502 THE RUINS OF AK-SIPIL AND RAWAK [Chap. XIV
appliances, to improvise which I had neither time nor the technical means, would have meant only Vandal destruction. In any case it would have been a practical impossibility to arrange for the safe transport of such loads over the mountains, whether to Europe or India. All that could be done in the case of these large sculptures was to bury them again safely in the sand after they had been photographed and described, and to trust that they would rest undisturbed under their protecting cover. Of the sculptural pieces already detached and of the smaller relief plaques, I succeeded in bringing away a considerable number as described in the list below 2. I felt greatly relieved when I found, on my arrival at Kâshgar, and later also in London, that the great trouble and labour which the safe packing of these extremely fragile objects had cost me, was rewarded by their having accomplished the long journey without any serious damage.
By April 18 those portions of the Vihâra had been explored which were not actually buried under sand-dunes. A careful examination of the surrounding area revealed no other structural remains ; broken pottery found here and there between the swelling sand-dunes was the only trace left of what probably were modest dwelling-places in the vicinity of the great shrine. The sandstorms which visited us daily, and the increasing heat and glare, had made the work very trying to the men as well as to myself. It was manifestly time to withdraw from the desert. Before, however, leaving the ruins I took care to protect the sculptures of the quadrangle and the foot of the Stûpa base by having the trenches which had exposed them completely filled up again. It was a melancholy duty to perform, strangely reminding me of a true burial.
While it was still proceeding I took occasion to visit 7umbe-kum, the only remaining desert site to the north-east of Khotan from which occasional finds had been reported to me by Turdi and others. I reached its southern edge after going about four miles to the north-east of Rawak. One mile of march had taken us out of the heavy sands surrounding the ruins, and the rest of the way led over low dunes amidst which tamarisks and young Toghrak shoots were growing in plenty. The site itself proved to be a débris-strewn Tati, with only the scantiest of structural remains. Not far from its southern edge I was taken to what Turdi had previously spoken of as a ` Potai'. I found there a small mound of hard stamped loess, about 6 ft. high and some i 5 ft. long, surmounted by a foundation of brick masonry, about 2 ft. high and much broken. It was quite impossible to ascertain the original shape and object of the structure that once stood here. The sun-dried bricks appeared to have measured about 19 by 14 in., with a height of 4 in. The ground for about half a mile around was plentifully covered with coarse broken crockery ; here three small and much worn copper coins without legend were also picked up.
The pottery débris ceased about half a mile to the north-east of the small mound and did not appear again until, continuing in that direction for about two miles further, we neared Turdi's second Potai '. For the greater part of this distance the Kumush-covered Sai which comes from near Tam-öghil, undoubtedly an old river-bed, kept close on our right. About half-way a wu-chat coin was picked up by one of the men. The only conspicuous ruin of yumbe-kum or 7umbe-Kalmak, as according to Turdi the site is also called 3, proved to consist of a broken mass of masonry about 8 ft. high, which seems to have formed originally a square