NO. 2041. THE INSECT FAMILY DELPHACIDAE— CRAWFORD. 559 



reason. In all this the personal element is very strong in deciding 

 what constitutes a genus and what a species. The number of keels 

 present on the head and thorax is a very reliable character. The 

 form of the hind tibial spur (calcar) is also a constant and easily 

 appreciated character, but unfortunately it has been overlooked by 

 many and not described. The relative lengths of the tibiae and tarsi 

 was used by Ashmead as a generic character, but this is impossible. 

 The number of lateral spiniform spurs or spines on the hmd tibiae, 

 however, does afford a good character for separating some genera. 



The female genitalia are very similar throughout the group, except 

 in one genus, Stenocranus, but the genitaha of the male present good 

 specific characters. No doubt the latter might be used for generic 

 diagnosis, as Earkaldy has done quite extensively, but I have hesi- 

 tated to use them for such simply because of the difficulty of identi- 

 fying an isolated female. 



Coloration, when rightly used and ample allowance made for 

 variation, is a good accompanying specific character, useful more for 

 final identification than for synopses. Unfortunately, it has been 

 used too commonly as an unaccompanied specific character, the result 

 of which has been confused synonymy. 



In the large group of genera more or less closely related to the old 

 genus "DelpJiax" or Liburnia we find it stiU more difficult to discover 

 good usable generic criteria. There are plenty of good variational 

 characters which wiU distinguish species, but it is very difficult to 

 resolve these into generic groupings. If one were working with a 

 more hmited collection, I have no doubt that he would easily be able 

 to separate and relate subgroups, as has, in fact, been done already. 

 But with a collection large both in species and in number of specimens 

 in a species one is sure to find species and individuals which fall 

 between two estabhshed genera in tliis group. The cause of this is 

 evident. Genera have been estabhshed on characters winch com- 

 pletely mtergrade, and the dividing fine has depended solely on the 

 personal judgment of the various students. Furthermore, it is 

 impossible for one man to convey to another, either by writing or 

 illustration, exactly what he has in mind by such arbitrary ideas as 

 "keels distinct" or ''keels evanescent," and where he draws the line 

 between the two. 



There is such a thing, too, as ''sphtting hairs" when it comes to 

 classification, that is, drawing the confining fines of generic groups 

 and species so narrow that it precludes any variation and results in 

 numerous genera and species with the minutest and often absurd 

 differences separating them. 



After a very careful study of the large series before me, and with 

 the above ideas in mind, I have arrived at the only logical conclusion, 

 namely, that many of the estabhshed genera are not warranted and 



