the colours of certain Leindoptera. 469 



GeometrcE, and specialised, like the latter, for conceal- 

 ment among twigs and on bark. It was the knowledge 

 of this specialisation which led me to test for that further 

 protection which is afforded by the power of colour- 

 change. 



The other larvffi (Smennthns, Sphinx, Aglia) which I 

 have tested are very inferior to the genus Catocala in 

 this respect, but from what Col. Swinhoe tells me, it is 

 evident that some of the Indian Sj^hingidcs are highly 

 susceptible. 



There may be a most extraordinary fluctuation in the 

 amount of susceptibility within the limits of the same 

 genus {Catocala, and in the pupfe of Paj^ilio). 



In Geometra; alone have distinct green larvae been 

 produced by these experiments. Prol)ably the great 

 majority of these larvae are sensitive. Out of 11 species, 

 many of which were selected at random, all but one 

 have proved to be so. 



There is no evidence that the results acquired by one 

 generation can be transmitted to the next {Ilitmia, Cro- 

 callis). The susceptibility is essentially an adaptation 

 to the fact that the individuals of each such species are 

 liable to find themselves in different environments, so 

 that any bias from the experiences of the past would 

 of course be injurious, unless the earlier and later 

 surroundings happened to correspond. 



In the case of R. cratcegata the test for hereditary 

 effects was as complete as it could be in one generation. 



Concerning the time which is necessary before the 

 colour-changes begin to appear — 



Some effect was produced in 8 days in young G. pa}}ilionana. 



8 „ C. electa. 



Much „ ,, 12 ,, C. elhiguaria. 



,, ,, ,, 14 about ,, 31. vunttunata. 



„ „ „ 11 „ ,, C. elocatii. 



,, ,, „ 13 (or less) ,, H. ahruptaria. 



,, ,, ,, 17 ,, R. cratii'ijata. 



„ ,, „ 8 ,, A. betularia. 



When carefully watched for, the changes are sometimes 

 seen to occur quite suddenly (C. elinguaria, R. cratcegata, 

 1886, II.). 



The effects cannot be reversed by reversing the sur- 

 roundings for a short time (6'. elinguaria, H. ahruptaria, 

 A. hctularia). 



When the conditions are uniform, the response to 



