ICE WORK IN SOUTHEASTERN MICHIGAN 199 
the time the lake is reached, forming the western slope of a 
very much larger trough underlying the western end of Lake 
Erie. An inspection of the Lake Survey chart shows that the 
eastern slope of this broad trough is terminated by a rock-ridge 
upon which are located West Sister Islands and Middle Sister 
Islands, running S. 35° W. Northeastward, towards the Ontario 
shore, occur Colchester reef and other rock. Toward the south- 
west this broad trough is continued beneath Toledo for a long 
distance as the Maumee trough, which would be filled with 
water were the drift removed, standing one hundred feet deep 
at Toledo, and extending seven miles further west than at pres- 
ent along the Ohio-Michigan line. Eight miles east of the 
ridge above-mentioned there occurs a second one running across 
the lake S. 32° W., and indicated by Niagara reef, East Sister 
Islands, and several other shoals and reefs. Between these 
ridges a second trough reaches across the lake, having its axis 
parallel to all the others. Beyond the second ridge, and 
between it and the string of islands which stretches continuously 
from Catawba Island northeasterly to Pointe Pelée, there is 
a narrower and less well-defined trough, but parallel to the two 
other Erie troughs. The position of this series of troughs, 
their direction and approximate parallellism, their shape, and, 
so far as known, their direction of slope toward the northeast, 
all suggest that the primary agency was ice, acting with excep- 
tional vigor straight out from the center of Laurentide accumu- 
lation. It so happened that the strike of the rock-strata in this 
region coincided with the main axis of ice movement, and the 
troughs in all cases were formed in the softer beds. It is not 
improbable that preglacial streams occupied some of these 
valleys, as in the case of the Sylvania, and that where the valley 
had the same trend as that of the ice movement it received the 
maximum ice erosion. The ice mass appears to have been too 
great at this stage to permit of its following the valleys when 
they turned aside from its general course. We seem to have 
here, but on a much larger scale, a similar phenomenon to that 
observed by Gilbert in western New York.' 
* Bull. Geol. Soc. Am., Vol. X, p. 121, 1899. 
