618 



In ever}' set, i. e. every measurement of llie t'oi'ce corresponding 

 to a definite value of tlie magnetic field and a definite temperature, 

 the intensity of tiie current in the coil D necessary to bring the 

 carrier into a chosen zero position was read the magnetic field lieing 

 "of" and "on". These readings were taken for both directions of the 

 currents in the coil and in the magnet. 



^ 3. Corrections, auxiliary measurements and sources of errors. 



a. Forces on the carrier without sample. These forces appeared to 

 be not quite negligible and they increased with decreasing tempera- 

 ture. Investigation of the different parts of the apparatus showed 

 that those forces were caused especially by a small screw at the 

 bottom of the carrier (near V,). The comparatively large increase 

 of these forces when the temperature falls from 20° to 14° K. is 

 very striking, e. g. 70 amp. passing through the electromagnet the 

 attraction amounts to 



0,259 gr. at atmospheric temp. 



0,326 „ 20° K. 



0,350 „ 14° „ 



This is not what would be expected if the brass of the screw 

 mentioned contained iron as an impurity. Further, such acompara- 

 tively very large increase in the liquid hydrogen region would 

 give reason of suspecting much larger forces in the range of 

 helium temperatures. Howevei', tiiey are then not large as a|)pears 

 from there being no systematic difference between the observations 

 in which the mentioned parts of the carrier were certainly below 

 and those in which they were certainly at some distance above the 

 liquid helium leveP). Particular circumstances |)revented determining 

 those forces (whose comparatively large increase in the hydrogen 

 region appeared firstly afterwards) at helium temperatures and in the 

 light of the foregoing remark it seemed not absolutely necessary. 

 In the following observations the correction for the forces on the 

 carrier without sample has been applied for the hydrogen tempe- 

 ratures only. 



b. Correction for demagnetisation. This con-ection may attain 

 considerable values at the temperatures of liquid helium. In the 

 case of a sphere of a homogeneous substance of density d in a 

 homogeneous tield the demagnetising field is — ^.tad. In our expe- 

 riments the cirumstances did not correspond exactly to these con- 

 ditions. The sample is a powder in the shape of a small cyli^ider 



^) Gf. the following communication § 3 note. 



I 



