The Descent of Man and Selection in Relation to Sex Volume II Part 5

[Ill.u.s.tration: Fig. 57. Basal part of the Secondary wing-feather, nearest to the body.]

I will next describe the other extreme of the series, namely the first trace of an ocellus. The short secondary wing-feather (fig. 57), nearest to the body, is marked like the other feathers, with oblique, longitudinal, rather irregular, rows of spots. The lowest spot, or that nearest the shaft, in the five lower rows (excluding the basal row) is a little larger than the other spots in the same row, and a little more elongated in a transverse direction. It differs also from the other spots by being bordered on its upper side with some dull fulvous shading. But this spot is not in any way more remarkable than those on the plumage of many birds, and might easily be quite overlooked. The next higher spot in each row does not differ at all from the upper ones in the same row, although in the following series it becomes, as we shall see, greatly modified. The larger spots occupy exactly the same relative position on this feather as those occupied by the perfect ocelli on the longer wing-feathers.

By looking to the next two or three succeeding secondary wing-feathers, an absolutely insensible gradation can be traced from one of the above-described lower spots, together with the next higher one in the same row, to a curious ornament, which cannot be called an ocellus, and which I will name, from the want of a better term, an "elliptic ornament." These are shewn in the accompanying figure (fig. 58). We here see several oblique rows, A, B, C, D (see the lettered diagram), &c., of dark spots of the usual character. Each row of spots runs down to and is connected with one of the elliptic ornaments, in exactly the same manner as each stripe in fig. 56 runs down to, and is connected with, one of the ball-and-socket ocelli. Looking to any one row, for instance, B, the lowest spot or mark (_b_) is thicker and considerably longer than the upper spots, and has its left extremity pointed and curved upwards. This black mark is abruptly bordered on its upper side by a rather broad s.p.a.ce of richly-shaded tints, beginning with a narrow brown zone, which pa.s.ses into orange, and this into a pale leaden tint, with the end towards the shaft much paler. This mark corresponds in every respect with the larger, shaded spot, described in the last paragraph (fig. 57), but is more highly developed and more brightly coloured. To the right and above this spot (_b_), with its bright shading, there is a long, narrow, black mark (_c_), belonging to the same row, and which is arched a little downwards so as to face (_b_). It is also narrowly edged on the lower side with a fulvous tint. To the left of and above _c_, in the same oblique direction, but always more or less distinct from it, there is another black mark (_d_). This mark is generally sub-triangular and irregular in shape, but in the one lettered in the diagram is unusually narrow, elongated, and regular. It apparently consists of a lateral and broken prolongation of the mark (_c_), as I infer from traces of similar prolongations from the succeeding upper spots; but I do not feel sure of this. These three marks, _b_, _c_, and _d_, with the intervening bright shades, form together the so-called elliptic ornament. These ornaments stand in a line parallel to the shaft, and manifestly correspond in position with the ball-and-socket ocelli. Their extremely elegant appearance cannot be appreciated in the drawing, as the orange and leaden tints, contrasting so well with the black marks, cannot be shewn.

[Ill.u.s.tration: Fig. 58. Portion of one of the Secondary wing-feathers near to the body; shewing the so-called elliptic ornaments. The right-hand figure is given merely as a diagram for the sake of the letters of reference.

A, B, C, &c. Rows of spots running down to and forming the elliptic ornaments.

_b_. Lowest spot or mark in row B.

_c_. The next succeeding spot or mark in the same row.

_d_. Apparently a broken prolongation of the spot _c_ in the same row B.]

Between one of the elliptic ornaments and a perfect ball-and-socket ocellus, the gradation is so perfect that it is scarcely possible to decide when the latter term ought to be used. I regret that I have not given an additional drawing, besides fig. 58, which stands about half-way in the series between one of the simple spots and a perfect ocellus. The pa.s.sage from the elliptic ornament into an ocellus is effected by the elongation and greater curvature in opposed directions of the lower black mark (_b_), and more especially of the upper one (_c_), together with the contraction of the irregular sub-triangular or narrow mark (_d_), so that at last these three marks become confluent, forming an irregular elliptic ring. This ring is gradually rendered more and more circular and regular, at the same time increasing in diameter.

