These groups of different species, some belonging to palatable and some to unpalatable groups, all exhibiting a close resemblance in colour and pattern, are far more strikingly developed in S. America than in either Asia or Africa, and it is not uncommon for eight or ten species to enter into such an association. A group of this sort which possesses unusual interest is the so-called "Transparency Group" from certain parts of the Amazon region. It was originally described by Bates with seven species belonging to six different genera. To-day it is said that no less than 28 species of this peculiar facies are known, though some are excessively rare. The majority are Ithomiines, but two species of the Danaine genus Ituna, the Pierine Dismorphia orise (Pl. XII, fig. 2), the Swallow-tail Papilio hahneli, and several species of diurnal moths belonging to different families (cf. Pl. XII, fig. 4) also enter into the combination. In connection with it there is a feature of peculiar interest in that the transparent effect is not always produced in the same way. In the Ithomiines such as Thyridia, where there are normally two kinds of scales, the wider ones for the most part lose their pigment, become much reduced in size and take on the shape of a stumpy V (Pl. XIV, fig. 3). Also they stand out for the most part more or less at right angles to the wing[19], and the neck by which they are joined to the wing membrane is very short. The longer and narrow form of scales also tend to lose their pigment and become reduced to fine hairs. In Dismorphia the scales, which are of one sort, are also reduced in size though apparently not in number. Like the wider scales of the Thyridia they tend sometimes to project at right angles to the wing membrane, though not to the same extent as in the Ithomiine: possibly because the neck of the scale is not so short. As in Thyridia these reduced scales lose their pigment except in the transition region round the borders of the transparent patches. In Ituna there is a difference. The scales are not reduced to the same extent in point of size. Their necks are longer as in normal scales and they lie flat on the wing membrane. The majority of the scales, as in the preceding cases, lose their pigment, but mixed up with them is a certain proportion, about one-quarter, in which the pigment is retained. In Castnia and in Anthomysa the scales on the transparent parts which are without pigment are also somewhat reduced in size, being stumpier than the normal ones. At the same time they tend to stand out at right angles to the wing membrane[20]. The neck here again is shorter in the transparent than in the pigmented scales. A good deal of stress has been laid upon this case by some supporters of the theory of mimicry, since it is supposed to shew that a similar effect can be brought about in a variety of ways; consequently the existence of this assembly of similar transparent forms belonging to various families cannot be put down as due to the effect of similar conditions, but must be regarded as having arisen in each instance in a different manner through the independent action of natural selection[21]. It is doubtful, however, whether such a conclusion necessarily follows from the facts. In all of the cases the process would appear to be similar: loss of pigment, reduction in the size of the scales, and eventually a tendency for the scales to stand at right angles to the wing—this last part of the process apparently depending upon the reduction of the neck of the scale. It has been said that greater transparency is brought about by the scales standing out at right angles in this way, but as the scales themselves are already transparent there would appear to be no reason why this should be so. Of course the process has not proceeded in all of the forms to the same extent. There is least change in Ituna where the scales are not much reduced in size and where a fair proportion are still pigmented. There is probably most in an Ithomiine such as Thyridia, where the scales are not only small and entirely without pigment, but also are for the most part neckless so that they stand out at right angles to the wing. Having regard to the fact that several widely separate genera with different types of scaling formed the starting points, the final results do not seem to preclude the supposition that the transparency has arisen through a similar process in all of them.
It is somewhat remarkable that no Satyrine exhibits mimicry in S. America, in spite of the fact that transparency of the wings, as in so many of the butterflies of this region, is quite common in the group. On the other hand the relatively large number of more or less mimetic Pierines is a striking feature of S. America. For the most part they belong to the genera Dismorphia and Perrhybris, and resemble the yellow, black, and brown Heliconines and Ithomiines, though some of the former genus are mimics of the small transparent Ithomiines. Some of the species of Pereute with their dark ground colour and the bright red bar across the fore wing (Pl. XI, fig. 6) resemble Heliconius melpomene, as also does Papilio euterpinus. But some of the most interesting Pierine mimics are several forms belonging to the genus Archonias (Pl. XI, fig. 10) which exhibit the simple and striking arrangement of black, red and white so characteristic of the Swallow-tail Poison-eaters of S. America. They form one of the rare instances of a Pharmacophagus Papilio being mimicked by a butterfly which does not belong to the Swallow-tail group.
As everywhere in the tropics the Papilios of S. America supply a goodly proportion of the mimicry cases. A few, such as P. zagreus (Pl. X, fig. 8), enter into the black-brown and yellow Ithomiine-Heliconine combination; P. euterpinus resembles Heliconius melpomene (Pl. XI, fig. 5); P. pausanias is like Heliconius sulphurea (Pl. XI, figs. 1 and 2). But this practically exhausts the list of Papilios which mimic Heliconines and Ithomiines. The great majority of mimicking Swallow-tails in S. America find their models among the Poison-eaters of their own family, offering in this respect a contrast to those of Asia where the majority of models are among the Danaines and Euploeines, and of Africa where they are exclusively Acraeines or Danaines.
The Poison-eaters of S. America fall into two well-marked groups which we may call the red-spotted and the dark green groups respectively. The red spotted group form a remarkably compact and uniform assemblage. The general ground colour is a deep black-brown (Pl. XI, figs. 8 and 9), the hind wings are almost invariably marked with red near the centre or towards the outer margin, and the fore wing may or may not bear a patch which is generally whitish in the female, though often of a brilliant blue or green in the male. This simple colour scheme with variations runs throughout about three-quarters (some 40 species) of the Poison-eaters. The same general colour scheme is also found in about two dozen species of the unprotected Swallow-tails. As the total number of the unprotected species is placed by Seitz at less than 100 this means that fully one-quarter of them fall into the general colour scheme adopted by the majority of the Poison-eaters. In many cases the resemblance between mimic and model is so close as to have deceived the most expert entomologists before the structural differences between the groups had been appreciated (cf. Appendix II). The matter is further complicated by the fact that polymorphism is not uncommon, especially among the females of the mimetic forms. Papilio lysithous for instance has no less than six distinct forms of female, which differ chiefly in the extent and arrangement of the white markings on the wings, one form lacking them entirely. Several of these forms may occur together in a given locality, and may resemble as many distinct species of Poison-eaters. Thus the three forms lysithous, with white on both wings, rurik, with white on the fore wing only, and pomponius without any white, all fly together in Rio Grande do Sul and respectively mimic the three distinct Pharmacophagus species nephalion, chamissonia, and perrhebus (Pl. XIII). It is worthy of note that mimics are provided by both unprotected groups of Swallow-tails in S. America, whereas in Asia the Cosmodesmus division never provides mimics for Pharmacophagus models (cf. Appendix II).
