“…In fact, we actually prove a more general result which does not assume separability. The proof is based on a refinement of a known lemma (see [5, p. 281 ], and [2] or [4] for the nonseparable case) that every nonempty closed subset of a metric space X of covering dimension 0 is a retract of X. The refinement, which seems to be new even in the separable case, asserts the existence of a closed retraction.…”