Traces of the junction of all three elongated spots or marks, especially of the two upper ones, can still be observed in many of the most perfect ocelli. The broken state of the black ring on the upper side of the ocellus in fig. 56 was pointed out. The irregular sub-triangular or narrow mark (_d_) manifestly forms, by its contraction and equalisation, the thickened portion of the ring on the left upper side of the perfect ball-and-socket ocellus. The lower part of the ring is invariably a little thicker than the other parts (see fig. 56), and this follows from the lower black mark of the elliptic ornament (_b_) having been originally thicker than the upper mark (_c_). Every step can be followed in the process of confluence and modification; and the black ring which surrounds the ball of the ocellus is unquestionably formed by the union and modification of the three black marks, _b_, _c_, _d_, of the elliptic ornament. The irregular zigzag black marks between the successive ocelli (see again fig. 56) are plainly due to the breaking up of the somewhat more regular but similar marks between the elliptic ornaments.

The successive steps in the shading of the ball-and-socket ocelli can be followed out with equal clearness. The brown, orange, and pale-leaden narrow zones which border the lower black mark of the elliptic ornament can be seen gradually to become more and more softened and shaded into each other, with the upper lighter part towards the left-hand corner rendered still lighter, so as to become almost white. But even in the most perfect ball-and-socket ocelli a slight difference in the tints, though not in the shading, between the upper and lower parts of the ball can be perceived (as was before especially noticed), the line of separation being oblique, in the same direction with the bright coloured shades of the elliptic ornaments. Thus almost every minute detail in the shape and colouring of the ball-and-socket ocelli can be shewn to follow from gradual changes in the elliptic ornaments; and the development of the latter can be traced by equally small steps from the union of two almost simple spots, the lower one (fig. 57) having some dull fulvous shading on the upper side.

[Ill.u.s.tration: Fig. 59. Portion near summit of one of the Secondary wing-feathers, bearing perfect ball-and-socket ocelli.

_a._ Ornamented upper part.

_b._ Uppermost, imperfect ball-and-socket ocellus. (The shading above the white mark on the summit of the ocellus is here a little too dark.)

_c._ Perfect ocellus.]

The extremities of the longer secondary feathers which bear the perfect ball-and-socket ocelli are peculiarly ornamented. (Fig. 59.) The oblique longitudinal stripes suddenly cease upwards and become confused, and above this limit the whole upper end of the feather (_a_) is covered with white dots, surrounded by little black rings, standing on a dark ground. Even the oblique stripe belonging to the uppermost ocellus (_b_) is represented only by a very short irregular black mark with the usual, curved, transverse base. As this stripe is thus abruptly cut off above, we can understand, from what has gone before, how it is that the upper thickened part of the ring is absent in the uppermost ocellus; for, as before stated, this thickened part is apparently formed by a broken prolongation of the next higher spot in the same row. From the absence of the upper and thickened part of the ring, the uppermost ocellus, though perfect in all other respects, appears as if its top had been obliquely sliced off. It would, I think, perplex any one, who believes that the plumage of the Argus pheasant was created as we now see it, to account for the imperfect condition of the uppermost ocelli. I should add that in the secondary wing-feather farthest from the body all the ocelli are smaller and less perfect than on the other feathers, with the upper parts of the external black rings deficient, as in the case just mentioned. The imperfection here seems to be connected with the fact that the spots on this feather shew less tendency than usual to become confluent into stripes; on the contrary, they are often broken up into smaller spots, so that two or three rows run down to each ocellus.

We have now seen that a perfect series can be followed, from two almost simple spots, at first quite distinct from each other, to one of the wonderful ball-and-socket ornaments. Mr. Gould, who kindly gave me some of these feathers, fully agrees with me in the completeness of the gradation. It is obvious that the stages in development exhibited by the feathers on the same bird do not at all necessarily shew us the steps which have been pa.s.sed through by the extinct progenitors of the species; but they probably give us the clue to the actual steps, and they at least prove to demonstration that a gradation is possible.

