402 DZUNGARIA 



width, and we were obliged to search carefully before we 

 could find a place suitable for crossing without harm to 

 our baggage. The water was clear and blue, margined 

 by sand-banks, which showed that in time of flood the 

 river would be at least a hundred and fifty yards wide. 



Beyond the Irtish extended a barren hill-country 

 leading over the Naryn Kara and the Kara-adir to the 

 Kobuk Valley, another centre of nomadic life. These 

 isolated localities, inhabited, probably, by people of 

 different races, are typical of Dzungaria ; to-day we 

 may be with the Turki-Kirei, to-morrow with a Mongol 

 colony ; while a week's journey over uninhabited 

 steppes may separate these from a Chinese town or a 

 Dungan settlement. The unusual variety of races in- 

 habiting Dzungaria gives an interest to a country 

 which might otherwise, from lack of character, be mono- 

 tonous and tedious to traverse. 



This stage in our journey led us past Lake Ulungur, 

 which stores a large volume of water descending from 

 the slopes of the Altai, and apparently absorbs the same, 

 having no visible outlet. We caught glimpses of this 

 lake from the Naryn Kara range, and noted its char- 

 acter, as being that of a typical desert-lake : barren, 

 yellow, desert hills surrounded it ; no margin of green 

 bordered its shores, the only vegetation being a fringe 

 of dead reeds. A Russian explorer, Miroshnishenko, 

 took some trouble in 1873 to prove that the Lake 

 Ulungur must drain into the Irtish, which flows past 

 only four miles away. There is, however, no visible 

 communication between the two, though a Kirghiz legend 

 speaks of a subterranean passage connecting the lake 

 with the river. By measuring the volume of the 

 Irtish some way above the lake, and again below 



