744 LT.-COL. N. MANDEES ON THE 



of June, the broods become larger in numbers and more frequent, 

 and this goes on until the end of August or beginning of 

 September, when there is a further spell of dry weather similai' to 

 but not so pronounced as that in the early part of the year, when 

 the broods again become smaller and less frequent, but at the same 

 time produce the individuals which take part in the migratory 

 flights of the monsoon in November and December. 



Strictly speaking, P. hector and P. aristolochice, though following 

 the above sequence of events, do not usually form part of the 

 flights, but they are nevertheless at their maximum at this time ; 

 the Eupteas and Polytes undoubtedly do so. We have now to 

 judge what influence the inexperienced young birds off the nest 

 in May, and theii' experiments concluded in October, can haA^e on 

 these species. A butterfly the size of Ewploca core pairs during 

 its first flight, if' we may judge by the cabinet condition of those 

 ovipositing, and begins to lay its eggs three days afterwards. The 

 usual number is about two hundred and fifty, which are deposited 

 according to the weather in about ten days (I have known one 

 hundred eggs laid in five days). What becomes of the parent 

 after this? Whether she dies a natural death or becomes the 

 victim of a tasting experiment is immaterial, her time of danger 

 is a brief fortnight. As the females are less in evidence than the 

 males, fewer of them would be captured, especially if we agree with 

 Professor Poulton's opinion that the Miilleriantheoiy presuppo.ses 

 that only young birds test the palatability of a few members of 

 each convergent group in their locality. To bring to such per- 

 fection the cases of mimiciy I have selected, we must assume that 

 such a victim would be one having less converging characteristics 

 than the others ; and it must also be borne in mind that unless 

 she is killed within three days of her emergence, she will have laid 

 a certain number of eggs which will produce buttei'flies similar to 

 herself. It is difficult to understand how the broods of butterflies, 

 numberine- some thousands of individuals, born between Octobei' 

 and the following nesting season, would be in any Avay affected 

 except in the very smallest manner. No doubt Nature is infinitely 

 slow in her methods, and we have no reason to suppose that these 

 cases of mimicry have been produced, otherwise than by a very 

 lengthy process of weeding out ; but even if we giunt this, there 

 is a still greater difliculty in the case of Hypolimnas misippus, the 

 well known mimic of Danais chri/sijipics. In Cejdon the formei- 

 appears on the wing in October, when as I have said tasting 

 experiments are over. It remains on the wing until the end of 

 the year, when it disappears until the following aufumn. There 

 are so fa.r as I can see only two ways of getting over this diffi- 

 culty — either by assuming that the inherited tendency to produce 

 this form of mimicry has become so fixed that the withdrawal of 

 the factor that produced it is immaterial, though there is no reason 

 for this supposition, or that there is a more or less constant influx 

 of the species from India. There is very little doubt that a certain 

 number of Ceylon buttei^flies in their annual migratory flights 



