the useful sciences and arts in which he acquired proficiency.
This is followed by the first year's record containing an account of how, in the very first year of his reign, King Khāravela undertook to repair the capital of Kalinga, and to do all in order to please bis subjects. The record expressly mentions the large amount of money spent by the king to do this work.
The second year's record contains an account of how the king, in the second year of his reign, defied King Sātakarni in marching towards the western quarter with a large army to attack the city of Asvaka or Rşika.
The third year's record contains an account of how the king, in the third year of his reign, entertained the city of Kalinga with musical performances and general feasts and festivities.
The fourth year's record contains an account of how the king, in the fourth year of his reign, consolidated his rule in Arakatapura or some other place which was inhabited by the Vidyādharas, an aboriginal race of men,
The fifth year's record contains an account of how the king, in the fifth year of his reign, facilitated communication by effecting an extension
**

Page #205

NOTES
177
of the old canal opened out by King Nanda so as to lead it into the capital city from the Tanasuliya or Tanasuli road. This record indefinitely hints at the amount of money spent for the purpose.
The sixth year's record contains an account of how the king, in the sixth year of his reign, showed unprecedented favours to inhabitants of the towns and districts by remitting all taxes and duties.
The seventh year's record contains an account of how the king, in · the seventh year of his reign, performed all the ceremonies of victory as a public demonstration of his royal pomp and power.
The eighth year's record contains an account of how the king, in the eighth year of his reign, after having killed or stormed Gorathagiri, brought a heavy pressure to be brought upon the people of Rājagļha and effected a timely retreat to Mathurā in order to relieve his troops terrified by the uproar of counter-attacks, and sumptuously feasted, in Mathurā as well as in Kalinga, all sections of people, those who kept to household life or those who turned ascetics, those who belonged to the Brabmanical orders or those who belonged to other religious orders. This record, too, mentions the large amount of money spent for the purpose.
The ninth year's record contains an account of how the king, in the ninth year of his reign, undertook to build a magnificent royal palace known by the name of “Great-Victory-Palace.” This record expressly mentions the large amount of money spent for carrying out the project.
The tenth year's record contains an account of how the king, in the tenth year of his reign, paid proper homage to the memory of the former kings of Kalinga at the cost of a large amount of money.
The eleventh year's record contains an account of how the king, in the eleventh year of his reign, reclaimed and rehabilitated Pșthudaka, founded by the former kings of Kalinga, by arranging to drive out its watery jungle of grass into the Lârgala river.
The twelfth year's record contains an account of how the king, in the twelfth year of his reign, terrorised the rulers of Uttarāpatha, subdued Bșhaspatimitra, the king of Magadha, forced the inhabitants of A nga and Magadha into submission, brought back the throne of Jina from Anga-Magadha to Kalinga, and made improvements of the capital city by opening new roads and squares and adding gate-bars, gate-houses and towers. This record also contains an accouut of how the Vasukis and the king of Pāņdya sent valuable presents to Khāravela.
The thirteenth year's record contains an account of how the king, in the thirteenth year of his reign, completed the excavation of 117 caves
23

Page #206

178
OLD BRĀHMI INSCRIPTIONS
under joint auspices of himself, his queens and others on the Kumāri hill for the resting of resident Jain saints.
The fourteenth year's record contains an account of how the king, in the fourteenth year of his reign, excavated a special cave for the residence of the recluses, ascetics and hermits from all quarters, caused stonepillars and shrines to be made with stone-slabs quarried out of best quarries, and set up ornamented pillars in beryl chambers at the cost of a huge sum of money.
The concluding paragraph which is but a long string of nicely woriled and choicest adjectives heaped upon the name of King Khāravela-siri is evidently a literary device to represent the sovran lord of Kalinga as the noblest type of kingly personality and the greatest and best of earthly warrior heroes. He is styled Khemarāja, the Lord of Security. He is styled Vadharāja, the Lord of Prosperity. He is styled Indarāja, the Lord of Kingly Power. He is styled Dhammarāja, the Lord of Justice. He is represented as a person who had the ripeness of understanding and judgment of the nature of what is conducive to human welfare. He is represented as a person gifted with special qualities; as one who honoured all sects and denominations; and as one who repaired all religious temples. He is represented as a descendant of a family of royal sages. He is represented as a great conqueror. He is represented as a most powerful king who maintained the prestige of his illustrious predecessors, who had the ability to protect his kingdom, and who secured services of the ablest and fearless ministers and officers. The Pāli Upāli-Sutta in the Majjhima-Nikāya goes to show that this mode of praising the acknowledged high personality or this exaggerated mode of bero-worship was Jaina, as will be evident from the following quotation :
The householder Upāli who happened to be formerly a Jain by faith is represented as saying to Nigaạțba-Nātaputta, the historical founder of Jainism, in praise of the Buddha in whom he subsequently took refuge :
Dhirassa vigatamohassa pabhinna-khilassa vijita-vijayassa | anigbassa susamacittassa vuddhasilassa sādhupaññassa || vessantarassa vimalassa Bhagavato tassa sävako' ham asmi
Tanhacchidassa buddhassa vitadhumassa anupalittassa āhuneyyassa yakkhassa uttamapuggalassa atulassa || mahato yasaggappattassa Bhagavato tassa sävako' ham asmi 11
1. Majjhima-Nikāya, Vol. I, p. 315. Upāli calls himself sävaka in the sense of " a lay disciple," which accords with Jain phraseology, and not with Buddhist,

Page #207

NOTES
179
It is interesting indeed that most of the high-sounding adjectives in the Hathi-Gumphā record are to be found in Upāli's eulogium of the Buddha : khemarāja=khemankarı; vadkarāja =vuddasila ; imdarāja= purindada sakka ; dhammarāja=dhammattha; mahāvijaya =vijitavijaya ; apatihata-caka-vāhana-bala= appatipuggala; rājisi-vamsa-kula-vinisita= isisattama ; guna-visesa-lensala=visārada nipuna ; pasamta sunamta anubhavamta kalānāni = satimā vipassī viditaveda.
The concluding paragraph clearly brings out the fact that Khāravela's autobiographical epigraph was composed for him by a skilled composer, to whom the task of composition was entrusted. The composition must have received the warm approval of His Majesty before it was incised on the rock and set up on a hanging brow wherefrom it might attract the visitors and pilgrims of the Kumāri hill. Thus in one important
respect the Hathi-Gumphā record of Khāravela differs from the edicts · of Asoka, and the Dialogues of the Buddha, namely, that in it one misses
the personal touches of the personage in whose name it stands. Both the edicts of Asoka and the Dialogues of the Buddha make one feel as though their texts were written to dictation by some reporters and were afterwards edited by certain agents with slight changes here and there, either in the sound-system or in the expressions, the general method of the editorial agents having consisted in (1) the use of Devānampiye Piyadasi Lājā hevam āha or a similar set clause as a literary device for paragraphic divisions ; and (2) the conversion of a direct narration into an indirect one by substituting Devānampiye for lājā, Devānampiyena Piyadasinā läjina: for me, mayā and mamayā, and Devānam piyu sa piyadasino lājino for me and mama, precisely in the same way as in the Dialogues of the Buddha Tathāgato is substituted for aham, Tathāgatena for me, mayā, and i Tathāgatassa for me, mama.
The Artha-Sāstra of Kautalya-Kautilya (II. 10, 31) prescribes certain rules of composition to be observed in drafting the royal writs. It specifies the principal qualities and defects of composition and hand. writing. In accordance with the prescription in the Artha-Šāstra, the defect of composition (lekha-doşa) consists in (1) vyāghāta—irrelevancy; (2) punaruktam--repetition ; (3) apaśabda--non-grammatical and unidiomatic use of words and expressions ; and (4) samplava-irregularity. According to the same authority, the qualities of composition (lekha-guna) consist in (1) arthakrama—the maintenance of syntactical order and logical sequence ; (2) sambandha-consistency of meaning ; (3) paripūrņatā-sufficiency; (4) mādhurya-elegance or exquisiteness ; (5)

Page #208

180
OLD BRAHMI INSCRIPTIONS
audarya-refinement or propriety; and (6) spaṣṭatva-lucidity or expressiveness. The fourteenth rock edict of Asoka goes to show that, as early as the third century B, C., the Maurya emperor was aware of these merits and defects, at least of the fact that repetition was a defect of composition in a royal writ which might be excused only if it was needed for emphasis and elegance.
Explanations were required for repetitions and certain omissions, errors and irregularities in the edicts of Asoka. But as regards Kharavela's inscription, the composition of its text is free from all the defects mentioned above and is bright with the required qualities.

Page #209

NOTES
181
9. THE RELATIVE CHRONOLOGICAL POSITION
The Hathi-Gumphã inscription was heretofore judged as an epigraph, the final record of which did not extend beyond the thirteenth year of Khāra vela's reign. But a weighty reason has now been found compelling us to believe that its final record extends beyond the thirteenth year, though by no means beyond the fourteenth. Its record of the thirteenth year opening with the words Terasame ca vase begins in the middle of the left half of 1, 14 and continues to the end of the same. This particular record cannot be said to run as far as I. 15, nay, to extend as far as the left three-fourths of the total length of l. 16 as previously supposed. The thirteenth year's record in 1. 14 contains a statement about the construction of 117 caves on the Kumāri hill, the twin-bills of Udayagiri and Khandagiri, a joint work of Khāravela and others, while the record in 1. 15 and l. 16 contains a statement about some costly works of art and architecture done by Khāravela himself on the defilement of the mountain, in the proximity of the resting place of the Jain saints. Further, the thirteenth year's record in I. 14 presents a grammatical construction in Passive Voice, while the record in 1. 15 and 1. 16 presents a construction in Active Voice.
1. L 14--Terasame ca vase......satadasa-lena-satam kārāpitam. 2. L 15 and 16 kūrayati......patithāpayati...upādayati.
If the foregoing reasoning be sound, there is little doubt that the final record of the Hathi-Gumphā inscription extends as far as and not beyond the fourteenth year of Khāravela's reign. And in the absence of any. thing to prove the contrary, there is little difficulty in representing the Hathi-Gumphā inscription as an epigraph, which was incised or put up in Khāravela's fourteenth regnal year. That is to say, this inscription was set up on a hanging brow of rock on the Udayagiri hill, in front of the Hāthi-Gumphā, about a year after the excavation of 117 caves with their inscriptions and architectural constructions.
In accordance with the thirteenth year's record, some out of 117 caves were excavated by Khāravela’s queens, some by his sons, some by his relatives, some by his brothers, some by the royal servants, and the rest by himself. Going by this statement, we cannot help thinking that the caves and inscriptions standing in the name of Khāravela's chief queen, of King Kadampa-Kudepa, of Prince Vaçlukha-Varikha, of the Town-judge Bhūti, of the menial Kusuma, and of others were works, which were

Page #210

182
OLD BRĀHMI INSCRIPTIONS
commenced and finished in Khāravela's thirteenth regal year. If so, is it correct to argue on the peculiarity of certain exceptional letter-forms and to draw such a conclusion therefrom as that one among the fifteen old Brāhmi inscriptions containing an exceptional letter-form reminding us of one in the inscriptions of Asoka is earlier by a century than the other ? Will it be correct, for instance, to say with Mr. R. D. Banerji that inasmuch as in the inscription of Prince Vaukha-Varikha (No. IV), the u-sign is quite distinct, though very small in size, and the letter kha “has neither a triangle nor a circle at its base," the inscription itself is anterior to that of King Kadampa-Kudepa (No. III)? We may venture to think that Mr. Banerji can no longer maintain his position, and that he will be prepared to appreciate the force of our argument pressed in favour of accepting the inscriptions standing in the names of Khāravela's chief queen, King Kadampa-Kudepa, Prince Vaļukha-Varikha, and the rest as (contemporary votive records, the records incised in Kbáravela's thirteenth regal year.

Page #211

NOTES
183
10. THE SLOKAS CONCERNING KING AIRA OF UTKALA
Mr. Jayaswal's search for light from literature on the history of Khāravela has proved a success, at least, in so far as it has enabled him to obtain a few Sanskrit slokas from an Oriyā Ms., which is lying unedited in the archives of the Indian Museum, Calcutta. These are seven in number, and the MS. itself, though described as “old,” may, in the opinion of experts, be safely assigned to the 16th century A.D., and can by no means be dated earlier than the 14th century. The slokas, as cited by him in JBORS, 1917, p. 482, and reproduced by Dr. Sten Konow in Acta Orientalia, Vol. 1, read as follows :
MS. LEAF B (OBVERSE)
1. Abiro nāma rājábhūt côtkale vidyate purā
abiņsā-dharmamāśệtya Buddha dharına-parāyaṇaḥ # 2. Nandarāja suvikhyātaḥ Magadhe vidyate tadā
sakāra-pāśako Nandaḥ Veda-dharma-parāyaṇaḥ || 3. Nandasya sahito yuddbe Airo jitavän bhavet,
Airo jayam āpnoti mabāhsstena mänasaḥ 1 4. Svadharma côtkale khyātiḥ Veda-dharma-vināśakaḥ
Asokasya mahāmittrah Airaḥ Utkaleśvaraḥ # 5. Eka-prastara-khande tu purāṇaḥ parvatôttamaḥ |
Khandagirîti nāmåsan pavitra côtkale bhuvi 6. Nivāsa-karaṇårthāya daiva-bāņi tu prāptavān
“Asmin nivasatu rājan yāvat tişthati medini #
tāvat kālasu paryyantaḥ tava kirittiḥ virājate " ! 7. Daiva-bāní srute Airaḥ harsa-nirbhara-mānasaḥ |
Kosalā-nagaram tyaktyā Khandaśailā-samipatu ! The substance of the verses, as made out by Mr. Jayaswal, is as follows:
"(a) That Kalinga had been conquered by the kings of Magadha,
and that it was liberated by one Aira (king) who defeated a
Nanda king of Magadha. (6) That the Nandas were Vedic, orthodox Hindus ; and the Aira
was heterodox (Jaina or Buddhist). (c) That the Aira was a great enemy of Asoka.

Page #212

184
OLD BRAHMI INSCRIPTIONS
(d) That the former capital of the Aira was Kosalā (South Kosala),
and that the Aira removed his capital to the Khandagiri at 'Eka
prastara' spot.” We have nothing to say against or to add to Mr. Jayaswal's fourth point. But his first three points are misleading, and need correction.
As to his first point, we find in the verses no indication whatsoever of Kalinga having been conquered by the kings of Magadha and liberated afterwards by King Aira. The implication of the verses obviously is that King Nanda of Magadba who tried either to conquer Utkala or appeared as a rival in his attempt to conquer some other country, was defeated by King Aira of Utkala in a battle.
As to his second point, what the verses state is that King Nanda of Magadha who was defeated by King Aira of Utkala in a battle, which ensued between them, was an adherent of the Vedic system. The verses are reticent about the religion of other Nanda kings of Magadha.
As to his third point, it does not appear from the verses that King Aira of Utkala was a great enemy of Asoka. Rather as a destroyer of Vedic religion and a promoter of his own faith, which was non-Vedic or anti-Vedic, he bore comparison with King Asoka, and the greater probability is that by mahāmittra, as pointed out by Dr. Sten Konow, the verses were intended to represent King Aira as “a great friend of Asoka."
In the inscriptions No. 1 and No. 3, Aira occurs as one of the distinctive royal titles prefixed, in the same manner, to the names of both Khåravela and Kadampa-Kudepa, the remaining titles being Mahārāja, Mahāmeghavāhana and Kalimgadhipati. Seeing that both Khāravela and Kadampa, two kings of one and the same dynasty, bore these titles in common, especially the epithet Aira, it may be surmised that Aira was a hereditary royal title or epithet of all the kings of this dynasty reigning in Kaliaga, in the same way that Brahmadatta was a hereditary royal title or epithet of all the kings of a particular dynasty reigning in Benares. Bat we are unable to see how such a surmise as made by Mr. Jayaswal could reasonably be made from the Sanskrit verses in the applauded Oriya MS. In support of this surmise, he argues, saying, “It is apparent that the Aira, who lived from the time of Nanda up to Asoka could not have been one and the same Aira. Aira tłerefore indicates a series of kings."
This ar zument of his, the force of which has freely been acknowledged by Dr. Sten Konow, is, in our opinion, unconvincing. For, in the first place, from the quoted verses, it is not at all clear that King Nanda of Magadha, referred to therein, was a pre-Mauryan Nanda king. And,

Page #213

NOTES
185
secondly, in describing King Aira of Utkala as a mahamittra (great-friend) of Asoka, the intention of the author of the verses does not appear to be to say that he was also a contemporary of the latter. The purport of the verses seems to be to the effect that in so far as King Aira of Utkala was a destroyer of Vedic religion and promoter of his own faith, which was nonVedic or anti-Vedic, he deserved, in the opinion of the author of the verses, to be described as “a great friend of Asoka."
As to the genuineness of two traditions embodied in the Sanskrit verses -(1) that King Aira of Utkala removed his capital from the Kosalacity to the Ekaprastara-tract around the Khanoagiri bill in Utkala, and (2) that the former capital of King Aira, who was a contemporary and victorious rival of King Nanda of Magadha, was in Kosalā or South KosalaMr. Jay aswal has urged two separate arguments, which are as follows :
(1)
(2)
That the capital of Kalinga before Asoka and after the Nandas is called Parthali (by Megasthenes) which corresponds with the Prastara of the Sanskrit verses, the Parthali which, by its location in the Khaņdagiri, seems to have been identical with Dhauli (Toşali).
That, according to the Purāṇas, amongst the local dynasties which arose during the Andhra-period, there was :(a) the dynasty of Kosalā (South Kosala, which, as described in
the Vişnu-Purāņa, adjoined Udra or Orissa, and, as described by Hwen Theang, adjoined Kalinga to the north-west and
above the Andhra country); () who were commonly known as the Megbas, Megha having
been a shortened form of Meghavāhana ; (c) who were very "powerful" and "wise”; and (d) whose kings were nine in number. 1
It is possible that Megha was a shortened form of Meghavāhana or Mahāmegharāhana. It is not impossible that nine kings of the Megha or Meghavāhana dynasty, including Khāravela and Kadam pa-Kudepa, reigned in Kalinga during the Andhra-period of Indian history. But there is
1. Kosalāyam tu rājāno bhavisyanti mah abalah "Meghā” iti samākhyāta buddhimanto navaiva tu
24

Page #214

186
OLD BRĀHMĪ INSCRIPTIONS
nothing in the inscription of Khāravela and Kadampa-Kudepa to corroborate the tradition that the Mahāmeghavāhana kings of Kalinga removed their capital from Kosala to Ekaprastara-tract.
It might be, if the information supplied in the Indika of Megasthenes be at all reliable, that Parthalis corresponding with Ekaprastara of the verses, was the capital of Kalinga before the reign of Asoka. But there is nothing in the Indika of the Greek ambassador to the court of Chandra. gupta to indicate that Parthali was the capital of Kalinga when King Aira of Utkala was its lord. The occurrence of the name of the hill as Khanc'agiri is enough to show that, whatever the source, the tradition is of a late origin. Mr. R. D. Banerji has conclusively proved by the evidence of the inscription of Udyotakeśarī that the ancient name of the Khandagiri hill, up to the 10th or 11th century A.D., was Kumāra-parvata, the expression embodying the name of the hill being Sri-Kumāraparvata-sthāne. It is for Mr. Jayaswal to say when the name of the sacred hill changed into Khandagiri or Khandaśaila.
This is not all. In the Hāthi-Gumphā inscription of Khāravela, Kalimga-nagara," the city of Kalinga," occurs as the name of the capital of Kalinga. This city, as the description goes, was not far removed from the Tanasuliya or Tanasuli road. That Tanasuliya or Tanasuli is the same geographical name as Tosala or Tosali is a conjecture, which awaits confirmation from independent evidence. Mr. Jayaswal has still to prove that Khandaśaila in the Ekaprastara-tract is identical with Kalimga. nagara, which is mentioned in Khāravela's inscription as the capital of Kalinga.
Lastly, we notice a wide discrepancy between what is stated in the Sanskrit verses and what in Khāra vela's inscription concerning King Nanda of Magadha. In accordance with the statement in the Sanskrit •verses, King Aira of Utkala defeated King Nanda of Magadba. And from the three statements in Khāravela's inscription, (l) that there was an aqueduct in Kalinga opened out by King Nanda 103 years back (Namdarājativasasata-oghātitam panādim), (2) that the Throne of Jina belonging to Kalinga was carried off by King Nanda (Namdarāja-nitam KālimgaJindsanam), and (3) that Pșthudaka, founded by the former kings of Kalinga, became a dark pool overgrown with a jungle of grass in 113 years (Kalimga-puvarāja-nivesitam Pithudaga-dabham anupa-dabhavanam ca terasa-vasa-sata-katam Ta(i)mira-daha-samghātam), it is clear that the then reigning king of Kalinga was defeated in a battle by the then reigning King Nanda of Magadha,

Page #215

NOTES
187
It seems that the story in the Sanskrit verses is a curious medley of a certain result of the misreading of Kharavela's inscription and a certain legend in the Puranas. We are, no doubt, at one with Mr. Jayaswal to think that these verses are important as confirming the correctness of the reading of the first word of Khāravela's inscription, after the invocation formula, as Airena instead of Verena or Kharena.

Page #216

188
OLD BRAHMI INSCRIPTIONS
11. THE GEOGRAPHICAL ALLUSIONS
The old Brahmi inscriptions in the Udayagiri and Khandagiri caves contain certain geographical allusions, the value of which has not, as yet, been systematically discussed and fully ascertained. What are these allusions?
Of the fourteen Brahmi inscriptions, the first ten are attached to the caves which were excavated by different donors on the hill now known by the name of Udayagiri, and the remaining four are attached, together with the table of Brahmt alphabet, to the caves which were excavated on the bill now known by the name of Khandagiri The two hills "form," says Major Kittoe, "part of a belt of sandstone rock, which, skirting the base of the granite hills of Orissa, extends from Autgar and Dekkunal (in a southerly direction) past Kurda and towards the Chilka lake, occasionally protruding through the beds of laterite." 1
"C
These belong, according to Mr. Stirling, to a group of four small hills, which are severally called the Udayagiri, Dewalgiri, Nilgiri and Khandgiri, are composed of a silicious sandstone of various colour and texture, and are all curiously perforated with small caves, disposed in two or three irregular stories," This group of four hills, from 150 to 200 feet in height, is situated "about five miles west of Bhobanesar, near the village of Jaymara, in the Charsudhi Khandaiti of Khurda, and still within the limits of the Khetr."
The Khandagiri is just four or five miles north-west of Bhuvanesvar and nineteen miles south-west of Cuttack, and is separated from the Udayagiri by a narrow glen about a hundred yards in width.3 Mr. R. D. Banerji has rightly suggested that the twin hills of Khandagiri and Udayagiri were known to the authors of the cave-inscriptions as Kumara-Kumāriparvata, the Kumara and Kumārī mountains. The ancient name of Khandagiri as Kumara-parvata is met with in Udyota-Kesari's inscription in Lalatendu-Kesari's cave on the Khandagiri, and that of Udayagiri as Kumari-parvata is met with in the Hathi-Gumpha inscription of Khāravela.*
1. JASB, Old Series, Vol. VI, p. 1079.
2. Extract from Stirling's Report on Orissa in JASB, Old Series, Vol. VI, p. 1076. 3. Kittoe's letter to Prinsep in JASB, Old Series, Vol. VI, p. 1079.
4. EI, Vol. XIII, p. 166.

Page #217

NOTES
189
Mr. Banerji has made a mere suggestion without bringing forward any arguments to prove his point or offering any explanations for its orientation. That the ancient name of Khaņdagiri was known to the author of Udyota-Kesari's inscription to be Kumāra-parvata is beyond dispute. The internal evidence of this inscription which must be assigned to the 10th or 11th century A.D. is enough to prove the identity between Kumāra-parvata and the modern Khandagiri. The inscription, dated in the 5th regnal year of Sri-Udyota-Kesari, records the re-excavation of an old tank, the repairing of an old cave-temple and the installation of the images of the twenty-four Tirthankaras on the sacred site of Kumāraparvata as meritorious works of the donor. The text of the inscription embodying this record reads as follows :
Sri-Udyota-Kesari-vijaya-rājya-saqavat 5 Sri-Kumāra-parvata-stbāne jirnna vāpi jirnna īsana udyotita Tasmina thâne caturvimšati Tirtham kara sthāpita.
Here the tank referred to is no other than the one which exists up till to-day on the eastern face of the Khaņdagiri, hewn out of the rock and sacred to both the Jains and Hindus, and the caves with images of the twenty-four Tirthankaras installed therein are no other than those which are known now-a-days as Navamuni, Durgā and Hanumān on the same hill.
In the Hathi-Gumpha inscription of Khāravela (1. 14), it is stated that His Majesty caused one hundred and seventeen caves to be made on Kumārī-pavata in the well-run realm of victory (supavata-vijayacake). Mr. Jayaswal and other scholars have sadly missed the real sense of the phrase supavata-vijaya-caka. As we have sought to show, this is just another and earlier form of the phrase sri-vijaya-rājya. We mean that supavata is not to be equated with suparvata, “the noble mountain," but with supravartta or supravrtta, “ well-run," " wellconducted," and that here the word caka is not to be taken in the sense of a belt, but in that of a realm of command (ānā.cakka). The manipulation of such an expression as supavata-vijaya-caka is perfectly in accord with the two of the royal epithets, pavata.caka and mahavijaya, attached to the name of siri-Khāravela in the Hāthi-Gumpbā inscription (I. 16). It cannot finally be decided whether the designation Kumāri-pavata was restricted to one hill, the Udayagiri, or was used to denote the twin hills of Udayagiri and Khandagiri. It is more probable that the author of the inscription intended to denote both the hills by a general name, just in the same way that the fashion

Page #218

www. kobatirth.org
190
· OLD BRĀHMI INSCRIPTIONS now-a-days is to denote both the hills, if brevity demands it, by one name, the name of Khandagiri. But this is not to deny that the Khantagiri was possibly known, as early as the Hathi-Gumphā inscription, by a separate name such as Kumāra-pavata. Here the case made out of the twin hills designated by a group name is on a par with that of Nārada and Parvata, the twin hills or mountain peaks designated by a common name as Närada. If it be true, as we believe it is, that the earlier group-name was Kumāri parvata, that is, Udayagiri, and the later group-name was Kumāraparvata, that is, Khandagiri, the historian has to find out a solution of the problem how was it that the name of Khaņdagiri came to receive a greater sanctity than, and gain precedence over, that of Udayagiri.
Thus it is clear that Khāravela's kingdom has been praised in his inscription as a supavata-vijaya-caka or “well-run realm of victory” (I. 14), an expression, corresponding to sri-vijaya-rājya in Udyota-Keśari's inscription, and a grandiloquent substitute for such a simple and earlier expression as vijital in Asoka's R. E. II, rāja-visaya in Asoka's R. E. XIII, or raja (rājya) in Dhanabhūti's inscription on the Barhut E. Gateway, in such a phrase as Suganam raje (Sungānām rājye). What was the kingdom for which the high-sounding phrase supavata-vijaya-caka was coined ?
One of the royal titles attached to the name of King Khāravela in the Hathi-Gumpbā inscription is Kalimoddhipati, “the Sovran Lord of Kalinga" (I. 1). The same royal title adorns the name of King KadampaKudepa in his inscription (No. III). Kbāravela figures in the inscription if bis chief queen as Kalimga-cąkarati, “the (King) Overlord of Kalinga” (No. II). These go at once to show that the intended kingdom was no other than Kalinga. .
The Hāthi-Gumphā inscription strikes throughout a patriotic note about Kalinga. In it, Khāravela is styled Kalimgadhipati, “the Sovran Lord of Kalinga” (I. 1). In it, he is said to have been consecrated as a Mahārājā or “Great Majesty" in the third generation of two kings in Kalimga-rājavamsa, “the royal dynasty of Kalinga” (I. 1). In it, he is said to have caused terror to the Assaka or Asika city with the aid of the army from Kalinga (Kalimgá gatāya senāya, I. 3). In it, he is said to
1. Cf. Nandarāja-vijitasmim in the Petavatthu, II. I, and III, 2. 2. Note that Jayaswal reads Musika. 3. Note that Jayaswal reads Kanihabenågatāya.

Page #219

NOTES
191
have governed Arakatapura,' inhabited by the Vidyād baras, in the manner of the former kings of Kalinga (Kalimga-puvarājānam dhamena va nitinā va, I. 4). In it, he is said to have built the Great-victory-Palace which was the Kalinga-royal-residence (Kalimga-rājanivāsa, 1. 10). In it, he is said to have done due honour to the memory of the former kings of Kaliaga (Kalimga-Puvarājānam-yasasakāram, I. 11). In it, he is said to have rehabilitated Pithudaga or Pithuda, founded by the former kings of Kalirga (Kalimga-puvarāja-nivesita, I. 12). Lastly, in the same record, he is said to have triamphantly brought back to Kalinga (Kalimgam āneti) the Throne of Jina belonging to Kalinga which was carried away by King Nanda (Namdarāja-nītam Kalimga-Jinásanam, I. 13). And in the inscription of Khāravela's chief queen, a cave is said to have been made by Her Majesty for the residence of the honoured recluses of the Kalingas (Kalimgānam samanānam, II)
The Hathi-Gumphā inscription clearly shows that the capital of Kalinga during the reign of Khāravela was Kalimga-nagara, " the city of Kalinga," which has been satisfactorily identified with Mukbalingam on the Vamsadbārā and the adjacent ruins in Ganjam District, Madras Presidency." The Purle Plates of Indravarman, dated in the Ganga year 149, go to show that the kings of the Ganga dynasty had generally granted their donations from Kalinga-nagara, 8 the self-same city of Kalinga which Prof. Sylvain Lévi seems inclined to identify with Kalinga patam,* 20 miles north-east of Srikakola or Chikakol.
We read in the Hathi-Gumphā inscription that as soon as he was anointed, in the very first year of his reign, King Khāravela repaired the gates, walls and buildings badly damaged by stormy wind in the city of Kaliiga (Kalimga-nagari), raised the embankments of the deep and cool tanks, and restored all the gardens at the cost of 35,00,000 (coins), thereby enabling his subjects to be pleased (1.2). We also read in the same inscription that His Majesty spent the first fifteen years of his life playing the princely games, and nine as a crown-prince, well-versed in the art of writing, coinage, and the rest ; and also that he belonged to the third generation of two kings belonging to the royal dynasty of Kalinga (tatiye Kalimga-rājavamse purisa-yuge), the kings of which may be
1. Note that another possible reading is ahatapuva or āhatapuva. 2. EI, Vol. IV, p. 187. 3. EI, Vol. XIV, p. 36. 4. JA, 1925, T. CCVI, pp. 50, 53, 57. 5. Cunningham's Ancient Geography of India, S. N. Majumdar's edition, Notes, p. 735

Page #220

192
OLD BRAHMI INSCRIPTIONS
supposed to have been all distinguished by such royal titles as Aira-Siri, Mahā meghavāhana and Kalimgâdhipati (I, II, III). It may be safely inferred from these that the city of Kalinga was the capital of Kalinga also during the reign of the preceding two generations of two AiraMeghavāhana kings each, at least, during the reign of the second generation of two kings. The city of Kalinga has been referred to in the records of the third, fifth and eighth years of Khāravela's reign, and there is nothing in any of the fourteen old Brāhmi inscriptions to suggest that this city ceased to be the capital of Kalinga during the reign of the third generation of two kings, of whom Khāra vela was one.
What was the capital of Kalioga, the land of the Kaliogas, when King Asoka of Magadha conquered it and permanently annexed it to his empire, no one knows. What is manifest from his edicts, especially the two copies of his two Separate Rock Edicts, is that Tosali and Samāpā were two principal towns in the Kalinga Province of his empire, of which the former was a seat of Maurya viceroyalty. One set of his Rock Edicts and one copy of his Separate Rock Edicts have been found “inscribed on a rock called Aswastama, situated close to the village of Dhauli," and the modern village 07 Dhauli which is no other than the town of Tosali is "about seven miles to the south of Bhuvanešvar,"1 though one must endorse the opinion of Prof. Vincent A. Smith that the exact position of Tosali, as known then," has not been ascertained."
A second set of Asoka's Rock Edicts and a second copy of his Separate Rock Edicts have been found “engraved on the face of a picturesque rock
in a large old fort called Jaugala (Lac-fort), near the bank of Rșikulyā | river, about eighteen miles to the west-north-west of the town of Ganjam."?
Though we have no means of ascertaining the exact position of Samāpā, this much is certain that it was a District town of the second division of Asoka's Kalinga Province, and that it was situated somewhere in the vicinity of the Jaugała Fort in the Ganjam District.
The ancient name of the rock on the face of which the Dhanli version of Asoka's Rock Edicts and Separate Rock Edicts was engraved was embodied in the Dhauli copy of his Rock Edict I. But, as bad luck would have it, exactly that portion of the inscribed surface is broken off which bore the four letters inscribing the name of the rock, and there is no means of restoring the lost name. In the present state of our knowledge, we have to be
1. D. R. Bhandarkar's Asoka, p. 255. 2. Ibid, p. 256,

Page #221

NOTES
193
satisfied with the fact that, whatever was the actual name of this rock, it had consisted of four syllables. We are very fortunate to have intact the Asokan name of the Jaugada rock in the Jaugada copy of Asoka's Rock Edict I, in the statement: Iyam dhammalipi Khapimgalasi pavatasi Devanam piyena Piyadasinā lājinā likhāpita.
"This Edict of the Law was caused to be inscribed by His Gifted Majesty and Grace the King on Mt. Khapimgala." Thus the recorded name is found to be Khapimgalapavata, "the Khapingala mountain."
Now, comparing the two statements, one in the Dhauli copy containing the lost Asokan name of the Dhauli rock and the other in the Jaugada copy containing the Asokan name of the Jaugada rock, and noticing how closely these agree with each other, and bearing in mind that the missing name in the Dhauli copy, precisely like Khapimgala in the Jaugaḍa version, consisted of four syllables, one cannot help feeling inclined to think that the same also was the name in the Dhauli copy. If these were correct, one might have reasonably taken Khapimgala to be the name, not of an isolated rock, but that of a range or group of hills representing the northern extension of the Eastern Ghats and extending along the sea-coast of Kalinga in its eastern extremity. We are not pressing this as an established fact, but just what is possible, if Khapimgala occurred as a common name in both the copies. Even if this were an established fact, the problem would have remained, whether the Kumaraparvata of UdyotaKesari's inscription and the Kumaripavata of Kharavela's inscription would have been included in the Khapingala range known in Asoka's time. The Fragment LVI of the Indika of Megasthenes mentions Parthalis as the royal city of the Calinga representing the tribes that dwelt by the Ganges, nearest the sea. Partualis is the spelling of the name which appears in one of the foot-notes of the Fragment XX. B in Prof. McCrindle's translation. Portalis is evidently a simpler form of the spelling Partualis, which has been suggested in the second foot-note of the Fragment LVI. M. de St. Martin would identify the royal city of the Calinga with Vardhana (contraction of Varddhamana), now Burdwan. Prof. McCrindle thinks that the Calinga were a great and widely diffused tribe that settled mainly between the Mahanadi and the Godavari, and that their capital was situated "on the Mahanadi, higher up than the site of Katak." And Mr. Jayaswal, as we saw, takes Parthalis in good
1. Dhauli Iyamsi pavatasi.
Jaugada Iyam dhammalipi Khapimgalasi pavatasi.
25

Page #222

194
OLD BRĀHMÍ INSCRIPTIONS
faith to be the correct and only spelling, and identifies Parthalis with the Ekaprastara tract which, according to the story of the Sanskrit verses quoted from an old Oriyā MS., was the site of the new capital of King Aira of Utkala, around Khaqdagiri. Because Megasthenes, the Greek ambassador at the court of Chandragupta Maurya, happened to mention Parthalis as the royal city of the Calinge, he safely assumes that Parthalis was the capital of Kalinga in the time of Chandragupta Maurya, nay, also during the reign of King Nanda who is mentioned in the Hathi-Gumphã inscription, and who, according to him, was no other than Nanda referred to in the Sanskrit verses as the king of Magadha defeated by King Aira of Utkala in a battle fought between them. The implication of this is that Pithudaga or Pithuda became abandoned to its fate 102 years (113-11) previous to the consecration of Khāravela.
In the same inscription (I. 6), we read that His Majesty brought into his capital, from the Tanasuliya or Tanasuli road, the canal which was opened out by King Nanda 103 years back (Nandarāja-tivasasata-oghātitam panādim). This canal must have been opened out 98 years (103.5) previous to the consecration of Khāravela.
In the same inscription (I. 18), we also read that His Majesty brought back to Kalinga, from Anga-Magadha, the throne of Jina which had been carried off from Kalinga by King Nanda (Nandarāja-nītam KālimgaJināsanam).
Now squaring up these three statements, it becomes easy to understand (1) that Kalinga was under the sway of King Nanda of Arga-Magadha, at least, from the 102nd to the 98th year previous to the reign of Khāravela ; (2) that Pithudaga or Pithuda, founded by the former kings of Kalinga, became abandoned to its fate with the advent of King Nanda in Kalinga ; (3) that here by the former kings of Kalinga Khāravela wanted to mean those kings of Kalinga who had reigned before Kalinga was conquered by King Nanda ; and (4) that the rule of King Nanda in Kalinga ended when the dynasty of Kalinga kings to which Khāravela himself belonged came into power.
It cannot be confidently maintained that Parthalis or Partualis, mentioned in the Indika of Megasthenes as the royal city of the Calinga, was a Greek pronunciation of the name of the tract called Eka-prastara or Prastaru which, according to the Sanskrit verses in Mr. Jayaswal's Old Oriyā MS., became the site of the new capital of King Aira of Utkala, whose former capital was the Kosalā-city, and that in other words, Parthalis or Partualis was the capital of Kalinga in the days of King Nanda of

Page #223

NOTES
195 Magadha who is supposed to have been a pre-Mauryan Indian monarch and a contemporary of King Aira of Utkala.
A presamption without convincing proofs may be utterly devoid of truth in sober history. In order to maintain (1) that Parthalis or Parlualis in the Indika was a Greek spelling of Elea-prustara or Prastara in the Sanskrit verses, (2) that Kiny Nanda of Magadha mentioned as a temporary of King Aira of Utkala was a pre-Mauryan Nanda king, and (3) that Parthalis or Prastara was the capital of Kalinga when King Aira of Utkala reigned there in the days of Chandragupta Maurya when Megasthenes was in India, one must be sure (1) that Parthalis or Partualis is mentioned in the Indika as a tract, like Eka-prastara, around the Khandagiri, and (2) that there is mention of any Aira King of Utkala or Kalinga as a contemporary of a Nanda king of Magadha who was a precursor, á posteriori, of Chandragupta Maurya. But nothing is surer than that one cannot be sure about these two points. We are entirely in the dark as to who, among the kings of Kalinga, were contemporaries of the pre-Asokan Maurya kings and pre-Mauryan Nanda kings of Magadha.
In Khāravela's inscription (I. 12), we have mention of a place founded by the former kings of Kalinga and known by the name of Pithudaga or Pithudu, which had become, in 113 years, a watery jungle of grass.
The city of Kalinga could not have been very far from the Panasuliya or Tanasulī road wherefrom the canal opened out by King Nanda 103 years back was brought into it by King Khāravela in the fifth year of his reign. The reading Tanusuliya is certain. The plaster casts and estampages of the Hāthi-Gumphā inscription leaves no chance for the reading of the second letter as u. If the inscribed name might be read as Tausuliya or Tausalī, it could have been easily equated with Tosali. But read as Tanasuliya or Tanusuli, it remains to be seen how the name could be equated with Tosali (passim). We have noted that tana, the first member of the compound, occurs in one of the verses of the Mahāvamsa as the opposite of mahā: Muhāsivu, Sivu, Tanusivu. And suliya, the second member of the compound, must be treated either as an equivalent in an eastern dialect of the Pali suriya, or of the Sk. sūryya, or as a form of suli conjoined with the suffix yu. The first alternative is less likely for the reason that the general tendency of the dialect of the Hathi-Gumphā inscription is to replace l-sound by r-sound, unless it be supposed that the name has been retained as it was locally pronounced. If Tanasuliya be regarded as an equivalent of Tanasuriya, it must be rendered in English : "the Little Sun-temple (road).” If, on the other hand, it be regarded as a form of/

Page #224

196
OLD BRAHMI INSCRIPTIONS
Tanasuli conjoined with the suffix ya, it must be rendered: "the Little Siva temple (road)," sul being one of the epithets of Siva. Anyhow, it is certain that the opposite of Tanasuliya or Tanasule is Mahasuliya or Mahasuli. If so, it may be shown that Pithudaga or Pithuda, the site of the former capital of Kalinga, was just in the neighbourhood of the city of Kalinga.
Pithudaga is the same name as Prṛthudaka in Sanskrit, and Pithuda is but a shortening from Pithudaga, a word which is the same in meaning as bahudaka, "abounding in many waters," "the watery." The same significance of the name may be gathered, I think, from a legend in the Visņu-Purana regarding the origin of the name of Pṛthudaka or Peboa, "Prthu's pool," an old town near Thaneswar. 1 The Gandavyuha which is a Buddhist work in Sanskrit and counted among the maha-vaipulyasutras by the Buddhists of Nepal, contains an interesting account of the wanderings of a Buddhist seeker of truth in the Deccan (Dakṣinapathe). In it, the Buddhist learner concerned has been represented as travelling from Mulaka (on the bank of the Godavari, near Patiṭṭhāna or Paithan2) to Naladhvaja, from Naladhvaja to Suprabha, from Suprabha to Sarvagrama of Tosala in Amitatosala, and from thence to Prthurastra. This Prthurastra is apparently not different from what Ptolemy in his work on geography (VII. 1. 93), calls Pitundra-metropolis, Pitundra, the capital, Pitundra which was a Greek spelling, as shown by Prof. Sylvain Lévi, of Pihumda. The following note on Pitundra from the pen of Prof. Sylvain Lévi is worth quoting in this connection :
3
"Ptolemy," says Prof. Sylvain Lévi, "describing the towns situated in the interior of the country of Maisoloi (VII. 1. 93), designates its capital Pitundra-metropolis. The country of Maisoloi or Maisolia (VII. 1. 15) lent its name to the river Maisolos which represents the group of the mouths of the Godavari and the Krsna. The Periplus writes Masalia instead of Maisolia. Since a long time the scholars have equated Maisoloi or Maisolia with Masuli, the denomination in the first word contained in the well known name of Masulipatam (patam=pattana, the town)
1. Cunningham's Ancient Geography of India, edited by S. N. Majumdar, p. 335. 2. Sutta-Nipata, Parayana-Vagga, Vatthugatha, verse 977. Assakassa visaye Alakassa (Mulakassa) samäsane, Godhävari-kule. For the discussion of the information contained in the Sutta-Nipäta-Commentary, see passim.
3. JA, 1925, T. CCVI, p. 61.

Page #225

NOTES
197
near the mouths of the Krsnā).1 Maisolia extended in the north up to Paloura, or more accurately, up to the equator in the neigbbourhood of Paloura....... Ptolemy locates Pitundra in the interior of Maisolia, between the mouths of the two rivers) Maisolos and Manadas, to put it otherwise, between the delta of the Godavari and the Mahānadı, nearly at an equal distance from both. It would, therefore, be convenient to search for its location in the interior of Chikakol and Kalinga patam,...... towards the course of the river Nāgāvali which bears also the name of Lārguliya, the
River of the Plough.' The Imperial Gazetteer of India itself indicates this etymology: lāngala, Sanskrit ; nagula, Telugu. This denomination evokes, bringing nearer the souvenir of Pitundra, the text of the Hātbi-Gumphā inscription in which Khāravela flatters himself having ploughed with the plough the soil of Pithuda,"? or as we prefer to read and interpret it, “ having let out the grassy jungle of Pithudaga into Namgala, the river Lāngala (Länguliya).”3
Prof. Sylvain Lévi draws attention to the story of Samudrapāla in Lec. XXI of the Jaina Uttaradhyayana-Sūtra, in which there is mention of Pihumda as a sea-coast town reminding us at once of Khāravela's Pithuda. Pithudaga and Ptolemy's Pitundra.4 This story clearly shows that Pithuda was an emporium of trade which could be reached from Campā by the merchant vessels that had to follow a sea-route to complete their voyage. Campā, as we all know, was the capital of Arga, situated on the lower course of the Ganges, and the story in question relates that a Jain merchant named Pālita, who was a native of Campā, had a son born to him at sea (samudra) wben he was returning home with his wife from Pihumda, where he went for the purpose of trade and happened to win the hand of the daughter of a merchant of the place. The father chose Samudrapāla, " the Protége of the Sea," as the name of the boy then born at sea. There can be little doubt that Pihumda in Ardha-Māgadhi is the same geographical name as Pithuda-Pithudaga in Khāravela's inscription, Ptolemy's Pitundra, the capital of Maisoloi-Maisolia, and Pythurāştra in the Gandavyūha.
1. For the historic name and etymology of Masulipatam, see Yule-Burnell, sub voce. 2. Translated from the French in JA, 1925, T. COVI, pp. 60-61. 3. Named Längalini in the Märkandeya-Parāņa. 4. JA, 1925, T. CCVI, pp. 57-58. 6. Jacobi's Jaina Sūtras (B, E.), Part II, p. 108.

