SURFACES OF DISCONTINUITY 703 



component of the mercury, while assuming that the Mr's of the 

 other ions do not alter). This would require that when da is 

 zero q + ^iTi should be zero, and would imply that in this 

 condition the charge on the mercury is just balanced by the 

 electric charge on the excess mercury ions in the solution part 

 of the region of discontinuity, and that there are no anions in 

 this part, or if there are, their charge is neutralized by other 

 cations situated there also. The truth is, however, that such 

 assumptions are not a necessary feature of the analysis. In 

 simple electrostatic theory, a change of electric potential 

 involves a difference of "charge" on the surface of a metal. 

 We make the hypothesis that this is occasioned by excess or 

 defect of electrons. Such electrons are a component of the 

 mercury alone. The mercury ions may travel in or out of the 

 solution across the interface. Other ions do not leave the 

 solution. The change in the concentration of the mercury ions 

 in the solution occasions changes of concentrations in the other 

 ions in the solution, but it does not necessarily follow that these 

 changes produce a change in each Mr which is exactly 

 equal to /3r5F. Indeed, electrocapillary curves constitute the 

 experimental evidence which should enable us to trace the actual 

 changes in the MrS, had we sufficient knowledge of the distri- 

 bution of the various ions in the solution layer adjacent to 

 the electrode. It may seem peculiar that changes in the very 

 small region adjacent to the capillary cathode of the electro- 

 meter should be responsible for changes in the Mr throughout 

 the whole solution, for of course the M of any ion in the solution 

 must equal its M in the surface layer; but we must not overlook 

 the fact that the electrometer is only part of a complete circuit 

 containing a voltaic cell, and we must not forget the existence 

 of the large mercury anode. It is assumed that it is not 

 polarized, i.e., that its surface has on it the normal positive 

 charge; it is not electrically changed. Now this might be quite 

 consistent with a different distribution of ions in the layer of 

 solution adjacent to it; fewer cations and fewer anions in this 

 layer could still provide just the correct negative charge in this 

 layer to balance the unchanged positive charge on the mercury- 

 anode. On changing the external E.M.F. by 8E(=—8V) 



