PHVLOGENY OF THE NEMOCERA. 



257 



I have now completed my analysis of the groups suggested by 

 Williston from the antennal characters, so far as my imperfect 

 data will permit, and I think that it shows that they furnish 

 important hints as regards the phylogeny, particularly of t he- 

 older flies, though the light they throw is intermittent and 

 flickering. From their exposed position and frail structure the 

 antennae are subject, in the swiftest flying insects, to more than 

 the usual risk of injury, which is possibly one reason for the 

 extreme variation they show, ranging from three joints in the 

 Scenopinidae to thirty-nine in Cerozodia. 



TABLE V. 



STRUCTURE OF EYES IN NEMOCERA. 



1 Chironomus. 



2 Exceptions exist. 



To recapitulate, I think that group I, with the inclusion of 

 the Simulidae. is a strong natural group, and even stronger tribe. 

 Groups 3 and 4 fail altogether. Group 5 is fairly strong, but 

 suggests the inclusion of the Stratiomyidae, Xylophaginae, Taba- 

 nidae and Leptinae, Lonchopteridae and Phoridae. As regards 

 the Acanthomeridae I have nothing to suggest. 



