92 
Indiana University Studies 
we might recognize a third species, because the bisexual in- 
sects of divisa in Central Europe are easily enough distin- 
guished; but the situation remains as shown in the following 
diagram where similarities are indicated by arrows, the 
agamic insects by ® , and the bisexual insects by $ $ . 
Folii Longiventris Divisa Agama Disticha Cornijex 
| f < — > i f < — > I « — > f < — > I 
southern f 9rfl< _ >9d , 9rf , 
insects J 
Northern 
insects 
Central 
and 
On the other hand, from galls alone, European Cynips would 
be granted six or seven species, which would recognize no dis- 
tinctions (on galls alone!) between northern and more south- 
ern material. Graphically, this may be shown (where © is 
the gall) as follows: 
Folii Longiventris Divisa Agama Disticha Cornijex 
Northern 
e 
0 
0 
galls 
A 
i 
A 
A 
Central 
1 
V 
1 
V 
1 
V 
European 
© 
0 
0 
galls 
A 
Southern 
V 
galls 
If we take into account both insects and galls, which should 
we use for establishing specific lines, and which for varietal 
lines ? Our procedure may be determined by the fact that be- 
tween the galls of the six types there are no intergrades, while 
between the northern and Central European insects there are 
numerous intergrades. This suggests that each northern in- 
sect has been derived by variation and geographic isolation 
from a Central European insect, and that the galls remain as 
indicators of the ancestry of these variants, i.e. that the galls 
in the subgenus Cynips should be taken to indicate the spe- 
cific, and the insects the varietal lines of a natural classifica- 
tion. The conclusion finds abundant corroboration in our 
American experience with geographic variation; and Euro- 
pean students are not likely to dissent with a solution that does 
no more damage than this to the previous classifications. The 
