In the autumn of 1827, a man was riding his horse along the mountain road above a settlement called Ashgrove. He looked toward a farm he had passed forty times in his life and saw four young women standing at the front window, their hands pressed flat against the glass.
He had seen them before. Nearly everyone in that valley had seen them at one time or another. That part was ordinary. What was not ordinary was that no living person in Ashgrove had ever seen those four women outside that house. Not once. Not on the porch, not in the yard, not on the path, not at the well, not at gatherings, not at funerals, not at weddings, not in 18 years.
Their father was the Reverend Amos Thorne, who had preached in that valley since 1809. By every surviving account, he was a mild man. He was careful. He did not raise his voice. He did not drink. He did not borrow money. He buried the valley’s dead. He married its young people. He sat with its sick. And through all of that, he never once explained why his four daughters were not allowed to leave the house.
He was asked repeatedly. Neighbors asked. Two traveling doctors asked. The county judge asked. Once, the head of his denomination asked in writing. His answer was always the same eight words, recorded identically in four separate documents over three decades.
They must not be looked upon under an open sky.
That was all. That was the complete explanation Reverend Amos Thorne gave to anyone for 18 years. He did not say they were ill. He did not say they were simple-minded. He did not say they were disfigured or afflicted or in mourning or under discipline. Only that they must not be looked upon under an open sky.
On November 9, 1827, eleven men from the Ashgrove congregation climbed that hill in the dark and broke through the front door of the Thorne house. The house was empty.
The word empty carries heavy meaning. The house was not abandoned. Abandonment leaves marks. Abandonment means open drawers. It means fine linen gone and cheap linen left. It means horses missing from the stable and money gone from the box. None of that was present.
There was bread in the oven still hot enough to burn the fingers of anyone who pulled it out. There was a kettle on the stove with water in it. A fire burned in the front room, described by a constable as strong, meaning it had been lit within the last hour. A chair was pulled up to a writing desk. A pen rested on an open book. The ink on the last word had not yet dried.
Every possession in that house was in its place. Two horses stood in the stable, fed and watered. The dog lay on the threshold. But Reverend Amos Thorne, aged 51, and his four daughters, aged by the last count anyone cared about 22, 21, 19, and 17, had vanished without a trace.
That was not what made the house feel empty. What made the house feel empty was the front window.
There were four sets of handprints on that glass. Eight hands. Small hands. Women’s hands. Fingers spread wide. They were on the inside of the pane, which is where you would expect them to be. Any sane person will tell you that handprints on a window are the least surprising thing in the world. Any sane person would be right.
Except these prints could not be wiped away. They were not grease. They were not soot. They were not smoke or shadow or oil or wax. Constable White Sample, aged 39, tried first with his shirtsleeve, then with a wet cloth, then with sand, then, by his own account, with the edge of his knife. He left no mark on them at all.
The handprints were not on the glass. They were inside it.
Sample wrote in his statement, given 11 days later, and the surviving text is exact: The marks lie within the substance of the pane as an impurity lies within ice. I could feel nothing beneath my fingers. The glass is smooth. The marks are within the glass. I do not know how to write this so that anyone will believe it, and I will not attempt to write it a second time.
He never did write it a second time. White Sample lived another 44 years. He died in 1871 at the age of 83. Throughout that time, according to his son, his daughter-in-law, and a neighbor interviewed by a local historical society in 1906, he never spoke of Ashgrove. He never returned to that hill. He never allowed a pane of glass to be installed in any window of any room in his house that faced west.
This is the story of who Reverend Amos Thorne really was and where he came from, and why there is not a single scrap of paper documenting his existence before the spring of 1809. It is the story of the twelve window panes in that farmhouse, shipped to a barren valley 11 weeks before the house that would hold them was built, which a laboratory determined in 1958 could not have been manufactured anywhere on this continent in 1809, or anywhere on Earth. It is the story of four young women whose ages did not change across three consecutive federal censuses. And it is the story of one missing page from the county record of 1827. One page. Not a file. Not a folder. Not a box. One page removed from a bound volume in a courthouse that never reported any theft.
Let us begin with the ordinary question. The ordinary question is: where did Reverend Amos Thorne come from?
The answer, as far as the documentary records of the United States of America are concerned, is that he came from nowhere. Amos Thorne first appears in writing on April 4, 1809, in the daybook of a general merchant at a crossroads about nine miles from the site of Ashgrove. The entry is short and strange. It survived because that daybook was donated to a county historical collection in 1934 and has been photographed twice since.
The entry reads: Mr. A. Thorne, preacher, four daughters, no wife with him, bought flour, 100 lbs, salt, 20 lbs, lamp oil, 4 gallons, assorted nails, one axe, did not ask for a Bible. Paid in coin, gold, foreign, unknown to me. I weighed it and took it.
There are three things in that entry you should not ignore. The first is the coin. Gold. Foreign. Unknown to me. A merchant in a border region in 1809 handled plenty of foreign currency. Spanish milled dollars circulated everywhere. So did French, Portuguese, English coins, and private issues from dozens of banks that would fail within a decade. That merchant would have recognized any of them on sight. He noted that he did not recognize these. He noted that he weighed them rather than counted them, which is what a careful man does when he trusts the metal but not the mint. Then he accepted them anyway, because gold is gold.
The second thing is the refusal of a Bible. A clergyman arriving in a desolate valley to establish a parish buys a Bible. He buys several Bibles. He buys hymnals if he can get them. Reverend Amos Thorne declined. He already had books. We will return to the books because they are the center of this entire story.
The third thing, which takes the longest to understand, is four daughters, no wife present. Not a widower. Not a deceased wife. Not a wife who would follow. No wife present. That is the phrasing of a man writing exactly what he saw and refusing to speculate about the rest.
There is no marriage record for Amos Thorne in any jurisdiction searched. There is no death record, no burial, no probate, no headstone, no mention of a wife in any letter, statement, sermon, or account book connected to this man. For 18 years of residence in a valley of about 410 people who talked about everything, no one on record ever asked him about the mother of his children. Think about that for a moment. Think about how impossible that is.
