244 THE HISTORY OF THE NIAGARA RIVER. 



reaching Lake Huron. Superior, Michigan, Huron, and Georgia con- 

 Btitute a lake system by themselves, indei)endent of Erie and Ontario, 

 and the channel of the Detroit River is dry. Lake Erie and Lake Ontario, 

 both greatly reduced in size, constitute another chaiij, but their con- 

 necting link, the Niagara Eiver, is a comparatively small stream, for 

 the diversion of the upper lakes robs the river of seven-eighths of its 

 tributary area. 



Whether this hypothetic state of drainage ever existed, whether the 

 ice retreated from the Nipissing pass while still the changitig attitude 

 of the land was such as to turn the Georgian outlet in that direction, 

 are questions not yet answered. But such data as I have at present 

 incline me to the belief that for a time the upper lakes did discharge 

 across the Nipissing pass. 



Professor Spencer has decribed a channel by which Georgian Bay 

 once drained across a more southerly pass to the valley of the Trent 

 River, and thence to Lake Ontario.* He states that there is an ancient 

 shore line about Georgian Bay associated with this outlet, and that he 

 has traced this line westward and southward until it comes down to 

 the shore of Lake Huron, demonstrating that during the existence of 

 that outlet also, the Detroit River ran dry. The Trent pass is much 

 higher than the Nipissing pass, so that it appears necessary to assume 

 that during the history of the Trent outlet for the upper lakes the 

 great glacier still occupied tlie region of Lake Nipissing, preventing 

 the escape of the water in that direction. 



The map in PI. v represents the system of lakes and outlets at that 

 time. It is largely theoretic, but at the same time I believe its general 

 features consistent with our present knowledge of the facts. 



Unless I have misunderstood Professor Spencer, Lake Ontario was 

 at high stage in the lirst part of the epoch of the Trent Valley outlet, 

 and was afterwards at low stage. 1 have selected as the date of my 

 map the epoch of the high stage, with the outlet of Ontario at Rome, 

 and have indicated an ice sheet so extensive as to block the way not 

 only at Lake Nipissing but at the pass of the Thousand Islands. The 

 date of this map is earlier than the other ; it belongs to a time when 

 the northward depression of the land was greater. Lake Erie is repre. 

 sented as less in extent, for its basin in that position would hold less 

 water. Huron and Ontario would likewise be smaller were their 

 waters free to esca'pe over the lowest passes ; but the ice blocks the 

 way, and so their waters are raised to the level of higher passes. Of 

 the contemporaneous relations of the upper lakes we know nothing at 

 present. They are drawn as though communicating with Lake Huron, 

 but it is equally possible that tliey fell into some other drainage 

 system. Here again the Detroit channel was not in use, and the 

 Niagara River was outlet only for the waters of the Erie basin. 



* Proceedings Am. Assoc. Adv. ScL, 37th Meeting (Cleveland), pp. 198-199. 



