The best way of learning to judge correctly is the one which tends to simplify our experience, and enables us to make no mistakes even when we dispense with experience altogether. It follows from this that after having long verified the testimony of one sense by that of another, we must further learn to verify the testimony of each sense by itself without appeal to any other. Then each sensation at once becomes an idea, and an idea in accordance with the truth. With such acquisitions I have endeavored to store this third period of human life. To follow this plan requires a patience and a circumspection of which few teachers are capable, and without which a pupil will never learn to judge correctly. For example: if, when he is misled by the appearance of a broken stick, you endeavor to show him his mistake by taking the stick quickly out of the water, you may perhaps undeceive him, but what will you teach him? Nothing he might not have learned for himself. You ought not thus to teach him one detached truth, instead of showing him how he may always discover for himself any truth. If you really mean to teach him, do not at once undeceive him. Let Émile and myself serve you for example. In the first place, any child educated in the ordinary way would, to the second of the two questions above mentioned, answer, "Of course the stick is broken." I doubt whether Émile would give this answer. Seeing no need of being learned or of appearing learned, he never judges hastily, but only from evidence. Knowing how easily appearances deceive us, as in the case of perspective, he is far from finding the evidence in the present case sufficient. Besides, knowing from experience that my most trivial question always has an object which he does not at once discover, he is not in the habit of giving heedless answers. On the contrary, he is on his guard and attentive; he looks into the matter very carefully before replying. He never gives me an answer with which he is not himself satisfied, and he is not easily satisfied. Moreover, he and I do not pride ourselves on knowing facts exactly, but only on making few mistakes. We should be much more disconcerted if we found ourselves satisfied with an insufficient reason than if we had discovered none at all. The confession, "I do not know," suits us both so well, and we repeat it so often, that it costs neither of us anything. But whether for this once he is careless, or avoids the difficulty by a convenient "I do not know," my answer is the same: "Let us see; let us find out." The stick, half-way in the water, is fixed in a vertical position. To find out whether it is broken, as it appears to be, how much we must do before we take it out of the water, or even touch it! First, we go entirely round it, and observe that the fracture goes around with us. It is our eye alone, then, that changes it; and a glance cannot move things from place to place.