154 /• BURR TYRRELL 



epoch a line of division in the till of western Manitoba, along 

 which stratified deposits may occasionally be seen, but which 

 is strongly marked in several places by beds of bowlders that 

 have been sunk in the surface of the earlier till, and being 

 there held firmly in place have been smoothed and striated by 

 the later glacier which passed over them. How far the Kee- 

 watin glacier retreated during this second epoch of deglaciation 

 is not known, but it is not improbable that it withdrew far north 

 of the present northern boundary of Manitoba. 



After this interglacial epoch the Keewatin glacier again 

 began to' increase, though its center of dispersion had gradually 

 shifted southeastward until it now rested over the country 

 between Doobaunt and Kazan rivers. From this center it flowed 

 outwards in all directions, and its striae may be seen on most of 

 the rocky knolls throughout that whole northern country, run- 

 ning southward towards Lake Winnipeg, westward towards the 

 Mackenzie River, northward towards the Arctic ocean, and east- 

 ward towards Hudson Bay. Everywhere the smooth-faced hills 

 give evidence of its presence, and even in the absence of striae, 

 the evenly-rounded surfaces facing the glacial center, and the 

 broken, craggy hills looking in the opposite direction, furnish 

 convincing evidence of the direction of flow of the ice. As the 

 glacier advanced southward it came in contact with the high 

 escarpment of Cretaceous shales forming the Porcupine and 

 Duck mountains in Manitoba, and part of it was diverted to the 

 east of south along the great valley of Lake Winnipeg. This lobe 

 appears to have extended southward into Minnesota, Dakota, and 

 Iowa, and to have deposited a till which probably corresponds 

 to what Professor Chamberlin has called the East Iowan Forma- 

 tion. The Palaeozoic limestones of western Manitoba have been 

 beautifully scored by the markings of this glacier, and its grooves 

 and striae were detected in many places around the shores of Lake 

 Winnipeg. East of the shores of Lake Winnipeg the exposed 

 surfaces of the Archaean rocks were carefully searched for this 

 set of markings, but none could be detected. It therefore 

 seems probable that the eastern edge of this lobe or portion of 



