I. The Roll on the Table

The translator opens the roll: what survives of Agrippina, what does not, and why one sentence in Tacitus is the reason for all of this.

Show Notes

There is a sentence in Tacitus that nobody reads. It sits in the middle of one of the most famous passages in Latin literature, and it says that he found a private scene between the emperor Tiberius and Agrippina’s mother in a book written by Agrippina herself.

She wrote a memoir. Pliny the Elder read it too, and cited it. Then it stops. This first episode is the case for the book that follows: the borrowed head in the Centrale Montemartini, a vanished dedication at Nîmes read from its nail holes, and what it means that everything we possess is what somebody chose to keep.

Source: front matter.

Free companion guide — a map of the novel, the family tree, a chronology and a pronunciation guide: pbmills.com

A work of fiction. Julia Agrippina left no memoir and there is no manuscript.

What is Augusta: The Lost Commentaries of Julia Agrippina?

AUGUSTA is a novel by P.B. Mills, read in full. It is presented as the recovered private commentaries of Julia Agrippina — daughter of Germanicus, sister of Caligula, wife of Claudius and mother of Nero. The manuscript does not exist. Tacitus really did read her memoirs, and really did cite them once, and they really are lost. Everything else here is invention.Twenty-seven episodes. A new one each week. Series guide, maps, and family tree available for free download at pbmills.com.

I. The Roll on the Table

Translator: Augusta. The Lost Commentaries of Julia Agrippina. Episode one. The Roll on the Table.

Prologue.

The Woman with the Borrowed Head.

If a name is what your parents call you, then Agrippina the Younger's name was Julia.

Agrippina came down to her from her mother, and to her mother from Marcus Agrippa, who won Actium for Augustus and was never allowed to forget that he had done it as a servant. It is a soldier's name worn sideways. It is also the name history filed her under, which is why we have her under it still.

Julia was the other one. Julia was the family. Julia was what the house said when it wanted her and not her mother, and what her father said, and what her women said in the dark when she was small, and what is written across the top of the first recovered leaf of the manuscript that has taken four years of my life. She was Julia Agrippina all her days. The Younger is a librarian's addition, hung on her centuries later by men who needed to keep two dangerous women apart on a shelf and could not be troubled to use both their names.

I want that on the first page, because everything that follows is a quarrel with the filing.

I am not a classicist. I run companies, and I have spent my working life reading contracts, which is a discipline that trains you to notice when a document is telling you something it did not intend to tell you. What I had, before any of this, was six winters of reading Roman history at night, which is a hobby and not a qualification and which I would not mention except that it is how I came to read one sentence properly.

Tacitus, Annals, Book Four, chapter fifty-three.

He is describing a scene between the emperor Tiberius and Agrippina's mother. It is private. It is a frightened woman asking a frightened old man for a husband and being refused in Greek, and it is one of the best things in the Annals, and I had read it perhaps a dozen times without once reading the sentence that comes after it.

Having told it, Tacitus does something he almost never does. He tells you where he got it.

Id ego, a scriptoribus annalium non traditum, repperi in commentariis Agrippinae filiae, quae Neronis principis mater vitam suam et casus suorum posteris memoravit.

This, not handed down by the writers of the annals, I found in the memoirs of her daughter Agrippina, who, as mother of the emperor Nero, recorded for posterity her life and the disasters of her family.

She wrote a book.

Not a letter. Not a note. Commentarii, the word Caesar used for his own account of his own wars, the memoir a Roman of rank leaves so that the history men cannot invent one for him. A full account of her life and of what happened to her family, composed by the daughter of Germanicus, the sister of Caligula, the wife of Claudius and the mother of Nero, who was born in an army camp on the Rhine and died on the shore at Baiae with soldiers standing over her.

And Tacitus, the greatest of them, the one who set the terms by which she has been despised for two thousand years, sat in a library and read it.

He is not alone. Pliny the Elder, in the seventh book of the Natural History, records that Nero was born feet first, and says he has it from the memoirs of his mother Agrippina. Pliny died at Vesuvius in the year 79. She had been dead twenty years. The book was on shelves and men were quoting it.

So it existed. Two independent Romans, working generations apart, held it in their hands and cited it by name.

And then it stops.

There is no third citation. There is no summary, no epitome, no papyrus scrap, no line quoted by a grammarian to illustrate an unusual ablative, which is the way half of Latin literature survives. There is no medieval library catalogue that lists it. The most quotable woman in Roman history wrote a book about herself, two of the most careful authors in the language used it as a source, and after that it leaves no trace of any kind.

