46 GEOLOGY OF OHIO. 



Second. That this agent must have been either water or ice. 



Third. That it was ice, because rivers never excavate such basins. 

 In a plateau country, as this was, rivers could only produce narrow 

 gorges, like the canons of the Colorado, or shallower and broader valleys, 

 widening towards their mouths, and with bottoms forming more or less 

 regular slopes. 



Fourth. That the ice vf as in the form of glaciers, and not icebergs, be- 

 cause the lake basins, wherever they can be examined, are found to bear 

 conspicuous marks of glacial action; the furrows and scratches having 

 the bearing of the long diameter of each, and flint nodules, with ridges 

 in their lea, and other signs, proving conclusively that the motion in 

 Lake Erie was from Buffalo toward Toledo. 



Since the publication of our second volume, the subject of the origin 

 of the great lakes has been discussed in a paper read by Prof. E. W. 

 Claypole before the Natural History Society at Cincinnati, and published 

 in the Canadian Naturalist, of April 6th, 1877. The theory advanced 

 by the author of this paper is, that the chain of great lakes are only 

 portions of the valley of the pre-glacial river, to which reference has 

 been made, blocked up in the ice period by beds of Drift. 



The considerations which oppose this theory are so apparent and for- 

 midable, that it never could have been proposed or accepted by any one 

 who had carefully studied the problem. In addition to those already 

 suggested, they are — 



First. That the lakes occupy a series of boat-shaped rock basins, which 

 have almost nothing in common with river valleys. The notion that 

 the valley of a river could be beaded in this way by the broad excava- 

 tion of such portions as lay in soft. rock, and the format-ion of canons 

 through hard strata, has no warrant in any facts yet observed on the 

 earth's surface. 



Second. The great and unequal depth of the lake basins renders it 

 impossible that they could have been excavated by a continuous flow- 

 ing stream. Lake Michigan is nine hundred feet deep to the silt which 

 covers its bottom ; it is excavated in rocks that are not softer than those 

 of the adjacent country ; is surrounded by a rocky rim, from which, it is 

 true, a narrow, buried channel leads, but that has, so far as known, no 

 greater depth than two hundred feet — the depth of the pre-glacial river 

 which drained this region before the formation of the lake. 



Lake Huron is eight hundred feet in depth, while the buried channel, 

 which connects it with Lake Erie, is not more than two hundred feet 

 deep. 

 Lake Erie is generally very shallow, and while its bed is no doubt 



