432 Professor E. 15. Poulton on 



and aJhimarulnta are so excessively abundant that his 

 native boy would get tired of catching tliem, whereas he 

 would secure every specimen of doniinicanus that came in 

 his way. But on the other hand, as Mr. G. A. K, Marshall 

 and Mr. Swynuerton have both pointed out to me, the 

 females of dardanm probably spend a great deal of their 

 time actually in the forest, and within that particular 

 environment there would be far less discrepancy between 

 the numbers of dominkanus and the two other species of 

 Amauris, than would appear from the above tabular state- 

 ment. Nevertheless, it has been already shown that echeria 

 and alhimaculata are exceedingly effective models for 

 dardanus in Natal, and the same is true of the south 

 coast of Cape Colony. Making every allowance therefore 

 for the above considerations, I am driven to conclude that 

 dominicanus possesses some special advantages as a model 

 over the other two species, which may compensate for a 

 numerical inferiority. These advantages may be conferred 

 by the far greater conspicuousncss which renders it visible 

 at a much greater distance than either Inhcngida or alhi- 

 macidata* It must be remembered furthermore that, 

 although so much less numerous than the other two, 

 dominicinms is quite a common butterfly at Chirinda, 

 whereas in Natal it is generally rare, and often altogether 

 unseen for long periods in many localities where Pcqoilio 

 dardanus is found. 



It is interesting to compare Papilio echerioidcs, Trim., 

 with P. dardanus. The former is also common at Chirinda, 

 39 males and 17 females having been so far counted. The 

 females, as is well known, are beautiful mimics on the 

 upper surface of Amanris echeria (or lobcwjiila) and albi- 

 tnaculata. For this Fapilio the latter are evidently very 

 effective models, but for dardanus, with another form, 

 mimicking the black and white Danaiues, their influence 

 in spite of preponderant numbers is entirely subordinate. 



The relative rarity of the trojdtonius form in the west 

 and east, as well as the south of Africa, in spite of the 

 general abundance and wide range of its model, has often 

 been remarked upon (see p. 430; also Trans. Ent. Soc. Lend., 

 1904, p. 688;. In this case the facts are probably to be ex- 

 plained by difference of habitat, chrysippus being an open 



* It M'ould also be very interesting to investigate the relative 

 efficiency of the .special means oi' protection possessed by these three 

 Danalnix. 



