“…25 The following considerations suggest that P5 is caused by a mechanism involving dislocations and interstitial hydrogen: ͑i͒ The process is more developed in sample Y0 where the dislocation density and the H concentration are higher than in Y1; ͑ii͒ the relaxation curve, about twice broader than a single Debye peak ͑Table I͒, is not compatible with pointdefect relaxation; ͑iii͒ a relaxation, with similar features, has been observed in plastically deformed polycrystalline Y. 8 The mechanism of peak P5 should be similar to those reported in transition metals ͑Nb, Ta, V͒ and attributed to the relaxation of dislocations dragging interstitial solute impurities, 26,27 which in the present case are the H interstitial atoms. According to this picture, the observed decrease of P5 after annealing can be explained by the concomitant reduction of the dislocation density and H concentration.…”