In 1926, Mally proposed the first formal deontic system. As Mally and others soon realized, this system had some rather strange consequences. We show that the strangeness of Mally's system is not so much due to Mally's informal deontic principles as to the fact that he formalized those principles in terms of the propositional calculus. If they are formalized in terms of relevant logic rather than classical logic, one obtains a system which is related to Anderson's relevant deontic logic and not nearly as strange as Mally's own system.