In 1829, at a bend in the Mississippi called Plum Point, a boat that looked like two boats joined together came upriver under full steam, heading straight for a tree. The tree was underwater. From the deck, the crew could see almost nothing of it—just a dark trunk breaking the surface and a twist in the brown current where the river curled around something solid. Below that trunk, a stem might run sixty feet or more, its roots buried deep in the silt of the riverbed, its head tilting with the current at exactly the level of a steamboat’s hull.
Plum Point was full of such trees. Rivermen called that stretch the most dangerous water in the lower Mississippi, and the wrecks were proof. Between the twin bows of the approaching boat, just below the waterline, ran a massive iron-clad beam. It struck the tree with all the weight of the vessel driving behind it.
A crack sounded from beneath the surface. The boat rose and passed over the spot where the tree had been. Then chains were lowered into the river, and on deck two enormous wheels, each eighteen feet in diameter, began to turn. Slowly, between the two hulls, the Mississippi gave up something it had been hiding for years.
The tree came out of the water heavy with mud, roots and all. Some of the trees pulled out this way weighed as much as seventy tons. There were no hydraulic cylinders on that boat. No pumps, no pressurized oil, no fluid-driven pistons.
All that lifted those trees was steam, wheels, chains, and leverage. The boat was called the Heliopolis, and the man who designed it had spent twenty years of his life dodging the very thing it was built to destroy. To understand why someone would build such a machine, you have to understand what the rivers of the West were in the 1820s. They were the road.
For the farmers and towns spreading across the middle of the continent, there was no other way to move anything heavy over a long distance. Grain, salt pork, lead, whiskey, timber, and cotton all went down the Ohio and the Mississippi, floated on flatboats or driven by steam, toward New Orleans. After 1815, more and more of that cargo moved by steamboat. The steamboats turned the rivers into a two-way highway, and the young nation poured its commerce onto that highway as fast as it could build the boats.
The problem was that this highway ran through a forest. The banks of the western rivers were lined with huge trees—cottonwood, sycamore, oak—and the rivers were constantly eating those banks. Every flood dug deeper under the roots, and sooner or later the tree fell in. A big tree falling into a big river did not simply drift away.
The root mass was heavy and full of dirt, so that end sank and drove itself into the mud of the bottom. The trunk swung with the current until it pointed downstream, and the water pinned it there, angled upward with its tip just below the surface. Rivermen called these obstacles snags. A snag that stayed fixed and firm was a planter.
One that swung with the current, showing itself and then sinking out of sight, was a sawyer, because it moved up and down like a man working a saw. The sawyer was what captains feared most, because a pilot could see it, lose it, and not find it again until it pierced the hull. And the hulls were thin. Western steamboats were built light and shallow to run in low water.
A loaded boat running upstream that hit a hidden log had almost nothing between its cargo and the river. A snag could open a hull the way a tin can is opened. Insurance rates on the western rivers climbed. Running at night became a gamble.
The government knew all of this. On May 24, 1824, Congress appropriated $75,000 for improving navigation on the Ohio and Mississippi, and a large part of that problem was the trees. The army’s chief engineer, Alexander Macomb, organized a competition for the best means of removing them. The winner, a man named John Bruce, received a prize of $1,000 and a contract to clear the Ohio.
His machine was a double-hulled boat equipped with pulleys, derricks, winches, and a capstan, all worked by hand. It worked in the sense that it could remove some obstacles. But it was, in the words of the army’s own history, of limited value. Before long Bruce set it aside and went back to ordinary boats and chains.
On the rivers, the old ways still held. A crew with ropes and hand tools might spend hours freeing a single tree, and the rivers held thousands of them. Henry Miller Shreve was born on October 21, 1785, in Burlington County, New Jersey. His father, Israel Shreve, was a colonel in the Revolution who had bought land in southwestern Pennsylvania from George Washington himself.
There Henry grew up, near the headwaters of the Ohio, and his father died when he was about fourteen. By 1807 he had built his own boat and entered the fur trade between Pittsburgh and St. Louis. By 1810 he had opened a lead trade on the upper Mississippi, and a single trip to Galena cleared more than $11,000.
