bV R. .t. TiLLYARD. ' /0< 



in any of the lower Plaiiipeiinia; the terniiiial gradate series is 

 absolutely peculiar to the Faychojjsidw. 



Of these series, only the discal is invariably present. The 

 internal series is usually present also, but not always. The costal 

 and terminal series are later developments, confined to the larger 

 species, and evidently correlated with the greater expansion of 

 the areas which they help to support. Further, there is great 

 \ ariability in the condition of these series in almost all species. 

 As an example of this, T would point out, in Text-fig. 1, the 

 presence of a single cross-vein in the forewing between the 

 anterior ends of the internal and discal series. This is frequently 

 absent; but, in some specimens, it may be replaced by either 

 half of an extra series, in others by a complete series of as many 

 as thirteen cross-veins, crossing the disc between the internal 

 and the discal series ! In one case, I noted the presence of this 

 extra series upon one forewing of a specimen, while it was absent 

 upon the other I It will be at once obvious that attempts to 

 define genera on such variable characters as these, are foredoomed 

 to failure. 



In conclusion, it is interesting to note that, regarded as a 

 whole, the wing of Psychopsis presents a very close analogy in 

 structure to the feather of the wing of a bird — an analogy which 

 is also suggested by such wings as that of Drepant'ptery.n^ but by 

 no means so closely. The vena triplica corresponds with the 

 rachis, forming the main support or midrib in both cases; the 

 position of this support is the same in both cases, viz., nearer to 

 the anterior than to the posterior margin. The numerous veins 

 radiating out from both sides of the vena triplica are the 

 analogues of the barbs of the bird's feather; and the series of 

 connecting cross-veins, or gradate series, are analogues of the 

 barbules, which serve to connect the barbs together. 



Classification of the Family. 



We are now in a better position to consider the true value of 

 the various venational structures which have so far been used in 

 subdividing it into tribes, and in defining genera. 



