vi THE ANATOMY OF HUMAN PERSONALITY 235
stitute for the natural stimulus. The first step was to discover how a reflex could become conditioned. He experimented to see whether he could attach the salivary-response to the sound of a buzzer or of a metronome. He let the metronome tick for a minute and then put food in a dog's mouth ; waited fifteen minutes, started the metronome again and after a minute again fed the dog. After this procedure had been gone through a number of times, the saliva began to flow before the end of the minute, anticipating the food stimulus. If the food stimulus was then omitted, a good flow of saliva would nevertheless be obtained. But if the experiment continued on this line, the metronome being sounded for a minute, but the food not being given at the end of that time, the flow of saliva diminished from trial to trial, and after a few such trials the conditioned reflex which had been established was extinguished.
Now, suppose the conditioned reflex had been established and the experiment was then discontinued till the following day. On the first trial the metronome gave no response; but the number of trials (metronome always followed by food) required to establish the reflex on the second day was less than on the first day ; and if the conditioning experiment, without extinction, was repeated day after day, the time soon came when the conditioned reflex held over from day to day. If%the extinguishing procedure metronome with no food were applied to such a thoroughly established reflex, the extinction was not so rapid and was only temporary, since the conditioned response would appear at once on the next day. Yet if the extinguishing procedure were applied repeatedly on a series of days, it finally eradicated the conditioned reflex. The processes of establishment and of extinction of a conditioned reflex were so closely parallel that Pavlov concluded that the brain mechanisms