374 
Indiana University Studies 
evidently derived from the same stock, but the varieties of 
gemmula are confined to the chestnut oaks from which pezo- 
machoides seems to be entirely excluded. 
We have galls of what is probably an undescribed variety 
of pezomachoides from Oklahoma (in Amer. Mus. coll.), and 
a few undescribed insects from localities in Utah. The latter 
material was bred with insects of Cynips hirta packorum , and 
this suggests that the galls of the two species are not separa- 
ble in Utah and that the insects of the two may even hybridize 
in that part of the country. The following published records 
seem to apply to undescribed varieties : 
Acraspis pezomachoides err. det. Thompson, 1915 ( Q . bicolor, Mass, 
record only), Amer. Ins. Galls: 15, 36. Err. det. Weld, 1926 (Tex- 
arkana record), Proc. U.S. Nat. Mus. 68 (10) : 58. 
The Rocky Mountain material of the species testifies to the 
origin of the group at a period before the Great Plains became 
so arid as to exterminate the oak forests that once connected 
the Southwestern and more Eastern relatives. The seven 
eastern American varieties of the species represent not more 
than three and possibly only two main stocks, a pezomachoides- 
derivatus stock, a wheeleri stock, and perhaps another stock 
which gave rise to variety ozark. Pezomachoides, existing in 
pure populations only on the rim of the Atlantic Coastal Plain 
today, is so widely represented in hybrids thruout the South- 
east that it may be taken as very close to the original southern 
form of the species. From it the variety derivatus is a dis- 
tinctive but not greatly divergent development. The northern 
variety wheeleri was undoubtedly forced southward during the 
Pleistocene glaciation, and as a result we have two varieties of 
hybrid origin. The well-known variety erinacei, found over 
most of the northeastern United States, seems to be of 
wheeleri x derivatus origin; and the variety advena, of the 
Southern Highlands, seems to be of ivheeleri x pezomachoides 
origin. The data are discussed under each variety. Variety 
ozark also seems to carry wheeleri blood, but I have not recog- 
nized the other possible parent. Variety echinoides, from 
Florida, is too poorly known to warrant analysis. 
Determinations of insects of this species are unusually diffi- 
cult, in part because of the few taxonomic characters available 
with such reduced wings and thoraces, partly because of the 
great variability of the hybrid populations erinacei and advena, 
