GREAT LAKES BASINS— SHEPARD 271 



higher than the bottoms of the Great Lakes." ^ In order to assume 

 some gradient toward the Mississippi, which was probably the pre- 

 glacial outlet of the Great Lakes drainage basin, the bottoms of the 

 preglacial valleys must have been almost as high as the present sur- 

 face of the Great Lakes. Furthermore, the area in question, being so 

 centrally located, must have been close to the drainage divide. It is 

 hard to conceive of river valleys in such a location with floors as 

 wide as those of the Great Lakes basins (fig. 2). Nor do the sides 

 of the basins have the typical sinuous outlines of river valleys. Under 

 the circumstances it seems unwise to attribute the present deep 

 basins of the Lakes to preglacial river erosion although the proba- 

 bility that small river-valley depressions existed along the lines of the 

 Great Lakes must be recognized. As a corollary, glacial deposition 

 must be relegated to the unimportant role of helping to limit the 

 size of basins formed in other ways. 



Figure 2. — Cross-sections of the Great Lakes. Note the relatively steep sides and broad base. 

 SIMILAR BASINS IN GLACIATED REGIONS IN GENERAL 



A second test may be made of the relative merits of the hypotheses 

 of glacial erosion and diastrophism. If the basins were due to glacia- 

 tion, similar basins might be found widespread over the glaciated 

 territories of the world; whereas, if they are diastrophic, such a wide 

 distribution in glaciated regions would be most unlikely. In consider- 

 ing this criterion three things must be borne in mind — first, that some 

 of the basins do not appear as lakes because they are completely sub- 

 merged by the ocean; second, that the basins in areas glaciated in 

 early epochs would have been filled to a considerable extent; and, 

 third, that the greatest gouging action of the ice would be expected 

 around or close to the margins of ice sheets, so that the absence of 

 basins in the interior would not be significant. With these points in 

 mind we may proceed to an examination of the principal glaciated 

 territories. 



A map of North America (fig. 1 ) shows a series of large lakes occupy- 

 ing basins in the soft rocks on the south and west borders of the 

 Canadian shield, and bathymetric maps show that a lowering of sea 

 level would leave similar lake basins in the gulfs and ba3^s southeast 



« Op. cit. (1st ed.), p. 43. 



