220 
The American Geologist. 
April, 1896 
denness of this change from active erosion to deposition in 
any area is indicated by the fact that, generally speaking, the 
ground moraine rests everywhere upon normally striated sur¬ 
faces. In other words, as soon as the ground moraine began 
to appear through basal melting it was essentially motion¬ 
less ; for ice moving over a bed of detritus in a way to impart 
motion to it would inevitably give rise to sidewise, oblique, 
and devious movements of individual stones which would tend 
to obscure and efface the rectilinear striation of the bed-rock 
surface. The ground moraine as it accumulated was pressed 
down by the ice to form the typical hardpan. In part it 
accumulated on stoss slopes through obstruction to, and in 
part on lee slopes through protection from, the forward move¬ 
ment of the ice, forming drumloid slopes and, later, drumlins. 
It is not a necessary deduction from this view that the bed¬ 
rock always rises to a good hight in drumlins, since the hard- 
pan character of the ground moraine and the tendency of the 
ice at this stage to flow over loose materials, as observed by 
Niles, Spencer, Chamberlin, and many others, makes of the 
initial knob or boss of the ground moraine an efficient gath¬ 
ering point for more material as fast as it is furnished by the 
melting of the ice. The main point may be presented in an¬ 
other way. When, as in the earlier and maximum stages of 
the ice-sheet, the basal temperature was below freezing, the 
freezing of subglacial waters made and kept the detritus a part 
of the ice-sheet; and when, as in the later stages of the ice- 
sheet, the basal temperature rose above freezing, the ice re¬ 
laxed its hold on the detritus and flowed over it, as attested 
by observations on modern glaciers. 
It is a necessary consequence or corollary of this view that 
transportation of drift by simple drag is relatively unimport¬ 
ant , 1 if not impossible. The transportation is almost wholly 
englacial, as insisted by Upham,* but highly differential, 
being extremely slow in the basal layers and more and more 
rapid at higher levels. The detritus reaching the highest 
levels in the ice is carried farthest, not only because of the 
higher velocity, but also because it remains for a longer time 
in the ice. 
*Bull. Geol. Soc. Amer., vi, 348. 
