466 J. D. Dana—“Kames” of the Connecticut River Valley. 
Conclusion.—The conclusion from the investigation is, as al- 
ready indicated, the following: that the suppose “Ikames” are 
portions of the terrace-for mation, with which they usually cor- 
respond re vey in height; and that their materials 
were the same in source with the rest of the stratified drift, 
and the beds the same in time of origin. 
The gravelly character of the terrace-formation off the 
mouths of the tributaries of the Connecticut is often mentioned 
by Mr. Upham; and, if the above conclusion is right, the coarse 
material of the “ kames” is to be explained on the same prin- 
ciple. The position of these coarsest deposits, near the borders 
of the flooded Connecticut, whether they make the lower or the 
upper beds, is a consequence of the rapid flow of the waters in 
this great stream, which drifted away much of the finer material 
within reach and left stones. The coarsest stone beds at the 
mouth of White River are located where the two streams —both 
great streams then—join, that is, where the great contributor 
of gravel and stones encountered the great distributor. 
The deposit of gravel and stones in the upper portion of a 
terrace I have attributed to the violence of the flood when at 
its maximum stage. ut in the region of the so-called 
“kames,” from Windsor to Thetford and beyond, floating ice 
was prone needed for much of the transportation ; and ice- 
have been abundant at the time, when the elacier- 
ice was in rapid process of dissolution about the slopes of the 
Green Mountains—the range at the head of the principal trib- 
utaries in this part of the Connecticut valley. At the same 
time, the Connecticut, by its rapid flow along its eastern side at 
one time and its western at another, would have determined an 
accumulation of stony material along its borders, as a great 
river now produces accumulations on its banks different from 
those more distant. Here the floating ice with its burden of 
earth and stones would have been stranided as well as other 
transported materials. Moreover such deposits might have 
been raised ten feet or more above the plain adjoining, as now 
happens on large streams from modern floods. But there is no 
occasion to account for a cobble-stone deposit along the whole 
top of any of the so-called “ kames;” for, only a small fraction 
of each has a crest of this kind; or ‘any difference in structure 
from the ordinary terrace- formation, except that in some cases, 
near tributaries, they have more of coarse gravel tye 
In Haverhill the angular stones and gravel, brought down the 
Ammonoosue on ice- -floes, made in one place a thick till-like 
deposit lying unconformably over the stratified drift and con- 
tinued some distance down the riverward slope of the terrace. 
This is an exceptional case, due probably to the fact that the 
White Mountains, the source of the stream, are near 
