848 



and the specific susceptibility / = 0.0036. The value obtained in 

 direct fields, according to unpublished results of Dr. H. R. Woltji<;h 

 is 0.0048 'j. 



^ 5. Results for Gadolinium Sulphate at the boiling poi7itof hydrogen. 



At tiie same frequency of 3.69 X 10'^ the gadolinium sulphate 



was balanced against a different combination of wires which when 



compared with 2 N had an effect smaller than 2 iV^ in the ratio 



1.48 



The length of the sample was 8.74 cm. and the corresponding 



10.1 



18.3 



correction -^r^r-i^r- The weight of the sample is 2.897 grams. The 



specific susceptibility is hence x = ^f*0051. The value obtained at 

 the same temperature for the same sample in a steady field was 

 0.0010 '). Measurements on manganic and nickel sulphate have 

 been also made and gave results of the same order of magnitude 

 as those for steady fields. 



^ 6. Conclusions and Discussion. 



A. Tiie order of magnitude of the susceptibility is unchanged if 

 tlie frequency is increased to 3.69 X 10'. 



B. The results seem to indicate that the susceptibility is smaller 

 than for direct fields. The values obtained in alternating fields for 

 CrClj and Gd sulphate are 0.75 ') and 0.51 *] respectively of what 



') This value is obtained by the method of weighing a rod of the material in 

 an inhomogeneous magnetic field (Kameblingh Onnes and Perrier, these Proc. 16, 

 p. 689, Leiden Gomra. N". 139fl). However, the susceptibility seems to depend on 

 the field strength, decreasing with increasing magnetic force. The value given 

 relates to a field ranging from 4500 gauss at the top of the rod to 220 gauss at 

 the bottom. The hmit for very weak magnetic fields may be about 20%, higher 

 (as found by extrapolating the susceptibility-magnetic force curves), so the ratio 

 0.75, given in § 6B for the susceptibilities in alternating and direct fields, may be 

 too large. 



') Kameblingh Onnes and Oosterhuis, these Proc. 15, p. 322, Leiden Comm. N**. 129?;, 

 § 6. However, it has to be pointed out, that it was not sure the sample was 

 really at the temperature of the bath, as it appeared afterwards, that in the 

 experiments of Kamerlingh Onnes and Oostkrhuis it took some 4 hours before 

 the susceptibility had taken a definite value, probably owing to lag in the tempe- 

 rature equilibrium. Even in order to avoid this difficulty and to ensure a better 

 temperature equilibrium of the powder and the bath, the tubes for the magnetic 

 investigations were later on not evacuated but filled with a small quantity of non 

 condensing gases (hydrogen or helium). The value 0.00051 obtained with the 

 present lube is probably too low. 



') See note § 4. 



*) See note § 5. 



