Older Than the Hills

Mountain View sits on sandstone. The flat brown slabs that weather out of the fields around town belong to the Roubidoux1 Formation, a layer of sand, lime and chert laid down along a tropical shoreline more than 470 million years ago2. About five miles north-northwest of town the two prongs of the Jacks Fork come together, and the river begins its run to the Current River in the same rock. Within about a mile3 it has cut down through the bottom of the Roubidoux into the gray dolomite beneath. Forty-six miles4 later, at its mouth, it flows past knobs of volcanic rock that are about a billion years older5 than anything in the bluffs upstream.

The river runs downhill faster than the rock layers do. The layers lie almost flat, while the river drops about 7 feet a mile on average6 (Chapter 2), so each mile downstream puts the water a little lower in the stack and among a little older rock. A float from the prongs to the mouth passes down through the rock record of the Jacks Fork country, from the Roubidoux to the Potosi Dolomite, with the volcanic basement rising beside the river at the end. Read from the bottom up, the record holds a volcano, a long gap, a warm sea, a shoreline, and then another long stretch of time that left almost nothing behind but chert.

Three clocks

The age of the rock and the age of the river are separate questions, and so is a third, the age of the channel a floater knows.

The first clock is the rock itself. The bluffs, ledges and boulders along the river are sedimentary rock from the Cambrian and Ordovician periods, roughly 500 to 470 million years old. Beneath them lies a basement of volcanic rock and granite close to a billion years older still. Up to 1,800 feet7 of the younger sedimentary rock has been mapped in and around the national riverways, a pile about five times the height of the tallest bluff8 on the river.

The second clock is the landscape: the high, rolling plateau, the steep hollows and the deep, winding valley cut into it. That clock started much later, once rivers began cutting down into the plateau, and it is still running (Chapter 2).

The third clock is the channel, the particular bends, holes and gravel bars of any given summer. It is measured in years and decades. Some of the low floodplain surfaces along the river are only a few decades old9, and the holes and bars move with every big flood (Chapter 15).

A Jacks Fork bluff is a cliff of very old rock that the river has exposed only recently, as geologists count time. Nobody has put a date on the river as a river, and the rock in its bluffs was already hundreds of millions of years old when the valley began to form.

Reading the age of a rock

Geologists put ages on rock in several ways, and each method checks the others.

The oldest method is order. Sediment settles in layers, each on top of the last, so unless something has overturned the pile, the lowest layer is the oldest. Geologists call this superposition. A layer that contains broken pieces of some other rock is younger than that other rock. A crack that cuts across several layers is younger than all of them. Rules this plain give the sequence of events without a single number.

Fossils narrow the sequence. Living things change over time, and some changed quickly enough that a particular species marks a thin slice of time wherever it turns up. The most useful fossils in the Ozark dolomites are conodonts, tiny tooth-like parts of an eel-shaped sea animal, most of them smaller than a grain of rice. They are abundant and they evolved fast. They are also made of calcium phosphate, the mineral of teeth and bone, so a block of dolomite can be dissolved away in weak acid10 in the laboratory and the conodonts picked out of what is left. Conodonts are how geologists pinned down the age of the Roubidoux11 Formation in and near the national riverways.

Numbers in years come from radioactive decay. Some kinds of atoms are unstable and break down into other kinds at a steady rate that heat, pressure and the passing seasons do not change. The rate is given as a half-life, the time it takes half of any amount to change. After one half-life half of the original atoms are left, after two a quarter, after three an eighth, so the share still unchanged works as a clock. Uranium-238 decays to lead-206 with a half-life of about 4.5 billion years. Uranium-235 decays to lead-207 with a half-life of about 704 million. When a volcanic rock cools, small crystals of the mineral zircon take in uranium and shut out lead, so any lead found inside a zircon crystal later was made there by decay, and the ratio of lead to uranium gives the time since the crystal formed. Each zircon carries both kinds of uranium, so it holds two clocks12 running at different speeds, and a date is trusted when the two agree. The answers come with an honest margin. A rhyolite lava from Shut-In Mountain, near Rocky Falls15 southeast of Eminence, has been dated at 1,470.4 million years, plus or minus 2.7 million, from six separate analyses14 of its crystals. The margin is less than a fifth of one percent16 of the age.

The dolomites of the bluffs cannot be dated that way. They formed from sediment on a sea floor, and their minerals did not start a clock when they formed. Their ages come from their fossils and from their position between rocks that can be dated13. A dolomite that lies on top of a dated lava is younger than the lava. At Rocky Falls the beds of Eminence Dolomite rest against the rhyolite of Shut-In Mountain17, so they must be younger than its 1.47 billion years, and the fossils in them narrow their age to the Late Cambrian18, about a billion years later. The mountain’s name carries an old Ozark word. Where a stream running in soft dolomite meets a ridge of the hard volcanic rock, it squeezes through a narrow, rocky gorge, and people in the hills have long called such a place a shut-in19 (Chapter 2). Rocky Creek drops over the rhyolite at the falls20.

Conodonts also record heat. As they are buried and warmed over millions of years they darken, slowly and for good, from pale amber toward brown and black, and geologists grade the change on a color scale. The conodonts in the Lower Ordovician rocks of the southern Missouri Ozarks score only 1 to 1.5 on that scale, the pale end, which means the rocks were never heated past somewhere between 50 and 90 degrees Celsius, about 120 to 190 degrees Fahrenheit22. Heat in the earth rises steadily with depth, so a low top temperature means shallow burial. The rock that once lay on top of these layers can never have been more than about 5,000 to 9,000 feet21 thick.

The mappers

Most of what is known about the rock under the Jacks Fork was learned by people walking the hollows with a hammer and a map, and many of the names they chose came from somewhere else in Missouri.

