This post is about what’s on the surface of the coated ones.

Not what the surface looks like, but what it is actually made of, and what that turns out to mean.

The Raman Question

When I examined the coating edge under the optical microscope, the boundary wasn’t a line. It branched. Fingers of material spread across the uncoated garnet face in a pattern that looked less like mineral film evaporating out of a solution and more like something advancing across a surface it hadn’t reached yet. Weathering rinds don’t have advancing fronts. They have boundaries that stop where the exposure stops. This one was still going somewhere, in a direction, with structure to the leading edge.

That was before any instrumentation. The shape of the question came before the answer.

Raman spectroscopy works like this: a laser fires at a sample, and almost all the light bounces back exactly as it arrived. But a tiny fraction comes back shifted — at a slightly different frequency. That shift happens because some of the laser’s energy was absorbed by the vibrating bonds inside the material and re-emitted. Different chemical bonds vibrate at different frequencies, so the pattern of shifts is a fingerprint. Not a rough one — specific enough to distinguish minerals that look chemically identical by almost every other method. On a coating this thin, the laser reads primarily what’s on the surface.

The uncoated garnet came back exactly as expected — a clean set of peaks belonging to almandine, the iron-rich garnet this locality produces. That’s the baseline. Everything else is what’s sitting on top of it.

Before we go further, a quick word about how to read a Raman spectrum, because the figure that follows is worth understanding. The horizontal axis is labeled in units called wavenumbers — a way of describing the frequency of the scattered light. You don’t need to know exactly what wavenumbers are. Think of the spectrum like a piano keyboard stretched out flat: different chemical bonds “play” different keys. Each mineral has its own chord — a specific set of positions where it shows up, and those positions don’t shift. The vertical axis just shows how loudly each key is being played.

The coating spectra came back playing something I didn’t recognize.

Spot 1, measured at the coating margin, showed peaks characteristic of hematite — an iron oxide, the rust-red mineral that gives the Painted Desert its color and covers much of Mars. That is not a mineral you expect to find as a coating on a garnet. The garnet’s iron is in a form called ferrous iron — the stable, unoxidized form, the same kind you’d find in unrusted steel. The hematite in this coating carries ferric iron instead, the oxidized version. Rust, essentially. Something at the garnet surface converted one into the other, leaving hematite as evidence.

Here is where it gets slightly technical, but it matters, so let’s slow down. Hematite produces a faint extra peak at a position on the spectrum that happens to be very close to where organic carbon produces one of its characteristic signals. It is an overtone — the same phenomenon as the faint harmonic notes that ring alongside the main note when you pluck a guitar string. In spot 1, what might appear to be an organic signal is actually hematite playing one of its overtones. I am flagging this now because confusing the two would change the interpretation of the data, and I want the logic to be visible.

Spots 2a and 2b showed something else entirely. No hematite at all. Instead, two broad peaks in the high-frequency region of the spectrum — and these are not hematite. They are the signature of carbon.

Carbon atoms bonded to other carbon atoms in flat, orderly sheets — the arrangement you find in graphite — produce a signal called the G band. G for graphite. When that orderly structure breaks down — disordered, irregular, never fully crystalline to begin with — a second signal appears alongside it, called the D band, for disorder. The ratio of the two tells you how chaotic the carbon structure is. In pure graphite, D barely registers. Biological material is different: D tends to dominate, often louder than G, because living things don’t build their carbon in neat sheets.

In spots 2a and 2b, D dominates — broad, not sharp, clearly not graphite, not carbonate, not any form of carbon you would expect to find sitting on a garnet on a mountain in Arizona. It is disordered organic carbon. And it has not been cooked — the signal shows it is thermally immature, meaning whatever produced it was not buried deep or heated. It is relatively close, chemically, to what it was when it formed.

The coating contains hematite near the garnet surface, and organic carbon above it. Whether those are two distinct layers or one intermixed phase was not something surface Raman could resolve on its own. That required breaking it open.

The Plywood Discovery

LIBS — laser-induced breakdown spectroscopy — works by firing a high-energy laser pulse at the surface, ablating (vaporizing) a tiny amount of material. The resulting cloud of vaporized atomic plasma glows, and the instrument reads that glow. Different elements emit light at different wavelengths — the same reason fireworks are different colors — and that is how the instrument identifies what was there. Because each pulse goes slightly deeper than the last, LIBS is a depth-sampling technique: it reads the sample layer by layer from the surface down.

On a coating this thin, the first pulse or two consume it entirely and begin reading the garnet below. That was expected. What was not expected was what happened when I looked at the ablation crater — the tiny pit left behind — under the microscope afterward.

The coating hadn’t peeled away as a single unit. It had delaminated in layers — the way old plywood separates along its ply boundaries when water gets in. Two distinct layers, each lifting cleanly from the garnet below. The LIBS experiment had not characterized the coating chemistry the way I had hoped. What it had done, accidentally, was section it.

With the layers physically separated and the interfaces exposed, I pointed the Raman laser at each layer individually. The lower layer, sitting directly on the garnet surface, matched hematite. The upper layer matched disordered organic carbon.

The stratigraphy — the layered sequence of what came first, read from the bottom up the way geologists read rock strata — is: garnet → hematite (Fe₂O₃, iron bonded to oxygen) → organic carbon.

Three layers with different compositions, each confirmed by independent measurement, in a sequence with a direction. The hematite formed first, directly on the garnet surface. The organic carbon came after, on top of the hematite. Whatever process built this coating did it in order.

7,500×

The scanning electron microscope (SEM) tells the story in a different register. Unlike an optical microscope, which uses light, the SEM uses a focused beam of electrons to trace the surface of a sample. Because electrons have a much shorter wavelength than visible light, the SEM can resolve structure at scales thousands of times smaller than anything you could see with a conventional lens.

The coated face looks uniformly covered at low magnification — the metallic surface you see when you hold the specimen in your hand. Move up to medium magnification and the texture becomes irregular, lumpy, three-dimensional, nothing like the flat surface of the uncoated face beside it. At 7,500×, the structure finally resolves into something specific.

The coating is not a film. It is a packed array of sub-spherical (nearly round) structures, each roughly 0.5 to 2 micrometers across. For scale: a human hair is about 70 micrometers wide, so these structures are somewhere between one-thirtieth and one-tenth of a hair’s width. They sit against each other with no angular geometry, no flat faces, no straight boundaries. Rounded, packed, uniform in scale.

This is not what minerals look like when they grow. When inorganic compounds crystallize out of solution onto a surface, they produce forms that reflect their internal crystal structure — flat faces, angular boundaries, layered or branching patterns with geometric regularity. That is the signature of chemistry following crystallographic rules. Rounded, sub-spherical, non-angular structures packed at this scale are not a crystallization texture.

Coccoid bacteria — the sphere-shaped variety, the most basic bacterial form — are 0.5 to 2 micrometers across.

The resemblance is worth flagging. It is not a conclusion.

What the Beam Found

Energy-dispersive X-ray spectroscopy (EDS) works like this: when the electron beam strikes an atom in the sample, it can knock loose one of that atom’s inner electrons. When the atom refills that vacancy, it releases a small burst of X-ray energy. The energy of that burst is specific to the element — each element has its own signature X-ray, the way each element has its own spectral color in a flame test. By measuring the energies of all the X-rays coming off the sample, EDS identifies which elements are present and in what proportions. This specimen was run without any preparation coating — no carbon or gold applied to the surface before imaging. The numbers are from the sample itself.

The point measurement on the coating surface returned the following: carbon at 31.49% by mass (41.88% by number of atoms), nitrogen at 7.53% by mass (8.59% by atoms), oxygen at 38.55%, phosphorus at 2.08%, along with silicon, aluminum, calcium, potassium, sodium, and chlorine in smaller amounts. The two percentage figures for each element — mass% and atom% — describe the same thing in different ways: mass% is the proportion by weight, atom% is the proportion by count. Both are reported here because they tell slightly different parts of the story.

Three elements can be accounted for immediately: sodium, potassium, and chlorine. Those are sweat. I handled the specimen. Sodium chloride — table salt — and potassium are the chemistry of a fingerprint, not a garnet surface. Silicon, aluminum, and calcium are expected signal from the garnet itself bleeding through: at 15 kilovolts, the electron beam penetrates deeper than the coating is thick, sampling the almandine garnet below. Those elements belong to the garnet, not the coating.

That leaves carbon, nitrogen, oxygen, and phosphorus. Those are the coating.

The carbon is not a surprise. Raman already identified organic carbon in the upper layer. EDS confirms it and puts a number on it.

The nitrogen is what matters.

No mineral phase expected at this locality would produce a nitrogen signal. The contamination routes don’t work either — fingerprints don’t uniformly cover a surface, sample prep involved no nitrogen sources, and the atmosphere wasn’t a factor. Nitrogen at 8.59 atom%, alongside carbon and phosphorus, indicates that this coating contains organic matter of biological origin. In living things, nitrogen is the structural backbone of amino acids and proteins — the building blocks and machinery of cells. Phosphorus at 2.08% is the backbone of nucleic acids (DNA and RNA) and of phospholipid cell membranes, the fatty envelopes that contain every living cell. The co-occurrence of carbon, nitrogen, and phosphorus in these proportions is not ambiguous about what class of material is present.

The EDS line scan at 134.1 μm and 190× plots phosphorus — carbon’s X-ray is too weak to map reliably at line scan acquisition speeds, so phosphorus serves as the spatial proxy. The signal runs low across the first half of the traverse (the uncoated garnet face), then rises sharply and remains elevated through the second half, with intense, irregular spiking. Carbon, calcium, and potassium follow the same trend. The rise marks the transition from uncoated to coated face. Within the coating zone, the spiking is irregular because the beam is crossing individual globules. Each spike is one structure. Phosphorus — which has no inorganic source at this locality — is not uniformly distributed across the surface. It is concentrated inside the globular domains, not in the spaces between them.

The carbon-to-nitrogen atomic ratio in this spectrum is approximately 4.9:1.

What the Sequence Means

The phosphorus resolves a question left open from Part 2. In the X-ray fluorescence (XRF) analysis in that post, phosphorus was elevated in the coated garnet population but ambiguous — the XRF beam is wide enough that it could have been reading the garnet itself, mineral inclusions inside it, or the coating. EDS can pinpoint the signal to specific locations on the surface, and it places the phosphorus in the coating. That changes what phosphorus means here. It’s in the coating, not the garnet below, and given everything else in that spectrum, the only reasonable description is biological.

The uncoated garnets from the same outcrop have none of this — no hematite layer, no organic carbon signal in Raman, nitrogen below detection. If this were simple weathering driven by exposure to air and water, both populations would show it. They live in the same mountain, experience the same conditions. Only the coated population shows the layered structure, and weathering doesn’t work selectively like that.

The most consistent interpretation of everything collected here is that iron-oxidizing bacteria — microorganisms that get their energy by converting ferrous iron (Fe²⁺) to ferric iron (Fe³⁺) — colonized the surface of these specific garnets, extracted iron from the crystal surface, and left hematite behind as the waste product of that process. The organic carbon layer above the hematite is what remained of the bacteria themselves: accumulated biomass that fossilized in place. The hematite isn’t incidental to the story. It is the receipt.

This is where the XRF chemistry from Part 2 becomes more than background. The garnet deposit occurs in organic-bearing host rock — the microbial community wasn’t introduced from somewhere else, it was already there, living in the surrounding material. What the XRF established is that the two garnet populations have different trace-element chemistries: the coated population carries elevated chromium, scandium, and zinc substituting into the crystal lattice, creating structural strain that makes its Fe²⁺ slightly less stable, slightly more available as an electron donor. The uncoated population, with a cleaner lattice, doesn’t offer the same opportunity. The bacteria didn’t colonize randomly. One population’s crystal chemistry made it worth the metabolic investment. The other didn’t. The coating is the record of that selection — not a weathering event, not something that arrived from outside, but the host rock’s own biology working on a specific subset of its own minerals.

There is a well-documented mechanism in Arizona desert mineralogy that may explain how the organic material got there in the first place. When a cactus dies, it doesn’t simply dry out and disappear. Its tissue is loaded with oxalic acid — a simple organic compound that cacti produce in large quantities during their lifetime — and as the plant decomposes, that oxalic acid and other organic compounds move downward through the soil and into the rock below. Mineralogist Anthony Kampf, Curator Emeritus at this museum, has spent years documenting what happens next: where those descending organic compounds meet iron-bearing and copper-bearing minerals in Arizona ore deposits, entirely new minerals form at the interface. Ferriphoxite, carboferriphoxite, and more than a dozen other new mineral species have been described from precisely this process at sites including the Rowley Mine in Maricopa County. It is a recognized pathway — surface biology writing a chemical record in the rock below.

The Aquarius Mountains garnets may be recording a version of the same story. A cactus — or many cacti, over time — died above this outcrop. Organic material moved downward. It contacted an iron-bearing mineral surface that was already in biological territory, already in contact with iron-oxidizing bacteria living in the organic host rock. What we measure in the coating is not directly what the cactus left behind. The C:N ratio of 4.9:1 tells us that. Cactus-derived oxalate contains no nitrogen — its C:N would be effectively infinite. Biological organic matter, the kind produced by living and dying microorganisms, runs between 4:1 and 6:1. Something processed the organic input. The coating is what that processing left behind: hematite as the metabolic product, biomass above it, the whole sequence preserved on the surface of a garnet that had the right chemistry to make it worth colonizing in the first place.

That interpretation is the most consistent reading of the data collected here. It isn’t a closed case. The sub-spherical morphology we see in the SEM is consistent with coccoid (sphere-shaped) bacteria and also consistent with other processes that produce rounded structures at that scale. The carbon-to-nitrogen ratio falls in the biological range, but it would also fall in that range for accumulated non-living organic material derived from biological sources. What would close the case — and what the instruments used here cannot provide — is molecular-level identification: lipid biomarkers (the distinctive fat molecules that form cell membranes), intact cell-wall chemistry, or preserved DNA.

What the instruments can say is that this coating has stratigraphy — a layered construction sequence running from garnet surface to hematite to organic matter — and that the organic matter contains nitrogen with no inorganic explanation, phosphorus that belongs in cell membranes and genetic material, and a carbon-to-nitrogen ratio of 4.9:1.

The garnet has been sitting in the collection drawer. The coating has been sitting on it. The question of what built it is still open.

Aaron Celestian is the Curator of Mineral Sciences at the Natural History Museum of Los Angeles County, a former scientist at NASA's Jet Propulsion Laboratory, and an adjunct professor at USC. The Aquarius Mountains garnets are part of the NHMLA research collection. He writes Pocketful of Χtals because mineralogy is stranger and more alive than most people have been told.

Two True Things

I watched it on a screen.

Julie Tolentino’s installation was already weeks into its transformation when we spoke — borax crystals she had grown onto the sculptures, beginning to dry, fracture, and, in her framing, die. That was the intent. Decay as artistic fact. She had designed a system for making something and then watching it come apart, and it was working exactly as she’d planned.

She was right.

What I couldn’t stop seeing was the tincalconite. Na₂B₄O₇·10H₂O shedding water molecules until the structure reorganizes into something else — Na₂B₄O₇·3H₂O, a different mineral, stable under the drier conditions of the gallery. Not a lesser borax. A translation. The transformation she had framed as ending was, crystallographically, a record of the environment writing itself into the material. The sculpture was not dying. It was indexing.

I watched the borax fracturing on screen and thought: she’s going to ask me what I see, and what I see is a birth, not a death. Not a correction of her reading — an addition to it. The borax was dying. That was real. And in the same motion, something that had never existed in that gallery before was forming. Tincalconite. Na₂B₄O₇·3H₂O, stable, new, a mineral record of the exact conditions of this room, this installation, this specific act of letting something go.

I wasn’t sure how she’d receive that. I wasn’t offering a better frame. I was offering a second one.

The borax didn’t have a preference. It just kept changing.

I held both readings at the same time and didn’t know what to do with that, which wasn't a feeling I expected to have while watching crystal chemistry on a laptop. But there it was. The borax died. Something formed in its place. The sculpture kept changing. All three are true. I’m still not sure what to do with that.

I’ve been thinking about that gap ever since. It turns out it isn’t new.

What the Hands Already Knew

The first thing that surprises most people about calligraphy is that the pen is not the point. The paper is.

I do pointed-pen work, which means the nib is split—two tines that flex under pressure and release ink through the gap. Get the paper wrong, and the nib catches on the fibers, and the line explodes. Too loose a fiber structure and the ink bleeds sideways before it can form. Too compressed and it sits on the surface and won’t soak in at all. The ink has to match the paper. The nib has to match both. All three have to arrive at the same moment, or nothing works.

Nobody told me the ratios. I learned them by ruining things.

I spent more than I should on a bottle of ink, watched it feather and bleed across four sheets of Japanese paper I’d bought by the sheet, and only understood what went wrong when I switched to hot-pressed watercolor paper, and it finally held — the ink was never the problem.

What I was learning, without calling it this, was materials behavior. The fiber structure of the paper, the viscosity and surface tension of the ink, the spring constant of the nib — these are not poetic considerations. They are physical properties that interact in ways that cannot be fully predicted from first principles. You have to feel them. And once you’ve felt them enough times, you carry the knowledge in your hands before it reaches your head. The medium doesn’t just carry your intent — it argues back. What you meant to say and what arrives on the page get negotiated by the material itself. Get that wrong, and nothing you meant to say arrives.

Cennino Cennini knew this in 1400. He just knew it about different materials.

Il Libro dell’Arte — The Craftsman’s Handbook — is a practical manual for painters working in Florence at the end of the medieval period and the beginning of something else. It is not a scientific document, and that is probably why it survived. It is a record of what happens when you pay close attention to materials for a long time and write down what you notice.

On azurite, Cennini is precise. Grind it too fine, he says, and it loses its color. The blue goes gray. Stop before that happens. What he is describing — without the vocabulary — is the relationship between particle size and light scattering in a crystalline pigment. Coarse azurite particles scatter light differently than fine ones. The color lives in the grain. Grind past a certain threshold, and you have destroyed the optical structure that made it blue in the first place. Cennini knew this. He learned it the way I learned whatever I learned at my desk. By watching it happen.

The painters of that period were working with a mineral palette that required this kind of knowledge for every color they mixed. Azurite — Cu₃(CO₃)₂(OH)₂, monoclinic, sensitive to grinding pressure and to the chemistry of the binding medium. Smalt — cobalt-doped potassium silicate glass, ground to a blue powder — was never as stable as it looked. The blue exists only while cobalt ions sit in tetrahedral coordination within the glass matrix. When potassium leaches out, driven by moisture and the oil medium itself, that geometry shifts to octahedral, and the blue doesn’t fade — it structurally ceases to exist. The dominant wavelength of Rembrandt’s Night Watch smalt has drifted from deep blue toward orange-yellow. Almost none of the original blue survives. Cinnabar — α-HgS, hexagonal, brilliant red, with a tendency under prolonged light exposure to reorganize into β-HgS, cubic metacinnabar, near-black. Vermilion passages in paintings that are now dark brown were once red. The crystal changed phase inside the painting.

The substitution they didn’t mention:

There is an economic layer to all of this that sharpens the picture considerably. Lapis lazuli — the source of ultramarine, the blue that medieval and Renaissance painters reserved for the Virgin’s robes — was worth more than gold by weight in 15th-century Florence. It came from a single source: the Sar-e-Sang mines in what is now Afghanistan. Every gram of it that arrived in a Florentine workshop had crossed mountains, trade routes, and hands that each took a percentage. Patrons who commissioned religious paintings specified lapis lazuli by name in the contracts. They could verify it by the quality of the blue. They were paying for a critical mineral, and they knew it.

Painters knew this too. So they used azurite — local, cheaper, optically similar enough to satisfy a patron who wasn’t looking closely — in the background passages, the secondary figures, the shadowed folds of the robe where the lapis would be less examined. The substitution is documented in both the contracts and the conservation records. It was an economic decision encoding a precise mineralogical judgment: this pigment will read as close enough in this context, under these conditions, at this viewing distance.

That is materials science conducted under financial pressure in a competitive market. It is also the kind of knowledge that does not survive in written form, because no painter was going to write down the substitution guide. You learned it by working alongside someone who already knew.

The room where it was transmitted:

Vespasiano da Bisticci ran the most important bookshop in 15th-century Florence. His clients were popes and princes. His books were made by hand — copied by scribes, illuminated by miniaturists who ground their own pigments and mixed their own inks, bound in materials that required their own set of mineral and chemical knowledge to work correctly. The shop was also a workshop. The knowledge it contained was not only in the texts on the shelves but in the hands of the people producing them.

I am reading Ross King’s account of Vespasiano right now — The Bookseller of Florence — and what strikes me is how precisely the material knowledge of that workshop maps onto what Cennini was describing a few decades earlier in a different room across the same city. The illuminators working for Vespasiano were making the same decisions about the same minerals as the painters. Azurite or lapis. Which blue for which passage. How fine to grind. Which medium. The manuscript tradition and the painting tradition were drawing from the same empirical well.

Then the printing press arrived.

By 1480, Vespasiano’s workshop was finished. The technology that replaced it was faster, cheaper, and capable of reaching readers who could never have afforded a hand-copied manuscript. What it could not do, at least not immediately, was transmit the embodied knowledge of the people who had been making those manuscripts. The scribes scattered. The illuminators found other work or didn’t. The accumulated material intelligence of several generations of practice — what pigment survives parchment, what ink eats the page, what blue holds and what blue shifts — began to exist only in the hands of whoever still had it, which was fewer people every decade.

The same rupture happened to the painting tradition in 1841, when Winsor & Newton began selling pre-ground pigments in metal tubes. The grinding room started to empty. The apprentice who had spent years watching azurite lose its color under the muller, who had learned by failing and adjusting and failing again, was no longer necessary. You could buy the color already made, already stable, already optimized by a manufacturer who had done the mineralogy so the painter didn’t have to.

Both disruptions were called progress. Both were. But both also severed something that had been accumulating for a very long time — quietly enough that most people didn’t notice until they went looking for what was gone.

You Already Know This

You have never ground a pigment. Probably. You have never matched a nib to a paper by feel, never watched a blue go gray under the muller, never held a piece of parchment up to the light to read the fiber before committing ink to it. Most people haven’t. That’s not a failure of education. It’s just the condition of living downstream of two disruptions — the printing press and the paint tube — that made those skills unnecessary for most purposes and invisible to almost everyone.

But you know something about this anyway.

You know it if you’ve ever chosen the wrong paper for a printer and watched the ink bleed. You know it if you’ve painted a wall and learned that the cheap roller leaves a texture the expensive one doesn’t. Or maybe you’ve just cooked with a cast iron pan long enough to stop treating it like every other pan — it holds heat differently, it needs different handling, and it rewards that attention in ways a non-stick surface never asks of you. None of that is mineralogy. All of it is the same epistemology — knowledge built through contact with a material that has its own properties, its own responses, its own way of recording what you did to it.

The painters knew this at a scale and depth most of us will never approach. But the knowing itself is not foreign to you. You’ve been doing a version of it your whole life. The formal vocabulary came much later, after most of the hard work was already done.

Here is what the hard work left behind. Not finished objects. Not frozen moments. The paintings in the great collections of Europe are still doing what the materials were always going to do — changing, reacting, recording. The painters set them in motion. They haven’t stopped.

They are running experiments. The lead white passages are slowly growing crystals that were never part of the original paint layer — lead soaps forming as the pigment reacts with the oil binder over centuries, migrating toward the surface and creating physical protrusions visible under microscopy. The smalt skies are losing their blue as potassium leaches out of the glass network and the cobalt coordination environment shifts. The vermilion passages in some paintings have been converting to metacinnabar for three hundred years and will continue to do so. The conservation scientists at the Rijksmuseum, the Getty, and the National Gallery are not preserving finished objects. They are monitoring ongoing transformations in materials that were set in motion the day the last brushstroke dried.

The painters knew transformation was coming. They didn’t know the mechanisms. But they made choices — about which pigments to trust, which grounds to use, which media to bind with — that reflected centuries of empirical observation about which materials held and which didn’t. The ones that held are the ones we still have. The record of their judgment is the survival of the object.

This isn’t a metaphor. A materials selection process, conducted over generations, tested by time — the proof is that you can stand in front of a Raphael Madonna more than five hundred years later and see more or less what he saw. The azurite has shifted in some passages. The lapis held. He knew it would.

Science builds on what came before. Art earns its keep by breaking it. I’ve spent my career on the first and watched the second from a safe distance — until I didn’t.

What I Took From It

I watched it on a screen. Julie’s installation, the borax dying, the tincalconite forming in its place. Two true things that didn’t add up to a single clean answer.

I’ve been thinking about that ever since.

Science taught me to build. Each experiment grounded in the last one, each claim supported by the previous one, the whole structure load-bearing, accountable, and slow. That’s not a complaint. That’s how it’s supposed to work.

But there’s a version of science I was taught that I’ve never actually seen practiced. The clean sequential steps. The neutral observer. The hypothesis that waits patiently for its test. No scientist I know works that way. You follow a hunch. You ruin an experiment and learn something the successful version never would have shown you. You find the answer to a question you weren’t asking. The method is the story you tell afterward, not the account of what it felt like to find it.

I was told to stay in my lane. My advisors weren’t wrong about focus. But everything that has mattered in my work has been at the edges — in the gaps between disciplines, in the places the map marked as uninteresting, in the questions nobody in my field thought to ask because they were arriving from the wrong direction. I kept going there anyway. I’m still going there.

I stopped shrinking the question to fit the lane.

Watching Julie’s borax become tincalconite, I felt something I hadn’t expected. Not recognition exactly. Something adjacent to it. The material followed its conditions somewhere nobody planned for it to go.

I don’t know what’s forming on the other side of this. I just know the old structure is losing water molecules.

I am the Curator of Mineral Sciences at the Natural History Museum of Los Angeles County, a former scientist at NASA’s Jet Propulsion Laboratory, and an adjunct professor at USC. I write Pocketful of Χtals because mineralogy is stranger and more alive than most people have been told.

If this post put something in your head you can’t fully resolve, that was the intent. Share it with someone who would argue with you about it.

Zhytomyr Oblast

In 1931, the Soviet state began mining a classified deposit in northern Ukraine. The mission was quartz — but not just any quartz. A radio transmitter and its receiver lock onto the same frequency because a precisely cut quartz crystal vibrates at an exact, unwavering rate — without it, one end drifts, and communication breaks down. Inclusions, internal stress, growth defects — any imperfection corrupts the signal. The crystals have to be nearly flawless and large enough to cut. Most quartz in the world isn’t. The deposit at Volodarsk-Volynski was. For decades it supplied the Soviet military with the raw material to keep its communications from drifting. The West had solved this problem with synthetic crystals by the 1970s. The Soviet Union kept digging.

The mine employed over a thousand miners and sixty geologists at its peak. Six main shafts. Three levels. 110 kilometers of cold, wet tunnels running beneath the farmland of Zhytomyr Oblast. It was a state secret — Western scientists weren’t allowed underground until 1995, just as the operation was shutting down, the men who had worked those tunnels earning between ten and twenty dollars a month for the privilege.

A pegmatite is what happens at the end of a granite’s life. As magma cools and crystallizes, the common minerals claim their elements first — feldspar takes the potassium, quartz takes the silica, mica takes what’s left. Everything that doesn’t fit gets pushed forward, concentrated into the remaining melt. Water. Fluorine. Rare metals. Elements that have been wandering through kilometers of rock for millions of years, never finding a home, now crammed together into a shrinking pocket of fluid that is becoming stranger and more chemically extreme with every passing degree of cooling. By the time this residual melt crystallizes, it is so enriched, so volatile-saturated, so unlike ordinary granite that the minerals it produces are outsized, rare, and occasionally extraordinary. The crystals don’t grow in millimeters. They grow in meters. This is a pegmatite — not a rock type so much as a last act, the Earth concentrating everything it couldn’t place anywhere else into one final, spectacular crystallization event.

Those cavities are called pockets. The name is accurate — most of them are roughly the size of the pocket in your pants. A few grams of mineral, a geological curiosity. But at Volodarsk-Volynski, the pockets were something else entirely. The Korosten pluton, the 1.77-billion-year-old granite body that hosts these pegmatites, produced cavities on a scale with no real parallel on Earth. Some were the size of a room. Some were the size of a ballroom — the largest measured pocket ran to nearly 8,000 cubic meters, with quartz crystals hanging from the ceiling like stalactites, some weighing a ton each. Sealed. Undisturbed. Growing in isolation for so long that the number becomes meaningless — hundreds of millions of years of nothing happening, and then a Soviet drill bit breaking through the ceiling.

Men who spent weeks underground at a stretch — working twelve hours, sleeping twelve hours, same clothes, cold water seeping through the walls — occasionally came out with something in their pockets. Not quartz. The quartz belonged to the state. But a piece of orange topaz, small enough to carry, extraordinary enough to remember — that was worth something to someone, somewhere. This is not ancient history. Miners at the Boulby potash mine in Yorkshire still pocket volkovskite, one of the rarest minerals on Earth, and sell it on Etsy. The director found out. He didn’t particularly care. Some things are too beautiful to leave in the ground.

The Soviet version carried different stakes. The mine was classified. Taking material from a state installation — even a crystal the state had no particular interest in — was theft from the people, which in Soviet legal terms was not a minor matter. There was no HR department. There was no grievance procedure. And the mine administration was not oblivious to the value of what was down there. Their cleverest tactic for controlling unauthorized collecting was to surround an empty pit with barbed wire and post signs reading “No Trespassing — Violators Will Be Prosecuted.” Collectors crawled under the fence at night and dug frantically for hours. The pit contained nothing. The real pockets had no signs. The people running this operation understood exactly what the gems were worth, which means the miner who slipped something orange into his pocket understood exactly what he was risking. A man caught stealing from a secret mine had limited options for explaining himself, and the apparatus available to deal with him was not known for proportionality. The crystals that reached the Western market in the 1980s arrived wrapped in towels, carried out by dealers who described themselves as visiting on other business, sold through back channels at flea markets in Moscow. The beautiful things that came out of those tunnels came out carefully.

