SEASONAL DIMORPHISM OP LEPIDOPTERA. 207 



very instructive to see that this does not in any way follow fixed 

 principles, but is in a certain sense without rule. It cannot be 

 said : the black increases, the yellow changes into white, but 

 " in this place the black expands, there it is changed into white," 

 as was shown above ; the white band of the hind wings origi- 

 nates from black in its lower part, from brownish yellow in its 

 upper part ; the interrupted band of white spots originates, on 

 the other hand, only out of the brownish yellow ground colour. 

 Many of these changes, therefore, cannot possibly be simply 

 chemical processes, induced by the action of a higher tempera- 

 ture on the "pigment-formers" (Pigmentbildner) of the wings 

 of the pupa, and comparable to the reddening of blue litmus- 

 paper in acids. All that I wrote on the subject twenty years ago 

 I still consider fully justified : *'A new marking has developed, 

 emanating from the existing one." But while I then thought 

 that this new formation must be always regarded as a reaction 

 of the specific levaiia-ovga^nism. to a higher temperature, I now 

 recognise that temperature does not in general take part here as 

 the actual cause, but that it is a question of a process of selection, 

 which goes on independently of the temperature, and which 

 gradually alters (umstempelte) some of the ids to the prorsa-ids. 

 But these prorsa-ids were at the same time so arranged that they 

 become active under the action of a higher temperature, if this 

 is acting at the beginning of the pupal period, while the levana- 

 ids become active at a lower temperature. Heat, therefore, is 

 only the excitant which sets free the _pro7^sa-determinants, while 

 cold sets free the levana-deievmrn^inis. 



But this does not exhaust the matter. As already explained 

 above, I formerly thought that the offspring of the prorsa brood 

 always assumed the levana-iovm, even if their pupae were sub- 

 jected to a high temperature : this was not, indeed, entirely 

 correct, but still it contained a germ of truth, so far as this 

 second brood has a stronger tendency to the levana- than to 

 the _p?'ors <2-form. This is convincingly proved by all the experi- 

 ments. They can be changed into prorsa, as one of my old 

 experiments of 1869 proves, in which the whole brood of a female 

 of prorsa again assumed the prorsa-ioxm. under the influence of 

 the unusually hot July sun. But in this case also the tendency 

 of this second brood to the activity of the levana-\d^ can be 

 •recognised from the fact that many butterflies exhibited a striking 

 amount of yellow, and were, indeed, almost porima. On the 

 other hand, the first brood of the year has just as pronounced a 

 tendency to the activity of the j9?^o?'sa-determinants, otherwise it 

 must be possible to always make all individuals of a brood into 

 levana by cold, which I, at least, have not succeeded in doing ; 

 but generally a not quite pure form of levana originates in this 

 way, often only approaches to it, viz. 'porima. If this were not 

 so, a number of levana butterflies would fly in July every cold 



