A TYPE OF LAKE FORMA TION IN CANADA 249 



Several exposures of the drift showed evidence of two ice 

 invasions. One of the clearest of these was discovered on the 

 banks of the Bow River, two and one half miles east of Banff 

 Station, on the Canadian Pacific road. Here the river sweeps 

 against its north bank, and has laid open a section of drift more 

 than 300 feet thick. About half way up the bluff the line 

 between two different kinds of till is clearly marked. 



The lower till is of unstratified drift, consisting almost wholly 

 of pebbles and gravel, with but very little clay and rock dust. 

 Quartzite, limestone, and argillite pebbles, many of which are 

 markedly striated, make up the principal mass. 



The overlying till consists almost wholly of clay, so hard as 

 to resemble sun-dried brick, which, when struck by a stone 

 resounds like solid rock. Interspersed at considerable intervals 

 are pebbles not differing much from those of the lower till. Like 

 them they are angular and striated. The bottom of these for- 

 mations is not exposed, as the river rests on drift. However, 

 two formations were later observed on the Cascade River two 

 miles distant, which were identified as the same, and these 

 rested directly on the Cretaceous sandstones of the vicinity. 

 Thus only two tills are represented in this region. 



The sides and summits of mountains must be examined for 

 evidence of greater depth in the ice currents than those given 

 by the drift formations. Near the station of Banff, which is 

 in the Bow or South Saskatchewan Valley, about twenty-five 

 miles from the point where the river leaves the mountains, there 

 is a low mountain whose summit is exactly one thousand feet 

 above the river. This mountain is of Devonian limestone 

 throughout, and in form is a blunt ridge running transversely 

 across the valley and partially blocking it. On the top of this 

 mountain there are many Cambrian quartzite bowlders and other 

 erratics which have been transported thither. The nearest point 

 at which these quartzite bowlders are found in place is at Castle 

 Mountain, seventeen miles up the Bow Valley. The limestone 

 ledges are channeled, grooved and striated, in a direction 

 exactly across this mountain, but parallel with the valley. 



