After the Dinosaurs: The Age of Mammals

1

Introduction

Fossil hunting is by far the most fascinating of all sports. It has some danger, enough to give it zest and probably about as much as in the average modern engineered big-game hunt, and the danger is wholly to the hunter. It has uncertainty and excitement and all the thrills of gambling with none of its vicious features. The hunter never knows what his bag may be, perhaps nothing, perhaps a creature never before seen by human eyes. It requires knowledge, skill, and some degree of hardihood. And its results are so much more important, more worthwhile, and more enduring that those of any other sport! The fossil hunter does not kill, he resurrects. And the result of this sport is to add to the sum of human pleasure and to the treasure of human knowledge.

George G. Simpson, Attending Marvels, 1934

Finding Fossils

The sun blazed down on the two men as they slowly walked up and down the ravines of the badlands. They walked stooped over with their eyes glued to the ground. The temperature was over 104°F (40°C), and there was no shade anywhere in the desolate landscape. They had been working like this all day and yet had only a few fossil jaws and teeth to show for their time and effort. Wide, floppy hats and loose, light-colored clothing kept off the sun, but they dared not wear dark glasses, despite the glare from the ground. To find the fossils they were seeking, they needed to detect subtle differences in the color and surface texture of the rocks on the ground, and dark glasses made this difficult. Many of the things they picked up were shiny black pebbles or concretions that resembled fossils. Frequently, they found chunks of fossil bone, which were clearly identifiable by their spongy texture in cross-section. Most of these pieces of bone were too broken to be identified. Others were scraps of fossil turtle shell, which had little scientific value. Occasionally they got lucky and found an isolated mammal tooth or two. These were worth saving, since the pattern of the tooth crowns of most mammals is distinctive. Fossil teeth are sometimes easy to spot, for instance, when the tooth enamel is black and shiny and stands out on the baked tan muds.

The men were hoping to find remains of the largest animals of the Eocene, the elephant-sized brontotheres, which were distantly related to horses and rhinos but had two blunt battering-ram horns on their noses (fig. 1.1). If the men were really lucky, they might find two or more brontothere teeth together, or a partial jaw with three or more teeth in it. Even a complete jaw and skull of a common animal, however, is not as valuable as a single tooth of a rare animal, which may be known only from a few scraps. Every isolated tooth of a rare fossil gets immediate attention when it is brought back to a museum. Sometimes it is described and published before anything else in the collection.

The two scientists were in luck today. One stooped down and noticed a small pile of bone fragments (fig. 1.2). In the midst of the pile, the skull and lower jaw of a fossil mammal protruded from the ground, lying on its side. Although the skull and jaw were nearly complete, they did not cause a lot of excitement. They belonged to a common fossil mammal, an oreodont (discussed in chapter 5), which must have roamed this area in herds of thousands over 30 million years ago (fig. 1.3). Oreodonts have no living descendants; they are distantly related to camels, yet they looked nothing like today’s ships of the desert. Although there were already hundreds of unstudied oreodont specimens back in the museum, this oreodont skull was worth collecting because it was so complete.

The collectors carefully dug a trench around the specimen until it rested on a pedestal of rock. Since the specimen was fragile, they made a cast of plaster bandages around the skull. Once the cast had dried, they carefully pried it up and turned it over. The skull had come out in one piece without breaking! After a few more strips of plaster bandage had been wrapped around the exposed surface, it was ready to carry back to the truck.

