1898] CLASSIFICATION OF THE DAY BUTTERFLIES 99 



the case of the Attacinae (I.e. 30, 31), and that we are not warranted 

 in using this character alone for associative purposes. 



In offering, in this paper, a reading of the facts presented by the 

 study of the structure of butterfly wings, I have tried to give a fresh 

 turn to the subject and to supply a firmer basis for classification. 

 And while the results I have reached can find in this place only a 

 partial application, and, indeed, while the neuration itself may not 

 ultimately in every case give clearly the clue to phylogeny, yet the 

 facts brought forward here cannot be disregarded in any system 

 purporting to arrange the butterflies according to a probable blood 

 relationship. 



A. Radcliffe Geote. 



RoEMEK Museum, Hildesheim. 



EXPLANATION OF PLATE IL 



The figures are obtained by photographic process. 



III = Radial veins ; IV=Median veins ; V = Cubital veins. 



Fig. 9. Pararge acgcria. Type of sub-family and genus. Size 1|/1. Attention is 

 called to position of cross-vein on secondaries, where vein IVs has effected 

 junction with cubital system as in the Nymphalidae, also relatively great 

 absorption of II and III at base. 



Fig. 10. Ercbia ligca. Type of genus of Agapetinae. Size 1^/1. Compare condition 

 of cross-vein of secondaries, typical of sub- family, the same as in Pieridae, 

 Heliconidae, Limnadidae, where vein IVs still leaves cross-vein. 



Fig. 11. Ocncis aello. Size 1|/1. Introduced to show its generalised character ; cells 

 strongly closed, absorption of II by III at base only partial as in Erebia, 

 which it most resembles. Compare text. 



Fig. 12. Lihythea celtis. Type of genus and family. Size 2/1. Compare text. 



Fig. 13. Eiodina lysippus. Type of genus and family. Size 2/1. Compare this with 

 figures of Lycaeni-Hesperiadae and especially with Thecla given here. Note 

 that pattern of wing is same and the main diff"erence is displayed by the 

 reduction of radial branches in which the Riodinidae are equalled by the 

 Zephyrini and beaten by the Theclini. Note retention of vein I on hind- 

 wings, which has become absorbed in the Lycaenidae and Hesperiadae. 



Fig. 14. Thcda spini. Type of genus and sub-family. Size 2/1. Compare above. 



Fig. 15. Lycaena endymion. Considered by Scudder type of genus and family. Attend 

 to four-branched radius, IVi not joined to radius, its base between junction 

 of punctated degenerate cross-vein and radius being morphologically part of 

 cross- vein. 



Fig. 16. Pusticus argus. Type of genus. Size 2/1. This common ' Blue ' differs a 

 little from Lycaena in the position and condition of vein IVi. 



Fig. 17. Pamphila palacmon. Type of genus and sub-family. Size 2/1. Note absence 

 of costal vein on forewings and presence of VIII. 



Fig. 18. Plathcsperia (olim Eantis^vel Achylodcs) husiris, Cram. sp. Type of genus. 

 Primary wing. Size 1/1. Introduced here to show more clearly the 

 typical hesperid pattern. Compare with Hesperia malvae. 



