674 



NA TURE 



[October 31, 1907 



resemblance to their more abundant models, the immunity 

 of which, he thought, was due to the possession of some 

 distasteful quality — probably a scent or flavour disliked 

 bv the birds. 



Accepting Darwin's view of evolutionary process, he 

 attributed the formation of these resemblances to the 

 accumulation by natural selection of variations in the 

 mimicking species that happened to point in the appro- 

 priate direction ; so that these mimics had gradually put 

 off the general aspect of the group to which they properly 

 belonged, and had become more or less completely 

 assimilated in outward appearance to the members of an 

 entirely dilTerent assemblage ; thus sailing, as it were, 

 under false colours, as if a peaceable merchantman were 

 to disguise itself under the rig and ensign of a man-of- 

 war. This is the well-known Batesian theory of mimicry. 

 It was at once, and cordially, accepted by Darwin ; while 

 the array of facts from South America on which it was 

 based was soon afterwards shown to be paralleled by 

 corresponding phenomena in the Malayan Archipelago and 

 in South Africa. This was the work of two great 

 naturalists happily still with us, Alfred Russel Wallace 

 and Roland Trimen. 



.An objection was raised in early times to Bates's view 

 on the ground that it was difficult to account for the first 

 advances towards the formation of a mimetic pattern. 

 This objection was felt in some degree both by Darwin 

 and by Fritz Miiller, of whom we shall hear more 

 presently. Darwin and Miiller thought that the objection 

 might be met by supposing a considerable original like- 

 ness between mimic and model ; it can, however, quite 

 easily be shown from forms actually at present existing 

 that a complete series of gradations may occur between 

 the ordinary type of a mimetic genus and its very distinct- 

 looking model or models. The transitional forms, even 

 those exhibiting the earliest stages of mimetic assimil- 

 ation, are evidently able to maintain themselves (how they 

 do it we shall see later), and they in many cases form 

 a perfect succession of links between extreme forms of 

 the utmost divergence in aspect. Hence it is unnecessary 

 to suppose that a considerable initial resemblance inust 

 exist between mimic and model, while the initial stages 

 of the mimetic pattern, however we are to account for 

 them, are not only theoretically possible, but are found 

 to be in actual existence. 



The beauty and simplicity of Bates's theory commended 

 it strongly to public acceptance, and it is probable that 

 to this day, when the subject of mimicry is mentioned, it 

 is the Batesian theory that presents itself to most people's 

 minds. 



But notwithstanding the immense value of Bates's 

 contribution to knowledge, it is now evident, as we shall 

 see, that he only touched the fringe of a great subject, 

 and that a much wider view is necessary before the facts 

 observed by him, and subsequently by others, can be fullv 

 explained. 



Those who read Bates's classical paper cannot avoid 

 remarking that he himself was not thoroughly happv 

 about all the facts there recorded. He directs attention to 

 the circumstance that not only do the mimics resemble 

 their models, but that the models themselves often show 

 an extraordinary resemblance to each other. He speaks 

 of " a minute and palpably intentional likeness which is 

 perfectly staggering." 



To take an instance : two species of the Ithomiine 

 genus Dircenna, D. epidero anti D. rhaeo, structurally 

 distinct, but almost indistinguishable on the wing, were 

 noted by Bates as being always found together where 

 they occur in the Amazonian region. A moth, Hvclnsia 

 tiresias, was regarded by Bates as a mimic of DiVroiiKT 

 epidero, but it did not escape him that his theorv failed 

 to account for the resemblance of the two Dircennas to 

 one another, the subfamily Ithomiina^. to which thev 

 belong, being on good grounds supposed to be generally 

 djstastpful. The difficulty becomes still greater when it 

 i<^ realised that not only members of the same presumably 

 dKlastcful genus, but also members of different genera, 

 all with the same habits and denizens of the same region, 

 Vvear the same extraordinary likeness to each other. 

 There are. for example, some twenty species of Ithomiines, 

 belonging to no less than seven different genera, all with 

 NO. 1983, VOL. 76] 



the same, or very nearly the same, external appearance. 

