94 W. UPHAM — THE SUCCESSION OF PLEISTOCENE FORMATIONS. 



Leaf Hills, Itasca, and Mesabi moraines.^ Beyond these, in the extreme 

 northern part of Minnesota, in Manitoba and onward, probably as many 

 more stages of the glacial retreat will be recognized whenever that wooded 

 and uninhabited country shall be fully explored. 



Loess and other Modified Drift Deposits. 



Much of the grist of drift produced by the ice-sheet was washed away 

 by the streams of its melting and of attendant rains, being deposited in 

 the ice-walled channels of the glacial rivers as eskers, at their mouths as 

 kames and kame-like plateaus, and along the valleys down which the 

 streams flowed on their way to the sea as stratified gravel, sand, clay, 

 and the fine silt called loess. All these genetically related formations, 

 supplied directly from the ice-sheet, especially during its final recession, 

 but modified by currents of water, being thus transported various dis- 

 tances, assorted in coarser and finer deposits, and laid down in stratified 

 beds, are collectively classed as modified drift, in distinction from the un- 

 modified glacial drift or till. If we consider the extensive deposits of the 

 loess and other finer and coarser beds of modified drift which were borne 

 beyond the boundary of the farthest ice advance and their continuation 

 northward upon the till along all the large waterways and on many 

 higher tracts of the glaciated area, it seems hardly too large an estimate 

 to assign a quarter or third part of all the glacial grist to this class of 

 drift modified by stream transportation and deposition. As noted in 

 my preceding paper, I believe that the streams in gathering this drift 

 from the ice-sheet flowed upon its melting surface, and that the modified 

 drift was nearly all derived from material that had been englacial and 

 became superglacial when the ice was melting away. 



In the loess of the Missouri and Mississippi valleys, a silt so fine that 

 the waters depositing it appear to have been very slowly flowing, broad, 

 but mostly shallow, river floods from the summer melting of the con- 

 tinental ice-sheet, we have a formation which proves that at the time of 

 maximum ice advance and during the early stages of its recession this 

 great river basin had somewhat less southward descent than now, prob- 

 ably with a lower altitude, yet not sinking to the sea level, and that its 

 floods spread over all the lowlands of the valley plain between its eroded 

 bluffs capped by the Lafayette gravels. During the cooler portions of 

 each year, however, the streams were restricted to comparatively narrow 

 channels ; and land shells lived, and probably grass and other vegetation 

 grew, on the bared flood-plain. 



* Descriptions of these moraines are in the Final Reports of the Geol. and Nat. Hist. Survey of 

 Minnesota, vol. i, 1884 ; vol. ii,1888 ; and vol. iv, describing northern Minnesota, j'^et to be published. 



