THE NAUTILUS. 



45 



itself. Later, another outlet was formed, known as the Nipissing 

 outlet, at a still lower level, which resulted in the closing of the 

 Trent outlet, and the establishment of a new one along the present 

 valley of the Ottawa into the St. Lawrence. When the Nipissing 

 outlet was first established, however, there had been a lowering of 

 the land toward the east, and the sea had invaded the region to a 

 considerable distance up the Ottawa Valley beyond the present city 

 of Ottawa, and into Lake Ontario. Later, with the subsequent 

 rising of the land, the Nipissing outlet flowed through the present 

 Ottawa Valley into what is now known as the St. Lawrence. It 

 seems reasonably certain that the western invasion of the Great 

 Lake region of the Atlantic fauna was through either the Trent or 

 the Nipissing outlet, and the probability is in favor of the Trent out- 

 let, because that was always entirely fresh water, and there would 

 seem to be every probability, from what we know of the inter-glacial 

 extension of the Mississippi fauna into this region, that the post- 

 glacial lakes were almost immediately invaded by the fish and with 

 them the Unionidae of the regions to the south and to the east. So 

 far as the particular question here involved is concerned, it is im- 

 material by which of these routes the invasion took place. Both of 

 them began on the west, at the Georgian Bay, and afforded a con- 

 tinuous waterway from the east to the northwest. Both of these 

 outlets were antecedent to the establishment of an outlet through the 

 Niagara River. That no invasion from the east of the Atlantic 

 fauna could have taken place by that route is clear for the reason 

 that there was always, to a greater or less degree, a falls in the 

 Niagara River, which was an absolute barrier to any migration of 

 the fish upstream from the east, and that there was no such invasion 

 from the east by that route is shown by the fact that in the case of 

 the Unio complanatus, there is no evidence to show that it ever 

 reached Lake Erie. The remarkable agreement between the pres- 

 ent range of Unio complanatus and the route of these earlier post- 

 glacial outlets is evidently more than a mere coincidence. If, then, 

 the invasion was through either the Trent or the Nipissing outlet 

 into Georgian Bay, it is easy to see how the species spread along 

 the north shore of the Georgian Bay into the St. Mary's, and from 

 thence into the eastern Lake Superior, without getting either into 

 Lake Erie, Lake St. Clair, or the lower part of Lake Huron. 



