On the Historical Misunderstanding That “Classes of Cones” Exist on the Retinal Surface – Focusing Now on the Non-Existence of “Green Sensitive Cones”

by Gerald Huth on December 13, 2005

I will begin by referencing Osterberg’s 1935 diagram of the distribution of cone and rod receptors on the retinal surface. I do not believe that there is any disagreement as to the validity of these data as it is appears in almost every textbook on vision and the eye. This diagram appears as Figure 2 of my original paper where it, in essence, forms the basis for this entire concept and where it is shown that this specific distribution of receptors is in exact consonance with the diffractive properties of the lens and structure of the eye.

But first, I have always been in wonder how anyone could ever reconcile this magnificent asymmetry of cones and rods with the idea that the retina was some sort of analogue of photographic film! Illustrations from vision textbooks repeatedly portray the retina (the “inverted arrow” diagrams) as the “image” plane of the eye. Such a retina would have to be composed of some sort of uniform array of RGB detection centers (or at least three separate color sensitive layers as color photographic film). It is just so obvious that this is is not the case! How might the vision science field have come to the conclusion that it is such a surface?

I will begin with the old saw, again repeated over and over, that “cones detect color” and “rods are the low level black and white receptors of the eye”. (I repeat a comment made to me by a vision science investigator that “no one believes this anymore”…what?) I have provided an explanation for the statement concerning rods elsewhere on the page – to the point that rod receptors that comprise the peripheral retina are linked together (i.e., linked “in parallel” which has been apparently experimentally verified) to form the “light meter” of the eye with this integrated array of receptors controlling pupillary constriction and thus the level of light entering the eye. So, it can be said that rods are involved in low level light detection but not at all as historically assumed. I can see no scientific basis for the assertion that single rod receptors, in themselves, have greater sensitivity to light (than cones).

But, to the historic belief that “cones detect color” and, further, that there are somehow three “classes”of cones (i.e., sensitive to separate red, green, and blue wavelengths). How on earth could this belief have come about?

I will digress here to make the point again that the term “color” itself should not even be associated with light interaction with retinal receptors. Rather this term implying a panoply of hues should properly be applied to a “synthesis of the three signals detected on the retina” appearing in the “color centers” of the brain – as so brilliantly deduced by Edwin Land. The retina itself detects only the “brightness” (or intensity) of three narrow RGB optical wavelengths) as shown in this work.

Further, doesn’t the idea of a single “color detecting cone receptor” imply that such a receptor is acting as the analogue of some sort of “complete laboratory spectrometer”? The answer is really much simpler, i.e., that an apposition of two cones forms a “tuned optical wavelength antenna” that has been located evolutionarily at a position on the retina that is in consonance with the diffractive properties of the lens and structure of the eye. The eye is then a biological structure that evolved using the principles of simple physics and geometry, with the subjective concept of “design” being not at all necessary. Science has done itself a great disservice by not getting at the fundamental principle involved here.

It is axiomatic in this work that it is the function of each cone-cone apposition to detect solely the long (red) wavelength. In fact, it is the dimension of this apposition that defines the long wavelength limit of visual response. (I have proposed that “blue sensitive cones” are actually rod-rod appositions and this is discussed in detail elsewhere on the page). It therefore seems logical how the idea that one of the classes of cones (i.e., the “red” sensitive cone) resides in the all-cone fovea. That much of the “class of cone” rationale is a true statement ..but for the wrong reason.

But where did the idea of the existence of “green sensitive cones“ come from? In this work “green” sensitivity derives from the dimensionality of cone-rod appositions (and, moreover, we show how it geometrically defines the exact mid-band point). Examining Osterberg’s data shown above it can be seen that rod receptors begin to “intrude” into the all-cone foveal region at very small retinal angles. It is not really obvious in the figure how small an angle this is but a statement from Pirenne (Vision and the Eye, The Pilot Press Limited, London, 1948) gives the magnitude…..”The first rod…..is situated at distance of 0.13 mm from the foveal center”. A small density of cone-rod appositions is therefore present immediately adjacent to (or, as I propose, “perceived as within”) the overall array of foveal cones. Dimensions here are extremely small leading to difficulty in differentiating light detection centers …and this has probably led to the misconception that a different type of cone is the source of green wavelength detection.

One can see a closer inspection of the above by studying Fig 28 from Pirenne that presents a diagrammatic representation of cone and rod densities in this crucial sub-millimeter region of the retina.



Leave a Comment

{ 7 comments… read them below or add one }

Joseph Carroll December 28, 2005 at 9:02 am

How do you explain color vision then, in non-foveate animals?

Moreover, your theory has many holes, and ignores a myriad of contradictory evidence. For example, you have ignored all of the recent literature describing quite thoroughly how (and why) the fovea forms. The mechanism of pit formation is the causative factor in the increasing cone density (i.e., decreasing cone diameter) in the fovea. This is easily seen if you look at the distrubution of cones/rods in a fetal retina and see no fovea, and see that the cones are all the same size. Its the physical process of the formation of the foveal pit (that isnt complete until about 4 years after birth in humans) that gives rise to the observed distribution of cones in the retina.

