the colours of certain Lepidoptera. 473 



Pupae may be observed more satisfactorily. Large 

 stones could be placed in a case with iifew mature larvae, 

 and when two or three (not more) pupae were suspended 

 to each, the stones could be removed to the borders of 

 some wood or field and noted twice a day. This would 

 be a fairer test than a garden. In this way the pupae 

 would be accessible to such probable enemies as insec- 

 tivora and rodents. In other comparison experiments 

 light pupae could be fixed to dark stones and vice versa. 

 This test could be satisfactorily applied to many species, 

 and other objects made use of as well as stones. 



With regard to imagines, we first require to find where 

 many of them conceal themselves at night and in 

 rainy weather. This could be accomplished by track- 

 ing the butterflies at dusk, marking the spot where they 

 finally come to rest, and again examining it at night 

 with a lantern. Butterflies bred in confinement could 

 then be placed at night in natural and unnatural situa- 

 tions, observed in half an hour to see whether they had 

 moved, and again observed and noted in the morning 

 before they begin to fly. The same kind of observation 

 could be made with hybernating species. 



In the meantime, however, there is some very strong 

 indirect evidence which is worthy of attention. Assuming 

 that a female V. urtica lays 300 eggs, every pair of 

 butterflies would be represented by 300 offspring in the 

 next generation, were it not for the deaths which ensue 

 at some period of development. Owing to this cause, 

 however, we know that, on the average, they only pro- 

 duce 2 mature offspring to take their place, and them- 

 selves become parents. The extinction of 298 out of 

 800 means a severe struggle for existence, and does not 

 support the assumption of "imaginary" enemies during 

 any stage or in any week, especially when we remember 

 that there are two or more generations in a year. And, 

 contrary to the commonly received opinion, I should 

 maintain that extinction is least during the first of 

 the three stages. The larvae are perfectly exposed and 

 obvious during their whole lives, and we know their 

 conditions fairly well ; of the pupae, in nature, and the 

 imagines, when concealed, we know comparatively little. 

 The larvae have been proved to be distasteful to certain 

 insect-eating animals, and the persistence of large 

 colonies through the whole of larval life proves that 



TRANS. ENT. SOC. LOND. 1892. — PART IV. (DEC.) 2 M 



