I92I.] N. Annandai,e & C. Dover : Fauna of Barkuda I. 351 



criminate feeders among birds, such as crows and mynas, which (as is 

 evident from their omnivorous habits) are indifferent to the precise 

 nature and taste of their food, will probably eat almost anything 

 not actuall}' poisonous when pressed by hunger. Moreover, though 

 the proof that birds do eat butterflies, unpalatable and otherwise, 

 is now convincing, the number of individuals they destroy must 

 be comparatively small, as is shown by the amount, and kind of 

 evidence it was necessary to collect all over the tropics in order 

 to prove that they did so. We have no evidence that the crows 

 and mynas seen attacking distasteful butterflies on Barkuda were 

 young birds. Nor were such attacks often observed, and it is by no 

 means improbable that creatures so perverse as the Indian crows, 

 in which something very like reason and almost what we may call 

 a sense of humour are strongly developed, may sjometimes attack 

 and even devour butterflies in mere wantonness. 



Lizards (Calotes versicolor major) were observed devouring 

 Danais chrysippus both on the island and on the mainland a few 

 miles away, and a tree-snake {Dendrelaphis trislis) was once seen 

 eating a specimen of Colotis Calais amatus. These reptiles, though 

 by no means abundant on Barkuda, are not actually scarce. 



Evidences of the Attacks of Enemies. 



In writing on butterflies showing evidence of the attacks of 

 enemies, it is necessary not to regard every damaged butterfly as 

 one which has been attacked, for it is probable that butterflies are; 

 often damaged in sudden gusts of wind while wending their way 

 through dense jungle, and that these damages sometimes look 

 like the injuries caused by enemies. As a general rule, however, 

 the results of wear and tear show mostly on the forewings, while 

 the injuries caused by birds or lizards are usually present on the 

 hind wings. The only instances in which it is reasonably certain 

 that a butterfly has been attacked by a vertebrate enemy are those 

 in which its injuries are quite symmetrical, but in others, with 

 caution and experience, a fairly accurate conclusion may be reached. 

 In drawing up the table on p. 352 we have been careful to include 

 in the " injury " columns only those specimens which have been 

 symmetrically injured, or, perfectly fresh specimens which have 

 undoubtedly been injured by a bird or lizard, as is shown by the 

 form of the injury. Worn specimens though apparently damaged 

 by an enemy have been included in the " perfect or worn" section, 

 as it is possible that their injuries have been caused by various 

 accidents. 



Only the commonest or more interesting species have been 

 included in the table. The data we have collected would seem 

 by themselves to show that the local Lycaenidae and Hesperidae 

 are either rarely attacked by enemies, or, are not able to escape 

 at the cost of a damaged wing, but it is significant that the 

 Pieridae would seem also to be more or less immune from attack. 

 This is probably due to a number of factors in environment and 



