48G Prof. E. B. roulton on the Mimetic N. American 



range of the models ;* The latter seems to be a far more 

 probable hypothesis ; for we have a good parallel example 

 in lorquini itself, where the mimetic influence has been 

 shown to lessen as the mimic passes northwards out of 

 the range of its model. At the same time it mnst be 

 remembered that species of Limcnitis (s. 1.) with an Adeljyha- 

 like pattern exist in the Oriental Region ; and it is a 

 possible hypothesis that these species and califarnica re- 

 present the ancestor of the Neotropical Adelphns. The 

 cream-colour of the band of californica is at any rate 

 an ancestral feature, unconnected with mimicry of the 

 Adelphas. 



We may hope with some confidence that this fascinating 

 but dirticult problem may receive a final solution when 

 the structural relationships of all the species of Limcnitis 

 and Adelpha in their broadest sense shall have been made 

 out in detail. In the meantime, as I have said above, 

 the origin of nearly the whole upper surface pattern of 

 ccdifoi'wica as the result of the spreading northward of an 

 influence exerted by the genus Adclj^ha in the south, 

 appears to be the safest provisional conclusion to adopt. 

 The following facts seem to support it and at the same 

 time to suggest that reciprocal mimicry of loQ^quini has 

 also taken place. 



The numlser of species of Limenitis in N. America points 

 to an ancient existence in this portion of the Holarctic 

 Belt. At the same time their niimetic relationships in 

 the eastern section of the Continent have been shown to 

 be extremely recent. The change in lorquini as it passes 

 north of the range of its model {californica), suggests that 

 its mimicry is also extremely recent. That the North 

 American Limcnitis are liighly susceptible to mimetic in- 

 fluence is shown in the fact that they contribute such a 

 high proportion of the butterfly mimics of the Continent, 

 and that they furthermore produce the most divergent 

 mimetic patterns. In the eastern section of the Continent 

 this recent development of mimicry has been shown to 

 correspond to a recent invasion of Danaine models and 

 to the influence of Papilios — themselves mimics of recent 

 date — exhibiting most beautiful examples of transition in 

 the stages of developing mimicry. It is therefore ex- 

 tremely probable that the very recent mimetic pattern 

 of lorquini — far more imperfect than that of any other 

 mimicking Limenitia in America — also corresponds to the 



