THE PHYSICAL BASIS OF FERROMAGNETISM 3 



electrical conduction. The electrons in the outer part of the third 

 shell are those responsible for the distinctive kind of magnetism found 

 in iron, cobalt and nickel. Some of these electrons spin in one direction 

 and some in the opposite, as indicated, so that their magnetic moments 

 neutralize each other partially but not wholly, and the excess of those 

 spinning in one direction over those spinning in the other causes each 

 atom as a whole to behave as a small permanent magnet. 



The well-established kinetic theory of matter tells us that if each 

 atom were to act independently of its neighbors, the atoms would be 

 vibrating and rotating so energetically that they could not be aligned 

 even with the strongest field that can be produced in the laboratory. 

 To explain the kind of magnetic properties found in iron, therefore, 

 it is necessary that there be some internal force capable of making the 

 magnetic moment of a group of neighboring atoms lie parallel to each 

 other — the small atomic "permanent magnets" of each group must 

 point in the same direction so as to provide a magnetic moment great 

 enough to permit a realignment when subjected to external fields. 

 Recently it has been shown by independent means that there is such 

 a force in just those elements which are ferromagnetic, and it is from 

 this force that the difference between magnetic and non-magnetic 

 materials arises. The force is electrostatic in nature and is called 

 "exchange interaction" by the atomic-structure experts, the wave 

 mechanicians, who have shown its existence and calculated its order 

 of magnitude. This force maintains small groups of atomic magnets 

 parallel against the forces of thermal agitation. (When the material 

 is heated so hot that the disordering action of the agitation becomes 

 strong enough to overpower the forces of "exchange interaction" the 

 material loses its ferromagnetism ; in iron this happens at 770° C.) 



But M^hy then is not every piece of iron a complete permanent 

 magnet? For some reason not understood at present, at ordinary 

 temperatures the electrostatic forces of exchange interaction maintain 

 the elementary magnets parallel only over a limited volume of the 

 specimen. This volume is usually of the order of 10~^ or 10~^ cubic 

 centimeters and contains a million billion atoms and is of course in- 

 visible. Such a volume is said to be saturated because the atomic 

 magnets are all pointing in the same direction, and has been given 

 the name "domain." Thus a magnetic material at room temperature, 

 before it has been magnetized by subjecting it to the influence of a 

 magnetic field, is divided into a great many domains each of which is 

 magnetized to saturation in some direction generally different from 

 that of its neighbors. The net or vector sum of the magnetizations is 

 zero, and externally the material appears to be unmagnetized but in 



