51 



335 



served hy Stahk. Thus the (lisplacciiuMils of these comjxiiu'iils from llie original 



positions of the lines are all entire multipla of the (luantitv „ , which is the 



same lor all lines of the spectrum, and which is proportional to the intensity of 

 the electric force. Moreover the Stark effect of the hydrogen lines was found to he 

 symmetrical as regards the displacements as well as the intensities of the compo- 

 nents. This agrees with the fact that to each component on one side of the position 

 of the undisplaced line there will correspond one on the other side which is oh- 

 tained hy interchanging the values of n\ and /»',, as well as of n" and /?'', and for 

 which the value of J will he numerically equal to that for the first component 

 but of opposite sign. That two such components will also appear w'ith the same 

 intensity is directly explained by observing that the motion of the electron in two 

 stales of the atom for which /.j and /j -j- /., are the same, but for which the value 

 of /j in the one is ecpial to that of /., in the other, will he symmetrical so that the 

 a-priori probabilities of spontaneous transition corresponding to the two compo- 

 nents must be expected to be equal'). 



Looking apart from the symmetry of the etïect, the relative intensities with 

 which the components appear on Stark's photographs vary in an irregular way 

 from component to component, but are independent of the intensity of the electric 

 force. Further, as regards the polarisation. Stark found that, when viewed in a 

 direction perpendicular to that of the electric field, the lines show a number of 

 components polarised parallel and a number of components polarised perpendicular 

 lo the direction of the field. When viewed in a direction parallel to the field, only 

 the latter components appeared, with the same intensity distribution, but without 

 showing characteristic polarisation. It was pointed out by Epstein that the polarisa- 

 tion of the com[)onents obeys the rule, that the components which, according lo 

 (111), would correspond to transitions for which n'^ — /?',' is an even number arc 

 polarised parallel to the direction of the field, while components, which would 

 correspond to transitions for which n'.. — /j" is uneven, are polarised perpendicular 

 to the field. On Hohr's theory this rule receives an immediate explanation because 

 according to this theory, as it has been discussed on page 45, only two kinds of 



M Here we liave looked apart Iroiii the interesting dissyinmeti y in tlie intensities of tiie eonipo- 

 nents of tlie hydrogen lines, wliicli under certain experimental conditions appears in Stauk s observat- 

 ions, and which consists therein that the components on the red side of the position of the original 

 line appear more, or less, intense than those on the blue side according as the direction of the electric 

 field is the same as, or the o|)i)osite of, the direction of propagation of the positive rays by means of 

 which the hydrogen lines are excited (see .1. Stauk, loc. cit. p. 4(l). This dissymmetry affords, as pointed 

 out by Bohr (Phil Mag. XXX, p. 404 (1915)), an interesting support for the general principles underlying 

 the application of relation (1), bccau.se it indicates directly that the different components correspond to 

 entirely different processes of radiation the relative occurrence of which may depend on the experimen- 

 tal conditions. Thus the dissymmetry in question must be ascribed to the fact that, under the men- 

 tioned conditions, the number of atoms in the vacuum tube present in a state (u, />. a and in a state 

 (/), a, c) will no more be equal to each other l)Ut will depend on the orientation of the electric lield 

 relative to the direction of the positive rays. 



43* 



