tJie colours of certain Lcpidopt'era. 331 



B. Artificial Dark Surroundings. 



A single experiment was made with sticks covered 

 with black tissue paper taking the place of twigs which 

 were naturally dark, like those of I. and 11. This and 

 some of the other experiments were begun rather later 

 than those just described ; for the whole stock of larvas 

 was not exhausted on July 15th. 



Experiment III. 

 Black-paper-covered sticks intermixed with food-plant. 



July 20 9 larvffi introduced. 



Aug. 17 The larvas were very large. All 9 were very black, at 



least as dark as those of I. and II. 



Aug. 30 4 ceased feeding. 



gept, 4 2 ceased feeding. They remained as dark as ever to the 



end ; 1 was dead by Sept. 5. 



This experiment shows that artificial may be as effec- 

 tive as natural surfaces. It is probable that the com- 

 parative failure in the case of li. cratcsgata (see p. 324) 

 was due to the growth of mould upon the black paper, 

 making it much Hghter in appearance. 



C. Very small proportion of dark twigs in sur- 

 roundings, AND COMPARISON EXPERIMENT WITH GREEN 

 ALONE. 



Two experiments were then made with the object of 

 testing the susceptibility of the larvae to a very small 

 proportion of dark material in the environment. Inci- 

 dentally the effect of green surroundings produced by 

 another food-plant (nut) was also tested, and found to be 

 as effective as that due to Popiilm nigra. 



The experiments were arranged as follows :— 25 young 

 larva? were introduced July 20 into a large glass lamp- 

 shade (about 165 mm. high, and the approximate capacity 

 of 1300 cc). On Aug. 21 they were removed to a larger 

 lamp-shade (204 mm. high, and the approximate capacity 

 of 1900 cc). These relatively large areas were kept filled 

 with green leaves and shoots of nut, intermixed with 

 which were 5 small dark pieces of dead twig. Three of 

 these were about 40, 65, and 75 mm. long respectively, 

 while 2 of them were about 80 mm. in length; the 

 diameters varied from 3 to 5 mm. They were unbranched, 



2 c2 



