Does The Concept Of ‘Classes’ Of Cones Have Any Merit? —I Think That The Answer Is No!

The following figure is from Cornsweet ( “Visual Perception”, Academic Press) portraying the absorption spectra from the “three classes of cones” of the retina. As C. states, these same curves have been reproduced in many places. Puzzling to me when I first looked at them was the question “where is the red response?” A wavelength of 570 nm corresponds to yellow-orange – not red! Moreover, if red lies at wavelengths from 650-700 nm the retina has zero response in this region! How can this be? Hasn’t anyone noted this?

I did not at that time realize that the “blue sensitive cone” curve was ENTIRELY FICTITIOUS! Although Cornsweet states the generality that the curves have been “arbitrarily adjusted” this in no way gets the message across that the blue sensitive curve just isn’t there!!! Any fragmentary “blue” sensitivty that has been measured (which I quarrel with) is “down in the noise” and doesn’t form a curve at all! How much misinformation has this curve caused having been reproduced over and over and is in textbooks worldwide.

The following text was taken at random from a website which starts by showing the same trichromatic cone response figure that Cornsweet used. I quote it here because I believe that it represents the “standard thinking” regarding the concept that “classes of cones” are thought to exist on the retinal surface:

“Boosting mechanism” “Set equal for lack of detailed data” Blue sensitive cones “have the highest sensitivity” and “constitute only 2% of the total number” – and these are found “mostly outside of the fovea” (where greater than 99% of all cones reside!) — what kind of nonsense is this!

The conclusion about being found outside the fovea is in consonance with the geometric model – but they aren’t cones at all – they are rod-rod appositions! THERE ARE NO BLUE SENSITIVE CONES.

The statement that the blue cones somehow have “higher sensitivity to overcome their low density” and some”boosting mechanism” – I would claim that the “boosting mechanism” is actually derived from the large area which the totality of rods occupy and the integrated signal derived therefrom.

“Red” cones are indeed packed into the fovea centralis – but this wavelength response is a result of the cone-cone appositions in this region. “Green” response should really be attributed to the density of cone-rod appositions that start to appear at retinal angles of ~5 degrees and peak at ~ 7 1/2 degrees.

Correctly noted is the exceptionally non-uniform distribution of cones over the entire retinal surface. Accepting this, how anyone conceives that the retina represents an intensity-only detecting,”‘photographic film-like” medium is beyond me!

My explanation as to how the geometric model explains these curves follows below (again, as noted elsewhere on this webpage, I propose to use wavelength and the physics deriving from that term and not “color” to describe the response of retinal receptors).

1. The short wavelength “blue cone” curve at ~440nm must correspond to some measurement inaccuracy (or a measurement of rod/rod appositions). As discussed elsewhere on this webpage I do not believe that it has ever been shown that a “blue”or short wavelength sensitive cone exists.

   It will be argued that measurements of single cones were probably used to formulate these spectra. But there are many difficulties in measurements made in this optical wavelength spatial domain. Because of the difficulty in illuminating a single retinal receptor “end-on” (following from the diffraction effects in collimation of light to micron retinal receptor dimensions) such measurements are usually made using a slit of light focused longitudinally to illuminate along the length of a single cone that has been “teased” from a retinal section. Such slit illumination, however, still must deal with diffraction in the width of the slit – which will probably extend beyond the intended micron or sub-micron dimension. One would then seem to be actually measuring optical absorption due to the refractive index of the intended receptor itself but also necessarily including the index of the surrounding medium. A gradient of refractive indices is thus measured which corresponds to a index-graded spatial dimensionality which is the premise of my hypothesis. I would guess that it would be possible to find in any single experimental configuration a range of absorbances.

   The logic for the validity of a slit measurement must be that what is being measured is some totality of the retinal/rhodopsin pigment complexes that are in fact uniformly distributed along the length of the cone. In this view the more pigment complexes that are illuminated the greater the sensitivity of the measurement – perhaps to surmount some statistical threshold.

   Additionally, such a side-looking, slit measurement must take into accont the observed dichroic effects in the illumination of retinal receptors. Rhodopsin/retinal complexes behave decidedly differently when illuminated in the plane of the thylakoid membrane in which they are embedded, i.e., from the side of the receptor, than when they are illuminated end-on (the longitudinal dimension of the receptor). The difference is in fact a factor of 6. Such dicroism with increased sensitivity “from the side” would seem to favor the geometric model proposed herein.

2. The “mid-band” ( or “G” ) cone response at 520 nm must correspond in the geometric model to a measurement of a cone/rod apposition (or it’s graded index equivalent) derived from the region outside of the central fovea. I would believe that this measurement is actually of a “spacing” between cones and rods that invariably appear in this area or alternatively, the measuremental consideration described above.
3. The questionable long wavelength curve – which actually does not correpond to red! It is my proposal that the only place where long wave response will be measured is in the very small, all cone region of the central fovea. Cone-to-cone appositions must be present to display absorption at 700 nm. I would propose that what is being measured here, based on the comments under A.) above, is “some chance spatial surround” that displays a slightly longer wavelength absorption – and thus finally “a peak”?

ADDED:

Since Cornsweet mentions Wald’s work, I would note that I have reviewed a great deal of that work. As noted elsewhere, Wald experimentally determined and reported that the fovea of the retina was “blue blind” (his term) which is axiomatic in the geometric model. He also, after discovering the chemical specie of only one retinal pigment, searched diligently for “the other two” that he felt must be present. He never found them. It is my assertion that they do not exist. The function of the retinal/rhopsin complex is to form an “electron sink” for the light absorption process and is the singular specie that serves this function in all retinal receptors.

It will be argued that in more recent times the genetic basis for the “other” pigments has been found. I would propose that what in fact might have been found is the genetic coding for proteins that in some way determine spatial size of the inner segment of retinal receptors – the portion of the receptor charged with setting intereceptor spacing and wavelength response.

FROM “ADDITIONAL THOUGHTS” OF 6/19/02… PLEASE READ!!!

A friend has given me the video of a documentary produced by the BBC in the 1980′s entitled “Colourful Notions”. The narrative is essentially divided into three parts: a.) an exposition of the traditional Young et al theory of color vision ending by presenting the color constancy delemma, b.) a beautiful sequence where Edwin Land personally demonstrates his color vision experiments, and, finally, c.) an uninteresting denouement wherein it is proposed that “the brain did it” using that organ about which we know so little to wrap up and explain everything (of course, the brain ultimately detects color!).

An interesting sequence is in the first part of the film where individual retinal receptors (or fragments thereof) that have been teased from dead retinal sections are spectrophotometrically examined under a microscope in an attempt to justify the assertion that “three classes of cones” exist. .I have written about these experiments elsewhere on this page noting the numerous problems with this type of measurement. But…one notes in the film at this point what can only charitably be termed “filmatic license”. Apparently, while an individual receptor is being measured and during wavelength scanning, the suspiciously overlapping middle “green” and long “red” (which is really “orange” containing no red wavelengths!) wavelength peaks magically rise to the what must be a normalized “100″ level on the oscilloscope presentation. I would emphasize again that it is purported that all of this is being done in real time.

BUT THEN… the measurement purported to be from “blue” sensitive cones – which is again termed a “peak” – looks suspiciously like noise lifted to the “100″ level – it is definitely not a peak!

What is this?

CAN ANYONE REALLY BELIEVE THIS?

Respectfully submitted,
Gerald C. Huth,
Ojai, CA
10/22/01