[ 109 j 



XIV. Radiation in a Magnetic Field. 

 By Albert A. v Michelson*. 



IN the interesting and important, paper of Zeeman " On the, 

 Influence of Magnetism on the Nature of the Light 

 emitted by a Substance " f, there is a reference to the work of 

 "the late M. Fievez, who found that instead of a broadening of 

 the spectral lines there were reversals and double reversals, 

 which Zeeman has not observed. 



In some cases the magnitudes to be observed are of the 

 order of a fortieth of the distance between the sodium lines, 

 and should be clearly seen in a good spectroscope under 

 proper conditions ; but others occur in which they are but a 

 third or a fourth as large, and in these cases all detail is lost 

 in diffraction effects and optical imperfections. 



For the investigation of just such cases the interferometer 

 is particularly adapted, and it was determined to investigate 

 the problem with the aid of this instrument. 



The first substance tried was sodium. A bead of sodic 

 carbonate was placed in the flame of a small hand blowpipe, 

 which could be kept under better control than a Bunsen 

 burner. This was placed between the flat pole-pieces of a 

 moderately large electromagnet in the manner described by 

 Zeeman, and the light after passing through a collimating 

 lens entered the interferometer. The difference of path 

 commencing at zero was increased by single turns of the 

 millimeter screw, noting at each turn the clearness or 

 visibility of the interference-fringes J, first without and then 

 with the magnetizing current. 



The curves A and D, fig. 1 §, show the results of this 

 experiment. They are the envelopes of the visibility curves, 

 the alternations of which are too rapid to show on this scale. 

 The abscissas are differences in path of the interfering pencils 

 in millimetres. From these the distribution of light in the 

 source is found as described in a previous article ||. The 

 results are shown at A and D (fig. 2), the first representing 

 the appearance of one of the sodium lines without, and the 

 second with the magnetizing current. 



It is evident from the figures that the broadening of the 



* Communicated bv the Author. 



t Phil. Mag., March 1897. 



j Phil. Mag., Sept. 1892. 



§ Negative ordinates indicate reversal of the fringes. 



|| Phil. Mag., Sept 1892. 



