20 W. UPHAM — ICE AND RIVER EROSION IN SAINT CROIX VALLEY 



preceded and followed by long stages of glaciation of this district, the 

 drainage from an embayment of the ice boundary, at the junction of 

 glacial currents flowing in Minnesota from the northwest and in Wis- 

 consin from the northeast, passed in a large river, the interglacial Saint 

 Croix, across the former watershed where we now have the gorges of the 

 Dalles. 



Separate preglacial streams flowing from this locality southward and 

 northwestward during many thousand years of the Tertiary era, in the 

 now continuous river course, had doubtless performed the greater part 

 of the valley erosion on each side of the old watershed, which itself, we 

 ma}"" also believe, was deeply indented here b}'' a col of the trappean 

 rocks in which the Dalles are channeled. The separate valle3''S leading 

 each way from the col, as eroded during the very long Tertiary era, 

 may have attained nearly the same size which they now have as })arts 

 of the present continuous valley, varying mainly from about a half mile 

 to one mile in width and from 75 feet to about 150 feet in de}>th below 

 the adjoining rock cliff's. 



In the Upper Dalles, at and just south of Taylors Falls, extending about 

 two-thirds of a mile, and again in the Lower Dalles, situated two miles 

 farther down the river and reaching one-third of a mile, immediately 

 above the village of Franconia, Minnesota, the rock cliffs of trap, Ke- 

 weenawan diabase, rise almost or quite perpendicularly on each side of 

 the river, inclosing it at each place by a very picturesque gorge. The 

 vertically jointed and castellated walls of the Upper Dalles form a gorge 

 from 200 feet to about 500 feet wide, which turns at a sharp angle in its 

 central part from a course nearly due south to another bearing west- 

 southwest. The course of the Lower Dalles, about 50Q feet wide, is also 

 west southwest, this direction being in each case determined by a prin- 

 cipal system of parallel and nearly vertical joint planes. 



Between these diabase gorges the valley widens to about a mile, its 

 western rock wall being an escarpment of almost horizontally bedded 

 Cambrian sandstone and shales, easily eroded, while on the east it is 

 inclosed by irregular slopes of the igneous Keweenawan rocks. Con- 

 tinuing south from the Lower Dalles, the valley, a half mile to one mile 

 wide, is inclosed by escarpments of the horizontal Cambrian sandstone 

 capped by dolomitic limestone, with overlying glacial drift. Returning 

 and going up the river from Saint Croix Falls, we find its valley there 

 inclosed chiefly by eroded drift bluffs. 



Glacial erosion in this part of the Saint Croix valley is supposed by 

 Doctor Berkey to have been an important factor in causing the river at 

 the end of the Ice age to take its present course. It seems to me, how- 

 ever, as before shown, that we may better regard the oppositely flowing 



