142 Prof. J. Zeleny on the Ratio of the Velocities of the 



And as the two ions travel with different velocities for the 

 same potential gradient there is a greater excess of positive 

 at the negative plate than there is of negative at the positive 

 plate. That this is the case is saen by referring to Table IV., 

 where it is shown that the amount of positive blown out by 

 the rapid blast always exceeds the amount of negative 

 obtained. 



Thus far the rate of ionization has been supposed uniform 

 for all points between the plates. This would be the case if 

 the rays acted uniformly upon all of the air between but did 

 not strike the plates themselves, and under those conditions 

 the free charges at the plates would be independent of the 

 kind of metal of which the plates are made. When, however, 

 the rays impinge upon a metal it is known that there is an 

 increase in the ionization of the gas at the surface which 

 depends in amount upon the kind and state of the metal used 

 (Perrin, Comptes Rendus, t. exxiv. p. 455). This increase in 

 the ionization at the surface increases the number of the ions 

 that are moving away from the plate, and as these mingle 

 with those travelling towards the plate w T hich usually produce 

 the excess there, this excess is decreased in amount, and in 

 some instances it is even reversed in sign. 



Again, if the rays are allowed to pass through but a small 

 central portion of the space between the plates, then the free 

 charges extend on either side at least from the plate to the 

 part exposed, for the electric forces bring ions of but one sign 

 to each of the unexposed portions of the space. 



It is clear from the nature of the polarization that has been 

 shown by the above experiments to be present, that it has 

 but a short duration after the electric field which produced it 

 has been removed. 



Still the following arrangement served to show its existence 

 directly by obtaining a reverse current. Two parallel, 11 

 centim. square, insulated, aluminium plates were used, one of 

 which was connected to a pair of quadrants of an electrometer 

 and also at first to earth, while the other was connected to 

 one pole of a battery of cells. A mercury commutator was 

 constructed by means of which, in rapid succession, the 

 charged plate could be disconnected from the battery and 

 connected to earth, and then the earth connexion leading to 

 the second plate be broken. This was all done in a small 

 fraction of a second, and the motion of any charges remaining 

 in the gas after that time would produce an effect upon the 

 plate connected to the electrometer. 



With the distance between the plates about 10 centim, and 