A complete oreodont skull was a good day’s work but nothing to write home about. As the men were working their way back to the truck, however, one of them spotted another ridge of fossil bone protruding from the ground. Although only a few inches were exposed, the thickness and curvature suggested that it was the back of a large jaw (fig. 1.4). A few minutes of careful excavation of the exposed part revealed that the specimen was indeed a very large one and that it continued into the hillside. The two men returned to the truck and carefully drove it up to the ravine as close as four-wheel drive could reach. First, they used the heavy-duty truck jack to lift a huge slab of sandstone from over the specimen and slide the slab off the cliff. Then they used picks and brooms to carefully dig a trench around the specimen, exposing it on all sides. When they were done, they could see that they had a complete set of the lower jaws of a fossil brontothere (fig. 1.1). The jaws were almost two feet long and in excellent condition, but still fragile. With all the surrounding rock, the specimen weighed several hundred pounds, so it could not be moved easily. To protect it for transport, the two scientists mixed up a small tub of plaster of Paris and tore burlap bags into strips. After dipping the strips into the plaster, they smoothed them over the fossil, overlapping each strip so that a solid bandage was formed. After about half an hour, the plaster jacket was finished and drying quickly in the hot sun. Next came the hard part. The jacket surrounded the specimen on nearly all sides, but it was still attached to the ground. More digging isolated the plaster jacket on a higher pedestal of rock. Carefully, the two scientists dug the pedestal from underneath the specimen. At last, they wedged the pickaxe underneath the cast, prying it from the ground and flipping it over. The underside of the jaw was revealed in almost perfect condition, with very few broken pieces. After carefully trimming the ragged edge of the jacket, they covered the exposed side with more plaster and burlap. This brontothere was ready to be transported to the museum for study.

Figure 1.2. (A) Digging an oreodont skull from the ground and (B) covering it with a plaster bandage. Photos by the author.

Figure 1.3. Reconstruction of the oreodont Merycoidodon. Although about the size of a sheep, it was more closely related to camels and was the most abundant animal in the Eocene and Oligocene beds of the Big Badlands. Painting by B. Horsfall, in Scott 1913.

Figure 1.4. Digging out and jacketing a brontothere jaw from the Eocene badlands near Lusk, Wyoming. (A) The jaw is fully exposed and trenched by many hours of hard labor with a pick. The skull is just visible in the quarry face behind the jaw. (B) The jaw is now covered by a thick jacket of plaster to protect it from damage during transport. (C) Once the jacket dries, the jaw is pried from the ground with a pick and turned over, so the underside is exposed and ready for the final plaster jacket. Back in the laboratory, the preparator will cut open the jacket with a saw (the same kind used by doctors to cut casts from broken bones) and then carefully scrape away all the sediment while preserving the bone. Photos by the author.

Not all fossils are so large or glamorous. In some areas, the fossils are so small that they cannot be seen from more than a foot away. The only way to collect them is to crawl on your hands and knees, with your eyes six inches from the ground. If the ground is rich in small teeth and bones, it is more efficient to use a large crew of students or volunteers. The greater the number of trained eyes covering the ground, the better. In such deposits, a few teeth are considered an excellent find since the fossils are badly crushed and seldom yield a complete skull. However, these tiny, isolated teeth are important because, for most mammals, teeth are our only record of their early evolution.

If fossil hunting sounds like grueling, backbreaking work, it is. Most fossil hunting bears little resemblance to the glamorous misconceptions we see in the movies. Scientist who study fossils, paleontologists, must put up not only with difficult conditions but also with days and weeks of looking without finding anything. To persevere in the face of such disappointment and discouragement, paleontologists must really love their work. However, one excellent find in a field season is often enough to make thousands of hours of toiling in the sun worthwhile.

Many a youngster has dug large holes in the backyard, unsuccessfully looking for the dinosaurs from the children’s books. How do paleontologists know where to dig? First of all, they must know where to look. Fossils are nearly always found in sedimentary rocks, which are formed from sand or mud or fossil shells. Only a small fraction of the earth’s sedimentary rocks carry fossils, so it helps to look in rocks that are known to be fossil bearing. Rock strata that were laid down in the ocean rarely produce fossils of land animals. Only sandstones and mudstones that were originally sands and muds on a river floodplain or in a lake will yield fossils of land mammals or dinosaurs. The rocks also have to be of the correct age. If they are more than 65 million years old, they will not produce many mammals, but they might produce dinosaurs. If the rocks are younger than 65 million years, however, no dinosaurs will be found, since they all became extinct at that time. Paleontologists must take all these factors is into account when they study the geology of an area, or learn of a fossil locality from some other collector.

