1886.] MR. H. J. ELWES ON THE GENUS PARNASSIUS. 17 



Though the genus Parnassius has hitherto been included in the 

 family of Papilionidae, yet it seems to me worthy to constitute a 

 separate family, comprising the genera Uuryciis, which is confined to 

 North Australia and New Guinea, Euryades, which is found in the 

 Argentine States, and Lulidorfia, which is an inhabitant of the 

 southern coasts of Amurland and probably North Japan. In makin" 

 this proposal, 1 do so on the ground that the extraordinary appendages 

 of the female abdomen, which are found in these four genera alone 

 among the Ehopalocera, and which, though very different in structure, 

 are apparently analogous, afford a character of at least as much, if 

 not of greater, value in classification than the characters drawn from 

 legs, venation, antennae, or larval structure. 



And though my ignorance of larval characters among Lepidoptera 

 generally, makes me unable to form an opinion as to their value for 

 purposes of classification, yet they apparently lead to the grouping 

 of very dissimilar forms. Mr. W. H. Edwards remarks on the 

 subject in ' Papilio,' vol. iii. p. 159 : — " I do not think, judging from 

 the egg and young larva as I know them, and by the mature larva 

 and pupa as figured in books, that Parnassius has any right among 

 the Papilionidee. Under a system in which the preparatory stages 

 were considered — and in the future we shall have to come to that — it 

 would stand a long way from the Papilionidse. The egg of smin- 

 theus is like Lyccena ; oihaldur like Chrysophanus ; the young larva 

 is like some Nymphalidae (and perhaps Erycinidee), the mature larva 

 more like a Heterocerous moth (in all but the tentacles), and the 

 chrysalis like a Hesperian, or also perhaps some moths." 



In 1870 the late Edward Newman published in the 'Entomo- 

 logist ' a system of classification for Butterflies, in which he places 

 Parnassius in the second division of the Ilhopalocera, which he 

 called Celantes, forming with Doritis the Group A, Bombyciformes. 

 The division is defined as follows : — " Celantes, or those in which 

 the larvae, prior to changing to pupae, envelope themselves in a 

 silken follicle or cocoon more or less compact ; the pupae are 

 generally without angles, like those of the genus Chelonia among the 

 Sessiliventres." The Bombyciformes are those in which the head 

 of the larva is smaller than the second segment, and the body is 

 altogether that of the Bomhyces. The C'apitati, which form the 

 second section of this division, are the Hesperidae, in which the head 

 of the larva is larger than the second segment. 



Now, without expressing any opinion on the propriety of such a 

 classification, it is clear that any arrangement which depends on 

 larval characters must in the case of very many genera be con- 

 jectural. Newman criticizes the classification of Kirby's catalogue 

 very unfavourably, saying that his subfamily Papilioninse is entirely 

 opposed to his own idea of natural arrangement. The genera 

 Kirby included are Mesapia, Calinaga, Hypermnestra, Ismene, 

 Doritis, Parnassius, Eurycus, Euryades, Sericinus, Thais, Teinoi>alpus, 

 Papilio, and Leptocircus. 



Staudinger includes Parnassius with Papilio, Thais, Ismene, and 

 Doritis, in his family Papilionidse, which comes at the head o[ the 



Proc. Zool. Soc— 1886, No. II. 2 



