Constraint on time-reversal tests in fully chaotic nuclear systems

In tests exploiting the unique sensitivity of a fully chaotic system like the compound nucleus, the possibility exists that unknown resonance parameters and interaction matrix elements taken on values which preclude the observation of a violation of time-reversal invariance independent of the strength of the time-reversal noninvariant interaction. A nontrivial constraint on time-reversal tests is implied, namely, that the observable has to be sampled a minimum number of times Mmin where Mmin>3. Several experiments purporting to be tests of time reversal do not satisfy this condition