70 THE JOURNAL OF GEOLOGY. 



have receded to a position still further north. If marine or 

 lacustrine beds lying far north of the later ice limit contain 

 proof of temperate climate, the argument becomes conclusive. 



The absence of marine and lacustrine deposits between beds 

 of drift, would be no proof that interglacial epochs, did not 

 occur. Lacustrine beds could be made only where there were 

 lakes, and lakes would be the exception rather than the rule. 

 Marine beds in similar positions would rarely be known, except 

 where a definite succession of changes of level has taken place. 

 Both classes of deposits, if once formed, would be subject to 

 destruction by the over -riding ice of a later epoch, if such there 

 were. Neither would be likely to be preserved at all points 

 where formed, and both may exist at many points where their 

 existence is not known. The absence of these beds is at best 

 no more than negative evidence. 



( 5 ) Beds of Sub aerial Gravel, Safid ajid Silt. Layers of stratified 

 drift between layers of ground moraine, are of common occurrence 

 in many regions. Under ordinary conditions their existence is 

 not regarded as evidence that the underlying and overlying tills 

 are to be referred to separate ice epochs. But it is conceivable 

 that beds of stratified drift may, under the proper circumstances 

 and relations, be strong evidence of separate ice epochs. The 

 last stages of ice work in the glacial period were accompanied, in 

 many regions, by the deposition upon adjacent land surfaces, of 

 extensive bodies of gravel and sand, washed on beyond the ice 

 by waters issuing from it. Except in valleys through which 

 strong currents coursed, such deposits were apparently not car- 

 ried far beyond the edge of the ice. But as the edge of the ice 

 withdrew to the northward, sand plains may have extended 

 themselves in the same direction, by additions to their ice-ward 

 faces. It is conceivable that the process of subaerial plain 

 building at the edge of a receding phase of ice, might be 

 carried so far under favorable circumstances, as to result in the 

 construction of plains' of great extent. In this event, a subse- 

 quent ice -advance might overspread such plains in such wise 

 as to bury, without destroying them, though such a course of 



