Drainage of Carboniferous Area, Michigan. — Mudge. 307 
this valley is rather flat, the center being depressed about 200 
feet below the marginal portions. The northwestern Carbon- 
iferous boundary is within this elongated valley for a distance 
of 75 miles. Our conclusions in regard to the origin of this 
valley and river are of a somewhat more speculative nature 
than the foregoing. Nevertheless, some interesting generali- 
zations may be drawn. 
It is worthy of notice, in the first place, that this valley oc- 
cupies exactly the place where one would naturally expect to 
find the backbone of the peninsula — the line of water-parting 
between the streams flowing west to lake Michigan and those 
flowing southeast to Saginaw bay and the central depression. 
It is as though the original watershed had been for many miles 
longitudinally cleft by the river into two roughly parallel 
watersheds. To learn the origin of these remarkable condi- 
tions is the problem before us, to which the circumstances 
suggest two possible solutions. 
The first is that the line of first separation between the 
eastern and western ice lobes was along the line of highest 
elevation ; and upon this highest ridge, flanked by the sepa- 
rating lobes, the river was formed by the glacial waste and 
established its permanent course. This theory is, in effect, a 
modification of that applied by Mr. McGee to certain rivers of 
northeastern Iowa (Eleventh Annual Report, U. S. Geol. Sur- 
vey, pp. 218, 219, 236). While this theory applies very nicely 
to the river, it seems quite plain that the valley did not orig- 
inate in this manner. The time since the Glacial period has 
been too short to permit the formation of so wide a valley; 
and besides, its surface is now covered with drift materials 
and dotted with small lakes, many of them being near the 
stream channel, indicating that it was ice-covered after its 
present form was attained. 
The second theory is more complex and difficult to work 
out in detail. It is suggested by the fact that for so long a 
distance the valley coincides with the Carboniferous bound- 
ary. If this valley is of preglacial origin we naturally assume 
that it was the result of stream erosion. The question then 
arises, Why should the ancient stream follow so persistently 
the strike of the undisturbed Carboniferous strata, rather 
than their natural inclination V This is a most difficult prob- 
