Comhinations of Tropical American Butterjiies. 603 



The whole problem, however, can only at present be one 

 of surmise, owing to the enormous amount of work still to 

 be done in these areas, and more especially in the 

 intervening districts ; for until far more data have been 

 accumulated than at present, we can only indirectly infer 

 that certain members of the associations are dominant as 

 compared with others : and it is impossible to feel much 

 confidence in the selection of any single pattern as the 

 ancestral type which has given rise to those of adjacent 

 areas. These questions must i-emain open until further 

 labours have thrown far more light upon this fascinating 

 subject. 



Certain Mullerian Combinations among the 



DANAIN.E OF the OlD WoRLD. 



The accompanying Plate XXXIV exhibits members of 

 three small associations from Southern India, Fiji and the 

 Solomon Islands respectively. The names and localities 

 are indicated on the plate itself, and, in greater detail, in 

 the explanation of plate. The colours and patterns are 

 those characteristic of an important Eastern section of the 

 Danciinx — the Euplceini. In the two first-named localities, 

 certain species of this section are seen to resemble one 

 another: in the third locality a species (Fig. 10) of the 

 other important section of the Danainm — the Dnnaini — 

 has assumed the superficial appearance of an Euploeine 

 (Fig. 5). It is unnecessary to speak in any detail of the 

 associations represented on Plate XXXIV : they are only 

 introduced on the present occasion for the striking compari- 

 son which they afford with the New World Combinations 

 exhibited on the four preceding plates. I may, however, 

 remark upon the interesting example of Fijian mimicry 

 in Figs. 4 and 9. It is here obvious, as Professor Poulton 

 pointed out to me, that the chief spot in the fore-wing of 

 Deragena proserpina (Fig. 9) has been lengthened inwards 

 so as to afford a superficial resemblance to the chief spot 

 of its model Nijmra eleutho (Fig. 4). Although the two 

 chief elements of the pattern in these two Euploeines have 

 thus attained a considerable degree of resemblance, it is 

 certain that they belong to a different series of white 

 markings, — sub-marginal in the mimic (Fig. 9), discal in 

 the model (Fig. 4). 



In these Old World Combinations no Jthomiina} lead 



