tlie colours of certain Lep'uloptera. 423 



black-lined compartments or cells, 3.0 centimelres wide, 

 3.8 high (six of them were 6 high), and 8 to 9 mm. 

 deep. The block tissue-paper lining sloped from the 

 sides to the back of the cells, so that there were no 

 sharp angles or corners. The frame, bearing 24 of 

 these black compartments, was placed vertically, and 

 covered in front with a sheet of white opal glass, which 

 was turned towards a strong light. Each larva was, 

 therefore, contained in a shallow black chamber with a 

 white front, both black and white surfaces being well 

 illuminated. When a larva, suspends itself to a surface, 

 it also rests upon il during Stage 11,, and even if dis- 

 turbed by its cramped position it must rest on the surface 

 for so much of this stage as is necessary for spinning the 

 silken boss. It is, therefore, safe to conclude that the 

 pupae suspended to the glass had spent Stage III., and 

 at any rate part, probably the whole, of Stage IL, 

 with the ventral surface closely applied to a white area 

 and the dorsal surface exposed to a black area, onl}^ 

 separated by a few millimetres from contact with the 

 larval skin ; and conversely with the pupae fixed to the 

 black compartments. 



Mature larvae belonging to the same company were 

 placed in 21 of the cells at the beginning of August, and 

 3 belonging to another company rather earlier. Both 

 companies were captured near Oxford. The pupae were 

 compared August 11. The results are shown below : — 



Experiment 102. 



{See Table, page A^4:.) 



The results are quite clear ; they prove that black is 

 far more powerful than white when the two conflict, that 

 there is no local effect of colour upon the skin, but that 

 the whole larval surface is uniformly sensitive, dorsal 

 and ventral alike. The two dull green pupae, (4)s, found 

 among the 9 dark ones which were attached to the opal 

 glass, may be most reasonably explained by individual 

 susceptibility to white rather than black, and to the 

 greater proximity of the surface which was in actual 

 contact with the larva. It is improbable that they 

 afford any evidence for a more sensitive condition of the 

 ventral area as compared with the dorsal, a view which 

 is hardly compatible with the other results. 



