§ 228. Next in the descending order of composition come compound leaves. The relative sizes and distributions of their leaflets, as affecting their forms as wholes, have to be considered in their relations to conditions. Figs. 206, 207, represent leaves of the common Oxalis and of the Marsilea, in which radial symmetry is as completely displayed as the small number of leaflets permits. This equal development of the leaflets on all sides, occurs where the foot-stalks, growing up vertically from creeping or underground stems, are so long that the leaves either do not interfere with one another or do it in an inconstant way: the leaflets are not differently conditioned on different sides, as they are where the foot-stalks grow out in the ordinary manner. How unlikeness of position influences the leaflets is clearly shown in a Clover-leaf, Fig. 208, which deviates from the Oxalis-leaf but slightly towards bilateralness, as it deviates from it but slightly in the attitude of its petiole; which is a little inclined away from the others borne by the same procumbent axis. A familiar example of an almost radial symmetry along with almost equal relations to surrounding conditions, occurs in the root-leaves of the Lupin, Fig. 209 b. Here though we have lateral divergence from a vertical axis, yet the long foot-stalks preserve nearly erect positions, and carry their leaves to such distances from the axis, that the development of the leaflets on the side next it is not much hindered. Still the interference of the leaves with one another is, on the average, somewhat greater on the proximal side than on the distal side; and hence the interior leaflets are rather less than the exterior leaflets. In further proof of which influence, let it be added that, as shown in the figure, at a, the leaves growing out of the flowering stem deviate towards the two-sided form more decidedly. Two-sidedness is much greater where there is a greater relative proximity of the inner leaflets to the axis, or where the foot-stalk approaches towards a horizontal position. The Horse-chestnut, Fig. 205, already instanced as showing how the arrangements and sizes of leaflets are determined by the incidence of forces, serves also to show how the incidence of forces determines the relative sizes and arrangements of leaflets. Fig. 210, which shows a leaf of the Bombax, further illustrates this relation of structure to conditions.

Compound leaves that are completely bilateral, present us with modifications of form exemplifying the same general truth in another way. In them the proximal and distal parts have none of that resemblance which we see in those intermediate forms just described. The portion next the axis and the portion furthest from the axis are entirely different; and the only likeness is between the wings or leaflets on opposite sides of the main foot-stalk or mid-rib. On turning back to Fig. 65, it will be seen that the compound leaf there drawn to exemplify another truth, serves also to exemplify this truth: the homologous parts a, b, c, d, while they are unlike one another, are, in their main proportions, severally like the parts with which they are paired. And here let us not overlook a characteristic which is less conspicuous but not less significant. Each of the lateral wings has winglets that are larger on the one side than on the other; and in each case the two sides are dissimilarly conditioned. Even in the several components of each wing may be traced a like divergence from symmetry, along with a like inequality in the relations to the rest: the proximal half of each leaflet is habitually larger than the distal half. In the leaves of the Bramble, previously figured, kindred facts are presented. How far such differences of development are due to the positions of the parts in the bud; how far the respective spaces available for the parts when unfolded affect them; and how far the parts are rendered unlike by unlikenesses in their relations to light; it is difficult to say. Probably these several factors operate in all varieties of proportion. That the habitual shading of some parts by others largely aids in causing these divergences from symmetry, is very instructively shown by the compound leaves of the Cow-parsnip. Fig. 211 represents one of these. While the leaf as a whole is bilaterally symmetrical, each of the wings has an unsymmetrical bilateralness: the side next the axis being larger than the remoter side. How does this happen? Fig. 212, which is a diagrammatic section down the mid-rib of the leaf, showing its inclined attitude and the positions of the wings a, b, c, will make the cause clear. As the wings overlap, like the bars of a Venetian blind, each intercepts some light from the one below it; and the one below it thus suffers more on its distal side than on its proximal side. Hence the smaller development of the distal side. That this is the cause is further shown by the proportion that is maintained between the degree of obscuration and the degree of non-development; for this unlikeness is greater between the two sides a and a´, than between b and b´ or c and c´, at the same time that the interference is greater in the lower wings than in the upper. Of course in this case and in the kindred cases hereafter similarly interpreted, it is not meant that this differentiation is consequent solely, or even chiefly, on the differential actions experienced by the individual plant. Though there is good reason to believe that the rate of growth in each part of each leaf is affected by the incidence of light, yet contrasts so marked and so systematic as these are not explicable without taking into account the inheritance of modifications either functionally caused or caused by spontaneous variation. Clearly, the tendency will be towards the preservation of a plant which distributes its chlorophyll in the most advantageous way; and hence there will always be a gravitation towards a form in which shaded parts of leaves are undeveloped.

§ 229. From compound leaves to simple ones, we find transitions in leaves of which the divisions are partial instead of total; and in these we see, with equal clearness, the relations between forms and positions that have been traced thus far. Fig. 213 is the leaf of a Winter-aconite in which, round a vertical petiole, there is a radial distribution of half-separated leaflets. The Cecropia-leaf, Fig. 214, shows us a two-sided development of the parts beginning to modify, but not obliterating, the all-sided arrangement; and this mixed symmetry occurs under conditions that are intermediate. A more marked degree of the same relation is presented in the leaf of the Lady’s Mantle, Fig. 215. And then in the Sycamore and the Vine, we have a cleft type of leaf in which a decided bilateralness of form co-exists with a decided bilateralness of conditions.

