110 BULLETIN OF THE ESSEX INSTITUTE. 



The picture presented by Professor Davis of the mar- 

 ginal portion of the ice broken up into isolated blocks 

 around and between which streams deposited gravels and 

 sands is again and again forced upon the mind in the low^- 

 land of the state and in the valleys in the uplands. These 

 ice-block holes as the bergs now present themselves to us, 

 like the sands which surround them, do not mark a single 

 phase of the retreat. As in the Narragansett Bay region, 

 the drift phenomena are increasingly newer as we go 

 northward. The repeated overlap of the lobate front of 

 one wash-plain upon the esker and kame deposits of an 

 earlier stage to the southward is sufficient evidence of the 

 general truth of this statement. This mode of retrogres- 

 sion of the front is what we should expect in the case of 

 an ice-sheet thinner on its margin than in its central part. 

 The existence of recessional wash deposits does not there- 

 fore of itself disprove the idea of a period of general and 

 complete stagnation of the ice over this area. But when 

 we consider the evidence of forward movement of the ice 

 at several successive lines across the eastern part of the 

 state as in the Middleboro, Providence-Bridge water, and 

 Cambridge moraines, it becomes evident that the ice-sheet 

 as a whole did not lie stagnant on the area. There were 

 periods of marginal inactivity, accompanied by the tunnel- 

 ing of running water, esker- building, terrace and plain 

 construction, with a general retreat of the main front, 

 followed by seasons of advance, with the shoving of drift 

 deposits, the spreading of till and boulders over wash- 

 plains. 



The occurrence of the several morainal patches with 

 wash-plains in lines which traverse the area between the 

 head of Narragansett Bay and the south side of Boston 

 Bay is further evidence of forward movement in the ice 

 sheet. These lines of frontage obey the law of marginal 