By 1991 the Soviet Union was gone. The mine kept running, the tunnels stayed cold, but the classified designation evaporated with the state that had imposed it. What came out of the ground after that moved differently — still informally, still through the networks that had formed in the dark years, but without the apparatus that had made those networks necessary. Exporting was never simple — inspections, mountains of paperwork, punitive payment laws — but the existential risk was gone. Dealers who had spent years building relationships in that world were perfectly positioned. The tunnels that had once required nerve to enter were now, simply, a mine.

Topaz forms in the most evolved, fluorine-rich part of these systems — late-stage fluids chemically aggressive enough to dissolve earlier-formed minerals and replace them. It crystallizes after most of the beryl is already gone, in solutions that are simultaneously the most extreme and the most productive in the pegmatite’s history. Most Volodarsk topaz is colorless to pale champagne. Blue occurs naturally in some crystals, its color the product of the same long irradiation that colors everything in these rocks. Imperial topaz — that specific orange-pink, saturated, warm — requires the right trace chemistry and the right radiation history and the right duration, all in the same crystal. Only about ten percent of the pockets at Volodarsk contained topaz at all. Imperial color represented a fraction of that fraction. The crystal that came out of Pegmatite No. 206 around 1993 was, by any measure, not supposed to exist.

Sometime around 1993, from a pocket in a pegmatite body that the Soviet geological survey had cataloged as No. 206, a crystal emerged from the Zhytomyr tunnels. It was imperial topaz — the rarest color variant of a mineral that forms at temperatures between 360 and 540 degrees Celsius, in fluorine-rich solutions so chemically aggressive they dissolve beryl on contact. The orange-pink color is not a coating or a treatment. It is the product of iron and chromium traces in the crystal lattice, irradiated by the surrounding granite for 1.77 billion years. You cannot manufacture that color. You can only wait for it.

Eye-clean. Its natural termination geometry — a form that would later matter enormously — created an interior that functioned almost like architecture when lit from below. And somewhere inside it, invisible until you knew to look, was a feather: a healed fracture plane, a record of a growth interruption that happened deep in the Archean. The crystal cracked, kept growing, sealed the wound. What remained was a thin planar inclusion, catching light differently than the surrounding topaz, floating inside the stone like a smear of breath on cold glass.

A physicist would eventually look at that inclusion and see a wing.

The Physicist

Konstantin Puzakov was not a gem carver by training. He was a physicist. He worked at the Moscow Engineering and Physics Institute — part of the Soviet state’s military-industrial atomic complex. In 1976, an accident left him unable to continue. He was in his early forties, with time, a small apartment on the outskirts of Moscow, and a background in optics that most artists don’t have and most jewelers wouldn’t think to use.

He turned to stone.

He couldn’t buy cameos in the Soviet Union — access to gems was tightly controlled, restricted to a licensed few. So he went to the Hermitage and studied what was there. He started collecting whatever he could find. His apartment filled up with stones. He taught himself mineralogy. He began carving with dental instruments because there was nothing else, then graduated to custom diamond-tipped tools, ordered one at a time from an acquaintance at a local institute — specifying the exact diameter and taper for each cut. The resulting collection, stashed in a corner of his apartment, looked like a dentist’s tray waiting for its next patient. Most of his work took place under a microscope, etching life into stone one grain at a time, which is not a metaphor.

What separated Puzakov from other carvers wasn’t patience — though he had extraordinary patience. It was that he refused to treat the stone as a surface. Most cameos work with color: you carve into a layered stone, and the image emerges from the contrast between strata. Puzakov wanted something the stone’s interior could do that its surface couldn’t. He had spent his career thinking about how light behaves inside materials — how it bends, reflects, focuses, scatters. He understood that a facet is a lens, that an etched surface is a diffuser, that depth perceived through a transparent crystal is not an illusion but an optical fact governed by refractive index and the geometry of the cut.

Under his hands, the interior of a stone became a space. Not a representation of space — an actual optical volume, with foreground and background, light entering from one angle and exiting from another, surfaces that caught and redirected illumination the way a physicist would calculate, not the way an artist would guess.

He worked almost exclusively in topaz, beryl, and ruby. Hard stones with high refractive indices, stones whose interior geometry could be controlled with precision. He had a clientele that ran from private collectors to the highest official of the Russian Orthodox Church. His secular subjects — Dante, Nefertiti, a Russian prima ballerina — carried the same quality the Lapidary Journal would later describe as an inner light. When he carved a scorpion into an orange topaz against a background of the constellation Scorpio, he put the scorpion and the stars on opposite sides of the stone and let the crystal’s own optics collapse the distance between them.

Once, carving a crucifixion into a small citrine, he looked up to find the stone filled with red light. He didn’t touch it. Sat very still. Then he traced it — the setting sun had reflected off a red-covered book on his bookcase, through his apartment window, and into the stone. He understood immediately what had happened, which didn’t make it less extraordinary.

No one has been able to find Puzakov in recent years. Death records have been searched. Contacts in Russia have looked. AI has been deployed. He remains, for now, unfindable. You have almost certainly never heard of him. The work is in museums. The artist is missing. These things are related.

Share

Golden, Colorado

In the mid-1990s, a man named Alex Grizenko was importing colored stones from the former Soviet republics into the United States. He had been doing it since 1991, building a business — Russian Colored Stone Company, based in Golden, Colorado — around materials the Western gem market had barely seen: alexandrite, chrome diopside, demantoid garnet, heliodor beryl, topaz from the Volodarsk-Volynski deposit. He was among the first Americans to recognize what was coming out of those tunnels after the Soviet state collapsed and the classified designation evaporated. By 1995, he was showing imperial and bicolored topaz at the Tucson Gem Show and drawing serious attention from gem writers and photographers.

By the 1980s, “blue topaz” had become shorthand for cheap — colorless stones irradiated in nuclear reactors, selling for a few dollars a carat. What Grizenko was showing at Tucson was something else entirely. You can see the difference immediately. One color came from a reactor. The other took a billion years.

He had also discovered Puzakov. The first piece Grizenko purchased was Venus on the Half Shell — imperial topaz, Puzakov’s work, acquired directly. His reaction, as he remembers it: “Let’s get into carvings.” He commissioned three more. Two more were discussed after that but never made — they couldn’t find imperial topaz of sufficient quality again. The third commission was The Resurrection.

The 1.2 kilogram imperial topaz crystal came to Grizenko through his gemologists in Moscow — he never held it himself before it went to Puzakov. He had only their word over the phone that it was the finest specimen of imperial topaz from this region they had ever seen. On that basis alone he acquired it. A month after acquisition, certain faces were polished to Puzakov’s requirements. Then it went to the artist with one straightforward request: use as much of the crystal as possible. Grizenko also asked for something non-religious — given Puzakov’s well-documented inclination toward religious subjects, it seemed worth saying. Whether that message was ever clearly conveyed to Puzakov is something Grizenko, thirty years later, is still not entirely sure about.

Puzakov studied the crystal for an extended period before he touched it. This was his method — he needed to understand the stone’s geometry completely before committing a single cut. Where the light entered, how it traveled, where it pooled. A mistake in a stone this size and quality was not recoverable. The crystal’s naturally etched, prismatic surfaces created a cavernous interior space when lit from the sides and below, with the color deepening toward the base. The natural indentation on one face read, under the right light, as an opening. And then there was the feather — and Puzakov, when he finally rotated the crystal into the right light, understood immediately what it was for.

From one angle it was nearly invisible, a hairline caught in the amber interior. Tilt it slightly, and it caught the light differently — a thin bright plane floating inside the stone, old as anything on Earth.

He built the entire composition around it.

The carving that emerged — executed in intaglio and reverse intaglio, the figures receding into the stone rather than projecting from its surface — depicted a scene set inside a cave. Three women approach an entrance. An angel sits on an open tomb. The steps, the tomb, the serpent beside it are all reverse intaglio, cut into the interior with diamond-tipped tools under a microscope, visible only because Puzakov had calculated exactly how light would travel through the stone to illuminate them. The angel’s wing is the healed fracture. He didn’t carve it. It was already there, waiting since the Archean, and he built everything else around it.

On the back face of the crystal, visible from the front through the stone, a figure rises. The composition is complete when you rotate it—front to back. The stone contains the whole narrative, each element placed where the crystal’s own optical geometry carries it to the eye.

The finished piece was shipped to Denver in 1996. Grizenko was shown it deliberately — from the bottom first, to extract bewilderment, then the back face where he met Jesus, then the front where he became, as he describes it, instantly amazed. His captivation has now lasted thirty years.

He had asked for something non-religious. He got the Resurrection. The carving was exhibited at several venues. In 1999, Lapidary Journal sent its leading writer to Moscow to interview the artist. The resulting piece called Puzakov “The Illuminated Man.” His works were in museums and with private collectors. He remained largely unknown outside a small circle of gem specialists — which is its own kind of crime, though a quieter one than what happened to that 117-kilogram topaz crystal they found in 1965 and didn’t bother to preserve.

What the Flaw Became

There is a version of this story where the crystal is just a crystal. Pulled from a sealed Archean pocket after 1.77 billion years in the dark, passed through the informal economy of a collapsing Soviet state, acquired by an American who trusted his gemologists’ word over the phone. It ends in a display case, and that’s fine.

But Puzakov looked at the flaw and saw the composition. Not metaphorically — literally. The healed fracture plane was the optical element around which everything else was organized. Remove it, and the carving collapses. The wing is not carved. It was there before Puzakov was born, before the Soviet Union existed, before there were humans to notice it. He found it, understood what it was doing inside the stone, and built around it.

This is what mineralogists mean when they talk about reading a crystal — not mysticism, not projection, but a trained understanding of how a specific material records its own history. In inclusions. In growth interruptions. In the geometry of fracture and healing. Puzakov had a physicist’s fluency with light and a carver’s patience with stone — a combination that the history of gem carving has produced in almost no one else who could claim it.

Alex Grizenko is now president of Lucent Diamonds, a company he founded in 2000 known for breakthroughs in diamond synthesis and the engineering of nanodiamonds for quantum sensing and processing. The applications depend on precisely engineered imperfections in the crystal lattice — defects placed at exact atomic locations — because a perfectly flawless diamond is useless for quantum sensing. The defect is the functional element. The diamond has to be nearly perfect everywhere except exactly where it isn’t.

Grizenko has thought about the parallel between the flaw in the topaz and the engineered defects in his quantum diamonds. He finds it resonant — and yes, a stretch, but good for storytelling. Which is honest, and worth saying. The connection he draws himself is cleaner: gemology, crystallography, physics. Adornment with crystals feeds the soul, he says. Engineered lattice defects in diamonds create new quantum sources of information. He sees no contradiction.

The arc is stranger than it looks. He started with colored stones from a classified mine in Ukraine, ended up growing diamonds atom by atom for navigation systems and medical diagnostics, and somewhere in the middle commissioned a carving that has captivated him for thirty years. The flaw in the topaz was accidental — a crack that healed in the Archean and waited. The flaw in the diamond is designed. The point, either way, is the flaw.

The Resurrection is coming to the Hixon Gem and Mineral Vault at the Natural History Museum of Los Angeles County. When you stand in front of it, you will see a carving of extraordinary technical and artistic accomplishment — intaglio and reverse intaglio figures inside a crystal that glows orange-pink from its own internal geometry, lit from within by a man who understood exactly how light moves through topaz, and all of that is worth the trip on its own.

Less visible is the old fracture floating inside the stone that started everything. It appears, from one angle, as a hairline. Rotate the crystal slightly, and it catches the light.

That is the wing — where Puzakov began, and where everything else followed from.

About 1.77 billion years in the ground. Thirty years above it, adorned with symbolic carvings. And now here, in Los Angeles, in a vault built for things that deserve to last.

Sources: Lyckberg, P., Chornousenko, V., and Wilson, W.E. (2009). Volodarsk-Volynski, Zhitomir Oblast, Ukraine. The Mineralogical Record, 40, 473–506. Katz, R.V. (1999). The Illuminated Man. Lapidary Journal, December 1999. Federman, D. (1995). Bi-Colored Topaz: Dual Delight. Modern Jeweler, June 1995. Personal communication: Alex Grizenko, 2025.

Not the crystal. The crystal, once you wipe the surface, is black. The green is on the outside — a film, nanometers thick in places, adhering to one population of garnets at this locality while leaving another population completely alone. Two garnets, side by side. Same habit, same hand-specimen color from a meter away — same approximate size if you were being careful about it. Put them under the microscope and one of them is doing something the other isn’t.

That asymmetry is what this post is about. Specifically: what an X-ray fluorescence spectrometer told me when I pointed it at both, and what it declined to answer.

Read Part I to Catch Up

What the Instrument Is Actually Doing

XRF works by firing X-rays at a sample and measuring the signal that comes back. Each element responds at a characteristic energy — iron at iron’s, calcium at calcium’s — and the instrument sums everything in the beam path into a single spectrum.

The beam on this instrument is 1.2 millimeters in diameter. It penetrates hundreds of microns into the sample. That matters here, because the coating on these garnets is extraordinarily thin — a film that produces optical interference colors, which means its thickness is on the order of visible light wavelengths: somewhere between 100 and 700 nanometers. A coating that thin is essentially invisible to XRF. The beam passes straight through it and reads whatever is below.

So when differences show up between the two populations, the coating isn’t what I’m reading. I’m reading the garnets.

I ran multiple measurements on each population. Seven spectra total — three on uncoated specimens, four on coated. The measurements are internally consistent within each population — which made the differences between populations harder to dismiss.

The Spectra Don’t Match

Here’s the spectral overlay. Blue is the uncoated population, averaged across three measurements. Orange is the coated population, averaged across four.

The dominant peaks — iron around 6.4 keV, manganese around 5.9 keV — are large in both spectra and broadly comparable. That's the garnet. Both populations are dominated by iron and manganese — the signature of Fe-Mn garnets in the pyralspite series (PYrope - ALmandine - SPessartine garnets). The Fe-to-Mn ratio differs between them, and that difference is one of the clearest signals in the dataset.

What’s different is everything else. Peaks around 2.0 and 2.3 keV are essentially absent in the uncoated spectrum; a feature at 3.7 keV that dwarfs the same region in the uncoated spectrum; a signal between 5.4 and 5.6 keV that the uncoated specimens don’t produce at all — four measurements, same result each time. Something is chemically distinct about these garnets.

The Complete Portrait of Population B

The quantified intensities put numbers to the spectral differences.

The uncoated garnets carry aluminum, silicon, chlorine, potassium, calcium, titanium, manganese, iron, and rubidium.

The coated garnets carry aluminum, silicon, phosphorus, sulfur, potassium, calcium, scandium, titanium, chromium, manganese, iron, zinc, and strontium. No chlorine. Rubidium either absent or dramatically depleted. K without Rb is a specific geochemical signature — more on that below.

Here’s what each element is telling me, because they are not all saying the same thing.

Phosphorus and calcium are elevated together — P at roughly 205 cps/mA in the coated population versus essentially zero in the uncoated population, and calcium roughly five times higher. Neither P nor Ca at these levels is surprising in the presence of apatite. Ca₅(PO₄)₃ is one of the most commonly documented accessory phases in vapor-phase and hydrothermal garnet assemblages. If Population B crystallized from a Ca- and P-bearing volatile phase that Population A never encountered, apatite micro-inclusions would explain both signals simultaneously.

Strontium tracks the calcium — 14 to 30 cps/mA across the four coated measurements, near zero in the uncoated population. Sr substitutes readily for Ca in apatite and Ca-bearing silicates. Its presence alongside the Ca anomaly is consistent: the same Ca-enriched volatile phase that produced elevated Ca would carry Sr in proportion. Sr has no structural site in almandine. Like K, it’s recording fluid history rather than garnet composition.

Sulfur is present at roughly 44 cps/mA in the coated population and only trace amounts in the uncoated. Sulfide phases are common accessories in vapor-deposited mineral assemblages in rhyolitic systems. But S can also concentrate in organic matter, and — I’ll say no more on this until the Raman post — there is reason to keep that possibility open.

Scandium at roughly 58 cps/mA in the coated population and zero in the uncoated is the most geochemically specific signal in the dataset. Sc has no structural site in almandine. It doesn’t concentrate in evolved felsic melts (think Mount St. Helens volcanic rocks) or sedimentary fluids in any meaningful amount. It partitions strongly into clinopyroxene in mafic rocks (think Hawaii-type volcanic rocks) and is mobilized by fluids that have equilibrated with those rocks. Its presence in Population B and its complete absence in Population A are strong evidence of different fluid-source chemistries.

Chromium at roughly 200 cps/mA in the coated, versus essentially zero in the uncoated, has the largest absolute difference among the diagnostic elements. Cr also has no structural site in almandine. Chromite (FeCr₂O₄), or chrome-spinel, as an accessory inclusion phase would point to a mafic component in the fluid or source-rock history. In a rhyolitic system where the volcanic suite runs from primitive basalt through to evolved rhyolites, a volatile phase that equilibrated with more mafic lithologies at any stage of that evolution could carry a Cr signature that a purely rhyolitic vapor would not.

Zinc at roughly 65 cps/mA in the coated versus zero in the uncoated. Zn is volatile-soluble and concentrates readily in magmatic vapor phases, particularly in evolved and rhyolitic systems. Zn-bearing accessory phases in vapor-deposited mineral assemblages are well-documented. Its presence in Population B alongside Sc and Cr points toward a vapor chemistry that Population A’s crystallization environment lacked.

P + Ca + Sr + Sc + Cr + Zn + S. Lined up like that, they tell a consistent story: Population B crystallized from a volatile phase that Population A never encountered. None of these elements has a structural site in almandine — they’re almost certainly sitting in accessory phases and inclusions, trapped during growth, recording the chemistry of that fluid.

The Fe and Mn signals deserve a direct comment, and they deserve it together because they are connected in ways XRF cannot fully resolve.

Mn averages roughly 27,150 cps/mA in the uncoated population and 21,850 in the coated — a 24% difference the coating cannot explain. That difference is in the garnets themselves.

The Fe picture is messier. The coated population averages higher but scatters widely across four measurements (20,800 to 26,150 cps/mA), while the uncoated population clusters tightly (21,932 to 22,472). That scatter is telling me something real about the coated garnets — but XRF has a specific limitation here that matters enormously for interpretation.

The leading hypothesis going into the next instrument is that the coated garnets are compositionally zoned — not homogeneous almandine, but almandine-dominant zones intercalated with zones carrying an almandine-spessartine component. Spessartine is Mn₃Al₂Si₃O₁₂; almandine is Fe²⁺₃Al₂Si₃O₁₂. If those zones are volumetrically minor relative to the almandine matrix, bulk Mn stays lower than in the uncoated population while the Fe signal scatters widely as the 1.2mm beam lands on different proportions of compositionally distinct zones across four measurements.

Why does the Fe/Mn ratio matter beyond accounting for the scatter? The coated population is more Fe-dominant — closer to pure almandine. The uncoated population is more Mn-dominant — closer to spessartine. Those are different surface chemistries. That difference may be exactly what the coating is reading.

That difference may be exactly what the coating is reading.

I will leave it there for now. The Raman post will say more.

K Without Rb Is a Very Specific Thing

Neither K nor Rb substitutes into almandine structure. Both signals are sitting in inclusions — K-feldspar, phlogopite, fluid inclusions — recording fluid history rather than garnet composition. Which makes the contrast between the two populations particularly telling.

The coated population has roughly four times more K than the uncoated, and essentially no Rb. The uncoated population has lower K and measurable Rb.

The K/Rb ratio is a standard geochemical discriminator between fluid sources. Primitive mafic rocks — basalt, mantle-derived melts — have high K/Rb ratios, typically around 1000, because Rb concentrates in felsic phases that are absent in primitive mafic systems. Evolved granitic and pegmatitic fluids have K/Rb ratios below 200, sometimes well below.

The coated population: mafic signature. The uncoated: more evolved, more felsic, K alongside Rb instead of K without it.

Two garnets, same locality, different fluid histories written into the trace elements. The coating finds one and leaves the other alone — and whatever it’s reading in the surface chemistry was decided when these crystals formed, not when the coating arrived.

Where the Elements Point

The Aquarius Mountains contain a mafic volcanic field. It is not a distant inference. The northern Aquarius Mountains volcanic field covers roughly 400 km² and includes basaltic flows and cinder cones dated to approximately 24–20 Ma. These basalts are primitive — high magnesium numbers consistent with direct mantle derivation — and the broader volcanic suite runs from them through latites and dacites to rhyolites, with K/Rb and Rb/Sr ratios that are distinctive and measurable at each compositional stage.

The garnets themselves are vapor-phase precipitates from a rhyolitic system. A volatile phase derived from a magma with any mafic input — mixing, contamination, or simply erupting through mafic crust — would carry Cr, Sc, and the high K/Rb character that Population B records. Population A, with its lower K/Rb and measurable Rb, appears to be a vapor pulse from a more purely evolved rhyolitic source. Two vapor pulses, two garnet populations, same cavity.

The spatial relationship between the volcanic field and the garnet locality needs field verification. The Aquarius Mountains are roughly 45 miles long, and establishing the fluid pathway between any mafic source rocks and the Lion Spring locality requires a map and boots on the ground, not just a range name. I am not claiming proximity. I am claiming that a mafic source consistent with the geochemical signal exists within the same mountain range, and that fieldwork now has a specific question to answer rather than a general region to survey.

That’s not a small distinction — going into the field with a specific geochemical target is different from going in to look around.

What the Microscope Added

Before the open questions, one more dataset.

The branching, dendritic front advancing into the bare garnet surface is not the morphology of mineral precipitation from solution, which tends toward uniform layers or crystallographic forms. It’s the pattern of systems growing under diffusion limitation — where the leading edge advances faster than the bulk material can supply. Two processes produce this morphology: diffusion-limited aggregation of large organic molecules in a viscous fluid, and biological colony expansion, specifically biofilm spreading across a substrate.

This is the fully coated face. The rainbow shimmer is thin-film optical interference. The coating is nanometers thick — in some patches perhaps 100 nm, in denser areas perhaps approaching 700 nm. This is not a crust, not a mineral precipitate in the conventional sense. It’s a film — and at the margin, it looks like something that spread.

What it’s made of is the subject of the next post. I will say only this: Raman spectroscopy found more than one phase in that coating, and at least one of them was not on my list of candidates going in. If you’re assuming iron oxyhydroxide, you’re not wrong — but you’re not finished either.

Where XRF Runs Out of Road

The instrument has a hard edge, and it’s worth knowing where it is.

XRF integrates everything in the beam path — garnet lattice, micro-inclusions, surface phases, fluid inclusions, all of it summed into a single spectrum. Phosphorus could be in apatite inclusions trapped during crystallization, or in the coating, or in fluid inclusions at grain boundaries. Chromium could be in chromite grains enclosed within the garnet, or concentrated at the surface. The Fe scatter across four measurements could be zoning — almandine-dominant zones intercalated with almandine-spessartine zones at a scale the 1.2 mm beam averages over rather than resolves. Each of these is a testable hypothesis. None of them can be distinguished from this data alone.

What it cannot tell me is how that iron is distributed spatially — which zones are Fe-dominant, which are Mn-dominant, and whether the coating preferentially covers one over the other.

That’s not a flaw in XRF. It’s a precise description of what it’s designed to do. Without it, there’s no chemical distinction to explain, no fingerprint to interpret, no reason to ask about iron oxidation states, no specific question to take into the field. The question doesn’t exist yet.

But XRF has now answered what it can answer. The question it leaves open — where exactly are these elements, in what phases, in what oxidation states, and how are they spatially distributed — requires tools that resolve chemistry at scales far below 1.2 mm.

Raman spectroscopy is next. It identifies mineral phases and organic compounds by their vibrational signatures and can target individual spots on a surface, separating coating from substrate and — as it turns out — revealing something about the internal structure of these garnets that the XRF beam was averaging over entirely. After Raman, SEM/EDS provides sub-micron spatial resolution, imaging the coating architecture in cross-section and identifying the chemistry of individual phases at the garnet surface and within the garnet interior. And field geology closes the loop on the fluid pathway: where, specifically, did these volatile phases come from, which rock did they equilibrate with, and why did one pulse grow a garnet that a later fluid would find and coat, while the other did not?

Each instrument resolves a different layer of the same question, and none of them alone gets there. The chain is the method.

The next post will cover the Raman and scanning electron microscopy data. I’ll warn you now: it does not simplify things.

Specimens from the Aquarius Mountains, Mohave County, Arizona. XRF collected on a Horiba XGT-7200, 30 kV, 30-second live time, 1.2 mm beam diameter.

Part 1 of this series: [link to Part 1]. Part 3 (Lasers): coming.

I was not expecting to stop.

Subnautica 2 had been out for only a few days. I was deep in a late-game biome called the Root Canyon — the Metal Farm — roughly 1,000 meters east-northeast of where the game starts, which, in an alien ocean, feels genuinely remote. You need special equipment to get there. The game does not hold your hand at this depth. I was scanning mineral deposits the way you do in survival games, half on autopilot, building inventory.

Then a word appeared on screen.

Troilite.

I stopped swimming.

Not because it was rare — though it is, the rarest mineral Unknown Worlds placed on this alien ocean floor. I stopped because I knew what that word meant. I knew it the way you know the name of something you once had a very bad night with.

The last time I encountered troilite’s close relative, I was a graduate student at Stony Brook University on Long Island, it was eight or nine o’clock at night, and by the following morning, I had very nearly burned the geology building to the ground. I cleaned up the evidence alone. I was terrified. If anyone noticed, they were generous enough never to say so.

That was decades ago. I had not thought about that night in a long time.

A video game brought it back in about four seconds.

(If you don’t play video games, stay with me. The game is just where the mineral turned up. The mineral is what matters. It always was.)

You have never seen it in a rock. Almost nobody has. It’s not a question of rarity in the universe — troilite is common in meteorites, a major constituent of lunar samples, and a likely component of planetary cores throughout the solar system. The problem is time, pressure, and where you are standing.

The FeS system is not a single mineral. It is a family of minerals — different crystal structures, different stabilities, different addresses in the phase diagram — each one stable under a specific set of conditions and unstable everywhere else. Mackinawite forms first at the surface in cold water under ambient pressure. It is where the system begins. Troilite comes later, deeper, under heat and pressure — the stable form that planetary interiors produce when they have the time and the weight to do it properly. Push troilite deeper still, and it transforms again, and again, through at many distinct phases, all the way down to conditions at the center of some rocky worlds.

Each phase is an address. The mineral that exists at that address is the one that physics allows.

Our atmosphere is the reason you can’t find these minerals here. Not because it destroys them violently — troilite is stable enough to sit in a museum collection, to survive handling, to hold its structure in a specimen drawer. The problem is geological time. Given enough of it, oxygen works its way into the surface, iron oxidizes, sulfur escapes as sulfate, and what was troilite becomes something else entirely — a weathering rind, a rust stain, a ghost of the original phase. High-troilite meteorites left in humid environments rust and disintegrate over years. Severely weathered ones show complete replacement of troilite by iron sulfate. The atmosphere doesn’t combust it. It simply, patiently, wins.

This is why you only find troilite in meteorites — objects that spent billions of years in the oxygen-free cold of space and arrived here recently enough that the erasure isn’t finished yet. The specimen in the NHMLA collection carries a fresh dark surface and, if you look closely, the beginning of a greenish weathering rind where the atmosphere has already started its work.

Our sky is not violent toward troilite. It is relentless.

Roughly 2.4 billion years ago, photosynthetic life flooded Earth’s atmosphere with oxygen — the Great Oxidation Event — and minerals like troilite retreated from the surface permanently. Not in an instant. Over millions of years, the argument tipped. Down into the interior. Into the deep ocean floor. Into space. To find troilite you have to go somewhere the atmosphere hasn’t reached, or hasn’t had enough time.

Subnautica 2 is set on an alien ocean world. The game placed troilite 2,000 meters from the surface, in the most remote biome, locked behind a tool you have to build specifically to extract it. No oxygen down there. No explanation offered to the player.

Whoever named it knew exactly where it belonged.

Here is what happened at Stony Brook.

Mackinawite is troilite’s close relative — the iron sulfide phase that forms at the surface, in cold water, at ambient pressure. The low-address member of the family. Where troilite loses the argument with oxygen over geological time, mackinawite doesn’t wait that long. It is genuinely reactive, genuinely dangerous under the wrong conditions, and I was making it by aqueous precipitation in a glove box because I needed to understand the iron sulfide family from the inside out. You cannot understand a mineral by reading about it. You have to make it. You have to be in the room with it.

The synthesis required a glove box — a sealed chamber filled with nitrogen, no oxygen, where you work through thick rubber gloves built into the walls. I had grown the mineral in solution, filtered it out as a fine black powder, and set the container aside inside the box, still under nitrogen atmosphere. Standard procedure. I went home.

I came back the next morning.

The glove box was full of smoke. The container had melted. The black powder — my mackinawite, hours of careful synthesis — had essentially combusted overnight while I was gone.

I stood there for a moment trying to understand what I was looking at.

The culprit was a pinhole leak in the glove box seal. Somewhere in the gasket, invisible, there was a gap. Atmospheric air had been infiltrating the nitrogen atmosphere all night. Not much. The leak was so small that the box had appeared to be holding pressure. But mackinawite doesn’t need much. Parts per billion of oxygen had been enough to trigger an exothermic oxidation reaction that built heat slowly through the night until the container couldn’t hold it anymore.

I cleaned it up alone. I told no one. I was a first-year graduate student, and I had just discovered, in the most direct way available, what it means for a mineral to be at the wrong address. Mackinawite belongs in cold anoxic water, in hydrothermal sediments, in the kind of oxygen-free dark where the atmosphere has never reached. The atmosphere had gotten in. The atmosphere always wins.

What no textbook quite captures — what you can only learn by standing in front of a melted container at eight in the morning — is that these minerals are not passive objects. They are phases in an ongoing chemical argument between a planet and its atmosphere. On the early Earth, before photosynthesis, mackinawite and troilite were stable at the surface. Life changed that. Life flooded the sky with oxygen, and these minerals lost the argument and retreated. They have been hiding ever since.

