masses, easily recognizable as far off as they can be seen. And the
almost entire absence of trees or other vegetation helps one to trace
out the relationship of these formations over immense areas with little
or no difficulty.

In the latitude of the Bow River, near the Canadian Pacific main line,
there is a long narrow valley of these Cretaceous beds some sixty-five
miles long, called the Cascade Trough, with of course pre-Cambrian
mountains on each side. Somewhat further south there are two of these
Cretaceous valleys parallel to one another, and in some places _three_;
while just south of the fiftieth parallel of latitude, at Gould's Dome,
there are actually _five parallel ranges_ of these Palæozoic mountains,
_with four Cretaceous valleys in between_, one of these valleys, the
Crow's Nest Trough, being ninety-five miles long.

But we ought to take a nearer view of these wonderful conditions. A
convenient point of approach will be just east of Banff, Alberta, near
Kananaskis Station, where the Fairholme Mountain has been described by
R.G. McConnell of the Canadian Survey. The latter remarks with amazement
on the perfectly natural appearance of these Algonkian limestones
resting in seeming conformability on Cretaceous shales, and says that
the line of separation between them, called in the theory the "thrust
plane," resembles in all respects an ordinary stratification plane. I
quote his language:

"The angle of inclination of its plane to the horizon is _very low_, and
in consequence of this its outcrop follows a very sinuous line along the
base of the mountains, and _acts exactly like the line of contact of two
nearly horizontal formations_.