Abstract
Scaling deviations of deep-inelastic lepton-hadron reactions are compared with predictions of various theories of strong interactions. It is shown that only asymptotically free theories are compatible with experiment.