The really astonishing thing about Kleiber's Law is that it holds good from the smallest bacterium to the largest whale. That's about 20 orders of magnitude. You need to multiply by ten 20 times or add 20 noughts in order to get from the smallest bacterium to the largest mammal, and Kleiber's Law holds right across the board. It works for plants and single-celled organisms too. The diagram shows that the best fit is obtained with three parallel lines. One line is for micro-organisms, a second for cold-blooded large creatures ('large' here means anything heavier than about a millionth of a gram!) and the third is for warm-blooded large creatures (mammals and birds). All three lines have the same slope () but they are of different height: not surprisingly, warm-blooded creatures have a higher metabolic rate, size for size, than cold-blooded creatures.

For years no one could think of a really convincing reason for Kleiber's Law, until a piece of brilliant collaborative work between a physicist, Geoffrey West, and two biologists, James Brown and Brian Enquist. Their derivation of the precise law is a piece of mathematical magic that is hard to translate into words, but it is so ingenious and important that it is worth the effort.

[image]

Tissues have a supply problem.

The complex supply system of a cauliflower.

The theory of West, Enquist and Brown, henceforth WEB, takes off from the fact that the tissues of large organisms have a supply problem. That is what blood systems in animals and vascular tubing in plants are all about: transporting 'stuff' to and from tissues. Small organisms don't face the problem to the same extent. A very small organism has such a large surface area compared to its volume that it can get all the oxygen it needs through its body wall. Even if it is multicellular, none of its cells are very far from the outside body wall. But a large organism has a transport problem because most of its cells are far away from the supplies they need. They need to pipe stuff from place to place. Insects literally pipe air into their tissues in a branching network of tubes called tracheae. We too have richly branched air tubes, but they are confined to special organs, the lungs, which have a correspondingly richly branched blood network to take the oxygen from the lungs to the rest of the body. Fish do a similar thing with gills: area-intensive organs designed to increase the interface between water and blood. The placenta does the same kind of thing for maternal blood and foetal blood. Trees use their richly dividing branches to supply their leaves with water drawn from the ground, and pump sugars back from the leaves to the trunk.

The cauliflower above, freshly bought from a local greengrocer and cut in half, shows what a typical stuff-transporting system looks like. You can see how much effort a cauliflower puts in to provide a supply network for its surface covering of 'flower buds'.4 Now, we could imagine that such supply networks tubes of air, pipes of blood or sugar solution, or whatever they are might compensate perfectly for increased body size. If they did, a typical cell of a modest cauliflower would be exactly as well supplied as a typical cell of a giant redwood, and the metabolic rate of the two cells would be the same. Since the number of cells in an organism is proportional to its ma.s.s, the scatter plot of total metabolic rate against body ma.s.s, with both axes on a log scale, would fall on a line with a slope of 1. Yet what we actually observe is a slope of . Small organisms have a higher metabolic rate than they 'should' have, for their ma.s.s, compared with large organisms. What this means is that the metabolic rate of a cauliflower cell is higher than the metabolic rate of an equivalent cell in a redwood, and the metabolic rate of a mouse is higher than the metabolic rate of a whale.

At first sight, this seems strange. A cell is a cell is a cell, and you might think there is an ideal metabolic rate which would be the same for a cauliflower as for a redwood, a mouse as for a whale. Perhaps there is. But what seems to happen is that the difficulty of delivering water, or blood, or air, or whatever 'stuff' it is, seems to place a limit on achieving that ideal. There has to be a compromise. The WEB theory explains the compromise, and why it ends up delivering exactly a slope of , and it does so in precise, quant.i.tative detail.

The theory consists of two key points. The first is that the branching tree of pipes, which delivers stuff to a given volume of cells, itself occupies some volume, competing for s.p.a.ce with the cells that it is supplying. Towards the tips of the supply network, the pipes occupy a substantial s.p.a.ce in their own right. And if you double the number of cells that need to be supplied, the network volume more than doubles because more pipes are needed to plumb the network into the main system, pipes which themselves occupy s.p.a.ce. If you want to double the number of supplied cells whilst only doubling the s.p.a.ce occupied by the pipes, you need a more spa.r.s.ely distributed plumbing network. The second key point is that, whether you are a mouse or a whale, the most efficient transport system the one that wastes the least energy in moving stuff around is one that takes up a fixed percentage of the volume of your body. That's how the mathematics works out, and it is also an empirically observed fact.5 For example, mammals, whether mice, humans, or whales, have a volume of blood (i.e. the size of the transport system) which occupies between six and seven per cent of their body. For example, mammals, whether mice, humans, or whales, have a volume of blood (i.e. the size of the transport system) which occupies between six and seven per cent of their body.

Taking these two points together, it means that if we wish to double the volume of cells to be supplied, but still keep the most efficient transport system, we need a more spa.r.s.ely distributed supply network. And a more spa.r.s.e network means that less stuff is supplied per cell, meaning that the metabolic rate must go down. But by how much, precisely, must it go down?

WEB calculated the answer to this question. Wonderful to say, the mathematics predicts a straight line with a slope of exactly for the graph of log metabolic rate against log body size! More recent work has built upon their initial theory, but the essential aspects still remain. Kleiber's Law whether in plants, animals, or even at the level of transport within a single cell has finally found its rationale. It can be derived from the physics and geometry of supply networks.

THE REDWOOD'S TALE People argue about the one place in the world that you must visit before you die. My candidate is Muir Woods, just north of the Golden Gate Bridge. Or, if you leave it too late, I can't imagine a better place to be buried (except I doubt that it's allowed, nor should it be). It is a cathedral of greens and browns and stillness, the nave lofted by the world's tallest trees, Sequoia sempervirens Sequoia sempervirens, the Pacific coast redwoods, whose padded bark damps out the echoes that would fill a man-made cathedral. The related species Sequoiadendron giganteum Sequoiadendron giganteum (see plate 47) (see plate 47), found inland on the foothills of the Sierra Nevada range, is typically slightly shorter but more ma.s.sive. The largest single living creature in the world, the General Sherman tree, is a giganteum giganteum over 30 metres in circ.u.mference and over 80 metres tall, with an estimated weight of 1,260 tonnes. Its age is not known for certain but the species is known to survive more than 3,000 years. General Sherman's age could be ascertained exactly to the nearest year if we cut it down a major undertaking, for the bark alone is about a metre thick. over 30 metres in circ.u.mference and over 80 metres tall, with an estimated weight of 1,260 tonnes. Its age is not known for certain but the species is known to survive more than 3,000 years. General Sherman's age could be ascertained exactly to the nearest year if we cut it down a major undertaking, for the bark alone is about a metre thick.6 Let us hope this will never happen, in spite of Ronald Reagan's notorious opinion, when he was Governor of California: 'If you've seen one, you've seen them all.' Let us hope this will never happen, in spite of Ronald Reagan's notorious opinion, when he was Governor of California: 'If you've seen one, you've seen them all.'

