396 The American Geologist. Jane, 1894 
us well as in friable sedimentary strata; and they occur on 
plains as well as in the hilly and even mountainous region of 
New England. One has no difficulty in understanding how- 
drift is accumulated in transverse valleys where the ice was 
nearly stagnant, nor the manner in which a drift slope may 
be accumulated on the southern end of a rock hill. There is 
here an apparent cause for deposition; but it seems to me a 
much more difficult task to conceive of irregular currents or 
supply, for which there is no apparent cause, but which, in 
some manner, shall accumulate narrow and short hills of drift 
and often a series of complex summits and valleys. 
Another objection to this theory is that there are so few in- 
dications of lenticular stratification. If these hills are built 
up by successive additions it would seem that there should 
be, at least commonly, some indication of this growth by a 
kind of stratification or lamination parallel to the drumlin 
outline. While in some sections this is apparently present, 
in general it is absent. Sections of drumlins usually exhibit 
either massive till or a mixture of till and stratified drift, 
morainic in aspect. 
The presence of stratified drift, which in the following pages 
is shown to be by no means uncommon, must be explained 
by any hypothesis which is put forward to account for 
drumlins. How is it possible by this theory to explain the 
alternation of loose stratified gravels and compact till? The 
compactness indicates a weight of superincumbent ice. Would 
a sheet of ice sufficiently massive to cause the compression 
thus indicated, change its nature so that flowing water should 
deposit modified drift? And how does it happen that this 
peculiarity exists in the drumlins and is so generally absent 
in the general drift sheet near by ? 
The peculiarity of distribution should also be explained. 
They are local deposits, sometimes isolated, but more com- 
monly in clusters, and over large areas of glaciated country 
they are entirely absent. These facts are certainly remarka- 
ble ; and if drumlins are the result of variation in ice currents, 
or drift supply, it evidently requires peculiar conditions for 
the existence of these irregularities. These conditions should 
be apparent, or at least indicated; but, so far as I know, no 
