PHYSICAL AND CLIMATAL CONDITIONS. 127 



origin of the coteau must also explain the transport of 

 these boulders so far above it and beyond its limits, as 

 well as the contemporaneous distribution of boulders 

 from the Kooky mountains to the eastward. These 

 phenomena are explicable on the hypothesis of a glacial 

 sea of varying depth, but not on that of land glaciation, 

 which would also be inapplicable in a region necessarily 

 of so small precipitation of moisture and occupied by soft 

 deposits so little suited to the movement of glaciers. A 

 fortiori the same explanation applies to that great tail of 

 d^hris extending from the southern end of the Missouri 

 coteau across the continent, and which forms the great 

 '' terndnal moraine " of the continental glacialists. The 

 fact that this so-called moraine sometimes occurs where 

 there is no elevated shore immediately outside of it con- 

 stitutes no objection to this, since there may have been 

 unequal elevation. There is, nevertheless, good evidence 

 of the action of glaciers on a large scale in certain portions 

 of the glacial periods, both on the Rocky mountains and 

 on the Laurentian hills and table-lands to the east. 



3.— ICE-FRESHETS IN RIVERS. 



3. A cause of boulder-drift to which too little import- 

 ance has been attached, is what may be termed " ice 

 freshets " in the rivers of northern latitudes. Lyell has 

 summed up some facts of this kind in relation to the 

 rivers of Siberia, and Belgrand has referred to the 

 evidence in the valley of the Somme. On a small scale, 

 I have noted the effects of these ice-floods in Nova Scotia 

 and New Brunswick. They occur in early spring, when 

 sudden thaws and violent rains sometimes occur before 

 the ice in the rivers has broken up. In these circum- 

 stances, the rivers rising break up the ice on their 



