406 The American Geologist. June, 1894 
drift? There are certainly reasons to suspect that this may 
be so; but there are few facts to indicate whether the origi- 
nal material may have been ground or terminal moraine. 
0. — Summary. 
Briefly stated in summary, the argument of the present pa- 
per is as follows : — The theory for the origin of drumlins which 
is most currently accepted has certain objections which must 
at least be considered and explained. These are, first, the 
fact that there has been no adequate explanation of the man- 
ner in which the irregular currents act in building these drift 
hills; second, that there is no apparent association with to- 
pography or supply; third, that the drumlins are so irregu- 
larly distributed ; fourth, that there are so few signs of in- 
ternal lamination parallel to the drumlin outline; and, fifth, 
the presence of stratified portions in many drumlins. 
A second theory, that of origin by a partial working over 
of morainic or other drift deposits, once proposed, but now 
more or less generally abandoned, has certain facts in support 
of it which appear to be overlooked. Attention is called to 
the fact that drumlins are locally and peculiarly distributed. 
They are in places parallel to moraines, and their irregularity 
is often similar to morainic irregularity. The drumlinoid 
outline of roches moutonne'es and of rock hills is also pointed 
out, as well as the presence of partly lenticular hills in moraines. 
These appear to be expressions of an effort on the part of the 
ice to erode a slope of least resistance which is strikingly like 
the curve of the drumlin. They are compared to the action 
of a meandering river which is carving ox-bow curves. 
Objections to the theory are of three kinds: first, the as- 
sumed absence of stratified drift ; second, the apparent ab- 
sence of boulders; and third, the peculiarities of distribu- 
tion. A number of instances are cited from Boston bay, New- 
Hampshire, central Massachusetts, and central New York, to 
show that the phenomena of stratified drift and abundant 
boulders are present in many drumlins. It is also shown that 
deep sections in these hills are rare, and that a moraine over- 
ridden by the ice would probably be coated with till, which, 
without a deep section, might easily be assumed to be typical 
of the entire drumlin, while, in reality, the core might be 
stratified moraine. Since many portions of moraines are 
