have existed at an early date several versions of the Buddhist canon, and that the Pâli and the Tibetan translations were most likely made from texts that differed to some

extent.

Our text affords no clue as to the language from which it was translated; it only gives the name of the work “in the language of the White Plain" or India. This expression is generally used to designate " Sanskrit;" it can, however, mean nothing more than Indian vernacular; and it is well known that at the time the Tibetan translations were made there existed Buddhist works in the language of Li (Khotan or Nepal), Zahora (Punjab), Kachmere,1 &c., and that Pâli works were also made use of by the Indian translators, for quite a number of volumes of the Bkahhgyur (notably vol. xxx. of the Mdo), contain texts directly translated from that language.2 I am, however, inclined to think that it was made from a Sanskrit version in the dialect prevalent in Kachmere in the first century B.C., at which period and in which place the compiler, Dharmatrâta, probably lived.

The Udânavarga is found also in the Chinese Tripitaka. The title of the work is there " Chuh-yau-king," or Nidâna Sûtra. It is also divided into thirty-three chapters, the titles of which agree with those of the Tibetan, with the following slight differences:-Chap. iv. is "Absence of Careless Behaviour;" chap. v. "Reflection;" chap. vi. "Intelligence;" chap. xxix. "The Twins" (Yamaka1). The contents of the two works, as far as has been ascertained, are identical.

Both the Chinese and the Tibetan version attribute the compilation of the Udânavarga to Dharmatrâta, and the

1 See Schlaginweit's Könige von Tibet, fol. 20b.

2 See L. Feer, Annales du Musée Guimet, ii. p. 288.

3 See Beal, Catalogue, p. 85.

4 I am indebted to Mr. Beal for this information. The title of chap. xxix. of the Tibetan version might also be translated by Yamaka, if

5

night and day are considered as forming a pair.

5 Csoma in As. Res. xx. p. 477, and M. Feer, Annales du Musée Guimet, ii. p. 280, call him Dharmarakshita, which in Tibetan would be Tchos skyong, whereas the name is translated by Tchos-skyob.

former says that he was the uncle of Vasumitra. If this Vasumitra was the one who was president of the Synod held under Kanishka, we might at once assign Dharmatrâta to the first century B.C. Unfortunately the question cannot be solved so easily. Târanâtha says that there was a Sthavira Dharmatrâta, who was one of the leaders of the Vaibâshika, "but," he adds, "one must not confound this Dharmatrâta with the compiler of the Udânavarga.” 1 So likewise Tchandrakirti, cited by Burnouf,2 mentions two Dharmatrâtas, a Sthavira and a Bhadanta, and it is highly probable, from Mr. Beal's catalogue of the Chinese Tripitaka, that both these Dharmatrâtas composed works. The Sthavira composed the Samyuktâbhidharma çastra (Catal. p. 82), but cannot have written the commentary on Aryadeva's Çataçastra Vaipulya,3 for Aryadeva must have lived later than he.

The date of the first translation into Chinese of a collection of gâthâs by Dharmatrâta (A.D. 221-223), enables us, however, to limit the period in which the compiler of the Udânavarga can possibly have lived, and to assert that the work was composed somewhere between 75 B.C. and

200 A.D.

If we compare the Udânavarga with the Fa-kheu-pi-u, it appears that in this latter work, out of 140 verses or parts of verses, there can be traced back to the Dhammapada about twenty-five which do not occur in the Tibetan version. This has led me to suppose that if the other works attributed to Dharmatrâta in the Chinese Tripitaka were examined, one might probably discover all the verses of the Dhammapada and quite a number of those of the Sutta Nipâta.

The Tibetan translation was made by Vidyaprabhâkara, who, from the fact that his name is frequently mentioned in connection with those of Çilendrabodhi, Dânaçîla, &c.,

1 See Târanâtha, p. 54, lig. 8.

2 Burnouf, Intr. à l'Hist., p. 566. 3 See Beal, Catalogue, pp. 76, 77,↑ 82, and 108. In the Bstan-hgyur

there are no works by Dharmatrâta except the present one.

4 Beal, Texts from the Buddhist Canon, Trübner's Oriental Series.

well-known translators of the ninth century, was most likely in Tibet at about the same time, when King Ralpa-chan (A.D. 817-842) was giving great encouragement to translators of Buddhist works.

