the colours of certain Lepidoptera. 361 



experiment were not uniform, but depended upon 

 averages. I was also very anxious to investigate the 

 pupa of Vanessa io as completely as that of V. iirtic(e. 

 Considering the importance of the conclusions which 

 seem legitimately to follow from the results of conflicting 

 colour experiments, I was desirous of repeating these, 

 and of devising some improved method by which the 

 larvae could be subjected to the conditions for the whole 

 of the sensitive period. Coloured glass screens have also 

 been employed in many of the experiments, especially 

 with the PieridcB. Attention was also directed to other 

 special points, some of which came out in the course of 

 the enquiry. 



Crowding the larvfe affects the colour, and therefore 

 the size of the receptacles becomes a matter of im- 

 portance. These are described in detail at the end of 

 the paper, and will be referred to by numbers, accom- 

 panied by a very brief description, under the experiments 

 themselves. 



Experiments upon Vanessa urtic^. 

 1887. 



In working at these pupae in the preceding year, I gained 

 a very strong impression that the pupae in darkness were, 

 other things being equal, formed later than those in the 

 light. If this were the case, it appeared possible that 

 time might be an element in the production of the dark 

 superficial pigment which prevents the golden appear- 

 ance. I had concluded that this protraction of the 

 period before pupation occurs, from the experiments on 

 Pieridce, as well as those on VanessidcB (Phil. Trans., 

 1887, B., pp. 339 and 432), and my friend Mr. G. C. 

 Griffiths had independently noticed the same thing with 

 the Pieridce (Trans. Ent. Soc. Lond., 1888, pp. 256, 257). 

 I was therefore anxious to make some experiments with 

 this special end in view, the impression I had gained 

 being merely the incidental result of experiments 

 intended for other purposes. 



I made three such sets of experiments upon V. urticce 

 in 1887, and a brief summary of the first is given in a 

 footnote to the paper referred to above (Phil. Trans., 

 Lc, p. 339). 



Before detailing these experiments, it will be necessary 



TRANS, ent. soc. LOND. 1892. — PART IV. (DEC.) 2 E 



