26 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 89 



in hollow enlargements, or terminal ampullae, situated at first in the 

 appendage rudiments of the tenth abdominal segment (fig. 3 B, e) ; 

 the female ducts, which likewise terminate with ampullae, end in the 

 appendage rudiments of the seventh abdominal segment. The female, 

 however, Wheeler observes, has also a pair of ampullae in the tenth 

 segment appendages, which appear to be homologues of the male 

 ampullae. The account given by Heymons (1890, 1895) of the genital 

 ducts in Dermaptera, Blattidae, and Gryllidae is essentially in agree- 

 ment with that of Wheeler for Conocephalus in so far as the male 

 ducts are described as terminating in mesodermal ampullae in the 

 tenth segment, and the female ducts in ampullae of the seventh seg- 

 ment. According to Heymons, however, there is evidence of a primary 

 branching of the ducts in each sex to both the seventh and tenth 

 segments. Wheeler, on the other hand, believes that male insects never 

 have ampullae or branches of the genital ducts in the seventh segment. 

 Heymon's illustration of Forficula (fig. 3D) gives a convincing ex- 

 ample of the branching of a genital duct to the two segments, but 

 his identification of the posterior branch {d) as the definitive oviduct 

 is evidently wrong, since the definitive female ducts in Dermaptera 

 open on the seventh segment. 



From the above we might deduce, as a plausible explanation of 

 modern conditions, the theory that at some period in the phylogenetic 

 history of insects, the mesodermal ducts of the gonads opened in each 

 sex on the bases of the appendages of the tenth abdominal segment, 

 and that later the ducts in the female developed branches to the sev- 

 enth appendages and lost the primitive connections with the tenth ap- 

 pendages. The reverse assumption, namely, that the openings on 

 the seventh segment are primitive in both sexes, and those to the 

 tenth segment secondary in the male, would appear to be disqualified 

 by the subterminal position of the genital apertures in both sexes 

 of Onychophora, Chilopoda, and Protura. It seems scarcely necessary 

 to accept the view of Heymons (1890) that primitive insects were 

 hermaphroditic. The theory of the branching of the genital ducts 

 offers a possible explanation of the differences in position of the dual 

 genital openings among the Insecta, and makes it possible to derive 

 this condition from the more primitive condition retained in Chilopoda 

 and Protura. The theory at least is in accord with the known facts of 

 embryonic development. 



III. THE OVIPOSITOR OF THYSANURA 



Among the Apterygota an ovipositor is well developed in the 

 Thysanura. In Dicellura and Collembola the organ is entirely absent ; 

 and the female genital armature of the Protura can scarcely be sup- 