The four daughters, by contrast, appeared constantly. They appeared in censuses. They appeared in three legal statements. They appeared in the private letters of a doctor, in a glazier’s account book, in the working notes of a portrait painter, and in the memories of at least nine people interviewed decades after the fact. Every one of those accounts places them in the same location. Behind a window.
To understand the house, you must understand the hill it was built on. To understand the hill, you must go back two years before Amos Thorne ever set foot in that valley.
In the summer of 1807, a five-man survey crew was working in that part of the hills under contract with a land company that was dividing tracts for sale to settlers. This was routine work. They measured distances, blazed trees, recorded data on soil, timber, water, and slope, then submitted a survey plat. On page twelve, describing the hills above what would become Ashgrove, the crew leader wrote a sentence the employers seem to have simply ignored.
He wrote: No settlement is recommended on these heights. The compass does not point north above its summit.
The compass does not point north. There is an ordinary explanation for that, and it should be stated before it is dismissed. Magnetic anomalies are real and common. Metal in the bedrock pulls at a compass needle. Surveyors of that era encountered local declination constantly, and they had a standard procedure for dealing with it. They recorded the declination, corrected for it, and moved on.
That is not what this man did. He did not record any declination. A declination is a specific number. You write that the needle stands at 11 degrees east of true north, then you adjust your bearings by 11 degrees. He recorded that the needle did not settle. Meaning it was moving. Meaning it did not settle anywhere at all. Meaning that in the summer of 1807, a surveyor’s compass on the top of that hill swung continuously and never stopped.
He noted this once. He recommended against settlement there. Then, in the private journal he kept alongside the official notes, which was donated separately and connected to the survey plat by no one for over a century, he wrote an additional line about that hill. On the night of July 29, 1807, he wrote: The summit is bare of trees in a very regular manner. I measured it with my steps. It is circular. Nothing grows there and nothing has been cut there. The men did not want to sleep on it and I did not press them to do so.
A bare circle on a wooded hill with no tree stumps. Meaning nothing had been cut. Meaning nothing had grown there in living memory. He measured it with his steps. He did not record the diameter in his journal, but the survey plat does, because the plat classifies the clearing as open undeveloped land and gives its width as 164 feet.
Keep that number. 164 feet. It will be needed twice more before this is over. The second time in 1958. The third time in 2019.
The survey crew left. The land company offered the tract for sale in 1808. It did not sell, because it was barren hilltop with thin soil and no water source above it. On May 11, 1809, a man who had been in the county for five weeks bought the entire hilltop. All of it. 190 acres. For $475. That was twice what the land company asked and about six times what it was worth.
He paid in gold. He paid at once. He did not inspect the land first. The clerk of deeds noted, with what reads as mild professional irritation, that the buyer declined an offer to view the property, stating he was satisfied with the location.
He was satisfied with the location. He had never been up there. That is the first impossible thing about Amos Thorne. It is simple, and it is placed first because what follows is greater. He bought a hilltop he had never seen at a price no one else would pay for a purpose unrelated to farming.
For 18 years on that land, Reverend Amos Thorne never planted a single field. He kept a quarter-acre kitchen garden and two horses. He bought everything else. This was noted repeatedly at the time. A preacher in a border valley in 1810 did not live on his salary, because there was no salary to speak of. The Ashgrove congregation paid their minister in kind. Flour, pork, firewood, labor. In a good year, it came to less than $60.
Amos Thorne spent, by the surviving merchant ledgers, an average of $430 per year for 18 consecutive years. In gold. Foreign, unknown gold. Weighed rather than counted. In a valley where most transactions were settled in eggs. No one ever asked him where it came from. That is what the records show. No one asked.
Now let us talk about the house, because the house is where this story stops being the story of a strange man and starts being the story of something else entirely.
The Thorne house was built in the late summer and fall of 1809 by a crew of local men, hired and paid in cash, which was itself a heresy. Frontier houses were built with neighborly help, through exchanged labor. Thorne paid wages. Eleven men worked on that house for about nine weeks, and because they were paid, there is an account book. Because there is an account book, we know things about that house we should not know two centuries later.
It was a two-story frame house, 32 by 24 feet, with a stone cellar, a central chimney, and a shingle roof. Every third house in that mountain region looked like that. There were three unusual things about it.
The first was that it was not built over the bear circle. It was built beside it, on the northern edge, close enough that the front of the house looked directly out over that open ground. The circle was left untouched. For 18 years, according to two separate witnesses, nothing was built on it, nothing was planted on it, nothing was stored on it. Thorne kept its grass cut. He cut it himself with a scythe twice a year, and he did it at night.
That is not an inference. That comes from the testimony of a man named Caleb Ivers, who was 26 in 1827 and who climbed the hill as a boy of 12 on a dare and watched the preacher cutting a circle of grass by lantern light at what he estimated to be two hours after midnight.
The second unusual thing was the cellar. The account book records that the stone cellar took four men 19 days, which is roughly twice the labor a cellar of that size should require, and one entry explains why. The entry reads: Extra pay for the deep end. The house was 32 by 24 feet. One end of the cellar had been dug deeper than the other. No one recorded how much deeper. The men who dug it were paid and went home. In 1827, eleven men entered that house. At least four of them went down into that cellar. What they found there is the reason a page is missing from the county record of 1827.
The third unusual thing about the house was the glass. The glass is why we are here.
On June 22, eleven weeks before a single timber of that house was raised, a wagon arrived at the river port dock carrying eight crates addressed to Mr. A. Thorne. The shipment was recorded by the port agent because it was heavy, because it was fragile, and because the freight charges were exorbitant. That record still exists.
Eight crates. Glass packed in straw. 22 panes. Addressed to A. Thorne. Freight paid in full at origin. Origin not stated.
Origin not stated. Every other shipment on that page has an origin. Philadelphia. Baltimore. Pittsburgh. Wheeling. This one has a blank space where the origin should be. The agent wrote beside the blank the words as delivered.
22 panes. A single window pane is a major item in the trade of that era. 22 panes of glass arrived wrapped in straw, in a barren valley, 11 weeks before there was a house to hold them, with the origin field left blank on the shipping manifest.
