Aurum Roma — 1815 to 1837 — the Age of the Black Crusades

Historical context

1815. The old world has just abdicated. In Aurum Roma the last surviving Veheranic nobles lay the crown of Teorbano III at the foot of the throne, and Aurelianus I — the First One, elected two years earlier on the news that a frontier garrison had held five months of winter with no angel in the field — proclaims the line. The Age of Ash closes here. Thirty-five years of collapse, forty million dead, and a continent that no longer resembles any map drawn before 1780.

What Aurelianus I founds next is, on paper, the most ambitious thing anyone has attempted since the Cataclysm: seven internal States under a single pontifical authority; the frontier of the Limbus Mundi fixed as the definitive limit of the human world; the Perpetual Vigil instituted in its canonical form; and the refounding of the Clerical Military Orders, decreed among the founding acts of 1815.

On paper.

In practice, in 1815, the Orders have no budget, no charter of recruitment and not one line of written doctrine. The secular crowns go on making their own war and buying their own purified mineral on the open market, as one buys lead. The Curia decrees and cannot enforce; declares a monopoly of revelation it has no means to exercise; and governs a continent that obeys it at mass and ignores it in supply.

Between that decree and its reality lie six years of nothing, and a disaster.

In the summer of 1821, nine secular houses raise nine thousand men and go into the Limbus Mundi on their own account, to prove that salvation has no sole supplier. It is the Crusade of the Nine Standards. The ammunition lasts eleven days. Not one standard comes back.

The Curia need argue nothing further. In the three years that follow, the Orders receive budget, recruitment and doctrine — because now there is an institutional vacancy for them — and the monopoly that could not be legislated in six years is handed to Aurum Roma by nine thousand men Rome did not kill.

This book does not go into the Limbo. It stays in Aurum Roma, and looks outward from a desk, a wall and a transcription table. It is the story of how a decree turned into an empire in twenty-two years, told by three people who built it without ever having decided to: the man who did the sums, the man who gave up his name and the man who wrote down what the Angels said.

Heaven answered a clergy with no head, and it was by answering that the clergy came by one.

"Decreeing is cheap. That is why we decreed so much, and so fast, and with so little shame. A decree costs ink and a man to write it. What costs is the second line: who pays, who goes, and by what authority he is told to go.

We spent six years writing the first line.

The second was written by other men, on a plain, and not in ink."

— Cardinal Amandus Krell, Memorial on the Means, 1831, preface (unpublished)

ACT I — THE DECREE

I. The Room of Means

The Decree of Refounding of the Clerical Military Orders had fourteen articles, took up two and a half sheets, and Amandus Krell read it for the first time on the 3rd of November 1815, standing, at the window of the second courtyard, with the afternoon light coming in from the side.

He read it slowly, as he read everything, and when he reached the end he went back to the beginning and read it again, and on the second reading he did what he always did and what was the one thing that set him apart from three hundred other men of the Curia equally competent and equally devout: he read it looking for the numbers.

There were none.

This is not a figure of speech. Krell went down the fourteen articles with his thumb and did not find a single digit that was not the numbering of the articles themselves. There were three Orders, written out in words. There was transverse jurisdiction over the seven States. There was direct and immediate answerability to the One, without the intermediation of any Cardinal-State. There was, in the ninth article, a sentence that made him stop and go back and read it three times, because it was beautiful and because it was, from where he stood, catastrophic:

The Orders shall be endowed with the means necessary to their end.

Shall be endowed. Passive voice, future tense, agent unspecified.

Amandus Krell was forty-seven, had spent nine of those years as the man who closed the accounts of the See of Aurum Roma, and knew — knew it in the particular and slightly humiliating way a bookkeeper knows things theologians do not — that shall be endowed is how a man writes I do not know without writing I do not know.

He folded the two and a half sheets. He went to work.

The Congregation of Means occupied, that year, three rooms of the third courtyard, and its name was a joke nobody said aloud, because the Congregation of Means was precisely the organ of the Curia that had none.

It had been created in 1814, in the second year of the pontificate, for a task Aurelianus I had described in an audience of eleven minutes as knowing what we have.

Eleven minutes. Krell had counted them, because there was a clock in the room.

And the task had seemed to him at the time modest and even faintly insulting — a man of forty-six asked to take an inventory — and it took him eighteen months to understand that it was the hardest task in existence on that continent, because the honest answer to the question what do we have was, in the year 1815, nobody knows.

Nobody. Not through negligence; through destruction.

The Ash had burned the diocesan archives of eleven of the nineteen ecclesiastical provinces of the east. It had killed, in thirty-five years, forty million people — and a population that loses forty million does not only lose people, it loses record: it loses the parish priests who knew how many souls there were in the parish, the clerks who knew what the tithe yielded, the old men who knew where the boundary stone of the holding stood. Krell received, every week, letters from bishops informing him that the diocese had been reconstituted and that the books prior to 1783 had been lost, and the wording varied and the fact did not.

Aurum Roma governed a continent whose exact size it did not know.

And it was for that reason — and Krell only put this in writing many years later, in the Memorial, and even there with care — that the pontificate of Aurelianus I decreed so much and so quickly. A decree is a thing that can be written in a morning, in a building that did not burn, by a man who is alive. A decree does not need to know how many souls there are in the parish.

There were, in November 1815, twenty-nine founding decrees published in thirty-one months.

Krell had them all in one folder, and the folder had a name he had written in pencil on the cover and showed to nobody:

To Do.

The first serious exercise he made on the Military Orders took him fourteen weeks and filled nine sheets, and the conclusion was on the first line of the ninth, because Krell wrote his reports with the conclusion at the end as the trade requires but wrote them knowing the conclusion from the third week.

The conclusion was this:

To endow the three Orders at the establishment foreseen in the Decree, the See would need to dispose, annually, of forty-one per cent of the total known revenue of the territory under obedience.

Forty-one per cent.

The See disposed, that year, of everything taken together — tithe, rent of estates, seal dues, the crowns' donative, and its share in mineral purification — of somewhere between eleven and thirteen per cent of that, and the interval was two points because Krell could not close the figure, and could not close the figure because three of the seven internal States had still not delivered their accounts for 1814.

Eleven to thirteen. He needed forty-one.

He took the report to the audience of February 1816 and read it aloud, entire, all nine sheets, and Aurelianus I heard him out without interrupting once.

The One was, that year, sixty-one years old and three years into his pontificate. He was a dry man with a long face, who had been Cardinal Sergio Hoppen and had refounded the See of Aurum Roma with forty men and a building without a roof, and who from the 17th of October 1813 had ceased to have a baptismal name. He received standing. Everyone in the Curia knew he received standing and nobody knew why, and the explanation — Krell only had it in 1830, from a chamberlain's stray remark — was that sitting hurt his back and that he considered it unseemly for this to be known.

When Krell finished, there was a silence of perhaps ten seconds.

Then the One said:

"Your Eminence is telling me the decree is impossible."

"I am telling Your Holiness that the decree is impossible on the known revenue. I am not saying it is impossible."

"It is the same thing."

"It is not, Most Holy Father." And Krell, who had rehearsed that part for four days, turned to the second sheet. "There are three ways to cover forty-one per cent. The first is to raise the revenue, and there is no way: the continent has lost forty million taxpayers and will not recover them in this generation. The second is to reduce the establishment foreseen, and Your Holiness will not wish to do it, because the establishment foreseen is published, and to reduce it is to admit in public that we decreed beyond the possible."

"And the third."

"The third is to charge those who today pay nothing."

Aurelianus I looked at him.

"Who today pays nothing?"

"Everybody who buys purified mineral and does not buy it from us," said Krell. "Which is to say, Most Holy Father: everybody."

Here it is necessary to say what the market in consecrated mineral was in 1815, and to say it in the tedious and exact manner in which Krell said it, because everything that happened in the twenty-two years that followed is contained in that description and in nothing else.

The revelation of the Hashmal had opened in 1787. Twenty-eight years on, the world knew seven things: that there was a family of minerals; that none of them worked raw; that purification was a rite; that the rite only worked if the celebrant had genuine faith; that genuine faith could not be verified by any instrument; that a man who purified sickened; and that nobody knew why.

And the world did with that what worlds do: trade.

They dug in Ferrum Alpina, in Kal'arn, in Pradel, in the islands, in the volcanoes of the south. The raw ore was sold to whoever wanted it — and raw ore was legally free across the whole continent, because it was, materially, stone. The stone was taken to a cleric. The cleric prayed over it. The woken stone was sold at double, at treble, at sixfold, according to the war.

And the cleric who prayed over it might be a canon of the Order of Aurum Roma; or a pastor of the Schism of the North, who answered to the King of Königsmark and did not recognise the One; or a minister of the Solar Church of Brytonia, who answered to Parliament and to the monarch and thought the Roman Curia a continental superstition; or a priest of the Riti Meridionali, who prayed in a Latin older than Rome's and whom Rome tolerated because it needed the Meridian fleet; or a village priest in the Cantons with no licence whatever, to whom the neighbours brought what they had because there was nobody else and because it worked.

Because it worked.

Krell wrote that sentence into the report of 1816 and then struck it out, and then wrote it again, and then struck it out again, and in the end delivered the report without it, and spent the rest of his life knowing he had struck it.

It was the central fact of the world and it did not fit in a report on means.

"Charge those who do not pay," the One repeated. "By what right?"

"By none, today."

"And tomorrow?"

"Tomorrow, Most Holy Father, by the right that follows from our being the only ones able to do that in quantity."

It was the first time the proposition was said aloud inside that palace, and Krell was never certain the One had taken it in the sense he had meant it, because the answer of Aurelianus I was theological and Krell's proposition was industrial.

The One said:

"The gift is God's. It is not ours."

"The gift is God's," Krell agreed. "The quantity is ours, or may come to be. A village priest blesses forty cartridges in an afternoon and has no hands until Thursday. A house of purification with sixty canons on shift blesses forty thousand in the same month and does it again the month after. Neither of the two is holier than the other. One of them arms an army."

Silence again.

"How many such houses have we?"

And Krell — who knew the answer, and had spent fourteen weeks making the figure, and hated saying it because it was the worst line of the nine sheets — replied:

"None, Most Holy Father. We have two in Ferrum Alpina with nineteen canons between them, and one in Sal Pradelica with eleven, and that is everything that exists in the world. The rest purify where they can, in a sacristy, alone, unrecorded, and die without anyone knowing they died of doing it."

The audience lasted fifty-one minutes. Krell timed it, because there was a clock.

He came away with three things.

The second was authorisation to draft a project for an imposition on mineral purified outside obedience — a tribute that could not be collected for six years and that in 1822 was collected without any resistance at all, and the difference between the one and the other was not juridical.

And the third was a sentence the One said to him at the door, with his hand already on the latch, without turning, and that Krell did not write down anywhere for fifteen years:

"Cardinal. You have made me a sum and the sum is right."

"Thank you, Most Holy Father."

"It is not praise." Aurelianus I turned then. His eyes were very pale, which showed the more because the rest of him was dark. "It is my difficulty. You make sums that are right and I have to govern with them, and the sums that are right all tell me the same thing, which is that I cannot. And I decreed that I could, and the decree is published, and there is Christendom that sleeps better for it."

He stood there a moment.

"Go on making them," he said. "And do not read them aloud in any room but this one."

II. The Man With No Surname

The quarters of House Solar, in the winter of 1816, were a disused Cistercian monastery eleven kilometres from Karentin, with a good roof, sixty-two men inside, and no paper anywhere stating whose it was.

Mattias was twenty-six and had been, since October, the most senior officer in the place, which signified nothing because there existed no legally constituted rank above his or below it. The men called him captain out of a habit inherited from another war. He did not correct them, because correcting them raised the next question.

The next question was: then what are you, sir?

And the honest answer, in 1816, was: I am a man of twenty-six who has sworn three vows to an institution that exists in a decree and nowhere else, and who eats here because the abbot of Vallia sends barley twice a year out of personal charity and not out of any obligation whatsoever.

He had given up his family name in 1809, at nineteen, and the story of it is short and is not edifying, and Mattias told it whole three times in his life.

His house was a valley house, small, with arms from 1640 and seven villages of rent, and the Ash had gone through it in the way it went through a thousand others: not with demons — the high valleys of Ferrum Alpina were never entered, the evacuation of 1784 had worked, the Solar Range held — but with hunger, and with typhus, and with twenty-nine years of a broken market.

In 1808 his father sold two of the seven villages. In 1809 he wanted to sell a third, and Mattias, who was nineteen and had the particular insolence of nineteen, told him at table that selling land to keep the house was keeping the house by letting it eat itself, and his father answered him something that Mattias never repeated to anybody, and Mattias got up from the table and walked out, and went on foot to the nearest high monastery, and arrived there after two days, and asked to be let in.

The monk who received him asked his name.

And Mattias — who was going to say the name, who had the name in his mouth, and who had in that half-second a clarity that lasted him his whole life and that he never managed to explain as devotion because it had been no devotion at all, it had been calculation: if I say the name, in three years my father writes to this monastery, and I go back —

Mattias said:

"I bring none."

The monk wrote Mattias in the book and wrote nothing else, and it is for that reason, and only that reason, that there exists today in the canon of the Clerical Order a vow called the Vow of Name.

In 1816 nobody knew this and nobody had any reason to. There was a dozen men in the Orders with no surname on record, for various reasons and nearly all of them worse than his — sons of pact-bound houses, sons of hanged men, sons of nobody — and it was not matter for a rule, it was matter for a biography.

What was matter for a rule, in 1816, was barley.

Sixty-two men eat, in a year, and Mattias had done the sum because he was the one who had to: a hundred and forty-three hundredweight of grain, plus salt, plus fat, plus whatever is shot. The abbot of Vallia sent ninety. The remaining fifty-three came from three places: from the wood the men cut and sold in the valley, from the smith-work they did for the villages, and — from the autumn of 1815, and Mattias authorised it and it took him eleven years to stop thinking about it — from hire.

Men were let out.

It was not called that. It was called escort. A merchant taking cargo from Karentin to Bernkanton hired four Knights of the Seal for twelve Aurelian ducats and a week's provisions, and the four went, and came back, and the twelve ducats went into the quarters' strongbox, and with forty-one escorts a year the barley shortfall was closed.

It was, in Mattias's view, the nearest thing to hell he had done up to then: men who had sworn to fight the demon walking beside bales of tin along a road where there was nothing, paid by the week, for a merchant who called them gentlemen to their faces and something else behind their backs.

It was also the reason they were alive in 1821 to be an Order.

Mattias understood this very late, and by the time he understood it he was Captain-General and had nobody left to say it to.

The worst was in September 1817 and it is the one he told when somebody asked him for a story of the old days.

A tin merchant of Bernkanton hired four men to take cargo from Karentin down the valley, eight days, twelve ducats. Mattias sent four of the best, because he always sent the best on the paid escorts — a captain of sixty-two men with no budget learns quickly that reputation is the inventory.

On the third day the merchant wanted the four to help unload.

It was not malice. He had paid for four men for eight days and the four men spent the day walking beside the cart doing nothing, and he was a practical man of a practical republic, and he asked.

The sergeant refused. Politely, and refused.

And the merchant said the thing that Mattias heard from the sergeant eight days later and that stayed with him twenty-seven years:

"Sir, I am paying you to be here. If this is a service of God, do me the kindness of telling me so, and I shall pray and save twelve ducats."

The four came back with the money and with the cargo not unloaded, and the sergeant told him everything, and Mattias listened standing in the yard and then told him he had done right and sent him to rest.

That night he went to the magazine and stayed there.

The magazine held, in September 1817, four hundred and eleven cartridges of Silver Chrism dead since 1814, sixty-one pikes, nineteen Choir-Iron swords in good condition and forty-two in bad, and three barrels of common powder.

Mattias sat down on the floor between the crates and made the sum he remembered better than any other of his life.

Forty-one escorts a year at twelve ducats is four hundred and ninety-two ducats. The barley shortfall cost four hundred and fifty-one. There were forty-one ducats a year left over for sixty-two men — which is what it costs to shoe two horses and buy salt.

And the forty-one escorts consumed, in man-days, the equivalent of seven weeks' work of the entire house.

Seven weeks a year walking beside carts of tin, to pay for the barley that allowed sixty-two men to go on swearing three vows to an Order that existed in a decree of fourteen articles published in Aurum Roma nine hundred kilometres away.

He wrote that night, in the notebook, the four bitterest lines that are in it, and wrote them at twenty-seven:

He is right and that is what is wrong with it.

If this is a service of God, one does not charge for it. If one charges, it is a trade, and a trade is defended by wages and not by vows, and in that case my three vows are a badly drafted contract.

I do not know how to answer him. I spent the night trying.

The only thing I have is that tomorrow morning I am still here, and he would not be.

In February 1822, when the endowment arrived with its twenty-two sheets and the annual sum written in figures and in words, Mattias read the article on pay and the first thing he did, before ordering the muster sounded, was to go to the magazine alone.

He stayed there as long as he had stayed in 1817.

He wrote nothing that night. He wrote four days later, and it is two lines:

The escorts are finished.

It took me five years to understand that the answer to the tin merchant was not theological. It was budgetary, and it came by post.

The monastery had a library of forty volumes and Mattias read them all over three winters, because there was snow from November to April and because the alternative was playing cards with sixty-two men for five months.

One of the forty was the Rule of Saint Benedict.

Another was a treatise on fortification of 1701, in bad Latin, with plates.

And the third that matters was a thin book with no cover, that somebody had left there and that had neither author nor date and that was, by Mattias's reckoning, the diary of some canon of the 1790s — a man who had been among the first to purify Choir-Iron after the revelation opened, and who had written a hundred and eleven pages about it, and who had died, presumably, because the diary ends in the middle of a sentence.

Mattias read it four times.

On the second reading he began taking notes. On the third, he began to experiment.

What he did that winter of 1817 was forbidden, and he knew it was forbidden, and the argument he gave himself for doing it was of a casuistry so frail it would have shamed any second-rate canon: the rite of purification is reserved to ordained clergy; he was not ordained clergy; but the quarters had had no cleric since 1814 and doctrine — the doctrine that did not yet exist, that he was at that moment inventing alone in a mountain monastery at twenty-seven — told him that a unit without a purifier withdraws, and to withdraw from there meant dissolving the only armed body the Order had in the field.

So he experimented.

He took a cartridge of dead Silver Chrism, from the four hundred they had in the magazine and that had been dead since 1814 because there was nobody to wake them. He did what the canon's diary described, which was seven steps and a prayer and a length of time.

And nothing happened, and he did it again the next day, and nothing happened, and on the eleventh day it happened.

There was no light, no sound, no heat. A brass cartridge does not change appearance when it wakes; that is what makes the whole business so hard to teach and so easy to counterfeit. What Mattias felt was something in his left hand, with which he was holding the cartridge, and which he described forty years later to a clerk of House Solar in the only words he ever found:

Like when you are lifting a weight and somebody takes the other end.

It took him four hours.

Four hours for one cartridge. He did it again the next day and it took three and a half, and by the end of the winter he was at three and a quarter, and there he stopped improving and never improved again in twenty-seven years of the trade.

A licensed canon of Aurum Roma did it in twenty seconds.

Mattias did not know that in 1817. He learned it in 1823, from an officer of the Congregation of Means who told him in a corridor, with no malice whatever, apropos of something else — and Mattias, who was a man of remarkable composure and of whom there is no record of a single raised word in his entire career, had to go out into the yard and stay there a while.

Twenty seconds.

He had taken eleven days to get the first and four hours to make it, and a canon of twenty-three in a house in Ferrumkar did a hundred and eighty an hour while thinking about something else.

And yet — and this is where the Vow of Measure comes from, and this is why it exists, and this is why it is the vow that defines the Order and not the other two — and yet Mattias went on.

He went on because in February 1818 one of the sixty-two, a lad from Bierkut named Halvard, went out with three men on a routine escort into the valley and found, nine kilometres from the quarters, in an abandoned shepherd's hut, a thing.

Nobody knows what it was. The report Mattias wrote — the first combat report in the history of House Solar, four paragraphs, badly drafted, which hangs framed in Karentin-Alta today — describes a form of the height of a man, with no constant face, and says nothing more, and says nothing more because the four men could not agree on anything more.

Halvard fired five shots.

Four were dead cartridges from the 1814 magazine, and they went through the thing like stones through smoke.

The fifth was one of the eleven Mattias had woken that winter, at four hours apiece, and which he had issued unwillingly and with express instruction not to spend them.

The fifth killed it.

Mattias went to the place the next day. He took two men and a spade.

There was nothing to bury — nothing is left, it is one of the first things one learns and one of the hardest for men who come from the secular war, where there is always a body — and so what he did was go into the shepherd's hut, sit down on the beaten earth floor, and stay there perhaps an hour.

And the thing he thought in that hour, and which became, by a route of twenty years that nobody planned, the doctrine of an Order of a hundred and ten thousand men, was this:

I took four hours to make that shot and it is worse than one made by a canon of twenty-three. And it worked all the same. And if it worked all the same, then the difference between me and the canon is not of kind — it is of speed. And if it is of speed, then never, in any circumstance, and however badly things go, can a man of this Order stay in a place without somebody to wake his ammunition. Because with somebody to wake it, badly, slowly, even at four a day, he kills. And with nobody, he is firing stones.

He wrote it that night, in four lines, in a notebook.

The four lines went, almost unaltered, into the written doctrine of 1824 as the third rule of the Knights of the Seal, and they are the ones that say that a unit without a purifier in condition withdraws even when winning, and that it is the only order of withdrawal a captain may give without consultation.

In May 1818 there arrived at the quarters of House Solar a dispatch from Aurum Roma, on good paper, sealed, addressed to the Command of House Solar of the Order of the Knights of the Seal, which made three men laugh because not one of those four words corresponded to an existing thing.

The dispatch requested information, for census purposes, on the purifiers in service with the unit.

There was a form. It had fields for name, age, house of formation, year of licence and declared monthly output.

Mattias looked at the form for a while.

Then he wrote, in the field for the name, none; and in the field for monthly output, eleven cartridges, by an unordained officer, by an unlicensed method; and signed it; and sent it.

And he was left thinking, as the knight rode down with the post, that it would either earn him a prosecution or it would come to nothing, and that of the two possibilities the second was the worse.

III. Another Man's Hand

A clerk of the Vigil works three paces from the chair, side-on, with the flame outside his field of vision, and never looks.