How is it that we can know the age of a large tree, even one as old as General Sherman, accurately to the nearest year? We count the rings in its stump. Ring counting, in a more sophisticated form, has given rise to the elegant technique of dendrochronology, by which the redwood's tale archaeologists working on a timescale of centuries can precisely date any wooden artefact.

It has been left to this tale to explain how, throughout our pilgrimage, we have been able to date historical specimens on an absolute timescale. Tree rings are very accurate but only for the nearest reaches of the historical record. Fossils are dated by other methods, mostly involving radioactive decay, and we shall come to them, along with other techniques, in the course of the tale.

Annual rings in a tree result from the unsurprising fact that a tree puts on more growth in some seasons than in others. But, by the same token, whether in summer or winter, trees grow more in a good year than in a poor year. Good years are two a penny, and so are bad years, so one tree ring is no good for identifying a particular year. But a sequence of years has a fingerprint pattern of wide and narrow rings, which labels that sequence in different trees over a wide area. Dendrochronologists compile catalogues of these labelled signature patterns. Then a fragment of wood, perhaps from a Viking longship buried in mud, can be dated by matching its ring pattern against previously collected libraries of signatures.

The same principle is used in dictionaries of melodies. Suppose you have a tune in your head and you can't remember its name. How might you look it up? Various principles are used, of which the simplest is the Parsons code. Turn your tune into a series of ups and downs (the first note is ignored because, obviously, it can be neither up nor down). Here, for instance, is the pattern of a favourite tune, the Londonderry Air, or Air from County Derry, which I have just typed into the Melodyhound website: UUUDUUDDDDDUUUUUDDDUD.

Melodyhound correctly sleuthed my tune (calling it 'Danny Boy' the name by which Americans know it because of some twentiethcentury words that were set to it). At first it seems surprising that a tune should be identified by such a short sequence of symbols, telling only the direction of movement, not the distance, and with no indication of durations of notes. But it really works. For the same kind of reason, a fairly short consecutive pattern of tree rings suffices to identify a particular sequence of yearly growth rings.

In a newly felled tree, the outer ring represents the present. The past can be exactly reckoned by counting inwards. So absolute dates can be put upon ring pattern signatures in recent trees whose date of felling is recorded. By looking for overlaps signature patterns near the core of a young tree that match the pattern in the outside layers of an older tree we can put absolute dates on ring patterns in older trees too. By daisy-chaining the overlaps backwards, it is in principle possible to put absolute dates on very old wood indeed in principle even from the Petrified Forest of Arizona, if only there were a continuous series of petrified intermediates if only! By this technique of overlapping jigsaws, libraries of fingerprint patterns can be built up and consulted to recognise wood that is older than the oldest tree we ever see alive. The changing thickness of tree rings can also, incidentally, be used not just for dating wood but for reconstructing year-to-year climate and ecological patterns dating from long before meteorological records were kept.

Dendrochronology is limited to the relatively recent time domains inhabited by archaeologists. But tree growth is not the only process that spurts and slows on an annual cycle, or on some other regular or even irregular cycle. Any such process can in principle be used for dating, aided by the same ingenious trick of daisy-chaining overlapped patterns. And some of these techniques work over a longer period than dendrochronology itself. Sediments are laid down on the sea bottom at an uneven rate, and in stripes which we can think of as equivalent to tree rings. These stripes can be counted, and signatures recognised, in core samples extracted by deep cylindrical probes.

Another example, which we encountered in the Epilogue to the Elephant Bird's Tale, is palaeomagnetic dating. As we saw there, the Earth's magnetic field reverses from time to time. What had been magnetic north suddenly becomes magnetic south for some thousands of years, then flips again. This has happened 282 times during the last 10 million years. Although I say 'flips' and 'suddenly', it is sudden only by geological standards. Entertaining as it might be if a polar flip today turned every plane and ship around in its tracks, that isn't the way things work. The 'flip' actually takes a few thousand years, and is much more complicated than the flip word suggests. The magnetic North Pole in any case seldom coincides exactly with the true, geographic North Pole (around which the Earth spins). It wanders around the polar region over the years. At present the magnetic North Pole is located near Bathurst Island in northern Canada, about 1,000 miles from the true North Pole. During a 'flip' there is an interregnum of magnetic confusion, with large and complicated variations in field strength and direction, sometimes involving the temporary appearance of more than one magnetic north and more than one magnetic south. Eventually the confusion stabilises again, and when the dust settles it may turn out that the previous magnetic north is now near the true South Pole and vice versa. Stability, with wandering, then resumes for perhaps a million years until the next flip.

A thousand years in geology's sight is but an evening gone. The time spent 'flipping' is negligible compared to the time spent in the rough vicinity of either the true North or the true South Pole. Nature, as we saw earlier, keeps an automatic record of such events. In molten volcanic rock, certain minerals behave like little compa.s.s needles. When the molten rock solidifies, these mineral needles const.i.tute a 'frozen' record of the Earth's magnetic field at the moment of solidification (by a rather different process, palaeomagnetism can be observed in sedimentary rock, too). After a 'flip', the miniature compa.s.s needles in the rocks point in the opposite direction compared with before the flip. It's like tree rings all over again, except that the stripes are not a year apart but of the order of a million years. Once again, patterns of stripes can be matched up with other patterns, and a continuous chronology of magnetic flips can be daisy-chained together. Absolute dates can't be calculated by counting stripes because, unlike tree rings, the stripes represent unequal durations. Nevertheless, the same signature pattern of stripes can be picked up in different places. This means that if some other method of absolute dating (see pages 531536) is available for one of the places, magnetic stripe patterns, like the Parsons code for a melody, can be used to recognise the same time zone in other places. As with tree rings and other dating methods, the full picture is built up from fragments gathered in different places.

Tree rings are good for dating recent relics to the nearest year. For older dates, with inevitably less fine pinpointing, we exploit the wellunderstood physics of radioactive decay. To explain this, we begin with a digression.

