54 Journal of the Asiatic Society of Bengal. [February, 1910. 



The modification in the shape of the filter partition thus 

 produced the desired effect. The values of the e.c.e. for the 

 cathode are on the whole quite concordant. 



The arrangement was, however, still unsatisfactory, 

 because in some cases the nickel carbonate got inside the en- 

 closure through holes in the filter paper. It was also difficult 

 to avoid contact with the filter paper altogether, specially at 

 the bottom. The deposit obtained in such cases of contact 



were as a rule bad. 



In experiment 8, the anode equivalent was exception- 

 ally high. This was because the filter bag was, by mistake, 

 put on the anode instead of the cathode. The result was, the 

 acid ions which collect round the anode were inside the parti- 

 tion, and thus, protected to a great extent from the diluting 

 and neutralizing action of the fluid circulating through the 

 voltameter, they attacked the anode with greater concen- 

 tration, and its loss was consequently much greater in this case. 

 The cathode plate got the characteristic dark and loose deposit 

 as it was exposed to contact with the carbonate particles. 



A further improvement was made upon the porous parti- 

 tion by mounting the filter paper on 

 frame. A piece of thread was wound on the frame underneath 



an U-shaped wooden 



sort of net- work, which prevented the 



the cathode. About 



the paper forming a 



walls of the filter paper from touchin; 



M litres of the solution largely mixed with nickel carbonate 



were run through the cell as before. The mixture was stirred 



from time to time with a glass rod both in the voltameter as 



well as in the bottle. The results are given below. 



Table VII. 



Final Results obtained with replenishing mixture and porous 



partition. 



Second . 



840" 



840 " 



840 

 840" 



840" 



840" 

 840" 

 840" 



Ampere. 



Ca. gain. 



An. loss. 



E.C.E. (Ca.). 

 0003023 



•8803 



•2209 t 



• * 



•8930 



•2275 



0079 



0003033 



•8902 



•2279 



0262 



•0003047 



•8986 



2281 



0231 



0003021 



•9028 



■2309 



0252 



0003044 



•8944 



•2290 



•0228 



•0003048 



•9069 



•2308 



! -0282 



0003029 



•8930 



•2286 



1 -0279 



0003047 



E.C.E. (An.) 



00001054 

 00003500 



00003057 

 00003328 

 00003032 



00003698 

 00003720 



Mean e.c.e. = 0003036. 

 Probable error of the mean 

 Mean error of onf> nhsprua t i 



0000003. 

 = •0000008 



