Quite dramatic progress has been made in the production and utilization of polarized e± beams at CERN's LEP, at HERA at DESY and at the Stanford linear collider SLC. The motivation for trying to overcome the tremendous technical problems involved derives from two sources:

1. (i) the realization that longitudinally polarized electrons permit extremely accurate measurement of the fundamental parameters of the Standard Model of electroweak interactions;

2. (ii) the discovery in 1987, by the European Muon Collaboration (Ashman et al., 1988), that only a very small fraction of the proton's spin appeared to be carried by its quarks, leading to what was characterized as a ‘crisis in the parton model’ (Leader and Anselmino, 1988). This made it important to carry out further studies of deep inelastic lepton–hadron scattering using longitudinally polarized leptons colliding with a longitudinally polarized proton target.

Though not a primary impetus, it turns out also that polarized e± permit an exceedingly accurate calibration of the beam energy at LEP and HERA.

The problems involved in having stable polarized beams are quite different in circular storage rings and in linear accelerators. Hence we shall discuss the two cases separately.

The natural polarization of electrons circulating in a perfect storage ring

As mentioned in the introduction to Chapter 6, in principle a circulating electron beam gradually acquires a natural polarization in which its magnetic moment μe becomes aligned parallel to the guide field B.