Books die all the time. Most of them do. They die of damp and mice and of nobody caring enough to recopy them when the roll wears out, and that is the ordinary, unremarkable, entirely sufficient explanation, and any classicist will give it to you in about nine seconds. Four of them have given it to me.

I believed it for about a month, and I still believe it, and this book is not an argument against it. It is what happened when I went on reading anyway.

You do not go to see Agrippina in Rome. You go to see what is left of her, and what is left of her is in a power station.

The Centrale Montemartini stands on the Via Ostiense, south of the old walls. It was built to give Rome electricity in 1912 and it did that job until it stopped doing it, and the city, having a decommissioned generating hall and several hundred pieces of ancient marble in storage, made the only decision anyone could love it for. The gods are in there with the machines. Diesel engines the size of railcars, painted deep red, and along the wall behind them, in a row, the Republic and the Empire in white and grey and black stone, lit like an operating theatre.

She is standing among them and she is the wrong colour. That is what takes you first. She is basanite, a hard dark stone the Egyptians favoured, nearly black in that light and green when you get close, and every other figure in the hall is pale, and the effect at forty feet is of a woman in mourning at a wedding.

Inventory S 1882. First century. She was pulled out of the ground on the Caelian hill in 1885, when they were digging foundations for the military hospital, and she is thought to have stood in the Temple of the Divine Claudius, which is to say the temple of the husband she is generally supposed to have poisoned. She is posed in prayer. Her arms are up. The body is quiet and beautifully done and it is unmistakably a body somebody paid a great deal of money for.

The head is a plaster cast.

It is a cast of another portrait, the one in Copenhagen, set on the ancient shoulders by a curator at some point in the last hundred and forty years so that the statue would have a face. Nobody hides this. It is on the label. I stood in front of it and read the label three times and then stood there a while longer with the engines humming behind me, because I had come four thousand miles to look at the most complete surviving image of Julia Agrippina and what I was looking at was a real woman's body with a copy of somebody else's head screwed onto it.

I know how that reads. A man who has flown to Rome about a sentence will find meaning in a museum label. But I did not invent the label and I did not put the head on the statue, and I would ask anyone who thinks I am reaching to consider what we actually possess of this woman.

We have Tacitus, who was a boy when she died. We have Suetonius, born after. We have Cassius Dio, writing a hundred and fifty years later in another language. Three men, none of whom were in the room, all of them working from other men who were also not in the room, every one of them inside a tradition that had decided in advance what an ambitious woman was for.

We have her body, and we have borrowed a head to put on it.

There is a better likeness upstairs at the Markets of Trajan, a portrait from the Forum of Trajan with nothing borrowed in it: the hair she made famous, two rows of tight curls over the brow, corkscrewed, absolutely rigid, a style that took a slave two hours a day and told every woman in the empire who set it. It is not a soft face. It is not a hard one either, which is the thing the histories never allow her. It is the face of somebody listening to a number she does not like.

That autumn I drove from Vienne to Nîmes, which is a hundred and fifty miles of the ordinary argument about what survives.

At Vienne there is a complete Roman temple standing in a square with a brasserie on one side. Six Corinthian columns across the front, early first century, raised to Augustus and to his wife Livia, who was Julia Agrippina's great-grandmother and, by every account we have, the woman whose career she studied. The thing you cannot get over is that it is there: not a footprint, not three columns and a diagram, the building, with its walls, with a bus stop near it. It is there because it was made into a church, and kept its roof for eleven hundred years because a congregation needed somewhere dry to stand on Sundays, and was afterwards a Temple of Reason, a commercial court, a museum, and the town library until 1852.

At Nîmes there is the Maison Carrée, the best preserved Roman temple in the world, and it is not a temple to a god. It was raised to two boys, Gaius and Lucius Caesar, the grandsons Augustus adopted and groomed and buried, whose deaths left the succession to Tiberius and, by that accident, produced everything that happens in this book.

Except that by the time anyone modern looked at it, there was no dedication. The letters had been bronze, pegged into the stone of the frieze, and bronze is worth money, and at some point in the long centuries when the building was a house and a stable and a church and a granary and the town archive, somebody prised every letter off the front and sold them for the metal.

What was left was holes.

