Comments on fits to nucleon form-factor data

Abstract
Recent investigations of nucleon electromagnetic form factors are criticized, because the parametrizations belong to spectral functions which, in the neighborhood of threshold, are not compatible with the unitarity relation used in dispersion theory.