Page #226

198
OLD BRÄHMI INSCRIPTIONS
According to the unanimous testimony of the Jātakas, particularly of two versions of the Mabāgovinda-Suttanta which represents one of the earliest forms in which one finds the Jātakas in Buddhist litera Dantapura was the earlier capital of Kalióga, as early as when Kasi an empire with Kalinga as one of its provinces. The couplet in the Suttanta-Jātaka” mentioning Dantapura as the capital of Kalinga, 9 the land of the Kalingas, is found to be a quotation from an earlier chronicle
en Purohitas in the Anguttara-Nikāya,+ embodied in a prose-story, which has not, as yet, assumed the character of a Jätaka. Dantapura, which is no other than Dantakura in the Mahābhārata and Dantagula in Pliny's Natural History, has been definitely identified by Prof. Sylvain Lévi with Ptolemy's Paloura and modern Paloura near Chicacole in the Ganjam District. When exactly Dantapura-Palonra ceased to be the capital of Kalinga we cannot say. But it is certain that it had been the capital of Kalinga before Pilhuda-Pitundra became the royal city. It may be safely concluded from the foregoing discussion that the transfer of the capital from Dantapura to Prthudaka must have taken place before the advent of King Nanda in Kalinga and before the establishment of the rule of the royal dynasty of Kalinga to which Khāravela himself belonged.
The Sarabhanga-Jataka (Faus böll's No. 522) refers to a time when Kāší was just an independent kingdom, which existed side by side with the kingdom of King Dandaki. The city of Kumbhavati was the capital of Dandaki's kingdom, of Dandaka, measuring 60 yojanas long. Dandaki was a powerful emperor, whose supremacy was freely acknowledged by Kālinga, the king of the land of the Kalingas (Kalinga-rājā). King Kāliiga is described as one of the lords of the subordinate kingdoms (antararatthādhipatino). The prosperity of the Danılaka empire and the
1. One in the Digha-Nikāya, Vol. II, and the other in the Mahảvastu. 2. Mahāgovinda-Suttanta enlisted as a Suttanta-Jātaka in the Calla-Niddesa, p. 80.
Dantapura Kalinganam Assakānam ca Potanum | Māhissati Avantinam Sovīrūnarn ca Rorukan! Mithilā ca Videhānañ, Campå Angesu māpitä |
Bārānasi ca Kūsinam ete Govinda-mapitā # 4. Auguttara Nikaya, Part III, pp. 371-373. 5. That is, it has not the concluding identification,
6. See for references and quotations, Sylvain Lévi's Notes Sur la Geographie Ancienee de L'Inde, I. Paloura-Dantapura in JA, 1925, T. CCVI, pp. 46-57.

Page #227

NOTES
199
subordination thereto of Kalinga are the annals, if we are to believe the Jātaka, of the political history of ancient India before the rise of Kāsi to the status of an empire. The Sarabharga-Jätaka which contains a pathetic story of the dire calamity that befell the kingdom of Dandaki and brought ut ter destruction upon it, indicates a turning point in the political history of ancient India, in the subsequent chapter whereof the historian is to look for the annals of the rise and influence of the Kāśi empire. If it be true that the Buddhist Birth-story in the Mahāgovinda-Suttanta is an annal of the full flowering of the Kāśi empire with Kalinga, Asvaka, Avanti, Sauvira, Videha and Anga as the six subordinate kingdoms under it, we have to understand that Dantapura became the royal city of Kalinga during the supremacy of Kāśi, and not before.
The Sarabhanga-Jätaka does not mention the name of the capital of Kalinga when it was a subordinate kingdom under Dandaka. The Mahābhārata speaks of a time when a matrimonial alliance came to be established between the Kuru kingdom and Kalinga by the marriage of the Kuru prince Duryyodhana with the Kalinga princess Citrāngadā, - when Hastināpura was the capital of the Kuru kingdom and Rājapura that of Kalinga.
Some of the Indologists are inclined to identify Rājapura with Rajmahendri on the Godāvarī river, "251 miles to the south-svest of Ganjam,' which became the capital of the junior or eastern branch of the Chalukya princes of Vengi from the time of the conquest of Kalinga by the Chalukya king of Vengi in circa 750 A.D. and the removal of the Chalukya capital from Vengipura to Rajamahendri. The remains of the former still exist at Vegi, 5 miles (more accurately, 7 miles) to the north of Ellur, and 50 miles to the west-south-west of Rūjamahendri." ?
Mr. Manomohan Ganguli doubts the soundness of the identification proposed evidently on the basis of an accidental coincidence of prefixes.' 3 Rājapura is phonetically the same geographical name as the modern Raipur. As regards the connection of Rājapura in the Mahābhārata with Rāja mahendri, some light might perhaps be obtained from the Mahāvastu which professes to be the first work of the Vinaya-Pitaka of the Lokottaravādi section of the Mabāsarghikas. This great Buddhist work in Sanskrit, dealing with a previous birth-story of three Kāsyapa brothers who are to be counted among the first Buddbist converts and
1. Mahābhārata XII, 4,3. 2. Cunningham's Ancient Geography of India, edited by S. N. Majumdar, pp. 590.592. 3. Manomohan Gangali's Orissa and Her Remains, p. 8,

Page #228

200
OLD BRĀHMI INSCRIPTIONS
immediate disciples of the Buddha Sākyamuni, relates that they were born, in a former life, as three half-brothers of the previous Buddha Puspa or Puşya and sons of King Mahendra of Hastināpura, by one mother (ekamātykā trayo bhrātaro), and reigned together amicably in the city of Simhapura, in the land of the Kalingas,' Simbapura which is identified by Prof. Dubreuil with modern Singupuram near Chicacole.2
The source from which the tradition in the Mabāvastu was derived is unknown. The Pāli version of the previous birth-story of the thr Kāśyapa brothers, as found in Buddhaghoşa's Manoratba-Pūrani, Part I, speaks of King Mahendra without any reference to llastināpuru, and speaks of the three half-brothers of the previous Buddha Pusya without any reference to Kalinga and Simhapura, and, curiously enough, in this respect, the Pāli version of the story is completely in accor with the narration in the verses quoted in the Mahāvastu.3 Moreover, the Buddhavamsa which is a Pāli canonical work belonging to the Sutta-Pitaka and the Nidāna-Kathā of the Pāli Jataka-commentary present a life of the previous Buddha Pusya, which differs entirely in its details from that in the Mahāvastu or in Buddhaghoşa's commentary on the Anguttara-Nikāya. In these circumstances, the Mabāvastu tradition of the reign of Rājā Mahendra in Hastināpura and of the reign of his three sons in the Simhapura city of Kalinga must be regarded as the growth of an age later than the date of composition of the Buddhavamsa and earlier than the date of composition of Buddhaghoşa's Manoratba. Pūrani and of the Mahāvastu in its extant form. Be that as it may, the importance of the prose version of the story in the Mabāvastu lies in the fact that it enables us to account for the foundation of royal city in Kalinga commemorating the name of Rājā Mahendra.
The historical fact behind the Buddhist story in the Mahāvastu is not probably far to seek. We learn from the Allahabad pillar inscription of Samudragupta that Kosala or South Kosala was the first kingdom in South India (Dakşiņāpatha) against which the great invader from the north directed his first attack, and that the first king in South India
endra.
1. Mahāvastu, edited by Senart, T. III, pp. 432-433. 2. Ancient History of the Deccan, p. 94.
Mahāvastu, T. III, p. 433 : Rajño Mahendrasya mahim praśāsato Dharmena jñänam ca samādāya-vartino || Trayo bhrātaro asya samānacārino 11

Page #229

NOTES
201
who was defeated by him was King Mahendra of Kosala. This Kosala or South Kosala as may be now ascertained, “comprised the modern Bilaspur, Raipur and Sambalpur districts, and occasionally even a part of Ganjam."! " Its capital was Srīpura, the modern Sirpur about 40 miles east by north from Raipur."'?
None need be surprised that the Buddhist story under notice3 grew up round the tradition of King Mahendra of Kosala or South Kosala who was a contemporary of Samudragupta, and that Hastināpura, which is said to have been the capital of King Mahendra, was just another name for Śrīpura. The story distinctly says that Kalinga was conquered and governed by the three sons of King Mahendra for their father. We may understand from this that Kalinga proper was treated as a seat of viceroyalty with its official headquarters at Simhapura or Singupuram near Chicacole.
Thus our enquiry concerning the successive capitals of Kalinga leads us to a point where we have to recognise (1) that the capital of the kingdom during the reign of Khāravela and other kings of the Aira. Meghavāhana dynasty was Kalinga-nagara, the modern Mukhalingam on the Vamśadhārā in the Ganjam district ; (2) that Pithudaga-Pithuda was the capital of the former kings of Kalinga before the advent of King Nanda of Arga- Magadha in Kalinga and the reign of the kings of the A ira-Meghavāhana dynasty ; (3) that Tosali-Dhauli and Samānā were respectively the official headquarters of two divisions of Kalinga during the reign of King Devānampriya Asoka of Magadha; (4) that the still earlier capital of the kingdom was Dantapura-Paloura near Chicacole; (5) that the capital of the kingdom under Dancaka is unknown; (6) that Śrīpura-Hastināpura and Simhapura-Singupuram were respectively the seats of government in South Kosala and Kalinga proper during the reign of Rājā Mahendra in the middle of the 4th century A.D.; (9) that Rajamahendri became the Chalukya capital in Kalinga in the 8th century A.D.; and (8) that the city of Kalinga again became the capital of the kingdom in the 11th century A.D. when the kings of the Ganga dynasty made themselves masters thereof.
Now we shall pursue an enquiry concerning the territorial extension and political divisions of Kalinga during the reign of Khāra vela, as well
1. & 2. 4. C. Raychaudhuri's Political History of Ancient India, 2nd edition, pp. 337-338.
3. According to the Buddhavamsa and the Jataka Nidana-Katha, the previons Buddha Pusya, was the son of King Jayasena of Kāśi.
26

Page #230

202
OLD BRAHMI INSCRIPTIONS
as concerning the ranges of Khāravela's military expeditions and conquests. It will be our interest also to indicate the regions where the influence of his military powers was felt and openly acknowledged. As regards all these points, we are to attempt to draw our corclusions from the following data that may be gathered from the Hāthi-Gumphā inscription and other sources, epigraphic and literary, which are now within our reach.
The old Brāhmi inscriptions are all found attached as labels to the caves or the Udayagiri and Khandagiri bills which are situated about five or six miles north-west of Bhuvaneswar and at a distance of a few miles from the village of Dhauli-Tosali that lies about seven miles south of Bhuvaneswar. That is to say, our old Brahmi inscriptions and the Dhauli copy of Asoka's Rock Edicts and Separate Rock Edicts are found engraved at a distance of a few miles on the rocks or hills that are included in the modern Puri District of Orissa.
It is clearly implied in the edicts of Asoka, especially in the two copies of his two Separate Rock Edicts, that his Kalinga province comprised two political divisions, the first or presidency division of which the official headquarters was Tosali-Dhauli, and the second division of which the official headquarters was Samāpā, and that adjoining thereto were the outlying unconquered tracts ("mtā avijitā). The names of the two divisions of Asoka's Kalinga province are not mentioned in his edicts. But we learn from the Gandavyūha account of the wanderings of a Buddhist seeker of truth in South India that Sarvagrāma was a locality in Toşala, while Toşala, undoubtedly a city corresponding to Asoka's Tosali, was located in Amitatoşala, and that beyond Amitatoşala was Pșthurāştra, which has been identified with Pithudaga-Pithuda in Khāravela's inscriptions, Pihumda in the Jaina Uttaradhyayana-Sūtra and pitundra in Ptolemy's work on Geography. It may be easily infer this that Asoka's Tosali-Toşala was the chief town of a division of the Ssame name, Amitatoşala, Tosali or Toşala. Here the crux is--does the Hāthi-Gumphā inscription offer us any information about the existence of a political division of Kalioga by the name of Amitatogala, Tosali or Toşala? Whether it does or not depends solely on the accuracy or inaccuracy of the identification of Tanasuliya or Tanasuli with Tosali.
It is evident from the record of Khāravela's fifth regnal year that Kalimga-nagara, the capital of Kbāravela's kingdom of Kalinga, was not far from the Tanasuliya or Țanasuli road wherefrom a canal opened out by King

Page #231

NOTES
203
Nanda was led by extension into the city of Kalinga. We have seen that Tanasuliya or Tanasuli is a name which stands in contrast to Mahāsuliya or Mahāsuli, tana or tanu being the opposite of mahā or maha. We are to bear in mind that Kalimga-nagara is the modern Mukhalingam
the modern Mukhalingam on the vaņsadhārå river and the adjacent ruins in the Ganjam district of Madras Presidenzy. We are to remember that Pithuờaga-Pitbula referred to ing the Hathi-Gumphà inscription as the earlier capital of Kalinga is not different from Pșthurāṣtra in the Ganda vyūha which lay just beyond Amitatoşala- Toşala, from Pihumda in the Jaina Uttaradhyayana-Sūtra which was a sea-coast town reachable by the Indian merchant ships sailing from Campā, the capital of Anga, and from Ptolemy's Pitundra which was the metropolis of Maisolia-Maisoli and is located by Prof. Sylvain Lévi somewhere near Chicacole in the Ganjam district. We are also to keep in mind that the Periplus writes Masalia for Maisolia-Maisololi. Thus in Ptolemy's Maisoilia-Maisoloi, written as Masalia in the Periplus, we trace the nearest Greek spelling of Mahāsuliya- Mahāsuli, which is the suggested antithesis of Tanasuliya-Tanasuli.
Apparently Tanasıliya-Tanasuli and l'ahāsuliya- Mahāsuli stand, in their contrast, on a par with Tosala and Mosila, which, as noted by Mr. Jayaswal, are mentioned as two among the three divisions of Kalinga in the Kāvyamāla edition of the Bhāratiya-Nāțya-Sāstra, Ch. XIV, Verse 40, where one reads :
Kosalās-Tosalāscaiva Kalinga yā ca Mosalāḥ.
The text of the quoted line may be so interpreted as to mean that the 1 kingdom of Kalinga consisted of these three divisions : Tosala, Mosala,
and_Kesala. We have in M. Joanny Grosset's edition Yavanosala instead of Mosala, which may not be copyist's mistake but ather an interesting variant indicating that Mosala was known to some of the copyists as a locality associated with the Yavanas. The importance of the text in the Kāvyamālá edition lies in the fact that here Mosala occurs as the final | form of phonetic decay: Mahāsuliya, Maisolia, Maisoloi, Masalia, Masuli,
Mosala, three 'missing links being Mahāsuli, Masali and Mosali. We maintain that in a similar manner Tosala may be treated as the final form of phonetic decay from Tanasuliya : Tanasuliya, Tosali, Tosala, the missing
links, among others, beiny Tansalia and Tansali. If so, we may draw this · conclusion that Tunasuli, Tosali or Tosala was that division of Kalinga of
which the official headquarters was somewhere near the Kumārl-parvata or Udayagiri, and that it extended along the sea-coast, at least, from the

Page #232

204
OLD BRĀHMĪ INSCRIPTIONS
Nilgiri to the Chilka lake during tho reign of Khāravela, and may be, also during the reign of Asoka.
What was precisely the southern limit of the second division, called Mosala or Mosali in the Bbāratiya Nātya-Sastra, whether during the reign of King Aśoka or during the reign of King Khāravela we cannot say. From the texts and location of the Rock and Separate Rock Edicts of Asoka, it may be clearly ascertained (1) that Samāpā was the official headquarters of the second division, (2) that the second division occupied, if it was not co-extensive with, the modern Ganjam district, and (3) that his Kalinga Province was guarded along its eastern sea-coast, from the Jaugada hill-fort to the rock at Dhauli, by a range of hills, which was perhaps known by the name of Khapimgala-pavata. This range of hills was nothing but a northern or upward extension of the Eastern ghats, and, probably no other than what is called Palapiñjara in the Bhārafya Natya-Šāstra.? If it be correct, as Prof. D, R. Bhandarkar suggests, that the Kalinga Province of King Asoka abutted in the south on the territory of the Andhras, counted among the vassal tribes within the dominions of the great Maurya emeror, and that the strips of territory occupied then by them included the Godavari and Kistna Districts, we may safely conclude that the second division of Asoka's Kalinga Province did not extend in the south beyond the Godavari river. And if King Asoka had anything to do with the canal referred to in the Hāthi-Gumpbā inscription as an excavation of King Nanda near the Tanasuliya or Tosali road, we may understand that the second division of Kalinga extended during the reign of the Maurya emperor from the mouth of the Vamsadbārā to that of the Godāvars. It is clear from the inscription of King Khāravela that during his reign the Tanasuliya or Tosali road was not far from the city of Kalinga which is identified with Mukhalingam on the Vamśadhārā. If it was possible for King Khāravela to cause the grassy jungle of a swamp in Pithuca-Pitundra to be let out into the Namgala or Lāngala river, we must understand that Pithuda-Pitundra was situated somewhere near the course of the river Nāgāvali which bears also the name of Länguliya, "the river of the plough," -- the Lāngālint river which finds mention in the Mārkandleya-Purāna along with the Vaņģadhārā. We have seen that
WM
1. Bharatiya Națya-Šāstra, XIV:
Mahendrah Malayo Sahyah Mekulah Palapinjarah.

Page #233

NOTES
205
Ptolemy locates Pitundra-Pithuda between the mouths of the rivers Maisolos and Manadas, which is to say, between the deltas of the Godāvarī and the Mahanadi, nearly at an equal distance from both. Seeing that both Kalimga-nagara and Pithu aga mentioned in the Hathi-Gumpba inscription lay outside of the Tanasuliya or Tosali division and fell within the second division, we may be pretty certain about the extension of the second division of Kalinga during the reign of King Kharavela from the mouth of the Vamsadhara to as far south as the mouth of the Godāvarī, if not further south.
The nanner in which King Asoka mentions in his edicts the kingdoms that lay outside of, and the territories that lay within, his empire enables us to conceive that, as early as the 3rd century B.C., just beyond the land of the Kalingas towards the south was the principality of the Andhras, that just beyond the latter was the independent kingdoms of the Cholas and Pandyas extending as far south as Tamraparni (nicam Coda Pamdiya avam Tambapamñiyā).
Our old Brahmi inscriptions are wanting in such clear data concerning the southern limit of the Mahasuliya-Mosali division of the Kalinga kingdom of King Kharavela. These are totally silent about the Andhras and the Cholas. But the Hathi-Gumpha inscription records that, in the twelfth regnal year of King Kharavela, the King of Pandya supplied him with the most valuable presents of pearls, gems and jewels and various kinds of apparels (Pamda-rajā vividhábharaṇāni muta-mani-ratanani aharapayati idha sata-sahasāni). In the absence of any mention of the Andhras and the Cholas, there may not be much difficulty in imagining that the second division of Kharavela's Kalinga stretched along the sea-coast even beyond the Godavari, and as far south as the mouths of the Krshna, if it was then known, as supposed, by the name of Mahasuliya-MaisoliaMosali-Mosala. Anyhow, the problem remains why the Andhras and the Cholas whose principalities lay to the north of Pandya have been passed over in silence in the inscription of King Kharavela.
A clear hint might be taken from one of the geographical allusions in the Hathi-Gumpha inscription in establishing the fact that the suzerainty of King Kharavela of Kalinga was felt in the south, along the eastern coast of the Deccan, into the very heart of the land of the Cholas, below the Krshna, below the Pennar, as far down as the northern limit of the kingdom of Pandya. The allusion referred to above is in the record of Kharavela's fourth regnal year wherein we read that His Majesty caused Arakatapura, the city of Arakata, inhabited by a race of magicians called

Page #234

206
OLD BRAHMI INSCRIPTIONS
Vidyādharas (Visādharadhivāsa Arakatapura), to be governed according to the established usage and administrative principle of the former kings of Kalinga, by the highest kind of law, and that he compelled all the tribal chiefs and village headmen of the place (sava-rathika-bhojake) who trembled in fear, whose royal insignia consisting of umbrellas and golden vases were taken away, and who were robbed of their jewels and riches, to bow at his feet (1.5). Arakata or Arakata is the same kind of geographical name as Parakata Bhojakata, or Bibikānadikata in the Barhut votive labels, and phonetically it is the same name as the modern Arcot which is wrongly "believed to be quite modern."} The Greek geographer Ptolemy speaks of a northern kingdom of Sorae Nomades, the Sora nomads, which lay between Mount Bettigo and Adeisat bros. With regard to the royal city of these nomads, in some editions of Ptolemy's work, the statement is sora regia Arcati, “Sora, the capital of Arcatos," and in some editions, it occurs as Arcati regia Sora, “ Arcatos, the capital of Sora.” The latter reading has come to be accepted as the correct representation of Ptolemy's statement, and Dr. Caldwell identifies Ptolemy's Arcati-Arcatos with the modern Arkad or Arcot. Sir Alexander Cunningham considers the Sorae Nomades of Ptolemy to be a branch of the Sauras, a race of aborigines, "who are still located on the banks of the Kistna river: while Prof. D. R. Bhandarkar is of opinion that the Sora of Ptolemy " can easily be recognised to be the Tamil Sora or Choda." 4 If Prof. Bhandarkar's suggestion has any truth in it, one has to understand that the dwellers of Ptolemy's Arcati-Arcatos were a race of the Chola nomads. Khāravela's inscription also distinctly says that the inhabitants of Arakatapura were the Vidyādbaras, an aboriginal people noted for their magical skill. The mention of Arakatapura in Khāravela's inscription and that of Arcati-Arcatos in Ptolemy's work set at once at nought the belief about the modern origin of the name of Arkad-Arcot, near Madras. If this identification be correct, it will be a mistake to suppose that the words rathika and bhojaka as used in Khāravela's inscription have precisely the same implication as that in Asoka's edicts. As used in Khāravela's inscription, the words are to be treated rather as titular designations than as names of any semi-independent tribes."
1. Cunningham's Ancient Geography of India, edited by S. N. Majumdar, p. 626. 2. JA, Vol. VIII, p. 362. 3. Cunningham's Ancient Geography of India, edited by S. N. Majumdar, p. 627. 4. Bhandarkar's Asoka, p. 89.
5. Anguttara-Nikāya, Part III, pp. 76, 78, 300. Maporatha-Pärani, Siamese edition, Part III, p. 119: Ratthiko' ti yo rattham bhunjati-Gāmayāmiko' ti gāmablojako.

Page #235

NOTES
207
But, upon the whole, one must take the river Godavari as the southern boundary of Khāravela's Kalinga, which appears to have extended along the Vindhya range as far as the western valley of the Godāvari. In the western direction, the only great royal power which King Kbāravela had to reckon with was that of King Sātakarqi, whose territories must have comprised a number of small states near about the western valley of the Godavari. In the Hathi-Gumphā inscription, in the record of King Kbāravela's second regnal year, we read that His Majesty, without taking King Sātakarņi into his thought (acitayitā Sātakaạim) caused a large army consisting of all the four divisions of Indian troops to move towards the Western quarter (pachima-disam......pathāpayati), and struck terror into Asakanagara, or it may be, into Asikanagara, with the army from Kalinga (Kalingágatāya senāya).
If Mr. Jayaswal's reading Kanhabernágatāya senāya vitāsitam Musikanagaram be accepted as correct, we have to say either (1) that King Khāravela succeeded in striking terror into Musika-nagara, the Mūsika. capital, with the aid of the army that advanced on the banks of the Kronavenā river, or (2) that he achieved this military feat with the army that advanced from the Kronavenā river, the expression Kañhabemnagatāya admitting of a twofold interpretation as suggested above. Taking the expression in the first sense, Mr. Jayaswal has sought to maintain that the presumed Musika-nagara was situated on the banks of the Kañhabenā or Kronavenā river.
Mr. Jayaswal's notes on the Mūşikas and the Krsmarenā, written in justification of his reading quoted above, are worth quoting in this connection. First, as to the Mūşikas, he has written: “They were a people of the south. The Mahābhārata (VI. 9. 58) mentions them in the company of the Vanavāsis. Their country could not have been far removed from Kalinga, for the Nātya-Sāstra (circa 100 B.C-100 A.C.) describes the Tosalas (the people of Tosali), the Kosalas (the people of Southern Kosala), and the Mosalas (the Mūşikas) as the Kalingas, implying that they comprised the Kalinga empire. This is a description naturally subsequent to the time of Khāravela. A more definite reference is in the Purāṇas (Wilson, Vişnu, IV, p. 221) where after a kingdom of some Vindhyan countries Stri-rājya and Müşika countries are mentioned as forming one princedom. According to the Käma-Sūtra Strī-räjya was a Vindhyan I country towards the West. The Mūşika country must have been between latitudes 20 and 22, between Paithana and Gondwana. As Kosala came

Page #236

www. kobatirth.org
208
OLD BRĀHMI INSCRIPTIONS
{next to Orissa (North-West), the Mūşika-land must have been contiguJous to it on the West."
Secondly, as to the Krenavenā, he has written : “ The Puranas place this river near the Godavari and treat it as distinct from the Southern Krshna. The Krshna-venā is mentioned in some authorities as if it were two rivers Krslinā and Vena.? Mārkandeya derives it from the Vindhya range.3 It is evident from these details that the river is identical with the modern · Wain River' or Wain-ganga' which has for its main tributary the Kanhan. The Kanban and the Wain unite in the district of Bhandara and the united stream comes down to meet the Wardha in the district of Chanda.. ...... The capital of the Muşikas, which the Hathi-Gumpbā inscription contemplates to be on the Kanhabenā, was situated somewhere within the districts of Bhandara and Chanda in the Central Provinces." 4
It cannot be denied that the Mūşikas were a people of the south. We are aware that the political division of Musika “ lay to the south of Kerala or Chera, the country south of Kupaka (or Satya), extending down to Kanheri in Central Travancore."5" Mr. Jayaswal is the best person to say if he has not misled his readers by stating that the Mahābhārata mentions the Müsikas" in the company of the Vanavāsis." The Mahābhārata, as will appear from the following quotation, has not only mentioned the Mūşikas in the company of the Vanavāsis, but also in that of the Drāvidas and the Keralas or Cheras :
Atbăpare jānapadā dakşină Bhāratargabhai
Drāvidāḥ Keralāḥ prācyā Mūşikā Vanavasikāḥ | We are ignorant of Mr. Jayaswal's authority for his identification of the Mosalas of the Natya-Sāstra with the Mūşikas. Vātsyāyana's Kāma-Sūtra, so far as we have read it, keeps us entirely in the dark about the location of Strīrājya. Yasodhara in bis commentary on the Kama Sutra, says that the Strīrājya lay to the west of the country called Vajravanta,
1. JBORS, 1917, Vol. III, Part IV, p. 442. 2. Vāyu-Purāņa, LV. 103. 3. Mārkandeya-Purāņa, LXVII. 26-27 :
Godavari-Bhimaratha-Kronavenya-tathapara
Vindhyapādā-viniekrāntā ityetā sariduttamā il 4. JBORS, 1918, Vol. IV, Part IV, pp. 374-375. 5. JRAS, 1923, p. 413.
6, Kāma-Sūtra, VI. 5.27 : "Stri-rājye Kosalāyām ca." Stri-rājya iti-Vajravanta. deáāt pascimena Strī-rājyam.

Page #237

www. kobatirth.org
NOTES
209
and that the Grāmanāri-vişaya was situated adjoining and on the other side of the Strīrājya. Hence, even if it be true that the Purāņas mention Strīrājya and Mūzika countries as forming one princedom, Mr. Jayaswal has yet to supply us with a definite evidence for locating Strīrājya as "a Vindhyan country towards the West."
It may be, as Mr. Jayaswal argues, that the Krşņaveņā is the same river as the modern Wain which unites with the Kanhan, its main tribu. tary, in Bhandara District, and that the united stream comes down to meet the Wardha in Chanda District in the Central Provinces. But is there any independent evidence, we ask, to prove that the Mūşika city or country was situated on the banks of the Kroņaveņā ? Even assuming his reading Kanhabenágatāya to be a correct one, can it be definitely suggested that the Hātbi-Gumphā inscription contemplates Musikanagara, the capital of the Mūşikas, “ to be on the Kañhabenā"? Can it not also be suggested that King Kbāravela struck terror into Muşika. nagara with the army that advanced from the Kanhabenā ? While Mr. Jayaswal's Hāthi-Gumphā inscription contemplates the capital of the Mūşikas to be on the Krşņaveņā, Dr. H. C. Raychaudhuri finds reasons to suggest that " the Mūşikas were probably settled on the banks of the river Musi on which Hyderabad now stands." 2
We have maintained that the intended reading is not Kañhabendgatāya but Kalimgågatāya; not Musika-nagara but Asaka-nagara or Asika-nagara. Had the letter been intended to rea i mu, there is no reason why the 1-mark should appear as the lower prolongation of a straight vertical line on the right. We have offered a good explanation for the probable appearance of the 2-mark along with sa, in case such a vowel-mark was not in the intention of the engraver. We also have shown how a few holes on the inscribed surface, on two sides of the letter li, are accountable for the production of a mirage of a letter, which Mr. Jayaswal reads siha in the estampages.
If our reading be correct, as we believe it is, a great advantage to be derived therefrom is that it does not compel us to resort to a number of assumptions without any proofs, that the Mūzikas “ were probably a subordinate ally of Sātakarņi," that the capital of the Müşikas was a city on the river Kronavenā, that Strīrājya and Masika countries formed at
Strirajya."
1. Käma-Sūtra, VI. 6.42: "Grāmanārīvisaye Stri-räjye ca Vahlike." samipa eva parato Grāmanārī viņayah.
2. Political History of Ancient India, 2nd edition, p. 59.
27

Page #238

210
OLD BRAHMI INSCRIPTIONS
one time or another one princedom in a Vindhya region towards the West, and so on and so forth.
The foregoing discussion concerning Arakatapura will have no effect and the question as to the non-mention of territories of the Andhras and Cholas will remain undecided if the reading proves to be otherwise. Mr. Vats' estampage reproduced by Mr. Jayaswal in JBORS, Dec., 1927, clearly presents a word consisting of five syllables, the letter representing the second syllable appearing as ha and the letter representing the fifth syllable as va. If so, we have to read the intended word either (1) as ahatapuva (=ahatapurva, unhurt before), or (2) as (a)hatapuva (=ahṛtapūrva, acquired beforehand). Mr. Jayaswal reads ahatapuva interpreting it in the sense of unhurt before.' Such an interpretation as this is utterly inappropriate in a statement where King Kharavela is said to have governed the Vidyadhara-abode (Vijadharádhiväsa) in the manner of the former kings of Kalinga compelling all the Rattbika-Bhojakas (sava RathikaBhojake) to pay due homage to him, no doubt as a test of their owing allegiance to him as they did to his predecessors. The statement, taken as a whole, indicates that the Vidyadhara-abode was an aboriginal tract of land, which was annexed to Kalinga previously, that is to say, by some Aira-Meghavahana king among Khāravela's precursors. This sense can be brought out only if one agrees to read the intended word as (a) hatapuva interpreting it in the sense of "acquired beforehand."
Having regard to the fact that the statement just referred to has been made immediately after the statement in which Kharavela is said to have struck terror into the heart of the city of Asaka (Asvaka) or Asika (Rṣika), we may maintain that the Vidyadhara-abode was situated somewhere near Agvaka or Ṛsika on the western border of the kingdom of Kalinga, and that this abode formed the territories of the Rathika Bhojakas who find mention in the edicts of Asoka (R. E. V, R. E. XIII) as semi-independent ruling races dwelling somewhere in Western India. The Raṭṭhika-Bhojakas figure in Asoka's edicts among westerners, the peoples of Western India (apalamta, aparanta). In R. E. V. the Ratthikas as a ruling race ale mentioned in combination with the Pettanikas (Lathika-Pitenika), while in R. E. XIII, we have the mention of the Bhojas or Bhojakas in the same combination (Bhoja-Pitinika). It would appear from this either (1) that the Raṭṭhikas and the Bhojakas were two confederate clans, or (2) that one and the same ruling race was known by two different names, such as Raṭṭhika and Bhojaka. That the Ratthikas and the Bhojakas were two separate clans is undoubted because, corresponding to them, we have

Page #239

NOTES
211
the mention of the Mahāratthis and the Mahābhojas in the records of the Sātavāhana kings. Prof. D. R. Bhandarkar has wrongly maintained Pitinika to be an adjective of Rathika or Bhoja on the strength of the Pāli passage in the Anguttara-Nikāya (III, pp. 76, 78, 300). It is clear from the Pāli passage, as well as from Buddhaghoşa's explanations, that Ratthika and Pettanika were two different or titular designations. Buddhaghoga, for instance, says: Ratthiko' ti yo rattham bhuñjati, "the Raţthika is one who enjoys the income derived from rāştra (as defined in the Artha-Sāstra, II. 6. 2 t)'; Pettuniko' ti yo pitarābhuttánubhuttam bhuñjati, " the Pettanika is one who enjoys the hereditary rights and privileges." The term Bhoja or Bhojaka may be interpreted either in the sense of rattha-bhojaka or ratthika, “ the enjoyer of rāștra," or in that of gamagā maņika or gāma-bhojaka, " the village headman.” If Berar and Konkan were the principalities of the Ratthika-Bhojakas, we can say that the Vidyadhara-abode was co-extensive with the ancient kingdom of Dandaka (Sarabhanga-Jataka, No. 520) bounded on the east by the kingdom of Kalinga, on the west by the kingdom of Avanti, on the north by the kingdom of Kaši, and on the south by the kingdom of Asvaka to the south of the Vindhya range. We might go further and maintain that this Vidyadhara-abode was no other than what is mentioned in the Purānas as StrIrāştra or Strirajya. Depending on the reading in one of the MS. of the Vişnu-Purāņa, Mr. Jayaswal says that in the Purāṇas, Strīrājya and Mūşika country are said to have formed one principality.3 The VişnuParāna is not all the Purā vas. Even all the MSS. of the Vişņu-Purāņa do not offer the same reading. The different MSS., as noted by Mr. Pargiter, give different readings, one giving Mūşiva, one Mūkhika... one Musith, and one Musika. 4 The Purāṇas as a whole telt altogether a different story. In them, we read that Strfrāstra and Bhokşyaka formed one principality under a king named Kanaka.5 Curiously enough, we have precisely Bhojaka as a variant for Bhokşyaka, which is significant as implying that Strīrāştra and Bhokşyaka formed the principalities of the Rathika-Bhojakas.
1. Sitā bhāgo balıḥ karo vanik nadipālastaro nāvah pattanam vartani rajjūs corarajjūs ca rāętram,
2. H. C. Raychaudhuri's Political History of Ancient India, 2nd edition, pp. 195.197.
3. Pargiter's Dynasties of the Kali Age, p. 54, f.n. 25: Strirājya-Trairāja-Müşika. janapadan Kanakāhvayā bhok yanti. Some of the MSS. omit Trairāja.
4. Ibid, p. 54, f.n. 25.
5. Ibid, p. 54; Strīraştram Bhokayakāms caiva bhokayate Kanakähvayah. For the variant Bhojaka see ibid, p. 54, f.n. 24.

Page #240

212
OLD BRAHMI INSCRIPTIONS
NY
HA
According to Khāravela's inscription, King Sātakarņi was the only powerful rival on the western border of his Kalinga kingdom whom he defied in striking terror into Asaka-nagara or Asika-nagura, the capital of Asaka or Asika. Asaka is just a Prākrit form of the Pāli Assaka and the Sanskrit Asvaka or Asmaka, and Asika is nothing but a Prakrit form of the Sanskrit Rşika or Arşikai met with in Patañjali's Mahābhāşya (IV. 2.2). It is clear from Ptolemy's Geography that the capital of Pulumāyi or Vāsisthiputra Sri Palumāyí of the Satakarņi Sātavāhana dynasty, was Baithan, Paithan, Patițýhāna or Pratisthāna, which became, according to the Prologue (Vatthugātha) of the Pārāyaṇavagga in the Sutta-Nipāta, the southern destination of a great trade-route called Dakkhiņāpatha or Southern Road.2 The Pārāyana-Prologue speaks of the hermitage of a hermit named Bavari as a religious institution founded on the bank of the Godāvari river (Godhāvarikūle), in the dominion of Assaka, in close proximity to Aļaka or Mulaka (Assakassa visaye Mulakassa samāsane). The author of the Sutta-Nipita-Commentary, writing in the 5th century, A.D., says that Bāvari's hermitage was situated in a strip of territory, which stood midway between the two Andhra kingdoms of Assaka and Alaka or Muļaka, that is to say, in that region where the Godāvarı having divided
into two streams, formed a doab, between them, and where in former times, the Hermit Sarabhanga and others lived. 3
In referring to the hermitage of Sarabhanga the author of the SuttaNipāta-Commentary kept evidently in his view the legend of th bhanga-Jātaka which points to a time when the high road called Dakkbiņā. patha had not come into existence, when a traveller undertaking a from Benares to the valley of the Godāvari had to proceed by a one-foottrack, or a "road for a foot-passenger (eka-padi-magga) under the escort of a forester (vanacaraka).
This Jātaka legend is on a par with the Rāmāyaṇa account of Rāma's wanderings from Ayodhyā in the north to Pañcavați on the bank of the Godavarī in the south, which too, points to a time when a traveller
1. Noticed by H. O. Raychaudhuri in his Political History of Ancient India, 2nd edition, p. 310, f. n. 2.
2. Rhys Davids' Buddhist India, Ch. V.
3. Paramatthajotikā, Vol. II, p. 581 : " Assakassa visaye Aļakassa (Muļakassa) samasane' ti Assakassa ca Alakussa ca ti dvinnam Andhaka-rājānam samásanna visaye, āsanne ratthe, dvinnam pi ratthanam majjhe ti adhippayo. "Godhavari-kūle ti Godhāvari. nadi-kūle, yattha Godhāvari dvidhā vijjhitvà tiyojanappamānam antara-dīpam akāsi sabbam kavitthavana-sañchannam, yattha pubbe Sarabhangadayo vasimsu tasınim padese,

Page #241

NOTES
213
from the north had to proceed to the south from hermitage to hermitage, following a forest track.
When exactly the high road was built and by whom we cannot say. But it is certain, as evidenced by several Pāli Dialogues, that some such high road was already in existence in the life-time of the Buddha, any. how from Rājagaha to Kusīnārā, fron Kusinārā to Pāvā, from Pāvā to Kapilavatthu, from Kapilavatthu to Säketa vid Setavya and Sāvatthi, and from Säketa to Kosambi and further south. And yet the Prologue of the Pārāyana-vagga in the Sutta-Nipāta must be considered a later addition and relegated to a post-Asokan age for the simple reason that it did not find any place in the Culla-Niddesa which is a Pali canonical exegesis on the Pārāyaṇa group of sixteen poems and undoubtedly a composition of a post-Asokan time. Even this Prologue has nothing in it to indicate that Assaka and Aļaka or Mulaka were two Andhra territories. It is only in the Sutta-Nipāta-Commentary that we have the mention of Assaka and Muļaka as principalities of two Andhra kings. As to who these two Andhra-kings were, this commentary does not supply us with any information whatever.
The Cullakālinga-Jätaka (Fausböll, No. 301) presents in its quoted verses an old chronicle of a war, which ensued between the Assakas and the Kāliigas when King Aruņa was the ruler among the Assakas, and ended in the victory of the former and the defeat of the latter. In explaining the historical allusions, the commentary version of the Játaka in prose introduces certain details that are beyond the scope of the chronicle in verse.
According to the commentary version, the war alluded to in the chronicle broke out between Kalinga and Assaka when King Kālióga reigned in Dantapura, the capital of Kalinga, and King Assaka reigned in Potali, the capital of Assaka. It represents the then reigning king of Kalinga as a ruler who was equipped with a powerful army (sampannabala-vāhano) and a mighty warrior who was unrivalled in military prowess (nāgabalo patiyodhum na passati), a description worthy of King Kbāravela of Kalinga as he appears in the Hāthi-Gumphā inscription.
These three points are clearly brought out in the commentary version of the Jātaka : (1) that the war which ensued was an agressive war inflicted on the king of Assaka by the then reigning king of Kalinga who suffered from the mania for war and love of conquests over the whole of India, a spirit so much extolled in Khāravela's inscription ; (2) that the battle was fought on a field between the boundaries of the two kingdoms

Page #242

214
OLD BRÄHMI INSCRIPTIONS
(dvinnam rajjānam antare), it implying that the kingdom of Assaka lay contiguous to the kingdom of Kalinga towards the west exactly as during the reign of King Khāravela ; and (3) that the king of Kalinga was ultimately compelled to return to his capital, acknowledging his defeat, in spite of the fact that he stacted with a mighty army (mahatiyā senāya nikkhami), a fate which is virtually the same as the final result of Khâravela's attack on the capital of Assaka, if it had been recorded in an inscription of the King of Assaka.
Whatever be the period of time to which the evidence of the commentary version of the Jātaka refers, it cannot be doubted that this evidence is in some important respects similar to that of Khāravela's inscription in the Hāthi-Gumphā. If its first verse, which appears to have been a later manipulation, be left out of consideration and be linked up with the prose story, the older chronicle in verse becomes reduced to a bardic narration of a tribal feud between the Assakas and the Kalingas, whereas the prose story in the Jātaka-commentary has for its theme the final result of a war, which ensued between two kings, viz, King Kālinga of Kalinga and King Assaka of Assaka. According to the first verse which we consider to be a later addition, Aruna was the family or personal name of the then reigning king of Assaka, and Nandisena was the name of his commander-in-chief feared for his valour.! The Jātaka as a whole keeps us in the dark as to the family or personal name of the then reigning king of Kalinga and the name of his commander-in-chief. It is highly improbable that any king of Kalioga before or after Khāravela, up till the 5th century A.D., is known to have grown so powerfal and ambitious as to make bold to stir out for effecting conquests all over India. As to the relationship of Aruņa, the then reigning king of Assaka, either with the Andhras or with the Sātakarni-Sātavāhanas, the Jātaka has nothing to say. But it definitely says that the king of Kalinya ultimately suffered a defeat and failed of his purpose in waging war upon the kingdom of Assaka, in spite of his being a mighty warrior unrivalled in military prowess and equipped with a powerful army, in spite of the fact that he marched towards Assaka with a mighty army. The Hāthi-Gumphā inscription, too, clearly brings out the same fact, namely, that King Khåravela of Kalinga could do no more than striking terror into the heart
1. Vivaratha imāsam dvāram, nagaram pavisantu 1 Arunarājassa sihenn susatthena surakkhitam
Nandisenenä til

Page #243

NOTES
215
WARENA
of the capital of Asaka or Asika in spite of the fact that he caused a large army consisting of all the four divisions of Indian troops to move on towards the west without paying any attention to the presence of sātakarņi's power. It cannot be inferred from this inscription that Asaka | or Asika was ever permanently annexed by Khāravela to his kingdom.
That the Sātakarạis and other kings of the Sātavāhana dynasty exercised suzerainty over Asaka, Asika and a number of other subordinate states, which clustered round the western Godavari valley to the west and south of Kalinga, cannot at all be doubted. For we find that in the Nasik cave inscription of Queen Gautami, dated in the 19th year of the reign of her grandson King Vāsişthiputra Sri-Pulumāyi, her son King Gautamiputra Sri-Satakarni has been eulogised as the ruler of Asika, Asaka (Asmaka on the Godāvari), Mulaka (the district round Paithan), Suratha (Kathiawar), Kukura (in Western on Central India, near the Pāriyātra or Western Vindhyas), Aparamta (North Konkan), Anupa (the district round Mahişmati on the Narmadā), Tidabha (Berar), and Akāra-Avati (East and West Malwa), and no less as the lord of the Vijha, Chavata, Pāricāta, Sahya, Kaņhagiri, Maca, Siritana, Malaya, Mahida, Selagiri and Cakora pavata, in short "of all the mountains from the Vindhyas to the Travancore bills."
There is little doubt that in Queen Gautami's inscription the same eulogium was meant also for her grandson King Väsişthiputra Sri-Pulumāyi. Even after King Mahākṣatrapa Rudradāman I of the Ksaharāta family had wrested some five of the western territories, after A.D. 130 and before A.D. 150, from one King Sri-Satakarai, the lord of the Deccan, who was perhaps King Vāsişthiputra Sri-Pulumāyi bimself, Asi ka, Asaka and the rest of the subordinate states mentioned in Gautami's inscription remained included in the dominions of the Satakarņi-Sätavahana kings up till, we venture to say, the reign of Yajña-Sri-Sātakarại, the last great king of the Sātavāhana family. Even going back to earlier times, to a period of the Sātakarņi-Sātavahana reign before the Kșaharāta rule in Western India and the Daccan and occupation of Mahārāztra in the
1. EI, Vol. VIII, p. 60: Asika- Asaka-Mulaka-Suratha-Kukura-Aparamta-Anupa-Vidabha-Ākāra-Avati-räjasa Vijha-Chava'a.Pāricāta-Sahya-Kanhagiri. Maca-Siritana Malaya. ifahida-Setagiri-Cakora prvata prtisa. See, for the identification of the places, H, C. Ray. chaudhuri's Political History of Ancient India, 2nd edition, pp. 310-311.
2. Namely, Ākāra- Avanti, Anupa, Surāştra, Kukura and Apzrānta. See EI, Vol. VIII, pp. 44-47.

