Chapter IV: Variation and Natural Selection (1)

Everything, so far as in it lies, said Benedict Spinoza, tends to persist in its own being. This is _the law of persistence_. It forms the basis of Newton's First Law of Motion, which enunciates that, if a body be at rest, it will remain so unless acted on by some external force; or, if it be in motion, it will continue to move in the same straight line and at a uniform velocity unless it is acted on by some external force. Practically every known body is thus affected by external forces; but the law of persistence is not thereby disproved. It only states what would happen under certain exceptional or perhaps impossible circumstances. To those ignorant of scientific procedure, it seems unsatisfactory, if not ridiculous, to formulate laws of things, not as they are, but as they might be. Many well-meaning but not very well-informed people thus wholly misunderstand and mistake the value of certain laws of political economy, because in those laws (which are generalized statements of fact under narrowed and rigid conditions, and do not pretend to be inculcated as rules of conduct) benevolence, sentiment, even moral and religious duty, are intentionally excluded. These laws state that men, under motives arising out of the pursuit of wealth, will act in such and such a way, unless benevolence, sentiment, duty, or some other motive, lead them to act otherwise. Such laws, which hold good, not for phenomena in their entirety, but for certain isolated groups of facts under narrowed conditions, are called _laws of the factors_ of phenomena. And since the complexity of phenomena is such that it is difficult for the human mind to grasp all the interlacing threads of causation at a single glance, men of science have endeavoured to isolate their several strands, and, applying the principle of analysis, without which reasoning is impossible, to separate out the factors and determine their laws. In this chapter we have to consider some of the factors of organic progress, and endeavour to determine their laws.

The law of heredity may be regarded as that of persistence exemplified in a series of organic generations. When, as in the am[oe]ba and some other protozoa, reproduction is by simple fission, two quite similar organisms being thus produced, there would seem to be no reason why (modifications by surrounding circumstances being disregarded) hereditary persistence should not continue indefinitely. Where, however, reproduction is effected by the detachment of a single cell from a many-celled organism, hereditary persistence[M] will be complete only on the condition that this reproductive cell is in some way in direct continuity with the cells of the parent organism or the cell from which that parent organism itself developed. And where, in the higher animals, two cells from two somewhat different parents coalesce to give origin to a new individual, the phenomena of hereditary persistence are still further complicated by the blending of characters handed on in the ovum and the sperm; still further complication being, perhaps, produced by the emergence in the offspring of characters latent in the parent, but derived from an earlier ancestor. And if characters acquired by the parents in the course of their individual life be handed on to the offspring, yet further complication will be thus introduced.

It is no matter for surprise, therefore, that, notwithstanding the law of hereditary persistence, variations should occur in the offspring of animals. At the same time, it must be remembered that the occurrence of variations is not and cannot be the result of mere chance; but that all such variations are determined by some internal or external influences, and are thus legitimate and important subjects of biological investigation. In the next chapter we shall consider at some length the phenomena of heredity and the origin of variations. Here we will accept them without further discussion, and consider some of their consequences. But even here, without discussing their origin, we must establish the fact that variations do actually occur.

Variations may be of many kinds and in different directions. In colour, in size, in the relative development of different parts, in complexity, in habits, and in mental endowments, organisms or their organs may vary. Observers of mammals, of birds, and of insects are well aware that colour is a variable characteristic. But these colour-variations are not readily described and tabulated. In the matter of size the case is different. In Mr. Wallace's recent work on "Darwinism" a number of observations on size-variations are collected and tabulated. As this is a point of great importance, I propose to illustrate it somewhat fully from some observations I have recently made of the wing-bones of bats. In carrying out these observations and making the necessary measurements, I have had the advantage of the kind co-operation of my friend Mr. Henry Charbonnier, of Clifton, an able and enthusiastic naturalist.[N]

The nature of the bat's wing will be understood by the aid of the accompanying figure (Fig. 12). In the fore limb the arm-bone, or humerus, is followed by an elongated bone composed of the radius and ulna. At the outer end of the radius is a small, freely projecting digit, which carries a claw. This answers to the thumb. Then follow four long, slender bones, which answer to the bones in the palm of our hand. They are the metacarpals, and are numbered II., III., IV., and V. in the tabulated figures in which the observations are recorded. The metacarpals of the second and third digits run tolerably close together, and form the firm support of the anterior margin of the wing. Those of the third and fourth make a considerable angle with these and with each other, and form the stays of the mid part of the wing. Beyond the metacarpals are the smaller joints or phalanges of the digits, two or three to each digit. The third digit forms the anterior point or apex of the wing. The fourth and fifth digits form secondary points behind this. Between these points the wing is scalloped into bays.

Hu., humerus, or arm-bone; Ul., conjoined radius and ulna, a bone in the forearm; Po., pollex, answering to our thumb; II., III., IV., V., second, third, fourth, and fifth digits of the manus, or hand. The figures are placed near the metacarpals, or palm-bones. These are followed by the phalanges. Fe., femur or thigh-bone; Ti., tibia, the chief bone of the shank. The digits of the pes, or foot, are short and bear claws. Ca., calcar.]

From the point of the fifth or last digit the leathery wing membrane sweeps back to the ankle. The bones of the hind limb are the femur, or thigh-bone, and the tibia (with a slender, imperfectly developed fibula). There are five toes, which bear long claws. From the ankle there runs backward a long, bony and gristly spur, which serves to support the membrane which stretches from the ankle to the tip (or near the tip) of the tail.

