of Binocular Vision. 191 



strong convergence, however (as, for instance, when the point 

 of sight is less than seven inches distant), the rotation continues 

 to increase as stated by Meissner. 



In all these experiments, in order to detect the true rotation, 

 it is absolutely necessary that the median line of sight should be 

 exactly at right angles with the plane of the diagram. The least 

 error in this respect will cause perspective convergence of the pa- 

 rallel verticals, or increase or decrease of the angle of inclination 

 of the inclined verticals. With the diagram three feet distant, if 

 my eyes look one inch above or below their true level, on combi- 

 ning the two halves of the diagram I can detect the perspective 

 convergence, upward or downward, with the greatest ease. In 

 all cases also, but particularly when the convergence is very 

 strong, it is necessary to fix the attention on that horizontal 

 which passes through the small circle ; for those above and below 

 converge by perspective. 



In these experiments the size of the diagrams is of little im- 

 portance. I have used them of every size from 5 by 10 inches 

 to 15 by 30 inches. 



My next desire was to determine how far these results were 

 general for normal eyes. The great difficulty was to find any 

 one who was able to repeat the experiments. Nevertheless I have 

 found four young persons with normal eyes who, with some 

 practice, have succeeded in all except the most difficult of them. 

 Their results agreed perfectly with my own. In a fifth case, how- 

 ever, in which great difficulty was experienced and the results 

 were uncertain, I was led to believe that the eyes in convergence 

 rotated inward. It is not improbable, therefore, that normal 

 eyes differ in this respect. 



We believe, therefore, that we are justified in the conclusion 

 that when the eye is in its primary position and therefore pas- 

 sive, the vertical line of demarcation coincides with the vertical 

 meridian, and the horizontal line of demarcation with the hori- 

 zontal meridian of the eye, and therefore these two lines of de- 

 marcation are at right angles to each other. But as soon as the 

 eyes begin to converge, the oblique muscles (particularly the 

 inferior oblique) begin to act, rotating the eye on the optic axis 

 and slightly distorting its form ; so that the vertical line of de- 

 marcation is now not only no longer coincident with the vertical 

 meridian, but also no longer at right angles to the horizontal 



is, that when my eyes are parallel or nearly so, elevation of the visual plane 

 causes inward rotation. In some other eyes, however, I have found that 

 elevation of the visual plane when the eyes are parallel causes outward 

 rotation as stated by Meissner. In these cases, therefore, Meissner's 

 results on this point are entirely true. 