Page #244

216
OLD BRAHMĪ INSCRIPTIONS
time of Nahapāna, nay, to the reign of king Sri-Sātakarņi I, the son of Sinuka who was the founder of the Sātavāhana branch of the Andhra family, the successor of King Krşna who was Simuka's brother, and the husband of Queen Nayanikā, we find the same state of things. The Nīnāzhāt inscription of Queen Nayanikā and the Sanchi inscription of Ananda, the leading artisan of one King Sr1-Sātakarņi bear testimony to the rise and existence of the first great Šātakarni-Sātavāhana empire in the western Godāvari valley “which," as Dr. Raychaudhuri acutely puts it, “rivalled in extent and power the Sunga empire in the Ganges valley and the Greek empire in the Land of the Five Rivers."'1
Thus the epigraphic records connected with the Satakarņi-Satavāhana and Ksaharāta Saka-Kşatrapa kings enable us to understand not only (1) why Ptolemy wrote to say that Baithan or Paithan (Patitthāna on the Godāvarī, the southern terminus of Dakkhiņāpatha, the Southern Road) was the capital of King Pulumāyi Vāsişthiputra Sri-Pulumāyi, and (2) why Assaka and Mulaka or Aļaka have been represented in the Sutta-Nipāta-Commentary as two Andhra principalities in the western valley of the Godāvari, but also (3) why the Hathi-Gumphā in scription records that King Kbāravela of Kalinga defying King Sātakarņi caused a large army consisting of all the four divisions of Indian troops to move on towards the west to strike terror into the heart of the capital of Asaka or Asika.
If it has been conclusively proved that Asaka or Asika, which was one of the subordinate states in the dominions of King Satakarņi, lay to the west of Khāravela's Kalinga, there can be no denying the fact that the Kalinga kingdom of Khāravela included in it the third division, Kosala or South Kosala, “ which comprised the modern Bilaspur, Raipur and Sambalpur districts, and occasionally even a part of Ganjam."?
Both Mr. Jayaswal and Dr. Sten Konow seem to think that Kosala or South Kosala was that division of the Kalinga kingdom in the time of King Asoka which comprised antā avijitā, the unconquered outlying tracts, referred to in two copies of Asoka's Separate Rock Edicts. Even if it be so, we find it impossible to say that any Aira King of Utkala, associated in
1. Political History of Ancient India, 2nd edition, p. 264.
2. H. C. Raychaudhuri's Political History of Ancient India, 2nd editon, pp. 337-338. fee, for extension of Kosala when Hwen Thsang visited it, Cunningham's Ancient Geography of India, S. N. Majumdar's edition, pp, 595, 735,

Page #245

NOTES
217 the Sanskrit verses from an old Oriyā Ms. with Kosala, was the then reigning monarch of Kosala.
The denotation of the term antas or pratyantas has varied with different authorities. As used in Buddhist literature, it denotes the regions that lay immediately beyond what is called Madhya-deśa, the Middle Country. As used in the Allahabad Pillar inscription of Samudragupta, it denotes such North-Eastern frontier states as those of Samatata, Davāka, Kāmarūpa, Nepāla, Karttạpura and the rest. And as used in Asoka's inscriptions, it denotes the unconquered territories or tracts that lay beyond his dominions, whether in the north-west, or in the south, or in the east. In his Rock Edict XIII, King Asoka has expressly mentioned the five Greco-Bactrian principalities as typical examples of antas in the north-west, and Chola, the land of the Cholas, and Pandya, the land of the Pāņdyas, as typical examples of antas in the south. In speaking of the independent dwellers of the antas with whom his Viceroy and High Functionaries at Tosali and Samāpā were to deal, King Asoka has not cared to name any ruling chiefs, peoples or powers as worthy of mention. If by antas in connection with his Kalinga Province he really meant Košala or South Košala, it must have then comprised the feudal states of a number of tribal chiefs, instead of being one kingdom under the rule of King Khāravela.
It is recorded in the Hathi-Gumpbā inscription that King Khāravela, in the eleventh year of his reign, caused the grassy jungle of Přthudaka to be driven out into the Länyala river and caused the watery jungle of grass to be destroyed, the watery jungle which was allowed to grow up in one hundred and thirteen years. The statement is important as giving us some glimpses into the physical features of Kalinga, of which we are offered the following interesting description in Mr. Mano Moban Ganguly's Orissa and Her Remains, pp. 6-7:
"It was always a terra incognita, by reason of its geographical position L and local circumstances. The whole country was a swamp intersected by a network of rivers with their feeders and tributaries carrying silt to the Bay of Bengal; impassable hilly jungles fringed its borders. The country was in a process of geological formation; there was a perpetual struggle between the different rivers and the sea, for the formation of the deltaic regions."
In the Hathi-Gumphā record of Khāravela's twelfth regnal year, we read (1) that His Majesty was able to strike terror into the heart of the kings of Uttarāpatha, (2) that in generating a great fear for the
28

Page #246

218
OLD BRĀHMI INSCRIPTIONS
people of Anga and Magadha he caused the elephants and horses to drink at the Ganges, (3) that he compelled King Bahasatimita of Magadha to bow down at his feet, (4) that he triumphantly brought back to Kalinga from Anga-Magad ba the Throne of Jina belonging to Kalinga which was carried away by King Nanda, (5) that be compelled the inhabitants of Anga and Magadha to bow down at his feet, (6) that a hundred Vāsukis sent him jewels, procured for him the elephants, horses and other animals capable of exciting curiosity, and made presents of deer, horses and elephants, (7) that the king of Pāņdya supplied him with many valuable presents of apparels, pearls, gems, and jewels, and (8) that he subdued the inhabitants of some other place.
The significance of the geographical designation Uttarāpatha is to be understood in contrast to that of Daksimāpatha. As we noticed, the Prologue of Book V of the Sutta-Nipāta represents Dakşivāpatha as a great trade-route, the Southern High Road, which starting from Rājagaha, the capital of Magadha (Magadha-pura), passed through Nālanda and Pātaligāma (where the city of Pāțaliputta was afterwards built), to extend northward to Vesāli across the Ganges and via Kotigāma, and extending as far as Kapilavatthu via Kusinārā and Pavā, turned southward to reach Sāvatthi via Setavya, and proceeding southward passed through Sāketa to extend to Kosambi across the Yamuna, and proceeding further south, reached Vana or Vana-Sāvatthi, at which point it turned towards the west to pass through Ujjeni and Māhissati, and extended as far south as Patiţthāna or Paithan across the Vindhya range. This Southern High Road lent its name to the region through which it passed, the whole tract of land lying to the south of the Ganges and to the north of the Godāvarī being known, according to Buddhaghoşa, as Dakşiņāpatha, the Deccan proper.1 That this region was known up to a late period as Dakşiņāpatha is evident from these two literary references : (1) Dākşinātye janapałe Pūtaliputram nāma nagaram, “ Pătaliputra a city on the south bank of the Ganges in a Deccan district,” occurring in one of the stories of the Pañcatantra,' and () Avantī Dakkhiņāpathe, “ Avanti in Daksimāpatha," occurring in one of the passages of the Vinaya-Pitaka. This original significance of
1. Sumangala-Vilasini, Part I, p. 256. 2. Paficatantra, Sreşthi-Kșapaņaka-Nāpita-Katha. 3. Vinaya-Malāvagga, V, 13, 0.' Vinaya-Callavagga, XI, 1,

Page #247

NOTES
219
the name Dakşiņāpatha was gradually lost sight of with the result that in later times the location of Daksiņāpatha came to be shifted to the south of the Vindhya range,? to finally cover the whole of the modern South India lying between the Vindhya range and the Cape Comorin, the Deccan proper still being the region between the Godāvari and the Krşıā. It may be easily inferred from the Nasik cave inscription of Queen Gautami, from the list of countries included in the dominions of Gautamiputra Sri-Sātakarņi, that, as late as the first two centuries of the Christian era, the kingdom of Avanti was located in Dakşiņāpatha.
A similar historical process can be conceived with reference to Utarā padha or Uttarāpatha. For Uttarāpatha, too, may be supposed to have been originally a great trade-route, the Northern High Road, whic extended from Sāvatthi to Takkasilā in Gandhāra, and have lent, precisely like the Southern High Road, its name to the region through which ity passed, the region, broadly speaking, covering the north-western part of the United Provinces, the whole of the Punjab Province and the NorthWestern Frontier Provinces. That this region was known up to a late period as Uttarāpatha is evident from these three literary references : (1) Uttara-Madhura-Uttarāpalhe," the Northern Mathurā, the Mathurâ proper on the Yamunā, in Uttarāpatha," occurring in Dharmapāla's ParamatthaDipaní, a commentary on the Petavatthu ; (2) Pșthudakāt paratah Uttarā. pathaḥ, “ Uttarāpatha lies (towards the west) beyond Pșthudaka (near Thaneswar)," occurring in the Kāvya-Mimāmsā ;? and (3) the following verse occurring in the Mahābhārata, XII, 207, 43 :
Uttarāpatha-janmanah kārtayisyāmi tān api Yauna.Kamboja-Gandhārah Kirāta Barbaraih saha || 8
Buddhaghosa in his commentary on the Kathāvatthu, attributes certain views in the Kathāvatthu to two Buddhist schools called Hemavatika and Uttarāpathaka. Hemavatika and Uttarā pathaka were the local or
1. Kāvya-Mimāmsā, p. 93. M&hişmatyaḥ parataḥ Dakşiņāpathaḥ, “ Dakşiņāpatha is the tract of land which lay beyond Māhişmati in the south." Cf. Bhāratiya Natya-Såstra, XIV, 39-41.
2. Kavya-Mimāınsă, p. 93, quoted in Canningham's Ancient Geography of India (8. N. Majumdar's edition), p. 965.
3. Quoted in H. C. Raychaudhuri's Political History of Anoient India, 2nd editio p. 83.

Page #248

220
OLD BRAHMI INSCRIPTIONS
territorial names of two post-Asokan Buddhist schools, the former being derived from Himavanta, and the latter from Uttarapatha. We have reasons to say that the Himavanta region consisted of a number of Himalayan states to the north of the Southern High Road in its extension from Vesali to Savatthi and beyond the northern boundary of the Middle Country as defined in Buddhist literature. Thus Uttarapatha may be accurately defined as a tract of land, which lay to the west of the Himavanta region, extending westward from Thaneswar, and which lay to the north-west of the Buddhist Middle Country and to the north of the Dakṣinapatha, extending north-west from Mathura, the capital of Surasena. Whatever be its later territorial extension, it is certain that Kharavela's Uttarapatha signified nothing but the region specified above, the region including Mathura in its south-eastern extension. Anyhow, from the record of Kharavela's twelfth regnal year, it is clear that Uttarapatha lay towards the west or north-west beyond Anga and Magadha. And should our reading be correct, from the use of the plural expression Uttarapadha-rajano, "the rulers of Uttarapatha," it may be inferred that when King Kharavela carried his campaign into Uttarapatha, it was parcelled, precisely as it was when Alexander invaded the plains of India, into a number of small independent principalities, although the Hathi-Gumpha inscription does not mention the names of their rulers.
According to the Mahabharata account of the journey of the Pandavas from Hastinapura to Kalinga, Kalinga proper could be reached by the travellers journeying along the sea-coast from the mouth of the Ganges, and it was the country through which the river Vaitarauf flowed.1 On the strength of this Epic description, Mr. Mono Mohan Ganguly assigns the following boundaries to Kalinga proper : "On the North, the Vaitarani; on the South, the Godavari; on the East, the Bay of Bengal ; on the West, the Tributary States of Orissa."2 The reader can judge for himself whether and how far the geographical extension of Kalinga proper,
1. Mahabharata, Vanaparva, Canto I:
Tataḥ samudra-tirena jagama rasudhadhipah | Bhrātṛbhiḥ sahito vīraḥ Kalingan prati Bhārata || Lomasa uvaca :
Ete Kalingah Kaunteya yatra Vaitarani nadi | Yatrayajata dharmopi devaccharanametya vai || 2. Orissa and Her Remains, p. 9.

Page #249

NOTES
221
as may be inferred from the Mahābhārata account, coincided with that of the Kalinga kingdom of Khāravela. Apparently, as we have sought to show, Khāravela's Kalinga kingdom with its three divisions comprised the Epic Kalinga as well as the modern Orissa with its Tributary States.
The countries of Anga and Magadha are too well-known to need any elaborate comments. The Hāthi-Gumphā inscription clearly indicates that both Anga and Magadha were situated side by side with each other on the south bank of the Ganges, which is to say, that the Ganges formed the natural northern boundary of both the countries. Anga lost her independence and became a dependency of the kingdom of Magadha as early as the reign of King Bimbisāra, who is said to have been one of the elder contemporaries of the Buddha. The Hãtbi-Gumpha inscription indicates that Anga continued to be the same old dependency of the kingdom of Magadha as during the reign of King Bahasatimita as during that of King Nanda.
The Vāsukis referred to were not fabulous beings. There would be no reason for mentioning them in a serious document if they were fabulous. The Vāsukis were the Nāga kings who had established principalities in all parts of India. "The prevalence of Nāga rule over a considerable portion of Northern and Central India in the third and fourth centuries A. D., is amply attested by epigraphic evidence." 1 The name of the Chota-Nagpur District situated between Bihar in the north and Orissa in the south may be taken as a reminiscence of the Nāga principalities that existed at one time between Anga-Mayadha, on one side, and Khāravela's Kalinga kingdom, on the other. There are inscriptional records to show that Mathurā and the countries around it were occupied in circa 300 A. D. by the Nāgas. The statement that the Vāsukis sent precious jewels to King Khāra vela, procured for bim the elephants, horses and other animals capable of exciting curiosity, clearly indicates that their principalities comprised those jungle tracts which abounded in the wealth of precious stones, and wild elephants, horses and other animals.
The statement that the king of Pāņdya supplied King Khāravela with many valuable presents of pearls, gems and jewels and various kinds of apparels, requires a word of explanation. One might say this could be explained without having to assume that the kingdom of Pandya was at any time invaded by King Kbāravela; the statement might be justified
1. H. C. Raychaudhuri's Political History of Ancient India, 2nd edition, pp. 809-304.

Page #250

222
OLD BRAHMI INSCRIPTIONS
either (1) on the ground that the kings of Pandya were the less powerful allies of the Aira-Meghavahana kings of Kalinga from earlier times, or (2) on the ground that alarming reports of the irresistible force of Kharavela's victorious arms compelled the then reigning king of Pandya to make an alliance, acknowledging the supremacy of King Mahāvijaya Siri-Kharavela of Kalinga.
As to whether Kharavela had at any time invaded Pandya, we should bear in mind that the fact of sending presents by the king of Pandya to King Kharavela is stated in the record of Kharavela's twelfth regnal year. Whether or no, the Hathi-Gumpha inscription bears testimony to the invasion of Pandya by King Kharavela depends, to a large extent, upon the nature of the reading and rendering of the record of Kharavela's eighth regnal year. The following are the three successive readings and renderings of this record that have been offered by Mr. Jayaswal :
I (4) First Reading: Athame ca vase (manatino ?) (dha ma ni ??) ......Goradhagiri ghatapayita Rajagaha-napam piḍapayati etinam ca kamupadana-panadena sabata sena-vahane vipamumcitam Madhuram apayato [.]
(6) First Rendering: "In the eighth year, the ministers (?)...at Goradhagiri.....having got killed, (he) causes oppression to the King of Rajagṛha who by the report of (Kharavela's offer of marching forward, was made to retire to Mathura, leaving behind everywhere his troops and vehicles."
II (a)-Second Reading: Athame ca vase mahati-senaya mahata(bhittim) Goradhagirim ghatapayita Rajagaham upapīḍāpayati [;] etina ca kamma' padana-panadena sambita-sena-vahinim vipamumcitum Madhurām apayato yeva narid(o) (nama)......(mo?) (yacati) vicha-palava-bhar(e)···[•] (6)-Second Rendering: "In the eighth year, he (Kharavela) having got stormed the Gorathagiri (fortress) of great enclosure by a great army, causes pressure around Rajagṛha (lays siege to Rajagṛha). On account of this report of the acts of valour, (i. e., the capture of Goradhagiri, etc.), the king (so-called) to forsake the invested division of his army, went away to Mathura indeed,.. ...."
III (a) Third Reading:
etian(am or a) ca kammapadanasamnadena samb(i)ta-sena-rahano ripamumcitum Madhuram apayato Yavana-raj(a)-Dimi(ṭa or oti) yachati...vi-palava....... [-]
(6)-Third Rendering
"on
account of the report (uproar) occasioned by the acts of valour (i.e., the capture of the Gorathagiri

Page #251

NOTES
223 fortress and the siege of Rājagļha, etc.) the Greek king Demetrius drawing in his army and transport, retreated to abandon Mathurā."
It will be seen that all the three renderings offered by Mr. Jayaswal show an agreement in representing Gorad hagiri not as a personal name but as the name of a hill-fortiess at Gorathagiri, the modern Barābar range of hills in the District of Gayã. These also show an agreement in represent. ing Rājagaha as the earlier capital of Magadha, and Madhurā as a city on the Yamunā in Northern India. The difference between his first and second renderings is that whereas in the former, accepting the reading Rājagaha-napam pidāpayati......, the king of Rājagsha is represented as the timid warrior who retired to Mathurā, in the latter, accepting the reading Rājagaham upapidāpayati......yeva narido, indefinitely the king so-called is represented as the warrior who went away to Mathurā. And the difference between his third and earlier two renderings is that while in the latter either definitely the king of Rajagrha or indefinitely and sarcastically the king so-called is represented as the warrior who went away to Mathurā, in the former Dimita- Demetrius is represented as the Greek king (Yavani-rājā) who, drawing in his army and transport, instead of advancing towards Räjagrha, retreated to abandon Mathurā.
Before we decide one way or the other, we ought to thresh out the matter carefully. We have nothing to say against Mr. Jayaswal's reading and rendering of the opening words of the eighth year's record up to senāya. There, taken by themselves, clearly indicate that King Kbāravela started on a campaign with a mighty army in the eighth year of his reign, His reading Goradhagirim ghātāpayitā cannot be challenged. This expression may be rendered either as (1) “ baving caused Gorathagiri to be killed," or as (2) “ having caused Gorathagiri to be stormed,” according as we take Goradhagiri to be a personal name or the name of a hill-for.) tress or mountain-fastness.
Whether Goradhagiri is a personal name or the name of a hill-fortress might have been decided at once if the words preceding it could be accurately read. From the faint traces of the letters, Mr. Jayaswal made out mahata-bhitti,“ of great enclosure," as an adjunct of Goradhagiri, keeping in his view Mr. V. H. Jackeon's identification of Goradhagiri or Gorathagiri with the modern Barābar hills. But he has given up at last this reading, declaring it to be uncertain. Mr. Banerji's estampage published in JBORS, 1917, enables us to make out Madhuram a[nupa]to, “ as he reached Mathurā."

Page #252

224
OLD BRAHMI INSCRIPTIONS
We say that the use of Goradhagiri as a personal name in Kharavela's Vinscription is not impossible. For in the Mabāvamsa, Giri occurs as the name of a Nigamtha or Jain contemporary of King Vattagāmani of Ceylon. In one of the ancient Brāhmi inscriptions, Yasogiri occurs as the name of a Buddhist monk. In Jain literature, Mahāgiri and Simhagiri occur as the names of two Jain apostles. And none need be surprised if Mahendragiri occurs as the name of the king of Pişthapura in the Alla habad Pillar inscription of Samudragupta. 3
The use of the verbal expression ghātāpayitā in the sense of “having caused to be killed " is appropriate, if Goradhagiri is a personal name. It may be argued that the use of the same verbal expression in the sense of “having caused to be stormed” is equally appropriate, if Goradhagiri is the name of a hill-fortress or mountain-fastness. We cannot but agree with Mr. Jayaswal when he maintains that the use of the expression Goradhagirim ghātāpayitā, even where Goradhagiri is the name of a hillfortress, is sanctioned by literary usage, and calls our attention to the Pali expression Dīpaghātakā Damilā, “the Tamil devastators of the Island of Larikā."4
The Mahābhārata is the only Indian work where we have the mention of Gorathagiri as a hill or mountain from which one might have a view of Girivraja or Old Rājagặha, the first or most ancient known capital of Magadha. Mr. Jackson is inclined to think that this Gorathagiri is no other than the mountain which Hwen Thsang saw between Pāțaliputra and
1. Lüders' List of Brāhmi inscriptions, No. 601,
2. P. O. Bagchi" On the Pūrvas", Calcutta University Journal of the Department of Letters, Vol. XIV, p. 5.
3. It is only by suggestio falsi that Fleet represents Mahendragiri of Samudragupta's inscription as the name of a hill or mountain. From the system of using the patronymio of a place name as an adjunct of a personal nume which has beeu consistently maintained in the whole list of the kings subdged by the Indian Napoleon, it is easy to understand, As D. R. Bhandarkar points out, that Mahendragiri is a ersonal name, nothing but a personal name. Kausalaka-Mahendra-Māhākāntā, aka-Vyaghrarāja-Kuuraluka. MaņţarăjaPaişthapuraka - Mahendragiri-Kautturaka-Svāmidatta, etu.
4. Attentio might be drawn to a better instance in the Däthāvamsa, III. 16 (JPTS, 1884, p. 126): Kutena ghātā payi dantadhatum," he caused the tooth-relio to be smashed with a hammer." 5. Mahābhārata, II. 19-30.
Goratham girim asadya dadráur Magadhan purap. 6. JBORS, Vol. I, Part II, p. 162.

Page #253

NOTES
225
Gayā, and on the eastern summit of which the " Tathāgata formerly stood for a time beholding the country of Magadha."1
Seeing that the names Goratha and Baithan are both connected with cattle and having regard to the fact that there is no hill near enough
jgir besides the Baithan hill, Mr. J. D. Beglar found reasons to think that the hill named Goratha in the Mahābhārata could not but be the same as the present Baithan hill.2 And with reference to Isigili or Rsigiri, Buddhaghoşa says that the Blessed One was once upon a time seated in that place sitting where he could behold all the five hills (that stood close to one another surrounding Rājagaha). The Isigili-Sutta of the Majjhima-Nikāya records the names of these five hills as the Vebhāra. the Pandava, the Vepulla, the Gijjhakūta and the Isigili. In the opinion of Mr. Oldham, Māgadhapura referred to in the Mabābhārata as the capital of Magadha which could be seen from the summit of the Gorathagiri hill might have been " a town close to the east of the Barābar hills, on the site of what is now known as Ibrahimpur." But the description in the Mabābhārata leaves no room for doubt tbat this Māgadhapura was no other than Girivraja, which precisely like Rājagaba of the Isigili-Sutta, was guarded by the five lofty-peaked and cool-planted hills, the vipula-Vaihāra, the Varāha, the V rşabba, the Rsigiri and the subha-Caityaka. 4 Further, the Sutta-Nipāta-Commentary takes Māgadhapura to mean Rajagaba,
We must carefully note that Gorathagiri, according to the Mahābhārata, was the hill which stood at some distance from Girivraja or Old Rājagặha and could be reached by the travellers journeying from the land of the Kurus through the Kuru-jungle, proceeding to the beautiful lotuslake and getting beyond the Kālakūta... mountain, and likewise going by degrees across the Gaņdaki, the great Sone and the Sadānīrā, the rivers taking their rise from one and the same mountain, crossing the charming
1. Beal's Records of the We.tern World, Vol. II, p. 104, 2. Arch. Sur. Reports, 1872-73, Vol. VIII, pp. 46-47.
3. Papasca-Sudani, Siamese edition, Part III, p. 536 : yasmim thane nisinnānam pañca pabhatā paññāyanti tattha. 4. Mahābhārata, II, 21-23 :
Vaihāro vipulah sailo Väräho Vrgabhas tatha | Tathā Rşigiris tābha śubhāś Caityaka-pañcamāh || Ete pañca mahāśrngāḥ parvatāḥ śitala-drumā)
Rakşantivabhisamhatyasamhatángã Girivrajam || 5. Para mattha-Jotikā, II, p. 584 ; Magadham puranti Magadhapuram Rajagahanti adhippāyo.
29

Page #254

226
OLD BRAHMI INSCRIPTIONS
Sarayu and beholding the Eastern Kosala, going beyond Mithila, crossing the rivers Mala and Carmanvati, crossing the Ganges and the Sone, and proceeding further eastward to get from Kusaciracchada to what was called Magadha-kṣetra, the Magadha-territory (literally, "the Magadha-field ").1
There can be little doubt, as Mr. Jackson has sought to prove on the strength of two short Brahmi inscriptions, that Gorathagiri was but one of the two ancient names of the Barabar hills, the other being Khalatikapavata, which latter is met with in two of the Barabar Hill-Cave inscriptions of Asoka and in Patanjali's Mahabhāṣya (I. 2. 2). As the ancient capital of Magadha was known at the same time by two names, Girivraia and Rajagṛha, so, Mr. Jackson thinks, the Barabar hills were known at one time by two names, Gorathagiri and Khalatika-pavata, while, later on, they came to be known by the name of Pravaragiri, Pravara wherefrom the modern name Barabar was apparently derived.
The two Brahmi inscriptions, relied upon by Mr. Jackson in proposing his identification, are engraved on rocks in two different places, both of which are not far from the well-known caves dedicated by Asoka to the Ajivikas," and consist each of five syllabic letters, one of them recording the name of the hill as Gorathagiri and the other as Goradhagiri, the letterforms of the former bearing a close resemblance to those of the dedicatory inscriptions of Asoka, and the letter-forms and spelling of the latter to those of the Hathi-Gumpta inscription of Kharavela. On palæographic grounds, Mr. R. D. Banerji assigns the former to the Asokan age and places the latter a century later, going so far as to believe that the latter was actually engraved by one of the men who accompanied King Kharavela and took part in His Majesty's first invasion of Magadha.5 If it be as Mr. Jackson presumes, that the earlier inscription recording the name of the hill on which it was engraved as Gorathagiri was of the same age as two of the dedicatory inscriptions of Asoka recording the name of the hill on which they were engraved as Khalatika-pavata, we do not see any necessity for
1. Mahabharata, II, 20, 26-29, quoted with Rāmavatar Sarma's translation in JBORS; Vol. I, Part II, p. 161.
2. JBORS, Vol. I, Part II, p. 169.
3. The Lomasa Rishi cave is called Pravaragiriguha in an undated Sanskrit inscription belonging probably to the 7th century A. D.. JBORS, Vol. I, Part II, p. 169.
4. JBORS, Vol. I, Part II, p. 162.
5. JBORS, 1917, Vol. III, Part IV, p. 500. See Plates in JBORS, Vol. I, Part II, published by Jackson.

Page #255

NOTES
227
speculating if Khalatika was a phonetic transformation of Goratha, or Goratha that of Khalatika. In such a case as this, one is compelled to think that neither Khalatika-pavata nor Gorathagiri was intended by Asoka to serve as a name for the whole of the present Barābar range but as dames for two separate hills of this range.
The close likeness between the Brāhmi letter-forms and the spelling Goradhagiri as found in the second inscription and as found in the Hāthi-Gumphā inscription is, beyond a doubt, a strong point in the argument pressed by Messrs. Banerji and Jayaswal for taking Goradhagiri of Khāravela's inscription to denote the Barābar hills. And yet in justifying the use of the verbal expression ghātā payitā in the sense of “having stormed," one has to prove that Gorathagiri or Goradhagiri was at all, at any time, a hill-fortress. It is certain that “the Mahābhārata which mentions the Gorathagiri in connexion with the earlier period of Rājagrha, does not refer to any fortifications there." “ Inside the Barābar enclosure stone foundations of buildings can be seen in various places, notably apon and to the east of tbe artificially raised area at the foot of the Siddheswarpath hill due north of the caves, and in the jungle towards the southwestern end of the valley. The whole enclosure is too small to bave been the site of any large town, but the natural strength of the position and the fact that the defences are strengthened at all vulnerable places by stone walls render it reasonable to suppose that it formed a refuge used in times of danger by the people who ordinarily lived in the plains and valleys outside." These reflections and the fact that the local tradition speaks of the hills as those which in ancient days formed a stronghold or fortress enable Mr. Jackson to opine “ that as Pataliputra was open to an attack from the south, the Gorathagiri which lay on the way to the capital was made a fortified position," and that " it must have been treated as the first line of defence on the southern side of Pāķaliputra." 1
We need not attach to the local tradition quoted by Mr. Jackson any more importance than what we should attach to his inference from a study of the ancient remains. The question still is whether the stone-enclosure and other remains are the relics of a rocky stronghold or hill-fortress or those of a rocky shrine, stūpa or temple. Evidently, the Barābar hills were a sanctified rather than a fortified position. For it is here that King Asoka dedicated certain caves for the residence of the Ājivikas. It is again here that Hwen Thsang saw a Buddhist stūpa on the eastern summit of the
1. Quoted by Jayaswal in JBORS, 1917, Vol. III, Part IV, pp. 469, 472.

Page #256

228
OLD BRAHMI INSCRIPTIONS
mountain where the Tathagata formerly stood beholding the country of Magadha. The Chinese pilgrim's mountain with the stupa on its summit is connected with "the country of Magadha." The Mahabharata connects Gorathagiri with Magadha-kṣetra or Magadha-territory. And we find that the Sutta-Nipata-Commentary locates Pasāṇaka-cetiya, the site of the rocky shrine, in Magadha-khetta (Magadha-khette Pāsāṇaka-cetiyam),1 and says that in spite of the fact that it became, during the dispensation of the Blessed One, a Buddhist retreat, the spot where there was formerly a shrine of Deva-worship continued to be designated by its ancient name as Pasānaka-cetiya or Paṣaṇa-caitya." The identification thus sought to be established between Gorathagiri and Paṣaṇa-caitya may receive its confirmation also from the fact that just as in the Mahabharata, Gorathagiri has been mentioned as the monntain standing on which the travellers coming from the Kuru-country to Mithila and finally getting into the Magadha-territory by a southward journey through Kusacfracchada, could have a view of Girivraja or Old Rajagṛha, so in the Sutta-Nipata and its commentary, Pasaga-caitya has found mention as a Buddhist retreat, which could be reached from Rajagaha by the travellers starting from Kapilavatthu along the great high road, reaching Pataliputta from Vesali by a southward journey and proceeding further southward, along the same high road, via Nalanda. Thus it may be shown that all the available evidences conspire only to prove the contrary of Mr. Jackson's inference that Gorathagiri was a rocky stronghold or hill-fortress. If Gorathagiri was a hill noted for its rocky shrine, stupa or Deva-temple, it is difficult to justify the use of the verbal expression ghātāpayita in the sense of " having stormed." Why should Kharavela storm Gorathagiri if it was not used as a fortress or stronghold of the Magadhan army?
Strangely enough, the Hathi-Gumpha expression following Goradhagirim ghatapayita is one involving a reference to Rajagaha. These are the three possible readings of this expression: (1) Rajagaha-napam piḍāpayati, "caused the king of Rajagaha to be harassed; " (2) Rajagahamupapīḍāpayati, "caused pressure around Rajagaha; or (3) Rajagahanam papīḍāpayiti, " caused a heavy pressure to be brought to bear upon the people of Rajagaha." The first reading does not suit the context in view of the
99
1. Paramattha-Jotikā, Vol. II, p. 583.
2. Paramattha-Jotika, Vol. II, p. 584: Pasānakam cetiyan ti mahato päsanassa upari pubbe devātṭhānam ahosi, uppanne pana Bhagavati viharo jato so ten' eva purima-vorhäena Pārāṇakam cetiyan ti vuccati,

Page #257

NOTES
229
fact that in the twelfth year's record, the king of Magadha is alluded to as "the king of Magadha " (Magadha-rājā), and not as "the king of Rajagaha," and no less for the reason that Rajagaha ceased long ago to be the capital of Magadha. The second reading, too, cannot justify the use of etinam, a pronominal expression with a genitive plural case-ending, immediately after upapīḍāpayati. The third reading alone fulfils all the requirements. There were at least two Rajagahas or Rajagṛhas in India, one which was the earlier capital of Magadha and the other which was the capital of the Kekayas "located in the Ramayana on the other side of the Vipasa (Beas) and stretching up to Gandhara."1 The latter Rajagṛha also bore a second name, Girivraja. By the people of Rajagṛha we may understand as well the Magadhas as the Kekayas. And there were Kekayas in the north-west, in Uttarapatha, as well as in the south, in the Dakṣiņāpatha.
"C
The Kekayas of Uttarapatha were settled in the Pañjab." It is evident from the Rāmāyaṇa that their territory "lay beyond the Vipasā (Beas) and abutted on the Gandharva or Gandhara Visaya."2 And it may be surmised from a number of inscriptions discovered on the ruins of the Jagayyapeta Stupa in the Krishna District and belonging, according to some of the archæologists, to so early a date as the second century A.D., that the Kekayas of Dakṣiuapatha who were matrimonially connected with the Ikṣvākus of South India were "probably a ruling family of ancient Mysore."3
These are then the three possible interpretations of the Hathi-Gumphā record of Kharavela's eighth regnal year: (1) That His Majesty having stormed Gorathigiri, a stronghold of the Magadhan army in the Barabar hills, caused a heavy pressure to be brought to bear upon the citizens of Rajagaha, the earlier capital of Magadha; (2) that His Majesty having caused Gorathagiri, the king or general of Mathura in Uttarapatha, brought a heavy pressure upon the Kekayas whose capital was Rajagṛha beyond the Beas; or (3) that His Majesty having caused Gora thagiri, the king or general of Mathura, the capital of Pandya in Dakṣinapatha, brought a heavy pressure upon the Kekayas who founded a territory in Ancient Mysore.
1. Canningham's Ancient Geography of India edited by S. N. Majumdar, Notes. 2. Rāmāyaṇa, II. 68. 19-22; VII, 113, 114, Raychaudhuri's Political History of Ancient India, 2nd edition, p. 36.
3. Dubreuil's Ancient History of the Deccan, pp. 88, 101; Raychaudhuri's Political History of Ancient India, p. 315, fn. 1.

Page #258

230
OLD BRAHMI INSCRIPTIONS
Each of these interpretations has its strong and weak points. The strong points of the first interpretation are obvious: (1) the contiguity of the kingdom of Magadha to that of Kalinga; (2) the discovery of two Brāhmi inscriptions as old as the Barābar Hill Cave inscriptions of Asoka and the Hathi-Gumphā inscription of Khāravela enabling us to identify Gorathagiri with the Pravaragiri or Barā bar hills; (3) the exploration of the lingering remains of an ancient enclosure; (1) the local tradition asserting that there was a fortress or stronghold in the Barābar hills; and (5) the Mahābhārata describing the Gorathagiri of Māgadhakşetra as the hill or mountain from which one could have a view of Girivraja or Old Rājagsba, the earlier capital of Magadha. Its weak points are: (1) the Mahābhārata describing Gorathagiri simply as a bill or mountain, and not as a stronghold or fortress; (2) the stone-enclosure and other supposed remains of an ancient fortress being explained also as vestiges of an ancient sbrine, temple or stūpa ; (3) it being inexplicable why, if Pătaliputra were at that time the capital of Magadha and Gorathagiri had served as its first line of defence on the south, Khāravela bad directed his attacks towards Rājagraba, unless it be presumed that the capital had to be removed from Pātaliputra to Rajagpha in fear of an attack from the north as formerly done during the reign of King Munda or that the army of Magadha having bien gradually driven southward, was finally concentrated in Råjagrha.
The strong point of the second interpretation is that it enables us to understand how Khāravela could venture to carry his expeditions into the very heart of Uttarāpatha after having made himself the master of Mathorā even before he planved his attacks on Angu and Magadha. The weak point of it is that there is no other authority but the Ramayana to establish that there was a Rājagļha in the Uttarāpatha as distinguished from Rajagļha, the earlier capital of Magadba.
The strong point of the third interpretation is that it enables us to understand how the king of Pāņdya was compelled to send valuable presents to Khāravela, which is to say, to ackaowledge the supremacy of the king of Kalinga. Its weak point is that there is yet no evidence to prove the identity of the Kekayas with the Rājagahikas. Another point in its favour is that if it could be established, we might have clearly seen that Khāravela turned his attention to Northern India only after he had subdued the powers in the Deccan.
But weighing the matter carefully, we feel that the balance of probablity lies, after all, with the first interpretation, and, to some extent, with the second interpretation. And it follows from either of these two interpretations

Page #259

NOTES
231
that Khāravela had to effect a timely retreat to Mathura to relieve his army terrified by the alarming reports of counter-attacks from the inhabitants of Rājagaba, while his twelfth year's record goes to show that he made a second attempt, which proved entirely successful.
12. KHĀRAVELA'S PERSONAL HISTORY
Many interesting facts concerning the personal history of Khāravela can be gleaned from the old Brāhmi inscriptions, particularly his own inscription in the Hathi-Gumphā, so much so that the Hāthi-Gumphā inscription itself may be judged as a Khāravela-Carita “ The Life of Khāravela," in Indian epigraphy, and no less as an Asokávadāna, “The Edifying Legend of Asoka," in Jain records. The text of the Hathi-Gumphā may be judged also as a kind of mahā. tīra-carita, “ the annal of a great hero," taking Khāravela to be the great warrior bero. There can be little doubt that the composer of the Hāthi-Gumphā text has sought all along to extol Khāravela as a mighty earthly hero who was destined to conquer, to rule, to protect and to please. Anyhow, no other inference can be drawn from the honorific epithets adorning the name of His Graceful Majesty Kbāravela.
In the Hathi-Gumphá inscription (I. 1)., Khāravela is described as pasatha-subha-lakhaņa, “one who bore the noble and auspicious bodily marks," and as caturumta-rakhaņa-guna-upeta, “one who was endowed with the qualities of a ruler capable of protecting the whole of this earth extending as far as the four seas." A less likely alternative of the reading of the second epithet may be caturamta-luthana-gunauprta, "one who was endowed with the qualities of a warrior capable of undertaking expeditions over the wbole of this earth bounded by the four seas." The expression caturantā' was the current old Indian idiom to denote indefinitely the whole extension of the earth conceived as an island in the seas or oceans.
Corresponding to caturamta-rakhana-guna-upeta or caturamtaluthana-guna-upeta, we have the familiar Pāli expression cāturanta-vijitāvi. janapadathāvariyappatta, “ the ruler of the whole earth bounded by the four seas, the upholder of the realm by the right of conquest and the
1. Here caturantā is the same expression as caturantā mahi (Artha-Sastra, III. 2.50), and asamudra-ksiti.

Page #260

232
OLD BRAHMI INSCRIPTIONS
consolidator of his hold on his territories," which is an oft-recurring epithet of a king overlord (räjä cakkavatti). Buddhaghosa explains caturanta as meaning "the lord of the earth bordering on the four seas and comprising the four island-like continents." He explains vijitāvi as meaning "one who has quelled the rebellious agitations within, overpowered the inimical rebels without and conquered all other kings."2 And lastly, he explains janapadathavariyappatta as meaning "one who has established so sure and permanent a hold on his territories that no one is able to move it an inch, or having retained a permanent hold on his territories, remains engaged in his duties unworried, unshaken and unmoved."8
of
That pasathasubha-lakhana and the other epithet are meant in the Hathi-Gumpha text to represent Kharavela as a king overlord is beyond dispute. For in the inscription of his chief queen (No. II), Kharavela is freely represented as Kalinga-cakavati, "the King Overlord Kalinga. But this is not enough to bring out the real significance of the two epithets. As used in the first paragraph of the HathiGumpha text, the epithets signify what the expert astrologers, palmists and diviners (lakkhana- paṭiggahaka, nakkhatta-pāṭhakā), after reading the bodily marks and making a thorough study of the birth-star and other factors and signs connected with the birth and person of the childprince Kharavela, declared him to be in future.1
Let us now see how Kharavela is represented in the HathiGumpha text and, to some extent, in some of the remaining old Brahmt inscriptions to have fulfilled other conditions of a person destined to become a king overlord, the conditions specified in such ancient texts as the Pāli Cakkavatti-Sihanada-Sutta in the DighaNikaya, and Rajavagga in the Anguttara-Nikaya, and the Lalita-Vistara (Ch. VIII).
1. Papañca-Sudani, Siamese edition, Part III, Brahmiyusutta-vanņaua; "caturantaya issaroti caturanto, catusamuddantaya catubbidhadipa-bhūsitāya ca pathaviya issaroti
attho."
2. Ibid: "Ajjhattam kopadi paccatthike bahiddha ca sabbarajāno vijetiti vijitāvī." 3. Ibid: "Janapade dhuvabhavam thavarabhavam patto, na sakka kenaci caletum janapade va tamhi thavariyappatto anussuko sakammani rato acalako asampabedhi."
4. Cf. Nidana-Katha in Fausböll's Jataka, Vol. I, p. 56, where the Lakkhanapaṭiggähaka Brahmin astrologers and diviners are said to have declared with regard to the future of Prince Siddhartha: "Imehi lakkhanehi samannagato agaram ajjhävasa. mano rajā hoti cakkavatti pabbafjamano Buddho' ti," "If one endowed with these marks choose to keep to household life, one is destined to be a king overlord, and choosing to renounce worldly life, one is destined to be an Enlightened Master."

