(Rewritten from “Land and Water,” November 6, 1897)

Many interesting problems constantly come before the fisherman, but certainly one of the most interesting which has recently attracted his attention is Sir Herbert Maxwell’s theory on the power of fish to discriminate between various colours.

His theory is, that though fish can undoubtedly discriminate between different shades of light and dark, they cannot distinguish one colour from another. The only conclusion that can be drawn from this theory is the conclusion at which Sir Herbert Maxwell has apparently arrived. This is, that if the same relations of light and shade be maintained in the artificial which exist in the natural fly, the colour of the imitation is quite immaterial.

The facts upon which he based this theory were (1) that during the May-fly season he used several artificial May-flies, some of which were coloured scarlet, some bright blue, and some coloured to imitate the natural fly, all of them being similarly graduated with regard to the shade of their various component parts; (2) that he caught trout with all these flies, no particular one of them being decidedly more successful than the others.

This experience of his no doubt would at first strike one as being very strongly in favour of his theory; but on going deeply into the matter, its bearing on the fish’s powers of vision is not so great as it appears.

To begin with, we must consider whether, judging from experience in the past, trout have been known to rise at things on the water which were not only unlike in colour to any flies on the water, but also unlike them in shape and gradations of shade. This we know they will sometimes do. I have on several occasions seen a trout which refused a fairly accurate imitation of the flies which were on the water rise at and take below the surface a swan’s feather. There are also many other much more extraordinary but similar cases on record. Thus, the fact that these trout took an abnormally coloured fly is not a conclusive proof that they mistook it for the natural fly, particularly as this experiment was made during the May-fly season, when the trout sometimes appear to be quite mad, but are at any rate always much less shy than at any other time of the year.

The experiment, too, was made upon a private water, and I think that there is great doubt that the same result would have occurred had it been made upon a well-fished water where the trout were more shy and better educated.

We must then consider whether, in what we know of the natural history of fish, there are any facts which point towards the probability of their being able to discriminate between different colours. Here we find that there are cases in which in certain species the males are more brilliantly coloured than the females, either at the spawning season or always. This is probably a process in evolution which tends to make them more attractive to the female. We also know that fish sometimes assume a colour similar to their surroundings. This colour is, no doubt, evolved for their protection from enemies, and surely a very large proportion of these enemies are other and larger fish. Many of the larvÆ of water insects and other creatures upon which fish feed are also coloured according to their surroundings, in order to facilitate their concealment. These facts would naturally lead us to come to a conclusion opposed to that of Sir Herbert Maxwell, as the probabilities all point towards the power of fish to discern various colours.

Another very important point is the structure of the fish’s eye in comparison with that of man, who we know has the power of discriminating between colours. This power is, in the human eye, probably situated in the layer of rods and cones of the retina. Had the fish’s retina not contained this layer, as is stated by Sir Herbert Maxwell, there would certainly have been most excellent grounds for supposing that his theory was true; but this layer is contained in the fish’s eye, though it is not the same as in man. If the fish’s eye did not contain it, fish would have been totally blind.

How far this difference in the retina of the fish bears on its sense of colour is, at present, a moot point, though I believe researches are being made in this direction. At present, our knowledge is too limited with regard to it for any definite statement to be made. The probability is, that fish have the power of distinguishing colour from colour. A probability, however, is not a certainty, though one is more inclined towards it than towards an improbability.

Even should Sir Herbert Maxwell’s theory prove true, in spite of probabilities to the contrary, I do not see that we should have progressed very much further with regard to facilities in imitating the natural fly. We know that the relative values of light and shade in various colours contiguous to each other, is not actually the same as the impression conveyed to our eyes. We have an example of this always with us in the photograph, where red and blue, in relation to each other, certainly do not produce the same effects on the plate as they do on the eye; and as we have no accurate knowledge as to the effect of contiguous colours upon a normally monochromatic eye, we could hardly be certain of producing an accurate monochromatic imitation of a multi-coloured object, which would deceive that eye.

The case of a colour-blind human being is certainly not a normal case, so the shade value of the various colours to this eye could hardly be taken as a safe standard.

Even if we assumed that all these difficulties had been surmounted, and that the exact relative shade values to this monochromatic eye of every colour were estimated, I think that there can be no doubt that it would be easier to imitate the colours, with the various shades in these colours, than to calculate out the relative shade values of the different colours, in one particular colour, and that the result of the former and easier, would be much more likely to be accurate than the latter and more difficult attempt.

Besides this, possibly, as the eyes of some families of fish are more highly developed than those of others, the relative shade values of colour might be different to the different families, so that if we eliminate colours from our lures, we must have different shading for different fish.

Having considered all these things carefully, I have come to the conclusion that it will be much safer and easier to keep on using colours in our imitations, even if we do present these imitations to a monochromatic eye.

Since writing the above article, I have been able to collect some further information with regard to the probable power of the trout’s eye to discriminate between colours.

These researches, though I have not yet had time to carry them as far as I had hoped, have led me to believe more firmly than ever that I am right in recommending the use of colours in our imitation flies. I have prepared some sections of the retina of the trout, and examined them carefully in comparison with the retinÆ of several other fish. A short account of what is known at present of colour-vision is, I think, advisable to make my meaning clear to those of my readers who may not be sufficiently well versed in this particular subject.

The sensation of an individual colour is produced by rays of light of a particular wave-length falling upon the retina. A sensation of “white” is produced by rays containing all the wave-lengths which are able to affect it. When, on looking at an object, we find that neither a colour nor white sensation is produced, this sensation is called “black.”

