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. iv] MSS. FROM THE HIDDEN LIBRARY, IN BRAHMI AND CHINESE 917

considerable time and labour to the examination of particular manuscripts, especially such as have a special bearing on Buddhist iconography and kindred studies.

Out of these thousands of manuscripts only two. texts, short but of distinct historical and geographical interest, have so far been published. Both owe their edition and translation with valuable notes to Dr. Lionel Giles of the British Museum. One, the Tun•huang lu, is a succinct treatise on the mirabilia of the Tun-huang district, dating from the close of the Tang period. It has proved useful by its topographical indications and been repeatedly referred to above.18 - The other, Ch. 922, contains a fragmentary original record of the official census of Tun-huang taken in A. D. 416.13 Preserved in the form of a small roll of which the reverse has been utilized in Tang times or later for some Buddhist text, it affords a good illustration of the valuable ` finds ' which may yet be expected among the masses of miscellaneous papers rescued from the ` mixed bundles'. From another and larger text, Ch. 917 (Plate CLXIX) written in A.D. 886 and containing notes on the geography of Central Asia, Professor Pelliot has used interesting extracts in his paper dealing with early Sogdian colonies in the Lop region."

Gratifying as these few publications are, they must make me wish more than ever that a systematic encouragement and expansion of Far-Eastern researches in England and elsewhere may soon provide an adequate number of Sinologists duly qualified by critical training and capable of turning to good use the wealth of these new materials, which in the end may prove not only the most numerous, but also the most valuable, of all literary remains I recovered from the walled-up chapel. Meanwhile I must feel glad that the following extract from Professor Pelliot's above-mentioned memorandum makes it possible to record here the opinion of a most competent scholar as to the extent and value of our Chinese manuscript collection from Tun-huang :

` Les manuscrits chinois rapportés de Touan-huang par le Dr. Stein peuvent se diviser au point de vue d'un inventaire en deux catégories

10 Les rouleaux manuscrits complets ou de dimensions assez considérables, environ 3,000 manuscrits.

20 Les pièces détachées ou fragments, de 5,000 à 6,000.

On pourrait être tenté de n'inventorier que la première catégorie. Mais ce serait retirer au travail toute partie scientifique, toute base sérieuse. En grande majorité les manuscrits complets appartiennent à des oeuvres bouddhiques que nous poss4dons dans les éditions chinoises et japonaises du Canon bouddhique, et ces manuscrits seront naturellement très précieux, étant les plus anciens, pour des études de détail sur ces ouvrages ; mais dans l'ensemble ils nous apportent relativement peu d'informations nouvelles immédiatement utilisables. Il en est autrement pour les pièces (actes d'ordination, baux, comptes), souvent datées, qui se rapportent à tous les actes de la vie locale, et qui représentent une catégorie de documents dont, avant les découvertes de Touan-houang, nous n'avons pour ainsi dire aucun spécimen.

Enfin, c'est parmi les fragments que se trouvent le plus souvent les textes de la littérature laïque, fragments historiques, géographiques, lexicographiques, etc., qui sont de la plus haute importance pour le progrès des études de sinologie érudite.

Même sommaire, l'inventaire, pour avoir quelque utilité, devra dans la mesure du possible indiquer la nature du texte quand le titre ne pourra être déterminé. Il devra utiliser tous les colophons et, en leur absence, indiquer une date approximative pour l'écriture du manuscrit. Ce travail, pour l'ensemble des documents, prendra certainement une année.'

It will help to illustrate Professor Pelliot's remarks, and to .show the wide range of subjects represented among our Chien-fo-tang texts, if I append here the notes which he and Dr. Lionel Giles have been kind enough to furnish regarding the manuscripts reproduced in Plates CLXVI—CLXIX. To their friendly help is also due the selection of these specimens which on account of their exactly fixed dates, their palaeographic features or contents, or for other reasons may claim some special

Chinese texts of antiquarian interest.

M. Pelliot's memo. on Chinese MSS.

Notes on specimens of Chinese MSS.

18 See Giles, Tun-huang Lu, J.R.A.S., 1914, pp. 703 sqq.; with supplements and corrections in J.R.A.S., r915, pp. 41 sqq. For references, see above, pp. 62o, 623, 716, 734 ; for a reproduction of two pages of the booklet, Ch. 1073, see Pl. Ci.xlx.

19 See Giles, A Census of Tun-huang, T `oung pao, 1915, pp. 468 sqq. Reproduced in Pl. CLXVI below, where by mistake mainly the later writing of the reverse instead of the original text of the obverse has been shown.

20 See J. Asiat., janvier—février-1916, pp. 120 sqq.