On the morning of October 21st, 1879, in a silver mine beneath the hills of Tombstone, Arizona, a Cornish miner named Elias Penrose felt the point of his drill steel strike iron. Penrose was not normally the man who held the steel that morning. He was the man who swung the hammer. But his shoulder was bothering him, so he had traded places with his partner, Michael Dulan, and now he was the one holding the bar against the rock, feeling every blow of the eight-pound hammer in the palm of his hand.

Without that small change, the hole would most likely have been put down to a hard streak in the rock, loaded with powder, and fired. That surface would have been buried under broken rock. No one would ever have seen it. Penrose brought his candle closer.

It was not quartz, and it was not a broken bit left in an old hole, because no hole had ever been drilled there. Ahead of him lay only limestone, a dry gray wall that had not moved in a hundred million years. Inside the rock, a dark, flat surface about a hand’s width across had come into view, and its edge was far too straight. The men cut the rock away with their picks.

It took an hour, then two. What came out of the wall was a cutting edge, straight, ground to a bevel on both sides. Every man in that drift knew the shape. He had held it in his hands that morning.

It was the bit of a drill steel. But this edge measured thirteen feet nine inches from corner to corner. It took seven weeks to find the other end, seventy-one feet away. Its head stood as high as three men.

Its shaft was so thick that two men could not have joined their arms around it. And its far end was battered and rolled outward like every steel that has ever been struck by a hammer. No human arm could have driven that steel. The foreman measured it and wrote it down in his shift book.

Three months later, the drift was closed off with timber. But the truly strange part of this story is not the steel itself. Fifty-eight years later, an Apache man of eighty-three, seven hundred miles to the east in Oklahoma, described it to an interviewer. He had never seen that steel.

He had never been inside a mine in his life, and he could not read a line of English. He said where it had been left. He said how far down it had gone. And he said whose it was.

To understand why that steel should not have been there, the hills have to be understood first, because Tombstone was not an ordinary mining camp. Tombstone stands in the southeastern corner of Arizona, east of the San Pedro River on a mesa called Goose Flats, forty-five hundred feet above the sea. To the east rise the Dragoon Mountains, and to the southwest the Whetstones. In 1877, when a prospector named Ed Schieffelin began working his way into these hills, he was told that the only stone he would find out there would be his own tombstone.

On September 3rd, 1877, he filed his first claim, and he named it Tombstone. The Lucky Cuss and the Tough Nut followed the next year. In March of 1879, the town site was laid out, and before that year was over, a few thousand people were living there under canvas. The mines here were not open pits.

The miners followed the ore-bearing fissure down into the ground on an incline, and at certain depths they drove drifts out to the side. Before rock could be blasted, it had to be drilled. And in 1879, the drilling was done by hand. It took two men.

One held a bar of steel about a yard long against the rock and gave it a quarter turn after every blow. The other swung an eight-pound hammer. The working end of the bar was ground to a wedge. The other end, under the hammer, slowly flattened and rolled outward like the cap of a mushroom.

Every morning the camp blacksmith dressed those ends again. In Tombstone, every miner knew both ends of that tool by touch, in the dark. The limestone of these hills is not young. Geologists date it to the Cretaceous, more than a hundred million years ago.

Back then, a shallow sea lay where this desert stands now. Mud settled to the bottom and hardened, and other layers piled on top. Much later, molten rock rose from below and carried silver and lead into the cracks. Solid limestone, therefore, is a kind of sealed book.

Nothing enters it afterward. If something does, it leaves a trace: a crack, a water channel, a cavity. And there is one more thing worth holding on to. Tombstone was a dry camp.

Water came in by wagon and was sold by the bucket. In 1879, nobody imagined there was water under those hills. The company’s Number Four prospect incline had been opened that summer, a mile southwest of the camp. The company was the Tombstone Mining and Milling Company, founded by the Schieffelin Brothers and their partner Richard Gird.

On the morning of October 21st, six people were working on the bottom level of that incline: two drillers, two muckers, a powderman, and a fourteen-year-old tool boy who carried the dulled steels up to the blacksmith. And only three years before the steel was found, this country had belonged to someone else. In the autumn of 1872, in the Dragoon Mountains, General Oliver Howard came to terms with Cochise, the leader of the Chiricahua Apache. On December 14th, 1872, an executive order created the Chiricahua reservation: the Dragoon and Chiricahua ranges and the valleys between them.

