In this erratum we want to point out that our interpretation and conclusion concerning the closing of our magnetisation curves is not correct. The fact that in our measurements the hysteresis curves close is due to the specific measuring method used in our Quantum Design MPMS SQUID magnetometer, combined with the small length of uniformity of the superconducting magnet. It is, therefore, not related to the fast relaxation of magnetisation with respect to the long measuring time of the SQUID, as we have argued in our paper. [...]

We present magnetic relaxation measurements on an epitaxial YBa2Cu3O7−δ -film. From these we have obtained the activation energy as a function of temperature for several initial conditions. The absolute value of the activation energy is different for the various cases and it increases with temperature. This last property provides evidence for the presence of a distribution of activation energies.

The critical current density, magnetic hysteresis and relaxation of the magnetisation have been measured on a 1000 ÅYBa2Cu3O7-δ-film in a field perpendicular to the plane of the film. The film was made by co-deposition in a UHV triple e -gun system, using BaF2 as one of sources. The results have been interpreted in terms of flux pinning and creep of curved flux lines with locally different pinning energies.