292 SMITHSONIAN MISCELLANEOUS COLLECTIONS [voL. 48 
Having seen the prevalence and primitive nature of this char- 
acter in the Rhopalocera, the question arises as to what extent 
radius is thus modified in the Heterocera. ‘To answer this I removed, 
photographed, and studied the pupal wings of the following moths, 
without, however, finding any trace of a split-back condition of R,,,: 
Carpocapsa pomonella of the Tortricina, Datana sp. of the Noto- 
dontidz, an undetermined species of the Geometrina, Plusia brassice 
of the Noctuide, Alypia octomaculata of the Agaristide, Pyrrharctia 
isabella of the Arctiide, Phlegethontius celeus and Ampelophaga 
myron of the Sphingide, Samia cecropia, Telea polyphemus, and Cal- 
losamia promethea, of the Saturniina, Clisiocampa americana of the 
Lasiocampide. Spuler has figured the pupal wings of Mamestra 
brassice, Harpya vinula, and Smerinthus ocellata, and none show 
any traces of this modification. Finding that tracheation showed no 
evidence of a split-back condition of R,,;, I turned to the adult 
wings and carefully examined for such traces as were so abundant 
among the butterflies, 287 cleared and mounted wings, representing 
216 genera and 35 families drawn principally from North America, 
but including such generalized forms as Sthenopis and Hepialus, and 
found no trustworthy evidence of any such modification. It is dif- 
ficult to believe that were such vestiges at all prevalent, they would 
have been overlooked. 
The great prevalence in Rhopalocera and the total absence in 
Heterocera of this split-back condition of R,,, affords a striking and 
genetic difference between butterflies and moths. It adds a new 
proof to that already existing for the accepted belief that the butter- 
fly groups are more closely related to one another than to any other 
living Lepidoptera. 
While butterfly wings conclusively show that a split-back condi- 
tion of R,,, was certainly characteristic of the rhopalocerous pro- 
genitors, moth wings hint at no such condition of radius in the 
heterocerous progenitors. In fact, everything shows conclusively 
that radius was of the type represented in Spuler’s lepidopterous 
hypothetical type, and Comstock and Needham’s hypothetical type 
for insects. Here, then, is the point of divergence between the 
heterocerous and rhopalocerous stems. This separation clearly did 
not occur until after the Lepidoptera had divided into Jugate and 
Frenate, for the reduction of radius of the hind wing to a two- 
branched condition, and the possession of a frenulum by the male 
of Euschemon rafilesie, shows that the butterflies clearly belong to the 
Frenate. 
