great antiquity. Neither within nor without the enclosed area could I trace any signs of serious wind erosion, that unfailing impress left by time upon all old sites in this region. In spite of very scanty protection by sand, the rough posts and roof beams exposed in the ruins showed little of that far-advanced bleaching and splintering which by experience I had learned to recognize in timber of old sites as the infallible mark of prolonged exposure to winds and climatic extremes on desert ground. The uniform roughness of construction observed held out little promise of finds archaeologically useful, and the number of the dwellings was embarrassing at the start. Luckily the adequate posse of labourers I had brought with me, reinforced as it was by nearly a dozen men raised at Yar-tungaz, made rapid progress possible in our experimental clearings.
Work was started on two of the less coarsely built structures northward (a, b in plan, Plate 19). The drift sand which filled them did not reach the low roofing made of rough Toghrak trunks, with a layer of brushwood and earth above. But it had sufficed to protect the walls which were formed here of vertically fastened bundles of reeds faced outside with a layer of mud. The timber framework supporting the whole showed none of the careful carpentry I had found at all ruins of the Buddhist period from Dandân-oilik to Endere and which is common also in all modern houses of substantial construction in the Khotan region. Here it consisted of mere unhewn Toghrak posts with other trunks laid across the gable ends, as seen in Fig. 76. The few rooms contained in each of these huts and in those subsequently cleared were found absolutely bare of fittings. There was not even the comfort of a mud-built sitting platform, such as even poor cultivators' houses in the modern oases ordinarily display, and which came to light in even the least spacious dwellings of the Niya Site. Nor did we come upon a single built fire-place ; but in one or two instances a sunk hearth in the rough floor, and a smoke-hole in the roofing above it, showed where the dwellers used to light their fires.
The experience was repeated when we continued the clearing at a series of the small ` houses' built with walls of stamped mud or rough lumps of clay used like bricks. Nowhere did we come upon the remains of furniture or household implements, however humble, with the single exception of the hollowed-out trunk of a Toghrak which might have served as a trough. Even for fragments of pottery, elsewhere the commonest marks of earlier occupation, I searched here in vain. But perhaps the most curious of these negative features was that nowhere, within or without the ruined dwellings and sheds, did we strike any of those accumulations of refuse or dung which in this region invariably adjoin any habitations tenanted for some length of time, whether ancient or modern. It was clear that there was little or no hope here of gaining datable archaeological evidence. So I did not think it advisable to sacrifice more than two days to the site. In making its plan, too, we had to be content with marking the position of each separate ` house and its approximate outside measurements, the division of individual rooms or sheds being indicated only quite roughly.
In the end, close observation of the general conditions prevailing helped to reveal some facts which have a definite bearing on the question as to the origin and character of this curious site. As already stated, I was struck from the first by the absence of marks of wind erosion. This fact was in itself a clear indication of relatively recent date for the ruins. But it helped also to emphasize the significance of certain other features. Near the circumvallation there were extensive patches of ground clear of drift sand. Yet in vain did I look on their flat expanse for any clear traces of the careful terracing and division of fields for irrigation purposes which ancient cultivated soil retains for long periods wherever surface erosion is absent.9 Nor could I find anywhere the remains of fruit-trees or cultivated poplars, though the trunks of dead Toghraks rose in plenty both within and without the enclosure. Many of these Toghraks had died while still young ; but others were big