All matter is made of atoms atoms. There are more than 100 types of atoms, corresponding to the same number of elements elements. Examples of elements are iron, oxygen, calcium, chlorine, carbon, sodium and hydrogen. Most matter consists not of pure elements but of compounds: compounds: two or more atoms of various elements bonded together, as in calcium carbonate, sodium chloride, carbon monoxide. The binding of atoms into compounds is mediated by two or more atoms of various elements bonded together, as in calcium carbonate, sodium chloride, carbon monoxide. The binding of atoms into compounds is mediated by electrons electrons, which are tiny particles...o...b..ting (a metaphor to help us understand their real behaviour, which is much stranger) the central nucleus nucleus of each atom. A nucleus is huge compared to an electron but tiny compared to an electron's...o...b..t. Your hand, consisting mostly of empty s.p.a.ce, meets hard resistance when it strikes a block of iron, also consisting mostly of empty s.p.a.ce, because forces a.s.sociated with the atoms in the two solids interact in such a way as to prevent them pa.s.sing through each other. Consequently iron and stone seem solid to us because our brains most usefully serve us by constructing an illusion of solidity. of each atom. A nucleus is huge compared to an electron but tiny compared to an electron's...o...b..t. Your hand, consisting mostly of empty s.p.a.ce, meets hard resistance when it strikes a block of iron, also consisting mostly of empty s.p.a.ce, because forces a.s.sociated with the atoms in the two solids interact in such a way as to prevent them pa.s.sing through each other. Consequently iron and stone seem solid to us because our brains most usefully serve us by constructing an illusion of solidity.

It has long been understood that a compound can be separated into its component parts, and recombined to make the same or a different compound with the emission or consumption of energy. Such easy-come easy-go interactions between atoms const.i.tute chemistry. But, until the twentieth century, the atom itself was thought to be inviolate. It was the smallest possible particle of an element. A gold atom was a tiny speck of gold, qualitatively different from a copper atom, which was a minimal particle of copper. The modern view is more elegant. Gold atoms, copper atoms, hydrogen atoms and so on are just different arrangements of the same fundamental particles, just as horse genes, lettuce genes, human genes and bacterial genes have no essential 'flavour' of horse, lettuce, human or bacteria but are just different combinations of the same four DNA letters. In the same way as chemical compounds have long been understood to be arrangements put together from a finite repertoire of 100 or so atoms, so each atomic nucleus turns out to be an arrangement of two fundamental particles, the protons protons and and neutrons neutrons. A gold nucleus is not 'made of gold'. Like all other nuclei, it is made of protons and neutrons. An iron nucleus differs from a gold nucleus, not because it is made of a qualitatively different kind of stuff called iron, but simply because it contains 26 protons (and 30 neutrons), instead of gold's 79 protons (and 118 neutrons). At the level of a single atom there is no 'stuff' that has the properties of gold or iron. There are just different combinations of protons, neutrons and electrons. Physicists go on to tell us that protons and neutrons are themselves composed of yet more fundamental particles, the quarks, but we shall not follow them to such depths.

Protons and neutrons are almost the same size as each other, and much larger than electrons. Unlike a neutron, which is electrically neutral, each proton has one unit of electric charge (arbitrarily designated positive), which exactly balances the negative charge of one electron 'in orbit' around the nucleus. A proton can be transformed into a neutron if it absorbs an electron, whose negative charge neutralises the proton's positive one. Conversely, a neutron can transform itself into a proton by expelling a unit of negative charge one electron. Such transformations are examples of nuclear reactions, as opposed to chemical reactions. Chemical reactions leave the nucleus intact. Nuclear reactions change it. They usually involve much larger exchanges of energy than chemical reactions, which is why nuclear weapons are so much more devastating, weight for weight, than conventional (i.e. chemical) explosives. The alchemists' quest to change one metallic element into another failed only because they tried to do it by chemical rather than nuclear means.

Each element has a characteristic number of protons in its atomic nucleus, and the same number of electrons in 'orbit' around the nucleus: one for hydrogen, two for helium, six for carbon, 11 for sodium, 26 for iron, 82 for lead, 92 for uranium. It is this number, the so-called atomic number, which (acting via the electrons) largely determines an element's chemical behaviour. The neutrons have little effect on an element's chemical properties, but they do affect its ma.s.s and they do affect its nuclear reactions.

A nucleus typically has roughly the same number of neutrons as protons, or a few more. Unlike the proton count, which is fixed for any given element, the neutron count varies. Normal carbon has six protons and six neutrons, giving a total 'ma.s.s number' of 12 (since the ma.s.s of electrons is negligible and a neutron weighs approximately the same as a proton). It is therefore called carbon 12. Carbon 13 has one extra neutron, and carbon 14 two extra neutrons, but they all have six protons. Such different 'versions' of an element are called 'isotopes'. The reason all three of these isotopes have the same name, carbon, is that they have the same atomic number, 6, and therefore all have the same chemical properties. If nuclear reactions had been discovered before chemical reactions, perhaps the isotopes would have been given different names. In a few cases, isotopes are different enough to earn different names. Normal hydrogen has no neutrons. Hydrogen 2 (one proton and one neutron) is called deuterium. Hydrogen 3 (one proton and two neutrons) is called tritium. All behave chemically as hydrogen. For example, deuterium combines with oxygen to make a form of water called heavy water, famous for its use in the manufacture of hydrogen bombs.

Isotopes, then, differ only in the number of neutrons they have, along with the fixed number of protons that characterise the element. Among the isotopes of an element, some may have an unstable nucleus, meaning it has an occasional tendency to change at an unpredictable instant, though with predictable probability, into a different kind of nucleus. Other isotopes are stable: their probability of changing is zero. Another word for unstable is radioactive. Lead has four stable isotopes and 25 known unstable ones. All isotopes of the very heavy metal uranium are unstable all are radioactive. Radioactivity is the key to the absolute dating of rocks and their fossils: hence the need for this digression to explain it.

What actually happens when an unstable, radioactive element changes into a different element? There are various ways in which this can happen, but the two best known are called alpha decay and beta decay. In alpha decay the parent nucleus loses an 'alpha particle', which is a pellet consisting of two protons and two neutrons stuck together. The ma.s.s number therefore drops by four units, but the atomic number drops by only two units (corresponding to the two protons lost). So the element changes, chemically speaking, into whichever element has two fewer protons. Uranium 238 (with 92 protons and 146 neutrons) decays into thorium 234 (with 90 protons and 144 neutrons).

Beta decay is different. One neutron in the parent nucleus turns into a proton, and it does so by ejecting a beta particle, which is a single unit of negative charge or one electron. The ma.s.s number of the nucleus remains the same because the total number of protons plus neutrons remains the same, and electrons are too small to bother with. But the atomic number increases by one because there is now one more proton than before. Sodium 24 transforms itself, by beta decay, into magnesium 24. The ma.s.s number has remained the same, 24. The atomic number has increased from 11, which is uniquely diagnostic of sodium, to 12, which is uniquely diagnostic of magnesium.

