The Jurassic Period: A Prehistoric Sleep Documentary

Full transcript

The hammer comes down, and the Jura limestone opens along a clean plane, a new face, the color of candle wax, cool against the palm on a warm afternoon, fine enough to take the print of a fingernail. Of all the boundaries between the periods of Earth history, the one at the top of the Jurassic is the only one that has never been agreed, and two centuries of patient work have not closed it. For a second or two, there is a smell to it, faint and mineral, the smell of a struck flint, and then it is gone into the air of the hillside. Alexander von Humboldt was twenty-five when he first held that stone in 1795 on a wooded ridge where France and Switzerland lean against one another, and he had no idea it would give its name to the Jurassic, to fifty-eight million years of the planet's history. If these journeys through deep time help you rest, subscribe, so the next ancient world can find you at bedtime. The beds under his fingers were carbonate, cream and gray, laid down grain by grain in warm, shallow water long before any alpine peak stood behind them. The outcrop itself was nothing much to look at, a bench of pale rock in a bank of beaches, damp where a cut had bitten into the hillside and left it open to the weather. That quiet band of stone now carries a name spoken on every continent, and the span it stands for is nearly as long as the whole interval separating the present day from the last of the dinosaurs. It is also, at its upper edge, unfinished.

The Jura Mountains keep their drama to themselves. They are a long, folded, wooded range lying against the Alps, more pasture and beach and fur than bare rock. And the limestone shows itself only where something has cut into the hillside, a road bench, a gorge, a quarry face, the occasional white scarp standing above the tree line like a bone through skin. It is walking country rather than climbing country. Cattle graze on the tops. The rock is mostly underfoot and out of sight, and a person can spend a day up there and see very little of it unless they are looking for it. Humboldt was looking for it, and he had been trained to. He was a mining official by education, still years away from the South American journeys that would make him famous across Europe, and at that stage of his life, he was simply a young man who examined stone with unusual attention. He carried the habits of the mine with him, the expectation that beds run in a definite order, that a particular rock has a particular company of other rocks around it, that a face can be read from bottom to top like a ledger. What he noticed was a difference. In southern Germany, there is a well-known limestone called the Muschel kalk, a shell limestone laid down in the Triassic and packed with the broken debris of a shallow inland sea. Humboldt knew it thoroughly. The carbonate beds of the Jura, he decided, belonged to another family entirely. They had a different character, a different order of layering, a different look and feel in the hand. In 1799, he gave them a name of their own, Jura Kalkstein, Jura limestone. Everything that follows grows out of that name, and the name arrived attached to a mistake. Humboldt concluded that his Jura limestone was older than the Muschel kalk of Germany.

In fact, it is younger, and by a wide margin. The rock he had separated so correctly belonged to the interval that came after the Triassic, not before it. The error tells you exactly what he had to work with. Radiometric dating lay more than a century in his future. There was no agreed sequence of periods into which a new rock could be slotted, and no chart on which a Swiss hillside could be set beside anything else on the planet and measured against it. The entire apparatus by which geologists now order time did not yet exist.

He had his eyes, his hammer, his memory of other outcrops, and a conviction that these beds were distinct from anything he had seen before. He was right about the thing that mattered and wrong about the direction. The distinction he drew between the two limestones survived every later correction and is still drawn today. The order he put them in did not last thirty years. Turning Jura Kalkstein into a period took another generation, and it happened in France. In 1829, the naturalist Alexandre Brongniart published a work with a sweeping title, Tableau des terrains qui composent l'écorce du globe, A Table of the Terrains That Compose the Crust of the Globe, and inside it, he used the phrase terrains jurassiques. His purpose was correlation. He was arguing that the limestone Humboldt had identified in the mountains on the French and Swiss border was of the same age as certain oolitic limestones in Britain, rocks made of tiny concentric grains that feel faintly like coarse sugar between the fingers and that lie across the English countryside from Dorset up into Yorkshire. In drawing that equivalence, in insisting that a Swiss hillside and an English quarry were recording the same interval of time, Brongniart coined and published the word Jurassic. Nothing was excavated to make that claim, and nothing new was found. Two piles of rock several hundred kilometers apart in different countries and different languages were declared contemporaries because one man had looked hard at both and recognized the same slice of time in each. Every later refinement of this science is that same act performed at a finer grain, bed against bed and shell against shell, until a level in a Portuguese headland could be matched to a level in a Yorkshire cliff to within the thickness of a few shells. The unit that gesture created is enormous. Following the timescale maintained by the International Commission on Stratigraphy, the Jurassic ran for roughly fifty-eight point three million years, from the close of the Triassic at two hundred one point four million years ago to the opening of the Cretaceous at one hundred and forty-three point one million years ago. It is the second of the three periods of the Mesozoic Era, which makes it the middle child, sitting between the Triassic before it and the Cretaceous after it. Counted across the whole Phanerozoic Eon, the long stretch of visible, abundant life, it is the eighth period. The number is easier to write than to feel. Fifty-eight million years does not quite reach from the present day back to the extinction that closed the Mesozoic and took the last of the great dinosaurs with it, which lies about sixty-six million years behind us, but it comes close. Everything that has happened since those animals died lies inside a span only a little longer than the Jurassic alone. The rise of every major group of modern mammals, the spread of grasslands across whole continents, the raising of the Alps and the Himalaya, the opening of the North Atlantic to its present width, and the whole cycle of ice ages that shaped the ground most people now live on.

All of that fits in a stretch of time barely longer than this one period. The Jurassic is a world in its own right, with its own beginning, middle, and end, its own climates and their reversals, its own reorganizations of life, and its own long, quiet centuries in which, so far as any rock can tell, nothing much happened at all. The Jurassic is also two different kinds of thing at once, and the distinction runs through everything that follows. As a period, it is a span of time, an interval measured in millions of years, an abstraction that exists in a chart. As a system, it is a physical object, an actual stack of rock lying in the ground, one bed resting on another that can be walked up and touched and struck.

When geologists argue about where the Jurassic begins and ends, they are arguing about both at once, which moment and which layer, and the moment is only reachable through the layer. That double nature settles where the great disputes in this science have to be conducted, not on paper, in quarries and sea cliffs with a hammer and mud on the boots. A period boundary has to be pinned to a real surface in real rock, in a place where a person can stand and put a hand on it, and the rest of the planet is then correlated back to that one spot. It is a strangely physical way to run a calendar.

It works better than any alternative anyone has proposed, and at the far end of the Jurassic, it has not yet been made to work at all. Before the period could begin, the land had to be gathered. Back in the Carboniferous, roughly three hundred and thirty-five million years ago, the separate continental units of the time drifted into one another and stayed. Gondwana came up from the south, and Euramerica came down from the north, with Siberia welding onto the northern margin as the process ran on, and out of that long convergence came a single landmass, Pangea.

Pangea was C-shaped, a great curved body of land whose bulk stretched between the northern and southern polar regions of the Earth, with the open mouth of the sea facing east, like a hand half closed around a bay. Around it, filling most of the planet's surface, lay the superocean Panthalassa. Into the mouth of the sea reached the Paleo-Tethys, and after it, the Tethys Ocean, a warm arm of water that would matter enormously to this story because so many of the finest Jurassic fossils were laid down along its edges and in its quiet margins. A single world ocean and a single world continent make for a strange planet.

Interiors that far from any coast receive very little rain because the air that reaches them has long since dropped whatever water it was carrying. Coastlines are long but simple, without the intricate fringe of seas and gulfs and island arcs that a fragmented world produces. A land animal, given enough generations, might in principle have walked from one polar region to the other without crossing open water. That connectedness is part of why life across Pangea looked so similar from place to place.

Break the continent up, and that sameness has nowhere to go but apart. One arrangement of animals, once cut through by water, becomes several arrangements, each drifting away from the others at its own speed and in its own direction, until the faunas of two coasts that were lately a single interior no longer match at all. Recognizing that any of this had happened took human beings a very long time. According to the record of the idea, the Flemish cartographer Abraham Ortelius may have been the first to suggest, in 1596, that the continents were once joined and were later torn apart. The suggestion sat largely unpursued for centuries, a curiosity in an old book. It was Alfred Wegener who assembled real corroborating evidence in three German language articles published in 1912 under the title of The Origin of the Continents, expanded into a book of the same title in 1915.

In that book, he called the single landmass the Urkontinent, the primeval continent. The famous name arrived almost by accident. Wegener used the word Pangea exactly once 1920 edition, in the phrase, "The Pangaea of the Carboniferous," written in the Germanized form Pangaea. The Latinized spelling entered German scientific literature in 1922 and reached English by 1926, notably at a symposium of the American Association of Petroleum Geologists held that November. The word itself is built from two ancient Greek pieces meaning whole or entire and earth or land. The whole land.

It is a good name, and it very nearly never became a name at all, having appeared a single time in a single edition of a book that most of Wegener's colleagues disliked. Pangaea began to come apart around 200 million years ago at the hinge between the Triassic and the Jurassic. That is not a coincidence of dates, and it is where the period opens. The Jurassic does not begin gently. Its base is placed at the Triassic-Jurassic extinction event, one of the major losses of life in the fossil record, a moment when a great many lineages present in the rocks below are simply absent in the rocks above. In the literature, this event is associated with the eruption of the Central Atlantic Magmatic Province, usually shortened to CAMP, an enormous outpouring of magma tied to the tearing apart of the supercontinent, whose remnants are now scattered across four continents on both sides of the young Atlantic.

