584 MR F. E. SMITH, MR. T. MATHER, AND DR. T. M. LOWRY 



It will be seen that with one exception all the eleven solutions gave, when 

 electrolysed under similar conditions, values for the electrochemical equivalent lying 

 between 1 '11826 and 1 '11831, the mean value being T11828 milligrammes per 

 coulomb. The average error for the nine concordant solutions is '00001, and is of 

 about the same magnitude as that observed in the case of duplicate determinations 

 with the same sample of nitrate. 



The only discordant value in the above series (No. VI.) gave on a first electrolysis 

 the figure I'll 816, but this may possibly have been due to some accident of manipula- 

 tion, e.g., to the loss of a trace of loose silver from the bowl ; a redetermination of the 

 electrochemical equivalent gave the value I'll 830, a figure which differs only by 

 '00003 from the mean. One solution prepared subsequently for use in a com- 

 parative test gave the value 1'11836 (52c), but the nitrate used for making the 

 solution had been exposed to the air for a long time and was considered not to be 

 sufficiently pure for an absolute determination. 



B. Tests of Commercial Silver Nitrate. 



Having established a definite figure for the weight of deposit obtained from highly 

 purified samples of silver nitrate when electrolysed under standard conditions, it was 

 desirable to ascertain how far commercial " pure " silver nitrate could be relied upon 

 to give a correct weight of deposit. The result was encouraging in so far as with 

 one exception all the samples examined gave figures agreeing with those obtained 

 in the preceding section A. The values for the different samples were as follows : 



Hj I'll 827^ Very large number of 



H 2 1'11827 / observations (see 



H 3 1-11827 J Table I.). 



W t T11819 (226, 236, 25a, 256, 316). 



G! 1-11827 (30a). 



M T11829 (30d). 



G 2 1-11827 (38a). 



W 2 1-11826 (48a). 



The abnormal specimen W\ gave a normal deposit (IX.) when recrystallised and 

 was therefore considered to contain a removable impurity. When the 15 per cent, 

 solution was tested with neutral litmus paper it did not show any acid reaction ; but 

 a more concentrated solution tested with blue litmus showed a marked acidity which 

 was absent from the solutions which gave normal deposits. The acidity of the 

 specimen was further established by precipitating the solution with neutral sodium 

 chloride and testing with methyl orange. The makers subsequently stated that the 

 nitrate had been crystallised from a slightly acid liquid in order to secure the forma- 

 tion of clear crystals, and there can be little doubt that a trace of acid mother-liquor 



