SCROLL_ANATOMY.md

Anatomy of a Scroll

Scope of this document

This document is not exhaustive. Carbonized scrolls may contain damage or material arrangements not described here.

The CT appearances described below are possibilities, not diagnostic rules. A single matching feature does not establish a phenomenon, and the absence of one described feature does not rule it out.

Papyrus sheets and their material

This document describes ancient papyrus sheets made from two layers of pith strips. Within each layer, the strips were laid side by side. The second layer was arranged roughly perpendicular to the first, and the two layers were pressed and dried to form a single sheet.

Sheets and fibers

• SHEET (kollema): A manufactured unit of papyrus formed from two layers of pith strips. It is not a winding or the complete roll.

• FIBER: In this document, a fiber is a fibrovascular bundle from a pith strip. In an intact strip, the fiber runs along the strip and is embedded in its soft ground tissue (parenchyma). Papyrus is not made entirely of fibers. The fiber is not the pith strip itself.

Surfaces and fiber families

• A manufactured papyrus sheet has two outer surfaces: RECTO and VERSO. Each is the outer face of one of the sheet's two layers. Each layer has thickness and is made of many pith strips. Each strip contains fibers surrounded by parenchyma. Neither surface is a single fiber. Where the material remains continuous, the surface can be traced by following the material and its fibers. Damage may interrupt or separate the material without changing which original surface it is associated with.

• In the manufactured sheet, each fiber lies within one of the two layers and runs along that layer. It does not pass through the sheet from RECTO to VERSO.

• Damage can separate a fiber from the surrounding parenchyma and from the rest of its sheet. The fiber may then move away from the surviving sheet material. The detached fiber still belongs to its original sheet and remains associated with the same RECTO or VERSO surface. It is not a separate papyrus surface, sheet, or winding. If direct continuity has been lost, its relationship to the surviving surface must be inferred from the surrounding papyrus geometry.

• The spacing between fibers associated with the same RECTO or VERSO surface can vary. Fibers may be widely spaced or grouped closely together, and one fiber may split lengthwise into smaller pieces that lie close to one another. Two fibers may also lie close together and run in the same direction even when they belong to different windings. Distance and direction alone therefore do not show which fibers are associated with the same surface. The continuity of the surrounding papyrus and the wider surface geometry provide the necessary context.

Roll geometry and windings

Roll structure

• ROLL: The complete papyrus support made from many sheets (kollemata) joined end to end and then rolled up.

• UMBILICUS AXIS: The central longitudinal axis around which the papyrus roll was wound.

• WINDING: One turn of the papyrus roll around the umbilicus axis.

• KOLLESIS: The overlapping joint between two consecutive sheets in a papyrus roll. It joins the sheets into the continuous roll.

  At a kollesis, two different sheets overlap within the same winding. Identifying another sheet there does not by itself identify another winding.

• OPPOSITE SURFACES OF ONE WINDING: The two surfaces that bound the papyrus material of a winding at a particular location. Where the geometry permits a local measurement, the separation between a sheet’s RECTO and VERSO surfaces is measured from RECTO through the sheet to the corresponding VERSO surface, perpendicular to RECTO at that location. Before measuring from RECTO, we need to identify the VERSO surface of the same sheet; it may not be the first papyrus surface we encounter along the direction perpendicular to RECTO. The distance includes any internal delamination gap. At a kollesis, two sheets overlap within one winding, so these surfaces bound the combined overlapping material rather than either sheet alone. The distance between these surfaces can change abruptly because of delamination, compression, damage, or other deformation. Their distance at one location may not match their distance nearby.

• Winding identity comes from the original rolled structure, not from present-day material connectivity. Delamination, cracks, and missing material can separate parts of one winding, while fusion can connect different windings. Connected material can therefore belong to different windings, and disconnected material can belong to the same winding. Their original relationship must be inferred from the surrounding surface geometry.

• In the original rolled structure, if three consecutive windings occur in the order A–B–C, then B lies between A and C. A and B are adjacent, B and C are adjacent, and A and C are two winding steps apart. Damage can change their positions, so windings that are close together now were not necessarily adjacent in the original roll.

• Sheets are the manufactured pieces joined to make the roll; windings are the turns formed when the roll is wound. They do not form a containment hierarchy: a single sheet can span several windings, and a single winding can contain parts of different sheets.

• Each surface of a manufactured sheet is associated with the principal fiber family of the layer that forms that surface. Before the sheet was rolled or deformed, the RECTO and VERSO fiber families were approximately perpendicular.

• By the convention used here, RECTO fibers run lengthwise in the unrolled roll, and VERSO fibers run across its height. In regularly rolled, undeformed geometry, RECTO fibers are approximately circumferential around the umbilicus axis, while VERSO fibers are approximately axial.

• These directions describe the manufactured sheet and the ideal geometry of a regularly rolled, undeformed scroll. The carbonized scroll is not preserved in that ideal geometry. Folds, compression, and other deformation can bend or rotate a surface locally, sometimes abruptly. After breakage, part of the papyrus may move from its original position. In either case, the surface and its associated fibers can have directions very different from the ideal axial or circumferential direction. In a particular CT view, the fibers may appear diagonal or extend out of the viewing plane.

