AMERICAN JOURNAL OF BOTANY 
[Vol. 8, 
bottles of resistant glass. It was necessary to have them well stoppered, 
and to keep the bottles well filled in order to avoid absorption of carbon 
dioxide and the consequent reduction of alkalinity and increase of buffer 
action when made alkaline again. 
Preliminary experiments showed that a concentration of o.i M sodium 
chloride was preferable for study. Anything above 0.2 M was found to 
give plasmolysis. A solution of 0.07 M of calcium chloride was taken as 
approximately isotonic with the o.i M sodium chloride, and all mixtures 
were made up with these concentrations. 
In order to avoid a possible error by the buffer action of the solutions, 
these were tested by bringing the distilled water and all the solutions to an 
alkalinity of pH 8 and then adding a few drops of a solution of CO2 in 
distilled water. All changed by approximately the same amount, which 
showed that there was no appreciable buffer action that would interfere 
in the measurements of production of CO2. 
The time curves of the production of CO2 were plotted in the manner 
explained by Osterhout^ and used in other papers in this series. Rate of 
respiration in percent is plotted against time in minutes, the normal rate (as 
determined before addition of salt) being taken as the reciprocal of the 
average period required to change the solution from PH 7.88 to 7.60; this 
rate was taken as 100 percent. 
The behavior of the plant was not quite the same in the fall and spring 
seasons. In early spring, while working with material kept in the green- 
house, it was found that the solutions of pure salts and the mixtures were 
giving different rates of production of COo than the same solutions had in 
the fall and winter. The shapes of the curves were not changed, but there 
was an increase in all the ordinates of from 5 to 15 percent. This prevented 
the inclusion of many data, for there were not enough experiments for 
certain points to give a complete curve by themselves. Nevertheless, they 
provide confirmation of the results presented; especially of the maximum 
points in figure 2. 
Figure i shows the production of CO2 in NaCl 0.1 M (curve A), in 
CaCl2 0.07 M (curve B), and in mixtures of these. It was found that all 
concentrations of NaCl (none above 0.2 M were tried) gave an increase,^ 
while all those of CaCU gave a decrease. In some of the weaker concen- 
trations of CaClo the rate showed a tendency to rise again after falling, but 
it remained below the normal. In low concentrations of NaCl the increase 
(of from 25 to 50 percent) lasted for at least 90 minutes, while in 0.1 M 
NaCl it fell off rapidly at first and then more slowly as shown in the typical 
curve. 
A remarkable behavior was observed when the molecular proportions 
^ Osterhout, W. J. V. Jour. Gen. Physiol, i: 171-179. 1918. 
^ B. Jacobi (Flora 86: 289. 1889) states that NaCl 0.0496 M causes an increase in 
the production of CO2 by Elodea, followed by a decrease. 
