Enfflacial Drift.— Crosby. 229 
referred to subglaeial streams. Superglacial and marginal 
streams have here and there notched the summits or terraced 
the sides of protruding drumlins: and frost and rain-wash 
doubtless accomplished some degradation of the drumlins 
after the disappearance of the ice and before the mat of veg- 
etation was spread over them. But when all these modes of 
erosion are taken into account, the waste which drumlins 
have suffered still appears so trifling that, if they were re- * 
garded as representative in this respect of the ground moraine 
in general, all subglaeial material might, perhaps, be safely 
neglected as a source of modified drift. In fact, the advocates 
of the derivation of the modified drift from the ground mo- 
raine should, it would seem, also espouse the theory that 
drumlins are mere erosion-remnants of a sheet of till of much 
greater average thickness than that which now encumbers the 
glaciated area. 
It may be noted, however, that, as I have elsewhere pointed 
out.* the ice-sheet was probably accompanied, at least in its 
later stages, by a more or less complete system of subglaeial 
drainage; and during all the time while the ground moraine 
or subglaeial till was accumulating through basal melting, 
and also while it was still englacial, through the agency of 
numberless shearing planes, it was undergoing modification 
by the washing out of its finer constituents, clay and rock- 
flour. Obviously, of this differential erosion no distinct trace 
or sear could be expected to survive the disappearance of the 
ice-sheet, especially since the action could not have been 
sharply localized, but must have affected in some degree 
almost the entire area of the ground moraine. 
While not denying or doubting that the ground moraine 
has made, in various ways, substantial contributions to the 
modified drift, I recognize that the relations of the still 
englacial drift to glacial drainage afford a more direct and 
satisfactory explanation of the main part of the modified 
drift, 
The real problem appears, then, to be as to the relative 
efficiency of subglaeial and superglacial streams. Upham, 
the foremost advocate of the efficiency and sufficiency of 
*Proc. Boston Society of Natural History, xxv, 117. 
