JOURNAL

OF

THE ROYAL ASIATIC SOCIETY.

ART. XIII.-The Tantrakhyāna, a Collection of Indian Folklore, from a unique Sanskrit MS. discovered in Nepal. By Prof. CECIL BENDALL.

CONTENTS.

I. Introductory Essay

II. (a) Index of Tales, with Comparative Notes

(b) Special Index to Tales corresponding to the Pañca-tantra

III. Notes on Variations from the Pañca-tantra in Tales generally corre

sponding to portions of that work................

IV. Abstract of Tales not in the Pañca-tantra.....

V. Extracts from the Sanskrit Text

.............

PAGE

465

470

473

474

478

485

497

I. INTRODUCTORY ESSAY.

THE first notice of the work that forms the subject of the present paper was given in Dr. Daniel Wright's "History of Nepal," where, at p. 322, the title of the book occurs in the list of Sanskrit MSS. procured for the University of Cambridge. In examining this collection in the years 1880-3, I noted the work as related to the Pañca-tantra. As, however, this MS. was (with the exception of some verses as to which I shall speak presently) entirely in Newari, and as I decided to issue at first only the catalogue of the Buddhist Sanskrit MSS., I have hitherto never published any further details, but only gave a passing mention of my discovery in my paper read at the Berlin Congress of Orientalists in 1881

VOL. XX.-[NEW SERIES.]

32

(Verhandlungen, Theil II. Hälfte ii. p. 204). When, however, I visited Nepal in 1884, I obtained the small palm-leaf MS., which I now exhibit, containing the work entirely in Sanskrit. The MS. is dated Nepal Samvat 604, or A.D. 1484, and was copied by one 'Jasavarman' svärthahetunā, by which, I suppose, is meant that his real name was Yaçovarman, and that he copied it for his own use. Perhaps this last intimation may account for the bad spelling, poor sandhi, and general corruptness of text that prevail throughout. There are also several perplexing lacunæ. For this reason I do not propose at present to publish the text in full, but, pending at all events the possible acquisition of another MS., I now offer such an account of the tales as may prove serviceable to the student of Indian, and of general, folklore, and subjoin (Pt. V. VI.) selections from the text.

Like most other Indian story-books, from the oldest known collection, the Pali Jātakas, downwards, each tale begins with a moral or text in verse. These texts are preserved in Sanskrit even in the Newari version; and this being so, I have collated for the present essay the MS. at Cambridge already cited (which I call 'A' below), as well as another ('B') in the same collection (Add. 1594 and 1613). Through the kind negotiations of my friend Professor Minaev, I have also been favoured with the loan of a third Newari MS. (which I call 'C'), belonging to the Imperial Academy of S. Petersburg, a body which I have found on a previous occasion most liberal in lending, and to which I desire to record my hearty thanks.

The general literary character of the stories is somewhat bald, mostly lacking the racy sense of humour that makes the Jātakas so delightful and exceptional in Oriental literature. Indeed many of the stories here seem to me to be mere notes for the viva voce telling of a story already more or less familiar to the speaker at least, if not also to the hearers. This theory seems confirmed by the very abrupt way in which many of the stories terminate; not by a leisurely ato

1 Called S in the critical notes to § IV. below.

'ham bravimi,1 as in the Hitopadeça, followed by a repetition of part of the verse text, but a curt phrase like evam buddhihinasya doshaḥ, so the fault lay with the witless wight,' where the story has been told in illustration of the advantage of buddhi or vous. Conversely in one or two cases the positive moral is pointed out by a compound ending in guna. The separate stories, moreover, are styled not akhyāna, but ākhyānaka, a diminutive form.

Having thus explained the general character of the work, I may now approach what is in fact the most important. question of the present paper: namely, what is the exact position of this collection in the general chain of Indian folklore, to which the poetry and fiction of our own middle ages are so largely indebted?

I am pleased to be able to exemplify this indebtedness by a small contribution to Chaucer-literature."

The book is, as I have already stated, closely allied to the Pañca-tantra. Of its 47 stories, about 25 Of its 47 stories, about 25 may be regarded as founded on tales in that collection. And what is important to note is, that several of them were put into their present shape from a recension of that work differing from any of those now extant. Thus, for example, Tale 38 in the present collection, that of the mouse and the cat, corresponds to chapter 5 of the Old Syriac version, which was made about 570 A.D.,3 but does not occur in the Sanskrit Pañca-tantra. Another tale, No. 16, the well-known story of the elephant freed by the mice, occurs only in the 'schmuckreichere Recension,' represented by the Berlin MS. used by Kosegarten in his unfinished text of the editio ornatior,' and. likewise in the Tamil Pañca-tantra accessible to European readers in the translation by the Abbé Dubois. It is, however, quite an old story, familiar to all in the Esopic fable of the lion and the mouse. On the other hand, Tale 24, the bird and the ape, belongs to the latest stage of stories in the Pañca-tantra, as it is not included in the Arabic nor even

1 The expression Tenaham in the introductory verse to Tale 21 cited below points to a very similar usage. Tale 28 ends "evam 'anyathā cintitam' iti." See the full verse below.

