1^6 Transactions of the Canadian Institute, [Vol. VII. 



In most of these the mouth of the valley is more or less drift- 

 encumbered, but it can be shown in several cases at least that they are 

 o-raded with respect to some level lower than that of the Medina 

 sediments immediately in front of the escarpment. This is definitely 

 proven for those which lie northwest of Collingwood, for the Dundas 

 valley, and for some others north of Burlington. 



Hence, the systematic form of each, their direction independent of 

 the ice movement, and other features cited, render it very improbable 

 that they are due to glacial erosion. On the contrary, they may all 

 except the Dundas valley, be regarded as due to the development of 

 obsequent drainage, tributary to some master stream or streams running 

 alono- the inner lowland. Some of them are, in their lower courses, 

 occupied by till, which in some cases is, and in others probably is, that 

 of the lowest till sheet ; main- of them are graded to a level on the rock 

 floor, which must have been deeply submerged at the time of the 

 deposition of the lowest interglacial beds. In the Dundas valley some 

 stratified deposits are found overlying the till. The similiarity of form 

 and development of the valle)-s whose relations to the lowest till sheet 

 and to the interglacial beds has been proven, to those in which the 

 relations are unknown, because not worked out, renders it probable that 

 none of them are of interglacial origin. It is possible, though very 

 doubtful, that the upper reaches of some of them may have developed 

 during interglacial time. 



Islands and Outliers. — In Lake Ontario towards the eastern end, and 

 extending as far west as Presqu' Isle, are a great many large and small 

 limestone islands and shoals, all lying north of the line between Stony 

 Point and Point Peter. Gull Island, four miles west of Cobourg, is also 

 a lime.stone island. In the northern part of Lake Huron, between Cape 

 Hurd and Grand Manitoulin Island, are a number of small rock islands. 

 Some of these are of rock fragments at water level, the bed rock not 

 being visible, but the large majority are composed of bed rock in situ. 

 The Manitoulin Islands are rock islands. In Georgian Bay many of 

 the islands are of limestone rock — attention will be specially called to 

 those along the Bruce peninsula (Map III). In Lake Simcoe, along the 

 east side, there are a few islands with limestone bases. Many of these 

 islands are unsubmerged portions of the higher irregularities of the 

 series of escarpments. Some of them lie in front of the main escarp- 

 ments, as indicated by soundings around them. 



On the Central Ontario lowland in Halton county, just west of 

 Milton, is an outlier, capped by Niagara limestone, severed completely 



