1907] 
MITCHELL --CULICID SUBFAMILY DEIXOCEIUTIXAE 
18 
Now in I), cancer the antennae of both sexes are nnusnally long and hear rela- 
tively short and few hairs; moreover the antennae of tlie male have neither longer 
nor more numerous hairs than those of the female. In all other Cnlicidae the 
antennae are coin])aratively short and with more and relatively longer hairs than 
in 1). cancer; and except in the Tricho])roso])ininae, (which have other distinctive 
subfamily characters), the male antennae are densely plumo.se, with much longer 
and much more numerous hairs than in the female. Therefore according to ( )sten 
Sacken’s theory, Deinoerrifes is the most primitive. 
By the above antennal characters the C'ulicidae may be divided into three 
sharply marked primary groups as follows: 
Antennae of males with much longer and more numerous hairs than those of the 
females. Antennae of females with long hairs. Second antennal joint in both 
se.xes le.ss than 8 times as long as wide. Metanotum nude. Outer caudal angle 
of mandible of larva not thumblike, or fle.xed, never with long spines or visible 
from above. Mandibular groove not angulated. Ventral brush ])rc,sent. 
Contains the subfamilies Anophelinae, Megarhininae, Psorophorinae, Culicinae, 
Uranotaeninae. 
Antennae of males not more plumose than those of females. 
Antennae relatively short, the hairs of the antennae of both sexes well developed, 
second antennal joint less than 8 times as long as wide. iNIetanotum bristly. 
Mandible of larva as in above .section. Mandibular groove sometimes angu- 
lated, (in L. hinala, according to Knab). Ventral brush absent. 
Subfamily Trichoprosoponinae. 
Antennae long in both sexes, the second joint over 14 times as long as wide, the 
antennal hairs in both sexes short and sparse. Metanotum nude. Outer 
caudal angle of the larval mandible visible from above, thumblike, flexed, 
.somewhat movable and with many long spines. IMandibnlar groove forming 
a distinct angle below ba.se of antennae. Ventral brush present. 
.Subfamily Deinoceritinae. 
Thus it would seem that the conclusions at which I arrived by a study of the 
larvae are borne out by the important antennal characters in the adults and that the 
subfamily Deinoceritinae is at lea.st as well marked as any other. 