The Roubidoux Formation and the Gasconade Dolomite were named in 1892 by the geologist Frank Nason, for Roubidoux Creek in Texas and Pulaski counties and for the Gasconade River23 in central Missouri, where the layers are well exposed. The Jefferson City Dolomite takes its name from bluffs along the Missouri River near the capital, and the Potosi Dolomite from the old lead town of Potosi in Washington County, both from work published by Arthur Winslow in 189424. The Gunter Sandstone takes its name from Gunter Spring, now Ha Ha Tonka Spring in Camden County, and from an abandoned post office in Miller County25. Of the formations along the river, only one was named for a place on it. In 1911 the geologist Edward Ulrich named the Eminence Dolomite for exposures at Eminence26, Shannon County. The town had been named for its high ground27, and the rock under that high ground was then named for the town.

The same paper of 1911 reached much further. In it Ulrich proposed a whole new system of geologic time, the Ozarkian, carved out of the upper Cambrian and the lower Ordovician, the span of the rock along most of the Jacks Fork29, and named for the Ozark uplift, where, with the valley of the Appalachians, he counted some 8,000 feet of limestone that belonged to it. He thought the Ozarkian, with the Canadian system proposed beside it, the crowning work of his career. Other geologists never took to the Ozarkian, and its rocks were handed back28 to the Cambrian and to the Canadian, a name some American age lists still use30 for the Early Ordovician.

Few geologists had walked the Jacks Fork country before the 1920s. The men of the state’s first geological surveys, George C. Swallow, Benjamin F. Shumard and Garland C. Broadhead, never mentioned Shannon County in their reports and do not seem to have visited it. Those who did come were after copper (Chapter 5). One saw the Slater mine on Shawnee Creek in 1841, after its first working had stopped, and published a short account of it the next year. In 1875–76 C. P. Williams, head of the new school of mines at Rolla and state geologist by virtue of that post, visited the same mine, mapped its workings, drew maps of Round and Alley springs, and put the porphyry in the Archean, the name then used for the oldest rocks of all. A short federal notice of Missouri’s copper deposits followed in 1905, written by Ulrich and H. Foster Bain, and in 1920–21 two investigators spent a few weeks in the region and wrote a report on the copper that was never published31.

Between 1922 and 1928 Josiah Bridge, an Ohio-born paleontologist then in his thirties33, mapped the Eminence and Cardareva quadrangles, which take in Eminence and part of the Current River country, at a scale of about a mile to the inch. The work began with copper. The old mines near Eminence had been reopened during the First World War, and the survey was undertaken largely to meet the demand for better knowledge of the deposits. It grew into a study of the whole rock column. In 192336 Bridge’s field assistant, E. T. Campbell, traced most of the contact between the Gasconade and the Roubidoux across the western and southern parts of the Eminence quadrangle, the Jacks Fork side of the map. A contact, in a geologist’s sense, is the line where one formation meets the next, and on a map it is the line between two colors. Bridge’s report, published by the Missouri Bureau of Geology and Mines37 in 193034, gave names to layers that later geologists renamed or folded into other units. Part of what is now the middle Gasconade Dolomite was his Van Buren Formation35. The year the report came out, Bridge moved to Washington as a paleontologist for the U.S. Geological Survey, and he worked there until his death in 195332.

In the 1920s the porphyry outcrops of the region were read as a block raised by faulting. With a single exception near Van Buren, they fall inside a rectangle about fourteen by eighteen miles, its sides running northeast and northwest, and edges that straight and parallel suggested faults, perhaps as old as Precambrian38 time. The volcanic rocks near the mouth were recognized as the remains of a collapsed volcano in the late 1970s, when the Missouri geologists Geza and Eva Kisvarsanyi described the Eminence caldera, in papers of 1977 and 198439. The faults that now mark the caldera run northeast and northwest40, the same directions as the sides of the old rectangle. The two had escaped from Hungary together in December 1956, after Soviet troops crushed the revolution42 there, and reached the United States in January 1957. Eva Kisvarsanyi took her geology degrees at the University of Missouri and spent thirty-four years with the Missouri Geological Survey, from 1959 to 1993, retiring as its assistant director with more than a hundred papers, maps and books41 on the state’s Precambrian rocks to her name.

Detailed modern mapping came later still. From the 1990s into the 2010s the U.S. Geological Survey, with support from the National Park Service, mapped the riverways quadrangle by quadrangle at a scale of about 2½ inches to the mile. Each of these quadrangles is a block of country about eight and a half miles from north to south and seven45 from east to west. Eminence was mapped in 1999 and Alley Spring in 2012. The Jam Up Cave and Pine Crest quadrangles of the upper river were mapped in the field in 2005 and 2006 as part of a study of the karst44, and the maps came out in 201343. The geologists traced contacts on foot and fixed them by GPS, logged drill holes and well records, and drove most of the roads in the region looking at outcrops and at the loose rock lying on the ground, which geologists call float. In 2014 almost fifteen years46 of that work came together in a single geologic map of the upper Current and Eleven Point basins, and in 2016 the Park Service published a geologic inventory of the riverways built on it.

Outcrops are scarce in this country. Soil, red clay and chert rubble cover more than 95 percent47 of the ground in the Current and Eleven Point basins, and most of it is forested. Where the bedrock does not show, the float tells what lies beneath, because chert and sandstone survive the weathering that dissolves the dolomite around them, and the pieces lying on a slope come from the bed under them or from beds just uphill. On the uplands of the upper Jacks Fork no outcrop of the Jefferson City Dolomite was found at all. It was mapped from its soil, which is yellowish and silty, and from its sandstone fragments, which are less flat48 than those of the Roubidoux below it. The top of the Roubidoux was placed from its known thickness and from the change in soil color49.

The volcano under Eminence

The oldest rock in the Jacks Fork country formed about 1.48 billion years ago, long before any plant or animal lived on land, when nearly all life on Earth was too small to see.

