98 NATURAL SCIENCE [February 



apparently in Leptidia, where it is central on fore and cubital on 

 hind wings, affording a parallel with Papilio. Otherwise the neura- 

 tion of Leptidia is so alien to the rest of the ' Whites ' as even to 

 recall that of Macroglossa. It is thus a distinguishing character of 

 the Pieridae that they offer a wide range of character, making them 

 difficult to diagnose and giving the group a certain incongruity. 

 The Pierids, then, may reasonably represent the matrix from which 

 the brush-footed type has formerly emerged. The practical identity, 

 in main features, of the wings of the Pieridae, Agapetidae, Heli- 

 conidae, and Limnadidae assists this conclusion. The inequality of 

 the specialisations makes a linear sequence very difficult to defend 

 the nearer the groups or forms come together. 



Since I believe that we must regard the three-branched Pontia 

 butterflies as a more modern and probably direct development from 

 the Anthocharini, which have a five-branched radius, it may be that 

 sometimes, at least, the loss of orange or yellow tints and the substi- 

 tution of white may be looked upon as a specialisation. There 

 remain the five-branched Anthocharis species which want the 

 orange colour. But these are also more specialised than Euchloe ; 

 in these a movement of III2 has taken place, carrying it at length 

 beyond the cross-vein and cell. A tendency is thus indicated, 

 prevented from equal step by the special causes determining 

 colour and pattern. Somewhat similarly we find in the * Large 

 Cabbage White ' a three-branched radius and in Pieris rapae a 

 four-branched. Thus, in the more modern Mancipium hrassicae, 

 the specialisation of the radius would also be accompanied by a 

 greater extent of white colour on the upper surface. On 

 another separate line of immediate descent we liave the North 

 American Nathalis iole with a three-branched radius and its more 

 generalised ally, Eurcma delia with a four-branched radius. Here 

 I find the yellow colour also to fade in the line of specialisation, 

 although it does not become white ; but the rosy fringes of 

 Eurema have also gone and there is less black on the wing 

 of Nathalis. There is perhaps not sufficient evidence amassed, 

 since no dependence can be placed upon the assumed relationship of 

 forms in literature where the specialisations of the wings have not 

 been minutely studied, but what I have gathered points in the 

 direction that increasing pallidity runs parallel with specialisation, 

 and is developed independently on the generic phylogenetic lines. 

 This is clearly the case with the Parnassiidae, where I find the white 

 forms to be the most specialised. Eoughly speaking, we might say 

 that the butterflies tend to losing colour. I may add here that the 

 examples of almost certainly direct descent, cited in the foregoing 

 paragraph, bear out the secondary nature of the reduction of the 

 radial branches, already mentioned and also pointed out by me in 



