615 



and its corrections form tiie subject of the followiiio- |)aragra|)lis; 

 we will consider more e.speciall}' tlie topographic calibration of the 

 electromagnet. It was can-led out partially by means of the iiivesliga- 

 tion of gadolinium sul|)liate in liquid hydrogen and so it furnished 

 new material for the knowledge of the susceptibility of this sub- 

 stance, confirming old results. This new material will be communi- 

 cated at the same time. 



^ 2. Ap/xtrntus mid method. The magnetisation was calculated 

 fi'om the force exerted bj' an iuhomogeneous nuignetic field on a 

 small quantity of the material. For the measurement of the force 

 the same apparatus was used as in the investigalioii of gadolinium 

 sulphate in 1914, except a small alteration in connecting the tube 

 containing the substance under consideration. At that time no de- 

 scription was given, so now some details may be mentioned. The 

 apparatus was constructed by Mr. G. J. Flim, chief of the Technical 

 Department of IJje Cryogenic Laboratory, mainly on the same prin- 

 ciples as the apparatus of K.amfrmngh Onnes and Pekrier ') for the 

 investigation of paramagnetic substances. The substance to be 

 investigated is placed at the bottom part of a long rod, the "carrier". 

 This carrier is suspended to one or two floats swimming on mercury. 

 The force exerted by the magnetic field on the substance is com- 

 pensated by a known force and the compensation is checked by 

 means of a telescope and a scale attached to the carrier (>S't'. (ig. 3). 

 Some modifications were required with a view to the special cir- 

 cumstances. The apparatus is introduced at the top of the helium 

 cryostat (C) and is supported by the rim R. It is counterbalanced 

 by weights acting on the connecting tul)e between cryostat and 

 liquefactor. Tiie weight of the apparatus has been minimised. Par- 

 tially for this purpose the ringshaped trough of the apparatus of 

 Kamerlingh Onnes and Pkrhier has been replaced by a small glass 

 reservoir (6r) with only one float (D?-). The comparatively large 

 forces occurring in the experiments (up to about 200 gr.) induced 

 to prefer magnetic compensation instead of electrodynamic compen- 

 sation by two coils, though the accuracy was diminislied theieby. 

 The compensating force comes from the attraction exerted by a 

 current of suitable intensity passing through a coil D at the top of 

 the apparatus on a weak iron rod «S at the top of the carrier; by 

 putting rings [Ri) under the coil D its height can be taken such 

 as to exert upward or downward forces, as appears convenient. The 

 distance of tiie weak iron rod to the interferrum of the electro- 



1) These Proceedings 16, p. 689 and 786. Leiden Gomm. N". 139a. 