A third kind of transformation is neutron-proton replacement. A stray neutron hits a nucleus and knocks one proton out of the nucleus, taking its place. So, as in beta decay, there is no change in the ma.s.s number. But this time the atomic number has decreased by one because of the loss of one proton. Remember that the atomic number is simply the number of protons in the nucleus. A fourth way in which one element can turn into another, which has the same effect on atomic number and ma.s.s number, is electron capture. This is a kind of reversal of beta decay. Whereas in beta decay a neutron turns into a proton and expels an electron, electron capture transforms a proton into a neutron by neutralising its charge. So the atomic number drops by one, while the ma.s.s number remains the same. Pota.s.sium 40 (atomic number 19) decays to argon 40 (atomic number 18) by this means. And there are various other ways in which nuclei can be radioactively transformed into other nuclei.

One of the cardinal principles of quantum mechanics is that it is impossible to predict exactly when a particular nucleus of an unstable element will decay. But we can measure the statistical likelihood that it will happen. This measured likelihood turns out to be utterly characteristic of a given isotope. The preferred measure is the half-life. To measure the half-life of a radioactive isotope, take a lump of the stuff and count how long it takes for exactly one half of it to decay into something else. The half-life of strontium 90 is 28 years. If you have 100 grams of strontium 90, after 28 years you'll have only 50 grams left. The rest will have turned into yttrium 90 (as it happens, which in turn changes into zirconium 90). Does this mean that after another 28 years you'll have no strontium left? Emphatically no. You'll have 25 grams left. After another 28 years the amount of strontium will have halved again, to 12.5 grams. Theoretically, it never reaches zero but only approaches it by successively halved steps. This is the reason we have to talk about the half-life rather than the 'life' of a radioactive isotope.

The half-life of carbon 15 is 2.4 seconds. After 2.4 seconds you'll be left with half of your original sample. After another 2.4 seconds you'll have only a quarter of your original sample. After another 2.4 seconds you are down to an eighth, and so on. The half-life of uranium 238 is nearly 4.5 billion years. This is approximately the age of the solar system. So, of all the uranium 238 that was present on Earth when it first formed, about half now remains. It is a wonderful and very useful fact about radioactivity that half-lives of different elements span such a colossal range, from fractions of seconds to billions of years.

We are approaching the point of this whole digression. The fact that each radioactive isotope has a particular half-life offers an opportunity to date rocks. Volcanic rocks often contain radioactive isotopes, such as pota.s.sium 40. Pota.s.sium 40 decays to argon 40 with a half-life of 1.3 billion years. Here, potentially, is an accurate clock. But it's no use just measuring the amount of pota.s.sium 40 in a rock. You don't know how much there was when it started! What you need is the ratio of pota.s.sium 40 to argon 40. Fortunately, when pota.s.sium 40 in a rock crystal decays, the argon 40 (a gas) remains trapped in the crystal. If there are equal amounts of pota.s.sium 40 and argon 40 in the substance of the crystal, you know that half the original pota.s.sium 40 has decayed. It is therefore 1.3 billion years since the crystal was formed. If there's, say, three times as much argon 40 as pota.s.sium 40, only one quarter (half of a half) of the original pota.s.sium 40 remains, so the age of the crystal is two half lives or 2.6 billion years.

The moment of crystallisation, which in the case of volcanic rocks is the moment when the molten lava solidified, is the moment when the clock was zeroed. Thereafter, the parent isotope steadily decays and the daughter isotope remains trapped in the crystal. All you have to do is measure the ratio of the two amounts, look up the half-life of the parent isotope in a physics book, and it is easy to calculate the age of the crystal. As I said earlier, fossils are usually found in sedimentary rocks, while dateable crystals are usually in volcanic rocks, so fossils themselves have to be dated indirectly by looking at volcanic rocks that sandwich their strata.

A complication is that often the first product of the decay is itself another unstable isotope. Argon 40, the first product of decay of pota.s.sium 40, happens to be stable. But when uranium 238 decays it pa.s.ses through a cascade of no fewer than 14 unstable intermediate stages, including nine alpha decays and seven beta decays, before it finally comes to rest as the stable isotope lead 206. By far the longest half-life of the cascade (4.5 billion years) belongs to the first transition, from uranium 238 to thorium 234. An intermediate step in the cascade, from bis.m.u.th 214 to thallium 210, has a half-life of only 20 minutes, and even that is not the fastest (i.e. most probable). The later transitions take negligible time compared to the first, so the observed ratio of uranium 238 to the finally stable lead 206 can be set against a half-life of 4.5 billion years to calculate the age of a particular rock.

The uranium/lead method and the pota.s.sium/argon method, with their half-lives measured in billions of years, are useful for dating fossils of great age. But they are too coa.r.s.e for dating younger rocks. For these, we need isotopes with shorter half-lives. Fortunately a range of clocks is available with a wide selection of isotopic half-lives.You choose your half-life to give best resolution for the rocks with which you are working. Better yet, the different clocks can be used as checks on each other.

The fastest radioactive clock in common use is the carbon 14 clock, and this brings us full circle to the teller of this tale, for wood is one of the main materials subjected to carbon 14 dating by archaeologists. Carbon 14 decays to nitrogen 14 with a half-life of 5,730 years. The carbon 14 clock is unusual in that it is used to date the actual dead tissues themselves, not volcanic rocks sandwiching them. Carbon 14 dating is so important for relatively recent history much younger than most fossils, and spanning the range of history normally called archaeology that it deserves special treatment.

Most of the carbon in the world consists of the stable isotope carbon 12. About one million-millionth part of the world's carbon consists of the unstable isotope carbon 14. With a half-life measured in only thousands of years, all the carbon 14 on Earth would long since have decayed to nitrogen 14 if it were not being renewed. Fortunately, a few atoms of nitrogen 14, the most abundant gas in the atmosphere, are continually being transformed, by bombardment of cosmic rays, into carbon 14. The rate of creation of carbon 14 is approximately constant. Most of the carbon in the atmosphere, whether carbon 14 or the more usual carbon 12, is chemically combined with oxygen in the form of carbon dioxide. This gas is sucked in by plants, and the carbon atoms used to build their tissues. To plants, carbon 14 and carbon 12 look the same (plants are only 'interested' in chemistry, not the nuclear properties of atoms). The two varieties of carbon dioxide are imbibed approximately in proportion to their availability. Plants are eaten by animals, which may be eaten by yet other animals, so carbon 14 is dispersed in a known proportion relative to carbon 12 throughout the food chain during a time which is short compared to the half-life of carbon 14. The two isotopes exist in all living tissues in approximately the same proportion as in the atmosphere, one part in a million million. To be sure, they occasionally decay to nitrogen 14 atoms. But this constant rate is offset by their continuous exchange, via the links of the food chain, with the ever-renewed carbon dioxide of the atmosphere.

