306 Annals Entomological Society of America [Vol. XII, 



jointed cerci, the loss of the ovipositor, etc., but the special 

 features of each group indicate a different line of descent. 



As regards the female genitalia, but little evidence in support 

 of either view is obtainable from this source alone. In Embia 

 major, e. g., (Fig. 68) the general appearance of the terminal 

 segments is much like that of a termite. The csrci are two- 

 jointed as in most termites, the tenth tergite is large and curves 

 downward, covering the supra-anal plate, to which it is adher- 

 ent. On the other hand, there is no extension of the seventh 

 sternum to form a subgenital plate, in fact, no modification 

 of either seventh or eighth sterna, and no lateral reduction of 

 tergites eight, nine and ten as in the Isoptera. The paraprocts 

 are large, but unchitinized and are distinct from the well- 

 developed cereal basipodites. 



Whereas in Termopsis there is evidence that the Isoptera 

 are descended from ovipositor-bearing ancestors, there is no 

 indication in E?nbia or any of the Embiidina that an ovipositor 

 was ever present, the eighth and ninth sterna being quite 

 simple. Since an ovipositor of primitive form is present in 

 some Apterygota {Macliilis, Lepisma, etc. of the Thysanura) 

 and is undoubtedly homologous with that of Pterygote insects, 

 it must have been present in the earliest representations of the 

 latter, unless we are to regard the ovipositor-bearing Thysanura 

 as descended from winged forbears, a view which I believe few 

 will accept. This being the case, the Embiids, Plecoptera, etc., 

 must likewise, (contrary to Crampton's opinion) be considered 

 as secondarily without ovipositors. The fact that Campodea 

 and other Thysanura have also no ovipositor does not affect 

 the question. There is evidence that some ' of the Palaeo- 

 dictyopteri had no ovipositor, while it is certain that some of 

 them had one, and it was probably among the former that 

 the ancestors of the Embiids and Plecoptera existed. The 

 ovipositor, which was probably never highly differentiated, had 

 already disappeared before these groups had acquired independ- 

 ent ordinal rank. 



It is scarcely profitable to make a comparison between the 

 Embiids and Grylloblatta until the structure of the male 

 genitalia has been considered, but the following characters 

 held in common between the females of the two orders may be 

 mentioned: (1) There is no reduction in the length of the eighth 

 and ninth tergites; (2) the spiracles are all situated in the 



