Insulating Properties of Indiarubber. 519 



static voltmeter, an E.M.F. as high as 20 volts (see Note on 

 the Electrification o£ Dielectrics by Mechanical Means*). 

 The energy thus existing as electric strain in the dielectric 

 is only given up very slowly, especially if the condenser is 

 not kept short-circuited ; and it appears possible that the 

 discrepancy noticed in the quantities absorbed and given up 

 by the okonite cable may be thus accounted for. The deter- 

 mination of the true conduction through a dielectric resolves 

 itself into finding the rate of conversion into heat of the 

 electric energy at any instant during electrification. The 

 value of the conduction-current after prolonged electrification 

 would be given by the constant c 0j assuming the equations 

 given above to hold when the time of charge is very great. 

 It does not necessarily follow, however, that, out of the whole 

 -current shown on the galvanometer after a few minutes' 

 electrification, c represents that portion which is being- 

 converted into heat. We might expect that the rate of 

 conversion into heat would be proportional to the rate of 

 production of strain in the dielectric (see Ayrton and Perry, 

 "On theViscosity of Dielectrics," Proc. Roy. Soc. vol. xxxvi.): 

 from this it would follow that the true conduction current is 

 greater than c at the commencement of electrification, but 

 .decreases to that value as electrification proceeds. 



The Alternating Pressure Experiments. 

 14. The experiments on the effect of alternating pressure on 

 indiarubber were commenced at the Durham College of 

 Science, under the direction of Prof. H. Stroud, and were 

 afterwards continued at the Pender Electrical Engineering- 

 Laboratory under Dr. J. A. Fleming. The material tested 

 at Newcastle was a pure rubber A r ery similar to No. 1 (see 

 Section 4), and was made into condensers in the manner 

 described in that section. The preliminary experiments 

 showed that breakdowns under alternating pressure occurred 

 generally at the edges of the tinfoil, and that pale lines of 

 light could be observed round the edges of the tinfoil when 

 under the action of 2000 alternating volts in a darkened 

 room. A set of five condensers, called A, B, 0, D, and E, 

 were then made from the same rubber ; of these A aud B 

 were shellacked round the edges to prevent the discharge 

 taking place from the tinfoil. The condensers were then 

 kept continuously on a transformer giving 2000 volts. When 

 a fault occurred in any condenser, the tinfoil was cut away 

 from the place were the rubber was pierced and the condenser 

 put to the transformer again. The results obtained are given 

 * Phil. Mag. Aug. 1901. 

 2 M2 