Thus the wing of the bat consists of a membrane stretched on the expanded or spread fingers of the hand, and sweeping from the point of the little finger to the ankle. Behind the ankle there is a membrane reaching to the tip of the tail. This forms a sort of net in which some bats, at any rate, as I have myself observed, can catch insects.

I have selected the wing of the bat to exemplify variation, (1) because the bones are readily measured even in dried specimens; (2) because they form the mutually related parts of a single organ; and (3) because they offer facilities for the comparison of variations, not only among the individuals of a single species, but also among several distinct species.

The method employed has been as follows: The several bones have been carefully measured in millimetres,[O] and all the bones tabulated for each species. Such tables of figures are here given in a condensed form for three species of bats.

Bat-Measurements (in Millimetres).
--------------------------------------------------------------------------
|R | P | 2nd | Third | Fourth | Fifth |T |
|a | o |Digit.| Digit. | Digit. | Digit. |i |
|d U | l |------|--------------|--------------|------------|b |
|i l | l | M | M | P | P | M | P | P | M | P | P |i |
|u n | e | e | e | h | h | e | h | h | e | h | h |a |
|s a | x | t | t | a | a | t | a | a | t | a | a |. |
| . | . | a | a | l | l | a | l | l | a | l | l | |
|a | | c | c | a | a | c | a | a | c | a | a | |
|n | | a | a | n | n | a | n | n | a | n | n | |
|d | | r | r | g | g | r | g | g | r | g | g | |
| | | p | p | e | e | p | e | e | p | e | e | |
| | | a | a | | s | a | | s | a | | s | |
| | | l | l | 1 | | l | 1 | | l | 1 | | |
| | | . | . | . | 2,| . | . | 2,| . | . | 2,| |
| | | | | | 3 | | | 3 | | | 3 | |
| | | | | | . | | | . | | | . | |
-------------------------------------------------------------------------
Hairy-armed bat (_Vesperugo leisleri_).
|41 |6.5| 38 |40 |16 |19 | 38 |14 | 7 | 32 | 8 | 7 |16|[Male]
|41 |6 | 38 |40 |16 |19 | 39 |15.5| 7 | 33 | 8 |6.5|16| "
|41 |6 | 39 |40 |16 |18 | 39 |16 | 6.5| 33 | 8 | 7 |16| "
|41.5|5 | 39 |40.5|17 |20 | 39 |16 | 7 | 33 | 8 | 7 |15| "
|40 |6 | 39 |37 |15.5|18 | 37 |14.5| 7 | 32 | 8 |6.5|15|[Fem.]
|41 |5.5| 38.5 |39 |16.5|20 | 39 |15 | 7.5| 33 | 8 |7.5|17| "
|41 |6 | 39 |40 |15.5|20.5| 39 |15.5| 7 | 33 | 8 | 7 |16| "
Horseshoe bat (_Rhinolophus ferri-equinum_).
|51 | 5 | 39 |36 | 19 | 29 | 40 | 11 | 18 | 40 |13 |15 |22|[Male]
|54 | 5 | 40 |36 | 18 | 32 | 40 | 11 | 19 | 40 |14 |16 |28|[Fem.]
|52 | 5 | 39 |36 | 18 | 31 | 39 | 10 | 19 | 40 |13 |14 |23| "
|54 | 5 | 39 |36 | 18 | 32 | 40 | 11 | 17 | 40 |13 |13 |25| "
|46 | 5 | 36 |34 | 16 | 29 | 36 | 10 | 19 | 36 |13 |17 |22| ?
Lesser horseshoe bat (_Rhinolophus hipposideros_).
|34 | 4 | 25 |23 | 12 | 17 | 26 | 6.5| 12 | 26 | 9 |13 |17|[Male]
|37 | 3 | 26 |24 | 13 | 20 | 28 | 8 | 13 | 28 | 9 |14 |17| "
|35 | 3 | 26 |24.5| 13 | 17 | 27 | 7 | 12 | 26 |10 |12 |15| "
-------------------------------------------------------------------------

It would be troublesome to the reader to pick out the meaning from these figures. I have, therefore, plotted in the measurements for four other species of bats in tabular form (Figs. 13, 14, 15, 16).

Fig. 13, for example, deals with the common large noctule bat, which may often be seen flying high up on summer evenings. Now, the mean length of the radius and ulna in eleven individuals was 51.5 millimetres. Suppose all the eleven bats had this bone (for the two bones form practically one piece) of exactly the same length. There would then be no variation. We may express this supposed uniformity by the straight horizontal line running across the part of the figure dealing with the radius and ulna. Practically the eleven bats measured did not have this bone of the same length; in some of them it was longer, in others it was shorter than the mean. Let us run through the eleven bats (which are represented by the numbers at the head of the table) with regard to this bone. The first fell below the average by a millimetre and a half, the length being fifty millimetres. This is expressed in the table by placing a dot or point three quarters of a division below the mean line. Each division on the table represents two millimetres, or, in other words, the distance between any two horizontal lines stands for two millimetres measured. Half a division, therefore, is equivalent to one measured millimetre; a quarter of a division to half a millimetre. The measurements are all made to the nearest half-millimetre. The second bat fell short of the mean by one millimetre. The bone measured 50.5 millimetres. The third exceeded the mean by a millimetre and a half; the fourth, by three millimetres and a half. The fifth was a millimetre and a half above the mean; and the sixth and seventh were both half a millimetre over the mean. The eighth fell short by half a millimetre; the ninth and tenth by a millimetre and a half; and the eleventh by two millimetres and a half. The points have been connected together by lines, so as to give a curve of variation for this bone.

