OF THE FALLS OF ST. ANTHONY. 891 



Mississippi is running in an ancient channel, and that for some 

 reason the course of that great river has been changed, the narrow 

 gorge that extends between Bassett's Creek and Fort Snelling being 

 of course the new cut. 



On tracing out the range of the rock-bluff on the west side of the 

 Mississippi above the falls, hidden as that bluff is by loam and drift, 

 it is found to fall rapidly away from the river near the railroad 

 bridge, turning south-westwards across the city, ascending the 

 south side of Bassett's Creek, which joins the river some distance 

 further up, and finally passing out of sight in a south-westerly 

 direction under a thick accumulation of drift. 



Going now across Bassett's Creek, and tracking the outcrop of 

 the limestone, we pass over a wide valley filled with alluvium or 

 brick-clay — much too large a valley to have been formed by the 

 sluggish creek that now flows through it. We find that the lime- 

 stone which, along the river, has a trend a little west of south, on 

 reaching the valley of the creek swings more westwardly, parallel 

 with the outcrop of the rock on the south side of the creek, and 

 thus encloses a valley, even a gorge, cut in the limestone and sand- 

 stone, much wider than the gorge now being cut by the recession 

 of the Falls, but in width corresponding with that between the rock- 

 bluffs above the mouth of Bassett's Creek, and comparable to that 

 below Fort Snelling. 



Here, then, we have an old drift-filled valley, evidently formed 

 at some more remote period than the present, which once held the 

 Mississippi as it ran between rock-bound bluffs towards the Minne- 

 sota and reached that great valley at some point between Fort 

 Snelling and Shakopee. Bassett's Creek, in making its way to the 

 Mississippi, falls into the depression caused by the old valley in 

 question, and follows it till it reaches the present river-channel. 

 This ancient drift-filled valley is over one hundred feet deep. This 

 has been ascertained by the borings made for deep walls, and the 

 materials that fill it up are found to be hardpan and fine alluvial 

 clay, from below which rises artesian water. 



Such ancient buried river-channels are not uncommon. A num- 

 ber have been described in various parts of the United States *. 

 It is not common, however, that circumstances have so combined 

 as to produce, by the change of course of a river and the burial of 

 its old valley, a retreating waterfall, which, by its uniform rate of 

 recession, fixes the date of such change. Niagara river has thus 

 been changed ; but its rate of recession has not been uniform, owing 

 to changes in the nature of the rock undergoing the process of 

 erosion, and to a dip in all the formations towards the south, which, 

 of course, gradually diminishes the height of the fall. There seems 

 also to be no recognized datum-point by which to establish a rate 

 of recession. That the Falls of Niagara have receded from Queen- 

 ston Heights has been sufficiently demonstrated by Lyell ; but the 

 rate assumed, based on observations of no very accurate kind, and 



* Compare ' The Geology of Ohio,' vol. ii. p. 12 et seq. 