Cochise died on June 8th, 1874 in his stronghold in the Dragoons. The place of his grave was never revealed. In the spring of 1876, the killing of two whiskey sellers was given as the reason, and the reservation was abolished. On June 12th, 1876, 325 Chiricahua were taken north to San Carlos.

Some did not go. They slipped away into the mountains of Mexico. Schieffelin filed his first claim fifteen months later. So the men working in that drift in the autumn of 1879 were digging in a country whose people had only just been taken out of it.

This matters because in this story, the company’s records and the memory of the people who were removed will travel by two separate roads. For generations the Chiricahua hunted in these mountains, gathered mescal in them, and gave the places their names. The refuge among the granite towers of the Dragoons is still called Cochise’s stronghold today. The Tombstone hills lie a day’s walk from it: low, gray, treeless hills between the mountains and the river.

And among the Chiricahua accounts of the time before people, there was one about someone who dug. No company record mentions that account, and no Apache account mentions the company. The two lines never crossed. One stayed in a shift book, the other in an interview transcript.

Fifty-eight years and seven hundred miles lay between them. The first record of this story is a mining company’s shift book. The shift boss at the Number Four incline was a man named Nathan Garrow. He had gone underground at sixteen and had been there for thirty-one years, fourteen of them in Nevada on the Comstock.

He had come to Tombstone in January of that year. In camp they called him Tally Garrow because he wrote everything down: how many cars of ore came out, how many candles were burned, which steel broke, and who was late. Which is why the page he filled on October 21st is so strange. The shift book survives.

It is the only book known from that incline, a clothbound book with pages warped by damp. Most of the entry dated October 21st, 1879, is routine: fourteen feet of advance, two broken steels, one box of candles. Then at the bottom of the same page, a single paragraph begins, written in a different ink. The gist of the paragraph is as follows.

About ten in the morning, at the face of the lowest drift, E. Penrose’s steel struck iron. The object is in solid rock; above it, below it, and on both sides there is no crack, no cavity, and no watermark. The part uncovered by evening is a straight cutting edge, thirteen feet nine inches from corner to corner.

Both faces are ground to the same angle. The edge points down. The edge stands 261 feet below the collar of the incline, measured straight down. And the final line: seated in its cut, not finished.

That is all Garrow wrote. He added no comment. He did not write strange, and he did not write unexplained. He did what a shift boss would do: he measured it, took its depth, and noted which way it pointed.

But he did not have to write those last two words. A shift book does not record whether an object is finished. Work is unfinished; an object is not. We do not know why Garrow wrote it.

Maybe he was simply meticulous. Or maybe he had noticed something. On the next page of the book, in the same ink, are the names of the men who were in the drift that morning: E. Penrose and M.

Dulan, the muckers Sam Tuttle and Joe Vance, the powderman Henry Lusk, and the tool boy, entered only as “Boy. ” This list is no coincidence. Under company rule, the names of witnesses went into the book for every unusual occurrence in the mine, for accidents and for cave-ins. In other words, Garrow had treated this as seriously as an accident.

That night, Garrow left two men on watch at the mouth of the drift. According to the book, the next morning the mine was as dry as it always was. Only the rock where the edge rested was damp. By the next evening the whole camp knew, and the seven weeks that followed made it more serious still.

The steel could not be brought out, so the men dug along it instead. Behind the cutting edge, the head swelled like an egg and filled the end of the drift like a door. Behind the head sat two thick collars. Then the shaft began: round, straight, running up into the hill at an angle of 31 degrees.

The steel lay inside a hole. The hole was round. Between its wall and the metal there was a hand’s breadth of space, and that space had filled with white crystal. As the digging went on, miners gathered at the head of the incline after their shifts and were let down a few at a time.

According to Penrose’s letter, one of them stood in front of the head and held his pick handle up as high as he could reach, and he could not touch the top of it. Nineteen men stood shoulder to shoulder along the shaft, and the shaft went on past them. On December 9th, the men reached the tail. The shaft ended in a heavy collar studded with bolt heads and a short, broad drum, and behind the tail the hole went on, but it was not empty.

For the forty feet the men were able to follow it, it was packed solid with white spar. The hole had closed behind the steel. The company’s surveyor, Edwin Coyle, measured it over three days with transit and chain: 71 feet 4 inches in length, 16 feet 2 inches across the head, and 5 feet 3 inches through the shaft at the collars. By Coyle’s arithmetic, if the thing was iron clear through, it weighed more than 700 tons.

