126 GLACIAL PHENOMENA IN MAINE. 
we have two successive sets of lines, the later 
ones having partially obliterated the first. The 
height of the ridge may have determined a 
change in the course of the ice, when it had 
diminished in thickness, and no longer acted 
with the same undeviating force. At Water- 
ville the facts are still more perplexing. On 
the road to Benton, near the house of Gf. W. 
Drummond, are slaty rocks striking northeast, 
upon the surface of which are again two sets 
of marks, — one consisting of large, distinct 
scratches and furrows trending due north, 
while the others are finer, less distinct, and 
point east-northeast. On the road to Winslow, 
near the house of Henry Gicliell, the same two 
systems of scratches may be seen on flat slabs 
of rock along the roadside. From the forma¬ 
tion of the land in this region, I am inclined 
to believe the second agent — namely, that to 
which the scratches bearing east should be 
ascribed — to have been icebergs. There is 
high land two or three miles beyond these 
rocky surfaces, in Benton township; and the 
flat over which the Sebasticook River flows 
extends to these heights. The ice is likely to 
have remained longer upon the higher ground, 
