334 SURFACE GEOLOGY. 
down of all the materials in the ice-sheet at the time of melting. The 
upper deposit would be loosely aggregated and more highly oxidized 
than the lower beds, because the opportunities for the addition of oxygen 
were more favorable. 
The very first question suggested by the position of the analyses is, 
whether the gray clay has not been derived from the upper and the blue 
from the lower till. The two classes of deposits agree in their chemical 
character, color, and superposition. I will advocate the affirmative an- 
swer to this question, and state how the conditions would favor such a 
derivation. 
The blue clay is not continuous throughout the Connecticut and Mer- 
rimack valleys, but occurs in definite and limited areas, as has been set 
forth in Chapter I. Its origin may be conceived to be like that of similar 
clays near the mouths of glaciers. The water gathering from the various 
sub-glacial rills gathers the fine clayey material pulverized by the motion 
of the ice, and transports it to some lake or place of comparative quiet, 
and the sediment is deposited as a blue clay. In Hooksett and Pem- 
broke is a mass of clay four miles or more long. Suppose that the gla- 
cier were still unmelted above Franklin, the pulverization of the rocks 
on the upper Pemigewasset would furnish an abundant supply of clay, 
which might have been transported to Hooksett before the current 
slacked sufficiently to allow the turbid water to settle. In this way the 
blue clay might have been formed, and it would have been derived alto- 
gether from the lower till, Subsequently, we may suppose, the ice be- 
gan to melt more rapidly, and the detrital residue, or upper till, would 
be subjected to carriage. The first result would be the soiling of the 
streams with the more oxidized clay, and there would be formed the gray 
deposit over the blue. In the continuation of the process, the streams 
would become still larger, and the material brought down would be sand 
covering the gray clay; or the gray clay may be conceived to be the re- 
sult of the washing of the upper terrace deposits. The finest particles 
in the case of a washing would be the first to show themselves low down 
the valley; and the material exposed is essentially the upper till. 
This view explains how the formation of clay may be connected di- 
rectly with the ice. The ice nearest the edge of the sheet, or on the 
ocean border, first melted, producing the upper till and a massive clay. 
