4 
In the same year (1890) J. B. Smith published his Revision of the Species 
of the Genus Agrotis (Bull. 38, U. S. N. M.) ; the system of classification 
proposed is vastly superior to any of the previous ones. The species are 
primarily separated on the armature of the fore tibiae and further sub- 
divided on characters such as wing-shape, tuberculate front, thoracic and 
abdominal tufting, male antennae, shape of palpi, and genital structure of 
the male. A number of new generic terms are introduced, the author 
himself stating that he “felt free to propose new names for the groups into 
which the species have been formed, quite satisfied to have the names 
relegated into the patient synonymy when the study of generic types 
proves their identity with others already proposed.” Smith’s classification 
was largely followed in Dyar’s List of North American Lepidoptera (1902, 
Bull. 52, U. S. N. M.) and also by Grote in his List of North American 
Eupterotidae, etc. (1895, Bremen), who, however, attempted to synonymize 
a number of Smith’s genera with older existing names, notably those of 
Hiibner’s Verzeichniss , and treated the majority of them in a subgeneric 
sense only. 
A very useful and important contribution to our knowledge of the 
group was made by P. C. T. Snellen in a paper entitled Eenige Opmerkingen 
over de nadere verdeeling van het genus Agrotis , appearing in the Tijdschrift 
voor Entomologie, 1896, page 142. The author makes no attempt at 
generic divisions, but is content to indicate the groupings into which the 
European Agrotids fall, based on a very careful study of numerous struc- 
tural details. Putting aside his first subdivision, based on the colour of 
the hind wings, as rather unscientific, his remaining sections show a very 
careful attention to the minute differences found in the spining of the fore 
tibia and its length as compared with the first tarsal joint, the varied 
vestiture of the thorax and palpi and the shape of the third palpal joint, 
and in the final instance the nature of the male antennse. As far as the writer 
has been able to check the work, Snellen’s grouping is the most natural of 
any proposed up to this time. 
Hampson’s work (1903), which has already been referred to, is the 
first that deals with the genera and species of the world and his generic 
keys are based on a combination of structural characters used by previous 
workers and numerous other ones employed for the first time. The least 
satisfactory feature of the work is the arbitrary selection of the genotype 
as the first species included in the original description and the consequent 
clashing with the International Code of Zoological Nomenclature. Though 
it was recognized that a careful generic revision would sooner or later 
prove a necessity, Hampson’s arrangement was more or less closely adhered 
to in the Barnes and McDunnough Check List, 1917. The following 
generic key has been adapted from Hampson and includes all the genera 
to which North American species are assigned; it will give a good idea of 
the structural characters employed by this author in separating genera. 
KEY TO NORTH AMERICAN AGROTID GENERA ACCORDING TO HAMPSON 
(1) Fore tibiae with spines or claws 2 
Fore tibiae without spines or claws 12 
(2) Frons with prominence . . 3 
Frons without prominence 5 
