REVIEWS. 745 



topography ; the production of vegetal beds and soils, and the deep 

 penetration of weathering." 



Following this period of deglaciation, the ice again advanced, mak- 

 ing, and finally leaving, the body of drift which it is proposed to 

 designate the East Wisconsin formation. As now exposed this for- 

 mation is much more extensive than either of the preceding, though 

 it is less extensive than either of the others were before they 

 were buried and disturbed or destroyed by later incursions of the ice. 

 It is characteristic of the East Wisconsin formation that it is bordered 

 by great moraine loops. It is also characteristic of this formation that 

 extensive gravel plains, in distinction from silt and loess plains, border 

 its moraines on the outside. It is in connection with this formation 

 that drumlins, kames, and osar are best developed. The East Wiscon- 

 sin formation is by no means simple. During its development there 

 were repeated oscillations of the ice edge. Some of them may have 

 been considerable, but they are not believed to mark more than 

 minor stages in the history of the glacial period. 



At several closely associated points in the vicinity of Toronto there 

 are fossil beds of stratified drift between beds of till. These fossil beds 

 have yielded a rich flora and fauna, which have recently been studied 

 by Messrs. Coleman, Townsend, and others. The character of the fos- 

 sils is unequivocal. They indicate a climate milder than that of the 

 present time in the same region. The molluscan fauna would be appro- 

 priate to Southern Illinois, the flora to Southern Ohio. Unfortunately, 

 the position of the fossil bed in the great drift series is not certainly 

 known. It is hardly probable that it belongs between the Kansan and 

 the East Iowan formations. It is more probable that it lies between the 

 East Iowan and the East Wisconsin formations. On the other hand, 

 it may lie between the East Wisconsin formation, and the deposits of 

 a later ice epoch which, within the United States, has not been differ- 

 entiated from the preceding. If the fossil bed occupies the position 

 last mentioned, there would be reason for separating the drift series 

 into four principal formations, rather than three. 



Professor Chamberlin is very conservative with reference to the 

 chronological importance of the several subdivisions which he pro- 

 poses. He does not assert that the three main subdivisions for which 

 he proposes names, are of equal importance. It is left an open 

 question which of the two deglaciation intervals between these forma- 

 tions is the more important. It is left an open question, so far as 



