J. D. Dana on Glacier movements along valleys. 239 
along the Connecticut river valley, but also of one along the 
upper Champlain valley, the Lamoille valley, the Winooski 
valley, the Otter creek valley, and probably the Merrimack 
valley ; and many of the courses of scratches observed in other 
parts of Vermont and New Hampshire have divergences from 
the normal course of the great glacier, which are probably due 
to the valley-depressions of the surface. Among these smaller 
valleys are perhaps those of the Queechee, Black, Middlebury 
and White rivers of Vermont and the Deerfield of Massachu- 
setts; for the existence of an independent glacier in each of 
these valleys is recognized as probable by Prof. Hitchcock, on 
the ground of the conformity between the direction of the 
scratches and the valley, although the iceberg theory is ado ted 
by him for all the rest, even the Lamoille, Winooski and Con- 
hecticut. The writer has elsewhere mentioned the evidence in 
favor of a Hudson river glacier movement, and of another in 
¢ Mohawk valley running easterly through central New York; 
and further, of one along the St. Lawrence valley, the scratches 
a following its course according to the observations of Dr. 
awson. 
The facts are sufficient to prove that examples of valley 
movements of glacier ice must have been common over the 
continent in the Glacial era, or rather the rule for all the larger 
valleys. Itis hence evident that no observations on glaciers 
ection, in his view, was that of icebergs. Besides the argu- 
ment against the iceberg hypothesis elsewhere presented, New 
“ngland affords another in the fact that if there were, at the 
time, a submergence to the depth required to overcome the 
Obstacles to a southeast movement offered by the poatnerly 
to have worked their way along them to do the scratching, 
atfords another strong argument against it. 
Continental clacier still had its continental, or at least its New 
tng rpscariso a, In Massachusetts and Connecticut the 
