138 THE CANADIAN ENTOxMOLOGIST 



unmcdified 8th sternum, and in the details of the ovipositor, which are yet to 

 be described. The immature ovipositor is of a distinctly more primitive type 

 than that of the most generalized Orthoptera of corresponding stage, e. g., 

 Ceiithophilus of the Tettigoniidae. This is seen in the form and position of the 

 valvulse, the presence of well-marked styli and the more definite basivalvulae. 

 In these respects it approaches the Blattida? and Mantidaj, but the valvula? 

 of the 9th segment have not the terminal position, nor have the dorsal valvulse 

 the broad, flattened form found in these groups, in which respects they are 

 apparently the more primitive, at least at this stage. The great reduction or 

 complete loss of the ovipositor in the Panplecoptera is, of course, a secondary 

 feature, in which thev are negatively specialized, as compared with the 

 majority of Orthopteroi 1 groups. 



The male genitalia considerably resemble those of certain Blattoidea and 

 Mantoidea, and also the Phasmoidea, and the fundamental plan of structure 

 seems to be the same in all of these groups, although I believe the asymmetry 

 has been independently acquired in some of them, at least. The true Orthoptera 

 seem at first sight to be constructed upon an entirely difTerent plan, but further 

 investigation tends to indicate that it is a very highly modified form of the same 

 plan. 



It is my present opinion that the "Panisoptera" (Blattoidea, Mantoidea 

 and Isoptera), which is unquestionably a natural assemblage, and the Orthoptera 

 together with the Phasmoidea, represent two main branches of the same stem, 

 originating as a section of the Palaeodictyoptera, and that Grylloblatta is the 

 sole survivor of a twig which separated from this stem before the two main 

 branches had become differentiated. The Phasmoidea also separated very 

 early from the Orthopteran branch. The characters in which Grylloblatta and 

 the Phasmoidea resemble the "Panplecoptera" are all primitive ones, and only 

 indicate the common origin of the two stems. 



The five-jointed tarsi, which are present in Grylloblatta, the Blattoidea, 

 Mantoidea and Phasmoidea, are characteristic of primitive Orthopteroid insects, 

 but I agree with Crampton's suggestion that the trimerous type may be, after 

 all, the more primitive for the Class Insecta in general. We never find penta- 

 merous tarsi in the Apterygota, nor in any of the Panplecoptera, so that they 

 were probably acquired very early in the Orthopteroid stock, and probably also 

 in other branches of the Palaeodictyoptera. The reduced number of tarsal 

 joints in the Isoptera and true Orthoptera is doubtless a secondary feature. 



Explanation of Plates VIII and IX. 

 Ahhreviations. 

 atg 10 — arm of 10th tergite. pm — chitinous process of right lobe, 



bc^ — basipodite of cercus. of phallus, 



bs — basivalvula. pr — paraproct. 



c — cercus. sa — supra-anal plate. 



cxr, cxl — right and left coxites. st 9 — 9th sternite. 



cxp— process of right coxitc. tg 9, tg 10 — 9th and 10th tergites. 



es — eversible sac. vd — dorsal valvula. 



Im — laminae forming part of clasping vf — valvifer. 



apparatus. vi — inner valvula. 



Ir, 11 — r'ght and left lobes of phallus. vv— ventral valvula. 



