374 THE BIOLOGY OF INSECTS 



Fieri dae and Lycaenidae were definitely selected for capture 

 *' from among a crowd of butterflies settled on the mud," 

 the noxious and mimetic insects present being left alone. 

 He also draws attention to the severed wings and fragments 

 of butterflies left by birds after hunting and eating them, 

 and the neat triangular nick often seen on both hindwings of 

 some butterfly that escaped from a pursuing bird at the 

 cost of those fractions of its wings caught by the bird's 

 beak. That choice is commonly exercised by insect-eating 

 birds is shown by Swynnerton's observations. He admits 

 that absolute inedibility among butterflies is very rare, and 

 that a hungry bird will seize a specimen of the large, con- 

 spicuous black and white- winged African danaine Ainauris 

 niaviiis in preference to the nymphaline Precis cebrene ; but 

 he insists that birds as a rule *' only eat Amauris when hungry, 

 but P. cehrene nearly to repletion-point." The advantage 

 of such comparative inedibiHty must be great, '' sufficient 

 to make the possessor worth mimicking," as " shown by the 

 immense meals that are sometimes eaten after the refusal 

 of a low-grade butterfly ; e.g. forty butterflies, including 

 fourteen large Char axes, by a roller after she had neglected a 

 MylothriSy and thirty-seven, including twelve large Charaxes, 

 after rejection of a Terias." Hale Carpenter (1921) records 

 the result of systematic trials with two monkeys, offered a 

 large selection of insects of various orders, some with the 

 concealing and others with the conspicuous type of colora- 

 tion. The two monkeys were offered altogether 375 insects, 

 of which 155 were ascribed to the former and 220 to the 

 latter type. Of the 155 procryptic insects 113 were eaten 

 and 42 refused, while of the 220 conspicuous insects only 

 44 were eaten, 176 being refused ; so that the reaction of the 

 monkeys with regard to the two types of insect agreed with 

 expectation in 73 per cent, of the specimens in the one 

 group and in 80 per cent, of those in the other. 



Of the many cases of mimicry now well known to 

 students of insects, that of Pseudacraea and Flanema, 

 elucidated in recent years by K. Jordan (1910), H. Eltring- 

 ham (19 10), and Hale Carpenter (1920), has the most 



