The Village That Is Not in the Gazetteer
The Bureau’s American desk has held the Rollandville file open for eleven months, and the difficulty with it is not that the story is extraordinary. Extraordinary stories are the desk’s ordinary business. The difficulty is that the story is circumstantial, consistent, populated with named neighbours and specific acreages, and attached to a village that does not appear in any gazetteer, any county history, any turnpike survey, any church register, or the returns of the first federal census.
That census is worth pausing on, because it is the tightest net ever thrown over the American countryside up to that date and it was thrown in exactly the right year. Census day was 2 August 1790 — some six weeks after the events described here. The marshals of the United States judicial districts went house to house and returned 3,929,214 inhabitants for the whole republic at a cost of $44,377.28. Their schedules for Delaware, Georgia, New Jersey and Virginia were lost outright in the decades afterwards, and near enough a third of the original returns are gone. New York’s survive, gaps and all. A marshal walked that country in the summer of 1790, wrote down the head of every household he found, and did not find Rollandville.
The rest of the year is well lit. It was the second summer of Washington’s first term; the seat of government was still at New York and about to move to Philadelphia; the post roads that carried news northward from the city were bad but they existed, and the printers along them copied one another’s columns wholesale, so that a curiosity from a back county could be in four papers inside a fortnight. Nothing about 1790 makes a burning sphere hard to report. It only makes it hard to report from nowhere.
A Summer’s Evening, and a Light That Kept Its Own Time
The account, in every version the Bureau has collated, opens the same way. On an evening in the summer of 1790, after milking and before full dark, the people of Rollandville saw a light in the northern sky that grew until it was brighter than the sunset it was competing with. It was not a shooting star. Witnesses are unanimous about that, and unanimous for the same stated reason: it was too slow. A star falls and is finished before a man can say that it fell. This one, they said, could be watched. It could be pointed at. It gave its audience time to be afraid of it, which is the one courtesy the sky does not usually extend.
It came down in the lower field of a rye farmer named Elijah Comstock, and it came down hard. The reports give a flash, a concussion that was heard — the phrase is always the same — “miles away”, and a fire in the standing hay that a dozen men beat out with sacking before it reached the wall. By morning there was an acre of burnt ground, a scatter of debris across it, and a village standing at the fence.
The Bureau notes the “miles away” with professional weariness. It is the most quoted measurement in the file and it is not a measurement. No version names the villages that heard it, no version names a person in them, and no version gives a direction. A detonation loud enough to carry across several miles of Hudson Valley farmland in 1790 would have been heard by hundreds of people who had nothing to do with Rollandville and no reason to keep quiet about it. The account has an explosion with an audience of exactly one village.
The Metal That Was Neither Wood nor Iron
What was on the ground is the heart of the account and the part that has changed least. Comstock’s diary — of which more shortly, since nobody living has seen it — is quoted as saying the material was “neither wood nor iron but something altogether different: smooth, shiny, and cool to the touch despite the heat of the crash.” His neighbours the Flynns added that some fragments carried marks: lines, and a figure like a wheel, and something between a letter and a brand.
The cool metal is the detail that has done the most work over two centuries. It is offered, in every retelling from about 1840 onward, as the proof that the thing could not be natural — a stone that has just burned a field ought to be too hot to hold. The Bureau is obliged to point out that this is precisely backwards. Accounts of freshly fallen meteorites vary: some are reported burning hot to the touch, and others have been described as cold enough to condense water and frost over, the glowing part being a thin fusion crust under which, on a stony meteorite, the heat reaches at most a few millimetres in. Of every element in the Rollandville story, the impossible cool metal is the one detail with a precedent in the record of real falls.
The marks are the opposite case. Not one person in the story made a rubbing, a tracing, or a sketch, in a decade when tracing a curious inscription was the standard reflex of every educated man who owned paper. The symbols exist only as words, and the words have not held still: lines and a wheel in the earliest versions, letters “of no known alphabet” by the 1840s, and by the middle of the twentieth century a face. Physical objects do not acquire new features while sitting in a barn. Descriptions do.