Across what is now southern Missouri the crust was stretched, and an enormous amount of thick, silica-rich magma rose toward the surface. Rhyolite, the rock this magma makes, is at least 72 percent silica51, the stuff50 of quartz and window glass. Magma that rich in silica is stiff, and the gas dissolved in it, mostly steam, cannot work its way out a little at a time. As the magma rises and the weight on it drops, the gas comes out of solution all at once and tears the magma apart into froth and ash. Much of the rhyolite here is tuff, made from that ash. Some of it fell from the sky, and some poured across the ground as ash flows, dense clouds of hot ash and gas that run downhill at highway speed52 and more. Where the ash was still hot when it came to rest, the grains welded together into hard rock. In Missouri rhyolite has long been called porphyry53, for the scattered crystals set in its fine-grained body.

One of the volcanic centers lay around where the Jacks Fork now meets the Current. Geologists call it the Eminence caldera, and the town of Eminence stands inside the western part of its outline54. A caldera, from the Spanish word for a cauldron, is what remains when an eruption empties the magma chamber beneath a volcano so fast that the ground above it has nothing left to stand on. The roof collapses, and a broad basin forms where a mountain might have stood. The eruption throws out sheets of hot ash that race across the land, and then the ground falls in. Yellowstone sits on a caldera of the same kind, and a caldera eruption made the basin that holds Crater Lake55 in Oregon.

The Eminence caldera did all of that, and its rocks still show it. The lower set of volcanic layers, erupted before and during the collapse, is probably 7 to 8 kilometers thick, four to five miles, and holds at least 360 cubic kilometers of rock, about 86 cubic miles. Spread evenly over the whole Jacks Fork watershed, that much rock would bury it about a thousand feet deep57. When the floor of the caldera sank, those layers were rotated until they stood at 65 to 90 degrees, tipped on edge like books leaning on a shelf. After the collapse, lavas spread across the floor of the basin. Those upper layers, about 1,000 feet thick, still lie nearly flat on top of the tilted ones.

Between the two sets lies a fine, layered ash called the tuff of Mule Mountain56. It holds small pellets called accretionary lapilli, which form when rain falls through a drifting cloud of ash and rolls the ash into balls, or when moisture in an eruption column58 gathers the ash together. The pellets are a record of water in a volcanic plume a billion and a half years ago.

Magma that never reached the surface cooled slowly underground into granite. Slow cooling gives crystals time to grow61, so granite is coarse, with grains that can be seen by eye, while the lava and ash that cooled in the open are fine-grained. Granite near Big Spring on the Current has been dated at 1,461.8 million59 years, give or take 5.5 million. Rocks from the collapse are dated at 1,480 million, plus or minus a broad 42 million, and the lavas that came after it at 1,470 million60. Volcanic activity went on in the region for about a hundred million years62.

The Eminence caldera was one center in a larger belt of volcanoes and granite that includes the St. Francois Mountains, about 45 miles northeast67 of Eminence, where Missouri’s highest point stands. Detailed dating in the St. Francois Mountains shows nearly continuous volcanic and granite-forming activity over thirty million years, between about 1.48 and 1.45 billion years ago, and a later, smaller round of intrusions between about 1.34 and 1.2763 billion years ago. Why the crust melted so thoroughly at that time and place is still argued. One reading ties it to a rift opening as an ancient supercontinent64 began to break up. Another puts the Eminence caldera in a pull-apart zone. Where a long fault on which the crust slides sideways jogs65 from one line to another, the block caught in the jog is stretched and drops, and magma can rise into the gap. A third reading points to a broad stretching of the continent’s interior66, far from its edge.

Where the old rock shows

The volcanic rocks do not reach the surface along most of the Jacks Fork. They survive as more than fifty knobs that stick up through the younger rock, all of them in the Current River basin68. Rhyolite is much harder than the dolomite around it. The knobs have probably been high points for more than 500 million years, and they still show from the air, with more bare rock and thinner woods69 than the dolomite hills around them. Knob is the old popular name in Missouri for a hill of this hard rock. The porphyry breaks along its joints into small, sharp-cornered blocks that weather very slowly and, being angular, do not roll far downhill, so jagged masses of rock stick out on the porphyry knobs, while granite weathers into rounded boulders that tumble into the creek beds70 below.

A few belong to the Jacks Fork’s own neighborhood. Near the mouth, knobs of welded tuff of the Coot Mountain units rise within a few hundred yards of the river71. A knob of the lower Coot Mountain unit, dark maroon, stands beside Missouri Route 10672 about a mile and a quarter east of Eminence. A single outcrop east of Storys Creek73, which enters the river just above Eminence, gave its name to a whole rock unit, the rhyolite of Storys Creek. Coot Mountain itself, on the Current, is made of the tilted tuffs of the lower sequence, and so is Jerktail Mountain74. Another rhyolite knob rises on the south bank of the Current about a mile above the confluence75.

Other old volcanoes lie wholly out of sight. Deep drill holes near Hurricane Creek, in Shannon and Oregon counties south of the river, have found a ridge of volcanic rock running northwest on the buried basement surface, cut off from the caldera’s own volcanic rocks by buried granite. It may be a separate volcanic center, or it may be an outlying piece76 of the Eminence caldera.

A billion years missing

After the eruptions ended, the land wore down for a very long time, and nothing that survives was laid down here. The evidence is in how the old rocks sit. Along the buried surface of the basement, volcanic rocks that formed in the open air now lie at the same level as granites that cooled at depths of more than 5,000 to 6,000 feet. For the two to end up side by side, the ground over the granite had to rise and lose a mile or more of rock before the sea came in and buried them both.

