the Winding of Voltmeters. 117 



Standard No. 3 is the one that we have used for some 

 time ; and as we have obtained very satisfactory results with 

 it, it may be well to describe our mode of using it a little in 

 detail. First, by means of a large number of direct com- 

 parisons with a copper voltameter, we determine the most 

 probable value in amperes of the indications in different parts 

 of the scale of an ammeter fixed permanently in position by 

 being screwed to the table. A current is now sent through 

 this ammeter in series with a long thick copper wire having 

 about 6 ohms resistance and wound in a large open coil. 

 This coil has attached to its two ends the standard voltmeter, 

 which is also fixed permanently in position. Simultaneous 

 readings of the ammeter and voltmeter are now taken by two 

 observers, when a signal is given and the bridge-piece of a 

 mercury commutator is moved so as to cut off the current 

 and connect the copper wire together with the voltmeter, 

 which is joined with it in parallel, to a differential galvano- 

 meter. Then time measurements, for about ten minutes, are 

 taken by a third observer of the parallel resistance of the 

 thick copper wire and the attached voltmeter. Next a curve 

 is drawn connecting the resistance of the cooling coil with 

 the time, and on completing the curve backwards we ascertain 

 with great accuracy, by interpolation, the resistance at the 

 moment of stopping the current, that is at the moment of 

 making the simultaneous observations on the ammeter and 

 voltmeter. Our copper wire is sufficiently thick that for a 

 current of some 15 amperes passing through it, corresponding 

 with about 80 volts maintained at its ends, the time-curve of 

 its resistance, obtained after stopping the main current, is 

 nearly a straight line with hardly any inclination to the axis 

 along which the time is measured ; or, in other words, the 

 change of resistance due to cooling while the measurements 

 of resistance are being made is very small and easily allowed 

 for. And this resistance, in legal ohms, multiplied by the 

 current passing in amperes, gives of course the potential 

 difference in legal volts. 



In using the copper voltameter to calibrate the ammeter it 

 is of course necessary, in order that a firmly adherent deposit 

 should be obtained, that the area of the immersed surfaces of 

 the anode and cathode opposed to one another should exceed 

 the limit given by Dr. Hammerl, viz. one square decimetre 

 per 7 amperes, or rather more than two square inches per 

 ampere ; and in our experiments we have taken care to use a 

 larger area in proportion to the current. Our negative plates 

 consisted of a number of separate sheets of very thin hard 

 copper, so that the weight of the deposit could be accurately 



