At 6:00 on the morning of September 9, 1818, Peter Hagener walked through the stone archway at Eleutherian Mills on Brandywine Creek in Delaware and took his position at the pressing wheel in the eastern rolling mill, the same spot he had occupied six days a week for eleven years. The mill was quiet except for the sound of the creek running fast after two days of rain, and the low rumble of the water wheel as it began to turn when the mill operator above opened the gate of the channel. Hagener had done this work long enough that he had learned not to think about it consciously. Thinking about it meant thinking about the gunpowder dust that no amount of sweeping ever fully removed from the stone floor, and the granite millstones turning slowly above the charge, and the fact that something in the mixture of charcoal, saltpeter, and sulfur moving between those stones carried the energy of a small cannon, compressed and patient, waiting for a single small mistake to set it off.

At 7:11, a mistake came. The building of the eastern rolling mill vanished in a fraction of a second. The blast wave crossed Brandywine Creek, shattered every window in the stone workers’ houses packed along the opposite bank, and was heard in Philadelphia, thirty miles to the north. Forty workers died.

Some were never found. The creek ran gray with gunpowder dust for three days. Peter Hagener’s name appears on no sign at the Hagley Museum, which now occupies the mill site on the Brandywine. His name appears on the company payroll for the week before the explosion, and then stops appearing.

The names of the other thirty-nine men who died that morning are in that same ledger, if the ledger survived, which it may or may not have. The September 1818 explosion was not the first at the Eleutherian mills, and it was not the last. Between 1804, when the mills first began production, and 1921, when the last powder was made there, more than two hundred explosions took place. The mills killed workers across six generations.

The names of most of those workers are not visible anywhere a visitor to the site might encounter. This is the story of the men who ground the powder, swept the floors, shaped the millstone covers, tended the charcoal retorts, and lived with their families in stone houses two hundred yards from buildings designed specifically to direct the energy of their own destruction through the water, and who came every morning anyway, and whose work made possible every major American military operation from the War of 1812 to the First World War, and whose names no one thought worth preserving in a form that anyone walking this site today could find. The problem was the powder. America in the 1790s had almost no powder that worked reliably.

Small local powder mills were scattered across the Atlantic states, a mill in Frankford, Pennsylvania, a mill in Milton, Massachusetts, operations in Maryland and Connecticut that had been producing black powder since the colonial era. But American powder in the last years of the eighteenth century had a catastrophic reputation in military circles. It was inconsistent. A barrel of American powder might fire a projectile the required distance, or it might produce a third of the expected force, depending on the quality of the saltpeter, the fineness of the charcoal, the humidity at the time of mixing, and dozens of variables the small local mills lacked the knowledge or equipment to control.

During the American Revolution, the shortage had been acute enough to change the course of military operations. Washington’s army at the Siege of Boston in the winter of 1775 had only about nine rounds per man. Nine rounds. The British garrison in the city could have marched out through the gates and scattered the Continental Army in an afternoon had they known how close the American ammunition stores were to empty.

The powder situation was so desperate that Benjamin Franklin, representing the American colonies in France, spent most of 1776 trying to arrange emergency powder shipments from European suppliers willing to risk British interception on the Atlantic. America had gotten most of its gunpowder from Britain before the war, but the supply stopped entirely during it. France provided what it could. The mills at Frankford and Milton worked around the clock.

It was not enough, and the product was not reliable enough for sustained large-scale military operations. Black powder consists of three components: saltpeter, known then as niter; charcoal; and sulfur. Standard proportions by weight were about 75 parts saltpeter, 15 parts charcoal, and 10 parts sulfur, though every powder maker of that era had his own formula which he considered the most accurate and guarded as a secret. Sulfur came from Sicily and was relatively easy to obtain.

Charcoal was produced locally from woods that every powder mill in Europe had learned to grade. Willow was considered the best, for reasons the men of that age understood practically but could not fully explain chemically. The problem was saltpeter. High-quality military saltpeter came mainly from India, where it formed as natural deposits in the soil and had been collected and refined for gunpowder since the Middle Ages.

The British East India Company controlled the most reliable supply in the world. Secondary sources, European niter beds made of animal manure and treated soil in shallow extraction pits, produced lower grades, smaller quantities, and higher cost. Domestic American saltpeter production was negligible. The small mills operating in the 1790s depended entirely on imported materials that cost more than the local market could bear and were subject to the same maritime disruptions that had nearly ended the Revolution.

