Darwinism (Wallace)/Chapter XI

by Alfred Russel Wallace

 

CHAPTER XI

 

THE SPECIAL COLOURS OF PLANTS: THEIR ORIGIN AND PURPOSE

The general colour relations of plants—Colours of fruits—The meaning of nuts—Edible or attractive fruits—The colours of flowers—Modes of securing cross-fertilisation—The interpretation of the facts—Summary of additional facts bearing on insect fertilisation—Fertilisation of flowers by birds—Self-fertilisation of flowers—Difficulties and contradictions—Intercrossing not necessarily advantageous—Supposed evil results of close interbreeding—How the struggle for existence acts among flowers—Flowers the product of insect agency—Concluding remarks on colour in nature.

The colours of plants are both less definite and less complex than are those of animals, and their interpretation on the principle of utility is, on the whole, more direct and more easy. Yet here, too, we find that in our investigation of the uses of the various colours of fruits and flowers, we are introduced to some of the most obscure recesses of nature's workshop, and are confronted with problems of the deepest interest and of the utmost complexity.

So much has been written on this interesting subject since Mr. Darwin first called attention to it, and its main facts have become so generally known by means of lectures, articles, and popular books, that I shall give here a mere outline sketch, for the purpose of leading up to a discussion of some of the more fundamental problems which arise out of the facts, and which have hitherto received less attention than they deserve.

The General Colour Relations of Plants.

The green colour of the foliage of leafy plants is due to the existence of a substance called chlorophyll, which is almost universally developed in the leaves under the action of light. It is subject to definite chemical changes during the processes of growth and of decay, and it is owing to these changes that we have the delicate tints of spring foliage, and the more varied, intense, and gorgeous hues of autumn. But these all belong to the class of intrinsic or normal colours, due to the chemical constitution of the organism; as colours they are unadaptive, and appear to have no more relation to the wellbeing of the plants themselves than have the colours of gems and minerals. We may also include in the same category those algae and fungi which have bright colours—the "red snow" of the arctic regions, the red, green, or purple seaweeds, the brilliant scarlet, yellow, white, or black agarics, and other fungi. All these colours are probably the direct results of chemical composition or molecular structure, and, being thus normal products of the vegetable organism, need no special explanation from our present point of view; and the same remark will apply to the varied tints of the bark of trunks, branches, and twigs, which are often of various shades of brown and green, or even vivid reds or yellows.

There are, however, a few cases in which the need of protection, which we have found to be so important an agency in modifying the colours of animals, has also determined those of some of the smaller members of the vegetable kingdom. Dr. Burchell found a mesembryanthomum in South Africa like a curiously shaped pebble, closely resembling the stones among which it grew;^[1] and Mr. J. P. Mansel Weale states that in the same country one of the Asclepiadeae has tubers growing above ground among stones which they exactly resemble, and that, when not in leaf, they are for this reason quite invisible.^[2] It is clear that such resemblances must be highly useful to these plants, inhabiting an arid country abounding in herbivorous mammalia, which, in times of drought or scarcity, will devour everything in the shape of a fleshy stem or tuber.

True mimicry is very rare in plants, though adaptation to like conditions often produces in foliage and habit a similarity that is deceiving. Euphorbias growing in deserts often closely resemble cacti. Seaside plants and high alpine plants of different orders are often much alike; and innumerable resemblances of this kind are recorded in the names of plants, as Veronica epacridea (the veronica like an epacris), Limnanthemum nymphaeoides (the limnanthemum like a nymphaea), the resembling species in each case belonging to totally distinct families. But in these cases, and in most others that have been observed, the essential features of true mimicry are absent, inasmuch as the one plant cannot be supposed to derive any benefit from its close resemblance to the other, and this is still more certain from the fact that the two species usually inhabit different localities. A few cases exist, however, in which there does seem to be the necessary accordance and utility. Mr. Mansel Weale mentions a labiate plant (Ajuga ophrydis), the only species of the genus Ajuga in South Africa, which is strikingly like an orchid of the same country; while a balsam (Impatiens capensis), also a solitary species of the genus in that country, is equally like an orchid, growing in the same locality and visited by the same insects. As both these genera of plants are specialised for insect fertilisation, and both of the plants in question are isolated species of their respective genera, we may suppose that, when they first reached South Africa they were neglected by the insects of the country; but, being both remotely like orchids in form of flower, those varieties that approached nearest to the familiar species of the country were visited by insects and cross-fertilised, and thus a closer resemblance would at length be brought about. Another case of close general resemblance, is that of our common white dead-nettle (Lamium album) to the stinging-nettle (Urtica dioica); and Sir John Lubbock thinks that this is a case of true mimicry, the dead-nettle being benefited by being mistaken by grazing animals for the stinging-nettle.^[3]

Colours of Fruits.

It is when we come to the essential parts of plants on which their perpetuation and distribution depends, that we find colour largely utilised for a distinct purpose in flowers and fruits. In the former we find attractive colours and guiding marks to secure cross-fertilisation by insects; in the latter attractive or protective coloration, the first to attract birds or other animals when the fruits are intended to be eaten, the second to enable them to escape being eaten when it would be injurious to the species. The colour phenomena of fruits being much the most simple will be considered first.

The perpetuation and therefore the very existence of each species of flowering plant depend upon its seeds being preserved from destruction and more or less effectually dispersed over a considerable area. The dispersal is effected either mechanically or by the agency of animals. Mechanical dispersal is chiefly by means of air-currents, and large numbers of seeds are specially adapted to be so carried, either by being clothed with down or pappus, as in the well-known thistle and dandelion seeds; by having wings or other appendages, as in the sycamore, birch, and many other trees; by being thrown to a considerable distance by the splitting of the seed-vessel, and by many other curious devices.^[4] Very large numbers of seeds, however, are so small and light that they can be carried enormous distances by gales of wind, more especially as most of this kind are flattened or curved, so as to expose a large surface in proportion to their weight. Those which are carried by animals have their surfaces, or that of the seed-vessel, armed with minute hooks, or some prickly covering which attaches itself to the hair of mammalia or the feathers of birds, as in the burdock, cleavers, and many other species. Others again are sticky, as in Plumbago europaea, mistletoe, and many foreign plants.

All the seeds or seed-vessels which are adapted to be dispersed in any of these ways are of dull protective tints, so that when they fall on the ground they are almost indistinguishable; besides which, they are usually small, hard, and altogether unattractive, never having any soft, juicy pulp; while the edible seeds often bear such a small proportion to the hard, dry envelopes or appendages, that few animals would care to eat them.

The Meaning of Nuts.

There is, however, another class of fruits or seeds, usually termed nuts, in which there is a large amount of edible matter, often very agreeable to the taste, and especially attractive and nourishing to a large number of animals. But when eaten, the seed is destroyed and the existence of the species endangered. It is evident, therefore, that it is by a kind of accident that these nuts are eatable; and that they are not intended to be eaten is shown by the special care nature seems to have taken to conceal or to protect them. We see that all our common nuts are green when on the tree, so as not easily to be distinguished from the leaves; but when ripe they turn brown, so that when they fall on to the ground they are equally indistinguishable among the dead leaves and twigs, or on the brown earth. Then they are almost always protected by hard coverings, as in hazel-nuts, which are concealed by the enlarged leafy involucre, and in the large tropical brazil-nuts and cocoa-nuts by such a hard and tough case as to be safe from almost every animal. Others have an external bitter rind, as in the walnut; while in the chestnuts and beech-nuts two or three fruits are enclosed in a prickly involucre.

