about a shadow is, that it
is always of some form which nobody would have thought of; and this
visible principle Turner always seizes, sometimes wrongly in
calculated fact, but always so rightly as to give more the look of a
real shadow than any one else.

CHAPTER XVI

RESULTING FORMS:--THIRDLY, PRECIPICES.

Sec. 1. The reader was, perhaps, surprised by the smallness of the number
to which our foregoing analysis reduced Alpine summits bearing an
ascertainedly peaked or pyramidal form. He might not be less so if I
were to number the very few occasions on which I have seen a true
precipice of any considerable height. I mean by a true precipice, one by
which a plumb-line will swing clear, or without touching the face of it,
if suspended from a point a foot or two beyond the brow. Not only are
perfect precipices of this kind very rare, but even imperfect
precipices, which often produce upon the eye as majestic an impression
as if they were vertical, are nearly always curiously low in proportion
to the general mass of the hills to which they belong. They are for the
most part small steps or rents in large surfaces of mountain, and
mingled by Nature among her softer forms, as cautiously and sparingly as
the utmost exertion of his voice is, by a great speaker, with his tones
of gentleness.

[Illustration: FIG. 73.]

Sec. 2. Precipices, in the large plurality of cases, consist of the edge of
a bed of rock, sharply fractured, in the manner already explained in
Chap. XII., and are represented, in their connection with aiguilles and
crests, by _c_, in Fig. 42, p. 195. When the bed of rock slopes
backwards from the edge, as _a_, Fig. 73, a condition of precipice is
obtained more or less peaked, very safe, and very grand.[80] When the
beds are horizontal, _b_, the precipice is steeper, more dangerous, but
much less impressive. When the beds slope towards the precipice, the
front of it overhangs, and the noblest effect is obtained which is
possible in mountain forms of this kind.

Sec. 3. Singularly enough, the type _b_ is in actual nature nearly always
the most dangerous of the three, and _c_ the safest, for horizontal beds
are usually of the softest rocks, and their cliffs are caused by some
violent agency in constant operation, as chalk cliffs by the wearing
power of the sea, so that such rocks are continually falling, in one
place or another. The form _a_ may also be assumed by very soft rocks.
Bu