270 STUDIES IN AUSTRALIAN NEUROPTERA, IV., 



any of the numerous well-defined families cut off from it. I 

 found, everywhere, that there seemed to exist no scientific con- 

 ception of what the family Henierohiid(e meant; no attempt has 

 been made to study its venation thoroughly; and the authors, 

 who used the family, either offered short, non-committal defini- 

 tions, which utterly ignored the fundamental characters of wing- 

 venation [see, for instance, Sharp(13), p.465], or merely used* 

 the name as a headline under which various genera might be 

 placed, without troubling themselves why or wherefore. It was> 

 therefore, of the utmost importance to carry out a thorough in- 

 vestigation of the venation of the genus Hemerohius and its true 

 allies, and to discover a clear and comprehensive definition for 

 the family Hemerubiidfe, which should not only embrace all the 

 members of one natural group, but should also exclude the 

 extraneous material that had, for so long a time, been allowed to 

 drift in and out of the family, with no apparent reason, other 

 than neglect of the fundamental principles of classification. 



The results of a close study of all the forms available to me 

 were originally collected together with a view to publication as 

 an extended introduction to this paper. I found, however, that 

 it would be necessary to refer back continually to the general 

 discussion, when I came to deal with the separate families. In 

 order, then, to save much repetition, I have decided to give the 

 main arguments under each separate family, where they will now 

 be found. Those who desire to trace the fate of the family 

 Hemerobiidce, its gradual narrowing down, and my own attempt 

 to give it definite form and function, as a unit embracing only 

 genera having true phyletic relationship with the type-genus 

 Hemerohius, will find their information under this particular 

 family, on pp. 282-293. Similarly, the vagaries suffered by Ithone 

 and Sisyra will be found under Ithonidre (pp.274-279) and 

 Sisyridcei^^.312-3U), together with analyses of the types of wing- 

 venation found in these families. In dealing with Spermoj^hor- 

 ella, gen.nov., I had to choose between forming a new family 

 for it, or including it within the Holarctic Berothidce. I chose the 

 latter, for the reasons given on pp. 3 15-3 16. No apology is needed 

 for the formation of the new family Trichomatido' (pp.324-32f)), 



