36 



tural features are not to be considered " well-marked," as, for 

 instance, the ordinary range of antennal peculiarity, slight changes 

 in shape of wing, tibial armature, etc. In fact all the characters 

 which, when tvcll-marked, are of generic importance, are liable to 

 slight modifications which are only specific. It is the kind, con- 

 stancy or amount of these modifications which decides the class in 

 all our artificial divisions of these natural objects. This really 

 decides i?tdividuals, varieties, aberrations, species, sub-genera, genera, 

 sub-families, families, sub-orders, orders and classes throughout the 

 animal kingdom. 



In the antennae of moths the range of variation is so great 

 that it is only upon an unusual departure (e.g., Renia) that we 

 can use them as generic characters. They are different in the 

 sexes, and here we come upon the question of secondary sexual 

 characters. It is clear that when the departure of one sex is as 

 strong as usually demanded by generic differences, we should use 

 a distinct title. Such genera are not equal to ordinary genera, 

 where the characters are shared by both sexes ; but neither are 

 all species equally strongly distinct. Just as on a former occasion 

 I drew attention to certain assemblages which I called PRO- 

 GENERA, of which Datana is an example, so here I call such 

 genera as are founded on peculiarities shown oftenest by the male 

 sex, ANTIGENIC. The genus HeliocJiilus is a case in point. 



Here we have such an unusual structure of <5 wing that it 

 has but but few parallels in the Sub-order. EiiscirrJiopterus is 

 another case in point. Now, it is clear that to throw Heliochiliis 

 into Heliothis is to contradict the diagnosis of the latter and to 

 prevent our recognition of a peculiar structure. There is another 

 species of //r//t'(;Z:////'5 apparently found in Asia. To refer both 

 of these io Heliothis is to obscure the facts which all our classifi- 

 cations are intended to make plain. We must not lose sight of 

 the fact that we are trying to fit plastic structures into unyield- 

 ing categories, " to put nature into a straight jacket," as Dr. 

 Packard calls it. Evidently we must increase our categories. 

 The number of purely antigenic genera is small, but of those of 

 a mixed character, in which features drawn from one sex alone 

 (more often the male), are important in the diagnosis, the number 

 is considerable. In Arzama, among the Noctuidce, the female 

 has a secondary sexual character in the large anal tuft. The 

 ordinary genus, as I have considered it in the moths, takes in 

 species agreeing in well-marked characters, allowing a certain range 

 in minor respects of structure. But the moment we put together 

 species with well-marked differences, we cover up the study of 

 these differences. Here is where the tact and talent of the student 

 come into play to decide upon the limits of ordinary genera. 

 The question, under the rules here presented, is really simplified 

 and the cases for a difference of " opinion " reduced to a mini- 

 mum. A generic character must be decided upon with a knowl- 