What happened at the tail we know not from Garrow’s book, but from a letter Penrose wrote to his brother in Cornwall. The company blacksmith was Otto Brandt. He was fifty-one. He was the tool boy’s father, and for thirty years he had been forging pick points and dressing drill steel.

According to the letter, Brandt scraped away the pale gray crust on the drum with a chisel. Underneath was a dark, dense, smooth metal. There was no rust. Then he ran his hand around the rim of the drum.

The rim was battered and rolled outward. Brandt knew what that was. It was what he put right on his anvil every morning of his working life: the end of a steel that has been struck. According to the letter, he turned to the men and said, “Something has been hitting this.

”

The next day, they took Brandt down to the other end, to the edge. In the rock where the edge was seated, there was a straight groove as long as the edge itself, and crossing that groove at a quarter turn, a second groove, an older one: the mark every driller leaves at the bottom of a hole, the mark of turning the steel between one blow and the next. According to the letter, the blacksmith said this: “There is no forge in this territory that could draw a bar that size. And if there were, there is no hammer to strike it.

And if there were a hammer, there is no hand to turn it between blows. ”

No one answered him. By then, the story had already leaked out. On November 15th, 1879, a nine-line item appeared on page three of the Arizona Weekly Citizen, printed in Tucson.

The headline read, “A curiosity at Tombstone. ” The item reported that a large iron bar of olden times had been found in a company shaft and that it might be a relic of the old Spanish mining of the region. Its length was given as about thirty feet. Nine lines.

It was never followed up. Three details stand out in those nine lines. First, the length. By that date, far more than thirty feet had been uncovered, so the story came not from anyone who had measured the object, but from talk going around town.

Second, the source. There is no reporter’s name under the item, only the phrase “as reported by one of our readers at Tombstone. ” Third, and most important, the place. The item says the object was found in a shaft.

It does not say it was inside solid rock. It gives no depth. It says nothing of a hole that had closed behind it. So, the story that leaked out in 1879 was a story with its strangest part left out.

The camp’s own paper, the Nugget, had started printing that autumn, and it printed nearly everything to do with the mines. In its November and December issues, there is not a single line about the steel. The company was already writing letters. On December 16th, 1879, the company’s agent in Tucson, Philip Dorance, wrote to the office of the United States Surveyor General in that same city, the federal office that recorded the surveys of mining claims in Arizona.

A copy of the letter is kept in the company’s letter book. Dorance describes the object, reproduces Coyle’s figures exactly, and asks two questions. First, is it possible for such an object to be found at that depth inside solid limestone? Second, does it affect the company’s title to the claim?

Dorance does not ask what the object is. He asks whether it could have been where it was found. The reply came on January 8th, 1880, signed by Eel Hume, a clerk in the office. It was polite, brief, and from his side of the desk entirely logical.

According to Hume, the object was most likely a relic of some old Spanish or Mexican working. Such workings were known in southern Arizona from the eighteenth century onward. An iron bar could have fallen into an old shaft, and the shaft could have filled with wash over time. As for the dimensions, they were most likely the ordinary result of measuring by candlelight.

The company’s title was not affected. A reasonable explanation, with one problem. Garrow’s book stated plainly that there was no crack, no cavity, and no watermark in the rock. The hole behind the steel had not filled with wash.

It had filled with crystal, and it was not a human eye that took those measurements. It was a transit. Hume’s letter touched none of these three points. Dorance pasted the reply into the letter book and added two words beneath it: “Matter closed.

”

On January 29th, 1880, the drift was closed with a timber bulkhead on account of bad ground. On the company’s map, that section was marked with a single hatched line. A few weeks later, the company was sold to investors in Philadelphia. In the schedule of the transfer, Number Four is a single line: “one prospect incline, abandoned.

” The drift was closed. The company changed hands. The newspaper item was forgotten. On paper, the case was closed.

But Elias Penrose was not done. Penrose left the company that spring, worked in other mines around the camp, and went on telling that morning for the rest of his life. According to his daughter, he repeated the same detail every time. His steel had not struck something.

It had come to rest against something, and the thing it rested against had not felt to him like a thing that was lost. It felt, he said, like touching another man’s tool on another man’s bench. This sentence is not a record. It is a memory.

