l among the mountains; if not, let him merely draw for
himself, carefully, the outlines of any low hills accessible to him,
where they are tolerably steep, or of the woods which grow on them. The
steeper shore of the Thames at Maidenhead, or any of the downs at
Brighton or Dover, or, even nearer, about Croydon (as Addington Hills),
are easily accessible to a Londoner; and he will soon find not only how
constant, but how graceful the curvature is. Graceful curvature is
distinguished from ungraceful by two characters: first, its moderation,
that is to say, its close approach to straightness in some parts of its
course;[249] and, secondly, by its variation, that is to say, its never
remaining equal in degree at different parts of its course.

This variation is itself twofold in all good curves.

[Illustration: FIG. 36.]

A. There is, first, a steady change through the whole line from less to
more curvature, or more to less, so that _no_ part of the line is a
segment of a circle, or can be drawn by compasses in any way whatever.
Thus, in Fig. 36., _a_ is a bad curve, because it is part of a circle,
and is therefore monotonous throughout; but _b_ is a good curve, because
it continually changes its direction as it proceeds.

[Illustration: FIG. 37.]

The _first_ difference between good and bad drawing of tree boughs
consists in observance of this fact. Thus, when I put leaves on the line
_b_, as in Fig. 37., you can immediately feel the springiness of
character dependent on the changefulness of the curve. You may put
leaves on the other line for yourself, but you will find you cannot make
a right tree spray of it. For _all_ tree boughs, large or small, as well
as all noble natural lines whatsoever, agree in this character; and it
is a point of primal necessity that your eye should always seize and
your hand trace it. Here are two more portions of good curves, with
leaves put on them at the extremities instead of the flanks, Fig. 38.;
and two showing the arrangement of masses of foliage seen a little
farther off, Fig. 39., which you may in like manner amuse yourself by
turning into segments of circles--you will see with what result. I hope,
however, you have beside you by this time, many good studies of tree
boughs carefully made, in which you may study variations of curvature in
their most complicated and lovely forms.[250]

[Illustration: FIG. 38.]

[Illustration: FIG. 39.]

B. Not only does every good curve vary in