G. H. F. ^S'UTTALL AND C. STRICKLAND 
95 
ferricyanide is introduced with a curved pipette, and care must be taken to see that no 
mixture takes place between the solution of ferricyanide and haemoglobin. The apparatus 
is now hung on the edge of a vessel full of water so that the bottles are completely im¬ 
mersed. After ten minutes the taps, which hitherto have allowed both the manometer 
and the bottles to be in communication with the external air, are closed so that the two 
former are in communication with one another but not with the air outside. The fluid 
surfaces in the manometer are read to ascertain if any zero error exists. They are now at 
the same level. The ferricyanide and the haemoglobin are mixed and the apparatus 
shaken for a minute and then replaced in the bath. Probably three such shakings sufiice 
to liberate all the oxygen. The observed difference of pressure is noted. The estimation 
is then repeated with the blood in the other bottle. 
Fig. 1. Sho\\dng the apparatus used in the method here described. 
To the right of the figure is seen one and part of another of the small flasks in which the 
mixture of blood, ammonia, and ferricyanide is made. The manometer and the flasks 
and the air are seen to be connected by three-way passages governed by three-way 
cocks. The dark line in the manometer represents the oil. (Figure reproduced 
from the Journ. of Physiol., bj’ kind permission of Professor Langley.) 