He was no surveyor or theorist. He was a working riverman who knew every bend of the Ohio and Mississippi from the pilot house, which meant he knew the snag grounds precisely—and he had watched them sink boats all his life. Then came steam. In January 1815, Shreve took a small stern-wheel steamboat called the Enterprise down to New Orleans, loaded with ordnance for the war against Britain.
In May of that year he brought the Enterprise back upriver against the current to Louisville, the first steamboat ever to make that passage north. A year later he built the Washington at Wheeling, the first steamboat with two decks, and its voyages are regarded as the beginning of the commercial steamboat era on the western rivers. When the Fulton and Livingston interests tried to enforce a monopoly on steamboat navigation on the lower Mississippi, Shreve fought them in court. On April 21, 1817, the suits against him were dismissed, and the western rivers were open to anyone who could build a boat.
By then, Shreve had been thinking about snags for years. In 1824 he wrote to the Secretary of War, John C. Calhoun, describing a steam-powered boat designed to catch them. In the winter of 1826–27, the government appointed him superintendent of western river improvements.
His first attempt at a snag boat, which he modeled on Bruce’s boats, failed. So he started over. The Heliopolis was completed in August 1829, and it looked like nothing else on the river. Shreve took two hulls, each about 125 feet long, and placed them side by side with a gap between them, binding them together on top with a series of heavy beams.
It was a catamaran built on the scale of a steamboat. The logic was balance. A tree weighing tens of tons, pulled to one side of a single narrow hull, would capsize it. Two hulls spaced apart and fastened together made a platform wide and rigid enough to take the strain from the middle, where the load could be hung evenly between them.
Across the bow, between the hulls and below the waterline, Shreve mounted his weapon. It was a massive frame of iron-clad timbers. The captain aimed the boat at the snag, the engineers opened the engines, and the Heliopolis drove straight at the tree at full power. The iron beam took the impact with all the weight of the boat moving behind it.
Sometimes the trunk snapped clean off below the surface. Sometimes the blow tore the roots out of the mud that held them. Either way, the boat kept going, passing over the obstacle until the loosened tree settled into the gap between the hulls. Then came the lift.
On deck, on both sides, stood two huge wheels, each eighteen feet in diameter, with an axle between them that acted as the boat’s winch. Each wheel was turned by a cable connected to one of the boat’s engines. The same steam that drove the Heliopolis through the water could be redirected to the hauling work. The chains went down and wrapped around the tree, the wheels turned, the axle wound, and the snag was pulled out of the river into the open space between the hulls.
There, the crew went to work with saws. The sound timber was cut up and burned in the boilers as fuel, or floated to the dock. The roots and the rest of the debris were sunk in deep pools where the current could not put them back in the channel, or hauled ashore. A crew with ropes and hand tools might need hours to free one big snag.
The Heliopolis could pull one in under forty-five minutes. And the trees were enormous. Shreve reported his boat could break and lift trees sixty feet long and three and a half feet thick, buried twenty feet deep in the riverbed, with perfect ease. Army engineer Captain Richard Delafield recorded that the Heliopolis lifted one tree 160 feet long in its first year.
The volume of one snag the boat pulled was estimated at 1,600 cubic feet of timber, and the report said it could not have weighed less than sixty tons. Accounts of the work describe snag boats hauling trees of up to seventy tons out of the western rivers. Here is the part that is hard to believe. Today, a machine asked to pull a tree like that out of a river would almost certainly do it with hydraulics.
Oil under immense pressure, driven into steel cylinders, pushing a piston out with force no chain could match. That is what moves the arm of a modern excavator and the boom of a modern crane truck. The Heliopolis had none of it. It had a ram, two wheels, an axle, some chains, and a pair of steam engines.
Every ton it lifted came from leverage and the turn of a drum. One source described these boats as possessing a permanent and powerful lifting force, and that is exactly what they were. No one had to invent a new kind of power. Shreve simply found a way to aim the weight of the boat and its engines at the tree.
