GLACIAL DRIFT. 225 
upper Pemigewasset valley are grouped together in the table (p. 185). 
The valley runs southerly, and the striz mostly conform to it; at the 
very entrance, high up, we have the course S. 8° W., above the Eagle 
lakes on Mt. Lafayette. On the summit of Bald mountain, in the Fran- 
conia Notch, the course is S. 2° E., with a large boulder of Bethlehem 
gneiss on its summit. The same force went over Mt. Profile, leaving 
many boulders of the same kind. It appears from these circumstances 
that a large body of ice entered the valley from above, which must have 
been a portion of the larger and older sheet. This and the later local 
masses scored the vertical east side of Mt. Pemigewasset, with the course 
S. 12° E., working southerly. Other examples occur through Lincoln 
and Woodstock, following the valley. Were the forests cleared upon 
both sides of the stream, many other interesting examples would be 
found. At North Woodstock, the hills upon the west side display a 
magnificent embossment on a large scale, as worthy of photographic 
reproduction as those in Rumney. There seems to be more easting in 
lower Woodstock and Thornton than is needful to conform to the valley. 
This may be explained partly by the fact of local turns in the course, 
not obvious, except on the ground. On the rim of the valley to the 
south-east is the course S. 70° E., which may have influenced some- 
what the direction of the ice lower down, yet is sufficiently different to 
satisfy us that the marks in the valley were not produced by the same 
movement. Through Campton the scratches essentially follow the val- 
ley. A single case is reported from Mad river, high up, of a descent 
N. 32° W., or down one of its tributaries. This will be an example of 
the local movement, or possibly the result of a recent slide, described 
in Volume I. Below Campton, the Pemigewasset glacier joined the ice 
movement down Baker’s river, and went with that over to Lake Winni- 
piseogee. 
Androscoggin Movement. Should a glacier start from the Rangely 
lakes and follow down the whole Androscoggin river, the course would 
be to the south-west, then south, and then a turn to the east as far as 
Bethel, in Maine. We have evidence of a strong south-east current over 
Rangely lake, and in Upton on the east side of Umbagog lake,—a move- 
ment not of a local character. Down the Magalloway river, close by the 
College Grant, there was a well defined valley movement, S. 4° E,, 
