326 Profs. J. A. Fleming and J. Dewar. 



general tendency for the susceptibility with large fields of the order 

 of 2500 to be less than the susceptibility for fields of the order 

 of 500. The average susceptibility in fields of from 2500 to 1900 

 is more nearly 310 x 10~\ and that in fields from 1100 to 500 more 

 nearly 330xlO~ 6 . 



The difference only amounts to about 10 per cent, of the lower 

 value, and it cannot be said that the observations are all of exactly 

 equal weight. 



The general result is to show that between the limits of H = 500 

 and H = 2500 the average magnetic susceptibility of liquid oxygen 

 has a value which does not differ much from 324 X 10~ 6 , but with a 

 small but decided tendency to decrease in strong fields. 



The determination of the variation of susceptibility in much 

 weaker fields is left undecided by these experiments, but by the 

 employment of a torsion balance we hope to be able before long to 

 give the ratios of the susceptibility in various fields much weaker 

 than those employed in the foregoing experiments. 



In connection with the determination of the absolute magnetic 

 susceptibility of liquid oxygen, our attention has been much directed 

 to the important matter of the determination of magnetic suscepti- 

 bilities of substances in general at very low temperatures. 



Having regard to the great loss in magnetic susceptibility 

 experienced by the ferromagnetic metals in heating beyond a certain 

 temperature, it has been frequently suggested tbat bodies of small 

 susceptibility might become strongly magnetic if cooled to a 

 sufficiently low temperature. Faraday made many experiments on 

 this question, using solid carbonic acid as a refrigerating agent, but 

 was not able to arrive at any conclusions. 



A difficulty which presents itself in the use of liquid oxygen as 

 a refrigerating agent for this purpose is the strongly magnetic 

 quality of the liquid itself. All bodies except iron, nickel, and 

 cobalt, and the strongly ferromagnetic bodies, become apparently 

 diamagnetic when placed in liquid oxygen and in a non-uniform 

 magnetic field.* 



Moreover, for obvious reasons it is easy to weigh a diamagnetic or 

 apparently diamagnetic body in a non-uniform field because the forces 

 restoring the disturbed body to its original position increase with 



* An interesting experiment was made with a ball of ebonite which illustrates 

 this fact. Ebonite is slightly magnetic in air, owing no doubt <o iron impurity. 

 Ebonite is denser than liquid oxygen. Accordingly, a small ball of ebonite 

 dropped into the liquid oxygen contained in the vacuum vessel of the pole of the 

 magnet sunk to the bottom of the vessel. On exciting the magnet the ebonite 

 became apparently strongly diamagnetic and was repelled by the pole. It floated 

 up in the liquid until it reached a level at which the diamagnetic repulsion just 

 balanced the apparent weight of the ball in the liquid oxygen. Instructive 

 lecture experiments can be made in this manner. 



