18 
Indiana University Studies 
irregularis and decipiens, and minutus and quercicola, are 
very evidently eastern and western equivalents of two primi- 
tive types. This phenomenon is found in several other in- 
stances in Cynipidse outside of this genus. The two areas, 
roughly divided by meridian 100, are not sharply separated 
by present-day geography ; in attempting to account for them 
we should consider differences in rainfall, in the species of 
the oaks, and particularly in the geologic histories. 
The distribution of varieties may be summarized quite 
strictly. Each faunal area has only a single variety of any 
species on any single host. There are no exceptions in the 
G3 varieties known, altho in Central Texas (as at Austin and 
Yoakum) where two areas meet, two varieties of a species 
may be found on the same tree. The general rule is abundant 
proof of the effectiveness of geographic and host (and other) 
isolations in preserving and fixing new species or varieties. 
My definition of faunal areas is based on my own studies 
of gall-making Cynipidse of all groups, and on the relatively 
small amount of published data which appears to be based 
on actual taxonomic work. But the rule as to the geographic 
isolation of varieties is only slightly dependent upon our 
definition of faunal areas. The abundance of geographic 
regions of restricted size which contain only a single variety 
of a species on a given host, the abrupt and completely changed 
characters of the clearly related variety in the next geographic 
area, and the uniformity of boundaries between areas for 
the varieties of practically all cynipid species, leave no room 
for doubting the reality of faunal areas. 
LIFE HISTORIES 
The alternating generations of 13 varieties of Neuroterus 
are known, representing 1 or 2 European species and 7 Amer- 
ican species. The European life histories are based on Adler’s 
experimental work, but no American life history has been 
experimentally determined except in the cases Dr. Patterson 
is studying, incomplete reports of a couple of these being 
given in this paper. It is possible to make predictions as 
to the alternation of 8 of the American species representing 
34 described varieties. 
The heterogeny of Neuroterus species is so very simple 
that it becomes possible to predict life histories on the basis 
