ESKERS IN THE VICINITY OF ROCHESTER, NEW YORK. 181 



growing- eskers which are composed largely of their fragments. 

 The streams mtist have been subglacial to get at these ridges, and 

 fm-ther, materials from the neighboring hills are absent from the 

 eskers showing that the esker streams did not have access to those 

 hills and did not get possession of drift they may have contributed 

 to the englacial and surface portions of the ice. This hypothesis 

 also accounts for the occurrence of eskers in troughs, so frequently 

 noted by Leverett. These troughs were eroded by the same sub- 

 glacial streams in which the esker ridges were deposited. 



Tributaries to eskers are rare, for if the ice possessed move- 

 ment it would obliterate such features inasmuch as they would be 

 for the most part transverse to the direction of ice flow. 



Double ridges may be accounted for by the formation of a 

 broad arch which, unable to support itself, bent downward dividing 

 the tunnel into two parts in each of which deposition took place, or 

 locally a deposit was formed in a superglacial or englacial channel 

 which, protecting the ice beneath from melting, slid down both sides 

 of the resulting ridge. 



Accordant levels of delta and feeding esker are significant. 

 Eskers are never greater in height, and rarely of less altitude than 

 the delta to which they are tributary. This is to be expected under 

 this hypothesis for aggradation would cease in the subglacial tunnel 

 when the upper level of the material clogging its mouth was attained, 

 the upper level of the delta (27, 28). 



Gaps in esker courses may be accounted for by subsequent 

 erosion, by glacial erosion of the once continuous ridge, by lack of 

 deposition resulting from lack of confinement of the subglacial 

 stream to a definite channel where the gap occurs, by an ice-block 

 falling into channel with deposition behind and in front of it, but the 

 water possessing too great velocity in passing around it or not 

 definitely confined to channels resulting in no deposition. Stone 

 suggests that the ratio of the volume of water and the size of the 

 tunnel varies in such a way that deposition takes place where the 

 stream is small and the tininel large and the velocity is therefore low, 

 and that deposition fails where the ratio is reversed and the velocity 

 high (98). 



The fact that the stream was under high pressure accotuits for 

 esker courses across divides, across valleys, over rough topography, 



