THE CAUSE OF MIMETIC B,ESEMBLA.NCE. 599 



natural selection. It will be shown below tbat this prediction 

 is abundantly justified. 



Although H. "W. Bates did not know all the species we 

 recognize now, he knew of all the chief sections of this group, 

 and described them (figuring certain of the species) in his paper 

 published in the Transactions of this Society for 1862. A rept-e- 

 sentation of the group as he describes it is shown in PI. 42. 

 fig. 1 ; but he also mentioned another set of rather smaller insects 

 which is really continuous with the other group and should 

 be considered with it. This included the Ithomiine species 

 Dircenna epidero, D. dero, and D. rhoeo, and the moth Hyelosia 

 tiresia. 



Owing to the kindness of Mr. Godman, Mr. Salvin, and 

 Mr. Herbert Druce, I am able to figure the group as it is now 

 known with over thirty species (PL 42. fig. 2). The simple 

 combination of transparency and black which makes up the 

 whole of the appearance of these species renders an uncoloured 

 representation entirely suflBcient. The strong general resem- 

 blance which runs through this large group is well seen in fig. 2, 

 which is taken from a beautiful photograph of the actual insects 

 made by Mr. Alfred Eobinson, of the Oxford University Museum. 

 The arrangement of the insects for the camera upon small 

 pieces of cork glued on to a sheet of glass, was a work involving 

 great care and skill, and in this, as in so many other sides of my 

 work, I wish warmly to thank Mr. W. Holland, of the Hope 

 Department, for his most efiicient help and ready kindness. 



Microscopic examination of the Itliomiince (Neotropince) 

 showed that the scales on the wings are of two kinds, broad 

 and narrow, which alternate more or less regularly. In the 

 transparent parts both kinds of scale can still be detected, the 

 narrow being frequently reduced to fine simple hairs, and, in 

 the most extreme cases, the broad scales being reduced to 

 Y-shaped hairs. The two commonest species of the whole 

 group, probably forming the centre towards which the others 

 have converged, are shown in PI. 43. fig. 1. These species, 

 belonging to very different genera, afford a good example of the 

 closeness of resemblance which may be attained by common 

 warning colours. The two forms are constantly found intermixed 

 in collections, and superficially are almost exactly alike. Never- 

 theless, a study of the transparent part of the wing under the 

 microscope reveals the fact that the degeneration of the scales 



