It must not be thought that the circle can be modified with impunity. My personal experience has shown me it is bad to frequently introduce strangers into the circle. It should be arranged that a series of at least six sittings will be held without modifying the group: that no new experimenter will be admitted: and that none of the original experimenters will miss even one seance. Then if at the end of six sittings nothing has been obtained, my advice is to change the circle, to eliminate certain elements, replacing them by others. It is preferable to change the sitters one by one, and to make a few experiments with the circle thus modified before making further changes. If interesting results be forthcoming, and a desire be felt to show them to other people, the new sitters must be introduced one by one, and, I repeat, at intervals of three or four sittings. Otherwise there would be a risk of compromising the success of the experiments. The personification sometimes asks for the addition to the circle of a certain person; it is then well to invite him to the sittings if circumstances allow of it. I now return to the seance which, I suppose, has begun. The sitters put their hands on the table; it is not generally necessary to ‘form the chain,’ that is to say, to establish contact between the sitters by linking the little fingers. The hands in position, and the room _well lighted up_, we wait. Talking or singing may be indulged in. The emission of the voice, especially rythmical emission, is an excellent condition: it is a good thing to play some music, organ-playing is particularly effective. Why is the production of sonorous rythmical waves favourable to these phenomena? I have no explanation to offer for this fact, which I am not the only one to have observed. At the end of a few minutes, the table often seems to be agitated. If we are experimenting with spiritists or with people accustomed to spiritistic proceedings, the table, raising itself, will be seen to strike the floor with one of its legs. I advise asking the table if it wishes to speak, and to arrange that two raps will mean ‘no,’ and three raps ‘yes.’ Of course any other numbers or signs will do equally well. The table, thus consulted, generally replies ‘yes.’ It can then be asked, if the sitters are well placed: if it indicates any other arrangement it is well to heed its advice. We should then make known to the table what kind of results are desired, and point out, particularly, that movements with contact, failing to carry conviction, are undesirable. I have already said that the personification—it is thus I call the entity, whatever it may be, who claims to be manifesting—is generally very open to suggestion; and it suffices to indicate, at the beginning of the experiment, the objection that is made to movements with contact to be almost completely rid of them.