330 Mr. G. A. K. Marshall on Birds as a Factor in the 



Wherever undoubted distastefulness has been demon- 

 strated by experiment, we find that the species are 

 invariably characterised by a slow, sailing or laboured 

 flight ; moreover they do not (except in a very few 

 instances) exhibit any procryptic coloration on the under- 

 surface. Thus the acquisition of nnpalatability, which is 

 probably a more specialised form of defence, leads to a 

 partial or complete abandonment of the other two 

 methods. Tijis is only what we should anticipate in 

 accordance with Wallace's law of warning coloration. 

 For distastefulness loses a great part of its protective 

 value unless accompanied by coloration or habits which 

 make for conspicuousness and thus advertise the un- 

 pleasant qualities of the insect. 



But the diminution of activity in flight can have 

 significance only in relation to winged enemies ; that is 

 to say, either birds or predaceous insects. Of the latter 

 the only ones of general occurrence which are known to 

 attack butterflies are Asilid flies, wasps and dragonflies. 

 With regard to Asilidae, Professor Poulton has clearly 

 shown in his admirable paper on predaceous insects 

 (Trans. Ent. Soc. Lond., 1906, p. 363) that these flies are 

 apparently indiscriminate in their attacks on other in- 

 sects and do not appear to be deterred by any nauseous 

 qualities. There seems to be a certain amount of evidence 

 to show that the same is true of wasps, including Belt's 

 record of the wasp which stocked its nest with Heliconii ; 

 and probably this applies to dragonflies also, though there 

 is very little evidence at present as to the nature of their 

 food (cf. Poulton, I. c. p. 399). But conspicuous colouring 

 and slow movements can be of no utility against enemies 

 which devour with equal avidity both nauseous and 

 normal forms ; indeed, rather the reverse. It is therefore 

 difficult to avoid the conclusion that the modification of 

 flight which is so constantly correlated with nnpalatability 

 in butterflies must have a particular signiflcance in regard 

 to the attacks of birds. 



The Influence of Lizaeds on Mimicry. 



That certain species of lizards will in a wild state eat 

 butterflies is well established, and it has often been 

 suggested that these reptiles play a considerable part in 

 the production of butterfly mimicry. Yet the evidence 



