________________
DECEMBER, 1909.] ARCHÆOLOGICAL NOTES DURING EXPLORATIONS.
301
trouble was amply repaid by the evidence that this small tribe in its alpine isolation had preserved remarkably well the main physical features of that race, represented by the present Galches of the Pamir region, and probably like those of Iranian speech, which in ancient times appears to have extended right through to Khotan and even further east.
5. Khotan and the Tatis.
By September 9 I had returned to Khotan, where preparations for my archaeological campaign and the examination of miscellaneous antiques brought in by t:easure-seekers detained me for some days. Hard at work as I was, I could not help attending a great feast which ChienDarin, the obliging prefect, was giving in my honour to the assembled dignitaries of the district. In spite of the time it cost to get through some thirty strange courses, I appreciated the attention the kindly mandarin desired thus to pay in acknowledgment of the labours I had devoted for years past to the elucidation of the history and geography of Khotan. Then I set out for the desert adjoining the oasis north-eastward, where I succeeded in tracing much-eroded, but still clearly recognizable, remains proving ancient occupation well beyond the great Rawak Stupa. I found the court of the latter even more deeply buried under dunes than when I carried on excavations here in 1901, and alas, the fine stucco relievos then brought to light completely destroyed by treasure-seekers in spite of careful re-burial. But when I subsequently surveyed the extensive dobris-strewn areas known as tatis fringing the north edge of the tract of Hanguya, where potsherds, fragments of bricks, slag, and other hard material cover square miles of ground once thickly occupied, but long centuries since abandoned to the desert, I had the satisfaction of recovering by excavation a mass of interesting small relievos in hard stucco, which had once decorated the walls of a large Buddhist temple, dating probably from the fifth to the sixth century A. D. In their style, uniistakably derived from models of Græco-Buddhist art, these relievo fragments closely resembled the Rawak sculptures. Curiously enough, of the temple itself and the larger sculptures once adorning it, but the scantiest remains had survived in the ground. The probable explanation is that the site hal continued to be occupied for some time after the temple had become a ruin, evidently through fire, and thnt only such smaller stuccoes as had become hardened by the latter into a likeness of terra-cotta could survive in soil constantly kept moist through irrigation.
The finds possessed special interest as proving that even sites so much exposed to erosion by wind and bavoc wrought by human agency, as tatis generally are, may preserve antiquarian relics of interest in lower strata, which neither the slowly scooping force of driven Band, nor the barrowings of treasure-soekers, etc., from the still inhabited area close by, had reached. Another important and curious feature was the prevalence of richly gilt pieces. This furnished striking confirmation of the hypothetical explanation I had given years before of the origin of the leaf gold washed from the culture strata of the old Khotan capital at Yotkan, I may notice in passing that, just as elsewhere along the edges of the Khotan oasis, cultivation in the fertile Hanguya tract is now steadily advancing in the direction of the areas previously abandoned to the desert. The present favourable economic conditions and the consequent increase in the population seem the chief cause for this extension of the cultivated area, which struck me again and again on revisiting portions of the oasis surveyed six years before, and which may yet, given a continuance of those factors, lead to the recovery of a considerablo portion of the desolate tati overrun by dunes and elsewhere undergoing wind-erosion. But it appears to me equally certain that the water supply at present available in the Yurung-kash could under no system whatever be made to suffice for the irrigation of the whole of the large tracts now abandoned to the desert, and for this broad fact desiocation alone supplies an adeqnate explanation.