328 On the Magnetic Susceptibility of Liquid Oxygen. 



298 and 91 on the absolute thermometric scale, and it is obvious 

 therefore that the paramagnetic susceptibility of solid maiiganons 

 sulphate is increased by cooling so that its value varies inversely as 

 the absolute temperature, and is increased threefold by cooling to the 

 temperature of liquid oxygen. 



It has been shown by P. Curie, that the above law holds good for 

 oxygen, palladium and other paramagnetic bodies at high tempera- 

 tures ; and E. Wiedeman'n (Pogg. Ann., vol. 126, p. 1, 1885) and 

 Plessner (Wied. Ann., vol. 39, p. 336, 1890) have shown that for 

 limited ranges of temperature, 16 to 60, it is true for paramagnetic 

 salts in the liquid and solid condition. 



Our experiments show that this law may be valid over very wide 

 limits, and to very low temperatures. We hope to fully examine this 

 matter shortly, and to make a full examination of paramagnetic 

 susceptibility at very low temperatures. 



We have also made measurements of the magnetic susceptibility 

 of bismuth at the liquid oxygen temperature. 



From the mean of three experiments, we find that the diamag- 

 netic susceptibility of bismuth is increased by cooling, and that it 

 has the following values : 



Diamagnetic susceptibility of bismuth at 15 C. = 13*7, 



- 182 C. = 15-9, 



thus showing an increase of 16 per cent, on the lower value. 



The diamagnetic susceptibility of bismuth is therefore not increased 

 inversely as the absolute temperature on cooling. 



The above considerations suggest that the very large paramagnetic 

 susceptibility of liquid oxygen, which is five or six times greater 

 than that of a saturated solution of ferric chloride, may in part be 

 due to its low temperature. 



Some experiments we have made on the susceptibility of nian- 

 ganous sulphate indicate that paramagnetic susceptibility varies 

 directly as the density. 



The density of liquid oxygen is 806 times that of gaseous oxygen 

 at C. and 760 mm. /and its absolute temperature is just one- 

 third. 



Hence if the law that susceptibility varies as the density and 

 inversely as the absolute temperature holds good down to the tem- 

 perature of 182 C., and over the physical change of state, we 

 should expect the susceptibility of the liquid oxygen to be 2418 

 (= 806 x 3) times that of the gas. 



The magnetic susceptibility of gaseous oxygen as determined 

 Faraday, E. Becquerel, and others is 0'13xlO- 6 . Now 2418x0' 

 X 10- 6 = 314 x 10~ 6 . 



