474 Mr. Vonlion^s further experiments upnn 



they are not subject to much extermination from this 

 cause. Their chief foes at this period are dipterous and 

 hymenopterous parasites, but the deaths are not nearly 

 so numerous as might be inferred from Mr. Bateson's 

 experience. In 1886 I experimented on 700 larva 

 belonging to many colonies (the exact numbers can be 

 ascertained in my paper, Phil. Trans., I.e.), and in 1888 

 upon the many hundred larvse tabulated in this memoir; 

 but I am quite sure that the deaths from this cause did 

 not come to anything like 10 per cent. The Tachina 

 larvie nearly always emerge before the Vanessa pupates, 

 and are quite obvious, together with the dead or dying 

 larvae. Tbe proportion of deaths in V. io was not widely 

 different. 



There are great differences between the colonies in 

 this respect, as is well shown by Mr. Bateson's experience 

 of 5 or 6 deaths to 1 survival. It would be very inte- 

 resting to observe whether there are any individual 

 differences in the methods employed by these larva) in 

 keeping off their insect foes, so as perhaps partially to 

 explain why some colonies are almost swept away, while 

 others are nearl}' untouched. I hasten to anticipate 

 Mr. Bateson's objection by stating that this suggestion 

 is not intended as a "basis for argument," but as a 

 stimulus to observation. Everyone who has observed 

 these larvae must have noticed the twitching of the wild 

 larvae when disturbed, and how readily the movement 

 becomes concerted and common to a whole detachment 

 of a colony. This is probably one of their methods of 

 defence against such foes. 



The numerical argument alone drives us to the two 

 remaining stages for the chief extermination, and it is 

 impossible, on these grounds alone, to admit that the 

 pupal stage, short as it is, can escape. 



Sir. Bateson considers that the theory of the protective 

 significance of colour has only been "applied to the case 

 of these pupae by an indiscriminate extension of de- 

 ductions made in other cases fairly enough, as, for 

 example, in that of the larvae of A. betularia." And yet 

 •we can only suppose that these latter are large enough 

 to be the prey of insect-eating vertebrates for about the 

 length of the joupal period of the Vanessa ; and, as for 

 concealment, the latter would be far more i)erfectly 

 hidden were it not for walls and palings, which are not 

 a strictly natural environment, 



