NO. O INSECT ABDOMEN SNODGRASS 25 



abdominal segments, are evidently the primitive paired openings of 

 the lateral genital ducts, which remain separate in the male, but unite 

 in a median exit in the female. The position of the gonopores in 

 this group might be taken as indicating that at some remote time in 

 the phylogenetic history of the Hexapoda the lateral mesodermal 

 ducts of both sexes opened just before the terminal body segment, 

 as they do in Qiilopoda. In this event, the paired openings on the 

 tenth and seventh segments of insects are phylogenetically secondary, 

 and, if so, it becomes an interesting subject for speculation as to 

 how they may have been established on these segments. 



Theoretical considerations. — Tillyard (1930) has pointed out that 

 under the term " opisthogoneate " there is included within the Hexa- 

 poda a variation of six segments in the position of the genital open- 

 ings, one extreme being represented by the Collembola with the gono- 

 pore on the fifth abdominal segment, the other by the Protura with 

 the gonopores on the eleventh segment. The discrepancy is somewhat 

 exaggerated, however, by including the Collembola, since the col- 

 lembolan gonopore is a median genital opening, which, as we have 

 seen, is secondary and subject to migration. The variation of the 

 primitive lateral gonopores in Hexapoda is thus only from the seventh 

 to the eleventh segments — a distance of four somites. To account for 

 this variation, and for the greater variation in other arthropods that 

 brings about the " progoneate " condition, Tillyard proposes the 

 assumption that the primitive arthropods had segmentally paired gon- 

 ads, and that each gonad had an individual segmental duct. The second 

 part of this assumption finds little support in the known facts of the 

 embryonic development of the insect gonads or their ducts. The in- 

 vestigations of Heymons (1892) and of Wheeler (1893), as shown 

 above, have made it clear that the only stage in which the reproductive 

 elements have a segmental arrangement is that in which the germ 

 cells become segmentally arranged in the dorsal part of the splanchnic 

 mesoderm. But at this stage there are no exit ducts. The condition is 

 in every way parallel with that in the Annelida, and it is only at this 

 early ontogenetic period that a comparison can be made between the 

 Annelida and the Arthropoda with respect to the reproductive organs. 

 The definitive gonads and the genital ducts of most insects are formed 

 as continuous mesodermal structures, and the later subdivision of 

 the gonads into tubules is generally a secondary process in ontogenetic 

 development. Though the tubules may take a segmental arrangement, 

 they never have separate, segmental ducts to the exterior of the body. 



According to Wheeler's (1893) concise account of the development 

 of the genital ducts in Conocephalus, the embryonic male ducts end 



