CENOZOIC TIME — QUATERNARY. 947 



and produced the excavation. The depth of Lake Ontario is 738 feet, 492 

 of which are below tide level; and hence the minimum elevation that 

 would give the same slope to the water as now was 738 feet. As shown on 

 the map on page 201, this Ontario River (or the line of greatest depth) was 

 near the south shore ; and the depression had a high declivity on that side 

 which was very steep for the first 500 feet. Similar conclusions may be 

 drawn from all the Great Lakes ; for they are generally believed to have 

 been excavated by running waters during the Glacial period. The map on 

 the page referred to has marked upon it the outlines of the drainage areas of 

 the several lakes, the deep-water line, and the position of the point of maxi- 

 mum depth ; and Schermerhorn remarks that the deep-water line of each is 

 near the center of the area of drainage. The Lake Superior basin descends 

 407 feet below sea level; the Michigan, 289 feet; the Huron, 121 feet. 

 For fluvial excavation, the elevation must have been not only that which 

 would raise the basins above sea level, but to a height above the surrounding 

 land that would enable even the bottom waters to flow out of the drainage 

 basins ; and to pass, not the existing drainage barriers, but the barriers of the 

 Glacial period, when the land in the vicinity was far above its present level. 



A change of level is also proved by the reversed flow of some streams. 

 Carll and others have shown that the Pennsylvania rivers, the Alleghany and 

 Beaver, then flowed northward into Lake Erie, proving that the land dipped 

 toward the Erie basin. In the Beaver Biver channel in western Penn- 

 sylvania, now a tributary of the Ohio, the filling of drift, according to Foshay 

 and Hice (1890), is only 60 feet deep at its mouth ; but 20 miles above, it is 

 200 feet, according thus with the view that its drainage, as shown by Carll 

 for the Alleghany, had been reversed. The Tionesta and Conewango basins, 

 according to Carll, participated in reversed Erie-ward pitch. Facts on this 

 subject of reversed drainage are presented by Chamberlin in a paper of 1894, 

 along with illustrating maps. Moreover, Gilbert pointed out in 1871, that the 

 Maumee River, now emptying into the west end of Lake Erie, then flowed 

 westward, and joined the Wabash, and thus made the lake a tributary to the 

 Ohio. He found the evidence both in westward glacial scratches and 

 moraines, and in lake terraces. It is possible that a Huron River made 

 another Ohio tributary. 



Again, Lake Winnipeg, as pointed out by G. K. Warren {Rep. U. S. 

 Engineer Dept, 1867, 1874, and Am. Jour. Sc, xvi., 417, 1878), which now 

 discharges into Hudson Bay by the short Nelson River, formerly discharged 

 into the Mississippi, and, with the Saskatchewan River, was its northern 

 head waters. At the present time, the level of the lake is about 260 feet ; 

 too low for a southward flow. The divide is in Minnesota between Big Stone 

 Lake, the head waters of Minnesota River, and Lake Traverse, the head 

 waters of Red River of the North, a Winnipeg tributary. These two little 

 lakes are but a few miles apart and differ but eight feet in level. The valley 

 of Red River and that of the Minnesota were found by Warren to be con- 

 tinuous, and to be a great valley across the divides, 125 to 150 feet deep, and 