And the rest of this story takes place precisely in that gap between record and memory. The second record of this story does not come from a mining company, nor does it come from a museum. It comes from nine typewritten pages written up from notes taken under a brush arbor in Oklahoma. The story of the Chiricahua who were taken to San Carlos did not end in 1876.

After ten years of breakouts and pursuit, in 1886 the government declared the entire Chiricahua people prisoners of war. That included the men who had served the army as scouts. All of them were put on trains and sent to Florida. From there they were moved to Alabama, and in 1894 to Fort Sill in Oklahoma.

The captivity lasted twenty-seven years. When they were released in 1913, about 180 of them went to the Mescalero reservation in New Mexico. Some eighty stayed in Oklahoma on small allotments around the towns of Apache and Fletcher. In the early 1930s, two young researchers from the University of Chicago came to Mescalero.

One was the anthropologist Morris Opler. The other was the linguist Harry Hoijer. Hoijer took the narratives down in the Apache language word for word. Those texts were printed in 1938 by the University of Chicago Press with notes by Opler under the title Chiricahua and Mescalero Apache Texts.

In 1942, Opler published a second volume, An Apache Life-Way, and similar accounts appear in his later collection of Chiricahua tales. Those books are still considered the most complete written record of Chiricahua oral tradition. Most of the people who spoke to them belonged to the generation that came back from captivity, people whose childhood had been spent on trains and in barracks and who had grown old at Mescalero. Those who had seen the country of Cochise with their own eyes, who had grown up in those mountains, were very few by the 1930s.

But they were still alive. And one of the accounts in those records is about someone whose arrows were the size of logs. According to the account, in the beginning, White Painted Woman was living, and one called Giant was living too. People could not increase because Giant hunted them.

White Painted Woman bore a son. His name was Child of the Water, and his mother hid him from Giant. When the boy was grown, he and Giant stood face to face. Each of them was to shoot four arrows.

Giant’s arrows were logs, whole tree trunks sharpened at the end. All four missed. The boy’s arrows were made of grass. Giant wore four coats of flint over his chest.

The first three arrows knocked those coats away one by one. The fourth went into his heart. According to the account, it was only after this that people were able to spread across the earth. Arrows the size of logs, in the printed text.

That detail does not read like an exaggeration. It reads like a measurement. In these records, Giant is not only a threat. He hunts, he wants meat, he bargains, and he agrees to a contest and keeps to its rules.

In this account, in other words, Giant has a daily life. Nor do these accounts take place in an empty geography. Chiricahua narrators name the mountains, the springs, and the passes. These are places that could be marked on a map.

But there is no digging in Hoijer’s book. There is no hole. There is no stick. Those come from another record.

And that record was written a year before Hoijer’s book was printed. In the second half of the 1930s, in the middle of the Great Depression, the federal government put unemployed writers and teachers to work on public projects. One of those projects in Oklahoma was called Indian-Pioneer History. It was Project S-149 of the Works Progress Administration, sponsored by the Oklahoma Historical Society and the University of Oklahoma, and directed by the historian Grant Foreman.

In 1937 and 1938, between eighty and one hundred field workers taken from the relief rolls traveled the state and wrote down what old people told them. Those interviews, bound into more than one hundred volumes, are still held in Oklahoma archives today. In the fall of 1937, one of those field workers was a woman named Hazel Pruit. She was thirty-nine, a former school teacher from Lawton who had lost her position in the Depression.

Her assignment was to interview the old people of the district. She was not looking for anything like this. She was filling in a form. One of those interviews took place on November 9th, 1937, a few miles north of the town of Apache, under the brush arbor in front of a house.

On the first page, Pruit noted the place and the weather: a clear cool afternoon with wind. The man across from her was eighty-three years old, one of the Fort Sill Apache. While he talked, he was mending a harness strap. The transcript runs to nine pages, and the man’s name does not appear in it.

Pruit wrote only two letters: B. T. And she added a note in the margin: “Speaks Apache and Spanish, does not read. Granddaughter interprets.

”

This note matters, and why will become clear shortly. Pruit asked the standard questions in order: old tales, place names, foods, removals. So B. T.

was not asked about the steel. He was asked about old stories. And there is one more detail in the transcript. The granddaughter was nineteen and had been a student at the Riverside Indian School in Anadarko.

