176 TROPICAL NATURE, AND OTHER ESSAYS. 



he had himself contributed some valuable facts and ob- 

 servations, Mr. Belt felt convinced that this frog must be 

 uneatable. He therefore took one home, and threw it 

 to his ducks and fowls ; but all refused to touch it except 

 one young duck, which took the frog in its mouth, but 

 dropped it directly, and went about jerking its head as 

 if trying to get rid of something nasty. Here the mi- 

 eatableness of the frog was predicted from its colours 

 and habits, and we can have no more convincing proof 

 of the truth of a theory than such previsions. 



The universal avoidance by carnivorous animals of 

 all these specially protected groups, which are thus 

 entirely free from the constant persecution suffered by 

 other creatures not so protected, would evidently render 

 it advantageous for any of these latter which were 

 subjected to extreme persecution to be mistaken for the 

 former ; and for this purpose it would be necessary that 

 they should have the same colours, form, and habits. 

 Now, strange to say, wherever there is a large group 

 of directly-protected forms (division a of animals with 

 warning colours), there are sure to be found a few 

 otherwise defenceless creatures which resemble them ex- 

 ternally so as to be mistaken for them, and which thus 

 gain protection, as it were, on false pretences (division h 

 of animals with warning colours). This is what is called 

 mimicry," and it has already been very fully treated 

 of by Mr. Bates (its discoverer), by myself, by Mr. 

 Trimen, and others. Here it is only necessary to state 

 that the uneatable Danaidae and Acrseidse are accom- 

 panied by a few species of other groups of butterflies 

 (Leptalidse, Papilios, Diademas, and Moths) which are 

 all really eatable, but which escape attack by their close 



