290 



STUDIES IN AUSTRALIAN NEUROPIERA, iv., 



The genera, Rapisma Siud Oliarces, induded in the H einerohiidce 

 by N. Banks(2), must be removed from that family as defined 

 above, on account of the presence of only one radial sector, and 

 numerous, unspecialised cross-veins. I think that a new family 

 will be required for these two genera. They ai^e not found in 

 Austialia. 



I have not attempted to divide the Uemerobiidfe into sub- 

 families, because it seems to me that a single phyletic line of 

 descent is clearly indicated, with all the intermediate stages still 

 present, from the comparatively large, densely-veined, and most 

 generalised form {Drejmnejjteryx) right down to the smallest 

 forms (Sympher'obius, Notiohidla, etc.) in which the venation is 

 comparatively simple, and in which all the signs of a liigh 

 specialisation by reduction are evident. Thus, if we attempt to 

 separate Drepanp,pteryx and its allies oif on a very important 

 venational character (as I had hoped to do), viz., the presence 

 of the original Cuj in the hindwing, we shall make an unnatural 

 grouping: for a new genus ( Drepanomina) has just come to light, 

 which is most certainly a close ally of Drepanepteryx, but lacks 

 this important vein. Nor does the falcate form of wing justify 

 us in elevating this same group into a subfamily, since, in other 

 respects, Drepanoniina, Megalomus, and Megalomiita are very 

 closely allied. Again, while we can pass in a descending scale 

 (by reduction) from Drepanepteryx to Drepanacra, thence to 

 Drejyauoinina, and thence directly to the pointed-winged forms 

 Megalomina and Oxybiella, the connection with the smaller, 

 round-winged forms is supplied by Micromns, which is clearly a 

 specialisation from Megalomina (loss of recuirent costal vein by 

 narrowing of costa), and in a somewhat different direction by 

 Psychobiella (fusion of the two basal radial sectors of forewing 

 into one). Thus we arrive, at last, at a form with only three 

 radial sectors in the forewing. The final reduction to two radial 

 sectors is actually accomplished, in the Paltearctic region, within 

 the range of the type-genus Hemerobms itself; while, in Aus- 

 tralia, the line of reduction passes on from Psychobiella to Motio- 

 biella, with Carobius as a side-branch. 



In the actual practice of determining genera of HemerobiidcH 