She put her grandfather’s words into English one sentence at a time. But on page five, there was one word she did not translate. According to Pruit’s note, the girl said, “That is a name from the old stories. There is no English for it.

” Pruit wrote the word down the way she heard it, in English letters: Gá’yé. And she put a question mark beside it. The first four pages of the transcript are the ordinary record of a life: a birth in the Dragoon Mountains and a childhood in the time of Cochise; beef issued at Sulfur Springs and the wagons going north in 1876; the train in 1886 and the days he counted on that train; Florida, Alabama, Fort Sill; hay cut for the army; the allotment he was given in 1914; the corn he planted on it; and a wife he buried. Short, calm answers to the questions.

On the fifth page, the subject changes. Pruit asks B. T. about the old stories he heard from his grandmother.

B. T. begins with White Painted Woman, the boy, and the arrows in a telling very close to the text Hoijer was at that moment preparing for the press. Then, according to Pruit’s note, he is silent for a while, and he tells this.

According to his grandmother, before the boy came for him, Gá’yé was digging in the earth between the mountains and the river, where the hills are low and gray and there are no trees, one day’s walk from the stronghold. Beside the word stronghold, Pruit wrote in parenthesis: “he means the mountains of Cochise. ” Pruit asks what he dug with. The granddaughter gives it this way: with a thing like the stick the women used to cut mescal.

Apache women cut the heart of the agave from its root with a hardwood stick shaped to a chisel point, and they drove the top of that stick with a stone. Then B. T. shows something with his hands.

Pruit’s note reads: “He closed one fist as if gripping a pole and brought his other hand down on top of it. I think he means a drill or an iron bar. ”

Pruit asks why he was digging. The answer is a single line in the transcript: he was going down to the water.

Pruit asks what happened then, and in the transcript the last lines of the fourteen-line passage attributed to B. T. are as follows. The boy called him out.

So he left the stick in its hole with its mouth down, and he went out. He meant to come back. He did not come back. The top of the stick is broken soft from the striking.

His work is not finished. My grandmother used to say, “Let no one finish his work. ”

Pruit made no note on that sentence. According to the transcript, when B.

T. had said it, he put down the strap he was holding and looked toward the west for some time. Pruit went on to the next question. The rest of the transcript returns to haying, a wedding, and a dry summer.

The interview ended that evening. Nothing else in the nine pages mentions it. We do not know whether Hazel Pruit was aware of the nine-line item printed in Tucson in 1879. She probably was not; she had not been born when it was published.

But we know that B. T. did not read that item, because Pruit had written it in the margin: he did not read. And we know one thing more.

B. T. was taken out of that country in June of 1876. Schieffelin filed his first claim fifteen months later.

B. T. never went back to those hills. First San Carlos, then the train, then twenty-seven years as a prisoner, then Oklahoma.

He never saw a single shaft sunk in those hills. And in eighty-three years, he had never once been underground. And that line in Garrow’s book, “seated in its cut, not finished,” was never printed in any newspaper, any correspondence, or any public record. It stood alone at the bottom of one page of a shift book, written in a different ink.

Here we need to pause for a moment and look only at the numbers, because numbers do not make judgments. The steel a driller held in his hand at Tombstone was one inch thick. The shaft Coyle measured was sixty-three inches thick. A hand that holds a steel has to close around it.

A hand that could close around a sixty-three-inch shaft with the same ease would have to be more than sixty times the size of a human hand. Not an interpretation: a ratio. The second number is harder still. A drill steel a yard long weighs about eight pounds.

The hammer that drives it weighs eight pounds. The hammer, in other words, weighs as much as the steel it strikes. By Coyle’s arithmetic, the big steel weighs 700 tons. At the same balance, the hammer that drove it would have to weigh 700 tons as well.

The third number is the turn: a quarter turn after every blow. The two grooves in the rock under the edge showed it. Seven hundred tons had been lifted out of its cut and turned between one blow and the next. And the rolled rim that Brandt found with his hand on the drum said the blows had come not once, but again and again.

The fourth number is the hole itself. For the forty feet the men could follow it, the hole behind the steel was packed with white spar. Calcite does not fill a cavity in a season. According to geologists, a filling of that kind grows drop by drop out of the mineral that water carries, over thousands of years.

The steel went into that hole before the hole closed. The calculation stops here. It does not provide a conclusion. It merely states that this object was not made to the measure of any arm we know.

