( xi ) 



species. One was compelled to believe that every perfect 

 resemblance began as an imperfect resemblance, and then 

 passed through stages in which the likeness was gradually 

 increased ; and it was only to be expected that examples of 

 all such stages should exist at the present day among the 

 numberless forms which exhibited mimicry and common 

 warning colours. 



Gradual changes in the geographical distribution of the 

 constituent species along the borders of groups would tend 

 from time to time to bring certain of them within the 

 influence of other groups, and so begin a change in another 

 direction. Furthermore, there was no reason for concluding 

 that the detached members of Milllerian groups must become 

 extinct, as in the case of the Batesian or true mimic. In the 

 presence of other dominant members of a group, any ten- 

 dency towards resemblance might well be of selection value : 

 in their absence it was by no means necessary to assume that 

 a species, which, ex hypothesi, was specially protected in some 

 way, must become extinct, although any further advance 

 towards the likeness would be checked, and the ground gained 

 in the past lost after a longer or shorter interval. To enter 

 the area of another group would tend towards a rapid modifi- 

 cation of the old appearance. 



The suggestion that parallel variation assisted in the for- 

 mation of these resemblances was strongly opposed by the 

 fact that the superficial characters were alone affected, and 

 that the closeness of resemblance bore no relation to degree 

 of affinity. For instance, the resemblance between a 

 Melhura and a Heliconius was frequently much closer than 

 that between the former and a Mechnnitis, or the latter and 

 an Eueides, m the same groups. Although the interesting 

 facts brought forward by Mr. Blandford with regard to Heli- 

 conius thelxiope and H. vesta, etc., would, taken alone, seem 

 to support his suggestion of parallel variation, yet when they 

 were considered as part of the whole phenomena of Miillerian 

 warning colours, as exhibited and as at present known and 

 understood, in the Neotropical Ehopalocera, one was led to 

 believe that one set of principles had been at work, and that 

 natural selection, which, he contended, offered the only hopeful 