I cleaned up the evidence of that argument and went back to work.

The mackinawite I made at Stony Brook eventually went to Sandia National Laboratory in New Mexico — to the LANCE beamline, a spallation neutron source. Neutrons see things X-rays cannot. I needed to know where the hydrogen sites were in the mackinawite structure, and how they changed under pressure. Specifically, deuterium — the heavy isotope of hydrogen, which neutrons can resolve precisely. Mackinawite at the bottom of a cold anoxic ocean, threaded with hydrogen, under pressure. These are the same conditions, the same minerals, the same chemistry that may have catalyzed the first biochemistry on Earth. Whether life began at hydrothermal vents is still an open question. The minerals were already there, doing what minerals do, whether or not anything noticed.

The near-disaster at Stony Brook didn’t stop the work. It focused it.

If these minerals were this insistent on belonging to specific environments — if mackinawite belonged to the surface and troilite belonged deeper — then the question of how deep troilite could go, and what it became when you pushed it further, was worth asking. The answer pointed inward. Into planets. Into the cores of rocky worlds where pressure is measured in gigapascals and oxygen has never reached and never will.

Which is how I ended up at Daresbury Laboratory in Cheshire, England, watching the status of an electron storage ring on a pub television instead of a football match.

Daresbury was one of the world’s synchrotron facilities — a particle accelerator built in a ring roughly the size of a city block. It accelerated electrons to near the speed of light and harvested the X-rays they shed in the process. The beam that came out was orders of magnitude more intense than anything a standard laboratory can produce — intense enough to generate a readable diffraction pattern from a sample the size of a printed period, squeezed between two diamond tips under pressures equivalent to a planetary interior. A hospital X-ray machine would see nothing. Daresbury saw a crystal structure.

My experiment used a diamond anvil cell — two gem-quality diamonds, polished to fine points, positioned tip-to-tip with the sample between them. The troilite I started with had been synthesized in the lab under vacuum and high temperature — you have to recreate something close to planetary conditions just to make the starting material. Then you squeeze the diamonds together and the sample — a grain of iron sulfide roughly the size of a period at the end of this sentence — experiences pressures equivalent to the interior of a planet. You shoot the synchrotron beam through the diamond and read the diffraction pattern. The way the X-rays scatter tells you the crystal structure. As you increase the pressure incrementally, you watch the structure transform in real time.

What I was looking for: the behavior of FeS at pressures corresponding to the Martian core. Because the question of what is inside Mars is not purely a geology question.

It is a physics question. The orbital mechanics of Mars — how it moves around the sun, how its spin axis precesses over time — can only be fully reconciled if you know the density of whatever sits at the planet’s center. Get the mineral phase wrong, and the numbers don’t close. Get it righ,t and you have visited the interior of another planet without leaving England.

You do the experiment in shifts. The synchrotron runs continuously, and time on the beam is allocated in blocks. You work through the night when you have to. And periodically, the beam dumps — the accelerator releases its stored energy, the ring goes dark, the X-rays stop, and there is nothing to do. The experiment is paused. You cannot hurry a particle accelerator.

So you walk to the pub.

It is a short walk. The pub knows the Daresbury crowd. And on the television, where a normal pub in Cheshire would be showing football, the screen displays the status of the electron storage ring. The beam energy in megaelectronvolts. Whether it’s recovering. How long until the experiment can resume.

The beer was flat and warm.

I remember standing there thinking: what the hell am I doing here.

The answer, when the beam came back up and we walked back to the lab, was: trying to find out what’s inside Mars.

My data matched previous published work. No new phases. No surprises in the FeS phase diagram at Martian core pressures. The existing map was correct.

In science, replication is not failure — it is the foundation that makes knowledge reliable. Someone had been right before me, and now there was one more set of hands confirming it. That matters. That is how the edifice holds.

But it meant the Martian core question, along this particular line of investigation, was answered. I packed up the diamond anvil cell. My master’s program would need a new direction. I went home from England with clean data, a confirmed phase diagram, and a new question to find.

Subnautica 2 sold a million copies in the first hour.

Somewhere inside that studio, someone got the geology right. Troilite, 2,000 meters from the surface, in the dark, in an anoxic ocean, inside a mineralized clinker, locked behind a tool you have to build specifically to extract it. The rarest mineral in the game. No explanation offered. No in-game text about iron sulfide chemistry or the Great Oxidation Event or what this mineral means to the question of what’s inside Mars.

Just the word, in the right place, waiting.

A clinker, for the record, is the fused, glassy residue of high-temperature mineral processes — exactly the geological environment where iron sulfides concentrate. Whoever made this decision didn’t just pick an obscure mineral name. They put it inside the right rock.

I sent a message on Bluesky the same evening I found it. I told them I was the Curator of Mineral Sciences at the Natural History Museum of Los Angeles. I told them I nearly burned down a building working with troilite’s close relative in grad school. I told them I had used it to try to see inside Mars.

I asked: who told you about troilite?

They wrote back within hours. Troilite had gone into a mineralized clinker because that’s where it belongs. The art had been built to match the geological intent.

I was very happy with its depiction.

I am writing this for anyone who has ever followed a hunch into territory that wasn’t guaranteed to yield anything.

The experiment at Stony Brook could have burned down a building. It produced a melted container and a very quiet cleanup and a precise, bodily understanding of what it means for a mineral to be at the wrong address. The experiment at Sandia found the hydrogen. The experiment at Daresbury confirmed what was already known. The data matched previous published work. My master’s program needed a new direction.

None of it was failure. All of it was science doing exactly what science does — not marching forward in a straight line toward predetermined answers, but feeling its way through the dark, occasionally setting things on fire, occasionally sitting in a pub in Cheshire watching beam current on a television instead of football, waiting for a particle accelerator to restart so you can go back and keep asking.

Sometimes it’s new. Sometimes it confirms what came before and that confirmation is quietly, unglamorously necessary — the kind of result that doesn’t make careers but keeps the map honest. And sometimes you come back to the lab in the morning and the glove box is full of smoke and what you learn that day, alone and terrified, is something no textbook manages to convey: that these minerals are not passive objects. That they are phases in an ongoing argument. That the argument is still happening.

Someone sat down to design the resource economy of an alien world and asked what minerals belong in an oxygen-free ocean. They followed that question somewhere it wasn’t supposed to go for a game designer. They came back with troilite inside a mineralized clinker at 2,000 meters depth, and built it so faithfully that a mineralogist stopped swimming.

I went to Stony Brook and learned what the wrong address looks like from the inside. I went to Sandia and found the hydrogen. I went to England and squeezed troilite between two diamonds to find out what’s inside Mars. Nobody knew about the glove box. I didn’t tell anyone for decades.

A video game came out three days ago and put the word on an alien ocean floor in exactly the right place, and I stopped, and everything came back.

That is what it means to pay attention. You never know where the mineral is going to turn up next — in a glove box, in a diamond anvil cell, in a video game three days after it ships. Someone else might have been following the same question through a completely different kind of dark, and you’d never know unless you said something. I said something. They wrote back in four hours.

Aaron Celestian is the Curator of Mineral Sciences at the Natural History Museum of Los Angeles County, former scientist at NASA’s Jet Propulsion Laboratory, and adjunct professor at USC. He writes Pocketful of Χtals because mineralogy is stranger and more alive than most people have been told.

I was not prepared for the rocks in his mouth.

The call came through guest relations, the way it always does — a visitor requesting to speak with the curator of minerals. Something important. I get these calls. I’ve learned not to anticipate what walks through the door, but I still do. I made my way out to meet him, and before I could finish introducing myself, he reached into his mouth and placed several small stones on the table between us.

He had been carrying them there for safekeeping. Nobody could take them from him that way. And now he was showing them to me, which meant he trusted me, and I understood immediately how much that cost him.

I looked at the stones. I knew what they were — gravel, honestly, the kind of thing you’d kick off a sidewalk — but I picked them up anyway and turned them over. I told him what I thought I was looking at, which wasn’t much, and I told him I had no idea where he’d found them or what context they came from, which mattered more than anything I could say about the mineralogy. He seemed okay with that. He put them back in his mouth. He left, and I stood there for a minute before I went back to my office.

I’ve thought about that encounter more than almost any other in my career.

I have been the Curator of Mineral Sciences at the Natural History Museum of Los Angeles County long enough to know what’s coming before the bag opens. The shape tells you. The weight tells you. The way someone holds it against their body — careful, a little defensive, closer than they need to — tells you before they’ve said a word.

They have a meteorite. They have a diamond worth a fortune. A jade carving from an ancient civilization. A crystal of cosmic origin that has been in their family for three generations, and nobody ever knew what to make of it.

I know, before they open the bag, that I am about to disappoint them.

The meteorite is slag — industrial waste from a smelting process somewhere, heavy and dark and full of bubbles mistaken for the regmaglypts of atmospheric entry. The diamond is quartz. It is always quartz. Quartz is the most abundant mineral in Earth’s continental crust and, somehow, also the mineral most persistently mistaken for something rarer. The jade is soapstone. The crystal of cosmic origin is fluorite from a gift shop, probably purchased in Sedona.

I bring them into the lab anyway. Every time. I run the analysis and show them the instrument — the XRF, the Raman spectrometer, the polarizing light microscope — and I walk them through what the data says and how I know it says that and what assumptions I’m making when I interpret it. The conclusion is not the point. What I want them to leave with is the method — not just what they were holding, but how a scientist arrives at that answer. What careful measurement actually looks like. That we took their rock as seriously as anything else that has come through this lab, because we did.

I still hate the moment before I tell them.

What I’ve never said out loud — not to the person across the table, not in any paper, not anywhere — is that I understand completely why they came.

They found something that felt wrong in the best possible way. Too heavy. Too dark. Too strange for the place they found it. And they did exactly what humans have done with that feeling for as long as humans have existed — they reached for the most powerful explanation available to them.

Something extraordinary came from somewhere else. Something fell.

They don’t have a mass spectrometer. They don’t have 20 years of isotope-geochemistry papers and an electron microprobe. What they have is the oldest and most persistent hypothesis our species has ever generated about unusual materials, encoded in the only format that survived long enough to reach them.

A story.

And I am standing there about to tell them their evidence doesn’t support their conclusion — which is true — without ever acknowledging that the instinct behind the hypothesis is not wrong.

It has never been wrong.

It just took the rest of us a very long time to catch up.

The Wrong Diamond

I want to show you something in my collection that I bought from a reputable dealer, because the science kept bothering me in a way I couldn’t set aside.

It is ugly. I want to be clear about that upfront. It is black and lumpy and looks approximately like something you would find at the bottom of a driveway. Nobody has ever held a piece of carbonado and felt that particular sensation that comes with a gem diamond — the cold, the weight, the geometry of light moving through it. There’s no geometry here. No light. Just a dull, porous, polycrystalline aggregate of millions of microscopic diamond crystals fused together into something that has no business being called a diamond at all, except that it is one, incontrovertibly, and that’s where the trouble starts.

And yet.

Here is the first thing that bothered me: carbonado has never been found in kimberlite. Kimberlite is the deep volcanic rock that carries every other diamond from the mantle to the surface — the delivery system that makes diamond mining possible. You want to find diamonds, you find the kimberlite. This has worked reliably for over a century of industrial mining.

Carbonado ignores this entirely. It turns up only in alluvial deposits — riverbeds, ancient sediments — with no primary host rock. Nobody has found where it came from. It simply appears, already separated from whatever brought it here, sitting in river gravel like it arrived and decided to wait.

It has been found in two places on Earth: the Central African Republic and the Bahia Province of Brazil. The two cratons these locations sit on shared a common geological setting for more than 200 million years — joined through the supercontinents Nuna and Rodinia before the opening of the Atlantic separated them roughly 180 million years ago. The distance between them is consistent with what you might expect from a large bolide that fragmented during atmospheric entry — or possibly two related events arriving in proximity to each other. We don’t know which. We don’t know when. There is no primary source rock to examine, no return address, nothing that tells us how two of the most geographically separated places on Earth ended up sharing the same impossible mineral.

The mystery doesn’t resolve. I want to be honest about that before I go further. What follows is the most complete story the published literature currently offers, and it is still incomplete.

What the Scanner Sees

I have a micro-CT scan of a carbonado specimen, and I want you to look at it carefully before I describe what you’re seeing, because the image does something the description can’t quite do.

A gem diamond is essentially a perfect lattice. Carbon atoms are arranged with almost no defects, almost no empty space. That’s why it’s hard. That’s why it’s transparent — light has nowhere to scatter.

The carbonado is full of holes.

Micro-CT movie of carbonado specimen. Voids are gray areas, which are difficult to see in the video. The CT shows density differences. Bright areas are likely metal inclusions. Faded areas are the pore spaces.

The internal pore structure runs through the entire specimen — a network of voids that looks, under imaging, like something between a sponge and a meteorite fragment. This porosity is one of carbonado’s most diagnostic features and one of the most argued-over. Because the question isn’t simply that the pores exist. It’s what used to be in them.

One interpretation, advanced by Haggerty and colleagues, is that the pores represent the loss of hydrogen — specifically, hydrogen implanted by solar wind exposure during the time before carbonado arrived on Earth. If that holds up, the pores aren’t defects or damage. They’re a record. The shape of something that was there once and isn’t anymore.

What the Literature Says

The published isotope data on carbonado — measurements I did not make, from studies I have been reading and re-reading — tell a specific story. The carbon and nitrogen isotope signatures documented across multiple independent analyses do not match any known terrestrial formation environment. This is not a matter of the numbers being anomalous or surprising. The numbers are exactly what they should be for the conditions under which carbonado formed. The problem is that we cannot place those conditions anywhere on Earth.

Nitrogen is particularly telling. In most natural diamonds, nitrogen atoms have had time to migrate and cluster — pairing up, then forming larger aggregates — a process that proceeds predictably with temperature and time and gives geochemists a reliable record of formation conditions. In carbonado, the nitrogen remains as isolated single atoms. Whatever environment produced this material was cold enough, or different enough, that aggregation never proceeded. The diamonds formed somewhere outside the thermal envelope of anything we recognize as a planetary interior.

Then there are the inclusions, which tell the same story from a different direction. Diamonds from Earth’s mantle carry a characteristic set of passenger minerals — chromium-bearing garnets, magnesium-rich olivines — things that make sense coming from deep inside a rocky planet. Carbonado carries none of these. Instead: pure metallic titanium, silicon carbide, iron-chromium alloys. Phases that form only under conditions of extreme chemical reduction with no recognized analog in Earth’s interior.

Infrared spectroscopy measurements — FTIR, using synchrotron radiation, published by Garai and colleagues in 2006 in The Astrophysical Journal Letters — found that carbonado’s absorption profile resembles two things. Diamonds produced by chemical vapor deposition in laboratory vacuum conditions. And presolar diamonds recovered from inside primitive meteorites. The hydrocarbon signatures in carbonado resemble those seen in spectral observations of stars and interstellar nebulae.

There are currently several serious competing hypotheses for carbonado’s origin — deep mantle formation, meteorite impact, subduction of ancient organic matter, radiation-induced crystallization, and direct extraterrestrial delivery. None has achieved consensus. Some require a cosmic component. Some don’t. What none of them can fully do is account for everything the instruments show. The mystery is real, the debate is live, and the rock in my collection sits at the center of it.

Video footage of the Enigma carbonado at Sotheby’s Beverly Hills, Rodeo Drive, 2022. 555.55 carats, 55 facets, shaped as a hamsa. Watch the surface — the pores are visible even on the polished faces. For a still image that captures the surface porosity in extraordinary detail, see photographer Michael Bowles’ work at Only Natural Diamonds.

And the moment I understood what they were pointing toward, I thought immediately of every person who has ever handed me a rock in a velvet bag.

Because they were asking exactly the right question.

Did this come from somewhere else?

For this one specific, ugly, porous, driveway-looking specimen in my collection — the answer the published literature gives back is: yes. One way or another. The debate is about the mechanism, not the fact.

That is as close to settled as science usually gets on something this strange.

The Oldest Hypothesis

The question that carbonado forces on the scientific literature — did this come from somewhere else? — is not a new question. It is the oldest question our species has ever asked about a material. And for most of human history, the answer wasn’t a hypothesis. It was a story.

Something extraordinary fell from the sky. It landed in one place. It was harder than anything else, stranger than anything else, or more powerful than anything else. The gods sent it. It belongs to whoever found it and to the civilization that grew up around it. Handle it carefully. It changes everything.

We have been telling this story since before we had writing to tell it in.

Meteoritic iron is the most literal example, and I have written about it elsewhere in this series. For most of human prehistory, the only iron available was iron that had already been smelted at the core of a protoplanet — metal from the sky, harder than bronze, unlike anything that could be produced on Earth at the time. The oldest iron artifacts are not from mines. They’re from meteorites. The Egyptians called it bja-n-pt. Iron of the sky. They worked it into beads and buried it with the dead thousands of years before anyone figured out how to smelt terrestrial ore.

They knew it was different. They didn’t need a mass spectrometer. It felt different, behaved differently, came from nowhere anyone could point to.

So they pointed up.

Around 1200 BCE, the city of Troy possessed an object called the Palladium — a wooden statue, ancient even then, said to have been thrown down from heaven by Zeus himself at the city's founding. As long as the Palladium remained within Troy’s walls, the city could not fall. When Odysseus and Diomedes finally managed to steal it, Troy fell. The object from the sky was also a protection. Remove it, and everything it sustained collapsed.

The story is a myth. But the structure of the story is identical to every other account of something extraordinary arriving from above and conferring power on whoever received it. The details change. The architecture doesn’t.

What I find remarkable — and I didn’t notice this until recently — is what happened to the word itself.

In 1802, an English chemist named William Hyde Wollaston was dissolving platinum ore in acid and examining the residue. The kind of painstaking bench work that produces discoveries nobody sees coming. He isolated a new element and needed a name. Two months earlier, an asteroid had been discovered and named Pallas — after Pallas Athena, the goddess whose most sacred association was the Palladium, the object that fell from Zeus. Wollaston named his new element palladium.

He almost certainly wasn’t thinking about Troy. He was honoring a timely astronomical discovery, the way chemists did. He had no idea he was completing a three-thousand-year chain that began with a Bronze Age story about something falling from heaven and changing a civilization.

The chain runs: sacred object falls from the sky → goddess associated with that object → asteroid named for the goddess → element named for the asteroid. Wollaston closed the loop without knowing the loop existed.

The story encoded itself into the language of science without anyone noticing.

In 1966, Stan Lee introduced a fictional metal called vibranium into the Marvel Comics universe. It arrived as a meteorite. It landed in Africa. It had properties no known material possessed, and the civilization that grew up around it became the most technologically advanced on Earth — its power depending entirely on protecting the metal from the sky.

Stan Lee almost certainly had never heard of carbonado.

He wasn’t making a scientific argument. He was reaching for the most powerful story available to explain why some materials feel different, why some things seem to arrive already charged with significance, already belonging to a different order of existence than the rocks underfoot. He was telling the oldest story there is.

And he got the structure exactly right — arrived from space, from one place, with impossible properties, changing everything for whoever receives it. This isn’t a coincidence. It’s not plagiarism either. It’s the same deep intuition surfacing again because it never went away. It has been surfacing in every culture and every medium available since before anyone wrote anything down.

And then there is the Enigma.

In 2022, the largest faceted carbonado ever offered at auction went on the block at Sotheby’s. 555.55 carats. Black, porous, unmistakably ugly in the way all carbonado is ugly. It sold for $4.3 million.

What I keep coming back to is not the price. It’s the shape.

The owner had held the rough stone for more than twenty years before deciding what to do with it. He didn’t cut it for jewelry. He had it shaped into a hamsa — the ancient Middle Eastern palm symbol of protection, power, and strength, one of the oldest protective emblems in human culture. The number five is sacred to the hamsa. So the stone was cut to 555.55 carats. Fifty-five facets. Every number is a deliberate echo.

Stan Lee reached for the cosmic origin story without knowing carbonado existed. The owner of the Enigma knew exactly what he had — a possibly extraterrestrial black diamond — and responded by giving it the form of an ancient sacred object.

Three thousand years of human instinct, meeting in a single stone.

This is what I think about when someone hands me a rock in a velvet bag.

They are not confused. They are not scientifically illiterate. They are participating in something ancient and — it turns out — not entirely wrong. A tradition of recognizing that some materials feel categorically different, and reaching for the most powerful explanatory framework available. For most of human history, that framework was divine. The gods sent it. The sky gave it. It means something.

The framework has changed. The instinct underneath it has not moved.

What carbonado suggests — tentatively, incompletely, with all the uncertainty the literature requires — is that sometimes the instinct was tracking something real. The oldest hypothesis our species ever generated about unusual materials turns out, in at least one case, to be pointing in the right direction. Something arrived from somewhere else. We can’t fully explain it. We’re still working on it.

The story beat the instruments to the answer by three thousand years.

What I’m Actually Giving Them

I want to go back to the lab. The other one — not the carbonado, not the CT scanner. The one where I walk a stranger in when they hand me something in a velvet bag, and I already know what I’m about to find.

I bring them in because the answer is not the point. I have known the answer since they walked through the door. What I don’t know — what I can’t know until I ask — is what brought them here. What they were hoping for. What the rock means to them and why they chose to carry it to a museum rather than keep the story they already had.

So I ask. And then I run the analysis. I show them exactly what I’m doing and why — the XRF, the Raman, the data coming back in real time — and I tell them what it says and how I know it says that and where the edges of my confidence actually are. Because what I am handing them is not a result.

It is a method. Twenty years of training distilled into instruments that can identify a mineral from how it scatters light, and I’m handing that across a lab bench to someone who walked in with a feeling they couldn’t name. I don’t know a better description of what a museum is actually for — not the cases, not the labels, this. The moment the vocabulary finally shows up for a question someone has been carrying around for years without knowing what to call it.

The person who carries a crystal in their pocket because it feels significant is not wrong. The feeling is real. It is the oldest human response to materials that seem to come from somewhere beyond the ordinary — harder than expected, stranger than expected, arriving without explanation. Every culture has had people who carried those objects and told those stories, and the stories survived because the feeling they described was accurate.

Some things really are different. Some things really did come from somewhere else.

The instruments just took a while to confirm it.

What I hope the person across the lab bench takes with them — even when the quartz is still quartz, and the slag is still slag — is the understanding that the question they came in with was the right question. Science is not a replacement for the story they were telling. It’s the extension of it. The next chapter of something that started the moment the first human picked up a piece of iron that fell from the sky and refused to believe it was ordinary.

They were right not to believe it.

We are still, thousands of years later, working out exactly how right.

The gift from the gods turns out to be this: the permission to keep asking. The tools to ask more precisely. The honesty to say what we don’t know yet, and enough history to understand that not knowing yet is not the same as being wrong.

Somewhere in the Central African Republic, and in the state of Bahia in Brazil, there is a black porous mineral sitting in river sediment that has been waiting, possibly for billions of years, for someone to pick it up and ask the right question.

The person with the velvet bag was already asking it.

They just needed someone to take them seriously enough to show them what it looks like to ask like a scientist.

Aaron Celestian, PhD Curator of Mineral Sciences Natural History Museum of Los Angeles County.

Press:

BBC - LUXUS Magazine - Reuters - Daily Mail - WhyNow - Taipei Times - TweakTown - KOHA - LiveMint - AisaOne

What follows is a professional assessment. I am going to treat a fictional crystal the same way I would treat an unknown specimen handed to me in the collection — systematically, seriously, without apology. I am going to read forty-eight years of canonical physical description as if it were field notes from a geologist I trust but cannot reach, and I am going to tell you what the mineral is.

This is not satire.

Episode I — A Confession at the Checkout Counter

Go to a gas station tonight. Buy a roll of wintergreen mints — the white ones, not the green. Go somewhere dark, let your eyes adjust for two minutes, and bite down.

You will see a flash of blue-green light inside your mouth.

This is not a trick and it is not your imagination. It is triboluminescence — light produced by mechanical stress on a crystalline material. When the sugar crystal structure in the mint fractures under your teeth, it separates electrical charges across the newly formed surfaces. Those charges recombine, exciting nitrogen molecules in the air, which emit ultraviolet light. The methyl salicylate in wintergreen flavoring absorbs that ultraviolet and re-emits it as visible blue-green fluorescence — a color your eye can actually see.

The crystal responds to being touched by emitting light.

I study minerals for a living. I work in a building that contains more than 150,000 geological specimens, including some of the finest crystal collections in the Western United States. I have access to Raman spectrometers, X-ray diffractometers, scanning electron microscopes, and over twenty-five years of accumulated mineralogical literature. I can identify most minerals by habit, luster, cleavage, and associated assemblage within about thirty seconds of picking them up.

I am telling you about a gas station candy because it is the most efficient demonstration I know of a physical property that sits at the center of one of the oldest stories humans tell about crystals.

The story goes like this: there is a crystal that responds to you. It knows when you are holding it. It recognizes something in you — worthiness, power, alignment with some force larger than either of you — and it responds. The response is light, or warmth, or the granting of extraordinary ability. The crystal that chooses its owner.

You know this story. You have seen it in a movie theater. You may have seen it in twelve of them.

But you have been hearing it for three thousand years. Fans have debated it. Physicists have taken runs at the lightsaber mechanics. There are YouTube channels dedicated to the engineering of a plasma blade. And yet the professional crystallographic question — the one my discipline exists to answer — has been sitting quietly in the background, waiting.

What mineral is it?

Episode II — 291 Games and a Question the Toolkit Can Answer

I have spent the better part of three years building a database of mineral systems in video games — 291 games spanning four decades, cataloguing every gem, ore, crystal, and mineral analog that appears in their crafting systems, economies, and lore. The goal was to understand what the entertainment industry has done with the same raw material that my discipline studies, and why games have succeeded at making minerals culturally compelling in ways that formal science education largely has not.

The database contains games from 1986 to 2024. It covers JRPGs, survival games, grand strategy, action RPGs, farming simulators, space exploration titles. It contains the entire Final Fantasy series, every Monster Hunter game, the Atelier franchise, Stardew Valley, Minecraft analogs, Dune adaptations. Every game in which a mineral of any kind appears as a meaningful object.

Star Wars is almost entirely absent.

Not underrepresented. Absent. Across 291 games with mineral systems, the franchise that contains one of the most famous fictional crystal in the history of popular culture appears in exactly one accidental false positive — a Zelda title that matched on the word “Force.” The actual Star Wars games — Jedi: Fallen Order, Jedi: Survivor, Knights of the Old Republic, The Old Republic MMO, Battlefront, LEGO Star Wars across its many iterations — are not in the database because they do not treat kyber crystals as minerals.

This requires a moment of examination.

In the database, minerals serve six functional roles: crafting and progression (54% of games), economic trade (36%), power augmentation (37%), cosmological significance (17%), collection and trophy (17%), and consumable currency (13%). When I cluster games by their mineral mechanics, the clearest grouping I find is what I call the Crystal Cosmology cluster — games where minerals are sacred objects, withdrawn from ordinary use, constitutive of the world’s structure. Final Fantasy’s elemental crystals. Dragon Quest’s Stones of Sunlight. The cosmological minerals of early JRPGs. These crystals are not crafted or traded. They are found, kept, and named.

Kyber crystals belong in this cluster. They have 100% fictional cosmology origin, chest/reward acquisition, and essentially no functional mineral role. They are not mined, traded, or combined. They are sacred objects — found, kept, named, and not touched again.

Now compare to Materia — Final Fantasy VII’s crystallized planetary energy, the closest structural analog to kyber in the entire database. Same premise: a crystallized form of the planet’s life-force, attuned to its user, capable of granting extraordinary ability. Materia has crafting roles, economic roles, collection roles, power augmentation roles. Materia gets mined, purchased, combined, leveled up, and traded on the black market. Materia is the most mechanically sophisticated mineral system in the entire database.

Kyber crystals and Materia start from identical premises and diverge completely in execution. The reason is not game design. It is that kyber crystals are treated, in every Star Wars game ever made, as what they are in the mythology: relics. Things you find and keep, not things you use.

This is accurate to the lore. It is also why the crystallographic toolkit has never been formally applied to them — sacred objects don’t invite the kind of analysis you’d run on a hand specimen. They invite reverence, not a Raman spectrometer.

Science fiction has a long tradition of invented crystals that turn out to gesture, however imprecisely, at real mineralogy. Dilithium — Star Trek’s warp drive crystal, introduced in 1966 — predates kyber by decades and established the cultural template for a mineral that mediates energy at civilizational scale. Real dilithium exists as a lithium salt; it has nothing to do with warp drives, but the archetype of the crystal that powers a civilization is the same one Lucas reached for. Kryptonite got its real mineral — jadarite, found in a Serbian mine in 2007, matching the formula on a Superman Returns prop label close enough to generate genuine scientific excitement. Dilithium got its cultural ubiquity.

Kyber got neither — despite having the most detailed physical description of any fictional crystal ever written.

In 2007, a mineralogist at London’s Natural History Museum was characterizing a newly discovered mineral from a mine in Serbia’s Jadar Valley. The mineral didn’t match anything in the literature. He searched its chemical formula — sodium lithium boron silicate hydroxide — and found it already described in existing literature. On a prop label in the film Superman Returns. The formula the screenwriters invented for the rock containing kryptonite was, within one element (fluorine), the formula of a real mineral that had been sitting in Serbian rock formations for millions of years before anyone found it. That mineral is now named jadarite.

That was accidental. A coincidence between a prop department’s invented formula and a genuine geological discovery.

What follows is deliberate. I am going to bring the mineralogical toolkit to kyber: read the canonical physical description as a field report, apply the standard analytical constraints, and tell you what the mineral is.

Episode III — The Draft That Didn’t Make It

They were not in Star Wars originally.

