252 



From llie remiltis (ililiiiiu'd tlic oüiiclii.sion may be drawn, that for 

 the aiiiiiioiiitical solutions "z'^- .,„0 . 'i''^' 'I'so the imnihers of uiagnetons 

 luv somewhat smaller than the corresponding (jnantities for the 

 a(|ueoas solutions. For the three salts investigated this difference is 

 the same within the limits of experimental accuracy, for instance 

 this difference is for the number of magnetons 0.47 0.48 and 0.49 

 respectively for the NiSO^, NiCij and Ni(N0s)5 solution. 



§ 4. Addition of H^SO, to a aqueous solution of NiSO^ and of 

 (NHJjSO, to an ammonical solution of NiSO^ evidently was without 

 influence on the number of magnetons: 



VoNiSO, VoH.ÖÜ, n 



3.619 16.01 



3.241 9.493 16.02 



"/„NiSO, 7„NH, 7„(NH,),S0, n 



3.659 8.308 15.46 



3.187 7.238 12.884 15.48 



§ 5. Finally it has been investigated how /"• depends on the 

 temperature, by measurements of a aipieous NiCl, -solution at 6°.0, 

 16°.7 and 89°.7 C. and of an ammoniacal NiCl.-solution at 6°.7, 

 18°.8, 20°.2 and 56 \2 C. 



As only that part of the tube which was in the magnetic field 

 had the temperature T, while the rest of the tube and the basin in 

 which the end of the tube had been immersed were at the tempe- 

 rature PL', of the room, a correction must be made for the inhomo- 

 geneity of the liquid in the tube and the basin ; therefore formula (I') 

 becomes : 



_ Tt 



where y^' indicates the density of the solution at the temperature T, 

 and Yi t'le density óf the solution at the temperature t. 



The coefficients of dilatation necessary for the calculation of yr 

 have been determined : 



Coefficient of dilatation of an aqueous NiCl,-solution containing 

 4.t514 7„ NiCl, between 5°.0 C. and 18^8 C. : 0.00021 

 between 5°.9 C. and 22°.0 C, : 0.00017 

 Average: 0.00019 



