J. F. McClendon ye 
In the above table the rate of rise of pH with washing out of 
COz was slower than in the previous experiment, probably due to 
less rapid flow of hydrogen. Somewhere between pH 5.4 and 
4.4 the proteins coagulated, presumably at the isoelectric point. 
This experiment shows the acid-binding power of the proteins 
after their isoelectric points have been reached by the slow change 
in pH on dropping in more acid, the last 0.1 cc. of acid changing 
the pH only 0.35. 
It seems to be impracticable to determine the alkaline side of 
the titration curve of plasma on titration with acid, and a far 
better way is the addition of an excess of acid at first and titration 
with CO.-free NaOH. In the following experiment, 0.4 cc. of 
0.1 n HCl were added to 1 cc. of plasma and titrated with 0.1 N 
NaOH. The blood had been exposed to air and hence the 
alkaline reserve of the plasma was not normal. 
NaOH pH NaOH pH 
cc. cc. 
0.00 5.40 0.25 9.02 
0.05 5.92 0.30 9.43 
0.10 6.57 0.35 9.76 
0.125 7.00 0.40 10.02 
0.15 7.42 0.45 10.24 
0.20 8.32 0.50 10.42 
On plotting these data a logarithmic titration curve is pro- 
duced which shows, however, considerable buffer effect. After 
the alkali is dropped in and well mixed by tilting the apparatus 
and by the rotation, a fall of potential of less than 2 millivolts 
is noted. This effect increases as the alkalinity increases and 
may be due to a slow process of combination of alkali with 
protein. No argument can be deduced from this curve to show 
that some other pH than 7 is a more logical end-point for titra- 
tion. It may be noted that when the alkali equivalent of the 
acid is dropped in the pH rises to 10. Plasma should reach this 
pH on driving out the CO, with hydrogen, but it would take a 
long time to accomplish it. 
On adding an excess of acid to drive out the CO, it is necessary 
to allow the vessel to rotate and a rapid stream of hydrogen to 