What the sources report, there is a coincidence in time, not a demonstrated cause. The eruptions and the extinction sit close together in the rock, and there are plausible chains of cause running from vast volumes of erupted basalt through changes in atmospheric chemistry to a biosphere under intolerable stress. What has actually been established, though, is the ordering and the overlap, and honest telling keeps the claim at that strength. Deep time is generous with correlation and stingy with causation, and the further back one looks, the more generous and the stingier it becomes. What is clear is the outcome. Whatever the exact sequence, the world that survived was rearranged.

On land, the Triassic had been shared between two great groups of archosaurs: the dinosauromorphs, the lineage leading to dinosaurs, and the pseudosuchians, the branch whose surviving members are crocodiles and their kin. After the extinction, that sharing ends. The Jurassic land fauna becomes a dinosaur fauna. Among the survivors, too, were the cynodonts, of which only a few lineages came through, including the mammaliaforms, the group that carried the mammal story forward. The Jurassic inherits both the animals that would dominate its landscapes and the small, quiet ones whose descendants are still here now, still small in the main, and still doing after dark most of what they were doing then. There is no single terrible day to picture in that transition and no sound to attach to it.

Extinction events in rock are mostly a matter of things that are there and then are not. A bed of shale full of a particular shell, and above it a bed of shale without it, and between them a surface no thicker than a sheet of paper, which under a lens turns out to be nothing more than a slight change in the size of the grains. A geologist working that section measures the loss by what stops appearing, marking absences upward through the mudstone, bed by bed, until the list of what is missing is longer than the list of what remains.

Into that emptied world came a geography in motion. By the opening of the Jurassic, Pangea had already begun to rift, and the single continent was separating into two, Laurasia in the north, Gondwana in the south, with new sea widening between them. The rifting ran too slowly for any lifetime to register it. It was steady spreading, a few centimeters in a year and no faster, an increase that would take a human life to amount to the width of a room, sustained without interruption across tens of millions of years until a seam had become a sea. Every separation of that kind changes the weather.

New seaways carry heat and moisture into places that had been dry continental interior. Coastlines multiply, and with them the shallows, the bays, the sheltered lagoons, and the muddy shelves. Water of that kind is extremely good at making fossils because it manufactures limestone and mud fast enough to bury whatever falls into it before decay has finished taking the body apart. The climate of the Jurassic was warmer than the present day, and the most important thing to say about it is a negative. There were no ice caps at either pole. The evidence describes forests growing close to the polar regions, trees standing at latitudes that now carry ice sheets a mile or more thick.

Anyone picturing this period, therefore, has to take the ice out of the picture and then find something to put in its place, and what the evidence puts there is timber. Wooded country ran to both ends of the earth, rooted in soil that thawed and drained, dropping needles onto a floor that went on rotting quietly all through the dark months. Rivers at those latitudes ran liquid in every season, through winters that never locked them. Whatever came down as snow in a polar winter was gone again by the following summer, and the ground beneath it never froze hard enough to split a root.

Meanwhile, across the lower latitudes, broad arid expanses spread over continental interiors, dry country of the sort that a large landmass generates when its middle lies impossibly far from any coast. Seasonality that near the poles had almost nothing to do with cold. It was governed by light. A tree standing within sight of the polar circle spent months in continuous darkness and months under a sun that never climbed far above the horizon, and it grew in surges and pauses accordingly. That rhythm survives in petrified wood as banding, wide increments alternating with narrow ones down the length of a trunk, and it is one of the very few places in this entire period where time can be read one year at a time. Between the wooded poles and the parched middle lay the country that produced most of the rock in this account.

Humid coastal lowlands, river deltas, muddy shelves, and quiet embayments along the edges of the widening sea. That is where sediment piled up fastest and burial was quickest. The great dry interiors kept almost nothing. The margins of continents kept nearly everything anyone now knows about the period. An interior loses material. A river leaving a dry heartland is carrying that heartland away by the ton. The coast is where the carrying stops, and a shelf that keeps sinking gently under the weight of what it receives will go on taking delivery for millions of years without a break in the sequence. A forest without flowers is a difficult thing to hold in the mind, and it is the first real demand the Jurassic makes of anyone who wants to picture it honestly. That absence removes an entire category of color from the world.

No blossom on a branch, no petal fallen on water and turning slowly in an eddy, no bright patch at the edge of a clearing where the light comes through. Nothing in that that whole landscape opened, offered, and closed again. The plants of the Jurassic reproduced by older means, spores let go into moving air and dry pollen shed from cones, and the record preserved in these sources shows ferns and gymnosperms in undisputed possession of the land, with no rival group anywhere in the sequence with the equipment to displace them. Grass belongs to the flowering plants, which means the Jurassic had none of that either. Whatever a Jurassic animal set its foot on was something else, bare soil, leaf litter, moss over a wet rock, a low mat of fern, or grit and stone with nothing growing on it at all. Nothing knitted the surface together the way sod does, so hillsides shed their soil freely under heavy rain, and the rivers ran brown and loaded with it, and the floodplains of this period buried a great deal, quickly and well. Every prehistoric scene that puts a long-necked animal knee-deep in grass has quietly imported a plant that would not exist for tens of millions of years. What lay between the trunks instead was fern in every size that ferns can manage, from low fronds spreading over damp soil to growth standing at the height of a person and higher.

Ferns are ancient and patient plants, tolerant of shade, quick to colonize bare ground after a flood or a fire. In this period, they were the carpet of the world, running unbroken from the wet floor of the polar woods to the last shaded soil at the edge of the desert, and they had held that ground for a very long time already. Above them stood the cone bearers. Conifers dominated the canopy, and one of the quietly remarkable facts about this period is that many modern conifer groups make their first appearance in it.

The families whose descendants now cover mountainsides in needle and resin were beginning here. That means a Jurassic forest, for all its strangeness, would not have smelled entirely alien. The chemistry of resin has not moved on. Needles crushed underfoot released the same volatile sharpness they release today. Warm bark on a still afternoon gave off the same faint turpentine, and a wound in a trunk wept the same slow amber that would occasionally close over an insect and keep it. Of all the sensory details of this world, the smell of a stand of conifers on a warm afternoon is probably the one a visitor from the present could recognize with real confidence.

The leaves were needles and scales, tough, narrow, waxy, built to hold water and to shrug off the abrasion of dry wind. Reproduction ran through cones with pollen carried on moving air and no animal courier anywhere in the arrangement. A conifer forest in pollen season releases clouds of it, drifting yellow across water and settling in the corners of everything, and pollen is one of the sturdiest objects a plant can leave behind. Grains survive in rock long after the tree that made them has vanished without trace. Whole vegetations have been reconstructed from sediments containing no wood at all, only dust. A hillside of forest recovered from a spoonful of gray mud and a microscope.

Sharing that canopy were the bennettitales, a group with no living members and therefore no living comparison. Everything known about them is read from stone. They are described in these sources as one of the dominant gymnosperm groups of the period, and their complete disappearance means that no reconstruction of them can ever be more than an argument. Paleobotanists infer form from fossil trunks, from fronds, from reproductive structures that match nothing alive, and the drawing is redone whenever a better trunk comes out of the ground. One of the commonest large plants of the Jurassic landscape is therefore also one of the least confidently pictured, and two illustrations of it made forty years apart may not resemble each other in outline, in height, or in the arrangement of the crown. So the standing landscape looks like this, green in gradations of frond and needle and scale, textures running from soft new fern growth to hard mature foliage, with fallen litter accumulating on bare soil rather than on a mat of roots. Add to that the plants not yet named, cycads holding stiff crowns close to the ground, ginkgos carrying leaves shaped like small split fans, horsetails standing in the wet margins in dense jointed stands, ribbed and hollow and green to the tip.

Light reaching the floor in the pattern that needle canopies make, dappled and shifting, never dense. And a particular sound in weather, because a canopy of needles hisses where a canopy of broad leaves rattles, so wind moving through a Jurassic wood ran thinner and steadier than wind in any wood a person could walk into now. In the higher latitudes, forest built for long darkness. In the low latitudes, dry country where the plants thin out and the bare ground shows through between them. Every large plant eater in this period had to make a living from that menu, and it is not a generous one. Fronds and needles are fibrous, tough, and slow to digest.

Paleontologists generally infer that the great Jurassic herbivores processed enormous volumes of low quality vegetation, and one line of reasoning behind their long necks and capacious bodies begins exactly there. That reasoning comes from anatomy and from the plants available, not from any observation of an animal feeding. The forest is what the evidence supports. The animal's response to it is reconstructed backwards from the bones of the animal itself, and any reconstruction has to accept the menu exactly as the plants leave it, all fiber and resin and cone, with nothing sweet anywhere in the picture and no fruit to be had at any season. About eighteen million years into the period, the sea did something serious.

Around one hundred and three million years ago, at the opening of the Toarcian age, the record registers a crisis that has collected an unusual number of names. It is called the Toarcian Extinction Event. It is also called the Pliensbachian to Toarcian Extinction Event, the Early Toarcian Mass Extinction, the Early Toarcian Paleoenvironmental Crisis, and the Jenkins Event. A pileup of names like that is itself a piece of information. It means that different research communities came at the same interval from different directions, one measuring extinction, another measuring environmental chemistry, and that the field has not yet settled on a single label. The name Jenkins Event is simply one of several currently in use. What the evidence describes is not one blow, but two main pulses separated in the rock. The first is called the Pliensbachian to Toarcian Boundary Event, and its position is given with a precision that repays a slow reading. It falls in the mirabile subzone of the tenuicostatum ammonite zone.

