JOO SURFACE GEOLOGY. 
found in Massachusetts is shown near the head of Nashua river, along 
the railroad between Fitchburg and South Ashburnham. These kames 
lie in north-west to south-east ridges parallel with the valley. When 
they were being formed we must suppose that the ice had gone from the 
lower east and north-east portions of the river’s course, and that the 
floods of water supplied from the melting ice-sheet at its source were 
then completing the deposition of these extensive plains at its mouth. 
At the same time floods were here poured into the Merrimack from 
the north-west, where no stream now exists. A continuous belt of allu- 
vium, upon which the Wilton Railroad is built, extends six miles from 
the Souhegan river in Amherst to the plains of the Nashua river. Its 
narrowest place, three miles from the city, is a third of a mile wide, 
while its widest portions, in the north-west corner of Nashua and south 
part of Amherst, are a mile and a half wide. These plains show a grad- 
ual descent from north-west to south-east, amounting to 75 feet in the six 
miles. They consist of levelly stratified sand and gravel, and in general 
have a very regular surface; but several ponds, often with no outlets, 
fill depressions upon their widest portions, as Stearns pond in Amherst, 
Pennichuck pond near South Merrimack, and Round pond in Nashua. 
Deposition probably took place very rapidly from floods which brought 
down the material from the melting ice-sheet. In some cases masses of 
ice may have remained where we now find these ponds, or they may be 
due to an unequal supply of material and varying currents. The waters 
of the Souhegan valley at this period found their way to the Merrimack 
by three routes. One was along the present course of this river, which, 
below its extensive plains in Amherst, is narrowly enclosed at two points 
by high land of till or ledges; a second, similar to the first, was along 
Pennichuck brook; while the third, which differs from the others in its 
ample. width and, direct course, brought the greater part of these floods 
to the same mouth with the Nashua river. In this way the flood-plains 
of the last route appear to have become slightly higher than along the 
present Souhegan river and Pennichuck brook, which therefore became 
the channels of drainage after the Champlain period. 
A half mile below the mouth of Salmon brook, hills approach nearly to 
the river, beyond which is a plain of similar height with that south of 
Nashua river. In the remaining three miles to Tyngsborough the allu- 
