In the autumn of 1883, two miners swung their picks into a four-foot coal seam near Birmingham, Alabama. The pick of David Proser struck something that rang like a bell underwater. By noon, they had exposed three straight, sharp edges and a toothed ring wider than a man. It was a machine, and it lay inside 300 million years of unbroken coal.
The Creek, removed from Alabama decades earlier, had already said what it was: something left behind. But no one asked them. For 140 years, two records—one in a field book, one in the mouths of living people—described the same thing, and nobody set them side by side. Until now.
The coal was clean on every side. That’s what geologist Amos Rudd wrote in his field book, underlining it twice. No crack, no clay vein, no fault, no old working. The seam’s own band of bone coal bent over the object like a blanket drawn over a sleeper. The metal was dry, and it was not cold.
Rudd measured the machine at 31 feet 4 inches long, heavier than a locomotive. Its barrel was sheathed in 499 rectangular plates, each set like a roof tile. In the ninth row, five seats were empty—filled with the seam’s own bedded coal. Those five plates were already gone before the coal was laid down.
The company physician, Hartwell Baskam, laid a thermometer against a blade: 57 degrees. The air was 51. He reached into the socket and waited five minutes: 63. He struck the barrel with a pick handle. The low note rang for nineteen counts.
An assay found 58% iron, 37% of an unidentified heavy metal, and almost no sulfur. The object was not pyrite. It was not a fossil trunk. And when the mine superintendent ordered the hillside stripped, daylight fell on the machine for the first time in longer than anyone could reckon.
Then came the rain. Village Creek flooded the cut, and the company ordered it filled with gob. The machine vanished under 1900 cartloads of mine waste. But Rudd kept a pencil rubbing of one loose plate—eleven marks in a single row. And in 1759, an old Creek chief had described seven plates taken from a great thing beside a creek, in a place where the ground is black. Three tongues at one end. An open mouth at the other. Teeth in the middle. Skin scaled like a roof.
The measurements match: 18 inches long, 7 wide, five shorter and narrower, two round seats 18 inches across. Four numbers agree at once, from two men who never met, 32 years apart.
But the plates of Tukabachi are copper and brass. The machine’s plates were neither. And those five empty seats were full of coal 300 million years old—coal that closed before there was a town or an ancestor anywhere on earth. The old people took as much as they could carry. The rest they left where it lay. That is where this story begins.
So the question is no longer whether the thing existed. The question is what it was, and why the two records refuse to close.
In 1899, an engineer named Harlon Vossberg put a churn drill down on the bank of Village Creek, testing ground for a coal washer. The first thirteen holes went as expected. The fourteenth stopped at 41 feet 6 inches. A steel bit that should have lasted a full shift went dull in four minutes. What came up the hole was not coal and not stone. It was bright, pale, gray, curled shavings.
Vossberg drilled around it—seven more holes, four of which struck the same thing at the same depth. He joined the points and drew a shape: 31 feet from end to end, lying 20 degrees east of north, tilted gently toward the creek. He had never seen it. He had found its length to within four inches of Rudd’s measurement.
At the seventeenth hole, the bit struck metal two and a half feet shallower than the rest. The log says the drill rods rang and went on ringing after the engine was stopped. Vossberg bailed water from that hole and took its temperature. It was March. The creek stood at 54. The water from the hole stood at 66.
He asked the old hands at the mine. Most knew nothing. One did. Isam Tarant was 45 now, still cutting coal at Pratt. He looked at the chart, put his finger on the seventeenth hole, and told the engineer he had found the west cut. He did not hesitate.
Then he told Vossberg the rest—the rain, the timber frame, the doctor’s thermometer, the man from the state who carried a piece away. And one more detail, which Vossberg set down in quotation marks: when the cut was being filled that spring, the last thing the gob covered was the toothed rings. For a week, the tips of the teeth stood up out of the fill like a comb. Then they went under too.
Vossberg’s report ran to two pages. It recommended moving the washer 200 feet to the north. The reason it gave was a hard mass in the ground of a nature that could not be determined. The recommendation was accepted. They moved the building, and they went back to work. In the margin, in pencil, Vossberg wrote one question: Where is the piece?
The answer is known, halfway. In January of 1884, Rudd sent a box from Tuscaloosa to the United States National Museum. The express receipt is still tucked into his field book—number 4,417, weight 44 pounds. Inside, wrapped in a sack, was a single plate.
The museum’s reply was a printed card, dated February 2, 1884. No signature, only a clerk’s two initials. It acknowledged receipt of one box, and for the contents it gave one word: iron.
There was no accession number on the card, and no such entry in the museum’s accession book for those three months. Nobody hid it. It is only that nobody opened it and looked, or somebody looked and did not write it down.
Rudd waited a year. In 1885, with his own money, he had an eleven-page pamphlet printed at a job shop in Tuscaloosa. Plain Notes on an Obstruction in the Pratt Seam. Two hundred copies. Three are known today, and the word machine appears in none of them. Rudd never once used it.
One copy is in Montgomery, in the state archives. It came in at the bottom of a donated box and was carded under the heading coal, miscellaneous. After that year, Rudd’s name does not appear in the survey’s lists. There is no record that he was dismissed. A volunteer is not dismissed. He is simply not sent for. He died in Utah in 1891, forty-two years old.
The photographs went by a shorter road. Elam Forner, the traveling photographer who exposed four glass negatives in the cut, died in 1908. The next year, in Columbus, Mississippi, his widow sold six hundred glass plates to a nurseryman. The nurseryman scraped the pictures off them and glazed his greenhouse.