The white sensation may be mixed with the sensation of any colour of the spectrum, as also may the sensation of black, and when these two are mixed they produce a sensation of “grey.” Some colours of the spectrum are probably produced by a mixture of various wave-lengths of different primary colours, and many colours in nature do not exist in the spectrum.

The word “tone” expresses variations of wave-lengths within a named colour, and “brightness” is used to indicate the intensity of the sensation produced upon the retina.

The enormous difficulty of working out into a monochrome the shade-values of a collection of colours, with several tones and shades of brightness in each of the variously coloured parts of the object we wish to imitate, can be imagined on considering these facts only; but there are more facts which lead me to believe that to do this is not only difficult, but impossible.

Two theories have been propounded to explain the sensation of colour produced upon the retina.

The Young-Helmholtz theory teaches that there are three primary sensations—red, green, and violet. Other colours are a mixture of these sensations; white is produced when all three sensations are excited together, and black is an absence of sensation.

Hering’s theory is that there are six primary sensations arranged in three pairs—white and black, red and green, and yellow and blue. He assumes the existence of three visual substances which undergo metabolic changes when subjected to the action of light. These are the red-green, the yellow-blue, and the white-black substances. The white-black substance is influenced by all the rays of the spectrum, while the red-green and yellow-blue substances are differently influenced by rays of different wave-lengths. When all the rays together fall upon the retina, no metabolism takes place in the red-green and yellow-blue substances, but only the white-black substance is affected. Thus the white-black substance is the most active.

Any discussion as to the relative value of these theories would in this work be out of place and unnecessary.

The ordinary form of colour-blindness in human beings is the inability to discriminate between red and green. This shows that the visual power of these people is dichromatic and not trichromatic, as their power is limited to two colours, or pairs of colours, and does not extend to three.

The individuals who belong to this class of the colour-blind may be divided into two sub-classes—those who are red-blind and those who are green-blind.

Those who are red-blind do not see the red end of the spectrum, and the blue-green appears grey, though they have distinct colour vision of the parts of the spectrum on either side of the blue-green. In matching red with a green, they put a bright red with a dark green.

On the other hand, those who are green-blind see the red end of the spectrum, while the green appears to them as grey. In matching a red with a green they put a dark red with a bright green.

No absolutely undoubted cases of blue-yellow blindness have been recorded, and only one of absolute colour-blindness; but one case is not sufficient to go upon.

According to the Young-Helmholtz theory, a case in which only shades of black and white were visible would be impossible, as it would not be shades of black and white which would be seen, but shades of either red, green or blue. According to Hering’s theory, of course, absolute colour-blindness would be possible.

In the normal human eye, only the central parts of the retina are sensitive to colour, the peripheral parts are practically colour-blind. AnÆmia of the retina, which may be produced by pressure on the eye-ball, will render the retina, first colour-blind and then insensitive to light. To me it appears that colours in relation to each other assume a grey tone, and the sensation of black and white disappears last.

The great difference which I have been able to observe between the human retina and the retina of the trout is, that while the human retina contains a layer of rods and cones, the retina of the trout only contains cones, or if it does contain rods, contains very few, as I have not found any as yet. There exists also at the back of the retina of the trout a “tapetum,” which extends over almost the whole of its posterior surface. This does not exist in the human eye, but is found in the eyes of some of the vertebrates. It consists of a layer of “guanin” crystals, and, presenting as it does a metallic appearance, and having great power of reflecting light, probably plays an important part in the visual power of the trout, particularly, I should think, in a dim light.

The fact that the rods are absent from the trout’s retina does not bear the important significance that one would imagine on first realising it. The fovea centralis of the human retina is the seat of most acute vision, and in the fovea centralis there are no rods. The cones in the retina of the trout are very closely arranged, so that they are practically in contact with each other, and their outer limbs are rather longer and finer than in the case of man. This layer of cones extends to the periphery of the retina, and the cones are just as closely arranged as far as they extend. These facts should lead us to believe that the vision of the trout is probably extremely acute, in fact, as we find in the retina of the trout, no material difference from the fovea centralis of the human retina, we have no reason to suppose that the visual powers of the whole of the retina of the trout, should differ in any way from the visual powers possessed by the fovea centralis, the seat of most acute vision both as to colour and light in the human retina. The retinÆ of other fishes which I have examined (none of them were SalmonidÆ) contained only cones; but these cones were some distance from each other.

The layer of pigment epithelium which is present in the human eye, is present also in that of the trout. It occupies the same position between the layer of rods and cones, or cones only, and the choroid. As in the human eye, it adheres sometimes to the choroid and sometimes to the retina, when the retina is removed, though perhaps it most often adheres to the retina.

My space is too limited to enter into any of the theories as to the possibility of the pigment cells playing a part in colour vision. It is quite sufficient to state that they undoubtedly do play some part in our sense of sight, and that they are contained in the eye of the trout.

The retina of a colour-blind person does not show any organic difference from the normal eye, so we cannot say to what cause colour-blindness is due; but so far as our knowledge goes, there is no reason to suppose that the trout is normally colour-blind.

As Michael Foster so ably put it, “No man can tell what are the sensations of his fellow-man,” still less I think can man say what are the sensations of a trout. All we can do with regard to this question of colour vision, is to find out all the facts we can relating to it, and working on comparisons, arrive, not at conclusions, but at probabilities.

The only thing of which I am sure is that we shall find it safe and comparatively easy to imitate flies in colours, but to make a monochromatic imitation of one, which would accurately represent it to a normally monochromatic eye (about which we know nothing), in a medium of which we know very little, is practically impossible.