The majority of us take our eyes for granted. We don't normally pay much attention to them unless we need new glasses, or get a speck of dust in one.

Biologists have a very different view. Eyes have been a great biological mystery for 150 years. They gave Charles Darwin fits when he first described the theory of evolution, and today they are in some ways a bigger puzzle than ever.

The trouble is, as Darwin clearly perceived, eyes require three different pieces in order to function. First, you need a small, transparent aperture through which light can enter (call it a pupil); second, you need a way of focusing incoming light (some kind of lens); and third, you must have a retina which can convert the collected light to nerve impulses, so they can be carried to the brain and there interpreted.

Now, according to evolution, any organism only develops things useful to its survival and propagation. Eyes are hugely useful, but only if you have all the pieces in place at once. But how could all three elements possibly develop in parallel, since without the third component any other two are totally useless?

The usual answer is, development came slowly and little by little. Some primitive animals have spots on their exterior slightly sensitive to light. Because that gives a small advantage in hunting food or avoiding danger, such creatures could prosper, and little by little the eye developed an increased sensitivity and focusing power, while at the same time the primitive brain learned how to interpret the images it received.

Put like this, it sounds very reasonable. We would all accept poor or limited vision over none at all. The problems come only when we examine other organisms and compare their eyes with ours. I'm not referring to what we might call "superficial" differences, such as chameleons having two eyes that move and see independently of each other, or spiders having as many as eight eyes, all located on their backs, or some mollusks possessing over 100. It's interesting to speculate on what the world must look like to a spider or a scallop, but once you have a basic eye design the number is simply a mass-production problem.

The difficulties arise when we explore the structure of the eye in different animals. For instance, the squid has eyes as complicated as our own, but the retina lies in front of the optic nerves, rather than behind them. Insects are even more different from us, with compound eyes made up of thousands of small separate light receptors.

Now, according to every biologist, humans, squids, insects, and spiders diverged from a common ancestor a long, long time ago. So long ago, in fact, that until recently it was assumed that eyes must have developed independently, many times over. The argument was, eyes are so valuable that they evolved from scratch, 40 or 50 separate times. That also accounts for the variety of forms we find in different animals: "pinhole camera" eyes, lensed eyes, compound eyes, even a reflector eye that focuses light as a satellite dish focuses radio signals. It would require many generations for all these diverse eye types to be perfected, but that's all right. We have billions of years at our disposal since life appeared on the Earth. Also, a single kind of eye would not suit every organism. People don't need to be able to focus on something a quarter of an inch in front of our eyeball, but to an ant or a bee a quarter of an inch is a long way.

So here was the picture by 1980: evolution of the eye little by little, starting with simple light-sensitive spots; and then evolution proceeding independently, to invent by separate patterns of descent the many kinds of eyes we find in today's animal kingdom.

Then, in 1993, came a major mystery. A gene was identified in insects whose presence was essential to the formation of eyes. It was named, backwards fashion, the "eyeless" gene, since without it, eyes did not grow. Also, if you introduced the gene into a developing insect where it was not normally present, more eyes appeared in odd places - on wings, or legs, or antennas. They didn't provide vision, because they were not "wired up" to the insect brain, but they were sensitive to light and had all the correct working parts.

And now for something even more curious. Genes, almost identical to the insect eyeless gene, were found to exist in other animals; mice had them, squid had them, even humans had them. And if you took the human gene (known as Aniridia) and implanted it in developing insects, lo and behold, those insects formed extra eyes. The same thing happened if you used the mouse or squid gene in the insects. All this suggests there is a single, basic gene, one needed to form eyes in any animal, and one already present so long ago that it appears today scattered across the animal kingdom.

Not all biologists accept this, pointing out that there's a lot more to development of an eye than the presence of the eyeless gene. However, the conclusion that eyes must have been developed independently, scores of times, in response to evolutionary pressure, has become a lot weaker.

The debate goes on: one-time eye development, or many times? Meanwhile, here is another and more disturbing thought. There is also a "wingless" gene in insects, required in order for wings to develop. What do you think would happen if you were to insert copies of this gene into a developing human embryo?

"Borderlands of Science" is syndicated by: