The Origin of Drumlins. — Tarr. 401 
fied ; and, even if this were necessaiy, we have a right to ask 
that proof be furnished that they are not commonly stratified 
in the core. 
Hi. Drumlins containing stratified drift. — We have very 
few descriptions of drumlin sections, but those which we have 
are mainly of drumlins containing stratified drift. This is 
natural since the current assumption is that these hills are 
composed of unmodified till; and variations from this are 
naturally noted, whereas verifications of the assumption are 
often not considered worth recording. It is, therefore, not fair 
to place too much stress upon the preponderance of this class 
of references. The author, in a rather extensive study of the 
drumlin areas of Massachusetts and Connecticut, has rarely 
seen deep artificial sections in drumlins; but such sections as 
have been observed are mainly in compact till. In the nu- 
merous sections naturally exposed in the sea cliffs of the Boston 
Bay region, till is, in many places, found even in the core. 
Nevertheless, where deep cuts have been made in drumlins, 
stratification quite unlike that found in true till has heen fre- 
quently observed. Upham* describes the two drumlins at 
Scituate, Mass.. which, where they are breached by t he sea, are 
called Third and Fourth cliffs. These are undoubted drumlins, 
yet they have a core of stratified gravel intimately mixed with 
unmodified drift arranged in more or less parallel layer-. 
There are no signs of shoving action, and Upham considers 
that these deposits furnish evidence of rapid subglacial (his 
belief is now that they are englacial) accumulations. But it 
appears difficult to account for the association of the very 
compact till with stratified <lri It. since the one must have been 
deposited under considerable ice pressure and the other under 
the opposite conditions and from running water. Thai these 
alternations of conditions should have taken place, and par- 
ticularly that they should have occurred in the narrow drum- 
lin areas is certainly remarkable. It would seem that there 
is a much better accordance of the facts with the theory thai 
these alternations an- morainic and that the hills are tnodifi- 
cal ions of a morainal ma<s. 
The same author says! that some of the New Hampshire 
*Proc. Boston Sue. Nat. Hist., 1889, vol. x\i\. pp. 228-242. 
fProc. Boston Soc. Nat. Hist., 1889, vol. xxiv, i>. 238. 
