Again, as regards the leaves there can, I think, be no doubt that
similar considerations of utility are applicable. Their forms are
almost infinitely varied. To quote Ruskin's vivid words, they "take all
kinds of strange shapes, as if to invite us to examine them.
Star-shaped, heart-shaped, spear-shaped, arrow-shaped, fretted, fringed,
cleft, furrowed, serrated, sinuated, in whorls, in tufts, in spires, in
wreaths, endlessly expressive, deceptive, fantastic, never the same from
foot-stalk to blossom, they seem perpetually to tempt our watchfulness
and take delight in outstepping our wonder."

But besides these differences of mere form, there are many others: of
structure, texture, and surface; some are scented or have a strong
taste, or acrid juice, some are smooth, others hairy; and the hairs
again are of various kinds.

I have elsewhere[21] endeavoured to explain some of the causes which
have determined these endless varieties. In the Beech, for instance
(Fig. 15), the leaf has an area of about 3 square inches. The distance
between the buds is about 1-1/4 inch, and the leaves lie in the general
plane of the branch, which bends slightly at each internode. The basal
half of the leaf fits the swell of the twig, while the upper half
follows the edge of the leaf above; and the form of the inner edge being
thus determined, decides that of the outer one also.

[Illustration: Fig. 15.--Beech.]

The weight, and consequently the size of the leaf, is limited by the
strength of the twig; and, again, in a climate such as ours it is
important to plants to have their leaves so arranged as to secure the
maximum of light. Hence in leaves which lie parallel to the plane of the
boughs, as in the Beech, the width depends partly on the distance
between the buds; if the leaves were broader, they would overlap, if
they were narrower, space would be wasted. Consequently the width being
determined by the distance between the buds, and the size depending on
the weight which the twig can safely support, the length also is
determined. This argument is well illustrated by comparing the leaves of
the Beech with those of the Spanish Chestnut. The arrangement is
similar, and the distance between the buds being about the same, so is
the width of the leaves. But the terminal branches of the Spanish
Chestnut being much stronger, the leaves can safely be heavier; hence
the width being fixed, they grow in length and assume the well-known and
pecul