126 THE entomologist's record. 



definite indication of their position in relation to Papilionidac and 

 Enjcinidac. 



It is on tlie distribution of hairs and on the grooving which is 

 more or less related to it, that the main portion of his argument rests. 

 " Fine sense hairs specialised as in Krucinidac, but more highly so." 

 " Configuration of ventral surface specialised, of the Erycinid type " 

 (p. 413). 



Now the plain meaning of these summaries, taken with the fuller 

 argument, is that after the En/cinidae had existed for some time with 

 the simpler antennae of Lycaenid type, which many of them still have 

 [('.//., " there is gradual transition from the not-grooved, to the grooved 

 antenna " (p. 380)], the grooving then began, and somewhere here 

 Plcridae branched oft'. Now there can be little doubt that if this were 

 so, there would be some, probably many, forms in existence shoAving 

 the transition. The double claw of Picridar probably resulted from 

 the impetus to tarsal development, resulting from the reacquirement 

 of anterior tarsi in the males. I do not know that this suggestion is 

 more improbable than the absence of the intermediate forms would be, 

 were the derivation as postulated. If the Pinidac are related to the 

 Piryrinidac, then they originated together with them before Pri/cinidae 

 had definitely grooved antennae, and the ancestral Erycino-Pierid 

 had smooth antennae, was an Erycinid, in fact, with very nearly 

 Lycaenid antennae. This is quite possible so far as antennae go, but 

 not, from an antennal standpoint, more probable than that P'ictidac 

 originated from the early Papilio along with XyuipliaUdae. 



If grooves originated in the Nymphalid stem, and also in the 

 Erycinid, why should they not also arise in the Pierid ? Three 

 independent origins of grooving are not more unlikely than two. The 

 Parnassiine groove must probably be added, making a fourth. 



There was clearly a strong tendency in butterfly evolution, for the 

 sense hairs to become restricted on the antennal surface, in favour of 

 pits, and for hollows to be developed in which they lay. Dr. Jordan 

 postulates that this tendency should take effect tAvice, in Erycinidae 

 and Painlhmidae, I ask the same, but I suggest that it took two or 

 three different directions in the Papilionid stirps. 



It seems to be possible to discriminate between the action of this 

 tendency in the two branches. In Erycinid ac the segregation of the 

 hairs from the punctures is vague and imperfect, and has only a 

 general relation to the grooving of the antenna. The grooving has 

 attained a very definite development, whilst the hairs and punctures 

 arc still mixed together, iDoth within and without the groove, though 

 the hairs have begun in some degree to mass themselves in the groove. 

 In the Papilionid stem there is no sign of a groove till the hairs have 

 massed themselves, and after the grooves have formed, no hairs are to be 

 found outside the grooves or amongst the punctures, even when these 

 have invaded the groove. 



When we come to the character of the groove, there is a great 

 dift'erencc between Pierids and Erycinids, proving that, whether the 

 families be or be not related, their antennal groovings had separate 

 origins. The Erycinid groove begins as a flattening on the antennal 

 clul), which becomes deeper like a channel, running along the whole 

 length of the club (except the tip) and, finally, so deep as to have 

 definite, almost carinate margins. It is the same width at the margins 



