¥ 
s 
African Species of the Genus Nepltis. 583 
I have now described at length all the species or forms of 
Neptis which can with moderate certainty be distinguished. 
There remains a residuum of forms which seem to merge 
one into another in such a way as to render it impossible 
to separate them specifically even by the male armature. 
If we examine a series of preparations of the genitalia 
taken from these forms, and then arrange our preparations 
in the order of their resemblance, we shall find a great 
difference in structure between the first and the last of the 
series, but no satisfactory distinction between any two or 
three consecutive examples. Moreover, if we then arrange 
the specimens themselves we shall find that their patterns 
do not by any means follow the resemblance of the arma- 
tures. Thus, as already stated, the armature of strigata 
is indistinguishable from that of a form very near Holland’s 
continuata. Some examples of Neave’s conspicua have 
armatures which resemble those of typical nysiades suffi- 
ciently to warrant us in agreeing with Aurivillius, who 
regards them as conspecific. But one of Neave’s co-types 
of conspicua has an armature of a slightly different pattern, 
whilst the type of conspicua itself has the two claspers of 
the same individual very different in structure. It is 
impossible to deal on ordinary taxonomic lines with an 
assemblage of this kind. It 1s useless to decide that a 
given example is the same as Holland's continuata because 
it resembles the published figure of that form. We do not 
‘know the structure of the armature in the type. A 
specimen outwardly indistinguishable from conspicua may, 
as I have shown, possess claspers of a different structure, 
or, since the armature of the type specimen is asymmetrical, 
we might have two examples outwardly resembling con- 
spicua one of which had claspers like the right one of the 
type, the other with claspers like the left. 
In our classifications we attempt to work in what 
may be metaphorically termed watertight compartments. 
Nature, though frequently conniving at our methods, refuses 
always to be bound by them, and we must, I think, decide 
more freely to admit the existence of indeterminate forms. 
We have an outstanding example of this in the many 
indefinite forms which are comprised under the specific 
name Acraea acrita, as set forth in my monograph of the 
genus Acraea (Trans. Ent. Soc., 1912). I propose there- 
fore to give a description of a typical example of Neptis 
nysiades, to detail also the other named patterns which 
