85 
certain stages of the progress from the primeval butterfly to Euselasia 
and certainly forms intermediate between the latter and the Nemeobiidi, 
which in solidity, hairs, and general form differs little from Lycamines. 
The Lycaenidae apparently made a short cut to a high form. This 
is to be taken less as my view of the facts, than as an expression of 
my ignorance of any of the possibly still, but no doubt once, existing 
intermediate forms with mobility of segments. The pupae of all 
the genera are very close together, very rounded and solid, but 
preserve the Papilionid arrangement of the leg-cases; the Lemoniidae, 
so far as the few which I have examined go, have attained the 
Nymphalid arrangement. This point in the structure of butterfly 
pupae, which seems to be of some importance, is one upon which I am 
sorry to say that I have only a few notes of pupae that I have 
examined but not possessed, so that I speak here from a smaller basis 
even than elsewhere. In all butterflies (including some Hesperids) the 
area between the head and antennae is occupied by the maxillae 
(proboscis) and the 1st and 2nd pairs of legs, and by these only. 
The point to which I wish to call attention, is as to the relative 
position of the two pairs of legs. In the Papilionids and Pierids the 
1st leg at its base extends from the maxilla on the inner side to the 
antenna on the outer, and so cuts off the 2nd leg from approaching 
the head (figs. 2-1). In the Nymplialids the 2nd leg extends upwards 
and reaches the eyes (head), and so cuts off the 1st from reaching the 
antennae (figs. 15, 16, 18, 20). We may here frame a pretty little 
hypothesis that this is the natural result of the atrophy of the 1st paii 
of legs in the Nymphalid imago ; and that this is no doubt an element in 
the case, is confirmed when we notice how small the 1st pair of legs 
becomes in some Sat.yrids —hyperantJius (fig. 22), for example. But 
when we extend our survey, we find that the reduced size of the fore¬ 
leg obtains in the Hesperids (figs. 29-33) from which the Papilionids 
are supposed to have had their origin, and when we come to the 
Lycaemds, we find the reduced size in the Lemoniidae (fig. 25), which 
possess forms nearest to (but also probably more recent than) the 
Papilionids, whilst in the Lycaenidae (fig, 28), where we might also 
look for a Nymphalid type in this matter, the fore-leg reaches the 
wing as in the Papilionids, as it does also in Parnassids. This would 
point to the commencing atrophy of the fore-legs of the imago in the 
Lycaenids as occurring quite independently of the same process in the 
Nymphalids. Another illustration is thus afforded of a condition of 
things of which I have already spoken as by no means rare in the 
Lepidoptera— viz., of similar stuctural developments taking place 
independently in different families, the common inheritance being, not 
the structure itself, but a tendency, or at least a capacity to 
develop it. 
There is at least one other distinction between Lycamid and 
Papilionid pupae which is strong!} 7 suggestive of their separate origin. 
The Papilionid and its derivatives, the Pierid and Nymphalid, always 
have smooth pupae, that is pupae which have no bristles or hairs ; their 
spines and processes are developments of the pupa itself, and cannot 
be injured without opening the general body cavity. Lycaenid pupae 
on the contrary, have hairs and bristles, that is, cutaneous appendages 
that can be removed without any substantial damage to the pupa 
proper. Both these forms occur among the Hesperids. 
