Journal of the Royal Horticultural Society 29: 93-101 (1904/5)
METAMORPHOSES OF INFLORESCENCES AND FLOWERS

By REV. PROF. G. HENSLOW, M.A., V.M.H.

INTRODUCTION.—One Of the most inexplicable things in plant-life is what may be called a normally localised energy concerned in making some one organ, taking up the function of another, so that a special structure is evolved, of a different kind, and altogether outside its usually perfectly limited sphere of action.

Hence, assuming that a whole inflorescence contains a peduncle and pedicels, bracts and bracteoles, the flowers being composed of sepals, petals, stamens, and carpels, all sorts of interchanges, very often only imperfectly effected, may occur between them.

If it be asked why any particular metamorphosis takes place, scarcely any satisfactory answer can yet be given at all. The reason is that it is a matter which is concerned with or executed by life; and we do not know what life is, nor the laws governing its activities. We see the effects of unknown causes in the changes of structures, but we know nothing of the nature of those causes and processes executed by life, which have brought them about. We may, however, always see an object in view, whether it be successfully obtained or not; but how it came about, for example, that the peduncle and pedicels of the inflorescence of the Vine could be converted into an excellent climbing organ or tendril, as perfectly adapted to its purpose as the stalk and branchlets are to carry grapes, is a mystery past finding out.

If a Vine be well searched, many examples can be found in which the tendril bears a few abortive buds revealing its true source. We must, however, distinguish between useful metamorphoses on the one hand, and more or less useless ones on the other. The former have resulted in well-defined and important organs. The latter remain as abortive attempts at changes only.

METAMORPHOSES IN PEDUNCLES AND PEDICELS

Useful Changes.—Besides the Vine, there are other plants of which the flowering branch has been changed into tendrils, but of different kinds. Ampelopsishederacea or Virginia Creeper and its ally A. Veitchii, the Passion-flower and Bauhinia, well illustrate Nature's power of adaptation of floral axes for climbing by tendrils, in different ways and in different families of plants.

Everybody is familiar with the method of climbing by adhesive pads in the two species of Ampelopsis, both belonging to the Vine family.

The Virginia Creeper turns its tendrils away from the light so as to reach, if possible, a 'rough surface. As soon as the hooks at the ends of the branchlets catch in any little depression, the irritation causes them to swell into adhesive pads, which are never formed previously to contact.

In the Japanese species, however, the pads are developed before any contact takes place, but are not adhesive till such is the case. Hence the former reveals the evolutionary history of the tendrils in the latter species, in which the effects of a mere irritation are hereditary.

The Passion-flower has only the peduncle transformed into a tendril, which behaves precisely like that of our common Bryony; but in this latter plant the tendril appears to be more comparable to a stipule, having, however, a somewhat anomalous origin.

Another form of a climbing organ constructed out of a flowering branch is that of a hook. This is seen in Uncaria, Artabotrys, &c., which enables the plants to scramble if not to climb.

A very different but useful metamorphosis of a peduncle is seen in the structure of certain inflorescences. Thus, in the Fig the peduncle forms the bag-like edible structure, the inner surface of which is densely covered with flowers. The forerunner of this is seen in the open dishlike form of Dorstenia, of the same family. It is comparable to the so-called "general receptacle" of the Compositae, well seen in the large plate-like structure which carries all the florets and fruits of a Sunflower. The same is edible in the Globe Artichoke.

In whole groups of flowers the pedicel forms what is called the "receptacular tube," such as the hep of the Rose and the cup round the pistil in a Cherry blossom. In both, the sepals, petals, and stamens arise from the rim, having been elevated by the growth of the tube. The tube or cup may be adherent to the ovary. Such is the case in the Apple, in which the two adjacent skins or epidermides are not developed, so that the two middle tissues become confluent; the core really representing the inner epidermis of the carpel only.

The use of the receptacular tube, and explanation of its origin, appears to have been as an organ for secreting honey. Such is well seen in the Raspberry, in which it forms a circular honey-trough round the pistil.

In the remarkable inflorescence of Marcgravia, the circlet of pitcher-like structures are abortive flowers, instead of which the peduncle has developed honey-pitchers. In several plants a not uncommon metamorphosis is for the pedicels to change into short thickened axes, and the parts of the flowers to be replaced by bulb-scales, so that veritable and propagative bulbils take the place of flowers, and in course of time fall to the ground. This is a peculiar habit in some Onions. In Alpine grasses little leafy buds often occur instead of flowers, and the panicle being heavily weighted falls to the ground, where each of the tufts may then strike root, and so form a new and independent plant.

