Sec. v] RELICS OF TEXTILE ART FROM THE TOMBS OF ASTANA 675
freehand and elsewhere by stamps or stencils.14 In two silks a spot pattern is produced by knot-dyeing, a technique still largely practised in north-western India and of which there is also an example in a fabric from the Tang fort of Endere.15
But, as in the case of the textile relics from Chien-fo-tung, the special archaeological and artistic interest of the Astana silks lies far more in the designs they exhibit than in the techniques and methods of ornamentation employed. Among their designs two main classes can readily be distinguished. The first class comprises designs either characteristically Chinese in style and execution or composed of motifs which, if not exclusively Chinese, may yet be reasonably assumed to have originated and been applied in the early textile art of China independently of Western influences. To the other class belong the designs which share the essential features of style peculiar to the decorated silks produced in Iran and the Near East during the period conveniently designated as ` Sasanian ', or in which the treatment shows them to have been imitated by Chinese hands from ` Sasanian ' examples.
I have had occasion, when dealing in Serindia with the decorated silks from the ` Thousand Buddhas ' of Tun-huang, to emphasize the importance attaching to certain fabrics found among them of which the designs belong to this second class. They serve as witnesses of that artistic penetration of Iranian art into the Far East in which textiles of ` Sasanian ' type were the transmitting agents, and of which the result is clearly reflected in the designs of some well-known silks preserved since the middle of the eighth century among the temple treasures of Japan.16 Some of the ` Sasanian ' figured silks from Chien-fo-tung described in Serindia and certain others subsequently recovered by M. Pelliot from the same hoard 17 can be recognized with certainty as direct imports from Western Asia, while some are undoubtedly Chinese imitations of similar fabrics. To these the tombs of Astàna have now added specimens, both of original ` Sasanian ' textiles and of fabrics produced under their influence, which are not only more numerous but also manifestly older. The terminus ad quern in the case of these Astàna specimens lies fully three centuries farther back than in that of the Chien-fo-tung materials, and they may safely be assumed to be approximately contemporaneous with the fabrics the designs of which had served as models for the Chinese silks imitating ` Sasanian ' style preserved at the temples of Nara.
Turfân, notwithstanding the important Chinese element in its population so strikingly attested by the discoveries in the Astana cemeteries, yet belongs to a portion of innermost Asia in which Iranian influences have strongly asserted themselves during a prolonged period. Hence we seem justified in first reviewing here those Astâna textiles which show purely Sasanian' designs and must therefore be considered as products of Western or Central Asia. Next we shall turn to those in which characteristic features of ` Sasanian ' textile style have been copied and adapted by the hands of Chinese workers. In adopting this order we may note the significant circumstance that whereas distinctively Chinese designs vastly preponderate among the many silks of Chien-fo-tung, and those of ` Sasanian' type are very few, the ` Sasanian ' specimens at Astana make up a considerable proportion of the total of figured silks and close upon one-half of those executed in colours.
In dealing with the designs of each class it will not be possible for me to attempt a systematic analysis of all details. I have before me at the time of writing neither adequate reproductions of the silks nor drawings of patterns, which are so necessary for the illustration and study of fabrics where these have for the most part been preserved only in small, and frequently much injured,
14 See ii. I. 014-15 ; vi. 2. 04, 3. 03, 07 ; ix. 2. 012 (freehand) ; Pl. XXXVI, LXXVIII, LXXXII.
15 See vi. I. o1—z ; Anc. Khotan, i. pp. 430, 442 ; ii. Pl. LXXVI.
16 Cf. Serindia, ii. pp. 907 sq. To the references there given in note 1 should be added von Falke, Seidenweberei (first ed.), i. pp. 87 sqq., Figs. IIO-19,
17 See Serindia, ii. p. 907, note 6.