2 See Tale 42.

3 Keith Falconer, Bidpai, Intr. pp. xiv, xlvi.

in the Tamil. Again, Tale 22, the story of the sage who changes a dog into a beast of prey, and then changes it back when attacked himself, is far closer to the Indian tale preserved in the Mahābhārata and even the Hitopadeça (IV. vi.) than to the Pañca-tantra version.

So much for the general relation of our book to the Pancatantra-cycle of story, which to the historian of European literature at all events constitutes the most important branch of Eastern folk-lore.

There are also a few ancient stories of Indian origin, but not included in the Pañca-tantra.

An example of these is Tale 25, where even the 'text' or introductory verse was evidently the same as that of an Indian story included in the Bkah-hgyur, the Tibetan Buddhist canon, translated in the ninth century A.D, if the reference in the note to Mr. Ralston's version denotes the fourth great section of the canon.

In subject, the tales present quite as much variety as the contents of most Indian story-books. Some are beast-fables, others turn on the relations of the sexes, others again look like mere incidents taken from historical legends or from romances.

The style is on the whole decidedly poor. Passive Past participles in -ta, for example, are used in an active sense. Cf. Tale 30, note 2, Tale 32, note 1. There are, however, some curious lexical forms, which seem to show that the book is of independent origin. Examples are: √kūt strike' (Tale 31), hitherto only found in the Dhātupāṭha, and consequently ignored by Professor Whitney in his "Sanskrit Roots." In Tale 43 (the Cat and the Mouse, lost in the Indian Pañca-tantra), in the introductory çloka, occurs the noun anupraveçaka side by side with the verb anupraviç; also the form 'ryayagat, for which I would read avyayayat, a causal form which is given in the Dhatupaṭha in the sense of 'motion' (gatau): here clearly of the wheeling of the hawk. In Tale 15 (not printed) occurs the form çakyāmi (=çaknomi) parallel to the Pali sakkāmi.1

1 In Tale 10 (not printed) we find the forms agrahārika for a brahman who has received an agrahara or royal donative: and just below, the form kulaputrikā

In spite of the odd forms that occur, I am not now inclined to the idea, which at first struck me, that the book is a mere local Nepalese production. This seems clear from several points in the tales. It is hardly likely, for example, that a Kathmāṇḍu pandit would take for the hero of an anecdote a king of so distant a people as the Kalingas, who appear to have lived between South Orissa and Madras. In the very next tale, the story of the Brahman and his wooden image (see abstract), we find that sums of money are mentioned as paid in darmmāḥ (Spaxuai) dirhams.' In another tale dīnārāḥ (denarii) are mentioned. Such coins would suggest that the stories in their present form originated not in Nepal, but rather in some part of India, such as the Panjab, in communication with Persia and other Muhammadan countries. The word darmma or dramma is of rare occurrence in Sanskrit literature. To the passages given in the lexicons may be added the X-XIth century inscription, which I discovered in Rajputāna, and published in the account of my journey.

I conclude, then, on the whole, that the Tantrākhyāna is one of the numerous independent workings-up of the tale-material current in India from an early date. It is parallel both to the Hitopadeça and to the portion of the Katha-sarit-sagara (chapters lx. etc., Tawney's translation, vol. ii. pp. 27, sqq.), which corresponds to the Pañcatantra, though it is not necessary to assume for it so late a date (eleventh century) as the latter of these books. I may here mention a compilation probably very similar to the present collection, as to which I have been kindly and most unexpectedly favoured with some private information by Dr. H. N. van der Tuuk, an eminent Dutch Orientalist, residing in the remote island of Bali, in which discoveries so important for Sanskrit literature have been made. This is a collection of tales called the Tantri, of which Dr. van der Tuuk gave some account in our Journal for 1881 (New Series, Vol. XIII. pp. 44, 45).

(ironically?) for a low-caste woman, analogous to the sense of kulaputra cited from Vaijayanti by Kavindra Sarasvati on Daçakumara-c° (p. 136, 1. 20, ed. Bombay, 1883). See also Pt. III., notes on Tale 42 A.