That phrase is an address as exact as a house number on a street. The tenuicostatum zone zone is a thickness of rock defined by the presence of a particular ammonite assemblage, and the mirabile subzone is a thinner slice inside it, defined the same way. Geologists locating this event are not primarily giving a date in years. They are giving a level, a specific band of stone that can be recognized in Yorkshire and in Portugal and in Germany because the same shells appear in the same order in all three.

Within that band, two things coincide. Ocean oxygen concentrations dip slightly, and the world begins to warm after a cool interval in the late Pliensbachian. Neither of those is catastrophic on its own. A small drop in dissolved oxygen is the sort of change that stresses some animals and passes others by entirely. The end of a cool period is, in isolation, an unremarkable turn of climate. But the pairing matters because warm water holds less dissolved oxygen than cold water and because the direction of travel would continue. Some authors do not read the first pulse as an opening act at all. They classify it as its own separate event, unrelated to the more extreme crisis that followed.

That disagreement should be stated plainly and not smoothed over. What would settle it is a single section somewhere that records both pulses in unbroken sediment with enough ammonites running through the interval to show whether the first slides into the second without a pause or stops, recovers, and starts again from the beginning. No section has yet been agreed on that does the job. The second pulse is the one that leaves the deeper mark. It is called the Toarcian Oceanic Anoxic Event, and it sits near the transition between the tenuicostatum and serpentinum ammonite zones, with additional smaller extinction pulses scattered across the same span. Anoxic means without oxygen, and the name simply states the single observation that the rocks of this interval make everywhere they are found.

The documented mechanism runs like a chain, each link pulling the next, and the sources are careful to hedge the first link as likely rather than proven. Volcanism from the Karoo and Ferrar Large Igneous Provinces released considerable carbon dioxide into the atmosphere. Global temperatures climbed into an interval named the Early Toarcian Thermal Maximum, described in the literature as a super greenhouse. Warmer air and higher carbon dioxide acidified the oceans. On land, a hotter and wetter world weathered rock faster, and rivers carried the products of that weathering to the sea, including phosphate, which is a nutrient. Then the chain does something that sounds gentle and is not.

Extra nutrients made the surface waters richly fertile, a condition called eutrophication. Plankton flourished. Plankton also dies, and dying organic matter sinks, and the decay of sinking organic matter consumes oxygen. Fertility at the surface therefore drained oxygen from the water beneath it. Eutrophic became oxygen-poor. Oxygen-poor became anoxic. And in water without oxygen, animals with gills cannot live. Biodiversity fell sharply. Every link in that sequence is an ordinary process, weathering, nutrient delivery, plankton growth, decay, and each of them leaves its own trace in the sediment so the chain can be tested link by link with instruments. None of it has to be argued in the abstract. In rock, the whole sequence reads as two signatures that a geologist can find with a hammer and a laboratory. The first is a high amplitude negative carbon isotope excursion, which is a way of saying that the balance of carbon isotopes in these sediments swings hard in one direction and then recovers. Carbon released from volcanic sources carries a different isotopic fingerprint from carbon already circulating through living systems, so a large injection changes the ratio recorded in shells and mud, and that swing can still be measured to a fraction of a part per thousand, a hundred and eighty million years after the fact. The second signature is simpler and visible to the naked eye. Black shale, thick beds of it. Sediment turns black when organic matter is buried without being oxidized, and organic matter escapes oxidation when the water above it holds no oxygen.

Those dark, finely layered rocks splitting into sheets that smell faintly of oil when broken are the physical record of a sea floor where nothing was breathing. The sources describe the Toarcian oceanic anoxic event as possibly the most extreme case of widespread ocean deoxygenation in the entire Phanerozoic, which is to say in the last half billion years of abundant animal life. That superlative is explicitly hedged and the hedging is forced by arithmetic. It is not modesty. Comparing severity across hundreds of millions of years means comparing archives of very different quality.

A basin that kept subsiding recorded the whole crisis in continuous mud, while a basin that shallowed recorded an opening chapter and then a gap where the rock was never laid down. The oxygen itself is never measured directly. It is read through proxies, through trace metals and the chemistry of pyrite, each of which answers a slightly different question about the water it grew in. The ammonites stand on both sides of the whole account. The zones by which the crisis is dated are defined by their shells, and a large share of the extinction being dated is their own, so the same animals serve as the calendar and as the casualties. There is more than an irony in that. It has a working edge where the shells thin out, the dating thins out with them, and the intervals in which the losses were worst are sometimes the intervals whose timing is least precisely known. Above the anoxic water, life on land went on reorganizing itself at a pace no single event can capture. The shift from a shared archosaur world to a dinosaur world was the headline, but it was not the whole reordering. The mammaliaformes, descended from one of the very few cynodont lineages to survive the end of the Triassic, went on diversifying through the period, and during the Jurassic, the sources record the appearance of the first crown group mammals. Crown group means inside the last common ancestor of the living members of a group and all of its descendants.

Everything outside that circle sits on the stem, related, often closely, sometimes with teeth and jaws a person would call mammalian at a glance, but standing outside the family that living mammals define. The Jurassic is where the circle closes, where the tree of mammals as they now exist stops being a set of relatives and becomes the thing itself. They were small. Everything about the Jurassic mammal story is small, and smallness is easy to overlook beside animals that could shake the ground. What survives of them is mostly teeth and fragments of jaw measured in millimeters, recovered by dissolving blocks of rock in weak acid or by washing sediment through fine sieves and picking through the residue under a lens.

It is patient, unglamorous work, and it is how an entire branch of the story has been assembled. The lineage that would eventually include bats and elephants was already present in those forests, moving through the fern layer, probably at night, leaving almost nothing behind but enamel. Alongside them, a set of animals a modern listener would recognize instantly begins to appear. Modern frogs make their earliest appearance in the Jurassic. So do salamanders. So do lizards. The sound of a wet Jurassic night, therefore, includes something entirely familiar, because frogs were calling in that fern understory, and whatever the calls were, they were being made by animals built on the plan frogs still use. The earliest crabs also appear in this period.

Crabs are one of evolution's most persistently repeated designs, arrived at again and again from different starting points, and the Jurassic is where the story begins. On a Jurassic shore, alongside the coiled shells and the unfamiliar reptiles of the shallow water, something small and sideways walking was already doing what crabs do, picking through rack at the tideline. There is a pattern running through that small bodied company, and it is worth naming. Each of those lineages had a way of living that no giant could adopt, a body that fits under a log or into a crack in a limestone pavement, a diet of insects and larvae and soft wet things that no drought or ash fall ever entirely removes, a habit of moving in the dark hours when the large animals were still. Smallness of that kind buys very little in easy conditions and very nearly everything in hard ones, and the role of survivors across the whole fossil record leans heavily toward animals that could hide.

And there is a second consequence, entirely practical for the people who came later to read the rock. Because such animals shed teeth and jaw fragments and single vertebrae by the hundred into pond mud and into fissure fills in limestone, they left their record in exactly the deposits where nothing large was ever buried. A bed containing not one bone worth mounting in a gallery can still be washed, sieved, sorted under a lens and matched tooth for tooth against another bed two valleys away. And then there is the air, which the Jurassic solved twice. Pterosaurs were the dominant flying vertebrates of the period. They had been aloft since the Triassic, and they held that dominance throughout the Jurassic, and precision matters here because popular usage muddles it constantly. Pterosaurs were not dinosaurs. They belong to their own group, a separate branch of the reptile family tree, related to dinosaurs but distinct from them. Their wing was built around one enormously elongated finger supporting a membrane, an architecture with no equivalent anywhere in the dinosaur lineage.

During the same period, from a branch of theropod dinosaurs, the first stem group birds appeared. Stem group is the opposite of the crown described a moment ago. These animals ran along the lineage leading toward birds without falling inside the group that living birds define, close relatives of the family without ever being members of it. Their flight surface was assembled from feathers rooted along the arm and hand, an entirely separate answer to the same problem of holding a body up in moving air. So the Jurassic sky held two independent inventions of flight among vertebrates, crossing the same warm air above the same fern forests and the same widening sea. The two solutions aged differently, and that is the more interesting fact about them. A membrane is one object. A tear opens across the working surface all at once, and the animal is grounded until the skin knits back together. A feathered wing is an assembly of separate parts, each replaceable on its own schedule, so wear is made good piece by piece while the animal goes on flying. That difference is not a matter of upkeep alone. It sets how often each kind could afford to be clumsy, how a juvenile could grow into an adult surface without ever being flightless, and how much of the wing an animal could risk in a bad landing.

Both designs were in the air above the same lagoons, and neither was available to the other. Holding those two lineages apart is not pedantry, and in the rock, it is not always easy. Both leave a faint field around the bones where the flight surface lay, and telling one field from the other is a matter of what is written inside it. A membrane preserves as a smooth sheet crossed by fine parallel fibers running back from the leading edge, stiffening it the way battens stiffen a sail. A feathered wing preserves as a series of separate quills, each with its own central shaft and its own spread of veins overlapping like tiles on a roof.

