xi RESPIRATORY EXCHANGES 383 



therefore, the 22 vols. of oxygen contained in arterial blood 

 must, to a large extent, be in a state of chemical combination 

 (Liebig, 1851 ; L. Meyer and Fernet, 185*7). We know, in fact, that 

 the oxygen absorbed by the blood is in loose combination with the 

 haemoglobin of the erythrocytes, which gives rise to the formation 

 of oxyhaemoglobin (Hoppe-Seyler, 1864; see Chap. IV. 7). 



The proof of this fact, one of capital importance in the 

 physiology of the respiratory exchanges, is that a watery solution 

 of 14 per cent pure haemoglobin (which corresponds to the normal 

 haemoglobin content of the blood) is capable of absorbing and 

 chemically fixing as much oxygen as an equal volume of blood, and 



FIG. 168. Curve to show percentage variations of oxyhaemoglobin in a solution of 14 per cent blood- 

 pigment, with variations of partial pressure of the atmospheric oxygen with which it comes 

 in contact. (Hiifner.) Quantity of oxyhaemoglobin in T J along axis of ordinates ; partial 

 pressure of oxygen in mm. Hg, along axis of abscissa. 



by means of the Torricellian vacuum it is possible to extract as 

 much oxygen from the same solution as from blood. 



It was fundamental to the conception of oxygen absorption in 

 the blood, to determine to what point the quantity that combines 

 with haemoglobin depends on its partial pressure in the atmosphere. 

 Bohr (1885) and Hiifner (1888) made a number of experiments 

 with this object. The method consisted in placing a given quantity 

 of defibrinated blood, or better, of 14 per cent solution of pure haemo- 

 globin (which, as we have said, corresponds with the haemoglobin 

 content of the blood) in contact either with normal air, or with 

 artificial air containing a considerably less amount of oxygen per 

 cent; and then shaking it. It is then determined how much 

 oxygen combines with, or is dissociated from, the haemoglobin 

 on a rise or fall of its partial pressure in the mixture of gases 

 Hufner's results are clearly expressed in the diagram (Fig. 168), 



