J. H. Austin and L. Jonas 95 
effect of administration of 10 gm. of sodium chloride by the 
stomach in two doses an hour apart upon the plasma chlorides, 
electrical conductivity of the serum, and rate of chloride excre- 
tion. A marked rise in the plasma chlorides resulted, amount- 
ing to from 0.5 to 1 gm. per liter. This was associated with a 
markedly increased chloride output in the urine. Inspection of 
the table shows that no formula could be applied which would 
express the relations between the increases in these two factors 
either with or without inclusion of the urinary chloride concen- 
tration in the formula. In Dog I the highest chloride output, 
absolute and percentile, was associated with the lowest plasma 
chloride observed in this animal in any of the examinations im- 
mediately after chloride administration. 
The electrical conductivity of the serum exhibited an increase 
about commensurate with that of the plasma chlorides. In Dog 
III, Experiment 7, when the dog had been on a diet of moderate 
chloride content neither the plasma nor the urinary excretion 
had returned to normal on the day following the administration 
of 10 gm. of sodium chloride (Experiment 7, ‘‘C’’). In Ex- 
periment 9, however, when the dog had been on a low chloride 
diet, the plasma chlorides had returned to normal the day fol- 
lowing the administration of 10 gm. of sodium chloride, although 
the urine still showed a somewhat increased output. 
In Experiment 4, in which the effect of a mild uranium nephri- 
tis was studied, the elevation of the plasma chlorides on an ordi- 
nary diet was very marked and the chloride excretion poor. 
Following the administration of 10 gm. of sodium chloride by 
stomach, although only a moderate increase occurred in the 
plasma chlorides, they reached the highest figure observed in any 
of these experiments, but almost no response resulted in the 
urinary chloride excretion. The behavior of the kidney in this 
experiment can clearly not be interpreted as due merely to the 
elevation of its threshold for chlorides as the result of the ne- 
phritis. Obviously in this experiment a given increment in the 
level of the plasma chlorides excited much less increase in the 
rate of the chloride excretion than was the case in the other ex- 
periments in which the kidneys were normal. 
McLean (7) has shown that in the normal animal the chloride 
threshold is constant and that a definite rise in plasma chlorides 