All this changes at the moment of death. A dead predator is cut off from the food chain. A dead plant no longer takes in fresh supplies of carbon dioxide from the atmosphere. A dead herbivore no longer eats fresh plants. The carbon 14 in a dead animal or plant continues to decay to nitrogen 14. But it is not replenished by fresh supplies from the atmosphere. So the ratio of carbon 14 to carbon 12 in the dead tissues starts to drop. And it drops with a half-life of 5,730 years. The bottom line is that we can tell when an animal or plant died by measuring the ratio of carbon 14 to carbon 12. This is how it was proved that the Turin Shroud cannot have belonged to Jesus its date is medieval. Carbon 14 dating is a wonderful tool for dating the relics of relatively recent history. It is of no use for more ancient dating because almost all the carbon 14 has decayed to nitrogen 14, and the residue is too tiny to measure accurately.

There are other methods of absolute dating, and new ones are being invented all the time. The beauty of having so many methods is partly that they collectively span such an enormous range of timescales. It is also that they can be used as a cross-check on each other. It is extremely hard to argue against datings that are corroborated across different methods.

1 The reason for this little hedge will emerge when we get to Canterbury. The reason for this little hedge will emerge when we get to Canterbury.

2 I would have included a tale about this if I had not already done it in two chapters of I would have included a tale about this if I had not already done it in two chapters of Climbing Mount Improbable Climbing Mount Improbable, 'Pollen Grains and Magic Bullets', and 'A Garden Inclosed'.

3 Apart from the rather insignificant 13 species of single-celled glaucophytes, which seem to be the outgroup. Apart from the rather insignificant 13 species of single-celled glaucophytes, which seem to be the outgroup.

4 Buds that have been grotesquely modified in this case by artificial selection under domestication, but the principle still stands. Buds that have been grotesquely modified in this case by artificial selection under domestication, but the principle still stands.

5 The actual percentage might differ a little depending on, say, whether you are warm-blooded or cold-blooded. The actual percentage might differ a little depending on, say, whether you are warm-blooded or cold-blooded.

6 Actually we don't need to cut it down. A core sample would be good enough. Actually we don't need to cut it down. A core sample would be good enough.

Rendezvous 37 UNCERTAIN.

The Microbe is so very smallYou cannot make him out at all,But many sanguine people hopeTo see him through a microscope.His jointed tongue that lies beneathA hundred curious rows of teeth;His seven tufted tails with lotsOf lovely pink and purple spots,On each of which a pattern stands,Composed of forty separate bands;His eyebrows of a tender green;All these have never yet been seen But Scientists, who ought to know,a.s.sure us that they must be so ...Oh! let us never, never doubtWhat n.o.body is sure about.HILAIRE BELLOC (18701953) (18701953)From More Beasts for Worse Children (1897) More Beasts for Worse Children (1897) Hilaire Belloc was a brilliant versifier but a prejudiced man. If there is an element of anti-scientific prejudice above, let us not play up to it. There is much that we are unsure about in science. Where science scores over alternative world views is that we know our uncertainty, we can often measure its magnitude, and we work optimistically to reduce it.

At Rendezvous 37 Rendezvous 37 we enter a world of microbes and also a realm of uncertainty: uncertainty not so much about the microbes themselves as about the order in which we are to greet them. I thought of making a guess and sticking to it, but that would be unfair on the other rendezvous points, about which we can be at least somewhat more certain. If this book's publication were delayed a year or two, the chances of resolution would be good. But for now, let us treat Belloc's verse as a Cautionary Tale for Scientists. We know whom we are to meet at the next one or two or three rendezvous points, but we don't know in what order, and we don't know how many separate rendezvous points there are. we enter a world of microbes and also a realm of uncertainty: uncertainty not so much about the microbes themselves as about the order in which we are to greet them. I thought of making a guess and sticking to it, but that would be unfair on the other rendezvous points, about which we can be at least somewhat more certain. If this book's publication were delayed a year or two, the chances of resolution would be good. But for now, let us treat Belloc's verse as a Cautionary Tale for Scientists. We know whom we are to meet at the next one or two or three rendezvous points, but we don't know in what order, and we don't know how many separate rendezvous points there are.

[image]

Remaining eukaryotes join. The high-level phylogeny of the remaining 50,000 or so described species of eukaryote is currently unresolved (see text). Faded lines indicate the current high level of uncertainty. The chromalveolate branch is often subdivided into the chromista (heterokonts) and the alveolates. The high-level phylogeny of the remaining 50,000 or so described species of eukaryote is currently unresolved (see text). Faded lines indicate the current high level of uncertainty. The chromalveolate branch is often subdivided into the chromista (heterokonts) and the alveolates.

Images, left to right: Giardia lamblia; Euglena acus; Giardia lamblia; Euglena acus; foraminiferan ( foraminiferan (Globigerina sp.); leather kelp ( sp.); leather kelp (Ecklonia radiata).

This uncertainty affects all the 'eukaryotes' who are yet to join the pilgrimage. This important word will be explained in the Great Historic Rendezvous. For the moment, know simply that one of the most momentous events in the history of life was the formation of the eukaryotic cell. Eukaryotic cells are the large and complex cells, with walled nuclei and mitochondria, that make up the bodies of all animals, plants and, indeed, all pilgrims who have so far joined us. That is, all living creatures except the true bacteria and the archaea, which used to be called bacteria. These 'prokaryotes' will const.i.tute the final two rendezvous points, and, as it happens, we are more certain about them. I shall arbitrarily number these final two 38 and 39. This means that the remainder of the eukaryotes all join us together at rendezvous 37 rendezvous 37, which is one of the possible theories at present. But please bear in mind that this is a toss-up: our final rendezvous, with the true bacteria, could be anything from 39 to 42.

Part of the problem is rooting. We met this in the Gibbon's Tale. A star diagram such as the one on the following page is compatible with many different evolutionary trees, and that means many different ways of organizing our rendezvous.

Before we move on to the main point, notice with proper humility, the tiny line labelled 'animals'. If you can't find them, look at the branch labelled 'opisthokonts' on the bottom left, where you'll find us as the sister group to the choanoflagellates. That is where you and I belong, together with the entire populace of pilgrims who joined us up to and including Rendezvous 31 Rendezvous 31.

Clearly there are many places where we could sling the root. The fact that the two most strongly supported hypotheses (indicated by the dotted arrows) are at two such distantly separated extremes contributed to the sapping of my confidence. But it gets worse. The positioning of the root is only the first of our problems. The second problem is that five of the lines meet at a single point in the middle. This doesn't mean that anybody thinks all those five groups burst forth from a single ancestor at the same moment and are all equally close cousins to each other. All it means is yet more uncertainty. We don't know which of the five are closer cousins to each other, so, rather than commit ourselves to what may be an error, justly to be lampooned by a latter-day Belloc, we draw them all as radiating out from a single point. The point where the five lines meet should eventually be resolved into a series of forking lines. Each one of those lines is potentially a place where we could sling our root.

