The Bionomics of South African Insects. 431 



colours which render the phase less conspicuous in the 

 stations it frequents. 



There does not seem to be any escape however from the 

 conclusion that the consjiicuousunder-sides of the wet phases 

 are relatively recent, and if this conclusion be considered 

 in relation to the comparison between the under-sides of 

 archcsia and 'pelasgls, it leads inevitably to the conclusion 

 that the cons'ineuous appearance of the one has been 

 modified out of the older cryptic appearance of the other, 

 and not vice versci. 



On what hypothesis can we believe that such a change 

 has taken place ? In the existing state of our knowledge 

 there are only two possible interpretations: (1) that the 

 modification is mimetic of some other conspicuous dis- 

 tasteful form ; (2) that it is a warning of some special 

 protection possessed by the Precis itself. The former 

 interpretation cannot apply to the case oi pclasgis, because 

 its pattern is so unlike that of the well-known distasteful 

 Ethiopian Rhopalocera, although some advantage may be 

 gained by Millie rian association with black and white 

 aposematic genera such as Amcmris, JYejytis, Flancma, etc. 

 Furthermore, it has been shown that there are important 

 elements in the conspicuous under-sides of the wet phases 

 of scsamus and antilopc which are not synaposematic, 

 although the appearance ns a whole is probably to be 

 explained in this way. I therefore firmly believe that the 

 conspicuous appearance of pclasgis has been produced by 

 selection from the cryptic archesia as a warning character 

 indicative of some special protection, an aposeme pro- 

 claiming that it is less palatable or in some way less suit- 

 able as the food of insect-eating animals than an immense 

 number of other species which abound during the wet 

 season in the same stations. 



I proved in 1887 (Proc. Zool. Soc, p. 191) that the likes 

 and dislikes of insect-eating animals are purely relative, 

 and that a conspicuous distasteful form will be freely 

 eaten under the stress of hunger, that tlie existence of 

 these forms depended entirely npon the co-existence in 

 their neighbourhood of an abundance of palatable species, 

 that under any other circumstances the warning colours 

 if freely exposed would be a danger and would lead to the 

 extermination of the species. As soon as I had studied 

 the case of archcsia and ]jclasgis I felt convinced that the 



