We cannot trace the history of the personal difference in detail. It is enough to say that the astronomers, having discovered it, were naturally anxious to get rid of it; and they presently found a way to relieve the observer of the task of listening; he simply pressed a key when the star crossed the line of the meridian, and the time of pressing was recorded automatically. This device did not eliminate the personal difference; but it was methodically of great importance. For the eye-and-ear method had now become, essentially, _a method of response to stimulus by movement_; and in that form it settled down permanently in the psychological laboratory. The stimulus for the astronomer was the star on the meridian, and the response was the pressure of his finger on a key. But it is clear that the stimulus need not be visual; the observer might just as well respond to a sound or a touch or a taste. It is clear, further, that the response need not be a movement of the hand; the observer may respond, just as well, by movement of the organs of speech, or of the foot, or of lip or eyelid. It is clear, finally, that if we know the actual time at which the stimulus is presented, and the actual time at which the movement of response takes place, we can measure the interval between the two. A little ingenuity makes this possible. If, for instance, the flash of light which serves as stimulus _makes_ an electrical circuit, and the finger-movement in response _breaks_ the circuit; and if an electrical clock is placed in the same circuit; then the clock-hands will begin to move when the flash comes, and will stop when the movement occurs, and we can read off the =reaction time= from the dial. In its simplest form, then, the reaction experiment takes shape as follows. We subject the observer to some prearranged form of stimulation (a flash of light, a sharp noise), to which he is to reply by some prearranged movement (perhaps, the slipping of his finger from the button of a telegraph key); and the instruments which we employ are so connected that we can measure the time elapsing between the exhibition of stimulus and the performance of answering movement. The experiment thus has two sides. It gives us _numerical results, the reaction times_ measured in units of our clock, in hundredths or thousandths of a second; but it gives us also a _complete impulsive action_, which we can observe as often as is necessary for analysis.