PALATABILITY OF SOME BRITISH INSECTS. 865 



(although nearly the same as P. rapce ih this respect), was 

 distinctlj" the least palatable of the foiu\ On the other hand, 

 P. napi and E. cardamines, in which the cryptic resemblance is 

 carried to its highest pitch, appeared to be most palatable ; but 

 a larger number of experiments is greatly to be desired. 



The results obtained in the two former species are of much 

 interest in relation to the experiments upon Mplanargia galathea 

 (p. 827). 



Pages 820-822. 



The evidence that T". urticce is not very palatable agrees with 

 mj' own experiments * with a Marmoset ; and I obtained the 

 same results with V. io when offered in considerable numbers to 

 lizards. It is probable that the procrj'ptic under surface of the 

 Vanessas is chiefly related to the attacks of mammals and of very 

 hungry birds during the long hibernating period. The special 

 interest in the eye- spots of V. io manifested by the Syrian Bulbul, 

 accords with previous observations on other insects and other 

 insect-eaters. Reptilian as well as Avian t. 



Pages 823-825. 



The evidence of a certain amount of unpalatability in 

 Brenthis (Argynnidae) is consistent with the degree of proeryptic 

 defence attained in this genus. It is also of much interest in 

 connexion with the experiments on Araschnia levana, the early 

 or levana form of which is probably a mimic of the species of 

 Brenthis. The examples of the Araschnia tested by the author 

 (pp. 823-824) were of the form pi^orsa, belonging to the later 

 brood, and generally looked upon as mimics of the White Admiral 

 [Limenitis sibi/Ua), which appears upon the wing at about the 

 same period. The expei-iments hei-e recorded prove that the 

 mimic is certainly unpalatable to several birds, and support the 

 conclusion that the resemblance is Miillerian or Synaposematic. 

 The evidence, so far as it goes, points indeed to the inference 

 that Araschnia is more unpalatable than its Brenthis model. A 

 few experiments on the imago of L. sihylla made by Mi-. Pocock 

 in the summer of 1910, also indicated that the prorsa form is 

 more unpalatable than the Lirnenitis. There is, however, nothing- 

 improbable in a Miillerian mimic being more highly protected than 

 its model. The role of model is related to many characteristics, 

 and relative abundance, conspicuousness, and extent of range 

 may play their part as well as relative unpalatability. Thus it is 

 probable, from its habits and flight, that the Eastern European 

 Neptis lucilla, W.Y. { = sappho Pall.) is more distasteful than its 

 Lirnenitis models, but the latter are widespread and abundant 

 species, and it is reasonable to suppose that the memories of 

 Eviropean insect-eating animals are more deeply impressed by 

 their pattern than by tliat of the Xeptis. 



* Trans. Eut. Soc. Loud. 1902, p. 442. 



t 'Essays on Evolution ' (Poulton), 1908, p. 210 : see also p. 326. 



