1919] Walker: Structure of Orthopteroid Insects 301 



In the third stage (Fig. 57) which represents the last nymphal 

 instar, the eighth sternum is still narrower and has lost its disto- 

 lateral angles. The ventral and inner valvulae have changed 

 but little, but the two lobes of the ninth sternum have greatly 

 decreased in width, though the styli are practically unchanged. 

 They are now clearly recognizable as the dorsal valvulae. 



The metamorphosis of this primitive condition into the 

 complerx structure of the adult is a transformation worthy of 

 a higher order of insects, but the most significant feature in 

 the entire process is the clear indication that the dorsal valvuke 

 are a part of the ninth sternum, and are homologous with the 

 flat, stylus-bearing sternal lobes or "coxites" of the Thysanura. 

 Their comparatively late development is no objection to their 

 interpretation as such primitive structures, in view of the well- 

 known fact that organs, which are not functional until adult 

 life is reached, are frequently retarded in development. 



The ovipositor of Cryptocercus punctulatus (Fig. 59) presents 

 an interesting structure intermediate between that of the adult 

 and late nymphal stages of more typical roaches, such as 

 Parcoblatta. The ventral and inner valves are quite like those 

 of an immature roach, while the dorsal valves have something 

 of the mature form, but are unusually simple. The valvifer 

 is of remarkably generalised form and together with the well- 

 marked intersegmental apodeme bears a closer resemblance to 

 these structures in the Mantids, than is found in any other 

 Blattid I have examined. 



The ovipositor of Cryptocercus is, however, very degenerate, 

 the valvulcB and valvifers being very feebly chitinized and their 

 inner connections greatly simplified. Thus, in view of the many 

 details held in common by the more complex types of Blattid 

 ovipositor and that of the Mantids, the simple structure found 

 in Cryptocercus must be looked upon as at least partly due to 

 arrested development, or the persistence of a larviform con- 

 dition, rather than a truly primitive one. It is closely paralleled 

 in the Phasmoidea by Timema californica. 



It may be added that in certain respects Cryptocercus is 

 highly specialized, namely in the enormous development of 

 the seventh tergite, which is prolonged backwards over the 

 remaining segments, so that the latter are completely concealed 

 .above and below. 



