16 
Eyes. Until quite recently the Agrotids were considered to possess 
non-hairy eyes. A few southwestern species, associated with Trichorthosia 
parallela Grt., show, however, very strongly haired eyes, forming a distinct 
group which may have Mexican affinities. In the northern Rocky moun- 
tains there is a further species, Trichosilia acarnea Sm., with weakly haired 
eyes, which is closely allied to the typical Agroiis group in all other char- 
acters and is very evidently an offshoot from this genus. 
A few genera, notably those with rough, hairy vestiture, show lashed 
eyes; as such may be cited Metalepsis Grt. and Pachnobia Gn., in which 
latter genus the eyes are also smaller and deep-set, a feature often noticed 
in species occurring in high altitudes or northern latitude. 
Antennae. No use has been made by the writer of the various forms 
of the male antennae as a means of generic separation, except in the case of 
Feltia Wlk. and this only with certain misgivings. As a general rule the 
male antennae, though forming often a good specific character, show no 
correlation with other characters which would even warrant their usage 
for subgeneric divisions. 
Front. The roughened or tuberculate condition as compared with the 
smooth front has been largely used by Smith, Hampson, and others as 
one of the primary means of subdividing Agrotid genera. As far as the 
writer is able to judge by his present researches, this has been a great 
mistake and has been responsible for more misplacements of species than 
any other character that has been used. The gradual forming of a raised 
tubercle on the front, as in Euxoa Hbn., is apparently a fairly recent 
acquisition in the Agrotids and should by no means be used as an infallible 
character for generic separation. Undoubtedly it is very useful in deter- 
mining Euxoa species, but even in this genus, as the writer has shown in 
previous papers, occasional species occur without the tubercle which on 
other characters are undoubtedly Euxoas. 
In a single instance, viz., longidens Sm.,the writer has erected a new 
genus on the strength of the frontal structure, but in this case a very high 
degree of specialization is shown and the type of prominence is quite unique 
in the Agrotids and reminiscent of Nocloa and its allies. 
Vestiture. The type of thoracic vestiture appears to be entirely an 
individual character and possibly considerably influenced by climatic 
conditions. The writer has noticed that the species indigenous to cold 
regions exhibit a marked tendency toward a rough hairy clothing, whereas 
in warmer regions the vestiture is more likely to be composed of mixed 
hairs and scales or scales alone; occasionally, as in the case of Pachnobia 
Gn., a genus entirely limited to semiarctic species, the hairy vestiture may 
be used as a final means of separation from allied genera, but in general it 
is not a safe character to employ. 
The writer has emphasized the thoracic tufting as little as possible 
in his keys, and then generally in conjunction with some other character. 
It can only be accurately determined in perfect specimens; in a few genera, 
such as Graphiphora Ochs., Eueretagrotis Sm., and Aplectoides Buth, tufts 
on both pro- and metathorax are quite evident, but in general the tufting 
is obscure and at times apparently merely individual; in the hairy species 
it is not at all evident. 
