Theoretically, I grant you, there is no possibility of error in necessary reasoning. But to speak thus theoretically, is to uselanguage in a Pickwickian sense. In practice, and in fact, mathematics is not exempt from that liability to error that affects everything that man does.