1922] Walker: Structure of Orthopteroid Insects 63 



primitive than in these orders, but this may be due to adap- 

 tation to a Hfe in passages or galleries, in which the flattened 

 form of the Blattids would be a disadvantage. The thoracic 

 sclerites are simpler and in some respects apparently more 

 primitive than those of the Blattids, as shown by Crampton, 

 but these more primitive features may have been present in 

 extinct cockroaches. The similarity of the two pairs of wings is 

 evidently secondary, due to reduction of the anal area of the 

 hind wings. This is indicated by the distinctly expanded anal 

 area of the hind wings in the primitive genus Mastotermes, in 

 which the venation also approaches more closely the Blattid 

 type. On the whole, however, it is perhaps best to consider 

 the Isoptera as springing from the Protoblattoidea, a palaeozoic 

 order from which probably all three orders of the Blattoid 

 group were developed. As far as the terminal abdominal 

 structures are concerned the Blattoidea and Mantoidea are 

 more like one another than are the various families of 

 Orthoptera. 



The Zoraptera appear to have affinities with the Gryllo- 

 blattoidea and the three orders just discussed. Like the former 

 they are plecopteroid in the structure of the cervical sclerites, 

 even more so, on account of the presence of a dorsal sclerite in 

 this region (Crampton, '20). The thoracic sclerites show 

 points of resemblance to Grylloblatta and the Isoptera, while 

 the coxse are large and much like those of Grylloblatta. The 

 asymmetrical penis also apparently belongs to the type char- 

 acteristic of the group of orders we have been discussing. On 

 the other hand there is a loss in the male not only of the styli 

 but of the 9th sternum itself (unless it be represented by the 

 membranous lobe shown in Crampton's figure) (1. c, PI. VII, 

 Fig. 2), a structure otherwise so conspicuous in this assemblage 

 of groups. The lack of modification of the 7th abdominal 

 sternum as a subgenital plate places it nearer Grylloblatta than 

 the other three orders, but Crampton finds important points 

 of resemblance with the Isoptera. The reduction of the tarsal 

 joints to two and the cerci to a short unsegmented structure, 

 and the modification of the anal plates are peculiarities which 

 set this order somewhat apart from the others, but I should 

 place it tentatively as a branch arising near the point of 

 divergence of the Grylloblattoidea from the Protoblattoidea. 



