280 THE GLACIAL LAKE AGASSIZ. 



season neany every summer, it ceases to outflow by the Bois des Sioux, 

 and that stream becomes reduced to a series of stagnant pools. The eroded 

 bluff noted, and others of the same character lying on each side of the 

 Bois des Sioux at a distance of 3 to 4 miles apart between Lake Traverse 

 and White Rock, were finished by the outflow of the glacial River War- 

 ren, but probably their erosion was begun by a stream outflowing here 

 from the Red River Valley during the Aftonian interglacial stage between 

 the Kansan and lowan stages of ice accumulation and extension.^ 



After following the old lake shore eastward to a distance of about 4 

 miles from Lake Traverse, the steep bluff gives place, in sections 2 and 11, 

 Walls, to a gentle slope of the surface, whicli allowed the accumulation of 

 a distinct beach ridge of gravel. This is smoothly rounded, 15 to 20 rods 

 in width, bounded eastward on the side toward the ancient lake by a mod- 

 erately steep slope which descends 10 or 12 feet, the land 1 to 4 miles 

 distant northeastward within the area that was covered by the lake being 

 20 to 40 feet below this beach. On the other side this ridge is succeeded 

 by a slight depression 2 to 5 feet deep, beyond which the land soon rises 

 10 to 15 feet above the beach. The material of the beach is gravel, con- 

 taining pebbles up to 2 or 3 inches in diameter, but all the surface else- 

 where on each side is till. The crest of the Ijeach here is 1,060 to 1,062 

 feet above the sea. 



The beach next passes southeastward through sections 30 and 32, 

 Croke, having in places a maximum altitude of 1,067 feet, being piled 

 several feet above its average height. 



Between 2 and 3 miles farther southeast, near the middle of section 

 9, Tara, the beach ridge sinks to the height of 1,057 feet. Its contour and 

 material, and those of the adjoining areas, are nearly the same as at the 

 locality first described. The width of the gravel beach here is 25 or 30 

 rods; the smoothed surface of till which descends thence northward is 10 

 to 20 feet lower in its first mile; on the south the sheet of till is at first for 

 40 or 50 rods about 5 feet lower than the top of the beach, but beyond this 



lAm. Geologist, Vol. XV, p. 281, May, 1895. The nomeuclature of these subdivisions of the 

 Glacial period was proposed by Prof. T. C. Chamberlin in Chapter XLII of James Geikie's "The 

 Great Ice Age," third edition, 1894, and more fully, first naming the Aftonian stage, in the Journal of 

 Geology, Vol. Ill, pp. 270-277, April-May, 1895. 



