THE OSAR GRAVELS OF THE COAST OF MAINE. 251 



region of the glacier furnishes the heat to enlarge the tunnels 

 within its own limits. This is the natural career of ordinary ice 

 sheets above the sea. 



An important law of the enlargement of subglacial tunnels 

 depends on the velocity of ice movement. Subterranean waters, 

 as those of the limestone caves, go on enlarging their channels 

 from age to age, because they act continuously on the same 

 body of rock. But the subglacial tunnel cannot become thus 

 enlarged, because of the constant renewal of the ice. Other 

 things being equal, the enlargement of the subglacial tunnels is 

 directly proportional to the time during which it is being 

 enlarged, and inversely as the rate of ice flow. Obviously, 

 when the flow is rapid the tunnel never becomes very much 

 enlarged, for before this can happen the ice at any given part of 

 the channel is pushed on to the distal extremity and disappears 

 by melting or by berg discharge. 



The details, here omitted, prove there was probably a small 

 acceleration of the rate of ice flow as the coast of Maine was 

 approached, hence the rate of enlargement of the ice channels 

 would not increase so rapidly as the supply of water of local 

 melting. But the surface of that region is much diversified with 

 hills and valleys. The rate of ice flow would be most rapid in 

 the deeper north and south valleys, and would be retarded in 

 the lee of the higher transverse hills, of which there are several 

 long systems. If differences in the rate of ice flow were the 

 only cause of different rates of ice channel enlargement, then we 

 ought, on such an uneven coast, to find evidence of the fact in 

 the distribution of the gravels. Examination shows that this 

 was a real cause of varying rates of enlargement, but it was a 

 minor cause. This cause alone could not have enabled all the 

 subglacial rivers to clear their tunnels of sediments at the same 

 or nearly the same horizontal line. It would have acted at vari- 

 ous levels, according to the conditions for rapid ice supply from 

 the north. 



We have seen that the ice in late glacial time flowed into the 

 sea in the coastal region of Maine. It remains for us to inquire 



