Supplement to ''Nature,'' Atigust 25, 1923 



o^/ 



along the zinc sulphide screen. Fig. 2 shows the 

 result when a thickness of mica corresponding in stop- 

 ping power to 3-5 cm. of air is placed over the source. 

 The main band^ due to He.,. ^. particles, is sharply defined 

 on the high velocity side, but there is evidence of some 

 heterogeneity produced in the beam by its passage 

 tlirough the mica. As we should expect, the midway 



Dkflexion in mm. by Magnetic Field 

 Fig. 2. 



l)and (He.,, particles) lies exactly between the zero 

 position and the main band and contains only about 

 1/55 of the particles in the main beam. Fig. 3 shows 

 the distribution when the thickness of mica is increased 

 to correspond to a stopping power of about 6 cm. of 

 air. Both the main and midway bands are no 

 longer sharply defined as in the first case, but each 



Dl'.II.i: \l(iN iN MM. 1:N MM.M.IK I'll. 1.1) 



^ consists of particles with a con.si(Ieral)l(> rnnLic of velori- 

 W ties. TIlc relative numlier of lie , and He. . particles 

 is about 1/8 for tlu- .swilici- parlidcs, but llii.s ratio 

 increases with dcrri'a.iiiiL; \cl(i('it\'. Tlic nrdwiM, hand 

 CKlends and jnin ^ i lie main baud whrrr it 1 ;iii 11 n j.in-rr 

 he 1m11ii\v((1. 'i'iic lu'iiJiMii ss of the scintillations due 

 1o lie, iKii'iiiIr-, i,dl ; ('!' ' ; K' and coiitinuonsly 



from .\lo 1'-. .\' liii-^^i. inif ncu'i i'mI par] iclcs 



inikr ilicir appcirance. This is shown by tlir lie,, 

 hand, which is not deflected by a maLiiciir field, but 



its intensity is small compared with that of the midway 

 band. There is also a sparse distribution of faint 

 particles between the neutral and midway band, 

 probably due in part to scattering of the a-particles by 

 the edges of the slit and possibly in part due to recoil 

 atoms of oxygen and other elements constituting the 

 mica. The distribution of the charged and uncharged 

 helium particles for a still lower velocity will be seen 

 in curves A, B^ Fig. 4, which will be referred to later. 

 It is seen that the relative number of He+ to He.,.+ 

 particles has increased ; similarly, the relative number 

 of neutral particles is much greater. 



We may now consider the interpretation to be placed 

 on these observations. It is clear that the particles 

 emerging from the mica consist of doubly charged, 

 singly charged, and neutral particles, but the relative 

 number of these three types varies markedly with the 

 stopping power of the mica plate. We may suppose 

 that the a-particle in passing through the outer electron 



Deflexion in mm. by Magnetic B'ield 

 Fig. 4. 



structure of the atoms in its path occasionally removes 

 and captures an electron. This electron falls into a 

 stable orbit round the doubly charged helium nucleus 

 and moves with it. 



This singly charged atom will, however, have only a 

 limited life, for in passing through other atoms the 

 electron is knocked off and the singly charged a-particle 

 reverts back to the doubly charged type. This process 

 of removal is analogous to the ordinary process of 

 ionisation where an electron is ejected from an atom by 

 a collision with an a-particle ; for as a singly charged 

 particle can remove electrons from another atom, so 

 there is a chance that the He.,, particle should lose its 

 attendant electron. We may thus consider that two 

 opposing processes are at work, one resulting in the 

 capture of an electron and the other leading to its 

 removal. From the data given later it will be seen 

 that this process of capture and loss may repeat itself 

 more than a thousand times in the flii^hl of an a- 

 particle, so that the a\('ra^e path lra\elk'd by an a- 

 pariiele before capture of an electron or before loss of 

 the captured electron is small compared with the total 



