1 Cha Uenger 1 Lep idoptera . 405 



with this question, and that is the fact of the apparent rarity 

 of these copying Papilionidae. If it be of great importance 

 for one species to resemble another, inasmuch as that thereby 

 the copying species shares, in common with its model, immu- 

 nity from evil, one would naturally suppose that this advantage 

 would be evidenced by abundance of specimens. It seems to 

 me, however, that, on the other hand, if the numbers of the 

 butterfly and moth were equal, many of the former would fall 

 victims to the inexperience of young birds before the associa- 

 tion of an evil taste or smell with such a type of coloration 

 was discovered ; this would quickly reduce the number of the 

 butterfly, whilst the moth escaped. On the other hand, many 

 of the butterflies which resemble Euploeinse appear to be abun- 

 dant ; and I think we must look for the explanation of this in 

 the abundance of examples in the species of that group, 

 coupled with the abundance of species, all much alike, and 

 therefore representing an army of unpalatable individuals 



greatly exceeding in numbers the so-called u mimicking " 

 species. 



The subfamily Euploeinas is largely represented in the 

 1 Challenger ' collections, no less than seventeen species being 

 referable to this group ; of these, one of the most interesting 

 to me is a species named by Mr. Moore Tirumala angustata, 

 and which is represented by eighteen examples from Tonga- 

 tabu ; in this series there is only one variety, a melanistic 

 example differing from the typical form in the want of the 

 irregular spot towards the end of the cell of the front wings, 

 but agreeing with it in every other respect. The point of 

 interest about this species is its marvellous similarity to T. 

 hamata of Australia, and nevertheless the absolute constancy 

 of the principal character by which it can be distinguished, 

 the dark brown band on the hind wings, separating the 

 greenish-white markings of the basal area from the inner 

 series of spots upon the external area, being invariably about 

 half as wide in T. angustata as in T. hamata. I regard the 

 constancy of the slight differences in these two locally sepa- 

 rated but nearly allied forms as important evidence against 

 those who assume that all differences of pattern which do not 

 at once arrest attention are due to individual variation, and 

 who consequently must not only be themselves disabled from 

 studying the geographical distribution of species in its finer 

 details, but must deter others from learning the exact truths 

 which a study of it is designed to teach. 



Another interesting form in which the characters are also 

 constant, though equally slight, is Saphara insula, a species 

 of the same subfamily, to which I shall have to call attention 

 later in this paper. 



