178 RALPH S. LILLIE. 



law). 1 The source of the current is thus the chemical changes 

 taking place at the electrodes; in these changes electricity is 

 added to the wire at one region and an equal quantity is simul- 

 taneously abstracted at another. Precisely the same chemical 

 change takes place and the same quantity of energy is freed 

 when the chlorine is brought directly into contact with the solu- 

 tion of ferrous salt; here the transfer of negative electricity 

 from the ferrous ions to the chlorine atoms takes place without 

 the intermediary of a metallic conductor. These considerations 

 show that the metallic conductor, by acting as a source of elec- 

 trons, merely enables the change of valence to take place without 

 the direct contact of the chlorine and the ferrous salt. This 

 effect follows because the affinity of the chlorine atom for a 

 negative charge (i. e., electron) is greater than that of the 

 ferrous ion ; the chlorine atom is thus able to abstract a negative 

 charge from the metal and impel the ferrous ion to yield up a 

 similar charge to it, i. e., to assume an additional positive 

 charge and become a ferric ion. The latter then interacts 

 chemically with whatever other substances are available. 2 



In the oxidation and reduction circuit considered above, the 

 current on the one side of the interface, i. e., in the metal, is con- 

 ducted by a simple movement of electrons without chemical 

 change; on the other side, i. e., in the solution, it is conducted 

 by the diffusion of ions and the charging or discharging of ions 

 at the electrode-areas, with consequent local chemical change. 

 Let us now suppose that instead of the combination of a metallic 

 conductor and an electrolyte-solution, we have two electrolyte 

 solutions separated by a thin semi-permeable partition the sub- 

 stance of which can combine chemically with the ions of one 

 of the solutions. Such an arrangement corresponds to the case 



1 Were this otherwise, electricity could flow through the solution by shifting of 

 electrons from ion to ion (or molecule to molecule, etc.) without the transport of 

 the ion as a whole. This, however, does not seem to occur (unless the current- 

 intensity is very high). It is only when the ion interacts chemically with the sub- 

 stances at the electrode, or with the electrode itself, that any flow of current can take 

 place. This is why electrolysis is always associated with the passage of a current 

 through a solution. The metallic state appears to be characterized by a relatively 

 high concentration and free mobility of electrons. 



2 The presence of potassium thiocyanate in the solution at the anode is a con- 

 venient way of demonstrating the production of ferric ions. The ferrous salt 

 should be as free as possible from ferric admixture. 



