Eskers of the Kansan Epoch. — Uershey. 197 
land itself at present. The direction of ice movement in Lab- 
rador was not determined excepting at Turnavik, where it 
was off shore. In Baffin land it was difficult to tell how the 
ice moved, because the strite were destroyed by post-glacial 
weathering; but on the southern side the ice movement was 
apparently southeastward, as if Baffin land had been a sepa- 
rate center of ice dispersion. 
The glacial action produced more effect in the down-cut- 
ting of the surface in Labrador than in Baffin land, for the 
surface features of the former are very much more rounded 
than those of the latter. Even near Turnavik, the ice cutting 
did not succeed in lowering the surface down to the level of 
the zone of weathering in the weaker rocks. In Baffin land 
the down cutting of the surface M'as even less pronounced. 
Both in Labrador and Baffin land there is evidence that the 
ice has withdrawn from these regions in very recent times. By 
comparison with studies that have been made in parts of New 
England it seems that the period since the ice left is much 
shorter in the northern lands. Comparing these northern 
regions with New England it seems certain also that the ac- 
tion of ice erosion has produced less effect upon the topog- 
raphy of the northern than the southern regions. On the oth- 
er hand, as we proceed northward the effects of post-glacial 
weathering become more and more pronounced, so that the re- 
cency of the ice uncovering is more and more masked. 
ESKERS INDICATING STAGES OF GLACIAL RE- 
CESSION IN THE KANSAN EPOCH IN 
NORTHERN ILLINOIS. 
By Oscar H. Hershey, Freepurt, 111. 
(Plate XI.) 
Introduction. 
The following pages present the result of an investigation 
into the character and origin of a remarkable series of ridges 
and knolls of stratified gravel and sand, situated mainly in 
the Peeatonica basin in northwestern Illinois and the adja- 
cent portion of Wisconsin. They constitute a very important 
part of a drift sheet which, although unevenly distributed, is 
very thin and presents no true moraines. It has now become 
well known as the Kansan drift sheet, and it is exposed in a 
