152 The American Geologist. September, 1897 
erent axes, uiid formed by a slightly diti'erent rate of supply. 
Another su])posed cause for irreguhirity of morainic topogra- 
phy was also seen illustrated. In the valley tongue of the 
south Cornell glacier, on the soutiiern side of mount Hope, the 
terminus is in a marginal lake whose waters are milk3' with 
sediment furnished from the glacier. Like the balance of 
the glacier, this tongue is retreating. So recently has it 
withdrawn to a notable degree, that at a distance of twenty 
or thirty yards from the margin there are two large blocks of 
ice stranded in the lake, each piece having an area of perhaps 
half an acre. These are grounded in the water, and are part- 
1}'^ or entirely cut off from the glacier. In any event they are 
separated from it by the lake water. Hence they are being- 
surrounded by sediment, and as they melt still further, their 
tops will come beneath the lake, and they will be buried be- 
neath the sediment. Even now they are evidently weighed 
down by the load of clay, for they do not rise more than a few 
feet above the lake surface, and not so high a? they would 
need to rise if they were afloat. When finally the lake dis- 
appears, and the ice withdraws, we shall have a pair of typical 
kettle holes. 
Variation in Lower Ice Currents. Along the glacier mar- 
gin, we may often seethe "effect of the irregularities of the 
bottom upon the movement of the ice. Ledges and small hills 
are often present beneath the glacier, and their influence- is 
to deflect the lower layers by their tiny irregularities. The 
lowest ice rests on the surface, occupying its depressions; but 
the layers above are curved only in accord with the general 
surface. In other words, the irregular surface is made more 
regular by the nearly stagnant bottom ice, which fills the de- 
pressions, and raises the lee side of the elevation toward the 
top of the higher part. This raises the irregular surface into 
a form having approximately the outline of the general up- 
per levels. The ice arches over the irregular land, making 
curves, which are plainly seen by the arch of the debris lay- 
ers, whose form is that of a generalized dome of a regular 
form. (Fig. 2, Plate XII.) Tliis curve frequentl}^ simulates 
the drumlinoid curve. Layers once disturbed and raised do 
not continue at this level in the ice, but return to approxi- 
mately the same elevation as that where they started. Here 