Useless Changes. —The peduncle or pedicels which terminate in flowers are sometimes "proliferous," in which case the axis does not cease to grow, but continues through the flower (instead of the pistil), and may either become a perfect leafy shoot, or attempt to form a flower, the parts being more or less foliaceous; then it may proceed further and finally terminate in a perfect flower.

Cultivated Roses and the wild Water Avens (Geum rivale) are peculiarly liable to this prceedure.

In the Feather Hyacinth (Muscari comosum syn. Bellevallia comosa) all the flowers are aborted. The upper ones of the raceme are normally so when wild—it is a common field weed in Malta—but under cultivation the peduncle and pedicels elongate, becoming pink or purplish, and assume large feather-like structure. Of course all functions of a reproductive character are entirely gone.

* For further details on this subject, 1 would refer the reader to my paper on "Fasciation and Allied Phenomena," Journ. R.H.S. vol. xxvi. p. 155.

Fasciation is another common affection among peduncles, but it does not reveal any special use to the plant. The flowers, however, remain perfect, perhaps sometimes increased in number, and their fruits are fertile.*

BRACTS AND BRACTEOLES

Origin of Bracts.—Bracts and their diminutives, bracteoles, are usually abortive or arrested leaves, from the axils of which peduncles, pedicels, and sessile flowers arise. In their origin, they may be homologous with the petiole, or leaf-stalk, alone, or with the lamina, or blade, alone. As an example of the former, perfect transitions can be found between the "pedate" leaf of the Hellebore and the small oval and pointed bract. This is obtained by reducing the blade to nothing and expanding the petiole.

In Buttercups, however, the bract will be seen to consist of rudimentary segments of the lobed blade, the petiole being suppressed.

In some flowers, as the Willowherbs (Epilobium), the bracts are normally sessile leaves merely reduced in size.

Useful Changes. —Though bracts are usually rudimentary leaves, they constitute useful bud-scales, and so protect the flowers or florets within them. Thus in Composites they form the "involucre," i.e. a "wrapper," but they may retain or re-acquire a foliaceous character. They can thus perform all the assimilative and other functions of leaves. Such is the case with the large "spathe" of most Aroids, as of our Arum maculatum, the " Lords and Ladies."

Another use to which they are sometimes put by Nature is to make them attractive by their being white or brightly coloured. Thus in "Everlastings," among Composites, they may be white, yellow, red, purple, &c., while the florets themselves are comparatively inconspicuous. In Hedaroma (Darwinia) and Haemanthus, the bracts take on a very bright coloration and assume all the appearances of a gorgeous corolla.

In some species of Cornus the four bracts are white, so that the head of flowers mimics a four-sepaled Clematis.

Useless Changes.—Several members of the Compositae growing in dry inhospitable conditions develop spines in place of the bracts of the involucre. This spinescence is a common feature in other organs on desert plants as the result of drought. Such occurs in Thistles, and especially Centaurea Calcitrapa, the 'Caltrop,' so called from the resemblance of a head to that instrument of warfare. Such is not, however (any more than spiny shoots and leaves of Thistles and Gorse), intended for a protection against browsing animals, but is simply a common result of a deficiency of water.

In many plants bracts and bracteoles are quite rudimentary and useless, as in Umbelliferae, or wanting altogether, as in Cruciferae.

FLORAL ORGANS

CALYX

Useful Changes.—The calyx as a rule is green in the bud and acts as a protection to the internal organs. It is mostly homologous with the petiole only of a leaf, as may be readily seen in a Rose-bud, in which the imperfect leaflets are still present, but as functionless remnants of the compound leaf.

In some few instances the blade is present, as in the Corn-cockle.

Sometimes the blades are restored abnormally as true leaves, when, of course, they re-acquire the functions of normal foliage. Thus, in the cultivated variety of Primrose known as 'Jack-in-the-Green,' the calyx develops the limb, or blade, thereby becoming foliaceous, so that each sepal is equivalent to a perfect leaf.

The calyx may be either normally or abnormally coloured. Clematis, Anemone, Caltha, and Hellebore have no corollas, the white or coloured calyx undertaking to be the attractive organ instead.

Abnormally, it is seen in the 'Hose in hose' form of Mimulus, Azalea, Campanula ('Cup and Saucer'), and Primrose.

In Hydrangea, the calyx is large, white or coloured; but this is secured at the expense of the other organs (corolla, stamen, and pistil) becoming abortive; so that no seed can be set, and the flowers are said to be neuter.

As a rule it is only the outermost flowers of the cluster which thus render the whole attractive, so that the perfect but very inconspicuous little flowers in the middle can be discovered and visited by insects.

