LYCANIDAL. @ (2) 
in this work, and by Westwood, Distant, and most entomologists) as a subcostal nervule ; 
so that his descriptions in respect to the number of these nervules are still more in (apparent) 
disagreement with those in the following pages, 
In looking for some other feature whereby to define main groups, the filamentous ap- 
pendages which often appear on the hindwing naturally suggest themselves. These ‘‘ tails ”’ 
however are sometimes somewhat uncertain characters, as genera occur in which obviously 
very closely related species differ fer se in the presence or absence of these delicate fila, 
mentous appendages ; in fact it would appear that the same species (¢.¢., Wecaduba ardates and 
Megisba malaya), may have both tailed and tailless forms. Regarding, however, such deviations 
from the ordinary type as occasional and abnormal only, the presence or absence of tails, their 
number, form, and size, seem to serve better than any other single structural feature as characters 
uniting naturally allied groups ; and as such they are used as of primary importance in the 
following table. 
Here however it should be recognized that the importance of this character, as of any 
other, must not be exaggerated: nor must it be assumed to be perfectly constant. The 
classifier is apt to become more or less a slave to his system, and to attach too strong a value 
to his characters: thus he is inclined to say that one species certainly strongly resembles 
another, but that the resemblance can only be superficial, as the first has three and the second 
only two subcostal nervules in the forewing ; whereas the fact perhaps may be that the manifest 
resemblance is real and betokens true affinity, while the structural difference in respect to 
the subcostal nervules may be really unimportant, having a value more or less fictitious, 
and dependent only on the classifier’s arbitrary and artificial system. 
With these reservations as to real value and constancy of the characters used in the 
following table, and as to the artificial nature at the best of any linear “ index methodicus,”’ 
it may be claimed that this Key seems to bring together insects which are naturally allied. 
A general consensus of collectors and students seems to recognize as real and natural groups 
the true tailless “ blues,” tailled “ blues,’ and “hairstreaks ”; or alliances, of which Portia: 
Lycena, Polyommatus, Amblypodia, Thecla, Mjrina, and Deudorix represent the main types. 
It is really immaterial with which group the series commences or ends ; nor, for reasons 
already given, does it seem possible to place each group so that it shall fit exacily and truly 
between those which precede and follow. It will be seen that at one end stands the Gerydus 
group with abnormal tibia or tarsus,and with apparent affinities to Geometers or other dceli- 
cate Heterocera: at the other Z7phyra, of which the stout build and other characters suggest 
relationship to the Mymphaline among butterflies and Castniide among moths. Between 
these lie the normal Zycenide, commencing with the brilliantly coloured Porttia group ; 
then follow the tailless “blues’’ or Zycena group, with their short and round hindwings, 
passing into the closely allied tailed “ blues” with very similar hindwings, the Polyommatus 
group ; these pass into the Amdlypodia group, whose hindwings are normally less round, and 
are more produced and sometimes lobed, varying much in respect to the single tail, which 
is generally much stouter than in the preceding group, but is sometimes filamentous, and 
indeed occasionally obsolete. These blend with the Coppers and Hairstreaks, the Zhecla 
group, with one or two tails and more elongated hindwings, and pass naturally into the Hair- 
streaks with two or three tails, the well-marked AZy7iza and Horaga groups. These are connected 
by a few genera with single tails of extraordinary length, accompanied by a distinct lobe (ZLoxura 
group), with the Dewdorix group distinguished by their elongate hindwings produced anally into a 
long deep and distinct lobe, and furnished with one tail. Here and there stand peculiar genera 
which diverge from the more distinct types and seem to mar the exact order based on the 
selected classificatory characters, but which appear to have natural affinities with the groups 
in or near which they are located in the Key. 
Admitting that a ‘natural system” must take cognizance of all or nearly all the different 
structural features, &c., of the several forms, it is manifest that it is in the egg stage that 
