The walls of a great canyon, and usually a small one, are terraced;
seldom are they wholly vertical for their entire height, though
occasionally they may approach this condition on one side or the
other, and more rarely on both sides at once, depending on the
geological formations of the locality. Owing to the immense height of
the walls of such canyons as those on the Colorado, the cliffs
frequently appear perpendicular when they are far from it, just as a
mountain peak often seems to tower over one's head when in reality it
may be a considerable distance off. In the nature of the formation
and development of canyons, they could not long retain continuous
vertical walls. What Powell calls the "recession of cliffs" comes
into play. The erosive and corrasive power of water being the chief
land sculptors, it is evident that there will be a continual wearing
down of the faces of the bounding cliffs. The softer beds will be cut
away faster than the harder, and where these underlie the harder the
latter will be undermined and fall. Every canyon is always widening
at its top and sides, through the action of rain, frost, and wind, as
well as deepening through the action of its flowing stream. EROSION
is this power which carves away the cliffs, and CORRASION the one
which saws at the bottom, the latter term, in geological
nomenclature, meaning the cutting power of running water.* This
cutting power varies according to the declivity and the amount of
sediment carried in suspension. It is plain that a stream having
great declivity will be able to carry more sediment than one having
little, and in a barren country would always be highly charged with
sand, which would cut and scour the bed of the channel like a
grindstone. As Dutton says, a river cuts, however, only its own
width, the rest of a canyon being the "work of the forces of erosion,
the wind, frost, and rain. That is why we have canyons. The powers of