Bearing in mind how carefully the male Argus pheasant displays his plumes before the female, as well as the many facts rendering it probable that female birds prefer the more attractive males, no one who admits the agency of s.e.xual selection, will deny that a simple dark spot with some fulvous shading might be converted, through the approximation and modification of the adjoining spots, together with some slight increase of colour, into one of the so-called elliptic ornaments. These latter ornaments have been shewn to many persons, and all have admitted that they are extremely pretty, some thinking them even more beautiful than the ball-and-socket ocelli. As the secondary plumes became lengthened through s.e.xual selection, and as the elliptic ornaments increased in diameter, their colours apparently became less bright; and then the ornamentation of the plumes had to be gained by improvements in the pattern and shading; and this process has been carried on until the wonderful ball-and-socket ocelli have been finally developed. Thus we can understand-and in no other way as it seems to me-the present condition and origin of the ornaments on the wing-feathers of the Argus pheasant.

From the light reflected by the principle of gradation; from what we know of the laws of variation; from the changes which have taken place in many of our domesticated birds; and, lastly, from the character (as we shall hereafter more clearly see) of the immature plumage of young birds-we can sometimes indicate with a certain amount of confidence, the probable steps by which the males have acquired their brilliant plumage and various ornaments; yet in many cases we are involved in darkness. Mr. Gould several years ago pointed out to me a humming-bird, the _Urosticte benjamini_, remarkable from the curious differences presented by the two s.e.xes. The male, besides a splendid gorget, has greenish-black tail-feathers, with the four _central_ ones tipped with white; in the female, as with most of the allied species, the three _outer_ tail-feathers on each side are tipped with white, so that the male has the four central, whilst the female has the six exterior feathers ornamented with white tips. What makes the case curious is that, although the colouring of the tail differs remarkably in both s.e.xes of many kinds of humming-birds, Mr. Gould does not know a single species, besides the Urosticte, in which the male has the four central feathers tipped with white.

The Duke of Argyll, in commenting on this case,[196] pa.s.ses over s.e.xual selection, and asks, "What explanation does the law of natural selection give of such specific varieties as these?" He answers "none whatever;"

and I quite agree with him. But can this be so confidently said of s.e.xual selection? Seeing in how many ways the tail-feathers of humming-birds differ, why should not the four central feathers have varied in this one species alone, so as to have acquired white tips? The variations may have been gradual, or somewhat abrupt as in the case recently given of the humming-birds near Bogota, in which certain individuals alone have the "central tail-feathers tipped with beautiful green." In the female of the Urosticte I noticed extremely minute or rudimental white tips to the two outer of the four central black tail-feathers; so that here we have an indication of change of some kind in the plumage of this species. If we grant the possibility of the central tail-feathers of the male varying in whiteness, there is nothing strange in such variations having been s.e.xually selected. The white tips, together with the small white ear-tufts, certainly add, as the Duke of Argyll admits, to the beauty of the male; and whiteness is apparently appreciated by other birds, as may be inferred from such cases as the snow-white male of the Bell-bird. The statement made by Sir E. Heron should not be forgotten, namely that his peahens, when debarred from access to the pied peac.o.c.k, would not unite with any other male, and during that season produced no offspring. Nor is it strange that variations in the tail-feathers of the Urosticte should have been specially selected for the sake of ornament, for the next succeeding genus in the family takes its name of Metallura from the splendour of these feathers. Mr. Gould, after describing the peculiar plumage of the Urosticte, adds, "that ornament and variety is the sole object, I have myself but little doubt."[197] If this be admitted, we can perceive that the males which were decked in the most elegant and novel manner would have gained an advantage, not in the ordinary struggle for life, but in rivalry with other males, and would consequently have left a larger number of offspring to inherit their newly-acquired beauty.

CHAPTER XV.

BIRDS-_continued_.

Discussion why the males alone of some species, and both s.e.xes of other species, are brightly coloured-On s.e.xually-limited inheritance, as applied to various structures and to brightly-coloured plumage-Nidification in relation to colour-Loss of nuptial plumage during the winter.

We have in this chapter to consider, why with many kinds of birds the female has not received the same ornaments as the male; and why with many others, both s.e.xes are equally, or almost equally, ornamented? In the following chapter we shall consider why in some few rare cases the female is more conspicuously coloured than the male.