Page #261

NOTES
293
- The nobility of origin is one of the primary conditions. Regard. ing the nobility of Khāravela's origin, we read in the Hathi-Gumpha inscription (1) that he was Cetarājavamsa-vadhana, “the increaser of the royal House of King Ceta” (I. 1); (2) that he was rājisivamsakula-vinisita, " emanated from the family and line of royal sages” (I. 16); and (3) that he was installed as a great king in the third generation of two kings belonging to the reigning dynasty of Kalinga (tatiye Kalimgārājaramse purisa-yuge mahārājábhisecanam pāpunāté).
Prof. Sten Konow prefers, in agreement with Mr. Jayaswal and Prof. F. W. Thomas, to read Cetirājaramsa-vadhana. We fully appreciate the force of his argument that the reading Ceti (Cedi) is quite in keeping with the adjective rājusivamsakula-vinisita whereby Khāravela claimed to have descended from an ancient family of royal sages. In adhering to the reading Ceta, our own point is that it is equally in keeping with the very same adjective, and does not at the same time make itself open to dispute from the palæographic point of view.
In the Vessantara-Jātaka (Fausböll, No. 547), Ceta occurs as an alternative spelling of Cetiya, Cetaratthn being the same kingdom as Cetiyarattha of the Cetiya-Jätaka (Fausböll, No. 422). In the saine Jataka, the ruling people of the Ceta or Cetiya Kingdom are represented as Cetiyā, Celā and Cetaputtā, which conclusively proves that Ceta was an alternative spelling of Ceti, and that the Cetas or Cetis as a ruling people were known as the descendants of King Ceta. The information thus derived from the Vessan.' tara-Jätaka is doubly significant inasmuch as the references are all found in the verse-quotations. The prose narrative wbich is peculiar to the commentary.version of the Birth-story contains an additional information concerning a land-route connecting Jetuttaranagara, the capital of the Sivi kingdom, with the Brahmin village Dunniviţtha, situated in the kingdom of Kalinga. We are told that the mountain named Suvannagiritāla was at a distance of five Yojanas from Jetuttaranagara, that the river named Kontimārā was at a distance of five yojanas from the Suvannagiritāla mountain, that the mountain named Arañjaragiri was at a distance of five yojanas from the Kontimārā river, that the Brahmin village named Dunni. vittha, situated in the kingdom of Kalinga, was at a distance of five yojanas, that the capital of the Ceta kingdom was at a distance of ten yojanas from the Dunnivițţha Brahmin village, and that the total distance
1. Hātbi.Gunnphá inscription, I. 1.
30

Page #262

234
OLD BRAHMI INSCRIPTIONS
between Jetiittaranagara and the capital of the Ceta or Catiya kingdom covered thirty yojanas.1
The Vedabbha-Jātaka (Fausböll, No. 48) points to the existence of a bigh road (mahāmagga) from Benares to Cetiyarattha, which was not safe in all parts, " being infested with roving bands of marauders.”
The Cetiya-Jataka (Fausböll, No. 422) represents Upacara or Uparicara as the king of Ceti (Cetiya-rājā) whose genealogy could be traced back to Mahāsammata, the first kaown king of the Solar line, who reigned in the kingdom of Ceti (Cetiyaratthe), in the city of Sotthivati, and who was gifted with four supernormal faculties. He was known also by the name of Apacara. His five sons are said to have built after his death five cities in five directions, to wit, Hatthipura in the east, Assa pura in the south, Sihapura in the west, Uttarapancāla in the north, and Daddarapura in the north-west. And the great Brahmin named Kapila is said to have been his family priest.
In referring to the above Jätaka, the Milinda-Pað ha mentions the name of the king as Suraparicara, and says that he reigned in the land of the Cetis (Cetīsu).
Dr. H. C. Raychaudhuri seems to be perfectly justified in identifying the Ceti-king Upacara, or Suraparicara with Vasu Uparicara, who is mentioned in the Mahābhārata (I. 63. 1-2) as the Paurava king of Cedi. And there is probably nothing to object to Prof. Rapson's identification of the Cedi-king Vasu Uparicara with Kasu Caidya, the king of the land of the Cedis, who is pra ised in the Rg-Veda (VIII. 5. 37-39), in a hymn containing dānastuti at its end. Dr. Raychaudhuri rightly observes that in the Mahābhārata account, too, the five sons of the Cedi-king are said to have founded the five different imperial lines of kings, 4 Bịhadratha among the Magadha people, Pratyagraha, Kuśāmba the Maņivābava, Māvella and the unconquerable Yadu. The tradition in the Rāmāyana (I. 32-6-9) "'associates the scions of Vasu's family with the cities of Kausāmbr, Mahodaya and Girivraja."
All these are legendary arcounts of Ceta and the Cetis or Cedis. But these are not without their bearing on the Hathi-Gumphā text. These enable us not only to grasp the significance of Khāravela's epithet
1. Fausböll's Jặtaka, Vol, VI, pp. 514, 521. 2. Milinda-Papiha, Trenckner's edition, p. 202. 3.4. H.C. Raychaudhuri's Political History of Ancient India, 2nd edition, p. 82.

Page #263

NOTES
235
Cetarājavamsa-vadhana or Cetirājaramsa-vadhana but also to appreciate why he has been praised as rājisivamsakula-vinisita.
The third statement as to Khāravela being installed as a mighty king in the third generation of two kings belonging to the then reigning dynasty of Kalinga, implies that be ascended the throne of Kalinga by the lawful right of succession, which is to say, that he was in no sense a usurper of the royal power. Differences of opinion are bound to be as to the correct interpretation of the phrase tatiye Kalimgarājavamse purisa-yuge. Mr. Jayaswal explains it as meaning "in the third dynasty of the Aira line of the kings of Kalinga," while with Dr. R. C. Majumdar it means “in the third generation of the Kalinga kings.” Both the interpretations have obviously missed the technical sense of the expression purisa-yuga. As for the ordinal tatiya meaning " the third," there is no doubt that it qualifies purisa-yuga. For in the Hāthi-Gumphā record of Khāravela's tenth regnal year (I. 11), we come across the expression Kalimgarājavamsānam tatiyayugasagāvasāne which, as a slightly different manipulation of the phrase taliye Kalimgarājavamse purisa-yuge, leaves no room for doubt that the numerical adjective tatiya was intended to be applied to yuga or purisayuga. Here the plural form of Kalim garājavamsa might seem to bear out Mr. Jayaswal's interpretation. But we must not forget that the word purisa is understood, and that the expression Kalimgarājavam sānam may be better interpreted as meaning "of those of the royal dynasty of Kalinga."
The yuga, saga (sarga) and vamsa are three of the main subjectmatters of a Hinda Parāņa. The term tatiya-yuga presupposes the Indian tradition of caturyuga, “the four yugus," each of the yugas implying, according to Hindu cosmogony, a distinct “age" in the development of the world-system, the term "age” being used in the same sense as in “ the Golden Age," " the Iron Age," and so forth, Thus it may be shown that the Hindu sense of ynga or of caturyuga is primarily cosmogonic.
The term saga or sarga signifies, according to Hindu cosmogony, a stage or landmark in, or a sectional presentation of, the creative evolution of the cosmos. It is precisely in the sense of sectional presentation or chapter-division that the term sarga has been used in both
1. Sridharasvāmi in his Țikā on the Vişnu-Parāņa, I. 2. 66, explains sarga as a synonym of sreţi : sargad ikartta ai esamürttih sretyä dirūpah cf. the terms Brahmasargo, dovasarga, bhūtasarga, and the like used in the Vişnu.Purāņa, I. 5. 18-24.

Page #264

OLD BRAHMİ INSCRIPTIONS
the Mahābhārata and the Rāmāyaṇa, it being conveniently rendered in English as “canto." Several sargas or cantos go to make a book called parva or kānda. Thus it may be shown that we are justified in interpreting the term saga in yugasaga in the sense of a "part," “portion " or "period” of a yuga.
Keeping the tradition of cosmogonic caturyuga as a presupposition, the Buddha or the Buddhist conceived four yugas ((cattāri yugāni) of Ariyapuggalas, “those of the Aryan lineage," i which is the same as to say, the Buddha or the Buddhist conceived four purisa-yugas (cattāri purisa-yugāni) in Ariyavınsa, "the Aryan lineage,"2 understanding the term yuga in the sense of yugga (couple) or yamaka (twin).3 Thus with the Buddhists the four yugas or purisa-yugas denote the four couples of Aryan personalities (purisa-puggalas) representing the eight notable stages in the progress of the Buddhist pilgrim towards Arahatship, which is his final destination. A notion of sequence or succession is implied in the Buddhist enumeration of four yugas or purisa-yugas as vathama (the first), dutiya (the second), tatiya (the third) and catuttha (the fourth). But each yuga or purisa-yuga considered by itself, eliminates altogether the notion of sequence or succession, for a yuga, to be worth the name, requires as a sine qnd non the co-existence of two persons, one representing, as the Buddhist put it, the stage of inception (maggattha), and the other that of fruition (phalattha). *
Corresponding to yuga or purisa-yuga in Pali, we have the use of yuga or purisa-ynga in the Hāthi-Gumphā text. And the expression tatiya-yuga or latiya purisa-yuga suggests the same kind of gradational enumeration as that of four yugas or purisa-yugas in Pāli. If so, there is no other alternative but to interpret the expression in the Hāthi-Gumphā text in the sense of the third couple of royal personages, one representing the fifth king and the other the sixth king of one and the same reigning dynasty of Kalinga. This is precisely the sense sought to be conveyed by the rendering " the third generation of two kings."
1. Ratana-Sutta in the Sutta-Nipāta and the Khuddaka-Pāgha, verse :
Ye puggalā archasata-pasatthā | Cattari etani yugāni honti || 2 Mahspartnibbāna-Suttanta, Digha-Nikāya, Vol. II : Cattāri purisa-yagāni atthapurisa-puggalā.
3. Cf. Páli sävaka-yuga. The chapter which bears the name of Yamakavagga in the Pali Dhammapada 's entitled Yugavarga in the Sanskrit Udanavarga.
4. Paramattha-Jotikā, Khuddaka-Patha-Commentary : Sotā pattimaggaţthophalaţtho'ti ekam yugam, evam täva Arahattamagghytho-phalaţtho'ti ekekam vuganti cattári yugāni honti.

Page #265

NOTES
237
Does it imply a conjoint rule of two kings of the same royal family reigning at the same time, and if so, in what sepse? In upholding the interpretation offered by Prof. D. R. Bhandarkar and Prof. R. C. Majumdar of the text of the Andhau inscription as implying a conjoint rule of King Chaştana, the grandfather, and King Rudradāman I, the grandson, of the same Kşaharāta royal family, Dr. H. C. Raychandhuri calls our attention to a number of facts deserving consideration : (1) the account given by Diodorus of the political constitution of Tauala (Patala), the Indus Delta, as having been drawn on the lines of the Spartan, enjoining the conjoint rule of two kings representing the two eldest representatives of the ruling clan and as vesting the command in war in two hereditary kings of different houses; (2) the mention of dvirāja in the Atharva-Veda (V. 20. 9) in the sense of a conjoint rule of two; (3) the danger of dvairājya, the conjoint rule of two kings, in the event of their disagreement and mutual enmity and hostility, discussed in Artha-Sastra (VIII. 2. 128);' (4) the system of dorajja (dvairājya), referred to in the Jaina Ayāramya-Sutta ; (5) the case of King Dhrtarāstra, the father, and King Duryodhana, the son, reigning together, to be cited from the Mahābhārata; (6) the case of Eukratides and his son reigning together to be cited from Justin's work ; and (7) tbe conjoint rule of Strato I and Strato II or that of Azes and Azilises, to be cited among other instances. 2
The overwhelming evidence thus produced goes to prove that there is no inherent improbability of a conjoint rule of two kings in each generation of the then reigning dynasty of Kalinga being implied in the two Hathi-Gumghā expressions : (') tatiye Kalimgarāja-ramse purisa-yuge, and (2) Kalimgarājavamsānnm tatiyayugasagåvasāne. But
1. N. N. Law in his well-informed article on “Technical Institutions” (Indian Historical Quarterly), maintains that dvairājya or “ the rule by two kings " was, according to the Artha-gåetra, a vyasana or " distress" of the royal state, it implying rather an abnormal than a normal state of things. The dvatrājya form of government must have been ushered in as a means of avoiding keeping the crown-prince waiting indefinitely till the death or retirement of the reigning king.
2. Political History of Ancient India, 2nd edition, p. 308. Attention has also been drawn to an anecdote in the Mahāvastu (III, p. 432), in which three sons of King Mahendra, the three uterine brothers, are said to have conjointly reigned in Simhapura, the capital of Kalinga. But attention might also be drawn to the Buddhist tradition of pine Nanda brothers, the nine kings of the pre-Mauryan Nanda dynasty, reigning conjointly in Magadha..

Page #266

238
OLD BRAHMI INSCRIPTIONS
to be sure about it, as well as to form an idea of the personal relation of the two reigning kings in each generation in the aforesaid sense, we must carefully take the Lalita-Vistara use of the term purusa-yuga into our consideration.
The Lalita-Vistara mentions the following characteristics, among others, of an ideal royal family fit to be graced by the birth of the Buddha:
Puruşayugasampannam ca tatkulam bhavati. Purvapuruşayugasampannam ca tatkulam bhavati. Abhijatapuruṣayugasampannam ca tatkulam bhavati. Abhilakṣitapuruşayugasampannam ca tatkulam bhavati.
Mahesakhyapuruṣayugasampannam ca tatkulam bhavati.
Rājā Rajendra Lala Mitra translates these :
"Such a family is......remarkable for having two generations (living at the same time). It has two predecessors, as well as two successors (living at the same time). It has two special generations living, and having a member of the name of Mahesa in two successive generations."
1. Literally, a coupling of.
We offer the following as a correct rendering of them :
"Such a family is noted for having throughout a connecting link between two successive representative men. It has in its previous generation a connecting link between two successive representative men. It has in its present generation a connecting link between two successive representative men. It has also in its coming generation a connecting link between two successive representative men. It has throughout a connecting link of two greatly powerful successive representative men."
The idea of an uninterrupted continuity of the royal line from father to son is contemplated in the Lalita-Vistara passage concerning purusayuga. In the life-time of the grandfather and father, the son and grandson are to be in full vigour of life ensuring the prospect of the birth of the great-grandson and great-great-grandson. Considered in this light, if the grandfather and father represent the first purusa-yuga, the son and grandson represent the second, and the great-grandson and great-greatgrandson represent the third. And if, as is recorded in the HathiGumpha text, the third purusa-yuga of the then reigning dynasty of Kalinga was completed with the birth of Prince Kharavela, it follows that Prince Kharavela was the great-great-grandson of the first king of this

Page #267

NOTES
239
dynasty, and that the part of the third couple (tatiya-yuga-saga) could be over only with the death of his father. Keeping consistency with this meaning of puruşa-yuga, as used in the Hāthi-Gumphā text, we might say that Khāravela's father remained joined as a king with his grandfather when he had been discharging the administrative functions as the crownprince for nine years from his sixteenth to his twenty-fourth year; that on the death of his grandfather be himself became joined with his father as a king as soon as he completed his twenty-fourth year; that his father died in the eleventh year of his reign in the record whereof we are told that the part of the third couple was over by that time and he paid proper homage to the memory of the former kings of Kalinga, which is to say, he performed the srādh ceremony; and that Kadampa-Kudepa came to be joined with him as a king after the death of his fatber in the eleventh year of his reign. If the cave standing in the name of Kadampa-Kudepa was one of the 117 caves excavated in the thirteenth year of Khāravela's reign, it is evident from the epithets, Aira, Mahārāja, Mahāmeghavāhana and Kalimngáilhipati, adorning the name of Kadampa-Kudepa in the old Brāhmt inscription No. III, that Kadampa-Kudepa as the son of Khāravela was then joined with him as a king of Kalinga in the fullest sense of the term.
Like the nobility of origin and ancestral line, the brightness, perfection and dignity of the bodily form and appearance is a primary condition to be fulfilled by a prince destined to be a king overlord. The Indian popular feeling on this point bas found a pithy expression in the following stanzas quoted in English translation from the Ulūka-Jātaka (Fausböll, No. 270):
“The owl is king, you say, o'er all bird-kind :
With your permission, may I speak my mind ?” “I like not (with all deference be it said)
To have the owl anointed as our Head. Look at his face! if this good humour be, What will he do when he looks angrily."
1. The meaning made out is this: A and B represent the first puruşa-yuga; after the death of A, the first yuga comes to be partly over and C comes to be joined as a king with B; after the death of B, C and D combine to represent the second purusa-yuga; after the death of C, the second yuga comes to be partly over and E comes to be joined as a king with D; after the death of D, E and F combine to represent the third puruşa-yuga; after the death of E, the third yuga comes to be partly over (tatiyayugasagavasāna) and G comes to be joined as a king with F.

Page #268

240
OLD BRAHMI INSCRIPTIONS
The Mugapakkha-Jataka (Fausböll, No. 588) bears testimony to the fact that bodily infirmity or deformity was considered an unbearable disgrace to a royal family. The Aśokan legends, as found in the Divyava. dana and the Mahāvamsa-Tikā, bring out the fact that the ugliness of appearance stood greatly in the way of Asoka when he was still a prince. The description in the Hathi-Gumpha text (I. 1) goes to prove that Khāravela as a prince had the very best bodily form "glowing with graceful majesty, so lovely as to captivate the heart of Grace herself, a veritable god Visga in the human garb." So much is implied indeed in the adjective siri-kaḍāra-sarīravā.
Proficiency in polite learning is another primary condition. Here the term 'polite learning' involves the study and practice of various useful sciences and arts. Proficiency in sciences implies the sound theoretical knowledge of the principles and details of the systems, and proficiency in arts implies the intelligent and skilful use or application of those principles and details. As to proficiency in polite learning in the above sense, the pronouncement of the Artha-Sastra (I. 5. 2) is:---
a
Vidya vinīto rājā hi prajānām vinaye rataḥ\ Ananyam prthivim bhunkte sarvabhutahite rataḥ ||
"The king who is well brought up in the discipline of sciences and arts, and engaged in bringing up his subjects in discipline, enjoys the earth without sharing it with others, doing good to all living beings under his rule."
The Artha-Sastra (1.5.2) prescribes the following courses of study for prince: (1) a prince should learn the alphabet and practise writing (lipi), and should learn counting and arithmetic (samkhyāna) from the ceremony of tonsure to that of investiture with sacred thread, that is to say, from the third or fifth year to the ninth or tenth year; (2) from the ninth or tenth year to the sixteenth year, a prince should study the Brahmanical treatises based upon the Three Vedas (trayi) and the systems of speculative philosophy (Anvikṣaki) under the teachers of acknowleged authority, should acquire the knowledge of the science of wealth (Vārtā) under the superintendents of various departments, and should acquire the knowledge of the science of government (Dandanīti) under those who are adepts in theories as well as in practical application; anl (3) from the sixteenth year onwards, a prince should spend the forenoon in receiving lessons in military tactics concerning the proper handling of troops and of weapons, and the afternoon in hearing the Purāṇa, the Itivṛtta, the Akhyāyikā, the

Page #269

NOTES
241
Udaharana, the Dharma-gāstra, and the Artha-śāstra, all of which go by the name of Itihāsa,
To the same effect we read in the Milindapañ ba (Trenckner's edition, p. 178) :
“Mahiyā rājaputtānam hatthi-assa-ratha-dhanu-tharu-lekha-muddasikkha-khattamanta-suti-muti-yujjha-yujjhāpana-kiriya-karaṇīyā."
“ The princes of the earth are to learn the arts of writing and counting and of handling the weapons and troops, and are to put into practice the principles of polity, Srali, Smrti and the sciences of war and warfare."
This is but a rough and ready way of enumerating in one breath the list of sciences and arts, which the Indian princes were required to learn, and make judicious and skilful use of. We may take it for granted that the list in contemplation of the author of the Milindapañha is virtually the same as that in the Artha-Šāstra. The Milindapanha furnishes us with a list of nineteen sciences and arts in all, in which King Milinda-Menander, its ideal Indo-Bactrian Greek prince, gained high proficiency. We read : “ Many were the arts and sciences he knew-holy tradition and secular law; the Sankhya, Yoga, Nyāya, and Vaišeşika systems of philosophy, arithmetic ; musio; medicine; the four Vedas; the Purāṇas, and the Itihāsas; astronomy, magic, causation, and spells ; the art of war; poetry ; conveyancing—in a word, the whole nineteen." 1
Mabākṣatrapa Rudradāman I is represented in his Jundgari inscription (circa 150 A.D.) as a prince who “gained fame by studying grammar (sabda), polity (artha), music (gāndharva), logic (nyāya), etc."
As to Prince Khāravela's education and ability, the Hāthi-Gumpha text (1.1) represents him (1) as one who was lekha-rūpa-gananā-vavahāra. vidhi-visārada, and (2) as one who was savavijáradāta. It will be a mis. take to suppose that the second adjective, savavijávadāta, "one whose self was purified by proficiency in all Indian polite learning," has been used in apposition with the first. The term savavijā ((sarva.vidyā), the whole of Indian polite learning, is meant to include lekha, rūpa, and the rest enumerated in the body of the first adjective but not to be exhausted by them. There are two very strong arguments against taking sava-vijā as limited or exhausted by lekha, rūpa and the rest. First, we find
1. This is Rhys Davids' rendering of the Milinda text (pp. 3.4) which reads: bahuni c'assa satthāni uggahitāni honti, seyyathidam ; suti sammuti sankhyā yoga niti visesikā ganika gandhabba tikiccha cătubbeda purāņā itihäsä jotina māyā hetu mantana yudddha chandasă mudda, vacanena ekūnavisati,
31

Page #270

242
OLD BRAHMI INSCRIPTIONS
that the Hathi-Gumpha text (I. 4) has praised him as Gamdhavavedabudha, "one who was versed in the science of music-the Gandharvalore." This goes at once to show that sava-vijā of Khāravela's inscription includes the science of music which is not mentioned in the first adjective.
Secondly, the fact that King Kharavela ventured, in the very second year of his reign, to defy so powerful a rival as King Satakarni in triumphantly marching with all the four divisions of his army amply attests that he excelled, even while he was yet a prince, in the art of war and warfare (yujjha-yujjhāpana-kiriyā), which is to say, that sava-rija in Kharavela's inscription is meant also to include yuddha-vidya. The same inference may be drawn from the many acts of valour recorded in the Hathi-Gumpha inscription.
Nevertheless, the expression sava-vijā, as employed in Kharavela's inscription, suffers from vagueness and indefiniteness. What was precisely the traditional total of vidya (sciences and arts) prescribed for the educa tion and training of Indian princes in the days of Kharavela we cannot say. The Milinda pañha (circa first century A.D.) mentions the total as nineteen (vacanena ekunavisati), while the Nidana-katha of the Pali Jataka-com. mentary (Fausböll, Jātaka, I. p. 58) speaks of twelve (dvadasavidham sippam), including archery (dhanuggaha).
The Vatsyâyana Kama-Sutra enumerates the ancient Indian sciences and arts called yogas under sixty-four heads (catuḥṣasthika yoga), implying that by the time the Sutra was compiled in the extant form (circa 3rd or 4th century A. D.), the traditional total came to be reckoned at sixtyfour. This total, once established, continued to be in use and gained a proverbial character in the later Hindu expression catuḥsaṣṭi-kala. The Kāma-Sutra enumerates, as pointed out by Raja Rajendra Lala Mitra,1 the sixty-four yogas as sciences and arts to be learnt and practised by "the young maidens aspiring for the position of court-ladies or for that of expert courtezans, either alone or in the company of their tutors, fellow students, friends of the same age, etc." Strangely enough, Sridhara Svami in his commentary on the Bhagavata-Purana, represents the Yadu princes Baladeva and Vasudeva as learning the sixty-four science and arts.
Although references to all or most of the sciences and arts can be traced in such an ancient Buddhist work as the Digha-Nikaya, Brahmajala-Sutta,
1. Rajendra Lala Mitra's translation of the Lalita-Vistara, pp. 186-8,

Page #271

NOTES
243 it is difficult to conceive the total sixty-four as coming into existence much before the third or fourth century A. D.
The Hāthi-Gumphi description of Prince Kbāravela's proficiency in I earning, lekha-Tüpa-gananā-ravahāra-ridhi-visārada and savavijávadāta, seems to correspond almost to a description of Prince Siddhartha's education in the Lalita-Vistara (Ch. X), which runs as follows :
Šāstrāņi yāni pracaranti manuşyaloke Sa khẽ lips ca gamam ăni ca dhātutantra | Ye silpayoga pythu laukika aprameyāḥ 1 Tegveşu śikṣitu purā bahukalpakotyaḥ 11
The parallel thus quoted from the Lalita-Vistara enables us to interpret the Hathi-Gumphâ expression savarijā (sarva-vidyā) as implying “whatever subjects of study are current in the world of men, the various multitudinous sciences and arts that are of use to the generality of mankind,”1 that is to say, " the whole of the Indian polite learning prevalent at that time."
Of the four enumerated subjects of study, samkhyā, lipi, gananā and dhālutantra, the first two are met with in the Artha-Šāstra (1. 5,2); and the second and the fourth are met with in the Milindapañha (p. 178) in the form of lekha and muddā. The Lalita-Vistara and the Artha-Šāstra have used the term samkhyā or samkhyāna in the simple meaning of counting the numbers. All the three texts bave used the term lipi or lekha in the simple meaning of the knowledge and writing of alphabet or alphabets. The term dhālutantra, as used in the Lalita-Vistara, does not seem to convey the idea of the knowledge of verbal roots, the meaning in which Rājá Rājendra Lāla Mitra has interpreted it. The term dhātutantra appears to have been used just as another expression for muddā (mudrā) in the Milindapanha. Buddhaghoşa has explained muddā as meaning hatthamudda-ganınā, which may be rendered as "counting the totals on one's fingers." The word hatthamuldā is generally taken to mean a seal. ring.' But none need be surprised if in hatthamuddā (hastamudrā) we trace a word similar to kilamudrā in the ancient Kharoşthi documents, and if
1. For the enumeration of the science and arts, see the Lalita-Vistara, Oh, XII, pp. 156-157 (Lefmann's edition).
2. The Lalita Vistara itself substitutes mudrā for dhātutantra in a second enumeration in Ch. XII, p. 166 (Lefmann's edition): lipi-mudra-ganana-samkkyā.

Page #272

244
OLD BRAHMI INSCRIPTIONS
the same was a current name for ' a punch-marked coin.' If so, the dhātu. tantra was no other than a mudrāśāstra dealing with the rules of calculation applied to monetary transactions,-to transactions by metallic media of exchange.
We think that the enumerated four subjects of study may be conveniently reduced to three to make them correspond to three in the HäthiGumpbā text, the term lipi corresponding to lekha, the term dhātutantra or mudrā to rūpa, and the terms samkhyā and gañanā to gañanā.
Mr. Jayaswal is in the right to suggest that the three terms lekha, rūpa and gañanā, as used in the Hathi-Gumphā text, were intended to have a deeper significance than what they generally implied in popular usage. The term lekha was not used to mean simply the knowledge of the alphabet and the practice of alphabet-writing. The learning and writing of alphabet has been proscribed in the Artha-Šāstra as a course of study for a beginner, for a prince of three or five years of age. Lekha in the sense of mere knowledge and writing of alphabet is evidently inconsistent with the adjective lekhavisārada, representing Prince Khāravela as an expert in the art of writing' in the Hāthi-Gumphā record giving an account of the nine years spent by Khāravela as a crown-prince, from his fifteenth to his twenty-fourth year. The Häthi-Gumphā inscription says that Khāravela passed the first fifteen years of his life just playing the games befitting his young age. But we shall misinterpret this statement to assume that Prince Khāravela commenced to learn ka, kha, ga just after he completed his fifteenth year and not before. The statement goes rather to show that he commenced his career as a crown-prince when he passed as “an expert in all matters relating to the art of writing,". The statement as to his spending the first fifteen years of his life in princely games has no meaning except as implying that he spent these years uomindful of and without being called to the responsibility of administration.
This may suffice to justify us in interpreting the term lekha in the Hāthi-Gumphā text in the same wider and deeper sense as lekha or sāsana (royal writs) in the Artha-Sāstra (II. 9-28).
Similarly we are not to take rūpa as a simple term for the counting of the totals of stamped coins but in the wider and deeper sense of all matters relating to coinage and currency, all transactions in which the medium of exchange is a factor, more or less in the same sense rūpa in the Artha-Šāstra.
In the same way we are not to take gañanā as a simple tarm for counting or calculation but in the wider and deeper sense of all matters

Page #273

NOTES
245
relating to accounting, more or less in the same sense as gananā in Asoka's R. E. III and in the Artha-Šāstra (II. 7. 25).
Over and above lekha, rūpa ang gananā, we have the use of two other words, to wit, vavahāra and vidhi. In the compound lekha-rūpa.gañanāvavahāra-vidhi, vidhi may be either combined with each of the preceding words and interpreted in the sense of "rule," lekha-vidhi, “the rule of writiny," rīparidhi, “the rule of coinage and currency," etc., or treated as a separate term per se. We have tentatively translated vavahāra-vidhi by "administration and procedures," which is somewhat vague and misleading. The Sanskrit term ryavahāra corresponding to darahāra has bee clearly defined in the Artha-Sa-tra (III. l. 58) as vyavahārika-śāstra, "judicial, administration and procedures in accordance with established cou Ventions." Treatiny ridhi as a separate term, Mr. Jayaswal has sought to interpret it in the sense of dharma-śāstra. There is no inberent improbability of tbis sense of ridhi. Tbe term vidhi has been used in the Artba-Sāstra in the sense of kriyā-vidhi, “the rule of action." But tanahāra, too, is just a “ rule of action," the difference between the two being that while vidhi implies state-action in accordance with the established laws of human conduct and duty, vavahāra implies stateaction in accordance with established conventions. In the two enumerations of four things in the Artha-Sāstra (III. 1. 58), caritra has been replaced by samsthā or dharmasāstra, and rājašāsana by nyāya or danda. It is quite possible that vidhi in the Hāthi-Gumphà text is just a synonym forge niyama (Artha-Sastra, I. 5. 2), or caritra or samsthā or dharmaśāstra (ArthaSāstra, III. 1. 58).
Varahāra of Khāravela's inscription is obviously the same word as vivohāla of A soka's P. E. IV, in which viyohala stauds in contradistinction to hamda : viyohāla-samatā ca damda-samatā. We fully agree with Prof. D. R. Bhandarkar in interpreting viyohāla-samatā in the sense of "uniformity of procedures" but differ from him, as well as from Prof. Bühler, both of whom take viyohāla to be a synonym of abhihāla ( Pāli abhihāra). Prof. Bühler seems, however, to be right in interpreting the Asokan expression abhihāle rā damớe vā as signifying " in the awards of reward or punishment” on the authority of the Sambhava-Jātaka (Fausböll, No. 515) where abhihāra is paraphrased by pūjā. We prefer to take damda of damda-samatā as an equivalent of nyāya or rāja-śāsana of the Arthasästra, to interpret damớa-samatā in the sease of " uniformity of decisions," and to explain abhihāla and damda as meaning respectively “ decisions for” and “decisions against."

Page #274

240
OLD BRAHMI INSCRIPTIONS
There is nothing in the Hathi-Gumpha inscription to indicate that Prince Kharavela was sent out of Kalinga for his education to such a place as Taxila, which was famous as an ancient seat of learning. On the other hand, the recorded facts go to show that he spent the first twentyfour years of his life in Kalinga. In all probability he was placed during the first fifteen years of his life under an experienced tutor,-a vṛddha in the language of the Artha-Sastra, just in the same way that Prince Siddhartha was placed, according to the Lalita-Vistara, under a tutor named Viśvāmitra, and the Yadu princes Balarama and Vasudeva were placed, according to the Bhagavata-purana, under a tutor named Sandipani. It also seems probable that, while a crown-prince, he received a practical training in the art of administration at the hands of the high functionaries in charge of various departments, and acquired the knowledge of the systems of religion and philosophy at the hands of the saintly and far-famed ascetic and recluse teachers in Kalinga.
The numerical strength of family members, relations and retinues is another primary condition to be fulfilled by a prince destined to be a king overlord. On this point we are supplied with some information in the Hathi-Gumpha record of Kharavela's thirteenth regnal year where we read that his queens, sons, brothers, relatives and the royal servants cooperated with him in excavating one hundred and seventeen caves on Kumārī-pavata, the Kumari hill, to provide resident Arhats (Jain saints) with accommodations and shelters for resting their bodies.
The old Brahmi inscription No. II records the upper cave of the Mañcapuri group as an excavation of Kharavela's chief queen and a dedication to the recluses in Kalinga. Thus this inscription furnishes us with a corroborative evidence of the truth of the Hathi-Gumph record of Khāravela's thirteenth regnal year. The expression "the recluses in Kalinga" may be viewed as explanatory of the Hathi-Gumpha expression "resident Arhats."
The very fact that she has been honoured as the chief queen (agamahisi) of His Graceful Majesty Kharavela, the king overlord of Kalinga, goes to prove that King Kharavela had queens more than one, that his chief queen was not the only queen, and that she was just the queen consort, the chief one among the queens, although we cannot definitely say among how many queens in all. Although this is a fact that Kharavela had queens more than one, we do not know altogether how many caves were donated by his queens. Asoka's "Queen's Edict" introducing Kaluvāki as his second queen (dutiya devi) enables us to say that King

Page #275

NOTES
247
Asoka had at least two queens of whom Kāluvāki was the second in rank. The old Brahmi inscription No. II, containing a reference to Kbāravela's chief queen, leaves the question as to the number of his queens entirely open.
Some of the caves on the Kumāri hill were excavated by Khāravela's sons in the thirteenth year of his reign, and King Kharavela became thirty-seven years old in his thirteenth regnal year. If he had married in the sixteenth year of his life and a son was born in that very year, his eldest son could not be older than twenty-one years when the caves were excavated. He had sons more than one, but altogether how many sons he then bad, and how many by each of his queens we cannot say. In the Mañcapuri group of three caves, the cave in the upper storey was an excavation of Khāravela's chief queen, the corresponding cave in the lower storey was an excavation of King Kadampa-kudepa, and the side cave in the lower storey is said to have been an excavation of Prince VadukhaVarikha. In the inscriptions of King Kadampa-Kudepa (No. III) and Prince Vadukha-Varikha (No. IV) their personal relationship with King Khāravela is not at all indicated. But the grouping of the three caves in the manner just described seems to indicate that even in the works of art and architecture the mother has sought to be combined with her two sons, of whom King Kadampa-Kudepa was the elder and Prince VaqukhaVarikha the younger.
The brothers of King Khāravela have been referred to as excavators of some of the caves in the record of his thirteenth regnal year. No caves have, as yet, been discovered with inscriptions recording their names. Had there been found out any such inscriptions, as princes of the same royal house, their names would have been recorded each as a “KumāraPrince," precisely as in the case of Prince Vaukha-Varikha.
As regards Khāravela's kinsfolk among the excavators of the caves we have no definite information from the old Brāhmi inscriptions. Two of these inscriptions (Nos. VI-VII) record the name of one Cūlakamma (Ksudrakarma, Junior Karma) as the excavator of two separate caves. The inscription No. VIII records the name of Kamma (Karma) as the gentleman who donated the chambers of one of the caves, and the name of Khiņā (Kșivā) as the lady who donated the frontal building construction.
1. R. D.'Banerji observes : "A minor inscription mentions a king of Kalinga named Kūdepa-siri who may be a descendant or a successor of Khāravela as he bears the same titles as Kbäravela" (JBORS, Vol. III, Pt. IV, p. 605).

Page #276

www. kobatirth.org
248
OLD BRAHMI INSCRIPTIONS
The association of the two names in one and the same inscription as well as in one and the same work of art and architecture is significant, and it may be taken to indicate that the gentleman was the husband and that the lady was his wife. It may be inferred from the two games, Kamma and Cūlakamma, that the gentlemen known by these two names were related to each other as brothers, of whom Kamma was the elder and Culakamma the younger. It seems probable that Kamma, Cūlakamma and Khiņā were counted among Khāravela's kinsfolk.
With regard to the royal officers and servants among the excavators of the caves, we have one inscription (No. IX), in which a gentleman named Bhūti has been distinctly mentioned as a Town-judge (Nagara-akhadamsa),--a designation corresponding to Nagalaka-mahāmāta or Nagalavi. yohalika-mahāmāta in Agoka's S.R.E.I., and to Nagaraka in the Artha-Sastra (11. 35. 56); one inscription (No. X), in which Nākiya of Bāriyā has been represented as a High-functionary with ministerial duties (Mahāmada); and one inscription (No. XIV), in which the donor Kusuma has been described as a Padamilika. Kusuma, as his inscription shows, donated more caves than one (leņāni).
It is snggested that Pādamūlika may be regarded as a local patrony. mic signifving a man belonging to a locality named Padamüla. But we must draw the reader's attention to the Asadisa-Jataka (Fansböll, No. 181) in which the Rāja-pādamülikıs (“Servers of the royal feet ") figure as the attendants who were in close touch with the king. Prof. Cowell and Mr. Ronse in their English translation have rendered Rāja-pādamülikā as “slaves," and we, following Mr. R. D. Banerji, have rendered Pūdamulika of the inscription as “Menial." Now it would seem that the Rāja-pāda. mülikas or Padamūlikawere not slaves or menials of an ordinary kind. For, according to the Buddhist Birth-story, they were the persons who slandered Prince Peerless to the king, big brother. In Mr. Francis' rendering of the Sarabhanga-Jätaka (Fausböll, No. 52?), the Raja-pādamülikas stand forth as "king's attendants." This Birth-story relates that on his return home from Takkasilā as a master of archery, Sarabhanga, the coval chaplain's son, was appointed a Rāja-pālamūlika bv the king of Benares in compliance with the request of his father. He daily attended on the king (upatthahi) and daily received a thousand pieces of money, a much higher honorarium than “ the king's attendants" could ordinarily expect. On the display of his skill in archery, he was soon promoted to the post of commander-in-chief, a fact which clearly proves that the king's attendant's rank was a lower status than that of a commander-in-chief,

Page #277

NOTES
249
The inscription No. X is capable of an interpretation, aocording to which Mahāmada, Bāriyā and Nākiya may appear as joint donors of one and the same cave. Should this be accepted as the correct interpretation, it may be inferred from the order of the three names that Mabāmada, the father, Bāriyā, the mother, and Nākiya, the son, were associated in one and the same work of art and architecture.
In the inscription of Khāravela's chief queen (No. II), Her Majesty has been represented as rājino Lālákasa Hathisā(ī) ha-sampanātasa dhuta, 1 " the daughter of the high-souled king Hastisāha or Hastisimba of rising glory.” In this reading and rendering, it is difficult to ascertain, whether Lāláka is a royal epithet derived from the name of the kingdom of which Hastisāba-Hastisimha was the king or it is just a title of praise similar to Yasalālaka in the name Yasalālaka-Tissa of a king of Ceylon mentioned in the Mahāvamsa. If it be a royal epithet derived from the name of the kingdom, it is easy to understand, first that Hastisāba- Hastisimha was the king of Lāla, and, secondly, that the royal families of Kalinga and Lāla were united by a matrimonial alliance. In accordance with the location suggested in the Mahāvamsa, Lāla or Lāla was a kingdom situated between Kalinga and Magadha, 1 in which case Lāļa or Lāļa cannot but be identified with Ladha or Ra lha.? Accepting the other interpretation of Lāldka, we fail altogether to understand of what kingdom Hastisāha-Hastisimba was the king. Whatever the correct interpretation, it is certain that Khāravela's chief queen was a princess born of a distinguished royal family.
The possession of a superb state-elephant is one of the tests for determining the status of a king overlord. The royal style Mahāmeghavāhana adorning the names of Khåravela and Kadampa-Kudepa goes to show that) the state-elephant of the kings of the royal family of which they were the descendants was known by the name of Mahāmegha, " the Gread Cloud.” Mahāmeghavāhana is also an epithet of Indra, the king of the gods. This royal epithet may accordingly be taken to imply that Khāravela and other kings of the Mahāmeghavāhana family were very powerful, each of them bearing comparison with Mahendra. Every Indian king was
1. Mahāvamsa, Ch. VI.
2. For the arguments in favour of the identification of Lāla or Lāla with lata in Western India, see S. K. Chatterji's “The Origin and Development of the Bengali Language, Vol. I, p. 72, f.n.
32

Page #278

250
OLD BRAHMI INSCRIPTIONS
regarded as an earthly representative of Indra or Mahendra. This is corro. borated by the royal title Imdarāja occurring in the concluding paragraph of the Hāthi-Gumphā text. The reading Bhikhurāja suggested by Mr. Jayaswal and others is out of the question. The elephants of Arga and Kalinga have been praised in the Artha-Sastra (II. 2.20) as those of the noblest breed. The Kurudbamma and Vessantara Jātakas (Fausböll, Nos. 276, 547) bear testimony to the fact that a kind of religious sanctity was attached by the peoples of India to state-elephants.
The possession of enormous wealth in the shape of a large amount of ready money, vast stores of food-stuffs, precious stones, rich apparels, horses, elephants and other live-stocks is a test of the high fortune and prosperity of a king overlord. Fortunately, the Hāthi-Gumpha text is not lacking in information on all these points.
First, as to ready money and solvency of His Majesty's government, we find that King Khāravela possessed a sufficiently large amount to be in a position to spend 35,00,000 pieces, in the very first year of his reign, to repair the capital of Kalinga (I. 2); to spend 1,00,000 pieces, in his fifth regnal year, to bring the canal near the Tanasuliya or Tanasuli road, into the capital (I. 6); to remit within the kingdom of Kalinya all taxes and duties in his sixth regnal year, the taxes and duties amounting to many hundred thousand pieces (I. 7); to spend some hundred thousand pieces, in his seventh regnal year, to organise a hundred kinds of pompous parade and to perform all ceremonies of victory (1. 8); to spend some hundred thousand pieces, in his eighth regnal year, to feast all sections of the people in Mathurā as well as in Kalinga (I. 9); to spend 38,00,000 pieces, in his ninth regnal year, to erect the "Great-victory Palace' (I. 10); to spend 1,00,000 pieces, in his tenth regnal year, to pay due homage to the memory of the former kings of Kalinga (I. 11); to spend 75,00,000 pieces, in his fourteenth regnal year, to erect certain religious edifices (I, 15); to carry out, in his eleventh regnal year, the costly work of reclamation of Pithuda-Pithu laga, which had become converted into a watery jungle of grass ([. 12); to excavate, in his thirteenth regnal year, as many as 117 caves on the Kumāri hill (I. 14); and, last but not the least, to finance, in bis second, fourth, eighth and twelfth regnal years, the expensive undertakings of military expeditions all over India (I. 3, 1. 5, I. 9, I. 13).
In this connexion, three points of importance deserve consideration : (1) that in the Hāthi-Gumphā records of Khāra vela's regral years, just the amounts, 35,00,000, 38,00,000 and 75,00,000, are mentioned without any indication as to what sort of money, Kāršāpaņa, Suvarņa or Satamānu,

Page #279

NOTES
251
was current at that time in Kalinga; (2) that the treasury of the government inherited by King Khāravela was full of ready money to enable him to spend 35,00,000 pieces, in the very first year of his reign, for repairing the capital city; and (3) that the annual income of His Majesty's government from the taxes and duties collected from his subjects in the kingdom of Kalinga, from the inhabitants of the towns and districts (pora-jānapada,),1 amounted to many hundred thousand pieces. Even leaving a good margin for hyperboles and exaggerations in the specified figures, one cannot but gather this impression from the Hāthi-Gumphā record, that Kbāravela was a fabulously rich king or that Kalinga was a prosperous kingdom under his rule.
Secondly, as to food-stuffs, we find that King Khāravela possessed vast stores to be in a position to sumptuously feast, in his ninth regnal year, all sections of the community—the religieux of Brahmanical and non-Brahmanical orders, the ascetics and householders, the officials and non-officials once in Mathurā and subsequently in Kalinga (I. 9); and to arrange for similar feasts on several other occasions, such as at the time when he was formally installed in the throne (I. 1), when he organized festiviti-s and merry gatherings (1. 1), when he performed all ceremonies of victory (I. 7), when he paid homage to the memory of his predecessors (I. 11), and, lastly, when he dedicated the caves and other religious edifices (I. 14, I. 15). King Asoka in his R. E. I, says that formerly many hundred thousands of living beings were daily slaughtered in his kitchen for dainty dishes, while subsequently the number was reduced to three, two peafowls and one deer. Although he has expressed in it a pious wish to stop even the daily slaugher of three living beings, there is nothing in his inscriptions to indicate that he was an advocate of vegetarian diet. What is clear from his edicts, especially R.E. I, is that he undervalued sacrificial slaughter of life, killin, living creatures in the name of religion. The Hāthi-Gumphā text is altogether silent on this point. If King Khāravela were a scrupulous Jain layman, it might have been expected from him! that he would strictly observe vegetarian practice.
1. Jayagwal has taken infinite pains to establish that in contexts, such as that of the Häthi-Gumphi inscription, where the terms pora and Janapada are used in singular num. ber, they are intended to denote two representative bodies of citizens and peoples. See, for a lengthy and detailed criticism of such an interpretation, N. N, Law's instructive paper-" The Janapada and the Paura," in the Indian Historical Quarterly, Vol. II, Nos. 2.3.

