BY R. J. TILLYARD. 783 



order to understand their position, it must be emphasised that 

 they have closest affinity witli those groups in which the cross- 

 venation is limited to gradate series, and Sc is not united with 

 R in a true anastomosis (i.«., by actual fusion of the two main 

 veins). Thus they show affinity, on the one hand, with the 

 Hemerobiidce (s.str.) and Dilaridce, but remain more primitive 

 than either of these, on account of the single radial sector of the 

 forewing; while, on the other hand, they are related to the Tri- 

 chomatidxfi^ Apochrysidce, and Chrysopidcf. The 7richoviatidce, 

 like the Psychopsidce, appear to be a direct offshoot of the old 

 Prohemerobiidcp; but their tendency has been towards reduction 

 and narrowing of the wings; whereas that of the Psychoj^sidm 

 has proceeded in the opposite direction. In the Apochrysidce, 

 we recognise a group of insects with certain cliaracters of the 

 wings strongly suggestive of Psychopsid affinities, viz., the widen- 

 ing of the costal area and the extreme differentiation of the disc. 

 Bearing in mind, however, the slender build and elongated 

 antennae, and their evident close relationship with the Chryso- 

 pidce, I am inclined to regard these venational characters as the 

 result of convergence, rather than of true relationship. Both 

 the ApochrysidcE and the Chrysojndce differ radically from the 

 Psychopsidce in the manner of development of the branches of 

 Rs, which tend to cross the wing transversely, and have thus 

 brought about new tendencies in the development of the gradate 

 series, and finally the unique formation of the pseudomedia and 

 pseudocubitus. 



The larval form of Psychopsis is such that, from it, there can 

 be equally easily derived the small-jawed series of larva? found 

 in the Hemerobiidce, Chrysopidw, etc., or the long-jawed series of 

 the Bei'othidce and Osmylidce, or the large-jawed series of the 

 Nymphidce, Myrmeleoiitidce, etc. In the last series, there has 

 been a gradual addition of internal teeth to the mandible, 

 beginning with one only in Nymphidce, and passing up to three 

 or more in MyrmeleoididcB. Hence we shall probably be on safe 

 ground, if we regard the larval type found in Psychopsia as 

 closely similar to that of the extinct Prohemerobiidce. 



From the Berothidct and Osmylidce, the Psychopsidce are at 



