860 Mr. 'Ponlton' 8 further experiments upon 



tree. For several years I have observed the corre- 

 spondence between wild larvaB and their food-plants. 

 Thus they are especially common in gardens feeding on 

 rose, and, so far as I have observed, these are invariably 

 green and well concealed among the abundant green 

 shoots on which, and not on the older brown wood, they 

 are found. I have found the same to be true of larvae 

 found on the green shoots of sallow and Ilibes ainericana, 

 while larvse found on the brown branches of cherry were 

 brown, and the same was the case with one found on apple. 

 Two green varieties found feeding on broom were brought 

 me by Miss Gould during the past autumn. The larvae are 

 very commonly beaten from birch and oak, and these are, 

 so far as my experience goes, always dark varieties. Mr. 

 Arthur Sidgwick, who has had a wider experience of the 

 wild larvae, not only agrees with this, but tells me that 

 be always notices a difference between the dark larvae 

 beaten from the two trees, corresponding to the difference 

 between the twigs on which the larvae rest in the two 

 cases. 



The most interesting example, however, was told me 

 by Dr. Stacey Wilson, of Birmingham, who beat the larva 

 from a lichen-covered food-plant, and found it so exactly 

 resembled the lichen that he thought it could not be this 

 species at all, and was only convinced when the moth 

 appeared. Had I known this earlier in the summer, 

 I should have tried the effect of lichen-covered sticks. 

 In a complex result of this kind it would be especially 

 interesting to attempt to determine the peculiar quality 

 in the reflected light which acts as the stimulus. 



There is thus a considerable body of evidence to prove 

 that the results obtained by breeding in confinement 

 under certain conditions, point to the existence of a 

 power of individual colour-adaptation which is possessed 

 and is widely used by the wild larvae in their natural 

 surroundings. 



C. Experiments on the Colours of Pup.e, 

 1887—1892. 

 These experiments were partly undertaken in order to 

 confirm the results of my previous work (Phil. Trans., B., 

 1887, p. 311), and partly to make out further details. 

 Professor Weismann had suggested to me that con- 

 firmation was desirable, inasmuch as the results of 



