periodic movement, such as the "sleep" of many leaves, or the nearly continuous rotation of the lateral leaflets of Desmodium. I propose, rather, to ask you to consider with me the structure, and especially the forms, of the common every-day leaves of our woods and fields.

In talking the subject over with friends, I have found a widely prevalent idea that the beauty and variety of leaves are a beneficent arrangement made specially with reference to the enjoyment and delight of man. I have, again, frequently been met by the opinion that there is some special form, size, and texture of leaf inherently characteristic of each species; that the cellular tissue tends to "crystallize," as it were, into some particular form, quite irrespective of any advantage to the plant itself.

Neither of these views will, I think, stand the test of careful examination.

In the first place, let us consider the size of the leaf. On what does this depend? In herbs we very often see that the leaves decrease towards the end of the shoot, while in trees the leaves, though not identical, are much more uniform, in size.

Again, if we take a twig of hornbeam, we shall find that the six terminal leaves have together an area of about 14 square inches, and the section of the twig has a diameter of 06 of an inch. In the beech the leaves are rather larger, six of them having an area of perhaps 18 inches, and, corresponding with this greater leaf-surface, we find that the twig is somewhat stouter, say '09 of an inch. Following this up we shall find that, cæteris paribus, the size of the leaf has relation to the thickness of the stem. This is clearly shown in the following table :

Impression of Stalk below the Sixth Leaf.

In the elm the numbers are 11 and 34, in the chestnut 15 and 72, and in the horse-chestnut the stem has a thickness of 32, and the six leaves have an area often of 300 square inches. Of course, however, these numbers are only approximate. Many things have to be taken into consideration. Strength, for instance, is an important element. Thus the ailanthus, with a stem equal in thickness to that of the horse-chestnut, carries a smaller area of leaves, perhaps because it is less compact. Again, the weight of the leaves is doubtless a factor in the case. Thus in some sprays of ash and elder which I examined of equal diameter, the former bore the larger expanse of leaves; but not only is the stem of the elder less compact, but the elder leaves, though not so large, were quite as heavy, if not indeed a little heavier. I was for some time puzzled by the fact that, while the terminal shoot of the spruce is somewhat thicker than that of the Scotch fir, the leaves are not much more than as long. May this not perhaps be due to the fact that they remain on more than twice as long, so that the total leaf area borne by the branch is greater, though the individual leaves are shorter? Again, it will be observed that the leaf area of the mountain ash is small compared to the stem, and it may, perhaps, not be unreasonable to suggest that this may be connected with the habit of the tree to grow in bleak and exposed situations. The position of the leaves, the direction of the bough, and many other elements would have also to be taken into consideration, but still it seems clear that there is a correspondence between thickness of stem and size of leaf. This ratio, moreover, when taken in relation with the other conditions of the problem, has, as we shall see, a considerable bearing not only on the size, but on the form of the leaf also.

The mountain ash has been a great puzzle to me; it is, of course, a true Pyrus, and is merely called ash from the resemblance of its leaves to those of the common ash. But the ordinary leaves of a pear are, as we all know, simple and ovate, or obovate. Why, then, should those of the mountain ash be so entirely different? May, perhaps, some light be thrown on this by the arrangement of the leaves? They are situated some distance apart, and though, as shown in the table, they are small in comparison to the diameter of the stem, still they attain a size of 15 square inches, or even more. Now, if they were of the same form as the ordinary pear leaf, they would be about 7 inches long by 2-3 in breadth. The mountain ash, as we know, lives in mountainous and exposed localities, and such a leaf would be unsuitable to withstand the force of the wind

in such situations. From this point of view, the division into leaflets seems a manifest advantage.

Perhaps it will be said that in some trees the leaves are much more uniform in size than in others. This is true. The sycamore,

for instance, varies greatly; in the specimen tabulated, the stem was 13 in diameter, and the area of the six upper leaves was 60 square inches. In another, the six upper leaves had an area of rather over 100 inches, but in this case the diameter of the stem was 18.

FIG. 1.

Another point is the length of the internode. In such trees as the beech, elm, hornbeam, &c., the distance from bud to bud varies comparatively little, and bears a tolerably close relation to the size of the leaf. In the sycamore, maple, &c., on the contrary, the length varies greatly.

