Plehtocene P(iper><^ Ottav:a 2leetin(/ <>f (i . S. A. 173 
referred to the observations of Torell, Hoist, and others, ou the 
exposure of the enghicial and superglacial till to full oxidation 
and weathering, while the subglacial till was protected by its 
position beneath the ice. The larger amount of erosion of the 
outer and older portions of the drift sheet seems attributable to 
its longer exposure where the drainage from the waning ice- 
sheet passed over it. 
Mr. W J McGrEE would add, as another criterion of an inter- 
glacial epoch, difference in the origin and size of boulders in the 
older and newer drift, as in northeastern Iowa. Nodules of bog 
iron ore in the Iowa forest bed, implying a long interval of forest 
growth, are striated by the succeeding ice incursion. 
Prof. C. H. HiTCFicocK remarked that the amount of oscilla- 
tion of the ice-front to be ranked as a division between glacial 
epochs must be a difficult question to be determined by every 
glacialist according to his own judgment, unless it can be shown 
that the ice-sheet was almost or entirely melted awa}' and after- 
ward was anew accumulated. In New England and the eastern 
provinces of Canada probably as many terminal moraines are 
recognizable as in the upper Mississippi region, but these all in 
both regions seem referable to stages of slight interruption in the 
decline of a single ice-sheet. No support for an interglacial 
epoch is found in fossiliferous marine beds between deposits of 
till on our Atlantic coast, for such beds are absent, excepting 
where, as at Portland, Me., and St. John, N. B., they imply 
only a moderate re-advance of the ice interrupting its general de- 
parture. Stream erosion during the Ice age may have proceeded 
very rapidly where the land was much higher than now. 
Mr. Warren Upiiam, referring to the first three of Prof. 
Salisbury's criteria, thought that the rapidity of departure of the 
ice-sheet, as shown by the eskers, associated plains and plateaus 
of gravel and sand, and the valley drift, implies for the closing 
stage of the Glacial period fully as warm summers close to the 
border of the retreating ice as at the present time in the same 
latitudes. Therefore a temperate flora and fauna would exist 
near the retreating ice. But the terminal moraines show that at 
various times the general glacial retreat was interrupted by secu- 
lar changes in the climatic conditions bringing increased snowfall 
and a halt or re-advance of the ice. If at such times it advanced 
only a few miles, the fossils of the beds covered by the later till 
