216 Dr. F. A. Dixey on 



out that T. .TtJiiojnca and hutlcri of Triiiien are respectively 

 dry and wet forms of the same species, and tlius, taking the 

 synonymy given in your- revision, hapale must fall as a 

 seasonal form of scnegalensis. I have not actually proved 

 the case by breeding, but I think you can take my obser- 

 vations on trust now,"* I may mention that I had some 

 time ago come independently to the conclusion that the 

 T. hajmh' forms could not be specifically separated from 

 the T. scnegalensis assemblage, and liad arranged the 

 examples in the Hope collection in accordance with that 

 view. But I do not think that even now the seasonal 

 relations of these forms are quite clear. 



In addition to the series just described, Mr. Marshall has 

 also presented to the Hope collection the greater number 

 of the specimens resulting from the experiments recorded 

 by him in the " Annals and Magazine of Natural History," 

 1901, vol.ii, p. 398. They exemplify the very slight effect 

 produced on tlie early dry-season broods by subjecting the 

 larvae and pupae to conditions of moisture without heat. 

 In Mr. Marshall's opinion, the amount of occasional inclin- 

 ation towards the wet-season form shown in this series is 

 no more than might have been met with in examples of 

 similar dates caught in the open. These specimens need 

 no further notice here, having been fully dealt with by Mr. 

 Marshall in his paper above referred to. 



5. Summary. 



The main points of the present paper may be sum- 

 marized as follows : — 



1. Catopsilia pomona, Fabr. (including C. catilla, Cram.), 

 and C. crocale, Cram, are phases of a single species. In at 

 least one part of its range, these phases appear to be in 

 relation with the seasons ; in other parts there seems to be 

 no such connection. 



2. In like manner Ckdo]i$ilia pyrantlic, Linn, is conspeci- 

 fic with C. gnoma, Fabr. Here the association of each 

 form with its own season is better recognized, but there is 

 reason to think that even in this case the relation by no 

 means obtains universally. 



3. There are many other instances on record of the 

 simultaneous occurrence in a given locality of forms of a 



* Proc. Zool Soc, 1898, p. 909. 