George Lucas wrote the first rough draft of what would become A New Hope in May 1974. The script went through at least four substantial revisions before the film shot in 1976. In the second draft — titled Adventures of the Starkiller, Episode I: The Star Wars — a crystal appears at the center of the plot. Lucas spelled it “Kiber.” It is a Force-amplifying MacGuffin: a small, diamond-like object that can enhance a Force user’s abilities on either side of the spectrum, granting immeasurable strength. The protagonist Luke Starkiller — not Skywalker, Starkiller, the name that would survive into production until Charles Manson’s shadow made it unusable — must deliver the Kiber Crystal to his father, an aging Jedi Master fighting on the side of the Rebellion. When Luke places it in his father’s hands, “years seem to drop from the old man as the Kiber’s force moves into his body.”

This is the crystal that heals. The crystal that recognizes the worthy. The crystal that amplifies rather than creates, that serves as a conduit rather than a source.

It is not a lightsaber component. The lightsaber — Lucas originally called it a “lazersword,” the word “lightsaber” was reportedly coined by Alan Dean Foster, who helped with the script — has no crystal in any of the early drafts. The two ideas are separate. One became the most iconic weapon in cinema. The other was cut.

Darth Vader in the second draft is not Luke’s father. He is a separate villain — Dark Lord of the Sith, partial to black robes and a grotesque breath mask, dangerous and nearly unstoppable. The revelation that Vader is Luke’s father does not exist yet. It will not be invented until The Empire Strikes Back is being written, years after the Kiber Crystal has already been removed from the story. The character relationships that define the franchise as we know it and the crystal that could have defined its mythology were never in the same script simultaneously.

By the third draft, the story has transformed almost beyond recognition. The Starkiller family is gone. Obi-Wan Kenobi appears for the first time. The Kiber Crystal is nowhere. The plot that would become A New Hope is taking shape, and there is no place in it for a healing artifact that amplifies Force connection. That’s not the story Lucas is telling anymore.

Before the film’s extraordinary success made an ambitious sequel possible, Lucas had been developing a contingency: a low-budget follow-up that could be filmed cheaply if A New Hope underperformed. The story centered on Luke and Leia searching for a powerful crystal. He gave it to Alan Dean Foster to develop as a novel, and when the box office made a cheap sequel unnecessary, Foster published it anyway: Splinter of the Mind’s Eye, 1978. The Kaiburr Crystal — Foster’s version — heals the sick, amplifies Force power tenfold, loses potency the further it travels from its native temple. It is one of the most interesting fictional crystals ever described, and almost nobody has read the book.

The mainstream Star Wars universe moved on without it. The crystal concept drifted through the Expanded Universe in various forms, mostly as an optional lightsaber component among many possible alternatives, never central. Then, in 2009, thirty-five years after Lucas first wrote the word “Kiber,” a character named Master Bolla Ropal is mentioned as guarding a “Kyber crystal” — not a lightsaber crystal, not a Force amplifier, but a data crystal containing a list of Force-sensitive children in the galaxy. The name is back. The concept has changed entirely.

It changed again in 2012. Season five of The Clone Wars animated series, episode seven: “A Test of Strength.” A group of Jedi younglings travel to the Crystal Cave of the planet Ilum to find their lightsaber crystals. For the first time in either Legends or canon, kyber crystals are explicitly identified as the component that powers a lightsaber. The youngling arc establishes The Gathering — the rite of passage in which each youngling is guided by the Force to the specific crystal that is meant for them. The crystal is colorless. When the youngling bonds with it, it takes on a color. The bond is described as profound, personal, permanent.

This is the modern kyber. It arrives 35 years after the original film. It is canonized in Rogue One in 2016, where it finally appears on the theatrical screen, named aloud, treated as the material the Empire is strip-mining Jedha to power the Death Star. The most famous crystal in popular culture receives its name in the 39th year of the franchise.

Lucas didn’t invent this. He invented something older.

The crystal that responds to its owner — that amplifies the worthy, that cannot be used by the unworthy, that knows something about you that you may not know about yourself — this is not a Star Wars invention. It is one of the oldest ideas in human storytelling, and it is always, in every culture that tells it, mineral.

The Urim and Thummim were worn over the heart of the Israelite high priest, housed in a pouch beneath the ceremonial breastplate — the Hoshen, set with twelve gemstones representing the twelve tribes. The Hebrew words translate roughly as “lights and perfections,” and the etymology is contested enough that serious scholars still argue about it: some read “urim” as deriving from “or” (light), others as deriving from “arar” (curse), making the pair mean something closer to “curse and wholeness” or “condemnation and innocence” — a binary oracle, yes or no, dark or light. The physical objects themselves are never described in the text. The Bible simply says they exist and that the high priest wears them. What they looked like, what they were made of, how they worked — Scripture does not say.

What we have instead is a tradition of interpretation stretching from the Talmudic rabbis through medieval Jewish scholarship, and in that tradition the objects are consistently imagined as stones — specific, mineral, responsive. The Talmud describes certain letters on the breastplate gems lighting up in sequence to spell out divine messages, the Urim and Thummim acting as the illuminating force that caused the letters to shine. Josephus, writing in the first century, described the oracle as operating through the brilliance of the stones. One persistent tradition holds that the stones glowed brightly to indicate affirmation and dimmed or darkened for negation. In every version of the tradition, the mineral responds. It does not speak; it shines.

The physical description that has come down through the Mormon tradition is unexpectedly specific: Joseph Smith’s mother, Lucy Mack Smith, described what her son received as being like “two smooth three-cornered diamonds.” Two transparent, faceted crystals, worn over the heart, used to translate hidden text. Whatever one makes of that history, the description is revealing. When nineteen-century Americans tried to imagine what the ancient oracle stones looked like, they reached for the clearest, most geometrically perfect crystals they knew. Diamonds. The obvious choice.

The Urim and Thummim appear in Exodus, Numbers, Deuteronomy, Samuel, and Ezra — spanning centuries of composition and redaction. They fade from the biblical record after the Babylonian exile, in the fifth century BCE, replaced apparently by prophecy. The last mention in Ezra is poignant: certain priests whose lineage cannot be verified are told to wait until a priest with Urim and Thummim can certify them. The oracle is absent. It is expected to return. It does not, in any text we have.

Three thousand years later, the crystal that glows when the right person holds it is at the center of the most successful entertainment franchise in history. The oracle never came back. The story did.

The sword cannot be drawn from the stone by the unworthy. The Palantíri of Tolkien are attuned to their users, and the most powerful can corrupt through that attunement — Sauron’s ring amplifies the holder, at a cost. The Echo Stone in Star Wars: The High Republic — Lucas’s original kiber concept finally canonized — brings prosperity and then slowly corrupts the leader who holds it.

Every culture, independently, reaching for the same object when the story requires recognition. And the object is always mineral.

This is not coincidence. It is observation. People have been handling crystals for as long as there have been people, and crystals do respond to being touched. They emit light when mechanically stressed. They generate electrical charge when deformed. They change color permanently when subjected to radiation or energetic stress. They encode their entire formation history in their internal structure, visible to anyone with the right instruments. The archetype is three thousand years old because the phenomenon is real. The mythology kept the observation alive until the science caught up with the mechanism.

Lucas inherited the archetype without knowing its geological basis. His writers, rediscovering it forty years later, described it in extraordinary physical detail without knowing what they were describing.

I know what they were describing.

Episode IV — Three Planets Walk Into a Geology Lab

The canon gives us three primary kyber-bearing worlds, and their geological settings are completely incompatible with each other. Not superficially inconsistent — structurally, fundamentally, mutually exclusive in a way that would be embarrassing in a planetary science paper.

Let me present them as field notes.

Ilum. A snow-covered planet orbiting a blue dwarf star, fifth from its sun, located near the system’s cometary cloud. The planet has a kyber crystalline core. Beneath the surface ice, geothermal geysers warm bodies of water, their size and duration variable due to volcanic episodes and mineral deposition. The Crystal Cave formations developed over billions of years from hydrothermal fluids. Kyber crystals are brought to the surface by tectonic movements along fault lines, having formed under high temperatures and pressures. They are found growing freely in lava tubes, caves, and fault zones within veins of kyberite host rock. The Empire, after Order 66, mines Ilum to exhaustion to power the Death Star, eventually hollowing out the planet’s interior to construct Starkiller Base.

Jedha. A small desert moon, 11,263 kilometers in diameter, orbiting the planet NaJedha. Permanent winter climate. Landscape dominated by jagged rock formations, broad mesas, narrow spires, and rocky dustbowls. Kyber crystals found within the moon’s sandstone crust. Rich deposits, sufficient to supply the Death Star project once the Empire begins full extraction. Among the oldest settled worlds in the galaxy, with some of the earliest known architecture.

Christophsis. A crystalline planet in the Outer Rim’s Savareen sector. Crystal forests cover the entire surface with tall, jagged peaks of turquoise rising from fields of hexagonal tiles. Breathable atmosphere. Kyber crystals found growing across the surface, holy to the native Christophsian people, stolen by the Empire for the Death Star.

Now. A mineralogist reads these three descriptions and immediately identifies the problem.

Ilum says kyber forms deep in the planetary interior under extreme pressure and temperature, then is delivered to the surface by volcanic systems — specifically by hydrothermal activity in a tectonic environment associated with fault zones and lava tubes. This is a coherent high-pressure, high-temperature metamorphic or deep hydrothermal formation story. It is the description of how garnet, tourmaline, and beryl actually form — in the deep crust or upper mantle, exhumed by tectonic uplift, exposed at the surface by erosion and volcanism.

Jedha says kyber occurs in sandstone. Sandstone is sedimentary rock. It forms at or near the surface, at low temperature and low pressure, from the compaction of transported sediment. Nothing that forms at the temperatures and pressures described for kyber on Ilum also occurs in sandstone. These are incompatible formation environments. A mineralogist does not find deep mantle metamorphic minerals in sandstone. They find sand-grain-sized quartz, feldspar fragments, and whatever happened to be nearby when the sediment was deposited.

Christophsis says kyber grows as planet-spanning surface crystal forests in hexagonal tile arrays. Surface crystallization. The hexagonal tile geometry is a cooling fracture pattern — the same geometry as columnar basalt and salt flat crystallization, produced when a crystallizing material contracts uniformly across a large surface. This is a surface evaporation or rapid cooling process, not a deep metamorphic one.

Three planets. Three completely different formation environments. One mineral.

In a lesser post, this is where I tell you the franchise is geologically inconsistent and leave it there.

Instead, I am going to do what a field geologist actually does when handed samples from three separate sites that supposedly contain the same mineral: I am going to find the formation story that makes all three true simultaneously.

The key is diamonds.

Diamond is the hardest natural material on Earth. It forms in one place — the deep mantle, under extreme pressure and temperature, at depths between 150 and 450 kilometers below the surface. Yet we find diamonds in completely different geological settings: in kimberlite pipes cutting through ancient cratons, in alluvial river gravels thousands of kilometers from the nearest kimberlite, and — critically — in the shocked rocks around meteorite impact craters. Same mineral. Radically different surface contexts. The surface geology tells you about the delivery mechanism, not the formation environment. The diamond doesn’t care how it reached the surface. It formed the same way regardless.

Kyber works exactly like this. The three formation environments the writers independently described across decades of worldbuilding are not contradictions. They are three different delivery stories for a single deep-origin mineral.

Ilum is a kimberlite system. Kimberlite pipes are formed by deep-source volcanic eruptions, originating at depths three times greater than ordinary volcanoes, carrying deep-mantle material to the surface in violent, gas-rich explosions that move faster than ordinary magma ascent. Diamonds arrive at the surface in hours via kimberlite, not millennia. Ilum’s ice caves are the eroded surface expression of ancient kimberlite-analog pipes. The hydrothermal geysers are the thermal afterglow of that system, still active after billions of years. The surface is frozen because the blue dwarf parent star provides minimal warmth, but the interior is hot enough to sustain the volcanic system that originally delivered the kyber. The Crystal Cave is not where kyber formed. It is where the eruption stopped.

Jedha is a placer deposit. In diamond mining, a significant fraction of all economically important diamonds are found not in their primary kimberlite pipes but in secondary sedimentary deposits downstream — alluvial gravels and placers where diamonds, extraordinarily stable and resistant to weathering, survived the erosion of everything around them and concentrated in sedimentary layers over geological time. Jedha is ancient. Its landscape of eroded mesas and dusty spires is what a billion years of erosion looks like. Its delivery system — whatever volcanic mechanism originally brought kyber to the surface — has been gone for geological ages. Only the crystals remain, concentrated in the sandstone the way diamonds accumulate in river gravel: denser, harder, and more chemically stable than everything that once surrounded them. The sandstone didn’t form with the kyber. The kyber outlasted everything that came before the sandstone.

Christophsis was hit. This is the scenario nobody in the Star Wars universe has apparently considered. Impact metamorphism is the geological process by which a hypervelocity collision delivers mantle-equivalent pressures to surface rocks instantaneously. When a large enough impactor strikes a planet, the shock wave compresses surface material to pressures that normally require hundreds of kilometers of burial — forming high-pressure mineral phases in rocks that have never been anywhere near the deep crust. Stishovite, a high-pressure polymorph of quartz, has been found in the sandstone around Barringer Meteor Crater in Arizona, formed by impact pressure in surface sedimentary rock. High-pressure minerals in sandstone, produced not by burial but by collision.

A sufficiently large impactor striking a kyber-rich planet could do something extraordinary: the impact pressure would shock-metamorphose the surface crust to mantle-equivalent pressures, nucleating kyber crystallization across the entire surface simultaneously, followed by rapid cooling that produces the hexagonal fracture geometry of the tile fields. The turquoise crystal spires are what grew in the shocked rock over millions of years of post-impact cooling. The hexagonal tiles are the contraction fracture pattern of impact melt, crystallized and now exposed.

Christophsis was not always a crystal planet. It was hit. The impact was so large that the surface became kyber-forming conditions across its entire extent. Christophsis is a geological crime scene, and the evidence has been growing beautifully out of the rock for millions of years, and nobody in the Star Wars galaxy has noticed.

One detail in the Ilum description deserves to stop us entirely.

Ilum has a kyber crystalline core.

Not deposits. Not veins. A core — the entire planetary interior composed of kyber crystal. Under this formation model, this is not strange. It is a planet where the deep-mantle kyber-forming conditions were so pervasive and so long-lasting, sustained across billions of years of planetary evolution, that the entire interior crystallized. The individual crystals in the caves are not deposits. They are fragments shed upward from the planetary interior through the kimberlite system — chips of the planet’s own body, worked loose by volcanic activity and hydrothermal circulation and offered up to the surface.

Every kyber crystal a Jedi ever carried was a piece of a planet. A piece of Ilum itself, shed upward over billions of years, waiting.

The Empire came, identified the source, and mined it to exhaustion. Then the First Order used what remained as the structural shell of Starkiller Base — a weapon that destroys star systems by draining the energy of its parent star. A planet whose entire interior was a crystal, born from stellar material under conditions that stellar bodies create, converted into a machine that consumes stars.

The geology completes the mythology without trying to.

Episode V — I Know This Rock

The canonical physical description of kyber crystals, assembled from across forty-eight years of films, animated series, novels, comics, and reference guides, reads as follows. I am treating this as field notes.

The crystal is colorless prior to bonding with a Force-sensitive individual. Upon bonding, it changes to a specific color that it retains permanently. It grows in an organized fashion, adding to its prismatic structure one piece at a time. It is composed of both organic and inorganic matter. It is stable at temperatures and pressures found in the cores of large stars. It is noted for intricate internal patterns. Its distinctive luster is called “the water of the kyber” by those who study it. It does not warm a sheath, a towel, or any inanimate object — it responds only to life, including plant life. When held by a living being, it warms without changing temperature. It has no discernible lifespan. True kybers are found only in veins of kyberite, the host matrix from which they grow; kyberite is the disordered, Force-inactive form of the same basic material. Crystals that have been “bled” — subjected to forced energetic stress by a dark-side user — turn permanently red, a condition that can be reversed by a sufficiently attuned light-side practitioner who “purifies” the crystal, turning it white.

Let me work through these constraints professionally.

Before I do, I should acknowledge where the fandom has already pointed. The most substantive real-mineral analysis that circulates in Star Wars communities identifies tourmaline as the closest real-world equivalent — and the reasoning is solid: tourmaline is piezoelectric, grows as prismatic crystals with organized layer-by-layer structure, occurs in nearly every lightsaber color across its many species, and forms in deep metamorphic and pegmatite environments consistent with kyber’s high-pressure origin. Tourmaline is the right answer if you read the physical description with conventional mineralogical tools.

It fails on one constraint, and that constraint changes everything.

Organic and inorganic composition. This single property eliminates tourmaline, quartz, beryl, and virtually every other conventional mineral that has been proposed, and points immediately to one material class: organic-inorganic hybrid crystals — biomineral composites. The most extensively studied example is bone. Bone is hydroxyapatite — calcium phosphate, formula Ca₁₀(PO₄)₆(OH)₂ — templated by collagen, producing a composite with mechanical and electrical properties that neither component possesses alone. The mineral in your skeleton is an organic-inorganic hybrid. It is piezoelectric. Its piezoelectric properties arise from the ordering of hydroxyl ions along the crystallographic c-axis — a structural feature that depends on the presence of the organic collagen framework during biomineralization. Remove the organic component and the crystal disorders. The organic component is not an impurity in hydroxyapatite. In bone, the collagen template organizes hydroxyapatite crystallites into a preferentially aligned composite — and it is that organizational relationship between organic and inorganic phases that produces the net polar, piezoelectric behavior of the tissue. The individual mineral and the individual protein are each insufficient alone. The response emerges from their arrangement.

Kyber is a planetary-scale biomineral. The organic molecular component — whatever Ω is, the Force-active channel occupant I will describe shortly — is not contamination. It is the structural template that makes the crystal respond to life. Its presence in the crystal is a record of the planetary bioelectric conditions during crystal growth.

Prismatic habit, organized layer-by-layer growth, intricate internal patterns. Prismatic habit with incremental, organized growth along a primary axis is the hallmark of the trigonal and hexagonal crystal systems. Tourmaline, quartz, beryl, apatite — all grow as prisms along the c-axis, adding layer by layer, incorporating environmental changes as compositional zones visible in cross-section. The “intricate internal patterns” are zoning — the organic and inorganic components vary in concentration as formation conditions change, producing optical banding that records the crystal’s growth history. The interior of a kyber crystal is a stratigraphic section. Every Jedi who held one was holding a geological archive.

Responds to life, not inanimate objects. Living organisms generate bioelectric fields — measurable electrical gradients produced by ion transport across cell membranes. Every living thing produces one. Inanimate objects do not. A crystal that responds to life but not to a sheath or a towel is a crystal with a response threshold calibrated to bioelectric field magnitude. This requires the crystal to be both piezoelectric — sensitive to applied fields through the converse piezoelectric effect — and non-centrosymmetric, since only non-centrosymmetric crystals can be piezoelectric. A sufficiently sensitive non-centrosymmetric crystal, with spontaneous polarization aligned along its growth axis, would respond measurably to the bioelectric field of a living hand and produce no measurable response to a leather sheath. The canon described a bioelectric field detector. The physics is consistent.

Warms without changing temperature. This sounds impossible and is actually a precise description of low thermal conductivity. High thermal conductivity crystals — diamond, most notably — feel cold to the touch because they draw heat rapidly from your hand. A crystal that feels warm without changing temperature is one that does not draw heat efficiently. It insulates rather than conducts, allowing your hand’s radiated warmth to accumulate at the surface. The subjective sensation of warmth is your own heat not being conducted away. This is a diagnostic physical property, not a supernatural one. It constrains the bonding character of the crystal: low thermal conductivity is associated with complex crystal structures with many atoms per unit cell, organic-inorganic hybrids, and structures with significant internal disorder. All consistent with the picture developing here.

Color change on bonding, permanent retention. This is color center chemistry, precisely described. Color centers are electron traps — structural defects in the crystal lattice that absorb specific wavelengths of visible light, producing permanent color without pigment. They form when a crystal is subjected to energetic stress: radiation, heat, or — in the case of kyber — bioelectric field exposure during bonding. The color is a scar in the lattice, a permanent structural rearrangement at the atomic scale. This is why the color persists after the Jedi dies. The crystal has been changed at the level of its electronic structure. It remembers.

The canon made this connection explicitly, though without the vocabulary: Galen Erso’s research notes describe incorrectly pumping laser energy into kyber crystals as weakening their lattice structure “in the same way living cells are affected by radiation.” That is a precise description of radiation-induced color center formation and lattice damage. The franchise arrived at the correct physical mechanism through a scientist character doing in-universe research. We are simply naming what he found.

The bleeding process — a Sith pouring rage and pain into a kyber crystal until it turns red — maps onto forced charge injection, a different class of color center formation. Clusters of multiple trapped electrons, driven together under intense energetic stress, selectively destroy the crystal's ability to transmit blue and green wavelengths. The red doesn't go in. The blue burns out. What remains visible is what the damage left behind. A bled kyber and a bonded kyber carry different color center signatures — a Raman spectrum would distinguish them — and the purification process maps onto controlled annealing: restoring the lattice to a lower-defect state, recovering the lost wavelengths, returning the crystal to white. The Sith isn't filling the crystal with darkness. The Sith is burning out the light.

The water of the kyber. “Water” in traditional gemological vocabulary refers to transparency and internal luminosity — the degree to which light penetrates the crystal, interacts with its interior, and re-emerges with optical complexity. High “water” gems are those with exceptional transparency and rich internal depth: diamond, fine aquamarine, exceptional tourmaline. The compositional zoning and channel structure of a prismatic organic-inorganic hybrid would produce exactly this optical character — light entering through a transparent exterior, scattering from compositional zone boundaries, and re-emerging with the internal complexity that gemologists call water. This is a diagnostic description of a specific optical behavior, not a poetic metaphor.

No discernible lifespan. A crystal with no lifespan is a crystal that is thermodynamically stable at surface conditions and kinetically inert — one that would require geological timescales to transform even if the thermodynamics eventually favored it. Diamond is metastable at surface pressure but requires billions of years to convert to graphite. The organic component in kyber — the structural template in the channel — is not a protein. Proteins denature. The Ω molecule is a simpler, more stable organic compound, a cyclic peptide or amino acid polymer, that acted as a crystallization template and is now effectively encased in the inorganic framework. Once the phosphate lattice set around it, the composite is as stable as the inorganic framework alone. The organic component is the mold. The crystal is what formed in the mold.

Kyberite as the opposing host. The relationship between kyber and kyberite is more precise than a simple ordered/disordered distinction. Canon describes them as two sides of a magnet — kyber grows in opposition to kyberite, the two materials in structural tension with each other. This maps directly onto a known phenomenon in materials physics: the ferroelectric/antiferroelectric relationship. In a ferroelectric crystal, the structural units — in our model, the Ω channel molecules — all align in the same direction, producing a net spontaneous polarization. This is kyber. In an antiferroelectric crystal, adjacent structural units point in opposite directions, canceling each other out, producing no net polarization. This is kyberite. Same compound, same chemistry, opposing structural geometry. The magnet analogy in the canon is not metaphor. It is an accurate description of opposing dipole ordering in the same material system. Kyberite is Force-inactive not because it is impure or disordered but because its dipoles cancel. It is the antiparallel configuration of a crystal that, in its parallel configuration, is alive.

The proposed crystal structure.

Crystal system: Trigonal.

Space group: P3₁ or P3₂.

These are chiral space groups — they contain no mirror planes, no inversion center, and no improper rotation axes. They are the most strongly non-centrosymmetric space groups in the trigonal system, producing the maximum possible piezoelectric response. And they have a property that no other space group assignment can explain as elegantly: P3₁ and P3₂ are enantiomorphic partners — mirror-image crystal forms that are physically identical in every measurable property but optically opposite, like left and right hands that cannot be superimposed. They are the same compound in opposite chiral forms.

This is the crystallographic basis for light-side and dark-side kyber.

Not different minerals. Not different compositions. Mirror-image polymorphs of the same structure, distinguished only by the handedness of their channel geometry. A Jedi’s bonded blue crystal and a Sith’s bled red crystal are the same compound. The chirality of the channel determines which way the optical rotation goes, which class of color center forms under bioelectric stress, and — in the model I am proposing — whether the crystal’s spontaneous polarization responds constructively or destructively to the specific electromagnetic signature of Force-sensitive biological activity. Spontaneous polarization is the crystal’s built-in electrical directionality — a permanent microscopic compass needle aligned along the growth axis, present without any applied field, a consequence of the chiral channel geometry alone.

The franchise already named this. Galen Erso, the scientist the Empire conscripted to weaponize kyber for the Death Star, described what he called the “day and night” lattices of the Force — two opposing structural orientations within the crystal that the Jedi had long recognized but never characterized in physical terms. Erso theorized that by forcing the lattice to realign along the dark or nighttime axis, the crystal’s energy yield could be dramatically amplified and its diffraction controlled. He was describing, in pre-crystallographic vocabulary, the reorientation of spontaneous polarization between enantiomorphic forms. The Jedi intuited the duality for thousands of years. Erso was the first in-universe scientist to propose a structural mechanism. This post is the second.

Light side and dark side, in this reading, are not metaphysical categories. They are enantiomeric crystal chemistry.

The proposed formula.

Ca₁₀(PO₄)₆(Ω)₂

Where the inorganic framework is hydroxyapatite — calcium phosphate, the mineral in your bones — and Ω is the Force-active channel occupant: a chiral organic molecule with no Earth analog, responsible for the crystal’s spontaneous polarization, its bioelectric field sensitivity, and its chirality-dependent response to the Force.

Every atom in the formula except Ω exists in your body right now.

The formula is almost certainly wrong in its specifics. The actual Ω molecule, if kyber were real, would require its own discovery and characterization. But the framework is constrained: it must be chiral, it must be small enough to occupy the channel without distorting the phosphate backbone excessively, it must be stable under geological conditions, and it must carry a permanent dipole moment that produces spontaneous polarization along the c-axis.

A cyclic dipeptide fits these constraints. So does a small aromatic amino acid. The exact chemistry is an open question.

What is not an open question is the structural logic. Kyber is a chiral, non-centrosymmetric, organic-inorganic hybrid phosphate in the trigonal system, with a polarized channel structure running along the c-axis, producing piezoelectric response, triboluminescent behavior, color center formation under bioelectric stress, and chirality-dependent Force coupling. It is a geological biomineral, formed under the influence of a planetary bioelectric field strong enough to order the channel occupant during crystal growth.

It shares hydroxyapatite's compositional logic and biological formation mechanism — but where bone mineral is centrosymmetric, kyber is not. The chirality is the difference between a crystal that exists and a crystal that responds.

It is a distant cousin of your skeleton — same family, different symmetry.

Episode VI — Why the Planet Had to Be Alive

Kyber is not rare because calcium and phosphate are rare. They are among the most abundant materials in the galaxy. Phosphate minerals occur on every rocky planet that has ever been surveyed. Calcium is the fifth most abundant element in Earth’s crust. If kyber were purely a matter of chemistry, it would be everywhere.

Kyber is rare because the planetary bioelectric field required to order the Ω channel occupant during crystal growth is rare.

Crystal growth is sensitive to its environment. The conditions present during crystallization are permanently encoded in the crystal’s structure — in its zoning, its defects, its isotopic ratios, its color center distribution. A crystal grown in the presence of a strong, coherent bioelectric field will incorporate that field’s influence into its channel structure. A crystal grown in the absence of such a field will not. The Ω molecule will crystallize, but it will disorder. The channel will form, but it will be centrosymmetric. The crystal will be kyberite — geologically interesting, structurally similar, functionally inert.

A planet has to be alive enough — Force-rich enough, in the canon’s vocabulary — to template the channel during growth. The Force, in this model, is not magic. It is a measurable planetary bioelectric phenomenon: a coherent field generated by the aggregate biological activity of all living things on a planet, amplified by whatever geological or cosmological conditions the Star Wars universe calls Force-richness, and strong enough during crystal growth to lock the Ω occupant into the aligned, polar orientation that makes kyber what it is.

Dead planets cannot grow kyber. They can grow kyberite. They cannot grow the living crystal.

This reframes everything that happens to Ilum.

The Empire did not merely mine Ilum. Imperial extraction operations disrupted the planetary surface, destroyed the Jedi Temple, eliminated the biological communities that had coexisted with the kyber-growing system for millennia, and ultimately hollowed out the planet’s interior to construct Starkiller Base. Each step of this process degraded the planetary bioelectric field that had been generating kyber for billions of years. By the time the First Order finished with Ilum, the conditions for new kyber growth no longer existed.

Every kyber crystal that will ever exist already exists. The mine is closed. Not because the ore ran out — because the planet died.

The Jedi Order understood this intuitively. They did not mine Ilum. They sent younglings to find single crystals in the Cave, guided by attunement to the planet’s field, taking only what the planet offered. The Gathering was not a harvesting ritual. It was a listening ritual. The youngling went into the Cave not to take a crystal but to find the crystal the planet had grown for them — a crystal whose formation history, encoded in its zoning and its color centers and the chirality of its channel, aligned with the bioelectric signature of that specific Jedi.

The crystal chose the Jedi because the planet had been growing that crystal in anticipation of that specific bioelectric field for as long as the Jedi had been alive.

This is not metaphysics. This is the geological record of a relationship between a living planet and a living person, encoded in the crystal’s atomic structure at the moment of their first contact.

The Sith cannot form this relationship because the relationship requires attunement — a slow, coherent resonance between the Jedi’s bioelectric field and the crystal’s spontaneous polarization. The Sith approach — forced charge injection, bleeding — bypasses the resonance and imposes change through energetic violence. It works. The color centers form. The crystal functions. But the relationship is between the crystal and the Sith’s pain, not between the crystal and the Sith. The weapon is powered by a record of suffering. The Sith is carrying a fossilized scream.

The Mint, Revisited

A gas station. A roll of wintergreen mints. The blue flash when you bite down in the dark.

That flash is triboluminescence — charge separation in a non-centrosymmetric crystal at the moment of fracture, recombined to produce light. It happens because the sugar crystal has no inversion center. Its structure is asymmetric in a specific, directional way. And when that asymmetry is broken by mechanical stress, the crystal responds. It emits light.

Every mineral that emits triboluminescent light does so for the same reason. The structure is non-centrosymmetric. The breaking of that structure separates charge. Charge recombination produces photons.

