88 University of California PuhUcations. [Entomology 



tion of generation, or may become associated Avith sex, pro- 

 ducing the so-called secondary sexual characters, together with 

 sexual dimorphism and such phenomena. Differences much 

 less than these, if they are such as to prevent crossing, are 

 sufficient to differentiate species. 



In the formation of higher groups, while there is no similar 

 criterion of separation such as that which decides the difference 

 between species, there is the same independence of the amount 

 of divergence between the members of the grovip, and the 

 division is determined solely by the extent of the gap sepa- 

 rating the group from its neighbors. No matter what may 

 have been the theory of systematists, it has been the uniform 

 practice in every department of biology to ignore the question 

 of phylogeny in the limitations of groups of all ranks, and 

 base them wholly upon the degrees of separation recognizable. 



In the arrangement of the groups thus obtained the attempt 

 is usually made, with greater or less consistency by modern 

 naturalists, to indicate the lines or times of their development. 



Applying these doctrines to the subject of ve'nation, we must 

 conceive that the first condition of the wing was one in which 

 we could recognize but a single type of structure — one subject 

 to considerable variation. Among these variations would exist 

 forms suggesting types that later became separated. Every 

 type of venation now existing is historically connected by 

 regular gradations with the primitive form. If all of these 

 intermediate forms existed at the present day we should have 

 to deal with an organ whose variation was great, but at the 

 same time of little significance. Only the fact of the isolation 

 of groups by the suppression of intermediate conditions gives 

 the variation taxonomic value. 



With structures, as with species, there are thus two problems 

 for study that should not be confused. One is the delimitation 

 of groups, and the other is the phylogeny of these groups. There 

 appears to be much reason for Scudder's Pahcodictyoptera. 

 Certainly, as far as the venation is concerned, there was not in 

 the Paleozoic time sufficient differentiation between the various 

 types of venation to consider them comparable with the divisions 

 between orders at the present day. We may consider that at 

 the close of the Paleozoic era there was only a single order of 

 winged insects, though the name Neuroptera seems preferable 

 to Piila'odictyoptera. At the same time the phyletic lines of 



