THE CHEMISTRY OF RESPIRATION 1109 



A large glass globe with a stop-cock at one or both ends (Fig. 507) is filled with a 

 gaseous mixture of known composition containing oxygen. Into it are introduced 2 or 

 3 c.c. of blood or of haemoglobin solution. It is then tightly stoppered and immersed in a 

 horizontal position in a pail of water kept at a constant temperature. In the pail it is 

 suspended between its two ends, so that it can be slowly revolved by means of a piece of 

 string passing round its neck. In this way the blood is continually spread in a thin layer 

 over the sides of the vessel. At the end of a quarter to half an hour it will have attained 

 equilibrium with the gaseous mixture. It is then turned into an erect position so that 

 the fluid can run down into the neck closed by a stop-cock, whence 1 c.c. may be drawn 

 off for analysis in a Barcroft apparatus. A further portion of the same blood may be 

 shaken up with air so as to saturate it completely, and the saturation of the two samples 

 may be compared in the differential gas apparatus. 



FIG. 507. Barcroft's apparatus for determining the curve of absorption of 

 oxygen by haemoglobin. 



Barcroft has shown that the dissociation curve of haemoglobin is largely 

 altered by slight variations in the fluid in which the haemoglobin is dissolved. 

 The most important of these conditions are (1) the saline content of the 

 fluid, (2) the reaction of the fluid. Under this latter heading must be classed 

 the, amount of carbon dioxide present, since its action on the dissociation 

 curve is similar to that produced by the presence of weak acids such as 

 lactic acid. The influence of dissolved salts on the dissociation curve is 

 shown in Fig. 508. 



It is interesting to note that the differences between the dissociation 

 curve of blood and of haemoglobin solution, as well as between bloods of 

 different animals, have been shown by Barcroft and Camis to be dependent 

 on the saline content of the solution in the various cases. Thus human 

 haemoglobin solution, with a concentration of salts similar to that of dogs' 

 blood, gives the same dissociation curve as normal dogs' blood. 



More important is the effect of reaction since, as we shall see, it is 'the 

 reaction of the blood, controlled especially by carbon dioxide tension, that 

 determines the activity of the respiratory centres. In Fig. 509 is repre- 

 sented the influence of varying tensions of carbon dioxide, and in Fig. 510 

 the effect of slight additions of lactic acid on the dissociation curve. It 

 will be seen that the more acid the blood, or the greater tension of carbon 

 dioxide it contains, the more readily does it undergo dissociation. This is 

 especially marked at the very high tension of 420 mm. carbon dioxide. It 