In the second and smaller group of the Pharmacophagus Swallow-tails the general colour scheme is a more or less dark metallic blue-green with a tendency towards the obliteration of light markings. Some idea of their appearance may be got from the figure of the Central and N. American P. philenor on Pl. XVI, fig. 1. Though one or two unprotected Papilios in S. America fall more or less into this colour scheme, the group, from the point of view of mimicry, is not nearly so important as the red-spotted one.
Nevertheless the blue-green Pharmacophagus group as represented by P. philenor is supposed to play a considerable part in mimicry in N. America. P. philenor is found throughout the greater part of the Eastern United States, straggling up as far as the Canadian border. On the west it is also found reaching up to North California. Over considerable parts of its range are three other Swallow-tails, belonging to the unprotected Papilios, which are regarded by Professor Poulton and others as mimics of philenor[22]. One of these, P. troilus, is dark brown with a dusting of blue scales over the hind wing (Pl. XVI, fig. 2). The sexes here are more or less alike. Troilus stretches up into North-west Canada some way beyond the limits reached by its model. P. glaucus is a black and yellow Swallow-tail with two forms of female. One of these resembles the male while the other is darker and is said to mimic philenor. It is known as the turnus form and is found more commonly in the southern part of the range of the species, i.e. in the country where philenor is more plentiful. The third species, P. asterius, has a more southerly distribution. Its female is darker and nearer to philenor than the male. It must, however, be admitted that none of the three species bears a very close resemblance to philenor. It is suggested that this is because P. philenor is a tropical form which has only recently invaded N. America. The crossing of philenor has, as it were, induced the three mimicking Papilios to turn dark, but the model has not been long enough in contact with them for the likeness to become a close one. The explanation, however, hardly accounts for the fact that the best mimic of the three, P. troilus, in which both sexes are dark, is found far north of philenor. Either the dark colour was established without the influence of the Pharmacophagus model, or else the species rapidly extended its range northwards after having been modified under the influence of philenor in the south. But in that case the critic may ask why it does not revert to the original pattern now that it has got beyond the model's sphere of influence. On the whole it seems at present quite doubtful whether any relation of a mimetic nature exists between P. philenor and these three species of Papilio.
P. philenor is also regarded as serving as a model for two Nymphaline butterflies in the United States. One of these is the large Fritillary Argynnis diana of which the dark female has a markedly blue tint on the upper surface (Pl. XVI, fig. 3). The other is a Limenitis[23] related to our own White Admiral. This form, L. astyanax (Pl. XVI, fig. 5), is a dark form with a bluish iridescence on the upper surface. It is found, like P. philenor, over the greater part of the Eastern States, while to the north, near the Canadian boundary, its place is taken by L. arthemis with prominent white bar across both wings (Pl. XVI, fig. 4). There is reason for believing that where the two overlap there is occasional inbreeding, and that the hybrid is the form known as proserpina, resembling astyanax more than arthemis. It must be admitted that in general appearance L. astyanax and Argynnis diana are more like Papilio troilus than P. philenor. In explanation it has been suggested that all the mimics are on the way to resembling P. philenor, and consequently we should expect them at certain stages to shew more resemblance to one another than to the form they have all as it were set out to mimic. On this view they will all arrive at a close resemblance to philenor in time. Another explanation is that favoured by Professor Poulton on which it is assumed that we are here dealing with a case of Müllerian Mimicry, all of the species in question being distasteful with the exception perhaps of A. diana. Thus troilus and astyanax though distasteful are less so than philenor. Hence it is of advantage to them to have even a chance of being mistaken for the more obnoxious philenor, and so the one has come from the black and yellow Swallow-tail pattern and the other from the white-banded arthemis form to what they are, i.e. more alike to one another than to philenor. They now form a Müllerian combination for mutual protection along with the dark females of glaucus and asterius. But they are themselves still moderately distasteful so that it is to the advantage of the female of Argynnis diana to mimic them. Whether they are all on the way to resembling philenor more closely, or whether they have sufficiently vindicated their inedible properties and are now stationary, it is for the future to reveal to posterity. Lastly we have the view that these different species have attained their present coloration entirely independently of one another, and that we are not here concerned with mimicry at all. Since the sole evidence available at present is that based on general appearance and geographical distribution, the view taken of this case must rest largely upon personal inclination.
Though the cases just quoted are only very problematically mimetic, N. America has yet several examples of resemblance between distantly related forms as close as any that occur in the tropics. In this region are found two species of the genus Danais—D. archippus occurring all over the United States and reaching up northwards into Canada, D. berenice found in the South-eastern States, e.g. in Florida, where it is said to be more abundant than archippus. D. archippus (Pl. XVI, fig. 8) is very similar to the oriental D. plexippus (Pl. IV, fig. 2), from which perhaps its most notable difference lies in the extent and arrangement of the white spots near the tip of the fore wing. D. berenice is not unlike archippus in its general colour scheme but is smaller and darker (Pl. XVI, fig. 9).