 But this is not all, for the same mimetic assemblage will 

 be found to include, not only these Ithomiines, but also 

 butterflies belonging to the group of Danaines (genus 

 Ituna) and Pierines (genus Dismorphia), as well as moths 

 of the two widely separated groups of Hypsidae and 

 Castniada;, all with a common facies. 



If it were merely a case of resemblance between two 

 or more species of the same genus, such as the Dircennas 

 that have just been mentioned, we might be tempted to 

 say that the resemblance was merely due to affinity, and 

 to explain, as Bates did, the circumstance of the constant 

 companionship of the two species by appealing to the 

 ■■ social and gregarious instincts of the group." When, 

 however, we see that not only Dircennas, but Ithomiines 

 generally, Danaines, Pierines, and moths all come into 

 the same mimetic assemblage, the explanation from 

 affinity breaks down. Affinity, no doubt, may help 

 mimicry, but there is no necessary connection between the 

 two. Some members of the company are closely related ; 

 others are widely distinct. Bates himself saw clearly 

 enough that his theory of one distasteful and immune 

 form sheltering others which would be attacked if de- 

 tected would not apply to cases of this kind. If all the 

 species but one of a " homoeochromatic " group are to be 

 considered as edible mimics, we should have to account 

 for the fact that they vastly outnumber the model, in 

 which case the mimicry would be more harmful to the 

 model than beneficial to themselves ; we should also have 

 to face the improbability of one species of a genus being 

 distasteful and immune, while other species of the same 

 and allied genera were edible and liable to attack. It 

 was plain that the distasteful models did really imitate 

 each other, but why? 



All that Bates could do in the face of this difficulty was 

 to fall back, somewhat doubtfully, on the hypothesis of 

 some local or climatic cause acting equally upon the 

 forms of different groups, and in some unexplained way 

 bringing about this strange resemblance between them. 

 In this supposition he was for a time followed by Wallace. 



It is not to be denied that there is a certain plausibility 

 at first sight in this view concerning the direct action of 

 external conditions. It is, for example, a striking fact 

 that the members of a mimetic group of very diverse 

 affinities will, as Bates says, every few hundred miles all 

 change their hue and pattern together, "as if by the 

 touch of an enchanter's wand." 



There is a well-marked assemblage of this kind, 

 generally characterised by a pattern composed of the three 

 colours red, yellow, and black. It contains, besides 

 moths, butterflies of many diverse group-;. Ithomiines. 

 Heliconiines, Danaines, Nymphalines, and Pierines — in 

 some of the latter the female only taking part in the 

 mimetic cluster, a point to which we shall return later 

 on. The members of this assemblage as it occurs in the 

 northern part of Central America — Guatemala to 

 Nicaragua — present in common a remarkable streakiness 

 of pattern, a feature that makes them easily recognisable 

 among the corresponding forms from other regions of the 

 same continent. Passing on to Venezuela, we find among 

 the geographical races, or, if we like to call them so, the 

 representative species, that there replace the Central 

 American forms, a tendency to the break!ng-up of the 

 streaks, and a slight encroachment of the red ground- 

 colour upon the yellow of the apex. In Trinidad there 

 occurs a general paling of the ground-colour, due to an 

 increase of yellow pigmentation, and running, as before, 

 through the entire group. Next, taking the correspond- 

 ing Guiana forms, we find a further breaking-up of the 

 streaks into spots, and also a general darkening, especially 

 of the hindwings, which gives a most characteristic 

 aspect to the whole assemblage. In East Brazil we have 

 a modification which somewhat recalls the Trinidad facies, 

 though here the yellow streak on the hindwing is better 

 defined, and the black of the apex is less broken up. At 

 Ega. on the Upper Amazon, a curious dark chestnut tinge 

 pervades the group, while in Peru a characteristic spotti- 

 ness takes the place of the streaky pattern we saw else- 

 where, and the apex becomes more uniformly dark. 

 Finally, in Ecuador the streaks have all but disappeared, 

 and even the spots have become almost blocked out by a 



