After the necessary preliminary rotations, the table is harnessed up in its second form. Equilibrium is first secured, then 3-5 of a pound is taken from the counterpoise. The force required to lift the top by its centre is then 4 ounces, and previous experiments have proved that the adherence of the fingers of the operators (the top was polished, and not varnished), together with the possible effects of elasticity, form a total lower than that figure. Yet the top is lifted by the action of the fingers placed lightly on its upper surface, at a certain distance from the edge. Then the counterpoise is diminished; the mechanical difficulty of lifting is augmented, yet still it takes place. The weight is again diminished, and more and more, up to the limit of the apparatus. The force necessary to lift the top is then 8 1-5 pounds, and the counterpoise has been relieved of 24 pounds; yet the levitation is easily accomplished. The number of the operators is gradually lessened from eleven to six. The difficulty goes on increasing, yet six operators still suffice; but five are not enough. Six operators lift 9 1-3 pounds,--an average for each man of about 1-1/2 pounds. We now possess, in the apparatus just described, a gauge or instrument of measurement. B. The following movements were produced without contact: The table on which were made the trials I witnessed has a diameter of 32 inches and weighs 31 pounds. An average tangential force of 4 2-5 pounds, which may be raised to 6 3-5 pounds, according to the greater or less inequalities of the floor, applied to the edge of the table, is necessary to give to it a movement of rotation. Ten is usually the number of persons who operate about this table. In order to assure ourselves of the absence of all contact, we placed our eye on a level with the table in such a way as to see light between our fingers and the surface of the table, the fingers themselves remaining a little less than an inch above the top. Usually, two persons would be observing at once. For instance, M. Edmond Boissier was observing the legs of the table, while I was watching the top. Then we exchanged roles. Sometimes two persons took places at the extremities of one and the same diameter, the one opposite the other, for the purpose of watching the top of the table. Several times we saw it move, although we could not detect the slightest touch by the fingers. According to my calculations, it would require the contact of at least 100 fingers, or the light pressure of thirty, acting voluntarily and fraudulently, to explain in terms of mechanics the movements we observed.