We show by counterexample that one of the results in the paper "On Pareto o
ptima, the Fermat-Weber problem, and polyhedral gauges", by R. Durier, Math
ematical Programming 47 (1990), does not hold. The stated characterization
of properly efficient points for a scalar location problem is only true in
dimension 1 and 2. For higher dimensions, the property stated is necessary,
but not sufficient.