Notwithstanding all these precautions, nuts are largely devoured by mammalia and birds; but as they are chiefly the product of trees or shrubs of considerable longevity, and are generally produced in great profusion, the perpetuation of the species is not endangered. In some cases the devourers of nuts may aid in their dispersal, as they probably now and then swallow the seed whole, or not sufficiently crushed to prevent germination; while squirrels have been observed to bury nuts, many of which are forgotten and afterwards grow in places they could not have otherwise reached.^[5] Nuts, especially the larger kinds which are so well protected by their hard, nearly globular cases, have their dispersal facilitated by rolling down hill, and more especially by floating in rivers and lakes, and thus reaching other localities. During the elevation of land areas this method would be very effective, as the new land would always be at a lower level than that already covered with vegetation, and therefore in the best position for being stocked with plants from it.

The other modes of dispersal of seeds are so clearly adapted to their special wants, that we feel sure they must have been acquired by the process of variation and natural selection. The hooked and sticky seeds are always those of such herbaceous plants as are likely, from their size, to come in contact with the wool of sheep or the hair of cattle; while seeds of this kind never occur on forest trees, on aquatic plants, or even on very dwarf creepers or trailers. The winged seed-vessels or seeds, on the other hand, mostly belong to trees and to tall shrubs or climbers. We have, therefore, a very exact adaptation to conditions in these different modes of dispersal; while, when we come to consider individual cases, we find innumerable other adaptations, some of which the reader will find described in the little work by Sir John Lubbock already referred to.

Edible or Attractive Fruits.

It is, however, when we come to true fruits (in a popular sense) that we find varied colours evidently intended to attract animals, in order that the fruits may be eaten, while the seeds pass through the body undigested and are then in the fittest state for germination. This end has been gained in a great variety of ways, and with so many corresponding adaptations as to leave no doubt as to the value of the result. Fruits are pulpy or juicy, and usually sweet, and form the favourite food of innumerable birds and some mammals. They are always coloured so as to contrast with the foliage or surroundings, red being the most common as it is certainly the most conspicuous colour, but yellow, purple, black, or white being not uncommon. The edible portion of fruits is developed from different parts of the floral envelopes, or of the ovary, in the various orders and genera. Sometimes the calyx becomes enlarged and fleshy, as in the apple and pear tribe; more often the integuments of the ovary itself are enlarged, as in the plum, peach, grape, etc.; the receptacle is enlarged and forms the fruit of the strawberry; while the mulberry, pineapple, and fig are examples of compound fruits formed in various ways from a dense mass of flowers.

In all cases the seeds themselves are protected from injury by various devices. They are small and hard in the strawberry, raspberry, currant, etc., and are readily swallowed among the copious pulp. In the grape they are hard and bitter; in the rose (hip) disagreeably hairy; in the orange tribe very bitter; and all these have a smooth, glutinous exterior which facilitates their being swallowed. When the seeds are larger and are eatable, they are enclosed in an excessively hard and thick covering, as in the various kinds of "stone" fruit (plums, peaches, etc.), or in a very tough core, as in the apple. In the nutmeg of the Eastern Archipelago we have a curious adaptation to a single group of birds. The fruit is yellow, somewhat like an oval peach, but firm and hardly eatable. This splits open and shows the glossy black covering of the seed or nutmeg, over which spreads the bright scarlet arillus or "mace," an adventitious growth of no use to the plant except to attract attention. Large fruit pigeons pluck out this seed and swallow it entire for the sake of the mace, while the large nutmeg passes through their bodies and germinates; and this has led to the wide distribution of wild nutmegs over New Guinea and the surrounding islands.

In the restriction of bright colour to those edible fruits the eating of which is beneficial to the plant, we see the undoubted result of natural selection; and this is the more evident when we find that the colour never appears till the fruit is ripe—that is, till the seeds within it are fully matured and in the best state for germination. Some brilliantly coloured fruits are poisonous, as in our bitter-sweet (Solanum dulcamara), cuckoo-pint (Arum) and the West Indian manchineel. Many of these are, no doubt, eaten by animals to whom they are harmless; and it has been suggested that even if some animals are poisoned by them the plant is benefited, since it not only gets dispersed, but finds, in the decaying body of its victim, a rich manure heap.^[6] The particular colours of fruits are not, so far as we know, of any use to them other than as regards conspicuousness, hence a tendency to any decided colour has been preserved and accumulated as serving to render the fruit easily visible among its surroundings of leaves or herbage. Out of 134 fruit-bearing plants in Mongredien's Trees and Shrubs, and Hooker's British Flora, the fruits of no less than sixty-eight, or rather more than half, are red, forty-five are black, fourteen yellow, and seven white. The great prevalence of red fruits is almost certainly due to their greater conspicuousness having favoured their dispersal, though it may also have arisen in part from the chemical changes of chlorophyll during ripening and decay producing red tints as in many fading leaves. Yet the comparative scarcity of yellow in fruits, while it is the most common tint of fading leaves, is against this supposition.

There are, however, a few instances of coloured fruits which do not seem to be intended to be eaten; such are the colocynth plant (Cucumis colocynthus), which has a beautiful fruit the size and colour of an orange, but nauseous beyond description to the taste. It has a hard rind, and may perhaps be dispersed by being blown along the ground, the colour being an adventitious product; but it is quite possible, notwithstanding its repulsiveness to us, that it may be eaten by some animals. With regard to the fruit of another plant, Calotropis procera, there is less doubt, as it is dry and full of thin, flat-winged seeds, with fine silky filaments, eminently adapted for wind-dispersal; yet it is of a bright yellow colour, as large as an apple, and therefore very conspicuous. Here, therefore, we seem to have colour which is a mere byproduct of the organism and of no use to it; but such cases are exceedingly rare, and this rarity, when compared with the great abundance of cases in which there is an obvious purpose in the colour, adds weight to the evidence in favour of the theory of the attractive coloration of edible fruits in order that birds and other animals may assist in their dispersal. Both the above-named plants are natives of Palestine and the adjacent arid countries.^[7]

The Colours of Flowers.

Flowers are much more varied in their colours than fruits, as they are more complex and more varied in form and structure; yet there is some parallelism between them in both respects. Flowers are frequently adapted to attract insects as fruits are to attract birds, the object being in the former to secure cross-fertilisation, in the latter dispersal; while just as colour is an index of the edibility of fruits which supply pulp or juice to birds, so are the colours of flowers an indication of the presence of nectar or of pollen which are devoured by insects.

The main facts and many of the details, as to the relation of insects to flowers, were discovered by Sprengel in 1793. He noticed the curious adaptation of the structure of many flowers to the particular insects which visit them; he proved that insects do cross-fertilise flowers, and he believed that this was the object of the adaptations, while the presence of nectar and pollen ensured the continuance of their visits; yet he missed discovering the use of this cross-fertilisation. Several writers at a later period obtained evidence that cross-fertilisation of plants was a benefit to them; but the wide generality of this fact and its intimate connection with the numerous and curious adaptations discovered by Sprengel, was first shown by Mr. Darwin, and has since been demonstrated by a vast mass of observations, foremost among which are his own researches on orchids, primulas, and other plants.^[8]

By an elaborate series of experiments carried on for many years Mr. Darwin demonstrated the great value of cross-fertilisation in increasing the rapidity of growth, the strength and vigour of the plant, and in adding to its fertility. This effect is produced immediately, not as he expected would be the case, after several generations of crosses. He planted seeds from cross-fertilised and self-fertilised plants on two sides of the same pot exposed to exactly similar conditions, and in most cases the difference in size and vigour was amazing, while the plants from cross-fertilised parents also produced more and finer seeds. These experiments entirely confirmed the experience of breeders of animals already referred to (p. 160), and led him to enunciate his famous aphorism, "Nature abhors perpetual self-fertilisation".^[9] In this principle we appear to have a sufficient reason for the various contrivances by which so many flowers secure cross-fertilisation, either constantly or occasionally. These contrivances are so numerous, so varied, and often so highly complex and extraordinary, that they have formed the subject of many elaborate treatises, and have also been amply popularised in lectures and handbooks. It will be unnecessary, therefore, to give details here, but the main facts will be summarised in order to call attention to some difficulties of the theory which seem to require further elucidation.