In 1758 a Nîmois named Jean-François Séguier, a botanist by preference and an epigrapher by obsession, stood in that square and read the holes. He measured them. He worked out which spacings belonged to which letters, how many pegs an M takes and how many an I, and he reconstructed the entire lost inscription from the pattern of the wounds the missing text had left in the stone. To Gaius Caesar, son of Augustus, consul. To Lucius Caesar, son of Augustus, consul designate. To the Princes of Youth. It is carved into the frieze today, and it is his.

He read a text that was not there by studying the shape of its absence.

Two buildings, a hundred and fifty miles apart, and between them the whole of the discipline. Everything Roman still standing in Europe is standing because somebody, at some point, decided to keep it. That is not a slander and it is not a conspiracy; it is a fact about who owned the roofs and the walls and the ink and the parchment and the labour between the sixth century and the fifteenth. The monasteries are the reason we have Virgil. They are the reason we have Tacitus at all, and we have him by a thread: the first six books of the Annals come down to us in one manuscript, one, out of the whole of the Middle Ages, and it sat at Corvey in Germany until the sixteenth century, and if it had gone into a fire the way its neighbours did we would not have her name in any depth at all.

Everything we possess is what somebody chose to keep.

Which means that everything we do not possess is, at least sometimes, what somebody chose not to. But it means that much less often than a man in my position would like, and I want to say plainly, here, that I do not think anybody suppressed this book. I thought so for a while. It is a better story and it is almost certainly wrong.

Parchment is skin, and skin is expensive.

When a monastery needed a book and had no hide to spare, a monk took a volume nobody in that house wanted any more, and he scraped it, and washed it, and pumiced it, and wrote the new book across the old one, frequently turning the leaves ninety degrees so that the two texts run at right angles through each other. The old text does not go. It cannot. The iron in the ink has bitten into the skin, and the skin remembers. Under the right light, in the right part of the spectrum, it comes back up out of the page like a bruise.

This is not speculative and it is not rare. It is how we have most of Cicero's De re publica, scraped clean in the seventh century, covered with Augustine on the Psalms, and read again in 1819 by a Vatican librarian with sharp eyes. It is how we have the Archimedes Palimpsest, sold at Christie's in 1998 in appalling condition with forged gold pictures painted over it, imaged over twelve years at three synchrotrons and a Baltimore museum, and found to contain two treatises no living person had read since the Middle Ages, including the one in which Archimedes does calculus eighteen hundred years early.

So here is the duller and worse thing I finally understood, at a kitchen table in Nashville with a legal pad, and it is the thing I would ask anyone who doubts this book to sit with.

Nobody has to burn a book in a world where parchment costs more than labour.

Burning is loud. Burning makes martyrs and rumours and lists. You do not need any of it. You take the volume off the shelf, and you give it to a house that needs a psalter, and in eleven months it is a psalter. No order has to be given. Nobody has to have an opinion about the contents. It was old, nobody in the house could read it, and they needed the skin.

What money is actually good for, in this field, is imaging.

There are somewhere north of eleven thousand manuscripts in French ecclesiastical and former ecclesiastical collections that have never been examined with any instrument more sophisticated than a good lamp, and the figure for Europe as a whole is a great deal worse, and nobody considers this a scandal because there has never been any money. In the second year I stopped writing letters to people who study this and started paying for a survey: multispectral capture across a regional collection, chosen because its catalogue was made in 1874 and never revised, and because the two curators there had been asking for exactly this for nine years and had been told nine times that there were no funds.

The survey found four palimpsests in two hundred and sixty volumes, which is about the expected rate.

Three of them are of no interest to anybody outside the discipline: a fragment of a legal compilation, thirty leaves of a grammar, and a set of estate accounts that will make one specialist very happy. I am told this is a good haul. I am told most surveys find nothing.

The fourth is a codex of two hundred and forty-one parchment leaves whose upper text is a passionary, a collection of the lives and deaths of virgin martyrs, written in a good clear hand in the second half of the thirteenth century.

Women who chose to die rather than yield.

It is written over a woman who chose neither. I have wanted very badly, at times, for that to have been a joke, because a joke would mean a mind in there with me. It was not a joke. It was on the shelf, it was old, nobody could read it, and the house needed a passionary.

The imaging took a great deal longer than anyone told me it would and cost roughly what I had been told a book like this would earn, which is a joke I am not tired of yet. Sixteen wavelengths, ultraviolet fluorescence, raking light, and X-ray fluorescence mapping for the iron where the ink was too far gone for the rest of it. Then processing. Then, for a long time, argument, because two people can look at the same recovered stroke and see two different letters, and when they do, both readings go in the file.

The undertext is in a Latin that made two specialists sit down.