It is the first thing taught and the only one taught with a threat. All the rest of the trade — the hand, the speed, the abbreviations, the way of turning a leaf without sound — is taught by patient correction. The do not look is taught by telling a boy of seventeen, in his first week, that there were four clerks before him who looked and that there is no record of any of them writing anything intelligible again.

Aelius de Sárvar had entered in 1790 and had never looked.

Twenty-six years, in 1816. Four Mothers Vigilant. Thirty-one Blancas. Two thousand four hundred transcriptions.

And he was, though nobody in the Curia had yet grasped it, the living man who knew most about the thing the whole Curia existed to administer — not by merit and not by study, but for a purely mechanical reason: he was the only one who heard them all.

The Blancas do not speak to one another. It is a rule of the Vigil from before there was a Vigil. A Blanca serves the shift, is released, the next comes in, and what one said on Wednesday the other does not know on Thursday. The Mothers Vigilant change every seven years. The attending canons rotate every two. The cardinals read the abstracts.

The clerk stays.

The clerk stays because a clerk decides nothing, and because replacing him costs two years of training, and because his hand must be comparable across decades so that old transcriptions remain legible by the same criteria.

Aelius stayed fifty-one years.

It is worth describing the room, because almost nobody has seen it and because almost everything the continent believes about it is wrong.

It is not a chapel. There is no altar, no image, no gold. It is a rectangular stone room, seven metres by eleven, without windows, with one door to a corridor and another to the women's house, and with the floor laid in cork so as to make no sound.

At the far end there are fourteen flames.

They are not the Seven Candles of the Perpetual Vigil, which stand in the Cathedral of the Choir and which everybody knows from the engravings and which mark the theoretical return of the Archangels. These are another thing, and that is why the confusion is permanent and why the Curia never took the trouble to undo it: fourteen Lumen lamps, mounted in a row a metre and a half from the floor, fed continuously since 1791, and carrying no liturgical significance whatever.

They are there to see by.

A Blanca in inscription does not speak loudly. She speaks at a normal voice or below it, and she speaks fast, and a clerk needs light enough to write by and not enough for anything else — and somebody in 1791 discovered, by trial, that fourteen Lumen lamps at a metre and a half give exactly that and that fifteen give too much.

It is the best-established number in the whole liturgy of the Clerical Order and the reason for it is that sixteen dazzled.

The chair stands three metres from the flames and faces them.

The Blanca sits at a quarter past six, with the shift cincture — a band of white linen tied at the waist, which has no ritual function at all and which serves so that the assistant may see, from a distance and from behind, whether the posture has changed.

She stays there two hours.

This astonishes everyone who hears of it for the first time. Two hours. It is not a night, it is not a forty-day vigil, it is not fast nor hair-shirt: it is two hours seated in a chair in a stone room with good light, four times a week, and for the rest of the day the woman lives as any other woman of a religious house — eats, sleeps, walks in the cloister, has an hour of reading and another of sewing.

And dies at twenty-four.

Aelius de Sárvar wrote about this once only, in a letter, and wrote the sentence that sums up what forty years in that room taught him:

There is nothing extraordinary happening there. That is what makes it unbearable to watch.

If there were a storm, if she screamed, if lightning fell — people would understand. A person understands that a violent thing wears out a body.

She sits, and says nine words on a Wednesday, and goes to supper.

And five years on she dies, and nobody can point to the day it was taken from her, because it was taken two hours at a time, four times a week, in instalments too small to be noticed.

The shift has three parts and the second is the one nobody can explain.

The first is the entry: the Blanca sits, the assistant ties the cincture, the clerk lays the sheet, and the three of them keep silence until the quarter hour.

The second is the waiting. Forty minutes, an hour, sometimes the whole two hours. Nothing happens. The woman sits looking at fourteen lamps and nobody knows what she is doing, and — this is the point and it is what the Curia never wished to hear — they do not know either.

Aelius asked, over fifty-one years, four different Blancas, against all protocol, what it was they did during the waiting.

The four answers are on a paper of his and they are these:

I do nothing.

I wait. Like somebody waiting at the door of a house.

I am listening for whether it comes. It is not listening with the ears.

I do not know. I never knew. At first I tried something and then I understood that when I tried it did not come.

The fourth is the Nineteenth Blanca's and was given in 1837, and Aelius laid it beside the four pages Mattias of House Solar had written thirteen years earlier about eleven men in an alpine monastery, and the two documents lie today in different archives, in different cities, catalogued by institutions that do not speak to one another.

They say the same thing.

The third part is the leaving, and it is called the Release, and it is the only part with a rite.

The assistant unties the cincture. The Blanca rises and says her name — her own name, the baptismal one, the one she stopped using on the day she entered — aloud, once.

That is all. There is no formula, no response, no blessing.

Aelius heard that four thousand six hundred times from thirty-one different women and wrote, at seventy-five, that it was the only part of the trade he never managed to hear with indifference:

They say the name to remind themselves that they have one.

Twenty-nine of the thirty-one I transcribed are dead, and I know all their names, because I heard them say them four thousand six hundred times, and I cannot write them down anywhere because in service they have no name and after death they have the service name on the stone.

I am the only place in the world where those names are together and I am blind and seventy-five years old.

When I die they die again, and the second time it is in earnest.

The transcription of the 14th of April 1817 is the most cited of his career and is the reason the seven rules of the Hashmal Syntax exist, although the rules were only written nine years later and by somebody else.

On that day the shift was held by a Blanca of twenty-one, in her third year of service, whom the Vigil recorded as the Fourth Blanca, because Blancas use no surname in service and because in that generation they were numbered by order of entry into the house.

The shift went as shifts went: forty minutes of nothing, then inscription.

And what she said — and Aelius wrote, as he wrote everything, unpunctuated, because to punctuate is to interpret and a clerk of the Vigil does not interpret — was a sequence of forty-one words of which thirty-four were technical terms of metallurgy that a girl of twenty-one born in a shepherds' village had no possible way of knowing, and the other seven were an instruction of proportion.

It was an alloy.

It is not the only thing the Blancas say, and it is not even the commonest. Most of what comes out of that chair is fragment, image, the name of a thing, a warning without a subject. But now and then — and the frequency was falling, and fell throughout the century, and by 1887 was what it was — now and then a recipe comes out.

Aelius wrote the forty-one words.

And then, because something had caught his attention, he did the thing he never did: he asked the Mother Vigilant, at the end of the shift, for authorisation to consult his own old transcriptions.

It took nine weeks. He had two thousand four hundred to go through and could only see them in free hours and could not take them from the room.

And what he found was this, and he wrote it on a half-sheet which he gave the Mother Vigilant in June 1817 and of which he kept no copy:

That those forty-one words had been said before.

Twice. Once in 1804, by a Blanca who had died in 1806. Again in 1811, by another, who had died in 1813.

Not identical. Partly identical — twenty-nine of the forty-one words coincided, in the same order, and the twelve that varied were all of the same kind: they were the numbers.

In 1804 the proportion was one. In 1811 it was another. In 1817 it was a third.

And the three proportions, set in sequence, were not random: they ran in a direction. They were converging.

Aelius de Sárvar, fifty-five years old, a clerk, with no formal education beyond the minor seminary and the hand, wrote on the half-sheet the sentence that is today the first of the seven rules of the Hashmal Syntax, and wrote it in the language of a clerk and not of a theologian, and that is why it still sounds strange in the manuals:

One does not receive the same thing twice. One receives the same thing corrected.

The Mother Vigilant of that year was called Irmgard and was sixty-seven with three months of life ahead of her, and she read the half-sheet twice and then asked Aelius a question he had not expected:

"Do you know what this means for the girls?"

Aelius did not know.

"It means," said Mother Irmgard, "that what they hear is not a message. A message is said once. That is a correction, and a correction presupposes that somebody on the other side is following what we did with the earlier version." She set the paper down. "Which is to say they are watching us all the time, and not only when they speak."

The two of them were silent.

"Write that," she said.

"I have written it, Mother."

"You wrote the rule. Write the consequence."

And Aelius took another half-sheet and wrote the consequence, and the consequence was far worse than the rule, and it was the consequence the Curia buried for nine years and that was only dug up again in 1826 by a woman of thirty-eight named Sofia Kranz who found it while tidying a cupboard.

The consequence was:

If they correct, it is because they saw it fail. And if they saw it fail and go on correcting, then we are not being rescued. We are being taught, and whoever teaches expects the pupil to end by doing it alone.

Aelius had, that summer, the only documented period of insomnia of his life, and it is documented because the Vigil noted the condition of its clerks in the same way it noted everything.

Eleven weeks of sleeping badly.

Then it passed.

What he did next, and did every day for another twenty-four years, was to go to the room at half past five in the morning, sit side-on three paces from the chair, with the flame outside his field of vision, and write what he heard without punctuating it.

Because it was what he knew how to do, and because somebody had to do it, and because — and this he said once, to a great-nephew, in 1839, already blind in both eyes, with a naturalness that struck the great-nephew at the time as almost disappointing — one does not need to understand a thing to copy it well, and copying well is rarer than understanding.

IV. What Rome Did Not Have

The census of purifiers closed on the 19th of December 1819, at nine in the evening, and Amandus Krell was present when the last bundle of returns came into the room, because he had waited forty-five months for it and wanted to see.

It was seven provinces, a hundred and eleven dioceses, four schisms that did not answer and three that answered badly.

The figure was three thousand and forty-one.

Three thousand and forty-one men and women, in the entire territory of the One's obedience, capable of waking consecrated mineral in measurable quantity.

Krell had estimated, in 1816, somewhere between six and eight thousand.

He spent that whole night in the room, and he did not spend most of it lamenting; he spent it making the other sum — the sum the figure made possible for the first time and which nobody in the world had been able to make before the 19th of December 1819, because before that nobody knew the denominator.

The sum was the capacity of the continent.

Three thousand and forty-one purifiers. Average declared output, discounting those at the end of their careers and those who declared too much out of vanity and those who declared too little out of fiscal prudence: nine hundred and ten cartridges per purifier per month.

Two million, seven hundred and sixty-eight thousand consecrated cartridges a month. Thirty-three million a year.

It was an enormous quantity. Written out like that, on a sheet, it was a quantity that made a man want to read it aloud to somebody.

Krell divided it.

The line army of the territory of obedience, with the seven state militias and the levies added in, ran to four hundred and eighty thousand men. Thirty-three million divided by four hundred and eighty thousand gives sixty-nine cartridges per soldier per year.

The Field Canon says five a day in contact.

Sixty-nine a year is fourteen days of contact per soldier per year.

Fourteen days. An entire army of the continent could fight, in total, fourteen days a year, and on the fifteenth it was firing stones.

Krell wrote the report in four days and showed it to nobody for two months, and when he finally took it to audience he took it with an annexed proposal, because he had learned in 1816 that one does not bring Aurelianus I a figure without bringing a door.

The proposal had three articles and was, in his own later judgement, the most important thing he did in his life, and no historian of the Order ever attributed the authorship to him because it went out in the name of the One, as everything did:

First: that the formation of purifiers should cease to be done by apprenticeship to an older canon, as it always had been, and should be done in dedicated houses, with a programme, an examination and a licence — and that three should be created, one in Aurum Roma, one in Ferrumkar and one at the Schola Maior.

Second: that a target be fixed. Not an ideal: a target, with a figure and a date, revised every ten years.

Third — and this is the article Krell wrote with a shaking hand, and which he knew to be, in strict theology, nearly indefensible: that the Curia should begin to measure the output of every licensed purifier and record it, by name, monthly, for life.

Because, Krell wrote in the justification, we cannot administer what we do not count; and if the gift is God's and the quantity is ours, then the quantity is matter for administration, and to administer without measuring is to pray.

The target was fixed in 1826, six years later, and it was eleven thousand.

The figure came from a calculation of need that Krell made on four sheets and that began from a simple question and an answer nobody wanted: how many purifiers are needed for the continent to be able to fight the whole year?

The answer was nineteen thousand four hundred.

Eleven thousand was what the Congregation of Means judged possible to train in thirty years, given the known mortality of the trade.

That is to say: the target was never sufficient. It was, from the day it was written, fifty-seven per cent of what was needed — and everybody who wrote it knew that, and the real figure never reached eleven thousand in seventy years, and in 1887 a diocese of Sal Pradelica drew back twenty-two kilometres because it could not garrison a sector.

Krell died in 1841 and never knew any of that, and it is as well that he did not.

There is one thing he did know, and which marked him more than the figure, and it is the last thing that matters in this chapter.

In the final bundle of the census, among the returns of the diocese of Solnica, there came a form filled in by a hand that was not a clerk's, in the upright, small writing of somebody who writes seldom and with care.

In the field for the name it said: none.

In the field for monthly output it said: eleven cartridges, by an unordained officer, by an unlicensed method.

It was signed Mattias, of House Solar.

Krell looked at it for a long time.

There was, on his table, a pile of thirty-one cases the Congregation had to refer to the Inquisition — the Inquisition that had not yet been reformed, that was still a slow and scrupulous and public tribunal, and that would have taken two years to try that and would have convicted, because the law was plain.

Purification by an unordained man is matter of functional heresy. It has been in the Canon since 1791.

Amandus Krell picked up the form, read it a third time, and put it in the other pile.

Then, because he was the man he was, he wrote in the margin, in pencil, for himself, a note of three words that stayed on the paper and that is on the paper still today, in the Archive of Means, and that nobody understood for a hundred and twenty years because nobody knew what he had in his head that December:

Eleven. Not none.

V. The Petition

The petition of the nine houses came into the Curia on the 11th of March 1820, by ordinary channel, docketed as a request.

It was an excellent document. Krell said so at the time and repeated it in 1831 in the Memorial, and it was excellent because it had been written by a jurist of Reichshamn who knew exactly what he was doing: it asked leave for nothing. It asked for provision.

The text, reduced to its nine useful lines, said this:

That the signatory houses proposed to take a host into the Limbus Mundi, in the summer of 1821, with nine thousand men, under the proper right of every crown and house to defend Christendom.

That, the gift of the Hashmal being universal grace and the property of no See, and the Curia being the de facto holder of the means of purification at scale, the signatory houses requested the provision of field chaplaincy in number sufficient to the sustenance of the host.

That, should such provision be refused, the houses would proceed by their own means.

Nine thousand men. Provision in sufficient number.

No figure came with it. The jurist of Reichshamn put no figure in the petition, and Krell understood why by the second paragraph: because they did not know the figure. They were not being clever. They were asking for sufficient because they believed sufficient was a thing the Curia would know how to estimate in the way a bishop estimates how many priests are wanted for a procession.

Krell did the sum that afternoon, at his table, in nine minutes.

Nine thousand men, five shots a day, forty-five thousand cartridges a day. A licensed purifier yields fourteen hundred on a good day. Thirty-two purifiers on the march, and thirty-two is the theoretical minimum, and the theoretical minimum assumes that none falls ill and that all work every day, which does not happen; in practice it was forty.

Forty purifiers.

The territory of obedience had three thousand and forty-one.

Here is the fact this book exists to establish, and it must be said with complete coldness because the pulpit spent seventy years saying something else:

Rome could have given the forty.

Forty in three thousand and forty-one is a little over one per cent. Was it a great deal? It was. It was a bleeding of a capacity already fourteen days below what was needed, and withdrawing forty licensed men from the houses of Ferrum Alpina for a summer campaign meant, in round numbers, four million cartridges not made and seven frontier garrisons on ration.

But it was possible. It was not impossible. Krell wrote, on the fourth sheet of the opinion, the exact sentence, and the sheet exists:

The provision is materially practicable at grave cost.

The opinion was discussed on the 2nd of April 1820 in a session of eleven cardinals, and Krell had twenty minutes.

He used them explaining the arithmetic. It was the third time in five years he had explained arithmetic to that room and he had already understood, by the second, that the room did not hear it as arithmetic; it heard it as a position, that is to say, it heard a man defending something.

He was defending nothing at all. He was saying that forty men could be done and that it cost seven garrisons.

It is worth saying who was in that room, because the room is the thing that decided, and because the way this is told from the pulpit — the Curia refused — attributes to an institution a decision taken by eleven men with eleven different histories on an April morning.

The Secretary of State presided, aged seventy-three, with a deafness of the left ear that nobody mentioned and that meant half the session reached him by halves. He voted with the majority, as he always voted.

There were the seven Cardinals-State. Five of them had arrived in Aurum Roma the evening before, from seven different provinces, and three did not know exactly what was to be discussed because the summary of the summons said petition of secular houses, provision.

There was the Cardinal-Prefect of Doctrine, a man of forty-nine, a rigorist, who had read the petition four times and who was the only man in that room besides Krell who had come with a written paper.

There was the Cardinal-Chamberlain, who did not speak.

And there was Krell.

The Cardinal-State of Ferrum Alpina spoke second, and Krell was right to call him political, but there is a thing he never said anywhere and that must be added for the portrait to be honest: that man also had a problem of means and could not say it.

Ferrum Alpina had, in 1820, one thousand one hundred and nine of the continent's three thousand and forty-one licensed purifiers.

All forty the petition implied would come out of there. There was nowhere else for them to come from. And the Cardinal-State of Ferrum Alpina had, in his pocket, a letter from two forge-masters of Ferrumkar, dated January, informing him that the State's output of Silver Chrism was already eleven per cent below its undertaking to the See for want of hands.

He did not read it to the room.

He did not read it because reading it would be to admit, before six Cardinals-State who were his peers and competitors, that the most powerful State of the Order was failing its undertaking; and because, if he read it, the argument for refusal would become we cannot instead of we ought not.

And we cannot is a sentence one says once and then pays for over thirty years, because whoever hears it knows thereafter where the limit lies.

He made the political speech instead, and spoke for forty minutes, and spoke very well.

The argument was this, and Krell recognised it by the third sentence as the one that would win: that to yield chaplaincy to a secular host which declared in writing that it went under the proper right of every crown was to recognise that proper right; that if Rome provided, Rome became the supplier of a war it did not command; that next year it would be eleven houses, and the year after seventeen; and that the Curia would end as a house of purification seated in Aurum Roma serving clients, and not the head of Christendom.

He did it brilliantly. And he did it because the other speech — the true one, the one of means, the one Krell could have made for him with his eyes shut — was unpublishable.

Krell only learned of the letter in 1827, seven years later, auditing the State for another reason. He wrote in the Memorial:

I was the only man in that room to vote in favour of giving the forty, and I spent six years thinking myself the only one who had seen straight.

The truth is that the man who defeated me had a worse figure in front of him than mine and could not say it, and went to theology because theology was there and was free.

That is how one governs. One does not lie. One chooses, from among the things that are true, the one that can be said aloud — and then history records that one, because it is the only one that was written down.

The others spoke.

The Cardinal-State of Sal Pradelica was for giving, and his argument was a frontier argument and it was good: if the nine houses go in and are destroyed, he said, the Limbo is left with nine thousand corpses, with four hundred and ninety-four thousand consecrated cartridges and with nine standards, and none of that is a matter of indifference. He was the second vote with Krell.

The Cardinal-State of Limes Orientalis was against, and his argument was the simplest of all and the hardest to answer: I have fourteen garrisons and six purifiers for them, and if forty men leave the system this year, they leave from somewhere, and I know which somewhere it is not.

The Cardinal-State of the Schola Coronata made a speech of twenty-two minutes on the nature of grace that nobody afterwards managed to summarise and that Krell described as impeccable and irrelevant.

And the Prefect of Doctrine spoke last, with the written paper, and said the thing that closed the session.

His paper had four paragraphs and the fourth is the one that passed into history as the position of the See.

But the first three did not pass, and Krell kept a copy, and they are far worse and far cleverer than the beautiful sentence that survived.

The Prefect of Doctrine began by conceding everything. He conceded that the provision was materially practicable — he cited Krell by name, which was a rare courtesy. He conceded that the signatory houses were Catholic, that the enterprise was pious in intention, and that to refuse it would carry the cost of scandal.

And then he said:

"Eminences, what is before us is not a request for priests. It is a request for capacity. And the petition takes care not to say so, because the jurist who wrote it knows that if he said so he had lost."

Krell straightened in his chair. It was the only time in twenty-six years that anyone in that house said it before he did.

"If we provide," the Prefect went on, "we establish that the capacity to purify at scale is a thing that is requested, that is allotted, and therefore that is apportioned. And a thing that is apportioned has a criterion, and the criterion will be argued over, and in ten years there is a table of apportionment between the See and the crowns, and the See is one of the parties to that table instead of being the one who writes it."

Silence in the room.

"If we refuse, we maintain the proposition on which all of this rests, which is that the gift is not capacity: it is ministry. A ministry is not apportioned. It is exercised or it is not."

He set the paper down.

"Eminences, I am perfectly aware that this is convenient. I am conscious of it and I ask God to forgive me the part of it that is only convenient. But it is also true, and the fact that it suits us does not make it false, and we cannot leave off saying what is true merely because we profit by it."

And it was then that he said the sentence, and said it with no emphasis at all, almost in passing, which is the reason it worked:

"If the gift is universal grace, as they write, then let them receive it universally; and if they do not receive it, perhaps the gift is not so universal as the petition supposes — and it is not for us to correct God for the convenience of nine houses."

The vote was taken in inverse order of seniority, which is the custom, which means Krell voted first.

He voted placet — in favour of providing — and sat listening to the other ten.

Nine to two.

And the thing he recorded about that morning, and recorded on the day, in three lines, on a paper he put into no file, was:

I lost nine to two and I was not defeated by cynicism.

I was defeated by a man who believes what he said, who said a true thing, and who was right at every step of the argument except one that nobody in that room could see, and it is this:

a ministry that only one house in the world can exercise in quantity is capacity, call it what you will.

The name does not change the thing. It changes who has to answer for it.

Krell was one of the two, and it was not out of generosity: it was because he thought the refusal would come dear and because his trade was estimating costs.

The reply of the Curia went out on the 19th of April 1820, in seventy-one words, and is one of the most-read documents in the history of the continent because it was reproduced in pamphlet by everybody, on both sides, for thirty years.

It said:

The Apostolic See of Aurum Roma praises the zeal of the signatory houses and commends it to the prayers of the faithful.

Field chaplaincy with faculty and provision of purification on the march is ministry, and not merchandise; it is not granted upon request, nor apportioned to any host not raised under the authority of the One.

The signatory houses are recommended to defer the enterprise and to place themselves under such authority, in which case they shall be provided for.

Krell read it on the day it went out, and the only thing he thought — and he wrote it in the Memorial eleven years later, and it is the bitterest thing that book contains — was:

It was all there except the number. If we had written forty, they would have understood. We wrote "ministry, and not merchandise", which is true, which is beautiful, and which teaches nobody anything.

