SOME OLAOIAL WASH-PLAINS. 87 



As yet the manner of flow, in the ice-sheet, of the 

 streams wliich produced the greater sand-plains, has re- 

 ceived little light from studies on the ground. This is 

 partly because the structure of our sand-plains is rarely 

 exposed at the head or terrace in a manner to show the 

 method of building. From studies conducted on the 

 Woodland plain it appears that building went on along 

 the entire front, quite regardless of the esker which joins 

 the plain on Beacon street. There is a very rapid passage 

 outward in the plain at the mouth of the esker channel 

 frojn coarse gravels to fine sands. The appearance of the 

 contact zone where seen in the plain is such as to show 

 that the esker built up^)«Kj;a6*.s'w with the plain, and that 

 there Avere streams flowing in or on the ice of which no 

 record now remains in the intraglacial field. From anal- 

 og}' with the conditions of discharge in Alaskan glaciers, 

 made known by Russell's studies, Ave might expect waters 

 under hydrostatic pressure bursting out as " springs " along 

 the marginal portion of the ice-sheet, thus breaking out on 

 the surface of tiie ice where it would be easier to main 

 tain an open passage than through the clogging sand in 

 the contact zone of the plain. An abandoned channel of 

 this sort, almost connecting with the plain but filling up 

 with gravel and sand, would present the " notch " Avhich 

 separates some eskers from their wash-plains, a feature 

 which forms at present the chief stumbling-block in ex- 

 plaining the relations of esker-channels to their fans. 



GEOGRAPHICAL DISTRIBUTION OF THE WASH-PLAINS OF 

 SOUTHERN NEW ENGLAND. 



(The numbers in parentheses refer to townships on the map, Fig. 7.) 



In the uplands of this region, sand-plains are practi- 

 cally wanting. If these deposits occur there at all it is 

 in narrow north and south valleys in association with rem- 



