ESKERS IN THE VICINITY OF ROCHESTER, NEW YORK. 179 



ing glacial ice. However, in the case of the longest eskers, eskers 

 100 miles or more in length, no part of the esker ridge was formed 

 until the ice front had receded to within a few miles of that part, in 

 most cases 2 or 3 miles at most, the part already formed extending 

 beyond the ice edge to its leeward termination. In the case of 

 segmentation, characteristic of practically all eskers of any length, 

 the ice may have stood at or near the end of each segment as it was 

 forming, see Fig. I. 



As already indicated patches of gravel and sand are very char- 

 acteristic of intervals between segments, and were carried out by the 

 subglacial stream in which the segments were forming beyond the 

 edge of the ice and spread out. 



Thus an esker 100 miles long doesn't mean that deposition was 

 going on in a subglacial channel beneath the ice for 100 miles back 

 from the ice edge, but instead it means that conditions were favor- 

 able for fairly continuous deposition near the edge of the ice while 

 the ice front was receding for a distance of 100 miles. 



Eskers are especially apt to be developed in rough, hilly, rugged 

 regions, such as Maine, for here the ice would be crevassed afford- 

 ing initial passage ways for the subglacial waters. Here also the 

 ice flow would be likely to cease sooner while the ice was still thick, 

 the melting of which would afford a large body of water to be con- 

 tributed to the subglacial streams and its thickness furnish the 

 requisite "head" for such streams. On peneplain tracts crevasses 

 might develop as a result of tension thus giving opportunity of exit 

 for subglacial waters. However this point is not of prime import- 

 ance, since the pent-up subglacial waters must find exit, crevasses or 

 no crevasses. 



That till beneath the ice was eroded by subglacial waters has 

 been repeatedly observed. Hershey has described the erosion of till 

 and water laid drift within a short distance of a glacial lake, where 

 the erosion must have taken place beneath the ice (48). Lack of 

 evidence of erosion of till between esker segments and back of the 

 point of origin of the esker by the subglacial waters may be expected, 

 for subsequent deposition of till would tend to obliterate all such 

 evidence. The above observer describes the erosion of rock ridges, 

 more or less broken by the ice probably, that existed just behind the 



