280 Dr. L. Silberstein on Fluorescent Vapours 



and if T be the fundamental period of the iodine series A 

 (X=5461 A.U.), 



- = 1 "7 . 10-*, if v electronic. . . . (27a) 



T 



On the other hand, if rj approaches more the value of the 

 electrolytic ratio, say, for hydrogen, then T/t would be of the 

 order 10 -s . In the former case t would amount to about 

 5^00 T, and in the latter to a hundred million T. Inter- 

 mediate and other values are also possible. The decision 

 must be left to the experimental physicists. 



In connexion with this subject one more remark. In 

 Wood and Bibaud's paper we read, p. 1016, Joe. cit. : 

 (i The effect of the magnetic field in reducing the intensity 

 of the fluorescence becomes more marked as the vapour- 

 pressure of the iodine is diminished." This would mean, 

 by (23), that with decreasing pressure the time t is lengthened 

 (or the " friction "-coefficient diminished), and this seems 

 quite plausible. In fact, there are reasons to believe that 

 extinction is due not only to emission but also due to the 

 encounter of the molecules or atoms, and then t ought to 

 contain a term proportional to the mean time elapsing 

 between successive encounters. Now, the latter time is 

 certainly lengthened when, cceteris paribus, the gas or vapour 

 is rarefied. 



7. Short Note on the Amount of Polarization. 



In the above treatment of the subject we have tacitly assumed 

 that the atoms or molecules of the fluorescent vapour, in which 

 the resonators are embedded, are fixed, i. e. non-rotating, 

 with respect to the apparatus, and therefore with respect to 

 the direction of the exciting electric vector E. The result 

 has been that rectilinearly polarized incident light has given, 

 in absence of magnetic field *, fluorescent oscillations r 

 parallel to E, t. e. totally polarized. On the other hand, 

 Wood has found experimentally a partial polarization only 

 (Phys. Zeitschr. xii. p. 1209), viz. for iodine vapour only 

 17 per cent, and for sodium vapour 20 per cent., " and in a 

 very rarefied vapour and at a low temperature even 30 per 

 cent," 



Now, it occurred to me that this state of things could be 

 accounted for by assuming that the atoms or molecules 

 carrying the resonators are endowed with rotational motion^ 



* To this simple case will our attention be confined here. The same 

 problem with a superposed magnetic field will be taken up at a later 

 opportunity. 



