A note on Branin's method for binding the critical points of smooth functions