Aqua Tofana, after the legend

The version that still moves online is a complete object. In seventeenth-century Italy, a woman named Giulia Tofana brewed a colorless liquid, sold it as a cosmetic, and taught wives how to use it. Hundreds of husbands died. The Church or the governor sprang a trap. She confessed under torture and was executed, sometimes with a daughter, sometimes with a coven, always as the spider at the center of a web.

That object is a nineteenth-century souvenir hanging on an older trial. It is not nothing. There was a poison associated with the name Tofana. There was a Roman investigation in 1659 that ended with women hanged in the Campo de’ Fiori. There were clients, intermediaries, and a fear — contemporaneous, not invented last year — that a plague year could hide a dose. What the souvenir adds is the part a thread needs: one named villain, a round body count, and a moral that women in locked marriages found a chemical exit and were punished for it.

Palermo first

The earliest solid public mention is not Rome and not a six-hundred-man census. In 1632–33, in Palermo, women including Francesca la Sarda and Teofania di Adamo were tied to a poison later remembered as Aqua Tofana. Teofania was executed in 1633. Later writers made her the mother of the more famous Giulia, or the inventor of the recipe, or both. Kinship of that kind is one of the first things to go soft when you leave the sentence and look for a parish book.

Arsenic is the usual inference for what the liquid was. It fits the period, the apothecary trade, and the way slow poisoning was described. It is still an inference. No one has produced a sealed flask from 1659, chemically assayed, and labeled in a hand everyone trusts. When a story says “the recipe was arsenic, lead, and the foam of a dying man,” you are listening to atmosphere, not a lab note.

The woman who is not on the scaffold

The name Giulia Tofana is itself a later convenience. Archival work that has reached a general audience in the last few years identifies a Giulia Mangiardi, arrived in Rome from Palermo, dead in 1651 — eight years before the hangings that the legend assigns to her. Historians who have read the death record treat her end as ordinary in the bureaucratic sense: a certificate, an age, no scaffold. The law, in that reconstruction, never caught up with her.

If that reconstruction holds, the popular execution scene is a splice. A Palermo poison case and a Roman poison case were stacked until they became one woman’s life. Splices are how folklore stays portable. They are also how a stepdaughter, a housemaid, and a set of clients get written out so the story can keep a single face.

Rome, 1659

What the Roman record will carry without much strain is this. In 1659 the papal city’s investigators moved against a network that mixed fortune-telling, the sale of a liquid, and the suspicion of murder. Gironima Spana — in some accounts Mangiardi’s stepdaughter and heir to a clientele — sits at the center of the later scholarly versions. Other names survive beside hers: Giovanna de Grandis among them. In July 1659, five women were hanged in the Campo de’ Fiori. Other people implicated in the file were imprisoned, disappeared from the popular versions, or were never as useful to the legend as a public rope.

A long notarial transcript of the investigation exists. Craig Monson’s work, using that transcript and related papers, is one of the reasons a careful modern account no longer has to pretend the only sources are nineteenth-century chapter books. The transcript is not a novel. It is a pile of testimony, contradiction, and official fear. Spana, in the versions that stay close to it, denied to the end that she had done the thing the city wanted a spider to have done. That denial is not proof of innocence. It is a fact about the file. So is the smaller, harder point Monson and others have pressed: the women on the scaffold were not the only people in the network, and some of the people who used the liquid were not the people who sold it.

Popes had a reason to keep the paperwork inconvenient. A poison that could be described clearly enough to copy was a public problem. Sealing a file is not the same as proving a recipe, but it is a contemporary signal that the case was treated as more than a street rumor.

Six hundred, five hundred, some

The number six hundred is the hook’s load-bearing wall. It is also the first number a cautious reader should set down. A contemporary rumor after the hangings — the kind of sentence that moves through a city in a week — already reached for a round figure: perhaps five hundred husbands, hidden by plague. Later popular histories made the count larger, cleaner, and more certain. Historians who have sat with the transcript do not treat either figure as a body count. They treat them as what a frightened city says when it wants the story to be equal to the spectacle in the square.

Some of the dead husbands of 1656–57 will have died of the diseases that were actually in Rome. Some of the living husbands of 1659 will have been hated for reasons that had nothing to do with a bottle. The archive can show a trade in a liquid and a set of accusations. It cannot, from the material now in public scholarly circulation, support a serial-killer census with a famous woman’s name at the top.

What to do with the leftover myth

None of this makes Aqua Tofana wholesome. A slow poison sold through intermediaries, in a city where a wife had few legal exits, is already a dark enough public fact. The leftover myth is the part that turns those women into a single legendary chemist and their clients into a statistic. That is a different product. It is closer to a cautionary tale about female conspiracy than to a reading of a notarial hand.

If you arrived from a sentence about six hundred corpses and a cosmetic vial, the honest remainder is smaller and stranger. A Palermo execution in 1633. A Roman woman later called Tofana who, on the better reconstruction, died in bed in 1651. A 1659 prosecution that did hang people, did frighten a government, and did leave a transcript long enough that popes preferred it shut. The almost, here, is not that the poison missed. It is that the legend missed the file, and then spent two centuries pretending the file had never been the point.