POSTGLACIAL DEVELOPMENT OP CONNECTING EIVEKS OF GKEAT LAKES. 475 



and contains gravel and sand bars in southwest sec. 24 and eastern sec. 26. At about mid 

 length, in sec. 13, it grows somewhat indefinite. Little narrow, creaselike channels branch off 

 on its west side, and it splits into a number of such channels toward its south end in sees. 23 

 and 26. In south sec. 1 two or three small channels start off to the west-southwest, but in south 

 sec. 11 they fade out on a sandy plain. 



Another better-developed channel lies east of those described and nearer St. Clair River. 

 It has two headward branches that start close together, the western one in southeast sec. 18 (east 

 tier) and the other in northeast sec. 19 (east tier). It starts at a lower level, its two heads 

 opening at an altitude of about 620 feet (aneroid), and is more deeply eroded than the others. 

 The eastern branch has a flat floor, swampy in places, and is 10 to 15 feet deep a mile south of 

 its head. The floors are 20 to 40 rods wide above the point of union, which is in southwest 

 sec. 30 (east tier). Throughout its whole length of 4 miles this channel is not over one-third 

 mile from the bank of either the present St. Clair River or that of the old river bed in the village. 

 It passes just west of the village and is the most striking old distributary channel of the localit} 7 . 



It seems certain that the water flowed longer in this channel than in the others. Its 

 altitude at its head is fully as high as the top of the higher ridge of the Lundy beach in this 

 vicinity, so that it seems hardly probable that it marks the Lundy level, although it lies so near it. 

 It seems rather to have been occupied until the lake level fell almost to the Lundy stage and then 

 to have been left dry. It was on a slightly lower part of the moraine and was probably a little 

 •deeper at the start than the others. But although it was cut deeper the stream hi it was rela- 

 tively small. 



CREASES. 



Several shallow creaselike troughs start in west sec. 18 (east tier) and run southwest, curving 

 south. Such creases are not known to occur except in association with the larger distributaries 

 where St. Clair and Detroit rivers first broke over morainic barriers. They lie closely parallel 

 to each other and 40 to 80 rods apart without branching or uniting* The intervening ridges are 

 sightly convex in cross section and the creases are narrow at their bottom with no perceptible 

 flat floor. Toward their heads some of them fade out on the nearly flat surface, but for most of 

 their length ihej are 5 feet or more in depth. Three of them are more than 2 miles long. The S. J 

 sec. 24 and all of sec. 25 are sprinkled with thin patches of gravel and sand, apparently deposited 

 by water coming from the north over the creased ground. Two other creases, rather more pro- 

 nounced, start in west sec. 13 and cross sees. 14 and 23 in the broader part of the till ridge between 

 the largest two channels. Another crease crosses sees. 19 (east tier) and 30 (east tier). 



It seems certain that the creases and the ridges between them are not original forms, but 

 were produced by the first flow of the river over the moraine, for they are transverse to its crest. 

 They appear to be the earliest recognizable product of erosion produced by the flow and to have 

 been made when the water was flowing over as a thin, even sheet, mainly before any of the other 

 channels had been eroded. Probably at first the whole slope now marked by the channels was 

 creased in this maimer. If the creases were made as supposed, by a thin water sheet, the level of 

 which at their heads was gradually falling, then the quantity of water passing through them 

 was diminishing while they were making and hi a very short time ceased entirely. The whole 

 process was too brief to permit the development of meanders, so the creases are straight or only 

 slightly curved and have no flat bottoms. The direction of the creases records the direction of 

 flow of the first water sheet as it passed over. On the lower parts of the moraine crest, where 

 the water flowed longer, the creases were apparently cut away by the widening of the deeper 

 channels. More and larger creases of this kind occur near Trenton. (See p. 489.) 



The channels that have been described are all that remain to-day as abandoned distributaries, 

 but at least one other distributary was deepened and became at last the channel of the master 

 stream. This was presumably the largest one at first and as it enlarged and deepened its bed, 

 the level of the waters north of the moraine was lowered and all the other distributaries were 

 abandoned. Thus the largest of the original distributaries became the course of the undivided 

 river. In its enlarged form it may include several originally independent channels. 



