1967] 
MacLeod — Berothidae 
349 
one can, I feel, maintain strong reservations about the importance of 
this set of alternatives as a taxonomic character. The degree of de- 
velopment of the basal piece of MA is subject to a good deal of 
variation, largely because of the weakening of its sclerotization which 
results from the line of flexion passing through it. This difficulty in 
detecting the vein is compounded when the vein is located in a 
relatively more proximal position between the approaching bases of 
R and M. * * * 4 As a result of these considerations, I believe that it would 
be premature to assume that Costachillea cannot be closely related 
(or perhaps congeneric) with Sphaeroberotha until, at a minimum, 
specimens of Costachillea have been examined to see if perhaps a 
poorly sclerotized basal piece of MA is not present after all which 
was missed by Navas. 
The final genus of berothines with rounded wing apices which 
should be considered is Cycloberotha. The characters which have 
seemed to separate this genus from Sphaeroberotha are the number of 
radial crossveins, the number of branches of the radial sector, and 
the presence or absence of squamae on the females. With respect to 
the number of radial crossveins, Navas (1929, 1930) 'has stated that 
Cycloberotha is characterized by the possession of three. Although 
this is true for the type species, C. ?njobergi (Esben-Petersen), Navas 
has erred in evaluating this feature in the second species of this genus, 
C. neuropunctata (Esben-Petersen), where only two radial cross- 
veins are present. Esben-Petersen (1917) states this quite clearly in 
his original description of this species and I have verified this fact 
from the examination of a series of specimens of this species in my 
collection. The number of radial crossveins is frequently rather con- 
stant within groups of closely allied species and it may be that C. 
neuropunctata has been incorrectly assigned to Cycloberotha. Prece- 
dent, however, exists in both Berotha and in the rather well-studied 
Lomamyia for including in the same genus species with differing num- 
bers of radial crossveins so that C. neuropunctata should not be 
excluded from Cycloberotha on the basis of this feature alone. In any 
Dilaridae is relatively distal in position and is slightly sinuate as it is in 
the fore wings of many berothids. Unfortunately, comparisons with the 
closest relatives of the Berothidae, the Mantispidae, cannot be drawn, since 
even in the most generalized of the Platymantispinae, the bases of R and M 
of the fore wing are too closely adjacent to permit study. 
4 In evaluating Navas’ descriptions an additional, essentially semantic 
problem is added to these sources of confusion: a vein is only an “internal 
radial vein” if it is located proximal to the fork of MP. Where the vein 
is located distal to this fork it no longer qualifies as an “internal radial vein” 
and the taxon is described as lacking the vein (cf. Navas, 1929, p. 23). 
