The paper shows that the conclusions drawn from recent experiments on electron irradiation of gold at about 100° K should be reversed, i.e. that these experiments speakagainstthe free thermally activated migration of interstitials at the irradiation temperature. It is shown that this viewpoint is in agreement with several related experiments, which may all be interpreted by a dynamic propagation of crowdions over distances between 1000 Å and 2000 Å. At intermediate and high defect concentrations the dynamic propagation mechanism is suppressed.