You have to understand what window glass meant in America in 1809. It was expensive. It was rare. It was of very poor quality. What glass existed came from a handful of eastern factories making crown and cylinder glass by hand. The product was thin, greenish, full of bubbles and ripples, and it came in small sizes because large panes broke on the road. A frontier farmhouse of that period had four to six small windows, and many used oiled paper instead of glass.
The Thorne house had 22 panes of glass. Every window in the house was glazed, including the cellar, which had two windows at ground level in the deep end, and a window on the upper floor that looked out at nothing but the ridgeline. And the panes, by every account that described them, were perfectly clear.
Let us pause on that point, because that single word carries all the strangeness of 1809. Perfectly clear. No green tint. No bubbles. No flaws. No ripples. No distortion at the edges. Flat. Colorless. And according to measurements taken in 1958, uniform in thickness to within 4/1000 of an inch across a pane measuring 18 by 24 inches.
There was no process on Earth in 1809 that could produce that. Not in America. Not in England. Not in France. Not in Bohemia. Nowhere. Flat glass of optical quality requires either grinding and polishing each pane by hand at a cost exceeding the price of the house, or it requires the float process, and the float process was not invented until 1952.
That is not a fringe claim. That is the standard, undisputed history of the glass industry. Flat glass of this quality did not exist until the mid-20th century. The Thorne house had 22 panes of such glass in 1809.
The man who installed them was named Tobias Merrick. He left a record of it. Tobias Merrick was a glazier and tinsmith from the county seat, aged 31 in 1809, an unremarkable craftsman who spent his career installing glass in frames for those who could afford it. He installed the glass in the Thorne house in October 1809 and worked on it for 6 days.
His account book survived because his grandson kept the trade and continued the bookkeeping, with the complete run, from 1802 to 1861, ending up in a regional archive in the 1920s, where it was indexed and forgotten. The Thorne entries occupy two facing pages. The invoice is ordinary and unremarkable. 6 days labor. Putty. Pins. Window cord. Lead. The total is $9.40, a fair price for the work.
At the bottom of the second page, below the ruled line for the total, in the same hand and the same ink, Tobias Merrick wrote four sentences that have nothing to do with accounting.
The glass panes were cut to size before they arrived. Each pane fit its frame to within 1/16 of an inch, and I did not cut a single one. Mr. T would not allow me to touch them with my bare hands and gave me gloves of a cloth I do not know. He stood behind me for 6 days and spoke only to tell me which side was up.
Read that again, because there are four impossibilities stacked in it, and they pass quickly. The glass panes were cut to size before they arrived. The 22 panes, shipped 11 weeks before the house existed, were pre-cut to fit 22 window frames that had not yet been built, by carpenters who had not yet been hired, to a design no one knew existed except in the head of Amos Thorne.
Each pane fit its frame to within 1/16 of an inch. In hand-made frames of softwood. Glass of that era was expected to be cut and fitted on site because nothing was perfectly square. Merrick cut nothing. 22 out of 22.
Gloves of a cloth I do not know. A craftsman who handled every material his trade required, from canvas to leather to wool to linen, was given gloves he could not identify.
And the side that goes up. 22 panes of flat, colorless, apparently identical glass, and each one had a correct orientation known only to Amos Thorne.
Tobias Merrick never wrote another word about the Thorne house in 52 years of bookkeeping. He installed glass in hundreds of homes. He died in 1864 at the age of 86.
There is another document from Merrick, from 1841, 32 years later. It is a letter to a supplier discussing an order for English crown glass and complaining about its quality. In the middle of that complaint, he wrote a sentence the supplier would have found utterly baffling. I have installed glass once in my life that deserved the name, and I will not install it again for any money you possess.
He did not say where. He did not say when. He did not have to.
Now let us talk about the daughters. There were four of them. Their names, as recorded in the 1810 federal census, the congregation membership roll, and two statements, were Silence, Charity, Patience, and Thankful Thorne.
Those are real names from that era. Puritan virtue names traveled west with New England emigration, and by 1810 they were old-fashioned but not remarkable. This is mentioned because people who encounter this case latch onto the names as evidence, and the names are not evidence. The names are ordinary.
Everything else about those four is not ordinary.
They arrived in the valley in 1809 as children. The 1810 census, taken in that county in August of that year, lists the Thorne household as one free white male over 45 and four free white females. Here the 1810 record becomes frustrating because it places ages only in categories, and all four girls fall into the same category. 10 to under 16.
Four girls, all aged between 10 and 15 in 1810. That is unremarkable. It is consistent. It means that by 1820, they should be between 20 and 25.
The 1820 census for that county lists the Thorne household as one free white male over 45 and four free white females aged 10 and under 16. Ten years later. The same age bracket.
Now, the first instinct is correct. Census takers copied data. They were lazy, underpaid, and in a hurry. A man riding through a rural district in 1820 with a previous record in his saddlebag might copy the old record rather than ride nine miles up a hill where preachers were known to be eccentric. That is the obvious explanation, and it should be put on the table because this is not about selling something the evidence does not support.
Here is why that explanation does not hold. The enumerator for that district in 1820 was a man named Rufus Dell, aged 44, a farmer and part-time constable. By the standards of frontier census work, Dell was obsessive about accuracy. His schedules are among the most complete in the state. He recorded people other enumerators missed. He went back a second time to households that were not home on the first visit and noted it.
And in his 1820 record, beside the Thorne entry, there is a mark in the margin. A small cross. Dell used that same cross in 11 other places in that schedule. We know what it means because it is explained in the cover letter he sent with his returns, which still exists. The mark meant they were counted in person by the undersigned.
He went up there. He stood before Amos Thorne. He looked at the four young women. And he classified them as between 10 and under 16. And he marked the entry to confirm he saw them himself.
The Thorne family does not appear in the 1830 census because the house had been empty for three years by then. But there is another census, not federal. In 1824, the denomination Amos Thorne nominally belonged to conducted a survey of its congregations in the border regions, requiring each minister to return a form listing parishioners by household with their ages.
Thorne returned his form. It survives in a collection held by a theological institute. In his own hand, in 1824, 15 years after arriving in the valley, Reverend Amos Thorne wrote the ages of his four daughters: 14, 13, 11, and 9.
