PHerc0139scroll

📍 Biblioteca Nazionale di Napoli ‘Vittorio Emanuele III’ (Officina dei Papiri Ercolanesi)

Recovered text

Philodemus, On the Gods (Περὶ θεῶν), Book 8 · Greek

The scroll's title and author were recovered, identifying it as Philodemus, On the Gods, Book 8 — a work by the Epicurean philosopher. Reading the title of a still-rolled scroll reveals what it contains before any of its body columns are studied.

⚠️ About 20% is virtually unrolled; the title and author are secure, but the body text is still being recovered.

3D model

loading…

Stats

Status: Title found · 20% unrolled
Segments: 23
Min pixel size: 1.129 µm
Scans: 4
Volumes: 5

Reached: Text recovered

SCANS

20250720235903-2.403um-0.2m-77keV ↗

20250720065842-9.362um-1.2m-113keV ↗

20251202104544-1.129um-0.2m-59keV ↗

20251205005141-2.399um-0.2m-78keV ↗

About this scroll

Title and author attribution recovered: Philodemus, On the Gods, Book 8 — a treatise by the Epicurean philosopher. About 20% virtually unrolled.

Data & access

webknossos viewer ↗CT in Neuroglancer ↗Browse files ↗

Voxel size (min): 1.129 µm
Energies: 59, 77, 78, 113 keV
Source / beamline: ESRF Grenoble
Scans / volumes: 4 / 5
Segments: 23
Formats: CT volumes (TIFF stacks · OME-Zarr) · surface segments
Predictions: 2 surface
License: CC BY-NC 4.0

HTTPhttps://vesuvius-challenge-open-data.s3.amazonaws.com/PHerc0139/

S3s3://vesuvius-challenge-open-data/PHerc0139/

Full dataset (HTTP)https://data.aws.ash2txt.org/samples/PHerc0139/

Full dataset (S3)s3://vesuvius-challenge/PHerc0139/

Model predictions (2)

Volume-level machine-learning predictions over this scroll's CT data — surface geometry and, where available, 3D ink detection. Each row opens the prediction volume in Neuroglancer or links its raw files.

Purpose	Base volume	Resolution	Model	Level	Threshold	Links
surface	`20250728140407`	9.362 µm · 113 keV	`20250701154204`	—	—	Overlay ↗Files ↗
surface	`20250728140407`	9.362 µm · 113 keV	`20260413222639`	0	0.2	Overlay ↗Files ↗

Ink predictions (38 segments)

Ink detected on each flattened segment surface. Thumbnails link to the full-resolution image; where available, each segment also links to its surface layers in Neuroglancer and its mesh.