( 407 ) 



a specialisation as the two grooves in ail the species of Xpn/j/nil/t/iit' that the 

 development of this character mnst have been at least incipient in the ancestor of 

 the family, the point of divarication of the Papilionidae and Nymphalitlw wns 

 probably an antenna with the extended area of sense-hairs partly divided in tlie 

 mesial line. The occurrence of that typically Nymphalid specialisation, the carinas, 

 fonnd nowhere else among Lepidoptera, in all the species — the few apparent 

 exceptions being explained by higher specialisation that has led to simplification — 

 makes it further highly probable, that this entirely new specialisation furnished 

 that character by which the- early Nymphalidae were distinguished from the 

 otherwise closely allied ancestors of Papilionidae, i.e. that the division of the area 

 of fine sense-hairs into two patches in the Nymphalid liranch of the Papilioni- 

 Nymphalid phylum was accompanied by the development of grooves separated from 

 each other by a mesial carina, and each bordered laterally by a ventro-Iateral carina. 

 As we learn from the most highly specialised Nymphalid antennae, for instance 

 from the series of Satt/riimc (figs. 57 — 59) Ipthima asterojic, Si/ncliula, and Triphijm, 

 that the carinae obliterate before the grooves have disappeared, we must conclude 

 that the carinae have always been in advance of the grooves, i.e. that the carinae 

 appeared in the ancestors of the N)/mphalidae before the grooves were developed. 



It is of interest to note that iu the simplified Satyrine antennae, alluded to 

 before, the not-scaled area is very much restricted, iu Triphi/sa to four joints. A 

 comparison of Ipthima asterope (f. 58) with the other species of this genus, which 

 have more generalised Nymphalid antennae with large grooves mostly extending 

 from the base to the apex of the joints and provided with strong carinae, most 

 obviously shows, that we have here to do with a highly specialised antenna, 

 specialised in the development of the grooves, carinae, and scaling. 



If we thus account for the similarity in the structure of the two-grooved 

 antennae of the yympkalidae and certain 'Papilionidae by deriving both families 

 from a common ancestral stock the antennae of which had a tendency to bilateral 

 separation of the area of fine sense-hairs and corresponding development of grooves, 

 it is clear that the presence of one series of grooves among Pariiassiinae would 

 either mean that these Butterflies do not belong to the Papilionidae, or that the 

 tendency of restricting the fine sense hairs laterad, not mesiad, was not common to 

 all Papilionidae, and hence could also not have been present in the ancestral 

 Papilioni-Xymphalidae ; or, in other words, that the deduced closer relationshi]) 

 between the Papilionidae and Ni/mplialidm' would not be established, if the series 

 of grooves of Parnassiinae is homologous to the mesial row of grooves of the odd- 

 grooved antennae of F.ri/cinidae and I'ieridac. Let us then consider tlie Paniassiine 

 antennae more closely. The development of the grooves can be studied at the 

 individnals of the same species, as the grooves are very variable in the individual 

 specimens. We meet sometimes on the club with grooves which are laterally more 

 extended apicad than mesially, and suggest a bilateral development, while in other 

 individuals they are indifferently shajjed. A comparison of sncli an irregular 

 groove on the club with the more regular grooves on the stalk brings a remarkable 

 fact to light: the grooves on the proximal joints of the club become more and 

 more concentrated the nearer we come to the stalk, but this concentration does 

 not take place equally from either side of the joint, as it must if the result should 

 be a mesial groove, but j)roceeds from tlie inner to the outer side, thus resulting in 

 a sublateral, not mesial, single groove. This groove is, therefore, not homologous 

 to the mesial groove of Pieridae and Erycinidae, but to the e.xterno-lateral groove 



