THE OSAR GRA VELS OF THE COAST OF MAINE. 253 



far from thirty miles wide. Without assuming any definite rate 

 or rates of ice movement we can at least all agree that it would 

 take many years for the ice to advance such a distance. An. 

 obstruction to the natural transfer of heat beneath the ice, and 

 consequent enlargement of the tunnels, though it might be slow 

 in its action, would, after a term of years, have a cumulative 

 effect on the development of the tunnels, at least "in cases where 

 the subglacial rivers flowed in channels parallel with the ice 

 flow. 



We have seen that the three features of the coastal gravels 

 above stated are associated together over a wide area, and would 

 appear to have a common origin. Glacial rivers of different 

 lengths, from five up to more than one hundred miles, all show 

 the same development. At almost the same elevation they all 

 were able to sweep their tunnels clear of sediments. We must 

 seek for some cause capable of acting along two hundred miles 

 of coast in lines approximately parallel to the surface of the sea. 

 What but the sea itself could do this under so many varying 

 topographical and glacial conditions? 



Rightly interpreted, it would appear that the termination of 

 the gravel systems north of the shore line is itself a proof of the 

 former elevation of the sea. We may leave it as an open ques- 

 tion how far the sea acted in other ways — such as by diminish- 

 ing the effective "head" of the subglacial streams, etc., but that 

 the sea was chiefly responsible for the peculiar development of 

 the coastal gravels, I am persuaded, is the best interpretation of 

 the facts. And of all the ways in which the sea or other body 

 of water that submerges the base of a glacier affects the sub- 

 glacial streams and their tunnels, I have been able to discover 

 none so potent as that which is above described, whereby the 

 enlargement of the tunnels is obstructed. 



Where subglacial rivers flowed up and over transverse hills, 

 as they often did in Maine, there would be a body of slack water 

 in the tunnels, like a sewer trap, on the north sides of the hills. 

 Some of these bodies of slack water or dams on the north sides 

 of hills were from five to ten or fifteen miles long, and in one 



