182 Mr Larmor, On the origin of magneto- optic rotation. 



This last result has been employed by H. Becquerel as a single 

 hypothesis (suggested by Maxwell's notion of a magnetic field in 

 this connexion as a vortex in the medium), from which to deduce 

 quantitatively both the Zeeman effect and the Faraday effect and 

 thus . correlate them, — " Sur une interpretation applicable au 

 phenomene de Faraday et au phenomene de Zeeman," Gomptes 

 Bendus, Nov. 8, 1897. He shows, employing chiefly the quantita- 

 tive results of his own previous experimental investigations, that 

 this hypothesis is capable of providing a satisfactory general view 

 of the whole range of the phenomena, and in particular that it 

 leads immediately to a simple law of dispersion for the Faraday 

 effect, namely, rotatory power proportional to XdnjdX when n is 

 the refractive index corresponding to the wave-length \ measured 

 in vacuum, a law which is in good agreement with Verdet's results 

 for carbon disulphide and creosote. 



The preceding argument forms a general dynamical justifica- 

 tion of this hypothesis for the case of all media in which the 

 ordinary gradient of dispersion is mainly controlled by one or 

 more powerful absorption bands beyond the visible spectrum, for 

 which the Zeeman constants are the same : it also shows that 

 Becquerel's hypothesis has an approximate validity when these 

 constants are nearly the same for all the effective bands. In the 

 immediate neighbourhood of any single band the dispersion is 

 anomalous, and is controlled practically by that band alone : the 

 application will then be exact, and in Becquerel's hands it has 

 given a complete account of the excessive and anomalous Faraday 

 rotation first observed by Macaluso and Corbino in sodium vapour 

 for light adjacent to the D lines. As was to be anticipated, these 

 simple general conclusions are consistent with the results of the 

 more special dynamical investigations by FitzGerald and Voigt. 



