over surfaces of varied shape.

4. Lines of Rest. Those which are assumed by debris when in a state of
comparative permanence and stability.

1. Lines of Fall.

1. Lines of Fall. Produced by falling bodies upon hill-surfaces.

However little the reader may be acquainted with hills, I believe that,
almost instinctively, he will perceive that the form supposed to belong
to a wooded promontory at _a_, Fig. 100, is an impossible one; and that
the form at _b_ is not only a possible but probable one. The lines are
equally formal in both. But in _a_, the curve is a portion of a circle,
meeting a level line: in _b_ it is an infinite line, getting less and
less steep as it ascends.

[Illustration: FIG. 100.]

Whenever a mass of mountain is worn gradually away by forces descending
from its top, it _necessarily_ assumes, more or less perfectly,
according to the time for which it has been exposed, and the tenderness
of its substance, such contours as those at _b_, for the simple reason
that every stream and every falling grain of sand gains in velocity and
erosive power as it descends. Hence, cutting away the ground gradually
faster and faster, they produce the most rapid curvature (provided the
rock be hard enough) towards the bottom of the hill.[90]

Sec. 22. But farther: in _b_ it will be noticed that the lines always get
steeper as they fall more and more to the right; and I should think the
reader must feel that they look more natural, so drawn, than, as at _a_,
in unvarying curves.

[Illustration: FIG. 101.]

This is no less easily accounted for. The simplest typical form under
which a hill can occur is that of a cone. Let A C B, Fig. 101, have been
its original contour. Then the aqueous forces will cut away the shaded
portions, reducing it to the outline _d_ C _e_. Farther, in doing so,
the water will certainly have formed for itself gullies or channels from
top to bottom. These, supposing them at equal distances round the cone,
will appear, in perspective, in the lines _g h i_. It does not, of
course, matter whether we consider the lines in this figure to represent
the bottom of the ravines, or the ridges between, both being formed on
similar curves; but the rounded lines in Fig. 100 would be those of
forests seen on the edges of each detached ridge.

Sec. 23. Now although a mountain is rarely perfectly conical, and never
divided by ravines at exactly equal distances, the law which is seen in
entire simpli