The first real test was Plum Point. It had a reputation as the worst stretch of the lower Mississippi, a bend so full of snags that captains dreaded it. The Heliopolis went in, and within a few hours of work, the worst snags at Plum Point were gone. Within two days, according to one account, the entire stretch was clear.
Word spread up and down the Mississippi and the Ohio. By one account, the boat made some 300 miles of river as safe as a millpond. In 1830, according to a river history, only one flatboat was lost to snags. Other snag boats followed.
A second boat, the Archimedes, was built in 1831 with the same pair of towering eighteen-foot lifting wheels. People who watched these strange twin-hulled boats gave them a name. They called them Uncle Sam’s tooth pullers, because that is what they looked like. A machine applied to a bad tooth buried in the riverbed, extracting it roots and all.
That was the whole machine. Two hulls, an iron beam, two wheels, a chain, and a river finally giving back its trees. In April 1833, the government sent the same kind of machine, with 159 men, to a different river, against something an official expedition had declared impossible to remove. The river was the Red River, and the thing blocking it was the Great Raft.
The Red River ran through what is now Louisiana before joining the Mississippi. For centuries, part of it had not been a river at all. Floods brought down driftwood, especially cedar and cypress. It gathered, jammed, and piled on itself until the channel was closed for miles.
Silt settled over the wood. Soil formed on top of it. Willow and cottonwood trees rooted on the surface, and in places the raft was a foot and a half thick. Vines stretched across the logs and bound them together.
A man looking at the raft from the bank might take it for a long island. Its length depended on who measured it and when. A report from 1815 estimated it at least 150 miles. An army officer in 1837 put it at 214 miles.
For the Caddo people who lived on its banks, the raft was sacred, and their lives were built around the water and soil it swept down the valley each spring. To the government, it was a wall blocking a river that could carry trade inland. In 1806, an American expedition under the surveyor Thomas Freeman worked its way up the Red River until it reached the raft and came away with a conclusion. Freeman wrote that no hope could ever be entertained of removing the Great Raft.
In 1832, Secretary of War Lewis Cass ordered Shreve to remove it. On April 3, 1833, a small fleet left Natchez. There were the boats Java, Souvenir, and Pearl, plus the snag boat Archimedes with its two eighteen-foot wheels above the water. At ten in the morning on April 11, the attack on the Great Raft began.
By the end of that first day, according to one account, the raft was five miles shorter. What followed was not a single battle. It was a six-year siege. The work went on section by section.
The snag boat rammed the front of the raft, the chains went down, the wheels turned, and the packed timber was torn out and hauled away piece by piece. Shreve had a problem with all that wood. If he simply let it go with the current, the slow Red River current would carry it a few miles and form a new jam somewhere downstream. So he had the crews push the logs into side streams.
The driftwood was stored out of the way, forcing the water back into the main channel where he wanted it. When a large section of the raft finally broke loose, the men said the sound it made was like human cries of pain. The conditions were brutal. The crews worked in heat and humidity that never broke, among mosquitoes, fleas, snakes, with alligators in the water around them.
Much of the work was done standing in waist-deep water, and boots were not for comfort but necessity. By June of that first summer, the workers could barely stand the heat and threatened to quit. In the end, Shreve had to promise his men the work would not continue through the summer months. But by then the results were already astonishing.
On May 8, the crews had cut about forty miles into the raft. On June 23, they reached a place called Coates Bluff. Seventy miles of the Great Raft had been pulled out in a single season. The fleet returned in the fall of 1834, leaving the Ohio on November 14 with 300 men.
That winter and spring were slower. By mid-April they had cut only a few more miles. Then, in November 1835, the Archimedes herself broke down. Her hull had rotted, and on the Ohio she finally gave out.
A new snag boat, the Eradicator, was built the following summer to replace her. Sickness was the worst of it. In the summer of 1836, yellow fever broke out among the crews. A report that month counted 9,006 sick days lost to disease.
That was the toll the river took. Even so, the same report showed another twenty-one miles of raft removed, with only nine miles left in that stretch. That spring, near the landing below Coates Bluff, Shreve laid out the streets of a small settlement. It was named Shreve Town.
