770 EXPLORATION IN CHINESE TURKESTAN. 



•But whatever revelations of interesting- detail may be in store for 

 us, one important historical fact stands out clearly already. The use 

 of an Indian language in the vast majority of these documents, when 

 considered together with the secular character of most of them, strik- 

 ingly confirms the old local tradition recorded by Hiuen Tsiang that 

 the territory of Khotan was conquered and colonized about two cen- 

 tuiies before our era by Indian immigrants from the northwestern 

 Punjab. It is a significant fact the Kharoshthi script used in our 

 tal)lets was peculiar to the very region of ancient Taxila, which the 

 above tradition names as the original home of those immigrants. It is 

 strange, indeed, that the ruined dwellings of a settlement far away 

 in the barl)arian north, overrun ])y what Hindu mythology knew as 

 the "great sand ocean," should have revealed to us, after nearly two 

 thousand years, the oldest written documents (as distinguished from 

 inscriptions), and of a type of which ancient specimens have never come 

 to light as yet in India proper. It is equally strange, and 3^et easily 

 explained by the historical connection of Khotan with China, that we 

 should find buried along with them what are likely to prove the oldest 

 written Chinese records actually extant. 



There is ample evidence to show that this remarkable site must 

 have been deserted already within the first few centuries of our era. 

 Apart from the Kharoshthi writing of the tablets and leather docu- 

 ments, which agrees closely in its palteographic features with the 

 Kharoshthi inscriptions of the Kushana kings of the first and second 

 centuries, there is the eloquent testimony of the coins. The very 

 numerous finds, extending over the whole area, which were made dur- 

 ing my stay include onl}- copper pieces of the Chinese Han dynastj^, 

 whose reign came to a close in A. D. 220. The use of wood as the 

 only writing material, apart from leather, is also a proof of great 

 antiquit}'. The use of paper for writing purposes is attested in Chi- 

 nese-Turkestan from at least the fourth century A. D. onwards; yet 

 among all the ruined houses and ancient rubbish heaps not the smallest 

 scrap of paper was discovered. 



After three weeks of almost incessant excavation work I left this 

 fascinating site which had yielded such rich antiquarian spoil in order 

 to visit, farther to the east, ruins I had heard of at Niya. A march of 

 about 100 miles through the desert, due east of Imam Jafar, brought 

 us to where the Endere stream is lost in the sands. After a day's 

 march farther to the southeast I found a ruined Stupa, and at some 

 distance from it a small circular fort filled with sand-buried buildings. 



My excavations at what proved to be a Buddhist temple, situated in 

 the very center, brought to light some interesting stuoco sculptures, 

 and, besides, a considerable quantity of manuscript leaves on paper. 

 They belong to a variety of texts in Sanskrit, Tibetan, and the 

 unknown language written in Indian characters, already referred to in 



