Thin Metal Partition in a Voltameter. 293 



Inspection of the Tables will show that in 30 per cent. 

 H 2 S0 4 the gold-leaf partition will pass about ten amperes 

 before gases are developed upon it. Platinum-leaf and 

 aluminium -leaf showed sensibly the same " critical current/'' 

 rather less for the aluminium, which was, however, thicker 

 though not so free from holes. 



A solution of Cdl 2 showed regular deposit of both Cd and 

 I upon the partition for the weakest current to 0*001 ampere. 

 This case was peculiar in having both ions solids and indicated 

 as a consequence that neither would pass through the gold- 

 leaf partition. 



Using 9 per cent. AgN0 3 solution, Ag 2 2 was deposited 

 on the anode, but this came about by secondary action at 

 the anode ) and though we have in this case also two solids 

 deposited on the electrodes as the result of the electrolysis, 

 one of them is not primarily a solid, and the critical current 

 is of the order of magnitude of that for other metallic salts. 

 A similar case was found in that of MnS0 4 solution, where 

 Mn0 2 was deposited on the anode. 



A solution of lead acetate gave Pb0 2 on the anode, but 

 showed a very low a critical current." 



An interesting fact is the dependence of the " critical cur- 

 rent " upon the concentration of the electrolyte, as shown in 

 Table V., indicating that the " critical current " is propor- 

 tional to the conductivity of the electrolyte. I have made 

 some experiments to determine whether variation of the tem- 

 perature has the same effect upon the ie critical current " as 

 upon the conductivity, though the results are as yet hardly 

 sufficient to justify a conclusion. I have been entirely unable 

 anywhere to get a series of platinum partitions of thicknesses 

 suitable to repeat the quantitative measurements of polariza- 

 tion which were made in Berlin for the gold plates. Table VI. 

 for 0'02 millim. palladium is interesting. Either gold, pla- 

 tinum, or aluminium as thick as this behaves like a very thick 

 plate, i. e. shows gas at once in amount corresponding to the 

 current-strength. When first set up, the 0'02 millim. pal- 

 ladium partition showed a " critical current " of 0*08 ampere. 

 After considerable use the "critical current" has gradually, 

 though not slowly, increased to 0*3 in 30 per cent. H 2 S0 4 . 

 This seems to be quite distinct from the well-known property 

 of palladium of occluding gases, especially H ; for this plate 

 will pass 0*2 ampere or more for hours without showing gas, 

 and develops gas at once for a little stronger current. The 

 gas on the partition of palladium will cease to escape if the 

 current be again reduced below the " critical " value. Cur- 

 rents of several amperes have several times been passed 

 Phil. Mag. S. 5. Vol. 37. No. 226. March 1894. X 



