( 414 ) 



the Jj/caenidae, En/cinidne, and Pieridae on the one haml, wliile other specialisa- 

 tions are found in Papilionidae and Nt/mphrilidae. Not a single specialisation 

 connects f,i/cne»id(ie, or F.rycinidae, or Pieridae closer with Pajnlionidne, or witli 

 Xymphalidae. The specialisation in scaling points to a connection between 

 yi/mphalidne and Papilionidae ; the fine sense-hairs speak for close relationship 

 between Li/caenidae-Erycinidae-Pieridae, and between PnpiUoi)idae-yi/mpli(di(liic : 

 the configuration of the ventral surface points very obviously in the same diri'ction: 

 and the sense-bristles speak again for the same connection. As, therefore, all the 

 characters which are not indifferent are for a phyletic connection between 

 L>jcaenidae-Erycii>idne-Pieridae on the one side, and between PapiUonidne- 

 2\>/mpkaUdae on the other, it is obvious from those statements that, leaving out 

 of consideration for the present the antenna of llesperiidae, the Bntterfly antenna 

 developed early in two directions, the development resulting on the one hand in the 

 odd-grooved Lycaeuid-Erycinid-Pierid antenna, and on the other in the even-grooved 

 Papilionid-Xymjilialid antenna.* Now, as not one of the four anteinial organs dealt 

 with is specialised in all the members of the Lycaeuid-Erycinid-Pierid branch of 

 Butterflies, every one being at least in some members of an ancestral type, it is 

 evident from the absence of a distinguishing character between the whole branch 

 and the ancestral Butterfly that this branch of Butterflies cannot have diverged from 

 the original stock before the Papilionid-Nym])halid branch had become specialised. 

 The divergent development of the two jihylums began witli a specialisation of the 

 ancestral Papilioni-Nymphalidae ; and as this specialisation must have taken place 

 in those autennal organs which show the same or similar direction of specialisation 

 in all members of the phylum, we have to conclude that tlie divergent development 

 of the early Papilioni-yi/mphitliddC began with a modification of the regular belt 

 of bristles into an irregular transverse ventral series, the aj)pearance of setiferous 

 punctures, and with a basad restriction of the sense-hairs on the distal joints. 



Now, where have the Ilesperiiilae to come in ? Did they branch off Ijefore or 

 after the divarication of the Butterflies into a Papilioni-Nym])halid and a Lycaeno- 

 Erycino-Pierid phylum took place ? If the Ilesperiidae are a development posterior 

 to the divarication of the two phylums, they must belong to the Lycaeno-Eiyciuo- 

 Pierid phylum ; for the Hesperids have not one of the specialisations by which the 

 P<ipilioni-Xymph<iUd(ie are characterised, and cannot, therefore, have depart eil with 

 them from the ancestral stock. Hence it remains only to consider from which puinl 

 of the second phylum the Hesperids shot ofl". The antennae of Hesperiidae have 

 most characters of an ancestral type, and hence resemble the antennae of Lycaenids 

 which are also ancestral ; this resemblance finds an explanation in both families 

 being derived from the common ancestor of all Butterflies. The specialisations are 

 such as occur in all or nearly all other families, or arc peculiar to the skipjiers (for 

 instance, the possession of three bristles in certain Hesperids, f. 4), and hence are 

 indiiferent in respect to the jjrescnt ((uestion: only the apical position of the ventral 

 belt of bristles points to a connection with Lymenidae. bnt not condnsively, as we 

 have seen. The large extent of ventral and dorsal scaling, and the specialisations in 

 the bristles, are equally well explained, if wo assume the Ilesperiidae to be derived 

 from the Lycaeno-Erycino-Pierid phylum before the odd-grooved Erycino-Pieridae 



* As this result is in the main point entirely at variance with the eurrcnt views of entomologists as to 

 the classification o£ the Butterflies, 1 expect to be criticised, and give expression to the hope that the 

 criticisms will be intrinsic. But I must st.ate here, that I shall not answer arguments from other organs 

 than antennae, a.s the other organs will be treated upon in subsequent instalments of these contributions. 



