Eye, Brain, and Vision
Home
Book
Illusions
Biography
Publications
David Hubel's
Home
Book
Illusions
Biography
Publications
HMS
 
     
                                                                                    
   
 
The recording electrode was close enough to three cells to pick up impulses from all of them. Responses could be distinguished by size and shape of the impulses. This illustrates the responses to stimuli to single eyes and to both eyes. Cells (1) and (2)
both would be in group 4 since they responded about equally to the two eyes.
Cell (3) responded only when both eyes were stimulated; we can say only that it was not a group i or a group 7 cell.

satisfied with the striate cortex and wanted to move on to the next area (in fact, we had moved on), we happened to record from a sluggishly responding cell in the striate cortex and, by making the slit shorter, found that this very cell was anything but a sluggish responder. In this way we stumbled on end stopping.
And it took almost twenty years work with the monkey striate cortex before we became aware of blobs—pockets of cells specialized for color, described m Chapter 8. Having expressed these reservations, I should add that I have no doubt at all that some of the findings, such as orientation selectivity, are genuine properties of these cells. There is too much collateral evidence, such as the functional anatomy, described in Chapter 5, to allow for much scepticism.



                              BINOCULAR CONVERGENCE
I have so far made little mention of the existence of two eyes.
Obviously, we need to ask whether any cortical cells receive input from both eyes and, if so, whether the two inputs are generally equal, qualitatively or quantitatively.
To get at these questions we have to backtrack for a moment to the lateral geniculate body and ask if any of the cells there can be influenced from both eyes. The lateral geniculate body represents the first opportunity for information from the two eyes to come together at the level of the single cell. But it seems that the opportunity there is missed: the two sets of input are consigned to separate sets of layers, with little or no opportunity to combine. As we would expect from this segregation, a geniculate cell responds to one eye and not at all to the other. Some experiments have indicated that stimuli to the otherwise ineffective eye can subtly influence responses from the first eye. But for all practical purposes, each cell seems to be virtually monopolized by one or the other eye.
Intuitively, it would seem that the paths from the two eyes must sooner or later converge, because when we look at a scene we see one unified picture. It is nevertheless everyone's experience that covering one eye makes no great difference in what we see: things seem about as clear, as vivid, and as bright.
We see a bit farther to the side with both eyes, of course, because the retinas do not extend around as far in an outward (temporal) direction as they extend inwardly (nasally); still, the difference is only about 20 to 30 degrees. (Remember that the visual environment is inverted and reversed on the retina by the optics of the eye.) The big difference between one-eyed and two-eyed vision is in the sense of depth, a subject taken up in Chapter 7.
In the monkey cortex, the cells that receive the input from the geniculates, those whose fields have circular symmetry, are also like geniculate cells in being monocular. We find about an equal number of left-eye and right-eye cells, at least in parts of the cortex subserving vision up to about 20 degrees from the direction of gaze. Beyond this center-surround stage, however, we find binocular cells, simple and complex. In the macaque monkey over half of these higher-order cells can be influenced independently from the two eyes.
Once we have found a binocular cell we can compare in detail the receptive fields in the two eyes. We first cover the right eye and map the cell's receptive field in the left eye, noting its exact position on the screen or retina and its complexity, orientation, and arrangement of excitatory and inhibitory regions; we ask if the cell is simple or complex, and we look for end stopping and directional selectivity. Now we block off the left eye and uncover the right, repeating all the questions. In most binocular cells, we find that all the properties found in the left eye hold also for the right-eye stimulation—the same position in the visual field, the same directional selectivity, and so on. So we can say that the connections or circuits between the left eye and the cell we are studying are present as a duplicate copy between the right eye and that cell.
We need to make one qualification concerning this duplication of connections. If, having found the best stimulus—orientation, position, movement direction, and so on—we then compare the responses evoked from one eye with the responses evoked from the other, we find that the two responses are not necessarily equally vigorous. Some cells do respond equally to the two eyes, but others consistently give a more powerful discharge to one eye than to the other. Overall, except for the part of the cortex subserving parts of the visual field well away from the direction of gaze, we find no obvious favoritism: in a given hemisphere, just as many cells favor the eye on the opposite side (the contralateral eye) as the eye on the same side (the ipsilateral). All shades of relative eye dominance are represented, from cells monopolized by the left eye through cells equally affected to cells responding only to the right eye.
We can now do a population study. We group all the cells we have studied, say 1000 of them, into seven arbitrary groups, according to the relative effectiveness of the two eyes; we then compare their numbers, as shown in the two bar graphs on the preceding page. At a glance the histograms tell us how the distribution differs between cat and monkey: that in both species, binocular cells are common, with each eye well represented (roughly equally, in the monkey); that in cats, binocular cells are very abundant; that in macaques, monocular and binocular cells are about equally common, but that binocular cells often favor one eye strongly (groups 2 and 5).
We can go even further and ask if binocular cells respond better to both eyes than to one. Many do: separate eyes may do little or nothing, but both together produce a strong discharge, especially when the two eyes are stimulated simultaneously in exactly the same way. The figure on this page shows a recording from three cells (1, 2, and 3), all of which show strong synergy. One

 

 

 

 

 

 


of the three did not respond at all to either eye alone, and thus its presence would have gone undetected had we not stimulated the two eyes together.
Many cells show little or no synergistic effect; they respond about the same way to both eyes together as to either eye alone.
A special class of binocular cells, wired up so as to respond specifically to near or far objects, will be taken up separately when we come to discuss stereopsis, in Chapter 7.
These hookups from single cells to the two eyes illustrate once more the high degree of specificity of connections in the brain. As if it were not remarkable enough that a cell can be so connected as to respond to only one line orientation and one movement direction, we now learn that the connections are laid down in duplicate copies, one from each eye. And as if that were not remarkable enough, most of the connections, as we will see in Chapter 9, seem to be wired up and ready to go at birth.


   
 
 
Contents
Next Page

 
 
 
 
 



In population studies of ocular dominance, we study hundreds of cells and categorize each one as belonging to one of seven arbitrary groups. A group i cell is defined as a cell influenced only by the Contralateral eye—the eye opposite to the hemisphere in which it sits. A group 2 cell responds to both eyes but strongly prefers the contralateral eye. And so on.

Next Page
Contents