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

Sec. Hi]   ACQUISITION OF MANUSCRIPTS AND ART RELICS   821

a safe cover of sand would have provided at a ruin definitely abandoned to the desert was not to be looked for in a hoard first opened and searched by an ignorant and careless priest like Wang Tao-shih.

It will be convenient to record here the scanty archaeological indications of the original character of this grotto which I gathered when it was completely cleared for a brief time. The exact measurements of its ground plan (Plate 43) were 9 feet from east to west and 8 feet 8 inches across. Opposite to the entrance and thus in front of the north wall there rose a plastered base, 5 feet long, 2 feet wide, to a height of I foot 8 inches. Its shape and position make it appear quite certain that it was intended as an image base. No remains of stucco sculptures or relievo haloes could be traced, and of mural painting there survived only decorative tracery work in faint colours on the upper portion of the north wall. Elsewhere the stuccoed surface of the wall was plain. Owing to the narrowness of the doorway, only 3 feet across, the light in the small chapel must at all times have been very poor. Hence it appears to me very unlikely that the large inscribed slab which reproduces the imperial edicts honouring Hung jên could have been originally placed there. At the same time the fact that it was carefully set up in a recess carved from the rock of the west wall does not seem to me to point to hurried removal under the stress of a sudden danger. It appears more probable that during a period of advancing decay, such as the sacred site might have witnessed during the decennia preceding the Hsi-hsia conquest, pious monks wished to assure here a better chance of protection for the inscription which glorified the influence and imperial honours enjoyed by a former ` head of the Buddhist religion at Sha-chou '. Whether the transfer took place when the manuscripts and other contents of the hidden deposit were collected here, or some time before, it is no longer possible to determine.

From the first it was obvious to me that the objects deposited in this chapel might often have been of considerable antiquity at the time when the deposit was finally walled up. But not until a year later, when Chiang Ssû-yeh was engaged on preparing a rough inventory of at least a portion of the Chinese manuscripts brought away from Chien-fo-tung and could find time for completely unfolding the Sûtra rolls, etc., in search of their colophons, did I receive the gratifying assurance that a considerable series among them showed exact dates which reach back as far as the beginning of the fifth century A. D. (see for specimens Plates CLXVI, CLXVIII). Thus, utter alia, the date corresponding to A.D. 416 of the interesting original record of a Tun-huang Census (Ch. 922, Plate CLxvI), which Dr. Giles has since published, was then correctly determined.2 That it would need protracted scholarly labours in Europe before the date of the earliest piece among the collection secured could be definitely established was then already certain ; 2a and even now, after another nine years, it is impossible to foresee when they may be carried to completion.

Not knowing how long we might rely on the Tao-shih's indulgence, all I could do during those first days at his cave was to work in great haste through the contents of the ' mixed' bundles. With the constant flow of fresh materials pouring down upon me, there was no chance of closer examination even in the case of art relics and of such manuscripts as were neither Chinese nor Tibetan and of which, consequently, I was able myself to estimate the full interest. All I could do was to assure their being put apart ` for further study', as we styled removal in diplomatic convention. More bitterly than ever did I regret the great hindrance created by my total want of Sinological training.. Amidst the smothering mass of Buddhist canonical literature Chiang Ssû-yeh's zealous

Archaeological indications from wàlledup chapel.

Earliest dated record (A. D. 416).

Hasty search of mixed ' bundles.

2 Cf. L. Giles, A Census of Tun-huang, T'oung pao, 1915, pp. 468 sqq. I regret that Pl. CLXVI, through a mistake explained by my absence in India, shows the reverse of this interesting roll and only a small portion of its obverse. The

text for which the blank paper of the reverse has been utilized is that of some later Buddhist manuscript.

2a [Dr. Giles has now found a Vinaya text with colophon dated A.D. 406.]