The commissioners grasped the crucial distinction at once: patients were plainly having experiences, and the question was not whether but why. Mesmer's associate Charles Deslon cooperated, which is why the tests could be run at all. In the best known of them, Deslon magnetised one tree in the garden at Passy, and a young man reported to be highly susceptible was blindfolded and led to embrace several trees in turn; he convulsed at the fourth, which had not been touched. In another, a woman was seated before a closed door and told that Deslon was magnetising her from the far side.
Nobody was there. She responded. The order was then reversed: Deslon magnetised her while she believed nothing was happening, and she did not respond. A series of variations on this pattern all produced the same asymmetry, effects following the subject's belief and never the treatment. The reasoning is worth stating plainly, because it is the whole of what modern trial design does. If a claimed agent produces its effect only when the subject knows it is present, the subject's knowledge is doing the work and the agent is not. Mesmer himself refused to take part, holding that Deslon's practice was a degraded version of his own, so the commission tested the method as Deslon administered it rather than as its originator would have wished.