Shortly after, it was named Shreveport. By the summer of 1837, only about a quarter mile of the original raft remained. Then the country went into a financial panic, and Congress grew reluctant to spend. The work stopped.
And while it was stopped, the river kept doing what it had always done. By the time the crews returned in the fall of 1838, that last quarter mile had grown to nearly two miles. They pulled it out again. On February 15, 1839, the last of the original Great Raft was pulled out.
Seven hundred fifty miles of the Red River, up to Fort Towson, were open to boats. The snag boats were busy everywhere. On the Arkansas River in the winter of 1834, Shreve’s crews worked upstream from Pine Bluff and reached Little Rock by February 22. On the Arkansas, the men removed 4,907 snags—an average of one snag for every 88 yards of river.
Across the western rivers by the 1840s, between six and ten government snag boats, each with a crew of thirty to forty men, had pulled more than 160,000 snags out of the water. It is hard to overstate what that did. A river that had destroyed boats since time out of mind became a road a captain could travel with a reasonable hope of arriving. But cleaning a river is not a one-time job.
The trees keep falling in. And the money that funded the snag boats was never guaranteed. In 1841, a new administration took power in Washington, and Henry Shreve was relieved of his post. He retired to St.
Louis. The appropriations for the work kept shrinking, and by 1844 four of the government’s snag boats were sold off. The river kept making snags. On the night of January 3, 1844, a steamer called the Shepherdess was working its way up the Mississippi toward St.
Louis. She had come from Cincinnati and was almost at her destination. It was a dark, stormy winter night. There was ice in the river, and she had close to sixty or seventy people aboard.
She was at the Cahokia bend, just above the mouth of Cahokia Creek, about three miles from downtown St. Louis. It was eleven o’clock at night. The Shepherdess struck a submerged snag.
The tree tore open her hull planking. Water came in so fast it reached the lower deck in under two minutes. In about three minutes, the upper cabin floor was flooded. In the darkness, passengers found themselves in freezing water.
The current carried the boat and pushed her back into the stream, where she struck a second snag. The Shepherdess lurched violently, and the shock threw her boilers, part of her engine, and both smokestacks over the side into the river. Her captain was Abram P. Hoyle of Covington, Kentucky.
He was the father of eleven children. When the boat struck, Captain Hoyle went to the alarm bell and began ringing it. He was still ringing it when the river swept him overboard. He was never seen again.
One of his children, who was with him on the boat, also died in the river. A passenger named Robert Bullock of Maysville, Kentucky, did his best. He helped people up to the hurricane deck, the highest part of the boat, which was by then the only part of the Shepherdess still above water. When the wreck broke up, Bullock made it to shore, and there, in the ice and cold, he kept two women alive.
The steamer Henry Bry and a ferryboat called the Islander came out to pull survivors from the river. Not everyone had a Bullock. A man named Levi Cradick lost three of his children that night. A man named Green lost his wife and three children.
A man named Wright lost two children. No one could fix the death toll exactly, because there was no accurate list of who was aboard. Early reports put it around forty dead. One account said no fewer than seventy.
The Shepherdess sank three miles from St. Louis—the city where Henry Shreve, the inventor of the machine that pulled those trees out of the river, was living in retirement. The Shepherdess was not an exceptional case. That is the hardest part of this story to accept.
In the decades before the Civil War, the most destructive cause of steamboat loss was not fire, nor the boiler explosions that made the headlines. It was the trees. By one statistic, between 1821 and 1860, snags wrecked 593 steamboats on American rivers. Fire destroyed 582.
Collisions wrecked 199. Boiler explosions, the disasters everyone remembers, wrecked 113. Even into 1849, snags and similar obstructions were behind about three-fifths of all steamboat accidents. On the Arkansas River alone, historians have documented 190 steamboat wrecks, 96 of them caused by snags—more than double the losses from fire.
An economic historian studying a sample of western boats found that about 40 percent of steamboats built before 1849 were lost, most of them to snags. The boats in his sample lost to accidents had been in service on average less than twenty-two months—less than two years. That was what a new steamboat could expect from the western rivers. The deaths from all of it were never accurately counted, because on most of these boats no one kept a precise list of who was aboard.
