266 THE JOURNAL OF GEOLOGY. 



It is probably safe to say that in selected instances at least 90 

 per cent, of the material was derived from the Paleozoic series 

 and more than half of this from the vicinity. This is, however, 

 too large an estimate for the average, but the local constituents 

 were never seen to be other than pronounced if not pre- 

 dominant. Material of local derivation also enters into the 

 constitution of the sand and clay as well as the coarser material 

 showing that the hillocks are made up not only of the glacially 

 ground rock -fragments, but of the glacial grindings. The 

 whole aspect of the material of these kames is so strikingly in 

 contrast with that of the superficial bowlder belt and points so 

 definitely to their derivation from the common sheet of sub- 

 glacial till as to seem to put beyond doubt the view that they 

 are quite strictly basal in formation. 



Osars of the typical variety have comparatively few repre- 

 sentatives in the plain tracts of the interior, but several well 

 characterized instances occur. It is notable that, in most of 

 these instances, as pointed out by Mr. Leverett, who has perhaps 

 carefully studied a larger number of them than anyone else, 

 they often lie in river -like channels cut into the till sheet of the 

 region. There are perhaps a dozen of these that have been 

 studied, varying in length from a few miles to about fifty miles. 

 These channels have the characteristics of river troughs, and 

 usually stand so related to the margin of the ice as to seem to 

 indicate that they were lines of subglacial drainage during the 

 same glacial stage as that in which the osars were formed. 

 These channels are so related to the surface slopes that they 

 could not have been formed by free open-air streams. The 

 restraining aid of ice seems necessary. While no demonstra- 

 tion of the history of their formation can be claimed, the most 

 plausible explanation appears to be that the river -like channels 

 were cut by subglacial streams at a time when the urgency of 

 the ice was such as to compel basal cutting, and that, sub- 

 sequently, when the pressure of the ice was less insistent, and 

 its motion feebler, the draining stream was permitted to fix its 

 channel in a tunnel cut in the under surface of the ice, which 



