594 PHYSIOLOGY 



100 mm. nearer to and 100 mm. further from the observer will lie in the thickness of 

 one and the same shell, and will therefore appear the same distance from the observer. 

 Objects between 100 and 300 mm. nearer to the observer will lie in the shell correspond- 

 ing to one cone discrepancy, and will therefore be appreciated as being at a different 

 distance from the observer, appearing nearer if crossed, and further if uncrossed. The 

 same reasoning applies to objects at other distances. If this calculation is correct 

 it should be necessary to place objects more than 100 mm. from a fixation mark, which 

 is itself placed at 10 metres, in order that a difference in the distance from the observer 

 should be appreciated. Greeff found by experiment that 5 l ^th the distance of the 

 fixation mark was necessary (i.e. 200 mm.), the observations being instantaneous ones. 

 If time be allowed for prolonged observation, greater accuracy in the appreciation 

 of distance is obtainable, because different points of fixation can be used. Suppose, 

 for example, that two objects 20 mm. apart be examined at a distance of 10 metres, 

 under instantaneous observation they will appear identical as described above ; but if 

 the examination be made more carefully, it will be found that, on fixating a point a mean 

 distance of 100 mm. away from the objects, the distance between the two is suddenly 

 appreciated because the demarcation between two shells now falls between them. It 

 is in this way that the accuracy of extended observation becomes greater than that 

 obtainable with instantaneous. The limit reached by experiment is stated to corre- 

 spond to a displacement at the retina corresponding to the ^th diameter of a cone. 

 The corresponding values for the acuity of stereoscopic vision would be y^th those 

 given above, namely -2 mm. at 1 metre, "20 mm. at 10 metres, and 1-7 mm. at 100 metres. 

 Bering's view would therefore appear to agree well with the results of experiment. It 

 remains to consider the type of cortical mechanism that would be necessary 

 for the estimation of the discrepancy between the images. One type may be briefly 

 described as follows : To a number of parallel planes hi the left side of the cortex are 

 connected the terminal ends of the nerve fibres from the left halves of the two 

 retinae. At the middle plane fibres from exactly corresponding retinal points are 

 connected together. At planes which lie superficially to the middle plane are connected 

 other terminations from the same fibres but with a crossed lateral discrepancy of one 

 cone in the 1st plane, two cones in the 2nd plane, three cones in the 3rd plane, etc.- At 

 planes which lie deep to the middle one, other Iberminations from the same fibres are 

 connected but with an uncrossed lateral discrepancy of one, two, three, etc., cones as the 

 case may be. On looking at a fixation mark on a uniform background therefore, a 

 series of impressions of the mark will be formed on all these planes, but in the central 

 one only will they exactly agree, for in all the others the lateral discrepancy will cause 

 the impressions to be duplicated. In all these other planes there will thus be antago- 

 nism, first one image and then the other predominating. When these images are com- 

 bined in consciousness, the stable image from the central plane will suppress the unstable 

 ones from all the other planes, the result being a single picture of the fixation mark. 

 If there are in front of the fixation mark other objects lying in planes at different 

 distances from the observer, the impulses sent up by the cones to the cortical plan 

 will not correspond at the central plane, because their images no longer fall on corre- 

 sponding points, but they will correspond in the superficial planes where the discrepancy 

 of their images agrees with the discrepancy of the nerve connections. These other 

 planes will therefore predominate according as each contains the identical images, 

 and when they are combined in consciousness these planes will suppress all 

 the others. Since each cortical plane represents a certain lateral discrepancy, it 

 must also represent a certain distance from the fixation mark. If consciousness recog- 

 nisrs the plane in which stable image is formed, it also must appreciate the distance 

 of the object lying in that plane from the fixation mark. This would not seem any 

 more difficult than the localisation of a touc*h on the skin. So far as we are able to judge 

 there is nothing inherently impossible in the arrangement of the hypothetical cortical 

 mechanism which has been outlined above, and therefore Bering's theory would appear 

 to be very plausible, 