Page #280

252
OLD BRAHMI INSCRIPTIONS
Thirdly, as to precious stones and the rest, we read in the Hathi. Gumphā text that King Khāravela received abundant supplies of pearls, gems, jewels and various kinds of apparels as tribute from the then reign ing king of Pāṇdya whose kingdom was noted for those products and received as well precious stones and horses, elephants and antelopes as presents from a hundred Vasukis (I. 13). The Vidyalhara-abode was apparently another territory wherefrom the precious metals were collected for filling the royal storehouse with the treasures of value (1.5). It may be inferred from the occurrence of such expressions as veļuriya ... Mahā. vijaya pāsāda (l. 10), veďuriya.gabha (1. 15) and veduriya-nila-vochimna (I. 15), that the precious stones treasured up in Kháravela's royal storehouse belonged to these two classes, viz., (1) the beryl (if lapis lazuli is not the correct rendering of veduriya), and (2) the emerald (nila). Lastly, from the Hathi-Gumphā record of Khāravela's fourteenth regnal year, it is evident that the kingdom of Kalinga abounded in the best quarries of stone extending over many leagues (varākarn-samuthāpitā aneka-yojanaāhitā silā, I. 15). We do not know, as yet, where these stone-quarries were to be found, whether in the Tosali division of his Kalinya kinydom or elsewhere. It is clearly brought out in the record of Khāravela's fourteenth regnal year that the stone-slabs quarried out of those quarries were best available materials for making stone-pillars, shrines and shrineposts.
The numerical strength and equipment of the army constitute a good test for determining the status of a king overlord. As to the numerical strength of Khāravela's army, we read in the Hāthi-Gumphā record of his eighth regnal year (1.9) that His Majesty marched out with a mighty army (mahati-senā) enabling him to bring a terrible pressure to bear upon the people of Rājagaha. The second year's record (I. 3) says that the Kalinga army of King Khāravela consisted of a multitudinous troop of the horses, the war-elephants, the foot soldiers and the chariots (haya-gaja. nara-radha-bahulam damdam), that is to say, of the traditional four divisions of an Indian army. The fact that King Khāravela was able to undertake, in the very second year of his reign, such a campaign towards the western quarter in defiance of so powerful a rival as King Sālakarņi, distinctly proves that his fighting army was, more or less, the fighting army of his predecessors, the kings of the Aira-Meghavābana dynasty.
Senā or army is a general term for denoting the fighting strength of a king. The horse (haya), the elephant (gaja), the footman (nara) and the chariot (radha) are the four terms to represent the four divisions of an

Page #281

www. kobatirth.org
NOTES
253
army, while senā and vāhana (troops and conveyance) are the two terms to distinguish the fighting warriors and soldiers from horses, elephants and chariots considered as vehicles and conveyances. We have the use of all these technical terms in Khāravela's inscripion.
As to the equipment of Khāravela's army, we find that, in the concluding paragraph of the Hāthi-Gumpbā inscription (I. 16), His Majesty has been represented as apatihata-caka-vāhana-bala, a mole of praise, which we have rather wrongly rendered (ante, p. 47) as a strong with undaunted carriers of the realm of royal command," suggesting in the footnote that by these carriers were intended to be meant " the ministers and other royal officers." We now detect that such an interpretation as this is not borne out by the expression actually used. Ca ka-vāhana is evidently the same expression as sena-vāhana, which occurs in the eighth year's record (I. 9). Can there be any doubt that by sena-vāhana, the drafter of the inscription intended to signify the troops and transports? In the Culla-Kalinga-Jātaka, King Kālinga of Kalinga, precisely as Kaliņgadhipati Khāravela in the Hāthi-Gumphà inscription, has been described as a warrior who was in a position to march out for conquest with a mighty army (mahatiyā senāya), equipped that he was with multitudinous troops and transports (sampanna-bala-vāhana). If this reasoning be at all sound, it clearly follows that undaunted' or 'invincible' (apatihata) was the character of Khāravela's fighting army:
There is nothing distinctly on record to indicate whether, when and how King Khāravela increased the number and fighting equipment of his army. It may be easily inferred, however, from the eighth year's record (I. 9) that the troops and transports with which he attacked the people of Rājagaha did not suffice to withstand the fear of counter-attacks. He must bave sufficiently reinforced bis fighting army and increased its equipment before he marched out again iu the twelfth year to produce consternation among the rulers of Uttarapatha, as well as to subdue Bahasatimita, the then reigning king of Magadha (I. 13).
Thus it may be shown that the Kalinga army of King Kbāravela was sufficiently well-equipped and enormously large. And yet the fact remains that the Hathi-Gumpbā inscription does not supply us with the actual figures relating to Khāravela's troops and transports. A tolerably clear idea of the numerical strength of his fighting army may, however, be formed from a few collateral evidences.
We know, for instance, “from the earlier account of Megasthenes (Indika, Frag. I. 6) that the king of the Kalingas was protected by a

Page #282

254
OLD BRAHMÌ INSCRIPTIONS
standing army, numbering 60,000 infantry, 1,000 cavalry, and 700 war elephants.” Prof. Radhakumud Mookerji rightly suggests that “this army must have been considerably expanded by the time of Asoka when the number of casualties alone is stated to be at least 4 lacs (taking the number of those who ultimately succumbed to the wounds of war to be, say, 3 lacs).”I
Prof. Mookerji's reference is evidently Asoka's R. E. XIII containing, as it does, an account of the heavy casualties suffered by the king. dom of Kalinga in the aggressive war waged by the Maurya emperor in the eighth year of his reign. This account goes to prove that the fighting army of Kalinga in Asoka's time could afford to suffer the losses of 150,000 men as deportees, of 100,000 men as those killed in action, and of " many times as many * men as 'those who died of wounds received in the fight.'2
Here the expression “many times as many" is vague and indefinite, and guilty, no doubt, " of an exaggeration.” The following seems to be a reasonable estimate of the fighting army of Kalinga, which has been recently suggested by Prof. Mookerji : “ If the number of those (who died of wounds received in the fight) be taken to be at least thrice that of the killed, the total number of casualties would be 4 lacs, and adding to this the number of the deportees, the number of the army that fought on the battle-field would be at least 5 lacs.:'3
Having regard to the fact that in the case of Asoka's Kalinga war, the army of Kalinga fought in defence against a foreign invasion, and that in the case of Kbāravelu's campaigns, the army of Kalinga marched out to produce a marked impression all over India, it may be safely presumed that the total pomber of the standing army of Kalinga during Khāravela's reign was by far the greater and by no means less than 51 (or 37 ?) lacs. In accordance with Plutarch's statement (Life of Alexander, Ch. XII), “Androkottos (Chandragupta Maurya) was able to overrun and subdue the whole of India with an army of 600,000.” If it was possible
1. Asoka, p. 16.
2. If Asoka's statement be taken to refer to the casualties suffered by two armies, the army of Kalióga and the army of Magadha (which is not likely), the number of the Kalinga army in Asoka's time must have been much less than 5 lacs, the number fixed by Radhakumud Mookerji
3. Asoka, p. 162, f. n. 3. It seems that Asoka's expression bahutāvamtake may be interpreted, with reference to the just preceding figure of 1 lac, also as meaning "as many as that (1 lao)," it which case the total will be 37 lacs (17 +1+1).

Page #283

NOTES
255
for Chandragupta Maurya to overrun and subdue the whole of India with an army of 6 lacs, there is no reason why it would be impossible for Khāravela to undertake military campaigns all over India with an army numbering thereabout.
It cannot be supposed that King Khāra vela marched out with the whole of the standing army of Kalinga without leaving a fraction of it for the defence of his kingdom in his absence. This fraction must be added to the number of the units with wbich he marched out in order to determine the total number of the standing army of Kalinga during his reign. In the case of Chandragupta Maurya, we find that he proceeded to conquer the whole of India with an army of 6 lacs, while the standing Maurya army during his reign consisted of 700,000 men in round numbers, 600,000 infantry, 30,000 cavalry, 9,000 elephants attended by 36,000 men, and 8,000 chariots managed by 24,000 men. It is impossible to thivk that the total of the standing army of Kalinya during Khāravela's reign exceeded 6 lacs.
The idea of militia was not, perhaps, altogether absent. The account given by King Agoka of his Kalinga war tends to create an impression in favour of the opinion that the conquest effected by the Maurya conqueror proved ultimately to be a defeat to the people of Kalinga. If the general people of Kaliöga had not somehow taken part in the battle, there is no reason why King Asoka would feelingly dwell upon the suffering caused to the civilian population by " violence or slaughter or separation from their loved ones."
In the Hathi-Gumphā record of Kbāravela's eighth regnal year (I. 9), ve read that he effected a timely retreat to Mathurā in order to relieve the troops and trausports of his army terrified by the uproar of counter-attacks from the people of Rājagaha. If our interpretation of this particular record be correct, it clearly follows that the citizens of Rajagaba took up arms in defence of their city against a foreign invasion. The twelfth year's record (I. 13) shows that His Majesty's conquest of Anga-Magadha was not completed until he was able to force into submission the inhabitants of these two countries (Amga-Magadhavūsinam ca pāde vamdāpayatı) after subduing King Bahasatimita. It is evident from many of the records in the Hathi-Gumphā text that in undertaking military campaigns all over India, King Khāravela tried by all possible means to evoke the patriotic sentiments among his subjects, the inhabitants of the kingdom of Kalinga in general and the citizens of his capital in particular. The record of his seventh regnal year (I. 8) says that he caused a hundred kinds of pompous

Page #284

256
OLD BRAHMI INSCRIPTIONS
parade of swords, umbrellas, flags, guards and horses, and all ceremonies of victory to be performed. The eighth year's record (I. 9) says that he sumptuously feasted all sections of the people once in Mathurā and subsequently in Kalinga, and organised triumphal processions as a means, no doubt, of impressing the idea of victory on the minds of the people. The erection of a new royal palace known by the name of Jahā-vijaya-pāsāda, “the Great-victory Palace,” the assuming of the self-conferred title Mahā-vijaya, “the Great Conqueror," the bringing back by a triumphal procession from Anga-Magadha to Kalinga of the Kalinga Throne of Jina which was carried off by King Nanda as a trophy, the receiving of tributes and valuable presents from the king of Pāņdya, as well as from a hundred Väsukis, the entertaining of the citizens of the capital of Kalinga with feasts, festivities and musical performances, the remitting of taxes and duties, the adorning of the capital with new roads, squares, gate-bars and towers --- all helped him to keep the people always in excitement, and induce them, as we may say, to join the army to fight for the glory of their country.
It seems that Mr. Jayaswal' and Prof. Radhakumud Mookerji have tried the impossible in endeavouring to infer the total of the population of Kalinga from the total number of its standing army. Whether assuming with Goltz that "every 15th soul of the population can take up arms in defence against a foreign invasion," or slightly altering with Prof. Mookerji, the proportion of its fighting strength to its total population from six per cent. to eight per cent. one counts 75 or 60 lacs as a reasonable figure of the population of Kalinga in Asoka's time against the present population of 50 lacs, we must treat it as nothing but an ingenuity without proofs. To infer the total of the population from the total of the standing army of a country, no matter whether it is Kalinga or any other land, is to forget that history is neither logic nor mathematics. Anyhow, we may assure Prof. Mookerji that Khāravela's inscription keeps us entirely in the dark about the population of Kalinga.
The soundness of administrative policy and method is another test for determining the status of a king overlord. The Hāthi-Gumphā inscription bears a clear testimony to the fact that it was a declared policy of King Khāravela to govern his kingdom in accordance with established customs
1. JBORS., Vol. III, Part 1V, p. 440. 2, Asoka, p. 162, f, n. 3,

Page #285

NOTES
257
and not departing from the traditional methods of his forefathers. In order that his subjects might bave no misgiving on this point, he did not forget to remind them of the fact that whatever he did, he did in consonance with the noble tradition of the former kings of Kalinga. For instance, the fourth year's record of his reign (I. 5) says that he governed the Vidyād hara country in accordance with the principle and usage of the former kings of Kalinga, everywhere by the highest kind of law (Vijādharddhi-vāsam... Kalimga-puvarājānam dhamena va nitinā va pasāsayati savata dhamakūtena). Similarly in the tenth year's record (I. 11), we read that he caused proper homage and honour to be paid to the memory of the former kings of Kalinga, as a public demonstration, no doubt, of his loyalty to the tradition of his royal predecessors (Ralimga-purarājānam yasa-sakāram kārāpayati). The eleventh year's record (I, 12) says that he reclaimed and rehabilitated Pfthudaka, founded by the former kings of Kalinga, by arranging to drive its grassy jungle into the Lāngala river (Kalimga-puvarāja-nivesita-Pithudaga-dabham Nagale nekāsayati). In the very opening paragraph of the Hāthi-Gumpha text (I. 1), he has been represented as "an increaser of the fame and prosperity of the royal House of King Ceta” (Cetarājavamsa-vadhana). The same opening paragraph goes to show that he felt much pride in declaring his connection with the reigning dynasty of Kalinga. And the same inference can be drawn from the concluding paragraph (I. 16) where he figures as a king who descended from a family of the dynasty of royal sages (rājisi-vamsa-kula-vinisita).
It is clear from the records in the Hathi-Gumphā text that King Khāravela successfully followed all the traditional methods of Indian kings to please his subjects. For instance, the first year's record (1. 2) says that as soon as he was anointed, in the very first year of his reign, he undertook to repair the capital of Kalinga, to build up the embankments of the deep and cool tanks and to restore all the gardens, avowedly as a means of pleasing the people, his subjects (pakatiyo ca ramjayati). In the sixth
ar's record (1.7). we read that he showed a great favour to the inhabitants of towns and districts by remitting all taxes and duties, which, too was one of the traditional methods of gaining popularity by a king with his subjects. The third year's record (I. 4), goes to show that he tried to entertain the citizens of the capital of Kalinga by musical performances and festivities and merry gatherings.'
It is equally clear from other records in the Hāthi-Gumphā text that King Khāravela spared no pains and left no stones unturned to win the
33

Page #286

258
OLD BRAHMI INSCRIPTIONS
heart of his subjects by convincing them of the fact that he gave them the full benefit of a good and efficient government, and that he enhanced their happiness by administeriny justice with an even hand, by increasing the wealth and prosperity of his kingdom, by the maintenance of internal peace and order, by guarding the kingdom well against foreign invasions, by facilitating communication with the capital, and no less by securing for the people a fertile tract of land for agricultural purpose.
The details of Khāravela's administrative machinery are unknown The Hathi-Gumpbă text records (I. 14) that the rayal servants (rāja. bhatakā) co-operated with him in excavating caves for the Jain saints and recluses on the Kumāri hill. The reference to the royal servants is to be found also in the record of his eighth regnal year (I. 9.). As regards these royal servants, we have mention of a Nagara-akhadımsa (Town-judge) in the old Brāhmi inscription (No. IX), of a Mahāmada (if it is the same official designation as Mahāmatta) in the inscription (No. X), and of a Padamulika (if it is the same official designation as Rāj pādamūlika in the Jātakas) in the inscription (No. XIV). We have no information as to whether there were any separate boards and departments, and as to whether any innovations were introduced by King Khāravela in the existing administrative system. Khāravela was evidently too much preoccupied with the ideas of military campaigns and expansion of dominions to be able to think of administrative changes, and Kalinga was too small a kingdom compared with the Maurya empire to require any very elaborate administrative arrangements.
As regards attitude towards the tradition of the former kings, there are some important points of difference between King Khāravela and King Asoka. As is well-known, King Asoka has not cared at all to refer to his royal pedigree in his edicts. His reference to his brothers, sisters and relatives (R. E. V) has no bearing upon the question of his ancestry. Even the name of the Maurya royal dynasty to which he belonged does not find mention. In his famous Bhabru Edict, be haş simply introduced himself as "the king of Magadha" (lājā Māgadhe). In the inscription of Khāravela's chief queen (No. 11), Her Majesty has been represented not by her personal name but as the chief queen of Khāravela-siri and the daughter of the high-souled King Lālárka Hastisäha-Hastisimha (rājino Lāldkasa Hothisäha-sampanātasa dhutā), while in Adoka's Queen's Edict," he has issued instructions to his high functionaries to commemorate all the works of benevolence done under the auspices of his second queen by inscriptions recording them as " donations of his second queen

Page #287

NOTÉS
269
Kāluvāki, the mother of Tivala” (Dutiyāye deviye ti Tīvala-mātu Kāluvākiye [dāne]). Thus the same difference of Asoka's mental attitude has been clearly brought out in his "Queen's Edict,” in his instructions to commemorate his second queen's donations by inscriptions representing her by her personal name and as his second queen and the mother of Prince Tivala.
In as many as five of his edicts (R. E. IV, R. E. V, R. E. VI, R. E. VII, P. E. VII), King Asoka has discussed his own position as a ruler with reference and in contrast to that of the former kings " who reigned in the past, during many hundred years." Like all great reformers, Buddha, Christ, and others, he has declared himself with reference to those who had gone before him to the effect that he came rather to fulfil than to destroy the Law. He says (P. E. VII) that he was able to recognize that the underlying motive of the former kings was to see the Law prosper sufficiently among the people, but the means adopted by them was not well-suited to the end. True to this underlying motive or spirit, he proceeded to devise, as a ripe fruit of his own reflections, certain new methods, such as the appointment of Dharma-mahāmatras, the public proclamations conveying happy royal messages, the formulations of moral principles, the nactment of many legislations, the introduction of quinquennial and tri. ennial tours of official inspection, the arrangements for a prompt despatch of business, the granting of the power of discretion to the high judicial officers, whereby he could see the Law prosper sufficiently among the people. So far as those methods went, he claimed that he had played the most difficult part of a pioneer (ādikara).
On the other hand, King Kbāravela, as he is represented in his inscription, appears to have followed the traditional methods of the former kings of Kalinga without exercising any discrimination on his own part. He did not play the rôle of a critic and reformer. He carried out a set programme of royal duties under the prompting of noble instincts and impulses, rather mechanically, without ever realising the need of a conscious adjustment of new methods to changed conditions of a progressive people. Thus in spite of the splendid success attending his administration, the credit must go to the time-honoured tradition which, as one might say, he had the prudence enough not to question...
The ability to foster all religions and to vouchsafe protection and extend patronage to all religious sects and institutions constitutes just another test for determining the status of a king overlord. If the invocation formula of the Hathi-Gumphã inscription, Namo arihamtānam, namo

Page #288

260
OLD BRĀLMI INSCRIPTIONS
sara-sidhānam, be indecisive as to whether Jainism was the religious faith of its composer or it was the religious faith of King Khôravela, for whom | the text was composed, the contents of the Hāthi-Gumpha and other old
Brāhmi inscriptions clearly prove that Jainism was the religious faith of King Khāravela and other excavators of the caves on the Kumāri hill. For instance, the inscription of Khāravela's chief queen (No. II) records that the cave commemorating her name was made for the sake of the Kālinga recluses of Arhata persuasion (Arahanta-pasādānam Kālimgānam samanānam). Similarly, the thirteenth year's record of Khāravela's reign (I. 14) says that 117 caves were built on the Kumāri bill to serve as resting places of the Arhats or Jain saints residing there (Arahato parinibãoato hi kaga-3%8&dagãya).
Buddhism became the state religion of India when King Asoka embraced it and vigorously espoused its cause. His father and grandfather were supporters of religions other than Buddhism. Even when he espoused its cause, it is difficult to prove that all the members of his royal family favoured Buddhism. If there are Buddhist legends in Pali as well as in Sanskrit asserting that some of his wives, sons, daughters, brothers and relatives had become Buddhists, there are both inscriptions and legends to prove that some or most of them had not become Buddhists.
On the other hand, the Hāthi-Gumphā inscription goes to prove that Jainism had become the state religion of Kalinga even long before the reign of Khāravela. With the royal support at its back, it had become the predominant faith in Kaliåga. The thirteenth year's record (I. 14) clearly brings out three facts of importance in this connection : (1) that when King Nanda had conquered Kalinga, he carried off the throne of Jina belonging to Kalinga as the highest trophy (Namdarāja-nita Kālimga-Jinásana): (2) that King Khāravela signalised his conquest of Anga-Magadha by bringing back that Throne of Jina to Kalinga by a triumphal procession ; and (3) that King Khāravela professed Jainism in common with his queens, sons, brothers, relatives and royal servants. One cannot mistake that, somehow or other, the affection and honour of the royal family, as well as of the people of Kalinga, became bound up with that Jipa-throne.
1. Regarding Bindusära, Asoka's father, Buddhaghoga in his Samanta Pászdikā (Oldenberg's Vinaya-Pitaka, Vol. III, pp. 390-301), says: Bindusāro brāhmana-bhatto ahosi : 80 brāhmaṇānañ ca brāhmanajātiyapåsandānañ ca pandaranga-paribbājakādinam (pandarangaparibbājaka-Gjivika-Niganchadinam) nicca-bhattam pattha pesi.

Page #289

www. kobatirth.org
NOTES
261
This is not to say there were no other religions and religious shrines in Kalinga. The Hāthi-Gumphã inscription clearly proves that there were other religions and religious edifices. If the remaining religions, including the different forms of animism, be designated in the lump as Hinduism, one can say that during the reign of King Khāravela, as also during the reign of his predecessors, Hinduism flourished side by side with Jainism, and the Hindu temples dedicated to various deities shone forth along with the cave dwellings of the Jain saints and recluses.
The royal epithet sava-pāsamda-pūjaka occurring in the concluding paragraph of the Hāthi-Gumphā text (I. 16), attests, beyond doubt, that King Khāravela of Kalinga unknowingly followed in the footsteps of King Devanampiya Piyadasi Asoka of Magadha in declaring himself as a ruler “ who honoured all lenominations." There would have been no necessity for the use of such an epithet, if there were no adherents among the people of Kalinga of different denominations. King Asoka in his R. E. XIII, definitely says, while speaking of the spread of Aryan religions in India in his time, that there was no other place but the Yona-region, where the sects of the Brāhmaṇas and Sramaņas were not, nor was there any other place where the people had not adhered in faith to one or the other of those sects.
Similarly, there would have been no necessity for the use of the epithet sava-deváyatana-samkāra-kāraka, "the repairer of all temples of the deities", if there were no worshippers among the people of Kalinga of those deities at the temples dedicated to them. The second epithet representiug Khāravela not as a builder, but only as a repairer of those temples, goes rather to show that those places of worship had existed from an earlier time. What those temples could be and where they were actually to be found, unfortunately, the Hāthi-Gumphā inscription does not clearly specify. And no Hindu temples have as yet been discovered with inscriptions or tablets recording that they were caused to be repaired by King Khāravela.
Who could the recognised representatives of the different religious denominations be is also a question at issue. Fortunately, the answer is not far to seek. For the fourteenth year's record of Khāravela's reign (I. 15) says that apart from 117 caves excavated on the Kumāri hill for the residence of the Jain saints and recluses, King. Kbāravela constructed a separate cave for the accommodation of the honoured recluses of established reputation (sakata-samaņa-suvihita), as well as for the accommodation of the yatis, hermits and sages hailing from a hundred directions

Page #290

262
OLD BRĀHMI ÎNSCRIPTIONS
(satadisānam yatinam tāpasa-isinam leņam kārayati). Here the yatis, hermits and sages must be taken as representatives of the fourth and third Brahmanical stages of effort (āšramas), that is to say, of the orders of Brahmanical ascetics. It is somewhat difficult to decide whom the king actually meant by referring to them as sakata-samana-suvihitā," the honoured recluses of established reputation."
Mr. Jayaswal contends for the reading sukata-samaņa-suvihitā, taking sukata to be the same word as sukrta,“ virtuous." Kata-samaņa occurs indeed in the Upåli-Sutta, Majjhima-Nikāya, as a Jaina mode of praise applicable to a recognized recluse teacher. Buddhaghoșa explains it as meaning "a recluse who has fulfilled the aims of recluse life." I Taking sukata to be the correct reading, it is easy to tend to equate it with Sugata, which is a well-known epithet of the Buddba. But it would be risky enough, in the absence of clear evidences, to suggest that the Buddhist teachers gained a foothold in Kalinga either during the reign of Kbāravela or before. We bave reasons to think that by sakata-samaņa-suvihitā or sukata-samaņa-suvihitä the composer of Khāravela's inscription rather kept in his view the Jain recluses who, as occasional visitors, had to be distinguished from those who permanently resided on the Kumari hill, Anyhow, the thirteenth year's record of Kbāravela's reign (I. 14) says that 117 caves on the Kumārī bill were excavated to serve as resting places of the Jain saints who resided there (Kumārā-parate arahato parinivāsato hi kaya-nisidiyāya). In the edicts of King Asoka, notably in his P. E. VII, the Jains (Nigamthā), the Ājivikas and the Buddhists (Samghathā) have been mentioned as typical representatives of the orders of the recluses as distinguished from those of the Brahmanical ascetics, while in Khāravela's inscription, the yatis, tāpasas and isis are mentioned as representatives of the Brahmanical orders, and, as regards the representatives of the śramaņas, the. Buddhists and the Ajivikas are passed over in silence. The eighth year's record of Khāravela's reign (I. 9) says that His Majesty feasted all the inhabitants (gharavāsi), all the royal servants (rājabhatakā), all the householders (gahapatayo), all the Brahmins (bamhanā), as well as the Jain recluses (arahata-samaņā) once in Mathurā and subsequently in Kalinga. It will be noticed that in the above enumeration, the royal servants have been
1. Papañca. ûdani, Siamese ed., Part III, pp. 92-93 : kata-samarassđti kata-sämaññassa, samana.dhainmassa matthakampattassa.
2. Barua's Religious Policy of Adoka' in the Maha Bodhi Journal, Vol. XXXV, No. 11, pp. 349.650.

Page #291

NOTES
268
distinguished from the general population (gharavast), in the same way that the Brahmins and the Jain recluses have been distinguished from those who kept to household life (gahapatayo). Here the Jain recluses have been specially mentioned as representatives of the sramanas, while the Buddhists and the Ajivikas have found no recognition. Seeing that the same sort of statement occurs in the Hathi-Gumpha record (I. 9) with regard to feasting in Mathura and to that in Kalinga, we may be led to think that even Mathura proved 'to be up till the reign of Kharavela an impenetrable region for Buddhism, although this faith was destined to thrive there together with Jainism during the reign of the Kusana kings.
It is quite clear, we think, from the foregoing discussion that King Kharavela was a Jain from his very birth. King Aśoka was not born in a Buddhist family. It can be established by the evidence of his own inscriptions and Buddhist legends that he was converted to Buddhism, his conversion itself being a gradual process of mental change,1 and that he possessed and displayed all the zeal of a new convert. Kharavela does not appear to have taken religion so seriously as Asoka.
If Kharavela was a Jain, what sort of Jain was he? The education which he received was purely secular and did not differ from that received by other Indian princes. His coronation ceremony was celebrated, as may be easily imagined, in accordance with Brahmanical rites. The principles and methods which he adopted in governing his kingdom were precisely those prescribed in the Brahmanical treatises on Hindu royal polity. Jainism did not compel him to exercise any scruples in undertaking military expeditions and aggressive wars for territorial expansion and world domination. The patriotic spirit which underlay all his activities was not inspired by Jainism. As for Jainism, he caused a large number of caves to be constructed on the Kumari hill to provide the resident Jain saints and recluses with resting places, and erected ornamented stone-pillars, shrines and pillared halls on a slope of the same hill. As for Hinduism, he made donations for repairing the temples dedicated to various gods and goddesses, and feasted alike the Brahmin ascetics and Jain recluses. Is it not to be inferred from all these that so far as this world was concerned he was a Hindu, and that so far as the other world was concerned, he was a pious Jain ?
1. Radhakumad Mookerji's Asoka, pp. 109-112, f. n. 3.

Page #292

264
OLD BRĀHMI INSCRIPTIONS
Kbāravela differed indeed from Asoka in beating the drum of victory by sword (bherighosa) rather than in proclaiming the glory of conquest by Dhamma, the higher and higher ideal of progressive humanity. He was proud to give out to the world that he excelled in his knowledge of the science of music (Gamdhava-veda-budha) rather than in that of the deeper truths that Jainism had to teach. He caused to be organized a pompous parade of swords, umbrellas, flags, guards and horses, in short, of emblems of royalty (kakudhas). He differed from Asoka also in his endeavour to entertain the citizens of his capital even by pandering to their taste, by dampa-dapa, by dances, songs and instrumental music, and by festivities and merry gatherings (dampa-nata-gita-vādita-samdasanāhi usava-samāja-kārāpanāhi). If dampa be the correct reading, it is, without doubt, the same word as darpa, which is mentioned in the Artha-Sastral as a sport and pastime (krīdā) along with madya-krīdā. The Artha-Šāstra prescribes a fine of three paņas for the ladies of good society going to witness these two sports and pastimes. There must have been something inherently wrong in them for which the Artha-Sāstra found it necessary to prevent the ladies of good society under penal laws from witnessing them.
The word madya (“spirituous liquor") suggests that in the sport od pastime bearing its name, there was a good deal of drunkenness, a good deal of licentiousness, a good deal of intemperance, a good deal of middight revelry. The exact significance of dampa or darpa-krīda is unknown. Bhattasvāmi's commentary on the Artha-Šāstra does not throw any fresh light. It may be a general name for a number of sports and pastimes in which challenging, boasting, competing and bettings play an important part. The name dampa or darpa seems to convey the idea of combating such as in wrestling, boxing, mock-fighting and gladiatorial feats. Anyhow, there must have been in this kind of sport and pastime a good deal of excitement, a good deal of noise and tension, for which the ladies of good society were prevented from witnessing it.
If dapa be the correct reading, it is, without doubt, the same term as dava kamma, which is mentioned in the Maha-Nid desa (p. 379) as an example of vācasikā khiddā along with nātika (dramatic acting), gita (singing) and lāsa (dancing by women) precisely in the same way that in the Artha-Šāstra darpa and madya-krīdā are mentioned along with preksā
1. Artha-Sastra, III. 358: Pratigiddh a-stri darpa-madya-kridayām tripanam dandam dadyat.

Page #293

NOTES
265
(dramatic performances, operas). The commentary explains davakamma in the sense of comics' (hasa-karaṇa-kīļā).
Whatever be the correct reading of the word, it is certain that Kharavela did not refrain from pleasing the citizens of his capital by 'combats' or 'comics,' and by dancing, singing and instrumental music, which were against the doctrine of Jainism. There is no evidence to prove that he exercised discrimination, like Asoka, in selecting and encouraging only those samajas' or joyous gatherings' which were approved ways of educating the people even through amusements and festivities.
It is true that Kharavela, too, honoured all denominations, which is to say, that he, too, observed the principle of religious toleration. But his idea of religious toleration was essentially of a Hindu nature. In his case, just as in the case of a Hindu, toleration implied the idea of noninterference, non-intervention, not meddling in another man's religion. He found it to be a wise policy on his part to leave each sect to follow its own creed without taking the trouble of considering the details of each faith. He does not appear to have made an attempt to bring all sects on a common platform for a free and frank discussion, and an interchange of ideas for discovering the common ground and mission of all religions, as well as for determining the merits and defects of each religion. Asoka and Akbar had their own ideas and programmes of religion. But Kharavela had no such ideas and programmes. How Aśoka's idea of religious toleration differed from that of Kbaravela will be manifest from Aśoka's R. E. XII, the relevant portion of which is quoted below:
"His Sacred and Gracious Majesty the king is honouring all sects, both ascetics and house-holders; by gifts and offerings of various kinds is he honouring them. But His Sacred Majesty does not value such gifts or honours as that how should there be the growth of the essential elements of all religious sects. The growth of this genuine matter is, however, of many kinds. But the root of it is restraint of speech, that is, there should not be honour of one's own sect and condemnation of others' sects without any ground. Such slighting should be for specified grounds only. On the other hand, the sects of others should be honoured for this ground and that; concord alone is commendable in this sense that all should listen and be willing to listen to the doctrines professed by others. This is, in fact, the desire of His Sacred Majesty, viz., that all sects should be possessed of wide learning and good doctrines."
1. For the meaning of this term, see Radha Kumud Mookerji's "Asoka," p. 129, f. n. 1.
34

Page #294

266
OLD BRĀHMĪ INSCRIPTIONS
13. SOME POINTS CONCERNING PERSONAL
HISTORY Our rendering of Aira as “ Lordly” requires a word of explanation. The first letter as it appears on the stone or in the estampages cannot but be read as ai. The correctness of such a reading is confirmed, no doubt, by the Sanskrit slokas quoted by Mr. Jayaswal from an old Oriya MS. But one must not lose sight of the fact that in these slokas, Ahira has been used as an alternative spelling. We have, moreover, sought to show that a similar letter occurring in the two Pabhosā inscriptions of Aşadhasena has been read by Dr. Vogel as va, the letter serving as the initial of the personal name Vaihidari. Should Aira be read on this ground as Vera, we have mentioned that there is no other alternative but rendering it " Hero” or “ Heroic," vera being the same word as vīra. Even if we adhere to the reading Aira, we do not quite understand why it should be interpreted as signifying " a descendant of Ila.” Rather keeping the two alternative spellings Aira and Ahira in the Sanskrit ślokas from the old Oriyā MS. in view, we have to think of a royal title, which can account for both of them. Such a royal title is undoubtedly Ayira which has been explained in the Jātaka-Commentary as meaning sāmi, “master” or “ lord," "a master as distinguished from a slave," that is to say, “an Arya, whose condition, according to the Artha-Šāstra, is not servitude.” 1 We still feel that Mr. Jayaswal's first note on Aira (JBORS, Vol. III, Part IV, p. 434) is something to the point : “The first word of the royal style is Aira. This word occurs in a Sātavāhana inscription and has been translated by M. Senart as 'noble' (Arya). I am inclined to take it as indicating the ethnic difference of Khāravela from his subjects (who were mostly Dravidians, or the mixed Aryo-Dravidians, for according to the Nátya-Sāstra, the people of Kalinga wore dark but not black)."
The personal names occurring in our old Brāhmi inscriptions have an importance of their own. We find that most of them are obviously Aryan names, viz., Kamma, Culakamma, Kusuma, Nākiya (Nāgita), Hathisaha. Hathisiha and Khiņā (misread Halakhina-Slakonā). Vadukha or Varikha, too, appears to be an Aryan name, Vadukha being a Prākrit equivalent of Vadavākşa, and Varikha that of Varekşu. As regards Kadampa or Kudepa (misread Vakadepa by Dr. Indraji, and Kudepa by Mr. Banerji),
1. Barhut Inscriptions edited by Barus and Sinha, sub poce" Aya."

Page #295

NOTES
267
we are inclined to think in the same way. Kadampa may be equated either with Kadamba or with Kandarpa. Kudepa as an alternative reading may be treated as an equivalent of Kudeva (the “ Lord of the Earth").
Mr. Jayaswal explains the name Khāravela as meaning “ the Ocean" (lit.“ one whose waves are brackish").1 Prof. S. K, Chatterji prefers to explain it in the sense of Kāda-vilvan, “ the Black-lancer," kāda being the same word as the Sk. kršņa, meaning "black." Kāļavela occurs in the Mahāvamsa2 as the name of a Yakkha, and in the Jätaka-Commentary as the name of a village in Ceylon, the spelling in the latter case being Kalavela.3 The word kālavela is met with in the Mahāniddesa where it is explained in the sense of " one who speaks words befitting the time." 4
We can quite see that khāra is the same word as kāla or krena, and vela is an equivalent of vilva, cf. Uruvela-Uruvilva. Whatever the sense in which the name is interpreted, Khāravela may be equated with Kronavilva. But, as suggested in the Mahāniddesa, vela of Khāravela may have been derived from velā meaning " the shore ” or “the wave breaking upon the shore." If so, Khāravela must be equated with Kronavela meaning “ the Sea," “ the Ocean" (lit., “that which is girt by watery black shores"). Kālidāsa's famous description of the sea or ocean may serve, it is hoped, to clear up this meaning of Khāravela or Kronavela :
Dūrādasyaścakranibhasya tanvi
tamāla-tālı-vanarāji-nila / Abhāti vela lavaṇamyurāśer
dhāràpibadd heva kalan ka-rekhā 5 !
1. JBORS, Vol. III, Part IV, p. 434. 2. Mahavamsa, IX-23. 3. Faugböll's Jataka, Vol. VI, p. 30 : Kalavelavasi.
4. Mahápiddesa, p. 504 : Katamā kālavelā? Kālātikkantam vācam na bhāseyya kalamı asampattam vācam na bhäseyya.
8. Raghuvamsa, Canto XIII. Vela tirabhumih dharanibaddha cakrāórita kalankarekha mälinyarekha iva abhāti (Mallinatha). “Velā syåt tiranirayoh " iti Visvah.

Page #296

268
14.
OLD BRAHMI INSCRIPTIONS
KHARAVELA'S PLACE IN HISTORY
Kharavela is one of the most striking figures in the annals of Indian kings. Although he cannot claim the proud position enjoyed by an Aśoka or an Akbar as a world-figure, even as a provincial figure, he represents a remarkable and charming personality. He maintained the noble tradition of Asoka as a successful builder of such sacred and artistic monuments as rock-cut caves, stone-pillars, shrines and ornamented shrine-posts, as a ruler who honoured and favoured all religious sects, and no less as a king who did his level best to work and strain all his resources for the good and happiness, or as is put in his inscription, for the pleasing of the people, his subjects, while he became, by his warrior-like spirit, valour and victory, the worthy precursor of Samudragupta. As an expert in the science of music (Gandhava-veda-budha) and a patron of fine arts, too, he played well the rôle of a precursor of the imperial Gupta monarchs. From the chronological point of view, too, he stood just midway between Aśoka, on one side, and Samudragupta, on the other. In respect of its style and contents, his inscription in the Hathi-Gumpha, too, must be accorded a similar intermediate position between the notable inscriptions of Asoka and the Allahabad pillar inscription of the lion-like Samudragupta. But unique is his posi tion in Indian history as an unsurpassed patriotic king of Kalinga, just in the same way that unique is his inscription of seventeen lines on the hanging brow of the Hathi-Gumpha roof in its presentation of systematic records of his successive regnal years. King Aśoka followed, as shown elsewhere,1 a quinquennial system in issuing his edicts and in all of his other undertakings from his twelfth regnal year onwards. So far as the edicts of Asoka go, these are far from presenting a systematic record of the events of his successive regnal years except, as one might say, with regard to the release of prisoners from the imperial jail (P.E. V). But, nevertheless, the chronological system of presentation followed in Kharavela's inscription goes to connect it historically with the edicts of the Maurya emperor rather than with the inscriptions of the Gupta emperors. It is the Jains and Buddhists who dated their votive offerings in the term of the successive years of the reign of the Kuşana kings, precisely in the same way that in
1. Barua's Asoka Edibts in New Light, pp. 63-66.

Page #297

NOTES
269
the subsequent phase of Indian History the universal custom became to date all public and private records and works in the term of an era known as the Saka-era.
V Kbāravela was the greatest known king among the monarchs of the Meghavāhana or Aira-mahāmeghavāhana dynasty who exercised their suzerainty over the kingdom of Kalinga, nay, among all the monarchs who reigned in Kalinga before King Aśoka and after the Meghavāhana kings. The memory of his predecessors in the Meghavähana line would have been completely obliterated from the pages of history but for allusions to them in his inscription as his predecessors.
With regard to the part played by his predecessors in the Meghavahana line, it may be clearly inferred from certain statements in his inscription that since the first king of this line succeeded in freeing the kingdom of Kalinga from the yoke of a foreign rule, they successfully maintained its independence till they safely handed it on to him in the third generation of two kings (tatiye purisa-yuge). If our reading āhatapuva in the sense of āhrtapūrva, "previously annexed," be accepted as correct, the HāthiGumphā record of his fourth regnal year (I. 5) may be so interpreted as to imply that the Vidyādhara-country was already annexed to the kingdom of Kalinga by some one of his predecessors, while the task that was left to him was just to consolidate the Meghavāhana rule over it. In the absence of any clear record in his icscription as to his conquest of Pāņdya kingdom, the record of his twelfth regnal year stating that the king of Pāadya supplied him with pearls, gems, jewels and rich apparels cannot be accounted for without some such supposition that either the king of Padya was an old ally of the Meghavāhana kings or the king of Pāņdya was compelled to acknowledge his supremacy, even as an ally, in fear of the consequences of an invasion contemplated by him. Further, he could not have used Mathurā as a military base of his attacks on Uttarāpatha, on one hand, and on Anga-Magadha, on the other, as it appears from the records of his eighth and twelfth regnal years (I. 9, 1.13) that he did so, if either he himself had not conquered it or it had not been an old dependency of the kingdom of Kalinga.
It is very clear from the record of his first regnal year (I. 2) that the city of Kalinga with its gates, walls, residential houses, deep and cool tanks and all kinds of gardens was used as the capital of the Kalinga kingdom also by his predecessors. If Danta pura-Paloura was the most ancient known capital of Kalinga, and no evidence be forthcoming as to the city of Kalinga being used as the capital by any pre-Meghavāhana

Page #298

270
OLD BRAHMI INSCRIPTIONS
kings, some one among his Meghavahana predecessors must be credited with the building of this as the capital city.
The Hathi-Gumpha inscription contains the following two significant statements indicating under whose sway the kingdom of Kalinga had remained before the Meghavahana kings rose into power: (1) Pamcame cédani vase Namdarāja-tivasasata-oghatitam Tanasuliya-vāṭā panāḍim nagaram pavesayati, "And then in the fifth year, King Kharavela caused the canal opened out by King Nanda 103 (or 300) years ago to be extended into the capital city;" and (2) Bārasame ca vase......Namdarāja-nītam Kalimga-Jinasanam (Amga-magadhato) Kalimgam aneti, "And in the twelfth year, he caused the throne of Jina belonging to Kalinga which was carried off by King Nanda to be brought back (from Anga-Magadha) to Kalinga."
To us the two statements are important as indicating (1) that some king associated with Anga-Magadha and known as Nandaraja had invaded and conquered Kalinga before the Meghavahana kings established their sovereignty over it; and (2) that that Nandaraja opened out a canal or aqueduct near the Tanasuliya, Tanasuli or Tosali road while the kingdom of Kalinga was under his sway.
Now, the Sanskrit slokas from the Old Oriya MS. tell us that King Aira or Ahira of Utkala proved to be the victor in a battle, which ensued between him, on one side, and King Nanda of Magadha, on the other. These represent King Nanda, as we saw, as a staunch supporter and King Aira-Ahira as a ruthless destroyer of the Vedic religion or Brahmanism. As a destroyer of the Vedic religion, King Aira-Ahira is said to have been a great friend (mahāmitra) of King Asoka. The slokas credit King AiraAhira with the transfer of the capital of Utkala from the main city in South Kosala to the Ekaprastara-tract around the hill of Khandagiri. Thus Utkala itself is described as a kingdom, which extended so far as to include in it Kosala and Toşala, forming two main divisions of Kharavela's kingdom of Kalinga. And if we are to believe the story of the Sanskrit slokas, it is from the hands of King Nanda of Magadha that King AiraAhira of Utkala freed the kingdom and people of Kalinga. All these may be true, but until we have the opportunity of examining the nature and credibility of the original source of information, it is likely to be injudicious to attach undue importance to them.
It is too late in the day of Indian historical research to attach any importance whatever to the opinion of Raja Rajendra Lala Mitra assigning Kharavela to the 4th century B. C. and claiming him, á fortiori, to

Page #299

NOTES
271
be a pre-Asokan king. It is impossible to regard Khāravela as a preAsokan king in the face of these two facts : (1) that his inscription in the Hāthi-Gumphā distinctly represents him as a contemporary and rival of Satakarņi, whose territories lay contiguous to the western border of his Kalinga kingdom (I. 3), and (2) that whatever the actual date and identification of this Satakarni, he was undoubtedly one of the Andhrabbstya Sātavāhana kings who rose into power on the decline of the power of the earlier Andhra kings and years after the reign of King Asoka. It is equally impossible to regard Kbāra vela as a pre-Asokan and not as a postAsokan king in view of the fact that his inscription distinctly alludes to Bahasatimita (Bșhaspatimitra) as a contemporary king of Magadha, subdued by him in the twelfth year of his reign, while there is neither tradition nor inscriptional evidence as to Bahasatimita's rule in Magadha before or immediately after Asoka's reign.
If our interpretation of the two expressions, tatiya-purisa-yuga and tatiya-yugasagávasāna, be correct, Khāravela was unquestionably the sixth king of the Mahāmeghavāhana family with whose accession to the throne of Kalinga the third couple of its two successive representative men was completed, and with whose father's death, the reign of the third couple of kings was at an end, and his son and successor Kadampa-Kudepa came, as the seventh king, to be joined with him. The records of the HathiGumpbā inscription take us as far as the fourteenth year of his reign and, by no means, beyond it, and there are neither inscriptions nor traditions to inform us who among the Mahāmegbavāhana kings and how many kings of the Mabāmeg havāhana family reigned in Kalinga as successors of Khāravela and Kadampa-Kudepa, when actually the reign of Khāravela came to an end, or what befell the Mahāmeghavābanas after Khāravela's death.
Looking back, we find that Khára vela remained an ordinary prince for fifteen and a crown-prince for nine years, which is to say, the tenor of his life as a prince and crown-prince covered full twenty-four years of the reign of the Mahāmeghavābana kings. From the foundation of the rule of the Mahāmeghavāhana dynasty right up to the accession of Khāravela we have to conceive the successive reign of (1) the first couple, (2) the first couple and a half, (3) the second couple, and (4) the second couple and a half. Even allowing twenty years as the average period of each reign, we do not get more than eighty years to represent the total length of the reign of Khāravela's predecessors in the Mabāmegbavāhana lipe.

