80 REMAINS OF KASHGAR, OASES OF YARKAND AND KARGHALIK [Chap. IV
present ground-level shows a circumference of about 35o feet, could easily be recognized as a ruined Stûpa. The ravages of time had completely effaced the outlines of its constructive features ; but the masonry of sun-dried bricks, of which it was formed, was plainly revealed at several points below the covering crust of crumbling débris, and its interior appears to have remained undisturbed. The bricks, closely resembling in make those of the Kashgar mounds, measure on the average i42 inches square, with a thickness of 3 inches.
The quadrangular structure adjoining measures about 26o feet along its two longer sides, running in the direction of north-west to south-east, and about i 70 feet along its shorter sides, from north-east to south-west. Its eastern corner is separated from the foot of the Stupa mound by a space of only 45 feet. Being unable to trace any masonry in sitzt along the low mounds which mark the sides of the quadrangle, I concluded that these represent the remains of broad walls or ramparts, which were built solid in stamped clay (loess), after the fashion still prevailing throughout Turkestan, and were once surmounted by superstructures now completely destroyed 3. The plentiful fragments of fairly hard bricks, charred wood, and plaster which mingled with the sand covering the surface of these low mounds, seem to be distinct traces of such superstructures. The extant remains of the quadrangular enclosure rose 8 to ro feet above the sandy level of the interior, which evidently had been an open courtyard ; the breadth of the mounds measured on the top varied from about 4o feet on the north-western face to less than 20 feet on the south-eastern face, where the entrance gate may possibly have been situated. Judging from the position and size of the quadrangle, it appears highly probable that it was intended for the accommodation of the monastery once attached to the Stupa.
From Hasa-Tam I had already sighted, through the dust-laden atmosphere, a second group of ruins known as Maur -Tim, situated almost due north. I reached them on the 5th of September, after a ride of about four miles to the north-north-west of Topa-Tim. On the way a deep-cut river-bed was crossed, about 500 feet broad, but now almost completely dry. I was assured that only after exceptionally heavy rain in the mountains does this Sai', which descends from the west and is the direct continuation of the stream passing Osten (or Little) Artush, carry any water so far down its course. Ordinarily its water, mainly supplied by the drainage of the Kara-Jilga Valley above that large village, is completely absorbed by irrigation channels taken off higher up. The fields of Khan-ui to the south, and those of Eski to the north of the river-bed, are now the lowest cultivated areas supplied by these canals. In ancient times canals from the same source undoubtedly reached the ancient sites now under discussion, and probably extended yet further to the south-east. Even at the present day there must be abundance of water in the area actually irrigated to account for the swamps which Colonel Trotter noticed at the point where the line of drainage of the Artush stream joins the Kashgar river, some twenty-five miles below Bû Mairyam Khanum 4. Apart from the water, which at times of exceptional floods may reach that point in the actual river-bed just described, I have no doubt much of the water absorbed in irrigation comes there to the surface again in the form of springs. We shall have occasion to note the same process of percolation and its attending result, the formation of marsh-Iand, in the case of all the rivers reaching the southern edge of the Taklamakan Desert from the mountains above Khotan and Keriya.
g Compare Yarkand Mission Report, p. 26r, where it is noticed that no water ever reaches Khush-toghrak, the point of junction above referred to, from the Artush river during the hot and irrigating season.