OF WASHINGTON. 41 



If I say an insect belongs to the genus Cicindela, anyone who 

 has ever collected insects knows what I mean ; but let the genus 

 be divided, and sub-divided again and again, and Cicindela 

 will mean nothing, except a type-species. Emphasizing the 

 genotype means splitting genera; emphasizing the generic 

 characters means better-defined genera. But what characters 

 are of generic value ? No one man can answer with authority, 

 yet we may consider a few points. 



CRITERIA OF SPECIES. 



First, what are the structures that should define species? 

 The nearest we can come to the criterion of a species is its 

 capability of interbreeding. Therefore the structure of the 

 genital organs, or such secondary characters as are associated 

 with them, must be the best specific characters. This includes, 

 of course, all recognition marks or structures. No one, I 

 think, studying insects to-day, can long resist the conclusion 

 that sexual characters, although often more or less variable, 

 and at times but slightly developed, are the fundamental char 

 acters of species. This vital difference in genitalia may find 

 expression in more visible signs, recognition marks, index 

 characters, by which the insect may know its mate. These are 

 of many sorts, such as punctuation, color, sculpture, etc. With 

 these necessary differences there may go accidental differences, 

 which, although not essential, may aid in the separation of 

 species. For example, in one fly a wing-cell may be longer than 

 in a closely allied species. This character cannot be essential ; 

 there must be accompanying it differences in genitalia or 

 recognition marks, or some point of vital interest to the fly. 

 These accidental specific characters may be of much use in 

 the tabulation of species, but we must not lose sight of their 

 intrinsically slight value. 



WHAT CHARACTERS ARE OF GENERIC VALUE? 



Genera should not be based on the characters used in the 

 separation of species, no matter how widespread or constant 

 they may be for groups. Although we expect that the species 

 of a genus shall have a certain uniformity in genitalia and 