Once you’re in the right place, you have to know how to look. Slowly scanning the ground a few inches at a time is a suitable pace, even if it takes tremendous patience. Finally, you have to know what to look for. Paleontologists develop a mental filter, known as a search image, that screens out all the nonfossils and fossil-like objects they see. Only the genuine glint of enamel or spongy texture of bone catches the eye among all the objects on the desert floor. Once paleontologists have spotted bone or enamel, they must also have the training to recognize and identify what they’ve found. If it’s really worthwhile, it deserves special treatment. To develop this kind of skill generally requires years of education and many more years of practice in the field, collecting and identifying hundreds of specimens. Since most finds are fragmentary, paleontologists must know the skeleton of each animal so well that any piece is instantly recognized. Only a few of the handful of paleontologists employed today have all these skills so well developed that they are master collectors. Good fossil collectors are a rare breed these days, but what they have found is extremely impressive considering their small numbers. From their years of collecting, we have fossils in museums that tell us the story of the evolution of dinosaurs, elephants, horses, rhinos, and many other important fossil animal groups.

These methods are standard for collecting fossil vertebrates (animals with backbones, like fish, reptiles, amphibians, birds, and mammals), which are generally rare and difficult to collect. By contrast, invertebrates (animals without backbones) are generally much more common—at least those with hard shells or skeletons, like clams, snails, sea urchins, and corals. Obviously, soft-bodied animals without skeletons, like worms and jellyfish, seldom fossilize. In many places, fossils occur as dense shell beds with thousands to millions of shells packed in close together. Here, collecting is much easier, and the collector need worry only about damaging the fragile shells as they are collected, and about keeping good records of everything that is collected. More often, however, marine shales and sandstones have relatively few fossils, so collecting in these locations is the same kind of backbreaking work I have just described, hiking over miles of landscape, looking for the rare shell.

Yet another set of conditions applies to microfossils, the skeletons of tiny organisms usually less than a millimeter in size (fig. 1.5). Most microfossils are the shells of single-celled organisms, such as the amoeba-like foraminifera and radiolaria that float in the plankton and settle on the sea bottom. Other microfossils come from plant-like single-celled organisms (such as diatoms). Still others are from multicellular animals that happen to be microscopic in size, such as the tiny snail-like pteropods that float in the plankton, or the minute crustaceans known as ostracodes, which litter the sea bottom with trillions of their tiny kidney-bean-shaped shells that hinge over their backs. In any case, microfossils are usually not rare. Some oceanic sediments are composed of nothing but microfossils, so even a sample of a few grams yields thousands of shells. In most marine sediments around the world, microfossils are abundant, so the experienced micropaleontologist need scoop only a few grams of sample into a bag, take it back to the lab, and look through the microscope. Better still, microfossils are so abundant that they can even be recovered from samples drilled from deep underground in the search for oil. For years, oil companies hired micropaleontologists because they could use the tiny microfossils found throughout the deep drill holes to determine how old the sediment was, or how deep the water once was at the site of deposition. In addition, many microfossils are sensitive to the oceanic conditions in which they lived. They often track changes not only in water depth but also in oceanic temperature and chemistry. As we shall see later in this book, the study of microfossils and the chemicals trapped in their skeletons is the key to understanding how ancient oceans and climates have evolved over time.

Figure 1.5. Microfossils. The large shells made of bubble-shaped chambers are planktonic foraminifera, and the smaller fossils with the mesh-like skeletons are radiolaria. Photo courtesy of Scripps Institution of Oceanography.

Dating Rocks

Paleontologists work in a world with a time frame completely different from ordinary everyday history. From various methods, we now know that the earth is about 4.6 billion years old, a staggering number in human terms. It is such an immense amount of time that some sort of analogy is necessary to make it comprehensible. Suppose we were to compress all 4.6 billion years of earth history into a single calendar year. On this scale, each of the 365 calendar days equals 12 million years, and each minute of the calendar is 8561 years long! The earth forms on New Year’s Day in this calendar. The first recognizable life—consisting of tiny, single-celled bacteria and blue-green cyanobacteria—does not appear until February 21. Complex, multicellular life, such as jellyfish, trilobites, and corals, does not appear until November 12. The first amphibians crawl out on land on November 28. The first tiny mammals and the first bird, Archaeopteryx, appear during the peak of the Age of Dinosaurs, the Jurassic Period, on December 17. The final extinction of the dinosaurs and the beginning of the Age of Mammals occur on the day after Christmas. The first ape-like primates that are members of our own family, the hominids, do not appear until eight hours before New Year’s Eve. Neanderthal Man, the classic Stone Age caveman, appears ten minutes before New Year’s Eve, as the countdown begins at parties everywhere. Recorded history begins less than one minute before New Year’s Eve, as the conductor raises his baton to start Auld Lang Syne. Within a second before midnight, Charles Darwin’s On the Origin of Species is published, and the American Civil War is fought. Virtually all of human history, especially the last few millennia, is drowned out by the drunks who blow their noisemakers a fraction of a second too early!

