49 
latter remaining constant and apparently forming the final or main 
extension of the radius, and corresponding to III of the hind wings. 
This process of absorption seems to be carried furthest in the 
common Mancpium brassice, for which we have to use the three- 
branched formula: III 1, III 2, III 3+ 4+ 5. Again, sometimes 
III x and III 2 are represented by a single vein, III 1 + 2, as in 
Farnassius and certain emperor moths. On the lower side of the 
radius the frst direction comes into play. The first medial branch, 
IV 1, which, in Vymphais and the four-footed butterflies generally, 
is still attached to the cross-vein, now leaves this position, and travels 
along the lower side of the radius, asin /veris and Vemeobius, until 
the point of its emergence therefrom approaches the extremity of the 
nervure. In this movement of IV 1 we have also a gauge of the 
amount of specialisation, and /vev7s and Memeobius are undoubtedly 
in this respect highly specialised. We are here not concerned 
with the physiological process underlying all these changes. We 
are endeavouring to determine the phylogeny by means of ascer- 
taining the amount of progress in the specialisation of the neuration. 
Here, as elsewhere, the direction must be first established, then the 
order in which the changes in the position of the veins follow one 
another, and lastly the mode and method of evolution. 
Leaving these two prime directions, the fzs¢ of which is still in 
action on both wings, whereas the second would seem to have 
culminated for the hind-wings by the attainment of a single branched 
radius in the bulk of the Lepidoptera, we may briefly consider other 
changes by absorption in the network of the wings. On the hind- 
wings the veins II and III are more or less fused at base. Evidently 
the forms in which this fusion is slight or even wanting (Left:dia) 
are less specialised than those in whic. it is carried up to the point 
of emergence of the short spur representing vein I. On this excess 
of specialisation I have founded the sub-family Nymphaline, and I 
see by the figure of Basz/archia astyanax in Prof. Comstock’s beauti- 
ful ‘‘ Handbook,” that this character obtains for the North American 
genera also. It has, therefore, a meaning, and this meaning must 
be expressed by a classificatory term. 
On the hind-wings, also, the point of juncture of the cross-vein 
with the cubital system varies. In the more specialised four-footed 
butterflies, the Nymphalide and the first sub-family I have established 
in the Meadow Browns (Agapetidze = Satyridee of authors, the latter 
name being preoccupied according to Scudder), the Pararginee, this 
juncture is effected at or excessively near the furcation of the first 
cubital branch, V 1. In the more generalised Agapetidz and in 
the Limnadidz the juncture is effected with the lowest branch of 
the medial system itself, vein IV 3, and this method is characteristic 
also for the Pieridz ; so that here again we have a character of 
specialisation by which to guide our phylogenetic researches. Other 
points are offered by the suppression of the internal veins, such as 
we find in the absence of vein VIII of the hind wings in the 
4 
