210 The American Geologist. April, 18% 
snow and to prevent excessive waste by summer melting — a 
mean annual temperature a few degrees only below freezing, 
like that on the Mount St. Elias coast of Alaska — is all that 
is required tor active glaciation. In harmony with this con- 
clusion we find that the extremely frigid areas, such as the 
interior of Alaska and northern Siberia, are relatively dry 
and non-glaciated. For this reason, and also because the in- 
creasing thickness of a sedentary ice-sheet would tend to neu- 
tralize the downward penetration of frost, we need not suppose 
that the ground beneath the ice would be frozen to any great 
depth or far below the detrital layer. 
The relatively high conductivity and ditf'usivity of pure 
ice, in comparison with the covering of neve and snow, would 
tend, through the steady efflux of the terrestrial heat, to raise 
the temperature of the lower portion of the ice-sheet to the 
melting point. On the other hand, the high latent heat of fu- 
sion would tend to prevent extensive melting of the ice until 
the entire basal portion of the sheet had attained approxi- 
mately the melting temperature. The effect of the w r eight of 
the ice itself in lowering the temperature of fusion of its base 
would be too small to require consideration here. 
The interesting question now arises as to the most probable 
plane of shearing when the sedentary ice-sheet finally begins 
to move. We may assume that the initial basal shearing or 
gliding plane will be approximately the same whether the ice- 
sheet begins to move in obedience to its own weight or through 
the overriding thrust of a thicker northern sheet. In any case, 
or whatever the cause of movement, it is, as we have seen, 
most likely to be inaugurated during a period of climatic 
amelioration. Above the bed-rock, three layers of material 
require consideration: (1) The ice-sheet proper: (2) the fro- 
zen soil beneath it, to which it is still firmly united; (3) the 
unfrozen soil resting upon or passing downward into the solid 
rocks. Under the assumed conditions of climatic ameliora- 
tion and a basal rise of temperature in the ice-sheet, we may 
suppose that the frozen soil, in consequence of its lower posi- 
tion and the relatively low specific heat of earthy and stony 
substances, would tend to rise in temperature and to thaw 
earlier than the pure ice. Hence the frozen soil may, perhaps, 
be regarded as weaker and more susceptible of shearing than 
