comprised in the acquisitions from Yotkan and Khotan and partly reproduced in Plates V, VI, VII. But this does not detract from their interest, since, if not actually produced in the Khotan region, they are tangible proofs of that importation of models through which the artistic influence from India and the West must be supposed to have directly asserted itself. Thus it may be considered certain that the remarkably well-carved small steatite sculpture Yo. 00121 (Plate VI), showing the crown of a miniature Stûpa, with its succession of umbrellas resting on figures of Buddhas and grotesque animals, came to ancient Khotan from Gandhâra. The same conclusion is probable of the small steatite relievos Yo. 00138 and Khot. 04. e, and of the well-modelled nude figure, Khot. 02. o, in a slatey kind of stone (Plate VI). In the case of other steatite carvings (Yo. o5. b–d, oo82, o0120, 00134, 00165 ; Khot. oo6 ; Plate VI) the indication of style is uncertain, while for Yo. oo166 the monkey figure points distinctly to Khotan. Local origin is made probable both by the jade material and the carving for the small monkey figure, Yo. o6. a ; the bird, Khot. 04. f (Plate VI) ; and the engraved slab, Yo. 0091. a (Plate VII). Of other jade objects a ring, Khot. oo61, a miniature vase, Yo. o6. f (Plate IV), and a fragmentary buckle, Yo. 00152, may be mentioned.
Of miscellaneous metal objects, mostly in bronze, of which Plates vi and VII show a number (Knot. 02. C, h, 005, oo8, 009, 0020, 0025, 0046, 0050; Yo. 05. a, 00105, 00118, 00139), it is still more difficult to determine the place of production. But in none of them does the style of modelling or ornamentation differ noticeably from that otherwise known to us in the art work of ancient Khotan. Of distinct interest are the leg of a small stand in bronze, Yo. co' 73 ; the octagonal ferruled ornament resembling a mace-head, Yo. oo8i, in the same metal ; and the curious object resembling a Janus-head, Yo. 00174, the purpose of which remains doubtful—all shown in Plate VII. Of the arrowheads, triangular in section, Khot. 003, 0017 (Plate VI), 0047, 0048, it is noteworthy that their shape closely approaches the type which specimens like T. x11. 0020 (Plate LIII), from the ancient Limes of Tun-huang, prove to have been in use during Han times. Instructive as showing the methods by which some of these small bronze objects were cast, are the clay moulds which were brought to me as having been found at Tain-öghil (Tam. 001-004). As detailed in my former Report, gold-washing operations are carried on at this place, near the north-east edge of the Khotan oasis, among ` culture-strata ' closely resembling those of Yotkan.6
The tiny pieces of gold jewellery, Yo. 00127 (Plate vI), acquired at Yotkan and showing in part very fine work, illustrate a class of ` finds ' which, unfortunately for archaeological interests, is but very rarely preserved from the melting-pot. Among the numerous beads in glass and stone, those inlaid with designs in white (Yo. 00114, 00125 ; Khot. 02. q, r, 0069 ; see Plate Iv) may be singled out for brief mention, as their technique is still in need of expert determination. Of distinct interest is also the bead of millefioriglass mosaic, Khot. 0072 (Plate Iv), as it shows a regular Western type which was common in the Roman Empire. Imports of Western glass ware into China are attested down to relatively late times, and while it is still uncertain whether glass was manufactured in Central Asia also, such indubitable proof of imports direct from the Mediterranean region has its value.
Of the seals acquired from Yotkan or Khotan town most will be found reproduced in the lower half of Plate V. Whether in metal or stone, they all show so close an approach in type to the corresponding finds and acquisitions made during my previous journey at sites of the Khotan region (see Plates XLIX, L of Ancient Khotan) that their local origin can scarcely be doubted. It is very different with the intaglios, of which fifty are reproduced in the upper half of Plate V, and for the determination of which Professor Percy Gardner has been kind enough to give help. Here we meet with pieces which are undoubtedly late classical work and others which, though Oriental,
6 See Ancienl Kholan, i. p. 472.