On the scale of geologic time, human affairs appear pretty insignificant. Geologists are accustomed to dealing with such large amounts of time and routinely deal with thousands and millions of years. For most geologic problems, events of less than thousands of years in duration cannot even be distinguished in the layers of sedimentary rocks. When dealing with events that occurred hundreds of millions or billions of years ago, even a million years here or there is negligible. A sense of deep time (as John McPhee labeled it) is important to all of us, not just to the geologists. Most geologists, however, find it practical to deal with time not in millions of years but in relative time terms. Just as historians use Elizabethan or Edwardian to refer to periods in English history, so geologists use Cambrian and Cretaceous to refer to distinct episodes in earth history.

For the purposes of this book, most of these time terms will not be necessary. The last 65 million years, known as the Age of Mammals in popular parlance, is formally known as the Cenozoic Era. The Cenozoic is divided into a number of epochs (fig. 1.6), beginning with the Paleocene approximately 65 million years ago and running to the present. The Paleocene, which lasted from 65 to 55 million years ago, is followed by the Eocene (55–34 million years ago), the Oligocene (34–23 million years ago), the Miocene (23–5 million years ago), the Pliocene (5–1.8 million years ago), and the Pleistocene, or ice ages (1.8 million years to 10,000 years ago). The period since the last retreat of the glaciers, which includes the present interglacial warming, is called the Holocene, or Recent (10,000 years ago to present). Although these terms may seem intimidating at first, using them is much easier than trying to talk about the age of an event in terms of millions of years.

Figure 1.6. Cenozoic timescale. Abbreviation: Quat. = Quaternary.

How did we establish these divisions, and where did these terms come from? Since the late 1600s, geologists have been able to establish the relative ages of fossils and rocks (i.e., this fossil is younger than or older than that fossil) by the principle of superposition. First proposed by the Danish physician Nicolaus Steno in 1669, this principle states that in any layered sequence of rocks (layered sediments or lava flows), the oldest rocks are at the bottom of the stack, and the rocks get progressively younger as you move up the pile. Clearly, the rocks at the top of the stack could not have accumulated unless there were already rocks on the bottom of the stack to build upon. A good analogy is a stack of papers on a messy desk. Those at the top were put there recently, whereas those at the bottom of the stack may have been laid there months ago and have been gradually buried by more recent activity.

The next breakthrough came in the late 1700s, when geologists began to try to decipher the superposed stacks of sandstones, shales, and limestones in England and Europe and to reconstruct the history of the earth. Some thought that the entire stack was produced during the biblical creation week and then modified by Noah’s flood. In 1760, Italian geologist Giovanni Arduino referred to the ancient granitic rocks and metamorphic rocks found at the bottom of the stack in most places in the world, and in the cores of uplifted mountain ranges, as Primitive or Primary, since they were supposedly produced in the original creation of the earth. Above the Primary rocks were layered sedimentary rocks, usually tilted and deformed and found in mountain ranges, which were called Secondary and were supposedly produced as Noah’s flood retreated from the mountains. Above these were Tertiary rocks, which were still horizontal and often poorly consolidated, supposedly produced from the final stages of the retreat of Noah’s flood. Of these terms, only Tertiary survives in modern timescales as a term for the first 63 million years of the Cenozoic, although some authors have tried to replace it with a less archaic term. For example, many geologists prefer to use Paleogene for the first 42 million years of the Cenozoic (the Paleocene, Eocene, and Oligocene epochs) and Neogene for the last 23 million years (Miocene, Pliocene, and Pleistocene epochs). However, the terms Tertiary and Quaternary persist, even though their flood-geology connotations are no longer considered valid.

