( 47 ) 



probably under the influence of the black Group I (= E) 

 derived from Hawaii. 



(6) It was not to be expected that the members of a Colour- 

 group formed on any island, would, after reaching another 

 island, always produce a group exactly like that in which they 

 originated. The immigrants would be working upon different 

 material, and would also be likely themselves to undergo 

 changes of pattern. The peculiarity of the red-marked 

 Group III (= C) on Oahu may perhaps be thus explained. 

 Especially may changes be expected to occur in an appear- 

 ance, like that of the wings, due to a uniform dark pigment 

 combined with the " structural colours " of thin plates. 



That there was undoubtedly a strong tendency to produce a 

 single group on a single island was shown by the condition of 

 Hawaii and Kauai, at opposite ends of the chain, and it was 

 suggested that the same tendency existed in the other islands, 

 but had been masked by the effect of accidental inter-island 

 immigration. 



Certain classes of facts established by Dr. Perkins, and 

 described in his paper, were only intelligible on the hypothesis 

 of mimetic approach due to selection by enemies attacking by 

 the aid of sight. 



(1) The persistence, in certain individuals belonging to the 

 black Group I, of yellow bands on the under-surface where 

 lxiv] 



they could not be seen. The same phenomena were found 

 independently in the Eumenidae and the Crabronidae. 

 Dr. Perkins, with whom these observations had been discussed, 

 entirely agreed that, as evidence, they were very important 

 indeed. He also stated that the yellow bands of the Kauai 

 Group II are clearly visible in flight. 



(2) Species belonging to the same Structure-group, and 

 therefore closely related, were distributed among different 

 Colour-groups. In like manner the Kauai Ci'abros resembled 

 its dominant, banded Eumenids, and the Hawaii Crabros its 

 black Eumenids. 



These facts fell into line with those which had been 

 observed in the mimetic associations of the same and other 

 groups of insects in other countries ; and this was equally true 