That distinction, read at the width of a hair on a freshly split slab, is the whole reason a particular block of Bavarian limestone quarried by men who wanted a smooth surface for printing became one of the most argued over objects in the history of biology. Before that slab, the water, the lagoon that would preserve it lay at the edge of a sea, and the Jurassic Sea has its own claim on the attention because it was doing at least as much invention as the land. With no polar ice to chill it, warmth reached surprisingly high latitudes, and the difference in temperature between the tropics and the far north was gentler than anything the modern ocean maintains, which slowed the great overturning currents and left large volumes of water standing still for long stretches. The rifting continents were low along their edges, and the sea came far inland over them, flooding ground that is now dry farmland across Europe and putting most of what is now England under a warm, quiet, sunlit shelf. Much of the rock geologists later used to define this period accumulated on those flooded shelves. Carbonate coming out of solution in water warm enough to make it all year round in places so shallow that a swimmer could have stood on the bottom with their head in the air a hundred kilometers from the nearest shore. Two groups of large marine reptiles dominate the picture. Ichthyosaurs swam these seas throughout much of the period, and plesiosaurs shared them.

Both were reptiles that had returned to the water long before, and both are known from skeletons alone and never from observation, which is a thing to keep in mind whenever a museum mural shows one arching through blue light. The outline of an ichthyosaur, the way a plesiosaur held its neck, the color of either animal, all of that is reconstruction built from bone shape, from proportion, from comparison with living swimmers that face the same physics. The skeleton is the evidence, and the attitude it has been given in the gallery is an interpretation laid carefully over that evidence by people who know exactly how much of it they are supplying. The crocodylomorphs were doing something more unusual, and the Jurassic catches them partway through it. That lineage had walked on land in the Triassic on limbs held under the body.

In the dry world, their relatives shared with the earliest dinosaurs. During the Jurassic, members of it made the transition from terrestrial life to aquatic life. This is not a small adjustment of habit. It is a whole body being renegotiated, the limbs, the tail that has to become a propulsive organ, the skull with its nostrils migrating, the way an animal breathes while mostly submerged, the places it can rest, and the places it can no longer go. Every crocodile lying in a river today belongs to a group whose ancestors made that commitment, and the Jurassic rock catches the commitment while it is still being made. Then there are the fish, and here the period starts to feel less like a foreign country. Modern sharks first appeared in the Jurassic and diversified through it. So did rays. The shark shape is very old, but the particular kinds of shark that patrol the present ocean trace their beginning to these warm waters, and the same is true of the flattened bottom-hugging design of the ray.

Sharks shed and replace teeth constantly through life, so a single animal seeds the sediment with hundreds of them over the course of a lifetime. That habit is what keeps the early history of the group legible at all, and it is what leaves certain beaches giving up teeth by the handful to anyone patient enough to look. Near the end of the period comes the quietest and most consequential arrival of the lot. The first known crown group teleost fish appear in the late Jurassic. Teleosts are the great bulk of what anyone means by the word fish. Herring, cod, salmon, carp, tuna, the small silver things flickering in a rock pool. Nearly all of it sits inside that one group. Their signature is a jaw arrangement that can be thrust forward, turning a mouth into something closer to a pipette, and that mechanism deserves a moment because it explains the numbers that followed.

In an ordinary fish, the mouth opens and the animal must swim onto its food. In a teleost, the front of the skull slides forward as the jaw drops, and the sudden increase in the volume of the mouth pulls water inward, so the food is drawn in rather than run down. An arrangement like that can be tuned in a thousand directions, drawn out into a tube for probing crevices, shortened and thickened into a crusher for shells, angled to work against a rock face or held level for open water, and the tuning is most of the reason the group would later occupy every kind of water there is. The Jurassic Sea holds the very beginning of it in animals no longer than a hand, which left behind not only bones, but the small paired stones of the inner ear, grown in fine concentric rings like the rings in a trunk, and identifiable to species from their outline alone.

But none of those animals is the one that geologists actually depend on. For the practical business of reading Jurassic time, the important creature in that water was smaller, far more numerous, and built like a tightly wound spiral of chalk. Ammonoids were cephalopods, and they are entirely extinct. Their nearest living relatives are not the chambered nautiluses they superficially resemble. They sit closer to octopuses, squid, and cuttlefish, which means the animal inside that hard coiled shell was kin to the soft, quick, large-eyed mollusks of the modern sea. The shell suggests a slow drifter. The family suggests something else entirely.

Anyone reconstructing one has to picture a squid that happens to be carrying a house with the eyes and the appetite that go with the comparison. The shells are usually plan spiral, every whirl staying in the same plane as the one before it so that a shell set down on a table lies flat and true with no tendency to tip onto a shoulder. Helically spiraled forms and strange uncoiled ones exist, but those heteromorphs belong mainly to the Cretaceous, later than the period in question. The Jurassic ammonoid is generally the classic shape, a flat spiral ribbed across the whirls, the outer chamber housing the living animal and the inner chambers walled off behind it.

Each wall meeting the outer shell along a line of extraordinary complexity that shows on a weathered specimen as a frilled branching seam. Those seams are diagnostic. A specialist can name a genus from the pattern alone, so a broken ammonite is often worth more to a stratigrapher than a perfect one since breaking is precisely what lays the seam open to be read. They are very old. The earliest ammonoids appeared in the Emsian stage of the early Devonian at around four hundred and ten to four hundred and eight million years ago. That comes as a range and not a single figure because a date that deep is anchored by a scattering of datable ash beds with everything between them interpolated and slack in the interpolation. It is more than two hundred million years before the Jurassic even opened.

Over ten thousand species have been described. The group survived catastrophe after catastrophe, and each time it came through, it came through narrower. By the Jurassic, only one order remained, Ammonitida. Everything that follows in this story is that single surviving branch radiating outward again into the warm shelf seas of a splitting world. The name they carry is a Roman inheritance. Pliny the Elder, the encyclopedist who died in seventy-nine AD near Pompeii while the volcano was erupting, is credited with calling these stones ammonis cornua, the horns of Ammon. Ammon, also written Ammon, was an Egyptian deity commonly depicted wearing ram's horns, and a ram's horn is exactly what a coiled ammonite looks like when it is weathering out of a gray cliff. The name stuck for nearly two thousand years.

It is still stuck. It also seeded a convention because a great many ammonite genus names end in the syllables ceras from the Greek word for horn. Anyone reading a stratigraphic table will meet that ending again and again, a small classical echo running through modern technical writing. Given that ten thousand species have been named and described in exhaustive detail, what remains unknown about them is startling. The shells preserve beautifully. The soft parts almost never do. Without soft tissue, there is no direct evidence of how the animal moved, what it ate, how fast it swam, or how it behaved toward its own kind.

Everything is inference from shell morphology, from the thickness of a wall, from the geometry of the chambers, from where the center of buoyancy must have sat relative to the center of mass, and from water tanks. One of the more charming methods in the field involves building physical models of ammonite shells and lowering them into tanks to see what they do. Which way up does this shape float? How stable is it when nudged? How much drag does a heavily ribbed shell carry compared with a smooth one? Does this particular spiral hang nose down or level or tip helplessly onto its side? What comes back out of a tank is never an animal.

It is a list of things the animal cannot have done. A shell that hangs nose down cannot have hunted level in open water, and a shell whose ribbing drags heavily cannot have chased anything quick. So each experiment strikes possibilities off the list and leaves the creature that used to live inside as unwitnessed as before. An animal whose way of life remains largely unknown nevertheless became the master clock of an entire geological period, and the reason has less to do with biology than with bookkeeping.

Ammonites were abundant. They were widespread across marine basins. They had hard parts that fossilized readily, and crucially, they changed fast. A species would appear, spread, persist for a geologically brief span, and vanish. Different species did this at different moments. Stack those appearances and disappearances in order, and the result is a sequence of markers laid through the rock like knots tied along a rope at uneven intervals. Jurassic stratigraphy rests primarily on ammonites as index fossils. The key concept is the first appearance datum, the lowest level in the rock at which a particular ammonite species is found.

That datum, and not any physical change in the rock, marks the beginning of a stage. It also marks the beginning of the smaller units within stages, which are called ammonite zones, and those are sometimes divided again into subzones. The result is a nested hierarchy of time built entirely out of shells, period, stage, zone, subzone, each boundary defined by a creature arriving. This has a consequence that catches people off guard. The boundaries of Jurassic time are biological events, not physical ones. A quarry face may show no visible change at all across a stage boundary. Same gray limestone below, same gray limestone above, the same beds running on without a flicker. The line is there because somewhere in that interval, a particular ammonite makes its first appearance, and by international agreement, that appearance is the start of a new stage.

Because the system was built in Europe, global stratigraphy uses the standard European ammonite zones as its reference. A sequence in South America or in Japan is correlated back to a standard established in English and French and German rock, and never the reverse. There is nothing mystical about that. It is simply where the work was done first, and the framework proved good enough that it was cheaper to calibrate to it than to replace it. Above all of this sits the International Commission on Stratigraphy, which ratifies each global stage by fixing a global boundary stratotype section and point.

The nickname is the Golden Spike, and the device is marvelously literal. Rather than defining a boundary by a description that different geologists might apply differently in different countries, the commission selects one particular outcrop in one particular formation at one particular place on the planet and declares that the boundary is there. Not approximately there, but there, at that level, in that bed, marked on the ground with a small plaque or a metal disc, and every other section in the world is measured against it.

