in Liquids free from Gas. 151 



an apparent decomposition of water. That this happens has 

 already been seen by previous observers, but without a very 

 exact determination of the conditions. 



The course of things is but little different when, without 

 alteration of the state of the electrodes, the electrolytic liquid is 

 exhausted of gas by keeping it for weeks in the vacuum of the 

 mercury pump. I succeeded in making greatly diluted sulphu- 

 ric acid so free from gas that, on pumping out, it would not free 

 itself from the vessel, and would not break up under the nega- 

 tive pressure of a mercury column of 60 millims. height. But 

 even when distilled water was employed, I have constantly been 

 able to bring it to this — that the traces of air at any time still 

 escaping from the water, in the course of three or four days no 

 longer perceptibly augmented the pressure in the vacuum, the 

 volume of which amounted to about one sixth of that of the 

 liquid, and which contained only aqueous vapour. 



Under these circumstances also, when the plates were copi- 

 ously charged with one of the two gases, currents entered, which 

 might last for several days, but finally sank to imperceptible in- 

 tensity. The multiplier here employed showed a deflection of 

 1 degree when traversed by a current which in twenty-four hours 

 was capable of developing only 0*03 cub. centim. of hydrogen. 

 A second difference was, that, as already remarked, motion of the 

 liquid failed to strengthen the current. 



On the other hand, the influence of the gases occluded in the 

 platinum plates showed itself very distinctly when I altered the 

 quantity of the store of gas contained in both. To this end, at 

 the original charging of the plates with oxygen I conveyed by 

 electrolysis small quantities of hydrogen to the two plates in 

 conducting connexion. The second electrode was the mercury 

 alloyed with a little zinc ; the electrolytic liquid was distilled 

 water. The oftener this was done, the briefer became the dura- 

 tion both of the current produced in the voltameter by a Daniell 

 element and of the depolarization- current after the removal of 

 the element. The same course of current-intensities which at 

 the commencement, with a copious charge of oxygen, was ac- 

 complished in twenty-four hours, was finally, with the plates 

 cleared as much as possible, run through in eighteen minutes. 

 But if, after this course was entered, more hydrogen was brought 

 to the plates, the duration of the current again increased, because 

 now the plates began to be charged with hydrogen. I believe, 

 moreover, that herein the minimum of gas-charging of the plates 

 has not yet been reached, because even with the minimum 

 attained of duration of the current there remained a small dif- 

 ference of time in favour of the polarizing compared with the 

 depolarizing current. It is, however, a very tedious labour to 



