Abstract
We show how an incorrect prediction of the t-dependence of diffractive deep inelastic scattering in the Anisotropic Chromo-Dynamics ACD framework was due to the erroneous evaluation of colour-traces through the perturbation theory Ž . prescription. By employing the ‘‘colour filtering’’ required by colour confinement, good agreement with experiment is obtained. As a result we extend our discussion to the diagrams of PQCD and we argue that submitting them to ‘‘colour filtering’’ removes their inconsistency with quark confinement and with the basic ideas of ACD.