Everything above the point belongs to one stage. Everything below it belongs to the stage before. Choosing such a place is harder than it sounds, and the requirements are unforgiving. The section has to record continuous deposition because a gap of even a few thousand years destroys the very thing being defined. A first appearance cannot be located in rock that was never laid down. It has to contain fossils that occur widely enough to be recognized on other continents. It has to be well exposed, physically accessible, and studied thoroughly enough that its contents are not going to surprise anyone later. Most outcrops fail on at least one of those counts, and they fail for reasons a visitor would not guess. The sea is the awkward case, and it cuts both ways.

A cliff being trimmed back by the waves renews its own face, keeping the beds clean and legible year after year instead of letting them disappear under soil and scrub. A surprising number of ratified points therefore sit within sound of surf. The price is that the same water can put the defining level under weed for half of every day and will eventually take the section away altogether. The commission accepts that bargain because completeness comes first and convenience comes last. A gap in the sediment cannot be argued around. A difficult walk at low tide can.

So the ratified list runs to both extremes, scoured coastal ledges at one end and quiet inland slopes that nobody would ever photograph at the other, each chosen for nothing but the steadiness of what it contains. Getting to that level of precision took a very long time, and it began with a word borrowed from quarrymen. In 1822, the English geologists William Conybeare and William Phillips used the term Lias for a set of strata in England. The word was already in use among workmen, applied to the alternating limestones and shales that split so cleanly along the southern English coast. Conybeare and Phillips gave it a place in a formal description of British strata, and it came to cover rocks equivalent to what would eventually be recognized as the lowest Jurassic. Lias is still used, still taught, and still turns up on geological maps. It arrived seven years before the word Jurassic did, a reminder that the labor of describing these rocks was well underway before anyone thought to call them by the name of a mountain range in another country.

Ten years after Brongniart's tableau, the material got sorted by color. In 1839, the German geologist Leopold von Buch established a threefold division of the Jurassic, which is still echoed in German usage. Black Jurassic at the bottom, brown Jurassic in the middle, white Jurassic on top, oldest to youngest. The names are literal. They describe what the rock actually looks like in the field in southern Germany. The lower beds are dark clays and shales, heavy with organic material laid down in poorly oxygenated water. The middle beds run to clay sandstones and ferruginous iron-rich oolites, which weather to a rusty brown. The upper beds are pale limestones, bright in sunlight, the same family of rock that eventually yielded the Bavarian slabs. A geologist walking uphill through a German section could therefore watch the period change color under their boots, from dark to rust to white, and read a rough position in time from that alone.

And the English word Lias, in use since 1822, turned out to correspond to von Buch's black Jurassic. Two men in two countries working in two languages on two piles of stone had drawn the same line without consulting each other, and that kind of accident is what reassures everybody involved that the framework has hold of something real. Color coarse instrument, though. Three divisions across roughly fifty-eight million years is not much resolution, and it fails the moment one leaves the country where the colors hold. The next generation went after finer subdivision, and they did it with ammonites. Between 1842 and 1852, the French paleontologist Alcide d'Orbigny divided the Jurassic into ten stages based on assemblages of ammonites and other fossils in England and France. He was not tracing color bands or rock types. He was tracing which creatures occurred together and where the membership of that company changed. Seven of the names he introduced are still in use today.

Not one of them retains its original definition. That last point is easy to skim past, and it captures how this science actually behaves. The Sinemurian, which d'Orbigny named in 1842 after Semur-en-Auxois near Dijon, originally included everything that is now called the Hettangian. The stage was later cut in half and the older portion given a separate name, and what happened to the Sinemurian happened, in one form or another, to every one of them. A stage in d'Orbigny's hands was a bundle of fossils that occurred together, with edges as soft as the collections available to him. A stage now is a single surface in a single named bed on a named coast.

The words traveled that whole distance intact while everything underneath them was dismantled and rebuilt. A geologist today can use a term coined in the 1840s and mean by it something its author would not recognize at all. In 1858, Friedrich August von Quenstedt worked over the German side of the problem. Taking von Buch's three-color series in the Swabian Jura, he divided each of them further, producing six subdivisions defined by ammonites and other fossils. The Swabian Jura is a plateau of pale limestone in the south of Germany, cut by valleys and pocked with caves and old quarries, and Quenstedt knew it in the kind of detail that only comes from decades of walking the same ground with a hammer in all weathers. His subdivisions gave the German scheme something closer to the resolution the French stages had. Running in parallel between 1856 and 1858, Albert Oppel took d'Orbigny's arrangement and reworked it. He altered the scheme where he thought the evidence demanded it, and then he went finer still, subdividing the stages into biostratigraphic zones defined chiefly on ammonites. Oppel's zones are the direct ancestor of the ammonite zones and subzones still doing the practical work of correlation across continents. After him, the Jurassic had a scale of divisions fine enough that two geologists working in different countries could disagree about which of their beds was the older and expect the disagreement to be settleable by evidence.

What followed was more than a century of quiet argument. Stages were proposed, contested, redefined, merged, split, and occasionally abandoned without ceremony. Most of the stage names in use today were only formalized at a meeting held in 1962, the Colloque de Jurassique, a Luxembourg, where the community sat down together and settled on a working scheme. That is one hundred and thirty-three years after Brongniart printed the word Jurassic, and more than a century and a half after Humboldt first walked those Jura carbonates and misjudged their age. Even then, the settlement was not final. Golden spikes continued to be driven long afterwards, one stage at a time, and a few of the Jurassic stages carry no ratified spike at all to this day. Four of those points now anchor the early Jurassic, and each one arrived by its own crooked route.

The oldest is the Hettangian, the stage that opens the period. Eugène Renevier named it in 1864 after Hettange-Grande, a town in northeastern France, and for well over a century, the base of the stage was defined by the first appearance of an ammonite called Cephaloceras planorbis, a marker chosen by Oppel in the 1850s. The trouble with planorbis was geography. It was reliable in the rocks where Oppel had worked and much less reliable elsewhere. A global boundary needs a fossil that turns up globally, and planorbis simply did not travel far enough to do the job. So the boundary moved in the sense that its definition was rebuilt while its age stayed much where it was.

In 2010, the commission ratified a new point, and it is not in France at all. It sits at Kuhjoch Pass in the Karwendel Mountains of Austria in the northern calcareous Alps, a high saddle of pale rock and thin alpine air. The marker there is a subspecies of ammonite named Cylotera spillei tirolicum, and it occurs in a unit called the Kendlbach Formation. The base of the Jurassic, and therefore the top of the Triassic, and therefore the line drawn under the extinction associated with the Central Atlantic Magmatic Province, is now fixed at that Austrian pass. A walker could cross the boundary between two periods of Earth history in a single stride and notice nothing but a change in the color of the ground. The next stage up is the Sinemurian, one of d'Orbigny's originals from 1842, named for Semur-en-Auxois near Dijon.

Its golden spike went in earlier than the Hettangian one in the year 2000, and it is on the coast of Somerset in England at a cliff face just north of East Quantoxhead, about six kilometers east of the harbor town of Watchet. The rock there is the Blue Lias, a rhythmic stack of gray limestone and darker shale that weathers into flat ledges running out across the beach. The very alternation that gave the quarrymen their word two centuries ago, now carrying the base of a global stage. The defining fossil is an ammonite named Vermiceras quantoxense, its species name taken from the place where the boundary was fixed. The tide crosses that section twice a day, and that constant scouring is half the reason the beds are so readable there because nothing is allowed to grow over them or slump across them for long. The definition of a global unit of time therefore sits in the splash zone under seaweed for half of every day with nothing bolted to the ledge to say so. The authority for it lives in a published ratification and in a measured log of the beds drawn to the centimeter so that a visitor standing there at low water has to take the boundary on trust from a printed diagram and count upward from a bed they can recognize.

The Pliensbachian follows, and its name belongs to Oppel, who introduced it in 1858 after the hamlet of Pliensbach at Zell unter Aichelberg in the Swabian Alb near Stuttgart. It is a small settlement, the sort a traveler passes without registering, and its name is now spoken by geologists on every continent. Once again, the naming place and the defining place are not the same. The spike was ratified in 2005 at Winehaven in Robin Hood's Bay on the Yorkshire Coast within the Redcar mudstone formation. The marker is an ammonite called Bephceroceras donovani. Robin Hood's Bay is a broad curve of cliffs and wave-cut platforms, and the platforms are laid out in great concentric arcs of eroded mudstone that look from the clifftop path above like the growth rings of something enormous.

Walking outward across those arcs is walking upward through time. Then comes the Toarcian, the stage that opens with black shale and warm, oxygen-starved water. d'Orbigny named it in 1842 after Thouars in the Loire Valley, south of Saumur, a town known to the Romans as Toarcium, and his original locality was the Vrines Quarry, roughly two kilometers to the northwest. The stage waited a long time for a formal point. Its spike was ratified only in 2014, and it is at Peniche in Portugal, on a rocky headland pushed out into the Atlantic.

The marker is an ammonite of the genus Dactyloceras in the subgenus Eodactylites. Standing at Peniche with the swell breaking below and the tilted beds running out into the sea, a visitor is looking at the agreed beginning of one of the harshest intervals the Jurassic oceans ever experienced. Three more spikes hold the middle Jurassic, and the pattern of small towns lending their names to global time continues without interruption. The Aalenian was named by Karl Mayer-Eymar in 1864 after Aalen in Germany, and its origin is a lovely piece of continuity with the older color scheme. The boundary was originally drawn where the dark clays of the black Jurassic gave way to the clayey sandstone and ferruginous oolite of the brown Jurassic.