That left the field book, and the field book had a weakness of its own. The man who wrote it was dead, and the man who might have read it was busy. Eugene Smith stayed at the head of the survey for fifty-four years, and in that time he drew up very nearly the whole geological map of Alabama by himself. He had no reason to go back to a volunteer’s field book about a cut that had been filled. He did not go back.
In 1910, the survey moved into the university’s new building, Smith Hall. The field books were shelved by county. On the spine of the seventh, somebody pasted a label: Walker County coal sections. The Pratt mines are not in Walker County. They are in Jefferson County. A book put on the wrong shelf is not lost. It only stands where nobody who is looking for it will ever look.
It stood there for sixty-four years.
In 1974, a graduate student named Dolores Kirkland was mapping the convict camps at Pratt. For comparison, she asked for the Walker County sections. In the box that was set in front of her, she found the seventh field book, and tucked inside it, Kesler’s letter, the express receipt, and the museum’s card. All of it.
She wrote a chapter of her thesis about it. Her adviser said the chapter lay outside her subject, and he was right. The chapter came out. Kirkland’s photocopies stayed among her own papers. She did one more thing. In 1975, she wrote to the museum. The answer was courteous and short. The accession records for 1884 showed no such piece from Alabama, and there was no means of tracing a piece that had come in without a number.
No number. That letter lies in the same folder as the card of ninety-one years before. The two of them say the same thing.
And inside the back cover of the field book, folded once, there was one more sheet of paper. Eleven marks in a single row, rubbed in pencil. The plate is gone. The negatives are gone. The cut is full. But that sheet is the exact impression of something a man held in his hands in the rain in November of 1883. And today, in a reading room, it is laid in front of anyone who asks for it.
Nobody has yet been able to read it.
There is a version of this story in which everything resolves into comfortable categories. It deserves to be told fairly.
In that version, the thing in the Pratt cut was the trunk of a scale tree, just as Hadley said. The only geologist who ever saw it in daylight was not even a salaried geologist. He was a school teacher, well-meaning, careful, and alone.
The warmth has an explanation, and coal miners know it well. When pyrite meets the air for the first time, it oxidizes, and the reaction gives off heat. Most of the fires that start by themselves in coal mines start that way. A mass of pyrite sealed for 300 million years, opened to the air, could easily stand twelve degrees warmer than the morning. The warm water Vossberg bailed up sixteen years later is the same chemistry, only in groundwater instead of in air.
The period has to be taken into account as well. The America of 1883 was used to wonders coming up out of the ground. Fourteen years earlier at Cardiff in New York State, a ten-foot stone figure of a man had been dug up. Tens of thousands of people paid to look at it, and within a few months it was shown to be a fake carved out of gypsum. People remembered that. And the only account of the Pratt cut that reached the public that autumn came from a weekly paper that lived by its circulation.
Kesler’s letter is a letter from an assay office. A metal that an assayer could not identify in 1883 is not an unknown metal. It is only something that could not be named on that bench on that day. And one pinch of filings does not stand for a whole mass.
The measurements are a subtler matter. Pickett’s book was printed in 1851, thirty-two years earlier. Nothing would be more natural than to find it on the shelf of a science teacher in Tuscaloosa. 18 by 7 may have passed from the book into the field book. Not as fraud—a man tends to measure the number he already knows.
Yah’s account was written down in 1912, twenty-nine years after the column in the Iron Age. In one generation, a piece of news can turn into a telling. That may be exactly why Swanton did not print the page. He did not judge it to be old.
And the plates themselves—most historians think the brass and copper plates of Tukabachi are European work. The Roman letters people glimpsed may be the stamp of a Spanish founder.
Every one of these points is reasonable. Not one requires bad faith. Nothing here was suppressed. It was simply not followed up.
But this also has to be said. The way to test all of these points at a single stroke is known, and it is not expensive. The place is known, and Vossberg’s chart fixes it to within a few feet. A single core 42 feet deep will tell a scale tree from a machine.
Since 1899, nobody has put a drill down there. The point on Vossberg’s chart falls inside an empty green strip where no one lives. There is no archaeological survey on record and no geophysical one.
Vossberg’s report lies in an archive in Birmingham. The inventory of the folder has a one-line entry: bottle, metal filings, one. Those filings have never been analyzed. Every instrument Kesler lacked in 1883 stands today in any university laboratory.
In Oklahoma, the Muscogee towns still keep their square grounds and still light their fires at green corn. Whether and where the plates of Tukabachi are kept today is not something the town has ever chosen to tell outsiders.
In 1937, an eighty-one-year-old woman named Lucinda Tiger, a granddaughter of one of the men who carried them, told a fieldworker a single thing. The carriers never said that they owned the plates. They said that they held them. They did not say for whom.
And one thing was never done. In 140 years, nobody took Rudd’s sheet of paper and carried it to Oklahoma. Nobody laid those eleven marks in front of the only people who might know them and waited without saying a word.
Rudd did not write an answer. Yah did not give one. He only raised his two hands. Each of them left what he had seen or what he had heard exactly as it was and stepped back.
Both records are still where they were—one in a reading room, the other on a square ground. Between them lie nearly a thousand miles and a single question.
Which should be done first: put down the drill, or carry the rubbing to Oklahoma?
Report this article
Select a reason for reporting:
