Wonder how this might be applied to perturbation methods, which often (nearly always) involve an expansion with natural-numbered derivatives.....