National Institute of Informatics - Digital Silk Road Project
Digital Archive of Toyo Bunko Rare Books

Sec. i]   FIRST OPENING OF HIDDEN CHAPEL   809

placed in a more accessible position on the left-hand, or southern, wall of the passage. To the interesting contents of this inscription, of which M. Chavannes was kind enough to furnish an annotated translation,' I shall have occasion to refer further on. Through Chiang Ssti-yeh's reading I was made aware at the time that this inscribed stone was dated in the year corresponding to A. D. 851. Thus the important fact became clear to me from the first that the deposit of the manuscripts must have taken place some time after the middle of the ninth century.

Except for any dated records that might be found among the contents of the deposit itself, there was no other indication of a lower limit to the date than the style of the frescoes which covered the walls of the passage. According to the Tao-shih's explicit statement, which the actual condition of the surface of the wall around the opening into the side chapel bore out, mural painting of the same type had also covered the plaster facing the brickwork which closed the opening. The frescoes represented over life-size Bodhisattvas marching in procession with offerings (Fig. 200). They were very well painted, and fortunately the Tao-shih's restoring zeal had not interfered with them. Their style was met with again in a number of caves, as shown by the passage walls of Ch. vii and Ch. ix (Figs. 214, 225), the mural decoration of which had not suffered from any recent restoration. It seemed to me difficult to believe that these frescoes could be later than the period of the Sung dynasty. But obviously at a site where so much of old art work survived to inspire and guide successive generations of local artists, and where restoring activity can be proved to have continued intermittently for centuries, from Tang to Mongol times, no absolute reliance could be placed on indications of style only.

Thus archaeological evidence from the first gave encouraging hope that this big hoard would prove to contain manuscripts of importance'and interest beyond the range of Chinese translations of Buddhist canonical literature ; for during the period which is indicated by the extreme limits mentioned above, and for some time before it, this westernmost portion of Kan-su had been the meeting-place of varied races and contending political p6wers, and among all of them we know Buddhism to have been widely spread. All the more I felt the misgivings which the very hugeness of the deposit was bound to inspire as to the possibility of a thorough methodical search. The limitations of my philological knowledge would not permit of rapid selection of what might be of special interest amidst these masses of Chinese texts, and without adequate time it would be difficult even to pick out any non-Chinese materials that might be hidden away among them. But foremost of all was the apprehension that the timorous shifty priest, swayed by his worldly fears and spiritual scruples, would be moved in a sudden fit of alarm or distrust to close down his shell before I had been able to extract any of the pearls. Eager as I felt to push on with all possible energy and speed, there were obvious reasons on the other side to display studied insouciance and to avoid whatever might cause 'the Tao-shih to attach exaggerated value to his treasures.

The interest and fascination of these novel labours of excavation may justify my presenting first a quasi-personal record of the discoveries, and of the impressions which accompanied them. A general survey of the different classes of manuscripts and other antiquities brought to light will be attempted in subsequent chapters. It was in keeping with the prevalent character of the old monastic library which had found here a safe place of refuge that the first bundles which the Taoshih brought us from it consisted of thick rolls of paper, from about 92 to Io; inches in height, evidently containing Chinese translations of canonical Buddhist texts or Chinese treatises on them. Most of them were in very good preservation, and yet showed in paper and details of arrangement unmistakable signs of great age. The jointed strips of smooth yellowish paper, very close in texture and hence remarkably strong and tough, were usually found, as the specimens reproduced in

See below, Appendix A, III.

Wall-paint. ings of passage.

Difficulties of searching hoard of manuscripts.

Chinese manuscript rolls first examined.

1114   5L