Modes of securing Cross-Fertilisation.

When we examine the various modes in which the cross-fertilisation of flowers is brought about, we find that some are comparatively simple in their operation and needful adjustments, others highly complex. The simple methods belong to four principal classes:—(1) By dichogamy—that is, by the anthers and the stigma becoming mature or in a fit state for fertilisation at slightly different times on the same plant. The result of this is that, as plants in different stations, on different soils, or exposed to different aspects flower earlier or later, the mature pollen of one plant can only fertilise some plant exposed to somewhat different conditions or of different constitution, whose stigma will be mature at the same time; and this difference has been shown by Darwin to be that which is adapted to secure the fullest benefit of cross-fertilisation. This occurs in Geranium pratense, Thymus serpyllum, Arum maculatum, and many others. (2) By the flower being self-sterile with its own pollen, as in the crimson flax. This absolutely prevents self-fertilisation. (3) By the stamens and anthers being so placed that the pollen cannot fall upon the stigma, while it does fall upon a visiting insect which carries it to the stigma of another flower. This effect is produced in a variety of very simple ways, and is often aided by the motion of the stamens which bend down out of the way of the stigmas before the pollen is ripe, as in Malva sylvestris (see Fig. 28). (4) By the male and female flowers being on different plants, forming the class Dioecia of Linnaeus. In these cases the pollen may be carried to the stigmas either by the wind or by the agency of insects.

FIG. 28. Malva sylvestris, adapted for insect-fertilisation. Malva rotundifolia, adapted for self-fertilisation.

Now these four methods are all apparently very simple, and easily produced by variation and selection. They are applicable to flowers of any shape, requiring only such size and colour as to attract insects, and some secretion of nectar to ensure their repeated visits, characters common to the great majority of flowers. All these methods are common, except perhaps the second; but there are many flowers in which the pollen from another plant is prepotent over the pollen from fertilisation, the same flower, and this has nearly the same effect as self-sterility if the flowers are frequently crossed by insects. We cannot help asking, therefore, why have other and much more elaborate methods been needed? And how have the more complex arrangements of so many flowers been brought about? Before attempting to answer these questions, and in order that the reader may appreciate the difficulty of the problem and the nature of the facts to be explained, it will be necessary to give a summary of the more elaborate modes of securing cross-fertilisation.

(1) We first have dimorphism and heteromorphism, the phenomena of which have been already sketched in our seventh chapter.

Here we have both a mechanical and a physiological modification, the stamens and pistil being variously modified in length and position, while the different stamens in the same flower have widely different degrees of fertility when applied to the same stigma,—a phenomenon which, if it were not so well established, would have appeared in the highest degree improbable. The most remarkable case is that of the three different forms of the loosestrife (Lythrum salicaria) here figured (Fig. 29 on next page).

(2) Some flowers have irritable stamens which, when their bases are touched by an insect, spring up and dust it with pollen. This occurs in our common berberry.

FIG. 29.—Lythrum salicaria (Purple loosestrife).

(3) In others there are levers or processes by which the anthers are mechanically brought down on to the head or back of an insect entering the flower, in such a position as to be carried to the stigma of the next flower it visits. This may be well seen in many species of Salvia and Erica.

(4) In some there is a sticky secretion which, getting on to the proboscis of an insect, carries away the pollen, and applies it to the stigma of another flower. This occurs in our common milkwort (Polygala vulgaris).

(5) In papilionaceous plants there are many complex adjustments, such as the squeezing out of pollen from a receptacle on to an insect, as in Lotus corniculatus, or the sudden springing out and exploding of the anthers so as thoroughly to dust the insect, as in Medicago falcata, this occurring after the stigma has touched the insect and taken off some pollen from the last flower.

(6) Some flowers or spathes form closed boxes in which insects find themselves entrapped, and when they have fertilised the flower, the fringe of hairs opens and allows them to escape. This occurs in many species of Arum and Aristolochia.

(7) Still more remarkable are the traps in the flower of Asclepias which catch flies, butterflies, and wasps by the legs, and the wonderfully complex arrangements of the orchids. One of these, our common Orchis pyramidalis, may be briefly described to show how varied and beautiful are the arrangements to secure cross-fertilisation. The broad trifid lip of the flower offers a support to the moth which is attracted by its sweet odour, and two ridges at the base guide the proboscis with certainty to the narrow entrance of the nectary. When the proboscis has reached the end of the spur, its basal portion depresses the little hinged rostellum that covers the saddle-shaped sticky glands to which the pollen masses (pollinia) are attached. On the proboscis being withdrawn, the two pollinia stand erect and parallel, firmly attached to the proboscis. In this position, however, they would be useless, as they would miss the stigmatic surface of the next flower visited by the moth. But as soon as the proboscis is withdrawn, the two pollen masses begin to diverge till they are exactly as far apart as are the stigmas of the flower; and then commences a second

FIG. 30.—Orchis pyramidalis.

movement which brings them down till they project straight forward nearly at right angles to their first position, so as exactly to hit against the stigmatic surfaces of the next flower visited on which they leave a portion of their pollen. The whole of these motions take about half a minute, and in that time the moth will usually have flown to another plant, and thus effect the most beneficial kind of cross-fertilisation.^[10] This description will be better understood by referring to the illustration opposite, from Darwin's Fertilisation of Orchids (Fig. 30).

The Interpretation of these Facts.

Having thus briefly indicated the general character of the more complex adaptations for cross-fertilisation, the details of which are to be found in any of the numerous works on the subject,^[11] we find ourselves confronted with the very puzzling question—Why were these innumerable highly complex adaptations produced, when the very same result may be effected—and often is effected—by extremely simple means? Supposing, as we must do, that all flowers were once of simple and regular forms, like a buttercup or a rose, how did such irregular and often complicated flowers as the papilionaceous or pea family, the labiates or sage family, and the infinitely varied and fantastic orchids ever come into existence? No cause has yet been suggested but the need of attracting insects to cross-fertilise them; yet the attractiveness of regular flowers with bright colours and an ample supply of nectar is equally great, and cross-fertilisation can be quite as effectively secured in these by any of the four simple methods already described. Before attempting to suggest a possible solution of this difficult problem, we have yet to pass in review a large body of curious adaptations connected with insect fertilisation, and will first call attention to that portion of the phenomena which throw some light upon the special colours of flowers in their relation to the various kinds of insects which visit them. For these facts we are largely indebted to the exact and long-continued researches of Professor Hermann Müller.

Summary of Additional Facts bearing on Insect Fertilisation.

1. That the size and colour of a flower are important factors in determining the visits of insects, is shown by the general fact of more insects visiting conspicuous than inconspicuous flowers. As a single instance, the handsome Geranium palustre was observed by Professor Müller to be visited by sixteen different species of insects, the equally showy G. pratense by thirteen species, while the smaller and much less conspicuous G. molle was visited by eight species, and G. pusillum by only one. In many cases, however, a flower may be very attractive to only a few species of insects; and Professor Müller states, as the result of many years' assiduous observation, that "a species of flower is the more visited by insects the more conspicuous it is."

2. Sweet odour is usually supplementary to the attraction of colour. Thus it is rarely present in the largest and most gaudily coloured flowers which inhabit open places, such as poppies, paeonies, sunflowers, and many others; while it is often the accompaniment of inconspicuous flowers, as the mignonette; of such as grow in shady places, as the violet and primrose; and especially of white or yellowish flowers, as the white jasmine, clematis, stephanotis, etc.

3. White flowers are often fertilised by moths, and very frequently give out their scent only by night, as in our butterfly-orchis (Habenaria chlorantha); and they sometimes open only at night, as do many of the evening primroses and other flowers. These flowers are often long tubed in accordance with the length of the moths' probosces, as in the genus Pancratium, our butterfly orchis, white jasmine, and a host of others.

