Premising that all rules are but general, especially those which relate to the mind, I observe that indefectibility may at least serve as a negative test of certitude, or _sine quâ non_ condition, so that whoever loses his conviction on a given point is thereby proved not to have been certain of it. Certitude ought to stand all trials, or it is not certitude. Its very office is to cherish and maintain its object, and its very lot and duty is to sustain rude shocks in maintenance of it without being damaged by them. I will take an example. Let us suppose we are told on an unimpeachable authority, that a man whom we saw die is now alive again and at his work, as it was his wont to be; let us suppose we actually see him and converse with him; what will become of our certitude of his death? I do not think we should give it up; how could we, when we actually saw him die? At first, indeed, we should be thrown into an astonishment and confusion so great, that the world would seem to reel round us, and we should be ready to give up the use of our senses and of our memory, of our reflective powers, and of our reason, and even to deny our power of thinking, and our existence itself. Such confidence have we in the doctrine that when life goes it never returns. Nor would our bewilderment be less, when the first blow was over; but our reason would rally, and with our reason our certitude would come back to us. Whatever came of it, we should never cease to know and to confess to ourselves both of the contrary facts, that we saw him die, and that after dying we saw him alive again. The overpowering strangeness of our experience would have no power to shake our certitude in the facts which created it.