Page #300

272
OLD BRAHMI INSCRIPTIONS
6
It is very likely, as suggested by Mr. Jayaswal, that Megha' in the Purānas is but a shortening from Meghavahana' or Mahāmeghavahana,' which is the high-sounding epithet whereby Kharavela and other kings of the same royal family were designated. The traditions in the Puranas unanimously assert that among the different Indian kings who reigned in various parts of India as contemporaries of the Andhra or Andhrabbṛtya-Satavahana rulers were the kings who reigned in Kosala or South Kosala, who were just nine in number, very powerful, intelligent and well-known as 'Meghas.' And it is expressly stated in the Bhavisya-Purana that seven Kosala kings of the Megha-Meghavahana dynasty and seven AndhraAndhrabbṛtya-Satavahana kings reigned as contemporaries.2
Thus the traditions in the Puranas lead us to understand that altogether nine kings of the Megha-Meghavahana family reigned in South Kosala, which, as we have sought to show, formed one of the three principal divisions of Kharavela's Kalinga kingdom. And if it can be elicited from the Hathi-Gumpha and other old Brahmi inscriptions that Khāravela reigned as the sixth king and Kadampa-Kudepa as the seventh king of the Mahāmeghavahana family, we can say that just two other kings of this family reigned after their death, which is to say, the rule of this family came to an end within thirty or forty years of Kharavela's death.
The determination of the age of Kharavela depends, to a large extent, on a satisfactory identification of King Nanda, Satakani and King Bahasatimita of Magadha, all of whom find mention in the Hathi-Gumpha inscription. As to the identification of King Nanda, this inscription furnishes us with a key in the two facts stated by it: (1) that a canal or aqueduct was opened out near the Tanasuliya, Tanasuli or Tosali road by him 103 or 300 years ago, the years being counted back from the fifth year of Kharavela's reign (pamcame cédani vase Namdaraja-tivasasataoghatita-panaḍim tanasuliya-vāṭā); and (2) that the Kalinga throne of Jina, carried off by him (Namdaraja-nita Kalimga-Jindsana) was brought back from Anga-Magadha to Kalinga by King Kharavela by a triumphal procession. Here ambiguity arises from the possibility of two alternative explanations of the compound ti-vasa-sata as meaning either 103 or 300
1. Pargiter's Dynasties of the Kali Age, p. 51:
Kośalāyām tu räjāno bhavisyanti mahabalaḥ | 'Megha' iti samakhyātā buddhimanto navaiva tu |
2. Pargiter's Dynasties of the Kali Age, p. 51, f. n. 16; Eka kaldḥ ime bhupaḥ sapta Andhraḥ sapta Kausalaḥ.

Page #301

NOTES
273 years. Admitting both the alternatives to be equally possible, we have to look out for a King Nanda of Magadha who conquered Kalinga, carried away the Kalinga throne of Jina as a trophy and opened out a canal in the Tosali division and not far away from the city of Kalinga, either 98 (103-5) or 295 (300—5) years before Khāravela's accession.
The only key to the identification of Sätakani furnished in the HathiGumphā inscription is that he held territories contiguous to the western border of Kbāravela's kingdom of Kalinga, comprising, as it did, South Kosala as one of its three main divisions. The tradition in the BhavişyaPurāna leads us to think that the first seven Meghavābana and the first seven Sātavābana kings reigned as contemporaries, in which case if Khāravela was the sixth Meghavāhana king, Satakani must have been the sixth Satavahana ruler. It remains to be seen whether the tradition in the Parāņa can be so rigorously interpreted as to mean that the first king of one dynasty was a contemporary of the first king of the other, the second of the second, and so on.
Regarding Bahasatimita our information from the Hāthi-Gumpbā inscription is that he is the king of Magadha whom King Khāra vela subdued in the twelfth year of bis reign. In Yasamita's Brick-tablet inscription, found in Mathurā, Queen Yasamitā is described as the daughter of Brhās vātimita, the royal personage whose name is taken by Dr. Vogel to be the same as Bahasatimita or Bșhaspatimitra. In one of the two Pabhosā inscriptions of Aşāờhasena, King Āşadhasena, the king of Adhichatrā, is represented as the maternal uncle (mātula) of King Bahasatimita. The same Pabhosã inscription records the construction of a cave in the tenth year of Udāka whom Mr. Jayaswal identifies with Odraka, Odruka or Ardraka, mentioned in the Purāṇas as the fifth king of the Sunga dynasty, suggesting that King Āsādhasena of Adhichatrā was feudatory to the Sunga kings of Magadha. A coin, too, has been found with a Brāhmi legend recording the name of Bahasatimita. The legend in the Divya. vadāna speaks of a Brhaspati as a Maurya king among the successors of Samprati, the grandson of King Asoka. It remains to be seen whether King Bahasatimita, mentioned as a contemporary of King Kbāravela, is the same personality as Bahasatimita of the coin, King Bahasatimita of the Pabhosā inscription, Bệhāsvātimita of Yaśamitā's Brick-tablet, and Bịhaspati of the Divyâvadāna, or he is a king of any neo-Mitra dynasty, which came possibly into existence sometime after the Kāņvas.'
The data of chronology which may be collected thus from the HathiGumphā inscription and other sources are insufficient to determine
35

Page #302

274
OLD BRĀHMI INSCRIPTIONS . undisputably the date of Kbāravela. In the present state of our knowledge, We can do no better than stating the three views put forward by three emi. nient Indian scholars, drawing the reader's attention to certain strong and weak points in the assumptions and arguments of each of them.
First Mr. Jayaswal's latest publication on the subject 1 assigns Kbāravela's accession to 182 B. C., taking him to be a contemporary of Puş yamitra, the founder of the Sunga-Mitra dynasty, whose accession is placed in 188 B. C. The validity of this view is claimed primarily on the soundness of identification of Bahasatimita or Bșhaspatimitra, mentioned in the Hāthi-Gumphā text as a contemporary Magadhan king whom Khāravela subdued in the twelfth year of his reign, chiefly on the ground that Babasati or Bșhaspati finds mention in the Sänkhyâyana Gșhya-Sūtra (I. 26. 6) as the presiding deity of the Puşya constellation of stars. King Bahasa timita or Bphaspatimitra is sought to be connected with the Sunga-Mitra kings of Magadha by the argument that a king of this very name figures in a Pabhosā inscription as the nephew of King Āşādhase na of Adhichatrā (in North Pancāla), while the main text of this inscription records the excavation of a cave by Aşāò hasena for the Kasyapiya Arhats in the tenth year of Udāka (Odraka, Odruka or Ardrakas) who happens to be counted in the Purāṇas as the fifth king of the SungaMitra dynasty. 4 The reign of a king known by the name of Brhaspati among the successors of King Asoka and of his grandson Sam prati can be inferred as well from a legendary list in the Divyâvadāna. The reign of a king known by the name of Brhaspatimitra towards the end of the Maurya rule can be inferred equally from Yasamitā's Brick-tablet, in which Queen Yasamitā (of Mathurā ?) is described as the daughter of Bșhās vātimita, a name that can be equated either with Bphaspatimitra or with Brhatsvātimitra, the Brāhmi letter-forms of this record appearing to be in their essential features still Mauryan. The coin-name Bahasatimita for Puşyamitra is explained by the fact, that in other Sunga-Mitra and Sungabhịty a-Kāņva coins, the names of some of the Sunga and Kāņva kings agree with and differ from those in the Purāņas :
1. JBORS, Vol. XIIJ, Parts III-IV, pp. 236-245. 2. JBORS, Vol. III, Part IV, pp. 477-478.
3. The Jains commentator Silänka equates U daka with Ardraka. See Jacobi's Jains Sūtras, Part II, p. 417.
4. Pargiter's Dynasties of the Kali Age, p. 31; JBORS, Vol. III, Part IV, p. 474. - 5. JBORS, Vol. I, p. 96.
6. EI, Vol. II, pp. 242-243,

Page #303

NOTES
Coin-name. Agnimitra Bhānumitra
Jethamitra
Purāņa-name. Agnimitras Vasumitra. 5 Vasujyestha.
Sujyestha. Ghoşa. 3 Ghoşavasu. (Vajramitra. Devabhūti. Bhumimitra,
Ghosa Bbadragboşa Indramitra Devamitra Bhūmimitra
The objection arising from the Hathi-Gumphā statement Namdarājativasasata-oghāțitam Tanasuliya-vātā panādim nagaram pavesayati which must be ordinarily rendered as “ He (King Khāravela) caused the canal opened out by King Nanda 103 (or 300) years ago to be extended from the Tanasuliya or Tanasuli road into the city (of Kalinga)” is sought to be met by the argument that here the expression Namdarāja-tivasasata oghāțita should be taken to mean "opened out in 300 Nanda-era (the era founded by King Nandavardhana, the grandfather of Mahāpadma Nanda, in 485 B. C.)." 8 The existence of such an era is sought to be proved by the fact that Alberuni, writing his Indika in 1030 A. D., 'found this era in actual use in Mathurā and Kanauj,' and 'heard the local tradition that the founder of the era lightened the taxes by obtaining wealth from the undersoil (which reminds us of the story of the buried treasures of the Nandas).' 4 The same is sought to be proved also by the evidence of the Yedarava inscription of the Chalukya king Vikramāditya VI, stating : “ Having said, why should the glory of the Kings Vikramāditya and Nanda be a hindrance any longer ? he with a loud command abolished that (era) which has the name of Saka, and made that (era) which has the Chalı counting " (Bombay Gazetteer, Vol. I, Part II, p. 437).
In accordance with Mr. Jayaswal's chronological interpretation of the fact of Khāravela's ' marching against the west disregarding Satakarņi' one must identify . Khāravela's Sātakarņi 'with Satakarụi I, who is no other than Sri-Satakarni mentioned in the Parāṇas as the third king of the Andhrabhftya-Sātavāhana dynasty founded by Simuka, the father of
1. JBORS, Vol. III, Part IV, p. 479. 2. JBORS, Vol. I, pp. 78, 106. 3. JBORS, Vol. III, Parts III-IV, p. 240. 4. JBORS, Vol. III, Parts III-IV, p. 240. 5. JBORS, Vol. III, Parts III-IV, p. 238.

Page #304

276
OLD BRĀHMÍ INSCRIPTIONS Sri-Sātakarņi, and Satakarņi mentioned as the Lord of the Deccan (Dakşināpatha-pati) in the Nānāghāt Cave inscription of Queen Nāyanika, the daughter-in-law of Simuka. Further, one must assign this Satakarņi's accession to 172 B. C., placing the foundation of the Sātavāhana dynasty in 203 B. C., on the strength of the expert opinion of Dr. Bühler who found Nāyanika's Nānāghāt Cave inscription to be, on palaeographic grounds, slightly older than Khāravela's inscription in the Hāthi-Gumpbā.
Our difficulty in accepting his interpretation of Namdarāja of the expression Namdarāja-tivasasata-oghātita in the sense of "Nanda-era” is that the same construction cannot be applied to Namdarāja in Namdarāja-nita Kālimga-Jindsana, which, too, is an expression in the Hāthi
umphā text and must be rendered as “the Kalinga throne of Jina carried away by King Nanda." We cannot appreciate the presumption in the Cambridge History of India suggesting even the possibility of Namdarāja being a local ruler, that is to say, one of the earlier kings of Kalioga, in view of the clear suggestion in the Hathi-Gumphā record that this King Nanda took away the throne of Jina from Kalinga and was somehow connected with Anga and Magadha.
Thus the interpretation of Namdarāja as Nanda-era failing to hold its ground, the only alternative left is to ex pound the compound Namda. rāja-tivasasata-oghātita-panādi as signifying "the canal opened out by King Nanda 103 or 300 years ago." Though there is much to be said in favour of the presumption that "103 years" is the prima facie construction of tivasasata, and that if " 300 years ” were the intended sense, 'the expression would have been differently worded, the compound, as it is, must not be expounded excluding the possibility of the sense of “ 300 years."
Proposing to expound the compound tivasasata in the sense of “ 103 years," counted back from Khāravela's fifth regnal year and to identify Khāravela's contemporary King Bahasatimita of Magadha with Puşyamitra, the founder of the Sunga dynasty, none can escape from the conclusions that "Khāravela's Nanda' was no other than King Asoka Maurya who conquered Kalinga in 261 B. C., in the eighth year of his reign, and that Khāravela's accession took place in 163 (261–98) B.C. But
is no other plausible reason for the supposed representation of King Asoka as King Nanda except the fact that his grandfather, Chandragupta
1. JBORS, Vol. III, Part IV, pp. 441-442.

Page #305

NOTES
277
NOMOR
Maurya, is described in the Vāyu-Purāņa as a scion of the Nanja family (Nandendu).
To us, the proposed identification of Khâravela's Nanda' with Nandavardhana or Nandivardhana, the first king of the pre-Mauryan Nanda dynasty, is arbitrary. It is difficult to prove that Nanda vardhana either conquered Kalinga or reigned in 458 B.C. to be able to found an era in that year.
We find it difficult to agree with Mr. Jayaswal in thinking that “Khāravela's Sātakarņi” holding territories that lay contiguous to the western border of Khāravela's Kalinga kingdom, was a contemporary of Puşyamitra, the founder of the Surga-Mitra dynasty, in the face of these two facts: (1) that the traditions in the Purānas assert that the Andhra. bhrtya-Sātavāhana rulers established their suzerainty after the reign of the Sungabhịtya-Kāṇya kings had come to an end, 2 as well as after the destruction of the Andhra power ;) and (2) that the territories to the west of Khāravela's Kalinga kingdom were included in the Sunga empire during the reign of Pusyamitra and governed by the Surga Viceroy Agnimitra, the son of Puşyamitra. : Vidišā was governed, according to the Mālavikågnimitra, by Prince Agnimitra acting, no doubt, as the Viceroy of his father, King Puşyamitra. One of the Barhut inscriptions records the first pillar of the Barhut stonerailing as a gift from Cāpādevi, the wife of Revatimitra of Vidišā,5 Another inscription records another gift from Vāsişthi, tbe wife of Veli. mitra of Vidisā. There can be little doubt that both Revatimitra and Velimitra were connected with the Mitra family in Vidisã. The Barhut E. Gateway inscription clearly proves that even when King Dhanabhūti invested the Barhut stone-railing with the gateways, Barhut continued to be included in the Sunga dominions (Suganam raje).6
The Purāṇas definitely state that Siśuka (Simuka of the inscription), the founder of the Andhrabhstya-Sātavāhana dynasty, came to rule the
1. B, C. Mazumdar's Orissa in the Making, p. 56. In the Mudrā-Rakşasa, too, Chandragupta Maurya is described as a son of the Nanda king by a Sūdra woman. In the Buddhist Pradition, the origin of the Moriyas is traced to the Moriya warriors of Pippalivana.
2. Pargiter's Dynasties of the Kali Age, p. 35. 3. Pargiter's Dynasties of the Kali Age, pp. 45.46. 4. H. C. Raychaudhari's Political History of Ancient India, second edition, p. 236. 5. Barhut Inscriptions (Barua and Sinha), No. 4. 6. Barhut Inscriptions (Barua and Sinha), No. 1.

Page #306

278
OLD BRAHMI INSCRIPTIONS
earth by destroying the lingering remains of the Sunga power represented by Sušarman, the last of the Sungabhịtya-Kāņva kings. On the termination of the Sunga reign, Vidiśã came under the rule of Siśunandi who was succeeded by his younger brother Nandiyasa or Yasonandi, and Nandiyaga was succceded in his turn by three kings of his line, wbile his daughter's son Sisuka became the ruler of Purikā.! Curiously enough, as Dr. Raychaudhuri observes, Sisuka is precisely the Purāņa spelling of the name of Simuka, the founder of the Andhrabbộtya-Sātavāhana dynasty.2 Here two important facts can be gleaned from the Purāņa chronicles : (1) the rise in Vidiśā of a neo-Nandi or neo-Nanda family of kings who struggled for supremacy within the Surga dominions, it being quite possible that either Siśunandi or Nandiyaśa temporarily seized the throne of Magadha, and even conquered Kalinga and opened out a canal 98 years prior to Khāravela's accession and after the termination of the rule of the Sunga family, and (2) the rise of the Sātavāhana dynasty after destroying the vestiges of the Sunga power represented by the last of the Kārvas, say, in 27 B.C.
The evidence of the Pabhosā inscription of King Āşādhasena of Adhichatră, representing Āşā!hasena as the maternal uncle of King Bịhaspatimitra and recording the construction of a cave for the use of the Kāśyaplya Arhats in the tenth year of Udāka who is mentioned in the Parānas as the fifth king of the Suóga-Mitra family, is really a strong point in favour of Mr. Jayaswal's identification of Khāravela's Magadhan rival Bphaspatimitra with Puşyamitra. But there are certain inscriptional evidences to dispute the correctness of the inclusion. It is difficult moreover, to account by Mr. Jayaswal's theory of two sets of names in literature and epigraphy for a number of kings bearing 'Mitra' as a common surname, e.g., Vişnumitra of Gautamimitrā’s inscription, Brahmamitra whose wife Nāgadevi donated a corner pillar of the old stone-railing at Buddha-Gayā, and Indrågnimitra whose wife Āryā Kurangi donated a large bulk of the Buddha-Gayā stone-railing. There may seem to be much truth in Dr. Raychaudhuri's suggestion that these Mitra kings belonged, perbaps, together with Bșhaspatimitra, to a neo-Mitra dynasty, which arose sometime after the destruction of the Kāņva power. We
1. Pargiter's Dynasties of the Kali Age, p. 49. 2. H. C. Raychaudhuri's Political History of Ancient India, second edition, p. 254.
3. Sir John Marshall's "A Guide to Sanchi," p. 11, f. n.; Raychaudhuri's Political History of Ancient India, p. 252.

Page #307

NOTES
279
must bear always in mind that so far as Puşyamitra is concerned, he is designated Pusyamitra both in literature and in inscription (Jayaswal's
Sunga Inscription of Ayodhyā '), and so far as Bșhaspatimitra is concerned, he is designated Bahasatimitra in the inscriptions and in the coin-legends. Secondly, in the opinion of Rai Bahadur Ramaprasad Chanda, Namdarāja of the Hathi-Gumphā inscription should be identified with Mahāpadma Nanda or with one of his sons on the ground that "it is not Nandivardhana but Mahāpadma Nanda who is said to have brought all under his sole sway' and uprooted all Kşatriyas or the old reigning families ; " Sātakani mentioned in the Hāthi-Gumpha inscription as a contemporary of Khāravela should be identified with “ Sätakarni II whose reign may be tentatively dated between B.C. 75.20," partly on the ground that this Satakarņi finds mention in the Purānas without a qualifying adjective sri' characterising his name, just in the same way that Sātakani finds mention in Kbāravela's inscription withouc such a qualifying epithet; King Sri Sātakarņi of the Sanchi Gateway inscription should be distinguished as Satakarni II from Simuka's son, Sātakarņi I, referred to in the Nānāghāt Cave inscription of Nāyanikā for the simple reason that on palaeographic grounds the HāthiGumpha inscription of Khāra vela must be placed between the Nānāgbât Cave and Barbut E. Gateway inscriptions, on one side, and the Sanchi Gateway inscription, on the other. 1
Identifying Khāravela's Nanda' with Mahāpadma Nanda or with one of his sons who reigned in the 4th century B.C. and explaining ti-vasasata in the sense of “300 years," it is easy to conclude that the rise of Khāravela “probably synchronised with the fall of the Sunga dynasty and the consequent weakening of the power of Magadha." 2
Our main objectiou to the proposed identification of Kharavela's Nanda' with Mahāpadma Nanda or with any other pre-Mauryan Nanda king is that it is almost conclusive from the statements in Asoka's R, E. XIII that Asoka was the first among the Indian kings reigning after the Buddha's demise to conquer the theretofore unconquered land of the Kalingas (avijitam vijinitum) and to annex the same to his own kingdom. Kalinga has been described as 'a recently annexed territory' (adhuna-ladha Kalimgā), it being conquered just in the eighth year of his reign (261 B.C.). Serious reflections on the terrible consequences of his war with
1. Memoirs of the Archaeological Survey of India, Vol. I, pp. 10, 11, 14-15. 2. Ibid, Vol. I, p. 11,

Page #308

280
OLD BRAHMI INSCRIPTIONS
Kalinga are said to have brought about a great turning point in his life and career. It is evident from his two Separate Rock Edicts that he governed the province of Kalinga by his viceroy and high functionaries. It is certain that this province remained under his rule for at least twentynine years (261-232 B.C.). It is manifest from all his statements that he really found the conquest of Kalinga a hard nut to crack, and that, in spite of his most determined attacks, he failed to conquer and annex all the tracts covered by Kharavela's kingdom of Kalinga. He had to leave out certain parts demarcating them as unconquered outlying tracts (amtā avijita) and placing them in charge of his high functionaries known as Wardens of the Marches' (Amtapālā-mahāmātā). The digging of a canal near the Tosali road may be regarded as a possible work of his. And, lastly, there were two sets of inscriptions, one set incised on the Dhauli rock separated by an open plain from the hills of Udayagiri and Khandagiri and the other incised on the Jaugada rock within a few miles from Khāravela's capital to remind the inhabitants of Kalinga of the fact of its conquest by an outsider. The personal name of King Asoka having not been recorded in these inscriptions, it was not impossible for the local people to identify the author of the inscriptions and the excavator of the Tosali road canal with a Magadhan king Nanda whom the growing spurious Brahmanical traditions made them familiar with.
Two of the stories of hell in the Petavatthu which is one of the postAsokan Pali works included in the Buddhist Tipitaka alludes to the kingdom of a fabulously rich Nanda king abounding in the wealth of all kinds of apparels. 1 The commentary on the Petavatthu identifies Nandaraja of these stories with a pre-Buddhistic king of Kasi.2 The Jaina BhagavatiSūtra (XV. 1) speaks of a powerful King Jayasena Vimalavahana Mahapadma (Mahapauma) reigning in the city of Satadvara at the foot of the Vindhya mountain long long after the death of the Ajivika teacher Gosala who predeceased Mahavira by sixteen years. As a mighty persecutor of the Jains, this King Mahapadma is represented as an embodiment of the evil spirit of Gośāla. None need be surprised if this Jayasena
1. Petavatthu, II. 1, III. 2 :
Yavata Nandarājassa vijitasmim paṭicchada.
2. Paramattha-Dipani, the volume containing the Petavatthu-Commentary, p. 73. Somehow, it is to this King Nanda that the legend in Alberuni's Indika of the Nanda king, rich with buried treasures, may be traced.
3. None need be surprised if some such king was behind the story in the Sanskrit slokas from an Old Oriya MS. of a vedadharma-parāyaṇa Nanda rival of the Jain King AiraAhira of Utkala.

Page #309

NOTES
281 Vimalavāhana Mahāpadma who held his kingdom at the foot of the Vindbya mountain was the same king as Pra vīra, the son of Vindhyasakti, whom the Purāņas praise as a ruler who reigned in the city of Kāncanaka for sixty years and performed a horse sacrifice. 1
Under the circumstances, particularly having regard to the hard facts recorded by King Asoka, it may be far safer to err by identifying • Khâravela's Nanda' with Asoka Maurya and assigning Khāravela's accession to the second quarter of the ist century A.D., interpreting ti-vasa sata in the sense of " 300 years," or by identifying him with Siśunandi or with Yaśonandi who snatched away Vidišā from the Sunga dominions on the fall of the Sunga power and assigning Khāra rela's accession to the same second quarter of the 1st century A.D. interpreting ti-vasa-sata in the sense of “103 years," than falling into endless difficulties by proposing to identify Khāravela's Nanda' with Mahāpadma Nanda and to assign Kbāravela's accession to the period covered by the reign of the Kāņvas in the 1st century B.C. and identifying Khāravela's contemporary King Bphaspatimitra of Magadha with none.
Khāravela's Sātakarni may have been the sixth Satavāhana king mentioned in the Purānas, that is to say, Rai Bahadur Chanda's Sātakarni II. This identification is borne out by the tradition in the BhavisyaPurana representing the first seven Meghavāhana and the first seven Sātavāhana kings as contemporaries. We may even concede that Sāta. karņi II was King Sri-Satakarņi during whose reign and within whose kingdom the Sanchi gateway was donated by a donor. But how can we think that this Sātakarņi reigned in the Vidiśā region after the fall of the Sunga power when it is expressly stated in the Purāņas that King Sisunandi and his four successors reigned in Vidiśā contemporaneously with the Kānva rulers of Magadba, and that Sisuka-Simuka founded the Sätavāhana dynasty only after he had succeeded in destroying the lingering remnants of the Sunga power represented by Sušarman, the last of the Kanva kings, --Siguka-Simuka of Parikā who is described as Yaśonandi's daughter's son ?
Thirdly, in the opinion of Dr. Raychaudhuri, Khāra vela's Nanda' should be identified with Mahāpadma Nanda on the authority of the Purāṇas extolling Mahāpadma, the third pre-Mauryan Nanda king, as "the sole monarch who brought all the ruling powers under his way," 2 and Khāravela's
1. Pargiter's Dynasties of the Kali Age, p. 50. 2. H. C. Raychaudhuri's Political History of Ancient India, 2nd edition, p. 141.
36

Page #310

www. kobatirth.org
282
OLD BRÄHMI INSCRIPTIONS Satakarņi' should be identified with Sātakarņi I, while Satakarņi I himself should be identified with (1) Simuka's son, King Satakarni, the lord of the Deccan, mentioned in the Nānāghāt Cave inscription of Nāyanikā, (2) King Sri-Satakarņi of the Sanchi Gateway inscription, (3) the elder Sarganus mentioned in the Periplus, and (4) Sāta karvi, the lord of Pratisthāna and father of Saktikumāra mentioned in Indian literature,i Regarding Mahāpadma Nanda, he seems to think that the traditions in the Purānas can be liberally interpreted so as to imply that even the old ruling family of Kalinga was subdued by the all-powerful Nanda monarch. In short, he does not think it improbable that Mahāpadma Nanda effected a conquest in Kalinga,” which is said to have been ruled by thirty-two or a similar number of kings who reigned independently as contemporaries of the Saišunāgas. At the same time he seeks to maintain that Sisuka-Simuka, the first king of the Sātavāhana family, was able to found the Sātavāhana dypasty only after the destruction of the Kaņva power in 28 B.C.
We have already disposed of the boasted claims in the Puranas for Mahāpadma Nanda. If Khåravela's Nanda' be Mahāpadma Nanda, the all-powerful pre-Mauryan Nanda monarch of Magadha, if 'Khāravela's Satakaroi' be the third king of the Sāta vāhana family, and the Satavāhana dynasty were founded not earlier than 27 B.C., and if “ 300 years" be the correct rendering of ti-vasa-sata, it is for Dr. Raycbaudhuri to explain how Kbāra vela's accession can be placed in the last quarter of the 1st century B.C. We must still hold that one should take the presumption created, on the strength of the tradition in the Purānas, in favour of the possibility of Mahāpadma Nanda's conquest of Kalinga with a grain of salt, first, because this tradition does not tally with the more reliable tradition in the Pāli chronicles, including the Mahābodhivamsa, and, secondly, in the face of a clear evidence of Asoka's R. E. XIII tending to prove just the contrary of what is elicited from the Parāṇas. Dr. Raychaudhuri bas attempted indeed in vain to invest the tradition in the Purāṇas with cogency and authenticity by reducing the length of Mahāpadma Nanda's reign from 88 to 28 years on the strength of the reading astāvimšati in a solitary manuscript of the Vāyu-Purāņa. So long as the reading about the total length of the reigns of Mahāpadma and his successors remains satam samāḥ ("one bundred years "), the reduction of the length of Mahāpadma's
1. H. C. Raychaudhuri's Political History of Ancient India, 2nd edition, p. 263 foll. 2. Ibid, pp. 142. 3. Pargiter's Dynasties of the Kali Age, p. 24.

Page #311

NOTES
283
reign from 88 to 28 in a single manuscript of a particular Purāṇa is unavailing as a proof of authenticity and cogency of the tradition.
We cannot, however, help appreciating the general tenor of Dr. Raychaudhuri's arguments persuading us to assign Khāravela's accession to the second quarter of the 1st century A.D. by placing the commencement of the Sātavāhana rule in 27 B.C., identifying Satakarņi of the HathiGumphā text with the third king of the Sātavāhana dynasty' and relegating King Bșhaspatimitra of Magadha to a neo-Mitra dynasty.
This chronological conclusion may derive support from a few collateral evidences, wbich are as follows:
(1) The style of the Hāthi-Gumpha text is closely similar and slightly anterior to that of the Milinda-Pañha which is one of the extra-canonical! Buddhist texts quoted by Buddhaghoşa in bis commentaries (ante, pp. 172), which, in its turn, presupposes all the Pāli canonical texts as we now have them,—the texts which were committed to writing during the reign of King Vattagāmani towards the close of the 1st century B.C., and which according to a tradition embedded in it, was not compiled earlier than 500 years from the Buddha's demise,- an event placed now-a-days in 483 or 484 B.C.: The Milinda-Papiha was compiled when the memory of King Milinda-Menander was still green in India with that of his courtiers Anantakāya-Antiochus and Damayanta-Demetrios, his capital Sāgala or sākala, and his birthplace Alasandadipa (Alexandria). The political position of the Greco. Bactrian king is well brought out in a passage in which he is represented as comparing himself to a lion in a golden cage surrounded by many enemies. 4 Just as the authenticity of the tradition about the compilation of the Katbāyatthu as a Pāli canonical text in the 18th year of Asoka's reign and 236 years after the Buddha's demise 6 may be proved by the close
1. According to the Parāṇas, Simuka, the first Sātavahana king, reigned for 10, Krsna, the second king, for 10 or 18, and Sri-Satakarņi, the third king, for 10 years.
2, 3. Geiger's Translation of the Mahāvamsa, Introduction.
4. Trenckner's Milinda-Patiha, p. 88 : Seyyathä pi scho migarājā suvanna-pañjare pakkhitto bahimukho yeva hoti, evam eva kho' ham kiñcapi agāram ajjhāpasāmi bahimukho yeva acchāmi.......bahu me paccatthikā.
5. Trenckner's Milinda-Pañiha, p. 13; Buddhaghoşa's Atthasālipi, p. 6, wrongly places the event 218 years, while the Mahāvamsa (V. 278-280) rightly places it 236 years, after the Buddha's demise.

Page #312

284
OLD BRĀHMI INSCRIPTIONS
resemblance, among other evidences, 1 between the most peculiar dialectical style of the Kathāvatthu and that of Asoka's R.E. IX, so, perhaps, the authenticity of the tradition about the compilation of the Milinda-Pañha some 500 years after the Buddha's demise may be proved by the observed close resemblance of its style with that of the Hāthi-Gumphā text. As regards the similarity of style between the Kathāvatthu and Asoka's R.E. IX, the following quotations may here suffice :1. Katbāvatthu, 1.6—Atthi siyā atītam siyā nvātītanti. Hañci
atītam atthi, althi siyā atitam siyā nvātilam tendtītam nvātitam
nvātītam atitanti, etc. 2. Asoka’s R. E. IX—Siya va tam atham nivateyā siyā punā no
hidalo kike ca vase. Tyam puna dhamma-magale akālikye. Hamce pi tam atham no nite ti hida atham palata anamtam punā
pasavati. Hamce punā tam atham nivate ti, etc. (2) The invocation formula, Namo arihamtānam namo sava-sidhānam, used at the commencement of the Hathi-Gumphā text is closely similar to and somewhat simpler than the formula, Namo samma-samouddhānam paramatthadassīnam silddigunapāramippattānam, occurring at the commencement of the Petakôpadesa. It may be noted that the Sātavāhana inscriptions have mostly Siddham for their benedictory formula. Another form of the convention characterising such works as the Netti-Pakarana, the Milinda-Pañba and the Kauţiliya Artha-Sāstra is the epitome of the text presented in one or more verses. This twofold convention 3. which became a universal phenomenon in Indian literature dating from
1. Kathāyatthu, I. 3, stating : Majjhimesu janapadesu atthi brahmacariyavāso, Puccantimesu janapadesu n'atthi brahmacariyavāso Milakkhesu aviññātāresu. yattha n'atthi gati bhikkhūnam bhikkhuninam upāsakānam upāsikānam. 2. Netti-Pakaraṇa, p. 1 :
Yam loko pūjayate salokapālo sadā namassati ca
Tass'eta sasanavaram vidūhi neyyam naravarassa | Milinda-Panha, p. 1:
Milindo nāma 80 rājā Sāgalāyam puruttame
Upagañchi Nāgasenam Gangā va yatha sāgaram | Artha-Sāstra, p. 6:
Sukha-grahana-vijñeyam tattvärthapada-nis citam
Kautilyena krtam śāstram vimuktegrantha-vistaram || 3. The rule that a treatise should be commenced with either benedictory invocation or the specification of the thesis (asirnamaškriyā vastu-nirdéso vāpi tanmūlam) which is prescribed in the Kivya-Prakāśa and other Indian works on poetics must have been sug. gested by the twofold convention that came into vogue.

Page #313

NOTES
285
the 2nd or the 3rd century A.D. cannot be traced either in Indian literature or in Indian epigraphy prior to the beginning of the Christian era. The Petakôpadesa and the Netti-Pakaraņa are the two extra-canonical Pāli companion works ascribed to Mahākaccāna. Of them, the former may be shown to have been quoted by name in Buddhaghoşa's Atthasālint,' and the latter may be shown to have been presupposed by the Milinda-Pañha and Buddbag hoşa's Atthasālini. Prof. Hardy, the editor of the Netti-Pakarana, fixes the 2nd century A.D. as the date of composition of this work. We have taken these two Pali works with the ArthaŠāstra because, in our opinion, the Artha-Sastra, as we now have it, cannot be dated earlier than the 2nd century A.D. The concluding chapter of the Artha-Šāstra deals with 32 Tantra-yuktis representing the terminology of textual methodology. Curiously enough, a precisely similar treatment has been accorded to the same number of Tantra-yuktis in the Susruta-Samhita (Uttaratantra, Ch. LXV) which is, in its extant form, a compilation of the 2nd or the 3rd century A.D., while the total number of the Tantrayuktis has been increased to 34 in the Caraka. Samhitā.3 So far as Pāli literature is concerned, the Netti-Pakaraṇa and the Petakôpadesa are the two well-known treatises on textual methodology. It will be noticed that the twofold convention combined in the opening verses characterises also the Mādhyamika-Kārikā of Nāgārjuna who was associated with one of the Sātavāhana rulers :
Anirodham anutpādam anucchedam aśāśvatam Anekártham anānártham anāgamam anirgamam # Yaḥ pratitya-samutpādam prapañcópašamam sivam Desayāmāsa sambuddhas tam vande vadatām varam #
(3) The Barhut stone-railing which is a purely Suriga architecture bears two sculptural representations of Asokan pillars. 4 There are instances of imitation of Asokan pillars in Sunga architecture among the ruins of Buddhist monasteries and monuments at Sarnath. The pillars
1. Atthasālini, p. 165.
2. See the characterisation of saddha in the Netti (p. 28), the same in the Milinda (pp. 34 foll.), and the quotation of both in the Atthasalini, pp. 119.120. See, also, how the simple characterisation of sati by apilāpana-lakkhanā satîti in the Netti has been elaborated in the Milinda (pp. 37-38).
3. Caraka Samhitā, Siddhisthāna, Ch. XII. 4. Cunningham's Mahâbodhi, Pl. III.

Page #314

286
OLD BRAHMI INSCRIPTIONS
with octagonal shafts and Asokan capitals as typified by the pillars of the Barhut E. Gateway represent the final stage in the development of the Sunga style of architecture. The old stone-railing at Buddha-Gaya, mostly donated by the queens of the later Mitra dynasty, bear sculptural representations of the typical Sunga pillars at Barhut, while the typical Buddha-Gaya pillars go to show a considerably modified form of the Sunga pillars at Barhut.2 The shrine-posts on the hills of Udayagiri and Khandagiri, ornamented according to their description in the HathiGumpha text (I. 15), alternately with two different flower-designs, indicate a wide departure from the Sunga style established at Barhut.
Thus we dispose of the vexed question of Kharavela's date, remembering all the while that Indian chronology is a house of cards, which may break down at any moment. But it cannot be doubted that the sun of the royal power of Kalinga reached the zenith during the reign of King Kharavela, though the light which dazzled the eyes was destined to set for ever. The warrior-like spirit of Kharavela and his bold undertaking of military campaigns all over India clearly prove that militarism was in full vigour in the country in spite of Aśoka's advocacy of the ideal of conquest by the dhamma. But what has been the final result of the wars and warfares that served to keep Kharavela ever busy and the people of Kalinga always in excitement? The final result has been this that Mahāvijaya Kharavela disappeared completely out of sight after the 14th year of his reign and the Meghavahana dynasty came to an end within a few years of his death. The fateful career of Kharavela is enough to prove that the arms that hurl missiles may strike terror, but the arms that embrace that conquer for good. It is Asoka who set up the ideal of conquest by the dhamma. Kharavela upset this ideal only to be forgotten even in the literature of the Jains, while the memory of Asoka has all along been adored by the whole Buddhist world. The sigh of monarchs found its fitting expression in the verse singing the death-ode of the royal state and the everlasting glory of the nobler deeds of men :
Jiranti ve rajaratha sucitta, atho sariram pi jaram upeti
Satan ca dhammo na jaram upeti, santo have sabbhi pavedayanti || 3
1. Canningham's Mahabodhi, Pl. IV, two square panels.
2. Cunningham's Mahabodhi, Pl. IV, the pillar in the middle.
3. Dhammapada, Jaravagga. Such is indeed the moral or central idea of the Maha. vamaa which is the greatest known epic chronicle of the kings of Ceylon, and of all the epics of the Hindus,

Page #315

NOTES
287
"The painted royal chariot falls indeed into decay ; even the body (which passes as one's own) is sure to approach the same fate. The noble tradition of the virtuous alone does not go to decay. This is indeed the proclaimed opinion of those of tranquil heart who discuss with good men." Kharavela, the sovran lord of Kalinga, was wise enough to beware of his royal state betimes and to take steps, when opportunity occurred, to build the costly works of art and architecture in glorification of his religion. And it is the lingering rock-cut caves on the hills of Udayagiri and Khandagiri which have immortalised him and raised up the people of Kalinga in the estimation of civilised humanity.
15. THE CITY OF KALINGA
Kharavela's inscription is not without a few interesting hints as to the plan, picturesqueness, internal life, joy and prosperity of the city of Kalinga which was the capital of his Kalinga kingdom.
As regards the plan and picturesqueness, we have mention, in the firstyear's record (I. 2), of (1) gopura (gate-house), (2) pākāra (wall), (3) nivesana (residential building), (4) taḍāga (tank), and (5) uyana (garden) as features associated with the city of Kalinga. As to other features, there are allusions, in the ninth year's record (I. 10), to (6) raja-nivāsa Mahāvijayapāsāda (the royal residence, the Great-victory palace); in the twelfth year's record, to (7) vithi (road), (8) catara (square), (9) palikha (gate-bar), and (10) sihara (tower), besides gopura (gate-house); and in the fifth year's record (I. 6), to (11) panāḍi (canal). A few terms preceding vithi and denoting other features are now missing from the twelfth year's record.
So far as the twelfth year's record goes, the terms withi, catara, palikha, gopura and sihara are all used in plural forms (vithi-catarapalikhani gopurani siharāni). We may with good reasons maintain that the terms gopura, pākāra, nivesana, tadaga and uyana have been used in a plural sense also in the first year's record. The plural sense of the first three terms is not inconsistent with the copulative compound gopura-pakara-nivesanam. The plural sense of tadaga may be easily derived from the compound taḍaga-padiyo which is a plural expression. The plural sense of uyana is conveyed by the pronominal adjective sava which is the first member of the compound savuyana (sarvódyāna). it may be established that the city of Kalinga, even as King Kharavela found it at the time of his anointment, contained many a gopura, many a
Thus

Page #316

(1) gopura
(2) taḍāga
(3) uyāna
(4) vithi-catara
(5) nivesana-sihara
288
OLD BRAHMI INSCRIPTIONS
But as
pākāra, many a nivesana, many a tadaga, and many a uyāna. regards Mahavijaya-pasada and panadi, the implied sense is singular.
The general features implied by these terms go to show that the city of Kalinga was built, even before the reign of Kharavela, on the same plan, more or less, as other Indian cities, e.g., the city of Sagala of which we have a vivid description in the Milinda-Pañha (Trenckner's edition, pp. 1-2). It may be inferred from the hints given in Kharavela's inscription that the palace used as the royal residence was the main centre of interest in the city of Kalinga, precisely as in other cities; that the residential buildings were all inside a city-wall provided with gateways and gate-bars, and themselves were provided with enclosures and gate-houses; that the various gardens, parks and groves added to picturesqueness of the city; that the roads, squares, court-yards and canals facilitated easy communication and intercourse; that the deep and cool tanks (gabhira-sitala-tadaga) served as reservoirs of water for drinking, bathing and washing purposes, while the spire-temples towering above the house-tops stood out as various centres of religious worship.
Comparing the plural expressions in Kharavela's inscription with those in the Milinda description of the city of Sagala, we cannot fail to notice that each of them falls short of the expected expression by one or two words, as the following table will indicate :
K. I.
M.
gopura-torana talaka-pokkharani
ārāma-uyyāna-upavana
vithi-caccara-catukka-singhataka
antarápāņa-vividha-dānaggasataHimagiri-sikhara-saṁkāsa-vara
bhavana
Reading between the lines, it is easy to understand that the purpose of Kharavela's inscription was far from giving a connected description of the city of Kalinga, and that the features implied by the additional words in the Milinda description were meant to be covered by the plural sense of the words in the former. We say that where gopura in the sense of gate-house or gate-tower is employed, torana in the sense of gate or gateway is implied there. Similarly where tadaga in the general sense of tank is

Page #317

NOTES
289
employed, pokkharaņi in the sense of small tank need not be separately mentioned. The Hāthi-Gumphā expression savúyāna, interpreted in the sense of all gardens or all kinds of garden, may be said to include in it both ārāma (park) and upavana (grove). As regards vīthi and catara, it is not unlikely that some term corresponding to the Pāli catukkasinghātaka meaning cross roads bas vanished with the missing words preceding vīthi. Anyhow, where vithi in the general sense of road is employed, catukka-singhātaka in the sense of cross-roart need not be separately mentioned. And so as to the varieties of nivesana-sihara.
For details of the plans of Indian cities, the reader is referred to Dr. B. B. Dutt's “ Town-planning in Ancient India.” As for the technical architectural significance of the above terms, Dr. P. K. Acharya's "Dictionary of Hindu Architecture” is sure to prove to be the best help. It will be evident from Dr. Acharya's quotations that gopuras in the sense of gate-houses or gate-towers were not peculiar to religious edifices; but these formed the paraphernalia of religious temples, as well as of resi. dential buildings. It is equally manifest from Dr. Acharya's article on Prāsāda that the Great-victory-palace as a literal rendering of Mahā. vijaya-pāsāda in Khāravela's inscription does not bring out the technical architectural significance of the term. The buildings of the Vijaya class were all two-storeyed. His quotations from the literary texts and the inscriptions make it clear that sihara or sikhara as a tower or turret was as much a crowning construction of a spire temple as that of a palatial building. Nevertheless, the terms gopurāni and siharāni, as used in the twelfth year's record of Khāravela's reign, would seem to be associated with certain religious temples within the city of Kalioga.
As for the existence of temples dedicated to various deities, we have to look just into the concluding paragraph of the Hāthi-Gumphā inscription (I. 16) in which Khāravela has been praised as sava-deváyatanasamkāra-kāraka, " the repairer of all abodes of the gods.” Devayatana is a technical term, the significance of which may be made clear from Dr. Acharya's quotations sub voce--Āyatana, Devayatana and Devālaya. It is not difficult to understand that the so-called abodes of the gods' were in reality nothing but the Hindu shrines which stood in the name of different deities or the Hindu temples in which the images of different gods, demi-gods, goddesses and demi-goddesses were enshrined for worship by the people. These shrines and temples were to be located, as Dr. Dutt's book will show, in every Indian city. We shall entirely miss the force of sava (all) in the expression sava-devd yatana, “all abodes of the
37

Page #318

290
OLD BRAHMİ INSCRIPTIONS
0
gods," not taking it to denote all kinds of shrines and temples : those in which there were images and those in which there were no images.
If it can thus be proved that the Häthi-Gumpbā text clearly refers, in its concluding paragraph, to the existence of “Deva-temples," we may be justified in thinking, first, that the terms gopura and sihara in its twelfth year's record were intended to be interpreted as well in connection with nivesana and pārāda preceding them as with devayatana following them; and secondly, that the worship of idols in the Hindu temples and shrines had come into existence in Kalinga before the reign of Khāravela, and long before the Buddhists made the Buddha-images for worship during the Kuşāna rule,
With regard to the part played by Kbāravela in the building up of the city of Kalinga, we read in the Hātbi-Gumphā text (I.2) that imme. diately after his consecration, in the very first year of his reign, he spent 35,00,000 (pieces of money) in thoroughly repairing the gate-houses, walls and residential buildings damaged by stormy wind in his capital, in raising up embankments of the deep and cool tanks, and in restoring all the gardens. From this it is clear that his first year's work was just a work of reparation and restoration. This record clearly proves that the royal city of Kalinga was bedecked with many gardens and many kinds of gardens ; that it shone forth with its high walls, gate-towers and buildings; and that it abounded in the deep and cool tanks serving as reservoirs of good drinking water and as bathing places. So far as these tanks go, we may say that the kingdom of Kalóga is remarkable in its modern identity precisely as it was two thousand years ago.
King Khāravela did not, however, stop at the work of reparation and restoration. The extension of Nanda-king's canal from the Tanasuliya or Tanasuli road into the heart of the city was a costly work, which was accomplished by him in his fifth regnal year as a meaus of facilitating communication and irrigation, among other advantages. The two-storeyed new royal residence known by the imposing name of “Great-victory palace" and decked with beryl work, for which he is said to have spent 38,00,000 (pieces of money), was, undoubtedly, a very costly addition made by him to the city architecture. But even this was not all. The
1. The architectural significance of the expression veduriya-Mahā-vijaya-pasāda may be understood in the light of veluriya.phala-santhata pasāda in Fausboll's Jātaka, Vol VI, p. 279.

