However, although we are not generally conscious of it, colour plays a much more important part in our lives than this. It can be important for our very survival. For example, we use colours as a diagnostic tool for our health. Not only do we use pale skin to diagnose anaemia and yellow skin as a symptom of liver disease but the bruising from subcutaneous bleeding after a blow can also be observed. Bad teeth can also be diagnosed from their colour.

The use of colour is also important for warning purposes in the case of red traffic signal lights, for example, and the red colour of certain very poisonous toadstools and frogs. We also use colour diagnostically to tell us when our bread is baked, or tell a ripe apple from a unripe one. Colour is very important cosmetically when used both as a sexual signal and as a means of camouflaging bad skin. And colour is very important when used for identification purposes, such as the colour coding of electrical resistors and for product identification, the Kodak yellow film packaging being a typical example. And one could go on, as there are many other important functions of colour for us humans; whole books have, in fact, been written about this subject.

But what about the submarine world of fishes? Have colours any importance for them? Or are the beautiful colours that divers see in fishes and coral reefs only observed by them and have no function at all for fishes? For there certainly aren’t any fish-dentists examining their patients for teeth decay, or fish-bakers making bread.

Several theories have been put forward over the years regarding the wonderful colours of the fishes in and around the reefs.
 It was thought that the colours developed through natural selection in order that males will be attracted to females. However, males and females often appear to be the same with regard to colouration.
 Perhaps, it was supposed, the colours are warnings that certain fish are toxic or otherwise nasty to eat. This is true for example for the Box fish. However, many brightly coloured fish are excellent eating, not only for other fish but also for us humans.
 The gaudy appearance of the box fish reminds predators that poison is secreted through its skin when it is attacked. Predators associate the effects of its venom with the black and yellow combination of warning colours and learn to avoid them.

Konrad Lorentz, the Austrian animal behaviourist, proposed that the fish colours might be acting as identifiers of possible mates. For humans this is no problem, as we only form a single species. For reef fishes, however, this could be a real problem as there are so many different species present in the reefs.
 Or perhaps the colours are just a byproduct of fish-metabolism and have no real significance, neither at the present time nor over evolutionary time. However, from the point of view of evolution theory this seems unlikely.

 Although there might be some truth in all of these theories the current consensus seems to be that fish use their colourings mostly for camouflage purposes. At first sight this would seem to be paradoxical, for the fish appear to us to be clearly observable against the uniform blue-grey background of the water or even against the many bright colours of the coral reefs. The critical words here, though, are “appear to us”, for fish are clearly not camouflaged against human sight. So, we can ask ourselves, are they camouflaged against non-humans, i.e. against predatory fish species? What do predatory fish actually see when they look at their prey? To attempt an answer these questions we must first look at some of the factors involved in the perception of colour.

The perception of colour
Firstly, there must obviously be light present, without light nothing can be seen. Human beings have evolved their colour vision under sunlight. Sunlight consists of electromagnetic radiation ranging from the infra-red (wavlengths 700 nm and above), through the so-called spectral or rainbow colours red, orange, yellow, green, blue, violet, to the ultra-violet (shorter than about 400nm wavelength). Of course sunlight varies greatly during the course of the day, from the “white” light at noon to the reddish morning and evening lights. In order to talk about colour objectively it has thus been necessary to define a standard daylight. An example of such a standard is average noon sunlight in Washington, D.C. All objective comparisons of colour, both above the water and in it, are then made using such standards.
 Secondly, there must be cells in the eye to detect this radiation. Most vertebrates, including humans, use two systems of light-sensitive cells in their eyes. Two or more types of so-called cone cells (three in humans) produce a sensation of colour in abundant light, and a single type of rod cell detects light much more sensitively, but achromatically, under reduced lighting conditions. Thus, as a survival strategy, humans forgo their colour vision when it begins to get dark and switch over to their rod vision.

 The human eye is most sensitive to green (about 550 nm wavelength) in the middle of the spectrum, with the sensitivity falling to zero in the infra-red and ultra-violet i.e. we can detect neither ultra-violet light nor infra-red radiation. To enable us to perceive colour, the three types of cell in the human retina, are sensitive to the blue, green and red spectral regions respectively. The relative amount of different light radiation falling on these three types of cell give rise to the perception of colour. For example, if there is relatively a lot of light of, say, 600nm wavelength and above, and little of the shorter wavelengths, we will perceive an orange/red colour.

 So, when we humans perceive coloured objects we are using a specialised set of light sensitive cells under an illumination which preferably contains all the wavelengths to be found in sunlight at sea-level. Of course, we do observe coloured objects under quite different illuminations such as the strong yellow sodium light of some street lamps. But as we have all experienced, this form of lighting disturbs our normal colour perception, and the colours of a given object seen under such lighting will generally be much different to that seen under normal daylight. To take an example, an object that looks a pure bright blue under ordinary daylight can appear to be quite black under yellow light. The perception of the colour of an object therefore depends critically upon the type of illumination.

Light
 Now, to any diver it is obvious that the light penetrating below the surface of the seas is somewhat different to that of daylight above the surface. Although water is apparently quite transparent it does absorb red light weakly, and has a bluish colour – a white object under the surface looks bluish-green. In 15m of ...