Mimetic Relationships in the genus Heliconius. 133 



in hydarus chestertoni,cyrbia , s fore-wing red being obtained 

 from hydarus hydarus. The peculiar white-dusted hind- 

 wing border in cyrbia is more difficult to account for. It 

 would at first suggest affinity with sapho, but the structure 

 of the armature precludes this. The form is, of course, 

 either a mimic or a model of wlcanus cythera. 



5. amphitrite is merely a development of hydarus, the 

 armatures being practically identical. 



6. erato appears in many forms, some of which have 

 developed the flame pattern common in other species. 

 All kinds of intermediates are known, for which some 

 forty-three names are recorded by Seitz. Most of these 

 forms resemble others belonging to Section I, of which 

 they are either the models or mimics. Its variability is, 

 perhaps, even greater than that of melpomene, and there is 

 nothing improbable in supposing that the forms I have 

 associated with it are really the same species. 



H. hermathena. 



Closely related to the erato association, but apparently 

 distinct. A very remarkable species combining the fore- 

 wing spots of erato pkyllis with the hind-wing of charithonia. 



H. charithonia. 



The true relationship of this common form is rather 

 obscure. It is apparently a good species. 



H. nattereri, H. fruhstorferi. 

 Of these I have had no examples to examine. There 

 is, as already stated, some probability that they are male 

 and female of the same species, though whether distinct 

 or merely forms of charithonia must remain for the present 

 undecided. 



H. clysonimus, H. hortense, H. telesiphe. 



These three forms are probably distinct species and call 

 for no special comment, except that clysonimus is probably 

 related to antiochus through forms of sara, in which traces 

 of the hind-wing red band can still be seen. 



We thus see that examination of the patterns and colours 

 of the various forms dealt with supports in most cases the 

 relationships suggested by the study of the genital arma- 

 tures. We may, therefore, draw up the following table 

 embodying the results of the foregoing investigation. 

 Forms apparently not specifically distinguished are placed 



