Recent experiments seem to show that the sign of the effective pseudoscakr term in μ-caplure reactions is different from the theoretical resuk of Goldberger and Treiman based on V-A theory and dispersion relation technique. A possible explanation of this difference is proposed in the present note. It is shown that a very small p. s. tern in the HamiHonian (of the order 1/100 of the effective p. s. term) can change the sign of the effective p. s. term. If this small term is transformed and added as a part of the universal axial vector current, it will produce practically no effect in β-decay and π→ μ+ v decay. Finally the inclusion of hyperon pairs is considered.