That is von Buch's 1839 division still doing useful work a quarter of a century later, translated from a color band into a stage boundary. Ferruginous means iron-bearing, and it is the iron that gives those beds their rusty tone. Oolite means the rock is built of ooids, tiny concentric grains of carbonate that form in warm, agitated water on a shallow bank, each one growing like a pearl around a fragment of shell or a grain of sand. A hand specimen of oolite looks and feels remarkably like compressed fish roe, and the name says so, being built from the Greek word for egg.

The Aalenian spike was ratified in 2000, and it went not to Germany but to Spain at Fuente El Saz in the Iberian range near Guadalajara. The marker is an ammonite named Leoceras opalinum, and the Opalinum Zone had been a familiar unit in German stratigraphy for generations before the formal point was chosen. The section at Fuente El Saz is the opposite kind of place from a sea cliff. Inland, dry, a series of low marley slopes in open country with no drama in them whatever and never a wave to trim them or a tide to hide them. Its value lies entirely in the completeness of its record and the reliability of its ammonites. Above it sits the Bajocian, another of d'Orbigny's 1842 names, taken from Bayeux in Normandy, which the Romans called Bayeux.

Its spike was the earliest of all the Jurassic points to be ratified, agreed in 1997, and it is in the Murtinheira section at Cabo Mondego on the Portuguese coast. The marker is Hyperleoceras mundum. Portugal thus holds two Jurassic golden spikes at Peniche and at Cabo Mondego, and the reason is geological and not political. The Portuguese margin preserves thick, continuous, well-exposed marine sequences of exactly the right age, subsiding steadily while the sediment kept arriving so that the record there accumulated for millions of years at a stretch without a break in it. The third of the Middle Jurassic points belongs to the Bathonian, named by d'Omalius d'Halloy in 1843 after Bath in England. The original definition came from an incomplete section of oolitic limestones in regional quarries.

Incomplete because those quarries had been cut for building material and nobody had ever asked them to record anything. Driven into the hillside where the good free stone lay and abandoned where it ran out. The stage that furnished the honey-colored stone for an entire city was therefore first defined across several partial faces pieced together from whatever the masons had not yet carried away, with the missing intervals filled in by inference. The formal point ratified in 2009 is far from Somerset in the Ravin du Baise at Base Ouran in the Alpes-de-Haute-Provence in France, where the marker is the Ammonite Gonolchites convergence taken at the base of the zone named for Zigzagichera zigzag. Set the seven together and the pattern is hard to miss. Hettang, Grand, Semur-en-Auxois, a hamlet in the Swabian Hills, Toar, Aalen, Bajoc, Bath.

Each name was fixed by a 19th-century naturalist who knew one patch of country better than any other, often because he lived beside it or had walked it as a student. Each modern spike was fixed by a committee comparing sections from a dozen countries and choosing whichever one had the fewest gaps and the most trustworthy shells. Bath and Bas Ouran lie the better part of 1,000 kilometers apart, and the Bathonian is defined at the second and named for the first, which is very close to the situation of every stage on the list. The late Jurassic tells a slower story.

The Kimmeridgian was named by d'Orbigny in 1842 after Kimmeridge on the Dorset Coast and after the thick, dark oil-rich shale known as the Kimmeridge Clay. It had to wait until 2021 for its golden spike, and when the spike arrived, it went a very long way north to the Flodigarry section in Staffin Bay on the Isle of Skye. The definition there rests on ammonites marking the Boreal Bauhinii zone and the Subboreal Bayleysi zone, and that pairing is itself the explanation for the delay. Northern and southern seas held different ammonite faunas, so a marker that works crisply in Scotland may be absent from Mediterranean rock and the other way round. Correlating between those provinces is the hard part, and it took the better part of two centuries to settle. Three Jurassic stages still have no ratified spike at all. Their bases rest on working definitions, which are agreements to proceed sensibly until something better becomes available.

The Callovian is the first of the three. D'Orbigny introduced it in 1852 from Kellaways in Wiltshire, and he put its base at the contact between two named rock units, the Forest Marble and the Cornbrash, a junction visible in the field. The kind of contact a person can follow along a hillside with the eye and never once lose. Later work showed that the contact lies within the upper Bathonian, one full stage below where the base was meant to sit. So the name went on being used while the surface it had been fastened to turned out to belong to somebody else's stage. In practice, geologists now work from the first appearance of ammonites of the genus Keplerites, and they do so knowing perfectly well that the arrangement is provisional. The Oxfordian, named by d'Orbigny in 1844 1844 for Oxford and for the Oxford Clay rests on the proposals of a single worker.

J.R. Kell in 1939 and again in 1946 placed the base at the first appearance of Quenstedtoceras marii, a species then assigned to a different genus. Arkell's placement was never ratified. Later proposals have favored a different marker, Cardioceras redcliffeense, and the discussion continues. It is a good illustration of how much of stratigraphy runs on carefully maintained provisional arrangements rather than final answers, and how comfortable the discipline has learned to be with saying that a matter is not yet settled. The last of the three is the Tithonian, the final stage of the Jurassic, named by Oppel in 1865. Alone among the Jurassic stage names, it does not come from a town, a hamlet, a quarry, or a clay.

Oppel took it from Greek story, from the figure of Tithonus, who was bound up with the dawn, and he chose it because the Tithonian stands at the dawn of the Cretaceous. It is the one poetic name in an otherwise resolutely geographical list. The working marker is the ammonite genus Grevasia and, like the Callovian and the Oxfordian, the Tithonian has no point on any cliff or roadside anywhere that the world has formally agreed to call its beginning. While European geologists were arguing over centimeters of ammonite-bearing marl, the rocks that would supply the popular image of the Jurassic were sitting under the western United States, largely unexamined. The Morrison Formation is a body of upper Jurassic sedimentary rock made of mudstone, sandstone, siltstone, and limestone, colored light gray, greenish gray, and red. Those colors are a record in themselves, written by the environments that laid the sediment down, by how much oxygen and iron and organic matter were present in each patch of mud as it settled and dried. Radiometric dating places the base of the formation at one hundred and fifty-six point three million years, plus or minus two million, and its top at one hundred and forty-six point eight million years, plus or minus one million. Those plus or minus figures come out of decay statistics and laboratory calibration, and what they mean in practice is that the floor of the formation could sit anywhere within a four-million-year window and the measurement would be behaving exactly as advertised. In stage terms, the formation spans the earliest Kimmeridgian into the early Tithonian, which places it squarely among the stages that waited longest for their definitions.

Most of its fossils come from a narrow selection of its rocks, specifically the green siltstone beds and the lower sandstones. Those are the relics of Jurassic rivers and their floodplains, channels that shifted across broad lowlands, sandbars, overbank muds left behind after a flood receded. A river is an excellent burial machine. It moves sediment constantly. It drops that sediment abruptly when its energy falls, and it covers whatever is lying on its bed before decay and weathering can finish the job.

The formation preserves the late Jurassic of North America because late Jurassic North America had rivers doing that work continuously for close to ten million years. The scale is difficult to hold in the mind. The Morrison covers roughly one point five million square kilometers, or about six hundred thousand square miles. It is centered on Wyoming and Colorado, and it reaches out across a dozen states with outcrops in Montana, North Dakota, South Dakota, Nebraska, Kansas, the Oklahoma and Texas Panhandles, New Mexico, Arizona, Utah, and Idaho, while equivalent rocks under different names continue north into Canada. And yet only a sliver of it can ever be seen. More than seventy-five percent of the formation remains buried under the prairie to the east, sealed beneath younger rock and soil and a great deal of farmland.

Much of its Western extent no longer exists at all, having been stripped away during the exhumation of the Rocky Mountains when the range rose and erosion carried the upper layers off piece by piece. What is available for study is the thin margin between those two fates, the exposed edges in canyon walls and road cuts and hillside slumps, where the buried world briefly comes to the surface. Every dinosaur ever collected from the Morrison came out of that narrow band. The formation takes its name from Morrison, Colorado, a small town where Arthur Lakes found some of the first fossils in 1877. Lakes was a school teacher and clergyman with a serious amateur interest in geology, working the hogback ridges west of Denver, where the beds stand tilted on edge against the mountain front. What he uncovered in those tilted sandstones was large enough that he wrote to the leading authorities of the day. That letter landed in the middle of a feud. The year 1877 and the ground around Morrison became the center of the Bone Wars, the collecting rivalry between Othniel Charles Marsh of Yale and Edward Drinker Cope of Philadelphia. The two men raced each other for specimens, for names, and for priority in print, and the rivalry ran on crates.

Bone was dug, boxed, packed in straw and plaster, loaded onto wagons and then onto freight trains, and shipped east in quantities that museums took decades to work through. The haste produced errors, duplicated names, and some genuinely poor practice, along with an enormous body of material that the science is still drawing on today. The result of all of it is that the Morrison stands as the most fertile source of dinosaur fossils in North America. It rests variously on older middle Jurassic units: the Somerville, the Sundance, the Bell Ranch, and the Wannaka. So a section climbing through the sequence passes from an earlier Jurassic world into the river plains that would preserve the later one. The rock had one further career entirely unrelated to bone. In Colorado, New Mexico, and Utah, the Morrison became a major source of uranium ore, and for a time in the twentieth century, prospectors with counters walked the same outcrops that fossil hunters walked for completely different reasons. The chemistry of those old river sands is responsible. Groundwater moved easily through porous channel deposits and dropped its dissolved metals wherever conditions changed. So uranium concentrated in the very beds that concentrate fossils.