We have already had occasion to mention the common Nymphaline, Limenitis arthemis (Pl. XVI, fig. 4) which is found in Canada and the Northeastern States. Widely spread over N. America is a close ally of this species, L. archippus, which, though so similar in structure and habits, is very different in external appearance. As appears from Pl. XVI, fig. 6, L. archippus is remarkably like the Danaid which bears the same specific name. In the Southern States L. archippus is replaced by a form slightly different in details of pattern and distinctly darker, L. floridensis (= eros) (Pl. XVI, fig. 7). In Florida occurs also the darker N. American Danaid, D. berenice, to which the colour of L. floridensis approximates more than to D. archippus, and it is of interest that although the last named is also found in this locality it is said to be much less abundant than D. berenice. Nevertheless it appears to be true that the range of L. floridensis is much more extensive than that of its model; in other words, that there are considerable regions where L. floridensis and D. archippus coexist, and from which L. archippus and D. berenice are wanting.
CHAPTER V
SOME CRITICISMS
The facts related in the last two chapters are sufficient to make it clear that these remarkable resemblances between species belonging as a rule to widely different groups constitute a real phenomenon, and as such demand an explanation. One explanation, that in terms of the theory of mimicry, has already been outlined, and we may now turn to consider it in more detail. Some years ago Wallace[24], combating the suggestion that these instances of resemblance might be mere coincidences, laid down five conditions which he stated were applicable to all such cases, and rendered utterly inadequate any explanation other than in terms of natural selection. These five conditions are of historical interest and may also serve as a peg for sundry criticisms in connection with the mimicry theory. They are as follows:
(1) That the imitative species occur in the same area and occupy the very same station as the imitated.
(2) That the imitators are always the more defenceless.
(3) That the imitators are always less numerous in individuals.
(4) That the imitators differ from the bulk of their allies.
(5) That the imitation, however minute, is external and visible only, never extending to internal characters or to such as do not affect the external appearance.
In offering certain criticisms of the mimicry explanation it will be convenient to do so in connection with these five conditions which Wallace regarded as constant for all cases of mimetic resemblance.
(1) That the imitative species occur in the same area and occupy the very same station as the imitated.
This on the whole is generally true. It is well shewn in some of the most striking cases such as those of the Old-World Papilios that mimic Danaines, or of the Dismorphias and their Ithomiine models. In many of these cases the range of neither model nor mimic is a very wide one, yet the mimic is found strictly inside the area inhabited by the model. Papilio agestor, for instance, is only found where Caduga tytia occurs, nor is P. mendax known outside the area frequented by Euploea rhadamanthus. Even more striking in this respect are some of the Ithomiine-Dismorphia resemblances in the New World. The Ithomiine models are as a rule very local though very abundant. Two hundred miles away the predominant Ithomiine often bears quite a distinct pattern, and when this is the case the mimicking Dismorphia is generally changed in the same sense. But though mimic and model may be found together in the same locality, they do not always occupy the same station in the sense that they fly together. According to Seitz[25] the Dismorphias themselves do not fly with the Ithomiines which they mimic. The occurrence of butterflies is largely conditioned by the occurrence of the plants on which the larva feeds, and this is especially true of the female, which, as has already been noticed, is more commonly mimetic than the male. The female of Papilio polytes, for instance, is found flying where are to be found the wild citronaceous plants on which its larva feeds. On the other hand, its so-called models, Papilio hector and P. aristolochiae, are generally in the proximity of the Aristolochias on which their larvae feed. The two plants are not always found together, so that one frequently comes across areas where P. polytes is very abundant while the models are scarce or absent.
Though in the great majority of cases the imitator and the imitated occur in the same locality, this is not always so. The female of the Fritillary Argynnis hyperbius (Pl. IV, fig. 3), for instance, is exceedingly difficult to distinguish from Danais plexippus when flying, although when at rest the difference between the two is sufficiently obvious. Both insects are plentiful in Ceylon but inhabit different stations. The Danaid is a low-country insect, while the Fritillary is not found until several thousand feet up. The two species affect entirely different stations and hardly come into contact with each other. Where one is plentiful the other is not found. It has been suggested that migratory birds may have come into play in such cases. The bird learns in the low country that D. plexippus is unpleasant, and when it pays a visit to the hills it takes this experience with it and avoids those females of the Fritillary which recall the unpleasant Danaine.
Migratory birds have also been appealed to in another case where the resembling species are even further removed from one another than in the last case. Hypolimnas misippus is common and widely spread over Africa and Indo-Malaya, and the male (Pl. IV, fig. 8) bears a simple and conspicuous pattern—a large white spot bordered with purple on each of the very dark fore and hind wings. The same pattern occurs in the males of two other Nymphalines allied to H. misippus, viz. Athyma punctata and Limenitis albomaculata. The two species, however, have a distribution quite distinct from that of H. misippus, being found in China. It has nevertheless been suggested by Professor Poulton[26] that the case may yet be one of mimicry. According to his explanation, H. misippus is unpalatable, the well-known association of its female with Danais chrysippus being an instance of Müllerian mimicry. Migratory birds did the rest. Having had experience of H. misippus in the south, on their arrival in China they spared such specimens of Athyma punctata and Limenitis albomaculata as approached most nearly to H. misippus in pattern, and so brought about the resemblance. The explanation is ingenious, but a simpler view will probably commend itself to most. Other cases are known in which two butterflies bear a close resemblance in pattern and yet are widely separated geographically. Several species of the S. American Vanessid genus Adelpha are in colour scheme like the African Planema poggei which serves as a model for more than one species. The little S. American Phyciodes leucodesma would almost certainly be regarded either as a model for or a mimic of the African Neptis nemetes, did the two occur together. Nevertheless examples of close resemblance between butterflies which live in different parts of the world are relatively rare and serve to emphasise the fact that the great bulk of these resemblance cases are found associated in pairs or in little groups.
(2) That the imitators are always the more defenceless.
