STUDIES OF THE CYCLE OF GLACIATION 383 
remnants in occupation of the cirques for at least a brief period 
while the continental glacier was withdrawing from the region. 
These would presumably develop in much the same manner as 
those already described on James Ross Island, but with differences 
which will be pointed out in the next section of this paper. Gold- 
thwait is no doubt correct in believing that the mountain glaciers 
had a much longer life during the advancing hemicycle of glaciation 
and that the cirques were shaped at that time. It is even doubt- 
ful if any appreciable work of erosion or deposition was accom- 
plished in the later period of mountain glaciation, and _ this 
interpretation would be in harmony with Goldthwait’s observations. 
Ill. THE GLACIAL CYCLE ON THE MARGINS OF THE 
CONTINENTAL GLACIER OF ANTARCTICA 
It is a fundamental and prerequisite condition for the sequence 
of stages through which mountain glaciers pass during a receding 
hemicycle of glaciation that the areas of alimentation and ablation 
should be sharply separated from each other. The former is 
restricted to the upper levels, and alimentation is augmented in 
amount toward the top, whereas the area of wastage is found in 
the lower levels and the losses are increased toward the bottom. 
Such a distribution results principally from two conditions: (1) 
mountain glaciers are nourished by upwardly directed air currents 
which deposit their moisture as a result of progressive adiabatic 
refrigeration; and (2) they are wasted by contact with warm-air 
layers whose temperature rises progressively toward the bottom. 
It is a direct consequence of the combination of these conditions that 
mountain glaciers during a receding hemicycle of glaciation become 
reduced in area through withdrawal of the glacier foot up the valley, 
and even in its expiring stage the glacier head occupies essentially 
ithe same position thai it did at the beginning (Fig. 14). 
Were these two conditions affecting the size of mountain gla- 
ciers not realized, the results would be quite different. When we 
examine the glaciers on the margins of the inland ice of the Antarctic, 
we find they differ widely from those of moderate latitudes, which 
are the ones that are well known and have formed the basis of our 
classification. Within the Antarctic air temperatures do not rise 