4. Bright red flowers are very attractive to butterflies, and are sometimes specially adapted to be fertilised by them, as in many pinks (Dianthus deltoides, D. superbus, D. atrorubens), the corn-cockle (Lychnis Githago), and many others. Blue flowers are especially attractive to bees and other hymenoptera (though they frequent flowers of all colours), no less than sixty-seven species of this order having been observed to visit the common "sheep's-bit" (Jasione montana). Dull yellow or brownish flowers, some of which smell like carrion, are attractive to flies, as the Arum and Aristolochia; while the dull purplish flowers of the Scrophularia are specially attractive to wasps.

5. Some flowers have neither scent nor nectar, and yet attract insects by sham nectaries! In the herb-paris (Paris quadrifolia) the ovary glistens as if moist, and flies alight on it and carry away pollen to another flower; while in grass of parnassus (Parnassia palustris) there are a number of small stalked yellow balls near the base of the flower, which look like drops of honey but are really dry. In this case there is a little nectar lower down, but the special attraction is a sham; and as there are fresh broods of insects every year, it takes time for them to learn by experience, and thus enough are always deceived to effect cross-fertilisation.^[12] This is analogous to the case of the young birds, which have to learn by experience the insects that are inedible, as explained at page 253.

6. Many flowers change their colour as soon as fertilised; and this is beneficial, as it enables bees to avoid wasting time in visiting those blossoms which have been already fertilised and their nectar exhausted. The common lungwort (Pulmonaria officinalis), is at first red, but later turns blue; and H. Müller observed bees visiting many red flowers in succession, but neglecting the blue. In South Brazil there is a species of Lantana, whose flowers are yellow the first day, orange the second, and purple the third; and Dr. Fritz Müller observed that many butterflies visited the yellow flowers only, some both the yellow and the orange flowers, but none the purple.

7. Many flowers have markings which serve as guides to insects; in some cases a bright central eye, as in the borage and forget-me-not; or lines or spots converging to the centre, as in geraniums, pinks, and many others. This enables insects to go quickly and directly to the opening of the flower, and is equally important in aiding them to obtain a better supply of food, and to fertilise a larger number of flowers.

8. Flowers have been specially adapted to the kinds of insects that most abound where they grow. Thus the gentians of the lowlands are adapted to bees, those of the high alps to butterflies only; and while most species of Rhinanthus (a genus to which our common "yellow rattle" belongs) are bee-flowers, one high alpine species (R. alpinus) has been also adapted for fertilisation by butterflies only. The reason of this is, that in the high alps butterflies are immensely more plentiful than bees, and flowers adapted to be fertilised by bees can often have their nectar extracted by butterflies without effecting cross-fertilisation. It is, therefore, important to have a modification of structure which shall make butterflies the fertilisers, and this in many cases has been done.^[13]

9. Economy of time is very important both to the insects and the flowers, because the fine working days are comparatively few, and if no time is wasted the bees will get more honey, and in doing so will fertilise more flowers. Now, it has been ascertained by several observers that many insects, bees especially, keep to one kind of flower at a time, visiting hundreds of blossoms in succession, and passing over other species that may be mixed with them. They thus acquire quickness in going at once to the nectar, and the change of colour in the flower, or incipient withering when fertilised, enables them to avoid those flowers that have already had their honey exhausted. It is probably to assist the insects in keeping to one flower at a time, which is of vital importance to the perpetuation of the species, that the flowers which bloom intermingled at the same season are usually very distinct both in form and colour. In the sandy districts of Surrey, in the early spring, the copses are gay with three flowers—the primrose, the wood-anemone, and the lesser celandine, forming a beautiful contrast, while at the same time the purple and the white dead-nettles abound on hedge banks. A little later, in the same copses, we have the blue wild hyacinth (Scilla nutans), the red campion (Lychnis dioica), the pure white great starwort (Stellaria Holosteum), and the yellow dead-nettle (Lamium Galeobdolon), all distinct and well-contrasted flowers. In damp meadows in summer we have the ragged robin (Lychnis Floscuculi), the spotted orchis (O. maculata), and the yellow rattle (Rhinanthus Crista-galli); while in drier meadows we have cowslips, ox-eye daisies, and buttercups, all very distinct both in form and colour. So in cornfields we have the scarlet poppies, the purple corn-cockle, the yellow corn-marygold, and the blue cornflower; while on our moors the purple heath and the dwarf gorse make a gorgeous contrast. Thus the difference of colour which enables the insect to visit with rapidity and unerring aim a number of flowers of the same kind in succession, serves to adorn our meadows, banks, woods, and heaths with a charming variety of floral colour and form at each season of the year.^[14]

Fertilisation of Flowers by Birds.

In the temperate regions of the Northern Hemisphere, insects are the chief agents in cross-fertilisation when this is not effected by the wind; but in warmer regions, and in the Southern hemisphere, birds are found to take a considerable part in the operation, and have in many cases led to modifications in the form and colour of flowers. Each part of the globe has special groups of birds which are flower-haunters. America has the humming-birds (Trochilidae), and the smaller group of the sugar-birds (Caerebidae). In the Eastern tropics the sun-birds (Nectarineidae) take the place of the humming-birds, and another small group, the flower-peckers (Dicaeidae), assist them. In the Australian region there are also two flower-feeding groups, the Meliphagidae, or honey-suckers, and the brush-tongued lories (Trichoglossidae). Recent researches by American naturalists have shown that many flowers are fertilised by humming-birds, such as passion-flowers, trumpet-flowers, fuchsias, and lobelias; while some, as the Salvia splendens of Mexico, are specially adapted to their visits. We may thus perhaps explain the number of very large tubular flowers in the tropics, such as the huge brugmansias and bignonias; while in the Andes and in Chile, where humming-birds are especially plentiful, we find great numbers of red tubular flowers, often of large size and apparently adapted to these little creatures. Such are the beautiful Lapageria and Philesia, the grand Pitcairneas, and the genera Fuchsia, Mitraria, Embothrium, Escallonia, Desfontainea, Eccremocarpus, and many Gesneraceae. Among the most extraordinary modifications of flower structure adapted

FIG. 31.—Humming-bird fertilising Marcgravia nepenthoides.

to bird fertilisation are the species of Marcgravia, in which the pedicels and bracts of the terminal portion of a pendent bunch of flowers have been modified into pitchers which secrete nectar and attract insects, while birds feeding on the nectar, or insects, have the pollen of the overhanging flowers dusted on their backs, and, carrying it to other flowers, thus cross-fertilise them (see Illustration).

In Australia and New Zealand the fine "glory peas" (Clianthus), the Sophora, Loranthus, many Epacrideae and Myrtaceae, and the large flowers of the New Zealand flax (Phormium tenax), are cross-fertilised by birds; while in Natal the fine trumpet-creeper (Tecoma capensis) is fertilised by Nectarineas.

The great extent to which insect and bird agency is necessary to flowers is well shown by the case of New Zealand. The entire country is comparatively poor in species of insects, especially in bees and butterflies which are the chief flower fertilisers; yet according to the researches of local botanists no less than one-fourth of all the flowering plants are incapable of self-fertilisation, and, therefore, wholly dependent on insect or bird agency for the continuance of the species.

The facts as to the cross-fertilisation of flowers which have now been very briefly summarised, taken in connection with Darwin's experiments proving the increased vigour and fertility given by cross-fertilisation, seem amply to justify his aphorism that "Nature abhors self-fertilisation," and his more precise statement, that, "No plant is perpetually self-fertilised;" and this view has been upheld by Hildebrand, Delpino, and other botanists.^[15]

Self-Fertilisation of Flowers.

But all this time we have been only looking at one side of the question, for there exists an abundance of facts which seem to imply, just as surely, the utter uselessness of cross-fertilisation. Let us, then, see what these facts are before proceeding further.