Buried bone is porous, too, and takes up whatever the groundwater is carrying. So a number of Morrison specimens will still stir a counter held against them, and more than one prospector's trench went through a rib on its way to ore. Today, some eight thousand kilometers of ocean and continent separate those Colorado hogbacks from the quarries of Southern Germany. In the late Jurassic, the gap was narrower since the Atlantic was young and still opening, but it was still thousands of kilometers of water and land, and the two places were recording the same period by opposite methods. The Morrison buries fast and rough in sand and grit and delivers bones. The Altmühltal limestones of Bavaria, quarried in country roughly halfway between Nuremberg and Munich, gently in fine carbonate mud, and they deliver things that should never have survived at all. In the late Jurassic, this was an archipelago at the edge of the Tethys Sea. Warm water lay in broad sheets between low islands, nowhere deep, and among those islands sat a series of lagoons with only limited access to the open sea.

That restriction made the deposit. Water came in, water evaporated under a tropical sun, and the salt that arrived did not leave. Salinity climbed. It kept climbing past the level ordinary marine life could tolerate until the brine in those quiet basins could not support life at all. Then came the second condition. The lowest water in the lagoons held no oxygen. Dense, salty and undisturbed, it sat on the bottom without mixing, and nothing lived down there to stir it. The floor of a normal sea is a busy place. Worms burrow through it.

Small scavengers arrive within hours of any death and take a body apart with great efficiency, and the traces of all that activity churn the sediment until its fine layering is destroyed. In these lagoons, that entire workforce was missing, not merely thinned to a few hardy specialists, but gone from the bottom water altogether, so that anything arriving on the mud was left in precisely the attitude in which it landed. The consequences arrive on the surface of the water. A dragonfly, blown out over the lagoon on a hot afternoon, comes down and cannot lift off again. It sinks.

Nothing rises to meet it. It settles onto carbonate mud so fine that the mud takes the shape of the veins in the wing, and then more mud drifts down on top, and the wing is held there, flat and complete, the delicate lattice of its membrane still legible. The same sequence caught everything else that came down onto that surface. Fossils from these beds include the wings of dragonflies, the imprints of stray feathers that drifted in from somewhere out of sight, terrestrial plants washed down from the islands, and soft-bodied organisms such as sea jellies. A sea jelly is mostly water. In almost any other setting, it decays into nothing within a day, leaving no trace whatever. Here it left an outline in stone. The lagoons were not always full either.

At times, they nearly dried out, and the exposed carbonate mud lay in the sun as a sticky expanse across the flats. Insects landed on it and stayed. A few small dinosaurs, crossing that ground for reasons no one can now recover, became stuck in it and did not get free. Those animals are among the most valuable specimens in the deposit because they came from the land and were caught by what the sea had left behind. Over six hundred species have been identified from Solnhofen according to counts drawn together on the deposit. That figure includes twenty-nine kinds of pterosaur.

The smallest of those was around sparrow sized, and the implication is worth following. A flying animal that size has hollow bones no thicker than a pin, held together by ligaments that rot within days. Under ordinary circumstances, such a skeleton is scattered and destroyed long before it can be buried. In the lagoons, it was laid out whole on the mud, like a diagram drawn for the purpose. Fossils here are not numerous, and that qualification separates Solnhofen from the Morrison in kind as well as in method. A quarry worker might split limestone for weeks without finding anything at all. But the range of what does turn up, from insects to plants to jellies to reptiles, both flying and walking, gives a comprehensive picture of one local Jurassic ecosystem seen from above and below at once.

Very few places in the whole fossil record offer that, and the reason anyone was splitting the limestone in the first place had nothing to do with fossils. The beds were originally quarried as a source of lithographic limestone. Lithography works on the refusal of grease and water to mix. An image is drawn on the stone in a greasy medium. The surface is dampened. The ink adheres only where the grease is, and paper pressed against it takes up the image. The method demands stone of extraordinary evenness, fine-grained, flat, free of flaws, dense enough to hold a line and uniform enough not to distort it.

A stone flawless enough to carry a printed line is a stone flawless enough to carry the veins in a dragonfly's wing. The two demands turn out to be one demand. The printers of the nineteenth century were after nothing more than a smooth surface, and in getting it, they were quarrying the most detailed page of the late Jurassic that anyone would ever open. In eighteen sixty-one, in that same Bavarian limestone, someone found a feather. It was a single feather isolated in the rock with a shaft and veins and the asymmetry that flight feathers carry.

It was given a name that year, Archaeopteryx, from the ancient Greek words archeos, meaning ancient, and pterux, meaning feather or wing, the ancient wing. In German, it acquired a second name, Urvogel, the primeval bird, and that word has stuck to it ever since. The animal itself lived around one hundred and fifty million years ago in what is now southern Germany, at a time when Europe was not a continent, but an archipelago of islands scattered in a shallow, warm, tropical sea, sitting far closer to the equator than the same rocks sit today. The Bavarian quarries are, in a real sense, the floor of that sea, lifted and dried and cut into slabs. The same year the feather was named, the first complete specimen was announced.

Twelve more fossils have followed since. The feather that started it remains contested, examined, and re-examined for its proportions and for the structures that carried its color, and the argument over whether it belongs to the animal that carries its name has never closed. The name, in other words, may have been given to a fragment shed by something else in the same lagoon, and every skeleton that now bears it was found afterwards. The complete specimens allow a much clearer description.

Archaeopteryx was about the size of a Eurasian magpie. The largest individuals possibly reached raven size, and the largest species grew to about fifty centimeters in length, which is roughly twenty inches, tail included. Most of that length is tail. What sits at the front of it is a light-boned animal with broad wings and, by inference from those wings, an ability to fly or to glide. No one has watched it move. Wing shape and skeletal proportion allow the conclusion, and the conclusion remains a reconstruction laid over a skeleton, tested against the physics that any flying body has to satisfy, but never once observed. What the specimens show more plainly is how little of a modern bird there is in it. Archaeopteryx had more in common with other small Mesozoic dinosaurs than with anything now nesting in a hedge. Its jaws carried sharp teeth. Its wings ended in three clawed fingers.

Behind it ran a long, bony tail, not the short fused stub that modern birds carry. Its second toes were hyperextensible, able to fold up and back in the manner that has earned the informal name of killing claw, a feature it shares with the dromaeosaurids and troodontids, along with feathers and a whole list of quieter skeletal details. Set that inventory beside a living magpie and the resemblance thins to almost nothing. Feathers, wings, a wishbone, a general lightness of build. Everything else in the skeleton belongs to a small predatory dinosaur, and a small predatory dinosaur is what it was. This is where the two flying lineages have to be held apart, because the Solnhofen lagoons preserved both of them in the same quarries, sometimes on the same bedding planes.

Membrane-winged animals had been coming out of that limestone for decades before the feather turned up, so a quarryman splitting a fresh slab had every reason to expect one more of them. Had 1861 produced exactly that, the specimen would have been an unusually pretty example of a familiar kind and nothing more, catalogued, shelved, and mentioned once in a regional journal. What made it explosive was the architecture of the wing. The hand in that slab is short, three-fingered, and tipped with claws, and the flight surface is built of separate feathers rooted along the forearm and the hand, which is the theropod solution and not the pterosaur one. Nowhere on the bedding plane is there a fourth finger drawn out to the length of the arm, and nothing runs back from wrist to ankle waiting to be held open. So the animal could not be filed with the flying reptiles. Neither could it be filed with the birds, because of the teeth, the tail, and the fingers. It sat between a group of ground-living predatory dinosaurs and the birds themselves, carrying the marks of both, and it was published two years after Darwin's book into an argument already looking for exactly such an object. That argument was only possible because the two ways into Jurassic air had been recognized as separate inventions.

Feathers on an arm implied descent from one particular lineage of land animals. A membrane on a finger would have implied nothing of the kind. Everything that has been contested about Archaeopteryx since, and a great deal has, has been contested inside that first distinction. For roughly a century and a half, from the late 19th century into the early 21st, Archaeopteryx held the title of the oldest known bird. It was the specimen in every textbook, the animal that arrived at the exact moment biology needed something halfway between two groups.

That status has since loosened. Older potential avialans have been identified, among them Anchiornis, Xiaotingia, Aurornis, and Bamnorinis, and the Urvogel now shares ground it once held alone. It has not been diminished by that. It has simply been joined. For most of its history, the specimen carried the weight of an entire transition by itself because nothing else of the kind had been dug up, and the questions asked of it were correspondingly large and correspondingly unanswerable. That weight is now spread across a dozen animals from two continents, and the questions have become narrower and better. Not whether such a creature could exist, but exactly where among its many relatives this particular one belongs. From Solnhofen, the period runs on for a few million years more into rock that no one has yet agreed how to label. The Jurassic closes at about one forty-three point one million years ago, and here the whole apparatus of stages and zones and golden spikes arrives at a problem it has not solved. The boundary between the Jurassic and the Cretaceous is the only boundary between two geological periods that remains formally undefined. The reason lies in the rock itself, and it is a real difficulty.

