,9o6.] LORENTZ— ON POSITIVE AND NEGATIVE ELECTRONS 107 



of the electric fluids, or, as we are to say now-a-days, on the elec- 

 trons ; if, for instance, the positive particles are more attracted by the 

 metal M' than by M, this will of course tend to produce a positive 

 charge of the first metal. A wholly different explanation that has 

 been proposed by Riecke and Drude is based on the assumption that 

 the free electrons in a metal have their share in the molecular 

 agitation by which we account for the phenomena of heat, going 

 to and fro with velocities whose magnitude is a function of the 

 temperature. The consequence of this heat-motion must be a cer- 

 tain equalization of the density (measured by the number of particles 

 per unit volume) with which the electrons are distributed over ad- 

 jacent parts of space. Hence, if at the same temperature the metal 

 M contains a larger quantity of free positive electrons than M' , the 

 first metal will lose and the second will gain a certain number of them 

 and the potential of M' will be made to exceed that of M. 



We need not stop to consider in detail these theories ; it will 

 suffice to observe that, according to both, the causes which bring 

 about the difference of potential are confined to a very thin layer 

 near the surface of separation of the two metals. Now, whatever 

 may take place in this layer, it is clear that the transfer of electrons 

 from one body to the other will go on until the causes determining it 

 are balanced by the difference of potential that is established. A state 

 of equilibrium would soon be reached in this way if there were but 

 one kind of free electrons. But if there are two, the case will be 

 different. The causes by which the positive electrons are driven 

 across the junction being quite distinct from those on which the flow 

 of the negative particles depends, the value P of the difference of 

 potential which is necessary for preventing a further transfer of the 

 positive electricity will in general differ from the value Q that is 

 required for stopping the current of negative electrons. Hence, as 

 there is but one difference of potential, a true state of equilibrium 

 can never exist, unless there be some other process that has not as 

 yet been taken into account. The only state of things that could be 

 attained by the motion of the particles we are now considering would 

 be one in which the difference of potential has such a value, inter- 

 mediate between P and Q, that the two kinds of electrons flow in 

 equal numbers towards the same side. It would be a final state in- 