In the case of butterflies "defence" as a rule denotes a disagreeable flavour rendering its possessor distasteful to birds and perhaps to other would-be devourers. Feeding experiments with birds (cf. Chapter IX) suggest that certain groups of butterflies, notably the Danaines, Acraeines, Heliconines, Ithomiines and Pharmacophagus Papilios—groups from which models are generally drawn—are characterised by a disagreeable taste, while as a rule this is not true for the mimics. This distasteful quality is frequently accompanied by a more or less conspicuous type of coloration, though this is by no means always so. Many Euploeas are sombre inconspicuous forms, and it is only some of the Ithomiines that sport the gay colours with which that group is generally associated. The members of the distasteful groups usually present certain other peculiarities. Their flight is slower, they are less wary, their bodies are far tougher, and they are more tenacious of life. The slow flight is regarded as an adaptation for exhibiting the warning coloration to the best advantage, but from the point of view of utility it is plausible to suggest that the insect would be better off if in addition to its warning coloration it possessed also the power of swift flight[27]. It is possible that the peculiar slowness of flight of these unpalatable groups is necessitated by the peculiar tough but elastic integument which may present an insufficiently firm and resistant skeletal basis for sharp powerful muscular contraction, and so render swift flight impossible. It is stated that the flight of the mimics is like that of the model, and in some cases this is undoubtedly true. But in a great many cases it certainly does not hold good. Papilio clytia (Pl. I, figs. 7 and 8) is a strong swift flyer very unlike the Danaine and Euploeine which it is supposed to mimic. The flight of the female of Hypolimnas misippus (Pl. IV, fig. 7) is quite distinct from that of Danais chrysippus, while the mimetic forms of P. polytes fly like the non-mimetic one, a mode of flight so different from that of the two models that there is no difficulty in distinguishing them many yards away. Swift flight must be reckoned as one of the chief modes of defence in a butterfly, and on this score the mimic is often better off than the model. And of course it must not be forgotten that where the mode of flight is distinct the protective value of the resemblance must be very much discounted.
(3) That the imitators are always less numerous in individuals.
In the majority of cases this is certainly true. Probably all the Old-World Papilios that mimic Danaines are scarcer, and frequently very much scarcer, than their models. This is very evident from a study of the more comprehensive priced catalogues of Lepidoptera. The mimic is generally a more expensive insect than the model, and not infrequently it costs as many pounds as the model does shillings. But the rule is not universal. Papilio polytes is often much more common than either of its models. The remarkable Pierines, Archonias tereas and A. critias (Pl. XI, fig. 10) as a rule far outnumber the Pharmacophagus Swallow-tail which they mimic. Or again the Chalcosid moth Callamesia pieridoides[28] is a more abundant insect than the Bornean Pierine Delias cathara which it closely resembles.
It has sometimes been suggested in explanation of the greater abundance of the mimic that in such cases we are concerned with Müllerian mimicry, that since both of the species concerned are distasteful there is not, strictly speaking, either a mimic or a model, and consequently the relative proportions have not the significance that they possess where the mimicry is of the simple Batesian type. It is, however, very doubtful whether such an explanation is of any value, for, as will appear later, there are grave objections to accepting the current theory as to the way in which a resemblance is established on Müllerian lines (cf. pp. 72-74).
(4) That the imitators differ from the bulk of their allies.
What importance we attach to this condition must depend upon our interpretation of the word "allies"—whether, for example, we use it for a small group of closely connected species, for a genus, for a group of genera, or in an even wider sense. Perhaps an example will serve to make the difficulty more clear. As already noticed, the S. American genus Dismorphia belongs to the family of Pieridae or "whites." Also certain species of Dismorphia bear a close resemblance to certain species of Ithomiines, a noteworthy example being D. praxinoe and Mechanitis saturata (Pl. X, figs. 3 and 7), in which the pattern, colour, and shape of the two species are all far removed from what is usually understood by a "white." It must not be forgotten, however, that these matters are generally discussed by European naturalists who have grown up in a region where the majority of the "whites" are more or less white. For this reason the statement that D. praxinoe differs from the bulk of its allies is likely to meet with general acceptance, especially as some of the species of the genus itself (e.g., D. cretacea, Pl. X, fig. 1) are regular whites in appearance. But when we come to look at the genus Dismorphia as a whole the matter assumes another complexion. Seitz[29] recognises 75 species of which about a dozen are predominantly white. The rest present a wonderful diversity of colour and pattern. Black predominates on the fore wings, and the insect is frequently marked with gay patches of yellow, bright brown, scarlet, or blue. Forms which from their colour are clearly not mimics present nevertheless the general pattern and shape of other forms which bear a strong resemblance to some Ithomiine. Sometimes a change of colour in certain patches from blue or yellow to bright brown would make all the difference between a non-imitative and an imitative species. Moreover, the non-imitative forms frequently have the peculiar narrow wing, so unusual in a Pierine, which enhances the resemblance of the mimicking species to the Ithomiine model, and which to some extent occurs even in D. cretacea. Clearly we are not justified in saying that D. praxinoe differs from the bulk of its allies, for inside the genus there are many non-imitative species which differ from it in some particulars and are alike it in others. There is a distinct family resemblance among the bulk of the Dismorphias, including practically all the mimetic forms, and on the whole the resemblances between the imitative and the non-imitative forms are as noteworthy as the differences. Though not exhibited in so striking a fashion, the same is to a large extent true of a large proportion of the cases of mimicry. It is on the whole unusual to find cases where a single species departs widely from the pattern scheme of the other members of the genus and at the same time resembles an unrelated species. Two of the best instances are perhaps those of Limenitis archippus (p. 49) and of the Pierid Pareronia (p. 23). Of the total number of mimicry instances a high proportion is supplied by relatively few groups. In each region several main series of models and mimics run as it were parallel to one another. In Asia, for example, we have the Papilio-Danaine series where the colour-patterns of a series of Danaines, all nearly related, are closely paralleled by those of a section of the genus Papilio, and by those of the Satyrid genus Elymnias. In Africa there is a similar Papilio-Danaine series though of less extent. Africa has a group of models not found in Asia, and the Papilio-Danaine series is as it were curtailed by the Papilio-Planema series with which to some extent runs parallel the genus Pseudacraea. These phenomena of parallel series have been mentioned here as shewing that mimicry tends to run in certain groups and that in many cases at any rate little meaning can be attached to the statement that the imitators differ from the bulk of their allies.
