BY R. J. TILLYARD. 125 



in size, and of similar shape; probably, therefore, the venations 

 of fore- and hind wings were the same. 



(2) The insect had a small, elongated head and prothorax. 



(3) The head carried a pair of straight, slender, and fairly long 

 antennse, and also a pair of shorter, but very prominent, some- 

 what curved, slender appendages, which Bolton considers, with 

 some doubt, to have been the mandibles. 



(4) The three pairs of legs were well developed, and placed 

 far apart from one another. 



(5) The abdomen was also moderately short, and probably 

 carried two very short cerci. 



In the Perlaria, the head and prothorax are neither narrow 

 nor elongated, but are always more or less flattened down dorso- 

 ventrally, more or less widened; and the prothorax is never far 

 removed from the pterothorax. The antennae are very long, and 

 the cerci also, in all the most archaic forms; genero, in which the 

 cerci are shortened are demonstrably descended from forms which 

 had longer cerci. Further, fore- and hindwing are never equal 

 in size, or of similar shape; and forms in which there is an ap- 

 proximation to equality are demonstrably derived from forms in 

 which there has been greater inequality. Nor are the venations 

 of fore- and hindwing ever the same, but differ fundamentally, 

 as a study of the nymphal tracheation of the two wings clearly 

 proves. Finally, the Perlaria have mostly reduced, weak, non- 

 projecting mandibles; and it is demonstrable that these organs 

 were never, within the limits of this Order, slender and project- 

 ing, as seen in Megagiiatha. 



Even if it were to be admitted that the restored venation in 

 Bolton's Fig. 4 were correct, I fail to see in it any true Perlarian 

 characters. It much more resembles the venation of an archaic 

 Termite forewing, such as Mastotermes. 



I conclude, therefore, that no affinity between MeijayiuUha and 

 the Order Perlaria can be shown to exist. 



The only possible claim to affinity with the Order Megaloptera 

 rests upon the projecting mandibles, since these structures also 

 project strongly in the a,H!:haJ\c Corydalus and allies. But the 