Now, if, instead of looking merely at a single leaf, we consider the whole bough of any tree, we shall, I think, see the reason of their differences of form.

Let us begin, for instance, with the common lime (fig. 1). The leaf-stalks are arranged at an angle of about 40° with the branch, and the upper surfaces of the leaves are in the same plane with it. The result is, that they are admirably adapted to secure the maximum of light and air. Let us take, for instance, the second or third leaf in fig. 1. They are 44 inches long and very nearly as broad. The distance between the two leaves on each side is also just 4 inches, so that they exactly fill up the interval. In Tilia parvifolia the arrangement is similar, but leaves and internodes are both less, the leaves, say, 1 inch, and the internodes 6.

In the beech, the general plane of the leaves is again that of the branch (fig. 2), but the leaves themselves are ovate in form, and smaller, being only from 2 to 3 inches in length. On the other hand, the distance between the internodes is also smaller, being, say, 1 inch against something less than 2 inches. The diminution in length of the internode is not, indeed, exactly in proportion to that of the leaf, but, on the other hand, the leaf does not make so wide an angle with the stem. To this position is probably due the difference of form. The outline of the basal half of the leaf fits neatly to the branch, that of the upper half follows the edge of the leaf beyond, and the form of the inner edge being thus determined decides the outer one also.

FIG. 2.

In the nut (Corylus), the internodes are longer and the leaves

correspondingly broader. In the elm (Ulmus, fig. 3), the ordinary branches have leaves resembling, though rather larger than, those of the beech; but in vigorous shoots the internodes become longer and the leaves correspondingly broader and larger, so that they come nearly to resemble those of the nut.

But it may be said the Spanish chestnut (Castanea vulgaris, fig. 4) also has alternate leaves in a plane parallel to that of the branch, and with internodes of very nearly the same length as the beech. That is true; but, on the other hand, the terminal branches of the Spanish chestnut are stouter in proportion. Thus, immediately below the sixth leaf, the chestnut stalk may be 15 of an inch in thickness, that of the beech not much more than half as much. Consequently, the chestnut could, of course supposing the strength of the wood to be equal, bear a greater weight of leaf; but, the width of the leaf being determined by the distance between the internodes, the leaf is, so to say, compelled to draw itself out. In fig. 5 I have endeavoured to illustrate this by placing a spray of beech over one of Spanish chestnut. Moreover, not only do the leaves on a single twig thus admirably fit in with one another, but they are also adapted to the ramification of the twigs themselves. Fig. 6 shows a bough of beech seen from above, and it will be observed that the form of the leaves is such that, while but little space is

FIG. 6.

lost, there is scarcely any over-lapping. Each fits in perfectly with

the rest.

The leaves of the yew (fig. 7) belong to a type very different from those which we have hitherto been considering. They are long, narrow, and arranged all round the stem, but spread right and left, so that they lie in one plane, parallel to the direction of the branchlet, and their width bears just such a relation to their distance apart that when so spread out their edges almost touch. Fig. 8 repre

sents a sprig of box. It will be observed that the increase of width in the leaves corresponds closely with the greater distance between the points of attachment.

The leaves of the Scotch pine (Pinus sylvestris) are needle-like, 1 inch in length and in diameter. They are arranged in pairs, each pair enclosed at the base in a sheath. One inch of stem bears about fifteen pairs of leaves. Given this number of leaves in such a space, they must evidently be long and narrow. If I am asked why they are longer than those of the yew, I would suggest that the stem, being thicker, is able to support more weight. In confirmation of this, we may take for comparison the Weymouth pine, in which the leaves are much longer and the stalk thicker.

When we pass from the species hitherto considered to the maples (fig. 11), sycamores, and horse-chestnuts (figs. 9 and 10), we come to a totally different type of arrangement. The leaves are placed at right angles to the axis of the branch instead of being parallel to it, have long petioles, and palmate instead of pinnate veins. In this group the mode of growth is somewhat stiff; the main shoots are perpendicular, and the lateral ones nearly at right angles to them. The buds, also,