E. I. du Pont understood this problem with a precision no American powder maker of that era could match, because du Pont was not merely a manufacturer. He was a chemist.

He had trained in the French Gunpowder and Saltpeter Administration under Antoine Lavoisier himself, the man who had invented modern chemistry almost singlehandedly before the Revolution executed him by guillotine in 1794. Du Pont learned powder making as Lavoisier understood it, not as a craft handed down from master to apprentice but as a chemical process, every stage of which could be controlled to produce a predictable, consistent, high-quality product. When du Pont arrived in America in 1800 and toured the small local mills along the Atlantic coast, he did not see backward craftsmen in need of better tools. He saw men working with the right ingredients but the wrong understanding of what those ingredients required.

American powder makers made it the way it had always been made, inherited practice, rough estimates, and habits of the trade learned from English practice and never examined. What du Pont brought was not better equipment. It was a systematic understanding of what the equipment was supposed to achieve. He combined Lavoisier’s chemistry with American trade knowledge, American timber, and American water power, and from that combination grew the most important gunpowder manufactory in American history.

He chose Brandywine Creek in Delaware for geographic reasons before chemical ones. The Brandywine drops about 120 feet in its lower course before joining the Christina River, descending through a series of short, sharp falls that had powered water mills since the mid-seventeenth century. By 1800 the Brandywine Valley had been grinding grain for 150 years. The millwrights, wheel builders, and channel cutters who worked along this creek understood water power the way the Boston live-oak shipwrights understood timber.

With their accumulated multi-generational expertise, entirely practical and entirely reliable, du Pont obtained the water power his grinding and drying mills required. The Brandywine Valley had been building precisely this kind of infrastructure since 1680. He was not bringing his knowledge into an empty place. He was bringing his chemistry into a valley full of craftsmen who knew water and stone and wood at a level he could not match and did not need to learn.

He needed their hands. They needed his process. The Eleutherian mills were built at the intersection of French chemistry and American craft knowledge, and the credit for that intersection went almost entirely to the Frenchman. He bought land on the east bank of the creek from a Brandywine farmer named Jacob Broom in 1802, surveyed water rights, and began construction.

The first workers at Eleutherian mills were not powder experts. They were millwrights, masons, and carpenters who built the infrastructure, the water races directing the creek’s flow into the mill wheels, the granite mill buildings, and the stone-walled drying yards facing south. These were the men who had been building Brandywine Valley mills for two generations and brought to du Pont’s operation the same water-powered stone-work mastery that the Lancaster County timber framers brought to their great barns and the bridge builders to the bridges across Vermont’s rivers. The mill buildings they raised were not built in ignorance of what they would contain.

Consider what it meant practically to design those buildings, in granite and timber, at the hands of the men who built them and in the bodies of the men who worked inside them. A powder-grinding mill building, where mixed dry ingredients were crushed under heavy stone pressing wheels turned by a water wheel, was built of granite on three sides and left open, or framed in light timber, on the fourth side, the side facing the creek. Three heavy walls. One light wall.

The logic was absolute and brutal. When the mill exploded, which it inevitably would, the blast could not penetrate three feet of local granite. It would take the path of least resistance. The light wall facing the water would collapse.

The explosion would cross the creek and dissipate over the open bank on the far side. The immediately adjacent buildings, shielded by the granite walls, would survive. The worker inside the building at the moment of explosion would not survive under any design. The design was not meant to save the man at the pressing wheel.

It was meant to keep one explosion from becoming ten. Think what it meant for the stonemason to raise those walls, selecting, shaping, and mortaring three courses of Brandywine granite to the required thickness, knowing the fourth wall was deliberately and functionally weak, knowing the structure taking shape between his hands was not built to last but to direct force, a shaped charge casing, a sacrifice vessel built not to withstand the force inside it but to direct that force across the water away from buildings where other men worked. The masons who raised those walls were not ignorant of what they were building. Du Pont had specified the requirements.

And the men who would operate the mills once the buildings were finished understood what the design assumed about them. They built the buildings to the standards the design demanded, precisely, completely, without shortcuts, because the work required it and they were the men who knew how to do it. The workers who operated the mills operated under restrictions no other trade in America imposed on its men. Rolling-mill workers wore leather shoes with no iron nails.

