32 
Indiana University Studies 
Fulvicollis and its immediate relatives are short-winged in- 
sects (fig. 234) which we had concluded were Gynips of the 
subgenus Philonix, but no long-winged Philonix had been 
recognized among entomologists. But the bisexual insect, 
Gynips pallipes, came to our attention as a possible bisexual 
form of fulvicollis (for reasons given on p. 271) , altho pallipes 
had a peculiar wing-body ratio of 1.17 not then recognized in 
any subgenus of Cynips, and a hypopygial spine (fig. 251) 
distinct from that of fulvicollis (fig. 252) . There was only one 
other long-winged. Gynips which we had not placed at that 
time in one of the six subgenera, and that was the agamic 
C. plumbea of the southwestern United States. Upon re- 
examination of the data, plumbea proved to have the same 
wing-body ratio and the same spine as pallipes, while the gall 
of the agamic plumbea (figs. 225-226) was hardly distinguish- 
able except in color from the gall of the agamic fulvicollis. 
The peculiar spine of fulvicollis had to be taken as one more 
case of a modification accompanying wing reduction. Thru 
galls, alternate generations, wing-body ratios, and interpreta- 
tions of reduced wings and transferred spines the data became 
interpretable without contradiction in any part. 
For all of the 42 short- winged insects which we are con- 
sidering as true Gynips there is this same coordination of the 
evidence. Our conclusions are summarized in the following 
table, where the style of indentation will indicate the phylo- 
genetic affinities of each form, and the order will show the 
development from the more primitive to the more specialized 
members of each group. The number that follows each name 
shows the average wing-body ratio in that species. It will 
appear that there are at least 11 stocks of Cynips in which the 
subapterous condition must have arisen independently, and it 
is possible that a still larger number of the 42 subapterous 
species have arisen directly from long-winged ancestors. 
Wing-lengths in Agamic Cynips 
1. Subgenus Cynips longiventris 
folii 
longiventris 
... 1.50 
folii 
... 1.50 
bisex. form . . . . 
... 1.50 
bisex. form . . . . 
... 1.50 
forsiusi 
... 1.50 
flosculi 
... 1.50 
divisa 
bisex. form . . . . 
. . . 1.50 
divisa 
... 1.50 
ilicicola 
bisex. form . . . . 
... 1.50 
atri folii 
... 1.50 
atridivisa 
... 1.50 
