________________
216
REVIEWS
in Khotanese and written in Brahmi script of which 59 manuscripts and fragments contain Buddhist texts, etc.). The authors explain how the different collections were built up. They distinguish five kinds of documents:
(1) Manuscripts written in three different Brahmi scripts (Indian, Upright and Slanting Central Asian) and containing texts in Sanskrit, Khotane se and Kuchean. They date from the fifth to the ninth centuries.
(2) Kharosthī manuscripts (Dharmapada and economic documents) written on birch-bark and wood in North-Western Prakrit and dating probably from the first century to the second half of the third century. '
(3) Uigur manuscripts in cursive and semi-cursive scripts and Uigur blockprints dating from the thirteenth and or fourteenth centuries. The Uigur manuscripts from Tun-huang were probably copied in the period from the ninth to the eleventh century.
(4) Tibetan manuscripts from the eighth century (documents on wood), and manuscripts dating from the ninth to the eleventh century (Tun-huang and Khara-Khoto).
(5) Chinese manuscripts from Tun-huang (fifth to eleventh century) and from Khara-Khoto (eleventh to twelfth century).
The Central Asian fund contains about 700 items, about one third (50 items) of which has been published. The authors announce further publications of Sanskrit and Khotanese texts. They have added three tables which list all publications relating to the Sanskrit, Khotanese and Kuchean manuscripts and fragments (pp. 22-36).
Of the six Mahāparinirvāṇasūtra fragments edited by Bongard-Levin, five had already been identified by Vorob'ev-Desjatovskij. Bongard-Levin points out that only two other fragments are known: one edited by F. W. Thomas in 1916 and another edited by J. Takakusu in 1916 and recently studied in great detail by Akira Yuyama. According to Bongard-Levin, the six fragments date from the sixth and seventh centuries. Bongard-Levin gives a brief description of each fragment, followed by the edition and translation of it. The notes deal with graphical and grammatical particularities and refer to the Tibetan translation and Dharmaksema's Chinese translation. Corresponding passages in the Derge edition of the Kanjur, and the Taisho edition of the Chinese Buddhist canon, are indicated in a table and also in the description of each fragment. All fragments of the Mahāparinirvanasūtra and of the other texts edited by the two authors are reproduced in facsimile.
In his introduction Bongard-Levin remarks that a detailed comparison with the Tibetan and Chinese versions has still to be carried out. He received help from several Russian scholars in the study of these versions. It would have been useful if the edition of each fragment had been accompanied by an edition of the