Geologists call a buried erosion surface like that an unconformity, a page torn out of the record. This one, where the first sea-laid rocks rest on a far older basement, runs under much of the continent, and at the bottom of the Grand Canyon, where the same kind of contact is laid bare, it has long been called the Great Unconformity78. Beneath the Jacks Fork country the missing time is more than 800 million years77. During that time a supercontinent, Rodinia, was assembled between about 1.34 billion and 1 billion years ago, and the basement of this country lay within a few hundred kilometers of the collision along the continent’s southern edge. Some of the faults79 in the old volcanic rocks probably date from those movements. On Rocky Creek, a Current River tributary about nine miles east-southeast of Eminence, thick beds of dark-gray Eminence Dolomite lie directly against a knob of pale rhyolite. The contact between them is a seam a person can span with one hand81, and about 980 million80 years of time is missing across it, nearly twice as much time as has passed since the dolomite on top was laid down.

The surface that the sea eventually drowned was not flat. It had hills and knobs, with as much as 1,600 feet of relief82 over short distances, and those hills are still down there under the younger rock. Where erosion has cut deep enough, as at Coot Mountain and the Route 106 knob, the old hills are being dug out of their covering again, the way they have been in the St. Francois Mountains83.

The sea comes in

More than 500 million years ago the sea rose over the worn-down basement, and for tens of millions of years a warm, shallow sea covered what is now southern Missouri. The supercontinent was breaking apart by then, and as the pieces drifted away from one another the ocean spread over the low interior84 of North America. The continent then lay near the equator. Open ocean lay only a few hundred kilometers to the south, around what is now central Arkansas85, where a new ocean basin86 was opening off the edge of the continent. Closer by, to the southeast, another rift began to split the continent and failed. Its buried floor, the Reelfoot rift87, lies under the lowlands of the Mississippi.

Warm, clear, shallow water makes lime88. Shellfish, algae and microbes take it out of the water to build their shells and mats, and in the warmest shallows it settles out by itself as a fine mud. Buried under more of the same and pressed and cemented for millions of years, that mud and shell debris became limestone, and most of it here later became dolomite.

The old volcanic knobs stood up out of the water as islands. The first sand to wash around them, at the base of the Lamotte Sandstone, includes beds of rubble shed from the knobs, with boulders several feet across89. As the sea deepened and the islands drowned, layer after layer of sediment piled up around them. The layers are thinner close to the knobs and thicker away from them, and they still slope away from the old island hills, in places as steeply as 38 degrees. Part of the slope was there from the start, where sediment settled on the flanks of an island, and part came later, because the thicker pile away from the knob packed down more under its own weight90 than the thin skin on top of it. At Rocky Falls91 on Rocky Creek, east of Eminence, the beds of Eminence Dolomite can be seen dipping away from a knob of rhyolite that was an island when they were laid down.

The lowest of these sea-floor layers, the Lamotte Sandstone, the Bonneterre Formation and the shaly Elvins Group, lie deep underground along the Jacks Fork and are known there only from drill holes. Two Cambrian dolomites come to the surface.

The Potosi Dolomite is the oldest rock the Jacks Fork touches, in the last miles of its course. It is a brownish dolomite, massive and thick-bedded, full of small cavities lined with quartz crystals and with lumpy masses of chalcedony. A freshly broken piece of the brown dolomite gives off what the geologists who described it called a “fetid odor.” It ranges from about 50 to 575 feet92 thick in the Current River basin.

The Eminence Dolomite lies on the Potosi. It is light gray and massive, weathering to a bluish gray with a pitted surface, and it tends to stand out as small knobs and pinnacles. It is about 80 feet thick close to the old volcanic islands and as much as about 919 feet thick away from them. Many of its beds are stromatolites93, mounds built up by mats of microbes in shallow water. A mat trapped fine sediment, the microbes grew up through it toward the light, and the mat trapped more, layer on layer, until it had built a low mound or column of laminated rock. Stromatolites are among the oldest traces of life on Earth, and in the Cambrian sea they were still building reefs. The upper part of the Eminence is one of the two main levels94 in the riverways where caves form (Chapter 13).

Sea creatures more complex than microbes were spreading through the world’s oceans by then, in the burst of new animal groups known as the Cambrian explosion95, when animals with shells, jointed legs and eyes appear in the rock record within a few tens of millions of years96. Trilobites, snails and other mollusks left their remains in the Cambrian rocks of the riverways. Here and there sandstone layers in the Eminence97 record times when the sea drew back and beach-like conditions returned near shore, and the volcanic islands may have supplied some of the sand.

The Ordovician shoreline

About 485 million years ago the Cambrian Period gave way to the Ordovician. The sea stayed, but it rose and fell again and again, and the rocks of the next fifteen million years record a shoreline moving back and forth across southern Missouri. When the sea deepened, lime mud settled and the rock that came of it is dolomite. When it shallowed99, the beach and its sand moved out over the old sea floor, and the rock is sandstone. North America was drifting slowly north, away from the equator.

At the start of the Ordovician the top of the Eminence Dolomite was exposed to the air and weathered into an uneven surface. Then sand washed in over it, in quick cycles98 as the sea rose and fell. That sand became the Gunter Sandstone, 10 to 25 feet100 thick, the bottom member of the next formation. It matters far more than its thickness suggests. Sandstone barely dissolves, and where water is held against the underside of a sandstone layer, it dissolves the dolomite beneath. Below the Gunter lie the most complex cave networks in the riverways, and the flooded cave behind Alley Spring101 lies at that level, where the Gunter has been broken through (Chapter 13; Reach H). Where the Gunter is exposed on a hillside, its resistance to erosion makes benches, small waterfalls102, and streams that run along on top of it.