A man who hears "it is not for sale" looks for another seller. A man who hears "forty are needed and we have three thousand and there are seven frontiers" does sums.

We knew how to do the sum and we did not put it in the letter, because to put the sum in the letter was to admit that grace can be counted. And while we argued whether it could be counted, they bought.

The second movement of this chapter takes place six hundred kilometres away and seven weeks later, and Krell never knew it had happened.

In June 1820 there arrived at the quarters of House Solar, in the valleys of Ferrum Alpina, an emissary of the nine houses, with an escort of six secular knights and a letter of credit drawn on a Kantonal bank.

He came to buy men.

It did not say so, obviously. The proposal was for temporary incorporation of an allied contingent, for eighteen months, with pay, with provisions, with armament supplied and with — and this was the article that mattered and it was written third so as not to look first — recognition of the Order of the Knights of the Seal as a constituted body, for all legal purposes, by the nine signatory houses and by the crowns to which they are bound.

Mattias read it three times.

It was, from where he stood and from where any man would stand who had spent five winters hiring out escorts at twelve ducats to pay for barley, the only serious offer the Order had received since it had existed on paper.

Pay for sixty-two men. Armament. And juridical existence.

Juridical existence, offered by nine secular houses, five years after the One had decreed the Order and had failed to endow it with anything at all.

Mattias asked for a day. They gave him a day.

He spent it alone, climbing to the col above the monastery, which was a thing he did when he needed to think and went on doing until he was sixty.

And what decided him was not loyalty, and was not theology, and he was always honest about that with the two or three men to whom he told the story.

It was the third rule.

A unit without a purifier in condition withdraws, even when winning.

He had written that in a notebook in 1818, after a shepherd's hut and a cartridge that had cost him four hours. And the proposal on the table down below, in its nineteen clauses, did not mention purifiers once.

Not once. Mattias counted.

It spoke of armament supplied. It spoke of ammunition in quantity — it even gave a figure, and the figure was generous, and Mattias, who had spent his life counting cartridges by the unit, looked at that figure and understood in a second what it meant and what it did not say.

It said how many bullets. It did not say who was going to wake them again in September.

He came down at the end of the afternoon and refused.

The emissary asked him why, with a perfectly reasonable irritation, and Mattias gave him the answer he gave, which is fifteen words and which House Solar had cut in stone in 1863 without quite understanding why:

"Because you gentlemen counted the bullets and did not count who makes them."

The emissary left the next morning. He went next to Königsmark, where he had better luck, and then east, where he had a great deal better.

And in June 1821, nine thousand men formed up on a plain called Marnheim with four hundred and ninety-four thousand cartridges bought from a Kantonal house and one canon of Vallia on a white horse, and the sixty-two men of House Solar were not there.

Mattias learned of it in September, from a merchant.

He went on thinking about the conversation on the col for the rest of his life, and the conclusion he came to, and never said to anybody because there was no way of saying it without its sounding like something he did not want it to sound like, was that he had not saved them by any virtue at all.

He had saved them because, two years earlier, a lad from Bierkut named Halvard had fired five shots in a shepherd's hut and four of them had been good for nothing.

Sixty-two men stayed alive because one of them had spent four dead cartridges in February 1818.

That is how this universe distributes things, and there is no way of saying it that does not sound like blasphemy or like luck.

ACT II — MARNHEIM

VI. The Summer

The summer of 1821 was, in Aurum Roma, the finest in living memory, and everybody remarked on it at the time and everybody said so afterwards, and there is something obscene in the way years remember themselves.

Krell worked.

He worked because it was June and the accounts for 1820 closed in July and because the Congregation of Means had then eleven officials for a territory of a hundred and eleven dioceses, and he worked also because the alternative was to think, and he knew what he would think if he let himself.

He knew the date of departure. The whole Curia knew the date of departure. It had come in a dispatch from the nunciature of Reichshamn in May, with the information that the host would concentrate at Marnheim from the 8th of June and would cross the line at Vado Aureo in the second week of July, and the dispatch was written in that neutral tone of diplomats which is the one thing a person cannot bear.

Krell read it and then went to fetch the opinion of April 1820 and reread the fourth sheet.

The provision is materially practicable at grave cost.

He put it all away again. He went to close the accounts for 1820.

On the 4th of July there was a routine audience on the tithe of the southern provinces, and at the end of it, with the papers already gathered, Aurelianus I asked him:

"Cardinal. That sum you made last year. Forty."

"Forty, Most Holy Father."

"Is it still forty?"

"Today it would be forty-three, because we lost eleven licensed men in Ferrum Alpina over the winter and the average output has fallen." Krell hesitated, and then said it, because he had sixty-eight years of life behind him and no wish whatever to spend what was left not having said it: "But they are no longer asking, Most Holy Father. They marched on the eighth of June."

The One nodded slowly.

"I know," he said.

And then, after a time:

"You think I ought to have given them."

"I think it would have cost seven garrisons, and that I do not know what the opposite costs, and that nobody knows, and that in three months we shall know."

Aurelianus I went very still, with both hands on the back of a chair he did not sit in.

"We shall not know," he said. "We shall know what happened. That is not the same as knowing what the opposite cost, because the opposite did not happen, and what does not happen is not measured."

Krell had no answer to that, and the reason he had no answer was that it was true, and that it was, besides, exactly the same thing he himself said to bishops who asked him for funds for works that would prevent future disasters.

"You are a man of numbers and you will spend your whole life making this subtraction," said the One, already leaving. "Make it well. Somebody in this house has to make it well."

There was, between July and September of 1821, a phenomenon in Aurum Roma that has no name and that Krell described in the Memorial with a sentence the historians of the Order always quote without the source:

The city prayed for them with a generosity it would not have had if we had relieved them.

It was true. It was ostentatiously true and everybody took part.

There was a procession of rogation on the 14th of July, with thirty-one thousand people on the Via of the Choir, which was more people than had come to the coronation of Aurelianus I. There were daily masses in twenty-two parishes for those who crossed the line. There was, in the Cathedral of the Hashmal — which was then four walls and a scaffold and a provisional altar of planks — a vigil of forty hours in August, with canons in relay, and Krell went there at three in the morning of a Saturday and found the nave full.

And he knelt with the rest, and prayed, and prayed in earnest, because he was a man of genuine faith and it must be said because everything else in this book may make it appear that he was not.

And while he prayed, with his knees on the beaten earth floor of an unfinished cathedral, he thought — and could not help thinking it, and never forgave himself the thought, and confessed it twice to two different confessors and neither of the two understood what he was confessing:

If we had given them forty men, not one of these thirty-one thousand people would be here, because there would be nothing to pray for. We refused, and the refusal produced the greatest public demonstration of faith this city has seen in forty years. And in two months, if that goes badly, the same refusal will produce the greatest transfer of authority in the history of the See.

And I did not make this happen. And I shall profit by it. And there is not one line in canon law that accuses me of anything.

VII. The News

The first news came on the 9th of September 1821, by a courier out of Vykhradia, and it was not news: it was an absence of news, conveyed with the nervous precision of a man who understands that he is saying something grave.

It said that the host had sent no dispatch since the 22nd of July.

Forty-nine days.

Krell took the paper to the Secretary of State and the Secretary of State took it to the One, and that afternoon the Curia did the only thing it could do, which was to send men to look — two canons and an escort of twelve, mounted, by Vado Aureo, with instructions not to cross the line.

They left on the 11th of September. They came back on the 3rd of October.

Between those two dates, Aurum Roma learned it from the merchants, as always, because merchants are faster than nunciatures and do not have to write carefully.

The first version that went round the city was that the host had been broken in a battle. It lasted four days and was false.

The second version was that there had been treachery by one of the nine houses. It lasted three weeks, and there are still people who defend it, and the canon never closed it.

The third version was the true one and arrived on the 26th of September, in a third-hand account that had come from a forwarding agent of Solnica who had it from a shepherd of the frontier, and Krell heard it in the antechamber, standing, from a monsignor who was telling it to two others.

The host had gone in. It had fought for eleven days. On the twelfth day, the consecrated ammunition had run out.

And after that — and this is the part that made that version instantly recognised as true by everybody who knew anything, because it is a thing nobody invents — they did not withdraw.

They did not withdraw because they were four days' march from the line, with wounded, in ground they did not know, and because to withdraw with dead metal is the same thing as to stay with dead metal only walking.

They held on another nineteen days.

Nine thousand men with Choir-Iron swords that still served, because Choir-Iron does not go out in the same way — a blade blessed in 1819 cuts in 1821 — and with four hundred and ninety-four thousand cartridges that were, from the twelfth day, brass and silver with nothing in them.

Nineteen days at close quarters.

Krell left the antechamber and went to his office and shut the door and sat down.

And the thing that happened to him there, and which he recorded in the Memorial of 1831 with an almost clinical precision because it was the only way to bear it, was that his head did the sum by itself.

He did not want to make it. He made it.

Eleven days of ammunition. Forty-three purifiers would have sustained forty-five thousand a day. Thirty days of campaign required a million three hundred and fifty thousand cartridges, and they had taken four hundred and ninety-four.

With forty-three men of Ferrum Alpina on the march, that host had ammunition for thirty days.

It did not have victory. Krell did not deceive himself about that and no competent man deceived himself: nine thousand men in the Limbus Mundi take nothing, with or without bullets, because there is nothing to take and because the problem of the Limbo was never fire, it was ground. They would have gone in, marched thirty days, understood that there was no objective at all, and come back.

Come back.

Nine thousand men coming back, humiliated, without victory, with heavy losses from sickness and skirmish — and alive. And the nine houses would have been left with the certainty that it could be done without Rome, and with the proof of it, and with nine thousand veterans.

Krell sat at that table and understood, with a clarity that turned his stomach, that the refusal of the 19th of April 1820 had been, from the strict point of view of reason of state, the most effective decision taken in that palace in the century.

And that nobody had taken it for that reason.

Nine cardinals had voted against out of a sincere theological reason, expressed in a beautiful sentence about ministry and merchandise. He himself had voted in favour. The One had confirmed the majority. Nobody, at any moment, in any room, had said if we refuse them, they die, and if they die we get everything.

Nobody had said it. He would have known — he read all the minutes.

And yet there it was: exactly that happened, and the See of Aurum Roma was going to receive, in November 1821, an authority over the continent it had failed to build in six years of decrees, and was going to receive it for having done nothing, and was going to receive it with its hands clean of any intention and dirty with every consequence.

Krell wrote, that day, on a loose sheet he put into no file and that was found among his papers in 1841:

We did not sin. We inherited.

Let God tell me which of the two is the worse to carry, because I cannot decide.

The two canons came back on the 3rd of October and their report is four pages and is dry.

They had not crossed the line. They had spent eleven days at Vado Aureo questioning whoever came through, which was few people, and the third page contains the only new information they brought and that the Curia classified for thirty years:

That on the 14th of September there had been sighted, on this side, two men.

Two, out of nine thousand. They came on foot, unarmed, and one of them did not speak. The other spoke for six hours and then died the following day, and what he said is on pages three and four, and it is from that that everything the continent knows about the thirty days comes, and it is little, and it is contradictory, and a man who walked thirty days in that and six hours in talk is not a source to be handled severely.

The man was called Anselm and was an assistant farrier of the house of Reichshamn and was twenty-two, and the first thing he asked when they sat him down was whether there was bread.

They gave him bread. He ate it all and asked for more and could not eat the second.

Then he talked for six hours.

Deposition taken at Vado Aureo, 15 September 1821, by Canon Bertold, commissary of the See. Full record, without interruption of the deponent.

You ask me how many days. I say thirty-one and the other says twenty-nine and neither of us is right because after a certain point there is no day.

It is not dark. That is what everyone asks first and it is not that. There is light, there is an orange thing above, and it does not move. It does not go from one side to the other like the sun. It stays.

The first week was good. I say good and you gentlemen will write that down and think me mad, but it was good. We marched four days without seeing anything, and on the fifth there was contact and we won, and on the sixth there was another and we won, and on the seventh the men were singing. I saw nine thousand men singing inside that.

On the seventh day the Count ordered a halt and a council, because we had gone in four days and there was nothing. Nobody had explained to the men what it was we were going to take. I am a farrier, I am nobody, but I went about with the officers' horses and I heard, and on the seventh day I heard a captain of Velinia ask his captain which was the city.

There is no city, gentlemen. That is the thing. You march and there is nowhere to go.

On the tenth day the quartermaster came to the Count. I was ten paces off with two horses.

He said we were spending more than had been reckoned. The Count asked how much there was. He said the figure, and I do not know the figure, but I know the Count repeated it aloud and then was silent a long time.

And then the Count asked a thing that I understood was the wrong thing to ask, and I was twenty-two and a farrier and I understood, and so they understood too.

He asked the quartermaster whether the canon could not do more.

The canon was called Bonifácio and came from Vallia and was, I don't know, fifty. He rode a white horse because it was the horse they had given him and he thought it embarrassing and said so several times.

He was a good man. I say it because you will ask.

They brought him and explained it to him, and he listened, and then he said to the Count a thing I did not understand at the time and understood afterwards, which was:

"My lord Count, I make three hundred a day if I do nothing else. How many are spent in a day of contact?"

And the quartermaster said the figure.

And Canon Bonifácio sat down on the ground. A priest of fifty sat down on the ground in front of a count and four captains and stayed sitting there.

Nobody lifted him up.

From the eleventh day he did nothing else but that. They put a tent up for him in the centre of the square with two men at the door and he stayed inside it from first light until he could not.

I took him water four times because I was told to.

The first time he was working and he spoke to me. He asked me what my name was and where I was from and told me he had a niece in Reichshamn.

The second time he did not speak.

The third time he took hold of my arm and asked me whether the men were firing a great deal. I said yes, because it was true. He said: "Tell them not to fire." I said I was nobody to say that to anybody. He said: "I know. Forgive me."

The fourth time was on the fifth day and he was no longer speaking and his hand no longer closed, and I saw him, gentlemen, I saw him take a cartridge in both hands like a man taking up a chalice, and pray over it, and put it in the box on the right.

And I stood looking, and he saw me looking.

He knew.

This is what I came to say and afterwards you may write what you like. Canon Bonifácio knew that what he was doing on the fifth day no longer served. I saw it in his face. It is not a thing one can fail to understand.

And he went on.

And I came out of the tent and went to the sergeant and told him, and the sergeant took me to the captain, and the captain heard me out and then told me I was a farrier.

I said yes, I was a farrier.

And he said: "Then you know nothing of blessings." And sent me away.

And I spent three days thinking he was right, gentlemen, three whole days, because I was twenty-two and a farrier and who am I to know.

Only on the third day there was contact and I was in the middle of it, and I saw a rank of twelve men firing at a thing twenty paces off, and I saw the thing reach the twelve.

After that it is as I said. There is no day.

There were the swords. The swords served. Everyone thinks it is done with firearms and it is not, gentlemen, it is done with a man three paces from another thing with an iron in his hands, and the irons served, and that is why we held what we held.

We held nineteen days with swords.

Nine thousand men. Then seven. Then I do not know.

The Count of Reichshamn died on the twenty-second or the twenty-third and died on foot because there were no horses left, and died at the front and not behind, and I want that written down because in Reichshamn they will say otherwise.

Canon Bonifácio died before, I do not know when. Somebody said it was in the tent and that they did not kill him, that he simply did not wake.

I hope that is true.

You ask me how I came out.

I walked. I walked with Kaspar, who is the other one, and Kaspar has not spoken since the tenth day and I do not think he will speak again.

We walked seven days or nine. We ate what there was and I will not say what there was.

You ask whether we knew where we were going. We went west, because the orange light is always a little higher to the east and a little lower to the west, and it was Kaspar who understood that on the third day and it was the last thing he said to me.

One thing more and then I have no more.

On the twelfth day, when the ammunition ran out for good, the Count ordered the square formed and spoke to the men. I heard because I was close.

He told them the truth. He said the ammunition was gone and that we would come out by iron and that whoever wished to pray should pray.

And then he said a thing I did not expect and that nobody expected, and there was a great silence, and it was:

"And do not speak ill of Rome. Rome told us no. It was we who came."

Nine thousand men, gentlemen. He could have said the opposite and they would have believed him and would have died in anger instead of dying as they died.

He did not say it.

Canon Bertold wrote, at the end of the thirteen sheets, a commissary's note:

The deponent died at five in the morning of the 16th of September, of a cause the physician declares to be exhaustion and want of nourishment. He did not receive the sacraments, being no longer conscious when the chaplain was called, which is lamented and is recorded.

The second survivor, by name Kaspar, did not speak and continues not to speak. He goes to a house of mercy at Vallium.

Krell read the thirteen sheets three times and it was at the last that he understood which of them would stay with him.

It was not the canon praying with both hands. It was not there is no day. It was not even the Count telling nine thousand men that the fault was not Rome's, though that sentence ran through the whole Curia in forty-eight hours and earned Ottokar of Reichshamn, from a cardinal who had never known him, the epitaph that stands today in the basilica of Reichshamn.

It was the captain.

Then you know nothing of blessings.

Krell wrote, in the margin of sheet nine:

That captain was not lying and was not being cruel. It was what he knew. It was what we had taught him — we, for thirty years, from the pulpit: that the thing is a blessing, and that a blessing is not measured, and that to ask how many is matter of little faith.

A farrier of twenty-two, who saw the canon on Thursdays, understood in five days what our catechesis spent thirty years making unthinkable for a captain of horse.

And when he said it, the captain answered him with our sentence.

We loaded that sentence. It was primed and we put it in his mouth.

And there was Canon Bonifácio's fifth day.

Nobody told him. They let him do it, and they issued what he made, because the alternative was telling nine thousand men that there was no more.

Krell read that passage four times.

Then he went to the archive and fetched the census form of House Solar, of December 1819, with the upright small writing in the field for the name where it said none, and laid it on top of the report, and sat looking at the two papers side by side.

Eleven cartridges, by an unordained officer, by an unlicensed method.

Eleven that worked.

Against four days of a licensed canon making dead silver because he no longer had hands and nobody had the courage to tell him.

VIII. The September Shift

The Vigil knew nothing, because the Vigil never knows anything: it is a closed house with fourteen flames and a chair, and the news of the world comes in three weeks late and in diminished form, because the Mother Vigilant of the day understood — and all of them understood, before her and after — that a Blanca who knows what is happening outside listens worse.

Aelius knew. Aelius went out on Sundays and had a brother in the city.

And that is why the transcription of the 30th of September 1821 is the hardest document of his career, and the only one about which he lied.

The shift was held by the Seventh Blanca, nineteen, in her second year, a girl from Coronata who had entered at seventeen and who would die in 1826, which was about the term.

The shift began at a quarter past six. It ran forty minutes of nothing.

Then there was inscription, and the inscription lasted fourteen seconds, and the words were nine, and Aelius wrote them as he wrote everything, unpunctuated, with the flame outside his field of vision.

The nine words were:

it was not refusal it was a sum and the sum was right

Aelius set down the pen.

A clerk of the Vigil does not set down the pen during a shift. It is in the rule and it is in the training and it is in the head of a man who has done it for thirty-one years the way breathing is.

He set it down.

And he sat looking at the nine words on the sheet, in his own small upright hand, and the first thing he thought was a thought of the trade and not of theology, and it was:

This is the language of the city. This is a sentence one hears in a tavern on the Via of the Choir in September 1821.

Because it was. It had not one technical term. It had no metallurgy, no proportion, no name of a thing that a girl of nineteen from Coronata could not have heard in the street if she had been in the street, which she had not been for two years.

And it had — and this was what was keeping Aelius de Sárvar's hand from going back to the pen — the exact structure of an accusation.

The Mother Vigilant of that year read the sheet and asked him what he thought.

A clerk does not think. To ask a clerk what he thinks is so gross a violation of protocol that the Mother committed it on purpose, and Aelius understood that it was on purpose.

He said:

"Mother, this may be contamination."

"Explain."

And Aelius explained, and explained it well, because he had thirty-one years in the trade and because contamination was the central technical problem of the whole business and nobody spoke of it aloud: a Blanca is not an instrument. She is a girl of nineteen inside a house into which come canons, servants, laundresses, the physician, the sister who brings the food. What comes out of that chair comes out through her mouth, with her tongue, with the vocabulary she has. If she heard a sentence in the kitchen on Tuesday, that sentence is in her head on Thursday.

And if what the Hashmal does is use the vocabulary it finds, then there is no way at all, none, of distinguishing a true inscription from a kitchen sentence — except when the inscription says a thing the girl could not know.

A metallurgical alloy she could not know. That was why alloys were the gold standard.

It was not refusal it was a sum and the sum was right — that she could have had from a laundress.

"So you think it false," said the Mother.

"I think it is not verifiable, Mother, and that is a different thing, and it is the worse of the two."

The Mother Vigilant had to decide what to do with it, and what she did is what anybody in that place would have done and is the reason there exists today a specific article about it in the seven rules.

She ordered it not transmitted.

The transcription stayed in the house. It did not go up to the Curia. It did not enter the weekly abstract that went to the Cardinal-Electors. It stayed in the September 1821 bundle of the White Vigil, with the Mother's note in the margin — inscription of current vocabulary, not verifiable, retained — and that is why no cardinal, no Pontiff and no inquisitor ever knew that, thirteen days after the account of the two survivors reached Aurum Roma, a girl of nineteen who could know nothing said nine words describing exactly what Amandus Krell had written on a loose sheet in his office, alone, with the door shut.

Aelius de Sárvar copied the sheet before filing it.

It was the only time in fifty-one years that he copied a transcription for himself, and it is a grave disciplinary matter, and he did it, and kept the copy in a Bible, and it stayed in that Bible another twenty years.

The reason he kept it was not theological and was not curiosity.

It was this, and he wrote it on the back of the sheet, and the sheet exists because in 1841 his great-nephew found the Bible:

If this is from Heaven, then Heaven commented on an administrative decision of the Curia and said that it was right, and we retained the comment because we could not verify it.

If this is not from Heaven, then a girl of nineteen, shut up two years in a house without newspapers, arrived by herself at the same conclusion the nine best-informed cardinals on the continent could not bring themselves to formulate aloud in a room.

I do not know which of the two possibilities frightens me more, and I have worked at this for thirty-one years.

IX. What One Does With Nine Thousand

The Curia did nothing for seven weeks, and it was the cleverest thing it did in the whole pontificate.

Krell watched it happen from inside and understood the mechanism, and the mechanism was not cunning: it was paralysis, and the paralysis looked, from outside, like tact.

