a second Journey to the Yang-tze 161 
guide was waiting for me below, and we headed back to 
camp. 
At the very highest point to which I climbed there were 
still flowering plants in the shelter of the rocks, such as 
Meconopsis speciosa, cushion plants, and so on, but we may, 
I think, regard 18,000 feet as the extreme limit of flowering 
plants on this range. Above that the soil is probably too 
cold for the roots to function, since there is no reason why 
a plant should not obtain adequate protection from the wind 
at altitudes well above that limit. 
Early on the following morning we struck camp and set 
out for Tung-chu-ling, the weather being fine and bright. 
After crossing the pass the great buttress of Pei-ma-shan 
was occasionally visible to the south, though partially buried 
in cloud, and immediately above us to the north stretched 
a barren ridge of limestone crowned by picturesque towers 
and pinnacles. No better proof of the difference in rainfall 
on the two divides could be adduced than by comparing the 
pyramidal peaks of K‘a-gur-pu with their beautifully curved 
outlines due to water erosion, and the wall-sided, flat-topped 
buttress of Pei-ma-shan, plainly the work of dry denuding 
agents, particularly those associated with rapid oscillations 
of temperature and great extremes. 
The snow clung to the precipices of Pei-ma-shan in 
patches and streaks, following particularly the lines of 
stratification, which were thus clearly defined; and the 
bottle-nosed glaciers, though several miles distant, were 
plainly seen to be retreating, their snouts being some little 
way from the terminal moraines. 
Our route lay across a sort of plateau country, for after 
crossing the pass, instead of descending into the main valley, 
we kept on across the heads of several other valleys, ascend- 
ing and descending steeply. Masses of a small brilliant 
blue gentian (G. afvzca) covered the limestone rock, which 
here presented a remarkable vesicular structure. Very con- 
spicuous also at 14,000—15,000 feet was the curious glaucous- 
leaved Sfzvaea laevigata, which predominated in the shrub 
flora. It forms loose-growing bushes about six feet high, 
but unfortunately it was not now in flower. 
A terrific and sudden storm of rain coming on at mid- 
day, we sought the shelter of some tents occupied by yak 
W. Ts II 