The quite simple leaves to which we now descend, exhibit, very distinctly, a parallel series of facts. Where they grow up on long and completely-independent foot-stalks, without definite subordination to some central vertical axis, the leaves of water-plants are symmetrically peltate. Of this the sacred Indian-bean, Fig. 216, furnishes an example. Here there is only a trace of bilateralness in the venation of the leaf, corresponding to the very small difference of the conditions on the proximal and distal sides. In the Victoria regia, Fig. 217, the foot-stalks, though radiating almost horizontally from a centre, are so long as to keep the leaves quite remote from one another; and in it each leaf is almost symmetrically peltate, with a bilateralness indicated only by a seam over the line of the foot-stalk. The leaves of the Nymphæa, Fig. 218, more closely clustered, and having less room transversely than longitudinally, exhibit a marked advance to the two-sided form; not only in the excess of the length over the breadth, but in the existence of a cleft, where in the Victoria regia there is merely a seam. Among land-plants similar forms are found under analogous conditions. The common Hydrocotyle, Fig. 219, which sends up direct from its roots a few almost upright leaf-stalks, has these surmounted by peltate leaves; which leaves, however, diverge slightly from radial symmetry in correspondence with the slight contrast of circumstances which their grouping involves. Another case is supplied by the Nasturtium, Fig. 220, which combines the characters—a creeping stem, long leaf-stalks growing up at right angles to it, and unsymmetrically peltate leaves, of which the least dimension is, on the average, towards the stem. But perhaps the most striking illustration is that furnished by the Cotyledon umbilicus, Fig. 221, in which different kinds of symmetry occur in the leaves of the same plant, along with differences in their relations to conditions. The root-leaves, a, growing up on vertical petioles before the flower-stalk makes its appearance, are symmetrically peltate; while the leaves which subsequently grow out of the flower-stalk, b, are at the bottom transitionally bilateral, and higher up completely bilateral.

That the bilateral form of leaf is the ordinary form, corresponds with the fact that, ordinarily, the circumstances of the leaf are different in the direction of the plant’s axis from what they are in the opposite direction, while transversely the circumstances are alike. It is needless to give diagrams to illustrate this extremely familiar truth. Whether they are broad or long, oval or heart-shaped, pointed or obtuse, the leaves of most trees and plants will be remembered by all as having the ends by which they are attached unlike the free ends, while the two sides are alike. And it will also be remembered that these equalities and inequalities of development correspond with the equalities and inequalities in the incidence of forces.

§ 230. A confirmation that is interesting and important, is furnished by the cases in which leaves present unsymmetrical forms in positions where their parts are unsymmetrically related to the environment. A considerable deviation from bilateral symmetry may be seen in a leaf which habitually so carries itself, that the half on the one side of the mid-rib is more shaded than the other half. The drooping branches of the Lime, delineated in Fig. 222, show us leaves so arranged and so modified. On examining their attitudes and their relations one to another, it will be found that each leaf is so inclined that the half of it next to the shoot grows over the shoot and gets plenty of light; while the other half so hangs down that it comes a good deal into the shade of the preceding leaf. The result is that having leaves which fall into these positions, the species profits by a large development of the exposed halves; and by survival of the fittest, acting along with the direct effect of extra exposure, this modification becomes established. How unquestionable is the connexion between the relative positions of the halves and their relative developments, will be admitted on observing a converse case. Fig. 223 represents a shoot of Strobilanthes glomeratus. Here the leaves are so set on the stem that the inner half of each leaf is shaded by the subsequently-formed leaf, while its outer half is not thus shaded; and here we find the inner half less developed than the outer half. But the most conclusive evidence of this relation between unsymmetrical form and unsymmetrical distribution of surrounding forces, is supplied by the genus Begonia; for in it we have a manifest proportion between the degree of the alleged effect and the degree of the alleged cause. These plants produce their leaves in pairs, in such ways that the connate leaves interfere with one another, much or little according as the foot-stalks are short or long; and the result is a correlative divergence from symmetry. In Begonia nelumbiifolia, which has petioles so long that the connate leaves are not kept close together, there is but little deviation from a bilaterally-peltate form; whereas, accompanying the comparatively marked and constant proximity in B. pruinata, Fig. 224, we see a more decidedly unsymmetrical shape; and in B. mahringii, Fig. 225, the modification thus caused is pushed so far as to destroy the peltate structure.[34]

§ 231. Again, then, we are taught the same truth. Here, as before, we see that homologous units of any order become differentiated in proportion as their relations to incident forces become different. And here, as before, we see that in each unit, considered by itself, the differences of dimension are greatest in those directions in which the parts are most differently conditioned; while there are no differences between the dimensions of the parts that are not differently conditioned.[35]