It is not thirteenth century and it is not tenth. The consensus of the people who have examined it, none of whom will yet put their names to it and all of whom I understand completely, is that we are looking at a late antique book, uncial, sixth century or a little before, copied from something considerably older, at close to the last moment in European history when such a thing could still have been copied at all.

And it opens, in the third line of the first recovered leaf, with a woman putting her name down.

Ego sum Iulia Agrippina, Germanici filia.

I am Julia Agrippina, daughter of Germanicus.

What follows is a translation, and a translation is an argument. Every choice in it is contestable and a great many of them have been contested, loudly, by people I am paying. Where the manuscript is damaged I have said so. Where two readings are possible I have generally given the one that makes her look worse, on the principle that I am the last person on earth who should be trusted to flatter her.

You are going to be told that this book is a fraud. You should expect that, and frankly you should want it, because the alternative is a world where a claim like mine is accepted on a Tuesday. The codex is where it has always been, in the collection that owns it, and the imaging data is with the people who made it, and every stroke of the recovered text will be argued over for the rest of my life and, I would guess, for a good while after it. That is not a threat to what I have done. It is the only thing that could ever make it worth having done.

What I would ask, in the meantime, is this.

For two thousand years we have had one story about this woman and it was written by her enemies, in her enemies' language, for an audience that had already decided what to think. She is the poisoner. She is the mother the son had to kill. She is the cautionary tale a Roman father tells at dinner, and every historian since has been working from a portrait cast by somebody who never saw her face.

Here she is with her own head on.

She is not sympathetic. I want to be very clear about that, because I have watched three separate readers arrive expecting a vindication and leave shaken. She does things in these pages that she does not apologise for, that I could not defend, and that she does not ask me to defend. She is exact about money and rooms and rank and who stood nearest the emperor, and she is a liar about innocence, and she is entirely aware of the difference, and she says so, which is the single most disarming thing about her and the reason I could not put it down.

She is also, on the evidence of her own hand, one of the two or three most capable political minds of the first century, and she has been read for two millennia as a woman who slept her way up a staircase.

One last thing, and then I will get out of her way.

I want it to be genuine. I have wanted that since the first month and I have never stopped wanting it, and a man who wants a thing that badly is the worst possible judge of whether he has it.

I have put that sentence here, at the end, where it will be the last thing anybody reads before her voice starts, because leaving it out would have been the easiest edit in the book.

She taught me that. Put down both versions, where you can see what you are doing, and then choose the one you are going to live with.

P.B. Mills

Nashville, Tennessee

A Note on the Text.

The manuscript. The codex is a passionary of two hundred and forty-one parchment leaves, written in the second half of the thirteenth century over an earlier book that had been scraped, washed and cut down for reuse. The undertext runs at ninety degrees to the upper writing and is not visible to the unaided eye except at four places along the outer margins where the scraping was hurried. It has been recovered by multispectral imaging and, in the worst passages, by X-ray fluorescence mapping of the residual iron. Roughly seventy-one percent of the original text has been recovered to a standard the imaging team was willing to sign. A further twelve percent is partial. The rest is gone, and where it is gone I have said so in the margin rather than repairing the sentence for the reader's comfort.

The copy, and what its scribe did. The undertext is not the author's manuscript. It is a copy of a copy of something, made in uncial in the sixth century or a little before, and no leaf of it was ever touched by anybody who knew her. The reader needs that fixed in mind, because it governs everything the apparatus of this book is entitled to claim.

What makes the copy usable is that its scribe was a diplomatic copyist, which is rarer than it sounds and which he tells us himself. At the foot of the last surviving quire there is a colophon of four lines:

Haec ita transtuli ut inveni. Quae alia manu in exemplari scripta erant, signo apposito notavi. Quae eadem manu postea addita sunt, suo signo. Nihil correxi, nihil omisi. Lector aequus sit.

I have copied these things as I found them. What was written in the exemplar by another hand, I have marked with a sign. What was added afterwards by the same hand, with its own sign. I have corrected nothing and left nothing out. Let the reader be fair.

He kept his word, so far as anyone can check him, and the checking is not trivial: there are places where his fidelity has preserved an obvious error that any ordinary copyist would have silently repaired, and one place where he reproduces a sentence that breaks off in the middle of a word and notes in the margin that it broke off in his exemplar too.