The other curves in these four tables are drawn in exactly the same way. The mean length is stated; and the amount by which a bone in any bat exceeds or falls short of the mean can be seen and readily estimated by means of the horizontal lines of the table. Any one can reconvert the tables into figures representing our actual measurements.

Now, it may be said that, since some bats run larger than others, such variation is only to be expected. That is true. But if the bones of the wing all varied equally, _all the curves would be similar_. That is clearly not the case. The second metacarpal is the same length in 5 and 6. But the third metacarpal is two millimetres shorter in 6 than in 5. In 10 the radius and ulna are _longer_ than in 11; but the second metacarpal is _shorter_ in 10 than in 11. A simple inspection of the table as a whole will show that there is a good deal of _independent_ variation among the bones.

The amount of variation is itself variable, and in some cases is not inconsiderable. In the long-eared bats 4 and 5 in Fig. 14, the phalanges of the third digit measured 26.5 millimetres in 4, and 34 millimetres in 5--a difference of more than 28 per cent. This is unusually large, and it is possible that there may have been some slight error in the measurements.[P] A difference of 10 or 12 per cent. is, however, not uncommon.

In any case, the observations here tabulated show (1) that variations of not inconsiderable amount occur among the related bones of the bat's wing; and (2) that these variations are to a considerable extent independent of each other.

So far we have compared a series of individuals of the same species of bat, each table in Figs. 13-16 dealing with a distinct species. Let us now compare the different species with each other. To effect such a comparison, we must take some one bone as our standard, and we must level up our bats for the purposes of tabulation. I have selected the radius and ulna as the standard. In both the noctule and the greater horseshoe bats the mean length of this bone is 51.5 millimetres. The bones of each of the other bats have been multiplied by such a number as will bring them up to the level of size in these two species. Mr. Galton, in his investigations on the variations of human stature, had to take into consideration the fact that men are normally taller than women. He found, however, that the relation of man to woman, so far as height is concerned, is represented by the proportion 108 to 100. By multiplying female measurements by 1.08, they were brought up to the male standard, and could be used for purposes of comparison. In the same way, by multiplying in each case by the appropriate number, I have brought all the species in the table (Fig. 17) up to the standard of the noctule. When so multiplied, the radius and ulna (selected as the standard of comparison) has the same length in all the species, and is hence represented by the horizontal line in the table.

Compared with this as a standard, the mean length of the second metacarpal in the seven species is forty-three millimetres; that of the third metacarpal, forty-four millimetres; and so on. The amount by which each species exceeds or falls short of the mean is shown on the table, and the points are joined up as before. Here, again, the table gives the actual measurements in each case. For example, if the mean length of the third metacarpal of the greater horseshoe bat be required, it is seen by the table to fall short of the mean by four horizontal divisions and a quarter, that is to say, by eight millimetres and a half. The length is therefore (44 - 8-1/2) 35.5 millimetres.

Now, it will be seen from the table that the variation in the mean length of the bones in different species is much greater than the individual variations in the members of the same species. The table also brings out in an interesting way the variation in the general character of the wing. The noctule, for example, is especially strong in the development of the second and third metacarpals, the phalanges of the third digit being also a little above the average. Reference to the figure of the bat's wing on p. 64 will show that these excellences give length to the wing. It fails, however, in the metacarpal and phalanges of the fifth digit, and in the length of the hind leg as represented by the tibia. On consulting the figure of the wing, it is seen that these are the bones which give breadth to the wing. Here the noctule fails. Its wing is, therefore, long and narrow. It is a swallow among bats.

On the other hand, the horseshoe bats fail conspicuously in the second and third metacarpals, though they make up somewhat in the corresponding digits. On the whole, the wing is deficient in length. But the phalanges of the fourth and fifth digits, and the length of the hind limb represented by the tibia, give a corresponding increase of breadth. The wing is, therefore, relatively short and broad. The long-eared bat, again, has the third metacarpal and its digits somewhat above the mean, and therefore a somewhat more than average length. But it has the fifth metacarpal with its digit and also the tibia decidedly above the mean, and therefore more than average breadth. Without possessing the great length of the noctule's wing, or the great breadth of that of the horseshoe, it still has a more than average length and breadth.

The total wing-areas are very variable, the females having generally an advantage over the males. I do not feel that our measurements are sufficiently accurate to justify tabulation. Taking, however, the radius and ulna as the standard for bringing the various species up to the same level, the greater horseshoe seems to have decidedly the largest wing-area; the noctule stands next; then come the lesser horseshoe and the long-eared bat; somewhat lower stands the hairy-armed bat; while the pipistrelle and the whiskered bat (both small species) stand lowest.[Q]

Sufficient has now been said in illustration of the fact that variations in the lengths of the bones in the bat's wing do actually occur in the various individuals of one species; that the variations are independent; and that the different species and genera have the character of the wing determined by emphasizing, so to speak, variations in special directions. I make no apology for having treated the matter at some length. Those who do not care for details will judiciously exercise their right of skipping.