Flood geology began to break down when geologists looked closer at the rocks. Soon they began to find supposedly Primitive granites that had once been molten and had cut across and intruded through Secondary sedimentary rocks, showing that the granites had to be younger than the sedimentary rocks. In addition, sandstones, shales, and limestones all over Europe looked similar, so matching them up from one place to another was difficult. The next breakthrough occurred in the 1790s, when William Smith, an untutored engineer for a canal company in southern England, began collect fossils from the fresh canal excavations. He noticed that each rock formation had its own distinctive suite of fossils and that no two formations had identical fossil contents. He soon became so good at recognizing this pattern of faunal succession that he could amaze the wealthy gentlemen-collectors by telling them exactly where their fossil collections came from. More importantly, faunal succession helped him map the rock formations and determine their precise sequence, because each sandstone or shale or limestone had a different fossil assemblage from the formations above it and below it. By the 1820s, geologists had mapped most of the formations of England and Wales on the basis of their fossil content and had begun to coin the terms, such as Carboniferous and Cretaceous, that make up the geologic timescale we used today.

But faunal succession is not enough. The sequence of strata in southern England is fairly thick and complete, but there are still gaps in the record, known as unconformities. Even the mile-thick pile of sedimentary rocks in the Grand Canyon represents only about 25% of the time between 250 and 550 Ma (mega-annum, or million years before present), and none of the time before or after. Nowhere on earth is there a complete record that spans all of geologic time. Thus, geologists had to use faunal succession to correlate rock sequences from one place to another. This practice of using fossils to correlate strata is known as biostratigraphy. Each distinctive fossil assemblage is unique to a given period of geologic time and can be used to correlate one local stratigraphic section with another. For example, the Cenozoic sequence in the Rocky Mountain region is the most complete terrestrial sequence anywhere in the world, but nowhere is it complete. Over a century ago, paleontologists had to patch together local sections from different areas to give a complete timescale of mammal evolution in North America (fig. 1.7). For example, the upper part of the section in the San Juan Basin of New Mexico overlaps in age and fossil content with the Wasatch Formation in Wyoming; together these two sections give us a composite section spanning most of the Paleocene and early Eocene. The upper Wasatch Formation, in turn, overlaps with the lower part of the Huerfano section in Colorado, and the upper part of the Huerfano section overlaps in fossil content and age with the base of the Bridger Basin section in Wyoming. The top of the Bridger Basin section, in turn, overlaps the base of the Uinta Basin section. These sections, knitted together over a wide region using successions of land mammal fossils, represent an almost complete record of most of the Paleocene and Eocene in North America. By correlating these and several other sections across the region, we can get a detailed picture of the geological, climatic, and faunal events for this span of time.

How was the modern geological timescale developed? Unfortunately, it was a rather haphazard, unplanned process. The scale was not set up by a single person in a systematic fashion, so that everything would be organized and represent a complete sequence. The time terms were proposed for distinctive rock units at different times and places by different geologists, so the timescale just grew and evolved. For example, the now familiar term Jurassic was named by the explorer and naturalist Alexander von Humboldt in 1795 for the distinctive sequence of rocks in the Jura Mountains of the French Alps. The Cretaceous was named by William Conybeare and William Phillips in 1822 and is based on the Latin word creta, or chalk, since the Cretaceous beds include the famous chalk deposits of the White Cliffs of Dover. As we saw, the term Tertiary (third in Latin) was left over from Arduino’s flood geology of 1760, but in 1829, Paul Desnoyers proposed the term Quaternary (fourth in Latin) for the poorly consolidated post-Tertiary deposits of the Seine Basin in France. At the time, they were thought to be deposits formed after Noah’s flood; but by 1837, Louis Agassiz was attributing them to a great ice age, and since then the terms Quaternary and ice age have been closely linked.

Figure 1.7. Diagram showing the correlations of the deposits of the Rocky Mountain basins, with the temporal overlap between deposits of different basins shown (based on the biostratigraphy of similar mammals, shown in the righthand column). Most of these correlations still hold today, a century after they were pieced together, although what was then called basal Eocene is now Paleocene, and the Titanotherium beds are now late Eocene, not Oligocene, in age. From Osborn and Matthew 1909.