His two signs are a small a for alia manu and a small p for postea, and they appear in the margin of the undertext three hundred and sixty times between them. Where his exemplar carried its own description of a change of hand or of ink or of the size of the writing, he copied the description across as he found it. Where a passage had been cancelled he did not omit it; he reproduced it and marked the cancellation with a line of dots beneath the letters, which is how deletion was shown, and which I have rendered throughout as a struck line so that a modern reader sees at a glance what is going on.

The two hands. The a passages are the work of a second person in the household, a copyist who kept a parallel set of the same events, and they are in this book wherever the text says so. The p passages are eleven in number. They are revisions to entries already written, added at a considerably later date, and in every case they know something the entry they sit beside could not have known.

I want to be exact about what can and cannot be said here. Nobody has looked at Julia Agrippina's handwriting. Nobody living ever will. The claim that the eleven later additions are her own is an inference, and it rests on three things: the sixth-century scribe's own distinction between another hand and the same hand later, which he did not invent and had no reason to fabricate; the internal evidence of the additions themselves, which argue with the entries above them in the first person and in her voice; and the fact that a forger with the wit to construct this would have had no motive to mark eleven passages as the author's own second thoughts and then let half of them make her look worse. I have set them in italic and marked them later. She is arguing with herself across a decade. She loses about half the time.

What it is. It is not the published book, and that is the first thing the reader needs, because it explains nearly everything strange about what follows.

A Roman of rank did not compose in one pass. She dictated or wrote onto wax, which is cheap and erasable and meant for the stage where you are still deciding; the wax was worked up onto a roll by a secretary; the roll was corrected, cut, argued with, and only then copied fair for anyone else to see. Tacitus and Pliny read the fair copy. They read the version she was willing to have read.

What survived under the passionary is the other thing. It is the working state: the roll with the corrections still standing in it, including passages she plainly intended to remove, and including the parallel record kept by somebody else in the same house. This is why she can be caught, repeatedly, keeping two versions of one event side by side and telling us which she means to keep. She is not being candid with us. She is drafting in front of us because she did not know we were there.

That state is the part of this that I cannot account for and have stopped trying to. A working roll is the copy a household destroys. It is superseded by definition; it is the one nobody is meant to have; and the ordinary fate of such a thing is the brazier within a year of the fair copy being made. This one was not destroyed. Somebody kept it, in its drafting state, with the cancellations legible and the second hand distinguished from the first, and somebody else, four or five hundred years later, sat down and copied exactly what had been kept rather than tidying it into a book. I do not know who either of them was. The sixth-century man does not say where his exemplar came from, and his colophon reads like a man who has been handed something unusual and has decided that the unusual part is the point.

What he preserved, a thirteenth-century house scraped for a passionary, and multispectral imaging has now brought back up. Three acts of transmission, two of them anonymous, and the whole of this book depends on the first one.

Dates. She dates as a Roman dates, by consuls and by the Kalends, Nones and Ides, and I have left her alone. The year headings in the modern form are mine, supplied so that a reader can hold the sequence, and they are not in the manuscript in any form. Where a consular date appears in the text I have kept it and given the corresponding year in a note at the foot.

Names. She is Julia Agrippina, or Agrippina, or the daughter of Germanicus. She never once calls herself the Younger, a name invented long after her death, and I have not put it in her mouth. She calls her mother by the family name and calls her husbands by whatever the sentence requires, and the reader will find that which name she reaches for is frequently the most informative thing in the paragraph. Where she uses a title Rome did not yet have, she is being sarcastic, and I have not smoothed it.

Money. Sums are given as she gives them, in sesterces, uncorrected. I have resisted every temptation to convert. Her figures are not decoration. She is a woman who thinks in columns, and the reader who keeps half an eye on the arithmetic will be ahead of the reader who does not.

Structure. The recovered text begins in the consulship of Gaius Pompeius Longus Gallus and Quintus Veranius, which is the first year of her marriage to her uncle, and it opens with her looking at the record of her own birth. She is thirty-three. She is writing about her infancy from the archive in her own house, with two women in the room who were there and who do not agree with each other, and the reader should understand from the start that this is a book about the difference between what happened and what was written down, composed by a woman who is doing both.

Translation. I have translated for voice and not for the classroom. Where the Latin is plain I have been plain. Where she is being formal I have let her be stiff. She is fond of a sentence that arrives at its verb like a door closing, and English will not do that, and I have not pretended otherwise; what I have done instead is refuse to soften her when she is being cold, which is often, and which is the whole of her.

The rest is hers.

That was episode one, The Roll on the Table. Next episode is two, Born Among Eagles. Augusta is a translation of The Lost Commentaries of Julia Agrippina.