In my 'Origin of Species'[198] I briefly suggested that the long tail of the peac.o.c.k would be inconvenient, and the conspicuous black colour of the male capercailzie dangerous, to the female during the period of incubation; and consequently that the transmission of these characters from the male to the female offspring had been checked through natural selection. I still think that this may have occurred in some few instances: but after mature reflection on all the facts which I have been able to collect, I am now inclined to believe that when the s.e.xes differ, the successive variations have generally been from the first limited in their transmission to the same s.e.x in which they first appeared. Since my remarks appeared, the subject of s.e.xual coloration has been discussed in some very interesting papers by Mr. Wallace,[199]

who believes that in almost all cases the successive variations tended at first to be transmitted equally to both s.e.xes; but that the female was saved, through natural selection, from acquiring the conspicuous colours of the male, owing to the danger which she would thus have incurred during incubation.

This view necessitates a tedious discussion on a difficult point, namely whether the transmission of a character, which is at first inherited by both s.e.xes, can be subsequently limited in its transmission, by means of selection, to one s.e.x alone. We must bear in mind, as shewn in the preliminary chapter on s.e.xual selection, that characters which are limited in their development to one s.e.x are always latent in the other.

An imaginary ill.u.s.tration will best aid us in seeing the difficulty of the case: we may suppose that a fancier wished to make a breed of pigeons, in which the males alone should be coloured of a pale blue, whilst the females retained their former slaty tint. As with pigeons characters of all kinds are usually transmitted to both s.e.xes equally, the fancier would have to try to convert this latter form of inheritance into s.e.xually-limited transmission. All that he could do would be to persevere in selecting every male pigeon which was in the least degree of a paler blue; and the natural result of this process, if steadily carried on for a long time, and if the pale variations were strongly inherited or often recurred, would be to make his whole stock of a lighter blue. But our fancier would be compelled to match, generation after generation, his pale blue males with slaty females, for he wishes to keep the latter of this colour. The result would generally be the production either of a mongrel piebald lot, or more probably the speedy and complete loss of the pale-blue colour, for the primordial slaty tint would be transmitted with prepotent force. Supposing, however, that some pale-blue males and slaty females were produced during each successive generation, and were always crossed together; then the slaty females would have, if I may use the expression, much blue blood in their veins, for their fathers, grandfathers, etc., will all have been blue birds.

Under these circ.u.mstances it is conceivable (though I know of no distinct facts rendering it probable) that the slaty females might acquire so strong a latent tendency to pale-blueness, that they would not destroy this colour in their male offspring, their female offspring still inheriting the slaty tint. If so, the desired end of making a breed with the two s.e.xes permanently different in colour might be gained.

The extreme importance, or rather necessity, of the desired character in the above case, namely, pale-blueness, being present though in a latent state in the female, so that the male offspring should not be deteriorated, will be best appreciated as follows: the male of Smmerring's pheasant has a tail thirty-seven inches in length, whilst that of the female is only eight inches; the tail of the male common pheasant is about twenty inches, and that of the female twelve inches long. Now if the female Smmerring pheasant with her _short_ tail were crossed with the male common pheasant, there can be no doubt that the male hybrid offspring would have a much _longer_ tail than that of the pure offspring of the common pheasant. On the other hand, if the female common pheasant, with her tail nearly _twice as long_ as that of the female Smmerring pheasant, were crossed with the male of the latter, the male hybrid offspring would have a much _shorter_ tail than that of the pure offspring of Smmerring's pheasant.[200]

Our fancier, in order to make his new breed with the males of a decided pale-blue tint, and the females unchanged, would have to continue selecting the males during many generations; and each stage of paleness would have to be fixed in the males, and rendered latent in the females.

The task would be an extremely difficult one, and has never been tried, but might possibly succeed. The chief obstacle would be the early and complete loss of the pale-blue tint, from the necessity of reiterated crosses with the slaty female, the latter not having at first any _latent_ tendency to produce pale-blue offspring.

On the other hand, if one or two males were to vary ever so slightly in paleness, and the variations were from the first limited in their transmission to the male s.e.x, the task of making a new breed of the desired kind would be easy, for such males would simply have to be selected and matched with ordinary females. An a.n.a.logous case has actually occurred, for there are breeds of the pigeon in Belgium[201] in which the males alone are marked with black striae. In the case of the fowl, variations of colour limited in their transmission to the male s.e.x habitually occur. Even when this form of inheritance prevails, it might well happen that some of the successive steps in the process of variation might be transferred to the female, who would then come to resemble in a slight degree the male, as occurs in some breeds of the fowl. Or again, the greater number, but not all, of the successive steps might be transferred to both s.e.xes, and the female would then closely resemble the male. There can hardly be a doubt that this is the cause of the male pouter pigeon having a somewhat larger crop, and of the male carrier pigeon having somewhat larger wattles, than their respective females; for fanciers have not selected one s.e.x more than the other, and have had no wish that these characters should be more strongly displayed in the male than in the female, yet this is the case with both breeds.

