250 
Psyche 
[September 
about which the inner core of the spermatophore is secreted. How- 
ever this remains to be verified. 
On several occasions females of Lomamyia have been collected by 
us which carried spermatophores in various stages of disintegration, 
indicating that the fate of these structures in nature closely parallels 
what we have observed in the laboratory. Tjeder (1954, 1959) also 
studied field-collected females carrying fragments of spermatophores 
in two species of A croberotha in which the structure and destruction 
of the spermatophore seems to be about the same as in Lomamyia. He 
did not, unfortunately, realize that the hard core was an integral 
part of the spermatophore and concluded that it was a permanent 
component of the female genital segments in the genus A croberotha 
to which he applied the term “ovipilum”. 
When females have removed or lost the last traces of their sperma- 
tophores, the very extensive space of the confluent bursa and genital 
chamber is effectively closed to the outside by the close approximation 
of various sclerites, with members of different groups of berothids 
achieving this result by somewhat different mechanisms. The details 
of these differences have been noted in the descriptions of the sub- 
families which we have given below. 
The classification of the berothidae 
The taxonomy of the Berothidae is still in a rather rudimentary 
state and the data provided in the present paper can do little more 
than to suggest what the principal groups and adaptive trends might 
be. The only world-wide treatments at the species level have been 
those of Kriiger (1922) and Navas (1929). Both of these studies 
suffer from a heavy reliance on relatively minor details of the wing 
venation and wing shape, with the resulting generic and suprageneric 
classification bearing little resemblance to phylogenetic reality. In 
addition, with the exception of Nosy bus, all of the species treated 
in these revisions are relatively specialized forms belonging to the 
Berothinae, so that little insight into the overall evolution of the 
family has been gained by their study. 
In what is the best account of this family to date, Tjeder (1959) 
recognized three subfamilies : Rhachiberothinae, containing mantispid- 
like forms with raptorial forelegs ; Berothinae, containing species with 
elongate, frequently falcate wings and hypocaudae on the 9th gono- 
coxites; and “Sphaeroberothinae” (i.e. Nosybinae, see below) , with 
short, rounded wings and lacking hypocaudae. In spite of the limita- 
tions imposed on him by an inadequate literature and the availability 