Iron nails in shoe soles could strike a spark against a stone floor coated with black powder dust. Their shoes were fastened with wooden pegs or sewn, without metal. Their clothing had no iron or steel buttons. Wooden buttons, bone buttons, no metal in any pocket, and no metal tools carried into the mill buildings.

The wooden tools used inside, shovels, rakes, brooms, were made without metal fasteners of any kind. The floor sweepers had the most dangerous job in the entire operation. The floor of the rolling mill accumulated powder dust in every crack and corner of the stone, and this dust was as explosive as the charge being mixed in the press. The workers swept in near silence with wide, soft brooms, wearing leather shoes, because any sharp noise at the wrong moment, a dropped tool, a stumble against a wall, could set off catastrophe.

They did not think about it. They learned not to think about it. Peter Hagener spent eleven years at the pressing wheel, and he had learned not to think about it. Until 7:11 on the morning of September 9, 1818.

Workers in the rolling mills were trained to move slowly, speak quietly, and treat every surface and every motion as though a life depended on it, because a life did. Every morning, most mornings, across the ten thousand mornings before the one where everything stopped, the caution held. The sweeper swept. The pressing wheel turned.

The charge moved through the stones at the right rate. The powder emerging from the other end was what it was supposed to be, homogeneous, properly granulated, fully incorporated. The powder the American army rammed into its cannons at Fort McHenry and at Lundy’s Lane and in a hundred other battles where American soldiers held their positions and kept firing until the lines held. Most mornings, the situation held.

The men hired to run the mills in the first production years came mostly from the mill towns of England, Scotland, and Ireland. Du Pont needed men who had worked in powder mills before, and the powder trade in 1804 was a trade passed hand to hand the same way the knowledge of every other trade in that age was passed, from master to apprentice through years of standing beside a man who knew the work and watching his hands until your own hands knew the work too. The English powder mills at Faversham in Kent, operating since the Tudor era, trained the men who trained the men who crossed the Atlantic to the Brandywine. The knowledge was carried in bodies, not in manuals.

They were not famous men. They were merely mill workers. A rolling-mill worker at Eleutherian mills was paid a wage comparable to a skilled agricultural laborer, not a generous one. He worked six days a week in a building whose design assumed his possible death as a structural standard.

He lived, if he lived in company housing, in one of the stone row houses du Pont built along the west bank of the creek. Solid houses, well built by the standards of workers’ housing of that era, each with a kitchen garden, each with a clear view across the water of the mill buildings in which he worked. Each was positioned so families inside could see, hear, and feel what happened when a rolling mill went up. When the eastern rolling mill exploded on the morning of September 9, 1818, the residents of the row houses felt it.

The blast crossed the two hundred yards of creek, smashed the glass, tore doors from their hinges, knocked children from their chairs. Forty men died inside the mill buildings. Their families survived because the granite held, and because the light wall collapsed outward across the water exactly as designed. The design worked.

Forty men were gone. Du Pont rebuilt the eastern rolling mill within three months. The workers who survived and the workers who came after them returned to the pressing wheels the following spring. This is the fact that is hardest to absorb when standing on that site today.

Not that the mills killed forty men in a single morning, but that the men who knew what had happened, who felt it, replaced their window glass and gathered the remains of their coworkers from the creek bank, and returned every morning through the same stone archway to the same pressing wheel because the work required it, and they were the men who knew how to do it, and the country needed the powder. The powder produced at the Brandywine mills was not a single product. The operation made at least a dozen types of black powder designed for specific applications. Cannon powder, large-grained, dense, slow-burning for artillery where the projectile needs sustained push down the barrel rather than instantaneous force.

Musket powder, finer-grained, faster-burning for small arms where the charge must ignite completely before the ball leaves the short barrel. Blasting powder for mining and canal construction, coarser and cheaper than military grades. Fowling powder for the local market, where consistency mattered to a hunter as it never did to a man loading a cannon in the dark. Each type required different proportions, different pressing pressures, different granule sizes as it left the granulating mill, different drying times.

The rolling-mill workers did not work from written specification sheets. They worked from their knowledge of how the pressed charge looked and felt and smelled at each stage of production. How the color of the pressed cake indicated the degree of incorporation. How the feel of the granule surfaces leaving the granulating mill told you whether the moisture balance was right before the charge entered the drying house.

