80 w. UPHAM — dp:rivation of kames, eskers, and moraines. 



tion is doubted by some most experienced giacialists, I have subjected 

 it to critical reconsideration, and have thereby gained new confidence 

 that it is true. The examples which have been described and discussed 

 seem representative of all our kames, eskers, and moraines. Small 

 deposits belonging to these classes, not less decisively than the grand 

 instances reviewed, appear by similar features to be likewise of englacial 

 derivation. 



The drift on Long island may be thought to have become englacial 

 only by being eroded from the hills and mountains of New England on 

 the north, being carried forward in the portions of the ice-sheet passing 

 their sides and crests; but at Rochester, N. Y., and in the region of lake 

 Agassiz and of Devil's lake in North Dakota, it must have been carried 

 upward into the ice on a nearly flat country. All giacialists readily 

 admit that land ice flowing over a mountainous country must carry 

 more or less of its drift onward enveloped within the ice at great heights ; 

 but we have good evidence at Rochester that glacial currents on that 

 nearly level area rose from the bottom of the ice into its mass so rapidly 

 as to carry bowlders up more than 200 feet within a few miles, and at 

 Bird's hill we see that there was much englacial drift borne along in the 

 ice-sheet upon that vast flat region at greater altitudes than 500 feet 

 above the ground. Not only from these flat areas, but also from low- 

 lands among hills and mountains, I think that much drift was eroded 

 and carried upward into the lower quarter or third part of the ice-sheet, 

 being most abundant near the ground and gradually diminishing upward. 

 The steady accession to the mass of the ice-sheet over any place by on- 

 flow from its thicker central part and especially by the accumulating 

 snowfall doubtless forbade the drift of the upwardly moving basal cur- 

 rent from being carried far into the ice in comparison with its total 

 thickness. 



Bowlders and other drift becoming incorporated in the higher portion 

 of the zone reached by the currents flowing upward would be carried 

 forward far at that height in some regions, as from the Huronian and 

 Laurentian areas north of lake Huron to the bowlder belts in Illinois, 

 Indiana, and Ohio, described by Chamberlin, without intermixture Avith 

 other englacial drift brought into the ice by less powerful currents on all 

 the intervening extent, which in the case mentioned is about five hun- 

 dred miles. The englacial transportation proved for these bowlders 

 seems to me to imply that much drift of less distant origin was carried 

 in the ice at lower altitudes, and that the portion of the englacial drift 

 within a few hundred or probably a thousand feet above the land, where 

 the ice-sheet was a half mile to one mile thick, must have become well 

 intermingled. 



