MEASUREMENTS OF ELECTROLYTIC MIGRATION 7 



duced above for hj'drogen. For a metal dissolving at the anode, 

 where the electrode is not weighed, the correction is equal to 

 the correction for oxygen (but taken with opposite sign) multi- 

 plied by one-eighth the gram-equivalent weight. ^ 



INTERPRETATION OF THE RESULTS. 



The transport number of any component of a solution is the 

 number of gram-equivalents of that component which enters or 

 leaves one of the electrode compartments during the passage of 

 96,540 coulombs of electricity. 



Hittorf has shown how in certain cases (double cyanides, 

 stannic chloride, cadmium iodide, for instance) these transport 

 numbers may suggest definite views as to the constitution of the 

 electrolyte. In the following paragraphs I shall briefly discuss 

 the bearing of transport measurements on the hypothesis of 

 hydrated ions, and on that of the possible movement of undis- 

 sociated molecules of the electrolyte ; and I shall explain the 

 methods of calculating the concentrations and mobilities of the 

 complex ions or products of hydrolysis that may be assumed to 

 exist in the solutions of certain salts. A few measurements 

 which are not easily reconcilable with the prevalent theory of 

 solutions are collected under the heading " anomalous results." 



Hydrated Ions. 



In a solution of sodium chloride (to take a concrete example) 

 it is customary to assume the ions Na' and CF together with 

 water and undissociated molecules of NaCl. It is, however, 

 possible to assume the existence of hydrated ions : (Na-fxHgO)' , 

 and (CI -f yH20)', or of complex ions, or even that the undis- 

 sociated NaCl moves towards one of the electrodes, under the 

 influence of a fall of potential, as colloids do. 



If hydrated ions be assumed, it is evident that in calculating 

 the transport numbers from the experimental data, allowance 

 must be made for the water carried by them into the electrode 

 compartments and set free at the electrodes. I have shown 

 elsewhere ^ that this correction may be expressed by the formula 



(-^y ^xjC, in which 1^ and 1^ represent the mobilities of 



^Zeit. f. Elektrochemie, //, 216 (1905). 