The true number is certainly far higher than the records show. The Red River did not stay open. After Shreve left, a new raft began to form in the upper river, near where the town of Belcher, Louisiana, sits today. Over the following decades, through the Civil War and after, that second raft kept growing until it reached the Arkansas state line.
In the early 1870s, the job of removing it went to a young army engineer named Eugene Augustus Woodruff. Woodruff was from Connecticut, a West Point graduate, a lieutenant in the Corps of Engineers. On December 1, 1872, his crews began work on the new raft. His main vessel was a snag boat called the Eide, a modern descendant of Shreve’s Archimedes, working alongside two crane boats.
By the 1870s, the old methods had a powerful new partner. Beginning in May 1873, Woodruff put a carefully planned program into effect using nitroglycerin, blasting the packed timber so the boats could haul it away. It worked. On May 16, 1873, a steamer called the R.
T. Brearley steamed through the opening. That September, yellow fever struck Shreveport, the city that carried Henry Shreve’s name. The epidemic was brutal.
Woodruff could have gone back to his crew on the river. He chose not to, because he did not want to carry the infection to his men. Instead, he stayed in Shreveport and worked among the sick, helping bring some order to a city in chaos. In the midst of it, he caught the fever himself.
Eugene Woodruff died on September 30, 1873. He was two months short of his thirty-second birthday. His military commander wrote that Woodruff died a martyr to the cause of blessed good. He was buried in Oakland Cemetery in Shreveport, near the mass grave holding more than seven hundred victims of that epidemic.
His brother, George Woodruff, took over the work. On November 27, 1873, less than two months after Eugene’s death, the snags were gone and the Red River was open to navigation. A photographer named R. B.
Talfor documented the work, and his 113 hand-colored photographs of the rafts and the boats that fought them survive to this day. The snag boats outlived everyone. The Army Corps of Engineers kept building and sending them out, decade after decade. One of them was the Montgomery.
She was built in 1925–26 by the Charleston Dry Dock and Machine Company in South Carolina, a steam-powered stern-wheeler about 190 feet long. Starting in 1926, she worked the Alabama and Coosa rivers, then the Black Warrior and Tombigbee, then the Apalachicola, Chattahoochee, and Flint, pulling trees, logs, and submerged branches out of the channels for more than half a century. She was retired on November 8, 1982. By then she was the last working stern-wheeler on the southern rivers, one of only two steam-powered snag boats left in the United States.
The other is the W. T. Preston, which cleaned the rivers of Puget Sound and now rests in Anacortes, Washington. The Montgomery is a National Historic Landmark.
She is moored on the bank near Pickensville, Alabama, where she can still be visited. Think about the span of that era. The Heliopolis struck its first snag in 1829. The Montgomery, still running on steam with her stern wheel, was pulling trees out of American rivers in 1982.
More than 150 years after Henry Shreve’s first snag boat, at least one boat was still doing the same work in the same way, using steam. Go back to Plum Point in 1829. A 125-foot twin boat driving at full steam at a tree that cannot be seen. An iron beam taking the shock.
Two wheels, each eighteen feet across, beginning to turn. A chain tightening, and a tree that had been lying in the Mississippi—a tree that might weigh sixty or seventy tons—coming out of the brown water into daylight. There is no hydraulic pump on that boat. No pressure line, no cylinder, no oil-driven piston.
There is a river captain who spent twenty years dodging these trees and finally decided to stop maneuvering around them. There is a steam engine turned to the work of hauling, and a winch used patiently, tree after tree, until by the 1840s more than 160,000 snags had been pulled from the western rivers. When people imagine the great age of the Mississippi steamboats, they picture paddle wheels, tall chimneys, endless white decks. They do not picture the squat twin-hulled boat working the river ahead of them.
But every one of those steamboats was traveling a river someone had cleaned first. And on a dark night at a bend near St. Louis, a single hidden tree was enough to remind everyone that the work could never stop.
That history survives in the reports of army engineers, in a handful of hand-colored photographs from 1873, and in newspaper accounts of a winter night when survivors were pulled from freezing water.
