( 284 ) 



The neuration exhibits some variation iu several points. The length of cell of the 

 forewing is not constant in the genus, three being species in which the distance from 

 the lower angle of the cell to the base is greater than that to the end of W (castor, 

 etc.) ; while in other Charaxes the reverse is the case {Ch. polyxena, for instance). 

 Vein R^ of the forewing varies in position specifically and individnally, joining 

 R' sometimes near the point of origin of M', sometimes 5 or more mm. distally of it. 

 M- of the forewing stands generally about midway between M' and the base of M, 

 but in many species it is considerably nearer M'; The praecostal veinlet of the 

 hindwing is often two-branched {raranes, candiope, etc.), sending a more or less 

 long spur basad, whereas in most species it is simply curved distad, being, however, 

 thickened at the bent as a rule. The distance between SC^ R' and W of the hind- 

 wing is also not constant in the genus, and varies even individually ; on the whole, 

 Ri is midway between SC- and D^ As explained in vol. V., p. 557. D' of the 

 hindwing reaches M either at the point of origin of M' (or near it), or is more 

 proximal ; there is some individual variability in this respect, but the more obvious 

 differences are found in the sexes or in different species. 



WhOe in Eulepis the sexes do not differ very greatly, except in epigenes from 

 the Solomon Islands, sexual dimorphism is among Charaxes a much commoner 

 phenomenon than similarity of the sexes in colour and shape. The wings of the 

 ? are always broader than those of the S, and the hindwing is, as a rule, less 

 triangular ; the tails are broader, often widened at the end ; in some species the <J 

 has one, the ? two tails, in other the tails of the S are obliterated, while the ? 

 possesses such an appendage. 



The pattern of the upperside is often very different in the sexes and in the 

 various species. The ? ? can be arranged according to the pattern of the upperside 

 of the forewing into four groups : 



(1) The discal and jiostdiscal interstices form a forked band, which is generally 

 more or less interrupted at tiie veins into spots or patches. This is the normal 

 type of ? . There are many species in which the c? has a similar band, but in this 

 sex the patches composing the band are smaller than in the ? . 



(2) The upper discal interstitial patches become small or obsolete, while the 

 postdiscal ones form together with the posterior discal ones an oblit^ue band which 

 extends from the costal to the internal margins. This type is found in both sexes 

 of a number of species, and in the S 3 only of some others. 



(3) Tlie upper postdiscal interstitial spots are reduced in size or absent, while 

 the discal interspaces and the posterior postdiscal ones form a broad curved band. 

 This type does not occur in the t?-sex. 



(4) The band is similar in appearance to that of type (3), but it is composed of 

 the posterior postdiscal and discal interstices and the upper median ones, the band 

 often entering the cell. This type is also confined to the ?-sex, and is met with 

 only in one polychromatic species 1 , etheocles, of which other specimens belong to 



type (1). 



It is worthy of note that though types (2) to (4) are doubtless derivations 

 from type (1), the middle and upper discal interstitial patches are enlarged in type 

 (3) and reduced in type (2), while the reverse is the case in regard to the upper and 

 middle postdiscal spots. The same elements iu the pattern of the wing follow here, 

 in the same genus, opposite directions of development, a phenomenon which we 

 meet with commonly among Lepidoptera. It is obvious that the divergency of the 

 lines of development of such closely allied species is not explained by assuming that 