1. An immense variety of plants are habitually self-fertilised, and their numbers probably far exceed those which are habitually cross-fertilised by insects. Almost all the very small or obscure flowered plants with hermaphrodite flowers are of this kind. Most of these, however, may be insect fertilised occasionally, and may, therefore, come under the rule that no species are perpetually self-fertilised.

2. There are many plants, however, in which special arrangements exist to secure self-fertilisation. Sometimes the corolla closes and brings the anthers and stigma into contact; in others the anthers cluster round the stigmas, both maturing together, as in many buttercups, stitchwort (Stellaria media), sandwort (Spergula), and some willow-herbs (Epilobium); or they arch over the pistil, as in Galium aparine and Alisma Plantago. The style is also modified to bring it into contact with the anthers, as in the dandelion, groundsel, and many other plants.^[16] All these, however, may be occasionally cross-fertilised.

3. In other cases precautions are taken to prevent cross-fertilisation, as in the numerous cleistogamous or closed flowers. These occur in no less than fifty-five different genera, belonging to twenty-four natural orders, and in thirty-two of these genera the normal flowers are irregular, and have therefore been specially modified for insect fertilisation.^[17] These flowers appear to be degradations of the normal flowers, and are closed up by various modifications of the petals or other parts, so that it is impossible for insects to reach the interior, yet they produce seed in abundance, and are often the chief means by which the species is continued. Thus, in our common dog-violet the perfect flowers rarely produce seed, while the rudimentary cleistogamic flowers do so in abundance. The sweet violet also produces abundance of seed from its cleistogamic flowers, and few from its perfect flowers; but in Liguria it produces only perfect flowers which seed abundantly. No case appears to be known of a plant which has cleistogamic flowers only, but a small rush (Juncus bufonius) is in this condition in some parts of Russia, while in other parts perfect flowers are also produced.^[18] Our common henbit dead-nettle (Lamium amplexicaule) produces cleistogamic flowers, as do also some orchids. The advantage gained by the plant is great economy of specialised material, since with very small flowers and very little expenditure of pollen an abundance of seed is produced.

4. A considerable number of plants which have evidently been specially modified for insect fertilisation have, by further modification, become quite self-fertile. This is the case with the garden-pea, and also with our beautiful bee-orchis, in which the pollen-masses constantly fall on to the stigmas, and the flower, being thus self-fertilised, produces abundance of capsules and of seed. Yet in many of its close allies insect agency is absolutely required; but in one of these, the fly-orchis, comparatively very little seed is produced, and self-fertilisation would therefore be advantageous to it. When garden-peas were artificially cross-fertilised by Mr. Darwin, it seemed to do them no good, as the seeds from these crosses produced less vigorous plants than seed from those which were self-fertilised; a fact directly opposed to what usually occurs in cross-fertilised plants.

5. As opposed to the theory that there is any absolute need for cross-fertilisation, it has been urged by Mr. Henslow and others that many self-fertilised plants are exceptionally vigorous, such as groundsel, chickweed, sow-thistle, buttercups, and other common weeds; while most plants of world-wide distribution are self-fertilised, and these have proved themselves to be best fitted to survive in the battle of life. More than fifty species of common British plants are very widely distributed, and all are habitually self-fertilised.^[19] That self-fertilisation has some great advantage is shown by the fact that it is usually the species which have the smallest and least conspicuous flowers which have spread widely, while the large and showy flowered species of the same genera or families, which require insects to cross-fertilise them, have a much more limited distribution.

6. It is now believed by some botanists that many inconspicuous and imperfect flowers, including those that are wind-fertilised, such as plantains, nettles, sedges, and grasses, do not represent primitive or undeveloped forms, but are degradations from more perfect flowers which were once adapted to insect fertilisation. In almost every order we find some plants which have become thus reduced or degraded for wind or self-fertilisation, as Poterium and Sanguisorba among the Rosaceae; while this has certainly been the case in the cleistogamic flowers. In most of the above-mentioned plants there are distinct rudiments of petals or other floral organs, and as the chief use of these is to attract insects, they could hardly have existed in primitive flowers.^[20] We know, moreover, that when the petals cease to be required for the attraction of insects, they rapidly diminish in size, lose their bright colour or almost wholly disappear.^[21]

Difficulties and Contradictions.

The very bare summary that has now been given of the main facts relating to the fertilisation of flowers, will have served to show the vast extent and complexity of the inquiry, and the extraordinary contradictions and difficulties which it presents. We have direct proof of the beneficial results of intercrossing in a great number of cases; we have an overwhelming mass of facts as to the varied and complex structure of flowers evidently adapted to secure this intercrossing by insect agency; yet we see many of the most vigorous plants which spread widely over the globe, with none of these adaptations, and evidently depending on self-fertilisation for their continued existence and success in the battle of life. Yet more extraordinary is it to find numerous cases in which the special arrangements for cross-fertilisation appear to have been a failure, since they have either been supplemented by special means for self-fertilisation, or have reverted back in various degrees to simpler forms in which self-fertilisation becomes the rule. There is also a further difficulty in the highly complex modes by which cross-fertilisation is often brought about; for we have seen that there are several very effective yet very simple modes of securing intercrossing, involving a minimum of change in the form and structure of the flower; and when we consider that the result attained with so much cost of structural modification is by no means an unmixed good, and is far less certain in securing the perpetuation of the species than is self-fertilisation, it is most puzzling to find such complex methods resorted to, sometimes to the extent of special precautions against the possibility of self-fertilisation ever taking place. Let us now see whether any light can be thrown on these various anomalies and contradictions.

Intercrossing not necessarily Advantageous.

No one was more fully impressed than Mr. Darwin with the beneficial effects of intercrossing on the vigour and fertility of the species or race, yet he clearly saw that it was not always and necessarily advantageous. He says: "The most important conclusion at which I have arrived is, that the mere act of intercrossing by itself does no good. The good depends on the individuals which are crossed differing slightly in constitution, owing to their progenitors having been subjected during several generations to slightly different conditions. This conclusion, as we shall hereafter see, is closely connected with various important physiological problems, such as the benefit derived from slight changes in the conditions of life."^[22] Mr. Darwin has also adduced much direct evidence proving that slight changes in the conditions of life are beneficial to both animals and plants, maintaining or restoring their vigour and fertility in the same way as a favourable cross seems to restore it.^[23] It is, I believe, by a careful consideration of these two classes of facts that we shall find the clue to the labyrinth in which this subject has appeared to involve us.

Supposed Evil Results of Close Interbreeding.

Just as we have seen that intercrossing is not necessarily good, we shall be forced to admit that close interbreeding is not necessarily bad. Our finest breeds of domestic animals have been thus produced, and by a careful statistical inquiry Mr. George Darwin has shown that the most constant and long-continued intermarriages among the British aristocracy have produced no prejudicial results. The rabbits on Porto Santo are all the produce of a single female; they have lived on the same small island for 470 years, and they still abound there and appear to be vigorous and healthy (see p. 161).

We have, however, on the other hand, overwhelming evidence that in many cases, among our domestic animals and cultivated plants, close interbreeding does produce bad results, and the apparent contradiction may perhaps be explained on the same general principles, and under similar limitations, as were found to be necessary in defining the value of intercrossing. It appears probable, then, that it is not interbreeding in itself that is hurtful, but interbreeding without rigid selection or some change of conditions. Under nature, as in the case of the Porto Santo rabbits, the rapid increase of these animals would in a very few years stock the island with a full population, and thereafter natural selection would act powerfully in the preservation only of the healthiest and the most fertile, and under these conditions no deterioration would occur. Among the aristocracy there has been a constant selection of beauty, which is generally synonymous with health, while any constitutional infertility has led to the extinction of the family. With domestic animals the selection practised is usually neither severe enough nor of the right kind. There is no natural struggle for existence, but certain points of form and colour characteristic of the breed are considered essential, and thus the most vigorous or the most fertile are not always those which are selected to continue the stock. In nature, too, the species always extends over a larger area and consists of much greater numbers, and thus a difference of constitution soon arises in different parts of the area, which is wanting in the limited numbers of pure bred domestic animals. From a consideration of these varied facts we conclude that an occasional disturbance of the organic equilibrium is what is essential to keep up the vigour and fertility of any organism, and that this disturbance may be equally well produced either by a cross between individuals of somewhat different constitutions, or by occasional slight changes in the conditions of life. Now plants which have great powers of dispersal enjoy a constant change of conditions, and can, therefore, exist permanently, or at all events, for very long periods, without intercrossing; while those which have limited powers of dispersal, and are restricted to a comparatively small and uniform area, need an occasional cross to keep up their fertility and general vigour. We should, therefore, expect that those groups of plants which are adapted both for cross-and self-fertilisation, which have showy flowers and possess great powers of seed-dispersal, would be the most abundant and most widely distributed; and this we find to be the case, the Compositae possessing all these characteristics in the highest degree, and being the most generally abundant group of plants with conspicuous flowers in all parts of the world.

