Subm:rgeiice in the Narragansett Bay. — Fuller. 3 1 7 
tion. The easternmost lobe of ice covering the state of Maine 
and terminating along a line reaching from cape Cod east- 
ward to Georges shoal was, in its lower portion at least, almost 
entirely independent of the ice to the westward, and possessed 
without doubt a drainage system complete in itself. It could 
have furnished no part of the waters discharged through Nar- 
ragansett bay. 
The area of New England, excluding Maine, is approxi- 
mately 33.500 square miles. Assuming the whole to have 
been covered with ice, a melting of 3.49 cubic feet of ice per 
day for every square foot in this area would have been re- 
quired to furnish the flood demanded. 
It should not be forgotten, however, that the conditions for 
great floods were just as favorable in the valleys of the Housa- 
tonic, Connecticut and Thames, and the area drained by the 
streams entering Narrangansett bay would not, in all prob- 
ability, have been more than a quarter of the area mentioned 
above. Assuming the actual area drained to have been equal 
to that of the state of Massachusetts (8,315 sq. m.), we find 
the ablation required to meet the demands of the flood reach- 
ing the enormous figure (5f 14. i cubic feet per day for each 
square foot of surface.* 
Comparatively few measurements of the surface ablation 
of glaciers have been recorded, but enough is known to show- 
conclusively that at the outside, it cannot be more than a 
few inches per day. Reid, in his report on "Glacier Bay and its 
Glaciers"fstates that during the cloudy and rainy weather of 
the month of July the surface melting varied from 1.6 inches 
to 2.5 inches per day, while on a clear, bright day it reached 
as high as 2.75 inches per day. When it is remembered that 
during the .'Vlaskan summer the thermometer often mounts 
well up towards the 100° mark, it will be readily seen that 
under no combination of circumstances could such a melting 
as that postulated in the preceding paragraph take place. 
*If, instead of assuming the whole of New England to have been 
covered at this stage of the ice retreat, we should regard the margin 
as occupying the position assigned by Upham (Amcr. Gcol., vol. XVI, 
plate V.) to tlie Toronto boundary, the area supplying water by abla- 
t'on vouid he ninch decreased. .-Mid the amount of melting required to 
meet the demands of the hypothesis would be even more enormous 
t'lan that given above. 
ti6th Ann. Rept. U. S. Geol. Surv. ])art i. p. 450. 
