Even if they had gone down the valley, it isn't necessarily obvious that the gill peak, when they eventually climbed it, would have turned out to be higher. Gills are not necessarily better than lungs for water-dwelling animals. No doubt it is convenient to be able to breathe continuously, wherever you are, rather than having to break off what you are doing to go to the surface. But our judgement is coloured by the fact that we take a breath every few seconds and panic at even a brief interruption to our air supply. Having been naturally selected through millions of sea-going generations, sperm whales can submerge for fifty minutes before they have to breathe. Coming to the surface to breathe, for a whale, might feel rather like going off to urinate. Or for a meal. If you start to think of breaths as meals, rather than as a continuously vital necessity, it becomes less obvious that every underwater creature would ideally be better off with gills. There are animals, like humming-birds, that feed more or less continuously. To a humming-bird, which needs to suck {132} nectar every few seconds of its waking life, visiting flowers might feel rather like breathing. Sea-squirts, bag-shaped marine invertebrates remotely related to vertebrates, pump a never-ceasing current of water through their bodies, filtering out tiny particles of food. Such a filter-feeder indulges in nothing corresponding to a meal. A sea-squirt might suffocate with panic at the thought of having to search for the next meal. Sea-squirts might well wonder why so many animals go in for the absurdly inefficient and dangerous habit of searching for meals, instead of sitting back and breathing in food the whole time.
Be this as it may, there is no doubt that whales and dugongs come with their dry-land history written all over them. If they had been deliberately created for the sea, they'd be very different, and a lot more like fish than they are. Animals that have their history written all over them are among the most graphic pieces of evidence we have that living things were not created for their present ways of life but evolved from very different ancestors.
Plaice, sole and flounders have their history written all over them too, to the point of grotesqueness. No sane creator, setting out from scratch to design a flat-fish, would have conceived on his drawing board the absurd distortion of the head needed to bring both eyes round to one side. He'd surely, right from the start, have gone for the skate or ray design, the fish lying on the belly with the eyes symmetrically placed on the top (Figure 4.7). Plaice and sole are all twisted around because of their history; because their ancestors lay down on one side. Skates and rays are gracefully symmetrical because their history happened to be different: when their ancestors settled for sea-floor life they lay on the belly rather than on their side. When I say ‘happened to be’ different, I don't mean that there wasn't a good reason for the difference. Skates and rays are descended from sharks, and sharks are already slightly flattened compared with bony fish which are typically deep-bodied and blade-like. A deep-bodied blade of a fish can't lie on its belly, it has to flop over on its side. When they settled on the bottom, the ancestors of plaice raced up the nearest peak of Mount Improbable, regardless of the fact that a possibly {133}
Figure 4.7 Two ways of being a flat-fish: the skate, Raja batis (top), lies on its belly, while the flounder, Bothus lunatus, lies on its side.
higher peak of the mountain — the skate/ray symmetrical peak — was theirs for the taking, if only they could force their way down a little valley in order to reach the foot of the higher peak. To say it again, going down the slopes of Mount Improbable is not allowed by natural selection, and these fish had no choice but to restore their vision in a makeshift way by twisting one eye round to the other side of the body. The ancestors of skates also rushed up their nearest flat-fish peak, which in their case led them to their symmetrical elegance. Of {134} course when I speak of having no ‘choice’, and ‘rushing’ up mountain peaks, you understand, as usual, that it is not individual fish that are meant. It is evolutionary lineages, and ‘choice’ refers to available alternative routes of evolutionary change.
I've stressed that going downhill is not allowed, but not allowed by whom? And can it never happen? The answer to both questions is about the same as for the case of a river not being ‘allowed’ to run in any direction other than along its established watercourse. Nobody actually orders the water to stay within the banks of a river but, for well-understood reasons, it normally does. Just occasionally, however, it overflows the banks, or even bursts them, and the river may be found to have altered its permanent course as a result.
