408 Mr. G. A. K. Marshall on 



the frogs aucl the water tortoises experimented on by 

 Professor Plateau. He then offered the larvte to the 

 larger Triton aljycsio'is, which paid no attention to them. 



Professor Plateau's experiments on spiders are described 

 on pp. 825-7, where it will be seen that the larvae were either 

 neglected or treated as A. G. Butler described hy Amaicro- 

 hius ferox, Tcgcnaria domcstica, and Agelcna labyrinthica. 

 The imago on the other hand was constantly eaten by 

 T'cgcnaria, offered once to Upcira diadcma and eaten with 

 avidity, killed but rejected by Agelcna. 



His experiments on Carahits auratus and two species of 

 Dytiscus are described on p. 330 : they indicate clearly that 

 these predaceous beetles freely attack and devour both 

 larva and imago of Abraxas. 



The above is a resume of the evidence as Plateau gives 

 it, although he also includes the attacks of insect-parasites. 

 These I have quoted on p. 338; but the other observations 

 recorded in the same section of this memoir indicate that 

 insects Avith warning colours and distasteful to the (non- 

 parasitic) enemies of their class are specially liable to these 

 attacks, so that the results obtained by Plateau in this 

 section (§ 10) of his paper tend, so far as they go, to support 

 the conclusions he seeks to destroy. 



As regards the vertebrate enemies Plateau has supplied 

 a considerable amount of evidence in support of the con- 

 clusions of his predecessors. He shows that certain 

 spiders are probably, and some predaceous beetles almost 

 certainly, enemies of AhraMis in one or more stages. These 

 latter facts are most interesting and valuable, but they no 

 more controvert or throw doubt upon the behaviour of the 

 generality of insectivorous vertebrates than the admitted 

 exception of the cuckoo, and the pleasant impression pro- 

 duced upon the Professor himself. All the theory of 

 aposematic colouring re({uires is that the conspicuous form 

 should be uiilit as food for the majority of the enemies 

 of its class; and this has been abundantly proved in the 

 case oi Abraxas. It is unimportant whether our anthropo- 

 morphic terms unpalatability, distastefulness, etc., truly 

 express what an animal feels, if we admit the fact, as we 

 are bound to do after the experiments, that the mnjority 

 of insect-caters after trial do not eat the insect, and after 

 one or more trials do not seek to eat it unless driven by 

 starvation. Exactly what impressions they feel we can 

 never know, and it is quite unnecessary for the theory 



