2 () MAGNETISM 



it. Suspended in the second weaker solution it behaved again as 

 a paramagnetic. If a tube containing the weaker solution was 

 suspended in a vessel containing the stronger, it pointed 

 equatorially, behaving as a diamagnetic. Surrounded by a solu- 

 tion of its own strength it was neutral, and in a still weaker solution 

 it was paramagnetic. He also prepared a solution of protosulphate 

 of iron in water which was neither paramagnetic nor diamagnetic, 

 the iron salt being neutralised by the water. 



Explanation of the action of the medium by an ex- 

 tension of Archimedes' principle. In Chapter XX we 

 shall investigate the action of the medium in detail. For the 

 present we may regard the action as being in accordance with an 

 extension of Archimedes'* principle, the resultant force on a body in 

 a magnetic field being equal to that on the body minus that on the 

 medium which it displaces. We may put this extension of the 

 principle in the following form : Let N, Fig. 159, be a pole with a 



FIG. 159. 



magnetic field round it. We may regard all parts of the medium 

 as acted on by the pole, and as coming, under that action, into a 

 state of equilibrium in which the medium is in a condition of 

 strain. We may resolve the total action on any part of the 

 medium A into the magnetic action of the pole on it, and tin- 

 actions of its surroundings due to their state of strain, tin- resultant 

 of the latter being equal and opposite to the former. If now A 

 be removed and replaced by some other body, A', of exactly the 

 same shape, the magnetic action on A' differs from that on A, 

 while, unless A' is so strongly magnetic as appreciably to alter the 

 field, the magnetic action on the rest of the medium remains un- 

 changed and its strain is the same as before. Since the action of 

 the medium on A' depends on this strain it is unchanged. Hence 

 there is no longer equilibrium, and A' will be attracted to or 

 repelled from the pole according as it is more or less paramagnetic 

 than the medium. A difficulty in making this explanation per- 

 fectly general so as to account for all diamagnetic actions is that 

 we have to suppose a vacuum magnetic, for the diamagnetics in 

 the list given above are diamagnetic in a vacuum, and we should 

 have to suppose that their repulsion from a pole is due to the 

 stronger attraction on the surrounding vacuum. This is only 

 another way of saying that what we term a vacuum is not empty 

 space, but contains something capable of acting on and being 

 acted on by material bodies, a supposition to which we are forced 

 by the transmission of light through it. In a beam of light there 



