108 BULLETIN OF THE ESSEX INSTITUTE. 



submergence of over 650 feet at Bristol, Conn., while at 

 Bristol, R. I., in the same latitude, the nearest wash- 

 plains would indicate a submergence to a depth of about 

 50 feet. Other anomalies, if we hold sea-level to lie 

 rigidly at delta plain level, appear in the Narragansett Bay 

 region as I have pointed out in another paper. There we 

 have the Slocumville plain at 160 feet in the hills, fol- 

 lowed by plains at 50 feet in the low now open grounds ; 

 and the Attleboro deposit at 140 feet in the low grounds 

 with the Woonsocket deposit at 300 feet in the hills and 

 only five miles farther north. In this latter case, we 

 should have a tilt rate of 32 feet to the mile ! 



It may be objected to the above statement of the marine 

 limit hypothesis that the high plain at Woonsocket for 

 instance was built during the deeper submergence which 

 attended the going off of the ice, while the low level plain 

 at Attleboro was deposited later when the land, unladen 

 of much ice, had risen higher. But this argument is met 

 by the rather decisive facts in the glacial history, showing 

 that the Woonsocket deposit belongs to a line of retreatal 

 moraine formed later than the Attleboro accumulation. 

 The attempt, therefore, to interpret sea-level by a rigid 

 application of the criterion of wash-plain level involves us 

 in hopeless inconsistency, sudden changes of level, and the 

 need of having the sea at different levels at the same time 

 in the same region. 



If the water-level index afforded by delta fronts means 

 anything at all, it seems to point to local bodies of water 

 standing at levels dependent on local topographic condi- 

 tions as in temporary glacial lakes or flooded areas by 

 which I mean bodies of water formed in basins where the 

 rise of the water is due to the excess of inflow over out- 

 flow, however brought about. The occurrence of plains 

 in high grounds along the south coast as well as on the 



