Spectrum produced by External Magnetic Forces. 257 

 Distances between the Centra in divisions of screw-head. 



fori), 



forD a 



26 



36 



18 



30 



26 



32 



45 



37 



25 



46 



28 



36 



38 



46 



42 



26 



26 



33 



33 



25 



35 



32 



53 



28 



36 



31 



51 



21 



26 



34 



26 



35 



25 



37 



31 



25 



21 



25 



Mean 32-3 + 15 



32-4 + 1-0 



On a former occasion I have found by a rough measure- 

 ment for the magnetic change 45^5, the field being 10,000 

 C.G.S. Assuming proportionality between change and in- 

 tensity the present measurements give for the said field ^-g 00 . 

 The close agreement between the result of my rough mea- 

 surement and the new one is of course the effect of mere 

 chance, for the magnetic change in the case of sodium depends 

 on temperature*, which may have been different in the two 

 cases. Probably it was not very high in the case now con- 

 sidered. The order of magnitude of e/m is entirely the same 

 as the one formerly given. 



17. The great brightness of the Groningen grating makes it 

 possible to observe very nicely also with sodium the " tri- 

 plet a," which I described for cadmium. If no nicol is used 

 the exciting of the magnet seems to break up the sodium line 

 into two lines, the phenomenon somewhat resembling the one 

 observed if a nicol is used, but with this difference, that the 

 darker part is not very dark and not narrowed; hence the 

 appearance differs considerably from the one we are accus- 

 tomed to observe with reversals. 



The explauation will be (as was remarked by Prof. Haga 

 * Phil. Mag. March 1897, p. 227, §§ 3 and 4. 



