I do not say, "The tables turn because my fluid is rotative." I say, "The tables turn, because, when they receive an impelling force or undergo an attraction, they cannot help turning." Stated in this way, it is a little less naive. Consequently, I should be under no obligation to undertake the cause of the poor university scholar of the _Malade Imaginaire_, and defend his famous reply: "_Opium facit dormire quia est in eo virtus dormitiva_" ("opium puts people to sleep because it has the sleep-producing virtue or property"). Nevertheless, I can't help it, out it must come: I find the reply an excellent one. I doubt whether the savants have found a better one to this day, and I advise them to resign themselves sometimes to the following kind of reasoning: "Opium puts us to sleep because it puts us to sleep; things are because they are." In other words, I see the facts and do not know the causes. I do not know. "I do not know!" terrible words, which one finds difficulty in pronouncing! Now, I suspect very strongly that the sly roguishness of Moliere is for the benefit of the doctors, who pretend to know everything, invent explanations which do not explain, and do not know how to accept the facts while waiting for more light. But there is more to come. The hypothesis of the fluid (a pure hypothesis, remember) must still prove that it is a hypothesis reconcilable with the different circumstances of the phenomenon. The table does not merely turn: it lifts its legs up, it raps numbers mentally indicated to it; in a word, it obeys the will, and obeys it so well that the removal of contact does not terminate its obedience. The impelling force or lateral attraction which account for rotations cannot account for levitations. But why? Because the will directs the fluid now into one leg of the table, now into another. Because the table identifies itself with us, after a fashion, becomes a limb of our own body, and produces movements thought of by us in the same manner as our arm produces them. Because we have no conscious knowledge of the direction imparted to the fluid, and govern the movements of the table without imagining that any kind of fluid or force whatever is in action.