M 



the Absurdity of Diamagnetic Polarity. 195 



To make our argument simple, let us suppose that in our 

 ideal experiment the body M is a cube each of whose edges 

 is one centimetre, and let it be hung so symmetrically with 

 respect to the magnet pole P that it is only on the face Y, 

 next to P, and on the opposite face 

 Z, that the pressure of the air need Fig-. 5. 



be considered. Also, for the sake 

 of simplicity, let us suppose the 

 pressure to be uniform over each of 

 these faces. Then if the pressure of 

 the air on the face Y exceed that 

 on the face Z by the 1000th part of an atmo, the resultant 

 pressure tending to drive M away from P will be rather over 



1 gramme weight. If M be of bismuth, the most strongly 

 pronounced of the so-called diamagnetic substances, the mass 

 of the cube will be nearly 10 grammes. The resultant pressure 

 of the air is therefore about -j^th of the weight of the cube. 

 It will, therefore, be capable, in one second, of generating 

 in the cube a velocity of 98 centimetres per second, and of 

 causing it to move from rest through 49 centimetres. In £ 

 of a second, it would generate in the cube a velocity of 24*5 

 centimetres (9 # o inches) per second, and cause it to move 

 through 3 centimetres (1*2 inches). If the pressure on the 

 face Y exceed that on the face Z by the 10,000th of an atmo, 

 the resultant pressure tending to drive the cube away from P 

 will be about the y^ -th part of its weight. This would be 

 sufficient, in one second, to give the cube a velocity of 9*8 

 centimetres (3*8 inches) per second, and to cause it to move 

 through 4*9 centimetres. In -J- second it would give the cube 

 a velocity of 4*9 centimetres (1*9 inches) per second, and 

 cause it to move through 1*2 centimetres ("46 inch). 



The preceding calculations may well throw doubt on the 

 common theory of diamagnetism, but the theory has still one 

 apology left. It is found that diamagnetic bodies retain 

 their characteristic property in a comparative " vacuum " of 



2 or 3 millimetres of mercury (say the -gJoth of an atmo). 

 To explain this, we have only to observe that if, in our ideal 

 experiment, the attraction of the pole P on the body M were 

 strictly zero, the body M would be driven away from the 

 excited pole P by the pressure of the air however good the 

 " vacuum " might be, short of absolute perfection. Hence if 

 the attraction of P on M be not strictly zero but exceedingly 

 small, it will be necessary to reduce the density of the air 

 very nearly to zero before this attraction can make itself 

 manifest. 



Having now shown that the common notion of diamagne- 



02 