What might permit an evolving lineage to go into reverse for a brief while and so expose itself to the opportunity to ascend a previously inaccessible peak of Mount Improbable? This is the kind of question that interested the great geneticist Sewall Wright who, by the way, was the first to use a landscape metaphor for evolution, the progenitor of my Mount Improbable. Wright was the American member of the cantankerously warring triumvirate who, in the 1920s and 1930s, founded what we now call neo-Darwinism. (The other two were English, those incomparable but bellicose prodigies R. A. Fisher and J. B. S. Haldane, and it is only fair to add that the cantankerous-ness seems all to have originated from them, not from Wright.) Wright realized that, paradoxically, natural selection can be a force against extreme perfection. This is for precisely the reason that we have just been dealing with. Going down valleys is forbidden by natural selection. A species that is trapped on a small foothill of the mountain cannot escape to higher peaks as long as natural selection pens it in at the top of the foothill. If only natural selection would relax its grip for a short while, the species might edge its way down the foothill far enough to cross a valley to the lower slopes of a high peak. Once there, it is in a position, when natural selection starts biting again, to evolve rapidly up the higher slopes of the mountain. From a global point of view, then, one recipe for improvement in evolution is periods of strong selection interrupted by periods of relaxation. Maybe this kind of relaxation is actually important in real-life evolution. {135} When might ‘relaxation’ occur? One possibility is when there is a vacuum to be filled. In a small way this will happen whenever a population is growing because it is smaller than the area can support. There may be bonanzas of opportunity and relaxation of selection when a virgin continent is first colonized after being cleaned out by a catastrophe. Perhaps after the dinosaurs became extinct, the remaining mammals had such a field day of opportunity that some of their lineages ‘relaxed their guard’, went temporarily downhill, and thereby found higher peaks of Mount Improbable from which they would normally have been debarred.
Another recipe is transfusions of fresh genes from other places. This is the point that I said I'd return to from Chapter 2 on spider webs. In the NetSpinner model of spider webs, there was not just one sexual population of simulated web-spinners but three ‘demes’ evolving in parallel. These were thought of as evolving independently in three different geographical areas. But — here's the point — not completely independently. There is a trickle of genes, meaning that an individual occasionally migrates, from one local population to another. The way I put it was that these migrant genes were a kind of injection of fresh ‘ideas’ from another population: ‘almost as though a successful sub-population sends out genes that "suggest" to a less successful population a better way to solve the problem of building a web’. It is tantamount to being guided up the higher peak of the metaphorical mountain by a smuggled-in map.
We are ready to take up the creationists’ favourite target, and the star stumbling block for would-be believers in evolution, perched precariously on the summit of the most formidable cliff Mount Improbable has to offer: the eye.
Note: After this book had gone to be typeset, J. H. Marden and M. G. Kramer published a fascinating study of stonefiies, which suggests yet another possible route up Mount Improbable towards true flapping flight (Marden, J. H., & Kramer, M. G. (1995) ‘Locomotor performance of insects with rudimentary wings’. Nature, 377, 332—4). Stonefiies are rather primitive flying insects, where primitive means that, although
they are modern living insects, they are thought to resemble ancestors more than other modern insects resemble ancestors. The particular species that Marden and Kramer studied, Allocapnia vivipam, skims across the surface of streams by raising its wings and using them as sails to catch the wind. Sailing velocity is approximately proportional to wing length. Individuals with the smallest wings sail faster than {136} individuals who don't raise their wings at all. Those same smallest wings are roughly the same size as the movable gill plates of early fossil insects. Maybe wingless ancestors lived on water surfaces and raised their gill plates as sails. There would then have been a smooth ramp up Mount Improbable as the gill plates grew to become progressively more effective sails. As for the next step towards flapping flight on this hypothesis, Marden and Kramer have made another relevant observation. A different species of stonefly, Taeniopteryx burksi, also skims along the water surface, but it flaps its wings to do so. Perhaps insects, on their way up to the flying peak of Mount Improbable, passed through a sailing phase like Allocapnia, then a surface flapping phase like Taeniopteryx. It is easy to imagine that light flapping insects buzzing their way over the surface might occasionally have been lifted by gusts of wind. There would then have been a further ramp up the mountain as their flapping wings kept them aloft for progressively longer times. {137}
* * *
>>
CHAPTER 5
ALL ANIMALS HAVE TO DEAL WITH THEIR WORLD, AND the objects in it. They walk on objects, crawl under them, avoid crashing into them, pick them up, eat them, mate with them, run away from them. Back in the geological dawn when evolution was young, animals had to make physical contact with objects before they could tell that those objects were there. What a bonanza of benefit was waiting for the first animal to develop a remote-sensing technology: awareness of an obstacle before hitting it; of a predator before being seized; of food that wasn't already within reach but could be anywhere in the large vicinity. What might this high technology be?
