242 SKETCHES OF CUE ATI ON. 



bullrushes are boldly invading and occupying it on every 

 hand. A thousand incipient islands are breaking up its 

 continuity. Once it was fifty miles in width and a hun- 

 dred miles long. A rise of ten or twenty feet would make 

 it that again. 



But the whole series of lakes is nearly of the same level 

 from Chicago to Buffalo. The former high waters of Lake 

 St. Clair imply similar floods in the other lakes. Indeed, 

 we easily discover corroboration of this in the topography 

 of the country at Chicago, Detroit, and Toledo. These cit- 

 ies are built upon the slime of the lakes, and a slight eleva- 

 tion of the waters would bury them beneath a new deposit 

 of lacustrine mud. The artesian wells of Toledo are sup- 

 plied from some of the sandy beds of the ancient lake sed- 

 iment, which follow the general configuration of the under- 

 lying drift, and come to the surface at some higher level 

 back of the city. 



These evidences of higher waters lead us to inquire for 

 the cause. They could scarcely be occasioned by a great- 

 er volume of water, since the outlets are of sufficient ca- 

 pacity to prevent its accumulation. Nothing but an ob- 

 struction of the outlet can explain the phenomenon. This 

 obstruction must have existed at a point where the contig- 

 uous shores were sufficiently elevated to prevent a flank 

 movement of the water. It must also have existed at a 

 point beyond or to the eastward of all these obvious traces 

 of the inundation. It could not have been at Mackinac, 

 for that would not have flooded Canada West. It could 

 not have been at the foot of Lake Huron for the same rea- 

 son, and because the contiguous country is too low. It 

 could not have been at Buffalo for the last-named reason, 

 and also because the country between Buffalo and Lake 

 Ontario belongs to the submerged area. It must have 

 been at the mouth of the Niagara River. 



