feY R. J. TILLTARb. 287 



conspicuous in all of them. Tt is also present in all Australian 

 genera: though, in extremely small and reduced forms like 

 Carobius and Notiobiella (Plate xvi.), the jugal process is either 

 absent or only represented by a frenulum. 



That the apparatus is of very archaic origin is shown, not only 

 by its being most highly developed in the oldest and most 

 densely-veined forms of Hemerobiida', but also because it is pre- 

 sent and well formed in Ithone, which stands very close to the 

 ancestral stem of those in.sects. 



(5). The absence of any distal fusion betiveeii So and R. This is 

 the character relied upon by N. Banks. Unfortunately, in some 

 of the most reduced genera (e.g., Carobius), Sc and R are so close 

 as to be practically fused throughout their length. Also, in the 

 Chrysopid(e, there is no fusion between Sc and R distally. How- 

 ever, from a phvlogenetic point of view, the character is import- 

 ant, since it shows us a point in which the Osmylido' and 

 Fsychopsidre have undoubtedly progressed beyond the Hemero- 

 biidce. In the Chrysopida', Sc runs into the pterostigma well 

 above R, and ends weakly there. This appears to be most 

 certainly a specialisation from an original Hemerobiid-like con- 

 dition of Sc and R. Lastly, in Ithone, there is the same condi- 

 tion of Sc as in the Hemerobiidce., but Sc is deeply sunk under 

 the high ridge of R. 



(6) The archaic, unspecialised form of Rs. This character is 

 shared with most other families of Neuroptera, but serves to 

 enable us to distinguish the Hemerobiidoi from the Chrysopidce, 

 in which Rs is zigzagged. 



(7). The retention of the archaic branchings of the veins as they 

 approach the ivincj-horder. In the Hemerobiidce, these branchings 

 are present, and usually numerous, for all veins from the ptero- 

 stigma outwards to the apex, and round along the posterior 

 border to the base. In the costal space of the hindwing, from 

 base to pterostigma, the cross-veins are regular and unbranched. 

 In the forewing, however, these same cross-veins are elongated 

 and usually branched. Only in those genera, in which the costal 

 space of the forewing is not enlarged, do we find any consider- 

 able number of these veins unbranched. 



