72 The American Geologist. August, wn 
subject to change, the former by recurring glaciation, and the 
latter by oscillation of the coastal lands. It is therefore obviously 
a non-sequitur to base any statements respecting the age of the 
upper Mississippi on facts and features that may be observed in 
the lower portion of the valley, and this will appl}', with proper 
limitations, to a great many of the tributaries of the Mississippi. 
Such "base-leveled" channel, with steady slope and easy cur- 
rent, could have had no water- falls such as are produced by alterna- 
tions in the rock-strata with which streams come in contact in 
drift-covered latitudes. The occurrence of a water-fall in a river 
implies a comparatively recent change in the location of its bed. 
The same is true of rocky rapids. The drifted regions abound in 
water-falls. The non-drifted are without them; but vice-versa the 
drifted regions are scantily supplied with deep river-cut gorges, 
and the non-drifted are scored by deep gorges cut by the surface 
drainage. 
Such a valley is the gorge of the Mississippi, from St. Paul to 
the Iowa state line. Its depth is not measured by the hight of 
the present bluffs, for the excavation is found to extend several 
hundred feet below the present surface of tlie river. It has re- 
cently been partly refilled. Drift deposits (gravel and sand) lie 
upon the bottom of the rock-gorge and have a thickness of over 
two hundred feet. 
As the eroded valley is immensely deeper, within Minnesota, 
than the present bluffs, so it is also wider. The Trenton lime- 
stone w4iich was its bed in Upper Silurian time and perhaps in Car- 
boniferous, now forms bluffs along each side some miles distant, 
having been wasted away more rapidl}' than the other, lower, lime- 
stone strata, through the disintegrating action of the erosible St, 
Peter sandstone immediately underlying it. (See plate v, fig. 3). 
These distant Trenton- St. Peter bluffs approach the river toward 
the north further, and at the falls of St. Anthony the Trenton 
again forms the bed of the river, being the barrier at the brink 
over which the water plunges. (See plate iv, fig. 2.) 
The study of the falls and the surrounding region has revealed 
some earlier history of the river, and has brought to light some of 
the al)andoned gorges which the river formed in interglacial and 
pre-glacial times. The oldest valley seems to have been the most 
direct one, viz: that extending from the mouth of Rice creek 
above Minneapolis, to the mouth of Trout creek, at St. Paul. 
