272 Dr. L. Silberstein on Fluorescent Vapours 



frequency of the given spectral line in ordinary circumstances, 

 i. e. in absence of the magnetic field. Now the solution of 

 (12) is, apart from damped vibrations, independent of c, 



x = ke in \ y = Be int , (13) 



where A, B are complex constants. 



Thus we have, to begin with, the negative but nevertheless 

 noteworthy result that there should be no splitting of lines of the 

 fluorescent spectrum, i. e. no ordinary Zeeman effect. 



As far as I know, this conclusion is not contradicted by 

 experience. In Wood and Ribaud's paper on u The Magneto- 

 optics of Iodine Vapour" (Phil. Mag. xxvii. pp. 1009-1018, 

 1914), which to my knowledge is the only publication 

 bearing on the subject, there is no mention of any splitting 

 of the lines of fluorescent iodine, although the authors have 

 disposed of a field of 20,000 gauss *. 



The reader will notice that the ordinary equations of 

 Lorentz's elementary theory of the Zeeman effect for common 

 spectrum-lines differ from (12) by the absence of ce int (and 

 of the terms Jcx, ky). This is the reason why Lorentz's 

 equations give a splitting, while our equations refuse to give 

 any trace of it. 



However, although none of the lines of a resonance- 

 spectrum can be expected to be split by the magnetic field, 

 each of them will undergo a change of intensity and of 

 character, i. e. of its state of polarization. In fact, substi- 

 tuting (13) in (12) we have, for the constants A, B, 



A=^[W-n* + in(k + Z)], | 

 B=^[W-n? + in(k-Z)]J ' ' ' (U) 



where D = (W _ n 2 + ^ n ) 2 - Z 5 



J 



These are the (complex) values to be inserted in (13). 

 The intensities will be given by | A | and | B |, and this will 

 occupy our attention a little later (cf. Section 6) . The state 

 of polarization, which interests us here, is entirely determined 

 by the quotient y/^ = B/A, and this is, by (14), 



|~ N w+4*+Z) pg ' say ' • • (15) 



* No direct mention is made, in the quoted paper, of the absence or 

 presence of splitting- of the fluorescent lines. AYood and Ribaud's state- 

 ment (p. 1015) " that if the Zeeman effect exists, it is less than 0'01 A.U. 

 for a field of 20,000 gauss " refers to the absorption lines of iodine vapour. 



