SECRETION AND ABSORPTION OF GAS, ETC. 249 



following reason. We know that gases such as nitrogen 

 or hydrogen, which are indifferent in their behaviour to- 

 wards ordinary protoplasm, diffuse freely through the lung 

 epithelium, either inwards or outwards. Carbonic oxide 

 seems also to behave towards protoplasm as an indifferent 

 gas. It is poisonous to vertebrate animals only because it 

 combines with their haemoglobin and thus cuts off the 

 oxygen supply to their tissues ; but insects, which are not 

 dependent on haemoglobin, live perfectly well in an atmos- 

 phere in which carbonic oxide is substituted for nitrogen ; 

 and even in the case of mammals carbonic oxide is no 

 longer poisonous when the animal is placed in compressed 

 oxygen, so that it obtains in simple solution in the blood 

 sufficient oxygen to render it independent of the oxygen- 

 carrying power of haemoglobin. 



A further reason for believing that the lung epithelium 

 presents no obstacle to the passage inward of carbonic 

 oxide is afforded by the fact that when air containing a 

 small percentage of carbonic oxide is breathed nearly the 

 whole of the gas which reaches the alveoli is absorbed, 

 until the point is almost reached where absorption rapidly 

 ceases in consequence of the state of equilibrium being 

 attained. 



As there can thus be hardly any doubt that carbonic 

 oxide diffuses freely through the lung epithelium, we are 

 driven to the conclusion that the oxygen tension is raised 

 in the lungs to beyond that of the alveolar air. To explain 

 the results of the above-mentioned experiments on men we 

 must assume that the oxygen tension rose to an average of 

 38*5 per cent, of an atmosphere, which is about three times 

 the alveolar oxygen tension. 



A further series of experiments made on a number 

 of different mammals and birds have fully borne out the 

 conclusions we arrived at in the case of man. Although 

 in different classes of animals the oxygen tensions found 

 have differed considerably, yet in all cases they have con- 

 siderably exceeded the oxygen tensions of alveolar air. 

 The following table shows the average results for normal 

 animals : — 



