124 Dr. F. A. Dixey on the phylogenetic 



the common ancestors of the group, i. e., the blue-centred 

 series IV. and the white spots of series D. In this last 

 particular the form^. ^rorsa comes nearer the ancestral 

 type than A. lev ana ; in all other respects, however, it 

 bears marks of higher specialisation, and it is worth 

 notice, as a confirmation, that Weismaun* concludes on 

 quite other grounds that A. lerana is the older form. 



The last line leads off through the genus Grapta to 

 Vanessa polijcldoros, urticce, ichnusa, and milberti, no 

 doubt in the order given by Prof. Meldola.t The 

 nearest approach to the " Protovanessa " among the 

 Graptas is furnished, I believe, by G. c-aureum. This 

 interesting form by its aspect at once recalls the 

 Argynnids, both in colour and markings. As we have 

 seen, it is characterised by the possession of several 

 ancestral features, for example, the co-existence of 

 series II. and III. in the hinder part of the fore wing, 

 and the presence of blue centres in the latter through- 

 out its extent. The affinity of Grapta to Araschnia is 

 suggested by the markings (especially the occurrence in 

 both of the series II.'), and is supported by the fact that 

 the larvffi of both genera are distinguished among their 

 allies by the occurrence of two spines on the head. The 

 blue centres of series IV. in V. polychloros, improved 

 upon in V. urticce, are an ancestral feature that has 

 mostly disappeared from the upper side of the Graptas, 

 though still, as a rule, recognisable beneath. A notice- 

 able characteristic of the present group is the gradual 

 disappearance of series III. Well represented in G. 

 c-aureum, where it preserves the blue centres of the 

 " Protovanessa," it loses in the rest of the genus its 

 distinct character ; in V. polychloros it cannot be traced 

 in the hind wing or in the middle of the fore wing ; 

 while in V. urticce the outermost costal patch is its only 

 remaining vestige. 



Somewhere probably near the beginning of this third 

 stem came off a branch, which ends in the two species 

 r. antiopa and V. io. Both of these forms preserve the 

 ancestral blue-centred black border ; V. antiopa in com- 

 pleteness, V. io partly as a constituent of the ocellus of 



* ' Studies in the Theory of Descent,' vol. i., p. 19. 



f " C-album, i-album, v-alhum, fautius, comma, dryas, poly- 

 chloros, xanthomelas, cashmirensit, urticce, milhertiJ" Weismann, 

 ' Studies, &c.,' vol. ii., p. 447, note. 



