Abstract
A traditional derivation of the Fokker‐Planck equation is examined with emphasis on assumptions, especially sufficient inequalities between excitation and response characteristic times in a ``real'' system, i.e., a system with finite correlation time of the excitation. Comparison with experiment for an electronic ``cubic spring'' oscillator gives good agreement in both second and fourth moments of the response.