With Eusapia we succeeded in operating in a good light by appealing to ‘John’s’ vanity. We explained to him that obscurity stood in the way of the observation of the phenomena, that he was just as capable of working in the light as the ‘guides’ of other mediums were. In this way, we lead him to change his habits with us; the _meno luce_ to which those who have experimented with this medium are accustomed, was still demanded, but only when the seance was well advanced. At Bordeaux, where there was a large bay-window in the seance-room, the reflection thereon from the lights burning in the kitchen and winter-garden enabled us to see a little. In that case, Eusapia or John did not desire total obscurity, and we always had this feeble light, allowing a visual control which was sometimes satisfactory. When we are lucky enough to meet with an undeveloped medium, it is easy to give him the habit of operating in full light. This has occasionally happened to me. I need not enlarge upon the influence of obscurity upon error. With some very rare exceptions we can never be certain of the authenticity of a phenomenon obtained in a dark seance. Obscurity is, however, necessary for luminous phenomena. When once we have observed decided luminous forms, or really characteristic lights, it is easy to distinguish between them and illusion. A cool, calm observer does not make a mistake; it is not quite the same with excited experimenters. These latter give veritable suggestions to one another, and they end by having curious collective hallucinations. This is one of the most interesting facts of observation in spiritistic seances, so rich in purely psychological curiosities. I have frequently heard a sitter say that he saw a light in a given direction; the others looked in their turn and also saw it. Then one declared he perceived a form; soon others also saw a form. And from exclamation to exclamation the description of the form is completed. This is the genesis of a collective hallucination. I need hardly say, that experimenters who are so suggestible are not good elements: in purely scientific researches they should be reduced to a minimum. Personal experience has shown me, that of all the senses, that of sight is the most liable to imaginary impressions; after sight, the sense of touch is the most prompt to receive illusion. There are constant examples of this in spiritistic seances; the _cool breeze_, which is often really felt, is more often only imaginary. One person says he feels it; others at once imagine they feel it also. Sometimes it is not an error of imagination, but an error of attribution, the sensation of a cool breeze being caused by the breath. The sense of hearing has seemed to me to be refractory to suggestion in seances, though it does not altogether escape. I know of very few examples of imagined raps or noises.