HILL T ARK AND. 87 
passes tlirongli several snow-beds, or frozen lakes, is 
called the Upper Karakash. Just above our present 
camp it turns round at a right angle and flows nearly 
due west for seventy miles to near Shahidulla ; it then 
flows north for about thirty miles, and ultimately 
turns east or north-east to Khoten, but the last part 
of its course has not been explored. The altitude of 
the Karakash above the sea level, where we joined it, 
is about 15,600 feet. 
On the 30th July we made a short march of only 
three miles. A complete halt of even one day has a most 
demoralizing effect on one's camp followers, as they 
manage to get everything out of its place, so that on 
starting again after a halt it requires more than 
double the usual time to strike the tents and pack up 
the baggage. At the new camp there was plenty of 
grass and fuel. The Karakash valley here varies 
from one to two miles in width, and the river flows 
over shingle, in a great number of streams, and winds 
a good deal, owing to the number of rocky spurs 
which project into the valley. Here and there the 
ground is covered with much saline efflorescence, and 
there are numerous springs, some of which are warm, 
^ along the foot of the ranges which rise abruptly and 
sometimes precipitously on either side. All the 
higher peaks appeared to be composed of granite, the 
lower ones of gneiss and slate. 
In the shallow side streams and the pools fed by 
springs there were numbers of small fish, but I never 
observed any fish in the main stream. To-day I got 
from Dr. Cayley my first specimens of the Tibetan 
sand grouse, SyrUaptes tibetanus, which had been 
shot a few days before in the Upper Karakash valley. 
I again noticed, in the marshy ground along the river. 
