

Inductive Capacity of Electrolytes. 205 



says * that this inductive capacity probably exists in all 

 bodies f. The capacity of a condenser consisting of two 

 electrodes immersed in a conducting medium cannot be mea- 

 sured by discharging its accumulated charge through a 

 galvanometer, and noting the deflexion of a magnetic needle. 

 But it is as directly measured by noting the force which this 

 constant (but continually renewed) charge exerts as it tends 

 to flow down a given slope of potential. The conductivity of 

 water depends mainly, if not wholly, upon its impurities, which 

 form in the case of distilled water an indefinitely small part of 

 its bulk. The inductivity, on the other hand, is a property of 

 the water itself, and hence ought not to be appreciably modified 

 by small quantities of foreign substances, though it might be 

 greatly different if they constituted a large part of its bulk. 

 We may thus think of conduction and induction producing 

 their separate effects side by side, the former not affecting the 



latter. Conduction may, however, affect the force A X K I -^— I 



dV . . 



by modifying -*- through polarization, not by changing K. 



That Maxwell's criterion of the equality of the specific 

 inductive capacity to the square of the index of refraction is 

 so far from being fulfilled in the case of water and alcohol, is 

 not a serious objection to the notion that electrolytes possess 

 a real specific inductive capacity. For comparatively few 

 substances is K = n 2 within the limits of experimental error. 

 Glass among solid and castor oil and ether among liquid 

 dielectrics are notable exceptions. These are usually explained 

 by supposing that the electrical displacements in these di- 

 electrics are much greater in a second or even a millionth of 

 a second than they would be in periods of 10~ 15 sec, a 

 thousand million times shorter, such as the period of light. 

 J. J. Thomson J has recently shown that for periods of 

 4 x 10~ 8 sec. the value of K for glass is very much nearer to 

 n 2 than for longer periods, and there is good reason for 

 believing that K for periods equal to those of light-waves 



* Electricity and Magnetism, vol. i. p. 53. 



t It is scarcely necessary to point out that the habit of reckoning the 

 specific inductive capacity of metals as infinite when calculating the 

 " reduced thickness '' of the stratum between the plates of a condenser 

 does not at all imply that if metals have a specific inductive capacity 

 it must be infinite or even large. Because of their conductivity there is 

 an instantaneous distribution of electricity such that there is no fall of 

 potential in passing through the plate of metal ; and this is what would 

 happen if the metal had an infinite specific inductive capacity and zero 

 conductivity. 



X Proc. Roy. Soc. vol. xlvi. 



