TRANSACTIONS OF SECTION A. 



655 



at the juncture of each pair of coils and one at the termination of the series. This 

 represented a circuit of a total capacity (k) of 54 microfarads, and of a total resist- 

 ance (r) of 1782 ohms, the ke being 96228. This was equivalent to an ordinary 

 submarine cable of 162 nauts length. It was not sufficient for our growing wants, 

 and another similar set was recently made, of the same dimensions and materials, 

 and in the same way. But it did not give the same results. The resistances were 

 exactly the same, but the individual capacities differed. This difference, of nearly 

 5 per cent., was, of itself, of no practical consequence for the particular purpose 

 for which the cable was constructed, but, on carefully comparing the two ' cables ' 

 with telephones, it was found that with the same nominal kr the one gave 30 per 

 cent, better results than the other. This was a surprising and unexpected result. 

 I thought it might be due to inductance, but none, neither self nor mutual, could 

 be detected. 



The insulation, taken after one minute's electrification, of each was : — 



Old cable 172'8 megohms per naut, 

 New „ 82-8 „ „ „ 



80 that the difference was in favour of the old one. But the electrification of the 

 new one was much superior to that of the old one. The difference in resistance 

 reading between the first and second minute of the former was 5\,th of that of the 

 Ifittsr 



Each cable was then rapidly charged and discharged 1,200 times, at varying 

 rates, with an E.M.F. of 40 volts, and the cumulative remaining charge in each 

 case read by deflection on a Thomson galvanometer. This was done by passing 

 a 10-foot length of perforated slip through a Wheatstone's automatic transmitter. 

 The following are the results : — 



This clearly shows that the effect is due to absorption. Electrification, polarisa- 

 tion, and absorption are phases of electrolytic work done by the current upon the 

 dielectric. The material is slightly decomposed by the leakage current, a counter 

 E.M.F. is set up, which increases the apparent resistance (electrification), it gives a 

 current of discharge reverse to that of charge (polarisation), it absorbs a quantity 

 of electricity, the undissipated part of which is recovered as a discharge current. 

 The condenser acts, in fact, like a secondary cell. 



We are now using better paper than formerly— that made from the best linen 

 rags, thinner tinfoil, and better wax, the best refined unadulterated white paraffin, 

 having a melting-point of from 125° to 128° F. Resin was formerly mixed with 

 the wax, and when hot and liquid it was painted on the paper with a brush. The 

 paper is now thoroughly dried, and well saturated with wax, by soaking under 

 pressure, so as to get rid of all the moisture. Hence there is less electrification and 

 less absorption, and consequently less retardation, and hence the superiority of the 

 new ' cable.' In the old type of condenser we observed a variation of capacity due 

 to temperature, but in the new type this has disappeared. This effect of absorption 

 on the effective capacity of the condensers, and of insulated conductors, is new to 

 me, and, as it may be new to many others, I have thought it to be of sufficient 

 interest to bring it before Section A. . . 



It is a new term to be added to the equation of energy of compound circuits, 

 and it has evidently a material influence upon the retardation of signals in long 

 submarine cables. It has an equally important influence on the efficiency of tele- 



