NO. 6 GRASSHOPPER ABDOMEN SNODGRASS 49 



segmental appendages. Even the fact that the first rudiments of the 

 valvulae appear in some insects on the embryo in hne with vestigial 

 appendages on the pregenital segments is not necessarily evidence that 

 they are homodynamous with the latter, since secondary structures 

 arising in the same relative positions as the true limbs would be very 

 likely to assume the same form in early stages of growth. The best 

 evidence of the origin of the ovipositor from limb structures is fur- 

 nished by the Thysanura. in which the valvulae are outgrowths of 

 lateroventral plates of the genital segments that are clearly equivalent 

 to the stylus-bearing plates present in some forms on the preceding 

 segments, which plates, there seems little reason to doubt, represent 

 the bases of true abdominal limbs. The ovipositor of Thysanura, there- 

 fore, appears to be formed of mesal processes (gonapophyses) of the 

 coxopodites of the appendages of the eighth and ninth abdominal seg- 

 ments, and the fundamental similarity of the ovipositor in Thysanura 

 and Pterygota leads us to conclude that the organ is an homologous 

 structure in all insects in which it occurs. 



The facts of the development of the ovipositor in Orthoptera are 

 easy to ascertain and are in general well known. In a young nympth 

 of the cricket Nemobius (fig. 21 A) two small conical processes (iVl) 

 project from the membranous ventral part of the eighth segment 

 behind the eighth sternum (VlllStn), entirely free from the latter. 

 These processes are the rudiments of the first valvulae. The valvulae 

 of the ninth segment are not yet in evidence ; the sternal region of this 

 segment (IXS) shows no differentiation except two slight rounded 

 swellings of its posterior margin. At a somewhat later stage (B, C), 

 however, a pair of valvular processes is present on each genital seg- 

 ment. Those of the eighth segment (B, iVl) still arise from the mem- 

 brane behind the reduced eighth sternum (VlllStii). The processes 

 of the ninth segment (C, 3VI), on the other hand, which become the 

 third valvulae of the adult, arise directly from a median sclerotization 

 of the ventral wall of the segment, at the sides of which is a pair of 

 small but conspicuous oval lateral sclerites (x) . There is thus no ster- 

 nal plate in the ninth segment distinct from the bases of the valvulae. 



The two primary pairs of valvular processes increase in length with 

 successive instars (fig. 21 E), and the rudimentary second valvulae 

 appear ventrally between the bases of the third valvulae (F, i'F/), 

 but the relations of the valvulae to their respective segmental areas 

 remain unaltered. Up to a late stage there is no evidence of the pres- 

 ence of valvifers, except for the small lateral sclerites (x) of the ninth 

 segment, which increase in size and become more dorsal in posi- 

 tion (F). 



