20 The Ohio Journal of Science [Vol. XXI, No. 1, 



onine penes were " Calopterygid " rather than Coenagrionid. 

 Building up from such discoveries the writer has been forced to 

 the conclusions presented in this sketch. 



The primitive Zygopter must have been a small insect no 

 larger (if as large), than Ortholestes, Miocora, Argiolestes, or 

 Micromerus that had a reduced venation, in which there were 

 probably but two antenodals, in which M3 and Rs arose near 

 the subnodus, in which there may or may not have been a few 

 more extra sectors than Mia. That the breathing by caudal as 

 well as by ahdominal gills is a specialized method, the primitive 

 Odonate method being by rectal gills.^ That the development 

 of the families and subfamilies was radiate and not serial. 



These conclusions reverse some of our previous views. The 

 nodus has passed distad in the richly veined Calopterygine series 

 instead of having passed basad in the "reduced" Coenagrionid 

 wing. The forks of Ms and Rs have changed very little, which 

 agrees with venational studies in other orders. The wings have 

 increased in size merely by the lengthening of some or all the 

 veins and by the addition of extra sectors. Thus the nodus is 

 merely the apex of Sc, which has been free to lengthen just as its 

 sister longitudinal veins have lengthened and so has moved 

 out as the wing increased in area, which has made the forks of 

 Rs and M3 appear to have moved basad more than they really 

 have. 



Interwoven with these are the two elusive factors : 



1. Heredity, which tends to hang on to old structures; and 



2. Orthogenesis, which may increase a tendency beyond 

 the actual needs of the insect. 



The writer hopes that his friends will deal leniently with 

 these startUng innovations until he is able to present the evidence 

 in full. His own views have been completely reversed during 

 this study. 



The sixteen subfamilies of Zygoptera recognized by the 

 writer fall into four major groups, or families, the AgrionidcE, 

 LestidcE, HemiphlebidcB, and CcEnagrionidcE. The list of subfam- 

 ilies will be discussed as a list because the writer has been unable 

 to construct, without exceptions, a natural key based on either 

 wings or penes. Probably the major groups can be defined in 

 larval characters when these are better known. Drs. Calvert 

 and Tillyard are in possession of material that should solve this 

 problem. 



^ See page 23, 



