BINOCULAR VISION. 123 



Malebranclie : c Suspend by a thread a ring, the opening of 

 which is not directed toward us ; step back two or three 

 paces ; take in the hand a stick curved at the end ; then, 

 closing one eye with the hand, endeavor to insert the curved 

 end of the stick within the ring, and we shall be surprised at 

 being unable to do in a hundred trials what we should be- 

 lieve to be very easy. If, indeed, we abandon the stick and 

 endeavor to pass one of the fingers through the ring, we shall 

 experience a certain amount of difficulty, although it is very 

 near. This difficulty ceases at the instant that both eyes are 

 opened.' (MALEBRANCHE, Des sens, liv. l er , chap, ix.) 



" As regards precision, exactitude of information concern- 

 ing the relative distance of objects, that is to say, the idea of 

 the third dimension or of depth, there is then a notable dif- 

 ference between binocular vision and that which is obtained 

 by means of one eye alone. 



" This difference is brought strikingly into view by the 

 following remarks of Malebranche, remarks which themselves 

 imply an anticipated idea of the very principles of associated 

 vision. 



" Returning to the experiment with the ring, Male- 

 branche adds the following judicious remarks : ' But we 

 must certainly note that I have said that we endeavor to pass 

 through the ring from the side, and not by a straight line 

 from the eye to the ring ; for then there would be no diffi- 

 culty, and indeed it would be still easier to accomplish this 

 with one eye closed than with both eyes opened, because 

 that guides us.' (Principle of direction, without localization.) 

 f Therefore, one could say that the difficulty experienced in 

 passing through a ring from the side, with but one eye 

 opened, is because, the other eye being closed, the angle of 

 which I have spoken (the angle of convergence of the optical 

 axes) is not known ; for it is not enough, in order to recog- 

 nize the size of an angle (read in order to recognize a tri- 

 angle), to know the size of the base and of the angle f onned 

 by one of the sides with the base, which is known in the pre- 



