SUMMIT SWAMPS. 45 



said to come from swamps on the divide, but their heads were not visited. 

 There was probably a time when the overflow was escaping by two or 

 three of the spillways at once, for the Cumber and Tyre channels were 

 undoubtedly active together before the Ubly channel opened, and the 

 Cumber and Argyle may also have been active together before that. In 

 the finally established outlet of this period the Tyre and Ubly channels 

 were both active as head branches of one outlet stream. 



SUMMIT SWAMPS. 



Between the Argyle valley and the Toledo moraine at Marlette and 

 Brown City is a wide stretch of flat country with many swamps and few 

 uneven features. The water divide of the thumb passes across this re- 

 gion from the north-northeast toward Brown" City. A north-and-south 

 swamp passes five miles east of Marlette at an altitude of about 810 feet. 

 It extends south many miles, passing west of Omard, east of Brown City, 

 through Valley Center and Lj r nn to Capac. Its summit level was not 

 determined, but lies a few miles southeast of Marlette. Northward it 

 extends down the course of Nettle creek west of Juhl and Elmer and 

 Snover. Its north end was not seen. There are two or three swamps 

 coming from the east into this one across the summit level of the thumb, 

 but none of them, so far as seen, show the characters of abandoned river 

 courses. They have no banks ; the bordering lands are low and slope 

 into the swamps at very low angles, nor were any bouldery or gravelly 

 floors observed in them. In short, although there are a number of sum- 

 mit swamps, no plausible indication of a course of overflow was found 

 between the Argyle valley on the north and the Imlay outlet on the 

 south. The great north-and-south swamp referred to seems to be the 

 trough between the Toledo and Detroit moraines, although the latter on 

 this part of the thumb is but poorly developed as a surface feature. The 

 general level on the divide is approximately 820 to 830 feet. In some 

 places long stretches of it are so flat that it is conceivable that the over- 

 flowing water might have passed over it in a very thin sheet many miles 

 wide, to be gathered into narrower, distinct channels only on the western 

 slope, but no such channels were found, unless, indeed, the Argyle valley 

 served in this way. Such a condition, however, could have lasted only 

 a very short time at most, for any channel bed once begun would 

 speedily eat its way back across the divide. 



The long north-and-south swamp looked very favorable at first for the 

 course of an outlet, but it probably never served as such unless as a very 

 temporary spillway after or during the fall of Leipsic beach stage. At 

 a point about half way from Elmer to Shabbona, Nettle creek and the 



