500 Mr. R. M¢Lachlan on a new Species of Corydalis. 
dilated at the base, the apex slender, and curved downwards and 
slightly inwards; inferior pair rather shorter than the superior, sub- 
cylindrical, obtuse, curved strongly inwards. 
I obtained a pair of this magnificent insect from the late M. 
Deyrolle. They are from Brazil, but I am unable to indicate any 
more exact locality. The female exceeds in size any described 
species of the genus. Distinct from all previously described species, 
with short mandibles in the male. C. nubila, Erichson (Schom- 
burgk’s Reisen in Brit. Guiana, vol. iii.), perhaps somewhat resem- 
bles it, but is very much smaller, and has the antenne black for 
two-thirds of their length. Five undescribed species from South 
America are noticed, by name only, in the Appendix to Hagen’s 
‘Synopsis of the North American Neuroptera’ (p. 321); but of these 
I possess no information. 
In all branches of entomology we are constantly reminded of the 
uncertainty that attends our attempts to fix generic characters on 
some striking peculiarity; for experience frequently proves that the 
character which at first seemed most forcibly generic, is in reality 
only specific, and demonstrates the impossibility of fixing any general 
laws to govern generic division. The genus Corydalis is in this case. 
Formed by Latreille for the reception of the Raphidia cornuta of 
Linneus, the enormous cornuted mandibles of the male seemed to 
point emphatically to that character as generic ; but further materials 
have shown that it is merely specific: numerous gradations occur in 
species, which otherwise so precisely agree that they could not rea- 
sonably be generically differentiated. 
Corydalis belongs to the limited family Sialde, differing from 
Sialis in the possession of ocelli, in the structure of the tarsi, and in 
the position of the wings in repose, but closely allied to Chauliodes, 
in which genus the antenne of the males are pectinated or serrated, 
and the prothorax is broader. Whether the genus Hermes of G. R. 
Gray can justly retain its position as distinct from Chauliodes, is a 
matter of opinion; the structure of the antennse appears to present 
most undoubted gradations *. 
* Chloroperla prasina, Newman (Zoologist, vol. iii. p. 853, 2), from New Zea- 
land, has been wrongly placed in Hermes by Walker (Cat. Brit. Mus. Neurop. 
pt. ii. p. 206, 10). It undoubtedly pertains to the Perléde, in which family it 
was originally located. 
