356 Mr. G. A. K. Marshall on 



unobtrusive manner, which might account for the paucity 

 of records. The fact that birds have been seen to capture 

 moths more frequently than butterflies need not neces- 

 sarily imply a preference for the former insects, but might 

 be explained on the supposition that they are aware that 

 they can be captured more or Jess easily on the Aving, and 

 therefore that when a moth does happen to get well up 

 into the air in open country it is promptly pursued, 

 whereas under similar conditions a butterfly would be 

 allowed to pass unmolested. While on the subject of swift 

 flight I might mention that I was much struck during my 

 visit home with the slow flight of English butterflies as 

 compared vvitli the generality of South African species. 

 I am inclined to agree with Trimen in his Presidential 

 Address to the Entomological Society, that birds are 

 among the chief enemies of butterflies. That they have 

 been the chief, if not the only, agents in the production 

 of mimicry, whether Batesian or Mlillerian, I have little 

 doubt. It is highly significant that mimicry in its fullest 

 development is only to be found in forest-clad regions 

 where insectivorous birds arc most abundant. Moreover, 

 I am not aware of a single instance of true mimicry 

 among species which habitually settle on the ground. 



Salisbury, March 10, 1898. — It would seem that mere 

 unpleasantness of taste or smell would hardly be suflicient 

 to give so great an immunity from attack from birds as is 

 apparently enjoyed by the DaiiainiG and Acriviiuv, unless 

 accompanied by poisonous or unwholesome qualities — at 

 least, if we may judge by other orders of insects. A large 

 number of Rhynchota, for instance, possess a very un- 

 pleasant smell, and yet their colouring is procryptic instead 

 of aposematic. In the crop of the great spotted cuckoo 

 I have found a large green Pentatoinid, which in the 

 strength and unpleasantness of its smell is only beaten 

 by Petascclis reruijx-s, our largest Hemipterou. Again, in 

 the crop of the racquet-tailed roller (Coracias caudata, 

 Trim.) I have found a full-grown specimen of a large 

 Fhyiaatcus locust, which is a most evil-smelling beast. 

 This insect appears to combine procryptic and aposematic 

 colours; fi>r when settled its general green colour is 

 eminently lu'otective, but during its laboured flight it is 

 most conspicuous owing to its brilliant crimson and purple 

 hind- wings. If annoyed when settled on the ground they 

 often raise their wings over their backs (clearly to exhibit 