The same process would have to be followed, and the same difficulties would be encountered, if it were desired to make a breed with the females alone of some new colour.

Lastly, our fancier might wish to make a breed with the two s.e.xes differing from each other, and both from the parent-species. Here the difficulty would be extreme, unless the successive variations were from the first s.e.xually limited on both sides, and then there would be no difficulty. We see this with the fowl; thus the two s.e.xes of the pencilled Hamburghs differ greatly from each other, and from the two s.e.xes of the aboriginal _Gallus bankiva_; and both are now kept constant to their standard of excellence by continued selection, which would be impossible unless the distinctive characters of both were limited in their transmission. The Spanish fowl offers a more curious case; the male has an immense comb, but some of the successive variations, by the acc.u.mulation of which it was acquired, appear to have been transferred to the female; for she has a comb many times larger than that of the females of the parent-species. But the comb of the female differs in one respect from that of the male, for it is apt to lop over; and within a recent period it has been ordered by the fancy that this should always be the case, and success has quickly followed the order. Now the lopping of the comb must be s.e.xually limited in its transmission, otherwise it would prevent the comb of the male from being perfectly upright, which would be abhorrent to every fancier. On the other hand the uprightness of the comb in the male must likewise be a s.e.xually-limited character, otherwise it would prevent the comb of the female from lopping over.

From the foregoing ill.u.s.trations, we see that even with almost unlimited time at command, it would be an extremely difficult and complex process, though perhaps not impossible, to change through selection one form of transmission into the other. Therefore, without distinct evidence in each case, I am unwilling to admit that this has often been effected with natural species. On the other hand by means of successive variations, which were from the first s.e.xually limited in their transmission, there would not be the least difficulty in rendering a male bird widely different in colour or in any other character from the female; the latter being left unaltered, or slightly altered, or specially modified for the sake of protection.

As bright colours are of service to the males in their rivalry with other males, such colours would be selected, whether or not they were transmitted exclusively to the same s.e.x. Consequently the females might be expected often to partake of the brightness of the males to a greater or less degree; and this occurs with a host of species. If all the successive variations were transmitted equally to both s.e.xes, the females would be undistinguishable from the males; and this likewise occurs with many birds. If, however, dull colours were of high importance for the safety of the female during incubation, as with many ground birds, the females which varied in brightness, or which received through inheritance from the males any marked accession of brightness, would sooner or later be destroyed. But the tendency in the males to continue for an indefinite period transmitting to their female offspring their own brightness, would have to be eliminated by a change in the form of inheritance; and this, as shewn by our previous ill.u.s.tration, would be extremely difficult. The more probable result of the long-continued destruction of the more brightly-coloured females, supposing the equal form of transmission to prevail, would be the lessening or annihilation of the bright colours of the males, owing to their continually crossing with the duller females. It would be tedious to follow out all the other possible results; but I may remind the reader, as shewn in the eighth chapter, that if s.e.xually-limited variations in brightness occurred in the females, even if they were not in the least injurious to them and consequently were not eliminated, yet they would not be favoured or selected, for the male usually accepts any female, and does not select the more attractive individuals; consequently these variations would be liable to be lost, and would have little influence on the character of the race; and this will aid in accounting for the females being commonly less brightly-coloured than the males.