Quartz does this. Fluorite does this more dramatically. Sphalerite does it brightly enough to see clearly across a dark room. And the Uncompahgre Ute, long before the mechanism was characterized, put quartz crystals in rawhide rattles specifically to produce the flash — ceremonial light from the mechanical stress of the rattle’s motion. They had observed the phenomenon. They had given it ceremonial weight. They had not yet named the mechanism.

Francis Bacon noted in 1620 that crushing sugar crystals produces light and called it “well known.” The phenomenon was not named triboluminescence until the twentieth century. The observation preceded the explanation by three hundred years.

The observation that crystals respond to being touched — that they emit light, that they generate charge, that they record what happens to them — preceded the science by three thousand years. The Urim and Thummim, the scrying stone, the Palantír that knows its master, the kiber crystal in Lucas’s 1974 draft, the kyber crystal in the Crystal Cave of Ilum: all of them are cultural expressions of an accurate physical observation about crystal behavior, made and remade across millennia because the observation kept being made.

The crystal knows you were there. It records the encounter in its structure.

This is not mythology. This is mineralogy.

The most powerful weapon in the galaxy runs on a crystal that is, structurally, a relative of your skeleton. It forms on planets where the aggregate bioelectric activity of all living things is strong enough to order a chiral channel molecule during crystal growth, locking in a record of the planet’s life that persists in the crystal’s structure for geological time. When a Jedi holds that crystal for the first time, their bioelectric field resonates with the spontaneous polarization of the channel, and the resulting charge redistribution creates a permanent color center — a scar in the lattice that is also a record of that specific encounter, that specific person, that specific moment of contact.

The crystal knows you because it was grown by a living planet, and it remembered the first time it touched you.

Lucas felt his way toward this in 1974 without knowing the mechanism. The writers who built the kyber mythology across forty-eight years of films, animated series, novels, and comics elaborated it without knowing the chemistry. They were describing a mineral that behaves as they described because real minerals actually behave this way — approximately, imperfectly, in ways that require a fluorescent candy to demonstrate rather than a lightsaber, but genuinely. The intuition was right. The archetype was accurate. The physics was there all along.

The franchise has been sitting on a geological story richer than anything it has used. The field notes are already written. The formation environments are already described. The physical properties are already in the canon. They were waiting for someone to read them the way a mineralogist reads field notes — not for plot, but for what the rock is actually telling you.

Ca₁₀(PO₄)₆(Ω)₂. Trigonal. P3₁ or P3₂. Non-centrosymmetric. Chiral. Piezoelectric. Triboluminescent. Force-sensitive.

Found on planets that are alive enough to grow it.

Handled with care.

Aaron Celestian is Curator of Mineral Sciences at the Natural History Museum of Los Angeles County, former research scientist at NASA’s Jet Propulsion Laboratory, and adjunct professor at USC. He writes Pocketful of Χtals because mineralogy is stranger and more alive than most people have been told.

The author has not been contacted by Lucasfilm. He is available, and literally right across the street from the Lucas Museum of Narrative Art.

That’s how these things usually start — not with a mystery, but with a question so routine it barely registers as one. The Borghese-Windsor Cabinet had recently been acquired by the Getty Museum in Los Angeles: a piece of Renaissance furniture built around 1620 for Camillo Borghese, later Pope Paul V, its surfaces encrusted with agate and lapis lazuli and jasper in the Italian technique called pietra dura. Spectacular object. Well-documented history. They wanted to know more about the stones.

I said yes, because of course I said yes.

What nobody told me — what nobody knew — was that the cabinet had been waiting four hundred years for someone to ask the right question.

The Cabinet

Stand in front of the Borghese-Windsor Cabinet and your first instinct is to look for a door. Not because it has one — it does, a tiny arched niche at the center of the facade — but because the thing in front of you is not furniture in any sense you recognize. It is a building. Three stories of ebony and gilt bronze columns, fluted and capped with Ionic capitals, rising in tiers to a pediment where a gilded Roman emperor stands watch over everything below. Statuary occupies the cornices. A crowned figure commands the apex.

And everywhere, inlaid into every surface between the columns and beneath the cornices and framing the arches: stone. Hundreds of individually cut and fitted panels of agate, lapis lazuli, and jasper, set in the Italian technique called pietra dura — hard stone inlay, each piece shaped to fit its cavity with a precision that, four centuries later, still holds. Oranges and blues and smoke-grays and translucent whites. Oval panels displaying the banded rings of agate like tiny windows into geology. Rectangular fields of deep lapis. Jasper the color of dried blood.

The stones aren’t decoration in the way wallpaper is decoration. They are the object. The ebony frame exists to organize them. The gilt figures exist to celebrate them. The whole elaborate architectural fantasy — modeled on a Roman palace facade, built at a scale that fits inside a room — is a machine for displaying stone.

And not just any stone. The Getty’s conservation scientists have traced the provenance of nearly everything on this cabinet. The alabaster almost certainly came from Egypt. The travertine from Italy — local, ubiquitous, Roman. The lapis lazuli from Afghanistan, which is where virtually all European Renaissance lapis came from, via overland trade routes that stretched from Badakhshan to Venice. Even the ebony frame has been sourced: Getty scientists used DNA analysis to identify the wood as a Central American species — timber harvested somewhere in the forests of what is now Honduras or Guatemala, loaded onto Spanish ships at the mandatory colonial port of Veracruz, assembled with the annual treasure fleet in Havana, and then carried east across the Atlantic to Seville before making its way overland to Rome. By the time someone was sawing it into columns for a papal cabinet, that wood had crossed an ocean and two continents.

This is worth pausing on. Italy did not colonize the Americas. But the cabinet sitting in Rome was built from wood that came from them — because the men who organized the global trade networks of the early 17th century were, in substantial part, Italian. Genoese bankers had been embedded in Seville since the 1490s, financing Spanish Atlantic ventures and holding colonial trade contracts from the very beginning of the colonial era. The man who named the continents, Amerigo Vespucci, was Florentine, working for the Medici, sailing first for Spain and then for Portugal, and serving as Spain’s chief navigator for the last four years of his life. The first monopoly contract for colonial trade issued by the Spanish crown — in 1517 — was immediately purchased by Genoese merchants in Andalusia. Columbus himself was from Genoa. Italy didn’t plant a flag. It provided the capital, the navigators, and the commercial infrastructure through which everyone else’s flags were planted. The ebony in this cabinet moved through that infrastructure.

The agate is the one material that has never had a confident answer. Everything else on this cabinet has an address. The agate alone remained a question mark.

This was built around 1620 for Camillo Borghese, who became Pope Paul V — and it is named for him. Borghese is straightforward. Windsor requires more explanation.

The cabinet passed through several hands across four centuries, each chapter leaving its mark on the name. From Pope Paul V it moved, by routes not entirely documented, until 1820, when it entered the collection of King George IV of England, who kept it at Windsor Castle. That residency — a papal treasure sitting in an English royal palace — is where the second half of the name comes from. In 1959 it was acquired by Robert de Balkany, an international collector, before the Getty Center purchased it in 2016 and brought it to Los Angeles, where it now sits in a climate-controlled gallery on a hill above the Pacific Coast Highway, waiting for visitors to walk past it and read the label.

Four owners across four centuries. Pope to king to collector to museum. The stones traveled with it through all of it, carrying their geological biography in silence.

There is a companion piece, the Sixtus Cabinet, built around 1585 for Pope Sixtus V. Similar construction, similar stones, similar ambition — built a generation earlier for a different pope, and decorated with agate from the same era and presumably the same trade networks. The two objects exist in parallel, one analyzed, one not. That asymmetry is part of what makes the story unfinished.

That assumption about the agate’s origins is where the question began.

The Valley That Made the World’s Agate

For centuries, if you wanted agate in Europe, you went to one place: the Nahe River Valley, near the small German town of Idar-Oberstein.

The geology there is Permian — ancient volcanic flows, 280 million years old, riddled with cavities where silica-rich groundwater slowly crystallized into banded chalcedony. Agate. The Romans knew about it. Medieval lapidaries knew about it. By the late 1400s, the river itself had been put to work: water wheels powered massive sandstone grinding wheels, and craftsmen lay flat on boards, pushing agate against the spinning stone with the weight of their own bodies. Not sitting. Not standing. Lying prone on a wooden plank, face inches from the grinding wheel, holding the stone against it for hours. Generation after generation, the same families, the same river, the same stone. The Nahe Valley smelled like wet rock and hot dust for four hundred years.

By 1620, when the Borghese-Windsor Cabinet was being assembled in Rome, Idar-Oberstein was the source. This was not a question anyone thought to examine. It was simply known — the kind of knowledge that calcifies into assumption so completely that nobody thinks to test it.

But Idar-Oberstein’s identity is more complicated than “source.” It was also, crucially, a processing center — the place where rough stone became finished gem. Over centuries, those roles collapsed into each other in the historical record. “From Germany” came to mean both “dug from the Nahe Valley” and “cut in the workshops of Idar-Oberstein,” and nobody kept careful track of which was which. When Brazilian agate began arriving in the early 19th century — shipped as ballast on empty vessels returning from South America — it went to Idar-Oberstein to be cut. The label “German agate” started meaning two completely different things simultaneously, and the convention never caught up.

So when art historians said the agates in papal cabinets came from Germany, they were probably right. The question I wanted to answer was: which Germany?

The Problem With Looking

Before I tell you how we solved this, I need to explain why it was hard.

The obvious approach — look at the stones and tell me where they came from — doesn’t work. Not because mineralogists aren’t good at looking. It’s because agate is a profoundly uncooperative witness.

A single locality can produce agate in dozens of colors, patterns, and textures. Idar-Oberstein material runs from translucent gray to deep orange to banded white and rust. Brazilian agate from Rio Grande do Sul can look almost identical — or completely different — depending on which layer of which nodule you’re cutting. Indian agate, Uruguayan agate, agate from the Atlas Mountains: all of it, banded silica, all of it capable of presenting in overlapping ranges of color and pattern. An expert eye can make educated guesses. But educated guesses are not provenance, and on an object this significant, a guess is not good enough.

The deeper problem is bias. If you sort by eye, you sort by what you expect to see. I knew the historical assumption was Idar-Oberstein. I knew what German agate looks like. That knowledge would inevitably shape what I noticed and what I discounted — not through dishonesty, but through the ordinary mechanics of pattern recognition. I needed a method that didn’t know the hypothesis. Something measurable, reproducible, and agnostic to the question being asked. Something that would give the same answer whether I was looking or not.

That requirement led me to the clock.

A Clock Built Into the Stone

Here is the thing about agate that most people don’t know: it carries a clock inside it.

Agate is made primarily of quartz — silicon dioxide, crystallized. But it also contains a second phase called moganite, a metastable polymorph of silica that forms alongside quartz as the agate grows. Moganite is impatient. It doesn’t want to exist. Given enough time and the right conditions, it converts back to quartz, and once it’s gone, it’s gone. The older the agate, the less moganite survives.

This means you can read geological age directly from the crystal structure of the stone.

Agate from the Nahe Valley near Idar-Oberstein is Permian — roughly 280 million years old. By now, most of its moganite is long gone. Agate from Rio Grande do Sul in southern Brazil is Cretaceous — around 120 million years old. Young enough that moganite is still present in measurable quantities. The two localities produce agate that looks, to the untrained eye, quite similar. To a Raman spectrometer, they are completely distinct.

Look at that image for a moment. The agate is not uniform. Moganite doesn’t distribute evenly — it concentrates in the cryptocrystalline bands that form during specific growth stages, leaving the larger crystalline zones essentially clean. This heterogeneity is actually useful: it means you can isolate the moganite-bearing bands specifically and measure their ratios precisely, rather than averaging across the whole sample and losing the signal. The narrow-band approach we developed turned out to be more reproducible in distinguishing Brazilian from German agates than any whole-sample method, precisely because it works with the stone’s internal structure rather than against it.

The technique is called Raman spectroscopy. A laser illuminates the sample; the scattered light carries a fingerprint of the molecular bonds present. Quartz scatters differently than moganite. The ratio between them tells you, within limits, how old the stone is. And age, in this case, is a proxy for provenance.

Building the Reference Library

Here is the challenge: you cannot remove stones from the Borghese-Windsor Cabinet to analyze them in a laboratory. The object is too important, too intact, too irreplaceable — and that’s exactly as it should be.

The first attempt to work around this was to bring the instrument to the cabinet. We tried Raman spectroscopy directly on the stones in situ at the Getty — laser pointed at the surface, no contact, no sampling. It didn’t work. The geometry of the mounted stones, the surface polish, the ambient conditions in the gallery — the spectra came back unsuitable. Too much noise, not enough signal. Plan A failed.

So we built the library first and brought it to the cabinet, rather than the other way around.

Working at the Natural History Museum of Los Angeles County and in collaboration with the Smithsonian Institution’s National Museum of Natural History, we collected and analyzed dozens of agate specimens of known provenance — agates with documented localities, from museum collections built over more than a century. Brazilian agates from Rio Grande do Sul. German agates from the Nahe Valley.

We mapped them with three techniques: Raman spectroscopy (for the quartz-to-moganite ratios), X-ray fluorescence (for bulk geochemistry), and visible-to-shortwave infrared hyperspectral imaging (for a broader optical fingerprint that doesn’t require contact with the sample).

The results were clean. Brazilian agates consistently showed moganite concentrations of 8% or higher in their cryptocrystalline bands. Idar-Oberstein agates came in below 2%. No overlap. The two populations were distinct enough that a trained instrument could tell them apart without removing a single stone from its setting.

Then we took the library to the Getty.

With the help of Getty conservator Arlen Heginbotham, and using hyperspectral imaging equipment developed with colleagues at Caltech — a custom-built Headwall Photonics sensor that captures both visible and shortwave infrared wavelengths simultaneously — we imaged the entire surface of the Borghese-Windsor Cabinet. Every panel. Every fitted stone. The instrument sees what the eye cannot: differences in reflectance that correspond to differences in mineralogy, traced across the full surface in a single pass.

We compared what we saw on the cabinet to what we had measured in the laboratory.

The cabinet had something to show us.

Two Populations

There are two distinct populations of agate on the Borghese-Windsor Cabinet.

Share

Look at the image above. The cabinet is the same object you saw in the first photograph, but now the instrument has stripped away everything except chemistry. The ebony goes black. The gilt bronzes disappear. And the agate panels divide themselves, with no ambiguity, into two groups: those that fluoresce one way, and those that fluoresce another. The eye couldn’t see this. The instrument could.

The first population matches Idar-Oberstein cleanly. Low moganite. Old stone. Ancient Permian silica from the Nahe Valley, exactly as the historical record would predict. These appear as the darker, cooler panels in the hyperspectral image — spectrally quieter, geologically older.

The second population is brighter. Its spectral signature is not what we see from the Idar-Oberstein reference specimens. And here is where it gets interesting: it doesn’t clearly match Brazilian agate either. It’s brighter than the German material, but it doesn’t show the strong red hyperspectral colors we associate with Rio Grande do Sul agate in our reference library.

About half the agate on this cabinet carries a different geological biography than we expected. We just don’t yet know which one.

Where Did the Other Half Come From?

This is the part of the investigation that is still open. And I want to be honest with you about what we know, what we suspect, and what remains genuinely mysterious.

The moganite clock tells us the second population is probably younger than the Idar-Oberstein material — meaning it’s not from the Nahe Valley. But younger could mean a lot of things. It could mean a different European locality. It could mean India. And — here is the hypothesis I find genuinely plausible, if not yet provable — it could mean Brazil, arriving by a route that nobody has documented but that history suggests was entirely possible.

Before we get to the candidates, one alternative explanation deserves to be named and set aside: restoration. Over four centuries, a cabinet accumulates repairs. Could the second population of agate be replacement material — stones swapped in during some 19th or 20th-century conservation intervention, sourced from wherever was convenient at the time?

I went directly to the Getty conservation team with this question. The answer is no. Every restoration performed on the Borghese-Windsor Cabinet is documented in meticulous detail — what was done, when, by whom, using what materials. The conservation records account for the object’s condition comprehensively. There is no documented agate replacement that would explain a second geological population distributed across half the cabinet’s surface. The stones are original. The question of where they came from stands.

India: Agate from the Gujarat region, around the port of Cambay, was present in European trade networks by the 16th century. The Portuguese, who controlled the spice routes through the Indian Ocean, were moving material from India to Lisbon well before 1620. Indian agate is also Deccan Traps in origin — Cretaceous volcanic flows, younger than the Permian Nahe Valley material. It would show elevated moganite. It is a real candidate, and ruling it out requires reference specimens from the right Indian localities that I haven’t yet had the chance to analyze. That expedition is next.

Bohemia: There are other European agate localities — Bohemia produced agate worked into objects during this period. Whether the spectral signature would be distinguishable from Idar-Oberstein material at this level of analysis is an open question.

Brazil: Here is where the story opens up.

To understand why Brazil is a genuine candidate rather than a romantic guess, you have to go back to 1528, and to a debt.

The Holy Roman Emperor Charles V owed an enormous sum of money to the Welser family — a banking and merchant dynasty based in Augsburg, in what is now southern Germany. He couldn’t repay it in coin. So he repaid it in territory. The Welsers were given colonial rights to the province of Venezuela, in exchange for the debts Charles had accumulated financing his own imperial election. They named the colony Klein-Venedig — Little Venice. It existed from 1528 to 1546, and it was staffed, from the beginning, with German miners.

Not farmers. Miners. The Welsers sent organized extraction operations into South America at the same moment the stones for the Borghese-Windsor Cabinet were beginning to be sourced. Their contemporaries and rivals, the Fugger family — equally vast, equally German, equally Augsburg-based — were running what amounted to the largest mining corporation in Europe, with operations spanning the Tyrol, Hungary, Silesia, and Spain, and products moving through Lisbon to Africa and India. Both families had trade offices threading South America into the European mainland. The Welsers maintained outposts simultaneously in Lisbon, Rome, and Santo Domingo. Lisbon: the port through which any Brazilian material would have transited north. Rome: the city where the cabinet was being built.

And the Welsers weren’t just moving spices and sugar through Lisbon. Surviving business records from the early 1540s document Bartholomäus Welser’s representative in Lisbon, Konrad Stuntz, drawing 1,800 ducats from the firm’s account specifically to purchase precious stones — gemming thirty rings with diamonds and rubies for the European market. German merchant capital, flowing through Lisbon, buying stones. This was their business. This is what they did.

Klein-Venedig collapsed in 1546 when the Spanish crown revoked the Welsers’ charter. But the Welser firm itself didn’t die with the colony — it limped on, reorganized, continued trading through its Lisbon and Antwerp offices, and didn’t finally go bankrupt until 1614. Six years before the Borghese-Windsor Cabinet was built. The network outlasted the colony by seven decades.

Then came the Iberian Union.

In 1580, a succession crisis handed the Portuguese crown to the Habsburg king Philip II of Spain. Portugal and Spain — and all their colonial empires — came under a single ruler. The union lasted until 1640. The cabinet was built in 1620, squarely in the middle of it. This matters: during the Iberian Union, German merchants with existing Habsburg relationships had structural access to both the Portuguese and Spanish colonial worlds simultaneously. The wall between the Portuguese Atlantic and German commercial networks hadn’t just developed a door — it had been partially removed by the same dynasty the Welsers and Fuggers had been financing for a century.

German colonial presence near the agate deposits of Rio Grande do Sul in southern Brazil didn’t come until much later. The great wave of German immigration to that region — the settlers who would eventually discover the agate and ship it home as ballast in the 1820s — didn’t arrive until after Brazilian independence in 1822. Those settlers came from the same communities, carrying the same lapidary traditions, following the same gravitational pull toward extractable stone. They were the heirs of a relationship between German mining culture and South American geology that the Welsers had begun three centuries earlier.

What I have not found — and I want to be precise about this — is a specific cargo manifest documenting rocks from Brazil moving through Lisbon to Germany or Rome in the years between 1550 and 1820. That document may exist in the Arquivo Histórico Ultramarino in Lisbon, where Portuguese colonial shipping records from this period are held. Much of it has never been digitized, and none of it has been searched with this question in mind. What I have found is the people, the network, the motive, and the mechanism. The Welsers were buying stones through Lisbon in the 1540s. Their trade infrastructure survived until 1614. The Iberian Union opened the Portuguese Atlantic to Habsburg-connected German merchants at exactly the moment the cabinet was being built. The ballast mechanism is documented history for the 1830s.

When I find that cargo manifest, I’ll post it as a comment.

In the 1820s, German emigrants in Brazil — many of them from Idar-Oberstein itself, driven out by hunger and a collapsing local industry — discovered enormous agate deposits in Rio Grande do Sul. The deposits were unlike anything they had seen at home: nodules the size of cannonballs, banded in layers so even and regular they looked almost manufactured. Beautiful material. And worthless, in the middle of Brazil, without a way to get it home.

They couldn’t afford to ship it commercially. So they convinced ship captains to load rough agate nodules as ballast for the return voyage to Europe. Ballast is weight. Weight is necessary. Rocks are free. The agate crossed the Atlantic as a passenger nobody intended to carry, and arrived in Idar-Oberstein in 1834. It saved the industry. Within a decade, up to 300 tons of Brazilian agate per year were making that crossing, and Idar-Oberstein was cutting stone for the world again.

The cabinet’s second population is bright enough, spectrally, to be consistent with younger agate. Whether it’s Indian, Brazilian, or something else entirely, I cannot yet say. The analysis is ongoing. We’re expanding the reference library, adding specimens from additional localities, building a more complete picture of what early 17th-century agate trade actually looked like in three dimensions.

The Cabinet That Hasn’t Been Asked Yet

There is one more object in this story.

The Sixtus Cabinet — the companion piece, built in 1585 for Pope Sixtus V — has never been analyzed. It sits in its own institution, decorated with agate from the same era and presumably sourced through the same networks as the Borghese-Windsor. If that cabinet also has two populations, we have a pattern. If it has one, we have a clue about what changed between 1585 and 1620. If it has three, all bets are off.

I know what instrument would answer that question. Someday, I’d like to have one.

What the Cabinet Knows

Four hundred years ago, someone chose these stones. They cut them and fitted them into the surfaces of an object built for a cardinal who would become pope, and they did it well enough that the cabinet survived dynasties and wars and the entire arc of the modern era and ended up in Los Angeles, where a mineralogist could point a spectrometer at it and discover that half its stones came from somewhere unexpected.

The label still says what labels always say. The analysis is still ongoing. The reference library is growing.

The cabinet has been patient. It can wait a little longer.

Aaron Celestian is Curator of Mineral Sciences at the Natural History Museum of Los Angeles County, former scientist at NASA’s Jet Propulsion Laboratory, and adjunct professor at USC. He writes Pocketful of Χtals because mineralogy is stranger and more alive than most people have been told.

If something in this post changed how you see a stone you’ve been carrying around — that’s the whole idea.

Acknowledgements

Getty Museum - Arlen Heginbotham

Caltech - Bethany L. Ehlmann and Rebecca Greenberger

I have a garnet in my lab that I cannot explain. I collected it as a teenager. I have been a professional mineralogist for over two decades. I still cannot explain it.

The crystal faces are catching the light and throwing it back metallic green — not an internal color, not a gemstone’s green, but a cold, iridescent metallic green, the kind you associate with the wing of a beetle. I turn it in my hand. The color moves. It is not a reflection of the sky.

The Place

The Aquarius Mountains in western Arizona are not a place that announces itself. From a distance they look like every other range in the Mojave transition zone — brown, dry, relentless. Up close they are steeper than they appear, the rhyolite cliffs dropping in vertical columns above a scrub desert that goes on until it doesn’t. You find garnets here by walking the talus slopes below the outcrops, eyes down, until the light catches something that doesn’t belong to the dirt.

My family has been coming here for decades. We are not the only ones. The black andradite garnets of the Aquarius Mountains are known to collectors — not famous, not rare enough to command serious money, but known. You pick them out of the rhyolite matrix or find them loose in the soil, perfectly formed dodecahedra, black and mirror-bright, sitting in their rust-orange weathering pockets like they’ve been waiting. Most of the time, that’s exactly what you find. Black. Bright. Expected.

The first time I found one that reflected metallic green, I was probably fifteen years old.

My father found it first, or something like it. I don’t remember the exact moment — I remember the object. A garnet that shouldn’t look the way it looked. The crystal faces were catching the desert light and throwing it back metallic green. Not dark green. Not olive. A cold, iridescent metallic green, the kind of color you associate with the wing of a beetle, not with a silicate mineral sitting in volcanic rock in the middle of Arizona. I turned it in my hand. The color moved with the light. It was not a reflection of the sky.

We had no framework for it. We had a conclusion instead: this was a new mineral species. What else could it be? Garnets are black or red or orange or green in the gemstone sense — transparent, colored by chemistry, passive. They do not reflect metallic green. They do not behave like metal. Whatever this was, it was not in the field guides we had brought.

We collected every one we could find. They were rare even then — most of the garnets on those slopes are the ordinary black kind, and the metallic ones appeared only occasionally, scattered without obvious pattern among their unremarkable neighbors. We carried them home convinced we were carrying something that had never been described.

What the Worn Edges Showed

Years passed. I kept collecting. I kept coming back to the locality, or sending family members who still make the trip. More specimens accumulated.

And then something happened that should have been a disappointment: I noticed the worn edges — and some areas where the metallic surface looked like it had peeled away.

Some of the metallic garnets, on their most exposed faces, had worn through. The metallic green was gone on those edges. And underneath — black. Ordinary, mirror-bright, andradite-black. The same mineral underneath, with something thin and green on top.

This was not a new mineral species. It was a coating.

That should have closed the question. Instead it opened it into something stranger. If the metallic green is a coating, then what is it? How did it get there? Why are some garnets coated and others — sitting right beside them in the same pocket, in the same rhyolite, under the same desert sun — not coated at all? Was the coating deposited when the garnets formed, twenty million years ago? Or later, by something else entirely?

I became a mineralogist. I still had no answer.

What I Now Know About the Garnets Themselves

The Aquarius Mountains sit in Mohave County, Arizona, on the boundary between two of the West’s great geological provinces — the Colorado Plateau rising to the northeast, the Basin and Range dropping away to the southwest. The mountains themselves are the transition, the place where the high, ancient platform of the plateau gives way to the stretched and faulted terrain of the desert below.

The rhyolitic volcanic field here is roughly 24 to 20 million years old — Miocene, late enough that the landscape we recognize was beginning to take shape, early enough that the specific volcanic episode responsible for these garnets is essentially undescribed in the published literature.

That last part is worth pausing on. The garnets are known. Collectors have been finding them here for decades. But the published record on this locality is thin and confused — sources disagree on which garnet species is actually present, and the formation mechanism has not been properly documented.

What I believe I am looking at is andradite — calcium iron garnet, Ca₃Fe₂Si₃O₁₂ — formed by vapor-phase crystallization directly in the rhyolite. This is an unusual formation pathway. Most andradite forms in skarns, where magma contacts limestone, or in hydrothermal systems where hot fluids move through rock and deposit minerals in fractures. Vapor-phase crystallization in rhyolite is a different process entirely: minerals growing directly from volcanic gases, without a liquid intermediary, in cavities within the cooling lava.

If that formation mechanism is correct, these garnets grew from steam.

The metallic green coating formed on crystals that had already solidified from vapor. Whatever deposited that coating was a later event — a fluid, an alteration phase, a chemical reaction between the garnet surface and the environment around it after the volcanic episode had ended.

That sequence matters. The garnet is one record. The coating is a second record, written on top of the first, in a language I haven’t translated yet.

The Question, Properly Stated

I still have the original specimens from that first family trip. They are sitting in my lab at the Natural History Museum of Los Angeles County, next to newer ones collected by family members who still make the trip to the Aquarius Mountains. Thirty-five years of the same question, accumulated in a tray.

Here is what I know: the metallic green is a coating. It is nanometers thin in places — thin enough that the worn edges of some crystals have lost it entirely. It is not uniformly distributed across all the garnets at this locality, even among garnets sitting in the same pocket. It has not been published. Nobody, as far as I can determine, has identified it in the literature.

Here is what I don’t know: what it is. What phase it is. What chemistry. Whether it formed at the same time as the garnet or later. What chemical conditions were required to deposit it. Why some garnets have it and others don’t.

The lab now has instruments that can ask the question properly. What those instruments are, and what it takes to use them on a coating that may be only nanometers thick, is the subject of the next post.

This is the first post in an open investigation. I don't have the answers yet. Neither does the literature. The next post goes into the lab.

I am the Curator of Mineral Sciences at the Natural History Museum of Los Angeles County, former scientist at NASA's Jet Propulsion Laboratory, and adjunct professor at USC. I write Pocketful of Χtals because mineralogy is stranger and more alive than most people have been told.

All photos by Stan Celestian

I was doing trail maintenance at Gold Creek Nature Preserve — land maintained by the Los Angeles Community College District for faculty, staff, and student research — when I found them. Dried. Still on the branch. Each one about the size of a ping pong ball, hard on the outside and powdery within, each one with a clean exit hole where something had left. I knew what they were. I put ten of them in my pocket and kept working.

Oak galls. The exit wound of a wasp larva that had spent its entire early life inside a growth the tree made around it — a structure the tree built to contain something it couldn’t expel. The wasp is gone. The gall remains. Inside the dried tissue: tannic acid, iron compounds, and a chemistry that the Western world used to write down everything it believed was worth keeping for roughly fourteen hundred years.

The powder inside those galls is partly crystalline — gallic acid breaking down from the gallotannins, forming colorless needles as the structure collapses after the wasp leaves. I had crystals in my pocket on that hillside without knowing it. The ink begins before the ink begins.

I am a mineralogist. I also practice calligraphy, and I have spent years studying how letterforms change — how a mark made in the first century becomes something subtly different by the fifth, unrecognizable by the fifteenth, and then transforms again in the hands of someone working today who has decided the letter doesn’t have to look like anything that came before. My friend Jason Logan at the had been in the back of my mind for months. Jason forages all over the world — walnut husks, rusty nails, goldenrod, whatever a city offers up — and his book Make Ink had made me want to find my own source material. The galls were golf ball sized, hard on the outside and powdery within, each one with a clean exit hole where the wasp had left. I put ten of them in my pocket and kept working. If you want to try this yourself, start with Jason’s book.