The commentary which has been made use of in the present work was composed by Pradjñâvarman, who lived in Kachmere in the ninth century A.D.1 Besides the present commentary entitled the Udânavarga vivaraṇa, he composed a commentary on the Vices chastava by Siddhapati, and one on the Devâticayastotra by Sanskarapati. In the introduction to the Udânavarga vivaraṇa3 it is said that "Pradjñâvarman was an Indian of Bhongala (Bhangala ?), and a disciple of Bodhivarma of Kapadhyara (sic). He was born at Kava, in the country of Bhangala, and his fame was great; he was blessed with great steadfastness and sound understanding. Being blessed with the recollection of many of the flawless jewels uttered in the Dharma, he was of infinite service to the rest of mankind. His recollection of the many sayings of the holy law caused him to shine forth like the sun, and through the extent of his knowledge he dispelled the darkness that enveloped mankind, bringing them joy and confidence. . . . He composed, then, a commentary to help to set forth clearly the sayings which he used to speak to the multitudes. He kept the still beautiful cut flowers (of the Dharma) in their original form, but dispelled the obscurity of some of the utterances, making their perfections to burst forth like lotus flowers, and thus every one' of the utterances of the most excellent of Munis (ie., Gautama) has become as bright as the sun. This commentary was therefore composed to extract the essence of the utterances of the Tathagata, called 'words of great

1 See Târanâtha, p. 204 trans. 2 See Bstan-hgyur Stotra, Nos. I and 2, and Sûtra xxxiii., Nos. 100 and 101.

3 Bstan-hgyur lxxi, fol. 54a, and vol. lxxii., fol. 244b. I find also

mentioned in the Index of the Bst. a Pradjnavarma as a translator of Indian Buddhist works into Tibetan. It is most likely the same writer.

blessing,' or 'udânas' - fragments of the words of the Victorious One-and to teach their real signification."

This commentary is divided in thirty-three chapters, each one of which is devoted to a chapter of the text. Each verse is generally preceded by a short history of the events which caused the Buddha to speak it. In some cases these stories appear to have some historical value, but in the great majority they have evidently been invented to suit the text. In the Appendix there will be found a certain number of these "nidânas," as they are called by Pradjñâvarman, illustrating the different kinds of stories he joins to the text. In some few cases there is some analogy between the events related and those mentioned in Buddhaghosha's commentary on the Dhammapada, but it is unimportant.

The explanations of the words of each udâna are borrowed from the sûtras or the âgamas (lung); they are all of that literal description which one might expect of a "hinayanist" who did not pretend to extract from words anything more than their ordinary meaning. In the chapter on nirvâna he shows us that Tchandrakirti's observation about the Vaibâshika Dhamatrâta, that "he believed in the existence of things past and things future," is also applicable to him.

1

Let me here call attention to the use made throughout the Udânavarga of the word "nirvâņa.” In the greater number of cases it only implies the condition of Arhat, in which "sorrow has been left behind," "the nirvâņa of the Arhat" (xxvi. 10), or in Pâli, kilesanibbanam. On the other hand, it cannot be understood to imply anything else but annihilation in such verses as xxvi. 27, in which it is defined as "not existence, not to be born." Here, then, it is in the "destruction of every particle of the elements of being (skandhas) or anupadisesanibbanam."

It has frequently been asserted that the Northern Buddhist texts were of no value for a critical examination

1 See Burnouf, loc. cit., 566.

of early Buddhism; the Dhammapada was held up as being absolutely necessary for any correct understanding of the real ideas of Gautama.1 I think, therefore, that when there exists so exact a rendering of every important verse of that book as is found in the present work, we are entitled to score a point in favour of the much-abused Tibetan and Chinese texts. It is a matter of deep regret to me that M. Schiefner, to whom we owe the first knowledge of the real contents of the Udânavarga, did not live to publish the translation of this work which he had undertaken, for in his hands it would have escaped much of the ill-treatment it has here experienced. However imperfect this translation may be, it is to be hoped that it will prove of some use, and that some indulgence will be shown to the translator of verses which are thus spoken of by the Chinese editor of the Fa-Kheu-king-tsu: 2 "The meaning of these gâthâs is sometimes very obscure, and men say that there is no meaning at all in them. But let them consider that it is difficult to meet with a teacher like Buddha, so the words of Buddha are naturally hard of explanation."

It remains for me to express my thanks to Dr. R. Rost, through whose kindness I have been able to avail myself of the rich collection of Tibetan works in the India Office, and also to Professor S. Beal, to whom I owe much valuable information concerning the Chinese version of the present work.

MONTREUX, 6th November 1882.

1 See for example Oldenburg's Buddha, sein Leben, &c., p. 198, where it is called "die schönste und

reichste unter der Spruchsammlungen."

2

Beal, Dhammapada, p. 30.