RESPIRATION OF TETRAPODS 53 



bronchi are compressed together and diagrammatically they 

 may be represented as hexagons in cross section with an internal 

 tube of about 0-5 mm. diameter (fig. 16c). From the lumen of 

 these tertiary bronchi, fine blind tubes arise which are up to 10 /x 

 in diameter. These air capillaries may anastomose and are pro- 

 fusely supplied with blood and it is at their surface that the gas- 

 eous exchange takes place. The distances from the parabronchial 

 lumen to the end of the air capillaries are small and both Zeuthen 

 and Hazelhoff calculated independently that diff'usion was suffi- 

 cient to satisfy the birds' requirements, both at rest and in flight. 



The mechanism for ventilation of this complex system is 

 basically a costal suction pump as in reptiles. The movements are 

 shown diagrammatically in fig. 17; expansion of the thoracic 

 cage is produced by the action of intercostal muscles which 

 move the ribs forwards and outwards while the sternum is de- 

 pressed and moved slightly forwards. Thus, during inspiration 

 the vertical dimension of the thorax increases greatly and the 

 transverse dimension slightly. The consequent reduction in 

 pressure within the body cavity and pleural cavities results in 

 air being drawn into the system. Expiration mainly results from 

 relaxation of the intercostal musculature but there is evidence 

 for active contraction of the abdominal muscles and internal 

 intercostals. During flight, the rhythmic raising and lowering of 

 the sternum produced by activity of the flight muscles ventilates 

 the respiratory system. It is possible, however, that the bird is 

 also able to ventilate the lung by another mechanism, for it must 

 exhale air continuously when emitting continuous sounds dur- 

 ing flight. The respiratory rhythm is unchanged when a bird 

 makes its more rapid flight movements in the laboratory. 



The path of the air through the respiratory system during 

 expiration and inspiration has given rise to much conjecture and 

 several theories have been developed to account for the experi- 

 mental observations. No completely satisfactory solution has 

 yet been found; those features which are fairly definite and 

 common to many accounts will be summarised here, and the 

 two most probable alternatives will be indicated. There is no 

 evidence for the existence of passive flap-valves which might 



