282 THE RUINS OF DANDAN-UILIQ [Chap. IX
but convinced myself by the absence of any movable objects except a plain wooden cornice, as well as by other indications, that it had been ransacked long ago. On this account I did not consider it useful to extend the labour to a large apartment eastwards traceable below the sand. I had subsequently reason to regret this omission, for it was probably from among the rubbish which treasure-seekers may have thrown outside on occasion of previous diggings that the curious relics were obtained which some of my labourers brought to me after I had proceeded to Rawak $. The room cleared was provided with a large fireplace, but had no door, from which I concluded that it was entered from an upper floor and served as a basement, perhaps for use during winter.
The numerous remains of ancient trees, bleached and splintered, in the neighbourhood of this ruin may have belonged to the orchard or garden once surrounding the dwelling. Among the trunks still rising above the sand my labourers were able to distinguish apple and apricot, as well as the ` J igda' (Eleagnus angustifolia), with the wood of which all Khotan cultivators are familiar from their own homesteads. A little to the west of the ruin remains of a fence made of rushes indicated an ancient enclosure such as could be traced also near the building D. v, and such as I subsequently noticed often near the ruined dwellings of the Niya Site. By scraping the sand-covered bank of a small depression formed through wind-erosion near the edge of the area which this fence appeared to have enclosed, Turdi laid bare a closely-compressed mass of straw, evidently once deposited in the corner of a fenced courtyard. The straw, though darkened and, of course, completely dried by the long centuries that had passed since its burial under the drift-sand, was in a remarkably well-preserved condition, so much so that Turdi looked upon it at first as a providential means for saving the pony which he had been ill-advised enough to bring to the site without my knowledge. The poor famished animal at the beginning swallowed ravenously this most ` desiccated ' fodder stuff, without, of course, being saved thereby from the end related in my Personal Narrative 4.
The arrival of Ram Singh, who on December 24 had safely joined me (as previously arranged) from the side of the Keriya river, after completing the survey of the high snowy range between Karanghu-tagh and Polu, enabled me to leave, for short periods, the actual excavation work under his supervision, and to reconnoitre, under Turdi's guidance, the outlying portions of the ruined site. It was in the course of these excursions that I acquired reasonable assurance that, apart from the detached ruins to be noticed, there were no remains of any size
recognizable below the drift-sand. The small ruin D. xvt, nearly one mile to the south of
my camp, and at the southern edge of the area over which pottery débris could be traced, proved to be the remains of a little cella nine feet square, destroyed to within a foot of the ground. Nothing but a shapeless mass of plaster in the centre, probably the last trace of a statue base, was found by clearing it. Proceeding to the north-east of my camp, all traces of old occupation, such as pottery débris and remains of trees, ceased within half a mile, the dunes rising here to 15 feet and more. Crossing the western offshoot of the ` Dawân' marked on the plan, I came upon what looked like the remains of a circular mud wall cropping out in five or six places between the closely packed dunes. No exact survey could be made, owing to the height of the latter ; but from the measurements taken by me it appeared that the circular space enclosed by the walls had an approximate diameter of 26o feet. The walls, which could be traced only in the form of low mounds with horizontal layers of rushes between the stamped loess, seemed to have had a width of about 35 feet at their base. Owing to the conditions of the sand, excavation