And there is one more problem. The steel never moved from where it lay, but its trail on paper begins to disappear. When the company changed hands in 1880, a transfer inventory was drawn up. In the list of the map cabinet, there is a single line entry: “one roll, survey of obstruction Number Four.

” That is the sheet Coyle drew over those three days. The entry is missing from the inventory of 1883. In the years between, Tombstone burned twice. The fires of June 22nd, 1881 and May 25th, 1882 took most of the business district of the town.

Whether the roll burned or was thrown out or was left behind in a drawer, the records do not say. Coyle’s figures are known today only from the copy Dorance wrote into his letter. Coyle himself left the camp in 1881, and after that year his name appears on none of the company’s payrolls. One line was left of the survey sheet.

But it was not the only thing that vanished. In the account book of a photographer who came to the camp that winter, there is a line dated December 1879: “For the company, Number Four, the big steel, two plates. ” Neither of those two glass negatives is known today. In the company’s cash book for February 1880, there is a payment: “timber and labor bulkhead at Number Four, $212.

” Under the entry, in small letters and in a different hand, is a note: “Door locked, key with Mr. Dorance. ” This is the only written trace that there was a door in that timber bulkhead. And the trace points to one person.

The man who wrote the letter to Tucson, pasted the reply into the letter book, and noted beneath it, “Matter closed. ” Dorance left the company in 1881. The receipt for the keys he turned in when he left is among the company’s papers. The receipt counts nine keys.

The key to Number Four is not among them. No one has ever explained that. One of the letters Penrose wrote to his brother also gives a detail from those years. In a letter of 1887, Penrose says that when he last saw the steel, the week before the bulkhead was nailed up, the edge was still true.

The place Brandt had scraped with his chisel was still dark and clean. It had come out of rock a hundred million years old, and there was not a single spot of rust on it. This letter is not an official record; it is a family letter. But the measurements in it match exactly those in the company’s book.

Meanwhile, Tombstone was living through a disaster of its own. And the disaster did not come from fire. It came from water. In March of 1881, the shaft of the Sulfuret mine struck water at 520 feet.

There was water under the dry camp. Before long, every deep mine in the district was in water. In March of 1882, the new shaft of the Grand Central went into water at 620 feet. The companies sent for great Cornish pumps, the kind used on the Comstock.

By February of 1884, those pumps were lifting 576,000 gallons every twenty-four hours. In those same years, the town was bringing its drinking water in by pipeline from the Huachuca Mountains, more than twenty miles away, and it was spending a fortune every day to throw away the water under its own feet. On May 26th, 1886, the hoist and pump works of the Grand Central burned. The iron of the pump warped in the heat, the water rose, and all work below the water level stopped.

In a town where thousands had lived a few years before, the census of 1890 found 1,875. In 1900, a new company was formed, the Tombstone Consolidated Mines Company. The next year, new pumps started up and the mines were drained down to the thousand-foot level. In June of 1909, the pumps stopped, the boilers ruptured, and water flooded the thousand-foot level.

On January 11th, 1911, the mines were shut down. And it was in the middle of that second attempt that the steel surfaced one last time on paper. The last piece of paper in this story comes from one of the Consolidated’s engineers. His name was Carl Ostrander, and he was twenty-seven.

To work out how the water moved through the old workings, he had to copy every old mine map in the camp, because the abandoned drifts were the places where water found its way. Ostrander’s field notebook is now among the company’s engineering papers. On a page dated September 14th, 1907, he copied one of the old company maps. He drew a hatched line at the bottom of the Number Four incline.

Next to it he wrote, “Number Four, bulkheaded 1880. ” And below that, in quotation marks, he copied exactly a note that someone had written on the map in pencil: “where the big steel lies. ”

Ostrander added his own comment to that note: “Bulkhead sound, dry, not entered. ” Between 1880 and 1907, that brief note is the only trace of the steel that survives outside the company’s own books and one family’s letters.

The original map is lost, but someone at some point had written this in pencil next to that line. And an engineer in the middle of a fight against the water that was drowning the whole camp thought it worth copying down. Why? There is no answer in Ostrander’s notebook.

Now we have two records that were never meant to meet. The first is a company’s internal paper: a shift book of 1879, a copy of a letter, a line in an inventory, and a pencil note copied by an engineer in 1907. None of the people who wrote these records were interested in Apache narratives. Garrow was counting feet, Dorance was worried about a title, Hume was answering a letter, and Ostrander was fighting water.

