PLEISTOCENE EROSION OF THE DALLES 10 



and through the southeastern part of Chisago county, continuing thence 

 onward in Wisconsin. 



Terraces of sand and gravel, which are found in the Saint Croix val- 

 ley 4 to 10 miles north of Taylors Falls, mostly having a height of about 

 90 feet above the river, are remnants of valley drift de})osited during 

 the Wisconsin stage of the final departure of the ice-sheet. These 

 gravel deposits, continuous as one expanse of modified drift from the 

 "jack pine barrens " of northwestern Wisconsin, bear testimony that 

 a part of the floods from the dissolving ice then passed southward along 

 the present Saint Croix, and that the erosion of the valley in the vicinity 

 of the Dalles had been mainly accomplished previous to the Wisconsin 

 stage. We are led, therefore, to the conclusion that much channeling 

 of the valley here, enlarging it along all its course from the Dalles 

 southward to the Apple river, and eroding the drift bluff, an escarpment 

 of till, which rises steeply on the west side of the valley at Taylors Falls 

 and northward to the height of 200 to 220 feet above the river, took 

 place mostly during the prolonged Buchanan interglacial stage. It was 

 a nearly similar history with that of the Minnesota river during the 

 same Buchanan time in the reexcavation of its valley, which had 

 doubtless become chiefly filled with drift during the principal Kansan 

 stage of glaciation. 



When I wrote the chapter on this district for the final report of the 

 Minnesota Geological Survey (volume 2, 1888, pages 399-425, with map 

 of Chisago, Isanti, and Anoka counties), I believed that the preglacial 

 and postglacial courses of the Saint Croix were alike ; but I now attribute 

 the establishment of this great river course and valley at the Dalles, and 

 for many miles above and below, to the capricious outlines of the re- 

 treating ice-front in Buchanan time, probably sending a considerable 

 stream across the preglacial watershed and along this course at first 

 because the ice itself was still a barrier on the lower country westward. 

 The erosion by this stream had cut down this section of the valley and 

 the two gorges of the Upper and Lower Dalles so far before that lower 

 land was uncovered from the ice that the channel so begun still con- 

 tinued as the lowest then available for the river, and the erosion appar- 

 ently extended as deep as to the present river level before the renewal 

 of ice accumulation. 



The duration of the interglacial stage attended by great decrease of 

 this part of the continental ice-sheet has been estimated by Winchell, 

 from his investigation of the drift-filled gorge of the Mississippi west of 

 Minneapolis, to have measured about 15,000 years.* Within that time, 



♦Paper before cited in the American Geologist (vol. 10), estimating the interglacial stage as 

 9,750 years ; which is corrected to about 15,000 years in the same volume, p. 302, Nov., 1892. 