For one thousand four hundred years, the Western world wrote itself down in iron.

Every document that shaped how we understand law, science, faith, and music was made from the same recipe: tannic acid from a wasp’s nursery, melanterite — a hydrated iron sulfate mineral pulled from the ground and dissolved into solution — gum arabic to control viscosity and bind the ink to the page, and water. The Lindisfarne Gospels, copied by a monk named Eadfrith on a tidal island off the Northumberland coast around 715 CE. The Magna Carta. Leonardo’s notebooks. Bach’s cantatas, written in his own hand before the music was ever performed. Galileo’s notes on motion — scientists have since used the iron concentration in different ink batches to sequence the order he wrote them. And Isaac Newton, who mixed his own iron gall ink from oak galls, gum arabic, copperas, and beer, wrote the recipe in his laboratory journal and added a note at the bottom: With this Ink new made I wrote this.

Sumi and other carbon-based inks are pigments — fine particles of carbon physically trapped between the collagen fibers of parchment. They don’t react with the skin of vellum. A pumice stone could dislodge them, lifting the carbon free, leaving the vellum surface intact and ready for new text. This is how palimpsests were made — the same skin written on, erased, and written on again. The Archimedes Palimpsest. The Sinai Palimpsest. Centuries of reused vellum, the old text ghosting beneath the new.

Iron gall ink chemically bonded with the collagen fibers of the animal skin. Pumice couldn’t touch it. Repeated scraping damaged the vellum. Acidic washes were tried and rarely worked completely — the ghost of the mark remained in the skin long after the surface was gone. The monks chose this ink precisely because it could not be undone.

There was something else. Fresh iron gall ink is a pale grey solution that looks like weak tea. Within minutes of hitting the page, as the ferrous tannate complex oxidizes in air and converts to ferric tannate, it darkens. Purple-grey first, then deep black. The scribe watched the words arrive. Carbon ink gave you exactly what you put down. Iron gall gave you a mark that developed, that committed, that declared itself permanent by becoming darker as it dried.

What the monks did not know — what nobody knew until conservation scientists began examining degraded manuscripts in the twentieth century — is that the same chemistry making the ink permanent was also, slowly and inevitably, destroying the substrate it had bonded with. The document and its destruction are the same reaction on different timescales. The monks writing the Gospels were simultaneously preserving and consuming them. They just wouldn’t live long enough to see it.

The destruction was the cost of that permanence. For centuries it was a cost nobody knew they were paying.

Nobody handed the medieval scriptoria a user manual.

The recipes that survive — and many do, collected in dedicated treatises and recipe books copied across centuries — are not instructions so much as field notes. The first known reference to iron gall ink appears in Martianus Capella’s encyclopedic work on the seven liberal arts, written in Carthage around 420 CE — a passing mention of a mixture of galls and gum, enough to know the ink existed, not enough to make it. The full recipes come later. Theophilus, a twelfth century monk, dedicated an entire treatise — De diversis artibus — to the arts of painting, glassmaking, and metalwork, including detailed ink recipes. The Compositiones ad tingenda, an eighth century compilation, is the earliest known recipe collection in the Western tradition. By the Middle Ages the recipes are everywhere, each one a report from someone who found a ratio that worked and wrote it down before they died.

The proportions vary from monastery to monastery, century to century, one scribe’s hand to the next. Some monks used wine instead of water. Some used beer, as Newton did six centuries later. Some added vinegar to accelerate the iron reaction. Some steeped the galls for days, some for weeks.

Every variation was an experiment. None of them called it that.

What they were tracking — without instruments, without chemistry, without any framework beyond observation and result — was the same set of variables a conservation scientist tracks today. Iron concentration. Acidity. Viscosity. The ratio of tannin to iron that produces a stable, dark, permanent mark without accelerating degradation so fast the document fails within a generation. They were optimizing a chemical system by hand, recording their results in the system itself, and passing the knowledge forward through apprenticeship rather than publication.

When you write with iron gall, you put the mark down pale. Grey, like weak tea dragged across the page. Then you watch it happen. Within seconds — not minutes, seconds — it darkens. The ferrous tannate oxidizing in air, converting to ferric tannate, the iron committing to its new state. By the time your eye travels back to the beginning of the line you just wrote, the beginning has already changed. The mark is declaring itself while you’re still making it.

No other medieval ink does it. No other ink makes the commitment visible in real time quite like it.

I have studied ancient handwriting from the first century forward — how a letterform shifts across decades, how a hand trained in one tradition carries its habits into another, how the same letter looks different cut in stone, brushed with carbon, pressed in type, or drawn with a nib on vellum. What I keep returning to is this: the calligrapher working in iron gall is never working with a finished material. The ink is in process. The letter is in process. The hand adjusting viscosity for humidity, testing flow on scrap before committing to the page, watching the mark arrive rather than simply appear — that hand is doing what every empiricist does. Observing a system. Adjusting variables. Waiting to see what the material decides.

The historical calligrapher mastering a dead script and the contemporary lettering artist pushing a letter until it stops being legible are doing the same thing — testing what a mark can carry before it breaks. The monks were doing it too. The instrument changes. The question doesn’t.

Calligraphy is not static. It never was. The letter itself is not static. The ink least of all.

The ink never stops reacting.

This is the part nobody told the monks. When the pen leaves the page the chemistry doesn’t stop. The iron tannate complex that makes the mark permanent continues to interact with the vellum, with the oxygen in the air, with the moisture in the environment, with the trace metals in the ink itself. The manuscript sitting in a library archive is not a static object. It is a slow chemical system. It has been running the same reactions for centuries. It is still running them now.

The degradation happens through two mechanisms working simultaneously. Iron gall ink is highly acidic — pH between 1 and 3 — and that acidity attacks whatever substrate it contacts. On parchment it degrades the collagen structure of the skin. On paper, which became the dominant substrate as iron gall ink spread beyond the scriptoria into secular use, it hydrolyzes cellulose chains, making the paper progressively more brittle until whole sections along the written lines simply fall away. The second mechanism is more aggressive. The iron ions in the ink drive a Fenton reaction — a free radical process that generates hydroxyl radicals, among the most reactive oxidizing agents in chemistry. These don’t corrode the surface. They break molecular chains at the microscopic level, shredding the cellulose from the inside.

Rusting is the closest common analogy. But rust sits on the surface and slows itself down as it builds up. Iron gall ink degradation doesn’t slow down. The Fe²⁺ ions that drive the Fenton reaction are continuously regenerated in the cycle — the reaction feeds itself. And as cellulose breaks down, it produces acidic compounds that further accelerate hydrolysis. The two mechanisms reinforce each other. The manuscript becomes more vulnerable as it degrades, not less.

And there is one more thing the monks couldn’t have known. The villain isn’t primarily the iron. It’s the copper — trace impurities in the green vitriol, present in every historical batch of iron sulfate, invisible to any instrument available before the twentieth century. Copper is a more powerful Fenton catalyst than iron. The purity of the raw materials the scribe used to make his ink determined the fate of the manuscript he wrote. A batch of green vitriol from one mine versus another. The difference between a manuscript that survives eight centuries and one that eats itself within two.

As the ink ages, it mineralizes.

The iron sulfate migrates through the paper fibers and crystallizes. Conservation scientists examining manuscripts from the fourteenth through seventeenth centuries using XRD and Raman spectroscopy have identified the mineral assemblage forming inside them: melanterite, rozenite, copiapite, amarantite, jarosite, iron oxalates. Named mineral species. The same minerals that form in acid mine drainage. The same minerals that form on the walls of abandoned mine shafts. The same minerals that NASA’s instruments have identified on the surface of Mars as evidence of ancient acidic water.

The gum arabic contributes its own minerals. Over centuries it breaks down through glycine degradation, producing oxalic acid that reacts with whatever cations are available — calcium from the paper filler, iron from the ink, copper from the trace impurities — crystallizing as calcium oxalate, iron oxalate, copper oxalate. The ingredient added to make the ink flow and adhere to the page becomes, over time, a second wave of mineral crystallization inside the document. The binder unmakes what it helped make.

A mineralogist reading the assemblage in a degraded manuscript can determine its degradation state the same way a physician reads a blood panel. Calcium and iron oxalates in low concentration, pH above five — early stage, recoverable. Magnesium and ferric oxalates — severe degradation, the document in the late stages of its conversion from record to mineral deposit.

I have the instrument to read this. Raman spectroscopy can identify every one of these phases non-destructively, without touching the document. I have iron gall ink made from galls I found on a Los Angeles hillside. I have not yet run it aged through the instrument. That is the next step.

What I already know is this: the manuscript the monks wrote to last forever is becoming something else. Not decaying — transforming. The words are still there. The minerals are forming around them. The document is doing what minerals do — responding to its environment, recording its own history in its crystal structure, becoming a different kind of record than the one the scribe intended.

The Lindisfarne Gospels are mineralizing. Bach’s cantatas are mineralizing. Newton’s laboratory notebooks, written in ink he mixed himself from oak galls and beer, are mineralizing.

The words are becoming stone.

I make the same ink. From galls I found on a hillside in Los Angeles, the same chemistry the monks used, the same minerals forming in the jar on my desk. I write with it. I sign my name with it. I named this publication with it.

Look at the name of this publication.

The Ρ is the Greek letter Rho — the second letter of Christos. It has always been there. And next to it, the letter most people read as an X.

Not an X. The Greek letter Chi — the first letter of Christos, the Greek word for Christ. Together they are ΧΡ — Chi Rho — the oldest sacred monogram in the Western world, the abbreviation early Christian scribes used for the name they considered too holy to write in full. When you see it in Pocketful of Χtals you are looking at that monogram, reversed. When you see it in Χmas you are not seeing Christ crossed out. You are seeing Christ written the way a sixth century monk would have written him — in the oldest shorthand for his name that exists in the written record.

The monks who copied the Gospels in iron gall ink did not begin their manuscripts with the first word of scripture. They began with the Chi. A single mark, written before the text each day. The start of a prayer compressed into a letter. The acknowledgment that what followed was not their work but something they were vessels for. The Chi came first. Everything else came after.

This is the nomen sacrum — the sacred name. Early Christian scribes developed a system of abbreviations for the names they considered too holy to write out in full. Christos became ΧΡ — Chi Rho — or simply Χ. The practice spread from the earliest Greek manuscripts through the Latin scriptoria of medieval Europe. Constantine put it on his battle standard. It appeared on the walls of Roman catacombs. It is on the Lindisfarne Gospels, on the Book of Kells, on every manuscript Eadfrith ever touched. It is the first mark the scribe made. And it could not be erased. Carbon ink could be scraped from vellum with a knife — the text recovered, rewritten, repurposed. Iron gall ink bit into the skin of the animal and became it. The Chi written in iron gall was permanent by the same chemistry that made it sacred. The monks chose this ink because the permanence of the material matched the permanence of the name. The knife that could erase everything else could not touch it.

The letter was never lost. It was just waiting for someone to look at it.

I named this publication pocketful of χtals knowing what that letter was. Chi — the scientific shorthand for crystal, and the oldest sacred abbreviation for Christ in the Western written tradition. The monks who wrote it in iron gall ink before beginning their manuscripts knew it as the second.

Share

I am the Curator of Mineral Sciences at the Natural History Museum of Los Angeles County, adjunct professor at USC, and former affiliate research scientist at NASA-JPL. I have studied ancient handwriting and manuscript traditions at Berkley. I make my own inks, including iron gall. I’ve enjoyed doing many forms of calligraphy from ancient uncials to modern letter forms.

Every unification in physics has come from finding a deeper symmetry. The search for a theory of everything is a search for the symmetry large enough to contain everything. And crystals got there first.

The Ball

I spent the better part of a morning wrapping a crystal ball in tissue paper, which gave me plenty of time to think about why so many cultures throughout human history decided this particular object was a window into the cosmos.

That’s the kind of question that sounds mystical until you actually try to answer it. Then it gets worse.

The ball had been in our collection for fifty years. Not in storage. On display. Under lights, behind glass, in a case that thousands of people had stopped in front of — on school trips, on first dates, on the kind of slow Tuesday afternoon when you wander into a natural history museum because you don’t know what else to do with yourself. They looked into it and felt something they probably couldn’t name. I know because I’ve watched them do it. Something about the clarity. The weight. The way light moves through a perfect sphere of silicon dioxide and comes out the other side changed.

There so many cultures throughout history that felt the same thing. Not about diamonds, not about pearls, not about gold — about this. A transparent sphere, held up to the light. Across four thousand years of recorded history, on every inhabited continent, someone held up a piece of quartz and said: this. This thing points toward something larger than itself.

I used to dismiss that as aesthetics. People liked shiny things, and the story accumulated from there.

But standing at that packing table, wrapping fifty years of other people’s wonder in tissue paper, I kept coming back to the question: why always a crystal? Why not glass, once they had it? Why not water in a bowl? Why did cultures that had never met each other — separated by oceans and millennia — keep reaching for the same material, and why a mineral at all?

There is an answer. It took four thousand years to arrive, and it is stranger than anything the mystics imagined.

The Loophole

Before I can answer the question, I have to tell you something that will seem like a detour. It isn’t.

Here is the most reliable law in physics.

Everything falls apart.

Not eventually. Constantly. Right now, as you read this, every system you can name is moving toward disorder. The coffee cools. The iron rusts. The mountain erodes. Stars burn out. Galaxies drift apart. The universe itself is expanding into increasing emptiness, and the physics is unambiguous about where that ends — not with a bang, not with meaning, but with a cold, dark, featureless uniformity in which nothing interesting ever happens again.

This is the second law of thermodynamics. Entropy increases. Disorder wins. It has no known exceptions. It is, as far as we can tell, the most ironclad rule in all of science.

And then a crystal forms.

Right there in the cooling magma, in the hydrothermal vent, in the groundwater moving through limestone — atoms that were drifting in solution suddenly lock into place. Not randomly. Not chaotically. In a precise, repeating geometric arrangement that extends in three dimensions with a perfection no human hand could achieve. A lattice. Exact. Symmetric. Ordered down to the distance between individual atoms.

This should not happen. Or rather — it should happen so rarely as to be essentially impossible. And yet crystals form constantly, everywhere, in conditions ranging from the inside of a volcano to the inside of your kidneys. The mineral world is made of them. The fossil record is made of them. You are, in several important senses, made of them.

So what is going on?

The answer is one of the most elegant loopholes in nature. A crystal forming locally increases order — but it releases heat to its surroundings in the process. The environment gets slightly more disordered so that this one small region can become extraordinarily ordered. The second law is satisfied globally. The accounting balances somewhere else. And in this one place, in this cooling solution, in this particular arrangement of silicon and oxygen atoms, something perfect locks into place.

The universe trends toward disorder. The crystal forms anyway. Not by cheating. By finding the one door the second law left open.

Order, purchased at the cost of heat. Geometry, extracted from chaos.

So the crystal ball exists because the universe found a loophole. But that didn’t answer the question — it deepened it. If anything that forms a lattice qualifies, why a mineral at all, and why always this one? I kept pulling the thread.

230

What is a crystal, precisely?

Not poetically. Precisely.

A crystal is a lattice — a regular, repeating arrangement of atoms in space, the same pattern over and over in three dimensions, essentially forever. The unit cell, the smallest repeating unit, stacked in every direction like the world’s most perfect three-dimensional wallpaper.

The symmetry of that lattice is not arbitrary. It is constrained. Mathematically, rigorously, provably constrained. There are exactly 230 ways that a repeating pattern can be symmetric in three-dimensional space. Not approximately 230. Exactly 230. This was proven in 1891, independently, by three mathematicians working in different countries, and it has never changed. Every crystal that has ever formed — in every rock, in every organism, on every planet in the universe — fits one of those 230 patterns.

Not 231. Not 229. 230.

Pick up any mineral. Any one. The quartz on the windowsill of every crystal shop in America, the pyrite in the school geology kit, the halite you put on your eggs this morning. Describe its atomic arrangement precisely enough and you will find it sitting in one of 230 slots in a classification system that was complete before the airplane was invented. The universe, in this one respect, is not infinite. It is surprisingly, almost shockingly, finite.

Now let’s talk about quartz specifically. Because quartz is where I stopped being able to explain what I was looking at.

Quartz is silicon dioxide — one silicon atom bonded to two oxygen atoms, repeated. Simple enough. But its space group, P3₁21, means its silicon atoms arrange themselves in a helix — a spiral staircase of atoms winding through the crystal. And that staircase can wind two ways. Left-handed or right-handed. Two versions of quartz, mirror images of each other, identical in almost every way.

Almost.

Pass polarized light through a left-handed quartz crystal and the light rotates one direction. Pass it through a right-handed crystal and it rotates the other. The crystal is literally twisting light. Not metaphorically. Physically. The atomic geometry of quartz reaches into the electromagnetic structure of light and turns it.

Pliny the Elder knew something was happening. In the Naturalis Historia, written around 77 AD, he documented the use of crystallum orbis — crystal spheres — by Roman soothsayers, and described the material itself as a substance so perfectly congealed it would never change. He was wrong about the mechanism. But he was right that the object was doing something. Quartz specifically. Not glass, not water, not beryl — quartz, because of the geometry of that atomic helix, physically turns light in a way no other transparent material does. Seventeen centuries after Pliny wrote it down, a French physicist named François Arago put a number to it. The optical rotation Arago measured in 1811 was already there in every Roman crystal ball, bending light, waiting for the mathematics to arrive.

Four thousand years of people holding this object up to the light were responding to something real.

That should have been a satisfying answer. It wasn’t. Because then I found out what Southampton did with it.

In 2013, researchers at the University of Southampton realized that quartz’s relationship with light wasn’t just a curiosity — it was a writing system. Using a femtosecond laser, they encoded data not in three dimensions but in five: x, y, z position, plus the size and orientation of nanostructures created inside the glass. Those structures change the way light travels through the quartz — the same optical rotation Arago measured — and can be read back out with a polarizing microscope.

Capacity: 360 terabytes per disc. Thermal stability: up to 1,000°C. Projected lifespan at room temperature: 13.8 billion years — the current age of the universe.

They have since encoded the Universal Declaration of Human Rights, Newton’s Opticks, the Magna Carta, and the King James Bible into quartz glass. They called it the Superman memory crystal.

Pliny was watching five-dimensional data storage. He thought it was frozen water.

Quartz is already reaching beyond what three-dimensional geometry alone can describe. Which is where the thread led next — and where mineralogy ran out of road. The mineral kept going.

The Forbidden Mineral

The 230 space groups have a rule.

Crystals can only have certain kinds of rotational symmetry. Two-fold, three-fold, four-fold, six-fold. That’s it. Five-fold symmetry — the symmetry of a starfish, a sand dollar, a sea urchin — is forbidden. Strictly, provably, completely forbidden by the mathematics of repeating lattices in three dimensions.

In 1982, Dan Shechtman looked into his electron microscope at an aluminum-manganese alloy and saw tenfold symmetry.

He checked his instrument. He recounted. He was not wrong.

What he was looking at would eventually be called a quasicrystal — a new state of matter with long-range order, sharp diffraction patterns, and all the hallmarks of a crystal, except that it was doing something crystals cannot do. It had broken the rule. Shechtman’s colleagues told him he was mistaken. One prominent scientist suggested he go back and read a textbook. He spent two years unable to publish. When he finally did, in 1984, it triggered a complete paradigm shift in crystallography.

In 2011, Dan Shechtman received the Nobel Prize in Chemistry.

In 2009, a geologist named Luca Bindi found something in a museum collection in Florence — grains of aluminum, copper, and iron with fivefold symmetry, associated with a meteorite from a remote region of eastern Russia. A quasicrystal. Natural. Geological. Real.

The mineral was named icosahedrite. Analysis revealed it had not formed on Earth. The oxygen isotopes, the mineral assemblage, the chemistry — all pointed to a carbonaceous chondrite asteroid. Icosahedrite formed 4.5 billion years ago, probably during a high-velocity collision in the early solar system. Earth cannot make it. The conditions here cannot sustain it. The universe had to build it elsewhere, in violence, and deliver it.

To describe icosahedrite’s atomic structure mathematically — to write down what it actually is — three dimensions are not enough. The structure requires a six-dimensional lattice. Icosahedrite is, in the precise technical sense, a three-dimensional slice through a six-dimensional crystal. It exists on a shelf in Florence. Its full identity requires more dimensions than our world contains.

I kept thinking about what Pliny’s soothsayers said — that a crystal points toward something larger than itself. I’d assumed that was poetry. Then I found out about icosahedrite. I’m not sure “points toward something larger” covers a mineral whose existence requires six dimensions to describe. But the question kept going.

The Ladder

In eight-dimensional space there is a lattice called E8.

You cannot visualize it. Nobody can. But you can describe what it does: in eight dimensions, E8 is the most efficient way to pack spheres — not one of the most efficient, the most efficient. Uniquely, provably, perfectly optimal. Each sphere touches exactly 240 neighbors simultaneously. Mathematicians find it unreasonably beautiful, and it keeps appearing in places it has no business being.

The forbidden symmetry of icosahedrite — the fivefold rotation that breaks three-dimensional crystallography — is mathematically connected to E8. The atomic arrangements of icosahedral quasicrystals, the class of mineral to which icosahedrite belongs, can be understood as a projection of E8 down into three dimensions. The forbidden symmetry doesn’t come from nowhere. It comes from eight-dimensional geometry, casting a shadow into our world.

The mineral that fell from space was carrying something older than the solar system. Not metaphorically. Structurally.

In 2010, a team of physicists at Oxford University published a paper in Science. They had taken a crystal of cobalt niobate — not a quasicrystal, not from space, just a crystal grown in a laboratory — and cooled it to 40 millikelvin, forty thousandths of a degree above absolute zero. Then they applied a magnetic field and tuned it toward a critical threshold.

At 5.5 Tesla, the magnetic order of the crystal dissolved. The electron spins began fluctuating in what physicists call a quantum critical state. And the resonant frequencies of those fluctuations — the notes the crystal played at the edge of its own order — appeared in the exact ratios predicted by E8.

The first two frequencies were in the ratio 1.618. The golden ratio.

“It reflects a beautiful property of the quantum system,” the team’s leader, Radu Coldea, said. “A hidden symmetry. Actually quite a special one called E8 by mathematicians — and this is its first observation in a material.”

Not in a particle accelerator. Not in a theoretical calculation. In a crystal, in Oxford, measured with neutrons.

I hadn’t expected any of this when I started wrapping a crystal ball in tissue paper. I’d expected to think about optics for a few minutes and go back to work. Instead I found myself reading about quantum criticality in a cobalt crystal in Oxford, and then about the geometry that appeared in that experiment — E8 — and how the same structure sits at the foundation of the best current mathematical candidate physicists have for a unified description of all four fundamental forces of nature. The gauge structure that framework requires is E8 × E8. Two copies. Whether it describes our universe is genuinely unresolved. Whether it will ever be confirmed by experiment is uncertain. But it has been measured. In a crystal.

That’s where I had to stop and look back at where I’d started.

Here is the ladder, and at every rung the crystal ball is still in the frame.

Quartz — the most common mineral on Earth — whose full optical behavior exceeds what three-dimensional geometry can describe, and whose structure we have now learned to write in five dimensions for data storage that will outlast the sun.

Icosahedrite — a mineral the universe had to build in a collision 4.5 billion years ago and deliver by meteorite, because Earth cannot make it. Its atomic structure requires six dimensions to describe. Its forbidden symmetry is a projection of something eight-dimensional.

Cobalt niobate — a crystal in a laboratory, cooled to near absolute zero, whose quantum fluctuations resonated in the exact frequencies of E8. The first time that geometry appeared in physical matter.

E8 × E8 — the mathematical foundation of the best current candidate for a theory of everything. Not confirmed. Not abandoned. Still being worked out.

Every rung is the same kind of object. I didn’t arrange that — I just followed the question.

And when you’ve followed all of it, the question changes. It’s no longer why did people hold up crystals and feel they pointed at something larger? It’s: how did they know?

Before Entropy Won

Now I want to go back to the second law, and tell you what I left out.

Everything trends toward disorder. The crystal forms anyway — that’s the loophole. But the loophole is downstream of something much older.

The universe itself began in a state of extraordinary symmetry.

Not disorder. Symmetry.

In the first fractions of a second after the Big Bang, all four fundamental forces — gravity, electromagnetism, the strong nuclear force, the weak nuclear force — were unified into a single interaction. One force. Perfect, undifferentiated symmetry. As the universe expanded and cooled, that symmetry broke. The forces separated from each other, one by one, like crystals precipitating out of a cooling solution. Each separation produced structure. Particles. Atoms. The periodic table. Minerals. You.

This process has a name in physics: spontaneous symmetry breaking. And it is — precisely, technically, not metaphorically — the same mathematics as a crystal forming from a melt.

When a liquid cools and crystallizes, it breaks a symmetry. The liquid looks the same no matter how you rotate it — perfect rotational symmetry in every direction. The crystal that forms picks one specific orientation, one specific lattice, and locks it in. The infinite freedom of the liquid — its continuous symmetry — collapses into the discrete symmetry of the solid. Symmetry broken. Structure made.

The universe did this. Repeatedly. On the largest possible scale. The forces we experience, the particles we are made of, the minerals in the ground — all of it is the residue of symmetry-breaking events in the first moments of time. The universe cooled. It chose geometries. It locked them in.

A crystal is what that process looks like when you hold it in your hand.

Not a metaphor for the cosmos. Not a symbol of cosmic order. A literal, physical, chemical record of the same kind of event — a system with infinite freedom choosing one geometry and locking it in. Symmetry broken. Structure made. The universe having chosen, and held.

I think that’s the answer. Not the one I was looking for at the packing table — not optics, not aesthetics, not cultural accident. Every person who ever held a crystal up to the light and felt that it mattered was responding to something written into the object at the atomic scale, something that connects the thing in their hand to the event that made everything. A fossil of the moment the universe chose this geometry and not another.

I don’t know what else to call that except right.

Pliny documented the crystallum orbis in 77 AD — crystal spheres, used by soothsayers, pointing toward something the Romans couldn’t name. He was wrong about the mechanism and right about the object. Arago gave it mathematics in 1811. Southampton wrote five-dimensional human history into it in 2013. Coldea found its deeper symmetry resonating in a laboratory crystal in 2010. And underneath all of it — the Roman encyclopedia, the polarizing microscope, the neutron scattering data — the same object. The same quartz. The same lattice. One of exactly 230.

Four thousand years of human intuition, waiting for the physics to catch up.

The second law says everything trends toward disorder.

The crystal forms anyway.

It isn’t fighting entropy. It’s a record of something that happened before entropy had fully won — a fossil of the moment the universe chose this geometry, compressed into something you can hold.

Every crystal that has ever formed is a record of the cosmos cooling into structure.

We Have Another One

I sealed the box. I watched it go.

The Smithsonian’s crystal ball, fifty years on display, wrapped in tissue paper and shipped back to Washington. A perfectly ordinary act of institutional housekeeping. Two museums, one object, a transfer of custody. Nothing dramatic.

Except that I now knew what I was shipping.

Not a mystical object. Not a prop for seeing the future. A piece of quartz that twists light because of the geometry of its atomic helix — the same geometry Pliny’s soothsayers held up to lamps in the first century, the same geometry Arago measured, the same geometry Southampton used to encode the Magna Carta for the next thirteen billion years. A lattice that sits at the base of a ladder leading through five dimensions, six dimensions, eight dimensions, all the way to the geometry that may underlie the structure of everything. A fossil of the moment the universe chose this symmetry, and locked it in.

Four thousand years of people holding this object up to the light and feeling that it pointed toward something larger than itself.

They were right. Not about the magic. About the object.

I walked back through the collection.

And then I stopped.

Because we have another one.

About Me

Aaron Celestian, PhD is Curator of Mineral Sciences at the Natural History Museum of Los Angeles County, adjunct professor at USC, and was an affiliate research scientist at NASA-JPL. He studies minerals the way other people study languages — as records of events too large and too old for any other archive. Pocketful of Χtals is where the specimens talk back.

Subscribe now

She was making her own sunscreen.

Not buying it — formulating it, from scratch, mineral-based, because she’d done enough research to have opinions about particle size and UV coverage and the difference between a physical blocker and a chemical one. She knew I was the curator of minerals at the Natural History Museum, so when our kids’ soccer team took the field and we ended up standing next to each other on the sideline, the conversation went somewhere most sideline conversations don’t.

She showed me the bottle. I borrowed some — it was a warm day — and we talked about TiO₂ while the game ran. I told her everything I knew. I didn’t tell her everything I suspected.

I told her what I knew. Titanium dioxide comes in three mineral forms: rutile, anatase, and brookite — chemically identical, architecturally different, each with its own relationship to light and reactivity. Industry chose rutile for sunscreen deliberately. It’s the stable polymorph, the calm one, inert enough to put in food and cosmetics and paper without concern. Rutile powder is one of the whitest materials known — the same mineral that grows as deep red or black crystals in a collection becomes brilliant white when ground small enough. That’s what goes in the bottle.

She asked about the photocatalytic forms — she’d read about anatase generating reactive oxygen species under UV. I told her yes, that’s real, anatase is more active, which is why rutile is preferred for sunscreen. You can eat rutile, spread it on your skin, and it does nothing, which is exactly the point.

We talked for a while longer. The game ran.

The Mineral

Rutile rewards attention precisely because it refuses to look the same twice.

Three specimens, one mineral. The gold twin on hematite. The large silvery-black prismatic crystal from Georgia. The deep red transparent crystals from Minas Gerais. Same chemistry — TiO₂, titanium and oxygen, nothing else — but the color shifts from gold to silver to red depending on trace impurities, crystal size, and the geometry of how light moves through the structure. Collectors love it because it never looks finished. It has the quality of still becoming something.

Ground to powder, it becomes something else entirely: brilliant white. Rutile powder is one of the whitest materials known. It’s in white paint, white paper, white food coating. The bright white of a hospital corridor, the bright white of a wedding cake, the bright white of a printer page — much of that whiteness is rutile. The same mineral, just smaller, with the same crystal structure but a completely different relationship to light.

