294 STUDIES IN AUSTRALIAN NEUROPTERA, iv., 



wing, which can be seen in Newman's type. All the species 

 are, however, so exceedingly variable in ground-colour and mark- 

 ings, that it seems almost certain that Newman's type was an 

 extraordinary variation, which may not turn up again for a very 

 long time. As D. humilis McLach., is, at the same time, the 

 commonest and the least variable species, it seems wise to fix 

 this species as the type of the new genus Drepanacra. 



In this genus, variability is not confined to the colouration, 

 but extends also, within certain limits, to the venation of the 

 wings, while the limits of size for any given species are also con- 

 siderable. The task of working out the material which has 

 passed through my hands has been a very laborious one; though, 

 indeed, it would prove delightful enough to the confirmed 

 "splitter," who could describe new species to his heart's content, 

 on the extraordinary variations of colour-scheme. A peculiar 

 difficulty is met with in this and many other Hemerobiid genera, 

 viz., that, after death, the body, inconspicuous enough even in 

 life, shrivels up so completely that it is quite useless for specific 

 diagnosis. As for the colour of the antennje, this varies with 

 age, being palest in freshly emerged specimens. We must 

 have recourse, then, almost entirely to the wings for our specific 

 characters. Even here, we must proceed with great caution, and 

 define our species so as to take account of the extremes of vena- 

 tional variation. 



Two characters that are of the greatest importance generally 

 in the Hemerohiidce, and that have been used for de- 

 fining species in Drepanacra by McLachlan, are found 

 to be variable when any large number of specimens is 

 examined. These are (1) the number of cross- veins 

 in the gradate series, (2) the number of radial sectors 

 in the foi'ewing. In order to explain the reason for 

 ^^ ' ^' ' the variability in these characters, reference should 

 be made to Text-figs. 5 and 6. As regards the outer gradate 



* Diagrams to show inconstancy of number of erossveins in the outer 

 gradate series of Drepaneptery.v : a, usual condition; h, approach of a fork 

 to gradate series; c, interpolation of an extra erossvein in the gradate 

 series, 



^ 



