L. Harrison 
125 
fourth segment. In view of the enormous development of the tracheal 
trunks in connection with a change of habit such as we have not 
previously had to take into account, this is best looked upon as a 
correlated adaptation. The presence of combs upon femora and 
abdomen shows that the affinities of the genus lie with Col-pocejihalvm. 
rather than with Menojwn, with which it was for so long associated; 
but the genus may remain in the Menoponidae. 
Turning to the Ischnocera, we find the stigmata are uniform 
throughout the group, and the only difference of any importance is the 
occurrence of the posterior commissure in Nesiotimis and some species 
of Tricliodectes. 
Kellogg (1903) has already pointed to the fact that Nesiotinus 
possesses both Ischnoceran and Amblyceran characters. The additional 
evidence afforded by the primitive tracheal arrangement indicates 
that it must certainly be looked upon as a very primitive Ischnoceran 
form; and also that it must, as I have previously suggested (1915, 
p. 384), form the type genus of a family Nesiotinidae. 
The Trichodectidae require further examination. Stobbe (1913, 
1913a) has recently re^dsed the family, and established two new genera, 
but on the same somewhat superficial grounds that have proved such 
a stumbling-block to a proper understanding of Mallophaga in the past. 
From the variations I have noticed in some twenty species which I have 
had under my eyes, variations which include the presence or absence 
of a vestigial segment in front of the abdomen, and of a posterior 
commissure; the possession of three, four, or five antennal joints; 
the presence, in one species at least, of only five pairs of abdominal 
stigmata; and the very variable structure of the hind-abdomen and 
d genitalia, I am convinced that this apparently homogeneous group 
is capable of being split up upon sound lines, and that a comparative 
study of the tracheal and stigmata! arrangement will considerably 
assist such division. Such a task must, however, be left to some worker 
with a far more considerable material at his disposal than is available 
to me. 
The remaining six families of the Ischnocera (Lipeuridae, Docophor- 
oididae, Goniodidae, Philopteridae, Giebeliidae, Akidoproctidae) show 
such uniformity in their tracheal systems that nothing of help in 
taxonomy can be gained from a study of the respiratory apparatus. 
