96 Annals Entomological Society of America [Vol. VIII, 
The adult characters of the groups thus created need not be 
discussed in detail at this time; suffice it to say that in this light 
also the groups give the impression of naturalness. But even 
their superficial consideration forces upon us another fact, 
and that is that we have in these groups two stocks which have 
become dissociated in the remote past and are developing along 
independent lines. For the evolutional picture presented by 
the adults is just the reverse of the larval one. The group with 
the more primitive (peripneustic) larve shows far the greatest 
specialization in the adults. This finds its most tangible 
expression in the reduction of the wing venation—its stronger 
development in the costal region together with a well developed 
costal vein, and a weakening of the posterior veins—as we see 
it in the Scatopside, Simuliide, Chironomide and Cecidomyi- 
de. In this last group this line of development reaches its 
culmination, not only in the greatly reduced wing veins, but 
also in the reduction of the number of tarsal joints (Hetero- 
pezine). In the larve, too, we find other evidences of extreme 
specialization along independent lines. As is well known, the 
larve of the Cecidomyide have no distinctly chitinized head, 
there having been a reduction, as in the higher flies; but the 
persistence of the lateral spiracles, as well as the adult char- 
acters, demonstrates the close relationship with the others of 
the series. Thus we can safely say that in the Cecidomyide 
we have the evolutionary apex of the peripneustic stock. 
In the metapneustic series we find the adults exhibiting a 
far more primitive type. The wing-veins are more numerous, 
more evenly distributed over the wing-surface and all of nearly 
equal strength throughout. This condition alone makes it 
impossible to deduce the metapneustic forms from the existing 
peripneustic ones. They must have had their origin in peri- 
pneustic stock long since extinct, in which the adults likewise 
were still in an archaic state. 
It is thus evident that the time-honored group of Nemocera 
must be discarded as an unnatural one. The multiarticulate 
antenne of the imago, which have been relied upon as of pri- 
mary value, and the phylogenetic significance of which in a 
broad way can not be ignored, must be discarded as a group 
character. The two groups here indicated may be designated 
by already established terms, originally applied in a much more 
limited sense, Oligoneura and Polyneura. This seems far pre- 