[image]

Notice, with proper humility, where you and I belong. Unrooted phylogram or star diagram of all life, based on the consensus of currently available molecular and other studies. Adapted from Baldauf [ Unrooted phylogram or star diagram of all life, based on the consensus of currently available molecular and other studies. Adapted from Baldauf [13].

By now it will be clear why I backed away from committing myself to the details of the next few rendezvous points. Actually, if you look at the diagram, you'll notice that I was even a bit rash committing myself to Rendezvous 36 Rendezvous 36 as the place where the plants join us. The line of the plants is one of the five that radiates out from the centre of the star. Since decisions hereabouts are still so arbitrary, I decided to treat the plants as though they had a separate rendezvous with us, but only because they are such a huge and important group that they seemed to deserve a separate pilgrim band of their own. What I did, in effect, was draw out a line from the middle of the star diagram. We could make a similarly arbitrary decision over how to resolve the remaining trichotomy, but my courage finally deserts me. I shall leave it buried within the uncertainty of as the place where the plants join us. The line of the plants is one of the five that radiates out from the centre of the star. Since decisions hereabouts are still so arbitrary, I decided to treat the plants as though they had a separate rendezvous with us, but only because they are such a huge and important group that they seemed to deserve a separate pilgrim band of their own. What I did, in effect, was draw out a line from the middle of the star diagram. We could make a similarly arbitrary decision over how to resolve the remaining trichotomy, but my courage finally deserts me. I shall leave it buried within the uncertainty of Rendezvous 37 Rendezvous 37, the blind-date rendezvous.

Instead of committing myself to the order in which they join us, I shall simply go through the remaining groups of eukaryotes, briefly describing them. The Rhizaria include various groups of single-celled eukaryotes, some green and photosynthesising, some not. Among them are the foraminiferans and the radiolarians, notable for their beauty, never better captured than in the drawings of Ernst Haeckel, the eminent German zoologist who seems to keep cropping up through this book. The alveolates include some further beautiful creatures, including the ciliates and the dinoflagellates. Among the ciliates, or so it would seem, is Mixotricha paradoxa Mixotricha paradoxa, whose tale we shall soon hear. The 'so it would seem' and the 'paradoxa' form the substance of the tale, whose thunder I shall not steal here.

The heterokonts are another mixed group. They include some further beautiful unicellular creatures, such as the diatoms, again memorably ill.u.s.trated by Haeckel. But this group has also independently discovered multicellularity, in the form of the brown algae. These are the largest and most prominent of all seaweeds, with giant kelps reaching 100 metres in length. The brown algae include the wracks of the genus Fucus Fucus, the various species of which segregate themselves in strata up the beach, each being best suited to a particular zone of the tide cycle. Fucus Fucus might well be the genus upon whom the leafy sea dragon (see its tale) is modelled. might well be the genus upon whom the leafy sea dragon (see its tale) is modelled.

The discicristates include photosynthetic flagellates, such as the green Euglena Euglena, and parasitic ones, such as Trypanosoma Trypanosoma, which causes sleeping sickness. There are also the acrasid slime moulds, which are not closely related to the dictyostelid slime moulds whom we met at Rendezvous 35 Rendezvous 35. As so often in this long pilgrimage, we marvel at the capacity of life to reinvent similar body forms for similar ways of life. 'Slime moulds' pop up in two or even three different pilgrim bands; so do 'flagellates', so do 'amoebas'. Probably we should think of 'the amoeba' as a way of life, like 'the tree'. 'Trees', meaning very large plants stiffened with wood, pop up in many separate plant families. It looks as though the same is true of 'amoebas' and 'flagellates'. It is certainly true of multicellularity, which has arisen in animals, fungi, plants, brown algae and various other places, such as slime moulds.

The last major group of our unresolvable star consists of the excavates. These are single-celled creatures that would once have been called flagellates and united with Trypanosoma Trypanosoma, the sleeping sickness organism. Now separated off, the excavates include the nasty gut parasite Giardia Giardia, the nasty s.e.xually transmitted v.a.g.i.n.al parasite Trich.o.m.onas Trich.o.m.onas, and various fascinatingly complicated single-celled creatures found only in the guts of termites. And that is the cue for a tale.

THE MIXOTRICH'S TALE Mixotricha paradoxa means 'unexpected combination of hairs', and we shall see why in a moment. It is a micro-organism that lives in the gut of an Australian termite, 'Darwin's termite', Mastotermes darwiniensis Mastotermes darwiniensis. Pleasingly, though not necessarily for the human inhabitants, one of the main places where it flourishes is the town of Darwin in northern Australia.

Termites bestride the tropics like a distributed colossus. In tropical savannahs and forests, they reach population densities of 10,000 per square metre, and are estimated to consume up to a third of the total annual production of dead wood, leaves and gra.s.s. Their bioma.s.s per unit area is double that of migrating herds of wildebeest on the Serengeti and Masai Mara, but is spread across the entire tropics.

If you ask the source of the termites' alarming success, it is twofold. First, they can eat wood, which includes cellulose, lignin and other matter that animal guts normally can't digest. I'll return to this. Second, they are highly social and gain great economies from division of labour among specialists. A termite mound has many of the attributes of a single large and voracious organism, with its own anatomy, its own physiology and its own mud-fashioned organs, including an ingenious ventilation and cooling system. The mound itself stays in one place, but it has a myriad mouths and six myriad legs, and these range over a foraging area the size of a football pitch.

Termites' legendary feats of co-operation are possible, in a Darwinian world, only because the majority of individuals are sterile but closely related to a minority who are very fertile indeed. Sterile workers act like parents towards their younger siblings, thereby freeing the queen to become a specialised egg factory, and a grotesquely efficient one at that. Genes for worker behaviour are pa.s.sed to future generations via the minority of the workers' siblings who are destined to reproduce (helped by the majority of their siblings who are destined to be sterile). You will appreciate that the system works only because it is a strictly non-genetic decision whether a young termite shall become a worker or a reproducer. All young termites have a genetic ticket to enter an environmental lottery which decides whether they become reproductives or workers. If there were genes for being unconditionally sterile, they obviously could not be pa.s.sed on. Instead, they are conditionally conditionally switched-on genes. They are pa.s.sed on when they find themselves in queens or kings because copies of the very same genes cause workers to labour for that end and forgo reproduction themselves. switched-on genes. They are pa.s.sed on when they find themselves in queens or kings because copies of the very same genes cause workers to labour for that end and forgo reproduction themselves.