The second is a people’s oral tradition: texts taken down at Mescalero in the 1930s and printed in 1938, and nine pages written under a brush arbor in Oklahoma in 1937. None of those who produced these records had ever seen the company’s books. Hoijer never read a shift book in his life. B.

T. did not read at all. Neither side could have copied the other. That is exactly why what follows matters.

Let us place these two records side by side without adding anything. The company record states the object lay in solid rock inside a round hole. The oral record states he left the stick in its hole. The company record states the place is the Tombstone hills, between the Dragoon Mountains and the San Pedro River.

On a map, that is sixteen miles from the stronghold. The oral record states: “Between the mountains and the river, where the hills are low and gray and there are no trees, one day’s walk from the stronghold. ” The company record states the object is a drill steel, its end ground to a wedge, a tool that is driven from behind. The oral record states: a stick with a cutting point struck on its top, one fist closed around a pole, and a hand coming down on top of it.

The company’s blacksmith states the rim of the tail was battered and rolled outward: “Something has been hitting this. ” The oral record states: “The top of the stick is broken soft from the striking. ” And the text Hoijer printed in 1938 states: “His arrows were the size of logs. ”

Now, the last line.

The company record, that paragraph Garrow wrote in a different ink, says: the edge points down. “Seated in its cut, not finished. ” The oral record says: he left the stick with its mouth down. He meant to come back.

His work is not finished. These two lines converge on the same five details: a hole, a place, a cutting tool, a battered top, and a piece of work left unfinished. But there is one more page of the transcript. On the back of page five of Hazel Pruit’s nine-page transcript, three lines were added later in pencil.

They are in Pruit’s handwriting, and they are undated. Apparently after the interview, perhaps that evening, perhaps the next day, she added a detail she had recalled. The note reads: “I asked B. T.

how deep he had dug. ” “Half,” he said. I asked, “Half of what? ” The girl gave it as half the way to the water.

Half the way to the water. According to Garrow’s book, the edge of the steel stood 261 feet below the collar of the incline. In 1879, nobody knew there was water under those hills. Tombstone was a dry camp.

The water was found seventeen months later, in March of 1881, in the shaft of the Sulfuret, at 520 feet. 261 is half of 520, to within one foot. And the small detail Garrow wrote down the next morning is still on the page: in a dry mine, only the rock where the edge rested was damp. B.

T. did not give a depth. He did not point to a map, and he did not say the name Sulfuret, and Pruit did not connect those three lines to anything. But the note stands there: half the way to the water.

Now there are not five but six details: a hole, a place, a cutting tool, a battered top, work left unfinished, and a depth. And there is still one word left to compare. The word Hazel Pruit wrote down as she heard it on page five, with a question mark beside it: Gá’yé. In the Apache text Harry Hoijer printed in 1938, the name of the one whose arrows were logs is written in Apache spelling, in letters Pruit had never seen put together.

Said aloud, it is the same word: Gá’yé. Pruit could not have read Hoijer’s book; it was printed a year after the interview. If she had read it, she would have recognized the word, and she would not have put a question mark beside it. B.

T. could not have read any book at all. The word did not make its way from a book into a transcript. It passed from a grandmother’s mouth to a child’s ear, and from there, more than seventy years later, to a field worker’s pencil.

In other words, the one B. T. described bore the same name as the one in the text Hoijer printed. And the thing he was said to have left behind had the same shape as the thing that came out of the rock at Tombstone.

And Elias Penrose? After the fire of 1886, like many of Tombstone’s miners, Penrose went south to the copper mines of Bisbee. According to his daughter, he never again worked below the 300-foot level, though better wages were offered deeper down. To those who asked him why, he gave no answer.

Penrose died in 1931 at the age of eighty-one. His daughter says the last night her father spoke of the steel was a few months before his death. He told her that steel was never lost. Somebody set it down.

This is not a record. It is a sentence an old man spoke to his daughter at a kitchen table. But it sounds a lot like a sentence an old Apache man spoke in 1937. There is another version of this story, and to be honest, that version is much simpler.

Skeptics quite rightly turn first to the geology. The ore at Tombstone is made of dark oxides of manganese and iron that have replaced the limestone. Ore of that kind sometimes runs along a fissure in the shape of a pipe. Miners call it a chimney.

