﻿Mechanism of Electrical Conduction. 61 



in order to take a sufficiently unfavourable view of the 

 question, let us assume the value to be as high as 20. Taking 

 the specific resistance of iron in electromagnetic measure to 

 be 10,000, and remembering that 1 ampere ='1 absolute 

 unit, we have for the electromotive intensity 1000 electro- 

 magnetic units of potential per cm., i.e. 1000h-(3 x 10 10 ) 

 electrostatic units per cm. Hence the electrostatic energy 

 per c. c. due to the impressed E.M.F. 



_ 20 



"87r(3xl0 7 ) 2ergS; 



while to calculate the thermal energy per c. c. at "a bright 

 red heat " — the temperature of the iron in the British Asso- 

 ciation experiments — we have: — 



Temperature above absolute zero (say) = 727 + 273 



= 1000 Cent, degrees, 



Density of iron =7*8, 



Specific heat =*113, 



One gram-water-degree of heat . =42 x 10 6 ergs. 



Thus (roughly speaking) the thermal energy per c. c. reckoned 

 from absolute zero 



= 1000 x 7-8 x '113 x 42 x 10 6 ergs. 



A comparison of these results gives 



electrostatic energy due to impressed E.M.F. 1 



thermal energy ~~4xl0 22 



only, even on our assumption that the specific inductive 

 capacity of iron in electrostatic measure is as high as 20. If 

 we suppose that half the thermal energy is potential and half 

 kinetic, then the electrostatic energy would be 1-z- (2 x 10 22 ) of 

 the thermal kinetic energy ; that is, would be equal to the 

 additional energy required to increase the existing velocity of 

 every particle by one part in 2 X 10 22 . When due account is 

 taken of these results it is not surprising to find that in iron 

 at a given temperature the specific resistance for a current- 

 density of one ampere per cm. 2 differs from the specific 

 resistance for an infinitesimal current-density by less than one 

 part in 10 12 . 



The same remarks apply with even greater force to platinum 

 and German silver, the other metals examined by Prof. 

 Chrystal, since the magnetic influence of the current on the 

 resistances of these metals must be far less than even in the 

 case of iron. 



From considerations similar to these, we should expect in 

 all true conducting substances (even in those having marked 



