Mi in' He Relationships in the genus Heliconius. 125 



spot at the base of the hind-wing, and elevatus has an 

 exactly similar spot, though, of course, this feature is only 

 of value in conjunction with the anatomical similarity, 

 as so many of the non-silvaniform Heliconii have basal 

 red spots. 



These two specie-, tumatumari and elevatus, are of 

 extraordinary interest as showing that there is no marked 

 line of division between the Silvaniformes and the other 

 members of Section I. The hind-wing pattern of elevatus 

 raises another very interesting point. The " flame 

 pattern" so frequently found in Heliconius is of two 

 kinds. In melpomene forms (Section I) it is always cut 

 across in a nearly straight line and the rays are of the 

 " nail-headed " type, whereas in the species of Section II 

 in which it occurs, the rays at their inner ends follow the 

 contour of the cell. Why should the pattern of H. elevatus 

 be of the nail-headed type ? The reply seems to me to 

 be that the characteristic horizontal black bar in the hind- 

 wing of species of Section I is an ancestral pattern of 

 considerable stability. It appears in one form or another 

 in narcaea, silvana, numatus, ethilla, novatus, pardalinus, 

 quitalenus, paracusis, aulicus, forms of anderida, etc., and 

 its inner (upper) edge runs across at the level of the cell 

 end. Hence when a " flame " pattern is developed it is 

 cut off straight along the top by this characteristic bar. 

 Thus, if my suggestion be well founded, we should ex- 

 pect any species of Section I which developed a flame 

 pattern to produce the straight-cut, more or less nail- 

 headed type found in melpomene and elevatus, irrespective 

 of fascies of more recent ancestry. The nearest approach 

 to a flame pattern in any truly silvanil'orm-patterned 

 species that 1 know is in an example before me which 

 agrees with Weymer's H. fortunatus. Here the usual 

 black band is very distinct, and beneath it is a series of 

 orange internervular marks of the ground-colour which, 

 whilst running off to fine points marginally, are cut off 

 proximally by the black band and have the nail-head 

 pattern, though greatly foreshortened. 



Professor Poulton has suggested to me that the black 

 bar may have been developed in Section I of Heliconius in 

 mimicry of Melinaea forms, doubtless an association "I 

 greal antiquity. The idea has much to recommend it. 

 since if it were a character of still greater antiquity we 

 should expeel it to be common to both sections of the 



