Donald D. Van Slyke 273 
more acid than a smaller volume. We accordingly indicated this 
effect by introducing C into the formula, and indicated its com- 
paratively minor importance by using its fourth root. The for- 
mula thus involved enabled us to solve our main problem; viz., the 
question as to whether any quantitative relationship could be 
demonstrated between fall in alkaline reserve and rise in ammonia 
and acid excretion. The formulas which had been discarded in 
the preliminary tests we did not afterwards apply to the main 
body of our data. It is fortunate that Barnett has now re- 
tested the formula without C on our published data, and shown 
that this variable may be neglected. With the formula thus 
simplified it becomes possible, as indicated in the accompanying 
table, to interpret the 24 hour excretion of acid plus ammonia 
directly into terms of acidosis with the use of no more elaborate 
terms than = or ec. of 0.1 N acid plus ammonia per kilo. 
Barnett also calculates from the results of Paper IV that but 
little is gained in average accuracy by including variations 
in the body weight in the estimation. The data of Paper IV, 
however, taken without those of Paper VI, are not suited 
to decide statistically the question of the influence of body 
size. Of the 65 determinations on diabetics reported in 
Paper IV, 29 are on a single patient of 50 kilos weight, 
and of the others, only one determination was made on a sub- 
ject of less than 37 or more than 50 kilos weight who had a 
marked acidosis without bicarbonate dosage. In this one, a 
boy of 12in actual coma, both the original \ =a 4/ Cand the 5 \ > 
formula indicate a plasma CO, of 30 per cent, whichis a severe 
acidosis, though not so severe as that shown both by clinical con- 
much larger amounts sometimes formed, excretion of several liters of urine 
per 24 hours appears necessary. The formation of large amounts of ace- 
tone bodies, or of the sugar which accompanies them, apparently acts as a 
diuretic, and secures the necessary excretion. For example, Magnus- 
Levy reports that when a patient on 2 successive days excreted 109.5 and 
157 gm. of organic acid calculated as hydroxybutyric, the corresponding 
urine volumes were 8.0 and 9.2 liters (Magnus-Levy (4), p. 182). Similar 
high excretions are seen in the data of Papers IV and VI when acid excre- 
tion, as indicated by the ammonia plus titratable acid, was high. 
