284 STUDIES IN AUSTRALIAN NEUROPTERA, iv., 



Thus, N. Banks, in 1909, dealing with our Australian Hemero- 

 biid(e{i), included both Sisyra and Psychojjsis in that family. 

 But the same author, in 1913(2), in a more general paper, includes, 

 in the Hemerobiidce, four separate subfamilies, viz., Dilarince., 

 Psychopsince, Osmylince, and Hemerohiince,, without offering any 

 diagnosis of characters common to the four. It would, indeed, 

 be difficult to indicate any. Sisyra is here put into Osmylincf, 

 to which it more properly belongs, but the same subfamily is 

 made to include the whole of the Nymphidce, Myiodactyhis, Poly- 

 stcechotes, and the exceedingly archaic and isolated Ithone 1 On 

 the other hand, Handlirsch, in 1908(6), recognised the Sisyridcf, 

 Polystoechof.idce, Dilaridce, Nytnphidce, and OsmylidcB as separate 

 families, retaining, in the Hemerohiidce, only thi'ee subfamilies, 

 viz., Berothince, Psychopsince, and Hemerobiince. 



There can be little doubt that Handlirsch has most nearly 

 achieved the task of cleansing the Augean stable, though he is 

 certainly wrong in including Berotha with the Hemerobiidce, 

 since it is more closely allied to the Osmylidce, as Banks supposed. 

 I have already given my reasons for separating out the Psychojy- 

 sincE as a very distinct, archaic family. What, then, are the 

 characters by which the family Hemerobiidce, when reduced to 

 its correct limits, may be known ? Nowhere have I been able 

 to find any satisfactory diagnosis— chiefly because, as we have 

 already seen, the family has always been so cumbered with ex- 

 traneous elements that a diagnosis was impossible. 



I think that the family Hemerobiidce may be very clearly defined 

 by the combination of the following characters (Text-figs. 2-4) : — 



(1). I'he presence of more than one radial sector in the foretving. 

 This character appears to me to be of the utmost importance, 

 since it separates out the Hem,erobiidce at once from all the rest 

 oj" the Order, except the Ithonidce and Dilaridce. 



(2). The absence of unsj^ecicdised cross-veins. In the Hemero- 

 biidce, the few cross-veins left are all put to some important use, 

 either by forming gradate series (cf. Chrysopidce) or, in single 

 cases, as special supports or junctions for longitudinal veins. 

 Unspecialised cross-veins are present in the Ithonidce and 

 Dilaridce. 