Useless Changes.— Degeneracy of the calyx to useless rudiments is by no means uncommon. It may be even entirely suppressed. This is especially the case when flowers of an inflorescence are much crowded. Hence members of the families Dipsaceae, Valerianaceae, Rubiaceae, Compositae, and Umbelliferae are particularly characterised by having rudimentary sepals, or none at all. In the Compositae it is often represented by hairs or the "pappus," when it re-acquires a use for dispersal.

Some other useless changes or conditions of the sepals are seen in malformations when the sepals try to be stamens or carpels. Such are comparatively rare, but have been described in Philadelphus speciosus, in which the sepals attempted to become stamens, and in double flowers of the Garden Pea the sepals formed open carpels with rudimentary ovules.

COROLLA

Useful Changes.—The normal use of a corolla is as an attractive organ, being visible to insects. It may, however, become green and foliaceous, in which case its use is changed to that of leaves.

Such occurs in the cultivated Green Rose and Alpine Strawberry, and not infrequently with late flowers of the Wild Honeysuckle, &c.

Useless Changes.—The wild Guelder Rose resembles the Hydrangea in having the flowers on the circumference of the truss or corymb conspicuous but neuter, while the minute central flowers are perfect; this is not the petaloid calyx, but an enlargement of the corolla, which renders the corymb conspicuous. In the garden form all the flowers have enlarged corollas, by the sacrifice of the essential organs. Hence the globular mass of flowers is now of no use to the plant at all.

In some few plants the corolla has been known to bear abortive anthers. Such has occurred in Foxgloves, Campanula, and Begonias. There may be even an effort to produce or convert a petal into a carpel with ovules. Begonias not infrequently exhibit various kinds of such malformations, but they are always abortive.

A large number of plants have flowers which show an entomophilous ancestry, but are now in a state of degradation, having become self-fertilising and far more fertile than was their previous condition. Under this changed condition the corolla is greatly reduced or has vanished. Shepherd's Purse is a case of the first; cleistogamous flower-buds of violet, the second. Such degradations may be regarded as metamorphoses into rudimentary organs.

STAMENS

Useful Changes.—The change from stamens to petals has laid the foundation, as it were, for the whole floral world. The first flowers, as far as we know, had no corolla nor coloured calyx. The Gymnosperms certainly preceded both Dicotyledons and Monocotyledons; and if existing forms of Gymnosperms have retained their primitive and ancestral characters, nothing but yellow anthers existed to tell us both what was the primitive colour, other than green, and what was the origin of petals. That stamens of Gymnosperms originated out of green scales appears revealed in such types as Cypress and Juniper; while, turning to ordinary flowering plants of the above-named classes, we seem to see the process of petal-making at work in such, plants as Water-lilies and Cannas.

In the former, as the filament broadens, the anther-cells become arrested, till complete petals are formed, in tracing them from within outwards, as may he readily seen in dissecting the flowers of the Waterlily.

In Canna all stamens are petaloid with the exception of one.

As another useful change, we find the filaments broadening in Atragene, forming quasi petals, but really destined to secrete honey.

In several other members of the Ranunculaeeae, it is the anthers which have undergone a change and become converted into honey-secreting nectaries, as may be well seen in the Winter Aconite and Hellebore. In the Buttercup one half of the anther constitutes a petal, the other half remains as a minute flap at the base. Between the two is the honey-secreting surface. Besides the Hellebores, in the Aconite, Larkspur, and Columbine the anthers have become variously spurred nectaries.

Useless Changes.— Double flowers are usually composed of metamorphosed stamens and carpels, coupled with a great multiplication of their number. Thus a Wallflower has normally six stamens and two carpels, but a double flower of this plant may have more than fifty petals.

In such flowers, whenever all sexuality is lost the doubling is of no utility to the plant itself.

This kind of doubling is mostly the result of the formation of petals out of filaments; but a "petaloid" condition often first shows itself in the anthers. In some forms it remains with them as explained in the Buttercup. In the double Larkspur the concentric circles of spurred petals are the result of several series of stamens having their anthers converted into spur-like nectaries, the spurs fitting one into another in radial series.

DOUBLE FLOWERS

The word "doubling" may be applied to any organ which is multiplied; so that, for convenience, various kinds of metamorphosis, &c., may be gathered together under this head as follows.

Beginning with bracts, it sometimes happens that the flower is totally arrested. The diverted energy is now directed into bracts, which are thus abnormally multiplied to a great extent. Thus arises the "wheat-eared" Carnation, in which the two pairs of opposite bracts of a normal flower are multiplied so as to form a long series, roughly resembling an ear of wheat.

