20 Mr. J. O. Westwood on the Chalcididce. 



c, c, represent the antepectus longitudinally divided in the centre by a 

 deep groove. To the front part of this antepectus the head is attached at 

 a, and to its posterior part are attached the basal joints, or coxse of the fore 

 legs, d, d. 



The Abdomen, (upon w^hich as a generic character I think great stress 

 ought to be placed, inasmuch as its variations clearly indicate correspond- 

 ing variations in the method of oviposition,) is longer than the head and 

 thorax, and of equal breadth v^ith the head near the vs^hole of its length ; 

 the two last joints are attenuated, and rounded at their hinder margins; the 

 upper surface, in all the specimens Mrhich I have seen, is quite flat, 

 7-jointed, the fourth being as long as the three preceding, the fifth about 

 twice the length of the fourth; the tips of the organs of oviposition being 

 just perceivable, (G. andH. e.) The peduncle is very short and rather 

 thick, (G. and H. a.) I may here notice a singular character, namely, 

 that the segments of the abdomen are all completely covered with minute 

 punctures, except at the posterior margins of each segment. Fig. H. re- 

 presents a side view of the abdomen, and it will be seen that it is not near 

 so deep nor so much angulated as in some of the neighbouring genera. 

 Fig. G. represents the under-side view of the same part, and exhibits 

 several characters extremely worthy of notice. In this view, the rings 

 marked respectively d, are merely continuations of the dorsal segments, 

 and not distinct ventral ones. This formation has been partially observed 

 by Messrs. Kirby and Spence, Intr. Vol. III. 707, who remark, that " in 

 some Hymenoptera (Cimbex) the sides of the last dorsal segment turn 

 down and become ventral; on its lower side it has in these a longitudinal 

 cavity, which receives the ovipositor in repose;" and at p. 703, they re- 

 mark of Leucospis, a genus of Chalcididse, that " the ventral segments are 

 replaced by a long, narrow, central plate, succeeded by a minute one." 

 In Cleonjonus, however, there appear to be only two actual ventral seg- 

 ments, independent of the dorsal ones; they are situated at the base of the 

 abdomen, and are small, the second being the most minute (G. and H. b 

 and c.) ; the dorsal segments, therefore, become in fact ventral, but they 

 do not close; indeed the basal and apical ones are considerably asunder in 

 the specimens which I have examined, one edge only of the fourth segment 

 folding slightly over the other ; the last segment does not even turn down. 

 From die hinder margin of the last, or second ventral segment, the ovi- 