What the Learned Would Have Said in 1790
Here the file becomes genuinely interesting, and it is worth being exact about the year. In 1790, the position of learned opinion in Europe and America was not that rocks from space were rare. It was that there was no such thing. Reports of stones falling out of the sky were classed with reports of frogs falling out of the sky: peasant testimony, honestly given and wholly mistaken, to be explained away as lightning striking an outcrop, or matter thrown out of some volcano over the horizon, or a stone that had been lying in the field all along and that nobody had noticed before.
That position had four more years to run unopposed and thirteen before it gave way. In 1794 Ernst Chladni published at Riga and Leipzig the book that began to break it, arguing from collected reports of falls that fireballs and fallen masses are one event and that the masses come from beyond the Earth. He was laughed at, Lichtenberg among the laughers. Six weeks later a shower of stones came down at Siena, the first fall in modern times to land beside a European city of any size, and it was seen by too many people to be argued away. In December 1795 a fifty-six-pound stone came down at Wold Cottage in Yorkshire within a few yards of a ploughman, dug a crater a yard across, and was exhibited in Piccadilly by a landowner who later put a twenty-five-foot monument over the hole. In April 1803 several thousand stones fell on L’Aigle in Normandy, Jean-Baptiste Biot went to look, and the argument was over.
So the Bureau records the following, and asks the reader to hold it steadily. If the people of Rollandville said in 1790 that a burning mass had fallen out of the sky into a rye field, every qualified authority in the western world would have told them they were wrong. The villagers’ claim was rejected — and it was rejected for entirely the wrong reason. It was not rejected because the marks were unverified or the diary unproduced or the militia unaccounted for. It was rejected because stones do not fall from the sky, which was the settled science of the day and which was false.
There was also a second explanation available, and it was extremely fresh. Seven years earlier the Montgolfier brothers had flown a balloon two kilometres over Annonay, sent a sheep, a duck and a rooster up at Versailles in front of the king, and put two men into the air over Paris for twenty-five minutes. In the same season Jacques Charles let a hydrogen balloon go from the Champ de Mars; it flew for forty-five minutes, came down twenty-one kilometres away at the village of Gonesse, and the villagers — confronted with an unfamiliar object descending into their fields — destroyed it with pitchforks. Within three years of Rollandville, Jean-Pierre Blanchard rose from the yard of the Walnut Street Jail in Philadelphia and landed in New Jersey while President Washington and four future presidents watched him go. The Rollandville account contains no balloon. It is the one explanation with a living, reported precedent on both sides of the Atlantic, and it is the only one nobody in the story ever proposes.
The Map, and the Cross Upon It
There is exactly one document in the Rollandville file, and the Bureau has spent more time on it than on all the testimony put together. It is a map: a sheet of laid paper showing a river, a road, a broken ring of woodland, a compass rose of some ambition, the word ROLLANDVILLE in a display hand, and a small cross where the ground was burnt.
What is missing from it is more eloquent than what is on it. There is no scale. There is no surveyor’s name, which in 1790 was as much a matter of professional pride as a signature on a painting. There is no date in the cartouche and no reference to a warrant, a patent, a lot number or a range — the apparatus by which every real parcel of New York land was described to the men who taxed it. The roads run to no named place. The compass rose is far too large and far too finished for a working survey and would be perfectly at home on a title page.
Set against that absence is the ordinary machinery of American news. A fireball, a burnt acre and a militia turnout was exactly the class of item the country printers lived on; they had columns to fill and they filled them by scissoring one another. The Bureau has looked for a printed notice of Rollandville in 1790 and in 1791 and has found none, which is not conclusive on its own — much of the period’s ephemera is gone — but which sits badly beside a village of ninety roofs, twelve militiamen, two wagons and an explosion heard for miles.
