80 BOTANICAL GAZETTE [AUGUST 
conditions must have been preceded and followed by two compara- 
tively cool periods, characterized by changes between drought 
and wetness greater in degree than seasonal variations. For this 
interpretation, however, a series of various facts is doubtless 
required. A consideration of the structural appearance of the 
deposits in line north of Canton should give more adequate evidence 
of such alterations. They represent in part a contemporaneous 
and later age of peat formation which should bring into clearer 
perspective the probable climatic conditions during and after the 
close of the third glacial substage. 
The recession of the ice front marked by the Valparaiso- 
Kalamazoo-Mississinawa morainic system to near the border of 
the Huron and Erie basins initiated the development of the Kent 
(fig. 10) and the Mantua peat deposits (fig. 12) in the order named. 
An examination of the profile sections suggests a long interval of 
peat accumulation. In about the middle of the Kent deposit there 
is evidence that here also an unusual disturbance had affected the 
course of peat formation, and that a well marked climatic change 
had occurred. The position of the layer of forest peat in the Kent 
deposit suggests that the change is contemporaneous with the 
deposition of the Lake Border-Defiance system. 
At the bottom of the Kent deposit, overlying the bowlder clay, 
are shells of fresh water mollusks, and above them a layer of plant 
remains from aquatic vegetation. This is followed by macerated 
material, a part of which is distinctly gelatinous. The upper por- 
tion of the structureless débris merges into a layer of fibrous plant 
remains, showing that a mat of sedges and other marsh plants had 
covered the basin. When this stratum was formed, a mixed 
deciduous but predominantly coniferous forest appears to have 
been growing on the borders of the basin, which gradually 
encroached and finally occupied the entire peat-land area. The 
thickness of the layer of forest litter shows that the ground water 
level at that time was below the surface soil, and that the tract 
remained moderately moist for a considerable period of time. 
It can scarcely be decided in the present state of investigation 
whether or not the end of the Mississinawa glacial substage was 
accompanied by a widespread dispersal of forest trees from south- 
