THE EXCHANGE OF IONS 
before reading them, however desirable such a procedure might have 
been. It is well known that the temperature coefficients of conduc- 
tivity for very dilute salt solutions, in which carbon dioxide exists in 
considerable concentration, are not the same as the coefficients for 
concentrated, relatively uncontaminated solutions. It was within 
the limits of accuracy of our work to use for all the dilute salt solutions 
the coefficient 0.20. All conductivities are referred to 18° C. Since 
in conductivity work with solutions of even lf/50 concentration the 
impurity of the water may require a correction of o.i per cent., it is 
quite obvious that in some of our more dilute culture solutions the 
concentration increment between contiguous cultures was less than the 
initial concentration of impurities in the water. It was found, how- 
ever, that the effect on conductivity of adding small increments of salt 
to distilled water was additive. Consequently we have assumed that 
the changes in conductivity of the solutions during the period of the 
experiment were likewise due to addition or subtraction of electrolytes. 
In the presentation of our results we have translated the changes 
brought about by the roots in the conductivity of the culture solutions 
into terms of change in concentration of the particular salt used in 
each experiment. All concentrations were determined with reference 
to the conductivities of the culture solutions as originally prepared, 
before the roots had been placed in them. As already remarked, at 
least part of the initial rise in concentration of most culture solutions 
after the roots are placed in them may be ascribed to carbon dioxide. 
Presumably the absorption of salt from all solutions was greater than 
the figures show, by the amount of this initial rise in the CO2 con- 
centration. As we have already pointed out, the error from this 
source can never have been as great as would correspond to a change 
in concentration of 14.M X io~^, in the case of a salt like NaCl. 
In connection with the fact that not only absorption but also 
excretion was measured in terms of the original salt used in spite of 
the fact that we know nothing in regard to what the roots may actually 
have excreted, it may be stated that most of the commoner electrolytes 
have nearly enough the same equivalent conductance so that little 
error would enter into our results on this account. At the close of 
each experiment the following measurements were taken for each 
plant: The length of the primary root, the length of several of the 
longer lateral roots, the length of the hypocotyl, and the length of the 
first internode above the cotyledons. It was found that only a very 
