258 THE EFFECT OF TEMPERATURE 



ception to this rule. Also he has found that the 

 cocoons themselves may undergo protective colour- 

 ation. 



The first recorded instance of variable protective re- 

 semblance in Lepidoptera is due to T. W. Wood,* who 

 in 1867 demonstrated it for the pupae of the Large and 

 Small White butterflies (Pieris brassica? and P. rapes). 

 A few experiments were made from time to time by 

 other observers, but it was not until 1886 that they 

 were undertaken systematically on a large scale. This 

 was done by Professor Poulton, who obtained most 

 striking results.f Working upon over YOO chrysalides 

 of Vanessa uriicce (Small Tortoiseshell), he found that 

 pupae placed against black surroundings became as a 

 rule extremely dark, whilst against white surroundings 

 " not only was the black colouring matter as a rule ab- 

 sent, so that the pupae were light-coloured, but there 

 was often an immense development of the golden spots, 

 so that in many cases the whole surface of the pupae 

 glittered with an apparent metallic lustre." Against a 

 gilt background a much higher percentage of gilded 

 chrysalides was obtained, and this led Professor Poulton 

 to suggest that in its original habitat the larvae pupated 

 either against glittering micaceous rocks which had a 

 somewhat metallic appearance, or against dark rough 

 weathered rocks, and that they had acquired the power 

 of protectively resembling either of these surfaces. In 

 that such metallic looking rocks occur over a compara- 

 tively limited area, whilst the species has a consider- 



*Proc. Ent. Soc., vol. xiii., 3d series, p. xcix. 1867. 

 fPliil. Trans. 1887, B- p. 311, also " Colours of Animals," p. 119 

 et seq. 



