Phylogeny of the Pierine. 263 
it is large and crescent-shaped, extending for some 
distance along both costa and outer margin of the wing. 
In G. rape ? it is not as a rule prolonged very far on 
the outer margin, but it still reaches some way along the 
costa. In the male of the same insect it is generally 
small, and confined to the actual region of the tip. In 
neither of these species is there much indication of 
resolution, except that in most specimens of G. brassicex, 
especially in the females, the hinder arm of the dark 
crescent shows an irregularity due to the massing of 
black scales about the marginal terminations of the 
nervules. But in G.napi (Fig. 16) resolution is generally 
well marked; and it becomes evident that the dark 
crescentic or triangular area of the tip is really formed 
by the fusion of the anterior members of the series 8, 
which we have just been considering, with another 
series, which may be called M, occupying the actual 
margin of the wing, and consisting, in its fully resolved 
condition, of a row of dark spots, each of which is 
traversed by the peripheral portion of one of the ner- 
vules. It is noticeable that in this species the black 
marginal spots are often found extended, especially in 
the female, in the form of a powdering of dark scales 
along the course of the nervules towards the base of 
the wing. 
Turning to S. daplidice (Fig. 15), we find the consti- 
tuent elements of the apical patch still better shown 
than in G. napi. The marginal and submarginal series 
are always distinct, being separated from one another by 
portions of the general white colour of the wing, which 
take the form of a row of white spots on the dark apex, 
usually four in number. Either of these two species 
serves well to illustrate an important difference between 
the spots of the two series, S and M. The spots of the 
former series occupy, as we have seen, the interspaces 
between nervules; and when fusion between adjacent 
members of the series takes place, it does so by an 
extension of dark scales across a nervule. On the other 
hand, each of the spots of the latter series is centred, 
not in an interspace, but around the peripheral portion 
of a nervule or nervure. Thus the spots of the two 
series tend to alternate with one another, though this 
effect is to some extent interfered with by the frequent 
fusion of some of the spots with others of the same 
