350 
Psyche 
[December 
case, attempts to separate Sphaeroberotka and Cycloberotha on the 
strength of this single character is made impossible by the present 
heterogeneous composition of Cycloberotha . 
The number of branches of the radial sector is likewise unsuitable 
as a criterion for distinguishing these two genera, since C. neuro- 
punctata and S. dumonti both have an identical pattern of branching 
of the radial sector. It is true that the type specimen of C. mjobergi 
has one fewer branch in the radial sector of the fore wing, but as 
has been recently noted (MacLeod and Adams, 1967) slight dif- 
ferences in the number of branches of the radial sector are probably 
too variable a feature to use as a taxonomic character of generic 
distinction. The presence of squamae on the females of Cycloberotha 
is likewise an insufficient reason for separating these two genera, since 
otherwise closely related species are known in both Berotha and 
Lomamyia which differ in the presence or absence of these squamae 
on their females. 
With the present state of our knowledge concerning the genera 
presumed to be somewhat related to Sphaeroberotha , it is not possible 
to more than guess at what the diagnostic features of the genus might 
be, assuming that it is actually a valid entity. Certainly none of the 
taxonomic characters discussed above is sufficiently free from sus- 
picion to serve as a guide in this direction. As a tentative guess, since 
I have not observed them in the form seen in Sphaeroberotha else- 
where in the family, I might suggest that the sclerotized, “snail-shell” 
specialization of the proximal portion of the copulatory bursa and the 
presumptive glandular elements associated with the distal portion of 
the spermatheca (rather than in their more usual position along the 
fertilization canal) may perhaps prove of diagnostic importance. 
There is a similar sclerotization of this portion of the copulatory 
bursa in Cyrenoberotha of the distantly related Cyrenoberothinae, al- 
though here the sclerotized portion is not coiled and of course there 
are many other differences. Again, in the somewhat more closely re- 
lated Naizema (but not in Trichoberotha) , there is a tendency for a 
weak sclerotization of this portion of the copulatory bursa (MacLeod 
and Adams, 1967). Here again, however, there is no tendency for a 
coiling of this sclerotized portion. Finally, Trichoberotha (to be re- 
described in a subsequent paper of this series) has a superficially 
similar, snail-like coil in the sperm-conducting pathway; however, 
here it is the spermatheca itself, rather than a portion of the copulatory 
bursa, which shows the apparent similarity. There are, in addition, 