As before mentioned, Mr. Wallace has collected and tabulated other observations on size and length variations. And in addition to such variations, there are the numerous colour-variations that do not admit of being so readily tabulated. Mr. Cockerell tells us that among snail-shells, taking variations of banding alone, he knows of 252 varieties of _Helix nemoralis_ and 128 of _H. hortensis_.[R]

That variations do occur under nature is thus unquestionable. And it is clear that all variations necessarily fall under one of three categories. Either they are of advantage to the organism in which they occur; or they are disadvantageous; or they are neutral, neither advantageous nor disadvantageous to the animal in its course through life.

We must next revert to the fact to which attention was drawn in the last chapter, that every species is tending, through natural generation, to increase in numbers. Even in the case of the slow-breeding elephant, the numbers tend to increase threefold in each generation; for a single pair of elephants give birth to three pairs of young. In many animals the tendency is to increase ten, twenty, or thirtyfold in every generation; while among fishes, amphibians, and great numbers of the lower organisms, the tendency is to multiply by a hundredfold, a thousandfold, or even in some cases ten thousandfold. But, as before noticed, this is only a tendency. The law of increase is a law of one factor in life's phenomena, the reproductive factor. In any area, the conditions of which are not undergoing change, the numbers of the species which constitute its fauna remain tolerably constant. They are not actually increasing in geometrical progression. There is literally no room for such increase. The large birth-rate of the constituent species is accompanied by a proportionate death-rate, or else the tendency is kept in check by the prevention of certain individuals from mating and bearing young.[S]

Now, the high death-rate is, to a large extent among the lower organisms and in a less degree among higher animals, the result of indiscriminate destruction. When the ant-bear swallows a tongue-load of ants, when the Greenland whale engulfs some hundreds of thousands of fry at a gulp, when the bear or the badger destroys whole nests of bees,--in such cases there is wholesale and indiscriminate destruction. Those which are thus destroyed are nowise either better or worse than those which escape. At the edge of a coral reef minute, active, free-swimming coral embryos are set free in immense numbers. Presently they settle down for life. Some settle on a muddy bottom, others in too great a depth of water. These are destroyed. The few which take up a favourable position survive. But they are no better than their less fortunate neighbours. The destruction is indiscriminate. So, too, among fishes and the many marine forms which produce a great number of fertilized eggs giving rise to embryos that are from an early period free-swimming and self-supporting. Such embryos are decimated by a destruction which is quite indiscriminate. And again, to take but one more example, the liver-fluke, whose life-history was sketched in the last chapter, produces its tens or hundreds of thousands of ova. But the chances are enormously against their completing their life-cycle. If the conditions of temperature and moisture are not favourable, the embryo is not hatched or soon dies; even if it emerges, no further development takes place unless it chances to come in contact with a particular and not very common kind of water-snail. When it emerges from the intermediate host and settles on a blade of grass, it must still await the chance of that blade being eaten by a sheep or goat. It is said that the chances are eight millions to one against it, and for the most part its preservation is due to no special excellence of its own. The destruction is to a large extent, though not entirely, indiscriminate.

Even making all due allowance, however, for this indiscriminate destruction--which is to a large extent avoided by those higher creatures which foster their young--there remain more individuals than suffice to keep up the normal numbers of the species. Among these there arises a struggle for existence, and hence what Darwin named _natural selection_.

"How will the struggle for existence"--I quote, with some omissions, the words of Darwin--"act in regard to variation? Can the principle of selection, which is so potent in the hands of man, apply under nature? I think that we shall see that it can act most efficiently. Let the endless number of slight variations and individual differences be borne in mind; as well as the strength of the hereditary tendency. Let it also be borne in mind how infinitely complex and close-fitting are the mutual relations of all organic beings to each other and to their physical conditions of life; and consequently what infinitely varied diversities of structure might be of use to each being under changing conditions of life. Can it, then, be thought improbable, seeing that variations useful to man have undoubtedly occurred, that other variations, useful in some way to each being in the great and complex battle of life, should occur in the course of many successive generations? If such do occur, can we doubt (remembering that many more individuals are born than can possibly survive) that individuals having any advantage, however slight, over others, would have the best chance of surviving and of procreating their kind? On the other hand, we may feel sure that any variation in the least degree injurious would be rigidly destroyed. This preservation of favourable individual differences and variations, and the destruction of those which are injurious, I have called Natural Selection, or the Survival of the Fittest. Variations neither useful nor injurious would not be affected by natural selection, and would be left either a fluctuating element, or would ultimately become fixed, owing to the nature of the organism and the nature of the conditions."[T]

"The principle of selection," says Darwin, elsewhere, "may conveniently be divided into three kinds. _Methodical selection_ is that which guides a man who systematically endeavours to modify a breed according to some predetermined standard. _Unconscious selection_ is that which follows from men naturally preserving the most valued and destroying the less valued individuals, without any thought of altering the breed. Lastly, we have _Natural selection_, which implies that the individuals which are best fitted for the complex and in the course of ages changing conditions to which they are exposed, generally survive and procreate their kind."[U]

I venture to think that there is a more logical division than this. A man who is dealing with animals or plants under domestication may proceed by one of two well-contrasted methods. He may either select the most satisfactory individuals or he may reject the most unsatisfactory. We may term the former process _selection_, the latter _elimination_. Suppose that a gardener is dealing with a bed of geraniums. He may either pick out first the best, then the second best, then the third, and so on, until he has selected as many as he wishes to preserve. Or, on the other hand, he may weed out first the worst, then in succession other unsatisfactory stocks, until, by eliminating the failures, he has a residue of sufficiently satisfactory flowers. Now, I think it is clear that, even if the ultimate result is the same (if, that is to say, he selects the twenty best, or eliminates all but the twenty best), the method of procedure is in the two cases different. Selection is applied at one end of the scale, elimination at the other. There is a difference in method in picking out the wheat-grains (like a sparrow) and scattering the chaff by the wind.

