77 The Silver Voltameter .
Part I.\#151 ; By F. E. Smith , A.R.C.Sc .
, and T. Mather , F.RS .
Part II.\#151 ; By F. E. Smith , A.RC .
Sc. , and T. M. Lowry , D.Sc .
( Communicated by R T. Glazebrook , F.R.S. Received July 22 , \#151 ; Read November 21 , 1907 .
) ( From the National Physical Laboratory .
) ( Abstract .
) Part I.\#151 ; On a Comparison oj many Forms oj Silver Voltameters .
By F. E. Smith ; and a Determination of the Electrochemical Equivalent of Silver .
By F. E. Smith and T. Mather , F.R.S. It has been known for several years that the measurement of electric quantity by the deposition of silver is liable to inaccuracies which appear to be dependent on the size and nature of the anode and cathode and on the electrolyte of the voltameter employed .
Hence the necessity of an enquiry to ascertain the possibility of specifying a voltameter which is easily reproducible and in which an ampere-second always deposits the same mass of silver .
In some very early experiments it was found that the mass of the deposit was dependent on the mode of preparation of the silver nitrate , but on several crystallisations of the salt constant results were obtained .
Very large voltameters were experimented with .
Four of the cathode bowls had a capacity of 500 c.c. each , and in general from 300 to 400 c.c. of electrolyte were employed .
The anodes were coated with electrically-deposited silver .
With a Rayleigh form of voltameter containing an electrolyte of pure silver nitrate 52 determinations of the electrochemical equivalent were made , the current being indirectly measured by the British Association ( Ayrton Jones ) ampere balance .
The mean of the 52 determinations was IT 1827 milligrammes per coulomb , and the mean difference was 2'4 parts in 100,000 .
With a Richards form of voltameter , in which the anode liquid was separated from the cathode liquid by a porous pot , variable results were at first obtained , but this was found to be due to the presence of acid in the pots .
When the pots were baked in an electric furnace before their employment in a voltameter , constant values resulted , and the mean of these was 1T1828 milligrammes per coulomb , i.e. , practically identical with that obtained with the Rayleigh form .
Richards originally obtained a difference of eight 78 Messrs. Smith and Mather and Dr. Lowry .
[ July 22 , parts in 10,000 between the two forms , his form giving the smaller deposit .later he found the difference to be four parts in 10,000 , and recently van Dijk has found a difference of half this latter amount .
Further observations were made with a syphon and other modified forms of voltameter , and the same value , 1'11827 , was found , pointing to little or no irregularities in the large-size Rayleigh form of voltameter .
Deposits were made when the voltameter was subject to a gaseous pressure of 2'4 cm .
of mercury , and were found to be identical with those made under a pressure of 1 atmosphere .
We have thus failed to confirm the observations of Schuster and Crosslev and of Kahle .
At a temperature of 90 ' C. we found the deposits to be very slightly heavier than at 15 ' C. , but the calculated temperature coefficient was so small ( 1 x 10-6 ) that we believe the increase to be due to the action of the filter paper on the silver nitrate , as originally suggested by Kahle .
The range in the current intensities was from 05 ampere to 8 amperes , and for this range we found no appreciable irregularity .
We conclude that the Rayleigh form of voltameter as employed by us is .
reproducible to one or two parts in 100,000 , and that the electrochemical equivalent of silver is IT 1827 milligrammes per coulomb .
Part II.\#151 ; The Chemistry of the Silver Voltameter .
By F. E. Smith , A.RC .
Sc. , and T. M. Lowry , D.Sc .
Before a definite value could be assigned to the electrochemical equivalent of silver it was necessary to demonstrate the possibility of preparing again and again , from silver nitrate of different origins , solutions which should give identical weights of silver when electrolysed under identical conditions .
We prepared silver nitrate from electrolytic silver , from much used silver nitrate , and from commercial samples of the salt , and satisfied ourselves that by taking precautions in recrystallising , etc. , a sufficiently constant product could be obtained .
Tests on commercial silver nitrate were gratifying in so far that with one exception all the samples ( eight in all ) examined gave figures agreeing with those obtained from the samples which were specially prepared , and we conclude that , except in measurements of high precision , the commercial salt may be used without purifying .
Attempts to confirm the observations of Novak , Rodger and Watson , Kahle , van Dijk and others , on the effect of repeated electrolysis of a solution , show that in our voltameters there may be a very small increase in the deposit with continued use of a solution , but 'nothing comparable with that obtained by the observers mentioned .
We also fail to confirm the formation , at the anode , of a complex silver salt , giving rise to The Silver Voltameter .
1907 .
] heavy deposits at the cathode , as suggested by Rodger and Watson and by Richards .
High values are obtained for the electrochemical equivalent if the solution contains oxide , carbonate , chloride , nitrite or hyponitrite .
Low values are caused by acid .
The impurities which raise the electrochemical equivalent appear to be those which are insoluble in water but soluble in silver nitrate solutions ; they are , therefore , precipitated from the impoverished solution at the cathode .
There may be slight changes in the electrolyte due to its interaction with filter paper , but the mass of the deposit is not seriously affected thereby in our size of voltameter in the course of one electrolysis .
It is inadvisable , however , in measurements of high precision to use an electrolyte more than once .
Silver chlorate and silver perchlorate appear to give normal deposits , but are more troublesome in use and have no advantage over the nitrate .

