J. F. McClendon 25 
The whole of 15 minutes need not be spent in titration as it is 
possible so to regulate the readings that practically all of the 
last 10 minutes may be spent in doing other things. 
In the above titrations, the pH of pure water was taken arbi- 
trarily as the end-point. This method gives at least comparative 
values and probably is not far from the true end-point, since it 
should at least determine the bicarbonate. It is impossible to 
titrate the phosphates, as a pH of about 3 to 4 would be neces- 
sary to decompose them, and at this pH the proteins would bind 
. acid. The isoelectric point of serum albumin is at pH = 4.7, 
and of serum globulin pH = 5.4, and these proteins should not 
bind acid or alkali at their isoelectric points. These proteins 
may bind some alkali at pH = 7 and therefore it might seem 
-best to use the mean of their isoelectric points as the end-point for 
titration. Itis necessary, however, to use about a third more acid 
in order to do this and it seems improbable that the proteins 
bind a fourth of the alkali at pH = 7. There may be other 
ampholytes in the plasma with very different isoelectric points, 
since the concentration of diffusible phosphates is not sufficient 
to account for this large acid-binding power. According to a 
rough calculation, the pH of distilled water under an atmosphere 
containing 5 per cent CO, should be near the isoelectric points of 
serum proteins. It would seem of interest, therefore, to compare 
the titration of plasma under hydrogen with that under hydrogen 
containing 5 per cent COz. Owing to the fact that no tanks were 
at hand and my gas mixer holds only 1 liter, I was not sure that 
equilibrium was reached. A sample of plasma was titrated while 
H» was passing and pH = 7.00 was maintained after 0.33 cc. of 
acid had been run in. By means of a 3-way cock, H. + 5 per 
cent CO2 was substituted and the change in pH noted while a 
liter of the mixture passed through the electrode vessel. The 
pH gradually fell, finally reaching 6.3. Since this is still far — 
removed from the isoelectric point of the proteins, it seems 
probable that some other buffer action is present but whether the 
phosphates could account for this buffer action was not determined. 
In the electrometric titration of many solutions the end-point 
is marked by a more or less distinct angle in the pH curve and it 
was thought advisable to plot these curves for plasma. The 
curves plotted by Cullen are almost straight lines and therefore 
