WARNING COLORATION. 161 



appear to get stung, thougli the wasps had every chance of 

 -stinging. The second wasp was eaten much more slowly than 

 -the first, the lizard holding it for some time in its mouth with- 

 out chewing it. This experiment recalls that made with the 

 •drone fly and recorded above. They show how difficult it is to 

 interpret experiments, apart altogether from the formulating of 

 wide-reaching generalisations. The wasp presumably advertises 

 its sting by its bright colours, while the drone fly is supposed 

 to delude its would-be enemies into the belief that it could 

 produce a sting if it liked. There is not supposed to be any 

 question of uopalatability, and yet the hesitation of the lizard 

 looks niore like the recognition of a disagreeable taste 

 -than the fear of a sting. Another queen wasp was eaten by 

 :a Laughing Jackass, but was well crushed first, and the head 

 .bitten off. 



Lithobius forficatus is instanced by Mr. Poulton as an 

 •example of a protectively-coloured animal; it is like the soil in 

 ■which it lives, and is inodorous. Its mandibles would be hardly 

 formidable to a bird or to a lizard which does not care for the 

 •stmg 01 a wasp. 



Prof. Weismann found that it was greedily eaten by Lacerta 

 ■viridis. Mr. Finn and I found, on the contrary, that the same 

 species of lizard bit and refused one with what are generally 

 termed, by those who make such experiments as these, " signs 

 of disgust." The lizard rubbed its mouth on the ground as if 

 trying to get rid of some disagreeable substance. A larger 

 rspecimen then attacked and bit it several times, wiping its 

 mouth afterwards in the same way. The centipede ended by 

 getting away in safety. 



Another centipede was offered to a green woodpecker : the 



woodpecker picked up the centipede, but immediately flicked 



it away as if it were unpalatable. 



11 