The a.n.a.logy of insect colony to human body is often made, and it is not a bad one. The majority of our cells subjugate their individuality, devoting themselves to a.s.sisting the reproduction of the minority that are capable of it: 'germ-line' cells in the testes or ovaries, whose genes are destined to travel, via sperms or eggs, into the distant future. But genetic relatedness is not the only basis for subjugation of individuality in fruitful division of labour. Any sort of mutual a.s.sistance, where each side corrects a deficiency in the other, can be favoured by natural selection on both. To see an extreme example, we dive inside the gut of an individual termite, that seething and, as I a.s.sume, noisome chemostat which is the world of the mixotrich.

Termites, as we have seen, enjoy an additional advantage over bees, wasps and ants: their prodigious feats of digestion. There is almost nothing that termites can't eat, from houses to billiard b.a.l.l.s to priceless First Folios. Wood is potentially a rich food source but it is denied to almost all animals because cellulose and lignin are so indigestible. Termites and certain c.o.c.kroaches are the outstanding exception. Termites are, indeed, related to c.o.c.kroaches, and Darwin's termite, like other so-called 'lower' termites, is a sort of living fossil. One could imagine it halfway between c.o.c.kroaches and advanced termites.

In order to digest cellulose, you need enzymes called cellulases. Most animals can't make cellulases, but some micro-organisms can. As Taq's Tale will explain, bacteria and archaea are biochemically more versatile than the rest of the living kingdoms put together. Animals and plants perform a fraction of the biochemical mix of tricks available to bacteria. For digesting cellulose, herbivorous mammals all rely upon microbes in their guts. Over evolutionary time, they have entered into a partnership in which they make use of chemicals such as acetic acid which, to the microbes, are waste products. The microbes themselves gain a safe haven with plenty of raw materials for their own biochemistries, preprocessed and ready-chopped into small, manageable pieces. All herbivorous mammals have bacteria in the lower gut, which the food reaches after the mammal's own digestive juices have had a go at it. Sloths, kangaroos, colobus monkeys and especially cud-chewing ruminants have independently evolved the trick of also keeping bacteria in the upper portion of the gut, which precedes the mammal's own main digestive efforts.

Unlike mammals, termites are capable of manufacturing their own cellulases, at least in the case of the so-called 'advanced' termites. But up to one-third of the net weight of a more primitive (i.e. more c.o.c.kroach-like) termite, such as Darwin's termite, consists of its rich gut fauna of microbes, including eukaryotic protozoa as well as bacteria. The termites locate and chew the wood into small, manageable chips. The microbes live on the wood chips, digesting them with enzymes unavailable to the termites' own biochemical toolkit. Or you could say the microbes have become tools in the termites' toolkit. As with the cattle, it is the waste products of the microbes that the termites live on. I suppose we could say that Darwin's termite and the other primitive termites farm microorganisms in their guts.1 And this brings us, eventually, to the mixotrich, whose tale this is. And this brings us, eventually, to the mixotrich, whose tale this is.

Mixotricha paradoxa is not a bacterium. Like many of the microbes in termite guts, it is a large protozoan, half a millimetre long or more, and large enough to contain hundreds of thousands of bacteria inside itself as we shall see. It lives nowhere except in the gut of Darwin's termite, where it is a member of the mixed community of microbes that thrive on the wood chips milled by the termite's jaws. Micro-organisms populate the termite's gut as richly as the termites themselves populate the mound, and as termite mounds populate the savannah. If the mound is a town of termites, each termite gut is a town of micro-organisms. We have here a two-level community. But and now we come to the crux of the tale there is a third level, and the details are utterly remarkable. is not a bacterium. Like many of the microbes in termite guts, it is a large protozoan, half a millimetre long or more, and large enough to contain hundreds of thousands of bacteria inside itself as we shall see. It lives nowhere except in the gut of Darwin's termite, where it is a member of the mixed community of microbes that thrive on the wood chips milled by the termite's jaws. Micro-organisms populate the termite's gut as richly as the termites themselves populate the mound, and as termite mounds populate the savannah. If the mound is a town of termites, each termite gut is a town of micro-organisms. We have here a two-level community. But and now we come to the crux of the tale there is a third level, and the details are utterly remarkable. Mixotricha Mixotricha itself is a town. The full story was revealed by the work of L. R. Cleveland and A. V. Grim-stone, but it is especially the American biologist Lynn Margulis who has drawn our attention to itself is a town. The full story was revealed by the work of L. R. Cleveland and A. V. Grim-stone, but it is especially the American biologist Lynn Margulis who has drawn our attention to Mixotricha Mixotricha's significance for evolution.

When J. L. Sutherland first examined Mixotricha Mixotricha in the early 1930s she saw two kinds of 'hairs' waving on its surface. It was almost completely carpeted by thousands of tiny hairs, beating to and fro. She also saw a few very long, thin, whip-like structures at the front end. Both seemed familiar to her, the small ones as 'cilia', the large ones as 'flagella'. Cilia are common in animal cells, for instance in our nasal pa.s.sages, and they cover the surface of those protozoans called, not surprisingly, ciliates. Another traditionally recognised group of protozoans, the flagellates, have much longer, whip-like 'flagella' (singular 'flagellum' and, unlike cilia, they often are). Cilia and flagella share an identical ultrastructure. Both are like multi-stranded cables, and the strands have exactly the same signature pattern: nine pairs in a ring surrounding one central pair. in the early 1930s she saw two kinds of 'hairs' waving on its surface. It was almost completely carpeted by thousands of tiny hairs, beating to and fro. She also saw a few very long, thin, whip-like structures at the front end. Both seemed familiar to her, the small ones as 'cilia', the large ones as 'flagella'. Cilia are common in animal cells, for instance in our nasal pa.s.sages, and they cover the surface of those protozoans called, not surprisingly, ciliates. Another traditionally recognised group of protozoans, the flagellates, have much longer, whip-like 'flagella' (singular 'flagellum' and, unlike cilia, they often are). Cilia and flagella share an identical ultrastructure. Both are like multi-stranded cables, and the strands have exactly the same signature pattern: nine pairs in a ring surrounding one central pair.

Cilia, then, can be seen as just smaller and more numerous flagella, and Lynn Margulis goes so far as to abandon the separate names and call them all by her own name of 'undulipodia', reserving 'flagella' for the very different appendages of bacteria. Nevertheless, according to the taxonomy of Sutherland's day, protozoans were supposed to have either cilia or flagella but not both.

This is the background to Sutherland's naming of Mixotricha paradoxa Mixotricha paradoxa: 'unexpected combination of hairs'. Mixotricha Mixotricha, or so it seemed to Sutherland, has both both cilia and flagella. It violates protozoological protocol. It has four large flagella at the front end, three pointing forwards and one backwards, in the manner characteristic of a particular, previously known group of flagellates called the Parabasalia. But it also has a dense coat of waving cilia. Or so it seemed. cilia and flagella. It violates protozoological protocol. It has four large flagella at the front end, three pointing forwards and one backwards, in the manner characteristic of a particular, previously known group of flagellates called the Parabasalia. But it also has a dense coat of waving cilia. Or so it seemed.

