ir tendency to break at right angles to
the beds, modified according to the power of the rock to support itself,
and the aqueous action from above or below.

I have before alluded (in p. 157) to this somewhat perilous arrangement
permitted in the secondary strata. The danger, be it observed, is not of
the fall of the _brow_ of the precipice, which never takes place on a
large scale in rocks of this kind (compare Sec. 3 of this chapter), but of
the sliding of one bed completely away from another, and the whole mass
coming down together. But even this, though it has several times
occurred in Switzerland, is not a whit more likely to happen when the
precipice is terrific than when it is insignificant. The danger results
from the imperfect adhesion of the mountain beds; not at all from the
external form of them. A cliff, which is in aspect absolutely awful, may
hardly, in the part of it that overhangs, add one thousandth part to the
gravitating power of the entire mass of the rocks above; and, for the
comfort of nervous travellers, they may be assured that they are often
in more danger under the gentle slopes of a pleasantly wooded hill, than
under the most terrific cliffs of the Eiger or Jungfrau.

[Illustration: FIG. 89.]

Sec. 40. The most interesting examples of these cliffs are usually to be
seen impendent above strong torrents, which, if forced originally to run
in a valley, such as _a_ in Fig. 89, bearing the relation there shown to
the inclination of beds on each side, will not, if the cleavage is
across the beds, cut their channel straight down, but in an inclined
direction, correspondent to the cleavage, as at _b_. If the operation be
carried far, so as to undermine one side of the ravine too seriously,
the undermined masses fall, partially choke the torrent, and give it a
new direction of force, or diminish its sawing power by breaking it
among the fallen masses, so that the cliff never becomes very high in
such an impendent form; but the trench is hewn downwards in a direction
irregularly vertical. Among the limestones on the north side of the
Valles, they being just soft enough to yield easily to the water, and
yet so hard as to maintain themselves in massy precipices, when once
hewn to the shape, there are defiles of whose depth and proportions I am
almost afraid to state what I believe to be the measurements, so much do
they differ from any which I have seen assigned by scientific men as the
limits of preci