NO. 8 INSECT ABDOMEN SNODGRASS 73 



In Blatta orientalis the dorso-pleural line of the abdomen (fig. i6 B, 

 a-a, C, a) is contained in the narrow, infolded lateral membrane unit- 

 ing the paratergites {pt) and the large ventral plates. Since in the 

 male roach the ventral plate of the ninth segment bears a pair of styli, 

 it is probable that the definitive sterna of all the abdominal segments 

 are composite pleurosternal plates (PlStn) formed of the limb base 

 areas and the primitive sterna. The lateral abdominal muscles of 

 the roach are all functionally tergosternal (fig. i6C), though some 

 may be morphologically tergopleural. Two muscles of each segment 

 are attached above on the median tergite (T), and one on the para- 

 tergite (pt), the last being described by Ford (1923) as a " pleural " 

 muscle. The three are attached below on the lateral parts of the large, 

 strongly convex sternal plate (PlStn). The roaches do not make 

 any perceptible breathing movements of the abdomen under ordinary 

 conditions. The well-known diagram by Plateau (1884) purporting 

 to show the respiratory movements of Blatta orientalis undoubtedly 

 depicts the kind of movements a roach can bring about between its 

 dorsal and ventral abdominal plates ; but since Plateau's subjects were 

 induced to make them after being pinned down by the edges of the 

 prothorax, with the legs and wings cut off, and the abdomen propped 

 up on a bent pin, there is reasonable doubt that the observed contrac- 

 tions and expansions of the abdomen represent normal respiratory 

 movements. 



The genital segments of female Blattidae are almost entirely con- 

 cealed within the seventh segment (fig. 24 A). Beyond the seventh 

 tergum, and above the large terminal lobes of the seventh sternum, is 

 the proctiger carrying the cerci laterally, projecting from the concealed 

 ninth segment (B). The dorsal plate of the proctiger (XT) is evi- 

 dently the tenth tergum alone, as claimed by Walker (1919), the 

 epiproct being represented only by the membranous fold beneath its 

 base. The muscles of the cerci take their origin on the proximal part 

 of the plate, but the weakly developed muscles of the paraprocts are 

 attached dorsally in the ventral membrane. The writer in a former 

 paper (1931), without examining the paraproctial musculature, re- 

 garded the terminal dorsal plate of the cockroach as formed of the 

 combined tenth tergum and epiproct, and Crampton (1929) makes the 

 same interpretation. Walker (1919) points out, however, that "the 

 disappearance of the supra-anal plate (epiproct) in the blattids and 

 its substitution by the tenth tergite is the more complete expression 

 of the same tendency indicated in the mantids, where the true supra- 

 anal plate, though present, is reduced and entirely covered by the 

 tenth tergum." In Tettigonidae, Gryllidae, and Acrididae, on the 



