BV R. J. TILLYARD. 



379 



which the common B. germanica (Linn.) is the genotype. A 

 similar condition also exists in the very fragmentary Mesozoic 

 fossil Pachyneuroblattina rigida Handl. (l, p. 433), in which also 

 the primary branches of B, as far as preserved, exactly corres- 

 pond with those of our new species. Tt seems a pity that Hand- 

 lirsch should have founded a new genus on such a fragmentary 

 type as this, especially as the thickening of the veins may not 

 really be of generic value at all, any more than it is in Samaro- 

 blatta. Until a more complete specimen is found, it would not 

 be possible to decide the amount of relationship between Pachy- 

 newoblattina and Samaroblatta. 



Samaroblatta intercalata, n.sp. (Text-fig. 38). 

 Greatest length of fragment, 10 mm.; greatest breadth, 33 mm. 

 A fragment of a left tegmen of probably about the average size 

 of those already described in this genus. 



\ ( 



Text-fig. 38. 



Samardblatta interccUata, n.sp. ( x ~h). Lett tegmen. Upper Trias of 



Ipswich, Q. Drawn from Specimen Nb.262. 



The anal area is missing, and the costal area not clearly pre- 

 served, except for a small portion of the costal border. The 

 double-curving of R is strong, and is the one definite character 

 that allows of the species being placed in the genus Samaroblatta. 

 The limits of Sc and R basal ly are not clear; what appears to be 

 the first primary branch of R is itself three-branched, a very 

 unusual condition for Samaroblatta. The next branch of R is 

 forked, then follow in order two simple branches, a forked 

 branch, another simple branch and a final branch making a very 

 definite forl< with the main stem: thus there are seven primary 



