A fossil bone has two surfaces. One is the familiar exterior: shaft, joint, scar, crack, and patch of clinging rock. The other is hidden inside. There, microscopic channels trace former blood vessels, successive tissues record changing rates of growth, and remodeling shows where an animal's living skeleton replaced itself. Those structures can survive mineralization well enough to turn bone from a static prop into a partial life history.[3]

Reaching that second surface creates a genuine conservation problem. Traditional paleohistology makes it visible by removing a small piece of fossil, fixing it in resin, cutting it, and grinding it thin enough for transmitted light. X-ray tomography instead collects projections through the intact object and reconstructs virtual slices. One route spends material to gain an optical section; the other preserves the object and gains a three-dimensional volume, but only at the contrast and resolution the scanner, beam, and specimen permit.[3][6]

The two videos below make that tradeoff physical. PBS Eons visits Ellen-Thérèse Lamm's paleohistology laboratory at the Museum of the Rockies, where a saw and grinding equipment turn fossil bone into a microscope slide.[1] In the second, Vincent Fernandez of London's Natural History Museum speaks at a 2020 European Synchrotron Radiation Facility workshop about “digging” through large fossils with X-ray microtomography.[2] Read together, they are not an old-method/new-method contest. They show why the right question is: which observation justifies which kind of access to an irreplaceable specimen?

1. A physical slice trades material for optical detail

In the PBS Eons film, watch the sequence before the colorful microscope views. Lamm does not put an entire dinosaur limb under a lens. A selected portion is embedded for support, cut into wafers, attached to glass, and ground until a finished section is about 100 micrometres thick. That transformation is the method's power and its cost: light can now pass through the fossil tissue, but the removed piece cannot be returned to its original state.[1]

The finished slide is not simply a set of tree rings. Under ordinary and polarized light, a researcher can distinguish the organization of bone matrix, the abundance and direction of vascular canals, later secondary osteons, and pauses recorded as lines of arrested growth. Those features can constrain growth rate, maturity, disease, and the history of a particular skeletal element. They do not automatically yield one clean age. Growth marks may vary within one bone, differ among bones in the same animal, or be erased by remodeling; tissue must be interpreted comparatively and in anatomical context.[3][6]

That is why the quietest moment in the workflow—the choice of where to cut—matters as much as the microscope image. A spectacular section taken from the wrong part of an element may answer a different question from the one proposed. A small sample may preserve the outer growth record but miss variation elsewhere. A larger sample may be more representative while consuming more of a finite object. Histology therefore begins before the saw, with the specimen's identity, the research claim, the sampling location, and a plan for documenting what will change.[3][4][6]

Museum rules make that reasoning explicit. The Paleontological Research Institution requires prior written permission for any procedure that could irreversibly alter an object. Its policy asks reviewers to weigh the scientific question, why its material is necessary, whether a less intrusive method could work, specimen rarity, impact on future research, and the fate of any remnants.[4] The University of Colorado Museum of Natural History likewise treats destructive sampling as an irreversible intervention that must be justified and performed by qualified researchers. Its policy rejects destructive sampling of type or rare specimens, requires imaging—and reproduction when necessary—before a cut, and requires all remaining material to return to the paleontology collection.[5]

These safeguards do not imply that a fossil should never be cut. Refusing every irreversible analysis would protect the object while locking away some of the biology it preserves. The stronger principle is proportionality: consume the smallest, best-documented sample that can resolve a worthwhile question, preserve the slide and offcuts under the original catalog identity, and make the resulting data part of the specimen's future record. The video makes cutting look like skilled craft. The policies reveal the less visible craft of deciding whether the cut should happen at all.

2. A synchrotron lets the slice move through the specimen

Fernandez's ESRF lecture starts from a different ambition: keep a fossil physically whole while making its interior navigable. Computed tomography records many X-ray projections around an object and reconstructs them as a stack of cross-sections. Those sections can be viewed in any plane, combined into a volume, and segmented so that a concealed tooth, bone, or void can be followed through three dimensions. In that sense, the “slice” can move after the scan instead of being fixed forever at the place where a saw passed.[2]

The word “large” in the title carries much of the lecture's argument. Early high-resolution synchrotron work benefited small specimens because field of view and available X-ray energy constrained what could be penetrated and captured. Fernandez describes how new optics, scan protocols, and the use of several ESRF beamlines expanded whole-object imaging from specimens roughly 15 centimetres across toward objects about 25 centimetres wide and nearly 60 centimetres long, with selected regions then examined at higher resolution.[2] The advance is not merely a sharper picture. It lets anatomy remain inside its surrounding rock or burial structure, retaining spatial relations that physical extraction could destroy.

