the Winding of Voltmeters. 101 



instrument becoming heated by the passage of the current 

 through them. 



2. An error arising from the sensibility of the instrument 

 being temporarily varied by external magnetic disturbance. 



3. An error arising from certain types of voltmeters altering 

 in sensibility so that their constants vary from time to time. 



4. An error arising from the volt standards employed by 

 different makers of voltmeters differing several per cent, from 

 one another. 



The variation in the resistance of a voltmeter produced by 

 changes of temperature of the room can of course be made 

 very small by winding the instrument with German-silver 

 wire, or better with platinum-silver, or, best of all, with wire 

 made of platenoid, or it may be made absolutely nought by 

 use of an outside resistance of carbon ; or, lastly, if the volt- 

 meter be simply wound with copper wire, this error, although 

 large, can be accurately allowed for if the temperature of the 

 voltmeter be known. But the error arising from the heating 

 of the voltmeter by the passage of the current through it is 

 much more difficult to allow for, as this rise of temperature 

 of the instrument will be unknown. Hence it is important to 

 reduce this error to a minimum, and it does not at all follow 

 that the employment of German-silver wire or even platenoid 

 wire will effect this result; indeed we have already given, in 

 a previous communication made to this Society*, general 

 reasons showing that German silver is undesirable to be used 

 in the winding of voltmeters. When, however, the diameter 

 of the coil of the instrument exceeds a certain limit, it would 

 appear that German-silver wire might be better for the outer 

 coils: hence the following investigation has been made for the 

 purpose of ascertaining whether in an instrument of the size 

 of our Solenoid, or Long-Range Voltmeter, any portion of the 

 coil should be wound with German- silver wire, and if so, how 

 large a portion, also what should be the law of variation of 

 sectional area of the copper and German-silver wire with the 

 diameter of the convolution in which it is wound, so that the 

 heating-error produced by the passage of the current through 

 the voltmeter shall be a minimum. 



Although the dimensions of the instrument adopted in the 

 following investigation are those of this special form of volt- 

 meter, the mode of considering the problem will apply equally 

 well to any other galvanometer used as a voltmeter, or to the 



* " Direct-reading Electro-Measuring Instruments, and a Non-Spark- 

 ing Key," Proc. Phys. Soc. vol. vi. p. 69 [Phil. Mag. [5] vol. xvii. p. 304 

 (April 1884)]. 



