According to a recent theorem, for a general quantum-mechanical system undergoing a process, one can tell from measurements on this system whether or not it is characterized by a quantum number, the existence of which is unknown to the observer, even though the detecting equipment used by the observer is unable to distinguish among the various possible values of the ‘‘secret’’ quantum number and hence always averages over them. The present paper deals with situations in which this averaging is avoided and hence the ‘‘secret’’ quantum number remains ‘‘secret.’’ This occurs when a new quantum number is hypothesized in such a way that all the past measurements pertain to the system with one and the same value of the ‘‘secret’’ quantum number, or when the new quantum number is related to the old ones by a specific dynamical model providing a one-to-one correspondence. In the first of these cases, however, the one and the same state of the ‘‘secret’’ quantum number needs to be a nondegenerate one. If it is degenerate, the theorem can again be applied. This last feature provides a tool for experimentally testing symmetry breaking and the reestablishment of symmetries in asymptotic regions. The situation is illustrated on historical examples like isospin and strangeness, as well as on some contemporary schemes involving spaces of higher dimensionality.

• Received 19 June 1986

DOI:https://doi.org/10.1103/PhysRevD.34.1241