WARREN UPHAM — DISTRIBUTION OF ENGLACIAL DRIFT. 143 



Relationship of the Englacial Drift to the Terminal Moraines. 



The irregular distribution of the englacial drift, thus abundant and scanty on 

 adjacent areas, was not apparently dependent on the character of the underlying 

 formations, nor on the altitude or configuration of the land, but rather on the 

 course, intensity, and limits of the great currents of glacial outflow from the central 

 part of the ice-sheet. There is consequently a marked relationship between the 

 inequality of distribution of this ice-enclosed material and the development of the 

 terminal moraines or marginal accumulations pushed out by these glacial currents 

 along the irregularly lobate boundaries of the ice during its maximum stage and at 

 halts in its recession and departure. The axial portion of each ice-lobe was more 

 an area of glacial erosion and less of deposition than its borders; and where the 

 glacial erosion was most severe and prolonged until the departure of the ice the 

 amount of both subglacial and englacial drift is small, the product of erosion being 

 carried to the outer portions of the ice-lobe or district of glacial outflow and there 

 deposited. On this principle we account readily for the deficiency of drift in 

 the extensive region north of the Lake of the Woods, Rainy lake, and Vermilion 

 lake, where the ice pushed strongly southwestward ; and for the abundance of drift, 

 much of it modified and therefore known to have been englacial, upon the adjacent 

 belt described between St. Paul and Winnipeg, where the ice currents from the 

 northeast and northwest were pushed together. The same principle also explains 

 the scantiness of drift upon large portions of Labrador and of most arctic lands. 

 Powerful glacial erosion has removed their preglacial superficial detritus and much 

 of the solid rock, sweeping nearly all its drift beyond the coast line. 



Even where small tracts of very scanty drift occur, with contiguous tracts 

 deeply drift-covered, as the instances cited in Massachusetts and New York, this 

 explanation probably still holds good, but applies to the latest movements of the 

 ice-sheet on these areas. It has seemed to me probable that the border of the ice 

 during its recession, melted by the returning warmer climate, had generally a more 

 Steeply sloping surface than in its time of greatest extent, and that in consequence 

 the rate of motion of the outer part of the ice-sheet was even increased during its 

 final melting. This would account for exceptional erosion and scantiness of drift, 

 either subglacial or englacial, on such limited tracts, and for its being thickly 

 amassed, often in drumlins, not far distant. Whether we consider the inequalities 

 of the drift distribution upon the larger tracts where they were due to the grand 

 movements of outflow of the continental ice-sheet, or upon the smaller tracts 

 where the irregularities of erosion and deposition seem attributable to minor 

 movements during the departure of the ice, it is clearly indicated that the deposi- 

 tion of the drift took place largely beneath the ice and near its boundaries. For 

 example, \ find reasons for believing that the drumlins near Boston were chiefly 

 accumulated during the departure of the ice at distances of only a few miles inside 

 its retreating edge* At the same time, probably, the tract of very scanty drift 

 close to the north was undergoing severe erosion. 



Impressive as are the more massive portions of the belts of marginal morainic 

 drift, they musl have been far larger if the ice had home most of its englacial drift 

 quite to its margin, instead of depositing it as subglacial till beneath its compara- 

 tively thin border. Professor N. S. Shaler estimates the terminal moraine on the 

 northwest pari of Marthas Vineyard to be on the average L50 feel thick, its volume 



♦ Proceedings Boston Society of Nnturnl History, vol. sxiv, 1889, pp, 228 242. 



