422 Mr. Ngwponr on the Formation and Use of the Air-sacs 
course, while in Mammalia this is not the case, the vesicles in them being only 
at the extremities of the ramifications of the respiratory structures. 
In what way then will the anatomy of the structures lead us to a well- 
founded inference that is supported by direct observation on the function ? 
We must compare insects with those animals which approach nearest to them. 
in the function of these structures,—Birds. In Birds the respiratory organs 
are not only vesicular, but are more extensively distributed over the- whole 
body than in any other Vertebrata. These, as every anatomist knows, are not 
confined merely to the great cavities of the body, but are extended to every 
part of the skeleton, as in insects. "They communicate directly with the in- 
terior of the bones of the wings and legs, as the trachez of the thorax are 
extended also into these parts in insects. "This distribution in both is more 
extensive and complete in the most active species. In Birds which are unac- 
customed to flight, as in the Ostrich, as remarked by Mr. Owen*, the com- 
munications of the respiratory organs with the bones is imperfect; whilst in 
Insects, although trachez exist in all, the vesicles are found only in those of 
flight. This fact extends even to the sexes of the same species. "Thus vesi- 
cles exist in the male of the common Glow-worm, which is winged, and de- 
signed to search out the apterous female, in which the respiratory organs are 
simply tracheal. The like conditions exist in the common winter-moth, Geo- 
metra brumaria. Yn the male of this insect I have found the vesicles large 
and numerous, but not a trace of these occurs in the female. The trachez in 
this sex, which has only the rudiments of wings, are larger relatively than in 
the female Glow-worm, and are precisely in that condition in which I have 
found them in the Diurnal Lepidoptera shortly before changing to the pupa. 
These anatomical faets are inferential of the real use of the vesicles, and are 
supported by an observation which I have been able to make on the common - 
Dung-beetle, Geotrupes stercorarius, at the moment when it is preparing to. 
take flight. A specimen of this insect which had been in confinement for about 
twenty four hours, and consequently had not expanded its wings during that 
time, when placed on a table immediately prepared to escape. After walking 
VM quickly for a short distance it began to respire freely, alternately short- 
ening and elongating its abdominal segments at the rate of about forty respi- 
* Cyclop. of Anatomy and Physiology, art. “Aves,” vol. i. p. 341. 