Nobody knew what to say. A host of nine thousand men raised by nine Christian houses against Hell had been annihilated, and Rome had refused them chaplaincy seven months earlier in writing, and the letter of refusal was in circulation across the whole continent in pamphlet copy. Any declaration of the See in those weeks would be read as a defence.

So the See did not declare.

It ordered masses said. It ordered the bells of Aurum Roma rung for one hour on the 9th of October, which was decided in a meeting of forty minutes in which it was debated whether one hour was too little or too much and in which nobody wished to be the first to say too much.

And it waited.

The answer came from the houses, and it came broken, and it was that fracture that decided everything.

Of the nine signatory houses, four were extinct in the direct male line by October 1821. A fifth had an heir of eleven under wardship. The other four had living heads because the heads had not gone — and this must be said, because the pulpit spent seventy years making of the Nine Standards a poem about nine heroic counts, and the documentary truth is that six of the nine signatories were not at Marnheim. They sent sons, they sent brothers, they sent captains.

Ottokar of Reichshamn went. That is why his is the name that stayed, and it is a sufficient reason, and there are very few things in this universe in which fame and merit coincide so exactly.

The six who stayed spent the winter of 1821 accusing one another.

They accused one another of everything there was to accuse: of contingents promised and not sent, of money not delivered, of armament diverted, and — the most repeated accusation and the only one that mattered — of supply. Each house alleged that another had been charged with the second consignment. The second consignment existed in no contract. Krell had it verified, because the Congregation of Means was, by an administrative accident, the body that fell to gather the documents.

It did not exist. It had never existed.

Nine houses had sent nine thousand men into the Limbus Mundi with a single consignment, and not one of them had asked the other eight what happened on the twelfth day, because each assumed somebody would have thought of it.

Krell wrote the final opinion on the supply of Marnheim in January 1823, and it was he who drafted, with his own hands, the two lines of page thirteen that the whole continent learned by heart:

The expedition departed without field chaplaincy possessing faculty and provision for purification on the march, against custom and against counsel, and it is in this want that the Commission fixes the proximate cause of the disaster.

He knew it was insufficient when he wrote it. He wrote it all the same, and spent the rest of his life looking at that sentence, and in the Memorial of 1831 devoted eleven pages to it.

His argument, in 1831, was this:

That he had before him, in January 1823, three truths, and that only two of them fitted in a public document.

The first truth was that they had set out without chaplaincy. True, simple, publishable.

The second was that the chaplaincy they needed was forty men and not one, and that therefore the problem was not dogmatic but industrial, and that anybody with that figure in his head would immediately understand that the See's monopoly was not a privilege, it was a factory.

The third was that Rome had refused the forty and could have given them.

To publish the second obliged him to publish the third.

And to publish the third, in 1823, with the whole continent reading pamphlets, would have been — Krell wrote this and then underlined it, and the underlining is in the manuscript — to hand the secular crowns, in writing and under the seal of the See, the exact information of how much they lacked in order to dispense with us.

Forty men. It is a small number. A king can find forty men.

He could not have trained them — but he could have bought them, and the Curia had at that time no mechanism whatever to prevent a licensed man of Ferrumkar from accepting three hundred ducats a year from an alpine prince, because exclusivity of service for licensed purifiers was only instituted in 1824, and was instituted because of this, and Krell drafted it.

It was for that reason, and not out of cruelty nor out of cynicism, that page thirteen says without chaplaincy in the singular.

And it was for that reason that in 1890, sixty-seven years later, some bishop in a provincial pulpit could still say they bought the bullet and thought they had bought the grace and the whole congregation would nod — because the sentence is true, and is beautiful, and hides exactly the information the See decided, in a room, in January 1823, should not circulate.

Krell died convinced that he had done well and that he had done a bad thing, and did not consider the two propositions incompatible, and wrote it thus:

A man of means spends his life choosing between two arithmetical evils. My misfortune was not to have chosen badly. It was to have had, of the three truths, the only one that did not fit, and to have been the one who decided it did not fit, and to have been right.

The second movement is of November 1821 and lasts forty minutes.

Aurelianus I summoned eleven cardinals and did not summon Krell, which was noticed, and sent for him halfway through, which was noticed a great deal more.

When Krell came in, the room was silent and there was a map on the table.

The One said to him:

"Cardinal. What does it cost to endow the Orders?"

And Krell — who had had that figure in his head since February 1816, revised fifteen times, and had said it in that house twice and been politely ignored on both occasions — replied:

"Forty-one per cent of known revenue, Most Holy Father. Today, with the census closed and the revenue of 1820 established: thirty-eight."

"And if an impost were charged on mineral purified outside obedience?"

"Twenty-nine."

"And if licensed purifiers were forbidden to serve outside the See?"

Krell stopped.

Because the second question was his own proposal of 1816 and the third was not: the third was new, and it was brutal, and it had not occurred to him, and he understood in an instant that it had been thought of by somebody in that room in the last thirty days.

"That does not lower the cost, Most Holy Father," he said slowly. "That changes something else. If a licensed man cannot serve outside the See, then all the purified mineral of the continent passes through us, and its price ceases to be a market price."

"And becomes what?"

"It becomes whatever we say."

Silence.

"With that, what does it cost?" said the One.

Krell did the sum in his head, there, before eleven cardinals, and it took him perhaps eighteen seconds, and the answer was:

"It does not cost, Most Holy Father. It yields."

The winter of 1821 into 1822, at the quarters of House Solar, was the worst of the seven Mattias spent there, and it was not for want of barley.

The news of Marnheim came in September, by the merchant, and came in a short and mistaken version — the host was broken — and the true version only came in November, in a pamphlet printed in Bernkanton that somebody brought up from the valley and that went from hand to hand for three weeks until it fell apart.

The pamphlet had eight pages and the fourth carried, in capitals, the full text of the See's refusal of the 19th of April 1820.

The seventy-one words.

Mattias read them in the refectory, one night, aloud, because four men asked him to read and because refusing would have been worse.

Field chaplaincy with faculty and provision of purification on the march is ministry, and not merchandise; it is not granted upon request, nor apportioned to any host not raised under the authority of the One.

He finished reading. He set the pamphlet down.

And a man at the far end of the table — nobody knows which, and Mattias, who remembered everything, always insisted that he did not remember which, which is the one documented lie of his life — said:

"And what are we?"

That was the question.

Sixty-two men in a disused monastery, who had sworn three vows to an Order that existed in a decree of fourteen articles, who kept themselves by hiring out escorts at twelve ducats, and to whom the same See that had not given forty purifiers to nine thousand Christians had also not given, in six years, a sack of wheat.

A host raised under the authority of the One. They were that. That was exactly what they were, on paper, and had been for six years.

And they had received nothing.

Mattias had forty-two men under thirty in that room, and he understood that night, with a clarity that frightened him, that he was eleven words away from a mutiny of conscience — not of violence, something worse: of dissolution. Men who go away one by one over the winter because they have understood that they are guarding a thing that does not exist.

He made no speech. He was not capable of it and knew he was not.

He did something else, and it was the decision that founded the internal culture of the Knights of the Seal, and he took it in something like four seconds, standing, with the pamphlet in his hand:

"Tomorrow at six I begin teaching you to wake ammunition."

It took them the whole winter and he taught it to sixty-one men, one by one, by the method of the coverless diary.

It was illegal from the first day to the last. The opinion of 1823 that made it tolerable did not exist; the Vow of Measure did not exist; the doctrine did not exist. What existed was the Canon of 1791 and the words functional heresy, and Mattias knew them and knew that each of the sixty-one knew them.

Of the sixty-one, eleven managed it.

Eleven in five months. The other fifty did not — and not because they had less faith, and this is what makes that winter the most interesting thing that ever happened in the Order before 1836, and is the reason Mattias wrote four pages about it in 1824 and gave them to the canonists and the canonists ignored them.

There were among the fifty men of evident faith. There were two who fasted without being asked. There was one who prayed four hours a day and whom Mattias considered, without hesitation, the best man in the house.

None of the three managed it.

And of the eleven who did, one was Halvard of Bierkut, who blasphemed with an inventiveness that became famous in three valleys, and another was a former deserter from Vykhradia who confessed twice a year out of obligation and never out of desire.

Mattias wrote, in the four pages of 1824:

I find no pattern.

I ordered the sixty-one by everything it occurred to me to order them by: age, years in the house, frequency of sacrament, conduct, education, health, temperament. None of the orderings predicts anything.

The only thing that seems to me to separate the eleven from the fifty is that the eleven, at the third or fourth attempt, stopped trying to convince themselves that it was going to work.

I cannot write this better. The fifty were straining. The eleven, at a certain point, stopped straining and came simply to be present, and that was where it happened.

This is not doctrine and I do not propose it as doctrine. It is field observation by a man with no theological training and it is probably wrong.

But if it is right, then what we call faith, for this purpose, is not the effort. It is what is left when the effort stops.

The canonists did not use those four pages, and wrote the Vow of Measure without them, and the four pages stayed in the file and nobody read them for thirty-one years.

In 1855, a canon of the Curia of the Hashmal found them while investigating something else, and the note he wrote across the top, in pencil, is there:

This ought to have been published in 1824. How many things like this are in this archive.

He did not publish it.

The third movement is of February 1822 and takes place in a disused monastery eleven kilometres from Karentin.

A dispatch came. Good paper, sealed, addressed to the Command of House Solar, like the one of 1818 — but this one came with an annex, and the annex had twenty-two sheets.

Mattias read it in the chapter room, standing, with four men watching him and with no idea why he was taking so long.

It was the endowment.

Pay, by scale, with grades. Ration. Clothing. An annual sum for armament, with the figure written in numerals and in words. A sum for forage. Instruction that House Solar should receive, by June, four licensed purifiers in permanent service, their expenses borne by the Congregation of Means.

And, on sheet nineteen, a thing Mattias read three times and did not understand at first because he was not expecting it:

It is determined that the members of the Clerical Military Orders shall give up the family name and adopt that of the House in which they serve, being unable to bequeath rank or to transmit preference of entry to any descendant.

There was a footnote. The footnote said:

In accordance with usage already observed in certain servants since 1809.

Mattias of House Solar held that sheet in his hand for a while.

Then he folded the twenty-two sheets and ordered the muster sounded, and read the endowment aloud to the sixty-two men in the yard, and read it entire, including the figures, because they were men who had spent five winters knowing exactly how many hundredweight of barley there were.

There was no cheering. That House did not have the habit and never acquired it.

One of the men, the oldest, who was called Ruprecht and who had entered in 1811 and who would die at Karentin-Alta in 1836 at sixty-four, asked:

"Captain, sir. Why did they send this now?"

And Mattias — who knew perfectly well why, who had known since September, and who would carry it for the next thirty-one years with a sort of background nausea that never quite became guilt because there was nothing to be guilty of — told the truth, because he did not know how to say anything else:

"Because nine thousand men died last summer, Ruprecht, and we were not there."

He was silent a moment.

"And because I refused a contract in June of 1820 for a reason that seemed small to me at the time."

Then he dismissed the muster, and went inside, and signed the receipt for the endowment, and the Order of the Knights of the Seal began to exist in the material world at twenty past eleven on a February morning in 1822, in a disused monastery, with sixty-two men and a man with no surname signing a paper.

ACT III — THE VACANCY

X. The Three Houses

Between 1822 and 1824 the Curia did, in twenty-six months, more than it had done in the previous seven years, and Krell — who was at the centre of nearly all of it and who signed a considerable part — was never mistaken about the cause.

It had been no conversion. There had been no vision, no courageous decision, no strong man imposing his will on a reluctant house.

There had been a vacancy.

Nine secular houses had occupied, for a hundred and fifty years, the political space in which a Military Order might exist. Four were extinct. Two were under wardship. The three remaining spent the winter suing one another in three different courts, and not one of them had, in 1822, a single public argument available against the See, because any argument they made would be answered with a question of nine thousand syllables.

The space opened. The Curia walked in.

Krell wrote in the Memorial: It was not we who grew. It was the hole that came to us, and we had the merit of filling it quickly, which is a real merit and is the only one I claim.

The House of Purification of Aurum Roma opened on the 1st of September 1822 with forty-one pupils and three masters, in a building of the fifth courtyard that had been a granary until the year before.

Krell was at the opening, which was not required, and stood at the back.

It was the thing he was proudest of and it must be said carefully because he never said it aloud to anybody himself: of everything he did in twenty-six years of the Congregation, what he thought was worth something were the three houses of purification, and they were worth it for a reason that had nothing to do with output.

They had a programme.

Until 1822, a purifier learned from an older purifier, in the way a smith learns from a smith. He went into the sacristy with the parish canon, watched him do it, did it, was corrected, and after three or seven or eleven years did it alone — and what he learned was exactly what that man knew, with that man's errors, and when that man died there died with him everything he had discovered and not taught.

Krell had put, in the census of 1819, a supplementary question nobody had told him to ask: with whom did you learn?

And then spent six months drawing the tree.

Three thousand and forty-one purifiers on the continent descended, in line of teaching, from two hundred and eleven men alive in 1800. Two hundred and eleven masters. And of those, fourteen had trained more than thirty pupils each, and the pupils of those fourteen yielded, on average, nineteen per cent more than the general average.

Nineteen per cent.

It was not gift. It could not be gift — gift is not inherited down a line of teaching, or else it is not gift. It was method. There were fourteen men on the continent who knew how to do it better and knew how to teach it, and nobody had ever asked what it was they did differently, because the whole business was classified as grace and one does not conduct productivity inquiries into grace.

Krell asked.

He sent two canons to interview the fourteen. Nine were still alive. The report runs to a hundred and ninety pages and is the basis of the programmes of the three houses, and the principal conclusion fits in three lines and is, theologically speaking, a scandal nobody ever wished to discuss:

The most productive masters do not teach a man to pray better. They teach him to stop sooner.

The average purifier works until his hands shake and produces dead stone in the last hour without knowing it. The disciples of the fourteen stop before that, rest, and produce less per day and more per year, and live four years longer.

It was from this that there came, in 1824, the counting of the seconds — the practice, ridiculed for a generation and today universal, of a purifier raising his right hand before each shift and counting how many seconds it takes for the tremor to appear, and recording the number.

It does not measure faith. It measures nothing metaphysical. It measures a hand.

And it is, in 1890, the only thing in the whole liturgical system of the Clerical Order that produces a comparable time series across seventy years, and it is because of it that the Hospitallers of the Hashmal were able to establish, in 1881, that the curve was falling.

A man who had been a tenant farmer in Vallia Sacra put, in 1824, seven seconds between the Church and the lie it would want to tell itself sixty years later.

The Hospitallers of the Hashmal were formalised on the 4th of May 1823, and Krell had almost no part in it, and it is the chapter he most liked to talk about.

They were not founded. The opposite: they had existed since the 1790s, in a network, with nobody's authorisation, as houses of mercy set up by whoever happened to be there, and they had gone through the Ash treating fourteen million people who died of hunger and typhus and cold — more than the Cataclysm killed, more than the fighting, more than everything.

What the Curia did in 1823 was recognise them, and Krell had to negotiate that, and the negotiation was the most disagreeable of his career because across the table sat a woman of sixty-six named Sister Ottilie who directed nineteen houses in three States and who wanted nothing.

She wanted no grant. She wanted no habit. She wanted no status.

"We already do it," she said. "Your Eminence comes to offer us authorisation for a thing we have been doing for thirty years without it, and then you will want something in return, and I am trying to work out what before you tell me."

Krell, who liked her immediately and said so in the Memorial, answered with the truth:

"I want the books."

"What books?"

"All of them." Krell opened his hands on the table. "You write things down. I know you do; I have seen three of your houses and in all three there was a notebook. You write down who comes in, when, with what, and what happened. You have been writing it down for thirty years. Sister, there exists on the whole continent no other continuous series of records about what kills people, and the See needs it, and the See does not have it, and you have it in a dozen cupboards in Coronata."

Sister Ottilie looked at him for a long time.

"You want to count the dead."

"I want to count the dead."

"Why?"

And Krell gave the answer that came to him, which he had not prepared, and which became — by routes nobody designed — the Vow of Record of the Hospitallers of the Hashmal:

"Because until they are counted, Sister, every one of them died by the will of God. Once they are counted, some of them died by our fault. It is the difference between a religion and a government, and we are both those things and until today we have only done one of them."

It was signed in May. The Order kept almost total autonomy, which seemed to everybody at the time an excessive concession by Krell and which was, in truth, the price he paid without hesitating.

The article he wanted was at the end and had nine words:

Everything is written down. Also what does not suit. Especially.

And the article Sister Ottilie wanted, and which Krell accepted in four seconds, and which is today the Vow of Indistinct Treatment and the permanent scandal of the Order, was shorter still:

One treats whoever comes.

XI. The Doctrine

The written doctrine of the Knights of the Seal was drafted between October 1823 and March 1824, in a room in Aurum Roma, by seven men, and Mattias was one of them and was the youngest and was the only one who had been in the field in the last five years.

This is worse than it sounds. Of the other six, three were canonists, two were retired secular officers of sixty-odd who had fought the Veheran war and were said to understand order of battle, and one was a canon of the Congregation of Means who was there to ensure that nothing written should cost more than the endowment allowed.

Mattias brought the notebook.

Not the whole notebook — he brought six sheets copied out by hand, without the rest, because the rest of the notebook held five years of barley accounts and two pages about the death of a lad from Bierkut and he was not going to put that on a table of canonists.

On the six sheets were four things.

The first they accepted without discussion in twenty minutes, because it was militarily obvious and because the two retired officers recognised it at once: never to operate in a unit larger than a regiment.

Mattias had written it in 1821, on learning of Marnheim, and his reasoning was simple and was not about the Limbo: an Order of four thousand men that loses a whole corps loses the Order. A crown that loses nine thousand men loses nine thousand men and raises others the next year. The Order cannot raise others because it takes eleven years to make a knight and because it has no province to draw them from in bulk.

Our strength is the only one that is not replaced. Therefore let us never expose it entire.

It passed unanimously and is today the first rule, and everybody supposes it came from the Massacre of the Iron Line of 1848, and it did not; in 1848 it was confirmed.

The second was harder and Mattias lost it and won it twice: never to defend a position one did not choose.

The two retired officers hated it, and with reason, because it is an insubordinate rule: it means a captain of Knights may hand a position to line troops and reposition without asking leave of any marshal, and no army in the world works like that.

Mattias's argument, which he repeated four times and which ended by winning not through force but through exhaustion, was this:

"A line regiment that stays where it should not loses men. A regiment of ours that stays where it should not loses men and the purifiers. And the purifiers are not ours, gentlemen. They belong to the See, they are four to a House, and if I lose my four in a position a secular marshal chose, House Solar ceases to be able to fight for two years, and the four I lost came from a house of formation where there are forty-one pupils for the whole continent."

The canon of the Congregation of Means lifted his head at that point and it was the only time he spoke in five months.

He said:

"That is true."

And it was for that reason that it passed.

The third was the one that mattered and was the one that nearly did not pass.

A unit without a purifier in condition withdraws, even when winning.

The canonists had no difficulty at all. The retired officers had an enormous difficulty, and the elder — a man named Hedrich, seventy-one, who had commanded a flank in 1798 and who was, so far as Mattias could make out in five months, a good man and a good soldier — said the thing that had to be said:

"You are writing into a rule that your men abandon a victory."

"I am."

"It will kill the Order, Captain-General. Troops who have a legitimate order of withdrawal written into the regulations use it. Always. In every army I have knowledge of, and they are many."

"I know," said Mattias.

"Then why do you write it?"

And Mattias told them about the shepherd's hut.

He told it whole, which was the second time he had told it to anybody: the winter, the coverless diary, the eleven days, the four hours a cartridge, the lad from Bierkut, the five shots, the four that went through the thing like stones through smoke.

And then he said the part they had not grasped, and which is the reason the third rule exists and is not what it appears to be:

"Gentlemen, you are reading that as an order of withdrawal. It is not. It is an order to keep the purifier alive. If I write that a unit without a purifier withdraws, then no captain of mine will let his purifier go forward, nor put him on a flank, nor let him sleep badly, nor spend him on a fortnight of skirmishing that is worth nothing. The purifier becomes the best-guarded thing in the regiment, and that is what he has to be, because without him we are infantry with expensive spears."

Silence.

"And there is the other half," said Mattias. "If a captain of mine is left without a purifier and does not withdraw, his men go on firing. They do not know the ammunition is dead. Nobody knows: it does not change colour. They will find out on the third volley, at twenty paces."

Old Hedrich sat looking at the table.

"At Marnheim," he said after a while, "they let the canon work four days without sleeping and issued what he made on the fifth."

"They did."

"Because they had not the courage to tell the men there was no more."

"Yes."

Hedrich signed the third rule that afternoon and did not argue anything again in five months.

The fourth thing on the six sheets was not a rule and Mattias did not present it as a rule, and that is why it got in.

It was a paragraph about what to do when there is no purifier at all and no withdrawal possible.

It was written in the language of a man who does not know how to write documents, and the canonists rewrote it three times, and the final version came out worse to read and better to defend, and is called today the Vow of Measure:

Every Knight of the Seal shall learn to wake his own ammunition, by the lesser method, aware that he does it badly, slowly and in a quantity that sustains no unit whatever; and he shall not do it save in extremity.

The canonists took six weeks to pass that, and what held them was not the doctrine: it was the Canon of 1791, which reserves the rite to ordained clergy and which classifies purification by an unordained man as functional heresy.

An opinion was required. The opinion took eleven weeks and is a piece of casuistry of which Krell used to say, with sincere admiration, that it was the best thing the Curia produced in that decade — and what it did was to distinguish, across forty pages, sacramental purification from waking in extremity: the first reserved, public, productive, which is the Church's; the second private, inefficient, not transmissible, not teachable outside the Order, and valid only where the alternative was the death of Christians.

It is sophistry. Krell wrote it in the Memorial without evasion: it is sophistry, and it is the most useful sophistry this house ever bought, and I paid for the dinner of the man who wrote it.

Because the opinion establishes, in small print and so as to be able to deny it in large, the proposition the Reformed Inquisition would spend seventy years burying:

That it works without a priest.

Slowly. Badly. In a ridiculous quantity. But it works, and it is written, and it is signed, and it is in the Archive.

Mattias signed the doctrine on the 9th of March 1824, and that night, at the lodging where he was quartered, he sat up late with a sheet in front of him.

What he wrote on that sheet went into no document. It stayed in the notebook, which is today in Karentin-Alta, and it is four lines, and they are the best thing he wrote in his life:

I spent five months in a room with six men turning into rules the things I learned because a lad from Bierkut died and I did not know how to purify.

