310 
Indiana University Studies 
winged pezomachoides , they are fundamentally similar as an 
examination of a series including all of the species of Acraspis 
will show. Enlarged details of these galls are shown in fig- 
ures 325 to 331. In every case the gall consists of a thick- 
ened wall with the larval cell large and without a distinct cell 
wall. The surfaces of the galls are set with polyhedral bodies 
that usually terminate in blunt, hair-like spines. These spines 
are long and flexuous and much tangled when they form the 
woolly covering of the galls of nubila; they are less dense and 
shorter in some varieties of villosa and some galls of pezoma- 
choides; they are still shorter, stiff er, and more nearly erect in 
other varieties of villosa and gemmula; and they are even 
lacking, leaving the galls nearly naked in hirta and in some of 
the varieties of pezomachoides. These seem to be brilliant in- 
dicators of the relationships of these insects. The galls of 
Cynips pezomachoides erinacei offer the best proofs, for within 
that single variety they vary from the naked to the very spiny 
form. 
The final proofs of the relations of the blunt-spined, short- 
winged insects are the characters of the bisexual females ex- 
perimentally and circumstantially connected with two of the 
agamic forms (erinacei and prinoides) of eastern Acraspis. 
The bisexual females have hypopygial spines, wing-body ratios, 
and indications of cubital cell markings that clearly agree 
with the characters of the long-winged, agamic forms of nubila 
and acraspiformis. 
For these reasons, the name Acraspis, originally established 
for a group of short-winged, blunt-spined, agamic cynipids, 
is here applied to a subgenus of Cynips that includes both 
short-winged and long-winged varieties, the latter always with 
slender, plough-shaped spines. 
The assignment of the species mellea has involved other 
questions. The typical variety of the species is a short-winged 
insect that Ashmead originally considered a Biorhiza. Later 
he made it the type of the genus Sphaeroteras because of the 
character of the frons, the “13-jointed antennae”, the rounded 
scutellum, the length of the hind tarsus, and the indistinct 
tooth on the tarsal claw. Mayr, studying four of Ashmead’s 
types, pointel out (1902, Verh. zoo.-bot. Ges. Wien 52:289) 
that the presence or absence of a carina on the frons was a 
variable character not even of specific import, that the anten- 
nae of mellea are really 13-14 segmented with the 14th segment 