All in all, I think it is bad scientific practice to ignore large volumes of data for fear it might weaken your hypothesis. That such practice ever makes it into the heads of aspiring scientists would be a tragedy.

ghuth December 28, 2005 at 11:01 am

As I read what you write it seems to me that your comments reinforce rather than refute my view.It is obvious in physics terms that the retina evolved as an array of “generic nanowires and wavelength-defining spaces between them” positioned exactly where optical wavelengths at the Fourier plane would interact. I am well aware of the foveal pit, have not discussed it believing that it was just a structure whose purpose was to further improve optical acuity at the foveal center, I am speaking here in solely optical physics terms..do your comments pertain perhaps to the biological evolution of this structure? Your other reference to fetal development seems to come from this same direction. Be clear… I am always talking only of the light interaction PHYSICS within the eye…..and if you want to be more precise…”within the fully developed eye”. I do not understand how biological consideration of the neonatal (or otherwise) developing eye has anything to do with my premise.

Again, to be clear, it is my view that the biological structures of the eye evolved as a result of, or in consonance with, the fundamental PHYSICAL principles of the diffraction of light. Biological considerations of growth and form I just take for granted…they will follow evolution as did ordered arrays of lipid molecules in the formation of the membrane of cells. I note biological and/or vision science related points when they appear to agree with my model (or otherwise seem unexplained)….noting, for example, George Wald’s experimental finding of the “blue blindedness” of the fovea…..explained herein.

And I really don’t consciously ignore “large volumes of data….etc What I do like are IDEAS and..sincerely hope that they somehow make it into the heads of aspiring scientists.

And I certainly thank you for your comments!


Joseph Carroll January 3, 2006 at 12:42 pm

“I do not understand how biological consideration of the neonatal (or otherwise) developing eye has anything to do with my premise.”

Because on the one hand you say that the size of the receptors is genetically determined, and even have used this to explain away the genetic basis of color blindness. But what I’ve referenced is a purely mechanical process that defines the cone diameter as a function of retinal location. So this leaves unexplained the genetics of color vision that has been brought up before by me.

ghuth January 4, 2006 at 10:38 am

I have not “explained away” a genetic basis for color variant vision (“color blindness”). I have proposed, consistent with my hypothesis, that the size of an individual’s cone (and rod) receptor is (certainly) genetically determined..and, speculatively, that this specific gene must code for the protein structure that determines the STRUCTURAL diameter of the inner segments of these receptors ..since it is this dimension that sets the wavelength-defining spacing of the light interactive (“nanowire”) outer segments.

I have noted for you that it is the diameter of the cone receptor that determines the POSITION ON THE RETINA OF THE GEOMETRICALLY-DETERMINED MID-BAND POINT. Larger cones have the effect of moving the peak of the mid-band brightness peak (the point where rod density is first sufficient to completely surround each cone) outward from it’s referential 7-8 retinal angle point. Since there is no reason for the refractive properties of the eye to change, mid-band wavlengths will incident the retina “off-center” from it’s usual, crucially referential, point of interaction. Land showed beautifully how even a slight mismatch here drastically alters the perception of color. One MUST have some knowledge of Land’s work to understand this.

Regarding a genetic explanation for color blindness…my first thought when reading that genetics “explained” this condition was that specific genes had been identified that “coded for the three RGB pigments that are supposed to differentiate the three types of cones”. I believe that this is a logical asssumption and one that, for example, the public reading such scientific news would think, namely, that knowledge of a specific PHYSICAL link between gene and pigment had been found. In looking into this, as far as I can find, no such direct linkage has been identified rather an INDIRECT (that word “inferential” again!) linkage relating a specific gene and the medical condition of blue color blindedness. I am not a geneticist and would appreciate comments here.


Joseph Carroll January 6, 2006 at 12:30 pm

“I believe that this is a logical asssumption and one that, for example, the public reading such scientific news would think, namely, that knowledge of a specific PHYSICAL link between gene and pigment had been found. In looking into this, as far as I can find, no such direct linkage has been identified rather an INDIRECT”

Review the work of Dan Oprian (Neuron paper) who systematically changed the amino acids in the red pigment until its absorption was that of the green pigment, and vice versa. This is DIRECT evidence that the genes that Jeremy Nathans identified do in fact encode for visual pigments with 3 different spectral sensitivities.

ghuth January 6, 2006 at 1:20 pm

Would you please identify the specific Oprian/Neuron reference…..Oprian is an author on many vision papers and I cannot find the specific paper that you cite. A cursory review of Oprain’s and related color blindness/receptor papers using Google seems to indicate that most agree with my interpretation…. most start with a connecton between receptor pigment and color blindness (the red/green variety whatever it is termed) as I have noted, but go on to ASSUME a link between gene and specific receptor pigment without providing an actual physical mechanism for such a linkage. A paper on the subject that you describe would seem to provide such a direct link…and I would like to review it.



Joseph Carroll January 7, 2006 at 12:17 pm

Asenjo AB, Rim J, Oprian DD.
“Molecular determinants of human red/green color discrimination.”
Neuron. 1994 May;12(5):1131-8.

Previous post:

Next post: