FROM THE ALTAI TO THE ILI 399 



from the large oases of Southern Dzungaria, all of whom 

 had originally come from the towns of the Tarim 

 basin. Sharasume is an isolated settlement. Its nearest 

 neighbour is Kobdo, at a distance of a hundred and sixty 

 miles, but from Kobdo it is separated by the main chain 

 of the Altai ; Chuguchak in Northern Dzungaria is 

 about two hundred and forty miles distant, while Urumchi 

 and other towns of Southern Dzungaria are over two 

 hundred and sixty miles away. Over these wide stretches 

 of country there are neither towns nor villages. 



We now began to formulate plans for our intended 

 exploration in Dzungaria. Before us lay the fact that 

 the fast approaching winter would hinder all scientific 

 work, the birds were already migrating to warmer regions, 

 the mammals were hibernating ; and a country under 

 such conditions of snow and intense cold would render 

 impossible the carrying on of survey-work. 



Travelling, however, was comparatively easy, and 

 the itinerary drawn up gave us time to traverse Dzun- 

 garia and to bring us closely in touch with its winter 

 aspect before the coming spring allowed us once more 

 to start detailed work. 



Leaving Sharasume on September 24th, we were faced 

 by a forty days' journey along the northern frontiers 

 of Dzungaria, to Kulja in the Hi Valley, where we had 

 our winter quarters for a couple of months during the 

 reconstruction of our caravan and outfit. There, Price 

 left us for a journey through Russian Turkestan and the 

 Caucasus on his way to England, while early in January 

 Miller and I started again, and followed the trade-route 

 which runs the whole length of Southern Dzungaria. 

 During March and April we explored and mapped in 

 detail the extreme eastern portion of the Tian Shan 

 II — 6 



