tJie colours of certain- Lepidopterd. 353 



The results of these exi)eriments have heeu described 

 under 11. and XIY. respectively (see pp. 34JI and 348). 



It is interesting to note that the darkening of a larva 

 which has become green appears to be more difficult 

 than the converse change. Thus the effects, if any, in 



XXVI. were confined to the dorsal area (even in the 

 larva transferred on July 31), whereas the pigment in 



XXVII. had become somewhat lighter over the whole 

 surface. 



F. White Surroundings. 



{Sec Tabic, page 354.) 



The faintly greenish white larva is shown in Plate 

 XIV., fig. 15 ; the faintly brownish in fig. 16. Eeference 

 to these figures will show how completely the white 

 dominates the tendency to other colours ; and, as shown 

 above, there was no such tendency in most of the larvse 

 in XXVIII. 



The less marked effects witnessed in XXIX. and XXX. 

 were probably due to the facts that the experiments 

 began later, and tha,t the enamelled surface was less 

 congenial to the larvte than the paper. But it would be 

 well to repeat these experiments. It must be remem- 

 bered, too, that the sticks in XXX. were not so white 

 as the others. 



G. Surroundings of other Colours. 



{See Table, page o55.) 



The infiuence of blue (XXXI. and XXXII.) is evidently 

 strongly in the direction of dark forms. The uniform 

 purplish brown colour of all the larvfe in XXXI. must be 

 something more than a coincidence. It is clear that the 

 blue not only tended to produce dark larvaB, but dark 

 larvae of a certain kind. At the same time the larvae 

 did not resemble the blue spills, but were such as would 

 have been protected on dark purplish brown twigs. 

 Some quality in the light reflected from such twigs would 

 cause the larval adjustment, and this experiment sug- 

 gests that the proportion of blue rays may be the effec- 

 tive stimulus which causes the larvae to assume the 

 appropriate shade of brown. The appearance of thet-e 

 larvae is well shown in Plate XIV., fig. 17, where the 



