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

912   TEXTILE REMAINS AND MSS. FROM CHIEN-FO-TUNG [Chap. XXIV

forking at each end, forms an interlacing diaper passing over the whole surface. It is impossible not to realize the striking evidence of wholly Chinese style in the fantastic and yet strangely animated figures of, the beasts as well as in a number of minor details, such as the curious hooked. scroll which decorates the arches and closely resembles the conventional Chinese cloud type. If the arrangement of the pairs of confronting animals might at a first glance appear strangely reminiscent of some ` Sassanian ' design, there is yet plenty to warn us against such a derivation. It is enough to look at the animal figures and the architectural motif in the panels, which could not possibly have been evolved from the stiff circle or oval of a supposed Persian model. On the other hand, the general treatment both of figures and of ornamental details suggests a connexion, difficult to define but all the same distinctly perceptible, with the style of the two figured silk fragments from sites of the ancient Tun-huang Limes, T. xv. a. iii. ooio. a and T. xxii. c. ooio. a, both reproduced in Plate LV.18 A comparison of the Chien-fo-tung fabric with the design of the latter fragment, as shown in Plate CXVIII, with its strange figures of dragons and phoenixes and its wave-scroll border, will best explain what is meant. To this may be added the fact that all three show the identical technique of weave (a variation of ` warp rib '), not found else among the Chien-fo-tung silks, and the same restraint in the use of colours, one serving for the ground and a single other for the pattern.

We owe what is likely to prove the true explanation of the puzzle to fresh and far more abundant finds of decorated textiles of early Chinese origin and to the advantage which Mr. Andrews has already been able to derive from their study, preliminary as it is. It was he who first called my attention to the important fact that, among the figured silks discovered by me in 1914 among the early Chinese cemetery remains of Lou-lan which go back to Han times, there are quite a number with designs which, on the one side, clearly attach themselves to the style of the fragments just discussed and, on the other, appear the likely precursors and harbingers of features we have so far been accustomed to treat as originating in ` Sassanian ' textile style.19 In particular we find there the motif of confronting animals fully established as a feature of decorative textile schemes. In Mr. Andrews' opinion the design of Ch. ooii8, a unique piece in our collection of fabrics from Chien-fo-tung, is as it were a survival from, or descendant of, that earlier Chinese style of textile decoration which has been first revealed by the fabrics discovered at Lou-lan.

It is impossible to discuss or to illustrate their evidence here. Consequently, in accepting the view just expressed, I must in part presume what has yet to be proved. But even thus I may use the occasion to point out that those fabrics discovered in Lou-last grave pits are also likely to throw light on other problems of far wider interest connected with the ancient textile art of the East.

Thoroughly Chinese in origin .and style and showing remarkable perfection in technique and artistic taste, those figured silks afford ocular proof of the powerful influence which the products of early

Chinese textile art must have carried westwards. Of the great commercial and even quasi-political

importance which the trade with the silk brought from the distant Seres and exported to the Mediterranean regions acquired for the whole of Iran in Parthian times, we are abundantly informed.

from historical sources. We know that Chinese textiles, not raw silk merely, were carried to Syria and even further west, to be there eventually unravelled and rewoven in occidental designs.20 In the silks I discovered at Lop desert sites we possess actual specimens of the figured textiles which this trade brought from China. Archaeological evidence makes it certain that they belong to the

19 For descriptions, see above, pp. 783, 785 sq.

19 [For an analysis of selected specimens from these finds, see now Mr. Andrews' paper on Ancient Chinese figured silks, Burlington Magazine, 192o, xxxvii. pp. 6 sqq., with my account, ibid. pp. 3 sqq., of the first discovery of those fabrics

in grave pits to the north-east of the Lou-lan Site.] Cf. also Geogr. Journal, xlviii. pp. 123 sq.

2° The latter interesting fact is attested by an important passage of Pliny, Historia nat., xi. 76; cf. Dalton, Byz. Art and Archaeology, p. 584.

Design resembling silk fabrics from Han Limes.

Designs of figured silks from Loulan cemeteries.

Early Chinese textile art in Lou-lan finds.