This is one of the stranger things about minerals at the nanoscale: size changes everything. At 200 nanometers versus 20 nanometers, the same material scatters light differently. Surface area explodes. The mineral doesn’t transform — its chemistry stays identical — but what it does in the world changes entirely.

Industry understood this. That’s why rutile is in sunscreen, in food, in the white of this page. It’s chemically inert across all of those applications. You can eat it. You can spread it on your skin. It does nothing, which is exactly the point.

There is, however, one form of titanium oxide that does something. Something quite different.

That’s the one that had been turning up in the folder.

The Wrong Spectrum

The Peru samples arrived as filters.

A team of climbers and atmospheric scientists had spent weeks in the Cordillera Blanca collecting snow at altitude — physically difficult, logistically complicated work conducted far from any laboratory. What arrived for analysis was the result: membrane filters loaded with particles melted out of high-altitude snow, labeled by site and elevation, ready for mineralogical characterization. The team was looking for black carbon and light-absorbing aerosols. The question was what minerals were present and in what proportions.

The answer, overwhelmingly, was clay. Fine-grained silicate dust carried up from the valleys and páramo below — the expected mineral background of high-altitude Andean snow. Black carbon was present too, the soot signature of combustion at lower elevations transported upward on rising air. Expected things, in expected proportions.

And then, in specific layers, something almost right.

Mineral identification in this kind of work is done with a Raman spectrometer — an instrument that fires a laser at a sample and reads the way the mineral scatters the light back. Every mineral has a characteristic fingerprint: a set of peaks at specific positions, like a barcode. Match the barcode to a reference database, and you have an identification. The instrument is fast, non-destructive, and works on particles too small to see with the naked eye. It is also unforgiving of anything that doesn’t match a reference — if the barcode is slightly wrong, you’re left with a closest match and a note that something is off.

Certain layers contained something in the TiO₂ region of the spectrum — peaks in positions that suggested titanium oxide, in a part of the Raman fingerprint where anatase and brookite live. But the peaks didn’t match either. They didn’t match anything. Not anatase. Not brookite. Not rutile. Not any mineral in the reference database. The barcode didn’t scan wrong — it didn’t scan at all.

The closest match was anatase, so that’s what went into the notes. Not because it was right, but because the database had to return something and “unidentified TiO₂-like phase” is not a mineral name. The data the team had asked for was returned. They got what they needed. The unidentified signal went into a folder, labeled with the closest approximation available, which happened to be wrong.

This is what responsible science actually looks like. You don’t invent explanations for signals that don’t yet have enough context to support one. You document, you flag, you move on.

The folder sat there. The work moved on.

The Same Signal, Everywhere

Except the signal didn’t stay in the folder. It followed.

The next samples came from the Great Salt Lake — brine delivered by a colleague running halite crystallization experiments, asking whether bacteria could survive entombment inside salt crystals as they formed. The Raman work was to identify mineral phases in the brine. The minerals were found and identified. But there it was again: the same shifted TiO₂ signal, persistent across multiple samples, in water from a landlocked basin in the middle of Utah. Labeled anatase-like, noted as anomalous, and the halite mineralogy was delivered as requested.

The Pacific Ocean filters came from Will Berelson’s group at USC and Jess Adkin’s group at Caltech — a North Pacific transect from Hawaii to Alaska, studying how pteropod shells made of aragonite dissolve in the water column and what that means for the ocean’s carbon budget. The Raman contribution was to confirm the carbonate mineralogy on filters collected by in-situ pumps at depth: map the distribution of calcite and aragonite across the filter surfaces, verify what the X-ray diffraction had shown. That work was clean. The carbonate identifications were unambiguous. But the filters also carried non-carbonate particles, and among them — in samples from multiple depths, multiple stations, hundreds of meters below the surface of the open Pacific — the same shifted peaks appeared again. Flagged. Filed. The dissolution study got its carbonate mineralogy.

Above: This is the project the Pacific filters came from. The carbonate mineralogy was clean. The other signal is what this post is about.

The fish gut samples came from Bill Ludt, the ichthyology curator at the museum, who wanted to know whether Raman spectroscopy could identify plastic polymer types in historical fish specimens — a microplastics project using the collection, specimens going back decades. It could. The protocols were worked out. The plastics were there, accumulating through time in the collection just as they accumulate in the ocean. But in open Pacific fish, collected far from any coastline, the gut contents also carried the signal. The same signal. In a fish preserved in ethanol in an archival jar, pulled from a shelf in the collections.

Four separate projects. A glaciologist, a geobiologist, an oceanographer, an ichthyologist. None of them asking the same question, none of them looking for the same thing. Each got the answer they came for. And in the background of every dataset, the same anomalous spectrum — in snow from the Andes, in brine from Utah, in water from the North Pacific, in the gut contents of open-ocean fish.

These places share almost nothing. Different hemispheres. Different altitudes. Different ecosystems. The Andes and the floor of the Pacific Ocean and a landlocked Utah lake and the gut of a fish preserved in a jar — there is no obvious thread connecting them. And yet the same unidentified signal kept appearing in all of them, in samples delivered by people who had never spoken to each other, asking completely different questions.

The connection wasn’t visible yet. But the same material had been called by two different names — sometimes anatase, sometimes something closer to hematite, depending on which peaks dominated in a given sample, depending on which reference mineral happened to be nearest in the database on a given day. Two mineral names. One signal. Four projects. One folder of anomalies that nobody had asked to be explained.

Then, in 2017, a paper arrived.

What It Is

Yang et al., Nature Communications, 2017. A team from Virginia Tech investigating a coal ash spill into a river in North Carolina. They found unexpected particles in the downstream sediment. Almost like rutile. But not. Characteristic superfine striations under electron microscopy. Raman peaks shifted from TiO₂ in a specific, diagnostic way.

Magnéli phases — TiₓO₂ₓ₋₁

The name comes from Arne Magnéli, a Swedish crystallographer who first characterized this family of titanium suboxides in the early 1950s — decades before anyone knew they were being produced by the billions of tons and carried to the top of the world on the wind.

These are titanium suboxides — members of a family of materials formed when TiO₂ is heated above 900°C in a low-oxygen environment. The oxygen atoms are stripped from the rutile structure along specific crystallographic shear planes. What remains is a material with the same titanium backbone but a fundamentally different character. Not white. Black. Not an insulator. A narrow-band semiconductor with electrical conductivity roughly 100,000 times higher than TiO₂. Absorptive across UV, visible, and near-infrared light. A material that, in Parker and Siegel’s 1990 calibration work, produces exactly the kind of shifted, broadened Raman peaks that had been accumulating in that folder for years — peaks that drift between “looks like anatase” and “looks like hematite” depending on which suboxide stoichiometry dominates in a given sample.

The conditions required to form Magnéli phases — temperatures above 900°C, reducing atmosphere, TiO₂ as a starting material — describe exactly one widely distributed industrial process.

Coal combustion.

TiO₂ is a minor accessory mineral in coal worldwide. It sits there unremarkably in the rock for millions of years. Then it goes into a furnace at 1,400°C with almost no available oxygen, and it comes out the other end as something that has never existed in nature. Coal burning plants in North Carolina. In Kentucky. In Virginia. In Texas. In Shanghai. In Chongqing. Yang et al. tested ash from 22 plants across the United States and China. Every single one contained Magnéli phases.

Every anomalous spectrum in the folder resolved.

The Glacier

Here is where the story stops being abstract.

The Peru snow samples came from a specific location: the wall of a crevasse on Vallunaraju mountain — the glaciated peak nearest to Huaraz, the largest city in the region. The crevasse wall showed what glacier walls always show if you look carefully: layers. Dark bands separated by cleaner ice, each band recording a season of accumulation, a year of deposition, a chapter in the mountain’s slow autobiography.

The absorptive capacity of particles in each layer was measured using a technique that exposes filters to visible light and records the temperature increase — the Light Absorption Heating Method. The results showed distinct spikes at layers 4, 8, 12, and 16 in the sequence, with absorptive capacity dramatically higher than the surrounding ice.

The interpretation was black carbon. Coal combustion produces black carbon. Black carbon absorbs light. Black carbon on glaciers accelerates melt. The logic was sound. The data supported it.

Here is what the technique cannot do: distinguish between black carbon and other dark, light-absorbing particles. It measures heat. It was calibrated against black carbon because that was the dominant hypothesis, the reasonable hypothesis. The study was not wrong to reach that conclusion.

Here is what Magnéli phases are: black. Not dark gray, not brown — black, in the way that materials with near-zero band gaps are black, absorbing across nearly the full solar spectrum. Under the LACM technique, a filter loaded with Magnéli phases is indistinguishable from a filter loaded with soot. They produce the same temperature curve. They would be reported as the same effective black carbon concentration.

This is not a theoretical concern about absorption properties. Industry knows exactly what Magnéli phases do with light — they are actively engineered for it. Ti₄O₇, the most common Magnéli phase, is used in photothermal conversion applications specifically because it absorbs solar radiation across the full spectrum and converts it to heat with exceptional efficiency. Engineers are trying to put these materials to work as solar heat absorbers. The same property that makes them commercially attractive is what makes them concerning on the surface of a glacier at 6,000 meters, under some of the most intense UV radiation on Earth.

Coal combustion produces both black carbon and Magnéli phases simultaneously. They travel together in fly ash. They deposit together on glacier surfaces. And in the Raman data from those same Peru samples — the data labeled with the closest available mineral name and set aside — the layers with the highest absorptive capacity are precisely the layers where the unidentified signal appears.

At 6,000 meters, UV flux is among the highest on Earth. Magnéli phases, with their broad-spectrum absorption and their extraordinary conductivity, are photocatalytically active under those conditions in ways that TiO₂ is not. A small concentration of dark, highly absorptive particles on glacier ice, under intense UV, releasing energy directly into the snowpack — the heat capacity calculations are consistent with the observed temperature anomalies.

The Cordillera Blanca has lost 20-30% of its glacier area since 1970. Dry season discharge in some valleys is projected to fall by 70% compared to pre-warming levels. Millions of people depend on glacial meltwater for drinking water and agriculture.

The accelerated melt was attributed to black carbon.

That attribution was not wrong. But it was not complete.

What We Don’t Know

Here is what the data actually shows: Magnéli phases are present wherever Raman spectroscopy has been run on environmental samples from the industrial era — high-altitude glacial ice, the open Pacific Ocean, the gut contents of museum fish, a terminal lake basin in the middle of Utah. Coal combustion is the dominant known mechanism for producing them outside a laboratory, though recent work has identified combustion of TiO₂-rich synthetic materials in urban structure fires as an additional source. They co-occur with the high-absorption layers in the Peru glacier stratigraphy.

That much is solid ground.

What isn’t solid is what any of it means at scale.

The glacier melt question is the one that lingers. The LACM technique — used to measure light absorption in those Peruvian ice cores — cannot distinguish between black carbon and Magnéli phases. Both are dark. Both absorb broadly. Both are combustion products that travel together in fly ash and deposit together on ice. Every study that has used thermal absorption methods to quantify black carbon on glaciers has, without knowing it, been measuring a signal that includes Magnéli phases. How much of the attributed warming belongs to them? Nobody has separated the two contributions, because until very recently nobody knew to try.

Then there’s the biological question, which opens in a different direction.

Yang et al. tested Magnéli phase toxicity in zebrafish embryos and found something unexpected: significant mortality without light exposure. TiO₂ is primarily toxic through photocatalysis — it needs UV light to generate the reactive oxygen species that damage cells. Magnéli phases apparently don’t require it. The mechanism is unknown. The concentrations at which biological effects appear in environmental organisms — not zebrafish embryos in a controlled laboratory, but copepods, pteropods, the small things at the base of the food web that live in water loaded with atmospheric deposition — have not been measured. The fish examined in the microplastics study had Magnéli phases in their gut contents. What that means for those fish, beyond presence, remains an open question.

And then there’s the pattern visible only when the Peru stratigraphy data is laid out in full: the four layers with dramatically elevated light absorption — layers 4, 8, 12, and 16 in the crevasse sequence — fall at perfectly regular intervals. In glacier stratigraphy, regular periodicity usually reflects regular deposition — a recurring atmospheric transport event on a predictable cycle, something that happens on a schedule and leaves a mark in the ice at regular depth. What that schedule is, or what drives it, is not yet known. The pattern is there. The explanation isn’t.

That might be the most honest summary of where this research stands: a gigaton of synthetic mineral distributed across the entire planet, present in the bodies of ocean fish and the walls of Andean glaciers and the sediment of a landlocked Utah lake, and the early chapters of understanding what it does are still being written.

The Sideline, Again

She’d done everything right. Rutile was the correct choice — inert, stable, the same material that goes into food and paper and paint without a second thought. The TiO₂ in that bottle was exactly what it was supposed to be.

The TiO₂ that went into the coal furnace was never meant to be there. It was an accessory mineral, unremarkable, sitting in the rock for 300 million years. Then it went into a furnace at 1,400°C with almost no oxygen, and it came out as something that has never existed in nature — something black, conductive, photocatalytically active, and now present in every glacier and every ocean on Earth.

There’s a coda to that sideline conversation that I couldn’t have offered at the time.

She asked about the photocatalytic forms — anatase, the reactive one. I told her rutile was the safer choice, which is true. And I almost mentioned the one form of titanium oxide that’s genuinely worth paying attention to — not TiO₂ at all, but a family of titanium suboxides produced under extreme heat in low-oxygen environments. Black. Electrically conductive. Photocatalytically active under visible light in ways that TiO₂ is not. The kind of material where you’d know immediately if you were working with it — because nothing in a mineral sunscreen formulation is black.

I didn’t say any of that. I didn’t have a name for it yet. I had a folder.

Now I have a name. And the black ones are already everywhere — in the ice above Huaraz, in the water of the North Pacific, in the sediment of a Utah lake, in the bodies of fish that never came near a coal plant.

The TiO₂ she spread on her skin that afternoon protects her. The TiOx spread across the planet by two centuries of industrial combustion is doing real harm — to glaciers, to ocean chemistry, to organisms whose tolerance for a novel synthetic mineral has never been tested, in concentrations that have never been measured.

She made a good sunscreen. That part of the story is fine.

The other part isn’t.

And here is the difference that will matter long after the rest of this is forgotten.

Radionuclides from nuclear testing, the plastics in ocean sediment, the nitrogen fixed by industrial fertilizers — these are the proposed markers of the Anthropocene, the signals future geologists might use to locate our moment in the rock record. Each has a problem. Radionuclides decay. Plastics degrade over geologic time. Nitrogen cycles out. These are transient signals, legible for thousands or perhaps tens of thousands of years before they dissolve back into background noise. On the timescale of deep geology, they are whispers.

Magnéli phases are structurally analogous to rutile — one of the most geochemically persistent minerals in the rock record, recoverable from sediments hundreds of millions of years old. They are thermally stable, chemically inert, and structurally robust under the same diagenetic conditions that erase most mineralogical evidence of surface processes. They will not dissolve. They will not decay. They will sit in the rock exactly as they were deposited.

And they are everywhere. Not concentrated near industrial centers, not localized to continental margins — but present in high-altitude glacial ice, in the deep floor of the open Pacific, in a landlocked basin in Utah, in the gut contents of fish collected far from any coastline. Synchronous. Global. Diagnostic. A thin layer distributed across the entire planet, deposited within the same narrow window of industrial time.

A geologist reading the rock record half a billion years from now will find that layer. They will recognize it as synthetic — no natural process produces titanium suboxides, and no natural process deposits them simultaneously at every latitude and depth. They will know, from that layer alone, that something happened here: a civilization that burned enough carbon, at high enough temperatures, in low enough oxygen, to transform an accessory mineral that had sat unremarkably in rock for 300 million years into a gigaton of something that had never existed before.

That layer will be us. Permanent, unambiguous, and readable in the dark.

Because the material is black.

The sunscreen she made is white.

Crevasse stratigraphy and LACM absorptive capacity data are from Schmitt et al. (2015), The Cryosphere, 9, 331–340. Magnéli phase discovery and global distribution estimates from Yang et al. (2017), Nature Communications, 8, 194. Raman calibration from Parker & Siegel (1990), Applied Physics Letters, 57, 943. Aragonite dissolution data from Dong et al. (2019), Earth and Planetary Science Letters, 513, 129–141. Wildland-urban interface fire ash as an additional source of Magnéli phases from Baalousha et al. (2026), Environmental Science & Technology, DOI: 10.1021/acs.est.5c09885.

The observations described here span glaciology, oceanography, geobiology, and ichthyology — no single journal owns this pattern, and no single paper could show it whole. That’s why it’s here first. Publicly available and reader accessible.

I’ve spent an embarrassing amount of time thinking about mithril.

Not the movie version — the shiny plot device that saves Frodo’s life in the mines. The canonical mithril. Tolkien’s mithril. The one described with almost metallurgical precision in Gandalf’s words in Moria:

“Mithril! All folk desired it. It could be beaten like copper, and polished like glass; and the Dwarves could make of it a metal, light and yet harder than tempered steel. Its beauty was like to that of common silver, but the beauty of mithril did not tarnish or grow dim.”

Apparently no one truly knows what inspired that description. Mithril is one of the very few things in Middle-earth that Tolkien invented wholesale, with no acknowledged mythological source and no letter, note, or interview in which he identifies a real-world referent. The Tolkien fandom has thought carefully about this — it’s one of the genuinely open questions in a legendarium where most things have documented sources — and the scholarship on Tolkien’s methods of world-building is rich and serious. I’m building on that scholarship, not correcting it. The properties he chose — silver appearance, non-tarnishing, malleable yet strong — are specific enough to suggest something concrete was in his mind. So I went looking for what that something might have been.

My argument is not that mithril is any real metal. It’s simpler than that: Tolkien was a world-builder who transmuted his immediate environment into fiction — he did it with geography, with architecture, with people he knew. The question worth asking is whether he did it with materials too. And if he did, the answer is hiding in plain sight in the city where he grew up.

The usual suspects, briefly

Tolkien scholarship has proposed titanium, aluminium, platinum, and silver as candidates. These are serious proposals from people who thought carefully about the text, and the titanium argument in particular has real intuitive appeal — lightweight, strong, corrosion-resistant. Each fails, though, and not just on physical grounds.

Titanium is very improbable. Pure metallic titanium wasn’t produced until 1910, and when Tolkien began writing in 1937 the entire global supply consisted of small experimental batches produced in a Luxembourg scientist’s home laboratory — by 1938 Kroll had managed just 50 pounds. Industrial production didn’t begin until 1948, the year Tolkien finished writing, at a grand total of 2 tons — a classified strategic material with no public profile as a metal. If “titanium” meant anything to a general reader in the 1930s and 1940s it was as a white paint pigment (titanium dioxide), which had been entering household use since the 1920s — a soft white powder with nothing in common with the lustrous, malleable, silver-coloured metal Tolkien describes. Beyond that, titanium work-hardens rapidly under mechanical stress — it is essentially the opposite of “beaten like copper,” which implies the easy plastic deformation of a ductile metal. On every relevant axis — cultural presence, physical behaviour, visual character — titanium fails.

Aluminium has a more interesting history. For a brief window in the mid-19th century it was more coveted than gold — Napoleon III reserved aluminium cutlery for his most honoured dinner guests while others made do with gold, and France’s 1855 Exposition Universelle displayed aluminium bars in glass cases alongside the crown jewels. In that moment it had exactly the mythic quality mithril carries. But the Hall–Héroult electrolytic process arrived in 1886, and within twenty years aluminium had collapsed from precious curiosity to the metal of cheap pots and mass-produced goods. By the time Tolkien was growing up in Birmingham in the 1890s, aluminium carried precisely the wrong cultural associations — industrial cheapness, not nobility. Polished aluminium can be bright, but it reads distinctly cooler and more bluish-grey than silver; no Victorian would have mistaken one for the other. More fatally for the argument, by the 1890s aluminium was the metal of mass production — the opposite of mithril’s prestige.

Platinum is far too dense. Silver tarnishes — that’s almost the entire point of Tolkien’s comparison. None of these candidates were present in Tolkien’s daily life as a child in any culturally significant way.

German silver

There is one material that was present — ubiquitously, inescapably present — in Tolkien’s Birmingham childhood, that the standard candidate list consistently overlooks. It is called German silver, and it contains no silver whatsoever.

German silver — also marketed as nickel silver, Argentan, and Nevada Silver — is a copper-zinc-nickel alloy developed commercially in the early 19th century. Its name refers to its appearance, not its composition. It was engineered to look like silver, behave better than silver in everyday use, and cost a fraction of silver’s price.

The key ingredient that made it possible — nickel — was once known as the devil’s copper. Medieval German miners in the Ore Mountains kept finding a reddish-brown ore that looked exactly like copper, yielded no copper when smelted, and made the men who worked it violently ill. They blamed a mischievous underground goblin — Nickel, a German diminutive of Nikolaus, related to the English Old Nick — and named the ore Kupfernickel: the devil’s copper. The curse was real. The miners were being slowly poisoned by arsenic trioxide released during smelting, and they had named an entire ore after their supernatural explanation for why it was killing them. It took until 1751 for a Swedish chemist named Axel Cronstedt to isolate nickel as a distinct element — inadvertently preserving the devil’s name in the periodic table, where it sits today as element 28. The same ingredient that poisoned medieval miners is what gives German silver its silver-white colour, its workability, and — most importantly for this argument — its permanent resistance to tarnishing.

The properties that made German silver commercially revolutionary were exactly three: it had the visual appearance of silver, it could be worked and polished like silver, and — the property that drove the entire market — it did not tarnish.

That last point deserves emphasis because it was genuinely remarkable. Sterling silver tarnishes primarily through reaction with sulphur compounds in the air, forming black silver sulphide on the surface. German silver does not. Its nickel content makes the alloy surface electrochemically stable under ambient conditions, producing a self-protecting finish that holds its appearance indefinitely without polishing. Victorian trade literature hammered this point relentlessly because it was the commercial proposition: the appearance of silver, permanently. Hiorns’ 1890 metallurgical survey describes the alloy as valuable for “its white colour, lustre, hardness, tenacity, and power of resisting certain chemical influences.”

Now read Tolkien’s description again: the beauty of silver, could be polished like glass, did not tarnish or grow dim.

The visual-drama properties — the ones that would lodge in a child’s imagination, the ones that make a material feelspecial rather than merely perform well on an engineering test — match with a precision that is hard to dismiss as coincidence.

There is one further detail worth pausing on. “Mithril” is a Sindarin compound: mith meaning “grey” or “light grey,” and ril meaning “glitter” or “brilliance.” Tolkien named his fictional metal after its appearance — cool grey brilliance. This is a more accurate description of German silver’s optical character than it is of actual silver. Sterling silver has a warm, slightly yellow-white tone. German silver, because its nickel content shifts the alloy’s reflectance toward shorter wavelengths, reads distinctly cooler and greyer under most lighting conditions. The name Tolkien chose for his fictional metal describes the appearance of German silver more precisely than it describes the appearance of silver itself.

A history worth knowing

German silver did not begin in Germany, and it took Europe a thousand years of embarrassment to figure out how to make it.

The Chinese had been producing a white copper alloy called baitong — literally “white copper” — for centuries, reliably documented from around the 3rd century AD and possibly much earlier. It was a copper-nickel alloy, and Chinese craftsmen were smelting nickel-bearing ores and producing a lustrous, silver-white metal long before European metallurgists had any idea nickel existed as a distinct element. European traders encountered it through the East India trade routes and brought it back as a curiosity under the Cantonese romanisation paktong or packfong. They admired it, coveted it, bought it, and for centuries could not work out how to make it themselves. The alloy Europeans eventually called German silver was Chinese technology, independently rediscovered fifteen hundred years later.

The reason Europe couldn’t replicate it was nickel — the devil’s copper, as we’ve seen. Once Cronstedt isolated it in 1751 and the element was understood, German and French metallurgists began systematically developing copper-zinc-nickel alloys. The alloy was refined and commercialised through the 1820s and 1830s under a proliferation of names — Argentan, Neusilber (”new silver”), Maillechort, Alpacca, Nevada Silver, Victoria Silver, Alfenide, Electrum — each designed to evoke silver without legally claiming to be silver. The naming chaos was entirely deliberate. Every trade name was a piece of commercial camouflage, projecting nobility and preciousness onto an alloy whose whole point was that it was cheap. The marketing of German silver is one of the more brazen exercises in 19th-century commercial euphemism: an industry built on calling things by names that meant something they were not.

The name “German silver” itself followed the same logic, acknowledging the country of commercial development while the “silver” referred purely to appearance. By the time Birmingham took over as the world’s dominant production centre in the mid-19th century, the alloy was being sold under so many names that consumers had little idea what they were buying — only that it looked like silver, held its shine, and cost considerably less.

Birmingham in the 1890s

Here is why this matters beyond a properties checklist.

Birmingham in the 1890s was the world centre of German silver manufacture. The alloy had been commercially dominant in British silverware since the 1820s, and by Tolkien’s childhood it was the material substrate of Victorian household life: the base metal of domestic cutlery, decorative goods, plated ware, candlesticks, picture frames, and the hardware of daily existence. It was not a specialist or industrial material — it was what the spoons were made of. A boy growing up in Birmingham between 1896 and 1900 would have encountered German silver every day of his life without ever thinking about it consciously.

And that is exactly the kind of influence that shapes a writer’s imagination. Not a deliberate reference, not a conscious choice, but the ambient texture of the world — the look and feel of things that were simply there, absorbed without being catalogued. Tolkien was not designing an alloy when he described mithril. He was remembering, at some level he could not have articulated, what it felt like to live in a city where the most celebrated material was a silver-coloured metal that never lost its shine.

The pattern of transmutation

This argument does not stand alone. It fits within a pattern of local influence that Tolkien scholarship has long recognised, even where Tolkien himself rarely confirmed it directly.

The Shire is Sarehole — the small Worcestershire village where the Tolkien boys played from 1896 to 1900, which he described near the end of his life as “a kind of lost paradise.” The mill at Hobbiton is Sarehole Mill; Tolkien drew it with its distinctive chimney-tower in his own illustrations. The miller’s son (George Andrew) who chased them off the grounds and was nicknamed the “White Ogre” became Ted Sandyman. Moseley Bog, where they played, became the Old Forest. His aunt Jane’s farm in Worcestershire, literally called Bag End, gave Bilbo his address. The Black Country — the industrial region visible as a smouldering, smoke-stacked horizon from Sarehole’s fields — maps onto Mordor in vocabulary, atmosphere, and name (Mordor means “Black Land” in Sindarin). Perrott’s Folly and the Edgbaston Waterworks tower, two landmarks Tolkien passed daily on his route to school in Edgbaston, are associated with the Two Towers. The illuminated face of the Chamberlain clock tower at the University of Birmingham is linked to the Eye of Sauron.

Tolkien was careful, for the most part, not to confirm these correspondences. He objected to allegory and was protective of his mythology’s internal coherence. But the pattern is too consistent to be accidental, and Tolkien scholarship has long operated on the sensible premise that undocumented influence is still influence. He didn’t need to consciously encode Sarehole as the Shire — he needed only to have lived there intensely as a child.

The same logic applies to German silver. He didn’t need to think “I shall base mithril on this alloy.” He needed only to have grown up in the city where it was made, used it every day at the table, and carried somewhere in his imagination the concept of a silver-coloured metal that never tarnished and that everyone wanted.

The irony at the bottom of the Shire

There is one further connection worth noting, not as a causal claim but as a piece of historical irony that reflects how thoroughly Birmingham’s metalworking identity permeated even the places Tolkien loved most.

That chimney on Sarehole Mill — the architectural feature that made it visually distinctive, the one Tolkien drew faithfully into every illustration of Hobbiton, the emblem of his pastoral ideal — exists because Matthew Boulton put it there. Boulton, the founding figure of Birmingham’s industrial metalworking economy and business partner of James Watt, leased Sarehole Mill from 1756 to 1761 and converted its machinery for metal rolling to produce sheet metal for button manufacture. He later moved operations to the Soho Manufactory at Handsworth, which became the direct ancestor of the Birmingham firms that dominated German silver production throughout the 19th century.

Tolkien almost certainly never knew any of this — Boulton’s tenancy predated his childhood by 135 years, and he wasn’t encoding industrial history, he was encoding a place he loved. But the chimney he loved was a metalworker’s chimney. The “lost paradise” he set against the industrial horror of Birmingham was itself, in its own architecture, a product of Birmingham’s industrial tradition.

Tolkien constructed the Shire and Mordor as opposites. The physical record suggests they were always the same place, seen from different angles and different centuries.

Celebrimbor’s name

One quiet piece of supporting evidence, for those who follow Tolkien’s linguistics.

Celebrimbor — the Elven-smith of the Second Age, the master craftsman who forged the Rings of Power and is associated in the legendarium with mithril-working — bears a name Tolkien constructed in Sindarin to mean Silver Hand. The greatest metalsmith in Middle-earth, the maker of its most prized substance, is named from a root meaning silver.

Tolkien’s naming was never casual or accidental. Mithril itself translates as “grey glitter” — cool grey brilliance, not warm white. Celebrimbor translates as Silver Hand. The craftsman who worked the grey-glittering metal is named for silver. The entire linguistic cluster around mithril points toward silver as its conceptual origin, while the physical description points toward something that improves on silver. A material that looked like silver, worked like silver, but never behaved like silver — never tarnishing, never losing its grey glitter — is exactly what German silver was, and exactly what mithril became in the imagination of a writer who grew up surrounded by it.

What this argument is and isn’t

To be precise about the claim, because the Tolkien fandom deserves precision:

The Tolkien scholarship that takes this kind of argument seriously has good foundations. Tom Shippey, whose work on Tolkien’s sources remains the standard, has documented extensively how Tolkien transmuted real philological and cultural material into secondary-world invention — not as allegory, which Tolkien explicitly rejected, but as what he called applicability: the way real experience shapes imagination without dictating meaning. The literary source for mithril’s narrative function — impenetrable armour of miraculous lightness — is almost certainly the silken mail coat in the Norse Hervarar saga, a text Tolkien knew intimately. That source explains what mithril does in the story. What it cannot explain is why Tolkien described it the way he did: the specific cluster of visual and material properties that make mithril feel real rather than merely functional. That is where Birmingham comes in.

