ing too much to make his pupil, at starting, think
everything depends on black dots; still, the main lines are good, and
very characteristic of tree growth. Then, in Plate 2, we come to the
point at issue. The first examples in that plate are given to the pupil
that he may practice from them till his hand gets into the habit of
arranging lines freely in a similar manner; and they are stated by Mr.
Harding to be universal in application; "all outlines expressive of
foliage," he says, "are but modifications of them." They consist of
groups of lines, more or less resembling our Fig. 23 below; and the
characters especially insisted upon are, that they "tend at their inner
ends to a common center;" that "their ends terminate in [are inclosed
by] ovoid curves;" and that "the outer ends are most emphatic."

[Illustration: FIG. 23.]

130. Now, as thus expressive of the great laws of radiation and
inclosure, the main principle of this method of execution confirms, in a
very interesting way, our conclusions respecting foliage composition.
The reason of the last rule, that the outer end of the line is to be
most emphatic, does not indeed at first appear; for the line at one end
of a natural leaf is not more emphatic than the line at the other: but
ultimately, in Harding's method, this darker part of the touch stands
more or less for the shade at the outer extremity of the leaf mass; and,
as Harding uses these touches, they express as much of tree character as
any mere habit of touch _can_ express. But, unfortunately, there is
another law of tree growth, quite as fixed as the law of radiation,
which this and all other conventional modes of execution wholly lose
sight of. This second law is, that the radiating tendency shall be
carried out only as a ruling spirit in reconcilement with perpetual
individual caprice on the part of the separate leaves. So that the
moment a touch is monotonous, it must be also false, the liberty of the
leaf individually being just as essential a truth, as its unity of
growth with its companions in the radiating group.

131. It does not matter how small or apparently symmetrical the cluster
may be, nor how large or vague. You can hardly have a more formal one
than _b_ in Fig. 9, p. 47, nor a less formal one than this shoot of
Spanish chestnut, shedding its leaves, Fig. 24; but in either of them,
even the general reader, unpracticed in any of the previously
recommended exercises, must see that there are