91 



tlirough the Irnchea, a pipe Iiaving becn introduced into tlie upper 

 larynx : in a short time the integuments oFthe whole of the lateral 

 and inferior parts of the body rose, and the air-ceUs seemed coni- 

 pletely filled, especially that which is situated in front of the os 

 furciforme. Being thus satisfied that they all had a free commu- 

 nication with the chest, we next proceeded to see at what points 

 these Communications took place, and in what degree the air-ceUs 

 communicated with each other. For that purpose the air-cells on 

 the left side of the body were laid open, and shortly after those of 

 the opposite side collapsed, indicating the existence of apertures of 

 communication, although the septum vvhich ran along the middie 

 line of the body appeared at first sight imperforate. There was a 

 free communication between the lateral air-cells of the šame side of 

 the body from the os furdfoiine to tlie side of thepe/rii; but the 

 air-cell in front of the os furciforme remained still tensely inflated. 

 The lateral air-cells had a free communication with the cavity of the 

 chest at the axilla, at vvhich part the air had entered these cells 

 during the inflation. The pectoral museles and those of the thigh 

 presented a singular appearance, being as it were, cleanly dissected, 

 having the air-cells extended above and below them ; the axillary 

 vessels and nerves also passing bare and unsupported by any sur- 

 rounding substance through these cavities. We traced the air-cells 

 down the side of the humerus, ubia, and metacarpal bone, into all of 

 which the air entered, and even into the bone corresponding to the 

 first phalanx, which agrees with what Mr, Hunter has described 

 in the Pelican. ( Animal CEcon. p. 92.) 



" As none of these proceedings had any effect on the air-cell in 

 iront oi ihe os furciforme, vvhich still continued distended, it was 

 evident that inflation by the humerus could not have filled it except 

 through the medium of the lungs theraselves. We next proceeded 

 to detach the integument from this air-cell to see its shape and ex- 

 tent ; this reąuired to be done with great care, as it adhered pretty 

 closel}' to the skin and roots of the feathers : it was of a globular 

 form, about four inches in diameter, and communicated vvith the 

 thoraz at its anterior aperture belovv the trachea. 



" Numerous strips of muscular fibres passed from various parts 

 of the surface of the body, and vvere firmly attached to the skin ; a 

 beautiful fan-shaped musele vvas also spread over the external sur- 

 face of the air-cell anterior to the os furciforme. The use of these 

 museles appeared to be, to produce instantaneous expulsion of the 

 air from these external cells, and by thus increasing the specific 

 gravity of the bird to enable it to descend vvith the rapidity neces- 

 sary to the capture of a living prey vvhile svvimming near the surface 

 of the vvater. 



" With respect to the general anatomy of this bird, it may be 

 observed that vve found the tvvo small glands at the termination of 

 the trachea, vvhich are noticed by Montagu, and vvhich exist in ad- 

 dition to the ordinary pair lying above the bronchice. The stomach 

 corresponded exactly vvith the figure given by Sir Everard Home 

 (Comp. Anat, pi. xlvi.), the pyloric orifice being provided with the 



