ΤΟ

WILLIAM D. WHITNEY, Ph.D.,

PROFESSOR OF SANSKRIT IN YALE COLLEGE, NEW HAVEN,

This Work

IS RESPECTFULLY INSCRIBED,

AS A SLIGHT RECOGNITION OF HIS GREAT KINDNESS,

BY

THE TRANSLATOR.

INTRODUCTION.

THE text here translated is taken from vol. xxvi. of the sûtra section of the Bkah-hgyur, folios 329-400. This version has been revised on that of vol. lxxi. of the Bstanhgyur, folios 1-53, which, though generally very incorrect reproducing nearly all the errors of the Bkah-hgyur (besides many others of its own), has enabled me to correct and complete my text in many places where it was so much effaced in the copy I made use of (that of the National Library at Paris) as to be nearly useless.

The work is divided into thirty-three chapters and four books, each of which contains about the same number of verses. Book I. has twelve chapters and 260 verses, Book II. twelve chapters and 249 verses, Book III. six chapters and 248 verses, Book IV. three chapters and 232 verses, making in all 989 verses or udânas, the greater part of which are in verses of seven and nine syllables.1

The title, "Tched-du brjod-pai tsoms," is rendered in Sanskrit by Udânavarga, i.e., chapters of udânas, but the word udâna must not be understood to imply "joyous utterances, hymns of praise," but something nearly approaching "gâtha, verse, or stanza," 2 although in some cases, where certain virtues are extolled, the word is employed with its habitual acceptation of “hymn.”

Such verses are very generally found at the end of the sermons or sûtras of Gautama, and were probably intended

1 There is really no prosody in Tibetan, metres being only distinguishable by the number of syllables in each line. In the Rgya-tcherrol-pa (Lalita Vistara) I find 7, 9, 11, 13, 15, 17, and 21 syllabled lines, the first three being the most used.

2 It would be perfectly admissible to call this work "a sûtra," using, however, that word in its habitual sense of "series of aphorisms.' See L. Feer, Une Sentence du Buddha sur la Guerre, p. 34, note.

to convey to his hearers, in a few easily remembered lines, the essence of his teaching. It appears to me that the founder of Buddhism must have attached great importance to these verses, and that he advocated their use by all his disciples. Take, for example, the history of Çariputra's meeting with Açvadjit shortly after the former's conversion, and we see at once what a single gâtha was able to do in the eyes of early Buddhists, and what rôle these aphorisms undoubtedly played in the work of their missionaries. As a natural consequence of the importance attributed to these verses, it appeared desirable to the first successors of the Buddha to collect in separate works all such utterances of the Master as might prove especially instructive, and as best answering the purposes of their school. To this plan is undoubtedly due the fact that in both the Southern and Northern canons are numerous works which only contain the pith of more voluminous and older ones attributable to the Buddha. In the Northern canon we know of the Sûtra in 42 sections and the Udânavarga, besides several others in the extra-canonical collection (the Bstan-hgyur). The Southern canon offers us a much greater number of such works, the best known of which are the Dhammapada and the Sutta Nipâta,

The Udânavarga contains 300 verses, which are nearly identical with verses of the Dhammapada; 150 more resemble verses of that work; twenty are to be found in the Sutta Nipâta, and about the same number are very similar to parts of the same book. Thus more than half of the Udânavarga is found in works of the Southern canon, and it appears highly probable that if the Udâna, the Theragâthâ, Therigâthâ, &c., had been examined, many more of the verses of the Tibetan work would have been found in them.

If the Tibetan version has been constantly compared with the Pâli, it is not because I consider the latter as the text on which the Tibetan translation was made, but because it is the only term of comparison available.

Throughout this work there occur constant slight variations from the Pâli, and, knowing as we do, the scrupulous care of the Indian Pandits who supervised the Tibetan translation, it is not possible to admit that these differences are the result of carelessness, but rather we must explain them by the existence of different versions on which the Pâli and the Northern Buddhist translations were made. Now, for example, in the Pâli Nidânakathâ, p. 76, we find the two celebrated verses, 153, 154, of the Dhammapada. In both works these verses are the same in every respect, but in the Tibetan version of the Nidânakathâ, called the Jatakanidânam,2 we find a quite different version of them. The text of the Udânavarga (chap. xxx. 6, 7) closely follows, however, the Pâli version.

The version of the Jatakanidânam is as follows. "He sung this udâna, sung by all Buddhas:

"153 Through an endless circle of births

Have I sought to end, to destroy the poison,

Seeking the maker of the house;

Again and again (have I known) the sorrow of birth.

154 I have found the maker of the house;

No more shall (he) make a house for me;

All his grief is pulverised,

And the poison is destroyed with the house.

(My) mind is freed from the sanskâra;

Craving is ended and (I) shall be no more."

This differs too much from the other versions to admit of the supposition of the translator having misunderstood the Pâli "sandhâvissam, visamkhitam," &c., of the generally received text.

We might show similar discrepancies between gâthâs 133, 134, of the Nidânakathâ and the two Tibetan versions of the Jatakanidânam3 and the Udânavarga, but what has been said appears sufficient to show that there must

1 See Burnouf, Intr. à l'Hist., p. 25 et pas.; Foucaux, Rgya-tcherrol-pa, passim.

2 Bkah-hgyur, Mdo xxx. f. 520

3 Jatakanidânam, loc. cit., fol. 454b, and also fol. 458a, which corresponds with Nidânakathâ, p. 24, gatha 163.