design has suffered
grievously. You may ascertain, by experiment, that all beautiful objects
whatsoever are thus terminated by delicately curved lines, except where
the straight line is indispensable to their use or stability; and that
when a complete system of straight lines, throughout the form, is
necessary to that stability, as in crystals, the beauty, if any exists,
is in color and transparency, not in form. Cut out the shape of any
crystal you like, in white wax or wood, and put it beside a white lily,
and you will feel the force of the curvature in its purity, irrespective
of added color, or other interfering elements of beauty.

[Illustration: FIG. 35.]

206. Well, as curves are more beautiful than straight lines, it is
necessary to a good composition that its continuities of object, mass,
or color should be, if possible, in curves, rather than straight lines
or angular ones. Perhaps one of the simplest and prettiest examples of a
graceful continuity of this kind is in the line traced at any moment by
the corks of a net as it is being drawn: nearly every person is more or
less attracted by the beauty of the dotted line. Now, it is almost
always possible, not only to secure such a continuity in the arrangement
or boundaries of objects which, like these bridge arches or the corks of
the net, are actually connected with each other, but--and this is a
still more noble and interesting kind of continuity--among features
which appear at first entirely separate. Thus the towers of
Ehrenbreitstein, on the left, in Fig. 32, appear at first independent of
each other; but when I give their profile, on a larger scale, Fig. 35,
the reader may easily perceive that there is a subtle cadence and
harmony among them. The reason of this is, that they are all bounded by
one grand curve, traced by the dotted line; out of the seven towers,
four precisely touch this curve, the others only falling hack from it
here and there to keep the eye from discovering it too easily.

[Illustration: FIG. 36.]

207. And it is not only always possible to obtain continuities of this
kind: it is, in drawing large forests or mountain forms, essential to
truth. The towers of Ehrenbreitstein might or might not in reality fall
into such a curve, but assuredly the basalt rock on which they stand
did; for all mountain forms not cloven into absolute precipice, nor
covered by straight slopes of shales, are more or less governed by these
great curves, it b