A dark, heavy, round chimney of ore a hundred feet long could look like a bar of metal to men working by candlelight. The swelling at its end could look like a head, and the polished straight line of a fault surface could look like a cutting edge. This explanation is compelling. Second, there is what Hume wrote in 1880.

Spanish and Mexican workings really did exist in southern Arizona from the eighteenth century onward. There were old shafts that no one had recorded, and the survey figures in this story come from a single source: Coyle’s sheet is lost, and all that remains are the figures a company agent worried about his title copied into a letter. Third, narratives become intertwined. The talk of a big steel at Tombstone had been going around since 1879.

In the 1880s, Apache scouts served in that country at Fort Huachuca and in the field. B. T. may not have read the newspaper, but he could have heard that story from a scout, from a soldier, or in twenty-seven years of captivity, from a guard who had once served in Arizona.

Fourth, the printed texts themselves. Arrows the size of logs do not have to be a measurement. They may be a way of saying how hard the fight was. And there is no digging in any of the texts printed in 1938.

What B. T. told is one telling by one man. The next objection is sharper still, because it points at the one record this story leans on most.

Fifth, and perhaps most important, are the interview transcripts themselves. The Indian-Pioneer History interviews are recognized today as a valuable source, but historians are well aware of their limits. The field workers made no sound recordings. They took notes by hand and typed the transcript afterward in their own words.

And here an interpreter stands in between as well: a girl of nineteen giving the words of a man of eighty-three. When B. T. closed his fist and brought his other hand down on it, Pruit wrote, “I think he means a drill or an iron bar.

” The word drill was not B. T. ’s. And a woman who grew up in Lawton, right beside Fort Sill, could have heard an old Arizona story from a soldier, and she could have filled the space above an old man’s raised fist with that story.

The three lines added on the back of the page are undated, too. Sixth, Garrow’s book. That single paragraph was written in a different ink from the rest of the page. It could be a note set down in the rush of that morning.

But it could also be a line added days, months, or even years later, after people had begun to talk about the steel. The same goes for “half the way. ” A half is the most ordinary fraction in any story, and the water did not stand at the same depth in every shaft. In the Grand Central, it was struck at 620 feet.

Measured against 620, 261 is not even half. All of these explanations are reasonable. They all remain within the bounds of the known world. But each of them accounts for only a part of the story.

What was an edge ground to the same angle on both faces doing at the end of a chimney of ore? How did an old shaft that should have filled with wash come to be filled from end to end with crystal? Where did a story heard from a soldier get the words “not finished,” which were never published anywhere? And how could a man taken out of that country in 1876 have known there was water under those hills before the first shaft was ever sunk?

Hume’s letter did not ask these questions. No one did. In the official records, there is a single paragraph about this steel. In the oral tradition, there is a single sentence explaining why it was there.

The gap between the two has remained unbridged for nearly 150 years. This is not the story’s weak point. It is the story itself. In the autumn of 1879, under Tombstone, the steel in a man’s hands came to rest against something inside the rock.

It was not rock. Today, about 1,300 people live in Tombstone. Every year, the town fills with visitors who come to walk its old streets. The deep workings are still underwater.

The collar of the Number Four incline caved in long ago. The timber bulkhead, if it still stands, is somewhere under one of those low, gray, treeless hills. The company’s papers have been scattered. Some were lost in the fire of 1882.

Others were lost when the property changed hands. What remains lies in different archives, in different boxes, under different names. Garrow’s shift book is one of them, and Ostrander’s field notebook is another. Hazel Pruit’s nine pages are somewhere else entirely, in Oklahoma, among more than one hundred volumes of interviews.

Three records, three separate archives. None of them mentions the others. But the granite towers of the Dragoon Mountains are still standing, and the refuge among them still carries the name of Cochise. The descendants of the people who were taken out of that country live today in the Fort Sill Apache tribe in Oklahoma and the Mescalero Apache tribe in New Mexico.

And the account of White Painted Woman, the boy, and the four arrows is still being told. A people remembered one who was called away before his work was finished. A company measured a steel seated in its cut. And in 1937, an old man gave his grandmother’s words to a woman who was filling in a form: “Let no one finish his work.

” Perhaps that work was finished in March of 1881, at the bottom of the Sulfuret, at 520 feet. Perhaps it is not finished yet, because some records lie in a shift book, and others lie inside a mountain.