( 84 ) 



together with all the other bred specimens referred to in the 

 paper, including the parents of families B, C and D. The 

 pupal cases of the bred Paeudacraeae were also exhibited 

 beside the butterflies which had emerged from them, and, for 

 comparison, there was included a series of the pupal cases of 

 Pseudaoraea imitator, Trim., from Natal, presented to the 

 Hope Collection by the late Mr. A. 1 >. Millar, of Durban. It 

 was Been that the two flat dorsal processes were rather less 

 pronounced and the cephalic processes distinctly shorter in 

 the pupae of the Natal form. Comparing Dr. Carpenter's 

 pupae with the whole series of ">1 Natal specimens, it was 

 al so seen that the apices of the two dorsal processes of bhe 

 Uganda pupae tended bo be directed backwards more strongly, 

 and that the contour of the processes and of the segments 

 lirt ween them formed a festooned outline instead of one that 

 was oearly smooth. The cephalic processes of the Uganda 

 pupae tended m turn upwards (viz. dorsally) at the tip, those 

 of imitator downwards, while the two processes of the latter 

 were more frequently separated. The Uganda pupae showed 

 ihe darker pigmentation, but this effect was probably due to 

 conditions. It was highly probable that bhis procryptic pupa 

 is susceptible to the colours and degrees of illumination of its 

 normal environment. The method of suspension from the 

 edge, near the leaf tip or near BOOM angle of a partially eaten 

 leaf, was similar in both forms. 



It was not necessary to assume that differences of the kind 

 described above imply specific distinction. Dr. Carpenter's 



description of the way in which the hollow cephalic and dorsal 

 processes gained their shape iii the fresh pupa was an indict 

 tion that they were of no morphological significance but 

 merely an adaptation which promoted the concealment of the 

 pupa by making it more leaf-like. 



[cxvii 

 'The synonymy of the Pseudaoraeae required considerable 

 mollification if we might assume from Dr. Jordan's researches 

 and Dr. Carpenter's breeding experiments that the close 

 allies of Pa. ourytus., L., were an interbreeding community. 

 Prof. Poulton had consulted Dr. Jordan on the subject and 

 he had agreed that it would be desirable to iutroduce the 



