We Henderson and A. H. Smith 45 
of the acid after the ferricyanide. Practically, however, unless 
the quantity of blood available is extremely limited it is easier 
and more accurate to make the two .determinations separately. 
Duplicate determinations for both gases independently can be 
made by this method in half an hour. 
Solubility of CO. in Acidified Blood Solutions. 
As the correction for the amount of CO. which remains dis- 
solved in the fluid is considerable we have devoted particular care 
Temperature 
07 08 09 10 LU 2 13 4 
Ce, C0, in solution in Icc. blood mixture at given temperature and 760mm. pressureCO2, 
Vig. 4. Curves expressing solubility of COs. 
The acid-blood mixture of Curves 1, 2, 3, 4, and 7 was made as fol- 
lows: ammonia 1.5 cc., blood 1.0 cc., tartaric acid 0.25 cc. The mixture 
of Curves 5 and 6 was made as follows: ammonia 2.0 cc., blood 1.0 ce., 
tartaric acid 0.5 ce. 
Curves 1 and 2, fresh human blood; Curve 3, fresh cat blood; Curve 4, 
blood of dog anesthetized with chloratone and alcohol; Curve 5, fresh 
pig blood; Curves 6 and 7, human serum. 
to this matter. Various arrangements of apparatus were tried 
but finally one similar to that shown in Fig. 3 (except that the 
rubber tube connecting the diffusion tube and gas burette was 
4 or 5 em. long to allow shaking the tube) proved most con- 
venient, as an entire series for all temperatures could be run 
from a single filling of the tube. 