In the 'Green Dahlia' the florets of the heads are suppressed, so that the usual small chaff-like scales become enlarged into a mass of green bracts.

The multiplication of the sepals, with or without any other perceptible change, is not particularly common. An increase has been observed in the Plum, Elder, Fuchsia, and Water Dropwort. It is not uncommon in Tulips, Iris, Narcissus, &c.

The corolla may have its petals multiplied irrespectively of any conversion of stamens into petals. There may be two or three corollas, one within another, as sometimes occurs in Campanulas. Double Stocks will occasionally have many series of petals and yet retain the six stamens and pistil in the centre unchanged and effective for setting seed.

* I have described and figured this in the Journal of the Linnean Society, vol. xix. p. 214, pl. 32.

In some flowers both calyx and corolla are repeated together over and over again, being slightly separated by short internodes on the axis which has elongated through the flower. This is the case with the flowers of a cultivated form of Arabis albida. It has occurred in Helianthemum vulgare and in a remarkable Mignonette issued by Mr. Balchin, called Reseda odorata prolifera alba.*

+These numbers are accounted for by the law of phyllotaxis, which need not be here discussed.

CybeRose: The author is alluding to the Fibonacci series: 1, 1, 2, 3, 5, 8, 13, 21, 34, 55...

Many Composites are said to be "double," but it is really a false name. Thus when the "single" Dahlia was first cultivated it was described as having only five petals, but is said to have soon shown eight; probably thirteen+ and twenty-one followed suit until the whole head was a mass of "petals." The interpretation of this is that the "head" consists not of petals but of florets; those of the "ray," on the circumference, having broad-limbed corollas, and those on the "disk" having minute five-toothed corollas. The number of ray-florets became increased by the conversion of tubular disk-florets into broad-petalled ray-florets.

When the tubular corolla with five petals changes to a large ray-floret, two of the petals are suppressed, while the remaining three become greatly enlarged. Simultaneously, the five stamens are arrested, the pistil remaining in a normal condition so that a completely "double" Composite is entirely female, and can only set seed by the aid of pollen from some more or less unchanged flowers.

In Centaureas, however, the trumpet-like ray-florets are so much enlarged that the pistil is now sacrificed as well as the stamens; hence the ray-florets are all neuter and functionless, except so far as they enhance the attractiveness of the head.

PROBABLE CAUSES OF "DOUBLING."—The primary cause of doubling appears to be any check to the vital activities of a plant; the principal one, at least in the wild state, being a very poor dry soil and a dry atmosphere.

Darwin wrote as follows in 1843 to the "Gardeners' Chronicle" about some plants of Gentiana Amarella: "The plants grow on a very hard bare chalk bank, where it was surprising that anything could grow at all. I found on an adjoining field of wretchedly sterile clay great numbers of Ranunculus repens producing semi-double flowers. The partial or entire sterility of double flowers is generally attributed to their doubleness, but is not this putting the effect before the cause? It is well known that plants when placed out of their natural conditions become sterile. Linnaeus has remarked that most Alpine plants, when cultivated in the lowlands, are rendered quite sterile. Is it, then, too bold a theory to suppose that all double flowers are first rendered, by some change in their natural condition, to a certain degree sterile?"

Observations on the conditions associated with doubling seem to corroborate this idea of Darwin's, for several instances have been recorded of wild plants becoming double under impoverished conditions of the sort mentioned.

A drier climate than that of England probably accounts for more double flowers being raised on the Continent than in England, as I learnt from the late Mr. Bull.

The Rev. C. Wolley-Dod and others have pointed out that the wild English Daffodil often becomes double when transferred to gardens, but that in his own cold and wet soil double-flowering plants tend to become single. Similarly, in a garden in Italy the owner could "keep no single Daffodils for more than a year or two, as they all turn double." The change of climate and soil appears to be the cause of doubling in these cases. Another writes saying that "the Tenby Daffodils (Narcissus obvallaris) have nearly all become double in this dry sandy soil (near Hitchin). N. 'Golden Spur' has also become double this season, 1904." Early double Tulips are best associated with a dry poor soil, as it is conducive to doubling, according to Mr. Houston.

Mr. Douglas's experience with Carnations is that potted plants produce more "doubles" than those grown in the open. "The finest lot of choice varieties was in a hot dry season." "In a German establishment thousands of flower-pots about five or six inches in diameter were arranged on a wooden staging fully exposed to the open air, and I was also informed that it was necessary to grow the plants in this way, to make sure of the seed producing a large percentage of double flowers." Mr. Douglas also observes that "the pink has a greater tendency to produce double flowers than any other plant known to me, and certainly the tendency is greater in rather exhausted soil, as can easily be proved by allowing a bed to sow itself from the previous year's bloom, and the plants to flower on the same ground without manuring, &c." As another illustration, the second and weaker growth of Peas, issuing from the lower part of the stem, bore double flowers. Hence an enfeebled constitution also appears in this case to have been the cause. Similarly, weak plants of Lapageria bore double flowers, while stronger plants grown with them bore only single flowers.