Theories, in the Order They Were Offered
The file preserves three readings, and the Bureau attaches weight chiefly to the order in which they turn up. The first is a meteor, offered by the villagers themselves and thrown out by their betters. The second is a manufactured object from some foreign country, which was the respectable position for about fifty years and which quietly assumes that a foreign country could throw a thing across the Atlantic and into a rye field. The third is visitors.
The third is much the youngest, and the Bureau would like to be precise about why. There is a long tradition of things in the sky being read as significant — the wheel in the Book of Ezekiel, the spheres and crosses of Hans Glaser’s Nuremberg broadsheet of 1561 that fought each other until they fell smoking to earth. But those are portents. They are messages from a God who is presumed to be sending them. Nobody in 1561 concluded that the spheres were a conveyance with somebody sitting inside it, because the idea of a crewed vehicle crossing between worlds had no shape available to think it in.
That shape arrives late, and the Rollandville debris arrives at it on schedule. The marks are lines in the earliest telling. They are unknown letters by the 1840s, when unknown letters were the fashionable mystery. They are a wheel by the 1880s. They are a face by 1954. Every stage of the description is contemporary with the thing its own century found most interesting, which is a pattern the Bureau observes in objects that do not exist and never in objects that do.
The Rediscovery, and the Shape It Took
Rollandville was not a famous story for most of its life. It was a local one, thinly attested, mentioned in a county compilation or two, and it acquired national interest abruptly and in company. The precedent everyone reaches for is 1947, but the better parallel is fifty years earlier: the American mystery airship wave that began over Sacramento on 17 November 1896 and rolled east through the spring of 1897, generating a witness count the newspapers put north of a hundred thousand while the San Francisco Call and the San Francisco Examiner each accused the other of inventing the whole thing. That wave produced its own crash. On 17 April 1897 the Dallas Morning News printed S. E. Haydon’s story that an airship had struck Judge Proctor’s windmill at Aurora, Texas two mornings before, killing a pilot not of this world who was buried in the town cemetery under Christian rites. An Aurora woman of eighty-six, interviewed a lifetime later, said Haydon had written it to bring trade to a town the railroad had bypassed.
Then the vocabulary changed. On 24 June 1947 Kenneth Arnold saw nine objects near Mount Rainier and described their motion as like saucers skipping across water; the press turned a motion into a shape and by 26 June the phrase was in a Chicago headline. A fortnight later the press officer at Roswell Army Air Field announced that the service had recovered a disc, and was corrected within the day by a general holding up a weather balloon. The Roswell story then lay dead for thirty-one years until a 1978 interview with a retired major revived it, and in 1994 the Air Force conceded that the original balloon story had been a cover — not for a craft, but for Project Mogul, a classified array of microphones flown under balloon trains to listen for Soviet atomic tests.
The Bureau sets these dates beside the Rollandville accessions without further argument. The plates in this file that show a disc are dated from 1948 onward. The face on the metal is recorded in 1954. The historical society that holds the map was founded in 1951. Rollandville did not survive from 1790 to the present carrying its cargo intact; it was reloaded, in a single decade, by a century that had at last been given the words.
Plates: the Rollandville File
Nine plates received with the account, none of which the Bureau has been able to date from internal evidence. They are printed as received and endorsed by nobody.
The Eyewitness, and How the Bureau Reached Him
Elijah Comstock is the load-bearing member of the whole structure. Every physical detail in the Rollandville account descends from him: the slowness of the light, the burnt acre, the cool metal, the marks. Everyone else in the story is quoted describing what Comstock showed them. The Bureau therefore put its questions to him directly, under the standing arrangement it maintains with witnesses who cannot otherwise be produced, and found him a great deal more careful than his reputation.
He is careful in a specific and rather touching way. He will not be moved off the two things he says he saw, and he will not be led an inch past them. Where the literature has him identifying a craft, he identifies nothing. Where the literature has him filling a sack with fragments, he says he never picked up a piece. The extravagance in the Rollandville story is not in the witness. It is in everyone who has handled the witness since.