The fifth of Wallace's conditions is clear and needs no discussion.
It is evident that at any rate a large proportion of the instances of close resemblance do not fulfil all of the conditions laid down by Wallace. Nevertheless we should expect them to do so if the resemblance has been brought about by the cumulative effect of natural selection on small favourable variations. Clearly there is a prima facie case for doubting whether we must of necessity ascribe all resemblance of the kind to natural selection, and in the next few chapters we shall discuss it in more detail from several points of view.
CHAPTER VI
"MIMICRY RINGS"
Having reviewed briefly some of the most striking phenomena of what has been termed mimicry, we may now inquire whether there are good grounds for supposing that these resemblances have been brought about through the operation of natural selection or whether they are due to some other cause. If we propose to offer an explanation in terms of natural selection we are thereby committed to the view that these resemblances are of the nature of adaptation. For unless we grant this we cannot suppose that natural selection has had anything to do either with their origin or with their survival. Granting then for the present the adaptational nature of these mimetic resemblances, we may attempt to deduce from them what we can as to the mode of operation of natural selection. In doing so we shall bear in mind what may be called the two extreme views: viz. (a) that the resemblance has been brought about through the gradual accumulation of very numerous small variations in the right direction through the operation of natural selection, and (b) that the mimetic form came into being as a sudden sport or mutation, and that natural selection is responsible merely for its survival and the elimination of the less favoured form from which it sprang.
There is a serious difficulty in the way of accepting the former of these two views. If our two species, model and would-be mimic are, to begin with, markedly different in pattern, how can we suppose that a slight variation in the direction of the model on the part of the latter would be of any value to it? Take for example a well-known South American case—the resemblance between the yellow, black, and brown Ithomiine, Mechanitis saturata (Pl. X, fig. 7) and the Pierine, Dismorphia praxinoe (Pl. X, fig. 3). The latter belongs to the family of the "whites," and entomologists consider that in all probability its ancestral garb was white with a little black like the closely allied species D. cretacea (Pl. X, fig. 1). Can we suppose that in such a case a small development of brown and black on the wings would be sufficient to recall the Ithomiine and so be of service to the Dismorphia which possessed it? Such a relatively slight approach to the Ithomiine colouring is shewn by the males of certain South American "whites" belonging to the genus Perrhybris (Pl. X, figs. 4 and 5). But the colour is confined to the under-surface and the butterflies possessing it could hardly be confused with a Mechanitis more than their white relations which entirely lack such a patch of colour. If birds regarded white butterflies as edible it is difficult to suppose that they would be checked in their attacks by a trifling patch of colour while the main ground of the insect was still white. But unless they avoided those with the small colour patch there would be an end of natural selection in so far as the patch was concerned, and it would have no opportunity of developing further through the operation of that factor. This is the difficulty of the initial variation which has been clearly recognised by most of the best known supporters of the theory of mimicry. Bates himself offered no suggestion as to the way in which such a form as a Pierid could be conceived of as beginning to resemble an Ithomiine[30]. Wallace supposed that the Ithomiines were to start with not so distinct from many of the edible forms as they are to-day, and that some of the Pierines inhabiting the same district happened to be sufficiently like some of the unpalatable forms to be mistaken for them occasionally[31].
The difficulty of the initial variation had also occurred to Darwin, and he discusses it in an interesting passage which is so important that we may quote it here in full:
It should be observed that the process of imitation probably never commenced between forms widely dissimilar in colour. But starting with species already somewhat like each other, the closest resemblance, if beneficial, could readily be gained by the above means; and if the imitated form was subsequently and gradually modified through any agency, the imitating form would be led along the same track, and thus be altered to almost any extent, so that it might ultimately assume an appearance or colouring wholly unlike that of the other members of the family to which it belonged. There is, however, some difficulty on this head, for it is necessary to suppose in some cases that ancient members belonging to several distinct groups, before they had diverged to their present extent, accidentally resembled a member of another and protected group in a sufficient degree to afford some slight protection; this having given the basis for the subsequent acquisition of the most perfect resemblance[32].
Both Darwin and Wallace recognised clearly this difficulty of the initial variations, and both suggested a means of getting over it on similar lines. Both supposed that in general colour and pattern the groups to which model and mimic belonged were far more alike originally than they are to-day. They were in fact so much alike that comparatively small variations in a favourable direction on the part of the mimic would lead to its being confused with the unpalatable model. Then as the model became more and more conspicuously coloured, as it developed a more and more striking pattern warning would-be enemies of its unpleasant taste, the mimic gradually kept pace with it through the operation of natural selection, in the shape of the discriminating enemy, eliminating those most unlike the model. The mimic travelled closely in the wake of the model, coaxed as it were by natural selection, till at last it was far removed in general appearance from the great majority of its near relations.
In this way was offered a comparatively simple method of getting over the difficulty of applying the principle of natural selection to the initial variations in a mimetic approach on the part of one species to another. But it did not escape Darwin's penetration that such an argument would not always be easy of application—that there might be cases where a given model was mimicked by members of several groups of widely differing ancestral pattern, and that in these cases it would be difficult to conceive of members of each of the several groups shewing simultaneous variations which would render them liable to be mistaken for the protected model. The difficulty may perhaps be best illustrated if we consider a definite case.
