Condition of ruined
116 THE ROUTE FROM KARGHALIK TO KHOTAN [Chap. V
Antiquarian evidence of more extensive cultivation in former times is furnished by an old site called Kul-langar (` the rest-house by the tank '), which I was shown about two miles to the north-west of the central village. I found there the ground near the edge of the cultivated area strewn with pottery fragments over about half a square mile. I could discover no ornamented pieces among them, nor was there any striking evidence of erosion. Among the débris, on and below the ground, bones, layers of ashes, and charcoal were plentiful. Hence the soil is considered valuable for manuring, and is carried away from numerous pits. In one of these pits I found the strata of decayed refuse and ashes extending to a depth of six feet below the surface. At the northern end of the site, high above a watercourse now dry, are two empty tanks enclosed by circular mud bands. The larger one measures about 6o yards in diameter. Old copper coins are said to be picked up occasionally on the site, but I could not obtain any of them. The potsherds looked coarse, though little affected by erosion. The general impression I brought away from Kul-langar was that of a village site abandoned at a not very distant period.
The march from Zanguya to Piâlma, the next oasis towards Khotan, passes over ground which throughout shows the character of true desert. At a distance of about two miles from the edge of the cultivated area, low dunes of coarse drift-sand begin to cover the bare loess ; some distance further south-east it changes into a hard pebble ` Sai '. About six miles from the centre of Zanguya, at a spot called Kizil-Tam (` the red wall '), I found a small plot of ground thickly strewn with fragments of old pottery, of a fine red colour, very hard, but entirely without ornamentation. The surrounding drift-sand may cover more. Close by is passed the dry bed of a stream, which probably carries occasional flood-water from the outer hills of Duwa.
Already at Moji I had heard of a ruined mound called Karakir-Tim (` the mound of the black ridge '), which was to be seen not far from the road to Piâlma. At Sai-Langar, a lonely rest-house some i 3 miles from Zanguya, where water is obtainable only from a deep well, we struck off the caravan route, and proceeding east by north over heavy dunes of coarse sand for two miles reached the mound. It proved to consist of a solid mass of sun-dried bricks badly decayed on its surface, yet furnishing evidence by its size and proportions that it represented the remains of a Sttipa. Considerable portions of the masonry have broken away, especially on the south side, while elsewhere the outlines were hidden under deep masses of crumbling débris.
A careful survey of the mound by means of the plane-table (see Plate XI X) showed that at its foot it formed a square of approximately 65 feet. But owing to the sand which from a high dune close by reaches the northern and western sides, and on account of the débris-covered slopes on the other sides, the actual outlines of the base where it touched the ground could nowhere be traced with absolute clearness. Higher up the contours were still more irregular, as will be seen from the plan. Nevertheless, the outlines visible in a photograph (Fig. 19) taken from the west still give a faint indication of what must once have been a dome surmounting a square base. The highest elevation of the mound above what appears to have been the original ground-level, as measured at the south-east corner, is about 22 feet. The top now forms an elongated small flat, showing that a considerable portion of the original superstructure must have crumbled away. In this respect, as well as in its general decay, the mound closely resembles the Stûpas of Kurghan-Tim and Kizil-Debe. Owing to the large masses of fallen masonry, I doubt whether even an extensive clearing of the débris would permit of approximately exact measurements being obtained of the several stories, &c., of the base and dome.
The bricks, which were exposed in many places, showed a fairly uniform size of i 6 inches square, with a thickness of 31 to 4 inches. These dimensions agree accurately with those recorded