812 KUCHA AND SOME OF ITS ANCIENT SITES [Chap. XXIII
My local informants knew of no finds of antiques ever having been made in this area. The interest of these earthworks lies solely in their plan. This differs strikingly from that of any other of the ancient circumvallations which I have seen in the course of my Central-Asian explorations. In their irregularity and apparent roughness of construction, the triple ramparts of Öch-kat distinctly suggested to me an origin different from that of the walled enclosures examined at ancient sites in the Kuchâ territory, or elsewhere in the Tarim basin. My thoughts naturally turned to such prehistoric earthworks as the ` rings ' found in widely separate regions, from England to the Caspian and beyond.
Whether the ramparts of Och-kat are likely to date back to a period anterior to that with which Chinese historical records and extant remains in the Tarim basin acquaint us, or whether they may be the work of some later invaders less advanced in civilization, it is impossible for me to say. There is, however, a curious notice in the account of Kuchâ contained in the Chin Annals which deserves to be mentioned in this connexion. It states that the people of Kuchâ ` have a walled town with suburbs, and the ramparts there are triple '.5 It is impossible to assume that the capital of Kuchâ could ever have occupied the position of Och-kat, were it only for the reason that the text further observes that ` within it there are a thousand Stûpas and temples of Buddha '. But it is certainly interesting to find this particular feature of the defences of the Kuchâ capital of Chin times illustrated to the present day by the triple ramparts of Och-kat.
We camped that night two miles off, at Torpak-bâzâr, the principal market-place of the Yulduzbâgh tract, which, together with Chilan and other villages in the westernmost cultivated portion of Kuchâ, is now included in the hsien or district of Shahyâr. On the morning of April 24th we set out for the ruined sites of Tajik and Toghrak-akin, which had been reported to me as the most westerly that showed structural remains. The road leading to them was said to be followed at times by travellers who wished to use the direct route to Ak-su, which crosses the scrubby desert south of the barren hill range that bounds the subsidiary basin of Bai. After proceeding about four and a half miles we reached the edge of continuous cultivation, but patches of newly reclaimed land intervened between strips of steppe for another two miles. There the ruins known as Kosh-iura came into view, about a mile to the north of the last fields. We had already passed some ruined enclosures of small size, with walls of stamped clay, near the farms of Shar-yailak. At Kosh-tura I found a massive tower, 45 feet square at the base and still 34 feet high, built of rough slabs of clay (kisek). Some 86 yards north of it stands a ruined platform, measuring about 46 feet by 42 at its base, constructed of the same material but with thick layers of tamarisk brushwood inserted into this masonry at vertical intervals of about 3 feet. This indication of antiquity was confirmed by the examination of the top, about 18 feet from the ground, which appears to have once borne a shrine. Mir Sharif stated that, some eight years before, he had seen walls standing some 6 feet above the solid masonry base, and bearing traces of painting. Small fragments of coloured stucco were, in fact, still to be picked up, evidence of the utter destruction that the ruin had undergone. About 6o yards to the east the low remains of a wall, built of stamped clay and tamarisk layers, could be traced for a distance of some 70 feet, all that remained of what seems to have been an enclosure. There was evidence here that wind-erosion had some share in the work of destruction.
Moving from Kosh-tura to the west across a clay steppe covered with scanty tamarisk growth, and in places with low dunes, we gradually approached the foot of the bare serrated hill range already referred to ; it rises here to a height of about 2,000 feet above the alluvial plain. The site of Tajik, reached after a march of a little over six miles from Kosh-tura, lies, as the plan in Pl. 41
6 See M. Chavannes's translation from the notice in the Chin Annals on Kuchâ, in Anc. Kholan, i. p. 544 ; also S. Lévi, J. Asiat., 1913, Sept.-Oct., p. 333.