MODIFIED DRIFT ALONG CONTOOCOOK RIVER. II5 
the kames at the east side of Henniker village were formed before or 
after this change, cannot perhaps be determined. Their position, trans- 
verse to the Contoocook, shows that they were formed by streams from 
the melting glacier on the north in the valleys of Amy and Warner 
brooks, while the rapid retreat of the ice to the west and south-west ap- 
pears to have been delayed by the high hills which closely border the 
river. It is not improbable that, when these waters first flowed towards 
the north-east down the Contoocook valley, a barrier of till near West 
Hopkinton, afterwards eroded by the river, held back a shallow lake 
which extended to the kames last mentioned. The deposition of the 
low alluvium of this area was going slowly forward during all the time 
occupied by this history. 
The melting of the vast ice-sheet over New England proceeded from 
the coast to the north-west and north, so that lakes were temporarily 
formed in valleys which drain northward. The avenues by which the 
waters escaped from the upper portion of the Contoocook basin, or 
that part above Long fall in the west part of Henniker, appear to have 
been three in number, as follows: Southward, over the water-shed at 
the head of the valley in Rindge; towards the south-east, through Green- 
field; and northward, along the course of the river. The length of this 
area is nearly thirty miles; and the outlet in Greenfield is about equally 
distant from its south and north ends. 
The conspicuous kames, which extend five miles along the Vermont 
& Massachusetts Railroad between South Ashburnham Junction and 
Westminster, show that a large area of the ice-fields on the north-west 
poured their waters along this course. These kames are less than 200 
feet below the plateau in Rindge, twelve miles distant, which forms the 
water-shed at the head of the Contoocook valley. Although the present 
drainage of the south part of Rindge and of Winchendon is into Miller’s 
river and the Connecticut, there is no considerable depression; and the 
separation between this basin and the head of the Nashua valley, in 
which these kames are found, is not so high as the water-shed in Rindge. 
This area has not been explored; but the deposits of modified drift in 
Rindge make it probable that the melting of the ice-sheet, while its out- 
let continued in this direction, proceeded beyond this divide, including a 
portion of the Contoocook basin. 
