58 Studies in the Theory of Descent. 



stage of phyletic development. Were this possible 

 it would directly contradict the idea of reversion, 

 according to which new characters never make their 

 appearance, but only such as have already existed. 

 If, therefore, the ancestral forms of A. Levana 

 (which we designate as Porima) present a great 

 number of transitional varieties, this leads to the 

 conclusion that the species must have gone through 

 a long series of stages of phyletic development 

 before the summer generation had completely 

 changed into Prorsa. The view of the slow cumu- 

 lative action of climatic influences already sub- 

 mitted, is thus confirmed. 



If warmth is thus without doubt the agency which 

 has gradually changed the colour and marking of 

 many of our butterflies, it sufficiently appears from 

 what has just been said concerning the nature of 

 the change that the chief part in the transmutation 

 is not to be attributed to the agency in question, 

 but to the organism which is affected by it. In- 

 duced by warmth, there begins a change in the 

 ultimate processes of the matter undergoing trans- 

 formation, which increases from generation to 

 generation, and which not only consists in the 

 appearance of the colouring matter in one place 

 instead of another, but also in the replacement of 

 yellow, in one place by white and in another by 

 black, or in the transformation of black into white 

 on some portions of the wings, whilst in others 

 black remains. When we consider with what 



