BS R. J. TILLYARD. 615 



only resolve itself into a .study of the comparative morphology 

 of the 'imaginal venation, and would scarcely be productive of 

 any new evidence concerning the Archetype, or the Phylogeny oi 

 the Order as a whole; though it would, no doubt, help consider- 

 ably in elucidating the phylogeny of the various groups and 

 families within the Order itself. 



Jn constructing the Archetype of this Order, the question of 

 outstanding importance is the determination of the correct limits 

 of veins M and Cu. I have dealt with this under Section iv. 

 The archaic Rhyphus (Text-figs. 48, G8b) may be taken as the 

 test genus, although the same result will be obtained from any 

 other genus in which the macrotrichia are sufficiently well pre- 

 served, as in many of the older Tipulidae, Leptidae, etc. The 

 result of a study of Rhyphus, or any of these latter forms, on 

 the lines indicated in Section iv., is to prove conclusively that 

 Cui is a simple, unbranched vein, as in the Mecoptera and Para- 

 trichoptera, and that M 1—< is four-branched, the vein M being 

 the one which Comstoek (15, fig. 357 et seq.) has labelled Cui, 

 while the trae Cui is the vein which he has labelled Cu 2 . The 

 true cross-vein m-cu is devoid of macrotrichia, and connects the 

 two veins M. and Cu^ just distad from the origin of the former 

 in Rhyphus. The vein labelled m-cu by Comstoek is the basal 

 piece of M 4 as it arises from the median cell, and carries strong 

 macrotrichia in Rhyphus and other archaic genera. The differ- 

 ence between the condition of Cu x in the Diptera, and the con- 

 dition of the same vein in archaic Trichoptera and Lepidoptera, 

 where it carries a strong distal fork, has not been sufficiently 

 appreciated hitherto, and is most strongly emphasised here. 



In working back to the archetypic venation of this Order, it 

 is at once apparent that we may leave out of account not only 

 all the Cyclorrhapha, which are highly specialised, but also 

 the numerous groups, both in the Nemocera and the Brachycera, 

 in which the venation shows definite reduction, or specialisation 

 by fusion of the distal branches of the main veins. We are 

 thus left with the Tipulidae (s. lat.) and Psychodidae anlongst 

 the Nemocera, the Rhyphidae, Leptidae (s. lat.), Stratiomyiidae, 

 Therevidae, Tabanidae, A*ilidae and allies amongst the Brachy- 

 cera . 



Tlie Psychodidae are evidently a very archaic group, but spe- 

 cialised by reduction, and by the loss of almost all the original 

 connections between the main veins. At the base of the wings 