It is a feature of mimetic resemblances among butterflies that a given species in a given locality may serve as a model for several other species belonging to unrelated groups. Generally such mimics belong to presumably palatable species, but other presumably unpalatable species may also exhibit a similar coloration and pattern. In this way is formed a combine to which the term "mimicry ring" has sometimes been applied. An excellent example of such a mimicry ring is afforded by certain species of butterflies in Ceylon, and is illustrated on Plate IV. It is made up in the first place of two species belonging to the presumably distasteful Danaine group, viz. Danais chrysippus and D. plexippus. The latter is a rather darker insect but presents an unmistakable general likeness to D. chrysippus. Those who believe in Müllerian mimicry would regard it as an excellent example of that phenomenon. For those who believe only in Batesian mimicry D. plexippus, being the scarcer insect, must be regarded as the mimic and D. chrysippus as the model. In both of these species the sexes are similar, whereas in the other three members of the "ring" the female alone exhibits the resemblance. One of these three species is the common Nymphaline, Hypolimnas misippus, of which the female bears an extraordinary likeness to D. chrysippus when set and pinned out on cork in the ordinary way. The male, however (Pl. IV, fig. 8), is an insect of totally different appearance. The upper surfaces of the wings are velvety black with a large white patch bordered with purple in the middle of each[33]. The "ring" is completed by the females of Elymnias undularis and Argynnis hyperbius. The former of these belongs to the group of Satyrine butterflies and the female is usually regarded as a mimic of D. plexippus, which it is not unlike in so far as the upper surface of the wings is concerned. Here again the male is an insect of totally dissimilar appearance. Except for a border of lighter brown along the outer edges of the hind wings the upper surface is of a uniform deep purple-brown all over (Pl. IV, fig. 6). In Argynnis hyperbius the appearance is in general that of the Fritillary group to which it belongs. But in the female the outer portion of the fore wings exhibits much black pigment and is crossed by a broad white band similar to that found in the same position on the wing of D. plexippus (Pl. IV, fig. 2).
Of the five species constituting this little "mimicry ring" in Ceylon two, on the current theory of mimicry, are to be regarded as definitely unpalatable, one (H. misippus) as doubtfully so, while the Satyrine and the Fritillary are evidently examples of simple or Batesian mimicry.
Now such examples as this of simultaneous mimicry in several species are of peculiar interest for us when we come to inquire more closely into the process by which the resemblances can be supposed to have been brought about. Take for example the case of E. undularis. The male is evidently an unprotected insect in so far as mimicry is concerned, while the female exhibits the general pattern and coloration characteristic of the warningly coloured and presumably distasteful species D. plexippus or D. chrysippus. If we are to suppose this to have been brought about by the operation of natural selection it is clear that we must regard the colour and pattern of the male as the original colour and pattern of both sexes. For natural selection cannot be supposed to have operated in causing the male to pass from a protected to an unprotected condition, or even in causing him to change one unprotected condition for another. Probably all adherents of the mimicry theory would be agreed in regarding the male of Elymnias undularis as shewing the ancestral coloration of the species, and in looking upon the female as having been modified to her own advantage in the direction of D. plexippus. The question that we have to try to decide is how this has come about—whether by the accumulation of slight variations, or whether by a sudden change or mutation in the pattern and colour of the female by which she came to resemble closely the Danaine. It is clear that if D. plexippus were what it is to-day before the mimetic approach on the part of E. undularis began, small variations in the latter would have been of no service to it. The difference between the two species would have been far too great for individuals exhibiting slight variation in the direction of D. plexippus to stand any chance of being confused with this species. And unless such confusion were possible natural selection could not work. There is, however, an immediate way out of the difficulty. We may suppose that the coloration of the male of the mimic, E. undularis, is not only the ancestral colour of its own species but also of the model. D. plexippus on this supposition was very like E. undularis, of which both sexes were then similar to what the male is to-day. The pattern is, however, an inconspicuous one, and it can be imagined that it might be to the advantage of D. plexippus to don a brighter garb for the advertisement of its unpleasant qualities. Variations in the direction of a more conspicuous pattern would for that reason tend to be preserved by natural selection, until eventually was evolved through its means the well-marked pattern so characteristic of the model to-day. If in the meantime variations in the same direction occurred among the females of E. undularis these would tend to be preserved through their resemblance to the developing warning pattern of the distasteful Danaine model. The development of model and mimic would proceed pari passu, but if the sexes of the mimic differ, as in this case, we must suppose the starting-point to have been the condition exhibited by the male of the mimicking species.
But Argynnis hyperbius is also a species in which the female mimics D. plexippus; and by using the same argument as that just detailed for Elymnias undularis we can shew that the Danaine model, D. plexippus, must also have been like the male of Argynnis hyperbius. And if the resemblance of A. hyperbius was developed subsequently to that of E. undularis, then both D. plexippus and E. undularis must at one time have been like the male of A. hyperbius, a proposition to which few entomologists are likely to assent. Further, since the female of H. misippus also comes into the plexippus-chrysippus combine we must suppose that these species must at some time or another have passed through a pattern stage like that of the misippus male.
It is scarcely necessary to pursue this argument further, for even the most devoted adherents of the theory of mimicry as brought about by the operation of natural selection on small variations are hardly likely to subscribe to the phylogenetic consequences which it must entail in cases where a model is mimicked by the females of several species whose males are widely dissimilar in appearance.
Even if we suppose the two Danaines to have been originally like the male of one of the three mimics, we must still suppose that the females of the other two originated as "sports," sufficiently near to Danaines to be confused with them. But if such sports can be produced suddenly by some mutational process not at present understood, why should not these sports be the females of the three mimicking species as we see them at present? Why need we suppose that there were intermediate stages between the mimicking female and the original hypothetical female which was like the male? If a sport occurred which was sufficiently similar to an unpalatable species to be confused with it, it is theoretically demonstrable that, although relatively scarce to start with, it would rapidly increase at the expense of the unprotected male-like female until the latter was eliminated. We shall, however, return in a later chapter (p. 96) to the argument by which this view can be supported.
So far we have discussed what we called the two extreme views as to the way in which a mimetic resemblance may be supposed to have originated. Of the two that which assumes the resemblance to have been brought about by a succession of slight variations must also assume that model and mimic were closely alike to start with, and this certainly cannot be true in many cases. On the other hand, there is so far no reason against the idea of supposing the resemblance to have originated suddenly except what to most minds will probably appear its inherent improbability.
