L96 W. M. DAVI! GLACIAL SAND PLAINS. 



hillocks — eskers and kames. The other three-quarters of the perimeter, 

 turned to the Bouth, is even more Btrongly characterized by a number of 

 convex lobes and sub-lol iarated by re-entranl interlobate hollows. 



The internal Btructure of the plains, where revealed by railroad cuts and 

 Band pits, consists In greatest part of obliquely deposited beds of .-and. or 

 occasionally of sandy gravels, dipping towards the lobate margin al an angle 

 of from 20 to 25 c : Inn these are covered by gravelly or sandy cross-bedded 

 horizontal layer- to a depth of from five to fifteen or more feet; and the 

 thickness and coarseness of this cover appears to increase towards the esker 

 and kame margin. 



Hypothesis of Origin. — In view of these facts of form and Btructure, it is 

 difficult to find any explanation for our .-and plains other than the one gen- 

 erally current, which regards them as delta-like deposits of sand and gravel, 

 washed in the closing stages of the last glacial epoch from the irregular front 

 of the melting, stagnant ice-sheet into bodies of water that bathed its edge. 

 Before looking further at the facts, let us extend this hypothesis as far as 

 possible to its consequences, and then test its correctness by the complication 

 of correspondence between deduction and observation. 



Deductive Extension of the Hypothesis. — The former existence of an ice- 

 -heet over New England i- accepted as evidence that is entirely independent 

 of the occurrence of Band plains. The ice-sheet is now gone; and between 

 the times of its greatest thickness and fastest motion and of its 'entire disap- 

 pearance it must have been reduced to a thickness at which motion was im- 

 possible; then it lay passive and stagnant, as Chamberlin has pointed out," 

 for the remainder of its existence; during this time it must have melted irr< 

 ularly, presenting a very uneven, ragged front, from which residual blocks 

 may have been frequently isolated: and it must have endured longest in 

 the valhy-. where it was thickest, not only by reason of it- greater depth, 

 hut also because it- surface there, where motion had Keen fastest ami longest 

 maintained, must have been higher than on the hill- —this being homologous 



with the variation in the thickness of a Swi<- valley glacier from middle to 



-id. 



A melting ice-sheet must have frequently embarrassed the drainage of the 

 BUrface on which it lay; ponds would accumulate in hollows and vallt 

 sloping towards it. a- I rphara ami Met ice have indicated, and after Btanding 

 lor a time at a level determined by one line of overflow they must have sud- 

 denly fallen or drained away a- m-w outlet- were opened by the melting of 

 the ice, thus causing active floods. Near the coasl a moderate (relative de- 

 pression of the land seems to have brought standing sea water against tin' 

 ten or twenty mile- inland from the present shore-line. 



Wherever active, drift-laden Btreams ran from the melting ice into Btand- 

 ing water at it- margin, their velocity must have been checked and all but 



