Xll 1NTEODUCTION. 



b' . Front pair of legs perfect in "both sexes. 

 a 2 . Vein 1 a in hind wing wanting ; claws 



simple Papilionidae. 



b 2 . Vein 1 a in hind wing present ; claws 



bifid Pieridae. 



b. Precostal nervure in hind wing absent .... Lycaenidae. 

 B. Antennae wide apart at base ; hind tibiae 

 generally with a medial as well as a 

 terminal pair of spurs ; all the veins in the 

 fore wing from base or from cell, none 

 forked or coincident beyond Hesperiidae. 



Opinions vary as to the probable line of descent of the butter- 

 flies. Packard considers that the moths of the family Castniidce 

 are their predecessors ; Meyrick traces their descent from the 

 Thy rididce group Pyralidina of the Frenatse, Hampson from the 

 Zyr/cenidce. 



However this may be, a provisional genetic tree for the 

 butterflies can be constructed as below. 



Nymphalidce. Nemeobidce. 



Pierida>. 



Lyccenidcp. 



Hespeviidce. 



Hypothetical Moth Ancestor. 



The evidence of one family group of the butterflies being 

 derived from another is entirely inconclusive. Specialization of 

 forms has followed often parallel lines in all the families, but 

 similar specialization, when it occurs in different groups, is no 

 proof of genetic descent, but of independent development along 

 similar lines. 



The tendency in modern Systematic Natural History is to' 



