Let us go on, however. We have not yet reached the end of our proofs. One of these has always especially struck me: I mean the proof derived from failures.

It is claimed that the movements are produced by the action of our muscles, by involuntary pressure. Now here are the same operators who yesterday secured from the table the fulfilment of their most capricious desires; their muscles are as strong, their vivacity is as great, their desire to succeed is perhaps keener—and yet nothing! absolutely nothing! A whole hour will pass without the least rotation beginning; or, if there are rotations, levitations are impossible to procure; what little is done by the table is done feebly, dismally, and as if reluctantly. I repeat it again, the muscles have not changed; then why this sudden incapacity? The cause remaining identically the same, whence comes it that the effect varies to such a degree?

"Ah!" says an objector, "you are talking of involuntary pressure, and say nothing about voluntary pressure, of fraud, in short. Don't you see that the cheaters may be present at one sitting and not appear at another, that they may act one day and not give themselves the trouble on the next?"

I will reply very simply, and by facts.

"The cheaters are absent when we do not succeed!" But it has happened many a time that our personnel has not been changed in any way. The same persons, absolutely the same, have passed from a state of remarkable power to a state of comparative impotence. And that is not all. If there exists no operator whose presence has preserved us from failures, no more does any exist whose absence has rendered us incapable of success. With and without each one of the members of the chain we have succeeded in performing all the experiments,—all without exception.

But 'the cheaters do not take so much pains every day!' The pains would be great indeed, and those who infer fraud little think what prodigies they are invoking. The accusation is an absurdity which verges on silliness, and its silliness removes its sting. One does not take offense at things like that. But come now, let us suppose for the moment that Valleyres were peopled with disciples of Bosco, that prestidigitation were generally practised there, and that it had been thrust under our very eyes for five months, and under the eyes of numerous and very suspicious witnesses without a single case of perfidy having been pointed out. We have so well concealed our game that we have invented a secret telegraphic code for the experiment of reading numbers, a particular turn of the finger for moving the most enormous masses, a method of gradually lifting tables that we do not seem to touch. We are all liars, all; for we have been mutually watching each other for a long time now, and do not denounce anybody. Nay, more, the contagion of our vices is so swift to take that, as soon as we admit a stranger, a hostile witness, into the chain, he becomes our accomplice; he voluntarily closes his eyes to the transmission of signals, to muscular efforts, to the repeated and prolonged suspicious actions of his next neighbors in the chain! Well and good; suppose we grant all that, we shall not have got farther along for that. It will still remain to be explained why our cheaters sometimes do nothing at the very moment when it would be to their interest to succeed. It has happened, indeed, that a certain sitting at which we had many witnesses and a great desire to convince turned out to be a mediocre one. Such and such another, under the same conditions, was, on the contrary, a brilliant success.

There you have real and important inequalities, and they dare to talk to us of muscular action and of fraud.

Fraud and muscular action! Here for instance is a fine opportunity to put them to the proof. We have just placed a weight on the table. This weight is inert and cannot be accessory to any device. Fraud is all around it perhaps, but it is not in the tubs of sand. This weight is equally divided among the three legs of the table, and they are going to prove it by each one rising in turn. The total load weighs 165 pounds, and we scarcely dare to increase it, for, as it is, it was enough, one day, to break our very solid table. Very well; now let someone try to move this weight. Since muscular action and fraud must explain everything, it will be easy for them to put the mass in motion. Now they cannot do it. Their fingers contract and the knuckles whiten without their obtaining a single levitation, whereas, some moments later, levitations will take place at the touch of the same fingers, which gently graze the table's top and make no effort at all, as any one may easily convince himself.

Certain very ingenious scientific rules of measurement, for the invention of which I cannot claim the credit, put us in the way of translating into figures the effort which the rotation or levitation of the table demands, when loaded in the way just described. With the above-mentioned weight of 165 pounds, rotation is secured by means of a lateral traction of about 17½ pounds, while levitation is only obtained by a perpendicular pressure of 132 pounds at least (which I will reduce, however, to 110, in deference to the presumed wishes of the critic, and on the supposition that the pressure might not be absolutely vertical). Several deductions are to be drawn from these figures.

In the first place, muscular action may cause the table to turn, but it cannot lift it. As a matter of fact, the ten operators have one hundred fingers applied to its surface. Now, the vertical, or quasi-vertical, pressure of each finger cannot exceed twelve ounces on the average, the chain being composed as it is. They only develop, then, a total pressure of 66 pounds, which is quite insufficient to produce levitation.