• A fiber family remains associated with its RECTO or VERSO surface even when the material is bent, rotated, or displaced. Fiber direction can help distinguish RECTO from VERSO, but the axial-versus-circumferential relationship is a reference for the broader roll geometry, not a fixed rule for assigning RECTO and VERSO at a particular location.

Following surfaces through the volume

• A fiber traced continuously from one location to another provides evidence of local surface continuity. A detached fiber may remain unbroken even where the surrounding papyrus surface is interrupted. Fibers running in similar directions at two locations do not by themselves show that an unbroken papyrus surface connects those locations. The fibers may belong to different windings.

• After tracing a sheet’s RECTO and VERSO surfaces in one region, we may encounter more papyrus farther along a direction perpendicular to the local sheet surface. That material may be another part of the same sheet: the sheet can extend into another winding, or it can fold back so that we encounter it again within the same winding. Encountering more papyrus therefore does not, by itself, establish that it belongs to a different sheet or winding.

  Even if the additional papyrus has a RECTO surface with its associated fibers and a VERSO surface with its associated fibers, identifying that pair does not tell us how it relates to the sheet we traced first. It could belong to the same sheet in another winding, to a folded part of the same sheet within the same winding, or, at a kollesis, to a different sheet overlapping within the same winding. We need the surrounding papyrus geometry to determine which sheet and winding the additional material belongs to.

• An intact fiber follows a continuous three-dimensional course, although CT may not resolve all of it. Where visible, a single slice shows only the portion or cross-section intersected by that plane; following the fiber through a sequence of slices or other coherent planes can reveal more of how it runs through space. Where the fiber families are approximately perpendicular, a viewing plane perpendicular to the local papyrus surface can cut lengthwise through the fibers of one family and crosswise through the fibers of the other. One family may then appear as elongated traces and the other in cross-section. The plane that gives this view changes orientation as the papyrus bends or rotates within the CT volume.

Delamination, kollesis, folds, damage, and compression

• Delamination can open a gap inside a papyrus sheet. At the gap, the separated material still belongs to the same winding; the split does not create another winding.

• In delamination, the split lies inside the papyrus sheet, between its RECTO and VERSO surfaces. The split can occur between the two layers of pith strips or inside one of those layers. RECTO and VERSO still refer to the two outer surfaces of the original sheet, not to the sides of the internal gap.

• Only part of a sheet may delaminate. The sheet can remain intact around the delaminated region.

• Seeing two parts of the papyrus separate and then come together does not by itself show whether they belong to one delaminated sheet or to different windings. A delamination can open and close within a sheet, while two different windings can touch over a limited region and then separate.

• At a kollesis, two sheets overlap within one winding. The interface between the sheets lies inside the overlap. The opposite surfaces of the winding are the two outer surfaces of the combined overlapping material.

• The interface between the two sheets in a kollesis can open, making the overlapping sheets look locally like surfaces from different windings. They still belong to the same winding.

• Fibers can be interrupted by piercings, sharp bends, and fractures that cross multiple windings. The end of a traced fiber does not by itself show that the papyrus surface ends there.

  When we follow the roll lengthwise across a kollesis, we reach the next sheet before the previous sheet ends. Identifying another sheet at the kollesis does not by itself identify another winding.

  A tear or kollesis can explain why a traced fiber ends, but only when that explanation is supported by the surrounding papyrus geometry.

• A fold bends papyrus back onto itself and can involve one or several windings. Folding does not change which winding the material belongs to. The RECTO and VERSO surfaces keep their identities through the fold. A surface may face the opposite direction after the bend without changing from RECTO to VERSO or from VERSO to RECTO.

• A sharp bend is not necessarily a fold. In a fold, the papyrus doubles back on itself; compression or another deformation can bend the papyrus sharply without making it double back.

• A fold may be confined to part of a winding. Beyond the folded region, the papyrus can continue along the winding. Within the folded region, a straight line through the papyrus may encounter the same sheet or the same surface more than once. Repeated intersections do not by themselves indicate different sheets or windings.

• Near the core of a roll, papyrus may fold back on itself one or more times rather than following a simple spiral through that region.

• A tear or fracture breaks the papyrus material. The affected surface may end at the break or reappear nearby in a displaced position.

• Breakage and movement can disrupt the local arrangement of papyrus surfaces. Parts of one or more sheets or windings may shift, separate, fold, or come back into contact. Surfaces that are now close together were not necessarily adjacent before the damage, and surfaces that originally belonged together may now be separated. This movement does not change which sheet or winding each surviving part came from. Their original relationships must be inferred by tracing the surfaces away from the damaged region and relating them to the surrounding sheets and windings. The shape of a break can help, but damage can alter it, and unrelated breaks can look alike.

