The Origin of Drumlins. — Tarr. >399 
What the reason for this rhythmic action is, I am not pre- 
pared to say; but that ice erosion normally distributes itself 
unevenly, with the ultimate result of producing rounded, len- 
ticular forms with axes in the direction of ice motion, it seems 
hardly probable that any one can question after seeing the 
rock hills of New York and New England. Nor is this due 
to differences in hardness or attitude of the rocks, for in cen- 
tral New York the strata are of comparative uniformity and 
nearly horizontal. It appears to be an expression of the effort 
of the moving ice to produce a slope of least resistance; and 
this slope appears to be the flat drumlinoid slope. Depres- 
sions are filled with drift, and the projections are worn down : 
and when the ice disappears the hill imi3 r be veneered with 
drift. This curve of ice erosion is beautifully shown on the 
sides of some north and south valleys, one of the most typical 
instances known to the author being in the valley of Cayuga 
lake. I believe it can be stated that it is a law of ice erosion 
to produce rounded, lenticular slopes. 
(c) Drumlinoid Form of Overridden Moraine*. — In the mo- 
raine at Cape Ann, Massachusetts, there are drumlinoid forme, 
not perfect drumlins it is true, but forms simulating these 
with their axes in the direction of ice flow. That these are 
moulded by the ice as it formed the moraine seems evident ; 
and they appear to be illustrative of a stage between mo- 
raines arid drumlins. f While these are not as perfect as true 
drumlins nor as drumlinoid rock hills, being in a moraine the 
resemblance is noteworthy. Others have pointed out the same 
lenticular tendency in morainic hills. 
(d) Flutings on Drumlins. — One other point which favors 
the destructional rather than the constructional theory is tin- 
presence of fluting on drumlins. In Massachusetts, benches 
and grooves, which from their small size appear to he groove- 
in reality, are frequently found on one side of the hills. At 
least it would seem to the author very difficult to explain 
them by variations in current-. 
4. — Consideration of Objections to Destructional 
Hypothesis. 
There are certain objections which niust first be explained 
before this theory can be generally accepted. At present 
fNinth Annual Report. V . s. Gteol. Survey, p. 551. 
