XXVI 



Horsf. the heads remain the same after this moult in T. Hecabe L. they 

 become black. Not suddenly, however, for I observed that after this moult 

 had taken place the head of this larva was still green ; at first it became 

 gradually darker and not until after about an hour did it become black which 

 thenceforth remained constant; probably, however, it is such a dark green that 

 it appears black to the eye. In the larva of Papilio Agamemnon L. also I 

 remember making a similar observation. The pupa of the Java Acraea 

 Terpsichore Cram, has the ground colour of a beautiful milky white much 

 marbled with black and ochre-yellow. A couple of days before the imago 

 emerges this milky white changes into an intense rose and after the butterfly 

 makes its appearance no trace of either milky white or rose is visible on its 

 wings or on the pupa case. How is this to be explained ? I have been unable 

 to investigate it but it appears to me that changes occur here which have no 

 connection with the phylogenetic development of colour. It leads incontestably 

 to the assumption that a certain amount of influence of air or light is required 

 for the maturing of the colour as it occurs in the perfect insect; it may, 

 therefore, be doubted if such reaction really occurs in wings exposed artificially 

 during the experiment and whether the mature colours may be traced back 

 to the earlier stages rather than that they should originate in the present stage 

 of development of the chrysalis. For this reason I consider this phenomenon 

 deserving attention. 



On the other hand there is a great deal which, in my opinion, confirms 

 the propositions of the theory in question. The most important is undoubtedly 

 the fact that it is becoming more and more manifest that the numerous colour 

 phenomena observed in Lepidoptera can only be explained by means of it. 

 This is especially striking with respect to the remarkable cases of colour dimorph- 

 ism or polymorphism which cannot be reasonably explained otherwise although 

 there is no lack of phraseology in this connection. Where such cases occur 

 regularly side by side as in Pap. Memnon L. they follow the rules of colour 

 evolution ; likewise where they make their appearance contemporaneously but 

 in different districts, as in Pap. Dardanus Brown ; and again where they are 

 restricted to specific periods by climatic influences, showing clearly which is 

 the more and which the less developed form, the same occurs. Finally where 

 by the action of retarding influences, of whatever nature, only a lower, and 

 therefore older, stage of development than the present normal one is reached, 

 this older stage agrees also with the doctrine of this theory. This alone enables 

 us correctly to understand the different results obtained in experiments with 

 heat and cold conducted in connection with the development of butterflies, 

 which are for the most part wrongly interpreted. A comparative study as to 



