572 REPORT— 1903. 



be magnetic. The attempt to produce pure copper wire was then abandoned, and 

 in order to reduce the magnetic control a very narrow coil was procured, and by- 

 putting pole pieces on the permanent magnet its field was made as uniform as 

 possible, since in a perfectly uniform field magnetic particles on the coil would not 

 produce any control. In order to increase the moment of inertia of the system 

 a light copper disc was attached rigidly to the coil by a light bone rod. 



This arrangement when tried gave satisfactory results as regards sensibility 

 and periodic time, but did not damp well when short-circuited. This difficulty 

 was overcome by placing below the copper disc a small electromagnet, which 

 when excited quickly brought to rest the disc by reason of the eddy currents 

 induced in it. 



There is one point of interest about the method adopted in calibrating the new 

 instrument. This is done best by the standard solenoid, as the deflections are not 

 proportional to the quantity of electricity discharged through the coU when the 

 resistance in the circuit is small. Taking a series of readings, curves can be 

 plotted showing the relation between flux-deflection for constant resistance and 

 resistance-deflection for constant flux ; or the three may be plotted isometrically. 

 This last is the most convenient method for odd deflections, but when several 

 readings are taken with the same resistance it is better to plot a separate curve 

 connecting flux and deflection for each resistance from the set of curves connecting 

 deflection and resistance for constant flux. 



8. On the Use of Capacities as Multipliers for Electrostatic Voltmeters in 

 Alternating Current Circuits. By Professor E. W. Marchant, D.Sc, 

 and G. W. Worrall, B.Sg. 



This arrangement has been devised to enable an electrostatic voltmeter to be 

 used to measure any alternating voltage higher than that actually existing 

 between its terminals. 



Two capacities having a known ratio are placed in series with each other across 

 the circuit the P.D. of which it is desired to measure, and the voltmeter is 

 shunted across one of them. If the resistance of the circuit be very high there 

 will be an alternating current flowing through the condensers, whose R.M.S. value 

 for a uniwave is given by 



T EC, C.,» 



I = ' - " amperes. 



Oj + O^ 



where Cj and C^ are the magnitudes of the capacities in farads. 

 E is the Iv.M.S. value of the applied P.D. in volts. 

 And |) = 27r X frequency. 



The P.D. between the terminals of C, will be - ^ - ^^' - ^' 



C,j9 Cj + Ca 

 Hence ^= ^i . 



E c.+c; 



By adjusting the capacities Cj and C, it is thus possible to obtain any desired 

 ratio between the voltage at the terminals of the instrument and that on the 

 circuit. 



This relationship has been worked out for a uniwave, which, of course, rarely 

 exists ; but it may easily be shown that the same ratio between the voltages on the 

 circuit and the voltmeter holds for all wave-shapes. 



There are two conditions which have to be fulfilled to ensure accuracy : 



(1) The insulation resistance of tlie condenser must be so high as to make the 

 resistance current very small compared with the capacity current. 



(2) The shunting capacity must be large compared with that of the instru- 

 ment itself. 



