THE CAUSE OF MIMETIC KESEMBLANQE. 609 



while the remaining species (figs. 4, 6, & 7) are mimetic of slender Aculeata 

 with opaque, iridescent, or even particoloured wings. 



Hence in this group of allied species of beetles there is great diversity in 

 the methods by which the mimetic resemblance is attained and in the form of 

 the model which is imitated. One thing is common to the whole, viz. the 

 production of a strong superficial likeness to a stinging insect. Such com- 

 munity of end through diversity of means strongly supports an interpretation 

 based on the theory of natural selection. 



Plate 42. 

 The resemblances between the members of a large and characteristic group 

 of transparent and black-barred South-American Lepidoptera. 



Fig. 1. About ^ natural size. The group as described by Bates in 18fi2 (Trans. 

 Linn. Sec. xxiii. pp. 495-566). The names of the species are engraved 

 on the Plate. The species belong to very different divisions, families, 

 and sub-families of Lepidoptera. Thus the IthomiincB are represented 

 by two species in difi'erent genera, the Danaince by two species, the 

 FierincB by one, while the moths belong to the very difi'erent families 

 of the CastniidcB and Pericopidce, the latter including Anthomyza. 

 Bates also described a group of smaller Ithomiine species, Bircenna 

 epidero, D. dero, and B. rhoeo, as well as another smaller moth 

 (Hyelosia tiresia) belonging to the Pericopidcs. Of these he made a 

 second group, but they are in reality continuous with that shown in 

 fig. 1 , having a very similar arrangement of black bars and border on 

 their transparent wings. The group has in all probability converged 

 around the two common and widespread Ithomiine species of the 

 genera Methona and Thyridia. All members of the group, including 

 these latter, are to be looked upon as primitively opaque-winged. If 

 selection has been at work in the alteration of the scales so as to 

 render them transparent, we should expect that this common end 

 would have been reached by very different methods in the case of 

 species belonging to such widely separated divisions. 



Fig. 2. About ^ natural size. By the kindness of Messrs. Godman and Salvin I 

 have been enabled to figure, from their great collection, all the species 

 of butterflies which are now known to fall into the two groups described 

 by Bates 37 years ago. The BanaincB and PierincB remain unchanged, 

 but the lihomiincB have been immensely enlarged, containing 20 species 

 in seven genera. Many of these are extremely rare, and so complete 

 a set could not have been got together from any other collection. 

 I also owe to their kindness the loan of butterflies shown in fig. 1 and 

 of some in Plates 43 and 44. 



Many species of Castniidee and of Pericopidce have now been added 

 to the group. Those which are here shown, as well as in fig. 1, and 

 those in Plate 44 were in part obtained from the Hope Collection and 

 in part from the very fine collection of Mr. Herbert Druce, who kindly 

 lent thsm for the purpose of this memoir. The set is probably 

 complete, or nearly so, except in the genus Hyelosia. 