Under nature both methods are operative, but in very different degrees. Although the insect may select the brightest flowers, or the hen-bird the gaudiest or most tuneful mate, the survival of the fittest under nature is in the main the net result of the slow and gradual process of the elimination of the unfit.[V] The best-adapted are not, save in exceptional cases, selected; but the ill-adapted are weeded out and eliminated. And this distinction seems to me of sufficient importance to justify my suggestion that _natural selection_ be subdivided under two heads--_natural elimination_, of widespread occurrence throughout the animal world; and _selection proper_, involving the element of individual or special choice.

The term "natural elimination" for the major factor serves definitely to connect the natural process with that struggle for existence out of which it arises. The struggle for existence is indeed the reaction of the organic world called forth by the action of natural elimination. Organisms are tending to increase in geometrical ratio. There is not room or subsistence for the many born. The tendency is therefore held in check by elimination, involving the struggle for existence. And the factors of elimination are three: first, elimination through the action of surrounding physical or climatic conditions, under which head we may take such forms of disease as are not due to living agency; secondly, elimination by enemies, including parasites and zymotic diseases; and thirdly, elimination by competition. It will be convenient to give some illustrative examples of each of these.

Elimination through the action of surrounding physical conditions, taken generally, deals with the very groundwork or basis of animal life. There are certain elementary mechanical conditions which must be fulfilled by every organism however situated. Any animal which fails to fulfil these conditions will be speedily eliminated. There are also local conditions which must be adequately met. Certain tropical animals, if transferred to temperate or sub-Arctic regions, are unable to meet the requirements of the new climatic conditions, and rapidly or gradually die. Fishes which live under the great pressure of the deep sea are killed by the expansion of the gases in their tissues when they are brought to the surface. Many fresh-water animals are killed if the lake in which they live be invaded by the waters of the sea. If the water in which corals live be too muddy, too cold, or too fresh--near the mouth of a great river on the Australian coast, for example--they will die off. During the changes of climate which preceded and followed the oncoming of the glacial epoch, there must have been much elimination of this order. Even under less abnormal conditions, the principle is operative. Darwin tells us that in the winter of 1854-5 four-fifths of the birds in his grounds perished from the severity of the weather, and we cannot but suppose that those who were thus eliminated were less able than others to cope with or stand the effects of the inclement climatic conditions. My colleague, Mr. G. Munro Smith, informs me that, in cultivating microbes, certain forms, such as _Bacillus violaceus_ and _Micrococcus prodigiosus_, remain in the field during cold weather when other less hardy microbes have perished. The insects of Madeira may fairly be regarded as affording another instance. The ground-loving forms allied to insects of normally slow and heavy flight have in Madeira become wingless or lost all power of flight. Those which attempted to fly have been swept out to sea by the winds, and have thus perished; those which varied in the direction of diminished powers of flight have survived this eliminating process. On the other hand, among flower-frequenting forms and those whose habits of life necessitate flight, the Madeira insects have stronger wings than their mainland allies. Here, since flight could not be abandoned without a complete change of life-habit, since all must fly, those with weaker powers on the wing have been eliminated, leaving those with stronger flight to survive and procreate their kind.[W] In Kerguelen Island Mr. Eaton has found that all the insects are incapable of flight, and most of them in a more or less wingless condition.[X] Mr. Wallace regards the reduction in the size of the wing in the Isle of Man variety of the small tortoiseshell butterfly as due to the gradual elimination of larger-winged individuals.[Y] These are cases of elimination through the direct action of surrounding physical conditions. Even among civilized human folk, this form of elimination is still occasionally operative--in military campaigns, for example (where the mortality from hardships is often as great as the mortality from shot or steel), in Arctic expeditions, and in arduous travels. But in early times and among savages it must be a more important factor.

Elimination by enemies needs somewhat fuller exemplification. Battle within battle must, throughout nature, as Darwin says, be continually recurring with varying success. The stronger devour the weaker, and wage war with each other over the prey. In the battle among co-ordinates the weaker are eliminated, the stronger prevail. When the weaker are preyed upon by the stronger and a fair fight is out of the question, the slow and heavy succumb, the agile and swift escape; stupidity means elimination, cunning, survival; to be conspicuous, unless it be for some nasty or deleterious quality, is inevitably to court death: the sober-hued stand at an advantage. In these cases, if there be true selection at work, it is the selection of certain individuals, the plumpest and most toothsome to wit, for destruction, not for survival.

