376 



92 



paper the theory agrees with the observations also in the case of the fine structure of 

 the other Unes (8^4), (9-^4), ... of the series in question; the stronger of the 

 observed components may be ascribed to transitions to the final states (04), 03), (22), 

 the weaker to transitions to the final state (31). It need hardly be remarked that 

 it must be expected for all the lines in question that the details of the theoretical 

 üne structure are influenced and disturbed to a high degree by the presence of 

 electric fields in the vacuum tube. In fact, as seen from the observations on 4686 À, 

 the intensity of these fields seems, in case a continuous voltage was applied to the 

 vacuum tube, to have been of the order of magnitude of 100 à 300 ^'»'t cM, but fields 

 of this intensity will, as may be seen for instance from the formulae (126), be large 

 enough to change the character of the stationary stales corresponding to n = 6, 7, 8, . . . 

 almost completely. That it has been possible to observe a fine structure at all is, 

 just as in case of the fine structure of the lines (6-^3), (7^3), . . ., due to the fact 

 that the final states involved in the transitions (n"=3, n" = 4) are yet stable against 

 the perturbing influence of electric fields of this order of magnitude. 



Before leaving the comparison with the observations on the fine structure of 

 the helium lines it may be of interest to emphasize that a further test of the 

 theory maybe obtained by an examination of a possible characteristic polarisation 

 of these components with respect to the direction of the electric field in the dis- 

 charge. Thus from the tables XI, XII and XIII it will be seen that, if the phenome- 

 non is viewed in a direction perpendicular to the external electric force, we must 

 expect that all the new components will contain a greater percentage of light 

 polarised parallel to the electric force than of light polarised perpendicular to this 

 direction, and especially that the component (31-»21) of the line 4686 Â and the 

 component (41-^21) of the line 3203 Å should be completely polarised in the 

 direction of the electric field. The question of the polarisation of the fine structure 

 components seems not to have been examined by Paschen, and it also appears 

 doubtful whether such a polarisation would have been detectable at all with the 

 experimental arrangement used by this investigator, since the electric field may 

 have quite different directions at the different points in the luminous gas which 

 contribute to the formation of the spectroscopical image. 



We shall now briefly consider the hydrogen lines (iV=l) the fine struc- 

 ture of which has also been discussed in Paschen's paper. For this element we 

 must expect that, owing to the circumstance that the denominator N^' appearing in 

 formula (126) is only equal to 1, the effect of small electric fields in the vacuum 

 tube is yet much larger than for the helium lines discussed in the preceding. This 

 is in agreement with the well known experimental fact that in the case of the Balmer 

 series it is very difficult to obtain spectrograms which show a distinct fine structure. 

 In table XIII we have, as mentioned, given an estimate for the intensities of the 

 original and of the new components in case of (3 ^ 2), corresponding to an 

 intensity of the electric field of 300 Voit/cM. The values for s{R''^) in this table are 

 so large that, already for a field of löO Volt Y- m, the influence of the field on the 



