128 REMAINS OF THE KHOTAN OASIS [Chap. IV
and of the interesting specimens of Khotan sculptural art then unearthed, all that survives now, I fear, are my photographs.3 With this evidence of rapid destruction before me, I felt glad that the movement of the dunes just indicated meant increased protection for whatever sculptural work may lie buried under sand on the other sides of the Vihâra court.
This movement had changed also the aspect of the Stûpa itself. Five years before its imposing base of three stories had been almost entirely covered by drift sand except to the south-east. Now its upper portion, with most of the second story, emerged free on all sides (see Fig. 36). Thus the previous delineation of the ground-plan of the base, with its strongly marked cross-like projections bearing the four flights of stairs, could now be fully verified.4 On the newly exposed parts of the
base more burrowings, evidently of old date, had become visible. The height of sand filling the Stûpa court made a complete clearing of the latter now quite as impracticable, without a disproportionate expenditure of time and money, as before. But by a small excavation at the salient angle next on the south to the north-eastern flight of stairs it became possible to secure the section of the elaborate mouldings which the lowest base shows near its top and foot. On the other hand, the change of the sand conditions obliged me to renounce the hope of clearing the sculptures which probably still exist intact along the north-west wall of the Stûpa court ; for whereas the top of this was in 1901 just traceable above the sand, a big dune now completely covered it.
As another illustration of the change in the dunes I may mention that the relatively large patch of bare eroded ground westwards where I had camped during April, I90I, was now entirely buried beneath dunes. The exact comparison of the sand conditions with those observed more than five years before had a special interest. It distinctly supported the view, already suggested, that the high dunes about Rawak
are the direct product of the fine alluvial deposit left behind by the floods of the Yurung-kash River and carried from its banks into the desert in the direction where the alternating east and west winds of this part of the Taklamakan have most play under local conditions.5
Leaving my camp at Rawak I pushed on the same day to the south-west where two of my treasure-seeking' guides had reported some ` Tims ' and a ruined building. We had tramped for three miles up and down closely packed dunes rising up to about forty feet, with beds of reeds and some scrub in the depressions between them, when one of the men recognized the spot where he had seen some five years before a small ruined mound now hidden again by the sand. Two miles further
MOULDINGS or RASE or
AT CORNER NEAR
STAIRS OF N.E FACE
3 See Ancient Khotan, ii. Pl. XIII—XVIII.
4 This peculiar ground-plan of the base has acquired additional interest since I found it reproduced, though on a much smaller scale, at the ruined Stûpa G of the SahriBahlbl site in Gandhâra; see my Archaeol. Survey Report, Frontier Circle, 1912, p. 16. It deserves to be specially noted that the ground-plan of Kaniska's famous Stûpa at Peshawar, excavated by Dr. Spooner (see - the plan in Archaeol. Survey Report, Frontier Circle, I9IO—II), shows
also a close approach to the Rawak shape. Considering that its main features can be distinguished also on a number of the miniature Stûpa models in clay (e. g. So. A. oo6, in Pl. CxXXIX)excavatedbyme at the ruined site of Ch'iao-tzû, southeast of An-hsi, it is clear that this cross-shaped Stûpa base represents a development which was known from the Indus to Kan-su. But where was it originally started?
6 Cf. Ancient Khotan, i. pp. 483 sq.
Remains south of Rawak Stûpa.