§ 232. Following an order like that of preceding chapters, let us first note a few typical facts respecting the forms of clusters of flowers, apart from the forms of the flowers themselves. Two kindred kinds of Leguminosæ serve to show how the members of clusters are distributed in an all-sided manner or in a two-sided manner, according as the circumstances are alike on all sides or alike on only two sides. In Hippocrepis, represented in Fig. 226, the flowers growing at the end of a vertical stem, are arranged round it in radial symmetry. Contrariwise in Melilotus, Fig. 227, where the axillary stem bearing the flowers is so placed in relation to the main stem, that its outer and inner faces are differently conditioned, the flowers are all on the outer face: the cluster is bilaterally symmetrical, since it may be cut into approximately equal and similar groups by a vertical plane passing through the main axis.

Plants of this same tribe furnish clusters of intermediate characters having intermediate conditions. Among these, as among the clusters which other types present, may be found some in which conformity to the general law is not obvious. The discussion of these apparent anomalies would carry us too much out of our course. A clue to the explanation of them will, I believe, be found in the explanation presently to be given of certain kindred anomalies in the forms of individual flowers.

§ 233. The radially-symmetrical form is common to all individual flowers that have vertical axes. In plants which are practically if not literally uniaxial, and bear their flowers at the ends of upright stalks, so that the faces open horizontally, the petals are disposed in an all-sided way. Crocuses, Tulips, and Poppies are familiar examples of this structure occurring under these conditions. A Ranunculus flower, Fig. 228, will serve as a typical one. Similarly, flowers which have peduncles flexible enough to let them hang directly downwards, and are not laterally incommoded, are also radial; as in the Fuchsia, Fig. 229, as in Cyclamen, Hyacinth, &c. These relations of form to position are, I believe, uniform. Though some flowers carried at the ends of upright or downright stems have oblique shapes, it is only when they have inclined axes or are not equally conditioned all round. No solitary flower having an axis habitually vertical, presents a bilateral form. This is as we should expect; since flowers which open out their faces horizontally, whether facing upwards or downwards, are, on the average, similarly affected on all sides.

At first it seems that flowers thus placed should alone be radial; but further consideration discloses conditions under which this type of symmetry may exist in flowers otherwise placed. Remembering that the radial form is the primitive form—that, morphologically speaking, it results from the contraction into a whorl, of parts that are originally arranged in the same spiral succession as the leaves; we must expect it to continue wherever there are no forces tending to change it. What now must be the forces tending to change it? They must be forces which do not simply affect differently the different parts of an individual flower. They must be forces which affect in like contrasted ways the homologous parts of other individual flowers, both on the same plant and on surrounding plants of the same species. A permanent modification can be expected only in cases where, by inheritance, the effects of the modifying causes accumulate. That they may accumulate the flowers must keep themselves so related to the environment, that the homologous parts may, generation after generation, be subjected to like differentiating forces. Hence, among a plant’s flowers which maintain no uniformity in the relations of their parts to surrounding influences, the radial form will continue. Let us glance at the several causes which entail this variability. When flowers are borne on many branches, which have all inclinations from the vertical to the horizontal—as are the flowers of the Apple, the Plum, the Hawthorn—they are placed in countless different attitudes. Consequently, any spontaneous variation in shape which might be advantageous were the attitude constant, is not likely to be advantageous; and any functionally-produced modification in one flower, is likely to be neutralized in offspring by some opposite functionally-produced modification in another flower. It is quite comprehensible, therefore, that irregularly-branched plants should thus preserve their laterally-borne flowers from undergoing permanent deviations from their primitive radial symmetry. Fig. 230, representing a blossoming twig of the Blackthorn, illustrates this. Again, upright panicles, such as those of the Saxifrage exemplified in Fig. 231, and irregular terminal groups of flowers otherwise named, furnish conditions under which there is similarly an absence of determinate relations between the parts of the flowers and the incident forces; and hence an absence of bilateralness. This inconstancy of relative position is produced in various other ways—by extreme flexibility of the stems, as in the Blue-bell; by the tendency of the peduncles to curl to a greater or less extent in diverse directions, as in Pyrola; by special twistings of the peduncles, differing in degree in different individuals, as in Convolvulus; by unusual laxity of the petals, as in Lythrum. Elsewhere the like general result arises from a progressive change of attitude, as in Myosotis, the stem of which as it unfolds causes each flower to undergo a transition from an upward position of the mouth to a lateral position; or as in most Cruciferæ, where the like effect follows from an altered direction of the peduncle.

There are, however, certain seemingly-anomalous cases where radial symmetry is maintained by laterally-placed flowers, which keep their parts in relative positions that are tolerably constant. The explanation of these exceptions is not manifest. It is only when we take into account certain incident actions liable to be left unremembered, that we find a probable solution. It will be most convenient to postpone the consideration of these cases until we have reached the general rule to which they are exceptions.