Page #319

NOTES
291
twelfth year's record goes to show that he made a good deal of im. provement by constructing new roads and squares, by strengthening the gates with new burs, as well as by setting up new gate-houses and towers.
Now, with regard to the internal life, joy and prosperity of the city of Kalinga, it may suffice to observe that there are allusions, in the second year's record (1, 3), to (1) the location of multitudinous fighting army, consisting of all the four divisions ; in the eighth year's record (I. 9), to (2) triumphal processions of victory characterised by the carrying of the Wishing-tree at the head and by the marching of the four divisions of the army at the back, as well as to (3) the feasting of all sections of the people, the general population, the officers, the religious sects keeping to household life, the Brahmanical ascetics and the Jain recluses ; in the third year's record (I. 4), to (1) the entertainment of the citizens by combats or comics, by songs, dances, musical performances, as well as by feasts, festivities and joyous gatherings, the king himself being represented as an expert in the science of music ; in the seventh year's record (I. 8), to (5) a pompous parade of swords, umbrellas, flags, guards and horses, as well as to (6) the performance of all ceremonies of victory; in the sixth year's record (1. 7), to (7) the wholesale remission of taxes and duties and the dispensing of many bounties as a display of the royal fortune ; and in the twelfth year's record (I. 13), to (8) the receiving of jewels, antelopes, horses and elephants, as well as the elephants, horses and other animals capable of exciting curiosity (abhutam-achariya-hathisapasava) as presents from the Vāsukis, and no less to (9) the supply of pearls, gems, jewels and rich apparels as tributes from the king of Pandya ; and in the eleventh year's record (I. 12), to (10) the paying of homage to the former kings of Kalinga as a great public and social function.
The description is apparently incomplete. Nevertheless, it is full enough to portray a vivid picture of the joyousness of a happy, active, prosperous and crowded city. Here just one point calls for remark. In regard to the presents from a hundred Vasukis (1. 10), we have these two statements, (1) abhutamachariyan hathísa-pasavam pariharamti, and (2) miga-haya-hathi upanāmayamli, which would have been mistaken as overlapping. But from the adjective abhuta-achariya, "curious and wonder. ful," characterising the elephants, horses and such other beasts in the first clause, it appears that these were collected and preserved as curios, and reared in the royal enclosure for exhibition.

Page #320

292
OLD BRAHMI INSCRIPTIONS
16. THE CAVES, SHRINES AND PILLARS
Kharavela's hard-earned fame as a builder was not confined to the repairing of the city of Kalinga and the improvement of the same by the rebuilding of embankments of the deep and cool tanks, the restoration of all the gardens, the extension of the Nandaraja canal, the erection of Mahavijaya-pāsāda as a new two-storeyed and beryl-set palace, and the addition of new roads, squares, gate-bars, gate-houses, and towers. And we may note that his religious endowments were not exhausted by the repairing of Hindu temples and the occasional feasting of the Brahmins and Jain recluses.
The Hathi-Gumpha and other old Brahmi inscriptional records go to prove that he showed his royal munificence to the professors of his own faith, namely, the Jain saints and recluses who resided on the Kumari hill, in causing one hundred and seventeen caves (satadasa-lena-salam) to be made as joint excavations of himself, his queens, his sons, his relatives, his brothers, and his officers, sharing the merit and fame with the rest of the pious donors and glorifying the tradition of Jainism with the most ancient known landmarks of its art and architecture. All of these 117 caves were intended to serve as resting places of the resident Jain saints and recluses (Arahato parinivasato hi kāya-nisīdīyāya).
Apart from what he accomplished jointly with other excavators of the Jain cave-dwellings on the Kumari hill, the Hathi-Gumpha inscription clearly shows (I. 15) that he caused to be excavated under his own auspices and as the crowning glory of the recorded last year of his reign, one cave for the accommodation of the venerated (Jain) recluses and the (Hindu) yatis, hermits and sages visiting the place from a hundred directions. His last recorded munificence, amounting to seventy-five hundred thousand pieces of Indian money current at that time in his Kalinga kingdom, sufficed to enable him to make, along with the excavation of the last-mentioned spacious cave, a number of stone-pillars and shrines on a slope in the neighbourhood of cave-dwellings of the Jain saints and recluses and by means of some hundred thousand slabs of stone, quarried out of excellent quarries extending over several leagues (Arahato nisidiyasamipe pahhare varakara-samthapitāhi aneka-yojana-āhitahi silahi silathambhani ca cetiyāni kārāpayati).
The same amount just sufficed also to cover the cost of erecting an edifice, providing it with a canopied court-yard adorned with a pillared

Page #321

NOTES
293
beryl-hall, as well as of causing one-half-hundred shrine-posts, ornamented with the alternate settings of beryl and emerald, to be produced alongside (patalika-catare ca veduriya-gabhe thambhe patithāpayati panatariya-satasuhasehi, veduriya-nila-vochimnam cera-yathi-adhasati kam tiriyam upādayati).
With regard to the 117 caves jointly excavated on the Kumārī hill by King Khāra vela and others, the following questions are apt to arise here: (1) Why all of them were not inscribed ? (2) What was the system of counting them ? (3) What has befallen the missing caves ? (4) How to account for their modern names ? (5) What are the component parts of a cave? (6) What is the technical significance of the term lena ? (7) What are the purposes that these caves in particular were intended to serve ?
First, the question as to why all the 117 caves were not inscribed has been discussed at some length in connection with the problem of the relative total of the caves and inscriptions. We have sought hitherto to maintain that the caves excavated by the donors other than Khāravela himself were labelled with inscriptions, recording in each of them the name of the person or persons by whom the particular cave or group of caves was excavated or the component parts of a cave were donated, while in the case of the caves excavated by King Khāravela, as well as of other works of art and architecture done under his auspices, we have a departure from the general rule in that these, instead of bearing a separate inscription each, are all collectively referred to in the records of a single inscription, namely, the lengthy inscription of Khāravela incised on the hanging brow of the projected roof of the Hathi-Gumphā on the present hill of Udayagiri. We have also pointed out that the thirteenth year's record in this inscription alludes in a general fashion to the caves excavated by his queens, sons, brothers, relatives and officers, in which case the engraving of separate inscriptions would have been superfluous were it not for keeping them distinct from His Majesty's own works, and no less for eatisfying the natural but legitimate desire of the various donors to perpetuate their memory and offering an incentive to others to similar acts of piety.
Secondly, as to the system of counting the caves, the general principle seems to have been to count each of the one-storeyed caves consisting of one or more cells or chambers confronted by an open or pilla red verandah as one cave, and to count each such suit on each floor of a two-storeyed construction as one cave, e.g., in the example of the Mañcapuri group of caves representing a two-storeyed construction, the suit in the upper storey was labelled by an inscription recording it to be a cave excavated by Khāra. vela's chief queen, and the corresponding suit in the lower storey was

Page #322

294
OLD BRAHMI INSCRIPTIONS
similarly labelled by a second inscription recording it to be a cave excavated by King Kadampa-Kudepa. A puzzling complication has arisen from a third inscription incised over the doorway of a side chamber on the southern side of the verandah of Kadampa Kudepa's cave and recording it to be a cave excavated by Prince Vaḍukha-Varikha. Bat it is somewhat difficult to decide whether the side chamber confronting the verandah of Kadampa-Kudepa's cave or a separate chamber provided with some sort of a verandah and situated on the southern side of the open enclosure or courtyard in front of Kadampa-Kudepa's cave was intended to be recorded in the third inscription as Prince Vaḍukha- Varikha's cave. In discussing the problem of the relative total of the caves and inscriptions, we have sought to maintain that although the third inscription was incised over the door-way of the former, it was really intended to refer to the latter, that is, to the suite on the southern side of the court-yard, and that the object of engraving the third inscription over the door-way of the former was to draw the notice of the visitor or pilgrim entering the verandah of Kadampa-Kudepa's cave and moving towards the right in peeping into the three chambers including that on the south side with the possibility of coming out without minding what was on the south side of the courtyard. If this argument be sound, we may make bold to say that the system was to count each suit of one or more chambers, whether in one line or not, but surely confronted by some sort of a verandah, as one cave.
Thirdly, as to the fate of the missing caves, we cannot but be astonished that so many of the caves on the Kumari hill have vanished beyond recognition. Sir John Marshall says that taken together, the two groups of caves on the hills of Udayagiri and Khandagiri, "comprise more than thirty-five excavations." But we are not, as yet, aware what method he has adopted in counting the total of the surviving caves, and whether the number suggested by him includes the four or five caves which have suuk down showing still their roofs above the ground on two sides of the Udayagiri or not. We may, once again, draw the reader's attention to Mr. Stirling's paper in the Asiatic Researches, Vol. XV, containing, as it does, all that we yet know of the miserable fate that overtook the missing caves,
1
As for the names whereby the caves are known at present, it is obvious that we require more explanations than one to account for them. In the
1. Manomohan Ganguly, in his Orissa and Her Remains, p. 81, has produced the following list of caves on the two hills (1) Hathi-Gampha, (2) Vaikunthapura,

Page #323

NOTES
295
case of such names as Hāthi-Gumphā (the Elephant-cave), Chota-HāthiGumphā (the Small Elephant-cave), Vyāghra-Gumphā (the Tiger-cave), Sarpa-Gumphā (the Snake-cave), Ajagara-Gumphā (the Python-cave), and Ben-Gumphā or Bheka-Gumphā (the Frog-cave), we find that all of them were suggested by the general shapes of the caves designated by them, particularly their front views creating appearances of the foreparts of the animals or creatures indicated in them. It is very striking indeed that the front view of the cave known by the name of Hāthi-Gumphā suggests the appearance of just the forepart of a royal elephant sitting majestically on its front legs. Such an appearance was appropriate to the cave which was to bear on the hanging brow of its projected roof the famous inscription of His Graceful Majesty King Khāravela. Similarly, the appearance of a small elephant was appropriate to the Chota-Hāthi. Gumphā which was to bear an inscription of Prince Vaduka-Varikha's.
The most remarkable is the front view of the Vyāghra-Gumpbā sug. gesting the appearance of the head of a tiger with its distended jaws. No one can mistake that the front view of the Sarpa-Gumphā suggests the appearance of the upraised large hood of a cobra, or that the front view of the Ajagara-Gumphā suggests the appearance of the gaping mouth of a python. In the same way, none can mistake that the front view of the Ben-Gumphā or Bheka-Gumpbā suggests the appearance of the open mouth of a frog.
In the absence of all of the caves it is difficult to say if there were not, amongst them, some caves with their front views suggestive of the appearances of animals other than the elephant and the tiger, and of creatures other than the cobra, the python and the frog. One fact, however, is certain that one can still trace shapes of the same creature, whether it is snake or frog, in a set of two or more caves, say, for instance, in the roofs of four or five caves that have sunk down on two sides of the hill of Udayagiri.
(3) Pātālapura, (4) Mañcapuri (Yamapura ?), (5) Haridās Gumphā (Chota-Häthi-Gumphā), (6) Sarpa-Gumphā, (7) Vyāghra-Gumphā, (8) Ganesa-Gumphã, (9) Jagannāth-Gumphā, (10) Svargapuri or Alakāpuri, (11) Rāni-Gumpha, (12) Jaya-Vijaya, (13-20) eight nameless, these being the caves on Udayagiri ; (21) Ananta-Gumphã, (22) Navamuni, (23-24) Sātbakhrā or Sātgharā group of two caves-the Barbhuji and the Trisula, (25) supposed cave of Lālatendu, (26-27) two nameless, these being the caves on Khaņdagiri. Be it noted that one of Ganguly's nameless cave on Udayagiri is known by the name of Jambesvara or Ben-Gumphā, and that among the caves on Khaqdagiri, one is known by the name of Tattva-Gumphā I, another by that of Tattva-Gamphi II, a third by that of Durg&-Gumphä, and a fourth by that of Hanumān-Gumph&.

Page #324

296
OLD BRĀHMI INSCRIPTIONS
If an inference can at all be drawn from the last mentioned fact, it will certainly be this that the underlying motive in the fashioning of the animal and creature shapes of the rock-cut caves was to represent the principal denizens of the Kumāri hill and to limit these shapes to them. Anyhow, the elephants, tigers, cobras, pythons and frogs are still the chief denizens of the two hills of Udayagiri and Khandagiri.
We are unable to unravel the mystery of the Frog-cave being otherwise known as Jambesvara-Gumphā. But it is certain that the ChotaHāthi Gumpbā came to be otherwise known as Haridas-Gum phā on account of the fact that this cave was once tenanted by a Vaişņava ascetic known by the name of Haridās.
Among the remaining names, Ganesa-Gumphā may engage our first attention. This name which, like Haridās-Gumphā, is of Hindu origin, must have been suggested by the figures of two calves of elephants set up in the court-yard in front of the cave concerned. It was easy for the neighbouring Hindu inhabitants to associate these figures with Ganesa, the elephant-faced god of the Hindu pantheon.
The name of the Ananta-Gumphā on the Khandagiri hill appears to have been suggested by the figures of the pairs of crawling tripleheaded snakes lying over the arch-fronts of the cave which are ornamented with various reliefs containing, among others, one standing figure of Gaja-Lakşmi, the Hindu goddess of Luck. A Vaishnavite must have been easily led by their association with a figure of Lakşmi, the consort of Vişņu, to identify the trip le-headed snake with the famous Ananta or seşa-nāga of his legend. Thus it may be proved that the name AnantaGumphā, too, was of Hindu origin, and that it must have come into existence when the Vaishnavite Hindu ascetics occupied some of the Udayagiri and Khandagiri caves.
A more convincing proof of the Hindu origin of some of the names can be adduced from the modern designations of three caves forming a group on the Khaņdagiri hill. Of these three caves, one is known as Navamuni on account of the fact that the figures of nine Hindu sages were set up on the walls of the cave in parallel with those of twenty-four Jaina Tirthařkaras set up, as recorded in a medieval inscription of the 6th or 7th century A.D. or of still later period in one of these caves, by King Udyota-Keśari of the Kesari dynasty ; another is known as Durgā-Gumphā on account of the fact that the figures of the Hindu goddess Durgā came to be set up at the entrance of the cave containing on its walls in the interior the figures of twenty-four Tirtharkaras set up by

Page #325

NOTES
297
King Udyota-Kesari in a row; and the third is known as Hanumān. Gumphā on account of the fact that the figures of Hanumān of the Hindu legend came to be set up in a similar manner in the cave contain: ing the figures of the Tirthankaras set up by the same king on its walls. Here we may add that the name Tattva-Gumphā whereby two of the caves on the Khandagiri hill are designated can as well be treated as being of Hindu origin. The Vaişņava legend of Jaya and Vijaya as two attendants of Vişyu may be held responsible for the name JayaVijaya designating one of the surviving caves on the Udayagiri hill, although there is no difficulty to regard this, along with Tattva, also as a Jain name.
At all events, there can be no doubt that the name Mañcapuri designating a group of three caves, the name Alakāpuri designating one of the surviving sculptured caves on Udayagiri and the Rāņi-Nür or Rāņi-Gumphā designating another richly sculptured cave on the same hill, were of Jain origin, although we are unable to say wben these came actually into existence. The tradition about the Rāni-Gumpha is that it came to be known as Rāņi-Gumphā on account of the fact that a queen of the Kesari family of kings lived in it.
The idea of naming the cave of Khāravela's chief queen in the upper storey of the Mañcapuri group as Vaikunthapura (the Paradisiacal), of the cave of King Kadampa-Kudepa in the lower storey as Pātālapura (the Infernal), and the side cave of Prince Vadukha-Varikha on the ground floor as Yamapura (the Hidden) must have been suggested by the peculiarity of their relative position.
We have no mears, as yet, of ascertaining whether these caves bore any distinct names as at the present time, and if so, by what names before and immediately after the reign of the Keśari kings. But it may he noted here that in Udyota-Kesari's inscription recording the fact of installation of the images of twenty-four Tirthankaras in the caves now known as Navamuni, Durgā and Hanumān, the caves have not been referred to by any name.
Now with regard to the component parts of a cave (lena) considered as a distinct structure of architecture, we have just a few hints from the inscriptions of Cūlakamma (Nos. VI. VII) and the single inscription of Kamma and Khiņā (No. VIII). In the inscriptions of Cūlakamma, we find that the structure has been represented in terms of its component parts instead of being denoted by a term, such as leņa, representing the whole of it. Kothājeyā and pasāda or pasāta are the two expressions that
38

Page #326

a cave.
298
OLD BRAHMI INSCRIPTIONS
have been employed in these records to denote the component parts of a cave-structure.
So far as the term kotha goes, there is little doubt that it was intended to denote the cells or chambers' hollowed out at the back and at the end,' and that the term pasada or pasata was used to denote the open or pillared verandah presenting the appearance of the front view of a palace or building, of a façade as one might say.
The difficulty lies in clearing up the mystery of kothajeya as to whether it was to be a combination of the two words, kotha and ajeya (impregnable, unconquerable), treating the latter as an adjective of kotha, or it was to be a compound of the two words, kotha and jeya, treating the latter as a term denoting altogether a separate construction within
Seeing that in the inscription of Kamma and Khina, the two terms, koṭhā and pasāda, have been used without any word suggestive either of ajeyā or of jeyā, we may be justified in treating jeya even as a distinct technical term.1 If so, we have to decide what particular construction was intended to be denoted by it. Dr. P. K. Acharya's Dictionary of Hindu Architecture makes it clear that jeya in the sense of 'pent-roof' is a technical term, which may be equated with the Prakrit jeya. One may notice that in some of the Udayagiri caves, there is a pent-roof in the form of a small projection from the side-wall of the verandah and a little below its main roof. It is possible that this small projection was intended to be denoted by jeya in Culakamma's inscriptions.
As to the technical significance of the term lena, we are aware that each of the cave-dwellings on the Kumari hill which is designated lena in the old Brahmi inscriptions is known now-a-days as gumpha, a term equating with the Prakrit gumpha, the Asokan kubha, and the Sanskrit and Pāli guha. Guha has been distinguished in Pali literature as mattika-guhā (earthen cave) and giri-guha (mountain-cave). In the Barabar Hill cave inscriptions of King Asoka and the Nagarjun Hill cave inscriptions of King Dasaratha, the term kubha or guha has been employed to designate certain cave-dwellings of the Ajivikas prepared by dressing up the mountain-caves and rocky dens of animals, polishing their walls and
1. The text is-Kammasa kotha ca Khinaya ca pasado.
2. The term pabbatānam guha (cave formations in the mountains) is explained in the Maha-Niddesa (p. 466) as signifying (1) Kandară (caverns), (2) giri-guha (mountaincaves), and (3) pabbata-gabbhara (hollowed depressions in the mountains).

Page #327

NOTES
299
ornamenting their entrances with arches serving to lend to them the appearance of chapels. These cave-dwellings of the Ājivikas have neither any verandah nor any properly shaped chambers hollowed out in the interior, and still bears the form of natural cavities and animal dens. These ancient cave-dwellings with their arch-entrances represent the ruder form wherefrom the later Buddhist cave-temples, with their infinitely superior workmanship and ornamentation in the interior, may be said to have developed. And with these ancient Maurya cave-dwellings may be sharply contrasted the caves excavated on the Kumāri hill by King Khāravela and his coadjutors for the accommodation of the resident Jain saints and recluses as representing altogether a different architectural design and artmanship.
It is evident from their component parts mentioned in some of the old Brāhmi inscriptions, that each of the caves on the Kumāri hill was so designed as to make it appear like a regular cottage or residential building with an open or pillared verandah (pasāda) in front of one or more chambers (kotha) dug out in the interior in the shape of rooms, the verandah being provided, in some examples, with a pent-roof (jeyā). These cottage-like or palace-like cave-dwellings of the Jain saints and recluses may be judged with their peculiar roofs as Orissan patterns of the Buddhist monasteries at Sarnath and other places.
Now it remains to be seen whether the literary usages sanction the application of the designation leņa to the kubhā or cave-dwellings of the AjIvikas in the Barabar and the Nagarjuni hills, and the application of the designation gumpha-guhā to the lenāni or cave-dwellings of the Jains on the two hills of Udayagiri and Khandagiri. So far as the Pāli literary usages go, we find that the term leņa or lena has been employed in its generic, as well as in its specific sense. In its generic sense, it has served as a common designation for five kinds of monastic abodes, viz., (1) vihāra, (2) addhayoga, (3) pāsāda, (4) hammiya, and (5) guhā.1 And in its specific sense, it has served as a designation for a particular kind of construction, which is distinguished, in the Vinaya Texts,” from guhā (cave) and tiņakutikā (straw-thatched cottage), and in the Milinda-Panha, 3 from pāsāda (residential building), kuti (cottage) and guha (ordinary cave) on the one hand, and from pabbhāra (cavity with a sloping projection as its covering
1. Vinaya Mahāvagga, I. 30. 4; I. 77; II. 8.1; III. 5; VI. 33. 2; VIII. 7. 1; Vinaya Cullavagga, VI. 1. 2. See Pali English Dictionary, sub voce-lena.
2. Vinaya Sutta-Vibhanga, IV. 1.1; XIX. 1-2. 1. 3. Milinda-Papha, p. 151.

Page #328

300
OLD BRĀHMİ INSCRIPTIONS
roof), dari (grotto), bila (chasm), vivara (hole) and pabbatantara (crevice), on the other. It may be realised at once from the Milinda list that pāsāda, kuti and lena represent human art and architecture, while pabbhāra and the rest represent natural formations. As regards guhā, it may be treated either as a natural formation or partly as a work of human hand and skill. As for a natural cave, cavity or cavern improved by human hand, there can be no objection to applying to it leña as a designation or to applying the designation gumpha or guhā to a leņa in the sense of such a cave. It cannot be denied that the leņa on the Kumārī hill sought to combine in its architectural design the structures of a pāsāda, a guha and a kuti.
Another interesting point is that one of the Vinaya passages, stating, as it does, that a pabbhāra was being cleared for making a leņa (pabbhāram sodhāpeti leņam kattu kāmo), enables us to understand the connexion and difference between the pabbhāra and the leņa, the pabbhāra (HäthiGumphā pabbhāra) which ordinarily carries with it the idea of a slope or projection (ninna, poņa)? conveys also the idea of a pabbbata-kucchi,3 that is to say, of a mountain-cave with a sloping projection as its covering roof. We can say that the lena as an excavation and form of architecture made by human hand is just a pabbhāra in its natural condition. At any rate, no other inference can be drawn from Buddhaghosa's definition of a leņa as “a habitation with an opportunity for sitting and lying, made either by excavating a house in a mountain or raising a wall where a pabbhāra is insufficient (to serve as a dwelling)." 4
Lastly, as to the purposes to be served by the Orissan caves in particular, two years' records in the Hāthi-Gampha inscription state (1) that 117 caves were jointly excavated by King Khāravela and others to provide the Jain saints and recluses permanently residing on the Kumāri hill with shelters for resting their bodies (Kumāri-pavate Arahato parinivasato hi kāya-nisīdīyāya), and (2) that one spacious cave was excavated by King Khāravela himself for the accommodation of the distinguished visitors among the Jain recluses and the yatis, hermits and sages among the Hindu
1. Vinaya Mabāvagga, VI, 15. 1. 2. See, for references, the Pali-English Dictionary, sub voce-pabbhāra.
3. Niddega-Commentary, Siamese edition, Part II, p. 172: pabbata-gabbharati pabbata-kucciyo.
4. Pabbharam khanitvā vā pabbharassa appaho nakatthäne kuddam utthāpetrā tā kata-senasanam. Quoted in the Pali-English Dictionary, sub doce-lena.

Page #329

NOTES
301
ascetics, coming from a hundred directions (sakata-samana-suvihitanam ca sata-disānam yatīnam tāpasa-isinam lenam karayati). In the fourteenth year's record (I. 15), the cave-dwelling on the Kumari hill are collectively designated as Arahato nisidiya, "the Arhata (Jain) seats."
Mr. Jayaswal seeks to maintain on the strength of some later and mediæval Jain authorities that the term nisidiya, as employed in the HathiGumpha inscription, should be interpreted as signifying a "tomb." Accepting his interpretation of the term, we have to understand that King Kharavela and his compatriots excavated the 117 caves on the hills of Udayagiri and Khandagiri to provide the resident Jain saints and recluses with suitable places for entombing or burying their bodies.
We know perfectly well that his explanation of nisidiya, considered as another form of nişaddhi, nisidi, nisadyā, nisadyakā, nisidhiyā, nisidhikā. nisidhyalaya or nisidhigeha, may be justified by the use of the term in several mediæval inscriptions, mostly found in South India, the references whereto have been collected by Dr. P. K. Acharya in his Dictionary of Hindu Architecture. But the question is whether this later medieval meaning can be read into nisidiya of the Hathi-Gumpha inscription or not, whether, if at all, the idea of a tomb can be associated with lena in its generic or in its specific sense.
In the first two Barabar Hill cave inscriptions, King Asoka has not stated the purpose of the caves labelled by them and dedicated to the Ajivikas. But in the third inscription, if Dr. Hultzsch's reading of it be correct, the stated purpose of the cave was to provide its inmates with a retreat during heavy showers of rain (jalaghosagamathati). Anyhow, we have the Nagarjuni Hill cave inscriptions of King Dasaratha to clearly state the purpose of the dedicated caves to be to provide their Ajivika-dwellers with resting places during the rainy season (vaanisidhiya). As regards the thirteenth year's record in the Hathi-Gumpha text (I. 14), the stated purpose of the caves may be taken to be, first, to provide the Jain saints and recluses with permanent residences as may be inferred from the expression Arahato parinivasato, and, secondly to provide them with solitary retreats for rest, bodily as well as mental (kāya-nisidiyāya, here kaya denoting both rūpa-kaya and nāma-kāya as the Buddhists would explain it). We have, moreover, seen that the Pali commentator Buddhaghosa has explained lena as a synonym of sendsana, a term standing for all kinds of retreats of the ascetics and recluses affording an opportunity for lying and sitting. It may also be noticed that the various purposes of the five kinds of hermitages or

Page #330

302
OLD BRAHMI INSCRIPTIONS
monastic abodes, commonly designated as lena, do not include the idea of entombing the bodies. We read in the Vinaya Cullavagga (VI. 1. 3-5 :
Sitam unham patihanti tato vāļamígāni ca Sirimsape ca makase ca sarire cápi vutthiyo il Tato vātátapo ghoro saħjāto patihanñati i Lenatthañ ca sukhatthañ ca jhāyitun ca vipassitum || Vihara-dānam samghassa aggam Buddhena vannitam! Tasmā hi pandito poso sampassam attham attano # Vihāre kūraye ramme vāsay'ettha bahussute !
It is manifest from the quoted verses that the purposes of dedicating a lena to the ascetics and recluses were to provide the learned teachers of religion with retreats with an accommodation for their residence, to provide them with shelters, comforts and suitable places for meditation and introspection, as well as with the means of protection as against heat and cold, wind and sun-light, ferocious animals, reptiles, mosquitoes and showers of rain.
We may understand that all these were precisely the purposes that the Udayagiri and Khandagiri caves were intended to serve, and that in the curious animal or creature shapes of their roofs and front views one can trace the motive of keeping off the ferocious animals and the venomous reptiles.
Turning to the works of art and architecture done under the auspices of King Khāravela in the fourteenth year of his reign, we have to confess that we are incapable of identifying the spacious cave excavated by him as a sort of rest-house for the distinguished visitors among the Jain recluses, as well as for the Hindu ascetics hailing from a hundred directions. The fourteenth year's record (I. 15), stating that some hundred thousand slabs of stone were collected from the excellent quarries extending over many leagues for the erection of stone-pillars and shrines in the neighbourhood of the Jain cave-dwellings on the Kumāri hill, eloquently speaks of the use of stone as an art-material, as well as of the free development of the stonecutter's art in Orissa by the time of King Khāravela.
Although King Asoka is generally credited with the use of stone (pavata, selā-thambha, silā-phalaka) as an enduring art and engraving material, the general consensus of opinion among the Indologists is tending to be in favour of an earlier existence of the stone-cutter's art in India which was called in requisition and brought to perfection by the great

Page #331

NOTES
303
Maurya emperor.1 If King Aśoka can claim the credit for the rock-cut figure of an elephant on Dhavalagiri, the hill of Dhauli, as the earliest known stone-carving in Orissa, King Kharavela can equally claim the credit for employing the stone-cutter's art successfully on an extensive scale for conjuring up all manner of figures even on the roofs of the rock
cut caves.
In connexion with the stone pillars and shrines (sila-thambhāni ca cetiyani), the important point calling for attention is whether the shrines stood apart from the pillars or the pillars and the shrines were combined.
The fourteenth year's record goes also to show that the last great work of art and architecture done by King Kharavela was a pillared beryl hall erected in the centre of a canopied court-yard surrounded by fifty shrineposts, which were ornamented with the alternate settings of beryl and emerald (veduriya-nila-vochimnam ceca-yathi-adhasatikam). We are unable to say where exactly this hall was erected together with the court-yard and the shrine-posts, and whether any traces thereof can yet be found.
With regard to ornamentation of the shrine-posts with the alternate settings of beryl and emerald (veḍuriya-nila-vochimna), we have sought to maintain that here by the beryl and the emerald were intended to be meant two different kinds of designs or floral devices. It is impossible to think that the shrine-posts were inset alternately with two varieties of precious gems in view of the fact that no such shrine-posts or pillars are to be anywhere found among the lingering remains of the Udayagiri and Khandagiri works of art and architecture. On the other hand, there survive some of the shrine-posts with two kinds of floral devices alternating each other on their shafts.2
17. MARSHALL ON CHRONOLOGY OF THE CAVES 3
"Of the early caves along the east coast, the only ones that merit attention here are the two neighbouring and intimately connected groups
""
na
1. If the simile in the Anguttara-Nikaya (Part I, p. 283), Seyyatha pi bhikkhave pāsāne vātena vā, ciratthitika hoti," O bhikkhus, a writing on a stone is not quickly effaced either by the effects of wind or by those of water but becomes long-enduring," which is put into the mouth of the Buddha be earlier than Asoka's time, it is certain that the idea of employing stone or rock as a long-enduring material for engraving had not originated with Aśoka.
2. R. L. Mitra's Antiquities of Orissa, Vol. I, Pl. I.
3. Quoted from the Cambridge History of India, Vol. I, pp. 638-642, Sir John Marshall's article" The Monuments of Ancient India."

Page #332

304
OLD BRAHMI INSCRIPTIONS
of the hills of Udayagiri and Khandagiri in Orissa. Unlike the rockhewn monuments of Western India which were the handiwork of Buddhists, these Orissan caves were both excavated and for many years tenanted by adherents of the Jain religion, who have left behind them unmistakable evidences of their faith both in the early inscribed records and in the mediaeval cult statues which are found in several of the caves. To this sectarian difference is due many distinctive features of the architecture, including, among others, the entire absence of Chaitya halls, for which, apparently, there was no need in the ceremonial observances of the Jains.
Taken together, the two groups comprise more than thirty-five excavations, of which the more remarkable in point of size and decoration are the Ananta-Gumpbā and the Jaya-Vijaya caves on the Udayagiri hill. Besides these, there are two caves in the Udayagiri group, namely, the Hāti. gumphā and the Mañchapuri cave in which a special interest attaches by reason of the inscriptions carved on them.
Of the whole series, the oldest is the Hati-Gumphā, a natural cavern enlarged by artificial cutting on the over hanging brow of which is the famous epigraph recording the acts of Khāravela, king of Kalinga. This inscription was supposed by Pandit Bhagabanlal Indraji and others to be dated in the 165th year of the Maurya epoch, which, if reckoned from the accession of Chandragupta would coincide with 157-6 B.C. Other scholars have, however, since denied that any such date occurs in the inscription, and at the present time, there is still a sharp division of opinion on the point. In the absence of the undoubted date in this record or in the records of Khāravela's queen and of his successor in the Mañchapuri cave, we must endeavour to determine the age of these monuments from other sources of information. In the case of the Mañchapuri cave, the problem luckily derives some light from the style of the sculptured reliefs of the interior. This cave, erroneously called Vaikuntha or Pātālapuri by earlier writers, possesses two storeys, the lower consisting of a pillared verandah with chambers hollowed out at the end; the upper of similar design but of smaller dimensions and without any chamber at the extremity of the verandah.
It is in the upper storey of this cave that the inscription of Khāravela's queen is incised, while in the lower are short records stating that the main and side chambers were the works, respectively of Vakradeva (Vakadepasiri or Kudepasiri), the successor appareptly of Khāravela and of Prince Vadukha,

Page #333

NOTES
305
It may be presumed, therefore, that the upper storey is the earlier of the two. The rail pattern which once adorned the broad band of rock between the two storeys is now all obliterated, but in the ground-floor verandah is a well-preserved frieze which confirms by its style what the inscriptions might otherwise lead us to suppose, namely, that next to the Hāti-Gumphā, this was the most ancient cave in the two groups. Compared with some of the reliefs of the sculptures in the locality, they are of poor, coarse workmanship, but in the depth of the relief and plastic treatment of the figures, they evince a decided advance on the work of Bhārhut, and unless it be that sculptures in this part of India had undergone an earlier and independent development (a supposition for which there is no foundation) it is safe to affirm that they are considerably posterior to the sculptures of Bhārhut.
Next in chronological sequence comes the Ananta-Gumpbā—a singlestoreyed cave planned in much the same way as the Mañchapuri, which seems to have been the prototype of all the more important caves excavated on this site. Over the door-ways of this cave are ornamental arches enclosing various reliefs; in one standing figure of Lakşmı supported by the usual elephants on lotus flowers; in another is the fourhorsed chariot of the sun god (?) depicted en face, with the crescent moon and stars in the field; in a third are elephants; in a fourth, a railed-in tree, and figures to right and left of it bearing offerings in their hands or posed in an attitude of prayer. The arch-fronts themselves are relieved by bands of birds or of animals and Amorini at play or of garlands intertwined, and over each is a pair of triple-headed snakes, while in the intermediate spaces are flying Gandharvas disposed in separate panels. The last-mentioned are more stiff and schematic than the similar figures in the Mañchapuri cave. And this, taken in conjunction with other features such as the Chubby Amorini and the treatment of the sun-god's chariot. seems to indicate for these sculptures a date not much earlier than the middle of the first century B.C.
A further stage in the development of this architecture is reached in the Räni-Gumphā, which is at once the most spacious and elaborately decorated of all the Orissan caves. It consists of two storeys, each originally provided with a verandah—the lower 43 ft. in length with 3 cells behind, the upper 20 ft. longer with 4 cells behind; in addition to which there are chambers of irregular plan in the wings to right and left of the verandahs. In both storeys the façades of the cells are enriched with pilasters and highly ornate friezes illustrating episodes
39

Page #334

306
OLD BRAHMI INSCRIPTIONS
connected with the Jain religion of which unfortunately the interpretation has not yet been established. The friezes resemble each other closely so far as their general treatment is concerned, but the style of their sculptures in the two storeys is widely different. In the upper, the composition is relatively free, each group forming a coherent whole, in which the relation of the various figures to one another is well expressed; the figures themselves are posed in natural attitudes; their movements are vigorous and convincing; and from plastic and anatomical point of view the modelling is tolerably correct. In the lower, on the other hand, reliefs are distinctly elementary and crude. The best of them, perhaps, is the group reproduced in Pl. XXVIII, 77, but even here the figures are composed as independent units connected only by their tactile congruity; their postures, too, are rigid and formal, particularly as regards the head and torso, which are turned almost direct to the spectator, and in other respects the work is stiff and schematic. At first sight it might appear that in proportion as these carvings are more primitive-looking, so these are anterior to those of the upper storey, but examined more closely, they betray traces, here and there, of comparatively mature art, which suggest that their defects are due rather to the clumsiness and inexperience of the particular sculptors responsible for them than to the primitive character of the plastic art at the time when they were produced. Accordingly, it seems probable that in this cave, as in the Mañchapurf, the upper of the two floors was the first to be excavated, though the interval of time between the two was not necessarily a long one; and there is good reason, also, to suppose that the marked stylistic difference between the sculptures of the two storeys was the result of the influence exercised directly or indirectly by the contemporary schools of central and north-western India. In this connection a special significance attaches to the presence in the upper storey of a doorkeeper garbed in the dress of a Yavana warrior, and of a lion and rider near by treated in a distinctively western Asiatic manner, while the guardian door-keepers of the lower storey are as characteristically Indian as their workmanship is immature. It is significant, too, that various points of resemblance are to be traced between the sculptures of the upper floor and the Jain reliefs of Mathura, where, as we have already seen, the artistic tradition of the north-west, were at this time obtaining a strong foothold. The pity is that the example of these outside schools made only a superficial and impermanent impressions in Orissa-a fact which becomes clear if we consider some of the other caves on this site.

