The galvanostatic reduction of the nickel hydroxide electrode is known to proceed at two successive potentials without this phenomenon being well understood. In this paper we prove that this second discharge plateau is unrelated to oxygen reduction and that it can be observed in the absence of the gamma-NiOOH hydrated phase. Some earlier studies have connected it to the phase diagram of the beta(III)/beta(II) system. Using highly precise X-ray measurements, we demonstrate that the latter is quite different from the one commonly used. We then interpret the unusual shapes of potential relaxation curves during the lower discharge plateau. Measurements over a wide range of current density (four orders of magnitude) yield experimental clues for an original interpretation of the second discharge plateau based on the dynamics of the discharge processes. We suggest that the second plateau is due to the existence of a phase close to Ni(OH)(2), which is not ionically conductive but (poorly) electronically conductive, in the vicinity of the current collector.