LEWIS AND JACKSON. — POLARIZATION ON MERCURY CATHODE. 403 



(not shown in the figure) based on the Poggendorff method. In order to 

 avoid changes in the polarization the potentials were first compensated as 

 nearly as possible with a Lippmann electrometer, and then more exactly 

 with a delicate galvanometer. 



Since the internal resistance of the cell was low and the currents 

 small, the fall of potential in the cell never amounted to more than a 

 small fraction of a millivolt and could therefore be disregarded. Having 

 then a cell with an unpolarizable hydrogen anode and zero fall of potential 

 in the liquid, the total potential between the electrodes is a direct measure 

 of the polarization of the cathode, for the whole process of the cell con- 

 sists in hydrogen dissolving at the anode and forming again at the 

 cathode, and the whole counter electromotive force of the cell is therefore 

 merely a measure of the work lost in that irreversible process at the 

 cathode which is responsible for the phenomenon of over-voltage. 



The importance of using pure materials in this work is illustrated by 

 the experiments of Walker and Paterson*. According to these authors, 

 in the amalgam soda process where a solution of sodium chloride is elec- 

 trolyzed with a mercury cathode, the presence of minute quantities of 

 metallic impurities sometimes causes an enormous difference in the pro- 

 portion between the amounts of sodium and hydrogen deposited. Thus, 

 for example, the yield in sodium was lowered from 100 per cent to less 

 than 70 per cent by one part in 10,000 of iron, cobalt, or nickel. Iu 

 other words, the impurity in some way f assists enormously the formation 

 of hydrogen from hydrogen-ion at the electrode, the very reaction with 

 whose velocity we are concerned. 



The sulphuric acid used in our experiments was prepared by the distil- 

 lation of the best acid obtainable for analytic purposes. For preparing 

 the mercury a new method was employed, which seems more trustworthy 



* Trans. Amer. Elec. Chem. Soc, 3, 185 (1903). 



t Mr. R. A. Hubbard has been kind enough to make a few experiments to verify 

 the work of Walker and Paterson. He corroborates fully their main results and 

 finds, as they did, that the formation of hydrogen at the cathode is associated 

 with tlie formation of a scum on the mercury surface. In order to obtain some 

 idea of its nature, the experiments were repeated in a weak acid solution. In this 

 case the scum did not appear, nor was the yield of sodium appreciably affected 

 by the impurities. It is probable, therefore, that the scum is due to the hydroxide 

 of the metallic impurity and that it favors the formation of hydrogen in the same 

 way that platinum favors its formation from zinc in the presence of acid, by lowering 

 the over-voltage. Notwithstanding this evidence of the comparative harmlessness 

 of metallic impurities in an acid solution, we have deemed it advisable to take all 

 possible precautions to avoid impurities of every kind. 