§ 234. Transitions varying in degree from the radial towards the bilateral, are common in flowers that are borne at the ends of branches or axes which are inclined in tolerably constant ways. We may see this in sundry garden flowers such as Petunia, or such as Isoloma and Achimenes, shown in Figs. 232 and 233. If these plants be examined, it will be perceived that the mode of growth makes the flower unfold in a partially one-sided position; that its parts of attachment have rigidity sufficient to prevent this attitude from being very much interfered with; and that though the individual flowers vary somewhat in their attitudes, they do not vary to the extent of neutralizing the differentiating conditions—there remains an average divergence from a horizontal unfolding of the flower, to account for its divergence from radial symmetry.

We pass insensibly from forms like these, to forms having bilateral symmetry strongly pronounced. Some such forms occur among flowers that grow at the ends of upright stems; as in Pinguicula, and in the Violet tribe. But this happens only where, in successive generations, the flower unfolds its parts sideways in constant relative positions. And in the immense majority of flowers having well-marked two-sided forms, the habitual exposure of the different parts to different sets of forces, is effectually secured by the mode of placing. As illustrations, I may name the genera—Orchis, Utricularia, Salvia, Salix, Delphinium, Mentha, Teucrium, Ajuga, Ballota, Galeopsis, Lamium, Stachys, Nepeta, Marrubium, Calamintha, Melittis, Prunella, Scutellaria, Bartsia, Euphrasia, Rhinanthus, Melampyrum, Pedicularis, Linaria, Digitalis, Orobanche, Fumaria, &c.; to which may be added all the Grasses and all the Papilionaceæ. In most of these cases the flowers, being sessile on the sides of upright stems, are kept in quite fixed attitudes; and in the other cases the peduncles are very short, or else stiff enough to secure general uniformity in the positions. A few of the more marked types are shown in Figs. 234 to 241.

Very instructive evidences here meet us. Sometimes within the limits of one genus we find radial flowers, bilateral flowers, and flowers of intermediate characters. The genus

Begonia may be instanced. In B. rigida the flowers, various in their attitudes, are in their more conspicuous characters radial: though there is a certain bilateralness in the calyx, the five petals are symmetrically disposed all round. B. Wageneriana furnishes two forms of flowers. On the same individual plant may be found radial flowers like Fig. 242, and others, like Fig. 243, which are merging into the bilateral. More decided is the bilateralness in B. albo-coccinea, Fig. 244; and still more in B. nitida, Fig. 245. While in B. heracleifolia, Fig. 246, the change reaches its extreme by the disappearance of the lateral petals. On examining the modes of growth in these several species, they will be seen to explain these changes in the manner alleged. Even more conclusive are the nearly-allied transformations occurring in artificially-produced varieties of the same species. Gloxinia may be named in illustration. In Fig. 247 is represented one of the ordinary forms, which shows us bilateralness of shape along with a mode of growth that renders the conditions alike on the two sides while different above and below. But in G. erecta, Fig. 248, we have the flower assuming an upright attitude, and at the same time assuming the radial type. This is not to be interpreted as a production of radial symmetry out of bilateral symmetry, under the action of the appropriate conditions. It is rather to be taken as a case of what is termed “peloria”—a reversion to the primitive radial type, from which the bilateral modification had been derived. The significant inference to be drawn from it is, that this primitive radial type had an upright attitude; and that the derivation of a bilateral type from it, occurred along with the assumption of an inclined attitude.

We come now to a group of cases above referred to, in which radial symmetry continues to co-exist with that constant lateral attitude ordinarily accompanied by the two-sided form. Two examples will suffice: one a very large flower, the Hollyhock, and the other a very small flower, the Agrimony. Why does the radial form here remain unchanged? and how does its continuance consist with the alleged general law?

Until quite recently I have been unable to find any probable answers to these questions. When the difficulty first presented itself, I could think of no other possible cause for the anomaly, than that the parts of the Hollyhock-flower, unfolding spirally as they do, might have different degrees of spiral twist in different flowers, and might thus not be unfolded in sufficiently-constant positions. But this seemed a questionable interpretation; and one which did not obviously apply to the case of the Agrimony. It was only on inquiring what are the special causes of modifications in the forms of flowers, that a more feasible explanation suggested itself; and this would probably never have suggested itself, had not Mr. Darwin’s investigations into the fertilization of Orchids led me to take into account an unnoticed agency.