Page #335

NOTES
307
In the Ganesh-Gumpbā, for example, which is a small excavation containing only two cells, the reliefs of the frieze are closely analogous in style and subject, but, at the same time slightly inferior to those in the upper verandah of the Räni-Gumphā.
Then, in the Jaya-Vijaya, we see the style rapidly losing its animation, and in the Alakápuri cave, which is still later, the excavation has become still more coarse and the figures as devoid of expression as anything which has survived from the Early School. The truth appears to be that the art of Orissa, unlike the art of Central or Western India, possessed little independent vitality, and flourished only so long as it was stimulated by other schools, but became retrograde the moment that inspiration was withdrawn.”
13. STELLA KRAMRISCH ON RELIEFS IN THE CAVES
The reliefs that decorate the facades of all the Orissan caves have three main functions : (1) As friezes above a railing pattern, they stretch from tympanon to tympanon above the many small entrances into the caves. Mostly they are narrative. The frieze of the Ananta-Gumphā forms an exception. (2) As symmetrical compositions, on the other hand, they fill the intrados of the tympanon. This, however, is the case in the AnantaGumphā only. Otherwise the intrados are left plain. (3) As continuous rhythmical bands, containing animal, floral and human figures purely decorative, they rise as a lively decoration of the arch of the tympanon.
Besides these types of relief the Rāni-Gumphā has two walls of its laterally projecting mandapa-like cells, especially on the one to the left of a large relief composition, depicting a forest scene, which reminds one of the Indra relief from Bbāja. It is one of the earliest renderings of an extensive " landscape" scene in Indian art.
Single human figures accompany the curve of bracket capitals and of 1 he bracket supports of the railings. Single and coupled animal figures form the capitals of the entrance pilasters. Isolated male and female figures, standing as a rule and riding once (Rāni-Gumphā, upper storey), rendered on a large scale, and in a variety of types, guard the entrances.
The style of the Mañcapuri-cave reliefs puts them right at the beginnirg of artistic activities in the rock-cut caves of Orissa. The relatively well-preserved portion of a frieze shows, above a railing pedestal from left to right, a group of four walking figures in three-quarter profile, the hands folded in añjali-mudrā. All of them wear loin-cloths and scarves and

Page #336

308
OLD BRAHMI INSCRIPTIONS
heavy earrings. The first figure is considerably damaged, but seems to wear a crown on the head. The second is decorated by a heavy mural crown as worn by the Mauryan figures. The third has locks curling into the face, a fashion much favoured during the Gupta period and a heavy chignon to the left at the back of the head, as is frequently seen in Mediæval Indian sculptures. Above these four isocephalous figures, two fat-bodied gaņas disport themselves. Behind them appears a cross-bar, with a rope and an object suspended from it. Next to it is a lotus-petalled sun-disc with human-faced seed-pot. Divided from this group by a considerable interval, the figure of an elephant is seen coming forward, emerging from the background of the relief and turning into three-quarter profile, following the procession. It is followed by an inpetuously flying Gandharva, holding a basket.
The spacing of the figures has the restfulness and simplicity by which most of the Barhut compositions excel. A sameness of gesture makes the mood of adoration impressive in a concentrated manner. The diction is bold and simple, and denotes this relief as one instance in a long tradition, of which, however, nothing is known, as most of its work, most probably executed in a perishable material, has succumbed to the effects of time.
Besides these general features in common, what distinguishes this relief from any of the Barhut reliefs is more important than what links it up with it and with such other early Indian reliefs as those from Bhāja. Here, howerer, as well as in Barhut, the transition froin one artistic attitude into another of quite a different temperament and outlook, is well-marked. The four standing figures, short in stature, and treated as dense rectangular volumes, are faintly reminiscent of that trend of art which during the Maurya period was still alive and degenerated in the following century into a stiff and lifeless awkwardness.
All the other figures are animated considerably. This animation, playful and purposeless in the Gaņa-figures, is enhanced into energetic speed in the onrush of the Gandharva-figure. The transition from the static squareness of the Mauryan figures to linear vitality-peculiar to the 2nd century B.C.,-is marked bere as well as in Barhut. But there the movement is of a hesitating grace and reverential, whereas here it is not only variegated in speed and expressions, but is altogether more intense, untouched almost by any scruples of the religious mind. The craftsmanship is mediocre. Yet it took Indian sculpture eight centuries more to develop it, until at the end of the Gupta period the flying movement could be rendered as convincingly

Page #337

NOTES
309
as it is done here. The way in which the movement is enhanced from the “ kneeling" bent right leg of the flying figure to the raised and outstretched left in order to culminate in the graceful diagonal of the ends of the scarve is contrasted with the playful hovering of the gaņas with their enlarged, rounded and inarticulate limbs. Altogether the anatomy of the figures is more suppressed even than in Barhut in favour of an all-round smoothness of limbs, which may be seen in the treatment especially of the forelegs of the elephant, as also in the way in which the drapery is arranged into serpent-body-like “ folds " and streamers. This plasticity of limbs is subservient to an easy flow of movements. It gains in liveliness by addressing itself directly to the spectator. Whereas the Barhut figures, unconcerned about his presence, enacted their parts, intensely absorbed by them or by their own existence ; the figures of superhuman beings, of men and animals alike, address the spectator in three-quarter profile, so to say, or else they turn their faces in full front-view towards him. Yet in spite of forcefulness and agitation, the work on the Mañcapuri cave, with its halting and economical way as far as spacing and description go, is on one level with the work of Barhut. The features last mentioned belong to the diapason of Indian sculpture in the second century B.C.; whereas the direct emotional appeal, liveliness of gesture, and smoothness of limbs belong to a somewhat later period and are fully developed in the first century B.C. (cf. the relief in Mahabodhi and Sanchi) and destined to become more and more emphasized in the work of the other caves.
Although undated and without inscriptions, the reliefs of the AnantaGumphã on the Khaņdagiri, of the Râni Gam phà, of the Jaya-Vijaya and Ganesa-Gumphā in Udayagiri are further documents for the potentialities and the successive stages of early Orissan sculpture. If the Mañcapuri reliefs are somewhat posterior to Barhut, the friezes and tympanon fillings of the Ananta-Gumphā appear to be subsequent to the early portion of the Mabābodhi railing reliefs.
The Sūrya-reliefs, bere as there, apart from the similarity of subjectmatter and composition, betray in either case a further step taken in the direction indicated in Barhut as well as in the Mañcapuri reliefs.
Freedom from the shackles of the static cubic form adhered to in the Maurya period had been reached partly and gained in Barhut in a painstaking and careful process, whereas this selfsame freedom was reached in the Mañcapuri reliefs with bolder vigour and impetuousness. This freedom, once gained in all the reliefs of the 2nd century B.C. and

Page #338

310
OLD BRAHMİ INSCRIPTIONS
of the succeeding period, asserts itself with ever-increasing ease. It lends to the monuments of Central India and Bihar a plastic roundness of limbs, able to display themselves with ever-growing liveliness. It adds in Orissa to such reliefs as those of the Ananta-Gumphā, a plastic volume of infinite curvilinear possibilities. The cobra-body running parallel with the extrados is the most abbreviated formulation of this inherent tendency. The chariot on which Sürya with his two female companions are drawn is of similar construction in both the representations. The wicker-work front introduced in the AnantaGumphā relief facilitates the employment of smooth, elongated and rounded units. They appear again in the bent legs of the horses to which further company is added by the zigzag of the broader, serpent-like and rounded limbs of " Pingala " with staff and pot. The way in which his flabby body and face are modelled and are spread into the surface with great breadth, resembles the treatment of the ogress of darkness in the Sürya relief from Bhāja. It stands in striking contrast to the relatively firm plastic form of Barhut and Mahābodhi reliefs, but it shows how the artistic traditions of Central, Eastern and Western India touched each other in certain features while yet retaining their local inde. pendence in the main.
Nothing indicates the peculiar idiom of the Ananta-Gumpbā reliefs more distinctly than the sweeping variety of curves in which the necks of the prancing borses are bent hither and thither. In the corresponding Mahābodhi relief their movement is more uniform, more definitely connected with reality. The Ananta-Gumpbā composition adding, as it does, the lotus disc of the sun and the bead-accompanied crescent of the moon to the figures with elbows, splayed out into the relief, obtains a peculiar fulness of an almost flowing and flat relief, which again strongly contrasts with the firmer discipline of the higher and more energetic Mahābodhi relief.
The Mañcapurī relief was remarkable specially for the direct and bold vigour of its movements. This vigour is not muscular. It is not rendered by a firm modelling, but by linear curve. The method remain the same in the Ananta-Gumphā reliefs, but what formerly expressed itself as vigour has been transmitted now into a tough pliability. The speed has slowed down. The movement unfolds itself in breadth instead of being stated in an outspoken and sharp curve (cf. the Gandharva, Mañcapuri). It appears as a continuation of that hovering playfulness that had splayed out the

Page #339

NOTES
311
clumsy limbs of the gaņas there. But what was noticeable only in one type of figures and just in one variety there, is made now the exclusive principle of form. The boyish figures playing and rushing about in and between animal figures, amply betray this. The Lakşmi relief, in the second tympanon of the Ananta-Gumphā 'corroborates it. The flying Gandharvas filling the single rectangular compartments of the frieze show a sweep of drapery ; more to fill the surface than to express movement dynamically. The figures themselves in a flat, round and inarticulate modelling blissfully hover in their respective attitudes instead of rushing towards a goal.
The frieze of the Ananta-Gumphā deserves notice. Its main panels containing the flying Gandharvas suggests a barrel vaulted corridor supported in even intervals by the smooth shafts of round pillars stuck into pots. Above this runs a minor frieze, partly consisting of the usual railing pattern and partly of the combination of battlement and blue-lotus profile-pattern, well established in Barhut.
Next to these reliefs, those of the verandah, lower storey, left wing proper of the Rāni-Gumphā have to be placed. They stand in striking contrast to the few reliefs of the Mañcapuri lower storey that escaped defacement and specially to those of the upper storey. The complete scene with the dance, the other with the kingly person seat. ed with his companions, as well as the minor scenes, such as the one of adoration recall the plastic treatment of the Ananta-reliefs. The contour, however is somewhat hardened, the proportions are somewhat more elongated than there. But a sameness prevails in the wide open and almost expressionless glance of the faces, in the splayed-out plastic, awkward in its flabbiness on account of the hard contour that limits it. These reliefs appear to be the work of a less-gifted band and of a more harshly conducted tool as those of the Ananta-Gumpha, but otherwise they are the outcome of one and the same artistic mentality although they appear to be of later date.
Quite different are the scenes on the main wing and those on the corner portion of the left wing. The figures have grown in the height of the relief and in individual vivacity. Frequently one row is oversected by the other, groups of figures are formed as spontaneously and as rebellious with life as those in Sanchi. The faces have a happy smile, the movements are full of individual freedom. Their keen variety is supplemented by a factor that now becomes of a paramount importance.

Page #340

312
OLD BRĀHMI INSCRIPTIONS
This is the darkness differentiated in its depth that gives to the single figures and groups a texture elusive and rich. They form the light pattern on it.
This pattern effect of plastic forms surging into light from the darkness of the background is worked out in the most intricate manner on the walls of the projecting, portico-like side-cell. Above polygonal pilasters in pot-basis supporting a tympanon arch above animal capitals on the one façade and above less elaborate rectilinear pilasters supporting on capitals a tympanon on the adjacent side, and on the wall to its left proper a scene of forest life is carved with mangoe trees, rocks, waterfall and animals. Every single form is worked out by itself in bold relief. The forms are so densely set that the entire wall is dissolved into a deep and rough texture. Even the figure of a bird which by its smallness might become inconspicuous, is represented as a clearly visible entity by means of a receptacle, co to say, carved around it of circular or squarish shape. These very elaborate devices make up one of the earliest pure "landscape" scenes of Indian sculpture.
The forest here is rendered in the same way as the jungle grows. One forms next to the other, densely crowded, not allowing any hori. zon to be seen, any path to be found, an intricate pattern in which the single forms arrange themselves by crowding into every nook and corner. The heavier and more voluminous fall to the ground, the smaller and more slender forms rise up. The density of the pattern worked in high relief, pervaded by the contrast of light and shade, scarcely gives room for any rhythms except a perchance the rigorous discipline of densely set figures in superimposed rows; it is of the same type as the method of crowding with figures the reliefs of the Sanchi gateways.
Whatever the actual chronology of these undated caves may be, an inner process is at work, akin, nay parallel, to the changes that took place in the evolution of the reliefs from Barhut, through Mahābodhi, to Sanchi. These three stages are represented in Orissa by the work on the Mañcapuri, on the Ananta and Rāni-Gumphā.
Guardian figures are carved at the side of the cells of the Rāni. Gumphā or in front of the verandah pillars. Such guardian figures are to be seen on the Bhāja Vihāra rendered in a somewhat different idiom and in an earlier fashion. The figures from the Râni-Gumpha represent most varied types in body and dress, The treatment, however, remains one

Page #341

NOTES
313
and the same. It is a modelling with naturalistic aim, elaborating the anatomy in a broadly summarizing manner. The standing posture, though generally easy,-except the one figure clad with dhoti only, occupying the uttermost panel of the upper storey, left wing, betrays determination. The faces are damaged beyond recognition but from the minutely variegated physique and costume it appears as if these high reliefs were meant to be portraits. The tall figure, wearing a staff in the right hand, is reminiscent in its slenderness of the Southern type as may be seen in the figure of the worshipping prince from the Jaggayapeta relief ; remarkable again is the mighty bodily bulk of all the figures in the upper storey ; specially striking is the man, clad in Northern style, girded with a sword and wearing high boots. Just as the Sürya composition of the Ananta-Gumphā tympanon may be traced via Mahābodhi to Mathurā, 80 also is the costume of this figure to be met with among the Indo-Scythian statues from Mathurā.
Of greatest artistic value, however, amongst all the early Orissan reliefs, is the frieze decorating the upper storey, main wing of the Rāni-Gumphā. It begins with the frenzied onrush of a youthful male figure, carrying a tray with floral offerings in the left and a full blown lotus in the right. Its impetuous movement is a direct descendant of that of the flying Gandharva, Mañcapuri. The hovering of the Ananta-Gumphā flight has sunk into oblivion. Moreover, the movement which in the first storey amounted to a gliding just abyve the surface of the earth, now seems to emerge from it; the left foot is still half-covered by it; it is no longer a movement from above downward, nor a hovering in mid-air, nor a gliding above the surface, but a frenzied gesture of freedom ; uttered with the conviction of one who masters the decorative side of every experience. The following relief on the other side of the tympanon shows three elephants next to a prelude of a rocky landscape with cave and animals. Nothing could afford a stronger contrast to the perfectly decorative, two-dimensioned ma iner in which the figure with the offering was treated than this bold modelling, as impetuous in its realism and voluminous depth as the other is accomplished in width and complete organization of the surface. The daring oversecting, the mastery of mass, the agitation of light and shade and the one superb sweep by which the mighty bodies are composed into a circle have nothing in early Indian sculpture to compare with. The following scenes are more lyrical in their treatment, more bent towards the decorative, so that the animated figures,
40

Page #342

314
OLD BRĀHMĪ INSCRIPTIONS
carved in a high, yet sparingly modelled, relief become displayed lucidly over a flat background. The trend of art represented by the bulk of Barhut reliefs is still going strong. But a greater individual animation, a more dissected rhythm, a tougher sense of modelling are characteristic features. The modelling especially is conspicuous; the figures appear as if made of blown glass, so smooth is the texture of their rounded surface. This softness, however, is counteracted again by crisp movements, imparted to imbs, weapons, etc. The ambiguous character of the frieze may best be judged by comparing the elephant group with the group of girls surrounding a seated woman. Whatever has been depth there is surface here : volume is displaced by linear flow and calm surface, boldness by playfulness, altogether a perfect artistry, a last climax of a development.
Its mannerism, its being oversure of itself, may be seen in such forms as offered by the back-view of the musicians where anatomy and movement are mastered and distorted to make a pattern and nothing more.
The reliefs on the Jaya-Vijaya cave may be considered a second rate side-issue of the work on the Rāni-Gumphā. The worship of the tree scene, however, deserves notice, inasmuch as it contains the clearest formulation of the spatial conceptions of early Indian sculpture, widely made use of at the days of Barhut, matured into completion in Mabäbodhi and in this special example. One of the female guardian figures betrays in the slimness of her body and by the easy grace of her posture that selfsame affinity with southern Sāivite type that suggested itself already in view of the male standing figure of the Rāni-Gumphā (side wing, lower storey).
One of the latest excavations, the Ganesa-Gumphā, gives in the reliefs of its frieze a somewhat slipshod and abbreviated, though a more mature, version to the same scenes that figured with such ingenuous variety on the frieze of the upper story of the Rāni-Gumphā. From the style of the carvings the reliefs of the Udayagiri and Khaņdagiri caves, compared with the reliefs on the Buddhist railings in Bihar and Central India, cover a period of one and a half century approximately. The Mañchapuri relief and those of the Ganesa-Gumphā mark the beginning and end of this continuity.
19. RELATIVE CHRONOLOGY OF THE CAVES Rightly presuming that the bas-reliefs in the Ananta-Gumphā, the Rāni-Gumphā and such other caves on the hills of Udayagiri

Page #343

NOTES
315
and Khaņdagiri were not an independent local development in Orissa, Sir John Marshall finds reasons to maintain that in the depth of the reliefs and the plastic treatment of the figures, these show a development posterior to that of the sculptured reliefs on the Barhut stonerailing, and that judged by the relative inferiority and superiority of workmanship in the reliefs of different caves, these appear to reveal a successive chronology of the Orissan art. Dr. Stella Kramrisch agreeing with sir John Marshall in many of his main points, goes so far as to regard the bulk of the Orissan cave-reliefs as standing midway in some of their essential features between the sculptures on the stone-railing at Barhut and those at Buddha-Gayā. She inclines to think that the evolutionary process of plastic art in the Orissan caves took about a century and a half from its commencement to produce all the works that we find there.
So far as their conclusions with reference to the development of plastic art in the reliefs themselves go, we may have nothing to gain. say. Our difficulty is whether at all their conclusions as to the relative chronology of the reliefs are valid so far as to decide the relative chronology of the caves themselves.
The real point at issue is whether the arch-bands with their wealth of reliefs treating of various subjects and adorning the entrances of the Ananta-Gumphã and such other caves were at all there when the caves were excavated by King Khāravela and his compatriots. The point gains in significance as we find that the Hāthi-Gumphā and other old Brāhmi inscriptions allude to the architectural details, but not at all to any sculptured reliefs. And from the architectural point of view, the inferior workmanship and the superior workmanship as arguments respectively for the priority and the posteriority of the caves cannot bear scrutiny. It is evident from descriptions in the Hathi-Gumphā inscription itself that workmanship displayed in the costly works of art and architecture done by King Khāravela in the fourteenth year of his reign was far superior to that displayed in the 117 caves jointly excavated by King Khāravela and others in his thirteenth regnal year on the Kumāri hill for providing the resident Jain saints and recluses with suitable resting places. In one case, King Khāravela started with the distinct object of producing certain monumental works of art and architecture, and in the other, he started with the distinct object of comfortably housing the resident Jain saints and recluses.

Page #344

316
OLD BRAHMÍ INSCRIPTIONS
The differences in style and workmanship may as well be accounted for by the differences in the tradition, training and skill of the artists or craftsmen employed. To prove that the caves with superior style and technique of art in their reliefs were chronologically posterior to those without reliefs, as well as to those with inferior style and technique of art in their reliefs, it is necessary, first of all, to establish that any of the caves now found with highly ornamented reliefs was not counted among the 118 caves excavated in the thirteenth and fourteenth years of Khāravela's reign.
We have no evidence, as yet, to prove that. The inscriptions in the Udayagiri and Khandagiri caves standing next iu point of chronology to the old Brābmi inscriptions are a few Sanskrit inscriptions, including one incised by King Udyota-Kesari in the 7th century A.D. or even at a later period. King Udyota-Kesari's inscription, as we have seen, records the installation of the images of 24 Tirthařkaras in three of the caves on the Khandagiri bill which are known as Navamuni, Durgā and Hanumān, and the re-excavation of an old tank. None of these medieval inscriptions alludes to the excavation of a new cave. On the other hand, the installation of the images of 24 Tirtharkaras by King Udyota-Keśari and that of the images of nine sages, Durgā and Hanuman by the Hindus thereafter indicate a process of successive superaddition of stone-figures.

Page #345

-akhadamsasa, VII. aga-mahisina, II. Amga, 1.13. Amga-Magadhato, I.13. Amga.Magadha-vasinam, I.13. Amgānam I.13.
acitayită, 1.3.
-adhasatikam, 1.15. -adhipatina, I.1. -adhipatino, III.
-adhivasam, I.5.
-anugaha, I.7. anubhavamto, I.16. aneka, I.15.
anekāni, I.7.
anupato (?), 1.9.
-achariyam, I.13.
-ajeya (jeya), V, VII, XII. athatisaya, I.10. athame, I.9.
apayato, 1.9.
apatihata, I.16.
abbisitamato, I.2. abhisito, I.7.
Arahamta, II.
Arahato, I.14. Arihamtanam, I.1.
avadatena, I.1.
-asa, asam, I.13. Asaka (Asika ?), I.3. asi, 1.8.
A
-abhisecanam, I.1. abhutam, 1.13.
Arakatapuram (ahatapuvam ?) I.5.
Arahata-nisidiya, I.15.
-akara, I.15. -āgataya, 1.3. aneti, 1.13. -bharapani, 1.13. -ayatana, 1.16. aharapayati, I.13. -āhitahi, 1.15.
LIST OF INSCRIPTIONAL WORDS
Imdaraja (?), 1.16. -idäni, I.1, 1.6. idha, 1.13. -isi, 1.16. -isinam, I.14.
A
I
U
Utarapadha-rajano, I.13. upanamayamti, I 13. upadayati. 1.15.
⚫upetena, 1.1.
-uyana, I.2. usava, 1.4.
ekādasame, I.12. etinam, I.9. -eva, I.9.
Airasa, III, Airena, I.1.
-oghatitam, I. 6.
käritam, II. kareti, I.13.
E
Ai
0
kamma-padāna, I.9. Kammasa, VI. -kaḍāra. I.1,
Kadampa (Kudepa ?), III. Kapa-rukha, 1.9. -kara, I.7. kalānāni, I.16.
Ka
Kalimga-cakavatino, II. Kalimga-nagari. 1.2. Kalimga-puvaraja, I.12.. Kalimga-pavarajānam, 1.5, 1.11.
Kalimga-rajanivasam, I.10. Kalimga-rajavamsānam, I.11. Kalimgam, I.9. Kalimgagataya, 1.3. Kalimgâdhipatină, 1.1. Kalimgadhipatino, III. kaya-nisïdiyaya, I.14. karako, 1.16.
karayati, 1.2, 1.10, 1,15. kārāpanāhi, I.4.
karapayati, 1.8, I.11. I.15.
kārapitam, I.14.
karapeti, I.12.

Page #346

318
OLD BRAHMI INSCRIPTIONS
Cha
chadhe, 1.7. -chata, 1.5, 1.8.
Kālimga-Jinisanam, 1.13. Kalimgānam, II. kida payati, 1.4. -kidikā, 1.1. kidită, I.1. kumāra-kidikā, 1.1. Kumari-pavate, 1,14. kamaro, IV. Kusumasa, XIII. -kūtena, 1.5. kothā, VI. koțbājeyā, V, VII, XII.
janetum, 1.13.
jānapadam, 1.7. Jinâsaņam, 1.13. jeyā (ajoya ?), V, VII, XII.
Kha
Na
- ātibi, 1.14.
khanāpayati,1.6. Khāravela, 1.14, 1.16. -Khāravelasa, II. -Khāravelena, 1.1. Khīņāya, Vi Khemarājā, 1.16.
Ga
Gamgãyam, I.13. Gamdhava-veda, 1.4. -gaja, 1.3, 1.9, I.13. -gananā, I.1. gabhira, I.2. -gabhe, 1.15. -gahapatikānam, 1.9. -gita, I.4. guna, 1.1., I. 16. guta-cako, I.16. gopura, 1.2. gopurani, 1.13. Goradhagirim, I.9.
-tadăga-pădiyo, I.2. tatiya, I.11. tatiye, 1.1, 1.4. tato, I.1. tathā, 1.5. Tanaguliya, 1.6. tāpasa, 1.15. -tasite, 1.5. tiriyam, 1.15. -tivasa-sata, 1.6. -tisāya, 1.10.
tishi, I.2. -turamga, 1.8. terasame, I.14. terasa-rasa-sata-katam, 1.12.
Tha
-thambhāni, 1.15. thambhe, 1.15.
Gha
-ghatāni, I.8. -gharavāsinam, 1.9. -ghātāpayitā, 1.9.
Da
ca
damdam, 1.3. dampa (dapa ?), I.4. -dabham, 1.12. dadāti, 1.9. dasame, 1.11. -disam. I.3, -disānam, 1.15. dutiya, 1.3. devåyatana, 1.16.
ca, 1.2 foll. -caka, 1.16. -cakavatino, 11. -cake, 1.14. -cako, I.16. -catara, I.13. -catare, I.15. catu visati, 1.1. cavathe, I.5. Cülakammasa, V, VII, ceca-yathi, 1.15. Ceta-räjavamsa, 1.1. cetiyani, I.15. cêdâni 1.6.
Dha
-dhaja, 1.8. .dhamakūtena, 1.5. Dhamarājā, 1.16. dhamena, 15. -dharo, I.16.

Page #347

nagaram, 1.6.
-nagari, I 2. nagarim, 1.4. -nata, 1.4.
namo, 1.1.
nara, 1.3, 1.9.
narido, 1.9. nava-vasāni, I.1.
navame, 1.10.
Näkiyasa, IX. nikhita, I 5.
Nagale, 1.12. Namda-raja, 1.6, 1.13. nagara-akhadam sasa, VIII.
-nivasam, I.10. .nivesanam, I.2. nivesayati, 1.13. nivesitam, I.12.
-nisidiya, I.15. -nisidiyaya, I.14. -nītam, 1.13. -nila, 1.15. nekasayati, I. 12.
pamcame, 1.6. Papḍa-raja, 1.13. pardarasa-vasāni, I.1. pakatiyo, 1.2. pachima, 1.3. paṭalika, I.15.
patisamkharayati, I.2. pathapayati, 1.3. patiṭhāpayati, 1. 15. patisartha panam, 1, 2 -padaua, I. 9.
padhame, 1. 2.
panatariya, I. 15.
panatīsahi, I. 2, 1, 15. panaḍim, I. 6.
-panadena, I. 9. papiḍapayati, I. 9. pabhäre, I. 15.
palikhani, I. 13. parinivasato, I. 14. pariharamti, I. 13. palava-bhāra, I. 9. pavata-cako, I. 16.
-pavate, I. 14. pavesayati, I. 6. pasatha, I. 1.
pasamto, I. 16. -pasādânam, II. -pasasayati, I. 5.
pākāra, I. 2.
-pāḍīyo, I. 2. padamalikasa, XIV. -påda, I. 12 pade, I. 5, I. 12. pana-bhojanam, payayati, I. 13. pasato (pasado ?) VII.
I. 9.
Na
LIST OF INSCRIPTIONAL WORDS
Pa
-pasādam, I. 10. pasado (pasado), V, VI, Pithudaga, I. 12.
puram (pavam ?), I, 5. purisa-yuge, I. I.
puvarajānam, 1. 5, I, 11. pujako, I. 16.
-pesayamti, I. 13.
pora-jana padam, 1. 7.
-bhatakanam, 1.9. -bhatakehi, 1. 14. bhayam, I. 13. -bhatihi, I. 14. -bhimgare, I. 5. bhita, I. 5. Bhutino, VIII. -Bhojake, I. 5. -bhojauam, I. 9.
bamdhapayati, I. 1. bamhananam, 1, 9. -balo, I. 16.
-Bahasatimitam, 1. 13. -bahnlam, 1. 3.
barasame, I. 13. Bariyaya, IX. -budho, I. 4.
Ba
yamti, I. 9, J. 12. yatinam, I. 15. yasa-sakāram I. 11. yuga-saga, I. 10. -yuge, I. 1.
Bha
-mamgalani, I. 8. -mani, I. 12. Madhuram, I. 9. -Magadhato, I. 13. -Magadha-vasinam, I. 13. mahati, I. 9. Mahamadasa, IX.
Ma
Mahāmeghavahanasa, I. 1. Mahāmeghavahanena, III. maharajâbhisecanam, I, 1. Mahārājasa, III. mahārājena, I. 1. mahāvijayo, I. 16. -mahisinā, II. -mahisihi, I. 14. Magadhanam, I. 13. muta 1. 13.
Ya
Ar

Page #348

320
yojana-âhitahi, I. 15. yovanâbhivijayo, I. 1. yovarajam, I. 1.
ramjayati, I. 2. -rakhi, I. 8. rakhanu (?), I. 1. -rathika, I. 5. -ra tana, I, 5, ratanam, I. 13. ra tanani, I. 12. -radha, I. 3, I. 8, I. 9. -raja, I. 6, I. 12. Rajagahānam, I. 9. raja-ñatihi, I. 14. raja-nivasam, I. 10. raja-bhatakanam, I. 9. raja-bhatakehi, I. 14. raja-bhātihi, I. 14. raja-mahisihi, 1. 14. rajavamsa-vadhanena, I. 1.
-rajavamsanam, I. 11. rajavamse, I. 1.
raja-sirim. I. 7. raja-sutehi, I. 14. -raja, I. 16.
rājānam, I. 5, I. 11. -rājāno, I. 13.
rajino, II. .rukha, I. 9. -rapa, I. 1.
-lakhanena, I. 1. Lalakasa, II.
Ra
varåkara, I. 15. Vadha-raja, I. 16.
vavahāra, I. 1. -vasasata, I. 6.
La
lena, I. 14.
lenam, lenam, II, III, IV, VIII, IX, X, leṇāni, XIV.
-vasani, I. 1. vase. I. 2 foll.
-vaso, I. 1.
-vasinam, I. 13.
-văţă, I. 6. vāta-vihata, I. 2. .vadita, J. 4.
-vamsa, I. 5.
-vamdapayati, I. 5, I. 13. -vamsa, I. 1, I. 16.
-Vamsanam, I. 11. -vamse, I. 1.
Vadukhasa (Varikhasa?), IV.
-vana, I. 7.
-vadhanena, I. 1. vadhamana, I. t.
Va
OLD BRAHMI INSCRIPTIONS
-vasasatam, I. 12. -Vasuki, I. 13.
-vahana, I. 1. -vahanasa, II. -vahane, I. 9. -vahanena, I. 1. -vijaya-cake, I. 14. vijaya-pāsādam, I. 10. -Vijayo, I. 1, I. 16. vijadharâdhivasam, I. 5 vijâvadātena, I. 1. vitaseti, I. 3.
-vidhi, I. 1. -vinisito, I. 16. vipamumcitum, I. 9. vipulam, I. 13.
visajati, I. 7. -visaradena, I. 1. vividha, I. 13. -visesa, I. 16. -vihata, I. 2. vithi, I. 13. -veda-budho, I. 4. veduriya, I. 10, I. 15. -vochimnam, I. 15.
sa, I. 16.
-samkara-karako, I. 16, -samkharayati, I. 2. -samthapana, I. 2. samdamsanam, I. 8. samdamsayamto, I. 7. samdamsanahi, I. 4.
sampuna, I. 1. sambhita, I. 9.
Sa
sakata (sukata ?), I. 15. -sakāram, I. 11. sagâvasane, I. 10.. sata, I. 6.
-satam, I. 12, I. 14. -sataghatani, I. 8. satadasa-lena-satam, I. 14. sata-disānam, 1. 15. satame, I. 8.
sata Vasuki, I. 13.
-satikam, I. 15.
samana, I. 15.
sata-sahasani, I. 7, I. 13.
sata-sahasahi, I. 15.
-sata-sahasehi, I. 2, I. 6, I. 9, I. 10, I. 11,
I. 13, I. 15.
samananam, I. 9, II, XI.
samaja, I. 4. samipe, I. 15. samuthapita hi, I. 15. -sariravata, I. 1.
sava-kara-vaņa, I. 7. savata, I. 5, I. 8. sava-mamgalani, I. 8. sava-rathika-bhojake, I. 7. sava-vijâvadatena, I. 1. sava-sidhānam, I. 1.

Page #349

POETR
41
LIST OF INSCRIPTIONAL WORDS
savûyana, I. 2. -saha, I. 9, I. 2. -sahasani, I. 1, I. 11. -sahasahi, I. 15.
sitala, I. 2. .sihasa, II. -sutehi, I. 15. sunamto, I. 16.
-sahasehi, I. 2, I. 6, I. 8, I. 9, I. 10, I. 11, supavata, I. 1.
I. 13, I. 15. sātakamņim, I. 3. -sapateye, I. 5. sāsitam, I. 1. -sidhanam, I. 1. -siri, I. 16. siri-kaḍāra, I. 1. -sri-Kharavelasa, II. siiri-Khāravelena, I. 1.
-sirina, I. 14. -sirino (sirino), III. -sila-thambhani, I. 15. silahi, I. 1. siharāni, I. 13.
-subha-lakhanena, I. 1. -suliya, I. 6. sena-vahana, I. 13. sena-vahane, I. 9. so, I. 1.
Ha
hathi, I. 13. hathisam, I. 9. hathisa-pasavam, I. 13. haya, I. 3, I. 13. hita, I. 5.
321

Page #350

INDEX TO NOTES
Abhuta-achariya, a description of certain Beal, Samuel, p. 225.
elephants and horges from the land of the Beglar, J. D., p. 225. Vasukis, pp. 221, 291.
Ben-Gumphā (Jambeswara), name of a cave Acharya, P. K., pp. 289, 298.
on the Udayagiri, p. 295. ahatapuva (āhatapuva ? ), an adjective ap- Bhandarkar, D.R., pp. 204, 211, 224,
plied to the Vidyadhara-abode, p. 210. Bhojaka, pp. 210.1. Aira (Ahira), a king of Utkala, pp. 183-5. Bhūti, a town-judge of Kalinga and donor of aira, a royal style, p. 262.
a cave on the Udayagiri, p, 248. Ajagara-Ġumphā, name of a cave on the Bindusara as a Brāhmana-bhatta, p. 260. Udayagiri, p. 295.
Brahmamitra, & king of the neo-Mitra Akbar, pp. 265, 268.
dynasty (?), pp. 278-9. Alaka (Mulaka). one of the Andhra terri- Bphaspatimitra. identified by Jayaswal with tories, p. 212.
Paşyamitra, p. 274. Alakāpuri, name of a cave on the Udayagiri, Byhāsvātimita, father of Queen Yagamitā, pp. 295, 297.
p. 274. Amga, a country conquered by Khāravela, caka-vāhana-bala, an epithet of Khāravela, p. 231.
p. 263. Ananta Gumphā, name of a cave on the Caldwell, p. 206. Khaņdagiri, pp. 295-6.
cakaoati, a royal epithet, p. 232. antă avijitā, pp. 216.7.
caturamta-rakhana-guna-upeta, an epithet anupa-dabha.vana, a description of Pithu. of Khāravela, pp. 231-2. daga, pp. 197, 217.
ceca-yathi, p. 303. apparels, as presents from the king of Cetarājavamsa-vadhana, an epithet of Kbāra. Pandya, p. 252.
vela, pp. 233-5. apatihata-caka-vähana-bala, an epithet of cetiyāni, shrines, p. 303. Khāravela, p. 253.
Chakravarty, Niranjan Prasad, p. 149. Arahamta (Arihamta), a Jain designation, Ohanda, Ramaprasad, pp. 150, 180-1. pp. 262-3.
Ohatterji, S. K., pp. 164, 249, 267. Arakatapura (?), a city identified with Arcot, Chicacole, p. 201. pp. 205-6.
Ohola, pp. 205-6. Asaka (Asika), a city identified, pp. 212-6. Chcta-Häthigamphā, name of a cave on the Agoka, his Kalinga province compared with Udayagiri, p. 295.
Khāravela's Kalinga kingdom, pp. 192-3, Coomarswamy, p. 141. 202; prohibition of sacrificial slaughter, p. Cowell, p. 248. 251; Kalinga War, pp. 264 foll. ; popula. Cunningham, Alexander, pp. 206, 229. tion of Kalinga in his time, p. 256; atti. Dakkhiņā patha as distinguished from Uttaratude towards former kings, p. 258; patha, p. 218. religious policy and idea of toleration, dampa (dapa), pp. 264-5. pp. 264 foll.; conquest of Kalibga as a Dantagura-Paloura, old capital of Kalibga, p. landmark in the history and civilisa-! 198. tion of Kalinga, pp. 279-80, 302-3; Deb, Harit Krishna, p. 143. compilation of the Kathāvatthu during Dhanabhūti, donor of Barhut gateways, his reign, p. 284 ; ideal of conquest by p. 277. dhamma, p. 286 ; work of art on the Dubreuil, p. 228. Dhauli rock, p. 303.
Dutt, B.B., p. 289. Agādhasena, maternal uncle of King Bahasati- Durgã-Gumpbā, name of a cave on the Khan
mita and king of Abichatrā, p. 278; his dagiri, p. 296. inscriptions, pp. 147, 161.
ekapadimagga, p. 212. Bagchi, P.C., p. 224
Fleet, p. 144. Bahasatimita (Brhaspatimitra), king of Francis, p. 241.
Magadha and contemporary of Khāravela gaja, a division of Indian army, p. 252. identified, pp. 273-4.
gahapatayo, p. 262. bamhanā, p. 262.
Gamdhava-veda-budha, an epithet of Banerji, R.D., pp. 161, 160, 166, 248, 326,
Khāravela, p. 242. Barā bar hills, its different names, pp. 226-7. Gamgā, as a northern boundary of AmgaBāriyā, a place or personal name, p. 248. Magadha, p. 221.
322

Page #351

INDEX TO NOTES
323
kata-sama ini, p. 247. cells of a cave
ganană, an art mastered by Kh&ravela, pp. | Kamma, donor of the cells of a cave on the
241 foll. Ganesa-Gampha, name of a cave on the kata-samana, p. 262. Udayagiri, p. 296.
Khalatika-pavata, old name of Barabar Ganguly, Manomohan, pp. 217, 220, 294-5. hills, p. 226. gharavāsī, p. 262,
Khapingala-pavata, p. 193. Girivraja, old capital of Magadha, p. 224. Khāra vela, contents and importance of his gopurāni, pp. 287-90.
inscription, pp. 175 foll. ; king ove lord of Goradhagiri, a place or personal name, pp. Kalinga, pp. 231-2; significance of his 223-7.
name, p. 267; nobility of origin, pp. guta-oaka, an epithet of Khāravela, p. 178. 233-5; dignity of bodily form, p. 239; Hunumän.Gumphā, name of a cave on the education, pp. 240.6; family members, Khaņdagiri, p. 297.
relations and officers, pp. 246.9; state Haridās-Gumpha, another name of Păvana. elephant, pp. 249-50; wealth, pp. 250Gumphā, p. 296.
2; army, pp. 252-6; administrative Hathi-Gumphā, name of a cave on the method, pp. 256-9; religion and religious Udayagiri, p. 295.
policy, pp. 259 foll.; place in history, pp. Hathis, ha (Hathisīha), father-in-law of 268 foll.; date, pp. 272 foll. ; city of Khāravela, p. 249.
Kalinga, pp. 287 foll.; other works of art hattha-mudda, P, 243.
and architecture, pp. 292 foll. haya, a division of the Indian army, p. Khiqa, a female donor of the verandah of 252.
a cave on the Udayagiri, p. 247. Himavanta region as distinguished from Kramrisch Stella, pp. 305 foll.
Uttarā patha and Dakkhiņāpatha, pp. Košala, a division of Khāravela's Kalinga 219-20.
kingdom, p. 203. Imdagimita, a king of the neo-Mitra dynasty Kothājeyà, p. 298. (P), p. 278.
Kumāra, an epithet of Vadakha-Varikha, p. Indaraja, an epithet of Khāravela, p. 247. 250.
Kusuma, donor of caves on Khandagiri, p. Indraji, Bhagawanlal, p. 144.
248. isī, Hindu gages, p. 262.
Lālāka, an epithet of King Hathisäha, Isigili, one of the five hills round Rājagana, p. 249. pp. 164-5, 225.
Lekha, an art learnt by Khåravela, pp. 241 Jackson, V. H., p. 224.
foll. Jaya-Vijaya, name of a cave on the Udaya-Lena (cave), 117 caves on the Kumāri hill, giri, p. 297.
p. 292; one special cave, p. 292; why Jayaswal, K. P., his explanation of the all are not inscribed, p. 293; system
symbols, p. 142; fire corrected readings, of counting, pp. 293-4 ; fate of missing pp. 154-6, Sanskrit Slokas from an old caves, p. 294; modern names, pp. 294 foll.; Osiyā MS., pp. 183-7; notes on Mūzikas, component parts, p. 300; technical sig. Kronaveņā and Strīrājyā, pp. 207 foll. ; nificance, p. 301 ; purposes, pp. 301.2. readings and renderings of Khāravela's Lévi, Sylvain, pp. 196-7. 8th year's record, pp. 222-3; identifica. Law, N. N., pp. 237, 251. tion of Goradhagiri, pp. 225-8; explana. Laders, H., pp. 164, 224. tion of purisa-yuga, p. 235; explanation Madhură (Mathurā), identified, pp. 219 foll. of lekha-rūpa-ganana-vavahara-oidhi, pp. madya-kridā, p. 264. 244-5; inforence about population of Magadha, country, p. 234. Kalinga, p. 256 ; opinion about Khāra. Magadha-Khetta, p. 228. vela's date, pp. 274 foll.
Magadha-pura, pp. 218, 225, Kadampa (Kudepa), a king of Kalinga and Māgadha-räjā, p. 229.
donor of a cave on the Udayagiri, p. 239; Magadha-vāsi, p. 234, relationship with Khāravela, p. 247; Mahamada, an official designation or persignificance of his name, p. 267; se s onal name, p. 248.
venth king of Meghavāhana line, p. 239. māhāmagga from Benares to Cetiys-rattha, Kalinga, kingdom of Khāravela, pp. 189-90;
its different capitals, p. 201; three Mahāmeghav hana, an epithet of Khāravela divisions, p. 203 ; contiguity to Assaka, and Kadampa, pp. 249-50. pp. 212-5.
Mahāpadma Nanda, third king of the Nanda Kalinga-cakavati, an epithet of Kharavela, dynasty, pp. 279 foll. p. 232.
mahārāja, a common epithet of Kbāravela Kalingadhipati, an epithet of Khāravela and and Kadampa, p. 239. Kadampa-Kudepa, p. 239.
Mahāraţthi-Mahābhoja, p. 211. Kalióga-naga a, capital of Khāravela, pp. Mahāvijaya, an epithet of Khåravela,
195 foll. ; proximity to the Tanasuli road, p. 256; name of the palace built by pp. 202-51; plan, picturesqueness and Khāravela, p. 289. internal life, pp. 287 foll,
Majumdar, R. C., pp. 235, 237.
p. 234.

Page #352

324
INDEX TO NOTES
Mazumdar, B. C., p. 277. Mañicapuri, group of three caves, pp. 138, 247, 293.
mani, as a present from king of Pandya, p. 252. Marshall, Sir John, pp. 278, 304, 315. Megha, a shortening from Meghavahana, p. 272; nine Meghas, pp. 272, 281. Mitra Rajendra Lala, pp. 238, 243. Mosala-Maisolos, a division of Kalinga, p. 203. Mukhalingam, modern identity of Kalinga.
nagara, p. 191.
Mukherji, Radhakumud, pp. 254, 256, 263, muta, as a present from the King of Pandya, P. 252.
Nagala (Läñgala), a river in Kalinga, p. 204. nagara-akhadamsa, the same official designation as nagalaka-mahāmātā and nagaraka, p. 248.
Nakiya, donor of a cave on the Khandagiri, P. 248.
Namo arihamtanam, a Jain formula of invocation, pp. 170, 259-60. Nandarāja, pp. 275 foll. nara, a division of the Indian army, p. 252. Nava-muni, name of a cave on the Khanda-,
giri, p. 296.
nava-vasäni, nine years spent by Kharavela as a crown-prince, p. 280.
Nayanika a queen of the Satavahana family, daughter-in-law of Simaka and wife of Satakarni I, pp. 276 foll.
nigamtha, a Jain designation, p. 262. nila, an architectural ornament, p. 304. niti, principle of government, p. 257. nivesana, p 287.
Oldenberg, p. 162.
Oldham, p. 225.
Old Oriya MS., pp. 183. 266, 270. päkāra, pp. 287-8.
pakatiyo ramjayati, p. 257.
Palapiñjara, perhaps another name of Mt. Khapingala, p. 204. Pamḍarājā, pp. 221, 222. pamdarasa-vasäni, fifteen years spent by Kharavela as a young prince, p. 271. Pargiter, p. 211.
Pāsāņaka-cetiya, an ancient Hindu shrine, p. 228.
pasāta, pasāda, prāsāda, p. 298. Pasatha-subha-lakhana, an epithet of Kha
ravela, p. 232.
Patalapura, name of a cave on the Udayagiri, p. 295.
Pavana-Gumpha, name of a cave on the Udayagiri, p. 295.
pavata-caka, an epithet of Kharavela, p. 189. Pithuḍaga-Pithuḍa, a marshy tract reclaimed by Khāravela, pp. 196 foll. pora-janapada, p. 251. Pravaragiri, a mediæval name of the Barabar hills, p. 226.
purisa-yuga, pp. 235-9.
radha, a division of the Indian army, p. 252. Rāņi-Gumpha, name of a cave on the Udaya. giri, p. 294. Rapson, p. 235, Rathika-Bhojaka, pp. 210-1.
Raychaudhuri, pp. 201, 209, 211, 212, 216, 219, 221, 229, 234.
Rhys Davids, T. W., pp. 212, 241. Rules of royal writs, pp. 179-80.
rupa, an art learnt by Kharavela, pp. 241-2. Saja, pp. 225-6. Sakata-samana, pp. 262-3. samaja, p. 263. Samāpā, p. 192. Sarpa-Gumpha, name
of a cave on the
Udayagiri, p. 294. Satakani, king of the Satavahana family and contemporary of Kharavela, pp. 215. 6.
sava-devayatana-samkara-karaka, an epithet of Kharavela, p. 261. sava-pāsamḍa-pujaka, an epithet of Khāra. vela, p. 261.
sava-vijā, pp. 242 foll. savuyana, p. 287.
siharāni (sikharani), p. 289. Sisuka-Simuka, Nandiyasa's daughter's son and first king of the Satavahana family, p. 281.
Sisunandi, king of Vidiśa on the fall of the Sunga power, p. 281. Smith, Vincent A., p. 192. supavata-vijaya-caka,
a description Khāravela's Kalinga kingdom, p. 189. Sugarman, last king of the Kanva family, p. 281.
taḍāga-pāḍiyo, pp. 287-8.
of
Tamira-daha-samghata, pp. 156, 217. Tanasuliya-văţa, pp. 195 foll.
Tattva-Gumpha, name of two caves on the
Khandagiri, p. 294.
Udaka (Ardraka), fifth king of the Sunga family (?), p. 273. Uttarapatha, pp. 218-20. Uttarapatha kings, p. 220.
Vadukha (Varikha), prince of Kalinga, and donor of two caves on the Udayagiri, p. 247.
Vaikunthapura, name of a cave on the Udayagiri, p. 294.
varakara, excellent quarries, p. 252. vavahara-vidhi, certain sciences and arts learnt by Kharavela, pp. 241-2. Vogel, p. 163.
Vyaghra-Gumpha, name of a cave on the Udayagiri, p. 294.
Yasonandi (Nandiyaga), second king of Vidisa on the fall of the Suhga power, pp. 281 foll. Yavana-raja, p. 223.
1

Page #353

Some of the Publications of the Calcutta University on Ancient Indian Culture, Religion and
Philosophy. : A History of Indian Literature
BY M. WINTERNITZ, Ph.D. PROFESSOR OF INDOLOGY AND ETHNOLOGY AT THE GERMAN UNIVERSITY
OF PRAGUE (CZECHOSLOVAKIA)
VOL. I
INTRODUCTION, VEDA, NATIONAL EPICS,
PURĀŅAS, AND TANTRAS
TRANSLATED FROM THE ORIGINAL GERMAN BY
MRS. S. KETKAR AND REVISED BY THE AUTHOR
Only Authorised Translation into English Demy. 8vo, pp. 653.
Rs. 10-8.
Some Problems of Indian Literature
By the Same Author. Royal 8vo, pp. 130.
Rs. 2-€.
D9
Lectures on Comparative Religion (Stephanos Nirmalendu Ghosh Lectures delivered in the
Calcutta University in 1923) . Published in July, 1925
BY Prof. A. A. MACDONELL, M.A. (Oxon.), Ph.D. (Leipzig),
D.Litt. (Edin.), D.O.L. (Calcutta) Royal 8vo, pp. 194, 1925.
Agents for Sale.--Messrs. Longmans, Green & Co., 39 Paternoster Row, London, E.C. 4.
Also all leading book-sellers of Calcutta, Bombay and Madras.
Rs. 3.

Page #354
Shri Mahavir Jain Aradhana Kendra www. kobatirth.org Acharya Shri Kailassagarsuri Gyanmandir rving Jin Shasan 093298 gyanmandir@kobatirth.org For Private And Personal Use Only