By the 1830s, geologists noticed that there were dramatic differences between the oldest strata (then known as Transition and Carboniferous), with their peculiar fossils of brachiopods and corals and sea lilies, and the younger strata (already divided into the Triassic, Jurassic, and Cretaceous), which were full of ammonites. In 1838, Adam Sedgwick applied the term Paleozoic (ancient life in Greek) to these oldest fossiliferous rocks, which would soon be divided by him and by geologist Roderick Murchison into the Cambrian, Silurian, Devonian, Permian, and so on. In 1840, geologist John Phillips wrote an article for the Penny Cyclopaedia in which he used the term Mesozoic (middle life in Greek) for the ammonite-bearing beds of the Triassic, Jurassic, and Cretaceous, and the term Cenozoic (recent life in Greek) for the younger beds without ammonites that had been called Tertiary and Quaternary. This three-fold division of the fossil record into Paleozoic, Mesozoic, and Cenozoic eras was no accident, because the great Permian extinction at the end of the Paleozoic wiped out 95% of species on earth. This extinction radically changed the life on the seafloor that arose in the Triassic, producing a very different looking Mesozoic fauna. Likewise, the Mesozoic and Cenozoic are bounded by the second largest extinction known, the Cretaceous-Tertiary extinction. This event is abbreviated K/T in geological shorthand, because on geological maps Cretaceous is abbreviated with a K (from the German Kreide for chalk; the C was already preempted by the Carboniferous). The T is for Tertiary. (Recently, a number of geologists have advocated replacing K/T with K/P for Cretaceous-Paleogene, because the Paleogene has now been formally defined and Tertiary is an obsolete usage. However, in this book I will continue to use the more familiar abbreviation K/T.) The K/T event wiped out not only the ammonites but also the great marine reptiles, and the dinosaurs on the land (discussed further in chapter 2).

In contrast to this chaotic growth of most of the timescale, pioneering geologist Charles Lyell attempted to subdivide the Cenozoic in a planned, logical fashion. In the third volume of his revolutionary work Principles of Geology (1833), Lyell tried to replace Tertiary and Quaternary with a finer-scale subdivision of the Cenozoic based on the percentage of recent molluscan fossils in the fauna. The French conchologist Gérard-Paul Deshayes had studied more than 8,000 species (40,000 specimens) and noticed that the mollusks look more and more modern in younger strata. Lyell used this work to propose four periods (now called epochs) for the Tertiary. The Eocene (dawn of the recent in Greek) had only 3.5% of living mollusks; the Miocene (less recent in Greek) contained 17% modern species; the older Pliocene (more recent in Greek) had 33–50% modern species; and the newer Pliocene had 90% living mollusks in its fossils.

Rudwick (1978) has shown that Lyell was thinking of the change in molluscan fossils as a continuously ticking clock (fig. 1.8), so that by identifying the percentage of modern species in a fossil collection, one could subdivide the Tertiary into many fine numerical increments. But in practical terms, the system was flawed. First of all, molluscan turnover was not a continuous clock-like process but rather an episodic one, with periods of stability and mass extinction (as we shall see in later chapters). Second, Lyell and Deshayes’s species are difficult to use today because some have been combined, and other species have been split into many species by later scientists, or raised to higher (generic) rank. Even with up-to-date species lists, Lyell’s molluscan clock would be hard to use. Stanley et al. (1980) calculated that only 50% of the molluscan species (by modern definitions) were in existence at the beginning of the Pliocene, only 5% existed at the beginning of the Miocene, and almost none were present in the Eocene.

