ANCESTR Y AND CLA SSIFICA TION. 245 



which are widely separated at base, great length 

 of tongue, and the presence of a middle pair of 

 spurs on the front and usually on the hind legs, 

 in the former developed as a curious foliate mem- 

 brane ; their eggs are broadly truncate spheres, 

 sometimes ribbed ; their caterpillars have a large 

 head with a very thick skull and a very con- 

 tracted neck, formed of the first body segment, 

 and bearing a corneous shield above ; their chrys- 

 alids are smooth and uniform, like the pupae of 

 most moths, but in rare instances are pointed in 

 front. In nearly all these features they resemble 

 the picture we drew of the primeval type ; but 

 in the hooked antennae and foliate appendage of 

 the fore tibiae of the butterfly, the frequently 

 ribbed eggs, the constriction of the neck of the 

 caterpillar, and in the occasional projection of the 

 head of the chrysalis, they have departed from 

 that type, and most of these peculiarities they 

 share with no other butterflies. The other fam- 

 ilies appear to have diverged simultaneously from 

 each other shortly after their common separation 

 from the skippers ; and by means of the accom- 

 panying rough diagram [Fig. 171], which, like 

 most attempts at genealogical trees, possesses 

 anything but artistic merit, I have attempted to 

 exhibit the apparent relation of the different 

 groups to each other ; the position of the main 

 branches and their divisions is supposed to indi- 