The diary is the recurring embarrassment. It is quoted in four separate publications between 1841 and 1968, always at a slightly different length and never with a page reference. Comstock himself says it went north with his daughter’s family and he has not seen it since Adams was president. The Bureau has traced the family into Vermont, into Ohio and out of the record, and has found no chest, no barn and no diary. Every sentence attributed to it in two centuries of literature is quoted from a quotation.
Elijah Comstock
Farmer · Rollandville · First upon the ground
Deposed under the Bureau's standing arrangement with those who cannot be produced. Mr. Comstock's diary has been described by four writers and seen by none of them.
The Bureau
State your name, your holding, and what you had under crop in the summer of 1790.
Comstock
Elijah Comstock. The home lot, and eleven acre beyond the wall in rye, and the low field in hay that we had not yet took up. Two cows. A horse that was mostly my brother's. The dogs.
The Bureau
At what hour did you first see the light? You have given it three ways in three tellings — at dusk, at full dark, and near midnight.
Comstock
It was after the milking and before I banked the fire. There is no clock in my house. There was no clock in the village but the meeting-house clock and that ran slow. It was the same evening every time I told it, sir. A man may lose the hour and keep the evening.
The Bureau
You say the object moved slowly. Slower than what?
Comstock
Slower than a star that falls. A star is done before you have said that it fell. This one I watched. I had the time to be afraid of it, and that is not a thing a star gives you.
The Bureau
Was the metal warm?
Comstock
It was cool. That is the part folk will not have, and I have never once said different, and I cannot tell you why it was cool. It come out of a fire and it lay cool in my hand.
The Bureau
Describe the marks. Did you copy them?
Comstock
Lines. Something like a wheel. Something like a letter that is no letter I know. I meant to copy them. I had it in my mind the whole of that summer to sit down and copy them and I never did it, and that is my fault and nobody else's.
The Bureau
Who was upon the ground before the militia came?
Comstock
Flynn, and his wife, and the daughter. Ezra Peck, who was nearest. The Bakers came at first light, and by then there was such a deal of people I could not say who.
The Bureau
The diary. Where is it?
Comstock
It went up into Vermont with my daughter's people. I am told it is in a chest. I am told the chest is in a barn. I have seen neither of them since Mr. Adams was president.
The Bureau
Did any printer put the matter in his paper that year, or the next?
Comstock
We are not a paper sort of place. The post came when it came. A man rode through in the spring after and I told him the whole of it standing at that fence, and he wrote nothing down, and I did not think to ask him why.
The Bureau
What do you say it was?
Comstock
I say it come down. I never said what come down. Other people have said what it was, and they have said a great deal of it, and not one of them was standing in my field.
Plates: the Neighbours
The Flynns and the households along the Rollandville road. Every one of them is named in the account. Not one of them is named in a record.
The Officer, and the Removal
The official response is where the Rollandville story stops being a country curiosity and starts being a conspiracy, and it is also where it is weakest. The account has a militia detachment ride out from Albany, secure the field, collect the debris and cart it away to an undisclosed location, on the authority of nobody in particular. This is a twentieth-century plot with an eighteenth-century cast list.
The state of New York in 1790 had no body with either the machinery or the appetite to sequester physical evidence from a private field. Militia companies were local, part-time and chronically unpaid; they turned out for musters, fires, riots and floods, and they filed for their expenses afterwards, because the county was tight-fisted and the paperwork was the whole point. A four-day turnout by twelve men with two wagons would have left a claim for forage, rations and wagon hire in a county account book. Such books survive in quantity. This claim does not appear in any of them.