Lyell’s noble attempt at a logical, clock-like subdivision of the Cenozoic had a bigger problem: it was not compatible with the system of subdividing geologic time that was already in existence. The rest of the timescale was built by biostratigraphic analysis of local sections, so the Lutetian Stage of the Eocene is based on a set of strata in the Paris Basin with a distinctive assemblage of mollusks. Lyell’s clock model does not mesh well with this system. Rather than placing discrete boundaries on real rock units in the field, Lyell’s clock model had no real boundaries, only arbitrary subdivisions of a continuum of molluscan fossils. In Lyell’s mind, the Miocene was not a division of time between 5 Ma and 23 Ma (as we now define it) but a discrete moment when approximately 17% of the molluscan species were modern forms. Thus, there were no precise boundaries for his units. His system baffled geologists who tried to apply traditional stratigraphic methods to an essentially chronological concept. Lyell indicated that several areas and their fossils were typical of each of his periods, which led to much confusion as later stratigraphers quarreled over what typified the Eocene or Miocene. Although Lyell’s concepts originated in the Italian Tertiary section, Deshayes’s collections were from the Paris Basin. Much of the Paris Basin fauna is restricted to that area, so the type fauna is difficult to recognize outside France.

Figure 1.8. Lyell’s conceptions of the Cenozoic epochs as moments on the clock of molluscan turnover. The numbers indicate the percentages of modern species in fossil collections from each epoch. After Rudwick 1978.

For this reason, paleontologists argued for more than a century about how to define and subdivide the Eocene, Miocene, and Pliocene. In the Paris Basin, the upper Eocene was poorly fossiliferous and had been labeled the lower Miocene by some geologists. In 1854, Heinrich Ernst von Beyrich coined the term Oligocene (few recent in Greek, because there were fewer recent fossils than in the Miocene) for a sequence of rocks in northern Germany and Belgium that were more fossiliferous and apparently younger than the upper Eocene rocks of the Paris Basin. The fossils of von Beyrich’s Oligocene were more advanced than those of the French Eocene, but not as modern as those of the Miocene, so this epoch was placed between Lyell’s Eocene and Miocene. Unfortunately, the type Oligocene is in a different basin and does not overlie the type Eocene, so it is difficult to decide where one ends and the other begins.

The origin of the Paleocene was similarly confusing. In 1874, paleobotanist W. P. Schimper recognized a series of fossil plants in the Paris Basin that he decided were distinct from those of Lyell’s Eocene. He called these Paleocene (ancient recent in Greek). Unfortunately, fossil plants are relatively rare and difficult to correlate around the world, so the term Paleocene did not catch on until the mammals and marine mollusks had also been studied and compared. Most works published in the early twentieth century still used lower Eocene for what we now call the Paleocene (e.g., fig. 1.7), so the reader must be careful when interpreting these early figures and texts. The United States Geological Survey did not formally recognize the Paleocene until 1939. However, since that time, the standard epochs of the Cenozoic—Paleocene, Eocene, Oligocene, Miocene, Pliocene, Pleistocene—have become internationally accepted and are the most useful way to subdivide the last 65 million years.

These are relative time terms, established based on local stratigraphic sections in Europe and correlated with their molluscan fossils. How do we correlate them outside Europe? How do we establish their numerical age? Each of these problems was a major field of study unto itself, and the answers have been slow in coming; there were many false starts before geologists arrived at methods that can date almost any Cenozoic rock around the world.

The first problem is correlating the classic Eocene, Oligocene, and Miocene rocks of western Europe with the rocks rest of the world. How can we decide if rocks in Utah, or on the deep seafloor, are Eocene or Oligocene or Miocene? The classic type sections in Europe turned out to be rather poor choices for the foundation of a timescale. Most of the subdivisions, or stages, of the Eocene and Oligocene in western Europe were based on relatively thin, incomplete shallow marine strata, with large gaps, or unconformities, between them (fig. 1.9). For years, geologists argued about whether a particular stage was sequential with the next, or partially overlapped it, or whether there was a gap between the two. In many cases, two successive stages were not in the same basin, or did not lie superposed upon each other, so determining whether they were successive or overlapping in age was impossible. In other cases, there was a clear gap between the types of two stages. Should the lower stage be extended upward to cover the gap, or should the upper stage be extended downward? Compounding the problem was that the biostratigraphic index fossils for these stages were shallow-marine mollusks, which are mostly species restricted to Europe and rarely could be compared to mollusks (or any other fossils) elsewhere. As a result, even up until the 1970s, there was much confusion about what was Eocene, Oligocene, Miocene, or Pliocene outside western Europe. We have since learned that a lot of the estimates made in the early and middle part of twentieth century were far off.

Figure 1.9. Depositional history of the type areas of the Paleogene stages and ages in