ON ELECTEOLTSIS IN ITS PHYSICAL AND CHEMICAL BEAEINGS. 351 



been proved. I tried experiments by similar methods, and obtained yet greater 

 values of the polarisation ; it was calculated with a current density of 12 amperes 

 per square centimetre as 4'4 Daniells, and increased more and more with increasing 

 intensity of the galvanic current. It is very improbable that this can be right. 

 By supposing, however, that the resistance of the decomposition-cell is not inde- 

 pendent of the intensity, but decreases in a fixed manner with increasing intensity, 

 the calculation of the same experiments gives small and constant values of polari- 

 sation. 



II. Such a decrease of the resistance of a decomposition-cell with very thin 

 platinum wires with increasing intensity may be inferred from the following 

 reasons : — 



(a) The resistance of the cell has its principal seat in the fluid layers imme- 

 diately near the electrodes. These layers, however, grow warmer the greater the 

 intensity is, and the warmer these laj'ers are, the less their resistance will be. The 

 heating so produced can be very great. Calculation shows for my experiments 

 that (loss of heat not being taken into account), at the greatest intensities employed 

 (circa 1 ampere), the fluid immediately at the electrode is heated to boiling heat after 

 a seventh of a second. The heating is proved experimentally by the appearance of 

 a phenomenon very similar to Leidenl'rost's phenomenon, and depending on the 

 formation of a steam-case round the electrode. 



(h) With small intensities the evolution of gas does not take place over the 

 whole wire but only on some spots of it ; with the least intensities only on one 

 spot. 



(c) The diffusion in the fluid of the gases evolved by electrolysis corresponds to 

 a 'transition-resistance' (H. von Ilelmholtz, ' Sitzungsber. d. Berl. Akad.' 1883, 

 p. 604). That resistance must relatively be the more important as compared with 

 the other efl'ects in the cell, the smaller the latter are ; therefore it must appear as 

 a resistance increasing with decreasing intensity. 



These three causes produce a considerable decrease of the resistance with in- 

 creasing intensity. The hypothesis according to which the methods mentioned 

 under I. gave such very high values for the polarisation is thus not fulfilled. 



III. I have made measurements of the polarisation of very thin platinum wires in 

 sulphuric acid, by a new method, which is independent of the resistance of the 

 decomposition-cell. The circuit of the galvanic current, which contains the polaris- 

 ing battery B and the cell z, is closed by the first contact Cj of Helmholtz's pendu- 

 lum-interrupter (Wied. * Electr.' iv. p. 219, 227). A branch wire proceeds from 

 either side of Cj to a galvanometer g. This branch circuit includes the second contact 

 c„ of the pendulum-interrupter, and a very great resistance (40,000 Siemens units). 

 While Cj is closed the current is very weak in the galvanometer. In the short 

 time between the interruption of c^ and c.^ a current, strong for the sensibility of 



the galvanometer, goes through it. The impact of the galvanic current is propor- 

 tional to the electromotive force of the battery minus the polarisation. From the 

 deflection of the galvanometer I can calculate the mean value of the polarisation in 

 the time between the interruptions of e, and c.,, and by diminishing this time 

 more and more I can judge if the polarisation diiring this time has already decreased 

 considerably. I shortened that time to 0'00059 second, and diminished the surface 

 of the platina wires to 0'8 sq. mm., and increased the intensity at those electrodes 

 to 0-4 ampere. The values of polarisation which I found were never gi-eater than 

 nearly 2'4 DanieUs, and, taking account of the decrease of polarisation, its value before 



