400 PROCEEDINGS OF THE AMERICAN ACADEMY. 



find evidence of the break in continuity (Knickpunkt) of the polarization 

 curve, which has been described by numerous observers, but which has 

 been attributed by others to erroneous observations or to incorrect inter- 

 pretation of results. 



For measuring polarization two general methods have been employed. 

 One consists in measuring directly the potential of the electrode in ques- 

 tion against a standard electrode, either while the main current is passing 

 or immediately after it is broken. The other method consists in measur- 

 ing the total electromotive force between the two electrodes during the 

 flow of a given current. The difference between this electromotive 

 force and that which exists when an infinitely small current passes, may 

 be called the total polarization, and is the sum of three terms: the 

 polarization potentials at the two electrodes, and the fall of potential 

 through the electrolyte. By diminishing the resistance in the electrolyte 

 the latter term can be made negligibly small (except in case a resisting 

 film forms at one of the electrodes). If, then, we use at one terminal an 

 unpolarizable electrode, the total polarization is a measure of the polari- 

 zation at the other electrode. This was the method adopted. 



In order to exclude all side reactions, the liquid about the cathode 

 must be carefully protected from any contamination by substances pro- 

 duced at the anode. This might be accomplished by separating the two 

 electrodes so far as to prevent diffusion from one to the other during the 

 time of the experiment. This would not only have necessitated a very 

 complicated apparatus, but, by introducing a large resistance, would have 

 prevented the simple calculation of the polarization at the cathode from 

 the total polarization of the cell. It seemed better, therefore, to look 

 for a suitable anode which could be used directly in the same liquid with 

 the cathode without introducing any foreign substances into the electro- 

 lyte. The choice apparently lay between a lead peroxide electrode and 

 an electrode of hydrogen upon platinized platinum. Since, as will be 

 shown presently, the merest traces of impurity in the electrolyte may 

 vitiate the results, it was feared that even the minute quantities of lead 

 that would go into solution from the former electrode might be dangerous. 

 We finally adopted, therefore, the hydrogen electrode in the usual form 

 of a spiral of platinized platinum foil, half immersed in the electrolyte 

 and bubbled over with hydrogen. It was not certain in advance that 

 such an electrode could be regarded for our purpose as unpolarizable, 

 but preliminary experiments indicated this to be the case for such small 

 currents as were to be used. This conclusion was completely verified 

 by our later experiments, as will be shown. 



