214 METHODS FOR THE ANALYSIS OF BLOOD 



different for different species of blood; it probably depends upon 

 the facility with which the blood is laked. 



When the reading of the volume of the gas mixture, consisting 

 of oxygen, carbon monoxide, and a little nitrogen, is constant, 

 a solution of alkaline pyrogallate 1 is introduced into the cup of 

 the apparatus, is covered by a thin layer of paraffin oil, and is 

 allowed to flow slowly down the inner wall of the graduated part 

 of the apparatus. A little suction is produced during this part 

 of the procedure by lowering the leveling bulb slightly. 



The absorption of the oxygen is very rapid and is completed 

 in less than one minute; the reading is taken and the pyrogallate 

 solution introduced once more until a constant reading is obtained. 

 The gas is then measured under barometric pressure in the same 

 way as described by Van Slyke for carbon dioxide (see page 127, 

 this Manual) and oxygen. 



As the solution is very dark and it is a little difficult to get good 

 readings of the meniscus, a new meniscus is produced by letting 

 a little water flow down after the pyrogallate solution; the water 

 floats on the top of the fluid and one can get readings to about 

 0.002 c.c. Instead of water a few drops of octyl alcohol may be 

 used. 



The apparatus is washed out twice with dilute ammonia solu- 

 tion after each determination. 



Calculation. The gas measured is reduced to standard con- 

 ditions by multiplying by the factor (0.999-0.00460 X ai ?*". e , 



t being the temperature in C. If 2 c.c. of blood have been used, 

 the values of this factor in Column 3 of Table 1 of Van Slyke's 

 paper on oxygen may be used (see page 126, this Manual), the result 

 then being expressed in c.c. of CO per 100 c.c. of blood, when the 

 nitrogen correction, 1.2 c.c. is subtracted. 2 



1 Prepared by dissolving 10 gms. of pyrogallic acid in 200 c.c. of strong 

 potassium hydroxide (160 gms. of KOH dissolved in 130 c.c. of water). 



2 The nitrogen correction is 1.2 per cent, instead of the calculated value 

 0.9 per cent, when actually determined by Bohr and by the authors. 



