B36 Mr. Toulton's farther exj)^ rim cuts upon 



the last stage, while G were nearly full-fed, and these 

 latter were chiefly compared. The white points on the 

 skin appeared to be far more abundant in these 6, and, 

 so far as any green tint appeared, it was of a whitish 

 bluish shade, instead of bright and yellowish, as in all 

 the larvffi upon nut. 



The green blood from 2 white and 3 bright green 

 larvae was then compared, to see whether any difference 

 in the shade of green was due to its colour. There was 

 some individual difference in the tint of the blood, but 

 this was true of both sets of larvje. It was clear that no 

 explanation was thus to be found, and that the seat of 

 effective colour was in the skin and the structures 

 immediately below it. 



The three smaller larvre were also white; and although 

 1 tended towards a dark variety, the tendency was 

 obscured by the whiteness. 



These results were so remarkable that, although there 

 was no exception, 1 did not venture to publish them until 

 1 had obtained confirmation. This, however, has been 

 forthcoming in the experiments of the past summer 

 (189'2), and the results have now been seen by many 

 naturalists. 



G. Effect of unsuitable food upon colour-relation. 



On July 20, 21 young larvae were introduced into a 

 cylinder, and supplied with lilac-leaves, dark twigs being 

 abundantly intermixed with the food-plant. The larvae 

 were observed as follows : — 



July 20. — 21 young larvae introduced. 



Aug. 9. — 11 alive, but small for age. 



Sept. 4. — 5 alive, but quite small for age ; reddish 

 brown in colour. 



Sept. 25. — 3 still alive, although these subsequently 

 died. 



It is therefore probable that the unsuitable food, 

 which prevented the larvae from attaining maturity, 

 did not interfere with their susceptibility to the colours 

 of the environment. The results observed on Sept. 4 

 show that all the 5 surviving larvae harmonized with the 

 dark surroundings. In order to furnish conclusive proof 

 that this result was due to true susceptibility, and not to 

 pathological change, it will be necessary to repeat the 

 experiment, employing green surroundings alone. Such 



