306 Lord Kelvin on the Duties of 



§ 21* For atoms o£ electricity, which, following Larmor, I 

 at present call electrons, it inevitably occurs to suggest a 

 special class of atoms not fulfilling the condition stated in 

 lines 12-22 of § 5. 



Thus a positive electron * would be an atom which by 

 attraction condenses ether into the space occupied by its 

 volume ; and a negative electron would be an atom which, 

 by repulsion, rarefies the ether remaining in the space occupied 

 by its volume. The stress produced in the ether outside two 

 such atoms by the attractions or repulsions which they exert 

 on the ether within them, would cause apparent attraction 

 between a positive and a negative electron ; and apparent 

 repulsion between two electrons both positive or both 

 negative. 



§ 22. But these apparent attractions and repulsions would 

 increase much more with diminished distance than according 

 to the Newtonian law of the inverse square. This law, which 

 we know from Coulomb and Cavendish to be true for electric 

 attractions and repulsions, cannot be explained by stress in 

 ether according to any known or hitherto imagined properties of 

 elastic matter. But a very simple hypothesis, assuming action 

 at distances, between different portions of ether, explains it 

 perfectly. Consider two portions of ether occupying infini- 

 tesimal volumes V, V', at distance D asunder. My hypothesis 

 is that they repel mutually with a force equal to 



Q-i)V.( P '-i)V' 



where p, p' denote the densities of the two portions of ether 

 considered, and 1 is the natural density of undisturbed ether. 

 This makes the force repulsion or attraction according as 

 (p — 1), (p' — 1) are of the same or of opposite signs ; and 

 zero if either is zero, (which means that ether of undisturbed 

 natural density experiences neither attraction nor repulsion 

 from any other portion of ether far or near). 



§ 23. This closely resembles Aepinus' doctrine of the 

 middle of the eighteenth century, commonly referred to as 

 the " one-fluid theory of electricity " ; but now, instead of 

 electric fluid, we have " ether," an elastic solid pervading all 

 space. According to our present hypothesis, similar electric 

 atoms repel one another, and dissimilar attract ; in virtue of 

 force between each atom and the portion of ether within it, 

 and mutual repulsion or attraction of these portions of ether 



* It seems probable that this may be the resinous electrification, but 

 it may possibly be the vitreous. It must be remembered that vitreous 

 electrification has hitherto been called positive merely because it is it 

 -which is given by the " prime conductor " of the old ordinary electric 

 machine. 



