Of the collection of small Khotan antiques brought by Badruddin Khan four sets have been separately listed owing to the interest attaching to their alleged provenance. The stucco relief fragments, marked as from Chalma-kazan and Kara-sai, agree very closely in character and in their material, a peculiar hard white plaster, with the stucco remains recovered by me from these two sites in 1908.' The finely designed bronze vase (Chal. 017, Pl. X), showing low relief ornament in Chinese style, was also said to have been obtained from Chalma-kazan. The origin of the few miscellaneous relief fragments, & c., said to have been found at Kohmari, the ancient sacred site of the Go§ringa hill,8 cannot be verified in this way. But the attribution deserves nevertheless to be specially noted, as it suggests that possibly the site retains some relics of the Buddhist convent that Hsüan-tsang mentions as standing near its sacred cave, though on my visits in 1900 and 1908 to the Mazar that now marks the site, I failed to trace any.9
Finally I have deemed it advisable to keep apart the collection of miscellaneous small antiques acquired by me at Khotan from Tokhta-akhûn, a ` treasure-seeker ' already well known to me, as his statement that they were found by him mainly during searches of the adjoining ` Tatis ' of Arkalik and Hanguya is borne out by the general character of the objects, which agrees with that of my own finds at Ak-terek and elsewhere in this area.10 This collection is now marked Ark. Hang. The occurrence in it of objects undoubtedly belonging to the early Muhammadan period, such as the inscribed bronze fragment, Ark. Hang. 048, Pl. X, and the carnelian seal, o66, and of ceramic ware with green glaze (030, Pl. IX ; 061), is in full accord with my chronological observations regarding the ` Tatis ' at and near the Ak-terek site.11 At the same time sôme, if not most, of the terra-cotta figurines and fragments, among which the wrestling monkeys (010, Pl. II), the graceful miniature vase (025, Pl. XI), and the appliqué frog (031) deserve special mention, may well have been added from Yôtkan or Tam-öghil.'2
On the last day of my stay in the Khotan oasis, November 29th, I was able to examine a small site close to the south-eastern edge of the Sampula canton and about a mile to the south-west of Kotaz-langar (Map No. 14. A. 3). From there I had received in 1908 a few stucco relievo fragments, evidently from a Buddhist shrine, and again during my latest visit to Khotan town some fragmentary leaves of paper in Brahmi script and a variety of small remains to be noted presently 13 On the lowest of the narrow gravel ridges which mark where the foot of the westernmost spur descending from the Tikelik-tagh ends just above the cultivated area of the village of Jirak, Abbas, one of my ` treasure-seeking ' party of 1906-8, pointed out to me the spot which had yielded the above-mentioned relics. It proved to be situated about 35o yards to the SSE. of a Ziarat, known as Toghrak-mazar, and about 150 feet above the level of the adjoining cultivation. Small debris of stucco, evidently from relievo decoration of walls, and of the same type as the fragments received in 1908, showed that a small shrine similar to those found at Khadalik and Dandan-oilik had once stood here. But it had been destroyed so completely that no trace of its walls survived. From the plentiful reed straw and dung mixed with the tiny debris it appears probable that the ruin was at one time used as a sheep pen. That the posts, &c., of its walls had been subsequently worked up for use elsewhere, just as at Khadalik, was suggested by the abundance of wooden chips.
By clearing the ground about 20 yards to the south of the destroyed shrine we brought to light a rough pottery jar, 102 inches high, provided with two handles and closely resembling in shape
Miscell. an tiques from particular sites.
7 Cf. Serindia, iii. pp. 1266 sq., 127o sq., 1273, 1280 sq.
8 See Ancient Khotan, i. pp. 185 sqq. ; Serindia, iii. p. 1267.
9 The point is of some interest in view of the alleged discovery of the famous Dutreuil de Rhins MS. of the Dhammapada at the same site ; cf. Ancient Khotan, i. p. 188.
1° See Serindia, i. pp. 141 sqq.
11 Cf. ibid., i. p. 140.
12 See Anc. Khotan, i. pp. 472 sq.
13 The stucco fragments acquired in 3908 are described and their provenance briefly indicated in Serindia, iii. pp. 1266, note 9 ; 1270. The MS. fragments received in 1913 are marked Savrp. 08-44 ; see Mr. Pargiter's Appendix E.