160 



SCIENCE 



[N. S. Vol. XXXIV. No. 866 



deposited silver a strongly striated and non- 

 crystalline appearance. Other strong reducing 

 agents produce similar effects, but to a less 

 extent. Cane sugar and starch do not pro- 

 duce these effects. 



It is very probable that the final effect of 

 the filter paper in increasing the weight of 

 the silver deposit is principally due to the 

 deposition of the colloidal silver (by cata- 

 phoresis) and also some of the protective or 

 " schutz " colloid-oxycellulose upon the ca- 

 thode along v^ith the electrolytieally deposited 

 silver, resulting in a breaking up of the usual 

 crystalline form of the silver so that it oc- 

 cludes greater amounts of electrolyte. Thus, 

 an unweighable amount of colloid can in- 

 crease the weight of the silver deposit by a 

 quite appreciable amount. Of eoursej the 

 weight is further increased by the actual mass 

 of the colloid deposited, but this seems to be 

 a small per cent, of the total increase. 



Silk when first used produces an effect very 

 similar to that of filter paper, due to the ease 

 with which it is partially decomposed into 

 aldehydes. After repeated use in the voltam- 

 eter, this aldehyde decomposition ceases, and 

 it renders the electrolyte strongly acid, prob- 

 ably due to its decomposition into amino- 

 acids. Since, in general, the effect of acid is 

 to decrease the weight of the silver deposit, 

 this fact probably accounts for the pro- 

 gressively decreasing values obtained with a 

 voltameter with silk septa as compared to 

 those obtained with the porous pot type. 



Porous pots, when prepared according to 

 certain specifications, have practically no ef- 

 fect upon the electrolyte so far as could be 

 determined. If not so prepared, they render 

 the electroljrte slightly acid and very faintly 

 reducing in character. This very slight ac- 

 tion is probably catalytic in character, and 

 probably consists in a slight reduction of the 

 silver nitrate to colloidal silver with the 

 formation of an equivalent amount of nitric 

 acid. This action lies at the basis of electro- 

 stenolysis. Of all the septa, the porous pot 

 is by far the most inert. 



In addition to colloidal silver, certain other 

 colloids produce similar effects, as colloidal 



silica and colloidal silver hydroxide, whereas 

 other colloids like hydrocarbons and starch 

 do not. It is evidently a matter of whether 

 the colloid migrates to anode or cathode 

 under the conditions which oitain in the 

 voltameter. This direction of migration 

 may of course be different under other con- 

 ditions. 



Next to the presence or absence of strongly 

 reducing impurities or certain colloids, the 

 condition of neutrality of the electrolyte has 

 the greatest effect upon the results. A 

 method has been developed for defining the 

 neutrality to within a part in a million of 

 nitric acid. Fused silver nitrate when pre- 

 pared according to the methods described in 

 the literature is both slightly basic and also 

 contains minute traces of reducing impuri- 

 ties which are very objectionable in voltameter 

 work. Eecrystallized silver nitrate (from 

 neutral solution) is slightly basic, conse- 

 quently the recrystallization must be made 

 from acid solutions. This acid may be re- 

 moved by recrystallization or by careful 

 fusion, but the neutrality must be tested in. 

 either case. 



There is no evidence of any secondary re- 

 actions at the cathode. The anode solution 

 remains practically neutral, but if appreciably 

 acid at the start tends to become neutral. 

 The same is true when no septum is used. 

 This indicates a secondary reaction of some 

 sort at the anode. No evidence was found of 

 the formation of reducing substances at the 

 anode, such as a sub-silver nitrate. The evi- 

 dence obtained by other investigators of the 

 formation of such substances seems to be at 

 least partially vitiated by the fact that in 

 some cases at least the anode solution was 

 filtered through filter paper. In other cases, 

 the character of the filter is not mentioned. 



No considerable work has as yet been done 

 on the last question (7). 



In the future work, the electrolyte will be 

 subjected to ultramicroscopic examination 

 for the presence of colloids. 



A. S. McDaniel 

 BUEEATJ OF Standards, 

 Washington, D. C. 