There are writers on these questions of mimicry who adopt a view more or less intermediate between those just discussed. They regard the resemblance as having arisen in the first place as a sport of some magnitude on the part of the mimic, rendering it sufficiently like the model to cause some confusion between the two. A rough-hewn resemblance is first brought about by a process of mutation. Natural selection is in this way given something to work on, and forthwith proceeds to polish up the resemblance until it becomes exceedingly close. Natural selection does not originate the likeness, but, as soon as a rough one has made its appearance, it comes into operation and works it up through intermediate stages into the finished portrait. It still plays some part in the formation of a mimetic resemblance though its rôle is now restricted to the putting on of the finishing touches. Those who take this view hold also that the continued action of natural selection is necessary in order to keep the likeness up to the mark. They suppose that if selection ceases the likeness gradually deteriorates owing to the coming into operation of a mysterious process called regression. This idea involves certain conceptions as to the nature of variation which we shall discuss later.
Though it is difficult to regard Batesian mimicry as produced by the accumulation of small variations through natural selection, it is perhaps rather more plausible to suppose that such a process may happen in connection with the numerous instances of Müllerian mimicry. For since the end result is theoretically to the advantage of both species instead of but one, it is possible to argue that the process would be simplified by their meeting one another halfway, as Müller[34] himself originally suggested. Variations on the part of each in the direction of the other would be favourably selected, the mimicry being reciprocal.
Difficulties, however, begin to arise when we inquire into the way in which this unification of pattern may be conceived of as having come about. By no one have these difficulties been more forcibly presented than by Marshall[35] in an able paper published a few years ago, and perhaps the best way of appreciating them is to take a hypothetical case used by him as an illustration.
Let us suppose that in the same area live two equally distasteful species A and B, each with a conspicuous though distinct warning pattern, and each sacrificing 1000 individuals yearly to the education of young birds. Further let it be supposed that A is a common species of which there are 100,000 individuals in the given area, while B is much rarer, and is represented by 5000. The toll exacted by young birds falls relatively more lightly upon A than upon B, for A loses only 1%, whereas B's loss is 20%. Clearly if some members of B varied so that they could be mistaken for A it would be greatly to their advantage, since they would pass from a population in which the destruction by young birds was 20% to one in which it would now be rather less than 1%. Moreover, as the proportion of B resembling A gradually increased owing to this advantage, the losses suffered by those exhibiting the original B pattern would be relatively heavier and heavier until the form was ultimately eliminated. In other words, it is theoretically conceivable that of two distasteful species with different patterns the rarer could be brought to resemble the more abundant.
We may consider now what would happen in the converse case in which the more numerous species exhibited a variation owing to which it was confused with the rarer. Suppose that of the 100,000 individuals of A 10,000 shewed a variation which led to their being mistaken for B, so that there are 90,000 of the A pattern and 15,000 of the B pattern of which 10,000 belong to species A. A will now lose 1000 out of the 90,000 having the A pattern, and ⅔ × 1000 out of the 10,000 of species A which exhibit the B pattern. The toll of the birds will be 1⁄90 of those keeping the original A pattern, and 2⁄30 of those of species A which have assumed the B pattern. The mortality among the mimetic members of A is six times as great as among those which retain the type form. It is clear therefore that a variation of A which can be mistaken for B is at a great disadvantage as compared with the type form[36], and consequently it must be supposed that the Müllerian factor, as the destruction due to experimental tasting by young birds is termed, cannot bring about a resemblance on the part of a more numerous to a less numerous species. Further, as Marshall goes on to shew, there can be no approach of one species to the other when the numbers are approximately equal. A condition essential for the establishing of a mimetic resemblance on Müllerian lines, no less than on Batesian, is that the less numerous species should take on the pattern of the more numerous. Consequently the argument brought forward in the earlier part of this chapter against the establishing of such a likeness by a long series of slight variations is equally valid for Müllerian mimicry[37].
CHAPTER VII
THE CASE OF PAPILIO POLYTES
Many instances of mimicry are known to-day, but comparatively few of them have been studied in any detail. Yet a single carefully analysed case is worth dozens which are merely superficially recorded. In trying to arrive at some conception of the way in which the resemblance has come about we want to know the nature and extent of the likeness in the living as well as in the dead; the relative abundance of model and mimic; what are likely enemies and whether they could be supposed to select in the way required, whether the model is distasteful to them; whether intermediate forms occur among the mimics; how the various forms behave when bred together, etc., etc. Probably the form that from these many points of view has, up to the present, been studied with most care is that of the Swallow-tail, Papilio polytes. It is a common butterfly throughout the greater part of India and Ceylon, and closely allied forms, probably to be reckoned in the same species, reach eastwards through China as far as Hongkong. P. polytes is one of those species which exhibit polymorphism in the female sex. Three distinct forms of female are known, of which one is like the male, while the other two are very different. Indeed for many years they were regarded as distinct species, and given definite specific names. To Wallace belongs the credit of shewing that these three forms of female are all to be regarded as wives of the same type of male[38]. He shewed that there were no males corresponding to two of the females; also that the same one male form was always to be found wherever any of the females occurred. As the result of breeding experiments in more recent years Wallace's conclusions have been shewn to be perfectly sound.
The male of polytes (Pl. V, fig. 1) is a handsome blackish insect with a wing expanse of about four inches. With the exception of some yellowish-white spots along their outer margin the fore wings are entirely dark. Similar spots occur along the margin of the hind wing also, while across the middle runs a series of six yellowish-white patches producing the appearance of a broad light band. The thorax and abdomen are full black, though the black of the head is relieved by a few lighter yellowish scales. The under surface is much like the upper, the chief difference being a series of small and slightly reddish lunules running outside the light band near the margin of the hind wing (Pl. V, fig. 1 a). In some specimens these markings are almost absent. One form of female is almost exactly like the male (Pl. V, fig. 2), the one slight difference being that the lunules on the under surface of the hind wing are generally a trifle larger. For brevity she may be called the M form. The second form of female differs in many respects from the male and the M female. Instead of being quite dark, the fore wings are marked by darker ribbed lines on a lighter ground[39] (Pl. V, fig. 3). The hind wings shew several marked differences from those of the male. Of the series of six patches forming the cross band the outermost has nearly disappeared, and the innermost has become smaller and reddish. The middle four, on the other hand, have become deeper, reaching up towards the insertion of the wing, and are pure white. A series of red lunules occurs on the upper surface outside the white band, and the yellowish-white marginal markings tend to become red. These differences are equally well marked on the under surface (Pl. V, fig. 3 a). The colour of the body, however, remains as in the male. From the resemblance shewn by this form to another species of Swallow-tail, Papilio aristolochiae (Pl. V, fig. 5), we shall speak of it as the A form.