• Missing material is papyrus that has been lost or destroyed at a location. It cannot be observed or traced through the missing region. Continuity across that region can only be inferred from the surviving surfaces and fibers around it and from the wider arrangement of sheets and windings. The evidence may not be available immediately around the missing region; following the surviving papyrus farther through the volume can help us understand which parts were connected.

• Foreign material, such as mineral grains or unrelated debris, may be embedded in the papyrus or lodged between windings. It is not itself a papyrus surface, sheet, or winding.

• Compression can bring adjacent windings into contact, reduce or eliminate the space between them, and make their surfaces appear locally merged in CT. The interface may become difficult or impossible to resolve. The loss of a separable interface does not by itself show that the surrounding unbroken material has changed order.

• When CT shows no visible gap between two papyrus surfaces, it may not reveal whether they are separated below the scan resolution, pressed together, adhered, or fused. In all these cases, the surfaces remain associated with the same original windings. Material from two windings can fuse into one connected region, but the windings retain their original identities even when CT cannot show where the material from one winding ends and the material from the other begins.

• If two papyrus surfaces that are distinct nearby converge into a region where CT cannot separate them, this supports the inference that both surfaces are still present there.

CT visibility and contrast

• SCAN ARTIFACT: A feature introduced by CT acquisition or reconstruction rather than a physical feature of the papyrus.

• Fibers may remain visible in CT where the surrounding parenchyma has been damaged, lost, or become too faint to resolve. Their direction and continuity can help trace a papyrus surface. Fiber spacing can vary, so a small observation window may contain only one visible fiber. Its relationship to the surface must be understood from its continuation and the surrounding papyrus geometry.

• CT intensity is a reconstructed image value, not a material label or a direct measurement of mass density. It depends on the material, X-ray energy, phase effects, and reconstruction. The same CT intensity value need not have the same meaning in another scan or processed volume.

• As X-rays pass through material, their waves can shift in phase. Interference between these waves can produce bright and dark fringes, or halos, around material boundaries in CT. These fringes arise from the imaging process and do not represent additional papyrus material.

• Blur can make nearby structures appear merged, so a visible band or profile peak need not correspond to one physical structure. Densely packed fibers can contribute to this blur: their fine cellular structure may scatter or slightly deflect X-rays. The amount of blur depends on the papyrus structure and scan conditions.

• A winding is a turn of the papyrus roll, not an intensity band. Neither a bright region nor a dark region identifies a fiber, a winding, or empty space by itself. A continuous bright or dark region in CT does not by itself show that the papyrus is physically connected there, and a gap in that region does not by itself show a break in the papyrus. We need the surrounding papyrus geometry to interpret the contrast.

• Voxel size is not the same as effective resolution. Whether a structure can be distinguished also depends on blur, phase effects, reconstruction, contrast, and local preservation.

• At coarser spatial resolution, nearby structures may no longer be resolved separately and can appear joined even though they are physically distinct. Finer spatial resolution may show that they are separate.

• Fibers are among the most visually distinctive features of the papyrus in CT. We can generally follow them through the volume using their shape, course, and relationship to the surrounding papyrus. Brightness alone does not identify a fiber. Depending on preservation, compression, contrast, and spatial resolution, fibers may appear blurred, become difficult to distinguish from the surrounding parenchyma and other nearby material, or disappear from view locally. Being able to follow a fiber in CT does not mean that its width or internal structure can also be resolved.

  Where a fiber is no longer clearly visible, the parts we can still see and the surrounding papyrus geometry can help us infer whether it continues and what course it follows.

Ink

• INK: Material applied to a papyrus surface to form writing. It is not one of the sheet's two layers of pith strips and does not create another sheet or winding. Its presence or absence does not define RECTO or VERSO in this document.

• Herculaneum inks are generally carbon-based and may provide little absorption contrast against carbonized papyrus. Composition, preservation, and scan conditions vary, so ink has no universal CT intensity or appearance.

• Ink may appear as a surface deposit, raised texture, cracked texture ("crackle"), or subtler changes in contrast and texture. In some scans it is directly visible. None of these appearances identifies ink by itself, and their absence does not establish that a surface is uninked.

Cues that are not diagnostic by themselves

• The distance between two papyrus surfaces tells us only how far apart they are at that location. Two nearby papyrus voxels in the CT volume can map to widely separated positions in the unwrapped scroll. Even a very small distance does not show that they are the opposite surfaces of one winding. How the two surfaces continue through the surrounding papyrus provides the context for understanding their relationship.

• Similar local orientation does not show that two papyrus surfaces are the opposite surfaces of one winding. Surfaces from different windings can also be locally parallel.

• Two papyrus surfaces following the same local bend do not necessarily belong to the same sheet or winding. Surfaces from different windings can have similar local shapes.

• Similar texture or CT intensity does not show that two papyrus surfaces belong to the same winding. Surfaces from different windings can look alike.

• Thickness alone is not a reliable topology cue, and a thickness value from one region should not be assumed to apply in another.

• The number of peaks in one CT intensity profile does not by itself establish the number of material layers or windings.