KHOPALOCEEA. 
17 
time) fully exposed, while tlie upper side is concealed. In Moths the 
under side of the wings is nearly always duller and paler than the upper 
side, and any pattern or colouring presented more or less indicates 
what is on the upper side ; and these Lepidoptera, instead of holding 
the wings erect over the back, deflect them at various angles when at 
rest, and for the most part so dispose them that the longitudinally 
folded hind-wings and the abdomen are covered or roofed by the fore- 
wings. The only group of Moths containing numerous exceptions to 
this rule is the GcomctrcE^ and it is very noteworthy that these alone 
have the rhopalocerous habit of resting with wings erect. 
A third distinction of Butterflies is their diurnal flight ; but, as 
certain groups are crepuscular (many Hesperidce, some Morphitoe, all 
the Brassolince^ and a number of Satyrince)^ and as a great variety of 
Moths in all the great divisions, including some entire Families 
{Zyoenidce, Umnidm), are diurnal in their habits, this can only be 
taken as much more characteristic of Butterflies as a whole than of 
Moths as a whole. 
2. Classification. 
A satisfactory arrangement of the Sub-Order is admittedly most 
difficult to arrive at, the fundamental structure of its members (as 
indeed may be said of the entire Order Lepidoptera) presenting but 
comparatively slight modifications, and those being often inconstant in 
character. As regards the Imago^ the most important features from a 
classificatory point of view are (1°) the neuration of the wings, (2°) the 
condition of development of the first pair of legs, and (3°) the presence 
or absence of a second pair of spurs on the hind pair of legs. In the 
Pupa, the mode of its suspension affords the most trustworthy character, 
while in the Larva the general form, the nature of the dermal append- 
ages, and the smooth or spined condition of the head, are points of chief 
importance. It is remarkable that in proportion to the atrophy of the 
first pair of legs, which is the character in which the higher Butterflies 
differ most absolutely from all Moths, the chrysalis is free from silken 
attachments. Thus the pupa3 of the Hesperidce^ the group of Butter- 
flies in all respects most intimately related to Moths, and whose fore- 
legs are invariably fully developed in both sexes, are not only attached 
by the tail and girt with silk, but also secured by many silken threads, 
which in many cases form a slight cocoon. Those of the Papilionida3^ 
a family which also has the fore-legs perfect in both sexes, are always 
attached by the tail and also girt ; and, in the case of the remarkable 
Alpine genus Farnassins, the chrysalis, like those of the Hesperidce just 
mentioned, is further enclosed by many additional silken threads. It is 
when we come to the Zyca^nidm, in which, although the fore-legs of the 
female are perfect, those of the male begin to exhibit the failure of the 
tarsal articulation and claws, that we first find instances of chrysalides 
attached by the tail only, but still in a fixed horizontal or slightly 
B 
