80 THE JOURNAL OF GEOLOGY. 



Under these conditions, if a second sheet of ice invaded the 

 region in question after it had been long exposed, it would find 

 a surface prepared to yield large bowlders. The result would 

 be the deposition of a new sheet of drift containing bowlders 

 much larger than those which would have been proper to an ice- 

 sheet overspreading a surface but recently abandoned. If, there- 

 fore, in the upper of two layers of subglacial till, bowlders of 

 great size predominate, as compared with those of a lower homol- 

 ogous layer, they may be indicative of a great interval of time 

 between the deposition of the upper and lower beds of drift. If 

 the home of these bowlders be far north of the limit of the lesser 

 sheet of drift, the distance, as well as the duration, of the ice 

 retreat must have been great, and the reference of the two beds 

 of till to distinct ice epochs would be favored. The case might 

 be so strong as to make no other interpretation possible. Where 

 in itself inconclusive, this criterion would have corroborative 

 significance. In its application, the discrimination of subglacial 

 and superglacial till would be imperative. 



The absence of physical dissimilarity between superposed 

 layers of subglacial till would not be proof of the absence of 

 separate glacial epochs. The phenomena constituting the cri- 

 terion could hardly be expected to be of common occurrence. 

 They would never be obtrusive, and may easily have escaped 

 attention where they exist. ^ 



( 1 1 ) Varying Altitudes and Attitudes of the Land. Another 

 line of argument has to do with the altitude and attitude of the 

 land during the deposition of various members of the drift com- 

 plex. If during the deposition of one part of the drift that part 

 of the continent covered by the outer part of the ice was low, 

 the drainage from it would be sluggish. If the deposits of 

 this drainage persist to the present time, we may find in their 

 character evidence of the nature of the drainage, and therefore 

 of the attitude of the land. If at a later time of drift deposi- 

 tion the glacial drainage in the same region was more vigorous, 



'The loth criterion, in the order here named, was suggested by Mr. McGee in 

 the discussion which followed the reading of the paper at Ottawa. 