In 1810, they were aged 10 to 15. In 1824, their father wrote that they were 14, 13, 11, and 9. In 1827, when eleven men climbed that hill, the estimate in the constable’s report, based on what people in that valley had seen through the window for 18 years, was 22, 21, 19, and 17.
Three different sources. Three different sets of numbers. None of them advancing at the rate of years.
The ages in the official record for that family are not merely contradictory. They are contradictory in a specific direction, and it takes a long time to notice. In 1810, an official placed them in the 10-to-15 bracket. In 1820, an official who verified them in person placed them in the 10-to-15 bracket. In 1824, their father placed them at 14, 13, 11, and 9, younger than the 1810 reading and younger than the 1820 reading. In 1827, the men who had watched them at that window their entire lives placed them at 17 to 22.
Strangers looking in from outside saw them aging. The father, and only the father, recorded them growing younger.
Whatever Rufus Dell was actually looking at in 1820, when he stood in that house and confirmed the entry with his own mark, he saw the same thing he expected to see 10 years earlier. Which brings us to the peddler.
His name was Obadiah Cook. He was 33 in the summer of 1819, a traveling peddler dealing in needles, thread, tinware, ribbons, patent medicines, and news, working a circuit of about 200 miles. He is the first outsider whose direct testimony about the Thorne daughters survives in full, and it survives for a reason unrelated to them, which is exactly why it is trustworthy.
Cook later became a merchant in a river town and served two terms as a justice of the peace. In 1851, at age 65, he gave lengthy testimony in a property dispute over the old Thorne tract. Most of it concerns boundary lines and fence posts. It is dull. On page four, opposing counsel asked whether he had visited the property himself. Obadiah Cook answered the question, and the court clerk recorded everything.
Here is what he said. It is long, and it will not be shortened, because shortening is where stories like this go wrong.
I went up that hill on July 9, 1819, in the afternoon, about four o’clock, intending to sell needles and thread to a household of women, which is profitable trade. I called out at the gate. No one came. I approached the door. The door did not open. I knocked twice. Then I looked to my left toward the front window, which was large and very clean, and there were four of them standing behind it, all in a row, pressed against the glass with their hands upon it. I have thought about this a great deal in the years since and I will say it plainly. They were not looking at me. They were looking past me at the open ground, and they were looking at it the way a man looks at deep water.
I took off my hat. I said good day to them. Not one of them moved. Their hands never left the glass. I counted because I did not believe it, and I got past two hundred, and all that time not one of those four young women moved her hands or her head or blinked her eyes so far as I could tell.
Then the preacher came around the side of the house from the back, and he was not angry. That is the part I want on the record. He was not angry at me. He looked tired. He said, sir, you must go now. And he said it kindly.
I said I had needles and thread and would like to show them to the ladies. He said they will not come out. I said then let them come to the door and I will show them on the step. He said, sir, the step is under an open sky.
The step is under an open sky. That phrase is the center of this entire case, and it needs to be shown how it traveled. It appeared in Obadiah Cook’s testimony in 1851 describing a conversation in 1819. It appeared in a district physician’s letter in 1822. It appeared in a written inquiry from a senior minister in 1825 and in Thorne’s one-sentence reply. And it appeared in the only surviving portion of a sermon Amos Thorne printed.
Four sources. Three decades. Four men who never met. Two of them died before the third wrote. And this is what took a long time to accept. Every one of those four sources treats it as an explanation. None of them treat it as an evasion.
Cook, the shrewd merchant who would become a judge, heard it and understood it as a fact, and got on his horse. The physician heard it and ended his examination. The senior minister read it and did not write again.
Which means that in that valley, in that period, to those men, the sentence made sense. We have lost the frame that made it make sense. That is the actual subject of this story.
Before Obadiah Cook left the hill that evening in 1819, Reverend Amos Thorne paid him for his time. Cook had sold nothing. Thorne paid him anyway, and the testimony records the transaction, which is the strangest three lines in the document.
He would not let me go without compensation for my trouble. He went into the house, came back, and gave me a square of cloth about a foot across, folded, weighing almost nothing, and said it was called a veil. I asked him a veil of what, and he said only a veil. I carried it to town that October, and a merchant in that line gave me $9 for it without asking me a single question. I have not seen its like since, and I have never met a man who could tell me what it was.
$9 in 1819 for a square foot of unknown cloth. That is roughly two weeks of skilled labor. A merchant expert in fabrics took one look at a folded square of something a peddler could not name and paid without a single question. That tells you he knew what it was. And it tells you there was a market for it.
The word veil is the first of four words. The merchant records for that region are not complete, but what survives is enough to reconstruct a partial picture of what Reverend Amos Thorne actually did for money over 18 years.
He never sold a bushel of grain because he never grew any. He never sold cattle, timber, whiskey, potash, or hides, the five things mountain households sold. His name appears on the selling side of the surviving records exactly 19 times over 18 years, and in every case the commodity is recorded under a one-word name. There are four such names in total.
Veil. Milk. Thread. Hour.
The word veil appears seven times, always by the square foot, at $7 per foot. 31 square feet in 18 years. The word milk appears five times, always sold by weight in ounces, never by liquid measure. You do not sell milk by the ounce in 1820, and you certainly do not sell it by weight, at $4 per ounce. One record from 1816 specifies milk, 6 ounces, sealed.
The word thread appears six times, by the yard, at $2 per yard. A ridiculous price in a period when good linen thread cost pennies.
Then comes the word hour. Hour appears once. One time out of 19 recorded sales over 18 years. The entry is dated March 3, 1823, in the ledger of an agent working out of a river town. The ledger itself burned in a warehouse fire in 1868, but not before a clerk copying an inventory of unusual entries transcribed a series of them into a pocket diary that still exists.
The transcribed entry reads in full: A. Thorne, one hour, delivered sealed, $400.
$400 in 1823. Cross-century comparisons are unreliable, but by the simplest measure, wage equivalence, $400 in 1823 is roughly two and a half years of income for a skilled craftsman. For one hour. Delivered sealed.
Sealed in what? Delivered how? An hour of what? Unknown. No one knows. The agent’s records burned in 1868. The clerk who copied the entry died in 1891 without ever being asked about it.
But hold the number $400 and the date March 3, 1823. When we get to the physician’s letters, that date will appear again, and it will matter.
