J. F. McClendon 23 
bending of the small tubes, cannot be shoved straight in but 
must be started at an angle, and during the process the dise must 
not be allowed to dry or to touch any solid object. The operation 
is easier than it may seem. The 3 mm. glass tube is connected 
with a supply of hydrogen washed with HgCh, alkaline pyrogallol, 
and distilled water, and a vigorous stream of the gas passed 
through the electrode vessel. The Tiffany motor is started and 
the electrode vessel rotated so as to spread the plasma on its 
walls, at the same time leaving enough in the bottom to vover the 
disc. If any portion of the disc dries or gets a clot on it, it must 
be cleaned and recovered with palladium black. Readings are 
taken with the potentiometer and acid is run in while the vessel 
rotates. A 455 ohm galvanometer was used and a balance with 
the potentiometer took but a few seconds. The palladium black , 
was removed each day by inserting the dise for a moment in 
aqua regia and spraying with distilled water, and was then de- 
posited again. Since this can be done so quickly and without 
disconnecting the pipette or hydrogen tube or emptying the 
rubber tube of a saturated solution of KCl, I have not attempted 
to determine the length of life of a coating of palladium black. 
Platinum black has lasted longer in other electrodes but is not so 
easily removed. The gold disc may be dispensed with, but in 
order to get surface it is well to use a thick platinum wire, and a 
thick wire is liable to cause the 1 mm. glass tube to crack. I 
tried sealing a thick platinum wire into the glass tube with parafhn, 
and found that it worked for a few determinations. It could be 
coated with platinum black and cleaned by pulling it out of the 
paraffin and holding it in a flame. In putting it back, if great 
care was not used to see that the paraffin seal was perfect, mois- 
ture sometimes got into the glass tube and threw out the readings. 
Some platinic chloride solution that had been used for conductivity 
electrodes and had perhaps been contaminated with. heavy 
metals, when used to platinize the hydrogen electrode, gave errone- 
ous readings. | | 
The end-point chosen for the titration was the pH of pure water 
which is about 7.03 at 20°, 7.00 at 22—23°, and 6.95 at 25°. The 
titrations were made at 23°. 
The time required for a titration is not prohibitive for routine 
work. The collection, oxalation, and centrifugation of the blood 
under precautions to avoid the loss of COz must be done in any 
