TI-tE great oasis of Khotan, to which I returned on August 5, 1906, after an absence of over Stays at five years, had been the centre and base of my archaeological work on the former journey. After Khotan.
all the attention and labour which I then devoted to the study of its ancient topography, history, and extant remains, and of which the results have been fully recorded in Chapters VI–VIII of Ancient Khotan, it was obvious that the short stays which I made at Khotan town in August and again in September, 1906, could serve only for the collection of information from local ` treasure-seekers ' about possible desert sites awaiting exploration beyond the oasis, and for preparations for travel further afield. Hence what I have to record here is confined to a few supplementary observations about remains still visible above ground, and to a short account of such acquisitions of antiques as I was able to make from Yôtkan, the site of the ancient Khotan capital.
To take the observations first, I may mention that, when marching from Kara-Kash town to the city of Khotan by the direct route I had not previously visited, I came across unmistakable indications of a ` Tati', i. e. the wind-eroded ground of ancient occupation, by the eastern edges of the narrow tongue of sandy desert, known as Balamâs-kum, which projects into the oasis from the north between the irrigated lands of Sipa and Laskuya (Map No. 2o). The red pottery débris strewing the ground, where not covered by dunes, looked decidedly old, and proves that cultivation in ancient times was continuous in the northern part of the oasis between the Kara-kash and Yurungkash Rivers. It is probable enough that the cultivated area then extended in this direction considerably beyond the line marked by the present outposts of the main oasis, the villages of Yangi-arik on the Yurung-kash and Sarïgh-yez by the Kara-kash. At the time of my visit steady expansion was taking place in the cultivated ground of the oasis, mainly, it seemed, owing to improved economic conditions and increased population. It was significant evidence of this process, to which I have had repeated occasion to call attention in my personal narrative,' that the desert enclave of the Balamaskum was then undergoing rapid reduction by irrigation channels brought to its edges for the sake of new fields. Thus the 'Tati' referred to is bound to disappear soon under fresh cultivation and to be hidden more and more by the steady accumulation of fertile loess which accompanies all irrigation in this region.2
I could see this process illustrated also at the extensive ' Tati' area south of Jamada village, which a canal newly opened is helping gradually to reclaim. The same change was said to be taking place at the large site of Chalma-kazan some four miles higher up by the left bank of the Yurungkash, where the big débris-strewn waste I had seen in 1900 was being brought under cultivation again.3
See Desert Cathay, i. pp. 164, 169, 23o, 251 ; ü. 2 Cf. Ancient Kholan, i. pp. 196 sqq.
p. 416. 3 For these sites, cf. ibid., i. p. 233.