548 EXTERNAL ANATOMY OF INSECTS. 
of the piece behind the collar, and excludes it from 
having any share in that cavity. These arguments at 
first sight seem to prove satisfactorily the identity of the 
collar and prothorax. But audi alteram partem, and I 
- think you will allow that the scale containing the claims 
of the collar to be considered as a piece su? generis, dips 
much the lowest. And, first, I must observe, that 
though in Hymenoptera the collar seems to replace the 
prothorax by its situation, yet it is in fact a part of the 
alitrunk ; for, if the manztrunk be separated from the lat- 
ter, the collar remains, in most cases, attached to it 3, 
while the antepectus and arm, with the ligament that 
«overs its cavity above, the real representative of the 
prothorax, are easily removed, and this in recent indi- 
viduals: as a further proof of this, I must request you 
will examine a neuter Mutilla ; you will see that in this 
the collar is not separated from the alitrunk in any 
respect, but forms one piece with it, while the antepectus 
is distinct and capable of separate motion: further, the 
action of the collar is upon the alitrunk, it being of es- 
sential importance in flight, whereas the prothoraz is of 
no other importance than as a counterpoise to that part®. 
A further argument to prove the distinction of these parts 
may be drawn from the case ef Xylocopa, a kind of bee. 
In-this genus the collar forms a complete annulus or 
segment of the body; now, if it really represented the 
prothorax, the under side of the segment, as in those 
Coleoptera in which no suture separates the upper from 
~* In Chlorion, Ammophila, &c., this part separates more readily 
from the alitrunk. 
® Chabrier Sur le Vol des Ins. c. 1, 413—. c. iv. 54. This author 
seems to regard the collar as something peculiar to Hymenoptera. 
