THE OSAR GRAVELS OF THE COAST OF MAINE. 249 



In the coastal region we find numerous marine glacial deltas 

 deposited in front of the ice by glacial rivers that flowed into the 

 sea, but we do not find such frontal or overwash sediments as 

 naturally form in front of glaciers terrtiinating above sea level. 

 These and other facts prove that the ice had not all melted over 

 the coastal region before the advance of the sea. ^The subsidence 

 of the land (apparent advance of the sea) either preceded the 

 retreat of the ice over the coastal region or accompanied it in 

 such a manner that all the land free from ice was covered by the 

 sea as fast as the ice melted, up to the time when the sea had 

 advanced northward to the highest beach. That is, up to this 

 time, all the subglacial streams poured into the sea at the ice front 

 and not on land above the sea. It follows that the causes of the 

 ending of the osar systems north of the shore not only acted 

 parallel to the present and former surface of the sea, but also in 

 a region where the basal ice was bathed in sea water. 



The presence of deep glacial pot-holes in considerable num- 

 bers near the coast proves the existence of subglacial streams 

 in that region. Since there are no glacial gravels near these 

 pot-holes, we have proof that there were rapid subglacial 

 streams that left no gravels. Evidently their velocities were 

 such that they transported all their sediments beyond our field 

 (out into the region now under the sea). For years my con- 

 clusion has been that the osar rivers of the coastal region of 

 Maine were all subglacial. Assuming the subglacial streams, 

 the problem now resolves itself into this : How happened it 

 that as the subglacial rivers approached the coast, they all 

 found themselves able to sweep their channels free from sedi- 

 ments at nearly the same elevation? 



Without here pausing to consider the genesis of the sub- 

 glacial tunnels, we confine ourselves to the question, how are the 

 tunnels enlarged? Two physical agencies do most of the work. 

 First, mechanical erosion ; second, melting of the ice walls by 

 surface waters. In the case of ordinary mountain glaciers there 

 is usually considerable land on the mountains that is bare of ice, 

 and thus water warmed on the land passes beneath the ice and 