Now the books. It was said earlier that we would return to them, and here we are. On April 4, 1809, when Amos Thorne bought 100 pounds of flour, he declined to buy a Bible. He already had books. We know how many, because in 1812 a neighbor helped him move a chest and mentioned it. And we know something about them because of a man named Reverend Josiah Kemp.
Josiah Kemp was the elder in charge of that district from 1823 to 1831. He was 52 in 1825, a reader of Greek and Hebrew, and by all accounts a rigid, humorless, entirely conventional man. His importance lies precisely in his lack of imagination. He had no interest in mysteries, only in doctrine and paperwork.
In the spring of 1825, Kemp made a visit to the Ashgrove parish. It was routine. He examined the records, inspected the meeting house, interviewed the minister, and filed a report. His report on Ashgrove ran six pages. Five were ordinary. The parish was described as 141 souls. The meeting house was in good repair. The minister was diligent and well regarded.
On page six, under the heading Remarks, Josiah Kemp wrote three paragraphs. They are the most important document in this case, and they sat in a theological school archive from 1884.
I visited Mr. Thorne at his home on the 17th. I was received in the front room and given tea. I was not introduced to his daughters, and I did not press, having been told by three members of the congregation that they are never seen by anyone. Mr. Thorne keeps a large library for a man in his position, some 40 volumes. I examined it while he was out of the room, as any man in my position would. Thirty of them are exactly what you would expect. The rest, I can only say, are not printed books. They are written in a small, fine hand on paper that does not feel like paper, and they are not in Latin or Greek or Hebrew, nor in any script I have seen in 29 years of study. The lines run in the usual direction. The characters are regular and repeat as characters do. It is a language written by someone, and I could not read a single word of it.
I asked him when he returned what they were. He said they were his instructions. I asked instructions from whom. He said from the other party. I thought at first it was a joke, a term from contract law, and I said so, but he did not smile, and I understood it was not a joke, and I did not ask again.
The other party to the contract. That is a term from contract law. It is the standard formulation for the counterparty to any agreement. And a frontier preacher in 1825, asked to explain 10 handwritten volumes in an unknown script, told a supervising elder of his own denomination that they were his instructions from the other party.
Kemp’s third paragraph is four sentences. They are the reason he never returned.
Before I left, I asked him plainly, as I was required to do, why his daughters do not leave the house. He said they must not be looked upon under an open sky. I asked him what would happen if they were. He said it depends entirely on the number of observers.
It depends entirely on the number of observers.
Josiah Kemp filed his report. He did not return to Ashgrove. He did not recommend disciplinary action. He did not recommend an investigation. He never mentioned that visit again, so far as the surviving correspondence shows, for the remaining six years of his tenure. When the Ashgrove congregation wrote to the district in 1828, a year after the fact, requesting guidance on what to do about a meeting house whose minister had vanished, the reply came from a clerk, not from Kemp.
He lived until 1844. He died at 71. His widow donated his papers, and among them was a single sheet, undated, in his hand, that mentions nothing about Ashgrove and never uses the name. It reads in full: The question is not what he was hiding from them. The question is what he was hiding from us, and I am persuaded he was doing it for us, and that he was the only man doing it.
Ezra Bell Holloway was a traveling physician, 46 in 1822, trained in Philadelphia, covering an area of about 600 square miles. He was a clinical man with clinical skills. His letters to his brother, a physician back east, run over 300 surviving pages and are almost entirely about medicine, weather, and money.
On February 11, 1822, he was summoned to the Thorne house. The letter he wrote that night is four pages long. Thorne had called him because one of his daughters was ill. Holloway does not say which one. He refers to her only as the second. Elsewhere in the letter, he refers to them by number rather than name every time, which is worth noting because Holloway names everyone. His letters are full of names.
He was admitted to the house. They took him upstairs. Here the physician’s letter stops reading like a physician’s letter.
I was permitted to examine her in a room on the north side with the shutters closed and two lamps burning. Mr. T remained present throughout, which I allowed, the patient being unmarried and no other woman present but her sisters, who stood along the wall and did not sit. Pulse 62 and regular. Skin cool and dry. No fever. The complaint was a cough of about 3 weeks’ duration, and I heard it, a dry cough of no consequence, and I said so.
I will now record what I observed and ask you to read it as written without interpretation. Her color. I have treated pale-faced patients my entire career. Anemia, consumption, hemorrhage, the pallor of long confinement. I know them all and can tell them apart across a room. This is none of them. There is no yellow in it, no gray, no blue at the lips. The skin is uniform and lustrous. Held near the lamp, it does not take the lamp’s color as skin does. I placed my hand beside hers on the coverlet for comparison, and I tell you I withdrew my hand again.
Her eyes. Her pupils did not respond to the lamp. I brought the lamp near and took it away four times and they did not contract or dilate to any degree I could see, and yet she followed my finger with perfect accuracy and read a line of small print from 8 feet on the first attempt. I have no explanation and I offer none.
All four are alike in this. All four are so alike in size and build that from 10 feet and in that light I could not have said with confidence which one I had examined. They answered every question I put to them and answered politely, and in 11 years of practice I have never had four patients answer so competently and tell me so little. They do not volunteer. Ask and you receive an exact answer and no more.
Then the final page.
On leaving I asked Mr. T the question everyone asks, and I asked it as a physician, for it is my right. I said, sir, these young women need air and light and exercise, and they have had none of it for 13 years, and whatever your reasons, their health will pay the price. He said, doctor, look at them. Do they look sick to you? I said no. And that is the truth, and it is the worst thing in this letter. They do not look sick. They look better than any two young women in this county who go out under the sun every day of their lives, and they have not left that house since 1809, and I cannot hold those two facts together.
Then he said they must not be looked upon under an open sky. And he said it the way a man reads a label. I said, and if they are? He said a single man might do it and no harm would come. It has happened, and we were fortunate. But there is a certain number, doctor, and I do not know what that number is, and I have spent my life making sure we do not find out.
Brother, I have written this plainly and I will ask you now to burn it, and I know you will not, and I find I am glad of it.
He did not burn it. That is why we have it.
