DIPHYLBTISM IN THE LEPIDOPTERA. 121 



Sphingides, Saturniades, Bombycides (Agrotides) are possibly 

 converging groups. Now the hypothetical ancestor of the Papi- 

 lionides must have already differentiated in the direction of 

 retaining "Grote's vein " in the imago. But no other offspring 

 appears, except what seems the probable culmination of the line 

 in the existing swallow-tail group. If the fourth anal vein in the 

 pupa of Pieris or Vanessa were really related to " Grote's vein," 

 it would be succeeded in the imago by an arrangement of the 

 inner veins, recalling that of Papilio. But it is not ; and the 

 resultant imaginal wing agrees in this respect with that of 

 Hesperia. Perhaps survivals of the ancestry of Papilio may 

 turn up among the lower moths of the tropics in the Eastern 

 Hemisphere, where I look for the arising of the Papilionides to 

 have taken place. 



The neuration of the swallow-tails, especially that of the more 

 generalized Ornithoptera, has a peculiar impress when compared 

 with that of other butterflies. There is not the same tendency 

 to approach the Hesperid type with divided veins, which we find, 

 for instance, in Charaxes. The shape of the discal cell and 

 arrangement of the median branches is sid generis. The object 

 of my studies on the wings of the butterflies was really, however, 

 attained with a demonstration that, in a linear arrangement, the 

 swallow-tails cannot be intercalated between the "blues" and 

 the " skippers," as proposed by Mr. Scudder and other authors. 

 Also that the blues and skippers are, at whatever distance, con- 

 nected groups. To show that the Lycsenid wing was a natural 

 development out of that of Hesperia, and that the wings of the 

 Nymphalid and Pierid were both related and might have well 

 proceeded out of the wings of ancestral forms of the Lycseni- 

 Hesperid branch — this was the task set before me. If accom- 

 plished, it followed, without saying, that the Papilionides (which 

 I also showed to be more specialized than commonly believed) 

 should take the lead in our catalogues. But the matter has now 

 gone further, and out of it has grown a question of diphyletism 

 in the Lepidoptera, hitherto unbroached. 



I do not recognize the relationship of the Papilionides to the 

 Pieri-Nymphalidse, as urged by Dr. Chapman or Mr. Quail, either 

 on account of the retention of residuary characters, i. e. the 

 cubital cross vein of primaries, or the first radial branch of 

 secondaries closing outwardly the humeral cell, since these are 

 common to the order ; or on account of the retention of fourth 

 anal of primaries in pupal wing of Pieris and Nymphalis, 

 because this is also a common generalization, and in this case 

 progresses by a disappearance in the imaginal state, not shared 

 in by the Papilionides. The Nymphalids and Papilionides do 

 not fit together ; they appear disjointed, having attained diverse 

 total stages of specialization. But all the groups of the Hespe- 

 riades seem to fit together. 



