132 Comparative Classification of Cynipidae. 



The different form of the plates in these two ovi- 

 positors is especially striking. In Neuroterus they are 

 almost circular, and owing to their great curvature, there 

 is no room for the usually strong muscles of the 

 anterior plate (Fig. i, No. i). This muscle therefore 

 appears quite rudimentary, and the second muscle 

 which springs from the arch is wholly wanting. 



II. The Aphilotrix-Andricus Group. 



In this group also, we meet with differences in the 

 ovipositor, but in some cases they are very slight. If 

 we compare the two ovipositors of Aphilotrix radicis 

 (Fig. 4), and Andricus Jiodidi (Fig. 4^), a general 

 harmony is observed, but functional differences are 

 easily recognized. The radicis ovipositor ends in 

 a sharply curved point, and is consequently not adapted 

 to penetrate perpendicularly into plant tissue ; the fly 

 must drive it into the bud in a circuitous fashion. The 

 ovipositor first glides under the bud-scales to the base 

 of the bud-axis, and is then directed upwards. The 

 noduli ovipositor, on the contrary, can penetrate per- 

 pendicularly into the bark with its straight point. 

 But with ovipositors so much alike as these, it is well 

 to have some further marks of distinction. At the end 

 of the posterior plate, in all gall-flies, there is a little 

 papilla which is distinctly prominent, of slender form, 

 and thickly covered with tactile hairs. The corre- 

 sponding parts of the ovipositor, and of the two posterior 

 plates, lie closely on each other, but room for the 

 protrusion of the rectum must be left between them. 

 For that reason there is in each plate a little notch. 



