Herculaneum Scrolls: The Villa Library Was Mostly Philodemus

1752: lumps of charcoal that were books

The painter Camillo Paderni, curator of the Herculaneum finds, wrote from Naples on November 18, 1752, that many rolls of papyrus had been found less than a month before. In his letter they had turned into a kind of charcoal so brittle that it fell to ash at a touch. His attempts to open them on the king's orders, he wrote, had saved only some words (Philosophical Transactions 48, 1753, pp. 71–73). English Wikipedia, summarizing Antonio de Simone and Richard Janko, says the rolls were at first taken for charred branches, and some may have been thrown away or burned for heat.

Illustration

Blank places most of the books, probably all or nearly all the philosophical ones, on shelves in a small room not furnished for reading. A few more were in cabinets in the tablinum, a room off the square peristyle, and others in traveling boxes (capsae) in several parts of the villa.

How many rolls there were depends on what is counted. Parker and colleagues (2019) give more than 1,800 papyrus rolls, and English Wikipedia gives an official list of 1,814 rolls and fragments. Blank estimates the bookrolls dug up in 1752 and 1754 at perhaps about 800. The counts measure different things: the official list includes fragments, and Blank notes that each piece cut from a roll got its own papyrus number.

Whose library?

Blank lists who else is on the shelves. Besides Philodemus and Epicurus there are other Epicureans (Demetrius of Laconia, Colotes, Polystratus, Carneiscus, Metrodorus and Zeno of Sidon), at least two books by the Stoic Chrysippus, and a group of Latin rolls. The Latin group includes a poem on the battle of Actium (PHerc. 817) and a history by Seneca the Elder (PHerc. 1067), and K. Kleve identified a possible manuscript of Lucretius.

Philodemus's own titles named in the encyclopedia include On Music, On Poems, On Rhetoric, On Death, On the Gods, On Vices and On Anger. Those hoped-for lost books of Livy are tracked in the Lost Books of Antiquity.

The "mostly Philodemus" description comes with a caveat. Blank notes that many rolls have lost the closing page where a scribe wrote the author and title, and nearly all have lost the opening one. Many such rolls were assigned to Philodemus indirectly, from the handwriting, the kind of roll, and the style and content of the text, though Blank adds that most attributions to Philodemus are generally accepted by scholars.

Who owned the villa, and its books, is still argued. The three proposals below are set side by side, not ranked.

Unrolling by hand: Piaggio's machine

Paderni's own method, as Blank describes it, was to cut a roll lengthwise, copy the layer of writing that showed, then scrape it off to reach the next layer. The copying moved backward through the text and destroyed each layer once it was drawn. The inner cores that survived were opened on a frame built by Antonio Piaggio, a priest, which drew the roll open with silk threads attached to its edge (Ethel Ross Barker's 1908 account, as quoted by English Wikipedia).

The sources give two dates for the first result. Blank says that in 1754–55 Piaggio finished unrolling the end of one book, a work by Philodemus against the art of music. English Wikipedia says Piaggio began in 1756 and that the first roll took four years. The work was slow at first, with only 18 rolls successfully unrolled in the workshop's first phase, from 1754 to 1798; under John Hayter, an English clergyman paid by the Prince of Wales, over 200 were unrolled in the five years from 1802 (Blank 2019).

Reading without unrolling

Opened pieces were read under microscopes from 1970 and, from 2000, in digital photographs, most of them in the near infrared (Blank 2019). Closed rolls were harder. Their ink is a carbon ink, "lamp black," long held to be invisible in ordinary micro-CT scans, and the first such scans of closed rolls, in 2009, showed their inner structure but not clearly their ink (Parker and colleagues, 2019).

Parker's team lists phase-contrast X-ray studies that claimed Greek letters and words inside rolled papyri, among them a 2015 paper by Vito Mocella's team. Such results, they wrote, "have yet to be confirmed, extended, or used to reveal complete passages of text," and "no identifiable, readable text has been recovered from inside a closed Herculaneum scroll." In the same paper they trained a machine learning model, a 3D convolutional neural network, to detect carbon ink in micro-CT scans of Herculaneum material.

Timeline data: herculaneum-scrolls-reading-timeline.csv, with the source and link for every row.

The Vesuvius Challenge, a prize contest launched on March 15, 2023, by Nat Friedman, Daniel Gross and Brent Seales, released X-ray scans of Institut de France rolls. Its steps since then, as of October 19, 2026:

PHerc. 1667 is the inner core left after hand openings in the 19th century, in 1969 and in the 1980s. The team says it was scanned by phase-contrast X-ray microtomography at the European Synchrotron Radiation Facility in Grenoble. The researchers then traced the wound sheet and flattened it, picked out the ink with machine learning, and papyrologists transcribed the columns. Its last surviving column mentions Aristocreon, a nephew and student of the Stoic Chrysippus, and from that name and the text's wording and subjects the team places this ethical treatise among Stoic writings.

The project counts hundreds of rolls as still sealed. In October 2026 it paid a prize for letters found inside PHerc. 343.

What is still underground

As English Wikipedia summarizes it, the villa was explored through tunnels from 1750 until digging was halted in 1765. The Swiss engineer Karl Weber worked there under Roque Joaquín de Alcubierre, who headed the Herculaneum excavations. The site was then lost for two centuries until it was found again in 1986 through an ancient well, and excavations in the 1990s uncovered two lower floors. Wikipedia reports that as of 2012 some 2,800 square meters were still to be dug, and that most of the villa remains underground.

Whether more rolls lie in those levels is an open question. English Wikipedia quotes Mocella and colleagues hoping for papyri not yet found, perhaps a second library of Latin papyri, at a lower level not yet dug. The Latin rolls already known are the group Blank describes.