THE AIR-SACS IN BIRDS. 271 



bony rings, .dilates, and then traverses the lung, gradually 

 narrowing, to the posterior edge of that viscus, where it ter- 

 minates by opening into the posterior air-sac, which generally 

 lies in the abdomen. From the inner side of the bronchus, ca- 

 nals are given off, one near its distal end, and others near its 

 entrance into the lung, which pass directly to the ventral sur- 

 face of the lung, and tliere open into other air-sacs. Of these 

 there are four. Two, the anterior and ih& posterior thoracic, 

 lit! on the ventral face of the lung in the thorax. The other 

 two are situated in front of its anterior end, and are extra- 

 thoracic. The external and superior is the cervical, the inter- 

 nal and inferior, the interclavicular. This last unites into one 

 cavity with its fellow of the opposite lung. Thus there are 

 altogether nine air-sacs ; two posterior or abdominal, four 

 thoracic, two cervical, and one interclavicular. Other large 

 canals given off from the bronchus do not end in air-sacs, but 

 those which pass from the inner side of the bronchus run 

 along the ventral surface, and those on the outer side, along 

 the dorsal surface, of the lung. Here they give off, at right 

 angles, series of secondary canals, and these similarly emit 

 still smaller tertiary canals ; and thus the whole substance of 

 the lung becomes interpenetrated by tubuli, the walls of the 

 finest of which are minutely sacculated. The different sys- 

 tems of tubuli are placed in communication by perforations in 

 their avails. 



In most birds, the air-sacs (except the anterior and pos- 

 terior thoracic, which never communicate with any cavity but 

 that of the lungs) are in connection with a more or less exten- 

 sively ramified system of air-passages, which may extend 

 through a great many of the bones, and even give off subcu- 

 taneous sacs. Thus the interclavicular air-sac generally sends 

 a prolongation into each axilla, which opens into the proximal 

 end of the humerus, and causes the cavity of that bone to be 

 full of air. When the sternum, the ribs, and the bones of the 

 pectoral girdle, are pneumatic, they also receive their air from 

 the interclavicular air-sacs. The cervical air-sacs may send 

 prolongations along the vertebral canal of each side, which 

 supply the bodies of the cervical vertebras, and communicate 

 with elongated air-chambers in the spinal canal itself. When 

 the dorsal vertebrfe are pneumatic, they communicate with 

 the system of the cervical air-sacs. The abdominal air-sacs 

 send prolongations above the kidneys to the sacral vertebrae 

 and to the femora, whence these bones, when they are pneu- 

 matic, receive their air 



