We indicated in our previous work that for QED the contributions of the scalar potential, which appears at the loop level, is much smaller than that of the vector potential, and in fact negligible. But the situation may be different for QCD, the reason be