How the Struggle for Existence Acts among Flowers.

Let us now consider what will be the action of the struggle for existence under the conditions we have seen to exist.

Everywhere and at all times some species of plants will be dominant and aggressive; while others will be diminishing in numbers, reduced to occupy a smaller area, and generally having a hard struggle to maintain themselves. Whenever a self-fertilising plant is thus reduced in numbers it will be in danger of extinction, because, being limited to a small area, it will suffer from the effects of too uniform conditions which will produce weakness and infertility. But while this change is in progress, any crosses between individuals of slightly different constitution will be beneficial, and all variations favouring either insect agency on the one hand, or wind-dispersal of pollen on the other, will lead to the production of a somewhat stronger and more fertile stock. Increased size or greater brilliancy of the flower, more abundant nectar, sweeter odour, or adaptations for more effectual cross-fertilisation would all be preserved, and thus would be initiated some form of specialisation for insect agency in cross-fertilisation; and in every different species so circumstanced the result would be different, depending as it would on many and complex combinations of variation of parts of the flower, and of the insect species which most abounded in the district.

Species thus favourably modified might begin a new era of development, and, while spreading over a somewhat wider area, give rise to new varieties or species, all adapted in various degrees and modes to secure cross-fertilisation by insect agency. But in course of ages some change of conditions might prove adverse. Either the insects required might diminish in numbers or be attracted by other competing flowers, or a change of climate might give the advantage to other more vigorous plants. Then self-fertilisation with greater means of dispersal might be more advantageous; the flowers might become smaller and more numerous; the seeds smaller and lighter so as to be more easily dispersed by the wind, while some of the special adaptations for insect fertilisation being useless would, by the absence of selection and by the law of economy of growth, be reduced to a rudimentary form. With these modifications the species might extend its range into new districts, thereby obtaining increased vigour by the change of conditions, as appears to have been the case with so many of the small flowered self-fertilised plants. Thus it might continue to exist for a long series of ages, till under other changes—geographical or biological—it might again suffer from competition or from other adverse circumstances, and be at length again confined to a limited area, or reduced to very scanty numbers.

But when this cycle of change had taken place, the species would be very different from the original form. The flower would have been at one time modified to favour the visits of insects and to secure cross-fertilisation by their aid, and when the need for this passed away, some portions of these structures would remain, though in a reduced or rudimentary condition. But when insect agency became of importance a second time, the new modifications would start from a different or more advanced basis, and thus a more complex result might be produced. Owing to the unequal rates at which the reduction of the various parts might occur, some amount of irregularity in the flower might arise, and on a second development towards insect cross-fertilisation this irregularity, if useful, might be increased by variation and selection.

The rapidity and comparative certainty with which such changes as are here supposed do really take place, are well shown by the great differences in floral structure, as regards the mode of fertilisation, in allied genera and species, and even in some cases in varieties of the same species. Thus in the Ranunculaceae we find the conspicuous part of the flower to be the petals in Ranunculus, the sepals in Helleborus, Anemone, etc., and the stamens in most species of Thalictrum. In all these we have a simple regular flower, but in Aquilegia it is made complex by the spurred petals, and in Delphinium and Aconitum it becomes quite irregular. In the more simple class self-fertilisation occurs freely, but it is prevented in the more complex flowers by the stamens maturing before the pistil. In the Caprifoliaceae we have small and regular greenish flowers, as in the moschatel (Adoxa); more conspicuous regular open flowers without honey, as in the elder (Sambucus); and tubular flowers increasing in length and irregularity, till in some, like our common honeysuckle, they are adapted for fertilisation by moths only, with abundant honey and delicious perfume to attract them. In the Scrophulariaceae we find open, almost regular flowers, as Veronica and Verbascum, fertilised by flies and bees, but also self-fertilised; Scrophularia adapted in form and colour to be fertilised by wasps; and the more complex and irregular flowers of Linaria, Rhinanthus, Melampyrum, Pedicularis, etc., mostly adapted to be fertilised by bees.

In the genera Geranium, Polygonum, Veronica, and several others there is a gradation of forms from large and bright to small and obscure coloured flowers, and in every case the former are adapted for insect fertilisation, often exclusively, while in the latter self-fertilisation constantly occurs. In the yellow rattle (Rhinanthus Crista-galli) there are two forms (which have been named major and minor), the larger and more conspicuous adapted to insect fertilisation only, the smaller capable of self-fertilisation; and two similar forms exist in the eyebright (Euphrasia officinalis). In both these cases there are special modifications in the length and curvature of the style as well as in the size and shape of the corolla; and the two forms are evidently becoming each adapted to special conditions, since in some districts the one, in other districts the other is most abundant.^[24]

These examples show us that the kind of change suggested above is actually going on, and has presumably always been going on in nature throughout the long geological epochs during which the development of flowers has been progressing. The two great modes of gaining increased vigour and fertility—intercrossing and dispersal over wider areas—have been resorted to again and again, under the pressure of a constant struggle for existence and the need for adaptation to ever-changing conditions. During all the modifications that ensued, useless parts were reduced or suppressed, owing to the absence of selection and the principle of economy of growth; and thus at each fresh adaptation some rudiments of old structures were re-developed, but not unfrequently in a different form and for a distinct purpose.

The chief types of flowering plants have existed during the millions of ages of the whole tertiary period, and during this enormous lapse of time many of them may have been modified in the direction of insect fertilisation, and again into that of self-fertilisation, not once or twice only, but perhaps scores or even hundreds of times; and at each such modification a difference in the environment may have led to a distinct line of development. At one epoch the highest specialisation of structure in adaptation to a single species or group of insects may have saved a plant from extinction; while, at other times, the simplest mode of self-fertilisation, combined with greater powers of dispersal and a constitution capable of supporting diverse physical conditions, may have led to a similar result. With some groups the tendency seems to have been almost continuously to greater and greater specialisation, while with others a tendency to simplification and degradation has resulted in such plants as the grasses and sedges.

We are now enabled dimly to perceive how the curious anomaly of very simple and very complex methods of securing cross-fertilisation—both equally effective—may have been brought about. The simple modes may be the result of a comparatively direct modification from the more primitive types of flowers, which were occasionally, and, as it were, accidentally visited and fertilised by insects; while the more complex modes, existing for the most part in the highly irregular flowers, may result from those cases in which adaptation to insect-fertilisation, and partial or complete degradation to self-fertilisation or to wind-fertilisation, have again and again recurred, each time producing some additional complexity, arising from the working up of old rudiments for new purposes, till there have been reached the marvellous flower structures of the papilionaceous tribes, of the asclepiads, or of the orchids.

We thus see that the existing diversity of colour and of structure in flowers is probably the ultimate result of the ever-recurring struggle for existence, combined with the ever-changing relations between the vegetable and animal kingdoms during countless ages. The constant variability of every part and organ, with the enormous powers of increase possessed by plants, have enabled them to become again and again readjusted to each change of condition as it occurred, resulting in that endless variety, that marvellous complexity, and that exquisite colouring which excite our admiration in the realm of flowers, and constitute them the perennial charm and crowning glory of nature.

