THE ANCIENT MISSISSIPPI AND ITS TRIBUTARIES. 619 



dress. The principal discoveries were made only a few years since, by General 

 G. K. Warren of the Corps of Engineers, U. S. A. At Ft. Snelling, a short dis- 

 tance above St. Paul, the modern Minnesota River empties into the Mississippi, 

 but the ancient condition was the converse. At Ft. SnelHng, the valleys form 

 one continuous nearly straight course, about a mile wide, bounded by bluffs 150 

 feet high. The valley of the Minnesota is large, but the modern river is small. 

 The uppermost valley of the Mississippi enters this common valley at nearly right 

 angles, and is only a quarter of a mile wide and is completely filled by the river. 

 Though this body of water is now the more important, yet in former days it was 

 relatively a small tributary. 



The character of the Minnesota Valley is similar to that of the Mississippi 

 below Ft. Snelling, in being bounded by high bluffs and having a width of one or 

 two miles, or more, all the way to the height of land, between Big Stone Lake 

 and Traverse Lake, the former of which drains to the south, from an elevation 

 of 992 feet above the sea, and the latter only half a dozen miles distant (and 

 eight feet higher) empties, by the Red River of the north, into Lake Winnipeg. 

 During freshets, the swamps between these two lakes discharge waters both ways. 

 The valley of the Red River is really the bed of an immense dried-up lake. The 

 lacustrine character of the valley was recognized by early explorers, but all honor 

 to the name of General Warren, who, in observing that the ancient enormous 

 Lake Winnipeg formerly sent its waters southward to the Mexican Gulf, made 

 the most important discovery in fiuviatile geology, — a discovery which will cause 

 his name to be honored in the scientific world long after his professional succes- 

 ses have been forgotten. 



General Warren considered that the valley of Lake Winnipeg only belonged 

 to the Mississippi since the " Ice Age," and explained the changes of drainage 

 of the great north by the theory of the local elevation of the land. Facts which 

 settle this question have recently been collected in Minnesota State by Mr. Up- 

 hara, although differently explained by that geologist. However, he did not go 

 far enough back in time, for doubtless the Winnipeg Valley discharged south- 

 ward before the last days of the "Ice Age," and the great changes in the river 

 courses were not entirely produced by local elevation, but also by the filling 

 of the old water channels with drift deposits and sediments. Throughout the 

 bottom of the Red River Valley a large number of wells have been sunk to great 

 depths, and these show the absence of hard rock to levels below that of Lake 

 Winnipeg ; but some portions of the Minnesota River flow over hard rock at 

 levels somewhat higher. Whether the presence of these somewhat higher rocks 

 is due entirely to the local elevation, which we know took place, or to the change 

 in the course of the old river, remains to be seen. 



Mr. Upham has also shown that there is a valley connecting the Minnesota 

 River, at Great Bend at Mankato, with the head waters of the Des Moines Riv- 

 er, as I predicted to General Warren a few months before his death. At the 

 time when Lake Winnipeg was swollen to its greatest size, extending southward 



