STRUCTURE. 
551 
are known to hibernate as adults. Not much is known about the past 
history of Diptera, but from the study of insects preserved in amber it 
appears that the oldest forms are among the Nemocera, and that the 
Schizophora are of much more recent development. Now if we com¬ 
pare these two groups we shall see that development has largely taken 
the forms of reduction, shortening and simplification of parts, rather 
than the production of new ones. This does not apply to the ptilinum, 
which is an organ not possessed by any of the Nemocera, but it does apply 
especially to the antennae, palpi, and venation. Fig. 351 represents 
two antennae of Nemocera, two of Brachycera, and one of Aschiza and 
Schizophora. The possible course of development is clearly seen where¬ 
by a nemocerous antenna might in course of time be modified into the 
compact three-jointed Muscid pattern (/). 
So also with the maxillary palpi (labial palpi are very rarely 
if ever present) : those of the Nemocera generally have four joints, the 
Brachycera two or one, the Aschiza and Schizophora always one only. 
The question of venation is more complex (p. 553), but on com¬ 
paring the wings of the Nemocera, Brachycera, Aschiza, and Schizo¬ 
phora, it is clear that the main characteristic of the Schizophora is the 
comparatively few veins in the hinder part of the wing and the closing 
up of the cells near the base (basals and anal). In the Nemocera the 
hind part of the wing has several veins, and the anal cell is large and 
open. In the Brachycera the anal cell shows signs of diminution, and 
may be either narrowed at the margin or closed. In the Aschiza it is 
always closed, though often large, while in the Schizophora it is always 
closed and usually quite small. In this connection it should be pointed 
out that though what has just been said is true on the whole, yet indi¬ 
cations of this line of development by reduction of wing-veins can also 
be traced in each of the four groups taken singly, more particularly 
among some of the smaller Nemocera and Brachycera ; such are the 
Chironomidce, some Bihionidce and especially Simulium in the former 
group, and in the latter some Stratiomyidce, Cyrtidce, Dolichopodidce, 
and especially Phoridce. The same tendency towards simplification is 
met with in the larvae. Speaking in a general way one may say that the 
oldest families of flies are probably those which have aquatic or semi- 
aquatic larva) with distinct head and eyes. Such larvae are the rule 
among the Nemocera. Bibionid and Mycetophilid larvae, though not 
