ADVANCED EXPERIMENTAL PHYSIOLOGY 143 



tap is so manipulated that the gases only, and not the water which 

 condenses in B, are driven over into the eudiometer. The water 

 is returned back into F. Several exhaustions 

 are needed to extract the gases. The eudio- 

 meter tube is filled with mercury and surrounded 

 with a water-jacket to keep the temperature 

 constant. The eudiometer is transferred to a 

 vessel of mercury and the volume of gas read, 

 the level of mercury inside and outside the 

 eudiometer being the same. The temperature 

 of the water in the jacket of the eudiometer 

 is also read and the barometric pressure. Pot- 

 ash solution 20 per cent, is then introduced 

 into the eudiometer by means of a pipette pro- 

 vided with a bent end. The carbon dioxide is 

 thus absorbed and the difference in volume 

 read. Pyrogallic acid is then introduced and 

 the oxygen absorbed. The remainder is nitro- 

 gen. The temperature of the water-jacket is 

 kept constant by adding cold water during 

 the estimation. To correct the volume of gas 

 to and 760 mm. the following formula is 

 employed : 



V= V / H -/ 



1 + t . 0-00367 ' 760 



where H = the observed pressure, / the tension of aqueous vapour 

 at the observed temperature t. The value of 1 -j- t . 0-00367 and 

 of / are obtained from tables (cf. Sutton's Volumetric Analysis}. 



FIG. 138. U, mer- 

 cury vessel ; t, 

 eudiometer ; p, 

 pipette. 



CHAPTER XXV 

 THE OXYGEN CAPACITY OF BLOOD 



The Ferricyanide Method of Determining the Oxygen Capacity of 

 Blood. Haldane has introduced a simple method of determining 

 the oxygen in combination with the haemoglobin of the blood. It 

 depends upon the fact that the combined oxygen is liberated rapidly 

 and completely on the addition of a solution of potassium ferri- 

 cyanide to laked blood. The gas can be easily collected and 

 measured with apparatus similar to that of Dupre for the deter- 

 mination of urea in urine. 



The apparatus used by Haldane is shown in Fig. 139. 



The process is conducted in the following way : 20 c.c. of oxalated 

 or defibrinated blood, thoroughly saturated with air, are measured 

 from a pipette into the bottle A. To this are added 30 c.c. of a 

 weak solution of ammonia made from ordinary strong ammonia 

 solution, sp. gr. 0-88, by diluting with distilled water to 



