108 RESPIRATION 



Some blood is then withdrawn and a sample of alveolar 

 air taken. To this alveolar air is exposed in vitro blood 

 taken from the subject before the experiment began. 

 The amount of CO in combination with hgemoglobin in each 

 case is estimated colorimetrically. There are therefore 

 two samples of blood, both in equihbrium with CO and 

 oxygen, but with this difference, that in one case the 

 blood and alveoh were separated by the lung epithelium, 

 in the other they were in direct contact. In both cases 

 the CO and oxygen may be regarded as competing for 

 the haemoglobin. An imaginary example will make this 

 clear. 



Supposing the partition of the two gases in the blood in 

 vivo were — 



HbO.2 80 per cent. 



HbCO 20 „ „ 

 and in vitro 



HbOg 75 per cent. 



HbCO 25 „ „ 



Such a result would show that the intervention of the 

 pulmonary epithelium has caused more oxygen to enter 

 the blood than if no epithelium existed. In other words, 

 the epithehum has actively secreted oxygen into the 

 blood. 



The results which Haldane obtained by this method 

 suggested that diffusion occurs at normal oxygen pressure 

 when the body is at rest, but secretion when the oxygen 

 is rarefied. But Hart ridge, using a modification of the same 

 method, found that the process could under all circum- 

 stances be explained by diffusion. This method therefore 

 failed to decide the question. 



Barcroft's Experiment 



Recently a determination of the alveolar air and of the 

 oxygen in the arterial blood at low oxygen pressures has 

 been made by Barcroft in an experiment performed upon 



