46 The American Geologist. July, 1895 
eral localities. East of the river exposures are rare, but the 
few found give a direction east of south for the latest striae. 
Conditions of Formation. 
The changes from fine to coarse material in a direction 
lengthwise of the range of hills are too abrupt, too complete, 
and too frequent for production by a single continuous river. 
Objections to such agency are also found in the irregularity 
of the topography ; the inclination of the strata ; the direc- 
tion of current lamination; the distribution of the materials 
according to size; and the extreme diiTerence between the 
north and the south slopes. P^very one of these features, 
along with the topography, the distribution of till, and the 
pushing of the northern sides of the hills, is entirely ex- 
plained by supposing the beds to have been accumulated at 
the front of the ice sheet by the drainage from the dissolving 
glacier. One other condition is necessary to account for the 
peculiar structure, and that is a body of deep water into 
which the materials were thrown. It was the recognition of 
this condition that gave the writer the key to the problem. 
At the time when the Ontario ice-lobe deployed over the 
Rochester plain, the eastern or Mohawk outlet of the glacial 
waters must have been still closed by the Adirondack ice- 
sheet. The waters of lake Warren laved the front of the 
Rochester glacier to a depth between 350 and 400 feet. The 
evidences of this deep water through western-central New 
York are abundant and conclusive. The discussion of this 
subject will form another paper. To the writer it seems clear 
that the "Pinnacle hills" are a true kame series, forming part 
of the frontal moraine of the water-laved glacier. 
Over the flat area south of the kame hills thick silts and 
brick clays lie above the till, which are the finest deposit car- 
ried out into the lake by the slackened currents of the glacial 
rivers. North of the range the silts lie thick in the depres- 
sions of the plain, as the waters of lake Warren doubtless 
buried the region for some time after the retreating ice-front 
had abandoned the locality. 
An objection to the morainic origin of the hills might be 
offered, that the accumulation of such a mass of assorted ma- 
terials along three miles of the ice-front seems inconsistent 
with the diverging flow of the glacial streams, and more ex- 
