oe jng, 7 inches 
366 J. D. Dana—Flood of the Connecticut River Valley. 
to make aqueous deposits, which might or might not be free 
from stones or bowlders. 
Dr. Rink, the Greenland explorer, and for many years, Gov- 
ernment Inspector of South Greenland, states, in his “ Danish 
Greenland,” (1877) that the sub-glacial streams which flow out 
ity of the natives, as well as to the observations of explorers. 
He observes that, according to his estimate,* out of the 10 inches 
of annual precipitation—the mean amount—only’ 25 per cent, 
or 24 inches, are needed to supply the loss of ice from the dis- 
charge of icebergs ; and that the rest, apart from what is lost by 
evaporation, makes up the amount of loss from melting, indi- 
cated by the sub-glacial waters. An annual supply to this 
end of only 6 or 7 inches could not produce the largest of 
rivers.t  ; 
From these facts we may hence assume that the great glacier 
of Eastern North America would have had 
1) Crevasses, with only local exceptions ; 
(2) Surface streams flowing with nearly pure waters over 
nearly pure ice, suddenly becoming sub-glacial by descending 
the crevasses; anc 
3) Sub-glacial streams as much more extensive than those 
of Greenland as the precipitation was more copious and the 
England during the progress of the Glacial era; and this, with 
an annual fall of 120 inches, would have been 72 inches or 60 
inches with one of 100 inches. This estimate can hardly be 
over-large considering that tropical and warm-temperate regions 
were not far distant, and that the summers, if it were a fact that 
they were shorter than now, were proportionally hotter. 
When the glacier began its decline, the proportion of the precip- 
itation becoming sub-glacial wo:ld have gradually inereased, 
and so on to the end. .And, consequently, as the amelioration 
of the climate made progress, the melting would have augment 
in rate through the increasing warmth of the winds. The fact 
that glacial conditions tend to perpetuate themselves is well 
%* Based chiefly on the flow of ice in the Jakobshavn fiord, as measured by 
nore to Helland, the mean precipitation at Jakobshavn for the year 
between July 1873 and July 1874, was only 84 inches, and for the year follow- 
jng, 7 inches, 
