7ri2 EXPLORATION IN CHINESE TURKESTAN. 



in the desert. 1 also made a series of excursions for the purpose of a 

 close survey of the old localities within the Khotan oasis itself. This 

 enabled me satisfactorily to settle numerous questions bearing- on its 

 ancient topooraph}^ and in particular to locate almost all the sacred 

 Buddhist shi-ines which are described to us by the earl}^ Chinese pil- 

 g-rims. Their positions were invariablj' found to be occupied now by 

 Muhammadan Ziarats, or Saints' tombs, which form the object of 

 populai' pilgrimage. Local worship can thus be shown to have out- 

 lived the great change in religion consequent on the Muhammadan 

 conquest. Its tenacity has indeed proved quite as useful for the study 

 of the ancient topography of Khotan as it had proved to me before in 

 Kashmir and other parts of India. 



I must restrict m3^self here to a few remarks only concerning the 

 most interesting of those old localities — the site of the ancient capital. 

 Its debris layers, which have furnished b}^ far the greatest portion of 

 the Khotan antiquities, such as terra cottas, seals, coins, etc., acquired 

 by former travelers, lie buried deep below the fields of the little vil- 

 lage of Yotkan, some T miles to the west of the present town. Gold- 

 washing operations, originating from an accidental discover}^ of gold 

 some thirty-seven j^ears ago, have gradually led there to the excava- 

 tion of an area over half a mile s<iuare. The careful examination of 

 the banks thus laid bare showed me that the "culture strata," as I 

 should call them, of Yotkan are composed of the rn))bish that gradu- 

 ally- accumulated during the centuries while the site continued to be 

 occupied by houses, from about the commencement of our era until 

 after the advent of Islam (in the eleventh century of our era). These 

 "culture strata," themselves 5 to 14 feet thick at various points, are 

 covered by a layer of pure soil from 9 to 20 feet in thickness. This 

 layer, which shows no sign of stratification, is manifestly due to silt 

 deposit, the necessary result of intensive and long-continued irrigation 

 such as prevails all over the oasis. Owing to the disintegrated condi- 

 tion of the soil, all the water that is brought down from the mountains 

 1)3' the Yurung-kash and Kara-kash rivers, and subsequently distrib- 

 uted by innumerable irrigation channels, carries an excessive quantity 

 of sediment. The silt thus deposited over all cultivated areas is amply 

 s-ufficient to account for the gradual burying of the rubbish layers of 

 the ancient capital and for other curious observations I have made as 

 to the gradual raising of the ground level throughout the oasis. All 

 antiquarian and ph3"sical evidence combines to oppose the assumption 

 of a great flood or similar catastrophe, such as some earlier European 

 visitors of the site have suggested. 



Among the ancient sites in the Taklamakan Desert which are fre- 

 quented by Khotan "treasure seekers," and which the prospecting 

 parties sent out by me had visited, none seemed to oft'er better oppor- 

 tunities for systematic excavations than the one known to them as 



