( 552 ) 



as well as those species of Eulfpis and Charaxes which are, in this respect, the less 

 specialized ones, there are as many costal ridires or serrations as there are rows 

 of scales between the costal nervnre and the costal false vein. Bnt in other species 

 of Charaxes and Eulepis, for instance in C/t. polijxena and allies, there are less 

 costal ridges than tliere are rows of scales in front of the costal nervnre, as about 

 one-fifth of the nnmber of rows of scales do not reach the thickened edge of the 

 wing (PL XIII. f. ;")) in the middle of the costal margin. This further specialization 

 is found to a much higher degree in a number of I'huraxes, where of every two to 

 three rows of scales only one extends to the costal edge, and this row is placed there 

 in a deep groove (as it were), the hinder edge of which is very strongly raised, 

 forming a sharp ridge pointing distad (Pi. XIII. f. 1). 



On the underside the specialization is still more remarkable. The less 

 specialized form of the costal edge occurs, as above, in Palla, Euxantlu', Eulepis, 

 and a nnmber of smaller Charaxes, nearly every row of scales reaching the edge 

 of the wing : the serration aj)pears, however, stronger than above (PI. XIII. f. 4). 

 The more specialized Charaxes have tlie number of rows of scales tliat reach the 

 edge less, the sockets are less distinctly impressed on the false vein, tiie serrations 

 are stronger and farther apart : while in the extreme forms only oue out of every 

 two to four rows of scale-sockets reaches the edge of the wing, and this row is 

 very feebly or not at all marked upon the false vein (PI. XIII. f. 2). The less the 

 nnmber of costal ridges is, the higher they are. 



It is apparent from PI. XIII. f. 1.2, which show few scale-sockets on the false 

 vein, and these feeldy impressed, that the scaling of tlie costal edge of sncli a 

 butterfly must be different from that of the butterfly represented on PI. XIII. f. 8. 9, 

 where the rows of scales extend to the wing-edge having the sockets well imj>ressed. 

 And we find, indeed, a farther specialization in Chiiraxes awl allies. In all the 

 genera under consideration, Charaxes, Eulepis, Eiixnnthe, and /'alia, the scales at 

 the costal edge are rather loose, falling off easily, and the edge of the wing is, 

 therefore, more or less bare of scales : in species with the edge so much specialized 

 as shown in f. I. 2 (PI. XIII.), there are only a few long scales present close to the 

 costal teeth, while the false vein itself is naked, a cliaracter that is very remarkable. 

 As the scales are more easily rubbed off from a stiff part of the wing, like the costal 

 edge of Charaxes and allies, than from a soft jjart tliat gives way when touched, 

 the nakedness of the costal edge might be attributed to accidents to individual 

 specimens. But that is not the true explanation ; for, on examiuatimi of the wing 

 of a freshly emerged Ch. jasn»,I find tliat (lie costal edge is very sjiarsely scaled 

 below, and even on the edge of a wing examined before the specimen had emerged 

 from the chrysalis the scaling is sparse, though there are more scales than in 

 specimens that have been at large. 



The loss of scales in the costal region of the forewing below is in some 

 Charaxes (cutxliope for instance) still more obvious in consequence of the 

 obliteration of every second row of scales l)efore the costal nervnre, the remaining 

 rows giving a peculiar aspect to that part of the wing. Intermediate between the 

 serrate wing of Charaxes and close allies and the normal Nymphalid wing stands 

 that of Vrothoe ralednnia, in which the costal edge is not serrate, bnt more or less 

 denuded of scaling. 



The gradation from the wing of J'allu, Charaxes Jaljiua, Eulepis pi/rrhus, etc., 

 to that of Charaxes camliope, tiridates, etc., is not complete, there being a break 

 ill the series, iiiasmnch as there is no complete transition from the edge on whicli 



