.278 Dr. L. Silberstein on Fluorescent Vapours 



This seems very simple by itself. But the matter becomes 

 even more simple when it is remembered that, by (16 ), 

 ..Z/& = tan f , where f is the rotation or the angle E, r, so 

 that (23) becomes 



Jo = cos 2 ? . (24) 



Now, this is as familiar as the most common of our 

 everyday-life mechanical experiences. In absence of the 

 rotating-agent (field H) the force E pulls r in its own 

 direction, and is thus most effective ; and when these make 

 with one another an angle f , then E cos f only is operative, 

 whence, and so on. The limit of magnetic rotation is 

 ,f = 90°, but before this is reached, the fluorescence vanishes. 

 We might even have started the whole investigation by such 

 a line of reasoning. 



At any rate, it seems now indubitable that the magnetic 

 " destruction" of fluorescence (at least as far as the funda- 

 mental line is concerned) is intimately connected with a 

 magnetic rotation, and since the former is fully established, 

 I have but little doubt that the rotation will also be found 

 by appropriate experiments. In fact, if J =^, as in Wood 

 and Bibaud's experiments*, with H = 30,000, then, by (24), 

 the angle of rotation can be expected to be as huge as 



f = 71°-565. 



Again, by (23), we should have, for the same magnetic 

 field, (Z/ky = 9,i. e. 



Z=±M, 

 and therefore for H = 10,000 gauss, 



Z=±k. 

 Thus, our above H c would be 10,000 gauss, or, at any rate, 

 very nearly so. The signs correspond to rj JO, as in (17). 



For the relative intensity J x of the first line of the series 

 we have, by (14), writing n = ~Np, the rather complicate 

 formula 



_ [N 2 (l-^ 2 ) 2 + P][N 2 (l-^) 2 -P-Z 2 ] . 



1 ~ N 4 (l -p 2 ) 4 - 2N 2 (1 - P y . (Z 2 - P) - (P + Z 2 ) 2 ' { J 



for any magnetic field H. The corresponding values of J 2 , 

 J 3 , etc., for the second, third, etc., lines are obtained from 



* Properly speaking, Wood and Bibaud's experiments, dealing with 

 the whole fluorescence, authorize us to assume not J = l/10, but 



— T ( . = — - : but in absence of better information we assume here 

 2I(H=0) 10 



that the fundamental line taken by itself is reduced, nearly, in the same 

 ratio. It would be interesting to investigate spectrophotometi'ically the 

 diminution of intensity of each line separately. 



