( xvi ) 



of the Plebeiid stock that have appi^oximated as a matter of 

 mimicry, or since both are common, but rai-ely, if ever, common 

 together, that they have approximated as occurring under 

 vei-y similar conditions, and in a slight variation of those con- 

 ditions each capable of taking the place of the other. We 

 find that the dorsal armature differs very much, and in these 

 Lycaenids the dorsal armature usually has generic rather than 

 specific value. Taking icarus or hellargus as an average 

 Plebeiid, we find in argus each lateral member of the dorsal 

 armature is long, slender, curved and slightly hooked at the 

 tip, and the accessory hook is sharply bent with the terminal 

 portion long, straight and slender, except that it also is 

 slightly curved and hooked at the tip. Argyrognomon on the 

 other hand differs from icarus in the opposite direction ; it 

 has the lateral process, short, broad and blunt, and the acces- 

 sory bends in a prolonged curve, so as to give the impression 

 that it is curved throughout, and though it has a slight 

 terminal bend or hook, like argus, it differs in being broader 

 and more robust. 



" When we compare the ends of the clasps, the difference 

 is immense, argus with a few long teeth rather on the side of 

 the hard process, argyrogiiomon has a number of very fine 

 teeth on the rather broad end of the process. These, therefore, 

 seem to be almost unrelated (within the Plebeiid group) forms 

 of appendages. 



" To accept them as closely related, I have to give a some- 

 what exaggerated value to a consideration that is of much 

 importance in the study of these genitalia. 



"The genitalia are under entirely different conditions as 

 regards selection, to the external characters we more usually 

 examine. These external characters are subject to a constant 

 process of selection in relation to such items of the environ- 

 ment as enemies, climate, food-plant, etc. The genitalia, 

 though equally inclined to vary, and probably in response to 

 similar disturbing elements of environment, are under no 

 selective force exercised by the environment. Within the 

 limits of a species, variation from the mean value must not go 

 so far as to prevent the individual from finding a mate, but it 

 is quite open to the mean value gi'adually to change in any 