This was the same knowledge charcoal burners used when reading the color of kiln smoke to judge the stage of burning. The same knowledge shipwrights used reading the grain of planks to judge where they would split under stress. Knowledge that lived entirely in the body. Knowledge that could not be written down in any form that fully captured it.

Knowledge that died with its holder unless passed to an apprentice in time. The charcoal used at the Brandywine mills was not bought from outside suppliers. It was made on site from willow trees grown in managed stands along the creek, pruned on cycles that produced shoots of the right age and diameter for the roasting process. Willow charcoal had been understood in the European powder trade since the sixteenth century.

The properties of wood charcoal vary enormously depending on species, age, and burning conditions. The finest military powder required charcoal made from young willow shoots, burned in sealed iron retorts at controlled temperature to a specific degree of carbonization. The charcoal burners at the Eleutherian mills would not have known the carbon percentage or the Celsius temperature their process achieved. What they knew was how the charcoal looked and smelled at the right stage.

The shift from the chocolate brown of incomplete char to the blue-black of properly treated willow charcoal. The sound the retort made when the moisture had fully driven off. How long to hold the temperature before beginning controlled cooling. How slowly to open the retort after cooling to ensure the fresh carbon did not ignite on contact with air.

This was not knowledge found in an instruction manual. It was the same knowledge the live-oak purchasing agents carried when they walked the coastal Georgia forests reading tree shapes for the specific curves required by the hull of the Constitution. Knowledge in the eye and the hand and the nose. Knowledge that took years to acquire and died with its holder unless passed on in time.

Applied correctly, it produced the precise material the process required, consistent, active, properly carbonized willow charcoal that, combined with refined saltpeter and sulfur at exact weights, ground, granulated, and dried correctly, produced powder that ignited reliably every time, in every weather, at every range, in the cannons and muskets of the American army for 117 years. By 1811 the Eleutherian mills were producing about 220,000 pounds of powder annually. At the outbreak of the War of 1812, the mills were the largest, most reliable, most consistent powder manufactory in the United States. When the British attacked Washington in August 1814 and burned the Capitol and the White House, the American army’s ability to respond depended entirely on powder that Brandywine mill workers had ground, granulated, and packed into barrels inside those stone buildings along the creek.

Powder du Pont had supplied to the government under emergency terms and at a reduced price the family could not fully afford. In September 1814, one month after the burning of Washington, a British naval force attacked Fort McHenry at the entrance to Baltimore harbor. The bombardment lasted twenty-five hours. The American garrison kept firing its cannons through the night in response to the rockets and shells.

A Washington lawyer named Francis Scott Key, watching the attack from a ship in the harbor under a flag of truce, wrote a poem about what he saw from the deck. The rockets’ red glare was British artillery fire. The bombs bursting in air were British shells. The flag that was still there in the dawn’s early light was American.

The cannons that kept firing through the night to keep that flag up were loaded with powder ground by Brandywine mill workers under stone pressing wheels, wearing leather shoes with wooden pegs, in buildings built to sacrifice them if necessary, on the mornings Peter Hagener and the men beside him walked through the stone archway and took their positions at the pressing wheels and did the work the country required and that no one else in the country knew how to do. Francis Scott Key’s name is on a bridge in Baltimore. The mill workers who made the powder that made the anthem possible, their names are in a payroll ledger in Delaware, in ink fading since before the Civil War. Black powder production on the Brandywine continued through the Mexican War, through the Civil War, where both Union and Confederate forces relied on du Pont powder at different stages of the conflict, through the Spanish-American War, and into the early years of the First World War.

The specific men who worked the mills changed across those generations. Irish immigrants in the early decades, German workers in the middle period, Black workers in the post-Civil War years, Italian immigrants by the end of the nineteenth century. The explosions continued. Major disasters in 1847, 1854, 1890, and 1906.

Two hundred explosions across 117 years of production. Each explosion was followed by rebuilding. Each rebuilding was followed by new workers taking the same positions at the same pressing wheels in the same buildings rebuilt to the same design. Three granite walls, one light wall facing the creek, resumption of the same work under the same restrictions and the same leather shoes with the same wooden pegs.

