10 



Since then the dimorphism we are speaking of is evidently so 

 completely dependent upon meteorological conditions, it is natural 

 to enquire whether some of these conditions cannot he artificially 

 reproduced, and whether we cannot by imitating the change of sea- 

 son, induce a butterfly which was preparing to emerge in one phase 

 to adopt the other in its stead. A good many experiments have been 

 made by different observers with a view to answering this question, 

 and the results obtained by them are of great interest. An early 

 worker in this department of research was Dorfmeister, who found 

 that by subjecting the pupjie of the summer brood of Araschnia 

 leiana to a lower temperature than they would meet with under 

 normal conditions, he was able to induce several of the resulting 

 butterflies to emerge as an intermediate form between levaiia and 

 prorsa, instead of the ordinary prorm, as they would have done 

 under natural circumstances. Several experiments of a similar 

 kind were afterwards made by Weismann in Freiburg and by Stand- 

 fuss in Zurich, but the most completely successful results with this 

 species have been obtained by my friend Mr. Merrifield, of Brighton, 

 a former President of the Entomological Society of London, who is 

 certainly the greatest authority in this country on the artificial pro- 

 duction of seasonal variations. Mr. Merrifield finds, as others have 

 done, that by subjecting the summer pup^e of A. levana to a low 

 temperature, it is possible in many instances to ensure the emer- 

 gence of the butterflies as a second generation of levana, instead of 

 prorsa. But the converse change, that of the normal winter phase 

 into that of the summer, by applying warmth to pupa?., if possible 

 at all, is attended with very great difticulty. On the other hand, by 

 employing the treatment of high or low temperature in the larval 

 (or caterpillar) instead of in the pupal period, he has succeeded in 

 the complete conversion of either phase into the other. His experi- 

 ments shovv conclusively that in this species it is not the pupal, but 

 the larval stage which is most susceptible to the influence of tem- 

 perature. x\n interesting difference between the two kinds of con- 

 version made itself manifest in these investigations. The transfor- 

 mation of the summer into the winter phase was found to be readily 

 effected in all except the very last period of larval growth, while the 

 converse change of the winter phase into the summer was, as a rule, 

 only brought about when the larvfe were exposed to artificial 

 warmth from the time of hatching. 



In this species, and in others experimented upon by Mr. Merri- 

 field, there is no doubt that temperature, if not the only, is at any 

 rate the chief determining factor in the seasonal transformation ; but 

 there is much reason to think that this is not universally the case. 

 In tropical countries generally, the most obvious meteorological 

 difference between the seasons is the alternation between rains and 

 drought rather than between heat and cold; and it appears from the 

 combined testimony of many observers, and also to some extent from 

 actual experiment, that in such regions degrees of dryness or of 



