350 Records of the Indian Museum. [Vol. XXII, 



larger herbaceous plants are either absent, or occur merely as 

 occasional visitors in the imagine state. In many cases the food- 

 plants are not those with which the caterpillar is commonly asso- 

 ciated. The fact that a single larva of Papilio iristolockiae (usually 

 a rare butterfly on the island, on which Aristolochias do not grow) 

 was found associated with one of P. hector (a butterfly of fairly 

 common occurrence but not abundant on Barkuda) and feeding on 

 the leaves of the Sword-l-.ean Canavalia ensijormis is particularly 

 noteworth}- in this connection. The scarcity of the Satyrinae 

 and the comparative paucity of most Lycaenid and Hesperiid 

 genera are also noteworth)^ features, and are probabl}' due to the 

 absence of suitable food-plants. The few skippers that occur are 

 mostlj- immigrants. 



The habit of immigration is also prevalent among some of 

 the most abundant resident Nymphalidae and Papilionidae such as 

 Danais chrysippus , Papilio polytes and P. demoleus, while it is pro- 

 bably habitual among the larger Pieridae such as the species of 

 Catopsilia. No large flights of any butterfly were observed ; the 

 immigrants flew singty across the lake. 



The Enemies of Butterflies on Barkuda. 



Insectivorous mammals, birds and reptiles are scarce on 

 Barkuda, and man}' of the common species known to feed on butter- 

 flies, absent. Those enemies of butterflies that exist on the island 

 do not seem to be particularly discriminate in their choice of food, 

 as the remains of unpalatable butterflies such as the Danaines are 

 not infrequently to be found in circumstances that prove they have 

 served as food for vertebrates. On the few occasions that mynas 

 and crows were seen actually attacking a butterfly, it was either the 

 " distasteful " Danais chrysippus, or a L3'caenid. The " Blues", 

 however, seemed to form quite an appreciable part of the daily 

 diet of the mjmas, and these birds have been watched eating the 

 butterflies, frequently denuding them of their wings and legs 

 before doing so. 



Small flocks of Bee-eaters {Merops vtridis) often fly over 

 from the mainland and do much damage among Papilio polytes, in 

 spite of its skill in eluding pursuit among thickets of shrubs. The 

 remains of this butterfly can be seen on the ground, under the 

 branches where these birds have perched in the intervals of their 

 short and rapid flights.' 



In short, the butterflies most liable to attack by birds on 

 Barkuda are the commonest and most conspicuous species. 

 Conditions are peculiar, however, in that the two genera of birds 

 that most frequently attack butterflies are not habitual butterfly- 

 eaters in the sense that the bee-eaters, etc., are. The more indis- 



i Marshall, {Trans. Ent. Soc. Land., 1909, p. 339) remarks thai Bee-eateri 

 probably cut off the wings of the butterflies they capture before eating them. To 

 this view we ourselves incline as the wings only of P. polytes, in most cases neatly 

 severed from the humerus, were found. 



