49 
from my point of view they are not of necessity cause for making a 
genus of them, but as merely a section probably of another genus. As 
we shall presently see with species, we are too often, in our exactitude, 
led away into subdivisions which are of themselves inexact because the 
divisions, all going under one name have such a varying factor of fixity. 
It is already stated, and it must be obvious to every scientific worker 
that it is impossible for all our genera to have their being under the 
same conditions. Markings of egg, minor structural differences in 
larva, pupa, or imago, complete difference in habit, such as the larva 
of one species being an internal, the other an external feeder, are all 
severally or by combinations of two or more of these traits, held to be 
of sufficient specialisation for the formation of a genus. But the 
internal larva calls for a different structure from the external, and there¬ 
fore this habit is, or should be, only considered under the larval 
structural character and not stand alone as of value. As is true of 
everything in nature there is gradation, and we actually have lame 
that are external for part of their existence, and internal for the other 
part. Leucania brevilinea is a case in point which has a larva that feeds 
within the reed stem until it is attaining towards full growth, and then 
feeds on the leaves, but always still hiding by day within the reed. 
At first glance it would appear as if this insect was an internal feeder 
and of that stock, and was gradually becoming an external feeder, from 
the fact that generally we know of internal feeding larvfe as being 
generalised in structure. In the first place, the true or absolute 
internal feeding Noctuids, such as Nonagria, are not the lowest Noctuids, 
and it must be assumed that their internal habit is a recent one called 
into use for protection’s sake. Certain traits correlated with generalised 
internal feeders, such as a delicate cuticle and great elongation, are 
brought about from the habit of feeding within long narrow reeds.' Of 
course if one goes far enough back one must believe that these rather 
specialised internal feeders were remotely descended from internal 
feeders of a generalised type. To come back then to Leucania brevilinea, 
if it has not quite reached perfect protection by absolute internal feed¬ 
ing it has not advanced so far as true Nonagrias which spend their 
whole larval (and pupal) existence within the stem. Leucania phraq- 
mitidis has reached further than L. brevilinea, it feeding wholly within 
the stem, and only coming out to pupate in moss near the root. L. 
lutosa has advanced still further for protection’s sake. It feeds wholly 
within the stem and feeds down to the root where, it emerges to pupate 
without any wandering for a suitable place for pupation. 
It is obvious then that upon the feeding habit, as enunciated, we 
have not very clear grounds for working with. Are we to have two 
genera dominated by the characters of the two extremes, and to fit in 
all the intermediates according as the greater number of characters 
leans towards ? Or, again, shall we place all the intermediates in one group 
between the two as has even already been done ? What we have to 
answer in reality is: did those two or three groups have a common ances¬ 
tor more recently than the family relationship? If they did then they 
should have some grouping distinction, whatever it may be. A genus 
should occupy the same position to species as a species does to the various 
forms which are bred from a single individual which are known as varie¬ 
ties. As soon as any of those varieties become definite and fixed to pro¬ 
duce their kind they become species. If then we can satisfy ourselves 
