30 The American Geologist. July, i899 
In that event lake Warren must have continued to exist during 
the halts at the Lockport and Albion moraines, though at the 
time of the latter it was beginning to shift its outlet from west 
to east and soon fell away to a much lower level. 
The correlative relations of these features with the moraines 
in Michigan seem to leave little doubt that during the time of 
the Lockport and Albion dams there was an open passage from 
the Huron to the Michigan basins through the straits of Mack- 
inac. The writer has been of opinion heretofore that lake War- 
ren never extended to the basins of lakes Michigan or Superior. 
That opinion was based on the fact that careful search for 
shore lines above the Algonquin beach at Petoskey, Mich., and 
near Indian river and Rogers City farther east on the same 
slope had yielded only negative results, although these locali- 
ties were very favorably exposed. The ground at Petoskey was 
reviewed with special care in September, 1897, but without 
finding any evidence of a beach. At the same time, the writer 
had ventured to entertain some doubt as to whether Mr. Lev- 
erett's Gowanda and Alden moraines were really distinct mem- 
bers of the recessional series. It was thought that they might 
possibly be spurs or secondary ridges belonging to one of the 
other moraines near by — the Gowanda to the Dayton (Euclid) 
and the Alden to the Hamburg. While nothing has occurred 
to change the report of facts relating to the absence of high 
beaches at Petoskey and eastward through the straits, a careful 
reconsideration of Mr. Leverett's work on the moraines of 
New York suggests a contrary conclusion. It suggests that 
in its last stages lake Warren probably did extend westward 
through the straits, although not occupying their whole width. 
For, taken in connection with the clearly defined correlation 
of the moraines in the two branches of the diverging series, it 
seems plain that Mr. Leverett's positive evidence from the 
moraines of New York furnishes a safer ground of inference 
than the writer's negative evidence as to the extension of the 
Forest beach through the straits of Mackinac. Pending further 
research in the field, therefore, the conclusions suggested by 
Mr. Leverett's work should stand first. If lake Warren did in 
fact extend to the Michigan basin, then it blended with lake 
Chicago (the glacial lake of that basin) and the Chicago chan- 
nel became for a time its outlet. A slight drop of the water 