Frost affecting early shoots of Roses and early-blossoming Pelargoniums, as well as the reduced temperature of autumn, causes flowers of Apples and Pears to be more or less liable to doubling.

A check received by repotting Pelargoniums and weakly Cinerarias was noted to produce double flowers. Even too cold water, grubs of insects and minute fungi attacking flowers, thereby checking their development, have been credited with the same result.

Crossing and hybridising are said to encourage the appearance of double flowers. This is attributed to a slight want of harmony between the sexual organs of the parents, the consequence being deformed seeds. Such occurring in Leucojum, without any crossing, produced plants with double flowers. The well-nourished seeds of Leucojum developed only single flowers.

Stocks bearing defective sexual organs, due to drought, bore seeds which developed into double-flowering plants. Similarly, Camellia seeds raised in dry conditions gave the same result.

Hence it will appear that Darwin's acute surmise in 1843 is fully borne out by subsequent experiences. Doubling having been acquired, it may become congenital. Thus Lemoine crossed a semi-double Lilac with pollen from a single, and thirty out of forty seedlings were double or semi-double. Similarly, a hybrid between East Indian Rhododendrons showed a slightly petaloid anther. Mr. Heal pollinated it with its own pollen. The whole of the Balsamaefloral section is the result of that single impregnation.

As any influence which may affect the reproductive organs may cause them to attempt a metamorphosis, it is not surprising to find exactly the reverse conditions to drought may induce doubling where plants have been long habituated to conditions of dryness. Thus Kerria japonica has become double in Europe, including England, but is single in Japan. I have seen the single-flowered form in Sir T. Hanbury's garden at La Mortola, Italy, and in Tunbridge Wells (on sandstone).

Cardamine pratensis has been found to produce double flowers at two very wet places.

Reversions from doubles to singles occur when the conditions are changed. Thus double Daffodils become single in poor, cold and wet, or shaded soils.

Double Balsams when "starved" became single-flowered; but with a liberal treatment bore double flowers, &c. This shows that when "doubling" has been acquired it becomes constitutional, and is greatly improved by high nourishment. But this latter alone does not seem to he capable of bringing about the metamorphosis at first.

CARPELS

Useful Changes.—Wherever carpels change into any other organ they of course lose their function of setting iced as in completely double flowers, but if they become foliaceous they acquire all the functions of the leaves, as in the Green Rose, Double Cherry, &c.

Useless Changes.—Besides the carpels being replaced by petals, it sometimes happens that petaloid structures take the place of ovules. Such occurs occasionally in Cardamine pratensis and betrays itself by the fruit being globular instead of a long siliqua or pod. Again, stamens have been known to replace ovules, but these metamorphoses are by no means of common occurrence and are always abortive and useless structures.

SEXUAL CHANGES

Useful Changes.—In plants with distinct sexes in different flowers, whether they be both on the same plant (monoecious) or on separate plants (diocious), female flowers may appear mixed with or in place of male ones, and vice versa. Thus ripe grains of maize may be found intermixed with the staminate flowers of a male panicle. So too female flowers sometimes occur on the male catkins of the Spanish Chestnut. Conversely, anthers may occur in the female cones of Pines.

In unisexual plants external conditions may bring about a change of sex. Thus the males are usually borne by weaker shoots of trees, and a greater preponderance of males among herbs appears when there is a less amount of nourishment for all the seedlings growing thickly together.

Useless Changes.—The last kind of changes attempted by stamens is into carpels. This is not at all uncommon in certain plants. Thus in all large growings of Wallflowers, as for market purposes, there will generally be found "rogues "; they do not open their flowers, the petals being dwarfed and imperfect, while the six stamens are changed into abortive carpels, often partially open or more or less coherent with the pistil in the middle. They bear rudimentary but useless ovules.

Oranges grown in conservatories often bear small rudimentary carpels more or less coherent to the ovary at the base, and when this swells into the fruit the "pistilloid" stamens form small orange-coloured projections on the fruit.

*Other abnormalities will be found described in Dr. M. T. Masters's Vegetable Teratology.

Poppies, again, not infrequently have a circle of little abortive carpels round the base of the pistil. They have arisen by metamorphosis of the stamens.*