390 ON THE GENUS CORDULEPHYA, 



variations of the type-species itself. The locality (Melbourne') of 

 the types is unfortunately at the extreme southern end of the 

 range of the species. Hence the measurements given by de Selys 

 disclose the fact that practically every other specimen from more 

 northern localities is considerably larger than the type-specimens. 

 More than this, variations in venational characters, used to define 

 the genus, are not only due to difference in locality, but are an 

 essential characteristic of the species; so that a long series, taken 

 from any given locality, can be arranged to show a gradation from 

 one extreme to the othei-. The chief variation lies in the form of 

 the triangle of both wings. De Selys defines them as follows : — 

 "Le triangle discoidal des ailes superieures irregulier, le cote 

 superieur brise, ce 'c6te formant un angle obtus dirige vers la 

 cote;" and, as he says nothing about the hindwing-triangle, it is, 

 of course, to be assumed that the definition "regular" applies to 

 it in this as in all other genera of the subfamily. An examina- 

 tion of a long series of C. pygin(ea, however, discloses the fact 

 that a regular triangle in the hindwing is rather the exception 

 than the rule. In the majority of specimens, the superior (costal) 

 side of this triangle is distinctly broken near the distal angle 

 (Plate xii., figs. 3 and 4), and, in a few specimens, the break is 

 even more than one-third of the whole length of the side from 

 the distal angle. Another variable character is the position of 

 the arculus-sectors at their base. De Selys defines them in the 

 words (for the hindwing only) "I'arculusdont les secteurs naissent 

 separes." This is scarcely the case. In the forewing they are 

 frequently just separated, and as frequently just joined. In 

 the hindwing, however, they are usually just joined, and, in 

 some specimens, joined for a perceptible distance. This may be 

 clearly seen in the photograph of the wings given in Martin's 

 "Cordulines,"* though he does not comment upon it. [It should 

 be here noted that the drawing of the wings of C. pygmcea, given 

 by Needham in his " Critical Notes on the Classification of the 



*Coll. Zool. du baron Edni. deSelys-Longchaiiips, Fasc. xvii., Cordidinei, 

 Bruxelles, 1906. 



