( 439 ) 



band, whicli is proportional to it; this widlh Avas about 1 Angstrom 

 Unil ; cii was caiculakMl from lli(> maniiidido of (he ZEKMAN-clTcct ; 

 for tliis magnitude in the lield which I used — of 9000 C. G. vS. 

 Units — 1 took Vsn i>'^''t of the iuler\al between the two sodium- 

 lines; sh was calcnlated from the value of the rotation of the plane 

 of polarisation in tiie neighbourhood of the a!)sorp(iou-l)aud ; on the 

 magnitude of this rotation as a function of tlie wavelengtli, for dif- 

 ferent intensities of tho magnetic ticld and different widths of the 

 absorption-bands, I have made measurements of which the resnlts 

 have been recorded in the tables gi\cn in my tiiesis. From these 

 tables I siiall quote one series here, giving the numbers from wiiicli 

 the above-mentioned value of f/, was deduced {ride )>. 42 of my 

 tiiesis, table 24 1) : 



Here 6 is proportional to tlie ditference between the wavelength 



of a given kind of light and the wavelength wiiich cori-ospoiids with 



the middle of tiie line />, ; the coefticient of proportion may be found 



from the tact that the ditferencc between the wa\e]eiigths of the two 



sodiumlines corresponds with a value rf=:130; / repi-esents the 



rotation of that particular kind of light in my e.\|>erimeuts, expressed 



in a ujiit of which the value is determined by the fact tliat a rotatictu 



of 180^ corresponds witli a vabie x = iOo. Tiius we read from the 



series, gi\en above, tliat for a \abie of d con-espoiidiug with 



35 , 18 



-— X Ö A. U. the rotation of the [)lane of pohirisalion is — ^ X IBO"; 



from these corresponding numbei-s tiie vabie of f/, is deduced in tlie 

 way which I explained in my thesis. 



Drl'DE, ill his LckrJnich dcr Optik wiiich I mentioned above (p. 353), 

 in ills \('i'sioii of tlie theory of dis|»ersion gi\es the eipialions of 

 motion of an electron in the form : 



