« 
CROSBY : ORIGIN OF ESKERS. 
385 
value; after, for instance, the trough of the Boston Basin had 
deflected the basal portion of the ice decidedly to the eastward. 
This was the period of drumlin formation, which was probably 
followed closely by the esker and sand plain period. 
The absence of basal water during all but the latest stages of 
glaciation is easier to understand when we consider how low, 
probably, were the initial temperatures throughout the entire thick¬ 
ness of the ice sheet, and the great depth of frost penetration into 
the ground beneath the ice. Nowhere, probably, within the range 
of observation at the present time are the conditions of the Pleisto¬ 
cene ice sheet more nearly realized than in the interior of Greenland ; 
and in this connection Nansen’s experience in his memorable transit 
over the ice from the eastern to the western coast is of special 
interest. He says (’90, vol. 2, p. 480), “Some of the temperatures 
which we experienced were far lower than the established meteorolog¬ 
ical laws would have led us to expect. The temperature on certain 
nights, September 12 and 14, probably fell, according to the calcula¬ 
tions of Professor Mohn, to —45° Cent. (—49° Fahr.), while the 
mean temperature of certain days, September 11-16, when we were 
about in the middle of the country, or a little to the west of the high¬ 
est ridge, varied from — 30° Cent, to — 34° Cent. (— 22° to —29° 
Fahr.). This is at least 20° Cent. (36° Fahr.) lower than anyone would 
have been justified in expecting, if he had based his calculations on 
accepted laws, taking for his data elevation above and distance from 
the sea, as well as the mean temperature of the neighboring coasts.” 
Now, supposing similar climatic conditions to have prevailed during 
the entire period of the accumulation of the ice sheet, and consider¬ 
ing that the glaciation of Greenland is long past its maximum and 
the climate, therefore, probably somewhat ameliorated, we have 
indicated a degree of refrigeration in and under the ice which the 
slow upward flow of the terrestrial heat would require a long time 
to overcome. The low temperature of the ice is seen not to be 
necessarily inconsistent with flow through the medium of a granular 
structure, in accordance with the views of Klocke, Deeley, Fletcher, 
and others, as summarized by Upham (’96), when we consider that 
the progressive melting of a granule at one point and its growth at 
another point demand differential or localized pressures, and there¬ 
fore pressures the maxima of which must exceed the average of the 
combined vertical and flowage pressures for the same depth. In 
