parts of the range, and, as the sun rose higher, drew a veil also over the ice-covered ridges first sighted. A couple of miles further, after crossing a broad but now entirely dry river-bed which, lower down, receives the water of the Kara-su, I came upon the first of the old sites which earlier reports had led me to expect on the march from Guma to Moji. All along the right bank of the ravine the ground was thickly strewn with fragments of coarse red pottery. No ornamented pieces could be found, but the exceptional hardness and glaze of these potsherds showed that they belonged to a period far removed from the present. The extent of the area covered by these scattered fragments plainly indicated the site of a large and thickly inhabited settlement. But no other trace now remained of its existence. The innumerable potsherds invariably rested on the bare surface of loess, with never a trace of walls or more substantial remains below.
When to the east of this old site I had crossed the narrow belt of irrigated ground occupied by the hamlets of Mokuila and was passing once more over a barren scrub-covered Daslit, I sighted to the north-east the mound of which my Guma informants had spoken as Topa Tim. ` Tim ' is the designation given to all ruined mounds about Kashgar, and as the one now within reach looked through my glasses much like an ancient Stupa, I made haste to reach it. It proved a longer business than I anticipated. For our guide insisted on our first following the road towards Moji and crossing the deep-cut bed of a watercourse, now dry. I accepted his guidance much against my instinct, with the tantalizing result that when we had got abreast of the mound, after a two miles' ride, a cation-like ravine absolutely cut us off from it. In vain we searched for a place where the perpendicular banks of loess would admit of a descent to the bottom of this fissure, 40 to 50 feet deep. There was nothing for it but to ride back to the road and start afresh on the other side.
IIIFカラー高解像度
