128 Determination of Oxygen in Blood 
assure even distribution of the corpuscles, and a sample is drawn 
into a 2 ce. pipette and run under the ammonia in the cup of the 
apparatus. (The lower delivery mark of the pipette should be 
3 or 4 cm. above the tip. A pipette calibrated for complete 
delivery would be inconvenient for placing the entire sample of 
blood below the layer of ammonia.) The blood is now run from 
the cup into the 50 cc. chamber, the ammonia layer following the 
blood and washing it in. <A few additional drops of ammonia 
may if necessary be added from a dropper to make the washing — 
complete. 
The blood and ammonia in the chamber are mixed and all: wed 
to stand until the blood is completely laked. This requires about 
30 seconds when saponin is present, 5 minutes when it is not. 
After laking is complete 0.4 ec. of a saturated potassium ferri- 
cyanide solution is introduced to set free the oxygen combined 
with the hemoglobin. (The cyanide solution is made air-free by » 
boiling or by shaking in an evacuated flask and is kept in a burette 
under a layer of paraffin oil 2 or 3 cm. thick to exclude air.) 
The apparatus is now evacuated by lowering the levelling bulb 
until only a few drops of mercury remain above the lower stop- 
cock, and is shaken, preferably with a rotary motion, to whirl the © 
blood in a thin Jayer around the wall of the chamber. If the 
blood was completely laked before the cyanide was added, ex- 
traction of the oxygen may be completed by half a minute of 
efficient shaking. The extracted solution may be drawn into the 
bulb of the apparatus below the lower cock and the extracted gas 
measured over mercury as in the determination of carbon diox- 
ide. Or, since the water does not absorb oxygen rapidly enough 
to cause an error, the solution may be left in the 50 ce. chamber 
during the reading of the gas volume, the levelling bulb being 
held at a sufficient height to balance the column of water solu- 
tion.* Finally, in order to make certain that a]l the oxygen was 
obtained by the first extraction, the apparatus is evacuated once 
more and the blood shaken again for a half minute. If the read- 
ing shows no increase, it is evidence that all the oxygen was ex- 
tracted by the first evacuation. If there is an increase the 
extraction must be repeated again. If the blood is completely 
2 Van Slyke, J. Biol. Chem., 1917, xxx, 353. 
