Kelpin men who had just searched extensive débris areas of the ` Tati ' type in the wide belt of absolute desert between the arid outer hill chain of Kelpin and Tumshuk on the Ak-su-Kashgar road. As with the ` Taklamakänchis ' of Khotan, it was the spring season which had led them to indulge in this treasure-hunt after a succession of ` Burâns' had shifted the drift-sand and laid previously hidden patches of hard soil bare. Of finds of small objects in precious métal and the like, such as occasionally reward these ` treasure-seeking' expeditions, the men had nothing to tell. But the numerous copper coins which I acquired from them ranged from Han to Tang times and left no doubt about the antiquity of the settlements once occupying the area they had searched.a Further evidence of this was to be found in the character of the small objects of stone, metal, glass, and paste which they had brought away, as described in the List at the end of section iv. Among them it will suffice to mention the intaglio seal in cornelian, Kelpin. 003 (Plate V), with a poorly worked female bust ; the intaglio bronze-seals, Kelpin. 004-6 (Plate V), badly worn but showing the influence of classical models ; the open-work bronze pendants, Kelpin. 0014. a-c (Plate VI), in the form of dancing men, and paste beads of the millefiori type, Kelpin. oo9. a (Plate VI), ooio. a, b (Plate Iv). The technique of the last calls for investigation. Another set of similar small objects in metal-and stone which were subsequently brought to me at Kelpin as having been found on desert ground south of the Kudughun hill (Map No. 14. D. 5) came in all probability from the same débris area. Three bronze seals from among them, Kud. 006-8, are reproduced in Plate VI and include a satyr head in relief, recalling the grotesque heads in Khotan terra-cotta appliqués.
SECTION IV.—DESERT SITES NORTH OF TUMSHUK AND MARAL-BASHI
On May 17 I set out from Kelpin with a party of ` treasure-seekers ' and labourers for the old sites in the desert on the south. The heat of the season, already far advanced, and the difficulty of carrying an adequate supply of water—my brave camels had to be spared all work after Ak-su and were no longer available for transport—made the three days' exploration somewhat trying. On the first day we were obliged to make a long detour skirting the eastern extremity of the outer hill chain of Kelpin, which owing to the very rugged nature of its barren ridges could not be crossed with laden animals. When turning the end of the range not far from Achal, the outlying colony of Kelpin already mentioned, I sighted far away to the south-east a high mound known as the ` Tim ' of Soksuk-shahri. It was impossible to visit it from the route for which we were bound. But the Kelpin men talked of it as a very ancient tower, and let fall the shrewd guess that it might have stood on the line of an old road leading from Ak-su towards Kashgar.
We camped at the foot of the range near the debouchure of a narrow gorge, in which was said to lie higher up a kink, or rock-cistern, occasionally holding water. Next morning we left the desert track which leads on towards Marâl-bâshi past the,foot of the conspicuous Kudughun Peak, and struck to the south-south-west. The bare clayey glacis of the hill chain gave way first to ground covered with scanty tamarisk-cones and drift-sand, and further on to bare dunes rising up w feet or so. After covering close on 14 miles, we arrived at the ruined site of which my Kelpin guides had spoken as Chong-tim, ` the big tower '. About half a mile before reaching it, all the ground left bare between the dunes became thickly covered with potsherds, slag, and similar ` Tati ' débris, and this soon proved to extend over an area of more than two square miles. Towards its northern end, and surrounded by dead tamarisk-cones and dunes reaching to 15 feet or so in height, there rose the tower which has given the site its name ; adjoining on the west was a square and approximately orientated
gee Appendix .8, below. One of these coins, of Tang type, shows four non-Chinese characters which have not been read as yet; see Pl. CXLI, No. 25.