The sun provided not only the energy to drive the chemical cogwheels of life. It also offered the chance of a remote guidance technology. It pummelled every square millimetre of Earths surface with a fusillade of photons: tiny particles travelling in straight lines at the greatest speed the universe allows, criss-crossing and ricocheting through holes and cracks so that no nook escaped, every cranny was sought out. Because photons travel in straight lines and so fast, because they are absorbed by some materials more than others and reflected by some materials more than others, and because they have always been so numerous and so all-pervading, photons provided the {138} opportunity for remote-sensing technologies of enormous accuracy and power. It was necessary only to detect photons and — more difficult — distinguish the directions from which they came. Would the opportunity be taken up? Three billion years later you know the answer, for you can see these words.
Darwin famously used the eye to introduce his discussion on ‘Organs of extreme perfection and complication’:
To suppose that the eye, with all its inimitable contrivances for adjusting the focus to different distances, for admitting different amounts of light, and for the correction of spherical and chromatic aberration, could have been formed by natural selection, seems, I freely confess, absurd in the highest possible degree.
It is possible that Darwin was influenced by his wife Emma's difficulties with this very point. Fifteen years before The Origin of Species he had written a long essay outlining his theory of evolution by natural selection. He wanted Emma to publish it in the event of his death and he let her read it. Her marginalia survive and it is particularly interesting that she picked out his suggestion that the human eye ‘may possibly have been acquired by gradual selection of slight but in each case useful deviations’. Emma's note here reads, ‘A great assumption / E.D.’ Long after The Origin of Species was published Darwin confessed, in a letter to an American colleague: ‘The eye, to this day, gives me a cold shudder, but when I think of the fine known gradations, my reason tells me I ought to conquer the cold shudder.’ Darwin's occasional doubts were presumably similar to those of the physicist whom I quoted at the beginning of Chapter 3. Darwin, however, saw his doubts as a challenge to go on thinking, not a welcome excuse to give up.
When we speak of ‘the’ eye, by the way, we are not doing justice to the problem. It has been authoritatively estimated that eyes have evolved no fewer than forty times, and probably more than sixty times, independently in various parts of the animal kingdom. In some cases these eyes use radically different principles. Nine distinct principles have been recognized among the forty to sixty independently {139} evolved eyes. I'll mention some of the nine basic eye types — which we can think of as nine distinct peaks in different parts of Mount Improbable's massif — as I go on.
How, by the way, do we ever know that something has evolved independently in two different groups of animals? For example, how do we know that bats and birds developed wings independently? Bats are unique among mammals in having true wings. It could theoretically be that ancestral mammals had wings, and all except bats have subsequently lost them. But for that to occur, an unrealistically large number of independent wing losses would be required, and the evidence supports common sense in indicating that this didn't happen. Ancestral mammals used their front limbs not for flying but for walking, as the majority of their descendants still do. It is by means of similar reasoning that people have worked out that eyes have arisen many times independently in the animal kingdom. We can also use other information such as details of how the eyes develop in the embryo. Frogs and squids, for instance, both have good camera-style eyes, but these eyes develop in such different ways in the two embryos that we can be sure they evolved independently. This does not mean that the common ancestor of frogs and squids totally lacked eyes of any kind. I wouldn't be surprised if the common ancestor of all surviving animals, who lived perhaps a billion years ago, possessed eyes. Perhaps it had some sort of rudimentary patch of light-sensitive pigment and could just tell the difference between night and day. But eyes, in the sense of sophisticated image-forming equipment, have evolved many times independently, sometimes converging on similar designs, sometimes coming up with radically different designs. Very recently there has been some exciting new evidence bearing upon this question of the independence of the evolution of eyes in different parts of the animal kingdom. I'll return to it at the end of the chapter.
