Abstract: A hydrogen-filled ballon was sent up into the air to distances of 50, 100, 150, 200 and 250 metres from 0, and with the help of wind, was trailed laterally to various angles of elevation in relation to 0's horizontal line of regard. For each distance 0 perceived the stimulus flora two postures: 1) Sitting erect on a chair and viewing the elevated stimulus by raising his head and eyes; 2) Sitting with back and head against the pivoted and adjustable rear of the chair, which was then continually adjusted so as to remain at a right angle to the elevated stimulus, so that 0 always saw the stimulus straight ahead. The comparison stimuli were 22 discs of various sizes, placed near 0. It was found that when 0 sat in the erect position, the apparent size diminished gradually as the angle of elevation of the stimulus increased, and reached its minimum at the zenith; while when 0 sat in the straight ahead position, the apparent size diminished abruptly, was much smaller at 20° elevation, and reached its minimum at about 40° elevation, then remaining constant up to the zenith.When 0 looked downward from a high building at the stimulus on the ground, i.e. with an angle of regard below the horizontal, the apparent size also diminished, but to a lesser degree than it did with a raised angle of regard.As the crucial results of this experiment do not substantiate Boring's studies on the moon illusion, it is suggested that the raising of the eyes is not the cause of the shrinkage of apparent size, but rather that feedbacks from the movement of the head and the eyes, when looking upward at an elevated object, provide cues to the angular relation between the stimulus and ground, thus tending to yield a constant visual size as compared with that viewed along the ground, and that these cues decline gradually with increased elevation. In the straight ahead posture, 0 lost reference to the ground immediately as the stimulus ascended 20°—40° above the horizontal, whereupon the apparent visual size followed the law of visual angle.

Key words: NULL