This mode of elimination has been a factor in the development of protective resemblance and so-called mimicry, and we may conveniently illustrate it by reference to these qualities. If the hue of a creature varies in the direction of resemblance to the normal surroundings, it will render the animal less conspicuous, and therefore less liable to be eliminated by enemies. This is well seen in the larvæ or caterpillars of many of our butterflies and moths. It is not easy to distinguish the caterpillar of the clouded yellow, so closely does its colour assimilate to the clover leaves on which it feeds, nor that of the Lulworth skipper on blades of grass. I would beg every visitor to the Natural History Museum at South Kensington to look through the drawers containing our British butterflies and moths and their larvæ, in the further room on the basement, behind the inspiring statue of Charles Darwin. Half an hour's inspection will serve to bring home the fact of protective resemblance better than many words.

It may, however, be remarked that not all the caterpillars exhibit protective resemblance; and it may be asked--How have some of these conspicuous larvæ, that of the magpie moth, for example, escaped elimination? What is sauce for the Lulworth goose should be sauce for the magpie gander. How is it that these gaudy and variable caterpillars, cream-coloured with orange and black markings, have escaped speedy destruction? Because they are so nasty. No bird, or lizard, or frog, or spider would touch them. They can therefore afford to be bright-coloured. Nay, their very gaudiness is an advantage, and saves them from being the subject of unpleasant experiments in the matter. Other caterpillars, like the palmer-worms, are protected by barbed hairs that are intensely irritating. They, too, can afford to be conspicuous. But a sweet and edible caterpillar, if conspicuous, is eaten, and thus by the elimination of the conspicuous the numerous dull green or brown larvæ have survived.

A walk through the Bird Gallery in the National collection will afford examples of protective resemblance among birds. Look, for example, at the Kentish plover with its eggs and young--faithfully reproduced in our frontispiece--and the way in which the creature is thus protected in early stages of its life will be evident. The stone-curlew, the ptarmigan, and other birds illustrate the same fact, which is also seen with equal clearness in many mammals, the hare being a familiar example.

Many oceanic organisms are protected through general resemblance. Some, like certain medusæ, are transparent. The pellucid or transparent sole of the Pacific (_Achirus pellucidus_), a little fish about three inches long, is so transparent that sand and seaweed can be seen distinctly through its tissues. The salpa is transparent save for the intestine and digestive gland, which are brown, and look like shreds of seaweed. Other forms, like the physalia, are cærulean blue. The exposed parts of flat-fish are brown and sandy coloured or speckled like the sea-bottom; and in some the sand-grains seem to adhere to the skin. So, too, with other fish. "Looking _down_ on the dark back of a fish," says Mr. A. R. Wallace, "it is almost invisible, while to an enemy looking _up_ from below, the light under surface would be equally invisible against the light of clouds and sky." Even some of the most brilliant and gaudiest fish, such as the coral-fish (_Chætodon, Platyglossus_, and others), are brightly coloured in accordance with the beautiful tints of the coral-reefs which form their habitat; the bright-green tints of some tropical forest birds being of like import. No conception of the range of protective resemblance can be formed when the creatures are seen or figured isolated from their surroundings. The zebra is a sufficiently conspicuous animal in a menagerie or a museum; and yet Mr. Galton assures us that, in the bright starlight of an African night, you may hear one breathing close by you, and be positively unable to see the animal. A black animal would be visible; a white animal would be visible; but the zebra's black and white so blend in the dusk as to render him inconspicuous.

To cite but one more example, this time from the invertebrates. Professor Herdman found in a rock-pool on the west coast of Scotland "a peculiarly coloured specimen of the common sea-slug (_Doris tuberculata_). It was lying on a mass of volcanic rock of a dull-green colour, partially covered with rounded spreading patches of a purplish pink nullipore, and having numerous whitish yellow _Spirorbis_ shells scattered over it--the general effect being a mottled surface of dull green and pink peppered over with little cream-coloured spots. The upper surface of the Doris was of precisely the same colours arranged in the same way.... We picked up the Doris, and remarked the brightness and the unusual character of its markings, and then replaced it upon the rock, when it once more became inconspicuous."[Z]

Then, too, there are some animals with variable protective resemblance--the resemblance changing with a changing environment. This is especially seen in some Northern forms, like the Arctic hare and fox, which change their colour according to the season of the year, being brown in summer, white and snowy in winter. The chamæleon varies in colour according to the hue of its surroundings through the expansion and contraction of certain pigment-cells; while frogs and cuttle-fish have similar but less striking powers. Mr. E. B. Poulton's[AA] striking and beautiful experiments show that the colours of caterpillars and chrysalids reared from the same brood will vary according to the colour of their surroundings.

If this process of protective resemblance be carried far, the general resemblance in hue may pass into special resemblance to particular objects. The stick-insect and the leaf-insect are familiar illustrations, though no one who has not seen them in nature can realize the extent of the resemblance. Most of us have, at any rate, seen the stick-caterpillars, or loopers (Fig. 18), though, perhaps, few have noticed how wonderful is the protective resemblance to a twig when the larva is still and motionless, for the very reason that the resemblance is so marked that the organism at that time escapes, not only casual observation, but even careful search. Fig. 19 gives a representation of a locust with special protective resemblance to a leaf--not a perfect leaf, but a leaf with fungoid blotches. This insect and the stick-caterpillar may be seen in the insect exhibits on the basement at South Kensington, having been figured from them by the kind permission of Professor Flower.

