54 ROSA, VINAL, AND McDANIEL: SILVER VOLTAMETER 



0.00003 and the computed probable error of a single observation 

 is 0.000025. The probable error of the mean result is O.OOOOO2. 

 The result of 55 deposits in the Smith form is: 



1.018274 volts 



The average deviation of a single observation from the mean is 

 0. 00004 and the computed probable error of a single observation 

 is 0.000084. The probable error of the mean result is O.OOOOOs*. 



The mean of the porous pot form and the Smith form is there- 

 fore 



1.01827 volts 



which we give as the final result. The electrolyte was so pure 

 that no volume effect was observed in either form of voltameter. 



With other forms of voltameter we have obtained the follow- 

 ing results: Siphon form (relative experiments, 5 observations) 

 1.01832; Modified siphon (7 observations) 1.01835: Poggendorff 

 form (9 observations) 1.01830. We thus have results with five 

 forms of voltameter all confirming the decision of the International 

 Committee, but the difficulties arising from the anode slime in the 

 last two make them more uncertain and difficult to use. The 

 siphon voltameter requires so much electrolyte that the presence 

 of the slightest trace of impurity increases the ''volume effect" 

 noted previously in several other forms. 



We made a number of miscellaneous observations on the effect 

 of various impurities, but will mention here only the excessive 

 deposits obtained when considerable CO2 is present in the atmos- 

 phere above the voltameters. Two possible theories are given in 

 the complete paper to account for the slight difference occasion- 

 ally appearing in the deposit between the Smith form and the 

 porous pot form. 



A further study of the effect of acid on solutions known to he 

 impure has shown the effect to be quite variable depending on the 

 nature of the impurity. This fact we believe accounts for much 

 of the conflicting opinion found in the literature on this subject. 

 We have endeavored to find just what the action of acid is by 

 which it reduces the deposit from pure electrolyte, but our results 

 are chiefly negative. It seemed possible that in the case of an 



