are comparatively few, and the internodes, consequently, at greater distances apart, sometimes as much as a foot, though the two or three at the end of a branch are often quite short. The general habit is shown in figs. 9 and 10. Now, if we were to imagine six beech or elm leaves on these three internodes, it is obvious that the leaf surface

would be far smaller than it is at present. Again, if we compare the thickness of an average sycamore stem below the sixth leaf with that of a beech stem, it is obvious that there would be a considerable waste of power. Once more, if the leaves were parallel to the branch, they would, as the branches are arranged, be less well disposed with reference to light and air. A glance at figs. 9, 10, and 11, however, will show how beautifully the leaves are

adapted to their changed conditions. The blades of the leaves of the upper pair form an angle with the leaf-stalks, so as to assume a horizontal position, or nearly so; the leaf-stalks of the second pair decussate with those of the first, and are just so much longer as to bring up that pair nearly, or quite, to a level with the first; the third pair decussate with the second, and are again brought up nearly to the same level, and immediately to the outside of the first pair. there is often a fourth pair on the outside look at such a cluster of leaves directly from in front, we shall see that they generally appear somewhat to overlap; but it must be remembered that in temperate regions the sun is never vertical.

FIG. 11.

Moreover, while alternate leaves are more convenient in such an arrangement as that of the beech, where there would be no room for a second leaf, it is more suitable in such cases as the sycamores and maples that the leaves should be opposite, because, if, other things remaining the same, the leaves of the sycamore were alternate, the sixth leaf would require an inconvenient length of petiole.

Perhaps it will be said that the plane-tree, which has leaves so like a maple that one species of the latter genus is named after it (Acer platanoides), has, nevertheless, alternate leaves. In reality, however, I think this rather supports my argument, because the leaves of the plane, instead of being at right angles to the stem, lie more nearly parallel with it. Moreover, as any one can see, the leaves are not arranged so successfully with reference to exposure as those of the species we have hitherto been considering, perhaps because, living as it does in more southern localities, the economy of sunshine is less important than in more northern regions.

The shoot of the horse-chestnut is even stouter than that of the sycamore, and has a diameter below the sixth leaf of no less than 13 of an inch. With this increase of strength is, I think, connected the greater size of the leaves, which attain to as much as 18 inches in diameter, and this greater size, again, has perhaps led to the dissection of the leaves into five or seven distinct segments, each of which has a form somewhat peculiar in itself, but which fits in admirably with the other leaflets. However this may be, we have in the horse-chestnut, as in the sycamore and maples, a beautiful dome of leaves, each standing free from the rest, and expanding to the fresh air and sunlight a surface of foliage in proportion to the stout, bold stem on which they are born.

Now, if we place the leaves of one tree on the branches of another, we shall at once see how unsuitable they would be. I do not speak of putting a small leaf such as that of a beech on a large leaved tree such as the horse-chestnut ; but if we place, for instance, beech on lime, or vice versa, the contrast is sufficiently striking.

The lime leaves would overlap one another, while, on the other hand, the beech leaves would leave considerable interspaces. Or let us in the same way transpose those of the Spanish chestnut (Castanea) and those of Acer platanoides, a species of maple. I have taken specimens in which the six terminal leaves of a shoot of the two species occupy approximately the same area. Figs. 4 and 11 show the leaves in their natural position, those of Castanea lying along the stalk, while those of Acer are ranged round it. In both cases it will be seen that there is practically no overlapping, and very little waste of In Castanea the stalks are just long enough to give a certain to the leaves. In Acer they are much longer, bringing the leaves

space.

approximately to the same level, and carrying the lower and outer ones free from the upper and younger ones.

Now, if we arrange the Spanish chestnut leaves round a centre, as in fig. 12, it is at once obvious how much space is wasted. On the other hand, if we attach the leaves of the Acer to the stalk of Castanea at the points from which the leaves of Castanea came off, as in fig. 13, we shall see that the stalks are useless, and even

mischievous, as a cause of weakness and of waste of space; while, on the other hand, if we omit the stalks, or shorten them to the same length as those of Castanea, as in fig. 14, the leaves would greatly overlap one another.

Once more, for leaves arranged as in the beech the gentle swell at the base is admirably suited; but in a

crown of leaves such as those of the sycamore, space would be wasted, and it is better that they should expand at once as soon as their stalks have borne them free from those within. Moreover, the spreading lobes leave a triangular space (fig. 11) with the insertion of the stalk at the apex, which seems as if expressly designed to leave room for the pointed end of the leaf within.

FIG. 14.

