Mimetic Relationships in the genus Heliconius. 121 



In each of these cases, as in that of " silvana " ethra and 

 narcaea flavomaculatus, the forms which resemble one 

 another are separable on the anatomical structure, and the 

 resemblance may be regarded as mimetic without stopping 

 to define the particular class of mimicry to which each 

 case should be referred. 



Bearing in mind the above facts, we may now consider 

 the extent to which the patterns support the conclusions 

 formed on the anatomy. 



The narcaea association. 



The typical form of this species appears to be very 

 distinct from the other reputed species with which I have 

 associated it, and it cannot be said that pattern affords 

 much support for the conclusions based on the anatomical 

 structure. Other members of the association, however, 

 offer strong support to these conclusions. Comparison of 

 silvana silvana with the typical form of numatus shows 

 that there is really little difference between their patterns. 

 H. gradatus thielei is scarcely separable from forms of 

 numatus on pattern. As to ethilla, pattern would cer- 

 tainly support the view of a close affinity with numatus, 

 especially if the undersides of numatus numatus and 

 ethilla <t<comus be compared. H. ethilla claudia, which 

 resembles anderida melicerta, stands out rather distinctly 

 from the rest. As we have seen, the claspers of forms of 

 clJiilla vary greatly, but in ethilla claudia they so closely 

 resemble those of ethilla ethilla that there can be no doubt 

 of their specific identity. RifTarth's diagnosis of this 

 case was extremely shrewd. With reference to the other 

 reputed species, the claspers of which I have not been able 

 to examine, H. hippola must remain very doubtful, though 

 an example of narcaea satis with the apical spot suppressed 

 and the discal band darkened to the ground-colour would 

 be difficult to distinguish from hippola. It is difficult en 

 mere outward examination to appreciate the grounds on 

 which sulphureus has been separated from el Jul I a. 



The novatus association. 



The pattern exhibited by the reputed species here 

 associated are extremely diverse, though some of the 

 special features may be traced through several different 

 forms. 