Holloway visited the Thorne house twice more, in 1823 and 1825. The 1823 visit is the one to remember, and here is why. It is dated March 5, 1823. Two days after a factory in a river town recorded the purchase of one hour, delivered sealed, from Thorne for $400.
Holloway’s letter of March 5, 1823, is one paragraph. It is the only letter in 300 pages in which his handwriting deteriorates.
I was summoned to T, not as a patient, and he asked for a witness and would not say to what. There is a clock in the front room that was not there in February. It is a tall wooden clock from a maker I do not know, and it does not strike. He asked me to watch the hands 10 minutes before eleven, and again 10 minutes before twelve, and to witness in writing that one hour had passed between them. I did so. I signed a paper stating that one hour had passed in a room, the most absurd document of my life, and he paid me $11 for it, and I took it, and on the way down the hill I stopped and vomited in the grass, and I do not know why.
He signed a paper witnessing the passage of one hour. Two days after that hour was sold, sealed, for $400.
Every other commodity in those records, the veil, the milk, the thread, was sold by physical measure. Square foot. Ounce. Yard. Someone, somewhere, was buying an hour by the hour, and it required a physician’s signature to certify that the hour was real.
Dr. Ezra Bell Holloway died in 1839 at 63. He never wrote about Ashgrove after 1825. His brother kept the letters, and the family kept them after that, and they were delivered to a medical library in 1951. That is why the 1958 investigation happened at all.
Now the painting. In the spring of 1824, a traveling portrait painter was in that region, a man named Peregrine Coulter, aged 38. The itinerant portrait painter was a mobile photographer, the equivalent of a traveling photographer of his era, moving from house to house painting likenesses for $8 to $20, depending on size and whether you wanted hands included. Hands cost extra.
Coulter climbed the hill in April 1824 and asked for work. To everyone’s surprise, Thorne agreed, sort of. He would not let Coulter into the house. He would not let the daughters out of it. What he proposed, and Coulter accepted, was that the painter set up his easel in the yard at a distance Coulter recorded as 11 paces, and paint the four of them standing at the front window.
Coulter worked on that painting over four days. His working notes survive in a museum acquisition file, because the painting also survives. It is the only physical piece of evidence in this case that you can, in principle, go and see. It is a small oil on canvas, 22 by 28 inches, in a regional art museum since 1962, catalogued as an anonymous American portrait, circa 1825. It is not on display. It has not been on display since 1971.
Coulter’s notes describe an ordinary commission that went wrong in a very specific way. Day one, he laid in the house, the window, the four figures. Ordinary. Day two, he writes: The light on them does not come from the sky. I have painted through glass before and I know its effect. This is not that. The illumination is even, it is from inside the room, and there is no lamp in the room. I paint it as I see it, and it looks like a lie.
Day three, he writes: I have four faces and I cannot make them different. I have moved the eyes. I have adjusted the mouths. I have shortened one nose by a full quarter inch from life. When I step back 11 paces, it is the same face four times. I am beginning to think the fault is not in my hand.
Day four is a single line. Finished the work. Took my $8. I will not come back here again.
He charged $8, which means without hands. But there are hands in the painting. That is the thing about it. Peregrine Coulter charged for without hands. His own notes record the fee as without hands. And the painting in that museum storage room shows four young women behind a window with their palms flat against the glass, and those hands are the most accurate part of the entire painting.
In 1971, the museum sent the painting for cleaning and technical examination, a standard procedure, and the restorer’s report is in the object file. It is the reason the painting has not hung on a wall in over 50 years.
X-ray imaging of that painting shows the composition beneath the surface. Under the four figures, in the same layer of paint as the floor, there are not four figures. There are five. Five figures. Four painted over one.
The restorer’s language is dry, and it should be heard how dry it is. X-ray reveals a fifth figure standing, centered, taller than the four visible figures, occupying the space between the second and third figures. This figure is not an underdrawing. It is fully modeled in lead white and appears to have been executed at the same time as the visible figures, not beneath them as an abandoned first attempt. There is no separating layer of varnish or dirt. The artist appears to have painted the four figures over the fifth in a single working session.
In a single working session. He painted five, then painted four over the five while the paint was still wet, in the same afternoon, in the yard, 11 paces from that window. And he never wrote it down. Four days of working notes complaining about light and faces, and not one word about a fifth figure.
Coulter died in 1858 at 72. He painted for 34 more years after Ashgrove, 206 documented works. That is the only painting in which a figure was concealed. It is the only one he signed on the back of the canvas rather than the front. And the inscription on the back, revealed during the 1971 examination, is not a signature in any traditional sense. It reads: P. Coulter, painted from the yard, April 1824, in the presence of an uninvited party.
In the presence of an uninvited party.
Now we come to 1826 and a girl named Elspeth Finch. Elspeth Finch was 11 years old. She was the daughter of a widow named Marjorie Finch, aged 34, who had come to the valley from downriver in 1824 and was never fully accepted by the valley people, as strangers in closed valleys are. They lived in a rented two-room house at the foot of the hill road, three-quarters of a mile below the Thorne place.
On the afternoon of October 28, 1826, Elspeth Finch was sent up the hill road with a covered basket carrying a dish of preserved plums to the preacher’s house, an ordinary errand she had done at least four times before. She reached the house. We know she reached the house because the basket was found on the front doorstep, upright, the dish inside untouched.
She did not come home.
To be fair, the disappearance of a child in a mountain region in 1826 is not by itself evidence of anything but the nature of mountain regions. The valley searched for 11 days. Forty men and boys walked that hill and the two hollows below it. They found nothing.
What matters is not the disappearance. What matters is what Marjorie Finch did and what was said to her. She went up to the house on the evening of the 28th within four hours and knocked. Amos Thorne answered. According to her sworn testimony the following spring, this is what passed between them.
I said, where is my daughter? He said she left the basket and went again and I did not see her leave. I said, let me into the house. He said I cannot let you into the house. I said then bring the girls to the door and let them tell me. He stood there a long time and then he said what he always says, that they must not be looked upon under an open sky, and I said to him, sir, it is dark. There is no sky tonight. He said to me, madam, that is not how it works.
