343 



To what extent tliis bisegmental condition of the pro- 

 podeum is found in Hynienoptera generally I am unable to 

 say. In Nasonia there can be no doubt of it, but to deter- 

 mine its constitution in other groups would necessitate an 

 embryological examination of these, the study of mature 

 material being generally useless. 



It remains only to describe the segments of the rest of 

 the abdomen. The third segment is large and conical and 

 overlaps the second, which is rather shorter; the next is very 

 long, and the following two rather shorter. The last four 

 segments are small and together form the posterior quarter 

 of the abdomen (fig. 4). 



Chitinisation has taken place so as to produce distinct 

 terga and sterna. 



The sterna of the seventh and eighth segments do not 

 meet below, the body being protected here by the st-emum 

 of the ninth segment. The tenth segment is partly invaginated 

 into the preceding one and is represented by a terminal plate 

 which bears a long horizontal slit, the anus (figs. 21, 26). 

 Surrounding this plate are other podical plates, but whether 

 these are formed from the last, or from the antepenultimate 

 segment, I am not definitely able to say. The last segment is 

 interesting in that it bears in the female a pair of processes, 

 which grow out some ten hours before putation; in the newly 

 formed pupa they are in the form of short blunt appendages 

 a little longer than broad, quite prominent in ventral view 

 (fig. 21). 



In the early pupa the usual contraction takes place, ajid 

 they remain as short conical papillae at either side of, and 

 just below the anus; each is covered with long bristles, 

 developed early in pupal life. These are the tactile hairs, 

 and the structure serves as a delicate sense-organ for the 

 female in the examination of the surface of the fly-pupa for 

 a suitable spot to pierce with her ovipositor. 



The modifications which these structures undergo in the 

 male will be described in connection with the development of 

 the male copulatory organs; histological details will be given 

 in connection with the description of the integument. 



In the female, meantime, the ovipositor has been develop- 

 ing. This is represented in the feeding larva in its last instar 

 by three pairs of imaginal discs, situated in the twelfth, 

 thirteenth, and fourteenth segments, and identical in appear- 

 ance with the anlagen of the legs (fig. 2) ; there is no evidence 

 of a biramous structure. Some time before defaecation these 

 appendages grow out above the surface ; the upper pair grow 

 backwards along the ventral side of the insect as two hollow 