Above the Gunter lies the Gasconade Dolomite, between about 148 and 770 feet thick. It is thinnest next to the volcanic knobs and runs about 500 feet104 thick in the quadrangles of the upper river. The Gasconade is the main rock of the Jacks Fork’s bluffs, and the river runs on it for most of its length. Geologists divide it into three informal parts. The lowest is the Gunter. The middle part is light-gray dolomite with a good deal of chert in nodules, stringers and beds. Some of that chert is porcelain-smooth, some is speckled with tiny round grains, and some is made of fossil stromatolites. A bed of oolite, rock made of grains that grew in concentric layers as waves rolled them back and forth on a shallow bottom, runs through the middle part, and in the 1920s it marked the top of a Van Buren Formation since folded into the Gasconade. At the very top of the middle part runs a persistent bed of white chert 2 to 10 feet thick, the Cryptozoon chert. The upper part of the Gasconade is light-gray, coarse dolomite full of small cavities, which geologists call vuggy103. Its weathered surfaces are pitted. Most of the caves105 in the riverways are in the upper Gasconade.

Next comes the Roubidoux Formation, up to about 250 feet thick, the rock of the upper bluffs along the upper river and of much of the upland around Mountain View. It is a mix of sandstone, sandy dolomite and chert. The sandstone is white to pale orange and weathers light brown to reddish brown, and its sandstone layers are almost pure quartz107, the other minerals of the original sand having dissolved or washed away. Many of its layers carry ripple marks and crossbedding, the slanting laminations left by sand moving under water, and some show mud cracks108, where the bottom dried out in the sun when the sea dropped. Some of its chert holds the same tiny round grains106 as the oolite, a sign of wave-washed water near a shore. The Roubidoux records a coast, with rivers, beaches and shallows close by. Its sandstone breaks into flat, tabular slabs109 that weather out and lie on the surface long after the dolomite around them has dissolved.

On top of the Roubidoux lies the Jefferson City Dolomite, about 200 feet110 thick on average, a light-brown, earthy dolomite that weathers to a powdery pale yellow, with lenses of sandstone and beds and nodules of white chert. A conspicuous brown, pitted ledge 25 to 40 feet above its base, known to geologists as the Quarry Ledge, marks its lower boundary where the contact itself cannot be seen. Inside the national riverways the Jefferson City appears only on the uplands along the upper Jacks Fork111. On the uplands north of the river around Summersville112, the Jefferson City and the Cotter Dolomite above it are mapped together. They are the youngest bedrock in the Jacks Fork country, about 470 million years old.

Life in the Ordovician sea

Fossils are scattered thinly through the bluffs. The Gasconade, the Roubidoux and the Eminence have all yielded fossils inside the national riverways: stromatolites, coiled snails, brachiopods (shelled animals that look like clams but are not related to them), trilobites, conodonts, the broken stems of sea lilies called crinoids, and cephalopods, relatives of the modern squid and octopus that lived inside shells. One Roubidoux snail, Lecanospira, found as occasional molds and impressions, coils flat like a watch spring. Two species of nautiloid cephalopods and three species113 of the small, cap-shelled mollusks called monoplacophorans were described to science for the first time from specimens found in the riverways. Three more nautiloids, two trilobites and a snail114 may have been as well, from specimens whose find spots are less certain.

Earlier, in 1930, Edward Ulrich and Josiah Bridge named three new species of Cambrian snails from chert of the Eminence Dolomite on a hill north of the Jacks Fork about a mile and a half northwest of Eminence. One of them, Taeniospira eminencensis, became the type species of a new genus, and its holotype, the specimen on which the name rests, went to the U.S. National Museum115 in Washington.

Lecanospira turns up in a zone of chert 60 to 70 feet above the base of the Roubidoux, just over one of its sandstone beds. Aside from that zone and a thin fossil bed just above the basal sandstone, fossils are extremely rare in the Roubidoux. The same Lecanospira fauna turns up along the Appalachians from Alabama to New York, and again in northern Scotland and Scandinavia116.

The river’s marker bed

Along the upper Jacks Fork the most useful line in the rock is the top of the Gasconade, where the dolomite gives way to the sandstone at the base of the Roubidoux. It is also one of the hardest to see. The contact itself is rarely exposed, because the thin basal sandstone is so often buried under loose rock117 fallen from above. Geologists find it by working up from the white Cryptozoon chert. That bed is hard, white and persistent, and it holds its place in the stack from one bluff to the next. The distance from the chert up to the base of the Roubidoux is given as about 70 feet and as 80 to 100 feet118. In the field the chert bed is the marker, and the contact is drawn a set distance above it. The method is old. In the 1920s the base of the Roubidoux was mapped from the reef across much of the western and southern Eminence quadrangle. Where groundwater has dissolved the dolomite between the two, the interval shrinks, and in the east of that map blocks119 of Roubidoux sandstone lie mixed with blocks from the reef.

The name Cryptozoon, “hidden animal,” was given in the nineteenth century to fossils that were later recognized as stromatolites120. The Cryptozoon chert is a bed of those microbial mounds, turned to chert, and in places it stands out of the bluffs as a ledge. Near Blue Spring the bed is a massive reef about three feet thick, and a ledge of it juts into the river121 at the waterline (Reach D). Field observations of the chert run along the river for about eight miles, from just above Rymers to a couple of miles above Bay Creek122.

The contact shows plainly in the shape of the bluffs. Along much of the upper river the Gasconade makes the lower cliffs, and the Roubidoux makes the upper slopes. The two weather differently. The dolomite holds a sheer face, while the sandstone, sandy dolomite and chert above it break into ledges and overhangs, and slabs of sandstone wedged loose by frost and tree roots123 tumble down the slope (Chapter 2). Downstream from Jam Up Cave the contact runs about 100 feet above the river near the top of a cliff, and in 2007 a light-colored scar124 there marked where rock had recently fallen. Upstream, the contact shows not far above the Highway 17 bridge125 at Steel Bridge. When the road cut for State Highway 17 was made on the north bank of the river there in the 1920s, the fresh cut gave the best section of the Roubidoux in the vicinity. The hilltop stood about 230 feet above the river along the highway, the Roubidoux was 160 to 170 feet thick, its basal sandstone lay about an eighth of a mile north of the bridge, and below it about 58 feet of Gasconade126 showed down to the water. Far downstream the same contact can be seen beside Highway 106, about three-quarters of a mile west of Alley Spring127.

