20 TRANSFORMATIONS OF INSECTS. 
of five pairs of them. The absence of the central legs produces 
the curious appearance of these well-known insects. In the © 
engraving the fore-legs are also shown, but they are very small 
and pointed, whilst the two pairs of membranous legs are short 
and distinct. 
The wings exist in the great majority of perfect insects. They 
are four in number, and are in two pairs: one is attached to the 
mesothorax and the other to the metathorax. One of the largest 
divisions of the class of insects (the Dzftera) is supposed to contain 
species that only have two wings; but the wings of the second pair 
do exist, although they are in a very rudimentary condition. 
In most insects the wings are either like a transparent mem- 
brane, or are covered with dusty scales of microscopic size, the 
whole being stretched upon hard radiating nervures; but in one 
entire order the anterior wings have the consistence of the skin 
of the body, and forma sort of case to the upper parts of the 
insect. The membrane of the wings is to all appearance simple, 
but it is really double. There is no trace of a wing, or anything 
approaching to it, in the caterpillar, but the shape of the thorax 
of the chrysalis indicates that these wonderful organs are in course 
of formation whilst it is motionless and inactive. Immediately 
after the chrysalis is formed, and just before the caterpillar’s skin 
which covers it for a while is burst and detached, the wings and 
some other organs may be traced in a very rudimentary condition. 
Réaumur has given the two accompanying views of chrysalides, 
to show the wing cases in their natural position, and when removed 
by a little management with the antenne from the lately-formed 
pupa. When the butterfly bursts forth from the skin of the 
chrysalis, its wings are often so moist and so unfinished that it is 
easy to see they are formed of two membranes, between which 
run the nervures which enable the insect to move them in its 
flight. 
The heads of insects are composed, like the other portions of 
the body, of several segments; but, although this is a fact, its 
demonstration has not been satisfactorily completed. At the 
birth of the larva no satisfactory division of the head into seg- 
ments can be detected; but knowing, as we do, that each pair 
of appendages—legs and wings—is attached to a particular seg- 
