OF WASHINGTON. 99 
The Cynipidce through the Ibaliince and Oryssidce connect with the 
Tenthredinidcen, and a natural sequence of the other families follows, as is 
shown in the arrangement of the families above given. 
The Braconidce seem to me to be more closely allied to certain saw-flies, 
and in consequence are placed above the Ichneumonidce. 
Both of these families need revision, and no eifort is made to arrange 
the subfamilies. 
The Ichneumonidce connect naturally then through the Trigonalidce and 
Stephanidce with the Evaniidce, and the latter lead naturally into the 
Chalcididce through the genus Hyptia and the Eucharid genus Lophyro- 
cera Cameron, a species of which I have taken in Florida and named in 
MS. L. floridana. 
The sequence of the subfamilies in the Chalcididce as arranged above 
seems to be a very natural one, so gradually do they merge the one into the 
other. 
The Eucharince, Perilampince and Eurytomince are very closely allied, 
and the last through Aximince. form a very close connection with the CJial- 
cidince, Torymince and the following groups. 
I am by no means satisfied that Axima is entitled to subfamily rank ; it 
seems to me to belong to the Eurytomid group. 
The Blastophagce, or fig insects, Francis Walker placed in the Procto- 
trupidce ; according to Sir Sidney Saunders, they form a section in the 
Cynipidce ; while Prof. J. O. Westwood says they belong to the Chalcididce. 
At present engaged in preparing a monograph of the N. A. Cynipidse, 
I was verv desirous of satisfying myself on this point, and fortunately have 
been able to do so. from a study of possibly an undescribed species, now in 
the U. S. Department of Agriculture, recently collected by Mr. E. A. 
Schwarz, on Ficus aurea in South Florida. 
The species studied evidently belongs to Saunders' genus Kradibia, 
which Dr. Mayr, in his " Feigen-insecten," says is identical with Blasto- 
phaga Grav. ; at any rate it is no Cynips, and I agree with Prof Westwood 
in considering the Blastophagce as a group in the Chalcididce. 
They seem most closely allied to the Spalangince and a position is as- 
signed them next to that subfamily. 
In conclusion, I beg to say that the views herein set forth are based upon 
a tolerably close study of most of the families mentioned, all being known 
to me in nature but two Mellinidce and Ampulicidce. 
In discussing this paper Mr. Howard stated that it appeared to 
him that Mr. Ashmead's arrangement was, in the main, a one- 
character classification, and that no natural classification should 
be attempted without a thorough review of many characters. 
The ovipositor is an important organ, and its variations should 
have much weight, but an arrangement of the families according 
to such variations will clear up but a single factor in the problem. 