The third form of female is again very distinct from the other two. The fore wings are dark but are broken by an irregular white band running across the middle (Pl. V, fig. 4), and there is also an irregular white patch nearer the tips of the wing. The hind wings, on the other hand, are characterised by having only red markings. The yellowish-white band of the male is much reduced and is entirely red, while the red lunules are much larger than in the A form. The under surface (Pl. V, fig. 4 a) corresponds closely with the upper. The body remains black as in all the other forms. This type of female bears a resemblance to Papilio hector (Pl. V, fig. 6), and for that reason we shall speak of it as the H form. It should be added that these three forms of female are quite indistinguishable in the larval and chrysalis stages.
It was Wallace who first offered an explanation of this interesting case in terms of mimicry. According to this interpretation P. polytes is a palatable form. The larva, which feeds on citronaceous plants, and the chrysalis are both inconspicuous in their natural surroundings. They may be regarded as protectively coloured, and consequently edible and liable to persecution. The original coloration is that of the male and the M female. From this the other two forms of female have diverged in the direction of greater instead of less conspicuousness, although the presumed edibility of the insect might have led us to think that a less conspicuous coloration would have been more to its advantage. But these two females resemble the two species Papilio aristolochiae and Papilio hector, which, though placed in the same genus as P. polytes, belong to a very different section of it[40]. The larvae of these two species are conspicuously coloured black and red with spiny tubercles. They feed upon the poisonous Aristolochia plants. For these reasons and also from the fact that the butterflies themselves are both conspicuous and plentiful it is inferred that they are unpalatable. In short, they are the models upon which the two polytes females that are unlike the male have been built up by natural selection.
The suggestion of mimicry in this case is supported by the fact that there is a general correspondence between the areas of distribution of model and mimic. P. hector is not found outside India and Ceylon, and the H female of P. polytes is also confined to this area. P. aristolochiae, on the other hand, has a much wider range, almost as wide indeed as that of P. polytes itself. Generally speaking the A female accompanies P. aristolochiae wherever the latter species is found. Beyond the range of P. aristolochiae, in northern China, the M female alone is said to occur. On the other hand, as the matter comes to be more closely studied exceptions are beginning to turn up. The H female, for instance, is found on the lower slopes of the Himalayas, far north of the range of P. hector, and there are indications that a careful study of the distribution in China and Japan may prove of importance.
Moreover, the investigation of a smaller area may also bring to light points of difficulty. In Ceylon, for example, P. polytes is common up to several thousand feet, while P. hector is rare at half the height to which polytes ascends. Nevertheless the H form of female is relatively just as abundant up-country where hector is rarely found as it is low down where hector is plentiful[41]. On the other hand, P. aristolochiae may be exceedingly abundant at altitudes where hector is scarce. Yet the A form of polytes is no more relatively abundant here than elsewhere on the island. All over Ceylon, in fact, the relative proportions of the three forms of female appear to be the same, quite irrespective of the abundance or scarcity of either of the models. As, however, we shall have to return to this point later, we may leave it for the moment to consider other features of this case of P. polytes.
In collections of insects from India or Ceylon it is not unusual to find specimens of the A form of female of polytes placed with P. aristolochiae, and the H form with P. hector. When the insects are old and faded and pinned out on cork the mistake is a very natural one. But after all the enemies of polytes do not hunt it in corked cabinets, and any estimation of resemblance to be of use to us must be based upon the living insects. Are the resemblances of the mimics to the models when alive so close that they might be expected to deceive such enemies[42] as prey upon them and have no difficulty in distinguishing the male form of polytes from P. aristolochiae or P. hector?
To answer for a bird is a hazardous undertaking. We know so little of the bird's perceptive faculties whether of taste or sight. But on general grounds, from the specialization of their visual apparatus, it is probable that the sense of sight is keen, though whether the colour sense is the same as our own is doubtful[43]. On the other hand, the olfactory apparatus is relatively poorly developed in birds, and from this we can only argue that the senses of smell and taste are not especially acute. Really we can do little more than to describe how these mimetic resemblances appear to our own senses, and to infer that they do not appear very different to the bird. If there is any difference in keenness of perception we shall probably not be far wrong in presuming that the advantage rests with the bird. After all if there is any truth in the theory of mimicry the bird has to depend largely upon its keenness of sight in making its living, at any rate if that living is to be a palatable one. If natural selection can bring about these close resemblances among butterflies it must certainly be supposed to be capable of bringing the bird's powers of vision to a high pitch of excellence.
Returning now to the case of P. polytes, there is not the least doubt that to the ordinary man accustomed to use his eyes the A form of female is easily distinguishable from P. aristolochiae, as also is the H form from P. hector. The two models have a feature in common in which they both differ from their respective mimics. In both of them the body and head are largely of a brilliant scarlet, whereas neither of the mimics has a touch of red on the body. In the living insect when the body is swelled by its natural juices the effect is very striking[44]. It gives at once a "dangerous" look to the insect when settled, even at a distance of several yards, and this although one may be perfectly familiar with its harmless nature. The mimics on the other hand with their sombre-coloured bodies never look otherwise than the inoffensive creatures that they are. The "dangerous" look due to the brilliant scarlet of the body and head of hector and aristolochiae is reinforced by the quality of the red on the markings of the wings. In both models it is a strong clamorous red suggestive of a powerful aniline dye, whereas such red as occurs in the mimics is a softer and totally distinct colour. The difference in quality is even more marked on the under than on the upper surface (Pl. V, figs. 3 a—6 a), and the net result is that when settled, with wings either expanded or closed, there is no possibility of an ordinarily observant man mistaking mimic for model in either case, even at a distance of several yards.