On Newton’s method