The contact matters underground as well. The base of the Roubidoux holds water against the dolomite beneath it, and on the upper river caves turn up again and again just below it128 (Chapter 13). Most of the sinkholes in the riverways country form in the Roubidoux and the Jefferson City129, where sandstone layers collapse into hollows dissolved in the dolomite below them (Chapter 2).

Down through the rock

Mile by mile, the course of the Jacks Fork is a descent through the rock column.

The river leaves the Roubidoux a little past mile 1 and runs on the Gasconade for about the next 34 miles, to about mile 35, past Steel Bridge, Blue Spring, Jam Up, Rymers and Bay Creek, while the Roubidoux climbs higher and higher in the bluffs above it. The rock layers stay almost level. Geologists show this with structure contours, lines that trace the elevation of a single bedrock layer the way topographic contours trace the ground, so that a layer lying dead flat would have no contour lines at all. Along the upper river the mapped layer stays within about a hundred feet131 of the same elevation for more than twenty miles. Over the same distance the river, falling 7 to 9 feet to the mile, drops about 180 feet132 (Chapter 2). In most of the region the layers tilt no more than two or three degrees, and their gentle lean is to the south and southeast. On the upper river the water tends to flow in the direction of that tilt133.

The Gunter Sandstone at the base of the Gasconade does not show along the river until the neighborhood of Alley Spring134, and a mile or two below Horse Hollow the river cuts through it into the Eminence Dolomite. From there to the mouth the river runs on the two Cambrian dolomites, the Eminence and, in stretches from a couple of miles above Eminence down to the Current, the older Potosi130. Near the mouth the volcanic knobs of the caldera rise beside it. In order downstream, the four formations of the Jacks Fork’s bed are the Roubidoux, the Gasconade, the Eminence and the Potosi135 (Part Five).

What dolomite is

Nearly every bluff on the Jacks Fork is dolomite. Dolomite is a close relative of limestone. Limestone is calcium carbonate, the same material as seashells and chalk, with almost no magnesium in it. Dolomite is calcium magnesium carbonate136, with magnesium in place of roughly half the calcium. Both begin the same way, as lime mud and shell debris on the floor of a warm, shallow sea. The rock becomes dolomite when water rich in magnesium soaks through it and makes the exchange, atom by atom137, while the rock stays solid.

When that exchange happened under the Jacks Fork is still debated. Some of the dolomite probably formed early, while the sediment was fresh on or near the sea floor, and some138 much later. One reading puts the weight on the later event. Toward the end of the Paleozoic Era, roughly 300 million years ago, hot, mineral-laden water rose through the rock of southern Missouri. It left the lead and zinc ores of the Viburnum Trend, north of the riverways, part of the largest lead-mining district in the world140, and on that reading it turned limestone that remained in the riverways into the dolomite139 seen today. The water was probably driven out of deep basins to the south141 as the Ouachita Mountains rose. The ore lies deep underground in the Bonneterre Formation, some of it draped over buried knobs of the old volcanic basement and some along faults142.

Nothing shows of it at the surface, and it was found late, by drilling. At the Fletcher mine, in the southern part of the Viburnum Trend, the discovery hole went down in 1958 and the mill started in 1967. Ore at the Sweetwater mine, the southernmost in the trend, was found in 1962 and mined from 1968, and its underground workings lie inside the area that feeds springs of the Current River. Ore of the same kind lies under the basin of Blair Creek, a Current River tributary, where companies drilled it in the 1960s and kept the results to themselves143.

The hills had their own ideas about buried metal long before the drills came. Lost silver mines are a favorite story all over the Missouri Ozarks. In the usual telling, men from Hernando de Soto’s expedition of the 1540s found a rich vein in a cave, worked it, sealed the cave and fell to killing one another on the way back, and a prospector with a degree in geology is said to have spent years hunting Shannon County for such a mine. Missouri has no native silver. Its silver comes locked inside galena, the heavy, bright gray lead ore, or in the Viburnum Trend inside the zinc ore, and a geologist of the state survey has suggested that the legends began with lumps of galena that looked like silver to whoever picked them up. Such a lump might really have held a little silver, though only smelting144 would have shown it.

Three of dolomite’s habits shape the river. First, it dissolves. Rainwater picks up carbon dioxide from the air and much more from the soil and becomes a weak carbonic acid145, and over long spans of time that acid eats dolomite away along its cracks and bedding planes. That is the origin of the caves, springs, sinkholes and losing streams of the Jacks Fork country, and of a river that keeps running clear through a dry August (Chapters 12 and 13). About 99 percent146 of the national riverways is karst, the name for country shaped by dissolving rock. Second, dolomite stands up. It is strong enough to hold a vertical face for a very long time, so the river can undercut a bluff and leave a cliff behind. Third, it makes glades. Where dolomite lies bare on a sunny slope, it carries thin soil and plants built for drought, and the open glades of the riverways come at the contacts147 between formations, between the Eminence and the Gasconade and between the Gasconade and the Roubidoux.

Chert, the rock that stays

Chert is the rock that stays behind when dolomite dissolves. It is silica, the same substance as quartz, in a form so fine-grained that it looks glassy or waxy. Most people in the Ozarks call it flint. It formed inside the dolomite as nodules, lenses and beds, from silica gathered within the sediment, and in places it replaced stromatolites and oolite grain for grain, keeping their shapes. It is very hard, and its crystals are too small to see even under an ordinary lens, and because they run every which way, the rock has no grain to split along. A blow spreads through it in a smooth, curved, shell-shaped break148 that leaves a very sharp edge. Almost nothing in Ozark rain or groundwater dissolves it.

