72 Over the Riin-tst-la ; 
there any trace of vegetation, which seemed to end several 
hundred feet below the summit; consequently I think the 
Riin-tsi-la can hardly be less than 18,000 feet. But though 
I was very much out of breath and had to stop every few 
yards, I felt none of the effects of mountain sickness on 
this, the highest pass I crossed. 
The last vegetation consisted of patches of Potentzlla, 
dwarf Meconopsis (MZ. rudis), dwarf Primula, Scirpus, a 
few grasses, a gentian actually in flower (G. heptaphylla), 
a Senecio at the very head of the trickling stream, and 
several scree Compositae and Umbelliferae. 
Under the combined action of sun and wind the snow 
had assumed some picturesque forms, and in the shade of 
the rocks small ice-pillars stood up from the ground, evi- 
dently formed in the following manner: the snow on the 
boulders melted in the sunshine, trickled slowly over the 
rock and froze again as soon as it got into the shade, an 
icicle being thus formed in the usual way. Eventually it 
reached the snow, and then the end in contact with the 
rock thawed as the sun got round so that the pendent icicle 
became an erect ice-pillar. At the summit, the snow on 
the south-facing rocks stood out in horizontal fern-like 
plates and crystals of great beauty, often several inches 
long. The manner of their formation I could not make 
out, but it was probably connected with the almost simul- 
taneous melting of the snow in the warm sunshine, and its 
freezing and recrystallising in the bitter wind. 
The descent through deep soft snow was bad, and spills 
were frequent. At last we got down to a small lake in the 
stony valley and halted for lunch beside a party of Tibetans 
on their way up. The men now informed me that there 
was a choice of routes, either down the valley in front of 
us to the village of N goug-chi and so up the Yang-tze, a 
day’s march to Mo-ting; or over a second pass and down 
the next valley. However I was chiefly concerned with 
the mountain flora, so I decided to go straight on. 
This second pass, the Chnu-ma-la, seemed scarcely as 
high as the Riin-tsi-la, though the less amount of snow on 
the south face was easily accounted for by its much more 
exposed position ; while on the north slope the snow was 
almost as deep as on the Riin-tsi-la. From the summit we 