Ammonite correlation had been under strain for a long time before this. The Kimmeridgian waited until twenty twenty-one because northern and southern faunas had to be stitched to one another across a shortage of shared species. But through most of the period, the stitching was at least possible. Somewhere in the sequence, there was always a genus that occurred on both sides of the divide and could serve as a hinge between them. At the close of the period, that overlap thins away to nothing. The Boreal, Subboreal, and Tethyan realms come to hold ammonite faunas that barely acknowledge one another, and a species that appears to define the boundary in one basin is absent from the next or turns up at a different level relative to everything around it, and the correlation on which the entire scheme depends comes apart in the hand.

Nor is there a rescue from chemistry. At other boundaries, a sharp shift in carbon isotopes or a comparable chemo-stratigraphic event provides a line that can be traced globally, independent of what happened to be living where. The Toarcian crisis left exactly that kind of signal. At the top of the Jurassic, there is nothing of the sort to anchor to. The rock changes from basin to basin without leaving a signature that all of them share. So the search has moved to smaller organisms. The most promising candidates are the calpionellids, an enigmatic group of planktonic protists with urn-shaped calcitic tests, tiny hollow vessels of calcium carbonate drifting in the open ocean.

Because they drifted rather than swam, they scattered very widely, and that breadth is the first thing a global marker has to have. They were briefly abundant across the latest Jurassic and earliest Cretaceous, and then they were not, and that brevity is the second. The first appearance of one species, Calpionella alpina, coinciding with the base of what is called the alpina subzone, has been proposed as the definition. Against that stands the long-used working marker, an ammonite named Strambegella jacobi, formally placed in the genus Beri aquella. It has served for decades as the practical answer to the question of where the Jurassic ends. Its position has been questioned because its first appearance does not correlate with that of Calpionella alpina, and two markers that disagree cannot both define the same line. The matter is unresolved. Somewhere in a quarry face, at a level that shifts by meters depending on which authority is consulted, the Jurassic becomes the Cretaceous. The period does not finish with a click. It thins out, hands its ammonites and its warm, shallow seas across to the Cretaceous, and leaves the paperwork open. It does not, however, end in irrelevance. The Jurassic sits under the modern world in an extremely practical sense, and a great deal of it is being burned. The Kimmeridge Clay, the dark mudstone named after that stretch of the Dorset Coast and its equivalents elsewhere are the major source rock for North Sea oil.

The organic matter that settled into oxygen-poor Jurassic seafloor mud, protected from decay by the same absence of scavengers that preserved Solnhofen’s dragonflies, was buried, heated over tens of millions of years, and slowly converted into hydrocarbon. Every platform standing in that gray northern water is drawing on a late Jurassic seabed. The pattern repeats on a larger scale further south. The Arabian intrashelf basin, laid down through the middle and late Jurassic, hosts the world's largest oil reserves, including the Ghawar Field. In Iraq, the Saghelu and Neokhelikhun formations do comparable work as source rocks.

These are not incidental facts about the period. A significant fraction of the energy that moved the twentieth century came out of Jurassic marine mud. Coal tells the terrestrial half of the same story. More than one thousand five hundred gigatons of Jurassic coal lie in Northwest China, chiefly in the Turpan-Hami and the Ordos basins. That figure represents forest compressed. The conifers and ferns and bennettitales that covered the Jurassic land, generation after generation falling into swamp and standing water where decay could not finish them, became seams thick enough to be measured in gigatons.

And one further Jurassic feature sits in the ground with no economic use at all. Beneath the sand of the Kalahari, buried and invisible from the surface, lies the Morokweng impact structure, a crater seventy kilometers across. Nothing of it shows at ground level. No rim, no ring of hills, no bowl in the land, only flat pale sand and scrub in every direction. Something struck Southern Africa hard enough to leave a wound that wide, and the desert has covered it so completely that it was found by measuring the magnetic field above the sand and then drilling down to see what was making the anomaly.

One principle joins all of this to the lagoons. Decay is the ordinary fate of organic material, and decay requires oxygen and scavengers. Remove those two things and carbon stays where it fell. Anoxic lagoons, drowned Jurassic forests, North Sea source rocks, and the seams of the Ordos are the same process running at different scales. A still lagoon floor holds one dragonfly wing intact for a hundred and fifty million years, and a subsiding basin holds unmeasurable tonnages of dead plankton for the same length of time by the same means and for exactly the same reason. Seen from a distance, the period has the shape of something opening, and the opening can be measured with a hand. Lay a palm flat against a Jurassic shelf limestone, and the thickness under it stands for a few hundred years of quiet chemistry. Carbonate coming out of warm water, grain by grain, a little faster in the hot months, a little slower in the cool ones, with nothing dramatic ever happening to it at all.

Stack enough of those handspans, and the sequence runs to hundreds of meters, and even then it is only one basin's share of one period. The whole Jurassic is that arithmetic repeated across a splitting world, an ocean let out of a seam so gradually that nothing swimming in it could have sensed the widening, a polar wood laid down needle by needle through a hundred thousand winters of darkness, a coal seam gathered one drowned trunk at a time, and a stage boundary made by a single shell arriving in a single bed on an afternoon that nobody watched.

The people who came later read it at the same patience and in the same direction, from the coarse toward the fine, until an entire period could be pinned to a surface no wider than a handspan, and the reading is still unfinished at the top, where three stages are waiting for somebody to choose where they begin. Now let all of it go, slowly, one thing at a time. Start with the limestone, since that is where the name began. Take the word off it. Let the beds in the Jura be simply the local stone again, the material of field walls and barn footings and the low arch of a bridge over a stream, gray with lichen on the weather side, damp under a hand at dawn. Let the trees close over the road cut. Let cattle move across the pasture above, in and out of the shade, and let the hillside go back to being a hillside that a person walks over on the way to somewhere else, holding no title, settling no dispute, older than the argument and indifferent to it. The young man with the hammer walks on and out of the story.

His mistake goes with him, unembarrassed. Now the water where the shells were made. Put the sea back over the shelf, shallow enough for light to reach the bottom, warm as bath water in the long afternoons, moving in a slow swell that never quite breaks. Let the coiled shells go back into it and go back to being animals rather than instruments. Whatever they were doing, they are doing it again now, out past the reach of any tank or any lamp, hanging in the green light, tentacles gathered, chambered spirals turning slowly as the current takes them, rising a little when the day cools and settling again toward the dark. Nobody is measuring where they first appear, and nothing at all now depends on when they stop. They rise and drift and sink and eventually settle, and the mud closes over them without any ceremony at all. Now the lagoons in the islands. Set the water back exactly as it was, still and heavy and oversalted, warm at the surface, dead and quiet below. The sun goes across it all day without a ripple.

In the evening, the light comes low over the flats and turns the whole basin the color of weak tea, and the salt crust at the margin holds that color for a while after the sun has gone. Nothing moves in it because nothing can. Whatever comes down onto that surface, blown or drifting or falling, goes under and is held. The mud is soft as flour and takes the shape of everything. Let the wedges and the chisels be set down at the quarry face. Let the last cut slab stay unopened in the hillside, the wing and the feather still inside it, unread, in the dark where they have always been.

Above that water, in the warm air over the low islands, something small and toothed and clawed moves between the trees, and it has no name yet. Nothing depends on it. It drops into the leaf litter after something smaller still, mantles over it with the wings half spread, and feeds. It is simply an animal of that afternoon, going about the ordinary business of an afternoon, and then it is gone from view behind the foliage and does not come back out. Now the rivers, half a world west. Return the water to the channels and let the sediment be soft and moving again, sand shifting along a bar, a slow brown flood spilling out over the flats after rain and then draining away. The floodplains carry fern and horsetail and the fallen litter of cones, and the air over them smells of wet silt and resin, and there is nobody there to notice it. Animals cross the shallows and leave prints that the next rise of water erases.

One of them, in the ordinary course of things, does not get up, and the river covers it before morning without any particular attention. Then let the whole plain go under, prairie soil, a farm road, a hundred and fifty million years of quiet. Most of it was never dug and never will be, and there is rest in that thought, a whole buried country lying under the wheat with nothing whatever asked of it. Now the shape of the world. Bring the two great land masses back toward one another until the ocean between them narrows to a green channel, then to a rift valley with lakes in the floor of it, then to nothing but a seam in a single continent. Let the poles hold forest instead of ice, standing dark through the long night, breathing out slowly. Let the interior go dry and quiet as it was, with the dust moving across it and no coast within reach in any direction, and let the whole span close over the way water closes over a dropped stone, unhurried, taking its own time to still.

Fifty-eight point three million years, finishing at their far edge without a line. What is left at the end of all that giving back is a very small scene. Imagine a single candle standing beside a slab of lithographic limestone on a workbench. The slab is pale and fine-grained and perfectly flat, split cleanly along a bedding plane that was once the floor of a lagoon nothing could live in. The flame is low. Its light runs sideways across the stone, the way quarrymen and preparators have always held a lamp, because raking light is what makes a faint impression visible. In that shallow angle, the surface gives up its detail, the thin lines of a wing, the fine rows of a feather, the coiled edge of a shell that arrived here from a sea that no longer exists. Look at it a moment longer, then blow that candle out. The stone goes dark, and the impression on it becomes what it has been for a hundred and fifty million years, which is a shape in rock that no one is looking at. The lagoon holds still. The rivers stay under the prairie. The shells keep coiling in beds that nobody is splitting tonight. The wooded hillside stands above its village with the trees on it, and the period settles back into its own long span, unhurried, undisturbed, and still missing one line at its upper edge, which somebody someday will draw. Rest well, sleep deeply, and let the ancient earth fade softly into dreams.