国立情報学研究所 - ディジタル・シルクロード・プロジェクト
『東洋文庫所蔵』貴重書デジタルアーカイブ

Sec. ii]   FINDS IN A POLYGLOT LIBRARY   815

Dr. Hoernle as an upright Gupta of the Nepalese type, it appears highly probable that it was imported directly from the south, i.e. through Tibet. The fact that the palaeographic features point to the eighth or ninth century A. D. fully agrees with this assumption, as that was exactly the period of Tibetan predominance at Tun-huang.

In the case of the other Sanskrit manuscripts (see e:g. Ch. vii. ooi. a, Plate CXLIII) the writing in Slanting Gupta places their Central-Asian origin beyond all doubt, and at the same time renders it very probable that they were brought to this site from Eastern Turkestan and not copied locally. This conclusion seems justified in view of the fact that, wherever Brahmi writing is found on the reverse of Chinese rolls, it is invariably of the Cursive Gupta type, whatever the language, as a reference to the inventory in Appendix F will prove. That these rolls originally belonged to the old Chinese stock of the local monastic libraries is made obvious at first sight by their paper and general appearance, including the careful penmanship of the obverse. It is equally clear that it was merely the convenient writing-material offered by their blank reverse which led to this being utilized at a later period for Brahmi texts.* As will be noted further on, I saw at the time that a number of these contained alphabetic tables and writing exercises. The observation is of interest, as it serves to prove conclusively the presence among the local monks, at least, in that later period, of men familiar with Khotanese as well as with Sanskrit of a sort.

Both the Upright and the Cursive types of Central-Asian Brahmi writing are found in the Pôthis containing ' Khotanese ' texts, of which the complete version of the Vajracchedika, Ch. 00275 + Ch. xlvi. ooi 2. a (Plate CXLIX), in 44 folios, and the large manuscript of a medical formulary, Ch. ii. 003 (Plate CL) in 71 leaves, may be quoted as typical specimens. As both types of Brahmi writing are met with in manuscript remains from sites of the Khotan region which contain texts of this language, the question whether these Pôthis were brought thence to Tun-huang or written locally cannot be settled on that ground. Nor is it possible to make quite sure of the origin of the huge roll, Ch. c. 001 (Plate CxLVi), over 7o feet long and nearly a foot wide, which by its size and excellent state of preservation—in its 1,108 lines it is practically complete—was outwardly the most striking among the non-Chinese manuscript finds.5 The contents have proved to consist of Buddhist texts in corrupt Sanskrit interspersed with Khotanese ' statements, the former in Upright, the latter in Cursive Gupta. But both its shape and its paper, of a kind plentifully found among the later Chinese documents and texts of the walled-up deposit, raise a presumption that this monumental roll was produced by some pious local scribe. The presumption is supported also by the design of the painted silk cover, which agrees in style with motifs frequently shown by the paintings and decorated textiles from the same hoard.

At the time I was able to appreciate best the philological interest of these Brahmi texts. But, apart from this, they have an archaeological value as offering tangible proof that the monastic communities established at Tun-huang, among a population mainly Chinese, must have retained, until a relatively late period, direct touch with those in the Tarim Basin and particularly in the Khotan region. Considering that ever since Tang times the main line of communication connecting China with the Western regions led, not westwards through Tun-huang to Lop and Khotan, but through Hami towards the oases along the Tien-shan- range, it seems difficult not to recognize evidence of some special links between Tun-huang and Khotan Buddhism in the prevalence of Khotanese texts among the Brahmi manuscript remains of the walled-up chapel. There

Brahmi texts on reverse of Chinese rolls.

Brahmi alphabet tables.

Pôthis and Khotanese texts.

Links between Tunhuang and Khotan Buddhism.

' Pl. CXLV shows such a roll, Ch. lviii. 007, with the end of the Chinese Buddhist text on the obverse, and portions of the syllabaries written in Cursive Gupta on the reverse ; cf. for the latter Dr. Hoernle's description, j.R.A.S., 1911,

PP. 452 sqq.

PI. CXLVI shows one unfolded portion of this roll reproduced topsy-turvy, a mistake which owing to my absence in India at the time remained uncorrected.