upon lepidopterous larva, <#c. 815 



coverings the condition of the food-plant could not be 

 watched, and sometimes it withered, or was eaten earlier 

 than I expected. Thus the larvae were sometimes 

 without food for a few days, and it was then found that 

 S. ocellatus had lost their caudal horns, while in some 

 cases the dorsal surface of the posterior part of the body 

 had been also nibbled. This habit of nibbling off the 

 caudal horns of other individuals is well known in this 

 species and in D. vinula, but I do not think that it has 

 been recognised as the result of hunger and as an 

 obvious tendency towards cannibalism. This is proved 

 by the fact that other lots of larvae, always abundantly 

 supplied with food, were either not injured at all or only 

 in a very small proportion of cases. In one instance I 

 kept some larvae of D. vinula and S. ocellatus together in 

 a blue cage, and on one occasion the food-plant had 

 withered, and while it was being renewed I observed that 

 a D. vinula was gnawing the thoracic leg of an 

 S. ocellatus, and when the two were separated the former 

 soon returned and seized one of the claspers in its 

 mandibles, and bit it until it bled. Under similar 

 circumstances I have found an almost full-grown larva 

 of A. betularia which was engaged in swallowing a 

 small larva of the same species. The small larva was 

 held tightly in the clasp of the thoracic legs, and 

 nearly half of it had disappeared when the observation 

 was made. 



These uniform results of the absence of food in all 

 three species of purely phytophagous larvae which have 

 been placed under such circumstances seem to offer a 

 probable explanation of those instances in which canni- 

 balism is well know r n to occur. 



13. The young larv.e of Vanessa urticje and 

 Saturnia carpini seek light. — Dr. Dixey informs me 

 that his larvae of S. carpini, when young, always 

 assembled on the side of the cylinder which was turned 

 towards the light, and I have made a similar observation 

 in the case of the young larvae of V. urticce. In both 

 cases rotation of the cylinders was followed by a corre- 

 sponding change in the position of the larvae. Both 

 these larvae are dark coloured when young, so that the 

 observation, as far as it goes, supports Lord Walsing- 

 ham's conclusions as to the advantage gained by the 



