8 The American Geologist. July, i899 
cupy the interval between Cincinnati and the straits of Mack- 
inac, and there are seventeen between Cincinnati and Roches- 
ter, N.Y.* 
It is obvious that the rhythm of retreat was not local. It 
was not confined to any one lobe or basin, but was continental 
or hemispherical or, more probably, terrestrial in extent. 
Rightly understood, this rhythm of the moraines is of incalcul- 
able value in the study of glaciation, and nowhere is it more 
so than in the study of the glacial lakes. It furnishes the key 
for the correlation and identification of moraines and ice-dams 
in different basins and in areas far apart. Where two sets or 
series of moraines are clearly defined in two adjacent basins, 
if the contemporaneity of any particular moraine in one basin 
with a particular moraine in the other can be clearly made out, 
then the correlation of the whole of the two sets into contem- 
poraneous pairs may be accomplished. Of course, the simpler 
the series and the more clearly the relations are defined the 
more reliable will the correlation be. But there are two situ- 
ations in which, with the individual moraines of the series 
clearly marked, precise correlation becomes absolutely sure. 
One is where a symmetrical lobe of the retreating ice-sheet di- 
vides around some highland mass into two distinct and clear- 
ly separated lobes; the other is where two such lobes retreat in- 
to one basin and become united into a single symmetrical 
lobe. If the moraines, after separation in the one case or 
before union in the other, form two clearly defined series con- 
tinuous with that of the single, symmetrical lobe, then correla- 
tion between the members of the two branches is perfect. For- 
tunately, a good example of each kind occurs within the area 
here considered. The very symmetrical Maumee ice-lobe 
divided into two parts as it retreated, one receding down the 
basin of lake Erie and the other up the Detroit-St. Clair val- 
ley. The latter part united at a later stage with the Saginaw 
lobe in the basin of lake Huron. When the Erie lobe had re- 
treated as far as the western end of lake Ontario, it separated 
from the Georgian Bay lobe. The moraine series of the Sagi- 
naw and Georgian Bay lobes have not yet been fully explored. 
*The character of the oscillations and their probable cause are dis- 
cussed in a previous paper entitled "Moraines of Recession and their 
Significance in Glacial Theory." Jour, of Geol., vol. v, No. 5, July- 
August, 1897. 
