DISTRIBUTION OF SPIRACLES IN INSECTS. 533 
in the Neuroptera* there are the same number, but none on the 
prothorax. ; 
In the larva of Corydalus cornutus there is a pair of spiracles 
on the prothorax, but they are no larger than those on the basal 
segment of the abdomen. It is difficult to say whether they are 
situated on the prothoracic or mesothoracic segment, but I am 
inclined to regard- them as placed on the extreme hind edge of 
the prothoracic ring. 
A curious fact may be here mentioned, as I have not seen it 
noticed before, regarding the distribution of the tracheæ in the 
larva of Corydalus. The main traches suddenly enlarge from the 
second abdominal spiracle to the base of the head, when it sub- 
divides and distributes branches to the head. From the spiracle 
on the basal abdominal segment a trachea, as large as the an- 
terior swollen portion of the main trachea, takes its origin and 
passes directly under the main trachea. Now both trachee send 
a branch opposite to where the mesothoracic stigma should be, if 
Present, i.e., on the hind edge of the ring. Both branches of the 
trachese, the main one and its fellow, anastomose perfectly over 
the branch sent off to the prothoracic spiracle. 
This doubling of the tracheæ, which “are so very large, forms 
evidently an hydrostatic, as well as respiratory, organ and serves 
to lighten the anterior and heavier portion of the body, as in the 
dilated air sacs of the terrestrial insects. This fact seems to 
Sustain the view of Gegenbaur,{ that the trachese were at first 
closed, forming air-bladders, and afterwards performed the func- 
tion of respiration. 
It would appear from these facts that while no more than ten 
Pairs of spiracles are to be found on the bodies of any one spe- 
“es of the groups of insects above mentioned, yet that eleven 
Segments of the body, in different species taken collectively, bear 
them. Now if we turn to the Thysanurous genus Campodea, we 
Shall find on the authority of Meinert that it bears spiracles on 
Sach thoracic segment. From this fact we are inclined to regard 
eleven as the normal primitive number of pairs of spiracles. 
Probably the larvæ of the different groups of winged insects had 
originally a pair on each thoracic segment. Certainly at least on 
Tolutional grounds from the indications in existing caterpillars 
re are two pairs of spiracles i horacic and meta-thoracic rings 
Te: * 2 piracles in the mes s 
nove, tively in the Libellulidæ and Ephemerid (none abdominal, as the larvæ have 
), and in the My d Perlidz 
tc, Gegenbaur, Griindziige der Vergleichenden Anatomie, 2te Auflage, 1870. p. 437 
