( vii ) 



in want of further support from observation and experi- 

 ment, which would afford a large if arduous field of 

 work to the enterprising naturalist ; the difficulties of the 

 subject did not appear to him to be fully overcome by these 

 theories, which should not be pushed so far as to lead to the 

 disregard of other factors which might have influenced the 

 genesis of these groups. One difficulty was that of distribu- 

 tion. As before mentioned, the groups of the Upper Amazon 

 valley were often of limited range ; but if they were geneti- 

 cally connected and the conditions of their environment were 

 constant, the causes which brought about association under 

 a common type, if prohibiting deviation therefrom under 

 penalty of destruction, should have operated to extend the 

 limits of a group as widely as possible by acting as a check 

 to variation on its outskirts. If it were assumed that one 

 form were so far dominant as to drag its associates with it in 

 any given direction, it must be also recollected that the prin- 

 ciple on which these large " Miillerian " associations were 

 supposed to be based was prohibitive to variation of any 

 component species, either as a whole, or in any part of the 

 range of the system. The logical tendency of such a group 

 would be to extend its limits indefinitely and not to give rise 

 to repeated changes of the colour-type. 



Another difficulty was presented by the very close resem- 

 blance, at times amounting to identity in external characters, 

 between certain pairs in these groups, a resemblance to 

 which Brnnner's epithet, " hypertelic " might be applied. 

 Existing theories postulated a selective elimination by insecti- 

 vorous birds, etc. ; but the birds' discrimination of members of 

 a " Miillerian," i.e., protected group must be in relation to 

 its distinctness from the other insects co-existing in the same 

 region. If, as was frequently the case, such a group was 

 immediately recognisable by its broad features as protected 

 and inedible, such further discrimination between its members 

 as would be necessary to bring about the intimate likeness 

 found, e.ff., between many species of Melinaa and Heliconius 

 was not adequately accounted for by Mi'iller's hypothesis. 



With regard to " convergence," which had been put forward 

 as a necessary phenomenon in Miillerian mimicry, the 



