MIMETIC ASSOCIATIONS 



33 



from there. The observation throws much light upon the comparatively recent intrusion 

 of the species into South America, and its even later spread to the Canary Islands, and 

 goes far to explain its extraordinarily wide distribution in the Old World.' 



H. misippus presents a certain amount of difficulty when viewed as an instance of 

 mimicry, inasmuch as the geographical distribution of its varieties presents a noticeable 

 lack of coincidence with that of the varieties of D. chrysippus. Thus the inaria form of 

 misippus is commoner in South Africa than the type, while the dorippus form of chrysippus 

 is rare. In West Africa the alcippus form of chrysippus is predominant, whilst here the 

 inaria form of misippus is common. In British East Africa there is a better correspondence, 

 but on the whole the varieties of the two species do not appear to show any well-marked 

 coincidence. Those who are opposed to the theory of mimicry, and who consider that the 

 resemblance between butterflies of different genera is merely accidental, may consider the 

 present case as strongly corroborative of their views. A little consideration, however, will 

 show that this is not the case. In the absence of much needed experiment we must argue 

 from probabilities. We have then the following facts. A butterfly, the ancestral appearance 

 of which, as shown by the male and its allied species, is black with conspicuous white markings, 

 shows in the female an entire departure from the family facies, and becomes practically 

 indistinguishable, on the wing at least, from a widely distributed and apparently protected 

 species of an entirely different genus. Such a phenomenon alone could not be accidental, but 

 the resemblance goes much further. The model presents no less than three well-marked 

 varieties, each of which is faithfully copied by a corresponding form of the mimic, and it 

 would be straining a point to argue, because we are ignorant of the causes which have since 

 led to the present geographical distribution of the varieties, that therefore the extraordinary 

 resemblance between them is merely accidental. D. chrysippus is accompanied throughout 

 its range by H. misippus, and although the latter is a recent arrival in America where the for- 

 mer is unknown, there may be different natural conditions in that country which combined 

 with its migratory habits have enabled it to become established. Even if we were to assume 

 the highly improbable hypothesis of accidental resemblance between the two forms, we 

 must still find some explanation of the fact that D. chrysippus is more or less accurately 

 .imitated in outward appearance by a considerable number of other widely different species 

 of butterflies which will shortly be described. It may be argued that a tendency to albinism 

 is common in many butterflies, and that, therefore, the coexistence of chrysippus alcippus 

 and misippus alcippoides can be explained on the ground that the albinistic tendency has 

 produced in this case a kind of parallel variation. The extreme improbability of two species 

 arriving at the same result by a process of parallel variation is pointed out in a later portion 

 of this work. But if it were shown that a tendency to albinism exists in H. misippus, then 

 that tendency merely becomes the material out of which a mimetic resemblance has been 

 fashioned. Granted that H. misippus has produced its various forms in imitation of D. chry- 

 sippus, its wide distribution still, however, demands some explanation. Such an explanation 

 is rather difficult to find if the species be regarded purely as a Batesian mimic, but there is 

 considerable doubt as to whether it can be so regarded. Most species which are really 

 palatable to insect enemies and depend for their existence on resemblance to protected forms 

 are comparatively rare. They occur in isolated instances amongst swarms of their un- 

 palatable models. H. misippus is a notable exception. It has established itself in America, 

 where its model is unknown. The wide range of the insect may be due in some part to 

 the abundance of its model, but the ease with which it appears to maintain large 

 numbers of individuals points to the possession of at least some degree of unpalatability. 



1200 E 