Flowers the Product of Insect Agency.

In his Origin of Species, Mr. Darwin first stated that flowers had been rendered conspicuous and beautiful in order to attract insects, adding: "Hence we may conclude that, if insects had not been developed on the earth, our plants would not have been decked with beautiful flowers, but would have produced only such poor flowers as we see on our fir, oak, nut, and ash trees, on grasses, docks, and nettles, which are all fertilised through the agency of the wind." The argument in favour of this view is now much stronger than when he wrote; for not only have we reason to believe that most of these wind-fertilised flowers are degraded forms of flowers which have once been insect fertilised, but we have abundant evidence that whenever insect agency becomes comparatively ineffective, the colours of the flowers become less bright, their size and beauty diminish, till they are reduced to such small, greenish, inconspicuous flowers as those of the rupture-wort (Herniaria glabra), the knotgrass (Polygonum aviculare), or the cleistogamic flowers of the violet. There is good reason to believe, therefore, not only that flowers have been developed in order to attract insects to aid in their fertilisation, but that, having been once produced, in however great profusion, if the insect races were all to become extinct, flowers (in the temperate zones at all events) would soon dwindle away, and that ultimately all floral beauty would vanish from the earth.

We cannot, therefore, deny the vast change which insects have produced upon the earth's surface, and which has been thus forcibly and beautifully delineated by Mr. Grant Allen: "While man has only tilled a few level plains, a few great river valleys, a few peninsular mountain slopes, leaving the vast mass of earth untouched by his hand, the insect has spread himself over every land in a thousand shapes, and has made the whole flowering creation subservient to his daily wants. His buttercup, his dandelion, and his meadow-sweet grow thick in every English field. His thyme clothes the hillside; his heather purples the bleak gray moorland. High up among the alpine heights his gentian spreads its lakes of blue; amid the snows of the Himalayas his rhododendrons gleam with crimson light. Even the wayside pond yields him the white crowfoot and the arrowhead, while the broad expanses of Brazilian streams are beautified by his gorgeous water-lilies. The insect has thus turned the whole surface of the earth into a boundless flower-garden, which supplies him from year to year with pollen or honey, and itself in turn gains perpetuation by the baits that it offers for his allurement."^[25]

Concluding Remarks on Colour in Nature.

In the last four chapters I have endeavoured to give a general and systematic, though necessarily condensed view of the part which is played by colour in the organic world. We have seen in what infinitely varied ways the need of concealment has led to the modification of animal colours, whether among polar snows or sandy deserts, in tropical forests or in the abysses of the ocean. We next find these general adaptations giving way to more specialised types of coloration, by which each species has become more and more harmonised with its immediate surroundings, till we reach the most curiously minute resemblances to natural objects in the leaf and stick insects, and those which are so like flowers or moss or birds' droppings that they deceive the acutest eye. We have learnt, further, that these varied forms of protective colouring are far more numerous than has been usually suspected, because, what appear to be very conspicuous colours or markings when the species is observed in a museum or in a menagerie, are often highly protective when the creature is seen under the natural conditions of its existence. From these varied classes of facts it seems not improbable that fully one-half of the species in the animal kingdom possess colours which have been more or less adapted to secure for them concealment or protection.

Passing onward we find the explanation of a distinct type of colour or marking, often superimposed upon protective tints, in the importance of easy recognition by many animals of their fellows, their parents, or their mates. By this need we have been able to account for markings that seem calculated to make the animal conspicuous, when the general tints and well-known habits of the whole group demonstrate the need of concealment. Thus also we are able to explain the constant symmetry in the markings of wild animals, as well as the numerous cases in which the conspicuous colours are concealed when at rest and only become visible during rapid motion.

In striking contrast to ordinary protective coloration we have "warning colours," usually very conspicuous and often brilliant or gaudy, which serve to indicate that their possessors are either dangerous or uneatable to the usual enemies of their tribe. This kind of coloration is probably more prevalent than has been hitherto supposed, because in the case of many tropical animals we are quite unacquainted with their special and most dangerous enemies, and are also unable to determine whether they are or are not distasteful to those enemies. As a kind of corollary to the "warning colours," we find the extraordinary phenomena of "mimicry," in which defenceless species obtain protection by being mistaken for those which, from any cause, possess immunity from attack. Although a large number of instances of warning colour and of mimicry are now recorded, it is probably still an almost unworked field of research, more especially in tropical regions and among the inhabitants of the ocean.

The phenomena of sexual diversities of coloration next engaged our attention, and the reasons why Mr. Darwin's theory of "sexual selection," as regards colour and ornament, could not be accepted were stated at some length, together with the theory of animal coloration and ornament we propose to substitute for it. This theory is held to be in harmony with the general facts of animal coloration, while it entirely dispenses with the very hypothetical and inadequate agency of female choice in producing the detailed colours, patterns, and ornaments, which in so many cases distinguish the male sex.

If my arguments on this point are sound, they will dispose also of Mr. Grant Allen's view of the direct action of the colour sense on the animal integuments.^[26] He argues that the colours of insects and birds reproduce generally the colours of the flowers they frequent or the fruits they eat, and he adduces numerous cases in which flower-haunting insects and fruit-eating birds are gaily coloured. This he supposes to be due to the colour-taste, developed by the constant presence of bright flowers and fruits, being applied to the selection of each variation towards brilliancy in their mates; thus in time producing the gorgeous and varied hues they now possess. Mr. Allen maintains that "insects are bright where bright flowers exist in numbers, and dull where flowers are rare or inconspicuous;" and he urges that "we can hardly explain this wide coincidence otherwise than by supposing that a taste for colour is produced through the constant search for food among entomophilous blossoms, and that this taste has reacted upon its possessors through the action of unconscious sexual selection."

The examples Mr. Allen quotes of bright insects being associated with bright flowers seem very forcible, but are really deceptive or erroneous; and quite as many cases could be quoted which prove the very opposite. For example, in the dense equatorial forests flowers are exceedingly scarce, and there is no comparison with the amount of floral colour to be met with in our temperate meadows, woods, and hillsides. The forests about Para in the lower Amazon are typical in this respect, yet they abound with the most gorgeously coloured butterflies, almost all of which frequent the forest depths, keeping near the ground, where there is the greatest deficiency of brilliant flowers. In contrast with this let us take the Cape of Good Hope—the most flowery region probably that exists upon the globe,—where the country is a complete flower-garden of heaths, pelargoniums, mesembryanthemus, exquisite iridaceous and other bulbs, and numerous flowering shrubs and trees; yet the Cape butterflies are hardly equal, either in number or variety, to those of any country in South Europe, and are utterly insignificant when compared with those of the comparatively flowerless forest-depths of the Amazon or of New Guinea. Neither is there any relation between the colours of other insects and their haunts. Few are more gorgeous than some of the tiger-beetles and the carabi, yet these are all carnivorous; while many of the most brilliant metallic buprestidae and longicorns are always found on the bark of fallen trees. So with the humming-birds; their brilliant metallic tints can only be compared with metals or gems, and are totally unlike the delicate pinks and purples, yellows and reds of the majority of flowers. Again, the Australian honey-suckers (Meliphagidae) are genuine flower-haunters, and the Australian flora is more brilliant in colour display than that of most tropical regions, yet these birds are, as a rule, of dull colours, not superior on the average to our grain-eating finches. Then, again, we have the grand pheasant family, including the gold and the silver pheasants, the gorgeous fire-backed and ocellated pheasants, and the resplendent peacock, all feeding on the ground on grain or seeds or insects, yet adorned with the most gorgeous colours.

