108 LORENTZ— ON POSITIVE AND NEGATIVE ELECTRONS [April 17, 



asmuch as there would be no further change in the charges or the 

 potentials, but could not be called a state of equilibrium because there 

 would be a never-ceasing stream of neutral electricity. 



The question now arises, in this as well as in our first instance, 

 what will become of the accumulating neutral electricity ? We can- 

 not suppose this mixture or combination of positive and negative 

 electrons to be absolutely nothing, so that it might be drawn from a 

 body or heaped up in it for hours or days without any observable 

 change in its properties. We are therefore compelled to imagine 

 some new process by which the neutral electricity is carried back 

 from the places where both positive and negative electrons are con- 

 centrated towards those from which they are traveling away. More- 

 over, it is easily understood that a hypothesis of this kind can only 

 suit our purpose if the moving neutral electricity is not composed of 

 free electrons. If it were, all sections of the metallic system would 

 after all be traversed by the same number of positive particles and 

 also by the same number of negative ones and this is precisely what 

 we have begun by denying. Our conclusion must be that the neutral 

 electricity is to be regarded as a real combination, in pairs for in- 

 stance, of positive and negative electrons, a combination that is 

 formed in one part of the system and is decomposed again in an- 

 other part towards which it is carried by a kind of diffusion. 



Though this is rather complicated, we could be ready to admit 

 it, if in doing so, we could obtain a quite satisfactory theory. Un- 

 fortunately, this is by no means the case, for it can easily be shown 

 that the state of things we have now imagined would be in con- 

 tradiction with the second law of thermodynamics. Indeed, it may 

 be taken for granted that combination of a positive and a negative 

 electron will produce a certain amount of heat and that, conversely, 

 heat will be absorbed if the electrons are separated from each 

 other. If now, as we have been led to assume, neutral electricity 

 were built up in one of two metals which are in contact with each 

 other and decomposed in the other, heat would be continually devel- 

 oped in the first and consumed in the second body. By Carnot's 

 principle this can never be the case in a system that is kept at a 

 constant uniform temperature, as our two metals may be. 



If I am right in making this last remark, and if it cannot be 



