of Binocular Vision. 



181 



case, and in most other cases tried, the eyes in convergence rotate 

 on the optic axes outward, and that the amount of rotation in- 

 creases with the degree of convergence. Meissner* has attempted 

 to determine experimentally the position of the horopter, and 

 from the position thus determined he infers the rotation of the 

 eyes: my experiments prove directly the rotation of the eyes; 

 and from this, as well as from direct experiment, 1 hope to esta- 

 blish the position of the horopter. 



Helmholtz, it is true, admits some degree of rotation of the 

 eye on the optic axis, particularly when the eye makes wide ex- 

 cursions in the field of view ; but that he does not regard this 

 as sufficient to interfere seriously with the law of Listing is evi- 

 dent from the form of the horopter which he deduces. More- 

 over, according to Helmholtz, these slight rotations are con- 

 trolled by the law of Donders, viz. " that the eye returns always 

 into the same position when the visual line is brought into the same 

 direction" He regards this law as rigorously exact. "Every 

 position of the visual line," he says, " is connected with a deter- 

 mined and constant degree of rotation/'' But the experiments 

 about to be described prove that under certain circumstances the 

 law of Donders, too, is far from being true. 



We have already stated (p. 136) that when the squares of the 

 ruled diagram (fig. 5) are combined by converging the optic 



Fig. 5. 



i . ■ 



axes, if the amount of convergence be great, the horizontal lines 

 of the two images are distinctly observed to cross each other at a 

 small angle. After my attention was once directed to this fact, 

 I could see slight crossing of the horizontals for every degree of 

 convergence ; but the verticals seemed to coalesce perfectly. By 

 placing, however, both the diagram and the head perfectly 



* Bib. Un. Archiv. des Scien. II. vol. iii. p. 160. 