There is, therefore, no adequate basis of facts for this theory to rest upon, even if there were the slightest reason to believe that not only birds, but butterflies and beetles, take any delight in colour for its own sake, apart from the food-supply of which it indicates the presence. All that has been proved or that appears to be probable is, that they are able to perceive differences of colour, and to associate each colour with the particular flowers or fruits which best satisfy their wants. Colour being in its nature diverse, it has been beneficial for them to be able to distinguish all its chief varieties, as manifested more particularly in the vegetable kingdom, and among the different species of their own group; and the fact that certain species of insects show some preference for a particular colour may be explained by their having found flowers of that colour to yield them a more abundant supply of nectar or of pollen. In those cases in which butterflies frequent flowers of their own colour, the habit may well have been acquired from the protection it affords them.

It appears to me that, in imputing to insects and birds the same love of colour for its own sake and the same aesthetic tastes as we ourselves possess, we may be as far from the truth as were those writers who held that the bee was a good mathematician, and that the honeycomb was constructed throughout to satisfy its refined mathematical instincts; whereas it is now generally admitted to be the result of the simple principle of economy of material applied to a primitive cylindrical cell.^[27]

In studying the phenomena of colour in the organic world we have been led to realise the wonderful complexity of the adaptations which bring each species into harmonious relation with all those which surround it, and which thus link together the whole of nature in a network of relations of marvellous intricacy. Yet all this is but, as it were, the outward show and garment of nature, behind which lies the inner structure—the framework, the vessels, the cells, the circulating fluids, and the digestive and reproductive processes,—and behind these again those mysterious chemical, electrical, and vital forces which constitute what we term Life. These forces appear to be fundamentally the same for all organisms, as is the material of which all are constructed; and we thus find behind the outer diversities an inner relationship which binds together the myriad forms of life.

Each species of animal or plant thus forms part of one harmonious whole, carrying in all the details of its complex structure the record of the long story of organic development; and it was with a truly inspired insight that our great philosophical poet apostrophised the humble weed—

Flower in the crannied wall,
I pluck you out of the crannies,
I hold you here, root and all, in my hand,
Little flower—but if I could understand
What you are, root and all, and all in all,
I should know what God and man is.

1. Burchell's Travels, vol. i. p. 10.
2. Nature, vol. iii. p. 507.
3. Flowers, Fruits, and Leaves, p. 128 (Fig. 79).
4. For a popular sketch of these, see Sir J. Lubbock's Flowers, Fruits, and Leaves, or any general botanical work.
5. Nature, vol. xv. p, 117.
6. Grant Allen's Colour Sense, p. 113.
7. Canon Tristram's Natural History of the Bible, pp. 483, 484.
8. For a complete historical account of this subject with full references to all the works upon it, see the Introduction to Hermann Müller's Fertilisation of Flowers, translated by D'Arcy W. Thompson.
9. For the full detail of his experiments, see Cross-and Self-Fertilisation of Plants, 1876.
10. See Darwin's Fertilisation of Orchids for the many extraordinary and complex arrangements in these plants.
11. The English reader may consult Sir John Lubbock's British Wild Flowers in Relation to Insects, and H. Müller's great and original work, The Fertilisation of Flowers.
12. Müller's Fertilisation of Flowers, p. 248.
13. "Alpenblumen," by D.H. Müller. See Nature, vol. xxiii. p. 333.
14. This peculiarity of local distribution of colour in flowers may be compared, as regards its purpose, with the recognition colours of animals. Just as these latter colours enable the sexes to recognise each other, and thus avoid sterile unions of distinct species, so the distinctive form and colour of each species of flower, as compared with those that usually grow around it, enables the fertilising insects to avoid carrying the pollen of one flower to the stigma of a distinct species.
15. See H. Müller's Fertilisation of Flowers, p. 18.
16. The above examples are taken from Rev. G. Henslow's paper on "Self-Fertilisation of Plants," in Trans. Linn. Soc. Second series, Botany, vol. i. pp. 317-398, with plate. Mr. H. O. Forbes has shown that the same thing occurs among tropical orchids, in his paper "On the Contrivances for insuring Self-Fertilisation in some Tropical Orchids," Journ. Linn. Soc., xxi. p. 538.
17. These are the numbers given by Darwin, but I am informed by Mr. Hemsley that many additions have been since made to the list, and that cleistogamic flowers probably occur in nearly all the natural orders.
18. For a full account of cleistogamic flowers, see Darwin's Forms of Flowers, chap. viii.
19. Henslow's "Self-Fertilisation," Trans. Linn. Soc. Second series, Botany, vol. i. p. 391.
20. The Rev. George Henslow, in his Origin of Floral Structures, says: "There is little doubt but that all wind-fertilised angiosperms are degradations from insect-fertilised flowers.... Poterium sanguisorba is anemophilous; and Sanguisorba officinalis presumably was so formerly, but has reacquired an entomophilous habit; the whole tribe Poterieae being, in fact, a degraded group which has descended from Potentilleae. Plantains retain their corolla but in a degraded form. Junceae are degraded Lilies; while Cyperaceae and Gramineae among monocotyledons may be ranked with Amentiferae among dicotyledons, as representing orders which have retrograded very far from the entomophilous forms from which they were possibly and probably descended" (p. 266).
    "The genus Plantago, like Thalictrum minus, Poterium, and others, well illustrate the change from an entomophilous to the anemophilous state. P. lanceolata has polymorphic flowers, and is visited by pollen-seeking insects, so that it can be fertilised either by insects or the wind. P. media illustrates transitions in point of structure, as the filaments are pink, the anthers motionless, and the pollen grains aggregated, and it is regularly visited by Bombus terrestris. On the other hand, the slender filaments, versatile anthers, powdery pollen, and elongated protogynous style are features of other species indicating anemophily; while the presence of a degraded corolla shows its ancestors to have been entomophilous. P. media, therefore, illustrates, not a primitive entomophilous condition, but a return to it; just as is the case with Sanguisorba officinalis and Salix Caprea; but these show no capacity of restoring the corolla, the attractive features having to be borne by the calyx, which is purplish in Sanguisorba, by the pink filaments of Plantago, and by the yellow anthers in the Sallow willow" (p. 271).
    "The interpretation, then, I would offer of inconspicuousness and all kinds of degradations is the exact opposite to that of conspicuousness and great differentiations; namely, that species with minute flowers, rarely or never visited by insects, and habitually self-fertilised, have primarily arisen through the neglect of insects, and have in consequence assumed their present floral structures" (p. 282).
    In a letter just received from Mr. Henslow, he gives a few additional illustrations of his views, of which the following are the most important: "Passing to Incompletae, the orders known collectively as 'Cyclospermeae' are related to Caryophylleae; and to my mind are degradations from it, of which Orache is anemophilous. Cupuliferae have an inferior ovary and rudimentary calyx-limb on the top. These, as far as I know, cannot be interpreted except as degradations. The whole of Monocotyledons appear to me (from anatomical reasons especially) to be degradations from Dicotyledons, and primarily through the agency of growth in water. Many subsequently became terrestrial, but retained the effects of their primitive habitat through heredity. The 3-merous [sic] perianth of grasses, the parts of the flower being in whorls, point to a degradation from a sub-liliaceous condition." Mr. Henslow informs me that he has long held these views, but, as far as he knows, alone. Mr. Grant Allen, however, set forth a similar theory in his Vignettes from Nature (p. 15) and more fully in The Colours of Flowers (chap. v.), where he develops it fully and uses similar arguments to those of Mr. Henslow.
21. H. Müller gives ample proof of this in his Fertilisation of Flowers.
22. Cross- and Self-Fertilisation, p. 27.
23. Animals and Plants, vol. ii. p. 145.
24. Müller's Fertilisation of Flowers, pp. 448, 455. Other cases of recent degradation and readaptation to insect-fertilisation are given by Professor Henslow (see footnote, p. 324).
25. The Colour Sense, by Grant Allen, p. 95.
26. The Colour Sense, chap. ix.
27. See Origin of Species, sixth edition, p. 220.