Then he added, and I have not repeated this to anyone before this day. He said, go home. Go home tonight and keep going home. In a year I will be gone from this place and so will they, and you must not be standing here when it happens. I asked him what would happen. He said the reckoning will come.
I have thought about those four words every night since. The reckoning will come.
She went to the elders. They told her to go home and wait, which is what elders told strange widows in 1826. She did not go home to wait. On November 2, 1826, Marjorie Finch walked 22 miles to the county seat, arrived at the sheriff’s office shortly after five in the morning, and filed a complaint against a man of the cloth. In that county, in that year, that was an almost unbelievable act.
The sheriff was a man named Absalom Tegg, 57, 21 years in office, four children, known for doing nothing quickly. Tegg did not go up the hill. He sent a deputy to take a statement, opened a file, and the file sat.
That is the hardest thing in this case. Absalom Tegg had heard about the Thorne house for two decades. In his own words, from testimony given in 1831: I have heard of that house since I first took office. I heard of it from merchants, from the physician, from a clergyman, from people I do not trust. In 21 years, I never went up there once. I cannot tell you why I did not. I have asked myself, and I find no answer in myself. I cannot tell you why I did not.
For a year, nothing happened. Elspeth Finch was not found. Marjorie Finch stayed in the valley, which surprised everyone.
On November 8, 1827, 376 days after her daughter went up that road with a basket, two men from the Ashgrove congregation, coming down from the high pasture at dusk, saw something on the hill that ended 18 years of leaving that house alone. Their accounts agree in every detail. The shorter one is from Caleb Ivers, the same Caleb who as a boy watched Thorne mowing that circle at two in the morning.
We were coming from the north fence about half past five, and the light was nearly gone. The Thorne house was below us and to the left, and we could see its front and the open ground before it. The four of them were outside the house. They were standing in the open ground in the middle of the circle in a row, all facing the house. They were not moving. Mr. Thorne was standing on the threshold of the house facing them, and he was not moving either.
There was no sky. I want that recorded because I was asked about it. The fog was low on the hill that evening and all day. You could not have seen a star from that ground with a telescope. We stood watching them, I think for 4 minutes. Then Josiah said we must go down and tell someone, and we went down.
The four of them were outside in the circle under a closed sky, and their father stood on the threshold watching them.
The congregation met that night. Eleven men agreed to go up the hill in the morning. They went on November 9, 1827, at a time the police report gives as exactly 9:29 in the morning. They knocked. No answer. They broke the door.
At the start, it was said what they found. Bread still warm in the oven. Water in the kettle. A fire stoked within the last hour. Two horses in the stable. The dog on the threshold. A pen resting on an open book with the ink still wet on the last word. The book was one of the ten unreadable ones. The ink was wet.
They searched the house room by room. Eleven men in a six-room house. It took them less than a quarter of an hour to confirm the house was empty. Clothes were in the wardrobes. Shoes were under the beds. Four sets of shoes. All four pairs.
Then they found the front window. Eight handprints inside the glass, not on it.
Constable White Sample arrived at noon and spent, by his own account, about 50 minutes trying to remove them by every means available to him. He left no mark on the glass. He did not touch the marks because there was nothing to touch.
Then someone opened the cellar door. The Thorne cellar, as noted, was dug in 1809 by four men in 19 days, with an extra charge in the account for the deep end. The shallow end was an ordinary cellar. Jars and barrels. A potato bin. Two hams. Ordinary.
The deep end extended another 11 feet by rough estimate, down stone steps that no 1809 account book mentions and no local mason built. Four men went down. Sample went down.
What the official record says about what is at the bottom of those steps is one sentence: The lower cellar was found to contain a quantity of filled earth and certain fittings, regarding which see page 11.
Regarding which see page 11.
Constable White Sample’s report on the Thorne house, filed with the county on November 20, 1827, was entered into the county returns volume for that year in the usual way, sewn, not loose. The report runs from page 7 to page 14. The volume is intact. It has never been rebound. The sewing is original. The pages of that volume run: 7, 8, 9, 10, 12, 13, 14.
There is no page 11. No paper remnant. No cut edge. No evidence of excision. No gap in the sewing. The leaf of page 10 and the leaf of page 12 are two sides of the same continuous sheet, never separated. The county clerk who numbered the volume in 1827 simply skipped a number.
This was not a theft. No one removed page 11 from that book, because page 11 was never in it. Someone wrote it, and Sample’s report refers the reader to it four separate times, while it was never in the volume at all.
We know what was on it. Not all of it. Enough.
The deputy sheriff who went down those steps behind White Sample was a man named Ransom Pike, 24 in 1827. He served 11 more years, farmed for 30 years after that, raised six children, and never discussed the Thorne house with any of them, not once. In 1891, at 88 and dying, he asked his granddaughter to bring a young man from the county newspaper. He talked for under two hours. The young man recorded everything he said and published none of it because Pike made him promise. The notebook ended up among the reporter’s papers. The reporter died in 1946. The papers went to a university, and the notebook was indexed in 1968 under the title Farm Reminiscences.
Pike said: The lower cellar was a room built of dressed stone, and it was older than the house itself. I know that because the house was built over it badly. The room does not match the walls above it by more than a foot. Its floor was paved in a pattern of small stones set on edge, in a spiral running from the wall inward, and it was worn. Worn and smooth the way a doorstep wears. Someone had walked that spiral path so many times over so many years that the stones were hollowed in the middle of the path.
There were four places along the outer wall where the stone had been cut to form shallow recesses the size of a standing person, and they were smooth inside and bore no tool marks. Four recesses. Four girls. I counted them then, and I have counted them every day since, and I have told no living soul because the number is of no use.
In the center of the spiral, there was a fifth recess, in the floor, and it was empty, and it was larger.
Mr. Sample wrote all of that. I watched him write it. He wrote it that night at the table in the front room of that house, and he put a page number on it, and he put it in his coat, and he never put it in the book. I asked him why in 1829, and he said, Ransom, there are things that are true and are also a kind of instruction, and the man who publishes them is not reporting news, he is passing it on. He said the whole thing was a set of instructions, and he would not be the one to pass them on.
I have thought about that for 62 years, and I have come to believe he was right, and I will say now, at the end, that I am not sure.
