THE ERRATIC PHENOMENA. 415 



The lower course of Michipicotin River is for several miles clammed 

 up in that way by concentric walls, across which the river has cut 

 its bed, and winding between them, has repeatedly changed its direc- 

 tion, breaking through the successive walls in different places. The 

 largest and lowest of these walls, a kind of river terrace near the 

 margin of the lake, shuts at present the factory from the immediate 

 lake shoi'e and the river, which has cut its way between the rocks to 

 the right and the walls, has left a bold bank in this dam on its left 

 shore. 



An important question now arises, after considering these facts, 

 how these successive changes in the relative level of the lake and its 

 shores have been introduced. Has the water been gradually sub- 

 siding, or has the shore been repeatedly lifted up ? Merely from 

 the general inferences of the more extensive phenomena described 

 above, respecting the relative changes between land and sea, I 

 should be inclined to admit that the land has risen, rather than to 

 suppose that the waters have gradually flowed out. But there are 

 about the lake itself sufficient proofs, which leave in my mind not 

 the slightest doubt that it is the land which has changed its level, 

 and not the lake which has subsided. 



In the first place, to suppose that the lake had once stood as high 

 as the highest terraces, it would be necessary to admit that its banks 

 were, all round its shores, sufficiently high to keep the water at that 

 highest level, or, at least, that there were, at the lower outlets, bars 

 to that height, which have been gradually removed since. But 

 neither is the main land sufficiently high, at the western extremity 

 and along the southern shores, to admit of such a supposition, nor is 

 there about the outlet of the lake, between Gros Gap and Cap Iro- 

 quois, an indication of a barrier which has been gradually removed. 

 There, as everywhere along the lake shores, the loose movable mate- 

 rials consist of the same drift, the accumulation of which, at various 

 levels, we are aiming to account for. If, therefore, we consider this 

 same drift as the barrier under whose protection the lake modeled 

 other parts of its mass, we should be compelled to admit another 

 cause to remove the barrier, a supposition for which there is not the 

 slightest indication in the geological structure of the country. But 

 ifj on the contrary, we suppose the lake to have removed the barrier, 



