Pauli, an ergodic theorem and related matters

Time averages are important in physics and in statistical mechanics. An ergodic theorem is a way of justifying the replacement of time averages by averages in phase space. A derivation of this theorem from quantum mechanics was given by von Neumann in 1929. Pauli and Fierz found a shorter argument in 1937. They agreed that for ergodicity to hold there should be no energy degeneracy in the Hamiltonian. I trace the circumstances and consequences of a disproof of these arguments.