tn the Champlain and Terrace Periods. 461 
it seemed impossible to account thus for the kames in the Con- _ 
necticut and Merrimack Valleys, which, being bordered on 
both sides by high hills, would have been long estuaries open 
to the sea only at their mouths, and therefore not affected by 
oceanic currents. - From the position of these peculiar accumu- 
lations of gravel, which are overlain by the horizontally strati- 
fied drift, the date of their formation is known to be between 
the period when the ice-sheet moved over the land, and that 
closely following, in which this more recent modified drift was 
deposited in the open valley from the floods that were supplied 
by the melting ice. We are thus led to an explanation of the 
kames, which seems to be supported by all the facts observed 
in New Hampshire, and which appears to apply, also, to the 
similar deposits which have been described in other parts of the 
United States and in Europe. 
At the beginning of the Champlain period, or final melting 
of the great ice-sheet, its nearly level surface of pure ice lay 
above our highest mountains. That it overtopped Mt. Wash- 
ington in New Hampshire, has been recently discovered by 
Prof. C. H. Hitchcock, the State geologist, who has found trans- 
ported rocks, and shown that glacial drift, or .till, underlies the 
angular blocks at the summit. The melting of the ice-sheet 
appears to have taken place mostly upon the surface, which 
was moulded into basins and valleys; and near the terminal 
front of the ice, these came gradually to coincide with the con- 
tour of the land. Here the surface of the ice became covered 
with the abraded material which bad been contained in its 
mass, and which was now exposed to the washing of its innu- - 
merable streams. Its finer portions would be commonly carried 
away; and the strong current of the rivers which would be 
formed near the terminal front of the ice-sheet could transport 
coarse gravel or even boulders of considerable size. In the 
