( 554 ) 



the costal margin agaiust foreigu objects liail Ird to increased vigour of this part 

 of tliL' wing, and oonspqncntly to a thickening of the originally normal edge; but 

 this exiilauation would leave out of account the development of serration. On the 

 other baud, it might be advanced that the scaling offered a protection to the costal 

 edge, and that tliis protecting scaling becoming rubbed away and the edge exposed 

 to danger, those varietal specimens hence being the better provided in which the 

 edge was thicker and which had another kind of protection, namely the ridges or 

 serrations, selection had set in and led to the present stage of development. But 

 this would merely mean pushing the origin of the character back to the stage wheu 

 it was an individual varietal character for which Natural Selection does not give an 

 explanation. To understand the meaning of the serration it will be necessary to 

 refer to other wings with serrate costa: the wing of the Pierid genera Prioneris, 

 Beknois, ('allidnjas, in which the tnale has the costal margin of the forewiug 

 dentate, of the Papilios allied to phaeton of which both sexes have a serrate wing, 

 will perhaps help to elucidate thi' biological siguiticance of the serrations and 

 their causes, a question I shall have to enter upon more fully in another place. 



The wings of Uharaxes and allies exhibit a second structure which, although 

 not peculiar to these genera, is of importance as a striking and easily discernible 

 distinguishing character. In nearly all Lepidoptera the scaling of that portion of 

 the underside of the forewiug which is covered by the hindwing is markedly 

 different from the scaling of the disc, as every Lepidopterist knows : the scaling 

 near the base being, moreover, different from that of the outer three-fourths of the 

 internal marginal area. In y>jtnphfiU(lae we find always a basal patch of peculiarly 

 modified scales at the internal margin, variable in length and width and of a silky 

 appearance, consisting of triangular scales that stand more or less erect, or, at least, 

 are not lying flat as those on the disc, and are shorter and narrower than the normal 

 scales. Such a triangular scale is represented by f. 13 (PI. XIII.), showing the 

 peculiar character of the striation ; while f. 14 (PI. XIII.) rejjrcsents a more 

 normal, but non-dentate, scale from near the basal patch. The corresponding patch 

 at the base of the costal edge of the hindwing is well known. The patch of the 

 forewing extends in Symphnlkhie from the internal margin either to the submedian 

 nervure SM-', or beyond that vein to the submedian fold (SM'j. The difference in 

 the extent of the patch is easily recognizable with the naked eye on account of the 

 difference in the gloss of the triangular, half-ereet, and the more normal scales. In 

 all Eulepis, Ckara^-es, and Eu.canthe (and many other Si/mphalidae) the patch 

 reaches to (SM'), as shown in f 11 (PI. XIII.), while in Palla and Protho:- {a,\\<\. 

 other Symphalidae) the patch stops at SM-, as shown in i. Vi. (PI. XIII.). The 

 sexes are alike in the development of the patch. What the function of the patch is 

 I do not know. It is not a stridulating organ, as vein SM" of the forewing protrudes 

 too much, but it is possible that the scales are the covering of small glands which 

 produce a lluid of specific smell serving as a guide to the sexes of the same species 

 in recognizing each other. For the individual Imtterfiy on emerging from the pupa 

 cannot be supposed to have a knowledge of the distinguishing characters in pattern 

 of the species to which it belongs, and to ac(|nire this knowledge by means of the 

 eyes is certainly a pretty difficult task, considering that the individual looks at its 

 own wings under such a very small angle that it cannot recognize the exact iw-^ition. 

 size, and outline of the wing-markings; besides, in the case of sexually dichromatic 

 and dimorphic species the knowledge of its own wing-pattern and outline would not 

 help the male to find the cospecific female. It is certainly not conceivable that 



