LATEK MORAINES OF LAKE MICHIGAN, SAGINAW, AND HURON-ERIE LOBES. 253 



ville, a distance of about 45 miles from its head near Hadley. It is particularly well developed 

 in Holly Township. Its course is mostly through a jumbled, broken morainic country, some of 

 which has moderately strong relief. Through this the river wound deviously, in places through 

 narrow passages between high hills, as in the eastern and western parts of Holly Township, and in 

 places over broad flats where it laid down extensive deposits of gravel and sand, as in central 

 Holly and central Deerfield townships. Its character and continuity are unbroken to the point 

 near Fowlerville and it has evidently not been overridden by a later readvance of the ice east of 

 there. Its relation to the general slope of the country, which declines toward the northwest, and 

 to a number of streams which flow in that direction across it and through some of the moraines 

 north of it, establish its relation to the ice front very clearly. 



Kersley Creek, for instance, rises in the high morainic hills south and east of Ortonville, 

 flows northwest across the Holly channel, through the Hadley ridge and three other moraines 

 and across the drainage channels or lake beds in front of them, and finally reaches Flint River 

 3 miles north of Flint. Another branch of the same stream crosses the channel and the Hadley 

 ridge in western Groveland Township and joins the main creek near Goodrich. In Tyrone, 

 Deerfield, and eastern Cohoctah townships several of the headward branches of Shiawassee 

 River flow northward across the Holly channel and through the moraines to the main river, 

 which follows the course of glacial drainage westward along the front of the Fowler moraine. 

 East of Cohoctah Township the belt of jumbled moraines extends about to Goodrich and is 

 only 3 or 4 miles wide, being bounded on the north by the drainage channel along the front of 

 the Fowler moraine and on the south by the Holly channel. 



Some channels of glacial rivers throw a more certain light on contemporary positions of the 

 ice front than do the moraines, especially where the course and continuity of the latter are doubt- 

 ful. It seems entirely certain, for instance, that when Holly River was flowing the ice front 

 Tested close along its north side, at least from Lapeer County to the north-central part of 

 Livingston County, for if the ice had stood farther toward the south the site of the channel 

 would have been buried under ice, and if it had stood much farther north the river would have 

 turned in that direction along the course of some of the creeks that cross it and would have 

 escaped westward on a lower and more northerly line. 



The fragment of old channel at Lansing may be a continuation of the Holly channel, which 

 it resembles closely, especially as the Cedar Valley seems a natural course westward from Fowler- 

 ville. The Lansing channel was closed, as described above, by a readvance of the ice front to the 

 Lansing moraine, but the Holly channel in Lapeer County runs along the front of the Hadley 

 ridge, which seems to correlate with the Portland or fourth moraine farther west. Hence the 

 Holly channel can hardly be earlier than the Hadley ridge which supported it. For this reason 

 it seems certain that the Lansing channel is not a part of the Holly but is probably a fragment 

 of an earlier channel which has been overridden and obliterated entirely east of Okemos. 



The closing of the Lansing channel west of that city by a readvance of the ice to the Lan- 

 sing moraine indicates a preceding recession of the ice front for at least 2 or 3 miles. But the 

 character of the Lansing channel — its magnitude and relative strength of development — cou- 

 pled with the fact that it seems to be entirely overridden in the eastern part of the area, are still 

 more significant in their bearing on conditions immediately preceding the building of the Lan- 

 sing moraine. The size of this channel indicates a large river, as large or larger than the Holly. 

 Its source must have been as far back as Lapeer County, where the Holly channel gathered most 

 of its waters, and the Lansing fragment lies only a little lower. It looks like an earlier drainage 

 line of which the Holly is a later revival on a slightly different course. Lansing River probably 

 drained substantially the same area, but its course, not far north of the Holly, has been over- 

 ridden and obliterated. It evidently marks a somewhat longer interval of time than that which 

 separated the successive slender moraines of the deployed group, and it indicates a greater read- 

 vance than that which closed a few miles of the channel west of Lansing. 



The fixing of the time of the Holly glacial river as coincident with the formation of the Had- 

 ley ridge raises the question as to whether some or all of the slender moraines earlier than the 

 Hadley ridge or Portland moraine may not be present south of the Holly channel. But no 