The actions which affect the forms of leaves, affect much less decidedly the forms of flowers; and the forms of flowers are influenced by actions which do not influence the forms of leaves. Partly through the direct action of incident forces and partly through the indirect action of natural selection, leaves get their parts distributed in ways that most facilitate their assimilative functions, under the circumstances in which they are placed; and their several types of symmetry are thus explicable. But in flowers, the petals and fructifying organs of which do not contain chlorophyll, the tendency to grow most where the supply of light is greatest, is less decided, if not absent; and a shape otherwise determined is hence less liable to alter in consequence of altered relations to sun and air. Gravity, too, must be comparatively ineffective in causing modifications: the smaller sizes of the parts, as well as their modes of attachment, giving them greater relative rigidity. Not, indeed, that these incident forces of the inorganic world are here quite inoperative. Fig. 249, representing a species of Campanula, shows that the developments of individual flowers are somewhat modified by the relations of their parts to general conditions. But the fact to be observed is, that the extreme transformations which flowers undergo are not likely to be thus caused: some further cause must be sought. And if we bear in mind the functions of flowers, we shall find in their adaptations to these functions, under conditions that are extremely varied, an adequate cause for the different types of symmetry, as well as for the exceptions to them. Flowers are parts in which fertilization is effected; and the active agents of this fertilization are insects—bees, moths, butterflies, &c. Mr. Darwin has shown in many cases, that the forms and positions of the essential organs of fructification, are such as to facilitate the actions of insects in transferring pollen from the anthers of one flower to the pistil of another—an arrangement produced by natural selection. And here we shall find reason for concluding, that the forms and positions of those subsidiary parts which give their shapes to flowers, similarly arise by the survival of individuals which have the subsidiary parts so adjusted as to aid this fertilizing process—the deviations from radial symmetry being among such adjustments. The reasoning is as follows. So long as the axis of a flower is vertical and the conditions are similar all round, a bee or butterfly alighting on it, will be as likely to come from one side as from another; and hence, hindrance rather than facilitation would result if the several sides of the flower did not afford it equally free access. In like manner, flowers which are distributed over a plant in such ways that their discs open out on planes of all directions and inclinations, will have no tendency to lose their radial symmetry; since, on the average, no part of the periphery is differently related to insect-agency from any other part. But flowers so fixed as to open out sideways in tolerably-constant attitudes, have their petals differently related to insect-agency. A bee or butterfly coming to a laterally-growing flower, does not settle on it in one way as readily as in another; but almost of necessity settles with the axis of its body inclined upwards towards the stem of the plant. Hence the side-petals of a flower so fixed, habitually stand to the alighting insect in relations different from those in which the upper and lower petals stand; and the upper and lower petals differ from one another in their relations to it. If, then, there so arises an habitual attitude of the insect towards the petals, there is likely to be some arrangement of the petals that will be most convenient to the insect—will most facilitate its entrance into the flower. Thus we see in many cases, that a long undermost petal or lip, by enabling the insect to settle in such way as to bring its head opposite to the opening of the tube, aids its fertilizing agency. But whatever be the special modifications of the corolla which facilitate the actions of the particular insects concerned, all of them will conduce to bilateral symmetry; since they will be alike for the two sides but unlike for the top and bottom. And now we are prepared for understanding the exceptions. Flowers growing sideways can become thus adapted by survival of the fittest, only if they are of such sizes and structures that insect-agency can affect them in the way described. But in the plants named above, this condition is not fulfilled. A Hollyhock-flower is so open, as well as so large, that its petals are not in any appreciable degree differently related to the insects which visit it. On the other hand, the flower of the Agrimony is so small, that unless visited by insects of a corresponding size which settle as bees and butterflies settle, its parts will not be affected in the alleged manner. That all anomalies of this kind can at once be satisfactorily explained, is scarcely to be expected: the circumstances of each case have to be studied. But it seems not improbable that they are due to causes of the kind indicated.[36]

§ 235. We have already glanced at clusters of flowers for the purpose of considering their shapes as clusters. We must now return to them to observe the modifications undergone by their component flowers. Among these occur illustrations of great significance.

An example of transition from the radial to the bilateral form in clustered flowers of the same species, is furnished by the cultivated Geraniums, called by florists Pelargoniums. Some of these, bearing somewhat small terminal clusters of flowers, which are closely packed together with their faces almost upwards, have radially-symmetrical flowers. But among other varieties having terminal clusters of which the members are mutually thrust on one side by crowding, the flowers depart very considerably from the radial shape towards the bilateral shape. A like result occurs under like conditions in Rhododendrons and Azaleas. The Verbena, too, furnishes an illustration of radial flowers rendered slightly two-sided by the slight two-sidedness of their relations to other flowers in the cluster. And among the Cruciferæ a kindred case occurs in the cultivated Candytuft.

