Drainage of Carboniferous Area, Michigan. — Mudge. 305 
movement over the lake region was from northeast to south- 
west, with local variations caused by river valleys or other 
already existing surface irregularities. Thus, it is presumed 
that lakes Huron and Michigan mark the sites of former im- 
portant streams, in the valleys of which the erosive force 
exerted was necessarily much greater than on the highlands, 
and hence the formation of the broad lake basins. This in- 
tense erosive force was also felt in the valley of the eastern 
Carboniferous outlet, and Saginaw bay was the result. When 
the continental glacier receded and thinned out, its margin 
took the form of great lobes which occupied the lower por- 
tions of the land. A branch of the great lake Erie lobe occu- 
pied Saginaw bay and valley and extended far across the 
country to the southwest. In its retreat it gave rise to two 
whole families of rivers. 
The origin of Grand river was fully described by the writer 
in the American Geologist for November, 1893. It was briefly 
as follows : The Saginaw glacial lobe having retreated over 
the watershed near the southwestern boundary of the Carbon- 
iferous district the accumulating waters flowed to the north 
and west along its front before rising high enough to pass 
over the watershed. 
Retreating to the northeast down the gentle slope of the 
country, the waters were again unable to rise to their recent 
outlet, the Grand, and a second river, the Cedar, was formed. 
It flowed to the west, uniting with the Grand near Lansing. 
Leaving the Cedar in an established channel, the glacial lobe 
continued its recession down the slope, and in a similar man- 
ner gave birth to the Looking-glass river. This also flows to 
the west, reaching the Grand at Portland. Still later that part 
of the Maple south of the central depression was formed, 
probably in the same manner. All these streams now at 
an angle with the natural inclination of the surface. 
Though these conclusions have not been reached through ;i 
personal visit and examination of all the streams, it is be- 
lieved that they are entirely warranted. The problem in brief 
is this: Given a glacial lobe retreating slowly down a gentle in- 
cline, and hemmed in at one side and in front by an elevated 
ridge, to determine the direction of How of the waters result- 
ing from its waste. It is clear that they would escape in the 
