a 
a 
i 
. 
26 SEASONAL DEPOSITION IN AQUEO-GLACIAL SEDIMENTS. 
these lakes the beach must have been nearly horizontal, and the basin in the northern part of 
the State must subsequently have been elevated nearly 200 feet more than on the south line.” 
Emerson, 1917, p. 141, 142. 
In connection with the recession of the Connecticut River lobe of the 
Wisconsin ice sheet, a factor which has been too little considered must be 
mentioned. The late Professor Tarr, in a paper (1912) which was published 
posthumously, noted that the retreats of glaciers in which the ice fronts were 
in water retreated much faster than glaciers with ice fronts on the lands. He 
wrote: 
“Under the simplest of circumstances the advance or retreat of a series of glaciers is a 
complex phenomenon in which so many factors are involved that a full analysis of them can 
not be undertaken here. Yet some of the factors stand out with such distinctness that I 
may take time to briefly point them out. The nature of the glacier terminus is of fundamental 
importance. If the end of an ice tongue is in water it makes a great difference in the rate 
both of advance and recession whether the water is salt or fresh, whether it is deep or shallow, 
whether it is in active movement or is quiet, whether there is or is not a free escape for the 
icebergs, and whether the relative area of ice cliff is small or great. All these factors are 
effective in addition to the rate of supply of ice to be discharged. If, on the other hand, 
the terminus is on the land, there are influences of exposure, of elevation, and of amount of 
moraine cover, as well as the amount of ice supplied.***It is clear that there must be a 
very great difference, especially in recession, according to whether the ice front is on the 
land or in the sea, for in the latter position wastage is far more rapid than in the former.” 
Tarr, 1912, p. 254. 
If, during the closing stages of the Wisconsin epoch, the Connecticut 
River lobe had its ice front in water of considerable depth, as seems to have 
been the case, that part of the ice sheet would have retreated faster than the 
ice on the adjacent lands. If such was the case the ice front of the lake would 
have been located in a reéntrant of the glacier, and ice would have extended 
on the land southeast and southwest of it. Streams would have entered the 
Connecticut Lake from the east or west and possibly glacier tongues entered 
the lake from east or west after the lake ice front had retreated farther north. 
The latter may have been the case in the Wells River section. Here, it will be 
remembered that large sections of the clays were greatly contorted showing 
that the ice must have advanced over them from the northwest. Ice and 
glacial streams entering the lake from the side would complicate the deposition 
in the clays, so it is well to bear in mind the possibility of such occurrences. 
The locations of some of the clays and sands which at times appear to be anom- 
alous, might be explained in this way. 
The main fact to be noted is that while clays were being deposited in one 
part of the lake, sands or even gravels were perhaps being laid down in other 