In the chapter just referred to, instances were given, and any number might have been added, of variations occurring at different ages, and inherited at the same age. It was also shewn that variations which occur late in life are commonly transmitted to the same s.e.x in which they first appeared; whilst variations occurring early in life are apt to be transmitted to both s.e.xes; not that all the cases of s.e.xually-limited transmission can thus be accounted for. It was further shewn that if a male bird varied by becoming brighter whilst young, such variations would be of no service until the age for reproduction had arrived, and there was compet.i.tion between rival males. If we suppose that three-fourths of the young males of any species are on an average destroyed by various enemies; then the chances would be as three to one against any one individual more brightly-coloured than usual surviving to propagate its kind. But in the case of birds which live on the ground and which commonly need the protection of dull colours, bright tints would be far more dangerous to the young and inexperienced than to the adult males. Consequently the males which varied in brightness whilst young would suffer much destruction and be eliminated through natural selection; on the other hand the males which varied in this manner when nearly mature, notwithstanding that they were exposed to some additional danger, might survive, and from being favoured through s.e.xual selection, would procreate their kind. The brightly-coloured young males being destroyed and the mature ones being successful in their courtship, may account, on the principle of a relation existing between the period of variation and the form of transmission, for the males alone of many birds, having acquired and transmitted brilliant colours to their male offspring alone. But I by no means wish to maintain that the influence of age on the form of transmission is indirectly the sole cause of the great difference in brilliancy between the s.e.xes of many birds.

As with all birds in which the s.e.xes differ in colour, it is an interesting question whether the males alone have been modified through s.e.xual selection, the females being left, as far as this agency is concerned, unchanged or only partially changed; or whether the females have been specially modified through natural selection for the sake of protection, I will discuss this question at considerable length, even at greater length than its intrinsic importance deserves; for various curious collateral points may thus be conveniently considered.

Before we enter on the subject of colour, more especially in reference to Mr. Wallace's conclusions, it may be useful to discuss under a similar point of view some other differences between the s.e.xes. A breed of fowls formerly existed in Germany[202] in which the hens were furnished with spurs; they were good layers, but they so greatly disturbed their nests with their spurs that they could not be allowed to sit on their own eggs. Hence at one time it appeared to me probable that with the females of the wild Gallinaceae the development of spurs had been checked through natural selection, from the injury thus caused to their nests. This seemed all the more probable as the wing-spurs, which could not be injurious during nidification, are often as well developed in the female as in the male; though in not a few cases they are rather larger in the male. When the male is furnished with leg-spurs the female almost always exhibits rudiments of them,-the rudiment sometimes consisting of a mere scale, as with the species of Gallus. Hence it might be argued that the females had aboriginally been furnished with well-developed spurs, but that these had subsequently been lost either through disuse or natural selection. But if this view be admitted, it would have to be extended to innumerable other cases; and it implies that the female progenitors of the existing spur-bearing species were once enc.u.mbered with an injurious appendage.

In some few genera and species, as in Galloperdix, Acomus, and the Javan peac.o.c.k (_Pavo muticus_), the females, as well as the males, possess well-developed spurs. Are we to infer from this fact that they construct a different sort of nest, not liable to be injured by their spurs, from that made by their nearest allies, so that there has been no need for the removal of their spurs? Or are we to suppose that these females especially require spurs for their defence? It is a more probable conclusion that both the presence and absence of spurs in the females result from different laws of inheritance having prevailed, independently of natural selection. With the many females in which spurs appear as rudiments, we may conclude that some few of the successive variations, through which they were developed in the males, occurred very early in life, and were as a consequence transferred to the females. In the other and much rarer cases, in which the females possess fully developed spurs, we may conclude that all the successive variations were transferred to them; and that they gradually acquired the inherited habit of not disturbing their nests.

The vocal organs and the variously-modified feathers for producing sound, as well as the proper instincts for using them, often differ in the two s.e.xes, but are sometimes the same in both. Can such differences be accounted for by the males having acquired these organs and instincts, whilst the females have been saved from inheriting them, on account of the danger to which they would have been exposed by attracting the attention of birds or beasts of prey? This does not seem to me probable, when we think of the mult.i.tude of birds which with impunity gladden the country with their voices during the spring.[203]

It is a safer conclusion that as vocal and instrumental organs are of special service only to the males during their courtship, these organs were developed through s.e.xual selection and continued use in this s.e.x alone-the successive variations and the effects of use having been from the first limited in their transmission in a greater or less degree to the male offspring.

Many a.n.a.logous cases could be advanced; for instance the plumes on the head, which are generally longer in the male than in the female, sometimes of equal length in both s.e.xes, and occasionally absent in the female,-these several cases sometimes occurring in the same group of birds. It would be difficult to account for a difference of this kind between the s.e.xes on the principle of the female having been benefited by possessing a slightly shorter crest than the male, and its consequent diminution or complete suppression through natural selection. But I will take a more favourable case, namely, the length of the tail. The long train of the peac.o.c.k would have been not only inconvenient but dangerous to the peahen during the period of incubation and whilst accompanying her young. Hence there is not the least _a priori_ improbability in the development of her tail having been checked through natural selection.