That last property explains the ground underfoot. When dolomite dissolves, the chert inside it is left behind, along with the sand and clay. Over millions of years the hills have lost a great thickness of rock, and the insoluble leftovers have piled up as a blanket of red or reddish-orange sandy clay full of chert, from small chips to boulders six feet across. Geologists call this blanket residuum. In the riverways it is typically about 40 feet thick, about the height of a four-story building, and in places more than 100. Most of it came from the Roubidoux, and it is the greatest of all the loose deposits by volume. Being rich in clay, it is sticky149.

A geographer surveying the Missouri Ozarks in 1920 judged that the region probably held more chert than any other area of its size150, and that over nine-tenths of the surface it “covers the roads, chokes the stream beds, and in many places all but obliterates the soil.” He placed Shannon County in the Courtois flint hills151, his name for the rugged hill country of the Current and Black rivers (Chapter 2).

Chert makes the Jacks Fork a gravel river. Nearly every pebble on a Jacks Fork gravel bar is chert or sandstone from the rock column above, broken, rolled and rounded on its way downhill (Chapters 2 and 15). Chert was also the stone the river’s first people chipped into dart points152 and tools (Chapter 3).

The state has honored the stuff. In 1967 the General Assembly made mozarkite, a chert of many colors that takes a high polish, the official state rock. It is found mainly in west-central Missouri, as boulders weathered out on hillsides and in road cuts, and it comes from the Cotter Dolomite153, which with the Jefferson City caps the uplands of the upper watershed, as around Summersville154.

Faults and cracks

The rock under the Jacks Fork is broken by faults, fractures along which the rock on one side has moved against the rock on the other, and they run through the basement and up into the dolomites. Rock can move along a fault in three main ways. Where the crust is pulled apart, one side drops. Where it is squeezed, one side rides up over the other. Where two blocks slide past each other sideways, the layers on the two sides stay at the same height but no longer line up, and geologists call that a strike-slip155 fault. The mapping of the 1920s found no faults around Eminence156 at all. Most are hidden under soil and gravel, and the later mappers traced them by finding layers out of place from one hillside or drill hole to the next, and by bands of crushed rock and smeared surfaces157 in the outcrops they could see, even in loose blocks of sandstone158. Along the large fault zones the rock is broken into angular rubble, much of its carbonate dissolved and replaced by silica159.

The faults fall into three age groups. The oldest were active only in the Mesoproterozoic and deform only the old volcanic and granite rocks. A second group moved at least once before the Ordovician layers were laid down. The third and youngest group moved at one or more times during the last 540 million years and probably earlier as well, and some of that movement probably came during the mountain-building of the late Paleozoic. Two directions dominate, northeast and northwest, and the pattern in the younger rock follows the grain of the basement160 beneath it.

Several of the faults on the upper river carry the names of its hollows. Three short faults running north and northwest, the Fly Blow Hollow, Buck Hollow and Belew Hollow faults, were mapped between two longer ones, the Teresita fault on the southwest and the Houston fault on the northeast, and the three merge toward the southeast, where they either die out161 in the Ordovician rock or move it too little to map. On the geologic maps the Belew Hollow fault crosses and recrosses the winding river between about miles 3.6 and 6.5, above Steel Bridge, the Buck Hollow fault crosses it around Blue Spring, and the Fly Blow Hollow fault crosses just below Jam Up.

The Houston fault is the longest of them. It breaks the beds of the Gasconade in a road cut on U.S. 63 at the north edge of Houston, and farther northwest the ground on its southwest side has dropped far enough to set the Gasconade beside the Roubidoux. From Houston it has been traced southeast for about 41 miles164, to and past the Jacks Fork, which it crosses a couple of miles below Rymers. Farther downstream the inferred trace of the Low Wassie fault crosses the river just below Bay Creek. It belongs to a line of faults that geologists have read as a strike-slip zone, and drill samples from it show horizontal scratches165 on the fault surfaces, the marks left when one block ground sideways past the other. The Hartshorn fault, which crosses about a mile above Alley162, runs east into the western part of the old caldera. Along it, as along several others, the side that is up in one place is down in another, a reversal known as scissoring166 that also points to sideways movement. A fault named for Possum Trot Hollow crosses the river above Eminence163.

The rock also carries joints, cracks along which it has split without moving. In the riverways the main joint sets run north-south and east-northeast. They probably opened when the crust buckled during the building of the Ouachita167 Mountains to the south, about 300 million years ago. Water moving through the rock widens some joints, but in the riverways most joints break only one or a few beds before they stop, so they rarely link up into a path straight down through the stack, and they play a smaller part168 in guiding caves than the bedding planes between the layers (Chapter 13). The fault zones too are paths for groundwater, and in the past they were paths for the hot fluids169 that left lead ore behind.

The faults of the Jacks Fork country have been nearly still in historic times. Earthquakes recorded in the region from 1964 through 2003 were small and scattered, between magnitude 2.5 and 5.1, centered within about seven miles170 of the surface, and usually felt only locally. The great New Madrid171 earthquakes of 1811 and 1812, among the largest ever recorded in North America, happened on faults far to the southeast, along the buried Reelfoot rift172. By one account Alley Spring stopped for twelve hours in 1811, the year the New Madrid shaking began174, and fell five feet173 (Reach H). Old earthquakes can leave a record underground as well. A hard shake can snap stalactites and knock over stalagmites, and broken formations in the caves of the riverways have been studied175 to find out when past earthquakes struck.

The missing record

The top of the bedrock under the Jacks Fork is the Jefferson City and Cotter dolomites, about 470 million years old. Above that there is no bedrock at all. Everything from the middle of the Ordovician Period to the loose gravel of the last few million years is missing.

Missing means one of two things. Rock was laid down and later worn away, or nothing was laid down at all.

