342 ENTOMOLOGICAL NEWS [Oct., '15 



The character of the antennal segments, the shape of the 

 head, the nature and position of the eyes, the general shape of 

 the mouthparts, many of the thoracic sclerites, certain of the 

 leg structures, and many other features indicate that the Gryl- 

 loblattids are very closely related to the Dermaptera. Fur- 

 thermore, the fact that in the immature stages the previously 

 mentioned Dermaptera have segmented cerci (which later be- 

 come modified to form the forceps) clearly indicates that the 

 Dermaptera are the descendants of forms having segmented 

 cerci similar to those of the Plecoptera. The presence of seg- 

 mented cerci (which are quite similar to those of immature 

 Karschiella} in the Grylloblattidae, therefore, instead of mili- 

 tating against the view that the Grylloblattidae are closely re- 

 lated to the Dermaptera, would, on the contrary, indicate a 

 community of descent between the two groups ; and would 

 simply show that the Grylloblattidae have departed less from 

 the common ancestral type in this respect, than have the Derm- 

 aptera and, indeed, the Grylloblattidae are more "primitive" 

 in many other respects than the Dermaptera are. On the other 

 hand, the presence of a well-developed ovipositor in the Gryl- 

 loblattidae, might be regarded as a more highly specialized 

 feature (although the fore-runners of ovipositors are to be 

 found in Lepisma, Machilis, and other primitive insects). This 

 however, does not affect the general primitive character of 

 the Grylloblattidae, since it is a matter of common occurrence 

 that animals which are very primitive in some respects, may 

 develop certain other highly specialized features. 



The Isoptera have apparently descended from ancestors hav- 

 ing segmented cerci also, and, indeed, the cerci of Archoter- 

 mopsis are composed of exactly as many segments as those of 

 Grylloblatta namely, eight. I had long suspected that the 

 Isoptera were rather closely related to the Dermaptera (as 

 well as to the Blattidae) and upon examining the various 

 structures of Grylloblatta, it at once became apparent that we 

 have in the latter insect a form annectent between the Dermap- 

 tera and the Isoptera (and leading to the crickets and katy- 

 dids). I am convinced that Grylloblatta is descended from an- 



