We obtained an unprecedentedly large number of s-wave neutron widths through R-matrix analysis of neutron cross-section measurements on enriched Pt samples. Careful analysis of these data rejects the validity of the Porter-Thomas distribution with a statistical significance of at least 99.997%.