They will say this is doctrine. It is not. It is the list of the things I did not know in time.

A regulation is the memory of the dead written in such a way that the living man does not notice he is reading it.

God forgive me for being proud of it.

XII. The Reform

The first reform of the Inquisition came out on the 2nd of October 1824 and everybody read it as being about heresy, and it was not.

It was about contracts of employment.

Krell knew because it was he who asked for the reform, in a two-page note of March that year, and the note is in the Archive of Means and does not mention heresy once.

The problem was this:

In 1823 the Congregation of Means detected that four licensed purifiers of Ferrum Alpina had left the service of the See. They had not apostatised, had not been accused of anything, had done nothing wrong. They had resigned.

One was working for an alpine prince at three hundred and forty Aurelian ducats a year, which was eleven times what the See paid him.

Another was in Brithon, at an insurance company, waking the escort mineral of the Cross-Insular Company's routes, and had been received by the Solar Church without a single question about obedience.

The other two were in Königsmark and had been reordained in the Schism of the North in a fortnight.

Four men. Krell saw the trend before there was a trend, which was what he did best, and wrote the note, and the note said:

We trained, in 1822 and 1823, eighty-nine purifiers at the expense of the See, at a unit cost of eleven years of formation and a hundred and ninety ducats a year.

We have no instrument whatever obliging them to serve us.

A secular crown does not need to defeat us nor to buy us. It needs to wait while we train, and to pay eleven times more, and in twelve years it has forty, which is exactly the number the host of Marnheim required.

I recommend that exclusivity of service be instituted.

It was instituted.

And here is the part that matters and that no historian of the Order ever wrote in these words: to institute exclusivity of service required a fast tribunal.

Because a licensed man who resigns commits no crime that the canon law of 1824 recognised. A slow, public and juridically scrupulous tribunal — which is what the Inquisition had always been — would take two years to try each case, would produce consultable records, and the records would say, in black and white, what a purifier is worth on the open market.

Three hundred and forty ducats. Eleven times.

To publish that was to put a price on every canon of the continent.

So the reform of 1824 did three things at once, and the three are in the same decree and nobody noticed the connection because they stand in articles forty pages apart:

It created the summary tribunal, with trial in days and no publication of sentence.

It extended the concept of functional heresy to cover the service of the gift under an unrecognised authority, which turned a man who changes employer into a man who abjures.

And it reserved the records, in their entirety, to a single depository with no public catalogue.

Krell drafted the recitals of the first article. He did not draft the other two. And he wrote, in the Memorial, the one sentence in which he comes near to accusing himself without evasion:

I asked for a hammer to nail down a board and they gave me a hammer. I did not ask what else is nailed down with that, and I did not ask because I knew.

The Reformed Inquisition tried, between October 1824 and December 1826, one hundred and nine cases of unauthorised service.

None was published.

Of the hundred and nine, ninety-one returned to the service of the See with summary abjuration and loss of licence for three years. Eleven were not found. Six died in custody, and the declared cause in all six is exhaustion, and it is possible that in three or four of the cases it is true, because the trade kills.

One was executed, in 1826, and is the only case of the period of which a name survives — because he was a man of sixty-one, of Solnica, who was working for no crown and who had no licence and never had had, and who purified salt and cartridges for seven frontier villages for nineteen years, free of charge, because there was nobody else.

He did not understand the charge. It is in the record. He asked the tribunal three times what he had done wrong, and the third time he asked is recorded thus:

Called upon to confess the service of the gift under an unrecognised authority, the accused replied that he served no authority, that he served the villages, and that if they would tell him which was the right authority he would obey it willingly, but that meanwhile winter was coming on.

He was called Pankrác Jelen and the record runs to twenty-two sheets, and it is the most-read document of the Closed Archive after 1852, because the Reformed Inquisition adopted it as an internal instruction piece — it is given to inquisitors to read in their first year, without comment, and they are asked at the end what they would have done.

There is no right answer. That is the exercise.

What he did was this, and it is established on sheets four to nine without contest from anybody, because he confessed everything at the first question and with evident goodwill:

Since 1807, every Thursday, Pankrác Jelen blessed salt.

He did it in the churchyard at Solnica, which had had no resident parish priest since 1804 and received a canon once a quarter. He used a rite he had learned from his mother. They brought it to him in a sack and he laid on his hands and said what he said, and the people carried it back and put it at their doorways.

Since 1812 he had also done cartridges, because the seven frontier villages had been armed that year by the militia of Vykhradia and nobody had thought about who was to wake them.

He did, by his own reckoning, between forty and sixty a week.

He never charged anything. This point was verified with extraordinary thoroughness by the investigation, because charging would have simplified everything — a man who sells a sacrament is a simoniac and one pities him and hangs him — and not a single ducat was found.

He was the keeper of a mill and lived by the mill.

The investigation went well, in the technical sense. He answered everything, did not contradict himself, and did not understand the charge.

It is on sheets eleven to sixteen and it is the part the Reformada gives out to be read.

Asked whether he knew that the rite of purification is reserved to ordained clergy by the Canon of 1791, he replied that he did not know there was a canon and that had he been told he would have stopped.

Asked whether he would persist if ordered to stop, he replied that he would not, and that all he asked was that they send somebody else, because winter was coming on.

Asked whether it had ever occurred to him that what he did might be an offence against God, he replied that it had, at the beginning, and that he had gone about two years with a scruple, and that afterwards he had ceased to have one.

Pressed to say why he had ceased, he replied:

"Because it worked, sir. I thought a great deal about it. If it were an offence it would not work, and it did work, and so I was easy."

The inquisitor who conducted the investigation was called Oswald and was thirty-one, and the note on sheet seventeen is his, and the note ought not to be there, and it is because of it that this record exists as a teaching piece:

Note of the investigator.

The accused is not a heretic. He maintains no proposition against the faith, denies no authority, taught nobody, received no money, and submits to whatever he may be ordered.

The accused is competent. In the verification carried out on the 11th of March, of the thirty pieces treated by him and delivered to the investigation, twenty-eight were woken and two were not, which is a better proportion than that of a licensed man in his third year.

The accused is, from the point of view of the Congregation of Means, an asset.

I request the instruction of the Tribunal as to the ground of the charge, inasmuch as the rule of 1824 punishes the service of the gift under an unrecognised authority and the accused serves no authority, serving villages.

I do not see in the file the authority from which he withdrew himself.

The Tribunal answered in four days.

The answer is six lines and it is the most important thing the Reformed Inquisition wrote in the century, and it is not famous because it never left the Archive:

To the investigator.

The authority from which the accused withdraws himself is that which is constituted by the fact of his exercising.

Whoever purifies outside the See establishes, by the mere continued and known exercise before the settlement, a parallel authority, though without intention and though without doctrine; and it is of that we treat.

The accused need not desire it. It suffices that the villages recognise it, and seven villages recognised it for nineteen years.

Proceed.

Krell read this order twice and then sat holding it.

Because it was perfect. It was, juridically, impeccable — and he understood, with a clarity that chilled him, that it was exactly the argument he himself would use if asked for it, and that the only thing preventing that reasoning from applying to another two hundred thousand people on the continent was the administrative decision not to go and look.

Seven villages recognise a miller. That is a parallel authority.

How many villages on the continent, in 1826, had somebody like that? Krell did not know. Nobody knew. Nobody had asked, and the census of 1819 had asked who holds a licence, which is an entirely different question.

He made an estimate that afternoon, alone, on a half-sheet, by the crude methods one uses when there are no data: number of parishes without a resident priest; known incidence of domestic rite in the folklore of Sal Pradelica and Vykhradia; a plausible rate of success.

The figure that came out was between four thousand and eleven thousand.

Between four thousand and eleven thousand people, in the territory of obedience, doing that. Against three thousand and forty-one licensed men.

He burned the half-sheet. It is the only thing Amandus Krell destroyed in twenty-six years of archive and he confessed it in the Memorial without excusing himself:

I burned it because, if that sheet existed, somebody would read it; and whoever read it would have to act; and to act meant opening proceedings against a number of Christians that I cannot write here without getting up from the chair.

I did not destroy it to hide it from Rome. I destroyed it so as not to oblige Rome to know.

There is a difference and it is the only thing I have left.

Pankrác Jelen was executed on the 4th of November 1826, by elimination of the bearer, without long rite and without a public, as the rule requires.

He was sixty-one.

The settlement of Solnica was informed that the keeper of the mill had died.

And in March 1827 — and this is on sheet twenty-two, added four months later in the hand of the investigator Oswald, and it is the reason the record has twenty-two sheets and not twenty-one —

It is added, for future memory:

On a routine visit to Solnica on the 2nd of March last, it was found that the domestic salt of the seven settlements continues to be treated on Thursdays.

It is now treated by the daughter.

No action was taken.

Krell read this record in February 1827, three years after asking for the reform, in a routine bundle the Inquisition sent him over a question of expenditure.

He read it twice.

Then he shut the folder and went to work, because he had that morning the revision of the budget of the three houses of purification, and because the three houses of purification had trained that year a hundred and forty-one men, and because a hundred and forty-one men wake a hundred and fifty-three thousand cartridges a month, and because somewhere on the eastern frontier there were garrisons that were not going to die because of it.

He wrote in the Memorial, seven years later:

I counted the ones we saved. I counted them well, which was my trade, and they are many, and they are not invention nor consolation: they exist, they have names, and some of them are still alive.

My difficulty is that I know how to count them and I know how to count the man from Solnica, and that the same instrument serves both counts, and that the instrument has no column in which the second enters with the weight it deserves.

We made an empire with a ledger sheet. It is the most effective thing one can do and there is no absolution written on that sheet.

XIII. The Price of the Girl

In 1826 three Blancas died in the house at Aurum Roma.

It was a bad year and it was not an exceptional year; the decade's average was two point one. But 1826 is the year Mother Sofia Kranz took charge, at thirty-eight, and it is the year a woman looked at those three funerals and did a thing none of the four previous Mothers had done, and which was neither noble nor compassionate and was a great deal worse than either:

She counted.

She arrived in March. She was of Bernkanton, a clockmaker's daughter, and had entered the White Vigil at twenty-one not as a Blanca but as an assistant — she had not the gift, never had it, and that is canonically required of a Mother and is the cruellest thing in the arrangement: whoever governs the house is always somebody who does not pay the price.

Aelius met her in the first week, because she came to him.

She did not go to the attending canon, nor to the physician, nor to the eldest of the Blancas. She went to the clerk, and Aelius understood at her first sentence that she had grasped the same thing he had grasped in 1817 and had never told anybody: that he was the only one who heard them all.

She said:

"You have the transcriptions from 1790."

"I have, Mother. Three thousand eight hundred and something."

"I want all of them."

Aelius hesitated, because it was irregular, and then said:

"It will take years."

"I know," said Sofia Kranz. "I have seven."

She had seven because a Mother Vigilant serves seven years. She served twenty-two, because they renewed her three times, and she is the only one in the canon to whom that happened.

What they did — and it is they, because Aelius always insisted on it and because the reading was done by Sofia Kranz and four assistants and he only supplied and organised — took nine years and produced two things.

The first are the seven rules of the Hashmal Syntax, which are in every manual, which everybody quotes, and which are, honestly, a grammar: what comes out of that chair, in what forms, with what patterns of repetition, and how an inscription is distinguished from a sentence.

The first rule is Aelius's and is from 1817 and is that one: one does not receive the same thing twice; one receives the same thing corrected.

The fourth rule is the most used in practice and is purely technical: only that inscription is verifiable which contains information the Blanca could not have possessed, and every inscription of current vocabulary is retained.

The seventh rule is the one nobody likes to quote: the absence of inscription is not recorded as absence, it is recorded as a shift. That is to say: when nothing comes, one writes that there was a shift. One does not write that Heaven did not speak.

Sofia Kranz wrote it in 1831 and wrote it with intention, and the intention was good and the effect was catastrophic, and she died in 1848 without knowing the effect: she wrote it to protect the girls, because a Blanca whose shift is recorded as nothing feels guilty and listens worse on the next shift.

The effect, across sixty years, was that the silences ceased to be countable.

In 1887 the Inquisition asked the Vigil for the series of silences by decade and the Vigil did not have it, because since 1831 it had not been recorded. There exists an inquisitorial note of 1888 about this, of sixteen words, and it is the bitterest thing the Reformada wrote in the century:

The house that hears Heaven is the only one on the continent unable to say when He fell silent.

The second thing that came out of those nine years is the one that matters to this book and is the one that is not in the manuals.

Sofia Kranz made a table.

It had five columns: service name, year of entry, year of death or release, number of shifts served, number of verifiable inscriptions produced.

Aelius supplied her with thirty-six years of data. The other assistants dug out the earlier records back to 1774, which is where the house begins.

The table is in the archive of the White Vigil and has a hundred and nineteen lines.

And what it says, without one word of comment, is:

That a Blanca produces, in a whole career, an average of eleven verifiable inscriptions.

That the average career, counted from entry to death, is five years and seven months.

That the average age of entry is nineteen and of death twenty-four.

And that — and this is the column Sofia Kranz added at the end, in pencil, and which was never inked in — the number of inscriptions per year does not fall with wear. A Blanca in her fifth year produces as much as one in her second.

They are not used up. They die at the same rate at which they were producing.

Aelius de Sárvar was sixty-four when he saw the table finished, on an afternoon in 1835, and his reaction was a clerk's and not a theologian's, and Sofia Kranz never entirely forgave him for it, though she understood.

He said:

"Eleven each. A hundred and nineteen women. That is one thousand three hundred and nine."

"Yes."

"Mother, one thousand three hundred and nine inscriptions in sixty-one years is everything there is. That is what stands behind all the consecrated metallurgy of the continent, all the Chrism, all the Ardentine, the rails, the lamps. One thousand three hundred and nine sentences."

"Yes."

"And a hundred and nineteen girls."

"A hundred and nineteen," said Sofia Kranz. "Of whom four are alive, and one of them is twenty-three and enters her fifth year in November."

The one who was twenty-three was called, in service, the Nineteenth Blanca, and Aelius de Sárvar transcribed her four hundred and eleven times.

Four hundred and eleven. That is the number of shifts, not of inscriptions; her inscriptions were fourteen, which put her well above the average and which is, by itself, a hideous sentence.

He never saw her face.

This always surprises anyone who does not know the house and it is the simplest thing in the world: the clerk sits side-on, three paces behind and to the left, with the flame outside his field of vision — and the Blanca's chair faces the flame. Over eleven years of common service, a clerk sees the nape, the left shoulder and the hands of a woman, and nothing else.

In 1834 Aelius went blind and ceased to see even that.

And in 1835, when she entered her fifth year, he knew her voice better than he knew his own sister's, and knew another thing about her that nobody else in the world knew, which was the sound she made beforehand.

Every Blanca has one. It is the thing that is in no manual and that every clerk knows and that none of them can explain to a canon without sounding like superstition.

The Nineteenth drew a deep breath and stopped. It was not a sigh and it was not astonishment. It was the breathing of somebody preparing to lift a weight, and it always came between one and three seconds before, and in four hundred and eleven times it failed twice.

Aelius learned, from the year 1831, to set the pen on the paper at that sound.

He gained half a syllable. Across eleven years, half a syllable per inscription is the difference between a transcription with a word missing and a whole one, and there are at least two of her fourteen verifiable inscriptions that are complete only because a man learned to hear a woman breathe in.

That is in no document of the Curia and is, in Aelius's view, the most important thing he did in fifty-one years.

She was from a village on the coast of Coronata and had entered at nineteen, in 1831, and had been the eldest of six.

Aelius learned this after four years, and learned it by accident, from an assistant, and was troubled by it for a week — because a clerk should not know the biography of the woman he transcribes, and the reason is a good one: a man who knows that the woman behind him has five brothers listens with tenderness, and tenderness interprets, and interpretation is the one thing a clerk may not do.

He asked the assistant never to tell him anything more.

The assistant was nineteen and was called Doroteia and she kept to it, and kept to it so well that Aelius died in 1841 without knowing that the Nineteenth had died.

We shall come back to this.

Her fourteen inscriptions are in the archive and nine are technical and five are not.

Of the nine technical ones, seven are corrections — first rule, one receives the same thing corrected — and two are new, and one of those two is the reason mineral Ardentine came, from 1833, to be stabilised with slaked lime in a proportion of eleven to two hundred, which reduced refinery accidents at Stromakra by something like two thirds and which saved, over sixty years, a number of Meridian workmen that nobody ever calculated and that is certainly in the thousands.

She was twenty-one. She was a fisherman's daughter. She said nineteen words on a Wednesday in March and never knew what they had done, because the Blancas are not told what their inscriptions produce — a rule of the house, to protect them from vanity and from despair, and both reasons are real.

Of the five that are not technical, four were retained under the fourth rule and one was not.

The one that was not retained is of the 7th of September 1835, has seven words, and is one of the most argued-over inscriptions in the history of the Vigil because nobody can classify it:

the hand that measures is also measured

It is not technical. It contains no information she could not have possessed. It ought to have been retained under the fourth rule and Sofia Kranz decided, against the opinion of the attending canon, that it should not be — and her justification, in two lines in the margin, is the only time in twenty-two years as Mother that she invoked personal authority without an argument:

I retain those I cannot verify. This one I cannot verify and I do not retain it. I shall answer for it.

Aelius, who transcribed it, always said the Mother was right, and always said too that her reason was not theological.

It was that the Nineteenth, in saying it, had not made the sound.

She did not draw breath. She did not stop. She had been sitting forty minutes with nothing, and said seven words out of the middle of nowhere, in her ordinary voice, and then went on with the shift, and there was a technical inscription at twenty past four with the sound and everything.

Aelius wrote the seven words and wrote, beside them, a thing he had never written in forty-five years and which is a disciplinary matter: an observation.

Without audible preparation.

In November 1835 she entered her fifth year.

The table gives five years and seven months as the average, and everybody in that house knew how to read a table, and the Nineteenth did too, because the Blancas are not children and because the house does not lie to them — and this is a point on which the Vigil always refused to yield, against the opinion of three generations of attending canons: they are told.

They are told on entry. At nineteen, in the first week, with the Mother present.

Sofia Kranz said it in seven sentences she had fixed in 1827 and never altered, and the sixth was:

I cannot tell you how many years. I can tell you they will be few and that the house will not spend you faster than it must, and that on the day I see that I am spending you too fast I shall stop, and you will have to believe me because there is no way of proving it to you.

It was because of that sentence — and Sofia Kranz said so in 1836 at the audience, and it is the only time she used a sentimental argument before cardinals, and she used it last, after she had used the table — that the limit of two shifts a week had to exist.

Eminences, I say that to a girl of nineteen every year since 1827. I have said it four times. If you do not institute the limit, I must stop saying it, and the thing I shall have to say instead I cannot formulate.

The Nineteenth served four more years after the fifth, because the limit came into force in 1836 and changed her arithmetic.

She died in March 1840, at twenty-eight, which is two and a half years above the table's average, and Sofia Kranz recorded that in the column with a note of three words that is not a statistical note:

Limit. Two years.

She counted. She counted what the limit had given that woman, in years, and wrote it in the table, and went on writing it for every one who came after — and the column exists, and it sums, and after fifty-eight years of the two-shift limit the sum of that column is two hundred and eleven years of human life, and it is in a cupboard in Aurum Roma, and it is the only place on the whole continent where anybody ever managed to write down the number of a thing that did not happen.

And there is the last fact, and it must be told in the order in which it was known and not in the order in which it happened.

Aelius de Sárvar died on the 9th of March 1841.

The Nineteenth died on the 22nd of March 1840, a year earlier.

And nobody told him.

It was not cruelty and it was not forgetfulness: it was Sofia Kranz deciding, with Doroteia, that the blind clerk of seventy-eight who had been transcribing for fifty-one years and who no longer left the house was not going to be told that the woman he had transcribed four hundred and eleven times had died.

They put another in the chair. Aelius asked, in the second week, whether the Nineteenth was ill.

Doroteia said she was.

He asked twice more over that year and both times she said she was, and the third time, which was in November 1840, he did not ask again.

Doroteia had sixty-two years to think about that. In 1888, Mother Vigilant, at seventy-three, she told it once, to a Blanca of twenty-four from Cracoviana who was in her fourth year and who had six to eight months left, and she told it because the girl had asked her whether the house lied.

And the answer she gave, and which that girl carried into the Conclave of 1888 and into the rest of the little that was left her, was:

"Once. To an old man, about a death, to spare him a year. And if you ask me whether I did right I tell you I do not know, and I have sixty-two years of not knowing, and it is the only thing about this house I still remember every week."

It was from the conversation of 1835 — the one over the finished table, and not Doroteia's fifty-three years later — and from no commission, that there came in 1836 the limit of two shifts a week: the rule Sofia Kranz imposed against the opinion of three cardinals and of the house's own attending canon, and which cut the Vigil's output by nearly a third in the first year, and which she defended at an audience with an argument that was not compassionate, because she knew the compassionate one did not win:

"A Blanca at two shifts a week lives eight years instead of five and produces sixteen inscriptions instead of eleven. It is in the table. I am not asking for clemency, Eminences. I am telling you that we are spending them too fast and losing five sentences for every one."

She won with that. She would not have won with the other thing.

The limit held fifty-eight years and fell in 1884, in a winter, by need of the front, and it was the first of the house's brakes to break.

And there is one last fact, and it is what closes this act, and Aelius recorded it in the margin of his own copy of the table, in 1835, with his hand already large and unsteady because he had been losing his sight since the year before:

The Mother put five columns. The sixth was missing.

The sixth is what each inscription was worth — in ammunition, in rail, in lamp, in a city that did not fall.

Nobody can calculate it, because there is no way of counting what did not happen, and Cardinal Krell told me once that he spent his life trying.

But the sum exists, and somebody is making it, and it is not us.

Eleven sentences a girl. Five years and seven months. If this is commerce, then it is the most unequal ever transacted, and we are on the side that receives.

ACT IV — THE PRICE

XIV. The Table

The first tariffed price of consecrated mineral came out on the 14th of January 1826, had nine lines, and was posted in the market squares of thirty-one cities at the same hour.

Krell wrote it.

It took him four months and they were not four months of calculation — the calculation was trivial, and he did it in an afternoon. They were four months of deciding whom it should hurt.

Because a tariffed price is not a number: it is a map of who pays what, and a table of nine lines with the same figures for everybody is a table that destroys half a continent in two years. Krell had before him seven internal States, four independent kingdoms, two republics, thirty-one market squares and a fleet, all buying from the same factory, and each of them with a different elasticity and a different usefulness to the See.

