126 Dr. H. Eltringham on Specific and 



genus. However that may be, the bar is now a deeply 

 established factor of 'the pattern, -and sufficiently accounts, 

 I think, for the special characteristics of the flame pattern 

 as developed in Section I. 



The melpomene association. 



Some ten reputed species are here included, totalling, 

 with subspecies and varieties, sixty or seventy named 

 forms. On PI. XI I have figured twenty-six butterflies 

 which exhibit some of the forms included in this associa- 

 tion. All the figures are much reduced from natural 

 size, but for convenience of comparison it was desirable 

 to have them all on one plate. The forms of cydno exhibit 

 great diversity of pattern. White and lemon-yellow are 

 easily interchangeable in both wings. The broad yellow 

 fore-wing discal band of cydno is divided into two separate 

 bands in the form shown at fig. 3, and in fig. 4 there is 

 a submarginal row of white spots, the other fore-wing 

 markings having become white. On the hind-wing of 

 cydno there is a white border of medium width, almost 

 marginal. In fig. 3 it is twice as broad, and in fig. 5 

 it is broad and yellow. In hermogenes (fig. 4) it is yellow 

 and considerably separated from the margin, whilst in 

 fig. 7 it is vestigial, but providing a link with the peculiar 

 hind-wing marginal pattern in vulcanus cythera, fig. 8. 

 Fig. 2 represents the underside of fig. 1 . Note the peculiar 

 pattern of the hind- wing. Two red-brown bands are faintly 

 visible and the white sub-marginal border of the upperside 

 is repeated. In some cases the upper red-brown band is 

 practically horizontal, broad proximally and tapering 

 towards the hind-margin. It is important to bear these 

 variations in mind when comparing the patterns of the 

 other reputed species. 



Fig. 14 represents weymeri, Stgr., which is separated by 

 Riffarth, but regarded by Seitz as conspecific with cydno. 

 The latter view is undoubtedly correct. The fore- wing 

 pattern is only a slight modification of fig. 3, whilst the 

 hind-wing central yellow band is probably a development 

 from the red-brown band of the underside of cydno. This 

 central yellow band is a common feature in the genus ; 

 it occurs in both sections, and I shall have to refer to it 

 again. 



Fig. 15 is the form known as weymeri gustavi, Stgr. 

 It has lost the fore-wing markings altogether and become 