Then there is the sentence. Captain Winslow’s report is universally quoted as observing that “the nature of the wreckage defies all current understanding of our world and its workings” — a line that has been doing the heavy lifting for two hundred years and which no militia officer of 1790 would have written, because it is not a militia officer’s sentence. It is a magazine’s sentence. Officers’ returns of the period are dry to the point of rudeness: ground burnt over, so many acres; no person hurt; men discharged.
Captain Josiah Winslow
Captain of Militia · Albany · Officer of the removal
Deposed under the same arrangement. The Captain's report has been quoted for two hundred and thirty-six years and has never once been produced; nor has the order that sent him.
The Bureau
State your commission and the company you had at Rollandville.
Winslow
Captain, in the Albany militia. I had what men could be got upon the day, which was not the whole of a company and which was sufficient.
The Bureau
Under whose written order did you march? Produce it.
Winslow
The order came down. I do not keep the paper that sends me out, sir. I keep the paper that answers for what I did when I was there.
The Bureau
Then produce that one. How many men, how many days, and what did the week cost the county?
Winslow
I was not the paymaster. We were out four days. The men were fed, and I signed for the feeding of them, and where that paper has got to is not my affair.
The Bureau
What did you take up from the field, and by what conveyance?
Winslow
Burnt timber. Iron off a burnt cart. Such rubbish as a fire leaves behind it. Two wagons, because one would have made two journeys and I did not care to be there twice.
The Bureau
Where were the wagons taken? Name the place.
Winslow
To a place appointed. I am not at liberty to name it.
The Bureau
By what instrument were you placed under that disability? Name the act, the officer, or the oath. In 1790 the State of New York had no power to seal a militia captain's mouth.
Winslow
Then I am under nothing but my own word, and my own word is that I will not say it. You will find that answers to the same effect.
The Bureau
Your report is quoted as saying that the nature of the wreckage defies all current understanding of our world and its workings. Is that your sentence?
Winslow
It is not how I write. I write that the ground was burnt over about an acre, that no person was hurt, and that the men were discharged upon the Thursday. Somebody has been at my report with a better pen than mine, and I would like to know his name as much as you would.
The Bureau
You have since given the whole matter as a lightning strike and a burnt tree. That cannot stand beside a removal in two wagons under seal.
Winslow
It stands well enough. A man may know what a thing was and still not know what it was he was carting. I have said both, and I have been called a liar for each of them by different people, which I take to be some sort of proof of honesty.
The Bureau
Do you wish to correct anything you have said to the Bureau?
Winslow
I wish to be believed. I am aware that is a different thing, and that a man does not earn the one by asking for the other.
Plates: the Removal
The militia at Rollandville. The Bureau has asked for the muster roll, the order to march, and the county's account of what the week cost. It has had none of the three.
The Bureau’s Finding
The file is closed provisionally and against the Bureau’s inclination. There is no Rollandville in the census of 1790, no notice of the fall in any surviving paper of 1790 or 1791, no debris, no diary, no muster roll, no county claim for four days’ forage, and no order under which a militia captain removed private property from a private field. The one document is a map without a scale, a surveyor or a date, drawn in a hand that is not the hand of its stated century. The debris grew a face a hundred and sixty-four years after it fell. The Bureau is satisfied that Rollandville is a story about 1947 wearing the clothes of 1790, and that most of its wardrobe was acquired between 1948 and 1954.
And yet the desk declines to call Elijah Comstock a liar, for a reason it wishes to set down plainly. Suppose, for the sake of the argument, that a burning mass really had come down in a rye field in New York in the summer of 1790, and that a plain man had walked out at first light and picked up a cool, smooth, unfamiliar stone. Every learned authority alive would have told him that no such thing had happened, because stones do not fall from the sky. They were wrong, and they had thirteen more years of being wrong in front of them. It is a poor sort of scepticism that gets the right answer for the wrong reason, and a worse sort that mistakes the right answer for a method. The Bureau rejects the Rollandville account on the evidence, and notes for the record that the men of 1790 would have rejected it on principle, and that only one of those is a defensible way to run an archive.
