142 LISTING'S LAW. [BOOK in. 



with each of the crosses of the pattern on the wall, the hori- 

 zontal limb coinciding with a horizontal line and the vertical limb 

 with a vertical line. This shews that during the up and down 

 and during the side to side movement, during the rotation of 

 the eyeball round its horizontal or round its vertical axis, no 

 swivel rotation has taken place, for otherwise the negative image 

 would have been turned round, and its cross would make an angle 

 with the image of the cross on the wall. If the pattern on 

 the wall be changed so that the lines while still at right 

 angles to each other are oblique, not vertical and horizontal 

 (this is most conveniently done by using not a wall but a large 

 board and turning the board round), and the observation be 

 repeated except that the eye is turned not vertically or hori- 

 zontally but obliquely so as to follow the lines of the pattern, 

 it will still be found that the cross of the negative image coincides 

 with the cross of the pattern, and that whatever be the angle 

 round which the board has been turned. This shews that 

 Listing's law holds good not only for up and down and side to 

 side movements but also for oblique movements, for movements 

 of rotation round an axis which whatever its obliquity lies in a 

 plane at right angles to the visual axis. 



The same result as regards oblique movements may be ob- 

 tained in another way even while the lines of the wall or board 

 are allowed to remain vertical and horizontal. If in this case 

 the eye be directed not up and down, or from side to side, but 

 diagonally from the fixed centre in the middle of the board to 

 one of the corners of the board, the cross of the negative 

 image will not coincide with the cross of the pattern at the 

 corner but will appear to slant; it will appear to slant to the 

 right in the right hand upper and left hand lower corner, to 

 the left in the left hand upper and right hand lower corner. The 

 slanting cannot be due to a swivel rotation, since in that case the 

 slanting would be in the same direction at both right hand 

 corners, and would be contrary to that occurring at both the left 

 hand corners. The discrepancy between the cross of the negative 

 image and the cross of the pattern at the corner is to be explained 

 by the fact that a horizontal line in the extreme upper part or in 

 the extreme lower part of the field of vision, appears to us curved, 

 bent up or down at each right or left end of the line ; and a ver- 

 tical line in the extreme lateral part of the field of vision appears 

 also curved, bent to the right or left at each, upper or lower, end 

 of the line. Hence in the experiment in question we are com- 

 paring the cross of the negative image, riot with a rectangular 

 cross in the pattern, but with one, the arms of which seem to dip 

 one way or the other and the two crosses necessarily slant towards 

 each other. Our mind, however, corrects this dip, and regards 

 the cross of the pattern as still rectangular, and in so doing judges 

 the obliquity to belong to the negative image. The experiment 



