LEWIS AND JACKSON. — POLARIZATION ON MERCURY CATHODE. 405 



As a test of the serviceableness of our apparatus, it was important to 

 determine how quickly a constant current would be established after in- 

 creasing the external electromotive force. The results not only were 

 eminently satisfactory for our purpose, but gave evidence of a remarkable 

 and unique phenomenon. In cases where polarization depends upon 

 processes of diffusion either in the electrolyte or in the interior of the 

 electrode, experiments have shown that the current falls gradually after 

 any increase in the electromotive force and that it takes hours, days, or 

 even weeks for the establishment of a constant condition. In such cases 

 comparable results can be obtained only by reading the current always 

 at a definite, arbitrarily chosen time after the change in electromotive 

 force. In our present case we found that when the electromotive force 

 was suddenly increased the current became constant in less than a min- 

 ute, and that before reaching constancy the current was rising instead 

 of falling. Thus, in a typical case the current at the end of one-fifth of 

 a second was a few per cent below the final value, after a quarter of a 

 minute only about half of one per cent below. 



This unusual phenomenon could hardly be caused by the heating effect 

 of the extremely small currents, and must be due either to the breaking 

 down of some surface resistance or to the formation by the current of 

 some substance which accelerates the electrolytic reaction, that is, to a 

 kind of autocatalysis. We will presently describe another phenomenon 

 closely connected with this one. 



The constancy of the current with a given potential and the imme- 

 diate attainment of this constancy, together with the fact that neither 

 tipping the cell so as to bring a new portion of the hydrogen electrode 

 under the electrolyte, nor suddenly bubbling hydrogen over the electrode, 

 produced any change in the current, we regarded as convincing proofs 

 that the hydrogen electrode was perfectly unpolarizable for the currents 

 used. 



The method of finding the relation between potential and current 

 strength was to begin at a low potential and at the end of regular inter- 

 vals of time to raise the potential step by step to the highest value de- 

 sired, and then by similar stages to bring it back to a low potential. To 

 illustrate, let us consider a particular case, in which the external electro- 

 motive force was changed every ten minutes, the current being observed 

 at the end of this period both before and after removing the hydrogen 

 bubble. The following table gives the potential in volts and the cor- 

 responding currents in millionths of a milliampere. When the potential 

 of .652 volts was reached, the current began to fluctuate, and therefore, 