As it has turned out, Mixotricha Mixotricha's 'cilia' are even more unexpected than Sutherland realised, and they don't violate precedent in the way she feared. It's a pity she didn't get the chance to see Mixotricha Mixotricha alive, instead of fixed on a slide. Mixotrichs swim too smoothly to be swimming with their own undulipodia. In the words of Cleveland and Grimstone, flagellates normally 'swim at varying speeds, turning from side to side, changing direction, and sometimes coming to rest'. The same is true of ciliates. Mixotricha glides along smoothly, usually in a straight line, never stopping unless physically blocked. Cleveland and Grimstone concluded that the smooth gliding movement is caused by the waving of the 'cilia' but a far more exciting conclusion, this they demonstrated with the electron microscope that the 'cilia' are not cilia at all. They are bacteria. Each one of the hundreds of thousands of tiny hairs is a single spirochaete a bacterium whose entire body is a long, wiggling hair. Some important diseases, such as syphilis, are caused by spirochaetes. They normally swim freely, but alive, instead of fixed on a slide. Mixotrichs swim too smoothly to be swimming with their own undulipodia. In the words of Cleveland and Grimstone, flagellates normally 'swim at varying speeds, turning from side to side, changing direction, and sometimes coming to rest'. The same is true of ciliates. Mixotricha glides along smoothly, usually in a straight line, never stopping unless physically blocked. Cleveland and Grimstone concluded that the smooth gliding movement is caused by the waving of the 'cilia' but a far more exciting conclusion, this they demonstrated with the electron microscope that the 'cilia' are not cilia at all. They are bacteria. Each one of the hundreds of thousands of tiny hairs is a single spirochaete a bacterium whose entire body is a long, wiggling hair. Some important diseases, such as syphilis, are caused by spirochaetes. They normally swim freely, but Mixotricha Mixotricha's spirochaetes are stuck to its body wall, exactly as though they were cilia.

They don't move like cilia, however: they move like spirochaetes. Cilia move with an actively propulsive rowing stroke, followed by a recovery stroke in which they bend so as to present less resistance to the water. Spirochaetes undulate in a completely different and very characteristic manner, and that is just what Mixotricha Mixotricha's 'hairs' do. Amazingly, they seem to be co-ordinated with each other, moving in waves that begin at the front end of the body and travel backwards. Cleveland and Grimstone measured the wavelength (the distance between wave-crests) as about a hundredth of a millimetre. This suggests that the spirochaetes are somehow 'in touch' with each other. Probably they are literally in touch: responding directly to the movement of neighbours, with a delay that determines the wavelength. I don't think it is known why the waves pa.s.s from front to back.

What is known is that the spirochaetes are not just jammed haphazardly into the mixotrich's skin. On the contrary, the mixotrich has, in a repeat pattern all over its surface, a complicated apparatus for holding spirochaetes and, what's more, pointing them in a posterior direction so that their undulating movements drive the mixotrich forwards. If these spirochaetes are parasites, it is hard to think of a more remarkable example of a host being 'friendly' to its parasites. Each spirochaete has its own little emplacement, called a 'bracket' by Cleveland and Grimstone. Each bracket is tailor-made to hold one spirochaete, or sometimes more than one. No cilium could ask for more. It becomes quite tricky to draw the line between 'own' body and 'alien' body in such cases. And that, to antic.i.p.ate, is one of the main messages of this tale.

The resemblance to cilia goes further. If you look with a powerful microscope into the very fabric of a ciliate protozoan, such as Paramecium Paramecium, you'll find that every cilium has a so-called basal body at its root. Now, amazingly, although the 'cilia' of Mixotricha Mixotricha are not cilia at all, they do appear to have basal bodies. Each spirochaetetoting bracket has at its base one basal body, shaped rather like a vitamin pill. Except that ... well, having learned about are not cilia at all, they do appear to have basal bodies. Each spirochaetetoting bracket has at its base one basal body, shaped rather like a vitamin pill. Except that ... well, having learned about Mixotricha Mixotricha's idiosyncratic way of doing things, what would you guess those 'basal bodies' actually are? Yes! They too are bacteria. A completely different kind of bacteria not spirochaetes but oval, pill-shaped bacteria.

[image]

Arrangement of pill bacteria (b), brackets (br) and spirochaetes (s) on the surface of the mixotrich.

From Cleveland and Grimstone [49].

Over large parts of the body wall there is a one-to-one relationship between bracket, spirochaete and basal bacterium. Each bracket has one spirochaete stuck through it, and one pill bacterium at its base. Looking at this, it is easy to understand why Sutherland saw 'cilia'. She naturally expected to see basal bodies wherever there are cilia ... and when she looked, lo and behold, there were the 'basal bodies'. Little could she know, both 'cilia' and 'basal bodies' were hitchhiking bacteria. As for the four 'flagella', the only true undulipodia the mixotrich possesses, they seem to be used not for propulsion at all, but as rudders for steering the craft as it is propelled by the thousands of spirochaete 'galley slaves'. Much as I'd like to claim it, by the way, that evocative phrase is not my own. It was coined by S. L. Tamm, who found, after Cleveland and Grimstone's work on Mixotricha Mixotricha, that other termite-gut protozoa do the same trick, but instead of spirochaetes, their galley slaves are ordinary bacteria with flagella.

Now for the other bacteria in the mixotrich, the pill-shaped ones that look like basal bodies what are they doing? Are they contributing to the economy of their host? Are they getting something out of the relationship themselves? Probably yes, but it has not been shown definitely. They may well be making cellulases that digest wood. For of course, the mixotrichs subsist on the tiny chips of wood in the termite's gut, originally broken up by the powerful jaws of the termite. We have here a triple-decker dependency, reminiscent of Jonathan Swift's verse: So, naturalists observe, a fleaHas smaller fleas that on him prey;And these have smaller still to bite 'em;And so proceed ad infinitum.Thus every poet, in his kind,Is bit by him that comes behind.

By the way, Swift's scansion in the middle lines is (surprisingly) so ungainly that we can understand why Augustus De Morgan came behind for another bite, giving us the rhyme in the form that most of us know today: Great fleas have little fleas upon their backs to bite 'em,And little fleas have lesser fleas, and so ad infinitum.And the great fleas themselves, in turn, have greater fleas to go on;While these again have greater still, and greater still, and so on.

And finally we come to the strangest part of the Mixotrich's Tale, the climax towards which the narrative has been directed. This whole story of vicarious biochemistry, the borrowing by greater creatures of the b