HISTORY INFERRED FROM MIMICRY 155 



by Limenitis (Basihrchia) arcMppus was evidence 

 that the model had long resided in North 

 America, and that we might on this ground alone, 

 even if we had not abundant positive evidence of 

 its gradually increasing spread in the Old World 

 during the past half-century, infer that Anosia 

 had reached Fiji, Australia, Hong-Kong, &c., in 

 comparatively recent times. This conclusion can 

 hardly be doubted, and the argument might have 

 been extended to enable us to infer the ancestral 

 line of migration by which North America itself 

 had been reached by this form. But in 1897 I 

 followed what appeared to be the general view, 

 that, in the New World, the original stream of 

 Danaine invasion had run from the American 

 tropics northward,^ nor did I observe that the 

 evidence based on the growth of mimetic resem- 

 blance warranted the interesting conclusion that 

 its flow had taken the opposite direction, and 

 that the south had been peopled by way of the 

 north. Accepting this conclusion the question 

 arises : Whence came the Danaini of North 

 America? The answer requires a somewhat 

 careful comparison between the New and Old 

 World butterflies of this group. 



Among the commonest of the Old World 

 Danaini, are certain species with tawny colouring, 

 a black border, and black white-barred apex to 

 the fore wing. The under surface is even more 



' Verhandl. d. V. Internal. Zool. Congr. z. Berlin, 1901, Jena, 

 1902, 171. See also Essays on Evolution (1908), 274: also errata. 



