tolman. — electromotive force produced in solutions. 131 



7. Ratio of the Electromotive Force to tub Square of the 

 Number of Revolutions per Second, 



From equation (2) it is evident that the electromotive force produced 

 by the rotation should increase as the square of the number of revolu- 

 tions per second, that is, Ejn^ for a given solution should be a constant. 

 The degree of the constancy of this quantity is illustrated by the fifth 

 column in Tables I-IV which gives the values oi E/n"^ as calculated from 

 the data. Considering the separate runs in a series of measurements, we 

 see a tendency for the individual measurements of i\iQ first run to show 

 the largest deviations from the mean. There is probably some connec- 

 tion between this and the fact that the value reached by the residual 

 electromotive force is also largest after the first run in a series. In 

 many of the individual runs there is a tendency for the electromotive 

 force te decrease somewhat during the run. This would correspond to 

 the gradual production of a negative residual electromotive force. As 

 already pointed out, the nature of these residual potential differences 

 is too uncertain to permit of a trustworthy correction. 



8. The Partial Volumes of Iodine and the Iodides. 



Before making a calculation of the transference numbers from the 

 electromotive force data which we have just considered, a knowledge 

 of the partial volumes of iodine in iodide solutions and of the iodides 

 in aqueous solution is necessary. 



The partial specific volume of any constituent of a solution may be 

 defined as the increase in volume of the solution when one gram of the 

 constituent in question is added to a quantity of the solution so large 

 that the addition causes no appreciable change in concentration. In 

 the language of mathematics, if the addition of ^m grams of the 

 constituent at the concentration under consideration produces an 

 increase of £^v cc. in the volume of the solution, the partial volume of 

 the substance may be defined as the limit approached by t^v/Lm as 

 Am approaches zero. 



The quantity Ay/Am and its limit the partial volume dvldm may be 

 obtained by the same experimental methods used for the determination 

 of the specific gravity of solutions or, indeed, may be calculated from 

 specific gravity data if such are available. 



The purpose of this section is to show the method of calculating 

 partial volumes from picnometer weighings or from specific gravity data 

 and also to present the results of some experimental determinations of 

 the partial volume of iodine in potassium iodide solution as well as the 



