The magnetic moment of a particle of nonzero spin
S is computed by the introduction of minimal coupling into the appropriate Galilean-invariant wave equation. It is shown that the requirement that the differential equations be of first order, together with a minimality assumption on the number of components, uniquely implies a
g factor of
S1. The role played by this minimality condition is made explicit by means of a counterexample.
The Aharonov-Casher (AC) effect for quantum motion of a neutral magnetized particle in the electric field is believed to be a topological effect closely related to the known Aharonov-Bohm (AB) effect. We study how it depends on the spin of the particle involved. Duality of the AB and AC effects is demonstrated to exist only for two extreme spin (and magnetic moment) projections. Classical consideration confirms the conclusion. Motion of a classical magnetized particle with generally oriented magnetic moment in the AC field appears to be subjected to both forces and torques. Only for two special orientations of the magnetic moment (same as in the quantum case) the motion is effectively free and similar to the AB motion of a charged particle. Thus, the AC effect is not really topological. Presence of higher multipoles totally destroys the possible AB-AC duality.
A system of two particles of spin \textonehalf{} coupled by their dipolar interaction and in an arbitrary magnetic field is considered. It is shown that exact expressions for the energy levels of the system can be obtained. The density matrix is calculated exactly for the cases where a magnetic field is suddenly applied, parallel or perpendicular, to the line joining the dipoles. It is used in the evaluation of the magnetic moment of the system. Instead of there being a gradual approach to an equilibrium situation the magnetic moment varies harmonically in time. A partially successful attempt is made to calculate the density matrix when the functional dependence of the magnetic field on the time is more complicated than a step function. The exact calculations for this simple system are compared with the approximate calculations of Waller for a system of
N spins. It is pointed out that there may be no gradual approach to equilibrium in the
N-spin system, either.