Quarterly Journal of Experimental Psychology

Twenty-eight subjects were examined on a visual matching task for their ability to maintain an orientation with respect to a particular direction in the horizontal plane following a voluntary rotary body movement through 180 degrees. Each subject was examined with respect to eight different directions.

Numerous gross errors occurred when visual information was reduced to the display of an arrow indicating a direction and a second arrow manipulated by the subject. The magnitude and distribution of the errors suggest that, under the conditions of this experiment, visual information as to direction in the horizontal plane is analysed according to the two horizontal dimensions defined by the sagittal and coronal planes of the head. In correcting for the rotary body movement, failure may occur with respect to either or both of these two dimensions. The frequency of a failure to make any correction at all (i.e. 180-degree errors) is consistent with independent failure in each of the two horizontal dimensions.

Failure is markedly more frequent in the fore-aft dimension than in the left-right dimension. It is suggested that this may be explained in terms of the ambiguous spatial significance of vertical disposition on the retina and the possibility of contamination between the two systems of conceptual analysis which identify the vertical and the fore-aft dimensions of visual space.

It is demonstrated that when minimal “landmarks” are provided they tend to be utilized as reference points in attempts to maintain orientation, even when the subject is aware that the “landmarks” are misleading. Such a use of “landmarks” does not suppress the previously mentioned mechanism of dimensional orientation.

The relevance of these results to normal human orientation is discussed.

The time spent by a rat in a bar-pressing situation is made up of active time spent n pressing, eating time, and extra time spent in other activities. With a well trained rat, active time and extra time are small, and eating time mainly determines the rate of reward delivery. Active time is affected by a change of weight on the bar, the time between reward deliveries is affected by the amount of reward, and the extra time is affected by extinction conditions.

There is not a one-to-one correspondence between periods of activity at the knob and rewards.

The term “response” and some variables based on it are given empirical referents, which show that much research and theorizing on bar-pressing behaviour has been concerned with only a small selection of the rat's bar-pressing activities. Some reasons for this restriction are the use of the simple weighted bar, the lack of a rationale for bar-pressing research, and the practice of not watching the rat during an experiment.

The evidence pointing to the retinal origin of after-images is considered. The reports of the occurrence of after-images from visual images of hallucinatory vividness are reviewed.

Experimental results are presented to indicate that a complementarily coloured afterimage may arise following the exposure of the temporarily blind retina to a coloured stimulus.

After-images, or after-effects, from vivid images are described in seventeen persons (mostly possessors of “number-forms”). They are found to move with the eyes and to show, in some persons, a degree of conformity with Emmert's Law which, while considerable, is less than that of after-images of real stimuli. In the case of one “eidetic” subject, the after-images from neither real nor imaged stimuli conformed with Emmert's Law. In some persons, after-images of images occur in complementary colours.

The retinal origin of after-images is affirmed, but that they can occur occasionally as a purely central phenomenon is acknowledged. The possible learned or inherent nature of after-images of central origin is discussed.

Exact definitions and statements are given of some widely accepted concepts and assumptions in colour vision. Two theorems are stated, and one of them is proved (the other being self-evident). Two applications of the theorems are briefly discussed.