GANNET. 
15 
air-cells in the parts already mentioned, hare a free com¬ 
munication with the lungs, by means of openings on their 
surface, through which the air passes readily into them. 
Thus far have the scientific researches of that anatomist 
contributed to our knowledge on this subject. No one 
appears to have noticed the phenomena attendant on the 
construction of the Gannet, or to what further extent this 
circulation of aerial fluid is carried in some particular species 
of birds, a circumstance which demands our highest admi¬ 
ration, when we contemplate the advantages of such a 
structure in conducing to the comforts, and perhaps to the 
very existence of such animals. On each side nearly 
equidistant between this pectoral membrane and the back, 
is situated another longitudinal one, very similar to the 
last, but perforated; between this and the pectoral are about 
nine irregular transverse membranous septa that hold the 
skin firmly to the body, having a free communication with 
each other. The skin is also furnished with a transparent 
cellular membrane, the cells being regularly perforated close 
to the base of each feather; at the upper part of the breast 
is a large bag, which extends some way up the neck; this 
is attached to the skin by the septa of innumerable small cells, 
but no opening into this cellular bag could be discovered; 
the introduction, however, of a small pipe through an 
artificial aperture clearly demonstrated a passage to the 
lungs, as the whole internal cavity of the body was inflated, 
and the air issued from the trachea. Upon opening this 
bag, the passage of communication with the internal parts 
appeared to be under the shoulder blades, (clavicles,) as a 
thin perforated membrane was perceived at the bottom, 
leading to the thorax, not directly into the lungs, but near 
the part where the trachea divaricates, and afterwards com¬ 
municating with the lungs. It could not, however, be 
discovered where the air could find a passage from the great 
magazine into the cellular bag, and yet there is every reason 
to conclude that at this part some valvular passage exists. 
Pursuing our researches we observed at the bottom of each 
lobe of the lungs a considerable opening for the passage of 
air into the cavity of the body. 
From what has been already observed, it will not be 
unreasonable to conclude that the Gannet is endowed with 
such singular properties for very different purposes than 
those of long and continual immersion, of which we have 
