Operator product and vacuum instability (addendum)

This paper is an addendum to our previous paper of the same title [Phys. Rev. D 26, 499 (1982)]. We discuss an apparent scheme dependence of the results reported in a note added in proof. We find that, when the $q^2$ dependence of operators is carefully identified, the results are scheme independent. Thus, for any subtraction scheme, our conclusion is that the operator-product expansion gives different results at next-to-leading twist when made about the physical vacuum and when made about the unstable symmetric vacuum with nonvanishing vacuum expectation values allowed for nontrivial operators. We include also some errata for the original paper.