Spencer. J ' *^^ [March 18, 



de Noc, and also my theoretical divergence slightly from this view. 

 That an outlet did exist somewhere in the region between Lakes Superior 

 and Michigan, seems certain, of course, we do not know to what depth, 

 but from the evidence brought forward from the soundings in the northern 

 end of Michigan lake to at least half the depth of our greatest lake, and 

 probably to the whole depth. 



There appears also to have been an outlet from Lake Michigan, near 

 Chicago, emptying into the Mississippi river. These two lakes both lie 

 outside of the study of the present paper, but before passing to the other 

 lakes, let us note certain facts which present themselves to the view of the 

 writer. Without taking up the bibliography of the subject, it may be 

 stated that some have regarded the origin of the lakes as being due, to a 

 greater or less extent, to the geological valley, formed at an earlier date. 

 Of our five great lakes, certainly not more than one can possibly come into 

 this category — that is Lake Superior. Not even the northern side of the 

 north channel of Lake Huron, bordered by crystalline rocks, could be in- 

 cluded. The writer even doubts that Lake Superior is essentially anything 

 more than a valley of erosion, and if it be, it remains to be proven. Yet 

 its position was probably determined to a greater or less extent by the 

 orography of the region, and its excavation appears to have been principally 

 by erosion. Although the lake is in the region of very ancient continental 

 oscillations, there are no evidences of different elevations, and subsidences 

 in different portions of the lake basins. Five years ago, I pointed out that 

 Superior was eminently a region liable to atmospheric erosion, as the great 

 volcanic seat situated about the Keeweenaw Point, in Presilurian times, 

 would tend to weaken the strata. All who have made a study of the 

 three miles in thickness of the copper bearing rocks, pitched at various 

 angles, whether on the northern or southern shores, and crossed by numer- 

 ous faults, know that they are particularly liable to irregular atmospheric 

 decay. As an example, we have remaining an excellent case in that in 

 which Portage lake lies, now a valley transversely situated in the Keeweenaw 

 hills, and excavated to a depth of six or eight hundred feet. In the nar- 

 row channel between Houghton and Hancock the waters are 60 feet deep 

 and underlaid by a considerable depth of silt. 



According to both Professors Dana and Whitney's explanation, fiords 

 are valleys of erosion, when the land was at a higher level ; but as a neces- 

 sary condition, high hills or mountains should be near a shore, so as to 

 give pitch to the descending waters ; from this definition we can fairly con- 

 sider most of the bays of Lake Superior, Lake Huron, and those about the 

 Niagara escarpment in the region of the Green bay extension of Lake 

 Michigan, as belonging to this category. 



Lake Michigan is nearly deep enough to remove all the waters of Lake 

 Superior. Such deep places as the 171-fathom hole north-east of Duluth could 

 easily be produced by a cataract, the same as the water in the Niagara 

 river is so much deeper than its rock-bound outlet into Lake Ontario. 

 Moreover, the peculiar form of the south-western extension of Lake Supe- 



