MELANISM AND MELANOCHROrsM. 319 



all stages at a temperature of about 8o^ I brought out four 

 successive broods of illnnaria in the ten months, all of the 

 summer type.^ (6). After larval growth is completed, no 

 complete conversion of the one type into the other can be 

 effected ; it seems clear that such a conversion cannot be 

 made as regards size,- and but slightly, if at all, as regards 

 shape ; - and it seems that it cannot be completely made 

 as regards colour'- or markings.- This incapability as to colour 

 and markings certainly exists as respects illiistraria, also as 

 regards A. levana and P. ajax (Professor Weismann's experi- 

 ments) ; and I gather that, in the cases published as to 

 P. rapes, P. iiapi, P. pharos and P. inter rogationis, the butterflies 

 from the iced summer pupae presented some differences from 

 the normal form proceeding from the winter pupae " {Trans. 

 Ent. Soc. Lond., pp. 146, 147). Nothing could show better 

 than these conclusions, how utterly impossible it is to leave 

 heredity out of account when experimenting. 



The influence of heredity in causing variation is again well 

 recognised by Mr. Merrifield in the following : — " It will be 

 seen that there is some individual variation more partiailai'ly 

 in the pupae exposed to the ordinary temperature, and therefore 

 some of the colouring must be attributable to individual and 

 presumably hereditary qualities " (pp. 137, 138) ; and the 

 still further certainty of outside influences is shown in the 

 failure to produce similar results from tiliaria (presumably 

 not domesticated like the other species), " the results on the 

 colouring, though tending in the same direction, were by no 

 means so regular or so striking " (p. 138). "There is no doubt 

 a strong pre-disposition, in an individual belonging to a double- 

 brooded species, at some period of its development, towards 

 one of the two different destinations, i.e. the emerging in the 

 summer and with the summer colouring, or the lying over 

 until the spring and then emerging in the spring colouring. 

 The experiments lead me to think that in the species operated 

 on by me the predisposition has become so decided in the 

 larval stage ^ that no treatment of the pupa can afterwards 

 entirely alter it, but, that in the early larval stage, treatment 

 can — I do not say in all cases — either give the required pre- 

 disposition, or, where it exists naturally, can completely 

 reverse it " (pp. 14.2, 143). Here my experience with larvae 



1 By bringing out four broods in one summer at a high temperature, tliere was 

 delay in larval stage. 



* I go further, and suggest not at all in either direction as most probable. 



^ 1 feel satisfied that this is the stage which is affected by external circumstances 



