I906.] LORENTZ— ON POSITIVE AND NEGATIVE ELECTRONS 105 



electrons in the direction of the breadth from one edge of the plate 

 towards the other, so that, if two points of these edges, which would 

 be at equal potentials in the absence of a magnetic field, are con- 

 nected by a conducting wire, a current will be set up in the latter. 

 A similar effect, but in the opposite direction, would be produced if 

 the current were a flow of negative electrons. This is easily seen if 

 we keep in mind that the motion of the negative particles must 

 be supposed to be opposite to the nominal direction of the current 

 and that the force exerted by a magnetic field on a moving electron 

 remains the same if the sign of the charge and the direction of the 

 motion are reversed at the same time. On account of their motion 

 from right to left, the negative particles will therefore be driven 

 towards the same edge of the sheet as the positive ones in the former 

 case ; the direction of the current produced in the connecting wire 

 will therefore be reversed. 



Having got thus far, we can also see what effect will be caused 

 by the magnetic field if the current we send through the metallic 

 sheet consists of a flow of both positive and negative particles in 

 opposite directions, so that its intensity can be considered as the 

 sum of those of two partial currents, i^ and tg- We shall then have 

 a superposition of two opposite effects, either of which may pre- 

 dominate, according to the relative magnitudes of z^ and ^2. 



Now the phenomenon which the foregoing reasoning might lead 

 us to expect, has really been observed. I need scarcely tell you 

 that it was discovered by Professor Hall, then working in the 

 laboratory of the Johns Hopkins University, at a time when there 

 was hardly any question of a theory of electrons. The effect has been 

 investigated for a large number of metals and has been found to 

 have different directions in different substances. This is of especial 

 importance in our discussion, for it seems to prove that we must 

 indeed imagine two kinds of free electrons, the motion of the positive 

 ones predominating in one body and that of the negative ones in the 

 other. 



I shall now point out some difficulties which present themselves 

 in the further development of this conception of an electric current 

 as a double stream of electrons. Take for instance the simple case 

 of a current flowing across the junction of two pieces of different 



