And even in cases where the good faith--the _conscious_ good faith--of everyone concerned is above suspicion, it must be remembered that there are both unconscious actions and unconscious perceptions which may wholly vitiate an experiment. The rule should be so to arrange the experiment that the percipient _cannot_ profit by unconscious indications; that he cannot (for example) see the expression of the agent's face, or hear the sound of his pencil as he writes down a number to be guessed. Such precautions should be a matter of course; and when they are taken, these experiments near at hand are certainly the easiest and best for private experimenters to begin with, although the desirability of gradually increasing the distance between the persons concerned should always be kept in view. Let A and P begin their trial, then, in quiet and calm of mind; let A, the agent, sit behind P, the percipient, and not in contact. Let A be provided with a full pack of cards, in which he replaces the card drawn, after each trial, or with a bag of known numbers--say from ten to one hundred--a range convenient for computation--in which bag he replaces and shuffles up the number drawn, after each trial. Let him draw a card (to take cards as our example) say, "Now!" and gaze fixedly at it. Let P keep his mind as blank as possible, and make his guess only when some kind of image of color, suit, or pips, in some way floats into his mind. His first guess only must be counted, and must be received in silence. Let A continue this process for some prearranged number of times, say ten times, and record accurately all the experiments made. Let him renew the process, with intervals of hours or days between each batch of trials, until he has some hundreds of results to analyze. Then let him send his results, with description of the conditions under which the trials were made, to Dr. Richard Hodgson, 5 Boylston Place, Boston, Mass. Dr. Hodgson will tell him if it is worth his while to go on, and will advise as to modifications in the form of experiment. These hints must here suffice as to experiments made close at hand. But experiment, or observation verging into experiment, is often possible at long distances as well. It often happens that some one tells me that he (or she) has so peculiar a sympathy with some given friend that what one of the pair is actually feeling or thinking at a distance is reproduced by the sensation or thought of the other. To such communications my invariable reply is, "Keep a 'psychical' diary. Put down therein at once every incident which you intend to count, if it turns out (so to say) a telepathic success, and no incident which you do _not_ intend to count. Let your friend keep a similar diary, without showing it to you; after a few months let me compare the two diaries with one another."