Perhaps one of the most striking instances of special protective resemblance is that of the Malayan leaf-butterfly (_Kallima paralecta_). So completely, when the wings are closed, does this insect resemble a leaf that it requires a sharp eye to distinguish it. These butterflies have, moreover, the habit of alighting very suddenly. As a recent observer (Mr. S. B. T. Skertchly) remarks, they "fly rapidly along, as if late for an appointment, suddenly pitch, close their wings, and become leaves. It is generally done so rapidly that the insect seems to vanish."[AB] Instances might be multiplied indefinitely. Mr. Guppy thus describes a species of crab in the Solomon Islands: "The light purple colour of its carapace corresponds with the hue of the coral at the base of the branches, where it lives; whilst the light red colour of the big claws, as they are held up in their usual attitude, similarly imitates the colour of the branches. To make the guise more complete, both carapace and claws possess rude hexagonal markings which correspond exactly in size and appearance with the polyp-cells of the coral."[AC]

When the special protective resemblance is not to an inanimate object, but to another organism, it is termed mimicry. It arises in the following way:--

Many forms, especially among the invertebrates, escape elimination by enemies through the development of offensive weapons (stings of wasps and bees), a bitter taste (the Heliconidæ among butterflies), or a hard external covering (the weevils among beetles). The animals which prey upon these forms learn to avoid these dangerous, nasty, or indigestible creatures; and the avoidance is often instinctive. It thus becomes an advantage to other forms, not thus protected, to resemble the animals that have these characteristics. Such resemblance is termed mimicry, concerning which it must be remembered that the mimicry is unconscious, and is reached by the elimination of those forms which do not possess this resemblance. Thus the _Leptalis_, a perfectly sweet insect, closely resembles the _Methona_, a butterfly producing an ill-smelling yellow fluid. The quite harmless _Clytus arietis_, a beetle, resembles, not only in general appearance, but in its fussy walk, a wasp. The soft-skinned _Doliops_, a longicorn, resembles the strongly encased _Pachyrhyncus orbifex_, a weevil. The not uncommon fly _Eristalis tenax_ (Fig. 20), is not unlike a bee, and buzzes in an unpleasantly suggestive manner.[AD]

Mimicry is not confined to the invertebrates. A harmless snake, the eiger-eter of Dutch colonists at the Cape, subsists mainly or entirely on eggs. The mouth is almost or quite toothless; but in the throat hard-tipped spines project into the gullet from the vertebræ of the column in this region. Here the egg is broken, and there is no fear of losing the contents. Now, there is one species of this snake that closely resembles the berg-adder. The head has naturally the elongated form characteristic of the harmless snakes. But when irritated, this egg-eater flattens it out till it has the usual viperine shape of the "club" on a playing-card. It coils as if for a spring, erects its head with every appearance of anger, hisses, and darts forward as if to strike its fangs into its foe, in every way imitating an enraged berg-adder. The snake is, however, quite harmless and inoffensive.[AE]

Here we have mimicry both in form and habit. Another case of imperfect but no doubt effectual mimicry is given by Mr. W. Larden, in some notes from South America.[AF] Speaking of the rhea, or South American ostrich, he says, "One day I came across an old cock in a nest that it had made in the dry weeds and grass. Its wings and feathers were loosely arranged, and looked not unlike a heap of dried grass; at any rate, the bird did not attract my attention until I was close on him. The long neck was stretched out close along the ground, the crest feathers were flattened, and an appalling hiss greeted my approach. It was a pardonable mistake if for a moment I thought I had come across a huge snake, and sprang back hastily under this impression."

Protective resemblance and mimicry have been considered at some length because, on the hypothesis of natural selection, they admirably illustrate the results which may be reached through long-continued elimination by enemies.

Sufficient has now been said to show that this form of elimination is an important factor. We are not at present considering the question how variations arise, or why they should take any particular direction. But granting the fact that variations may and do occur in all parts of the organism, it is clear that, in a group of organisms surrounded by enemies, those individuals which varied in the direction of swiftness, cunning, inconspicuousness,[AG] or resemblance to protected forms, would, other things being equal, stand a better chance of escaping elimination.

Elimination by competition is, as Darwin well points out, keenest between members of the same group and among individuals of the same species, or between different groups or different species which have, so to speak, similar aims in life. While enemies of various kinds are preying upon weaker animals, and thus causing elimination among them, they are also competing one with another for the prey. While the slower and stupider organisms are succumbing to their captors, and thus leaving more active and cunning animals in possession of the field, the slower and stupider captors, failing to catch their cunning and active prey, are being eliminated by competition. While protective resemblance aids the prey to escape elimination by enemies, a correlative resemblance, called by Mr. Poulton aggressive resemblance, in the captors aids them in stealing upon their prey, and so gives advantage in competition. Thus the hunting spider closely resembles the flies upon which he pounces, even rubbing his head with his fore legs after their innocent fashion.

