12 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 89 



described by Nelson (1915), the genital ridges of the female at first 

 extend along the splanchnic walls of the mesodermal tubes from the 

 second to the seventh abdominal segments, inclusive. Later they 

 become shorter and thicker, and finally reach only from near the 

 anterior end of the fourth abdominal segment into the anterior end 

 of the seventh. 



Beyond the region of the germ cells the genital ridges are continued 

 posteriorly, but they are here much reduced in size and consist of 

 simple cellular strands. These parts of the ridges will form the meso- 

 dermal parts of the lateral genital ducts. It has been observed in many 

 insects that the primitive ducts of the female in embryonic and larval 

 stages turn downward posteriorly and are attached to the ectoderm 

 at the posterior end of the ventral wall of the seventh abdominal seg- 

 ment. The male ducts of Orthoptera, as described by Heymons and 

 Wheeler, extend into the tenth segment and are here similarly attached 

 to the ectoderm. 



Comparison with Annelida. — The early segregation of the germ 

 cells in Blattella into groups within the dorsal parts of the inter- 

 coelomic septa is highly suggestive, as Heymons (1892) points out, 

 of the similar arrangement of the germ cells in many adult Annelida, 

 in which the gonads are simple swellings of the dissepiments, retain- 

 ing the germ cells beneath a thin mesodermal epithelium. The gonads 

 of Annelida, however, may occur on almost any part of the coelomic 

 walls, but wherever they are formed, the germ cells are not long 

 retained within them, being soon thrown out into the body cavity, 

 where they mature, and from which they are eventually discharged 

 through the nephridia, through special genital ducts, or by way of a 

 pore or rupture in the body wall. The liberation of the germ cells 

 into the body cavity in Annelida, Heymons claims, is represented in 

 Blattella by the extrusion of the germ cells from the dissepiments 

 before their reentrance into the mesoderm, the only difference being 

 that in the cockroach the germ cells are not detached from the meso- 

 derm, but migrate upon its surface to their definitive positions in the 

 genital ridges. In Conocephalus, however, Wheeler (1893) observed 

 that some of the primitive germ cells of the segmental clusters " show 

 a tendency to leave the wall of the somite and to drop into the coelomic 

 cavity." These cells, he says, " sometimes enlarge considerably, become 

 vacuolated and take on the appearance of young ova." 



If the condition described by Heymons in Blattella is truly a primi- 

 tive one, it becomes evident that the definitive gonad of the Arthropoda 

 is a secondary structure produced by the enclosure of the previously 

 liberated germ cells in folds of the splanchnic mesoderm, which folds 



