On Edwards' model for polymer chains