140 M. I. NEWBIGIN [February 



and colour phenomena in general ; but the paper differs from the 

 preceding in that it is concerned with the broad interpretation 

 of facts observed in the field, rather than with the setting forth 

 of the results of special research. As readers of his previous 

 papers are aware, the author speaks with the authority of one 

 having large experience of butterfly -collecting, especially in the 

 Far East. Like many others, he has as a collector been greatly 

 impressed by the phenomena of sexual dimorphism and polymorphism, 

 and colour variation in general ; but, unlike many field naturalists, 

 he is unable to accept the suggestion that these are clue to secondary 

 causes. On the contrary, he believes that they prove that the 

 colours of butterflies tend to vary in a definite direction ; that 

 there is an evolution of colour. In other words, he holds that 

 if certain genera in a given family include species which exhibit 

 colour varieties, then, in the different genera, the colour varieties 

 will be of similar nature, certain colours being in each case 

 characteristic of the less specialised species or varieties, and other 

 colours characteristic of the more specialised forms. 



Beginning with this hypothesis, the author seeks, by a series 

 of examples, to determine the course of colour evolution in the 

 Pieridae, and to consider the bearing of his results upon various 

 problems. 



By a detailed study of various genera, notably Tachyris and 

 Hehomoia, he shows (1) that the amount of red in the wings varies 

 greatly ; (2) that red and black vary inversely in amount ; (3) 

 that in dimorphic species the male has usually more red and 

 less black than the female; (4) that where the fore wings contain 

 much red, yellow occurs on the hind wings, but where red is present 

 in small amount, yellow is replaced by white. To put this briefly, 

 red and yellow on the one hand stand in antithesis to black and 

 white on the other, and red tends to predominate in the males. 

 Further, the conditions under which red occurs, sometimes as a 

 ground colour, sometimes as spots of varying size, together with 

 its late appearance in ontogeny relatively to black, convince the 

 author that red in itself, or in the form of orange or yellow, is 

 the most primitive colour in the Pieridae ; that black and the 

 dark tints are more specialised, and that as they develop the 

 original red diminishes in amount and fades to orange, yellow, 

 and ultimately white. The process which results in this gradual 

 decoloration of the primitive pigment does not, however, end at 

 white, but ultimately results in the loss of the scales themselves ; 

 forms like Macroglossa fuciformis L., and M. bombijliformis Ochs., 

 with their glassy wings, being regarded by the author as terminating 

 the course of evolution along this line. Simultaneously with the 

 decoloration of the primitive red pigment, there may be a development 

 of structural colour, but the relation of this to the dark pigments 



