MODIFIED DRIFT ALONG CONTOOCOOK RIVER. 117 
gravel and sand, which remain nearly at their original height upon the 
hillside. 
The kames which we find south-west from Bennington, and a large 
portion of those north of Whittemore pond, are principally composed cf 
sand and fine gravel. They were probably deposited at the mouth of 
the glacial streams where these entered the lake, nearly all the modified 
drift which was brought from the melting ice being thus accumulated in 
mounds, ridges, and terrace-like banks. The want of continuity in these 
deposits appears to be due to the irregular rate of melting and to the 
varying slopes assumed by the terminal front of the ice-sheet, the latter 
being determined by this rate and by the contour of the valley. 
The lack of stratified drift in the valley west from the Greenfield plain 
seems to show that the ice over this area, while it still confined the little 
lake on the east, had been melted nearly to this level, sending its alluvium 
to form this plain; and that the remainder disappeared from the valley 
without sufficient currents to form alluvial deposits. All the material 
which it still held was dropped as unstratified till, unless we except rare 
instances of kames like the isolated banks of sand seen on the hillside 
east of the river near the north line of Peterborough. 
The first deposit belonging to this period that we meet in going up the 
valley is the high level-topped sand north-west from North Peterborough. 
This and other terrace-like deposits extending to Peterborough appear to 
be of similar origin with those already noticed south of Bennington sta- 
tion. Kames of the common type, composed of coarse gravel and sand, 
occur one and two miles farther up the valley; and they are increased in 
amount as we approach the line between Peterborough and Jaffrey, ap- 
pearing to have come principally from Sharon on the south-east. We 
may suppose that these, as in Bennington, were deposited at the same 
date with the sand which partly filled the opening channel below. This 
was a branch of the lake, and the sand fell in irregular and thin deposits 
with stratification conforming to the sloping sides of the valley. The 
occasional boulders which we find embedded in the alluvial deposits of 
this lake appear to have been dropped by floating masses of ice broken 
from the glacier which bordered its shores. 
Going down the valley we find evidence that the glacial melting ad- 
vanced beyond Hillsborough Bridge, while its outlet continued to be 
