47 
which the Noctuidae would probably be the nearest. Except for the 
Geometridae the Notodontidae are the most perplexing from their 
neuration. Apart from their general appearance that suggests the 
Sphinyidae, the neuration has precisely the same characteristics, includ¬ 
ing in two or three species the unusual short bar connecting vein 8 to 7 
in the hindwing, and the position of vein 5 which arises at equal 
distances from 4 and G. It is extremely interesting, however, to find 
that we have a link for this latter character so that we can trace even 
neurationally connection with the Noctuids. The genus Cyphanta, 
which contains two North Indian species, shows vein 5 arising almost 
from the corner of the cell with vein 4, but still has the small 
connecting bar between 7 and 8 of the hindwing, and other traits 
characteristic of the Notodontidae. It thus appears that, although no 
doubt can now be held in regard to the position of this family, viewing 
the case from all sides, there must have been further links that have 
died out and thus have made our task the more difficult. If it had not 
been for the knowledge of the embryological stages we should probably 
never have departed from the old idea that Sphingids and Notodonts 
were closely allied, and more than likely never have looked for a 
relationship in the neuration. But I am on ground on which I said I 
would not detain you. We already have grouped our insects into large 
divisions or families by essential structural affinities shown throughout 
their existence. If a structure in one stage is at variance with what 
we have laid down as rules, it must be that that particular structure 
is not fixed enough for family diagnosis, and is only suitable 
for minor subdivisions, such as we know as genera or even for species. 
It may be also that we interpret it wrongly, as has been the case with 
the neuration, for want of fuller knowledge of the insect’s existence. In 
the case of families it is found that no single character will hold for 
definition. It is also true of genera in a still greater degree. Inde¬ 
pendent working on any single character invariably breaks down in the 
face of other characters. The minor structural parts are, particularly in 
the imago, so numerous, that we find we are able to give genera very 
varying interpretations. We make these groups perhaps by the antenme 
of the imago differing only from those of another group, or, may be, the 
palpi, the leg structures, or the presence or absence of ocelli; in the early 
stages also, where any character holds for a number of individuals, we 
give a name to that group and call it a genus. Now it seems to me 
that although this is perhaps useful from the point of view of conveni¬ 
ence, it is not necessarily a statement of the true genealogy of the 
species. My point is this: x411 structures, even minor structures, are 
not of the same fixity. We can define some genera much better than 
others, some have a larger number of recognisable characters than 
others. We often speak of a very natural genus, meaning that the 
genus is well differentiated from all others. But in these latter days 
of entomology Avhen large genera are split up again and again by slight 
structural differences, surely these cannot he of the same rank as the 
well-defined or very natural genera ? If they are not of the same rank 
they should not go by the designation, and I am glad to see there are 
some entomologists who are recognising this, and dividing genera into 
sections A, B and C, or more according to the facts of the special case. 
The definition of a genus is hard if not impossible, but the cases are 
considerably lessened by the system of sections. The conditions for 