He made a table of nine lines and four columns of abatement, and the four columns were not posted in the square.

They were in a reserved annex, and the annex runs to fourteen pages, and it is the founding document of the foreign policy of the Clerical Order, and it has never been published in any language.

The four columns were these, and Krell gave them administrative names because to call them by what they were would have been impossible:

Column I — full obedience. Base price. The seven internal States, the state militias, the Military Orders. They paid cost plus twelve per cent.

Column II — alliance with chaplaincy. Plus twenty-nine per cent. Vykhradia and Meridia, which accepted chaplains of the See among their forces and which in exchange bought almost at house price.

Column III — recognised sovereignty without chaplaincy. Plus seventy-one per cent. The Kantonal Confederation, Velinia. They bought dear and bought freely, and they paid what they paid for a reason Krell wrote in three words in the margin: they have no alternative.

Column IV — schism. Plus two hundred and forty per cent, with an annual quota.

Königsmark and Brytonia.

Krell defended Column IV at a session in February 1826 and had, that time, the whole Curia on his side, which left him uncomfortable, because they were on his side for the wrong reason.

The cardinals wanted to punish the schismatics. Krell wanted to punish nobody; he wanted to hold the margin.

His reasoning — and it is in the annex, on page nine, and it is a reasoning of a coldness that still makes the Order avoid quoting it — was:

Königsmark and Brytonia purify. Not so well, not so much, but they purify, and the North has been purifying Hashmalite out of its own Kal'arn since 1819 without asking our leave.

Therefore we have no real monopoly over them, and never shall, and our advantage is one of scale and of unit price and not of dogma.

A very high Column IV does not prevent them from buying. It prevents them from buying in quantity sufficient to arm an entire front, which obliges them to produce the rest at home, which obliges them to burn their own purifiers at a rate we know and they do not, because we have Sofia Kranz's table and the Congregation's series and they have guesses.

In twenty years the schisms will be training men faster than we do and burying them faster still, and they will not know why.

Krell wrote this in 1826.

In 1855, Königsmark found in Kal'arn the largest deposit of Hashmalite in the known world, and the problem of the North ceased to be mineral and became exactly what Krell had foreseen: people. And nobody in Königsburg had a table of a hundred and nineteen lines, because the table of a hundred and nineteen lines was in a cupboard in Aurum Roma, unpublished, kept by a house of women.

The table did not collect itself, and the part Krell least liked to speak of is that of 1826 to 1829, which he called, in private, the three years of knocking on doors.

The resistance did not come from outside. It came from within, and it came from the seven States, and it is the only time in twenty-two years that the Council of Seven united against the Curia as a bloc.

The motive is simple and it is economic. Until 1826, each Cardinal-State sold the purified mineral of his State at whatever price he could get. Ferrum Alpina sold dear because it had the Choir-Iron and the Hashmalite; Sal Pradelica sold cheap and by volume; Maritima Coronata charged freight on top.

The table put a single price on all of them and sent the difference to the See.

What Krell did, in terms any bookkeeper recognises and no theologian would formulate, was to centralise the margin. He took the profit from seven cardinals and handed it to a strongbox in Aurum Roma that he himself managed.

The seven understood within three weeks.

Their offensive was excellent and was made on two fronts, and Krell wrote in the Memorial that he learned more about government in those three years than in the eleven before.

The first front was juridical. The seven argued, in a joint memorandum of forty-one pages presented in May 1826, that the right to dispose of the product of the State predated the table and derived from the constitution of 1815 — which was true, because Aurelianus I, in creating the seven States, had given them fiscal autonomy in terms nobody had revised since.

The second front was operational and was a great deal worse.

They did not refuse compliance. They complied — and they were late.

They were late with the monthly returns. They were late with the remittances. They discovered, collectively and without coordinating in writing, that a Congregation of Means with a hundred and twelve officials cannot process seven States working at half speed, and that the table, without monthly returns, is a sheet nailed to a door.

By the end of 1826 the See had collected nineteen per cent of what the table provided for.

Krell did not go to the One.

This point he made a particular effort to record and to explain, because it is the decision he was proudest of and the hardest to defend: he could have taken it to Aurelianus I in June 1826 and obtained a decree, and the decree would have settled the juridical question in one paragraph.

And it would have been the end.

A decree against the seven States obliges the seven States to choose between obeying and disobeying.

While there is no decree, they are being late. Once there is a decree, they are resisting.

The same conduct, the same man, the same fourteen days' delay on a return — and from the decree onward it becomes matter of obedience, and a Cardinal-State who disobeys once understands that he can disobey, and we discover at the same moment as he does that we have no way of compelling him.

I learned this late and I learned it from them: the authority that is not tested is greater than the one that is. I did not go to the Holy Father because I preferred a table at nineteen per cent to a decree at nothing.

What he did instead took three years and has nothing heroic about it.

First, he bought Sal Pradelica.

Not with money: with purifiers. Sal Pradelica had the worst ratio of licensed men to tonnage of the seven States and a Cardinal-State who had been complaining of it for nine years without success. Krell offered him thirty-one licensed men out of the houses of formation over four years, with priority over the other six, in exchange for punctual monthly returns.

They signed in October 1826. Sal Pradelica left the bloc.

Six.

Then he split Maritima Coronata off from Ferrum Alpina, which was cheaper than it should have been because the two had detested one another since 1819 over a freight charge.

Then he waited.

The Schola Coronata fell by itself in 1827, because the Schola lives on subsidy and not on production and had, in practice, no margin whatever to defend — the Cardinal-State of the Schola had joined the bloc out of solidarity and left it out of arithmetic, and Krell wrote beside his name, on the list he kept, a single word: predictable.

Vallia Sacra fell in 1828 for a reason Krell never foresaw and that amused him for the rest of his life: the Cardinal-State of Vallia was of a family of tenant farmers from near the village where Krell was born, and a cousin of his wrote him a letter.

That left Ferrum Alpina and Limes Orientalis.

Limes Orientalis was the only one Krell had to pay, and he paid dearly, and it was the best bargain of the three years.

The Cardinal-State of the Limes did not want margin. He wanted garrisons.

He had fourteen of them and six purifiers for them, and he was, in 1828, losing ground in fact — not in battle, but in the way that frontier has always lost ground, which is by villages emptying because they can no longer renew the household perimeter.

Krell gave him a thing that was not in the budget and that he had to invent: a permanent allocation of licensed purifiers to the frontier State, financed by the See and not by the State, revised every ten years, and indexed to the number of settlements and not to revenue.

He called it the frontier provision.

It was instituted in January 1829 and exists still in 1890, and it is the reason Limes Orientalis and Sal Pradelica, which are the two poorest States of the Order, have more licensed men per head than the Schola Coronata — and it is, in the whole administrative history of the continent, the only policy of redistribution that ever worked for sixty-one years running.

Krell was sixty-one when he wrote it.

He wrote in the Memorial:

It is the only thing I did that I would defend before God without preparing anything.

And I did it to break a political bloc. It was a bribe. It worked because it was a good bribe, and it is good because it went on working after the bloc had ceased to exist, and I do not know what to do with that.

Ferrum Alpina was left alone and capitulated in November 1829, without agreement and without any consideration, in a letter of four lines.

Krell did not celebrate and wrote nothing that day.

He wrote two years later, and wrote this, and it is the paragraph of the Memorial that the officials of the Congregation of Means of 1890 still copy out by hand to remind themselves:

It took me three years to collect a table of nine lines and I did not use the authority of the Holy Father once.

I used thirty-one purifiers, a freight quarrel of 1819, a cousin, and a frontier provision I invented in an afternoon.

An empire is not made with decrees. It is made with what people want and cannot ask for aloud, and the trade consists in finding out what that is, and in keeping it in stock.

That is why they gave me three rooms and eleven men in 1814 and called me the Congregation of Means, and I thought the name a humiliation for nine years.

The price table of 1826 is the date from which the Clerical Order ceases to be a See and becomes an empire, and Krell knew it on the day he posted it, and wrote nothing about it for five years.

When he finally wrote, in the Memorial of 1831, he wrote this:

I am asked when it was that we became what we are. I am offered the Decree of 1815, or the death of the Nine Standards, or the doctrine of 1824.

It was on the fourteenth of January 1826, at nine in the morning, when a man of mine nailed a sheet to a door in Bernkanton and another in Brithon and another in Vykhrad.

In that instant we stopped asking and began to charge, and nothing that came after — not the rails, not the victories, not the purple — is anything other than the compound effect of that morning.

A sheet nailed to a door. There was no trumpet.

XV. The City That Grows

Between 1826 and 1833, Aurum Roma grew from two hundred and eleven thousand inhabitants to three hundred and forty thousand.

Krell watched it from the window of the third courtyard for seven years, and the thing he remembered best was not the scaffolding: it was the noise.

The city had, from 1827, a permanent background noise that had not existed before and that never stopped again in sixty years — stone cart, hammer, pulley, and at six in the morning and six in the evening a thing that exists only in cities that are being built, which is the sound of ten thousand people walking at the same time in the same direction.

They came from everywhere. They came because there was work, and the work paid, and by 1828 the See was the largest employer on the continent by a margin no kingdom came near to matching.

And they came, above all, because the price table had meant that all the raw mineral of the continent now had to be routed, purified and taxed through circuits the See controlled — and a commercial circuit is not a line on a map, it is warehouse, it is quay, it is clerk, it is carter, it is inn, it is guard, it is exchange house, it is a court for contracts, and all of that is people, and people live where they work.

Krell had calculated the growth in 1825, when proposing the table. He was short by forty-one per cent.

It was the largest error of estimate of his career and the only one he took pleasure in.

In 1829 the Congregation of Means moved into a building of its own with forty-one rooms and a hundred and twelve officials. Krell hated it.

He hated it because he had spent his life doing sums with eleven people in three rooms, and because he knew — in a way he could demonstrate to nobody and that irritated him profoundly — that an office of a hundred and twelve people produces a different species of truth from the one an office of eleven produces.

He wrote in the Memorial:

With eleven men, I knew every number that reached my table and I knew who had made it and whether that man had a tendency to round.

With a hundred and twelve, a total reaches me. The total is better than mine of 1819, faster, more complete, and it obeys me.

And I no longer know where it lies. I know it lies somewhere, because they all lie somewhere. I have lost the nose.

There was, in those seven years, one thing that went wrong and that Krell never managed to solve, and it is the one documented administrative defeat of his career.

It is called the Ferrumkar leakage.

From 1828, the declared output of Silver Chrism from the houses of purification of Ferrum Alpina began to diverge from the declared consumption of the units they supplied. The divergence was small — between two and four per cent — and it grew, and by 1832 it was running at seven.

Seven per cent of the output of Ferrum Alpina is a hundred and ninety thousand cartridges a year.

Krell ordered three audits. All three came back clean.

There was no theft. There was no smuggling — that was the first hypothesis, and it was investigated with considerable thoroughness, and the Reformed Inquisition went so far as to detain four forwarding agents in 1831 and to release them all.

The mineral was not leaving Ferrumkar.

It was being consumed in Ferrumkar, and there was no record of who was consuming it, and the fourth audit — the one Krell ordered in 1833, by a young and obstinate canon named Vitale — came back with a sentence on page eleven that Krell read and reread and then filed without action:

There is in the sector no record of loss. There is, however, among the workmen of the galleries and the forges, a widespread practice of acquiring single cartridges for domestic use, within the household perimeter, in a quantity the forge-masters describe as ancient custom and the toleration of which is recommended for reasons of industrial peace.

Domestic use. Household perimeter.

A hundred and ninety thousand cartridges a year protecting miners' houses in a city that was on no front at all, two hundred and forty kilometres from the Limbo, behind a range that had never been entered.

Krell understood. He wrote it in the Memorial and wrote it without room for doubt, and it is the passage of that book that the Reformed Inquisition had copied in 1853 and that has circulated inside the Order ever since as a one-page extract:

They are not afraid of demons. They are afraid of the dust.

A man of Ferrumkar knows, at thirty, that he is going to die of it and that his son goes in after him. He cannot leave, because there is a waiting list to get in and because there is nowhere else.

So he buys a cartridge, and puts it on the kitchen shelf, and never fires it — and I saw three houses, and in all three the cartridge was on the shelf, and in all three the woman of the house showed it to me as one shows a relic.

It is not superstition. The thing works. It is woken. It cost twenty seconds of the life of a canon.

They bought a piece of a woken thing and put it on top of the dresser, because it is the only thing in that city that has been touched by Heaven and that belongs to them.

I spent four years auditing this as though it were shrinkage of inventory.

It was liturgy. It was liturgy happening in a kitchen, without a priest, and we have no name for that and nowhere to enter it, and so I ordered it filed.

The Serene Velinia asked in November 1829, and the request came to the Congregation of Means, and Krell recognised it in the first paragraph.

It was the petition of 1820 again.

Not literally — it was written in chancery Venetian and not in the Latin of a Reichshamn jurist, and it was a great deal better-mannered, and it came with an offer instead of a threat. But it was the same thing: a secular power asking Aurum Roma to give it capacity of purification for a military operation Aurum Roma did not command.

The Vigil of the Pit is decennial. Leviath rises in 1830 as he had risen in 1820 and in 1810, and the combined fleet goes out, and does not kill him, and comes back with fewer ships. Velinia arms forty-one hulls; the Maritima Coronata arms nineteen.

And a fleet in campaign spends consecrated ammunition at a rate that does not bear comparison with an army on land, because a gun does not fire five times a day.

Velinia asked for twenty-two licensed purifiers, for five months, and offered to pay them at four times the See's stipend, plus transport, plus life insurance for the families in case of loss.

Krell laid the request on the table and sat looking at it with a sensation he did not know how to classify and which he ended by classifying, in the Memorial, as administrative vertigo:

Ten years earlier the same question had split us down the middle in a room and cost nine thousand lives.

This time I answered it in nine days and it was settled by a secretarial order. It did not go to council. It did not go to conclave. It did not trouble the Holy Father.

And the difference between the two occasions is not moral, is not theological and is not one of courage.

In 1820 we had no procedure. In 1829 we had one.

The procedure had existed since 1826 and was called, in the jargon of the Congregation, the Detachment of Exterior Ministry, and Krell had written it with three canons over fourteen months, and it has twenty-nine articles and solves exactly the problem that had destroyed the Crusade of the Nine Standards.

It says how many. It says for how long. It says who pays. It says that the detached men remain under the jurisdiction of the See and answer to the chief chaplain and not to the commander of the operation. It says the operation is obliged to furnish them with separate quarters, a personal guard and — this is article seventeen, and it is Krell's, and it is the one he defended most — a limit of daily hours, enforced by the chief chaplain, with power to suspend the detachment in case of non-compliance.

Article seventeen exists because of Canon Bonifácio.

Krell never wrote that anywhere and it was not necessary: everybody in the Congregation knew why there was a limit of hours and why the power to enforce it lay with a cleric and not with the admiral.

It was granted.

Twenty-two licensed purifiers, for five months, under the twenty-nine articles, with a chief chaplain appointed by the See and with Velinia paying four times the stipend — which Krell accepted without haggling because the money was the least interesting part of that agreement.

What he wanted was in article twenty-six, and Velinia accepted it without discussion because in November 1829 it looked to them like a formality:

The records of production and of consumption of the operation are the property of the See and shall be delivered to it at the close of the detachment.

That is to say: the See was going to learn, cartridge by cartridge, what it costs in consecrated ammunition to contain Leviath for one night.

Nobody in the world had that figure. Velinia had it, in part, in the archives of the Arsenal, and had never given it to anybody, and the See had been estimating it by guesswork for thirty years.

Krell obtained it by a clause in a contract of employment.

The Vigil of the Pit of 1830 went badly for reasons that have nothing to do with any of this: the fleet arrived late, on a wrong tide and in closed fog, and the Green Eye was already out when the line formed.

Fourteen ships to the bottom in one night. Thirty-one of the eighty-four families of the Council without a male heir.

Of the twenty-two purifiers detached by the See, nine died.

Seven at sea, with the hulls they were in. Two in the four months after, of what the trade does.

Krell received the list of the nine on the 14th of October 1830 and read the names, and knew six of them because he had signed the licences.

And then he received, in January 1831, the records under article twenty-six — four bundles, from the Arsenal of Velinia, with the consumption of the operation hour by hour — and spent three weeks working them, and produced a document of eleven sheets that is today the basis of all consecrated naval doctrine on the continent and that contains, on sheet seven, the figure nobody had:

To contain Leviath for one night costs between a hundred and forty and a hundred and seventy thousand consecrated cartridges and does not kill him.

A hundred and fifty thousand, on average, once every ten years, for a thing that comes back.

It is the equivalent of a hundred and seven purifiers working a whole month.

Krell wrote the conclusions and then wrote, on a separate sheet, what he thought, and the separate sheet is in the Memorial and is the passage the Order's naval officers quote without knowing whose it is:

We did everything right.

There was procedure. There was a limit of hours and there was a chief chaplain with power to impose it, and he imposed it twice and there is a record of both. There were personal guards, separate quarters, contract, insurance.

Not one of the nine died as Canon Bonifácio died. Not one was spent. Not one made, on the fifth day, dead silver that somebody issued without telling him.

Seven went down with the hulls they were in, doing their office, in an authorised, paid, supervised and canonical service.

And I understood, at the end of eleven sheets, that I had spent fifteen years building a machine whose only demonstrable function is to ensure that people die correctly.

They do not die less. They die in the right place, with paperwork, and we are left knowing how many there were.

God forgive me if this is little. I have spent my life doing it and I do not know how to do anything else, and I have the reasonable impression that it is more than there was, and the reasonable impression has not been much company to me.

The second movement is of 1830 and lasts one night.

Mattias of House Solar was appointed Captain-General of the Order of the Knights of the Seal on the 9th of June 1830, at forty, and received the appointment at Karentin, and the first thing he did was to go to Aurum Roma to make his obedience, because it was in the regulation he had himself helped to write.

He stayed three weeks. In one of them he dined with Cardinal Amandus Krell, who was then sixty-two, and it is the only documented meeting between the two, and it is documented because Krell recorded it.

It was a bad dinner.

They did not quarrel. Neither of the two was a man for quarrelling and both were of humble origin and recognised it in one another immediately, which in general helps and that night did not help at all.

What went wrong was a question.

Mattias asked — and asked it in complete innocence, because he did not know, because nobody had told him and because the file had been closed for eleven years:

"Eminence. In nineteen, when the census was made, I sent in a form."

"You did."

"Did Your Eminence see it?"

"I saw it."

Mattias set down his glass.

"I spent three years waiting for a prosecution," he said. "I wrote that expecting a prosecution. I could not declare nothing, because it was a lie, and I could not declare what I did without accusing myself, and so I declared and waited, and nothing came. Eminence, I am forty years old and I have spent eleven of them thinking my Order had forgotten me in an archive."

Krell was silent a while.

Then he told the truth, because he was sixty-two and because he was the sort of man who told the truth when it no longer cost anything:

"I did not forget you, Captain-General. I put your form in the wrong pile on purpose."

"Why?"

"Because I needed it."

And here — and Krell wrote this with an honesty that cost him, and wrote it knowing that one day somebody would read it — here the conversation stopped being about 1819.

"Your Eminence," said Mattias, very slowly, "kept a crime of mine because you needed it."

"I kept a fact of yours. I had not, in nineteen, one single record, anywhere, that the thing could be done by somebody unordained. I had rumour. I had what they say in the villages." Krell laid both hands on the table. "You sent me a signed declaration, with the name of a unit, with a figure, with a method. Eleven cartridges. Not none."

"And what did Your Eminence do with it?"

"Nothing, for five years. Then I gave it to a canonist in 1823 and told him to find me a way for that to be legal, and he took eleven weeks, and what he wrote is the Vow of Measure you have in your regulation and that, as I am informed, has already saved four garrisons."

Mattias said nothing for a long time.

Then he said:

"So I am the proof."

"You are the only documentary proof this house has ever had, produced voluntarily, that the gift does not pass exclusively through us." Krell looked at him. "And it is in an archive to which only I and three men have access, and it will stay there, and you will never be troubled, and on the day I die somebody will inherit that key."

Mattias went back to Karentin three days later and never spoke to Krell again.

And what he carried from that dinner, and carried the rest of his life, was not fear and was not gratitude.

It was a thing he wrote in the notebook on the road, between Vallium and the pass, in three lines, with the hand shaking from the trot:

He did not protect me. He kept me.

There is a difference and it took me the whole dinner to find it: a man who is protected is safe. A man who is kept is available.

I have been in a cupboard in Aurum Roma for eleven years, with a key, and somebody is going to inherit it.

XVI. The Year In Which Nothing Happened

In 1830, nine days' distance from Aurum Roma, fourteen ships went to the bottom in one night in the Sea of the Choir, and thirty-one of the eighty-four families of the Council of Velinia were left without a male heir, and the Vigil of the Pit cost again what it always costs.

The White Vigil learned of it in October and learned it badly.

And what matters about that year, in the house of the flames, is not the Pit.

It is that in 1830 there was not one verifiable inscription.

Twelve months. A hundred and ninety-two shifts served, with four Blancas in service and the rule of two shifts a week still to come.

A hundred and ninety-two shifts. Inscriptions of current vocabulary: forty-one, all retained under the fourth rule. Verifiable inscriptions: none.

Aelius de Sárvar had thirty-nine years in the trade when that happened, and it was the third time he had seen a lean year, and the other two had been 1803 and 1819, and in both the following year had made it up.

This one did not. 1831 gave three. The average of the preceding decade was eighteen.

The Curia did not notice, and the reason it did not notice is the most sinister thing in this chapter and is purely administrative.

The Curia did not notice because 1830 was an excellent year.

The price table was yielding. The three houses of purification had trained, by that year, four hundred and eleven licensed men. The Knights of the Seal had nine regiments and a written doctrine. Continental output of Silver Chrism had risen sixty-one per cent in five years, not because there was new revelation, but because there were more people applying the revelation of 1817 and that of 1822 and that of 1826.

An inscription is not a machine. It is an instruction. And an instruction received in 1822 goes on yielding in 1830, in 1840 and in 1880, and yields more with every year in which more men are trained who can execute it.

So the continent, in 1830, was growing fast on the basis of things that had been said to it eight and thirteen and twenty years earlier, and nobody in the Congregation of Means had a line in the budget for quantity of new things said to us this year, because that column never existed and does not exist today.

