140 Decrease in the Paramagnetism of Palladium. 



The temperature throughout the experiment varied within 

 a few degrees o£ 15° C. 



It will be seen from the Tables that the figures for what 

 has been called the susceptibility-gradient show no systematic 

 variation, in spite of the fact that palladium itself shows a 

 marked decrease in susceptibility with increase of field. 

 This result might lead us to suppose that the diamagnetism 

 of the hydrogen absorbed is simply added algebraically to 

 the paramagnetism of the palladium, if it were not that the 

 value obtained for the susceptibility-gradient due to the 

 absorbed hydrogen, 7 x 10 -8 , is probably much greater than 

 the diamagnetism of hydrogen gas. It is true that the 

 latter has never been satisfactorily determined *, but it is 

 probably beyond the range of the methods used. Again, on 

 Langevin's (4) theory, if the diamagnetism of hydrogen itself, 

 whether in the form of molecules or atoms, were equal to 

 the value here found, the radius of the orbit of the electrons 

 in the hydrogen molecule or atom would be 2'4 x 10~ 7 cm., 

 which is certainly too large in either case. The observed 

 effect is therefore probably due to some interaction of the 

 hydrogen and the palladium, whereby the latter's para- 

 magnetism is reduced. It should be noticed that if the sus- 

 ceptibility-gradient due to the absorbed hydrogen remains 

 constant till saturation is reached, since palladium will 

 absorb 900 times its volume of hydrogen, the resulting com- 

 bination should be diamagnetic. Unfortunately no pains 

 were taken to saturate the palladium, and the apparatus 

 was taken down to make way for an experiment on the 

 susceptibility of hydrogen gas before this point was noticed. 



I wish to express my best thanks to Sir E. Rutherford 

 for allowing me the use of: all the apparatus required for 

 this research. 



* Kammerlingli Onnes and Perrier (Ak. Wet. Amst. xx. p. 81, 1911) 

 have made a rough determination of the susceptibility of liquid hydrogen ; 

 .also Pascal (Attn. Chim. Phys. viii. ser. 19, p. 28, 1910) deduces a value 

 for the susceptibility of hydrogen from that of its compounds. Both 

 these determinations give a value of about 2*4 X 10" for the suscepti- 

 bility per cm. 3 , which would mean that the radius of the electronic 

 orbit in hydrogen is about l # 2xl0~ 8 cm. Bohr's theory (Phil. Mag. 

 xxvi. pp. 5, 863, 1913) gives -5xl(T 8 cm. 



