on some Carnivorous Insects. 335 



to a growing Rutaceoiis leaf displayed by the pupa. Each 

 emerges. The Acraea's, fluid-supply has been much de- 

 creased, yet is efficient enough in relation to present enemies, 

 and its flight is sufficient to commonly save it from the 

 driver. The Papilio has flight, and, instead of a pro- 

 tective fluid, has polymorphic mimicry in the female 

 (compensated in the male by better flight and slightly 

 greater nauseousness), and a strong procryptic element in 

 the " matching " of the dulled underside by both sexes 

 in resting. I have already mentioned the native view of 

 nestlings, and I am publishing elsewhere (" Ibis ") the 

 results of some actual experiments with birds' eggs : a 

 stronger procryptic element, represented most usually 

 in the nests, is present here than in the case of butterflies' 

 eggs ; also parental protection ; so that nauseousness is 

 to a varying extent less necessary, though it is often, I 

 believe, in some degree present to complement or replace 

 these other defences. Those plants in which the seedlings 

 are less liked by herbivora than is the adult foliage (pro- 

 tected by height, etc.) afford a somewhat closer parallel, 

 though they trust more than even butterflies to their 

 reproductiveness. It would be very interesting, in view 

 of recent observations that have discredited the view that 

 it is myrmecophilous, to ascertain whether the bull's-horn 

 thorn acacia of America is not one of them. 



Further interesting points in the experiments were : 

 (1) the effect on survival of, apparently, variability in 

 juice production or conservation in individuals of Mylabris, 

 one such individual, a large one but amongst the first to 

 be inserted, remaining protected to the end, while others 

 were taken and one was definitely exhausted of juice and 

 killed; (2) the shearing of the spines of Acraea acara 

 larvae. This was improved on in a subsequent experiment 

 in which the juice was absorbed by the application of 

 earth-crumbs. I have on a few occasions seen Acraea 

 larvae feeding with similar earth-crumbs attached to their 

 bristles, and there can be no doubt that the ants' successes 

 against Acraeine larvae in these experiments were mainly 

 due to my replacing the escaped larvae amongst them 

 time after time ; (3) the behaviour of an Acraea larva 

 meeting successive ants, then not prepared to attack it ; 

 (4) the ants' general variabihty as to the food they would 

 accept, rather surprising and reminding one somewhat 

 of the different stages of hunger in a bird ; (5) the release 



