THE METAMORPHOSIS OF FLIES. 717 
organ lying next to the skin (ganglion or muscle) ; still this is ac- 
complished in quite a different way from the Tipulide, where the 
cavity of the appendages is traversed longitudinally by a nerve 
(at least in the legs), while in Musca the appendage is in fact 
wholly free from nerves, at all events with none of its primitive 
fibres present, and at best may be traversed by such as are formed 
anew during the outgrowth of the appendages. At all events the 
new formation of the tissues filling the cavity do not proceed from 
the neurilemma, but from the nuclei arising from the destruction 
of the fat body. 
But still in the manner and mode, in which the growing limb 
pushes out, there is no inconsiderable difference. 
In Corethra there is at first a completely unjointed cylindrical 
skin which lies wound up in a spiral on the piece of the thorax 
belonging to it; as soon as the new formation has evidently grown 
in length, when the differentiation of the cell-masses of the cavity 
into tissues has advanced, the first traces of joints appear. 
Quite otherwise in Musca, where the segmentation begins at an 
early period and is complete along the whole length of the pro- 
jecting appendage. Even before the growth outwards of the limb 
ever projects up prominently above the level of the basal mem- 
brane, the end of the limb (the fifth tarsal joint), is separated 
from the basal membrane; then with the elongation of the tarsal 
spines, the four other tarsal joints become interpolated, while the 
tibia and fibula still form an unjointed portion, whose complete 
division into the joints of the fly’s limbs occurs after the formation 
of the thorax. During the whole development within the disk the 
limb still remains as a short projection which, without turning, 
stands straight up over its thoracic piece. As soon as the new 
thorax is formed, it begins to grow moderately, and considerably 
later, in the second half of the pupal sleep, the muscles and other 
tissues: arise in the cavity of the appendages, and finally the limb 
assumes its external, definite form. From the last mentioned 
facts, it clearly appears that with the existence of imaginal disks, 
which the muscid larva brings with it out of the egg, no greater 
preparation in fact is made for building up the body of the imago ; 
that here a much more marked transformation must be undergone. 
When Corethra transforms into a pupa, the muscles of the wings 
and limbs are already formed, while in Musca there is not yet the 
first trace to be seen of either; the mouth parts undergo in Co- 
