A third type of result appeared in still another subject (Ch.), who normally expressed the concepts "right" and "left" by eye or head movements (never both kinds at the same time) to the right and left. Here my arm movements up and down caused the eye and head movements to be made simultaneously, so that the thought of "right" found expression in an upward movement of the head and an eye movement to the right, and the idea of "left" in a downward head movement and a movement of the eye to the left. The subject had no knowledge of this process, and it took six tests to bring about the new reaction. From that point onward the new movements were so well established that, depending upon them for my cue, I was able to make 32 correct inferences in a total of 40 tests. During the latter part of this series I blindfolded the subject, so that I could not see the movements of his eyes, and therefore had to base my inference entirely upon his head movements.--After removing the bandage, at the end of the series, I told the subject that I would go through another series, in which I intended to indicate his thought of "right" by an arm movement downward (instead of upward as heretofore), and his thought of "left" by a movement upward. (This he regarded as an idle whim of mine). It was only after the twelfth test that the former "association" which I myself had caused to be established, was completely displaced by the new. The thought of "right" was now accompanied by an eye movement to the right and instead of a raising there was a lowering of the head. A corresponding change occurred in the head movement expressive of the thought of "left". These responses were occasionally varied by some in which only the head movement or only the eye movement occurred. But these movements were always to the right, or downward and to the right, at the thought of "right",--and to the left, or upward and to the left, at the thought of "left". In ten tests I made ten correct inferences. After the new association appeared firmly established, I ceased responding by means of arm movements, and indicated my 'guesses' by word of mouth. At first the newly acquired movements continued to appear promptly in the subjects. But gradually they tended to become more uncertain and finally disappeared, as readily as they had appeared, and the normal conditions were once more established. Nor was there any tendency to reappear on the following day in another series of tests. (Those just described had been made on one day in the course of an hour or two). But as soon as I again used the earlier method of arm movement to indicate my inferences (raising the arm for "right", lowering it for "left"), the former artificial association was again established, although not until some 14 tests had been made,--during which the normal movements to the right and left were often inhibited and during which the conditions were, on the whole, chaotic.