But the females of various pheasants, which apparently are exposed on their open nests to as much danger as the peahen, have tails of considerable length. The females as well as the males of the _Menura superba_ have long tails, and they build a domed nest, which is a great anomaly in so large a bird. Naturalists have wondered how the female Menura could manage her tail during incubation; but it is now known[204] that she "enters the nest head first, and then turns round with her tail sometimes over her back, but more often bent round by her side. Thus in time the tail becomes quite askew, and is a tolerable guide to the length of time the bird has been sitting." Both s.e.xes of an Australian kingfisher (_Tanysiptera sylvia_) have the middle tail-feathers greatly lengthened; and as the female makes her nest in a hole, these feathers become, as I am informed by Mr. R. B. Sharpe, much crumpled during nidification.

In these two cases the great length of the tail-feathers must be in some degree inconvenient to the female; and as in both species the tail-feathers of the female are somewhat shorter than those of the male, it might be argued that their full development had been prevented through natural selection. Judging from these cases, if with the peahen, the development of the tail had been checked only when it became inconveniently or dangerously long, she would have acquired a much longer tail than she actually possesses; for her tail is not nearly so long, relatively to the size of her body, as that of many female pheasants, nor longer than that of the female turkey. It must also be borne in mind, that in accordance with this view as soon as the tail of the peahen became dangerously long, and its development was consequently checked, she would have continually reacted on her male progeny, and thus have prevented the peac.o.c.k from acquiring his present magnificent train. We may therefore infer that the length of the tail in the peac.o.c.k and its shortness in the peahen are the result of the requisite variations in the male having been from the first transmitted to the male offspring alone.

We are led to a nearly similar conclusion with respect to the length of the tail in the various species of pheasants. In the Eared pheasant (_Crossoptilon auritum_) the tail is of equal length in both s.e.xes, namely, sixteen or seventeen inches; in the common pheasant it is about twenty inches long in the male, and twelve in the female; in Smmerring's pheasant, thirty-seven inches in the male, and only eight in the female; and lastly in Reeve's pheasant it is sometimes actually seventy-two inches long in the male and sixteen in the female.

Thus in the several species, the tail of the female differs much in length, irrespectively of that of the male; and this can be accounted for as it seems to me, with much more probability, by the laws of inheritance,-that is by the successive variations having been from the first more or less closely limited in their transmission to the male s.e.x,-than by the agency of natural selection, owing to the length of tail having been injurious in a greater or less degree to the females of the several species.

We may now consider Mr. Wallace's arguments, in regard to the s.e.xual coloration of birds. He believes that the bright tints originally acquired through s.e.xual selection by the males, would in all or almost all cases have been transmitted to the females, unless the transference had been checked through natural selection. I may here remind the reader that various facts bearing on this view have already been given under reptiles, amphibians, fishes, and lepidoptera. Mr. Wallace rests his belief chiefly, but not exclusively, as we shall see in the next chapter, on the following statement,[205] that when both s.e.xes are coloured in a strikingly-conspicuous manner the nest is of such a nature as to conceal the sitting bird; but when there is a marked contrast of colour between the s.e.xes, the male being gay and the female dull-coloured, the nest is open and exposes the sitting bird to view.

This coincidence, as far as it goes, certainly supports the belief that the females which sit on open nests have been specially modified for the sake of protection. Mr. Wallace admits that there are, as might have been expected, some exceptions to his two rules, but it is a question whether the exceptions are not so numerous as seriously to invalidate them.

There is in the first place much truth in the Duke of Argyll's remark[206] that a large domed nest is more conspicuous to an enemy, especially to all tree-haunting carnivorous animals, than a smaller open nest. Nor must we forget that with many birds which build open nests the males sit on the eggs and aid in feeding the young as well as the females: this is the case, for instance, with _Pyranga aestiva_,[207] one of the most splendid birds in the United States, the male being vermilion, and the female light brownish-green. Now if brilliant colours had been extremely dangerous to birds whilst sitting on their open nests, the males in these cases would have suffered greatly. It might, however, be of such paramount importance to the male to be brilliantly coloured, in order to beat his rivals, that this would more than compensate for some additional danger.