That is the missing page. It was not lost. It was not stolen. It was not suppressed by any authority. One man wrote it, realized what he had written, refused to place it in the public record, and then lived 44 more years in a house with no windows facing west.
The Thorne house was left standing. It was not sold because the county could not determine an heir. It was not burned because no one wanted to go up there to burn it. The roof collapsed in the winter of 1841. The chimney stood until a storm in 1853. By the 1870s, nothing on that hill stood above a man’s knee except the edge of a stone cellar covered in dirt, and a circle that farmers kept mowed under an arrangement no one knew who had made or when.
The circle was still being mowed in 1938. That is documented in a state agricultural survey. 164 feet in diameter, on a hill no one farmed, kept clear by four generations of men who could not tell an interviewer why they took the glass.
Of course the glass was taken. 22 panes of the finest window glass anyone in that county had ever seen sat in an empty house, and by the time the roof fell in 1841, all of it was gone. It went to other houses. Two can be traced by deed and probate to a farm 8 miles east. One is in a church vestry window frame. The rest went where such things go.
In 1958, a materials chemist at a state university, working from Dr. Holloway’s letters that had reached the medical library seven years earlier, went looking for the Thorne panes. He found four. He obtained two. His report is 11 pages of ordinary analysis and one page that is not.
The glass is soda-lime silicate, which is ordinary. Its iron content is below 200 parts per million, which in 1809 is not ordinary but impossible, because iron impurities are what make that era’s glass green, and removing it below a tenth of a percent was not possible until late in the 20th century. The surfaces are flat and parallel to within 4/1000 of an inch. He measured it three times.
On page 11 of his report, which was promised to be a coincidence and was verified twice, he recorded the thing he could not explain. Both panes, distributed through the thickness of the glass rather than on its surfaces, contain a dispersion of very small impurities invisible to the naked eye, arranged in a regular three-dimensional grid with a spacing he measured and found consistent in both samples.
He wrote: This distribution is consistent with no impurity mechanism known to me. Impurities in glass are accidental products and are randomly distributed. These are not randomly distributed. I am required to record that they are arranged, and I decline to speculate as to why.
The report was submitted to the university and never published. He was 51 when he wrote it. He died in 1982 at 75. In 1979, in a recorded interview in which Ashgrove was never mentioned, he was asked if there was anything in 40 years of work he had been unable to explain. He said yes, and I returned it.
He returned it. He returned both panes to the two families who had lent them to him, and they are, as far as anyone knows, still in those houses.
In 2019, a university team conducted an aerial survey over those hills for reasons unrelated to this story. They were mapping historic charcoal kiln sites. The survey shows that hill with the circle exactly where the 1807 map placed it, 164 feet across, and the rectangular scar of the house on its northern edge. It also shows, beneath the circle, a subsurface anomaly.
That anomaly is not a cellar. It is not a mine. It is not a natural cavity. The report describes it in the most cautious language a university report can muster. It is regular. It is spiral in form. It descends. Its outer loop is 164 feet in diameter.
Which means the bare circle on the surface, the one measured by a surveyor in 1807, the one farmers mowed for 130 years, is not just an empty clearing. It is the top of that thing.
The survey report notes, in a single line, that the structure shows evidence of ongoing modification. It does not say by whom. It does not say when. The data is dated 2019.
Here the scope must widen, because Ashgrove is not unique. That is the actual point of this entire story.
In 1861, on a farm in Ohio, three sisters who had not gone outside in 9 years vanished with the doors locked from within, and handprints were found on a window that the county coroner described in writing as beneath the surface of the glass.
In 1904, in a house on the Georgia border, two girls whose grandmother had kept them inside since childhood vanished on a Tuesday morning with breakfast on the table, and the local paper noted that the family had for years refused to allow the girls onto the porch, quote, because of the open air.
In 1959, in a valley in the western highlands, a man was committed to a state hospital for refusing to allow his four daughters to leave the house. The daughters were removed by court order on June 11. Three were placed with relatives. The record for the fourth is one line, and the line says she was not in the dwelling. There is no follow-up. There is no missing persons file. The man died in 1966, still repeating the same eight words Amos Thorne used, in a state hospital 1,200 miles from Ashgrove, 132 years later.
They must not be looked upon under an open sky.
17 locations. That is the number as it stands. All share three things: a house built to hold specific glass rather than glass fitted to the house, a family with no documented origin, and a small number of people, always women, always young, kept from the open air by someone who loved them enough to be hated for it.
Here is where most writers on this subject land, and it should be acknowledged. The usual conclusion is that these families were imprisoning their daughters, and that folklore arose later as an excuse for it. That is the humane reading. It is the reading that requires nothing of you. But it does not fit the documents.
It does not fit a physician who examined them and found four young women in better health after 13 years indoors than anyone walking under the sun. It does not fit an elder who came to investigate and left persuaded that Amos Thorne was the only man in the valley doing something for others and not for himself. It does not fit a constable who wrote the whole truth and then decided that writing it down was equivalent to passing it on.
Read those documents together and a different picture forms. No one was being kept inside. Something was being kept out. Or more precisely, something was being prevented from completing an agreement that had run too long, paid in installments of veil and milk and thread and one sealed hour, by a man who called the other party the other party, and who spent 18 years and $430 a year of unaccounted gold making sure the final payment would never be demanded.
He told Marjorie Finch the reckoning will come. He was right. It came almost exactly a year later. He walked out to that circle with his four daughters under a clouded sky and settled it. The mercy in this entire story, if there is any, is that he settled it alone at dusk, with clouds over the hill, where only two men on the fence line could see them. No more than that.
It depends entirely on the number of observers.
Two images remain. The first is a spiral floor of small stones set on edge, eleven feet beneath a farmhouse, worn hollow in the middle of the path by someone who walked it for eighteen years. The second is four young women at a window, hands flat on the glass, looking past a peddler at open ground, the way a man looks at deep water.
If you have read this far, something has stuck. It will stay with you for a few days, and then it will surface when you least expect it. That is how these stories work. It is why the men who witnessed this did not speak for 60 years, and only spoke then from their deathbeds.
The next time you put your hands on a window, look at what is on the other side of it. Then ask yourself which side of that glass you are really standing on.