OF WASHINGTON. 69 
The antennce are very important organs in the classification of 
the group. They vary in number of joints from seven to thirteen, 
never exceeding the latter number. The bulb is not well separated 
from the scape. The scape is long, usually the longest joint, and 
slender; usually sub-cylindrical, sometimes distinctly fusiform, 
occasionally with a leaf-like ventral expansion (some species of 
Encyrtus and Aphycus}, or with a flattened tip (Melitto bia}, or 
flattened out and rolled from side to side ( Cerapterocerus] . Its 
insertion on the face varies considerably. \Vith the Pirenincz it 
is inserted near the border of the mouth, but usually it arises 
from a point nearer the middle of the face. The distance apart 
of the two bulbs also varies somewhat. The pedicel is usually 
obconical with its apex somewhat curved. It is not inserted 
directly upon the apex of the scape but upon an articulating facet 
ventrad of its distal end. Following the pedicel come, in many 
instances, two ring-joints. These reach their maximum of de- 
velopment in the Pteromalince and in some other subfamilies are 
not perceptible. Following the ring-joints come from one to six 
funicle- joints. These present much diversity of form. Their 
simplest form is that of a long, naked cylinder, as in the female 
Psilophrys. The joints of this form vary all the way from ten 
times as long as wide to much wider than long. They may present 
a serrate appearance collectively, as in Habrolepis ; they may 
become gibbous dorsally, as in the male Eurytoma, or each 
may give off a long branch, as with Eulophus. They may also be 
flattened out so that one face is several times wider than long, as 
in Mira and Anusia. One discouraging fact to the student of 
collected specimens is that the antennae of the two sexes of the 
same species may differ more widely from each other in form than 
do those of different subfamilies. This is markedly so in all En- 
cyrtid genera. The female antennae in this subfamily are usually 
regularly sub-clavate and naked, while those of the male are 
usually linear, with each joint strongly constricted above and 
furnished with one or more whorls of strong hairs curved at the 
tip. A similar difference may be seen with some of the Euryto- 
mince. The number of joints is quite constant in genera, and 
also with the two sexes. Excluding the apparent exceptions to 
this latter rule, in which the club is divisible in the one sex and 
not in the other, the only absolute exception which I have 
