THE GLACIERS OF MUS-TAGH-ATA 335 
manifestly been caused by the relative warmth of the 
earth. Four small glacial streams and several little rills 
dropped from the edge of the ice in pretty cascades. The 
largest had a fall of sixty feet, and had eaten into the 
edge of the ice, so that it did not come all the way from 
the top. Another had cut nearly twenty feet into the 
ice, and as the brash had got heaped up on top of it, the 
streamlet resembled a spring issuing from a hole in the 
level wall. It was a beautiful sight, to stand underneath 
the big waterfall, and watch it shoot out into the air as 
if from the gutter of a house, split into a thousand drops 
that glittered like pearls in the sun. The ice was every- 
where as soft as a sponge, so that we could actually 
make “snowballs” of it. Water was dripping, trickling in 
every direction ; no matter which way we turned we heard 
the sound of bubbling and running water. The ice was 
courting destruction in venturing down to tracts whose 
climate it could not endure. Beneath the face of the 
glacier were large detached blocks and heaps of extremely 
rotten ice, which had broken off and were melting rapidly. 
When the grotto, just mentioned, has become sufficiently 
hollowed out, and the superincumbent mass of ice too 
heavy, the latter will crash down, and contribute to the 
hastier decay of the glacier. 
Keeping immediately alongside the ice, we then rode 
round the face of the glacier, and continued up its right 
side. At one point at the foot of the glacier, where a 
cascade splashed noisily into a pool of its own making, 
the Kamper-kishlak approached so close to the lateral 
moraine of its neighbour, the Sarimek glacier, that we 
could scarcely get through the narrow passage. 
The surface of the glacier sloped at an angle of twenty- 
five and a half degrees, and was thus extraordinarily steep 
as compared with the glaciers of the Alps, which often 
have an inclination of less than one degree in their lower 
regions. The two contiguous glaciers of Sarimek and 
Kamper-kishlak approached each other at right angles, 
and between them, near the rocky spine which divided 
their common n^vd-h-A&xw, a brook issued and helped to 
