Metals under the Influence of Alpha Rays. 327 



swift electrons than it gains from the enclosure ; its potential 

 will rise until it can attract enough slow electrons to balance 

 this loss. The value of this steady potential will not depend 

 on the speed of the slow electrons so much as upon the size and 

 shape of the enclosure in relation to the insulated source. If, 

 as in Campbell's experiment,* the apparatus is arranged so that 

 all the electrons from either of two electrodes A and B strike 

 the other one, there is still a complication which renders 

 impossible any simple interpretation of the results. Compara- 

 tively slight changesf in the velocity of the swift secondary 

 electrons from A, for example, may alter considerably the 

 number of slow tertiary electrons emitted by B. Thus false 

 balances may be obtained which have little relation to the 

 speed of the slow electrons. In fact it seems possible that the 

 difficulties, which Campbell encountered when one of his 

 electrodes was covered with soot,:}: and which led him to adopt 

 the hypothesis that the S-electrons had no appreciable velocity, 

 may be accounted for in this way. 



When a metal is struck by a-rays it is evident that some, at 

 least, of the slow S-electrons which result are not due to the 

 direct action of the a-rays, but are produced through the inter- 

 mediary of the swifter secondary electrons generated in the 

 metal itself near its surface. The question naturally arises 

 whether the whole of the effect may not be produced in this 

 indirect way. A consideration of the results exhibited in curve 

 I of fig. 4 shows that this is not beyond the bounds of possibil- 

 ity. In this experiment we have secondary rays produced by 

 the same pencil of a-rays at two places, — the aluminium foil 

 on the wheel and the plate corresponding to E, fig. 1. The 

 S-ray current from the plate due to the a-rays is 1*24 ; let us 

 assume that it is produced as indicated above. The current pro- 

 duced by the secondary rays from the foil (with 40 volts on the 

 case) is ns = 014 ; the ratio of these currents is - ll. JSTow the 

 plate subtends at the foil a solid angle of 47r X - 15§ ; for an 

 electron generated in the place itself we may take the solid angle 

 as 4:tt. Taking the figures as they stand, this would indicate 

 that at least 11/15 of the supposed a-ray effect is produced indi- 

 rectly by means of the secondary electrons generated in the 

 plate. We may suppose that they give rise to the slow elec- 

 trons, either as they emerge from the plate itself, or as they pass 

 through the layer of gas which we have been led to postulate by 

 the results of the preceding section. 



If this should prove to be the mechanism of the emission of 

 S-rays by metals, it will at least render it probable that a simi- 



*Phil. Mag., xxi, 280, 1911. 



f See Gehrts, 1. c. 



% Phil. Mag., xxiii, pp. 54 et seq., 1912. 



§ The dimensions are different from those shown in fig. 1. 