Evidence of a somewhat different kind is offered us by clustered flowers in which the peripheral members of the clusters differ from the central members; and this evidence is especially significant where we find allied species that do not exhibit the deviation, at the same time that they do not fulfil the conditions under which it may be expected. Thus, in Scabiosa succisa, Fig. 250, which bears its numerous small flowers in a hemispherical knob, the component flowers, similarly circumstanced, are all equal and all radial; but in Scabiosa arvensis, Fig. 251, in which the numerous small flowers form a flattened disk only the confined central ones are radial: round the edge the flowers are much larger and conspicuously bilateral.

But the most remarkable and most conclusive proofs of these relations between forms and positions, are those given by the clustered flowers called Umbelliferæ. In some cases, as where the component flowers have all plenty of room, or where the surface of the umbel is more or less globular, the modifications are not conspicuous; but where, as in Viburnum, Chærophyllum, Anthriscus, Torilis, Caucalis, Daucus, Tordylium, &c., we have flowers clustered in such ways as to be differently conditioned, we find a number of modifications that are marked and varied in proportion as the differences of conditions are marked and varied. In Chærophyllum, where the flowers of each umbellule are closely placed so as to form a flat surface, but where the umbellules are wide apart and form a dispersed umbel, the umbellules do not differ from one another; though among the flowers of each umbellule there are decided differences: the central flowers being small and radial, while the peripheral ones are large and bilateral. But in other genera, where not only the flowers of each umbellule but also the umbellules themselves, are closely clustered into a flat surface, the umbellules themselves become contrasted; and many remarkable secondary modifications arise. In an umbel of Heracleum, for instance, there are to be noted the facts;—first, that the external umbellules are larger than the internal ones; second, that in each umbellule the central flowers are less developed than the peripheral ones; third, that this greater development of the peripheral flowers is most marked in the outer umbellules; fourth, that it is most marked on the outer sides of the outer umbellules; fifth, that while the interior flowers of each umbellule are radial, the exterior ones are bilateral; sixth, that this bilateralness is most marked in the peripheral flowers of the peripheral umbellules; seventh, that the flowers on the outer sides of these peripheral umbellules are those in which the bilateralness reaches a maximum; and eighth, that where the outer umbellules touch one another, the flowers, being unsymmetrically placed, are unsymmetrically bilateral.[37] The like modifications are displayed, though not in so clearly-traceable a way, in an umbel of Tordylium, Fig. 252. Considering how obviously these various forms are related to the various conditions, we should be scarcely able, even in the absence of all other facts, to resist the conclusion that the differences in the conditions are the causes of the differences in the forms.

Composite flowers furnish evidence so nearly allied to that which clustered flowers furnish, that we may fitly glance at them under the same head. Such a common type of this order as the Sun-flower, exemplifies the extremely marked difference which arises in many of these plants between the closely-packed internal florets, each similarly circumstanced on all sides, and the external florets, not similarly circumstanced on all sides. In Fig. 253, representing the inner and outer florets of a Daisy, the contrast is marked between the small radial corolla of the one and the larger bilateral corolla of the other. In many cases, however, this contrast is less marked: the inner florets also having their outward-growing prolongations—a difference possibly related to some difference in the habits of the insects that fertilize them. Nevertheless, these composite flowers which have inner florets with strap-shaped corollas outwardly directed, equally conform to the general principle; both in the radial arrangement of the assemblage of florets, and in the bilateral shape of each floret; which has its parts alike on the two sides of a line passing from the centre of the assemblage to the circumference. Certain other members of this order fulfil the law somewhat differently. In Centaurea, for instance, the inner florets are small and vertical in direction, while the outer florets are large and lateral in direction. And here may be remarked, in passing, a clear indication of the effect which great flexibility of the petals has in preventing a flower from losing its original radiate form; for while in C. cyanus, the large outward-growing florets, having short, stiff divisions of the corolla, are decidedly bilateral, in C. scabiosa, where the divisions of the corolla are long and flexible, the radial form is scarcely at all modified. On bearing in mind the probable relations of the forms to insect-agency, the meaning of this difference will not be difficult to understand.[38]

§ 236. In extremely-varied ways there are thus re-illustrated among flowers, the general laws of form which leaves and branches and entire plants disclose to us. Composed as each cluster of flowers is of individuals that are originally similar; and composed as each flower is of homologous foliar organs; we see both that the like flowers become unlike and the like parts of each flower become unlike, where the positions involve unlike incidence of forces. The symmetry remains radial where the conditions are equal all round; shows deviation towards two-sidedness where there is slight two-sidedness of conditions; becomes decidedly bilateral where the conditions are decidedly bilateral; and passes into an unsymmetrical form where the relations to the environment are unsymmetrical.