98 THE CANADIAN ENTOMOLOGIST. 



of the head, but resembles the Tettigonidae much more closely in 

 the structure of the ovipositor the inner valves of which, as in this 

 family, are nearly as long as the other two pairs. The antennae, 

 however, in length and in number, size and form of the segments 

 are most like those of the Locustida* (Acrididae). 



The presence of the Tettigonid type of ovipositor in so primi- 

 tive a form as Grylloblatla seems to indicate that this type of ovi- 

 positor was present in the earliest saltatorial Orthoptera and that 

 the Tettigonidae are therefore, in this respect, the most primitive of 

 the modern saltatorial families. 



If this conclusion is not accepted, then one at least of the 

 following assumptions must be made: 



(a) That Grylloblatta is a degenerate descendant of saltatorial forms, 



which, like the mole-cricket, has secondarily lost its power of 

 leaping. 



(b) That the adaptations for leaping have been independently ac- 

 quired in the Tettigonidae on the one hand, and the other 

 families of saltatorial Orthoptera on the other, i.e., Gryllo- 

 blatta and the Tettigonidae represent a distinct line of descent 

 from those which have culminated in the Gryllidae and 

 Locustidae. 



(c) That the resemblance between the ovipositors of the Tettigon- 

 idae and Grylloblatta is due to convergence. 



The first assumption is clearly untenable on account of the 

 evident relationship between the Grylloblattidae and the other non- 

 saltatorial Orthoptera, particularly the Blatticke, with which they 

 share the 5-jointed tarsi and jointed cerci, both of these being 

 primitive characters not present in any of the saltatorial Orthop- 

 tera. 



The second assumption is equally inadmissible, as the families 

 of saltatorial Orthoptera form a decidedly natural group, the 

 Gryllidae and Tettigonida' being particularly closely allied to one 

 another. 



The third assumption is less improbable, but the resemblance 

 between the ovipositors of Grylloblatta and the Tettigonidae is 

 very close, and it seems more natural to regard them as truly re- 

 lated in this respect. 



