312 
Indiana University Studies 
Were we to extend Ashmead’s conception of Sphaeroteras 
to include the whole species mellea, we could not make satis- 
factory assignments of the Southwestern species arida and 
conica which seem to me to be about equally related to mellea 
and to nubila and villosa, altho these latter species are unques- 
tionably true Acraspis. While recognizing the extreme posi- 
tion of mellea within the subgenus, I cannot see convenience 
in or phylogenetic basis for separating it from Acraspis. 
If our extension of Acraspis seems too inconvenient for 
practical use, it should be admitted that it is sometimes impos- 
sible to show true relationships by any simple scheme of classi- 
fication. Whenever these two functions of taxonomy prove 
incompatible, I have preferred to show the phylogeny of the 
group, believing that our data may thereby be best coordinated 
with the other fields of biologic science. For those who prefer 
a simpler classification, the older revisions of these groups are 
still available. 
The best structural characteristic of the subgenus Acraspis 
as here defined is the slender, drawn-out, plough-shaped hypo- 
pygial spine of some of the forms, or the uniformly wide spine 
with a blunt, rounded tip in other forms. All the long-winged 
species except mellea and conica show at least one smoky patch 
in the cubital cell, tho this mark is very faint in the bisexual 
forms. The basically spherical, thick-walled gall without dis- 
tinct larval cell is a better key to the group than any character 
of insect structure. 
Weld’s separation (1922, Proc. U. S. Nat. Mus. 61 (18) : 
10-12) of Acraspis and Philonix is good for the species 
covered. The statement that the adults emerge in the fall, 
usually before the leaves drop, does not agree with our records 
for emergence in the group (see the data accompanying any 
of the well-known varieties, e.g., erinacei). In my experience, 
the life history data for Acraspis are so nearly in accord with 
those for Philonix that they furnish some proof of the rela- 
tions of the two groups. 
The conspicuous galls of the agamic forms of Acraspis ap- 
pear late in June or early in July, reaching maturity late in 
July or in August. The insects pupate sometime later, usually 
in September, and adults emerge from early in October until 
the following January, mostly in November further north or 
December further south. There are no data to show that any 