As in the case of protective resemblance, so, too, in its aggressive correlative, the resemblance may be general or special, or may reach the climax of mimicry. And since the same organism is not only a would-be captor, but sometimes an unwilling prey, the same resemblance may serve to protect it from its enemies and to enable it to steal upon its prey. The mantis, for example, gains doubly by its resemblance to the vegetation among which it lives. Certain spiders, described by Mr. H. O. Forbes, in Java, closely resemble birds'-droppings. This may serve to protect them from elimination by birds; but it also enables them to capture without difficulty unwary butterflies, which are often attracted by such excreta. A parasitic fly (_Volucella bombylans_) closely resembles (Fig. 20) a bumble-bee (_Bombus muscorum_), and is thus enabled to enter the nest of the bee without molestation. Its larvæ feed upon the larvæ of the bee. The cuckoo bee _Psithyrus rupestris_, an idle quean, who collects no pollen, and has no pollen-baskets, steals into the nest of the bumble-bee _Bombus lapidarius_, and lays her eggs there. The resemblance between the two is very great, and it not only enables the mother bee to enter unmolested, but the young bees, when they are hatched, to escape. Another bee (_Nomada solidaginis_), which plays the cuckoo on _Halictus cylindricus_, does not resemble this bee, but is wasp-like, and thus escapes molestation, not because it escapes notice, but because it looks more dangerous than it really is.[AH]

Many are the arts by which, in keen competition, organisms steal a march upon their congeners--not, be it remembered, through any conscious adaptation, but through natural selection by elimination. Mr. Poulton describes an Asiatic lizard (_Phrynocephalus mystaceus_) in which the "general surface resembles the sand on which it is found, while the fold of their skin at each angle of the mouth is of a red colour, and is produced into a flower-like shape exactly resembling a little red flower which grows in the sand. Insects, attracted by what they believe to be flowers, approach the mouth of the lizard, and are, of course, captured."[AI] The fishing frog, or angler-fish, is possessed of filaments which allure small fry, who think them worms, into the neighbourhood of the great mouth in which they are speedily engulfed; and certain deep-sea forms discovered during the _Challenger_ expedition have the lure illumined by phosphorescent light.

a, b, _Bombus muscorum_; c, d, _Volucella bombylans_; e, _Eristalis tenax_; f, _Apis mellifica_. The underwings of the hive bee (f) were invisible in the photograph from which the figure was drawn. (From an exhibit in the British Natural History Museum.)]

We need say no more in illustration of the resemblances which have enabled certain organisms to escape elimination by competition. Once more, be it understood that we are not at present considering _how_ any of these resemblances have been brought about; we are merely indicating that, given certain resemblances, advantageous either for captor or prey, those organisms which possess them not will have to suffer elimination--elimination by enemies, or else elimination by competition.

The interaction between these two kinds of elimination is of great importance. Hunters and hunted are both, so to speak, playing the game of life to the best of their ability. Those who fail on either side are weeded out; and elimination is carried so far that those who are only as good as their ancestors are placed at a disadvantage as compared with their improving congeners. The standard of efficiency is thus improving on each side; and every improvement on the one side entails a corresponding advance on the other. Nor is there only thus a competition for subsistence, and arising thereout a gradual sharpening of all the bodily and mental powers which could aid in seeking or obtaining food; there is also in some cases a competition for mates, reaching occasionally the climax of elimination by battle. There is, indeed, competition for everything which can be an object of appetence to the brute intelligence; and, owing to the geometrical tendency in multiplication--the law of increase--the competition is keen and unceasing.

Such, then, in brief, are the three main modes of elimination: elimination by physical and climatic conditions; elimination by enemies; elimination by competition. Observe that it is a differentiating process. Unlike the indiscriminate destruction before alluded to, the incidence of which is on all alike, good, bad, and indifferent, it separates the well-adapted from the ill-adapted, dooming the latter to death, and allowing the former to survive and procreate their kind. The destruction is not indiscriminate, but differential.

Let us now turn to cases of selection, properly so called, where Nature is in some way working at the other end of the scale; where her method is not the elimination of the unfit, but the selection of the fit. Such a case may be found on Darwin's principles in brightly coloured flowers and fruits. "Flowers," he says, "rank amongst the most beautiful productions of nature; but they have been rendered conspicuous in contrast with the green leaves, and, in consequence, at the same time beautiful, so that they may be easily observed by insects. I have come to this conclusion from finding it an invariable rule that, when a flower is fertilized by the wind, it never has a gaily coloured corolla. Several plants habitually produce two kinds of flowers--one kind open and coloured, so as to attract insects; the other closed, not coloured, destitute of nectar, and never visited by insects. Hence we may conclude that, if insects had not been developed on the face of 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, spinach, docks, and nettles, which are all fertilized through the agency of the wind. A similar line of argument holds good with fruits; that a ripe strawberry or cherry is as pleasing to the eye as to the palate; that the gaily coloured fruit of the spindle-wood tree, and the scarlet berries of the holly, are beautiful objects,--will be admitted by every one. But this beauty serves merely as a guide to birds and beasts, in order that the fruit may be devoured and manured seeds disseminated: I infer that this is the case from having as yet found no exception to the rule that seeds are always thus disseminated when embedded within a fruit of any kind (that is, within a fleshy or pulpy envelope), if it be coloured of any brilliant tint, or rendered conspicuous by being white or black."[AJ]

Here we have a case of the converse of elimination--a case of genuine selection under nature. But even here the process of elimination also comes into play, for the visitations of flowers by insects involve cross-fertilization. The flowers of two distinct individuals of the same species of plants in this manner fertilize each other; and the act of crossing, as Darwin firmly believed, though it is doubted by some observers nowadays, gives rise to vigorous seedlings, which consequently would have the best chance of flourishing and surviving--would best resist elimination by competition. So that we here have the double process at work; the fairest flowers being selected by insects, and those plants which failed to produce such flowers being eliminated as the relatively unfit.