158 
Psyche 
[Sept. 
Withycombe placed the Dilaridae and Berothidae, to- 
gether with the Mantispidae, Osmylidae, Sisyridae, and 
Myiodactylidae, in a superfamily called the Osmyloidea. 
With the exception of the Myiodactylidae, I believe that 
this grouping is entirely correct. Larvae of the latter 
have been illustrated and described by Tillyard (1926, 
fig. U17), and, on the basis of larvae, it seems clear that 
the family is related to the Ascalaphidae and associated 
families. So far as I am aware, there was no description 
of myiodactylid larvae prior to 1926, though Tillyard 
(1917, p. 543) mentioned the larva of Myiodactylus and 
indicated it would later be described. It is probable, 
therefore, that no information concerning these larvae 
was available to Withycombe, and that the resemblance 
between Osmylidae and Myiodactylidae with respect to 
wing venation led him to associate closely the two 
families. 
Klingstedt (1937) has suggested that the Dilaridae may 
be closely related to the Raphidiodea, basing this view on 
chromosome structure. Tjeder (1937), who has studied 
the external and internal anatomy of adult Dilaridae and 
Raphidiodea, points out both similarities and dissimi- 
larities, and concludes that sufficiently close relationship 
is shown to justify transferring the Dilaridae from the 
Neuroptera (strict sense, Planipennia of authors) to the 
Raphidiodea. Larval characters of the dilarids, previ- 
ously unknown, impress me as more trustworthy indica- 
tors of relationship than the adult characters thus far 
studied, and consequently I believe the primary affinities 
of the Dilaridae are with the Osmyloidea. The Berothidae 
are close relatives. 
The presence of an ovipositor in both raphidiids and 
dilarids has led Tjeder and others to suspect close rela- 
tionship. A comparison of the adult female ovipositor 
of Nallachius (fig. 15) with those of mantispids of the 
genera Symphrasis and Plega, and with those of the 
raphidiid genera Agulla and Inocellia shows that there is 
an abrupt bend or elbow at the base of the ovipositor in 
the first two families which is absent from the Raphi- 
diodea. In several other ways the Mantispidae suggest 