Aelius wrote, in January 1831, a report of two sheets and sent it to the Mother Vigilant, who was Sofia Kranz, who read it and sent it up to the Curia with a note of support.

It came back in March with an order of seven words from a monsignor of the office of internal affairs:

No matter here. To be filed. Prayer recommended.

Aelius kept the order.

He kept it the way he had kept the transcription of September 1821, that is to say, in a Bible, and it is for that reason that there exist today two papers in the Bible of a clerk who died in 1841 and it is for that reason that the Reformed Inquisition, in 1888, reconstituting the series of silences that the seventh rule had made uncountable, found exactly two points earlier than 1850.

One is Aelius's report of 1831.

The other is the order of seven words.

And it is with these two papers, and with nothing else, that the Reformada established — sixty-two years later — that the curve was already falling in 1830, and therefore that the fading of the Angels did not begin in the Stagnation: it began before the Age of the Forge began, in the best year of the century, when nobody had any reason whatever to look.

What Aelius wrote in the margin of the order, with his hand already large because his sight was going, was:

Prayer recommended.

I recommend it too. I have prayed for forty years and have not left off one single day.

But I asked whether they were speaking less and the monsignor answered me that I should pray more, and the two things are not answers to one another, and I do not know whether he did not understand the question or whether he understood it so well that he could not answer it in writing.

I am a clerk. The only thing I know how to do is to note things down, and I noted, and they noted back at me that there is no matter here.

It will stay in this book. Somebody will open it.

ACT V — KARENTIN

XVII. The Fall

Karentin-Alta fell on the 3rd of November 1834, between two and six in the morning, and the thing nobody in Aurum Roma managed to understand for two years was that it did not fall from outside.

The north wall was not breached. It stands today. It has thirty-one metres of Choir-Iron facing and was built in 1811 by people who knew exactly what they were building it against, and it held twenty-three years of everything the fissure sent, and it was not by there.

It was by the aqueduct.

The fortress-city had a cistern, and the cistern filled by gravity from a spring at sixteen hundred metres, three kilometres to the north-east, along a stone conduit covered by thirty-one vents.

The vents had grilles. The grilles were of Choir-Iron in the first eleven and of common iron in the other twenty, and the reason for that is in a budget of 1809 and it is the reason it always is: Choir-Iron cost eleven times more and the conduit was above the snow line and nobody believed anything would climb up there.

Mattias had read that budget in 1831, when he became Captain-General and ordered a review of the fortifications of the seven Houses, and had written in the margin replace the twenty, and had sent the request to the Congregation of Means in March 1832.

The request was granted.

The work was scheduled for the summer of 1835.

This is what Mattias knew in November 1834 and it is what he never said aloud to anybody in his whole life, and it is known because it is in the notebook:

I asked in thirty-two. They granted it in thirty-three. They scheduled it for thirty-five because the work is summer work and because above sixteen hundred metres one does not build in October.

There was no negligence. There was no meanness. Everybody did his office competently and within the term, and the city fell by eleven months.

There is nobody to blame and that is what keeps me awake. If there were somebody I should have somewhere to set this down.

What is known of the night is known from eleven depositions, and the longest is that of Sergeant Vitor of House Solar, forty-one, taken at Vallium on the 19th of November 1834 by a canon who wrote down everything he said and did not ask him a single question for two hours, which was the most intelligent decision anybody took that month.

It is transcribed in Mattias's report between pages nineteen and thirty-one. It is the only part of the document Krell had copied into the Archive of Means, and the only part the Captain-General did not rewrite.

Deposition of Sergeant Vitor, House Solar, fourth company. Vallium, 19 November 1834.

I was on cistern duty. It is punishment duty and I was under punishment for a matter of drink in October, and I say so at once because I do not want it to look afterwards as though I was there on merit.

The cistern has a guard room with two men and a bell that rings if the level drops fast, because the level dropping fast means a rupture in the conduit and that is what one expects to happen one day. The bell had never rung in twenty-three years and everybody called it the bell of the Judgement, begging your pardon.

It rang at twenty to two.

The other man was Hansi, who has been in House Solar nine years and is from the valley of Karentin, and Hansi said at once that it was not a rupture because there had been no tremor and because the conduit bursts with a tremor. I know nothing of conduits. I went to look at the level and the level was rising.

Rising. That is what I want written down, canon, because when I said it to the lieutenant he made me repeat it three times.

The bell is made to ring when it drops. It was ringing because the float had gone to the bottom, and the float had gone to the bottom because something in there had pulled it, and the water was rising because something was in it and had a body.

The two of us went to the mouth of the cistern with a Lumen lantern. It is regulation to carry Lumen to the cistern and that night it was the only thing in the regulations that was any use.

With Lumen the thing does not come up. It does not come into the light. It is there, and one hears it, and the water moves, and so long as the lantern is lit and the hand holding it does not shake, it stays at the bottom. The two of us stood there, holding the lantern turned to the water, from twenty to two until a quarter past three.

One hour and thirty-five minutes, canon. I counted because there was a wall clock in the guard room and because there was nothing else to do with my head.

Hansi wept most of the time and there is no shame at all in it and I have told him so. He wept quietly and held the lantern and did not lower it once.

What came in by the cistern was not what took the city. That we learned afterwards. What came in by the cistern was two things and they stayed at the bottom because of the Lumen and died there when it was drained in thirty-six.

What took the city came in by the upper vents and went down the dry conduit, which is the service one and carries no water in November, and came out into the casemates of the inner ring by inspection shaft number nine, which is on the inside of the wall.

On the inside, canon.

Twenty-three years of a wall thirty-one metres high with a Choir-Iron face, and they came in at us through a manhole with a grille made by a village blacksmith, on the inside, two hundred paces from the governor's house.

The inner-ring alarm rang at nine minutes past two.

From there on it is confusion and I am not going to say I saw things I did not see. I was at the cistern and I stayed at the cistern until a quarter past three, which was when a corporal came to say the city was being evacuated and that we were going out with the fourth column.

I told him I could not let go of the lantern.

He told me the city was being evacuated.

And I told him again that I could not let go of the lantern, and he understood, and it must be said that he understood in two seconds, because if we let go of it the thing at the bottom of the cistern came up and came out behind the column that was leaving.

He asked how long we could hold.

I said the lantern had charge for four hours and that my hand had less.

We stayed another two hours and forty minutes.

I am not going to say it was courage because it was not. It was that going meant twenty men dying behind us and us knowing it and going anyway, and I am not one of those and neither is Hansi, and nothing more.

At five fifty-five Canon Petrek came himself with two men and a charged lantern and told us to get out.

I asked him who was staying to hold it.

And he told me — and this is it, canon, this is what I came here to say and all the rest is padding —

he told me:

"I am."

Canon Petrek was sixty-one and was from Solnica and I knew him eight years. He woke our ammunition on Thursday mornings and in the afternoons he played cards with the sergeants and always lost and always paid.

By then he had been waking cartridges since half past two in the morning, with the other three, for the rear of the column. Three and a half hours. His right hand was shut and would not open.

He held the lantern with the left.

I told him he could not stay alone and he told me he was not alone, and I did not understand at the time whether he was speaking of the two men he had brought or of something else, and I did not ask, and I have spent these sixteen days wishing I had.

I went out with the fourth column at ten past six.

The last thing I saw of Karentin-Alta was the mouth of the cistern with light inside it.

It was still burning when the road turned.

Mattias read this deposition for the first time on the 22nd of November 1834, at Vallium, and read it standing, and when he had finished he sent for the canon who had taken it and asked him one question.

He asked whether the sergeant had said anything more after the transcription ended.

The canon said he had. He said Sergeant Vitor had sat a while and then had asked whether what he had told them was enough to ask for something.

"Ask for what?" said Mattias.

"He wanted to know whether it was enough to ask the Order to have the city retaken." The canon hesitated. "I told him such decisions are not taken on depositions. He asked me what they are taken on."

Mattias was silent.

"And what did you answer him?"

"I answered nothing, Captain-General. I did not know."

There died, on the night of the 3rd of November, one thousand four hundred and eleven people: two hundred and nine knights, a hundred and forty-two of the garrison troops, and one thousand and sixty civilians of the three thousand who lived there.

One thousand nine hundred and twenty-seven came out, by the south gate, in four hours, and the evacuation was executed in a manner the military manuals of the continent have studied ever since, and it was executed by a lieutenant of twenty-six whose name is not known because he died in 1841 and House Solar does not keep surnames.

The garrison's four purifiers stayed. All four.

Not out of declared heroism and not by order — Mattias verified this, and verified it with an almost morbid care over the following winter, because he needed to know whether somebody had ordered them to stay, and nobody had.

They stayed because the third rule says that a unit without a purifier withdraws, and because a purifier who leaves with the evacuation column takes with him the one thing that allows the rearguard to fire, and because the four understood simultaneously, without speaking to one another, that a column of nineteen hundred people was going to need consecrated fire behind it for four hours.

They made four hours.

The last of them was a canon of sixty-one named Petrek, of Solnica, and the thing that is known of him is that in the last hour he was waking cartridges sitting on the floor, propped against a wall, and passing them to a sergeant who issued them, and that he could no longer lift his right hand and was performing the rite with his left, which is not permitted and is not taught and ought not to work.

It worked. It is in the sergeant's account, who came out alive.

Mattias wrote the report of the fall in eighteen days and it is the longest document of his career and runs to sixty-one pages.

There is not one word of justification in it. Krell read it in Aurum Roma in January 1835 and wrote on the cover, in pencil, two words the archive has preserved: honest man.

The part the Curia did not want was page fifty-nine.

On page fifty-nine, Mattias did a thing nobody had asked him to do: he calculated the cost of the grilles.

The twenty grilles of common iron, replaced with Choir-Iron, at the price of the 1826 table, Column I, with labour and high-altitude carriage: four thousand one hundred and ten Aurelian ducats.

And he wrote, beneath it, without comment, what the Congregation of Means had spent in the same financial year of 1832 on representational works in the fourth courtyard of Aurum Roma: eleven thousand two hundred.

The Curia ordered page fifty-nine withdrawn from the archive copy.

Mattias had a copy. The copy is in Karentin-Alta, framed since 1871, and most of the officers who walk past it suppose it is a relic of the defence and cannot read what is on it.

XVIII. The Dark

Aelius began to lose his sight in 1832 and went blind in both eyes in August 1834, three months before Karentin-Alta, and the thing happened slowly and then all at once, which is how it happens.

There was a physician at the Vigil. The physician called it double cataract with exhaustion and knew nothing further, and the only thing he said to Aelius that was of any use was that it had not been the flame.

"You never looked."

"Never."

"Then it was not that." The physician packed up his things. "It is age and it is forty-four years of writing by tallow candle in a hand of two millimetres. Forgive my frankness: you have worn out your eyes in the most banal way there is."

Aelius thanked him sincerely. It was the best news he received that year.

A clerk of the Vigil must never look at the flame.

It is the first rule one learns and the only one taught with a threat, and Aelius de Sárvar had spent forty-four years keeping it — and had kept it as a man keeps a hard thing, which is to say willingly, every day, fourteen thousand times, with the corner of the eye pulling and being pulled back.

And in August 1834 he no longer had to be willing.

He wrote about this once, to his niece, in a letter dictated in 1837, and the letter exists:

Everybody in this house thinks I am being brave and nobody has the courage to tell me so, and I let it be because explaining would be worse.

I am not being brave. I was, for forty-four years, a man who chose every day not to look.

Now I am a man who cannot look.

I have lost the one thing I did that was worth anything before God, and nobody notices, because from the outside the two things look the same.

He went on working seven years.

This seems impossible and is not, and the explanation is mechanical and without any romance in it: a clerk of the Vigil writes without looking at the paper from his second year in the trade, because to look at the paper means lowering the head and losing syllables. One writes side-on, with the sheet under the hand, looking at the wall.

Aelius had had his hand formed for forty years. The hand knew.

What he lost was not the writing: it was the rereading. He could no longer verify what he had written, and therefore could no longer correct it, and therefore the Vigil began placing an assistant beside him to check each sheet at the end of the shift.

The assistant was nineteen in 1834 and was called Doroteia, and she spent seven years reading aloud to Aelius de Sárvar what Aelius de Sárvar had just written, so that he might confirm it.

In 1868 she became Mother Vigilant, and in 1888 she still was, and she is the Mother Doroteia of the emergency conclave.

Neither of the two ever knew what the other would become.

The thing that matters to this book happened in 1836 and was of a completely administrative banality.

The Curia sent, in February, a dispatch to the White Vigil proposing the honourable retirement of the clerk Aelius de Sárvar, aged seventy-four, blind, forty-six years in service, with full pension and lodging for life.

It was a generous dispatch. It was, besides, entirely sensible: a blind clerk who needs an assistant to check his work costs two people and yields one.

Sofia Kranz refused it.

She refused it in writing, in three paragraphs, and the third paragraph is famous inside the Vigil and has never left it:

Most Reverend Eminence, the clerk Aelius de Sárvar entered this house in 1790 and has since transcribed thirty-one Blancas, of whom twenty-nine are dead.

There is in this house, nor in the Curia, nor on the continent, anybody else who has heard them all. The canons rotate every two years, the Mothers every seven, and the girls do not speak to one another by a rule of this house that Your Eminence knows.

He is the continuity. Not because he is wise — he is not wise, he is a clerk — but because he stayed.

If you retire him, this house comes to have a memory seven years long. That is what is being proposed, and I do not believe it is what is being intended.

He stayed.

XIX. The Eleven Weeks

The retaking of Karentin-Alta began on the 2nd of March 1836 and ended on the 19th of May, and that is seventy-nine days, and everybody calls it eleven weeks because eleven weeks is what is in the report and because it sounds better.

It was done with Knights of the Seal alone.

This is the fact the Order is proud of and it is necessary to say where it came from, because it did not come from pride: it came from Mattias having refused, in January 1836, an offer of seven thousand men of crown troops and of two alpine houses, and the refusal cost him a written reprimand from the Council of Seven and nearly cost him his post.

The reason he gave in writing was logistical and was true: seven thousand extra men in a glacial valley in March means supply the road will not bear.

The reason he did not give is in the notebook and is two lines:

If they go in with crown troops, the city is retaken by everybody, and in ten years half the continent remembers that Rome needed help to clear its own mountain.

It costs me six weeks more and I do not know how many men. I shall pay.

He took four thousand two hundred men. Seven regiments from the seven Houses, which was the first time in history that the seven operated together and the last until 1859.

He took nineteen licensed purifiers — nineteen, out of a continental establishment that in 1836 stood at about five thousand six hundred — and the Congregation of Means took five months to authorise it, and Krell signed the authorisation at sixty-eight and the signature is unsteady.

And he took, for the first time into the field, first-generation Ardentine artillery: eleven pieces, out of the southern refineries fourteen months before, with a handling instruction of forty pages that nobody had tested at altitude.

Two exploded. The second killed nine men.

It is in the report, on page eleven, with the nine service names, and Mattias insisted that it be there, against the opinion of two senior officers who wanted the thing recorded as an accident of materiel, and his argument was the one it always was:

"If we do not write it down, next year somebody does it again."

The campaign is not militarily interesting and Mattias was the first to say so.

It was a siege. It was a seven-week siege of a city the Order had built and whose plans the Order had, with Ardentine sapping at three points chosen by an engineer who had worked on the wall in 1811 and who was then seventy-three and who made the whole campaign in a litter.

The hard part was the aqueduct.

They went in by where they had come out. By the vents — by the twenty of common iron that Mattias had ordered replaced in 1832 and whose work was scheduled for the summer of 1835 and was never done, and which were exactly as they had been, and it was by there.

Mattias went personally in the third party. He was forty-six and was Captain-General and ought not to have gone, and went, and never gave any explanation to anybody.

The city was retaken on the 19th of May 1836, and what was inside it stays outside this book, because this book is about a desk in Aurum Roma and about what is decided in a room.

It stays outside, and there is one thing that stays inside, and it is a thing of bookkeeping.

In November 1834 there were in Karentin-Alta three thousand and forty-four civilians, by the parish census of August of that year.

One thousand and sixty died in the night.

One thousand nine hundred and twenty-seven came out by the south gate in four hours and were received and counted at Vallium between the 6th and the 11th of November, and the list exists, with names, and it is the document the diocese used for thirty years to settle inheritances.

One thousand and sixty plus one thousand nine hundred and twenty-seven is two thousand nine hundred and eighty-seven.

Fifty-seven are missing.

Mattias detected the difference in January 1835, working on the report, and the first thing he did was what any competent person does, which is to presume error.

He had the lists rechecked three times. He had them compared with the August census and with the parish baptismal registers. He had enquiries made in nine valley villages as to whether there were people from Karentin-Alta taken in by relatives and not declared, and there were — eleven such people were found, over fourteen months, and all eleven were struck from the difference.

Forty-six remained.

Forty-six people who entered a city in August 1834 and who did not die in the night, because the bodies were counted, and who did not come out by the south gate, because the column was counted at Vallium by name.

He spoke of it to nobody for sixteen months.

Not out of secrecy. Because there was nothing to be done with it and because to say it aloud, in a garrison of refugees lodged at Vallium with nineteen hundred people who had left everybody they knew inside, would have been a cruelty with no purpose whatever.

He carried the list of the forty-six names on the campaign of 1836. It is in the notebook, on a folded sheet, in his hand.

And on the 20th of May, on the first day inside the retaken city, he gave the order — the second order he gave, after the one about the purifiers — that a search be made.

They searched eleven days, with forty men, and this is the one subject on which the final report of Karentin-Alta is deliberately vague, and it is vague because Mattias wrote it vague and because it was the only time in his life he did so.

What is written is this, and it is all:

Of the census wanting, no whereabouts was established for forty-two. That of four was established, whose names are remitted separately to the Ordinary of Ferrum Alpina for such effects as he shall see fit.

There were found, in the inner ring and in the service galleries, signs of human habitation subsequent to November 1834, without duration or number being establishable.

No testimony was taken.

No further inquiry is recommended.

Signs of human habitation.

It is four words and Mattias chose them and argued them with his own chaplain for an afternoon, and what they cover — and this is what the officers of the Order who read the report at Karentin-Alta today do not grasp, because the report is framed and nobody frames things in order to read them — is that the city was occupied eighteen months and that somebody had been living there.

It is not known who. It is not known how many. It is not known in what condition, and the phrase in what condition is here the most important of the three and is the reason no testimony was taken.

The chaplain's opinion, which Mattias requested in writing and attached to the file, is six lines:

There is in canon law no presumption permitting us to treat as living one who has been eighteen months on the far side, nor any presumption permitting us to treat as lost one whom we have not seen.

I recommend that what was seen be recorded, that nothing be concluded, and that the four names be remitted to the Ordinary.

Concerning the forty-two, I recommend that they stand on a list of perpetual rogation of House Solar, without declaration of death.

It is little and it is what there is.

The list of perpetual rogation of House Solar exists.

It is read aloud once a year, on the 3rd of November, at Karentin-Alta, and it had forty-two names in 1836 and has, in 1890, six hundred and eleven, because other things happened in fifty-four years and because the Order, having once invented a category for people of whom one can say neither that they died nor that they live, went on using it.

It is the only liturgy of their own that the Knights of the Seal have.

It was not approved by Rome. It was made by an alpine chaplain on a May afternoon in 1836 to solve an administrative problem of forty-two names, and sixty-one years later it is the nearest thing to a soul that Order has.

Mattias wrote, in the notebook, on the night of the 31st of May 1836, the last entry of the campaign:

We retook the city in eleven weeks and it will go down in history as the first victory of a Military Order as an institution, and it is true, and the men deserve it.

I went in by a vent I ordered replaced in thirty-two and I came to look for forty-six people and I found four.

I shall write the report and I shall write it well and I shall not put in it what I think, which is this:

that the difference between this campaign and Marnheim was not courage, nor faith, nor competence, because those men had all three.

It was that I had nineteen purifiers and the Count of Reichshamn had one.

And I had them because a man in Aurum Roma spent twenty years building a machine that produces them, and because that machine hates me and I hate it, and because it works.

What matters is what Mattias did on the 20th of May, on the first day.

He ordered a search for the four purifiers.

There was nothing to find and he knew it, and he ordered the search all the same, and it was three days with forty men, and on the third day a sergeant found, in a casemate of the inner ring, a brass box with a hundred and eleven cartridges.

They were woken. All of them.

They were woken and had been there eighteen months and nobody had fired them, and that means a thing very simple and very hard to bear, which is that the four men made a hundred and eleven cartridges more than those who stayed managed to spend.

Mattias kept that box.

He carried it with him the rest of his life and it is today at House Solar and has never been opened since 1836, and there is an argument in the Order, running fifty-four years and already the subject of two opinions, as to whether those hundred and eleven cartridges still work.

Nobody has fired one to find out.

XX. The Vykhradian Knight

The man who carried the dispatches of Karentin-Alta to Aurum Roma, in June 1836, was a major of thirty-eight of the House of the Limes named Konstanty Vlassov.

Mattias chose him for three reasons and told him all three, which was his habit:

"You commanded the assault on the eastern vent and you are in the report by name. You are a Vykhradian, and I want the Curia to see a Vykhradian delivering this, because half the Curia still thinks the east is a place one goes to fetch people from. And you belong to no house."

Vlassov was the grandson of a general. His grandfather had commanded seventeen thousand Veheranics in a valley in 1786 and had died there, with nearly all of them, on the morning of the 17th of May, under a thing that is not described.

The grandson had been born a refugee in 1798, without land, without rent and without a surname worth anything outside the east, and had risen by career.

"I belong to none, Captain-General."

"I know. That is why you are going."

Vlassov stayed in Aurum Roma five weeks.

He was received, he was praised, he was promoted to lieutenant-colonel, and he spent thirty-three days doing what a field officer does in a capital, which is to wait in antechambers.

And in one of those waits — on the 11th of July 1836, in the outer corridor of the fifth courtyard, where one waits for the Congregation of Means — a very old blind man with an assistant sat down beside him and asked whether he was the officer from Karentin.

Vlassov said he was one of them.

And Aelius de Sárvar, seventy-four, clerk, said to him:

"You were at the aqueduct."

"I was."

"How many purifiers did you take on the campaign?"

Vlassov thought the question strange, because it was a quartermaster's question and the man was a clerk of a house of women.

"Nineteen."

"Nineteen," the old man repeated. And then, to the assistant, in an entirely different tone, like a man confirming a sum: "Nineteen into four thousand two hundred. One to two hundred and twenty."

He was quiet a while.

"In 1786," he said, "forty-three thousand men went into a valley and took nine."