The men who built the Brandywine mills, and worked in them, and rebuilt them after every explosion, were not naive about the terms of this arrangement. The design of the buildings made those terms perfectly clear. Three walls to contain the blast, one wall to release it. Families in stone houses in plain sight of the release wall.

If you lived in the row houses on the west bank of the Brandywine, you knew what the designer assumed about your husband’s life, and about your own. They came anyway, generation after generation, because the work required it, and they were the men and women and children of families who knew how to do this work, who inherited the knowledge exactly as the shipwrights of Hartt’s Yard had, through years of watching and practicing until the work was in the hands and the body, not just the mind. The Hagley Museum stands on the Brandywine. Its stone foundations remain fixed in the creek bank where the water races direct the flow.

The granite is local stone, cut and shaped by the masons who built these structures in 1802 and 1803, knowing what they were building. It has not moved. It did not move in September 1818 when the eastern rolling mill exploded forty yards away. The blast went the other way.

The granite held. The stone workers’ row houses stand on the opposite bank, two hundred yards across the water. Solid construction, stone laid with skill. The kitchen garden plots are now grass.

The window sills are the same granite cut from the same local outcrops as the mill foundations. You can walk the path from the row house doors to the stone archway of the mill complex in about two minutes. The men who lived in those houses walked that path every working morning and came back every evening of those mornings that did not end at the pressing wheel. In November 1818, two months after the explosion, A.

E. du Pont wrote a letter to his brother Victor describing the rebuilding of the eastern rolling mill and the resumption of production schedules. The letter still exists in the Hagley Library, which holds the du Pont company records. The letter discusses the new construction, the insurance settlement, expected production figures for the coming year, and the name of the supervisor who coordinated the rebuilding work.

It does not name the forty men who died in September. It does not name the men who returned in October. This is not an indictment of du Pont as a man. By the documented record, he was a more conscientious employer than most of his era.

The company housing was solid, the wages were competitive. He is documented as having provided direct financial support to the families of workers killed in the mills. The absence of their names in his letter is not evidence of indifference. It is evidence of a documentary culture that recorded workers when they were paid, sometimes when they died in numbers large enough to warrant insurance documentation, and at no other times.

That absence is not personal, it is structural, and it is just as destructive as if it were personal. The man who swept the floor of the eastern rolling mill in his leather shoes with wooden pegs, at six in the morning, knowing what the floor dust was and what the stones over him were doing to it, who had been coming to that floor for eleven years before that morning, is gone. That man knew his work with a precision and constancy that no engineering document of that era required anyone to acknowledge. The knowledge ran in his body.

The constancy was in his choice to return every morning. Acknowledgment is the one thing we owe him that we have not paid. Stand at the foundation of the eastern rolling mill and absorb the meaning of the work there, not as a historical exercise but as a series of factual calculations. Six thousand pounds of black powder in the granary sixty yards away.

A floor darkened with gunpowder dust that no amount of sweeping ever fully removed. Leather shoes with wooden pegs. A granite wall behind you built by masons who knew what they were building. A light wooden wall facing the creek, designed to fail toward the water.

And the absolute structural certainty that the people who designed this building planned for your death before the first stone was laid. Peter Hagener took his position at the pressing wheel at 6:00 on the morning of September 9, 1818. He had done this ten thousand times before. He knew the work as a craftsman knows it, through the hands, through years of practice until the knowledge was no longer in the mind but in the body where the mind could not interfere with it.

He was right about the powder. He was right about the process. On ten thousand mornings before September 9, the charge moved through the stones correctly, and the powder came out correctly, and the barrels were filled correctly, and shipped to government depots that supplied the cannons that held the lines that kept the country together. He was right on every one of those mornings.

On the morning of September 9, 1818, something went wrong in the eastern rolling mill, of which we have no precise record because the men present at the moment of failure did not survive to testify. But for ten thousand mornings before that day, Peter Hagener was right, and the powder held, and the cannons fired, and the country those cannons defended is still here, built in part on the work of men whose names are recorded in a ledger in a library in Delaware, in ink that has been fading for two hundred years. The flag was still there in the dawn’s early light. It was still there because men like Peter Hagener showed up every morning in their leather shoes to a building designed to kill them, and did the work the country required with a knowledge in their hands that no one else possessed.

Their names are not in the anthem. They are in the ledger, if the ledger still exists. They deserve better than the silence they were given.