Psychology needs the experimental method for both the reasons noted above: because the observed phenomena are elusive and slippery processes, and because the observer is warped and biassed by common sense. We may therefore show by an example _how psychological experiment is possible_. Suppose that we wish to find out how a printed word is perceived,—whether we read it letter by letter, or take in its form as a whole, or take in certain letters clearly and the general form vaguely. We first prepare our material. We print upon cards, or photograph upon lantern slides, a large number of words. We employ different printing types; different groups of letters; different lengths of words; single words and groups of words; words properly spelled, and words altered by mutilation or omission of particular letters at different parts of the word. Every one of these classes of stimuli, as the words may be technically called, is represented by a number of cards or slides. The stimuli are mixed in haphazard order, and are thrown upon the screen by a reflectoscope or projection lantern in an otherwise dark room; a pneumatic shutter before the lantern makes it possible to show them for a brief time, say, a fifth of a second. All this apparatus is put in the charge of an experimenter. When the material is ready, and the whole arrangement works properly, an observer is called in. He works for a limited time, at the same hour every day, and only after a certain time has been allowed for his eyes to accustom themselves to the dark. The stimuli are presented at regular intervals. The observer reports what he perceives at every exposure of a stimulus, and the experimenter writes down what he says.