through Exhausted Tubes ivithout Electrodes. 463 



When h vanishes the rate of heat-production, as given by the 

 preceding expression, is infinite. We have, however, in this 

 expression used approximate values for A and B. If we use 

 their accurate values we find that the rate of heat-production 

 vanishes when A = 0, but rises very quickly to a large value 

 as h increases. 



The slight increase in the brightness of the discharge in A 

 when a brass rod is placed in B is due, I think, to the dimi- 

 nution in the self-induction in the primary circuit produced 

 by this rod, whose conductivity is so good that it absorbs 

 practically no heat. 



We will now return to the case of bad conductors, where 

 na is small ; here the absorption of energy is proportional 

 to the conductivity, and we might use this method to com- 

 pare the conductivity of electrolytes for very rapidly alter- 

 nating currents. I tried a few experiments of this kind, and 

 found, as I did in the experiments described in the ' Pro- 

 ceedings ' of the Royal Society, xlv. p. 269, that the ratio 

 of the conductivities of two electrolytes was the same for 

 rapidly alternating as for steady currents. I was anxious, 

 however, to see whether these rapidly alternating currents 

 could pass with the same facility as steady currents from an 

 electrolyte to a metal. To try this, two equal beakers were 

 filled with the same electrolyte made of such strength that 

 when inserted in B they put out the discharge in A. I then 

 placed in one beaker six ebonite diaphragms arranged so as 

 to stop the eddy currents, and a similar metallic diaphragm 

 in the other. The ebonite diaphragm made the beaker in 

 which it was placed cease to have any effect upon the dis- 

 charge in A. I could not detect, however, that the effect of 

 the beaker in which the metal diaphragm was placed on the 

 discharge in A was at all diminished by the introduction of 

 the diaphragm. I conclude, therefore, that very rapidly 

 alternating currents can pass with facility from electrolytes 

 to metals, and vice versa. In this respect electrolytes differ 

 from gases, the currents in which, as we have seen, are 

 stopped by a metallic diaphragm in the same way as they 

 would be by an eboniie one. 



It may be useful to observe, in passing, that a somewhat 

 minute division of the electrolyte by the non-conducting 

 diaphragm is necessary to stop the effect of the eddy currents; 

 a division of the electrolytes into two or three portions seemett 

 to produce very little effect. 



Another point which is brought out by these experiments 

 is the great conductivity of rarefied gases when no elec- 

 trodes are used as compared with that of electrolytes, An 



