________________
318
REVIEWS
11
the Tanjur in which he is mentioned as translator." The SRKK must therefore have been composed before the beginning of the eleventh century. It is quite possible that both the first part (chapters 1-22) and the second part (chapters 23-27) have come into existence in the period between Santideva and the eleventh century. Zimmermann refers to Bendall's introduction to the Sikṣāsamuccaya for Santideva's date (seventh century). It has to be pointed out, however, that this date is based entirely upon Taranatha who wrote his history of Indian Buddhism in the beginning of the seventeenth century. Bendall draws attention to the fact that the Sikṣāsamuccaya was translated by Jinamitra, Danasila and Ye-ses-sde and that, consequently, 800 A.D. has to be admitted as the latest possible terminus ad quem. The dates of Santideva are as uncertain as those of Aryaśura who is usually said to have lived in the third or fourth century. The translation into Chinese of a work by Aryaśura in 434 A.D. seems to be the main reason for putting him in the third or fourth century. However, Lin Li-kouang has shown that this work, the Fen-pieh ye-pao lüeh ching (Taishō no. 723), is another recension of the Fen-pieh shan-o so-ch'i ching (Taishō no. 729) of which the translation is attributed to An Shih-kao (148170 A.D.). Both the attribution of this recension to An Shih-kao and the attribution of the work itself to Aryaśūra are very doubtful. The only reliable date in connection with Aryaśūra is furnished by the quotations of his Jātakamālā in the inscriptions in Ajanța. According to Lüders these inscriptions are from the sixth century.13
Zimmermann has found thirteen verses of the SRKK in other texts. The fact that out of 160 verses 13 are found elsewhere seems to indicate that the SRKK is a compilation largely based upon other texts. In an article which has escaped Zimmermann's notice V. V. Mirashi points out that verse 6 is quoted in two grants of the Maitraka-s of Valabhi, dated Gupta year 248 (A.D. 567-568) and Gupta year 269 (A.D. 588-589). According to Mirashi this proves that the SRKK was composed before A.D. 550.14 He adds: "We know of no Buddhist writer named Aryasura who flourished before this date except the well-known author of the Jātakamālā. It is not therefore unlikely that the SRKK also was the work of Aryasura who flourished in the fourth century A.D." It is much more likely that this verse was well-known in the sixth century and that the SRKK was not the source of the quotations in the Valabhi grants.
Zimmermann's edition of the SRKK gives for each verse: 1. The text as published by Banerjee; 2. Variant readings from the manuscript used by Zimmermann; 3. Indication of the metre; 4. Translation of the verse; 5. Notes to the translation containing the readings proposed by the editor; 6. The Tibetan translation with indications of the corresponding parts of the Sanskrit verse; 7. Variant readings; 8. Translation of the Tibetan version containing comments on all points which require attention. His main object is to show that the Tibetan translation, though very imperfect and unsatisfactory, is based upon the same text. According to Zimmermann's English summary the frequent divergences between the Sanskrit and Tibetan versions are due to the following causes: 1. There were considerable deficiencies in the text used by the translators; 2. The Tibetan text is not free of corruptions. One may occasionally wonder whether that could not be the result of subsequent attempts at retouching, undertaken without help from the Sanskrit original; 3. The translators occasionally failed to understand the Sanskrit text. Zimmermann calls the risk which has its origin in the aforementioned discrepancies 'lexical risk'. The other uncertainty factor is, according to him, the 'syntactic risk' due to the fact that the Tibetan translators often stubbornly kept the order of words and/or lines of the Sanskrit version. Zimmermann points out that in several cases there would have been little chance of understanding the text correctly, had the Sanskrit text not been at hand.
Marcelle Lalou, Répertoire du Tanjur d'après le catalogue de P. Cordier (Paris, 1933), p. 213. 12 L'aide-mémoire de la vraie loi (Paris, 1949), pp. 102 and 313.
13 Arya-Śūras Jātakamālā und die Fresken von Ajanta', Gött. Nachr. Phil.-Hist. Kl. (1902),
pp. 758-762 [= Philologica Indica (Göttingen, 1940), pp. 73-77].
14 'A note on the Subhāṣitaratna-karaṇḍakakatha of Aryaśura', Adyar Library Bulletin, 25 (1961), pp. 304-307.