Method of Measuring Energy dissipated in Condensers. 37 



sections, so that by choosing different sections or combina- 

 tions of sections, a wide range of inductance could be secured. 

 In this particular case the entire coil was used. The fre- 

 quency of the alternating electromotive force was varied by 

 varying the speed of the generator, complete resonance being 

 attained at a speed of 2175, for which the frequency is 145. 

 The current was 1*20 amperes, the resistance r c + r w was 

 9-82 ohms, I*{r a + rJ = 14'15 watts. The wattmeter gave a 

 deflexion of 188, corresponding to 37*6 watts. This leaves 

 23'45 watts for condenser loss, or EI cos <£. 



E being 1808, 1 = 1-20, EI = 2169, and cos 0= |^ ='0108; 



7r cot <j> = 3*39 per cent., e= 96'61 per cent. 



The quantity of the eddy-current loss in the coil does not 

 of course appear. From subsequent experiments we became 

 satisfied that it w r as large enough to cause a serious error in 

 the above value of the condenser loss. Hence we shall not 

 give any of the other values found using this coil. The 

 results obtained over a range of from 400 to 2250 volts 

 showed that the loss is sensibly proportional to the square of 

 the electromotive force. This conclusion is not seriously 

 affected by the presence of eddy-currents, since the latter are 

 themselves proportional to the square of the E.M.F. and yet 

 are not large enough to swamp the condenser loss. 



We therefore wound up a coil of nearly 2000 metres No. 14 

 wire (B and S gauge, diarn. '160 centim.), in 41 layers of 

 45 turns each, external diameter of the coil being 37 centim. 

 The eddy-current loss in this coil is less, owing to the smaller 

 diameter of the wire and the smaller quantity; a subsequent 

 measurement by an independent method gave 3*2 percent, as 

 the increase of the effective resistance by the eddy-currents 

 at a frequency of 120. At a lower frequency it would, of 

 course, be less. Its use will therefore illustrate the method 

 and give a fairly accurate value of the condenser efficiency. 



Measurements on Beeswax and Rosin Condensers. 



We give below a series of measurements on the efficiency 

 of a set of commercial condensers made of tinfoil and paper, 

 the latter being saturated with melted beeswax and rosin. 

 We understand that they are piled up dry, as we have done 

 with condensers made in our laboratory, and while immersed 

 in the melted beeswax and rosin placed in a receiver from 

 which the air is exhausted, to free them from air and moisture. 

 With the details of the process we are not, however, 



