118 On the Winding of Voltmeters. 



measured ; the current was supplied by accumulators and in- 

 creased or diminished by a carbon resistance worked with a 

 screw. By watching the ammeter and slightly turning this 

 screw one way or the other, the current could be kept very 

 constant for a long time. The value of the electrochemical 

 equivalent employed for copper was 0*3295 milligrammes per 

 coulomb, which is deduced from the mean of the electro- 

 chemical equivalents of silver, 1*118 milligrammes per coulomb 

 as given by Lord Rayleigh, and 1*1183 by F. and W. Kohl- 

 rausch, and by using 63*47, 107*56, 32*07, and 16 as the 

 atomic weights of copper, silver, sulphur, and oxygen re- 

 spectively. The solution of copper sulphate employed was a 

 saturated one, the crystals of copper sulphate used in making 

 it being chemically pure. 



To avoid errors being introduced by oxidation of the copper 

 that had just been deposited on the copper plates, they were, 

 on being taken out of the sulphate of copper bath, on the 

 stoppage of the current, immediately and quickly washed, 

 first in water, secondly in alcohol, thirdly with ether poured 

 over them, and then dried. 



The currents that a given set of accumulators will send 

 through a copper voltameter depends of course partly on the 

 distance between the alternate positive and negative plates, 

 and partly on the depths to which they are immersed in the 

 saturated solution of copper sulphate. The proper arrange- 

 ments of the plates to allow any particular current to pass 

 cannot conveniently be determined with the plates that are 

 going to be used for the experiment, as an unknown weight 

 of copper would be deposited on them during the process of 

 adjustment, and great delay would be introduced if the plates 

 had to be dried and weighed after the adjustment and before 

 making the final experiment. Hence after the proper plates 

 to be used as the anodes had been very carefully cleaned and 

 weighed, a spare set of exactly the same size and shape were 

 put into the bath, and their distance, as w T ell as the carbon 

 resistance, adjusted until the desired current was passing; 

 then the spare set was removed and the weighed set substituted 

 for them, and the experiment carried out. 



Having determined our volt-standard in the way above 

 explained, we tested various voltmeters of different makers, 

 and we found that whereas the indications of some of them 

 agreed very well with our standard, the indications of others 

 differed by several per cent. 



