Spencer.] *^*^^ [March 18, 



further westward to about the same height. The country gradually rises 

 from Lake Ontario to this ridge, but on passing the summit, it descends 

 nearly 300 feet. This ridge consists of drift to a considerable depth. I 

 have several profiles across it. Yet there are no indications that the rivers, 

 such as the Nottawasaga and others flowing northward into Georgian bay, 

 formerly flowed in the opposite direction, emptying into Lake Ontario by 

 the Humber. 



The evidences are not quite clear whether the Georgian bay always 

 emptied (except when the waters were at a much higher level), by the 

 present outlet, or by that just indicated. But from the soundings I am 

 inclined to favor the present route. It may be stated that the writer is 

 assuming too frequently that the present soundings are some criterions of 

 the original depths. This assumption I hope to prove in a subsequent 

 paper, when treating of the drift-deposits, and feel confident that outside 

 of confined channels of comparatively narrow width, or certain bays, that 

 the evidence adduced, with regard to the western end of Lake Erie, holds 

 still nearer to the truth when applied to the more northern waters. If 

 Georgian bay were so filled with drift we ought not to find the deep es- 

 carpment situated so close to Indian peninsula. 



One more remark is necessary with regai'd to Georgian bay and the 

 !Nortli channel, — that is concerning the deep bays or fiords. All the con- 

 ditions for the making of fiords as noticed under Lake Superior exist here. 

 Owen sound, one of the largest of those fiords, is situated at the junction 

 of the Niagara and Hudson River Formations, with a buried channel 

 emptying into it, and now occupied by the small Sydenham river. At any 

 rate the fluviatile origin of this rivulet is unquestionable (although Mr. 

 George J. Hinde asserts that it was made by glacial action), after the study 

 that we have made in the Dundas valley. The buried channel of the 

 Sydenham river is more than half a mile wide at the town of the same 

 name. 



Some of the indentations in Manitoulin island were probably formed by 

 rivers flowing across the island, but were closed by drift in portions of their 

 course, thus producing the lakelets and bays. That most of these bays are 

 fiords is apparent, as is also proven by the numerous islands north of Mani- 

 toulin islands, the whole being a perfect counterpart of Puget sound, or of 

 the fiords of the Scandinavian peninsula. 



Owing to the much greater depth, and other obstacles of the present 

 time, it does not seem at all likely that Lake Huron ever emptied by 

 Georgian bay, excepting possibly at the close of the great floods that made 

 the whole region from Huron to Ontario one body of water, even then the 

 present topography would not favor it. 



The Outlet of Lake Ontario. The three great questions, involved in the 

 sub-aerial and fluviatile origin of our three Great Lakes, are, where were the 

 outlets of Lake Ontario, Lake Erie and Lake Huron, at sufficient depths to 

 drain their basins. As shown, the outlet of Lake Erie through the Dundas 

 valley is sufficiently deep to empty the two upper lakes. Also, the outlet 