Virtual access also changes sampling logic. Instead of committing to one transverse cut, a researcher can inspect many planes through one element and compare different regions. A 2023 experiment at SPring-8 in Japan scanned Fukuiraptor femora roughly three centimetres wide at a voxel size of about four micrometres. The resulting virtual sections recovered vascular canals, secondary osteons, and lines of arrested growth comparable to features in physical sections. Because the data form a volume, the researchers could examine variation within and among elements without taking a new chip for every view.[6]

Yet “non-destructive” does not mean “complete.” In that experiment, the virtual sections did not resolve osteocyte lacunae or collagen-fibre arrangements; the effective voxel size was too coarse for those finer targets. The authors also identify the practical barriers of synchrotron cost and proposal-based access.[6] Different fossils create different contrast problems because bone, mineral infill, and matrix may attenuate X-rays similarly. Larger or denser objects demand more penetrating beams, while higher spatial resolution normally narrows the field that can be captured. Reconstruction and segmentation then add decisions about artifacts and boundaries after the beam time is over.[2][6]

The cover photograph makes the alternative concrete: narrow fossil eggshell pieces rest beside the blade of a specialized saw, aligned for a physical section rather than free to be inspected from any virtual plane.[7] That visibility should not be mistaken for simplicity. A scan must still match the specimen and question. If the goal is to map growth marks through a substantial volume, virtual histology can avoid an otherwise unjustified cut. If the goal depends on tissue structures below the achievable resolution, a carefully governed physical section may still provide evidence the volume cannot.

The best archive may contain both object and intervention

The videos converge on a less glamorous definition of progress. A new instrument is valuable not because it makes the older method obsolete, but because it changes which losses are necessary. Scanning before sampling can map internal structure, reveal whether a proposed cut is well placed, preserve a three-dimensional record, or show that no cut is needed. A later physical section can then target the feature whose optical detail justifies the intervention.[5][6]

That sequence also makes disagreement more productive. Future researchers can revisit the scan volume, inspect the retained slide, connect both to the cataloged bone, and test whether an apparent growth mark continues through the element or disappears with a change of plane. The scientific archive becomes more than an untouched fossil. It becomes the specimen plus its images, samples, preparation history, analytical settings, and explicit record of what was removed.

A saw opens one real plane and closes off the possibility of an unaltered specimen. A synchrotron opens many virtual planes but may stop short of the smallest tissue signal. Neither fact supplies a universal winner. Good paleontology puts the biological question, the instrument's limits, and the object's future in the same frame—and chooses the least loss that still produces evidence.

Sources

1. PBS Eons, “What a Dinosaur Looks Like Under a Microscope,” official YouTube video, March 19, 2018.
2. European Synchrotron Radiation Facility, Vincent Fernandez, “Digging virtually into large fossils using synchrotron X ray microtomography,” official YouTube lecture from the 2020 Cultural and Natural Heritage workshop.
3. Alida M. Bailleul et al., “Dinosaur paleohistology: review, trends and new avenues of investigation,” PeerJ 7 (2019), open full text.
4. Paleontological Research Institution, “Research Collection Policy,” section 8, “Technical Analysis, Destructive Sampling, and Preparation,” revised April 28, 2020.
5. University of Colorado Museum of Natural History, “Destructive Sampling, Reproductions, and Imaging,” paleontology policy covering authorization, pre-sampling imaging, protected specimens, and returned material.
6. Takuya Imai et al., “High-energy synchrotron-radiation-based X-ray micro-tomography enables non-destructive and micro-scale palaeohistological assessment of macro-scale fossil dinosaur bones,” Journal of Synchrotron Radiation 30 (2023), open full text.
7. Bureau of Land Management Utah, “Dinosaur Eggshell #10,” 2016 photograph of fossil eggshell material at a specialized saw, public domain, via Wikimedia Commons.