BY R. J. TILLYARD. 361 



show the subcosta reduced to a few veins — an intermediate stage 

 between Phyloblatta (Text-fig. 29a) and Rhipidoblattina (Text-fig. 

 29/>). This latter genus belongs to the Mesoblattinidce, which, 

 represented by only three genera in the Upper Carboni£erous ) 

 become more important in the Permian, and finally reach the 

 dominant position among Mesozoic Blattoids. 



Let us next consider the radius (R) In the type exhibited 

 by Phyloblatta, we can usually indicate the boundaries of the 

 main stem of R (Rj and the radial sector (Rs) respectively. 

 Thus these, cockroaches fall into line with almost all other insects 

 in this respect. But when the radius supplants the subcosta, as 

 in the Mesozoic and recent forms, it tends more and more to 

 take on the original character of that vein, viz., a single, gently 

 curving main stem, giving oft' only anterior branches to the wing- 

 border. There are many forms in which this process can be seen 

 uncompleted; yet, in general, we may say that the limits of Rj 

 and Ks become no longer clearlv evident, and it is best to con- 

 sider the radius, for purposes of description, as a single main 

 stem with many anterior branches. 



The media (M) tends to alter very little throughout all Palaeo- 

 zoic and Mesozoic forms. It always supplies about the same 

 area of the wing, viz., a more or less triangular area covering the 

 middle line of the tegmen, including the apex and portion of the 

 wing-border below it. In the recent form figured (Text-fig. 29c), 

 the media is shown much reduced, and fused basally with the 

 radius. A reduced media is found in a few Mesozoic forms, but 

 its basal portion is never fused with R in these fossils. There 

 are also many genera still existing in which the media remains 

 free and many branched. 



The cubitus (Cu) also tends to alter very little. In the great 

 majority of forms it supplies the broadly triangular area lying 

 below the middle line, sending branches to the posterior border, 

 from below the most posterior branch of M to near the anal area. 

 The manner and number of these branches is very variable in 

 all forms. 



The most characteristic structure in all Blattoid tegmina is 

 the sharply marked-off anal area or claims (Text-fig.29, aa). This 



28 



