SECRETION AND ABSORPTION OF GAS, ETC. 251 



falls to about that of the alveolar air. The following- table 

 gives the average arterial oxygen tensions found under the 

 conditions referred to. 



When instead of ordinary air a mixture rich in oxygen 

 was breathed by mice which had previously been exposed to 

 high-pressure oxygen, it was found that the arterial oxygen 

 was always distinctly below that of the air breathed, and not 

 much more than half as high as in normal animals breathing 

 the same atmosphere. 



The results of these experiments appear to exclude the 

 possible view that for some reason or other the carbonic- 

 oxide method gives values which are always too high in the 

 living body. They also afford very strong support to the 

 conclusion that the normal lung epithelium plays an active 

 part in the taking up of oxygen. It seems clear that abnor- 

 mal conditions, whether acting only locally on the lungs, or 

 on the tissues of the body generally, very readily interfere 

 with the normal activity of the epithelium, just as is known 

 to occur in the case of other secretory glands. 



It may further be remarked that experiments with car- 

 bonic oxide, apart from the "evidence which they furnish in 

 favour of active absorption of oxygen, strongly support the 

 view that diffusion alone is capable of bringing about a con- 

 dition of perfect equilibrium between the gas-tensions in the 

 alveolar air and the blood passing the lungs ; for even the 

 lungs with damaged or paralysed epithelium give an oxygen 

 tension about equal to that of alveolar air. Fredericq's con- 

 clusions that the oxygen tension of the blood leaving the 

 lungs is considerably lower than that of alveolar air would 

 present a good deal of difficulty even on the diffusion 



