POSTGLACIAL DEVELOPMENT OF CONNECTING KIVEES OF GREAT LAKES; 489 



The distributaries thus far described have their headward openings at a relatively high 

 level and must therefore have been among the first to be abandoned and were occupied for 

 relatively short times. 



The next distributary opens southwest about a mile north of Trenton. At its head the 

 floor is now about 4 feet above the level of the river. It is larger than the distributaries just 

 described, having a width of 400 to 500 feet on its floor. After passing the railroads it curves 

 gradually south, passing about half a mile west of Trenton, and after running about due south 

 through the east hah of sec. 24 turns gradually southeast through sec. 25, where, within half a 

 mile of Detroit River, it splits up into several smaller channels, some of which run southeast to 

 the river, and some continue south, the main one to join the previously described channel about 

 a mile west of Gibraltar. From this last two or three narrow, swampy channels turn southeast 

 to Detroit River. This channel heads north of Trenton and may be called the Sibley channel, 

 for the Sibley limestone quarries are near its headward opening. 



In southwest Trenton another slight depression, evidently cut by a distributary stream, 

 begins near the railroad station and extends for half a mile southeast to the channel behind 

 Slocum Island. 



CREASES. 



Besides the large distributary channels in the vicinity of Trenton, a number of the peculiar 

 crease type appear. (See p. 289.) The best occur between the larger channels near Trenton, 

 particularly between the Sibley channel and Brownstown Creek southwest of Trenton. They may 

 be seen from the train on any of the railroads running southwest from Trenton. These creases 

 show the same peculiarities as those of St. Clair River (pp. 474-475), having narrow bottoms 

 with no flat floors. They run in parallel lines 30 to 40 rods apart, curving gradually from south to 

 southeast. The ground between them hes in long billow-like, smooth ridges with crests gener- 

 ally 10 to 12 feet above the troughs. Their surface is rather bowldery, suggesting moraines, and 

 Sherzer has interpreted them as such. 1 But they are closely related to the general slope of the 

 ground and change their courses with its changes in slope. Besides, they are related to the 

 barrier which obstructed the river at this locality in precisely the same way as the creases near 

 St. Clair; and the latter are certainly not moraines, but are transverse ridges left between early, 

 immature fines of scour made by water currents flowing down the front slope of the main mo- 

 raine of the Port Huron morainic system at the earliest stage of river flow. 



LATE DISTRIBUTARIES. 



At the south end of Trenton a remarkable channel a mile long separates Slocum Island from 

 the mainland. It is about 500 feet wide and is floored by a marsh with aquatic plants. Its 

 bottom lies so close to the level of the river that some water formerly passed through in high 

 stages but not in low ones. The channel, which is now obstructed by a railway embankment, was 

 evidently actively occupied for a very long time. Farther down the river a number of islands, 

 evidently fragments of land left between a network of interlacing distributaries, are now at or 

 slightly below lake level. These fragmentary channels and others that are clearly traceable 

 below the present level of Lake Erie were made when the lake stood at least 12 or 15 feet lower 

 than now. (See p. 462.) Many of these later distributaries have been submerged by the recent 

 rise of the lake waters and the fragments of land between them appear now as islands. Humbug 

 Island is one of these; Gibraltar and Snake Island, just west of Gibraltar, are others. Smaller 

 ones are Calf, Horse, Big and Little Cherry, Oak, Sturgeon, and Peabody islands. These are 

 separated from the mainland and from one another mainly by very shallow channels, 300 to 800 

 feet wide and in few places more than 3 to 4 feet deep. The water passing through them is not 

 now effectively eroding. Indeed, many of these channels, especially those that are choked 

 with reeds, are probably being slowly filled. 



1 Sherzer, W. H., Geological report on Monroe County, Mich.: Michigan Geol. Survey, vol. 7, pt. 1, 1900, pp. 134-135. 