"Nine purifiers?"

"Nine." Aelius turned his face towards him, in the imprecise way of the blind. "Do you know how many came back, Colonel?"

"Three hundred and seventeen," said Vlassov. "My grandfather was not one of them."

And there was then a silence of a quality that Vlassov described, many years later, in a notebook he wrote himself and which ended in the hands of a cardinal in 1866, with a sentence that stayed:

I understood that that blind old man had spent his life comparing things nobody had told him to compare, and that I was, in that corridor, the first person in forty years to whom he could say one of them aloud.

EPILOGUE — 1837

XXI. The Candles

Aurelianus I entered liturgical coma on the 2nd of February 1837 and died on the 11th, and in those nine days his hands wrote.

It is not a figure of speech. It is documented by fourteen witnesses and there are thirteen concordant accounts: the One was without consciousness, without reaction to stimulus, without speech — and the right hand wrote, continuously, in Latin, and when the sheet ran out it stopped until another was put under it.

He wrote two hundred and eleven sheets.

They are the Golden Canons, and they are the technical foundation of the Curia of the Hashmal for sixty years, and Krell was present on four of those nine days and saw it happen and wrote in the Memorial that it was the only thing in his life he could not reduce to numbers.

What he could reduce to numbers, and did, and is his part in that room, was this:

The two hundred and eleven sheets contained, between theology and rite, forty-one new technical instructions.

Forty-one.

The annual average of verifiable inscriptions of the White Vigil in the preceding decade was nine.

A man of eighty-two, in a coma, in nine days, produced four and a half years of revelation — and died, and the interface closed with him, and the next One had nothing resembling it, nor the next, nor any of them until 1890.

Krell wrote out the sum and then wrote, beneath it:

I do not know what this means and I distrust everybody who says he does.

I record only this: the thing was available, at that intensity, for nine days, and the condition for its being available appears to have been a man dying.

If that is the price on the table, then we have spent twenty-two years arguing about budgets and never knew what the currency was.

The conclave opened on the 3rd of March 1837.

It lasted nineteen days and three turns of black ash, and on the fifteenth day two thirds converged on a cardinal of sixty-one, prefect of the Congregation of Doctrine, a man of austere reputation and rigorist line, whom the Curia had known as a candidate for four years and who would take the name Reformatus.

They led him in silence to the Perpetual Vigil, as the rite requires, and the Cardinal-Electors went with him and did not speak.

Krell was in the third row.

They stood before the seven flames of the Archangels for nineteen minutes.

Not one flickered.

There is no good description of this and Krell attempted it three times in the Memorial and struck out two.

What he left was:

Nineteen minutes is a long time.

By the end of the fourth, everybody in that chapel already knew. By the end of the seventh, the candidate already knew. And we stayed twelve more because nobody has authority to declare a rejection before the end of the canonical time, and the canonical time is nineteen minutes because somebody in 1790 wrote nineteen on a sheet.

Twelve minutes with that man on his knees and seventy-two cardinals behind him looking at his back.

I never saw anything like it and I pray not to see it again.

And there is one thing more, and I write it because I am what I am and because I cannot leave a room without having counted what was in it.

At the thirteenth minute, I looked about me.

Seventy-two cardinals, and I had known their faces for twenty-three years, and I saw in eighteen of them the same thing that was in mine and that I had hoped did not show.

Relief.

Not for him. For us. Because that man had a rigorist line and because half that room had, in the preceding fourteen years, signed something a rigorist pontificate would have gone looking for.

Eighteen out of seventy-two is a quarter. A quarter of the College of Cardinals was, at the thirteenth minute, with a man on his knees in front of them, silently thanking God for a thing God had just done that happened to suit them.

I do not know what to do with this and I shall not take it to any confessor, because my confessor was in the second row.

The rejected cardinal returned to the condition of cardinal, as the canon requires, and was definitively excluded, and lived fourteen years more, and never again spoke in public about anything.

The conclave began again the next day and on the fourth turn elected a cardinal of fifty-four attached to the line of the Military Orders, of whom the only thing everybody said was that it had been he who defended the allocation of nineteen purifiers for Karentin-Alta in the Council of Seven when the Congregation hesitated.

He was led to the candles on the 22nd of March.

The flame of Camuel flickered once and steadied.

Hashmalian I.

Krell walked back to his office that night and it is the only time in his whole work that he writes about what he felt, and he writes it badly, which is a sign:

We elected a Pontiff on eleven weeks of alpine campaign.

This is not impiety. The candle flickered; Heaven recognised him; I saw it and I believe what I saw, and I shall to the end of my life.

But I was in that chapel and I know how to count, and I know that if the city had not been retaken in May of thirty-six, that man would not have been on that list in March of thirty-seven.

And therefore Heaven recognised a man who arrived there by a military campaign that I financed with nineteen licensed men whom I authorised with my shaking signature in September of last year.

I do not know what to do with this sentence. I write it because it is true and I put it in this book that nobody is going to publish.

Twenty-two years. I began with a decree of fourteen articles without one number in it.

I end knowing that the numbers reached everywhere, including the places where I thought they did not go.

Hashmalian I sent for Amandus Krell on the 4th of April 1837, thirteen days after the election, and it was the first audience on means of the pontificate.

Krell was sixty-nine and had served Aurelianus I for twenty-three years, and he went into that room with the certainty — unformulated, but entire — that he was to be retired.

He prepared for it. He took the folder.

The folder was called To Do and he had written that name in pencil on the cover in 1815 and had never corrected it, and in the twenty-two years since it had ceased to be a folder and had become four volumes and an index to them.

He took them all. If he was to be retired, he would hand them over in person.

The new One received him seated, which was the first difference and which Krell recorded.

He was fifty-four and had a reputation as a practical man. He was a cardinal attached to the line of the Military Orders and was, it was said, the man who had defended the nineteen purifiers for Karentin-Alta when the Congregation hesitated — and Krell, who knew exactly who had hesitated in the Congregation because it had been himself, never denied that version and never confirmed it either.

Hashmalian I was direct.

"Eminence. How old are you?"

"Sixty-nine, Most Holy Father."

"And how many years in the Congregation?"

"Twenty-three in November."

The One nodded.

"I am going to ask you a question and I want the answer you would give your confessor and not the one you would give me." He laid his hands on the table. "If you could undo one thing of these twenty-three years, what would it be?"

Krell had the answer.

He had had it for eleven years and nobody had ever asked him for it, and he understood, with a kind of vertigo, that the most improbable thing that can happen to a man of his age was happening, which is to be asked, once, the right question.

And he understood at the same time, in the same second, that the true answer would not serve — because the true answer was the reform of 1824, and to say it meant asking the One to dismantle the Reformed Inquisition, and that was not going to happen, and the only effect of asking would be to end his career in an afternoon and change nothing.

So he gave the second.

"The fourth determination of the order on the table, Most Holy Father."

"I do not know it."

"It is of 1826. It is the one that orders the price posted in the square and that keeps the columns of abatement in a reserved annex." Krell opened the second volume, found the place, and read: "The coefficients of application are not matter for publication."

"And ought they to be?"

"They ought. And it is not out of virtue, Most Holy Father, it is out of arithmetic." He closed the volume. "Königsmark pays two hundred and forty per cent above Ferrum Alpina and knows it pays more, but does not know how much nor why, and has spent eleven years building theories about it. The theories are all worse than the truth."

"What is the truth?"

"That we charge the North dearly because the North purifies at home and we want that to cost them their purifiers instead of costing us ours. It is cold and it is defensible and I defend it. But in secret it turns into something else: in secret, a man of Königsburg who pays that price has only two possible explanations, which are either that Rome hates him, or that Rome is lying about what the thing costs."

Krell raised his head.

"We do not hate them and we are not lying. But we spent eleven years letting them choose between two false things, and they chose, and the Schism of the North is deeper today than it was in 1826, and part of that is the sheet I did not let be posted."

Hashmalian I was silent a while.

Then he said a thing Krell had not expected.

"Eminence, you have not answered me."

"Most Holy Father—"

"I asked you what you would undo. You have given me a foreign-policy error of 1826, which is a true error and which I shall have reviewed." The One leaned forward. "And you gave it to me because it is the second on the list. I want the first."

Silence.

"Eminence, I have read the minutes of 1824. All of them. It took me nine days." Hashmalian I did not raise his voice at any point of that audience and Krell recorded that too. "I know who asked for the summary tribunal and I know why, and I know the request runs to two pages and does not mention heresy once. I know you wanted exclusivity of service and that they gave you the tribunal, and I know you wrote the recitals of the first article and did not write the other two."

Krell said nothing.

"How many proceedings did the Reformada open last year?"

"Three hundred and forty-one, Most Holy Father."

"And in 1825?"

"Sixty-nine."

"In eleven years it has multiplied by five, and the population of obedience has grown sixteen per cent." The One sat up. "You know what that means and so do I. It is not catching more heresy. It is looking deeper, because a house with a budget and without public trial has nowhere to stop."

And here Amandus Krell did the one thing in that audience he regretted, and he regretted it for four years and wrote it down.

He defended it.

Not out of loyalty to the Inquisition, which he detested. For a worse and deeper reason, and he recognised it while he was saying it and could not stop:

"Most Holy Father, if Your Holiness opens the trials, the records become consultable, and the records contain what the accused did." He heard himself saying it and went on. "Half of them worked. It is in the files. It is verified by investigation, with testimony and with expert examination, because the Reformada is competent and verifies everything."

"I know."

"Then Your Holiness knows that to publish a trial for functional heresy is to publish the proof that the heresy works."

Hashmalian I looked at him for a long time.

And then said the sentence that closed the audience, and Krell wrote it in the Memorial that same night, word for word, with the note that he had heard it standing:

"Eminence, you have just given me, in three sentences, the reason this house cannot be honest. And you gave it to me well, and I have no answer, and it keeps me awake."

The One rose.

"Stay in the Congregation. I am not going to retire you; I have sixty-nine years of your work in four volumes and not one man who knows how to read them. But I want one thing, and I shall ask it once only."

"Say it, Most Holy Father."

"Write. Everything. The three truths, not the two."

Krell did not understand at once and the One saw that he did not.

"Not to publish. To exist." Hashmalian I went to the window. "Eminence, I shall govern this twenty years or two, and whatever comes after will inherit a machine nobody designed on purpose and that nobody can any longer explain entire. You are the only living man who saw it made piece by piece."

"Nobody is going to read it."

"Probably not." The One turned. "But if nobody writes it, in sixty years this house will believe it was born like this, and a house that believes it was born like this cannot change anything, because it no longer knows where it put the joints."

The Memorial on the Means was begun on the 9th of April 1837, five days after that audience, and Krell wrote it in three years and seven months, at night, in three volumes, with his hand steadily worsening.

He did not publish it. He left express instruction that it was not to be published while the table remained in force, which is a condition that never came about and that he knew would never come about, and which is therefore a man of seventy's way of saying never without saying it.

It is in the Archive of Means. It was consulted, between 1841 and 1890, by fourteen people.

One of them was a provincial canon of thirty-nine, in 1860, who was researching the history of the Congregations and who took eleven days to read the three volumes and who came away with what he himself described, many years later, as the only political education I ever received.

He was called Sebastien Vergesa.

He made his whole career in archive and commission. He reached cardinal without ever heading any State, and in 1866 he received from the hands of a retired Vykhradian marshal a package in oilcloth with an index copied out at night by an assistant of nineteen, and he did not ask where it came from because he already knew.

XXII. The Index

The last thing in this book takes place on the 4th of August 1837, in a corridor, and lasts two minutes.

Konstanty Vlassov was in Aurum Roma again — he had come back in July for the parade confirming ranks under the new pontificate — and was leaving the fifth courtyard when he was stopped.

It was the blind clerk's assistant. Doroteia, twenty-two.

She said that Master Aelius asked whether the colonel could spare a moment.

Aelius was sitting on a stone bench in the sun, which was where he spent his mornings since he had stopped seeing, and he had on his knees a package wrapped in oilcloth and tied with string.

"Colonel."

"Master clerk."

"You remember me."

"I remember the sum," said Vlassov. "One to two hundred and twenty."

Aelius smiled, and it was the only time Doroteia saw him smile that year.

"I want to give you a thing and you cannot refuse it, because if you refuse it I have to carry it back and I have nowhere to put it."

He held out the package.

"What is it?"

"It is the index."

It was the index.

Aelius de Sárvar had, over forty-seven years, kept a personal notebook in which he recorded, for each transcription, the date, the Blanca, the classification, and — and this is what made it a thing that ought not to exist — the cross-references: which transcription repeats which, with what variation, in which direction it corrects.

It was the first rule applied four thousand six hundred times.

Sofia Kranz and the four assistants had used it for nine years to write the Hashmal Syntax. Without that notebook, the Syntax would have taken thirty years or would not have been written.

The notebook was the property of the Vigil. The copy Aelius had in his hands that morning was authorised by nobody, and had been made over four years by the assistant Doroteia, at night, from dictation, and Doroteia knew exactly what she was doing and was twenty-two and did it anyway.

"This is irregular," said Vlassov.

"And now in two."

"Now in two."

Vlassov looked at the package and did not take it.

"Why me?"

And Aelius de Sárvar gave the answer he had prepared, because he was not a man to improvise such things, and which Vlassov wrote in his notebook that same night because he understood, on hearing it, that he would want to remember the exact words:

"Because you belong to no house. Because you are of the east and the Curia does not look at the east. Because you rose by career and have nothing to protect but your rank. And because you asked me how many came back from a valley in 1786 and told me your grandfather was not one of them, and I understood that you are one of the people who count."

He was still a moment.

"There are few of us, Colonel. There are very few of us and we do not know one another, and we spend our lives making sums in different places without knowing that other people are making them."

Vlassov took the package.

He kept it thirty-one years, carried it through the Age of the Forge, commanded at Cherniwicz in 1859 with it in a chest, refused in 1861 the elevation to Lord of Salvation with a sentence that stayed — a man who commands everything stops seeing something — and in 1866 handed it over, with a notebook of his own that nobody had asked him for, to a cardinal of forty-five who had made his whole career in archive and commission.

Aelius de Sárvar died on the 9th of March 1841, at seventy-nine, with four thousand six hundred transcriptions to his name and not one signed, because a clerk of the Vigil does not sign.

Amandus Krell died on the 2nd of December of the same year, at seventy-three, and the Memorial on the Means remained unpublished by his express will and is in the Archive of Means in three volumes with a note on the cover of the first, in his hand, which says:

Not to be published while the table is in force.

It is in force. It has been revised eleven times and never revoked.

Mattias of House Solar died in 1851, at sixty-one, at Karentin-Alta, and directed that he be buried without a name, as the vow requires, and House Solar complied, and the stone says only Captain-General, 1830-1848.

Somebody, at some point in the sixties, cut beneath it, without authorisation and with a poor chisel, the sentence of June 1820:

They counted the bullets and did not count who makes them.

The Order debated twice whether it ought to be effaced.

It did not efface it.

XXIII. What Remained

It is worth saying what became of the things, because the people ended and the things did not.

The key.

The Archive of Means has three keys and Amandus Krell had one. At his death, in December 1841, it passed to the next Prefect, who was a canon of forty-four whom Krell had trained and who was called Vitale — the same man who in 1833 had made the fourth audit of Ferrumkar and written the sentence about domestic use on page eleven.

Vitale was Prefect of the Congregation of Means for nineteen years and gave nobody the census form of 1819 with the upright small writing in the field for the name.

He kept it for the same reason Krell had kept it, and he kept it with one difference: he knew, from the beginning, that he was keeping it for somebody.

In March 1852, when the Broken Glass took Ferrumkar and held it forty days and purified ammunition with no cleric at all, and it worked, the Curia had to decide in nine days a thing it had not thought about in thirty-one years, which was what to say.

And the Reformed Inquisition asked the Congregation of Means for everything there was in the archive on purification by the unordained.

Vitale sent the opinion of 1823 — the forty pages of casuistry that distinguish sacramental purification from waking in extremity, and that establish, in small print and so as to be able to deny it in large, that the thing works without a priest.

He did not send the form of 1819.

He did not send Mattias's four pages of 1824 about the eleven and the fifty, which were in the same bundle and which would have told the Reformada, in a single document, that the phenomenon had been observed, measured and described thirty-one years earlier, by an officer, in a house of the Order, with nobody's authorisation.

He wrote, in the register of remittance, the justification, and it is nine words and it is the most Krell-like thing a disciple of Krell ever wrote:

What was asked is remitted. What was not asked is not.

The four pages.

They stayed in the bundle. They were found in 1855 by a canon of the Curia of the Hashmal who wrote across the top, in pencil, this ought to have been published in 1824, and did not publish it.

They were found again in 1871, by an archivist who catalogued them correctly and without comment.

And in 1889, a year after the Emergency Conclave, an inquisitor of forty-eight who was reconstituting the series of silences of the Vigil asked the Archive of Means for everything there was on observation of purification outside licence prior to 1852, and received three documents, and one of them was the four pages.

He read them in an afternoon.

What he did next is not in this book, because this book ends in 1837, but it is somewhere else, and his name was Reinhardt Voss, and the sentence of Mattias's he underlined in pencil — the only underlining he ever left in an archive document in twenty-six years in the trade — was:

The fifty were straining. The eleven, at a certain point, stopped straining and came simply to be present, and that was where it happened.

Voss had no visible personal faith. He believed in the method.

The table.

The price table of the 14th of January 1826 was revised eleven times between 1831 and 1878 and was never revoked, and the coefficients of application were never published.

Hashmalian I ordered the decision on the reserved annex reviewed in April 1837, as he had promised Krell at the audience, and the review took two years and concluded, in 1839, that publication of the coefficients was inopportune in the present state of relations with the schismatic Churches.

It is a perfect sentence. It says that publication follows when relations improve, and relations with Königsmark and with Brytonia never improved between 1839 and 1890, and therefore the condition never came about, and therefore the decision of 1826 remains in force in 1890 by virtue of a review that ordered it reviewed.

Krell died in 1841 convinced the review was going to happen. It is the last optimistic thing known of him.

The frontier provision.

Instituted in January 1829 to break a political bloc of seven Cardinals-State, indexed to the number of settlements and not to revenue, revised every ten years.

It survived everything. It survived five pontificates, the industrialisation, the whole Age of Empire and the Stagnation, and in 1887 it was the reason the diocese of Pradel-East had, on paper, four licensed purifiers for a sector of nine villages.

Four on paper. One and a fraction in stone.

The provision went on being paid. What it bought ceased to exist, and the mechanism Krell invented in an afternoon in 1828 went on working perfectly for sixty years distributing a thing that was no longer there, and it must be said that this is not irony: it is what happens to every good institution, and the alternative — no provision at all — would have been worse, and the withdrawal of twenty-two kilometres of October 1887 would have happened in 1879.

The Syntax.

The seven rules were published in 1839, under the name of Sofia Kranz, and they are the foundation of all Vigil doctrine on the continent.

The seventh rule — the absence of inscription is not recorded as absence, it is recorded as a shift — was written to protect girls of nineteen and succeeded, and made the silences uncountable for fifty-seven years.

Sofia Kranz died in 1848, at sixty, Mother Vigilant for twenty-two years, and never knew.

The table of a hundred and nineteen lines went on being filled in after her and stands in 1890 at four hundred and six lines, and the column she added in pencil at the end — the one that counts the years the two-shift limit gave back to each woman — was inked in by the next Mother and was never removed.

In 1884 the limit fell, in a winter, by need of the front.

The column went on being filled in. From 1884 the figures that go into it are negative, and Mother Doroteia, who took charge in 1868 and who had been the assistant of nineteen who read aloud to a blind clerk what he had just written, wrote them down all the same, every one, one by one, for six years.

The index.

Aelius de Sárvar transcribed four thousand six hundred times in fifty-one years and signed none of them, because a clerk of the Vigil does not sign.

The notebook of cross-references he kept alongside — the first rule applied four thousand six hundred times, the thing without which the Syntax would have taken thirty years or would not have been written — exists in two copies.

The other was handed over in a corridor, on an August morning in 1837, to a Vykhradian lieutenant-colonel of thirty-nine who belonged to no house, and afterwards crossed thirty-one years and a chest and a campaign and two men, and in 1866 came into the hands of a cardinal who had made his whole career in archive and commission.

In 1887 that cardinal entrusted a sealed box to a sergeant of the Limes with twenty-eight years' service, with instruction not to open it.

In 1890 the sergeant is fifty-one, the box is still shut, and he has not opened it.

And the sheet.

There is one thing that did not survive and it is worth saying so, because this book is about sums and a sum that is destroyed is also a fact.

In February 1827, alone, with the door shut, Amandus Krell estimated on a half-sheet how many people in the territory of obedience were doing what Pankrác Jelen did, and the figure that came out was between four thousand and eleven thousand, and he burned the sheet.

Nobody made that sum again in sixty-three years.

Not because it is forbidden — it is not forbidden, there is no canon prohibiting it, any canon of the Congregation could redo it in an afternoon with the public data.

Because nobody commissioned it.

What remained, then.

There remained a table nailed to a door in 1826 that still governs the price of everything.

There remained a rule of withdrawal written by a man of twenty-eight in a shepherd's hut and which has since forbidden four thousand captains to stay where they should not.

There remained a frontier provision that distributes purifiers to dioceses that no longer have any.

There remained a list of perpetual rogation that began with forty-two names of people of whom one can say neither that they died nor that they live, and which today has six hundred and eleven.

There remained a column in pencil, inked in, that counts years of life given back to women of nineteen, and which since 1884 counts backwards.

And there remained three volumes in the Archive of Means, with a note on the cover of the first saying not to be published while the table is in force, consulted by fourteen people in forty-nine years.

Not one of the three men in this book changed anything.

The decree of 1815 had fourteen articles and not one number in it. By 1837 the Clerical Order had budget, doctrine, recruitment, a table, a tribunal, three houses of formation and five thousand six hundred licensed purifiers, and a Pontiff elected on eleven weeks of alpine campaign.

Between the one thing and the other there were nine thousand men on a plain to whom nobody gave forty priests, a farrier of twenty-two who understood the arithmetic before the cardinals did and was told that he knew nothing of blessings, a miller of Solnica who blessed salt on Thursdays and whose daughter went on doing it, and a hundred and nineteen girls of whom four were alive in 1835.

The Church was not validated by having a head. It came by a head for having been validated, and then spent twenty-two years building the body, and the body was built by people who were solving small problems on a Monday.

That is how empires are made and it is the only way there is.

Let God judge us by the right sheet.