to limitations of the Milllcrian Hfpothesis of Mimicry. 119 



under-side colouring which prevails in that section of the 

 genus to which /. haliensis belongs, renders it very im- 

 probable that the Ixiafi has in any way influenced the 

 Huphina. Furthermore, so far as the $ sex is concerned, 

 the increase of the black border in the hind-wing of 

 H. nerissa corva is not particularly striking, and there are 

 two females in the British Museum which in this respect 

 are scarcely to be distinguished from a $ nerissa from 

 Sikkim. It is in the $ that the increase is specially 

 noticeable, and this sex can certainly not be reasonably 

 regarded as a mimic of /. haliensis $, for the entire absence 

 of the heavy black bar across the cell of the fore- wing 

 gives it a very diiferent appearance. 



Another serious difficulty, to which Dr. Dixey has him- 

 self referred, is the discrepancy in the geographical range 

 of the two species. To meet this it is suggested that 

 either H. corva has extended its range, or /. haliensis has 

 contracted its range, subsequently to the production of the 

 mimicry. The facts do not appear to justify either of 

 these assumptions. Dealing first with the Huphina, we 

 find that the continental form nerissa (with its var.^j/ir^/wc) 

 is represented in the Andamans by a slightly darker form, 

 lichcnoscc, leading up to the broad -bordered race coo^va 

 (including the scarcely separable sumatrana) which is 

 found throughout Sumatra, Java and up to the islands 

 of Bali and Lombok. On the other hand, so far as we 

 know at present, the range of Ixias reimvarcUii lies 

 entirely to the east of this, namely, in the islands of 

 Timor, Flores, Sumba, Sumbawa, Lombok, up to Bali, 

 which constitutes its westernmost limit. 



If we assume H. co7'va to be a mimetic modification of 

 nerissa which was evolved in the island of Bali and has 

 since extended thence, we must suppose that originally 

 typical nerissa occurred throughout Sumatra, Java and 

 Bali. If this were so, why should corva have so totally 

 replaced nerissa in Java and Sumatra, where Ixias 

 haliensis does not occur, and how are we to explain 

 the existence of an intermediate form in the Andamans ? 

 On the other hand, there is even more difficulty in accept- 

 ing the supposition that I. reimvardtii haliensis, which is 

 confined to a little island some 80 by 50 miles in extent, 

 should have originally ranged throughout Java and 

 Sumatra (with a combined length of some 1,700 miles), 

 from which it lias now completely disappeared. For the 



