( 470) 



qnantitatively much slighter; hence depilis aud novohibernicus must (on the ground 

 of external features) be considered as nearer related to one another than to the other 

 euchenor fw'ms. As in tlic present paper it is onr purpose to demonstrate the 

 variation of the genital armature within the limits of a species, and as therefore we 

 have to avoid, as far as possible, any error as to the actual specific identity of the 

 forms included l)y us in tlie limits of a resjiective species, we will, merely for the 

 sake of being on the safe side, assume that depilix is a species distinct from euchenor. 

 The external characters by which novohibeniicus is distinguished from depilis are 

 found in every specimen of our short series ; if, therefore, the constant presence of 

 a distinguishing cliaracter is considered sufficient to make the respective fonn 

 specifically distinct (as some naturalists do), norohibernicus is also a distinct species- 

 However, the distinguishing characters of nocohibernicus mnoxmi quantitatively not to 

 more— about the qualitative amount of these characters we know notlung — than the 

 differences do whicli are observed between New Guinea individuals, differences 

 in the extent of the yellow markings which are not tliought to indicate any- 

 thing else but individual variability within the same species ; consequently we 

 must assume that the differences between tlie individuals from New Britain and 

 those from New Ireland being quantitatively the same are also qualitatively the 

 same, i.e. do not indicate more than divergency of individuals of the same sjiecics. 

 Hence the characters distinguishing depilis and novo/iibernicus must correctly be 

 considered as" not being of specific value. The various forms in question are 

 therefore to be grouped as follows : — 



{euchenor ; New Guinea and islands near it. 

 (/odarti ; Woodlark Island. 

 ob.^olescens ; Aru and Kei Islands. 

 r, , -J. ( depili.i ; New Britain. 



' \novokibernicus ; New Ireland. 



The genital armature of the five forms is in accordance witli this division. 



The vahe of ]'. euchenor (f. 51) is very large, strongly convex outwardly, witli 

 the apical margin rounded, the ventri-apical angle not being triangularly produced as 

 in P. aecjeus and most other species ; it exhibits some individual variability in the 

 outline, especially in the ventri-ajiica! portion. The armature of the valve consists 

 of a fohl running along the ventral margin of the valve in a slightly oblique 

 direction, turning near the apex round towards the dorsal edge of the valve, running 

 from here as a thin fold backwards to the base, fir.st in a directly basal, then in a 

 dorso-ventral direction, and thus returning to the starting-point ; from the basal 

 dorso-ventral portion an oblique fold (f .51, tf) starts, traverses the (concave) 

 valve, and widening out joins the ventral longitudinal ridge. The ventral portion is 

 raised into a ridge, armed at both ends witli a process, of which the first (c) is here 

 called " basal hook," tlie second (/;) " ventri-apical hook "; the basal ridge leans over 

 ventrally, so that in a view perpendicular to the valve the dorsal snrfiic'e of the ridge 

 is visible (as in f. 51); it is higliest near the basal hook (f. 52); the outline is 

 not constant. The ventri-apical hook is directed in a basi-apical direction leaning 

 over ventrally, with the tip jirotruding above the elevated edge of the valve. The 

 vertical, ventri-dorsal, portion of ihc fold is less high than the ventral ridge, tliinuer 

 and denticulate ; it leans over ajiically, so tliat iu a perpendicular view the Imsal 

 surface of the ridge is visible ; at the dorsal end it is produced into a slender 

 and very sharp hook ('/), the " dorsal hook," wliicli is curved in an apici-ventral 



