In regard to ”Packing arrangement of the three cone classes in primate retina”, Roorda et al, Vision Research, Vol.41, 1291-1306, (2001)
February 2nd, 2006 | No Comments »In regard to ”Packing arrangement of the three cone classes in primate retina”, Roorda et al, Vision Research, Vol.41, 1291-1306, (2001)
The authors begin with a perceived dilemma in their Introduction. I will quote from the first paragraph:
“Spatial vision and color vision depend on the same mosaic of cone photoreceptors, yet their demands are different and to some extent in conflict. Fine spatial resolution is best served if photoreceptors all have the same spectral sensitivity, so that local variations in the signals originating in different cones represent only differences in intensity and not differences n spectral composition. On the other hand, to resolve fine variations in the spectral composition of a scene requires neighboring photoreceptors to have different spectral sensitivities. The visual system must therefore inevitably compromise in balancing good vision against good color vision.”
They go on in the second paragraph to relate, what I have already discussed, that:
“In the trichromatic eye of humans…..the spatial structure of images is conveyed predominantly via middle (M) and lon (L) wavelength sensitive cones…”
Now, besides commenting that I do not believe that the evolution of biological structures creates dilemmas, the first paragraph I believe rather unambiguously states the fundamental belief of the authors that some sort of logically organized retinal “mosaic” of color sensitive centers forms an intensity-only sensitive image plane of the eye. Ths means that the retina must consist of some kind of ordered array of RGB-sensitive light detection centers similar to the arrays that we use technologically to form the photo surface of imaging phototubes, solid state devices, etc. This paper goes on to,and consists primarily of, a protracted effort to find such a logical array in the measured retinal morphology.
As to the comment of the second paragraph cited above, they must somehow explain the presence of two (why two?) color sensitive sites (or types of “cones) within the foveal region ..the region associated with visual acuity.
I will immediately cut to their Conclusion - another dilemma - before my comments:
“The arrangement of the S. M. and L cones near the fovea in the human can be considered to be random. Given the size of our sample randomness represents a very narrow and restrictive condition among the range of possible mosaics that can be formed. Therefore, we conclude that the apparently small tendency toward aggregation…is likely an artifact…..”
My comments:
The author’s analysis of cone arrangement apparently relies on measurements made earlier that I have discussed in a previous Comment (Roorda et al, “The Arrangement of the Three Cone Classes in the Living Human Eye”, Nature, Vol.397, 11 February 1999). Their Figure 2 seemingly reports cone mosaics from a number of individuals from which they attempt to find some ordered color center structure. They do not specify the retinal angle that the graphics of Figure 2 represent but the distribution of centers appear similar to the figures of the earlier paper….and that was reported as being measured at one degree of retinal angle. They appear the same with a preponderance of red centers interspersed with green and a lesser number of blue.
Again, this is, from the perspective of Osterberg’s data and drawings , exactly what one should see at this retinal angle! There is no mystery about it! Rods are beginning to intrude into the hexagonal cone matrix of the fovea forming a few, and increasingly greater numbers as retinal angle become larger, green sensitive centers. These will be, and as found by Osterberg, statistically distributed..exactly as the authors observe. In this process of rod intrusion and increasing rod density, there will, again statistically, be a fewer number of rod-rod appositions and thus blue sensitive centers scattered about. The author’s bleaching and imaging methodology IS correct..and they are observing what they should be observing! For the life of me I do not see why, with all of the other data reporting cone and rod gradients from the foveal region, this persistence in wanting to see an ordered mosaic array of color centers!
Incidentally, my prediction of a totally green field at 7-8 degrees is based on the beautifully perfect arrangement of eight rods surrounding each cone at this angle. This is shown in Figure 7 of my original paper. (taken from M.H.Pirenne, “Vision and the Eye,” Pilot Press, 1948) At this point there is no statistical ambiguity…all cones and rods are used to form green sensitive centers (at eight angles of polarization?). Why this singular point on the retina where all of the rods and cones are purposefully used? I have proposed that this point uniquely defines the geometrically-defined mid band that Land deduced was required.
I again propose that this successful measurement employed by the authors for imaging the retina be made at larger retinal angles….first perhaps at 3-5 degrees and finally at the 7-8 degree angle. What I believe will be seen is an increase in the density of green centers culminating at the 7-8 degree point where response will be totally green. I would wish that these measurements be made.
Another possible experimental effort….instead of statistically analyzing the distribution of centers in an attempt to find order, perform a statistical analysis of the distribution of cones and rods at the one degree angle based on the measured densities of these receptors around this point. I would predict that the distribution found in the Roorda measurements would result.
GCH