As I survey the diversity of animal eyes, I'll often mention whereabouts on the slopes of Mount Improbable each type is to be found. Remember, though, that these are all eyes of modern animals, not of true ancestors. It is convenient to think that they might give us some clues about the kinds of eyes that ancestors had. They do at least show that eyes that we think of as lying half-way up Mount Improbable {140} would actually have worked. This really matters for, as I have already remarked, no animal ever made a living by being an intermediate stage on some evolutionary pathway. What we may think of as a way station up the slope towards a more advanced eye may be, for the animal itself, its most vital organ and very probably the ideal eye for its own particular way of life. High-resolution image forming eyes, for instance, are not suitable for very small animals. High-quality eyes have to exceed a certain size — absolute size not size relative to the animal's body — and the larger the better in absolute terms. For a very small animal an absolutely large eye would probably be too costly to make and too heavy and bulky to carry around. A snail would look pretty silly if its eyes had the seeing power of human eyes (Figure 5.1). Snails that grew eyes even slightly larger than the present average might see better than their rivals. But they'd pay the penalty of having to carry a larger burden around, and therefore wouldn't survive so well. The largest eye ever recorded, by the way, is a colossal 37 cm in diameter. The leviathan that could afford to carry such eyes around is a giant squid with 10-metre tentacles.
Accepting the limitations of the metaphor of Mount Improbable, let's go right
down to the bottom of the vision slopes. Here we find
Figure 5.1 Fantasy snail with eyes large enough to see as well as humans can. {141}
eyes so simple that they scarcely deserve to be recognized as eyes at all. It is better to say that the general body surface is slightly sensitive to light. This is true of some single-celled organisms, some jellyfish, starfish, leeches and various other kinds of worms. Such animals are incapable of forming an image, or even of telling the direction from which light comes. All that they can sense (dimly) is the presence of (bright) light, somewhere in the vicinity. Weirdly, there is good evidence of cells that respond to light in the genitals of both male and female butterflies. These are not image-forming eyes but they can tell the difference between light and dark and they may represent the kind of starting point that we are talking about when we speak of the remote evolutionary origins of eyes. Nobody seems to know how the butterflies use them, not even William Eberhard, whose diverting book, Sexual Selection and Animal Genitalia, is my source for this information.
If we think of the plain below Mount Improbable as peopled by ancestral animals that were totally unaffected by light, the non-directional light-sensitive skins of starfish and leeches (and butterfly genitals) are just a little way up the lower slopes, where the mountain path begins. It is not difficult to find the path. Indeed it may be that the ‘plain of total insensitivity to light has always been small. It may be that living cells are more or less bound to be somewhat affected by light — a possibility that makes the butterfly's light-sensitive genitals seem less strange. A light ray consists of a straight stream of photons. When a photon hits a molecule of some coloured substance it may be stopped in its tracks and the molecule changed into a different form of the same molecule. When this happens some energy is released. In green plants and green bacteria, this energy is used to build food molecules, in the set of processes called photosynthesis. In animals the energy may trigger a reaction in a nerve, and this constitutes the first step in the process called seeing, even in animals lacking eyes that we would recognize as eyes. Any of a wide variety of coloured pigments will do, in a rudimentary way. Such pigments abound, for all sorts of purposes other than trapping light. The first faltering steps up the slopes of Mount Improbable would have consisted in the gradual improvement of pigment molecules. There is a shallow, continuous ramp of improvement — easy to climb in small steps. {142}