WARREN UPHAM — DISTRIBUTION OF ENGLACIAL DRIFT. 135 



until we find its extreme boundaries marked by the outermost of its numerous 

 approximately parallel terminal moraines; but the inner moraines, of which no 

 less than ten have been mapped by the writer in Minnesota and North Dakota, 

 and a larger number by Mr. Frank Leveret t in Illinois, Indiana, ( >hio and Michi- 

 gan, tell of stages of temporary halt or re-advance interrupting the general reces- 

 sion of the ice. During the greater part of the second or last I rlacial epoch, through 

 the time of general growth of the ice-sheet, while it was performing most of its 

 work of erosion and transportation of the drift, with the incorporation of a large 

 volume of detritus in the lower portion of the slowly moving ice, the only deposits 

 which it made were the subglacial till, or ground moraine, and scanty stratified 

 drift in subglacial water-courses. Outside the glacial boundary during the same 

 time beds of gravel, sand, clay and silt were laid down in the avenues of drainage 

 from the ice-sheet by the streams of its scanty melting beneath throughout the 

 whole year, produced by the heat of friction and the slight access of heat from the 

 earth's interior, and of its plentiful melting above, near its edge, by the sun's heat 

 and by frequent rains each summer. But during the time of departure of this ice- 

 sheet, which is known as the Ohamplain epoch, very abundant deposition of the 

 drift that had been inclosed within the ice took place, partly as till or unmodified 

 glacial drift, and partly as stratified or modified drift, transported, assorted and 

 laid down by currents of water. Professor James D. Dana * first directed the atten- 

 tion of glacialists to this rapid formation of diverse drift accumulations of Cham- 

 plain age, and gave this name to the epoch from its fossiliferons marine beds 

 adjoining Lake Champlain, which had been described by Professor C. II. Hitch- 

 cock ;f and President T. C. ChamberlinJ has named the ice-held detritus " englacial 

 drift," this term being applicable to it wdiile it was inclosed and being borne for- 

 ward within the ice, from which during the final melting it was deposited in many 

 forms, as the upper till, perched blocks, kames, osars or eskers, valley drift, and 

 loess. Besides these deposits derived from the englacial drift when the retreat of 

 the ice set it free, the terminal moraines were formed chiefly or wholly from it 

 during stages of glacial growth and advance; and the drumlins and other masses 

 of subglacial till were also made mainly by gradual additions of material that had 

 been englacial. 



Nearly everywhere throughout the drift-covered areas glacial erosion has re- 

 moved all the preglacial residuary clay which more or less mantled the entire 

 country. This product of the preglacial denudation, and the gravel and finer 

 alluvial detritus of valleys, were plowed up by the ice-sheet and carried forward 

 in the direction of its motion; and portions gathered throughout great distances 

 along the path of the glacial current were mingled and thoroughly kneaded to- 

 gether. Occasional bowlders and rock masses were also supplied on the higher 

 lands by the irregular action of subaerial erosion and weathering before the ice 

 age, ready to he borne along and deposited in the glacial drift. Bui the ice-sheet 

 Commonly did more than to remove the loose material before existing, as is shown 

 by rock surfaces embossed, planed and striated by glacial erosion. In general, far 

 t he greater part of the drift was thus worn off, and most of its how iders were torn 

 and plucked away, from the rock floor over which the ice-sheet moved, grinding 



•Am. Jour. 8ci., :;<! Beries, vol. v, 1st.:, pp. 198 212, and numerous papers in vols. \. \ii. win, 

 xxiv, Kxvi, and sxvii, L875-1884. Manual of Geology, firs! ed., 1802, p, ">I7 : third ed., 1880, p. 543. 

 ; Gtoolog] of Vermont, vol. i. 1861, pp. L56 i > ,t . 

 ; I . 8. Geol. Survey, Third Annual Report, for l881-'82, p, 297. 



