114 K. H. WINCHEJ.L GLACIAL LAKES OF MINJllLSOTA 



at the time when this lake had its initiation and another when it had 

 its greatest extent. No attempt is made here to correlate this limit with 

 any of the moraines, although it is to be supposed that the main changes 

 and the principal epochs of duration of this and all the other glacial 

 lakes were determined very largely by the main steps of retreat and the 

 stationar}'^ epochs of the ice-border. 



2. Lake Undine 



The Undine region was named by Nicollet.''^ It embraces the valley 

 of the Mankato river. The later glacial lake to which the name is ap- 

 plied was the descendant of lake Minnesota, and covered part of the 

 same area, but extended from 35 to 40 miles farther north. The outlet 

 of lake Undine was through the Cannon valley in Le Sueur and Rice 

 counties at an elevation of about 1,070 feet, or 80 feet lower than the 

 level of lake Minnesota. It was very near the boundar}^ line between 

 Rice and Le Sueur counties that the waters of this lake began the de- 

 scent of the Cannon valley eastward to the Mississippi. At the present 

 time the surface of the Cannon river at that })oint is 994 feet above the 

 sea. The river lies in a valley which is enclosed b}'' gravel terraces which 

 afibrd an indication of the original height of the Cannon at that point 

 and of the erosion which the gravel has suffered since its first distribu- 

 tion. The highest gravel terrace at the commencement of the outlet of 

 lake Undine requires that the surface of the water at first was at least 

 1,070 feet, but as this terrace is rather broken and sometimes blends 

 with the till surfaces, it is ver\' evident that this elevation was not long 

 maintained. When the outlet became less obstructed by the glaciers 

 the water of lake Undine acquired a more settled stage, and a second 

 terrace, which is much more marked and persistent than the higher 

 terrace, affords an indication of the a])[)roximate level of the water dur- 

 ing a much longer period, namely, about 1,020 feet a])ove tide. It is 

 apparent, therefore, that between the date of first acquiring the outlet 

 by way of the Cannon valley and that of a settled stage of outflow the 

 level of lake Undine was lowered by the removal of obstruction by ice 

 and by erosion about 55 feet. 



No attempt has been made to trace out the beach lines of lake Undine, 

 but that they could be identified if search were instituted for them, there 

 is little doubt. They would be found running southwestwardl}^ from 

 the outlet in eastern Le Sueur county into the eastern borders of Blue 

 Earth county, westward across that county, where delta deposits would 

 probably be found at the points where the northward-flowing streams 



* Final report of tlie iMiunesota Geological Survey, vol. 1, p. 71. 



