334 Mr. Poulton's further experiments. 



Cojielusions. — The fact that one larva became brown 

 in VI. is not sufficient evidence that any results were 

 produced, except b}^ the light being somewhat dimmed 

 by the surrounding twigs, and especially by the effects of 

 crowding in a small cylinder. The experiments of 189'2 

 show that both these causes are effective in producing 

 dark larv?e. It is probable that the dark larva is to be 

 accounted for in one of these ways, inasmuch as green 

 surroundings in which such causes did not operate never 

 produced a single dark larva (see below). 



E. Green Surroundings. 

 {See Table, page 335.) 



The strong susceptibility to green surroundings when 

 nothing brown or dark is present is extremely clear in 

 these results, as well as in the comparison experiments 

 of C and D (V. and VII.). Among the 105 larvae which 

 matured in these 7 experiments not a single exception 

 occurs. 



A very characteristic green larva with a brownish 

 shade along the dorsal area was painted by Miss Cundell, 

 and is represented in Plate XIV., fig. 8. It is shown 

 in a xeA-j common attitude, resting on a green twig of 

 Populns nieira. 



Many of the green and dark larvae from one of these 

 experiments, and either I. or II., were interchanged for 

 a few days during the last stage. No effects were pro- 

 duced, and it was clear that the larvae are not susceptible 

 to a short exposure during this period of life. 



F. White Surroundings. 



Nine larv.'B were fed upon Populus nigra, the sur- 

 roundings being green, except for the presence of many 

 white paper spills. The experiment was arranged July 2(3, 

 and the larvre were compared with the others Aug. 17. 

 Some of them tended towards green, and some towards 

 light brown, but in both, these colours were, without 

 exception, almost hidden under the predominant whiteness 

 which gave the larvae a very remarkable appearance, 

 utterly unlike that in any of the other experiments. 

 They were again examined Aug. 24, when the whiteness 

 of the larger larva) was even more pronounced. They 

 were carefully compared Sept. 2 with 12 of the green 

 larvae upon nut (Experiment V.) ; 3 remained small in 



