48 Annals Entomological Society of America [Vol. XV, 



paragraph a deep V-shaped notch separates the two coxites, 

 though these are at all stages fused with the sternite. In 

 all these respects the Mantid n^^mphs resemble those of the 

 Tettigonoidea. 



The genitalia (PI. XI, Figs. 105-107) are strongly asym- 

 metrical and consist of two paramere lobes, whose relative 

 positions are similar to those of Periplaneta. 



The dextral lobe is very much smaller than the sinistral, 

 which it overlaps slightly, being little developed except dorsally. 

 It may be compared in some measure with that of such Blattids 

 as Blatella and Parcoblatta. A slender basal ventral piece 

 (pmb) evidently represents the basal apophysis of the Blattids, 

 although not actually internal; and as in Blattella and Parco- 

 blatta, though less distinctly so, it is bifurcated at base, the 

 lower branch curving sharply around and terminating in the 

 small, heavily chitinized ventral prominence, while the other 

 branch ends in a similar process just above it. These two 

 structures serve as a clasper in copulation, as I have ascertained 

 in the case of a female specimen, whose genitalia are still con- 

 nected with those of a male, although all but the end of the 

 abdomen of the latter is missing, having doubtless been eaten 

 by the female, as is the habit of many Mantids. The clasper 

 is possibly homologous with that which appears as such in some 

 Blattids, such as Periplaneta. 



The upper jaw of the clasper is continuous above with the 

 large, dorsal, shelf-like flap (df) which is readily compared with 

 the dorsal plate (base of shaft) in the dextral paramere of 

 Blattid£e (cf. Periplaneta fnliginosa, Fig. 100). 



The sinistral paramere lobe bears a distinct resemblance 

 to that of Periplaneta, especially as seen from the left side 

 (cf. Figs. 95, 108), in which position parts representing the 

 dorsal and ventro-lateral processes of Periplaneta are seen, 

 though in the case of the Mantid the latter would be more 

 appropriately termed the ventral process, as its broad chitinized 

 base forms the entire ventral surface of the lobe and extends 

 far over upon the right side, where it meets the narrow ventral 

 portion (apophysis) of the dextral lobe. The process itself is 

 bent dorsad in Stagmoma?itis, but in Mantis religiosa, in which 

 it is much longer and more slender, it is directed dextrad. 

 The dorsal process curves strongly to the left and in S. Carolina 

 is bifurcate. It appears to arise from beneath the dorsal plate, 