Hence we see how beautifully the whole form of these leaves is adapted to the mode of growth of the trees themselves and the arrangement of their buds.

Before we proceed to consider the next series of species to which I wish to direct attention, it will be necessary for me to say a few words on the microscopical structure of the leaf. Although so thin,

the leaf consists of several layers of cells. Speaking roughly, and as a general rule, we may say that on each side is a thin membrane, or epidermis, underneath which on the upper side are one or more layers of elongated cells known from their form as "pallisade cells," beneath which is a parenchymatous tissue of more or less loose texture. The leaf is strengthened by ribs of woody tissue. From this general type there are, of course, numerous variations. For instance, some water plants have no epidermis.

If the surface of the leaf be examined with a tolerably high power, small opaque spots will be observed, resembling a sort of button-hole, with a thick rim or border composed of two more or less curved cells, the concavities being turned inwards. When dry, they are nearly straight, and lie side by side; but when moistened they swell, become somewhat curved, and gape open.

It is difficult to realize the immense number of these orifices or "stomata" which a single bush or tree must possess when we remember that there are sometimes many thousand stomata to a square inch of surface. In a large proportion of herbs the two sides of the leaf are under conditions so nearly similar that the stomata are almost equally numerous on the upper and on the lower side. In trees, however, as a general rule, they are found exclusively on the under side of the leaf, which is the most protected; they are thus less exposed to the direct rays of the sun, or to be thoroughly wetted by rain, so that their action is less liable to sudden and violent changes. There are, however, some exceptions; for instance, in the black poplar the stomata are nearly as numerous on one side of the leaf as on the other. Now, why is this? If we compare the leaves of the black and white poplar, we shall be at once struck by the fact that, though these species are so nearly allied, the leaves are very different. In the white poplar (Populus alba), the upper and under sides are very unlike both in colour and texture, the under side being thickly clothed with cottony hairs. In the black poplar (P. nigra, fig. 15), the upper and under surfaces are, which is not frequent, very similar in colour and texture. The petioles or leaf-stalks, again, are unlike; those of P. nigra presenting the peculiarity of being much flattened at the end towards the leaf. The effect of the unusual structure of the petiole is that the leaf, instead of being horizontal as in the P. alba and most trees, hangs vertically, and this again explains the similarity of the two surfaces, because the result is that both surfaces are placed under nearly similar conditions as

FIG. 15.

regards light and air. Again, it will be observed that, if we attempt to arrange the leaves of the black poplar on one plane, they generally overlap one another; the extent is larger than can be displayed without their interfering with one another. In foliage arranged like that, for instance, of the beech, elm, sycamore, or, in fact, of most of our trees, this would involve a certain amount of waste; but in the black poplar, as fig. 15 shows, the leaves when hung in their natural position are quite detached from one another. Another interesting case of a species with vertical leaves is the prickly lettuce (Lactuca scariola), while those of L. muralis and L. virosa are horizontal. With this position of the leaves is connected another peculiarity, especially well marked in the so-called compass" plant of the American prairies (Silphium laciniatum), a yellow composite not unlike a small sunflower, which is thus named because the leaves turn their edges north and south. This has long been familiar to the hunters of the prairies, but was first mentioned by General Alvord, who called Longfellow's attention to it, and thus inspired the lines in "Evangeline:"

"Look at this delicate plant, that lifts its head from the meadow,
See how its leaves are turned north, as true as the magnet;
This is the compass flower, that the finger of God has planted
Here in the houseless wild to direct the traveller's journey
Over the sea-like, pathless, limitless waste of the desert."

The advantage of this position, and consequently the probable reason for its adoption, is that in consequence of it the two faces of the leaf are about equally illuminated by the sun; and in connection with this we find that the structure of the leaf is unusual in two respects. The stomata are about equally abundant on both surfaces, while pallisade cells, which are generally characteristic of the upper surface, are in this species found on the lower one also.

The leaves of the Lactuca scariola have also, when growing in sunny situations, a tendency to point north and south. Under such circumstances also they have a layer of pallisade cells on each side.

Hitherto I have dealt with plants in which one main consideration appears to be the securing as much light and air as possible. Our English trees may be said as a general rule to be glad of as much sun as they can get. But a glance at any shrubbery is sufficient to show that we cannot explain all leaves in this manner, and in tropical countries some plants at any rate find the sun too much for them. I will presently return to the consideration of the general characteristics of tropical vegetation. In illustration, however, of the present point, perhaps the clearest evidence is afforded by some Australian species, especially the eucalypti and acacias. Here the adaptations which we meet with are directed, not to the courting, but to the avoidance, of light.