The sea came back. On hilltops and slopes across this part of the Ozarks lie jumbled blocks of chert that do not belong to any of the local layers. They lie in oval patches from a few dozen to a few hundred yards across, and not always on the highest hills. The pattern suggests that the blocks settled into old sinkholes and were let down, little by little, as the rock beneath them dissolved. They contain fossils: pieces of bryozoans, which were small colonial animals, the stems of crinoids, the molds of brachiopods, and in one block, from the Birch Tree quadrangle south of the river, the mold of a trilobite. Identified in 1998, the fossils belong to the Mississippian Period, about 359 to 323 million years ago, more than a hundred million years after the Jefferson City Dolomite formed. They come from at least two chert-rich limestone formations, the same age as the Burlington and Keokuk176 limestones of the upper Mississippi Valley, that once covered the central Salem Plateau and have since been stripped away, leaving only their chert. On the Jam Up Cave and Pine Crest quadrangles of the upper Jacks Fork, cobbles of what appears to be reworked Mississippian chert177 and sandstone, with the molds of crinoids and snails, turn up in the loose deposits.

Sometime in the Mississippian a sea covered this country again and left limestone over it, and that limestone was later eroded almost completely away. The pale conodonts set the limit: whatever lay on top of the Ordovician rocks here was never more than about 5,000 to 9,000 feet178 thick in all.

Mountains rose to the south. In the late Paleozoic, roughly 300 million years ago, the continents were colliding to form the supercontinent Pangaea. The Ouachita Mountains were pushed up about 124 miles south of the riverways, the Appalachians rose in the east, and the Ancestral Rocky Mountains were forming in the west. The squeeze fractured the rocks of the Ozarks, moved them along faults, and drove the hot fluids179 that left lead and dolomite behind. What those events left here was fractures and minerals.

And then, for a very long time, erosion. At the dawn of the Mesozoic Era, about 252 million years ago, all the continents were joined in Pangaea. Within a few tens of millions of years they began to pull apart, opening the Atlantic Ocean and the Gulf of Mexico. During the Cretaceous Period the sea rose high enough to split North America in two along the Western Interior Seaway, and the crust between the southern Appalachians and the Ouachitas sank into a low trough, the Mississippi Embayment. The sea in the embayment reached to near Cape Girardeau, about a hundred miles east of Eminence181, but it never came back over the Jacks Fork country. No rocks180 of Mesozoic age have been mapped in the riverways. The area has been dry land, exposed to the weather, since at least the Cretaceous182, and weathering, erosion and the cutting of rivers have gone on through the 66 million years of the Cenozoic Era since then.

The dinosaur question

No dinosaur fossil has ever been found in the Jacks Fork country, and there is no rock of the right age here for one to be found in. Dinosaurs lived from about 230 million to 66 million years ago, in the Mesozoic Era. Every bedrock layer along the river is more than 200 million years older than the first dinosaur. The fossils in the bluffs are sea creatures that lived long before any animal with a backbone walked on land.

Dinosaurs may still have walked here. During the age of dinosaurs this was dry land, and land animals presumably lived on it. A bone becomes a fossil only if it is buried quickly under sediment that keeps piling up, in a delta, a swamp or a sea, and stays buried for tens of millions of years. Land that is wearing away is the opposite of that. Anything that died on the Mesozoic Ozarks was eventually washed downhill or broken down, and the record that might have shown what lived here is the part that is missing.

Missouri’s best-known dinosaur site shows where that kind of record did survive. In 1942 the Chronister family, near Glen Allen in Bollinger County, west of Cape Girardeau and about 80 miles east of Eminence185, turned up large bones while digging a well, which they abandoned for lack of water. A boy named Ole Chronister led Daniel R. Stewart of the Missouri Geological Survey, who was studying clay deposits in the area, to the site, and Stewart found more bones nine feet down in the clay of the well. In 1943 Lula Chronister sold most of the bones to the Smithsonian Institution for $50. They belonged to a Late Cretaceous plant-eater that lived along the eastern shore of the Western Interior Seaway, now classed as a duck-billed dinosaur with more than a thousand teeth. Stewart and the paleontologist C. W. Gilmore announced it in 1945, and it has gone through several scientific names since. It became Missouri’s state dinosaur in 2004, and in 2022183 the legislature changed its official name to Parrosaurus missouriensis. In 2021 an adult specimen was excavated at the site and taken to the Field Museum184 in Chicago. The bones lay in clay of Late Cretaceous age, preserved in a corner of Missouri where sediment of that age piled up and survived.

The Jacks Fork country holds fossils from both ends of its missing record. In the bluffs are the Ordovician sea creatures. In the caves of the riverways are the bones of Ice Age animals, all of them from the last major glaciation or later186, protected underground from the erosion that destroyed everything older on the surface (Chapter 13). Many were collected between the 1950s and 1980 by Oscar Hawksley, the Warrensburg biology professor who helped found188 the Missouri Speleological Survey, and are kept in the vertebrate paleontology collection of the Illinois State Museum187. Collecting fossils on park land is prohibited by federal regulation189 (Part Six).

The Ice Age, at a distance

The last episode of deep time began about 2.6 million years ago with the Pleistocene Epoch, the ice ages. Ice sheets spread across northern North America and drew back many times. The ice reached the Missouri River190 at its farthest but never came near the Jacks Fork.

The cold reached the Ozarks anyway, speeding the weathering of the slopes and spreading wind-blown silt191 over the hills (Chapter 2). Through the ice ages the Ozarks were a refuge192 for plants and animals pushed out of colder country. When the climate warmed, many of the northern plants died out across the south of their range and held on only where the ground stayed cool and damp193. On cool, north-facing cliffs in the Jacks Fork Natural Area grows harebell, one of the plants counted as relicts194 of a colder Missouri. Others on moist Ozark cliffs include red-berried elderberry and northern bedstraw195 (Chapter 16).

By about 12,000 years ago196 the ice ages were over. People had already come to the Ozarks (Chapter 3).

Deep time in order