NO. 14 INSECT ABDOMEN SNODGRASS 59 



original nature may have been, since the vasa deferentia come to open 

 into its anterior end at the bases of the glandular diverticula, the 

 median duct of the male becomes the exit both for the secretion of 

 the accessory glands and for the spermatozoa. The mouth of the com- 

 mon duct is the median gonopore of the male. The gonopore may be 

 exposed on a small phallic papilla arising from the ventral membrane 

 between the ninth and tenth abdominal segments, but generally this 

 membrane is invaginated to form the male bursa genitalis, or genital 

 chamber (C, GC), and the gonopore is then carried out upon a tubu- 

 lar evagination of the anterior wall of the chamber, which forms the 

 usual phallic organ. 



A study of the anatomical relations between the vasa deferentia and 

 the posterior segmental nerve trunks of the abdomen in Orthoptera 

 shows that the male genital ducts must originally have crossed over 

 the lateral nerves of the tenth abdominal segment and attached to the 

 ectoderm behind these nerves (fig. 21 A, Vd). The subsequent union 

 of the vasa deferentia with the median ductus ejaculatorius (B), 

 however, necessitated a median and forward migration of the posterior 

 ends of the lateral ducts. The correlated forward retraction of the 

 posterior part of the ventral nerve cord then drew the large nerves 

 of the cerci (CerNv), given off from the terminal ganglion, over the 

 incurved ends of the vasa deferentia (Vd). Hence, in the definitive 

 condition (C), the vasa deferentia are always looped beneath the 

 cereal nerves, but lie dorsal to the other lateral nerve trunks of the 

 abdomen. This condition could follow only from one in which the 

 primitive genital ducts turned downward to the body wall between 

 the nerve trunks of the tenth and eleventh abdominal segments (A). 

 Likewise, in female Orthoptera, the relation of the lateral oviducts 

 to the nerve trunks shows that the primitive lateral ducts must have 

 opened somewhere between the nerves of the seventh and eighth 

 abdominal segments. Incidentally, the nerve-duct relation in the male 

 demonstrates also that the cerci are appendages of the eleventh 

 abdominal segment. 



The paired penes of the Hexapoda are best shown as simple inde- 

 pendent structures in some of the Ephemeroptera (fig. 25 C, Pen). 

 They are supported on basal arms or a basal plate (C, F, x), which 

 possibly may represent the sternum of the tenth abdominal segment. 

 Since there are no true appendages on the tenth segment, there is no 

 evidence as to what relation the penes may have had to the limb bases, 

 but, judging from other arthropods, there is no reason to suppose 

 that the primitive penes of the hexapods were structures other than 

 papillae of the coxopodites containing the outlets of the genital ducts. 



