619 



aii(i ih placed in an iiiliomogeiieoii.s field. Dr. Bkk.it 'j has made a 

 careful investigalion in tlie case of a powder. According to liim a 

 first approxiraatioii for the demagnetisation is obtained if the formnia 

 mentioned is ap|)lied, talking for d not the density of the powder 

 itself, but of the snbstance. If necessary this correction has been 

 applied in that manner. 



dH . . . 



c. lo/iographfcni corrections. — is in tirsl approximation propor- 

 tional to the lield strength in the middle of the interferrnm : /ƒ,. 

 The factor of |)roportionality was calculated from a ballistic topo- 

 graphical calibration of the magnet '). At currents of 10 and 20 

 amp. no appreciable difference in the topography was stated and 



dH 

 for z =r 2.45 cm. ( r— being there a maximum) was found: 

 oc 



H = Q.S15.H„ 1^ = 0.199. //„ . . . (2,3) 



oz 



If however for gadolinium sulphate ') the force F is calculated 

 as a function of /ƒ„, no proportionality of F to //„' is found, as 

 might be expected on account of previous measurements *) (apart 

 from small corrections if Langevin's fornuila is followed) bnt devia- 

 tions occur up to 20 "/o- This appears from fable I and fig. 4. To 

 the observed value of F, given in the third column now first a 

 correction for the demagnetisation is applied: i^ is multiplied by 

 i. -{- i ^d^y,; according to the remark h (see above), rf, is taken 

 equal to 3 '), for ■/, the specific susceptibility, the value following 

 from the un-corrected measurements has been taken. At 20°.42 K. 

 this correction is 1.2 %, at 13°.98K. 1.8V,. '" the column headed L 

 the corrections for the deviations according to Langevin's formula 

 have been given. With those two corrections an apparent Curie- 

 constant 6" = x2' has l)een calculated. 



The values found for C' appear to be strongly dependent on the 

 field strength (cf. fig. 5). This may not be due to errors in the 



1) These Proceedings 25, p. 293; Leiden Comm. Suppl. N». 46. 



') The calibration really refers to a pole distance of 26 mm., not to 26.5 mm., 

 the distance occurring in the experiments described. 



The parameters of this field do not belong to those for which Forker has 

 given so much and such important data (J. Korrer, thesis Zurich, 1919). 



') The gadolinium sulphate, Gd^tSOJs . SHjO, originated fiom the supply previously 

 kindly sent by Prof. Urbain. Two tubes have been filled with it, Gd I and Gd II, 

 containing resp. 0.4735 en 0.4414 gr. of gadolinium sulphate. 



') H. Kamb;rlingh Onnes and E. Oosterhuis, these Proceedings 15, p. 322 

 § 6, Leiden Comm. N». I29h, § 6. 



5) P. Groth, Chem. Krystallographie 11 (1908), p. 460. 



