28 
This classification looks very nice on paper, but when we under- 
take to apply it to the specimens themselves, then the difficulty arises. 
In the first -place, in many species the macrochaetae are so reduced in 
size that it is no easy matter to decide as to whether they should be 
considered as macrochaetae or simply as bristly hairs. Then again, the 
males of several species have five visible abdominal segments, while 
their females have only four: and, conversely, the females of a few 
species have live and their males only four. This difference is due to 
the greater or less development of what, in those with only four abdom- 
inal segments, is the first segment of tin 4 genitalia, and as this varies 
in size in the different species, it is not always easy to decide as to 
whether to regard it as belonging to tin 4 genitalia or as forming a dis- 
tinct abdominal segment. In at least one genus (Hemyda), the fourth 
abdominal segment is so reduced in size as to appear as a part of the 
genitalia, and thus there arc apparently only three abdominal segments. 
These differences, therefore, are developmental rather than structural. 
If there were differences in habits between these subfamilies there 
would then be some reason for retaining these divisions, but the differ- 
ences are so slight that they hardly enter into this consideration. 
Thus the Tachininae are known to attack the Bymenoptera, Lepidoptera, 
I >iptera. < Joleoptera, and Orthoptera, but not the Hemiptera ; the Phasi- 
mc and ( >cypterinae attack the Coleoptera, Orthoptera, and Hemiptera : 
the Gymnosominae attack the Hemiptera only, and the Phaninae theCole- 
optera. It will thus be seen that all of these subfamilies, with the excep- 
tion of the Gymnosominae, attack Coleoptera; that all except the 
Tachininae and Phaninae prey upon the Hemiptera; while the Tachininae 
is the only one known to attack the Hymenoptera, Lepidoptera, and 
Diptera. In point of numbers the Tachininae outnumber all of the 
other subfamilies put together by at least fifteen species to one. and to 
this fact alone is apparently due the greater diversity in regard 
to their hosts. 
The most recent attempt at a classification of the Tachinidae of the 
world is by limner and Bergenstamm, of Vienna, Austria. These 
authors, who are new workers in this field, had access to the types of 
most of the species described byMeigen, Wiedemann, Schiner, Kondani, 
Jaenuicke, and several by Macquart, and their figures and redescrip- 
tions of many of the species can not but prove to be great aids to future 
students of this group. In the first part of their work 1 they threw 
together in one mass the live families: CEstridae, Tachinidae, Dexidae, 
Sarcophagidae, and Muscidae, and out of this chaotic mass they erected 
fifty-five families, which were duly given family names: but this classifi- 
cation did not prove satisfactory to the authors, so in the third or hoi 
part of their systematic work they again threw all the families into one 
and divided it into sixty groups, which they call sections. 
'Published in the Denkschriften tier Mathematisoh-Naturwissenchaftlichen ('lasso 
der Kaiserlichen Akademie der Wissenschaften for L889. 