This is not an argument that mithril is German silver, or that Tolkien consciously modelled one on the other, or that the physical properties match on every criterion. Some do not — mithril’s extreme lightness and its hardness exceeding tempered steel go well beyond anything German silver achieves, and that’s as it should be. Fantasy takes real things and removes their limitations. That is the direction of travel from German silver to mithril: take the material that amazed people in Victorian Birmingham — silver appearance, permanent lustre, workable and beautiful — strip away the mundane constraints of actual metallurgy, and you have a legend.

The argument is that German silver was the most culturally prominent material in the city where Tolkien spent his formative years; that its defining properties — the ones that made it remarkable to non-specialists, to households, to children — are precisely the properties Tolkien chose to mythologise; that Tolkien is independently documented as a writer who encoded his Birmingham environment into his fiction with extraordinary fidelity; and that no other proposed candidate for mithril’s inspiration satisfies all three of these conditions simultaneously.

Tolkien never said. But the city he grew up in did.

The minerals behind the alloy

German silver is built from three elements, each with its own story in the earth:

Native copper (Cu) — occasionally found as pure metal in nature, the first metal humans ever worked. Its malleability is what Tolkien captures in “beaten like copper” — a property so characteristic it became the benchmark.

Smithsonite (ZnCO₃) — the principal carbonate ore of zinc, named for James Smithson, founder of the Smithsonian Institution. Zinc contributes brightness and corrosion resistance to the alloy.

Pentlandite [(Fe,Ni)₉S₈] / laterite — the primary nickel-bearing ores. Industrial-scale nickel extraction was one of the defining metallurgical achievements of the 19th century. It was nickel that gave German silver its silver-white colour, its hardness relative to precious metals, and — most importantly — its non-tarnishing character. Without industrial nickel, there is no German silver. Without German silver, there is no obvious real-world seed for the most celebrated fictional metal ever invented.

Photos:

The Hill Illustration: Boston Globe

German Silver Cutlery: Bonhams

Nickeline: Natural History Museum Of Los Angeles County

Chinese Bowl: Metropolitan Museum of Art

Birmingham: Wiki Commons

Sarehole Mill: Sarehole Museum

Minerals: Stan Celestian - Flickr

“Yes — and actually, let me correct myself.”

The guest had asked a simple question. Is there native iron on Earth? I’d answered reflexively, gestured toward the meteorite case, and then heard myself do it. Close enough passing as correct.

I stopped mid-sentence.

“That’s a meteorite — iron-nickel alloy. It came from space. Native iron, pure terrestrial iron that formed right here on Earth, is something else entirely. We have some, but it’s among the rarest materials in this building.”

He nodded and we moved on.

I went back to the drawer after the tour.

The label read: iron (native).

And I thought about an anvil.

The word nobody looked up

The Greek word for anvil is akmon.

It also means meteorite.

This is not a coincidence that archaeologists have footnoted and moved past. This is the entire story. The ancient Greeks — who thought carefully enough about the world to give us democracy, tragedy, and the concept of the atom — looked at the thing Hephaestus built everything on, the foundation of all his work, and reached for the same word they used for rocks that fell from the sky. The anvil and the meteorite were, to them, the same category of object. Unworkable. Primordial. Arriving already made.

Hephaestus didn’t mine his anvil. He didn’t smelt it or forge it. In the oldest versions of the myth, he found it — or it found him. Cast off Olympus by Zeus, he lands on Lemnos, rescued by the Sintians, a local tribe. He builds his first forge from the volcanic fires of Mount Mosychlus. In the version from the Iliad, Hera throws him into the sea at birth, and he spends nine years in an underwater cave with the sea-nymphs Thetis and Eurynome, making jewelry in secret, proving something to no one but himself.

In both versions, the anvil comes first. Before the lightning bolts. Before the armor of Achilles. Before Hermes’ winged sandals and Aphrodite’s golden girdle and the bronze giant Talos who guarded the island of Crete. Before any of it — the thing everything gets made on.

The akmon.

A rock from the sky.

Nine years of homework nobody assigned

We don’t know exactly what Hephaestus made in that underwater cave. Homer gives us the forms: brooches, spiral armbands, necklaces. Simple inventory, no elaboration. But any metalsmith reading that list knows what it actually describes.

Brooches require spring tension — a pin that holds under pressure without snapping. You learn that by failing. By annealing the metal, working it, feeling where it wants to go and where it refuses. Spiral armbands require understanding work hardening — what happens to metal’s crystal structure under repeated bending, why it stiffens, when it becomes brittle, when to stop and reheat and start again. Necklaces require consistent, repeatable joins at scale. Every link identical. Every join strong enough to hold but small enough to move.

Nine years of those problems. In the dark. With no commission, no deadline, no one watching.

When Thetis arrives at his forge in Book 18 of the Iliad — sweating, surrounded by his golden automaton women, building something nobody has asked for yet — he already knows every material he will ever need. She has come to beg him for something impossible: new armor for her son Achilles, who has given his own armor away and is about to re-enter a war he cannot survive without it. He stops immediately for her. She was there in the cave. She knew him before Olympus wanted anything from him.

The brooches were the akmon. The spiral armbands were the akmon. The careful, unwitnessed, unfunded work of understanding what metal actually is before you ask it to become something else.

He was not the only one doing it.

They found it before they could make it

The first iron objects humans ever made were not smelted. They couldn’t be. Smelting iron — reducing iron ore with carbon at high temperature to produce workable metal — requires sustained temperatures above 1200 degrees Celsius and a technical understanding of carbon chemistry that took humanity thousands of years to develop. The first iron objects were found, not made. Picked up. Recognized as something different.

They came from the sky.

The iron beads of Gerzeh, Egypt, date to approximately 3200 BCE — more than two thousand years before the Iron Age. They were hammered, not cast. Shaped by hand from raw material that arrived already metallic, already workable, requiring no smelting because space had already done that work over millions of years. Analysis confirmed what the chemistry suggested: high nickel content, the isotopic signature of extraterrestrial origin. Someone picked up a meteorite, recognized it as metal, and made something from it.

The nickel is the diagnostic. Terrestrial iron ore, even when smelted at the highest temperatures achievable in antiquity, cannot produce the nickel concentrations found in meteoritic kamacite — the chemistry simply doesn’t allow it. High nickel content is a fingerprint that points only one direction: space.

Tutankhamun’s iron dagger, buried with him around 1323 BCE, is the same story. The blade is meteoritic — nickel content far too high for any terrestrial smelting process of that era. The Egyptians called meteoritic iron bja n pt — iron from the sky. They knew it was different. They knew it came from somewhere else. They didn’t have the crystallography to explain why, but they had the observation. They described it accurately before they could explain it.

That description was everything.

One degree per million years

Cut a meteoritic iron-nickel meteorite. Polish the surface. Wash it with dilute nitric acid.

What appears is not a texture or a pattern in the conventional sense. It is a crystal structure — interlocking bands of two distinct iron-nickel alloys, kamacite and taenite, grown together at the atomic scale over millions of years of cooling in space. The bands intersect at angles determined by the geometry of the parent crystal system. No two meteorites produce identical patterns. Each one is a fingerprint of its thermal history — how large the parent body was, how deep inside it the metal formed, how slowly it cooled as the body broke apart and drifted through space.

The cooling rate required to produce this structure is approximately one degree Celsius per million years.

You cannot replicate it. You cannot accelerate it. You cannot forge it into existence in any workshop on Earth, divine or otherwise. The Widmanstätten pattern — pronounced VID-man-shtet-en — was first observed by the English scientist William Thomson in 1804, and described systematically four years later by Count Alois von Beckh Widmanstätten in Vienna. One name stuck. The other didn’t.

Hephaestus didn’t make the akmon. He recognized it.

That is a different skill entirely. And it is the skill that made everything else possible.

The forge and the jewel in the same rock

There is a type of meteorite that stops people cold in the collection.

Not because it looks dangerous, or ancient, or alien — though it is all three. It stops them because it looks designed. Like someone made it deliberately, as an argument.

A pallasite is a stony-iron meteorite — part of an uncommon class that forms at the boundary between the metallic core and the silicate mantle of a proto-planet that no longer exists. That boundary is not just a physical interface — it is a chemical one, where two compositionally distinct worlds meet and interact. We study that same boundary in our own planet today, trying to understand what happens at the core-mantle interface thousands of kilometers beneath our feet. When that ancient proto-planet broke apart, catastrophically, billions of years ago, its interior was exposed. What you hold in your hand is a chemical record of a planetary interior — frozen at the moment of destruction and preserved in space for billions of years before falling to Earth.

Inside the iron-nickel matrix, suspended like insects in amber, are crystals of olivine — the gemstone peridot.

The forge material and the jewelry material. In the same object. Formed together at the boundary between core and mantle, by processes that had nothing to do with craft and everything to do with chemistry.

Hephaestus would have had to describe it before he could decide what to do with it.

The rarest iron on Earth looks nothing like you’d expect

Now open a different drawer entirely. Not the meteorite case — something closer, in a different cabinet. Back in the building. Back in Los Angeles.

The native metals drawer. Same label as before: iron (native). Different specimen entirely.

Disko Island, Greenland. A chunk of dark, rough, almost scorched-looking material — nothing like the metallic sheen of a meteorite slice, nothing like the structured elegance of a pallasite. It looks like something went wrong. Which, in a sense, is exactly what happened.

Disko Island is one of the only places on Earth where native iron — pure, terrestrial, formed right here on this planet without any contribution from space — occurs in significant quantities. The mechanism is specific and violent: basaltic magma, forcing itself upward through the crust, intrudes into coal-bearing sedimentary rock. The carbon in the coal pulls the oxygen away from the iron oxides in the basalt — a reduction reaction, driven by heat and pressure, that strips the iron down to its elemental form. No smelting. No human intervention. Just geology doing accidentally, under extreme conditions, what took humanity thousands of years to learn to do deliberately.

The result is iron. Pure, metallic, terrestrial iron. But it looks nothing like what anyone expects iron to look like. It is dark and rough and embedded in its host rock, without the Widmanstätten pattern, without the nickel content, without any of the signatures that would tell you it came from space. It came from here. From carbon and heat and a basaltic intrusion into coal-bearing rock at the wrong depth at the wrong moment.

It is extraordinarily rare. We have some. Not much.

And it sits in a completely different drawer than the meteorite that built civilizations — different classification, different community of scientists, different set of questions being asked of the same two words.

Iron (native). Two words. Three completely different formation stories. Three different drawers. One label that flattens all of them and moves on.

This is what descriptive science exists to prevent.

How iron finally won

Somewhere around 1200 BCE, everything changes.

Not because humanity discovered iron — they had been working meteoritic iron for two thousand years by then. Not because iron suddenly became available — it had always been there, locked in oxide minerals in virtually every rock formation on Earth. What changed is that someone, somewhere, figured out how to take it back out.

Smelting iron requires three things: ore, fuel, and the right temperature sustained long enough for the chemistry to work. Iron oxide plus carbon plus heat yields iron plus carbon dioxide — a reduction reaction not unlike what happens accidentally at Disko Island, but controlled. Deliberate. Repeatable. The earliest bloomery furnaces — simple clay structures packed with alternating layers of iron ore and charcoal, fed by bellows — could barely reach the temperatures required. The product was a spongy, inconsistent mass of iron mixed with slag, called a bloom, that had to be hammered repeatedly while hot to drive out the impurities.

It was brutal, slow, and transformational in the most literal sense. The Iron Age didn’t begin because iron was better than bronze in any simple way — early smelted iron was actually inferior to good bronze in several respects. It began because iron ore is everywhere and tin, the critical ingredient in bronze, is not. The person who could smelt iron didn’t need a trade network. They needed a hillside and a fire.

But the smiths kept working. Kept observing. Kept noticing what happened when you reheated the bloom and hammered it again, and again, and again — driving out the slag, aligning the grain structure, densifying the metal. Wrought iron, worked this way, became tougher and more reliable than the early blooms suggested was possible. Not as hard as good bronze, but far more available, far more workable, and capable of being produced anywhere there was ore and fuel. Over generations the techniques improved. The furnaces got hotter. The bellows got better. The smiths learned, empirically, what the metal wanted.

Iron didn’t win because it started better. It won because the people describing its behavior — feeling it, watching it, learning its limits — kept getting more precise.

And then someone noticed that some iron, under some conditions, came out harder than anything bronze could produce.

That was steel.

Why steel sounds like a mineral

Steel is iron with carbon incorporated into its crystal structure at precisely the right concentration — between 0.2 and 2.1 percent. Too little and you have soft iron. Too much and you have brittle cast iron that shatters under impact. The window is narrow, and for most of human history nobody knew it existed as a chemical phenomenon. They knew it experientially — that some iron, worked in some ways under some conditions, came out harder and sharper and more resilient than other iron. They developed techniques — carburization, quenching, tempering — that manipulated the carbon content without understanding why they worked.

What they were doing, without knowing it, was controlling a phase transition at the atomic scale.

When iron is heated above 912 degrees Celsius — the precise temperature at which its crystal structure reorganizes from one atomic geometry to another, a transformation as exact as a phase boundary can be — it shifts from a body-centered cubic structure called ferrite to a face-centered cubic structure called austenite. Austenite can absorb carbon in a way that ferrite cannot. When you quench austenite rapidly — plunge it into water or oil — the carbon gets trapped in the structure, forming martensite: a highly strained, extremely hard phase that gives steel its cutting edge. When you temper it — reheat it carefully to a lower temperature — you relieve some of that strain, trading a degree of hardness for toughness.

The Japanese sword makers who developed tamahagane — the folded steel of the katana — were manipulating martensite formation through empirical observation across generations, without the vocabulary of crystallography. The Damascus steel bladesmiths of the medieval Islamic world were doing the same thing with a different technique, producing a material whose microstructure — the precise role of carbide banding, wootz chemistry, and forging protocol in producing its characteristic properties — remains a subject of active investigation even now, centuries after the tradition itself was lost.

They were doing metallurgy. They called it craft.

But here is something worth pausing on. The names we use for these phases — ferrite, austenite, martensite — sound like mineral names because they were built by people who thought like mineralogists.

The -ite suffix is not accidental. It is the standard suffix of mineral nomenclature — siderite, ferberite, malachite, halloysite — a naming convention developed by systematic mineral describers who needed precise, stable, unambiguous language for what they were observing. When metallurgists in the late 19th and early 20th centuries turned their microscopes on steel and saw distinct phases with distinct crystal structures and distinct behaviors, they reached for the same framework. They described what they saw. They named it systematically. They built a classification structure that metallurgy still uses today.

Ferrite comes from ferrum — Latin for iron — the same root that runs through iron-bearing mineral species across the geological literature. Austenite is named after William Chandler Roberts-Austen, a British metallurgist working squarely in the tradition of careful materials description. Martensite is named after Adolf Martens, a German metallurgist whose instinct, when confronted with a new phase, was the same instinct that drives every new mineral species description: look carefully, describe precisely, name it so others can find it again.

The forge borrowed its vocabulary from the drawer.

Not metaphorically. Literally. The descriptive framework that makes modern materials science legible — the language that allows an engineer today to specify martensite content in a steel alloy, that allows a sword maker in feudal Japan and a materials scientist in a modern laboratory to be talking about the same thing across seven centuries — was built by people whose primary training was in the careful, systematic, patient description of what materials actually are before anyone asks what they can become.

The akmon again. Different century. Different laboratory. The same discipline doing the same invisible work.

It arrived already made

I put the Disko Island specimen back in the drawer.

It sits where it has always sat — dark, rough, unremarkable to anyone who doesn’t know what they’re looking at. Next to it, the Huntukunski Massif specimen from Siberia. Same label. Different story. A few drawers over, the meteorite collection. The pallasite. The Widmanstätten pattern frozen in metal that cooled at one degree per million years in space between worlds that no longer exists.

All of it waiting. All of it already made.

In collections and laboratories and field sites around the world, people are doing the work Hephaestus did in that cave — describing what is actually there before anyone asks what it can become. New mineral species. Crystal structures at the boundary between phases. The chemical record of a planetary interior preserved in a meteorite that fell into someone’s field. The careful, patient, precise work of getting it exactly right even when nobody needed the distinction.

There is a reason mineralogists say “described” rather than “discovered” when a new species is published. The mineral was already there. Already made. The scientist’s act is recognition — precise, careful, unrepeatable recognition of something that existed long before anyone had the language for it. That is what Hephaestus did. That is what the Gerzeh beadmakers did. That is what every person in this post did before anyone asked them for a lightning bolt.

The akmon doesn’t have a birth story because nobody thought to write one. It arrived already made, and the god who recognized it for what it was built everything else on top of it. That recognition — not the lightning bolt, not the armor of Achilles, not Aphrodite’s golden girdle — was the first and most necessary act of his entire existence.

I stopped mid-tour to correct myself because a meteorite and a native iron specimen are not the same thing, and I knew it, and the distinction mattered. That instinct — the one that won’t let close enough stand in for correct — is the same instinct. It is what Hephaestus did when he picked up the akmon. Not forging. Not transforming. Recognizing.

Every transformational thing that followed required it.

Every transformational thing that follows still does.

I closed the drawer.

Aaron Celestian is Curator of Mineral Sciences at the Natural History Museum of Los Angeles County, where he oversees one of the largest mineral collections in the western United States. He writes about minerals, science, and the things culture forgets to notice at Pocketful of Χtals on Substack.

I bought it for the color.

That’s the honest answer. I’ve been a mineralogist long enough to know better — to walk into a dealer’s room with a list, a budget, a set of analytical questions I’m trying to answer. I know that “pretty” is not a scientific criterion. I know it anyway, and I bought it anyway, because the specimen stopped me. It’s from Colombia, about the size of a dinner plate, and it is extraordinary in the way that only chaotic things can be: hundreds of prismatic quartz crystals radiating outward in every direction, sharp-terminated, a frozen explosion of silica arrested mid-burst. But that’s not what stopped me. What stopped me was the color at the tips. Each termination is stained a warm yellow-orange — not a surface deposit, not something applied after the fact, but halloysite included within the crystal itself during growth, captured as the quartz was forming, like an insect in amber but the other way around: the mineral swallowed by the crystal before either of them finished becoming what they are.

The color is the color of a mango in the last hour before you have to eat it. That particular yellow-orange at the edge of too ripe.

The specimen’s collector name is mango quartz. Its mineralogical name is halloysite-included quartz. But there is a third name — Corona Quartz — which comes from the Corona Mine in Boyacá, Colombia, the only place on Earth where this material has been found. The mine was rediscovered in 2017 after sitting lost in the jungle for nearly two decades, sealed by the man who found it. Specimens began reaching the market in volume around 2019 and 2020. I cannot find a single written record documenting the decision, but the timing is difficult to ignore: the name “Corona” became commercially complicated in early 2020, and somewhere in that same window the trade name shifted to mango quartz. It may be coincidence. I find that I don’t quite believe it is.

When I got it back to the lab I already knew what I was looking at. Halloysite. A clay mineral, common enough in weathered volcanic terrains, in soils above certain basalts, in the oxidized margins of hydrothermal systems. I’ve walked past halloysite specimens a hundred times in the collection. I’ve seen it on labels without stopping.

The mineral is named after Jean-Baptiste Julien d’Omalius d’Halloy — Belgian nobleman, statesman, pioneer of modern geology, born in Liège in 1783. He is the geologist who first defined and named the Cretaceous period, after the chalk strata in the Paris basin.

What I realized, holding the mango quartz under the light, is that I could not have told you what halloysite actually does. I knew the name. I knew the group. I did not know what it was for. What it had been for, across a history far longer than I’d considered.

So I started pulling the thread.

What I found at the other end surprised me enough that I’m still not sure how to say it simply. I’ll try: the same mineral sitting in my specimen case, the same tubular architecture that defines halloysite’s chemistry, has been recognized — independently, by organisms with no shared language and no common laboratory — as one of the most precisely useful structures in the biological world. The gorillas knew. The macaws knew. The oncologists are just catching up.

The Mineral

Halloysite belongs to the kaolin group — the same family as kaolinite, the clay in fine porcelain, in white paper, in the original formulation of antidiarrheal medications that sat in American medicine cabinets for most of the twentieth century. Chemically they are nearly identical: both are aluminum silicate hydroxides, both form in the same weathering environments, both leave the same faint peaks on an X-ray diffraction pattern. A non-specialist looking at the two side by side would struggle to tell them apart.

The difference is geometry.

Kaolinite forms flat, hexagonal plates — stacked sheets of linked silicon and aluminum tetrahedra, layered like pages in a book. Halloysite does something stranger. The same sheets, under the right conditions of hydration and surface charge mismatch between the silica and alumina layers, roll. They curl around themselves into hollow cylinders — tubes with an external diameter of roughly 50 to 70 nanometers and an internal diameter of 10 to 20 nanometers, ranging in length from a few hundred nanometers to about a micron and a half. To hold one in your hand, you would need to stack roughly a thousand of them to reach the width of a human hair.

What makes this architecture chemically interesting — what makes it, as it turns out, biologically interesting — is not just the shape but the charge. The inner surface of the tube, lined with aluminum hydroxyl groups, carries a slight positive charge at neutral pH. The outer surface, lined with siloxane groups, carries a slight negative charge. The tube is not simply hollow. It is a hollow cylinder with an inside that behaves differently from its outside. That asymmetry means it can selectively bind different molecules on different surfaces. It means things can be loaded into it, held, and released under the right conditions. It means that, at the nanoscale, halloysite is not just a container. It is a system.

The mango color at the tips of my specimen — the halloysite captured inside the quartz during growth — represents individual tubes too small to resolve without an electron microscope, present in sufficient concentration to tint the crystal like tea steeping in water. What I can see with my eye is the aggregate signal of a geometry I cannot see at all. The color is the evidence. The tubes are the argument.

The Ancient Pole

In the Virunga Mountains of Rwanda, mountain gorillas mine.

The behavior has been documented since Dian Fossey’s fieldwork in the 1960s. Gorillas in the dry season — when their diet shifts toward bamboo, lobelia, and senecio, plants higher in toxic secondary compounds than their usual forage — seek out specific outcrops of weathered volcanic rock. They loosen pieces with their teeth, grind the material to powder with their hands, and eat it. The clay fraction of what they consume has been analyzed. It is dominated by halloysite.

The timing is not coincidental. The shift to more toxic plant foods correlates with increased diarrhea — the body’s mechanism for clearing ingested compounds. The clay consumption follows. What the gorillas appear to be doing, with no laboratory and no periodic table, is adsorbing plant toxins onto the charged surfaces of halloysite nanotubes before those toxins can cross the gut wall. They are, in the functional sense, running a detoxification column in their digestive tract. The mineral is the column packing.

In the tropical forests of Peru, the evidence is more dramatic. On the eroding clay banks of Amazonian rivers, researchers have counted up to 900 parrots from 21 species gathering on a single outcrop — macaws, parakeets, and others arriving before dawn and working the clay face in dense, noisy crowds for hours. In 1999, a research team at UC Davis tested the hypothesis directly. They established that seeds eaten by macaws contain quinidine, a plant alkaloid toxic in sufficient doses. They then fed one group of captive Amazona parrots a mixture of the alkaloid plus clay from a preferred lick, and another group the alkaloid alone. The clay group showed significantly reduced alkaloid absorption — roughly 60% less quinidine in the bloodstream. The birds at the river bank were not eating randomly. The selection was chemically justified.

The clay mineralogy at macaw lick sites is dominated by smectite, with halloysite present in significant quantities. The birds appear capable of distinguishing between outcrops — returning consistently to specific sites even when other clay sources are available nearby. Whether the selection mechanism is taste, texture, or some more cryptic sensory signal is not yet established. That they are selecting is not in question.

These are not isolated cases. Geophagy — the deliberate consumption of earth, clay, or mineral-bearing soil — has been documented across virtually every class of terrestrial vertebrate. Elephants excavate mineral springs. Deer and ungulates seek salt licks with a precision that maps closely onto known mineral deficiencies in local forage. Chimpanzees in West Africa consume specific clay types during periods of intestinal distress. The behavior appears, conservatively, to be as old as the vertebrate gut.

In humans, the record is older than writing. Archaeological evidence of clay consumption appears at sites across sub-Saharan Africa, South America, and Asia. Hippocrates documented it. Avicenna described medicinal clay preparations in the eleventh century. In the American South, kaolin clay consumption — locally called “white dirt” — persisted through the twentieth century as a documented cultural practice, particularly among pregnant women and children. Across sub-Saharan Africa today, studies consistently report geophagy prevalence of between 30 and 65 percent among pregnant women, with consumption of between 100 and 400 grams of clay per day in some regions. The clay is sold in markets. It is mined from specific deposits. It is passed between women as knowledge, generation to generation.

The dominant clay mineral in geophagic soils consumed by pregnant women in Tanzania, Kenya, and Zanzibar has been characterized in multiple independent studies. The answer is consistent: kaolin minerals, frequently halloysite specifically, or a mixture of kaolinite and halloysite. Studies attempting to identify what these materials share — what property might explain why humans and animals converge on them — point repeatedly to the same suite of characteristics: fine particle size, chemical inertness relative to the gut lining, and high adsorption capacity for toxins, pathogens, and plant secondary compounds. The tubular geometry that makes halloysite a candidate for cancer drug delivery is the same geometry that makes it an effective gut-lining buffer. The mechanism is the same. The application is separated by a hundred million years of evolutionary trial and a few decades of materials science.

What the clinical literature calls pica — the consumption of non-food substances, classified as a disorder when it presents in a Western medical context — is, when examined mineralogically, a behavior with a chemically coherent rationale that spans the entire vertebrate lineage. The gorilla grinding clay in the dry season is not exhibiting aberrant behavior. It is running a protocol that works.

The Bridge

The mineral doing all of this is halloysite.

The same tubular geometry that a mountain gorilla selected empirically in the Virunga Mountains, that pregnant women across four continents have been mining from specific deposits and passing between generations as knowledge — that geometry is now the subject of several hundred peer-reviewed papers in materials science and nanomedicine. For most of the twentieth century, a version of this clay sat quietly in American medicine cabinets under a brand name that didn’t mention minerals at all. Nobody connected the dots. The discipline that owns the object wasn’t looking.

The Frontier

The first thing materials scientists discovered about halloysite was the same property the gorillas had been exploiting for centuries: the tube.

The hollow lumen of a halloysite nanotube — 10 to 20 nanometers in internal diameter, up to a micron and a half in length — turns out to be nearly ideal for loading and releasing chemical compounds in a controlled way. The inner aluminol surface holds a slight positive charge. The outer siloxane surface holds a slight negative charge. Those opposing charges mean the tube can be loaded with different molecules on different surfaces simultaneously — one compound inside the lumen, another bound to the exterior — and the two payloads can be engineered to release under different conditions. pH. Temperature. Enzymatic activity. The tube doesn’t just carry. It responds.

In cancer drug delivery, this matters enormously. One of the persistent problems in chemotherapy is that cytotoxic drugs don’t distinguish well between tumor tissue and healthy tissue — the drug goes everywhere, and the damage follows. Halloysite nanotubes offer a partial solution. Loaded with doxorubicin, one of the most widely used chemotherapy agents, halloysite nanotubes release their payload preferentially in acidic environments. Tumor tissue is acidic. The pH-triggered release means the drug concentrates where it’s needed and spares, to some degree, the tissue around it. Studies have demonstrated this effect in cervical adenocarcinoma cells, breast cancer cells, and leukemia cell lines. The nanotube has been decorated with folic acid to actively target cancer cells that overexpress the folate receptor. It has been conjugated with magnetic particles for guided delivery. It has been used to carry two different anticancer agents simultaneously — one inside the lumen, one on the exterior — with independent release profiles.

Halloysite nanotubes have also been shown to cross the blood-brain barrier, one of the most formidable obstacles in neuropharmacology. A sustained delivery system that can carry a drug payload across that barrier and release it gradually, without killing the cells it’s trying to treat, is not a minor achievement. The clay is doing this. The clay that forms in weathered volcanic rock, that a gorilla grinds with its hands in the dry season, that a pregnant woman in Zanzibar selects from a specific deposit and carries home.

The mechanism in every case is the same one Act 2 described: the tube, the charge asymmetry, the selective surface chemistry. What the researchers are engineering with functionalized polymers and pH-responsive bonds is a refined version of what the halloysite nanotube does naturally. The sophistication is real. But the architecture was already there, in the mineral, waiting to be described.

There are now several hundred peer-reviewed papers on halloysite nanotube biomedical applications. The literature is almost entirely in materials science, chemistry, and nanomedicine journals. Search “halloysite” in a mineralogy database and the results thin considerably. The object is the same. The disciplines looking at it are not.

The Gap

Here is a simple test.

Search “halloysite” in Web of Science. As of this writing, the database returns results across more than fifty subject categories. Chemistry leads with 3,193 papers. Materials Science follows with 2,870. Engineering contributes 1,291. Polymer Science adds 1,174. Combined, those four applied science categories account for roughly 9,500 papers about a mineral.

Mineralogy appears sixth on the list, with 843 papers. Geology adds 490. Together, the two disciplines that own the object — that named it, characterized its crystal structure, and maintain the reference collections — account for roughly 11% of all papers written about it. That figure is itself generous: the Earth Sciences bin includes structural crystallography, geological occurrence studies, and XRD characterization work — papers describing what halloysite is, not what it does. The literature engaging with halloysite’s biological interactions, its behavior in living systems, its potential in medicine, sits almost entirely outside those 1,333 papers. Notably, Oncology as a subject category returns just 2 papers — the cancer drug delivery research described in the previous section is being published and classified under Chemistry and Materials Science. The clinical literature hasn’t caught up yet, which means the gap may look different in a decade. But right now, the mineral is doing oncology work that oncology journals have barely begun to claim.

The geophagy literature barely registers. Nutrition and Dietetics: 20 papers. Zoology: 5. Veterinary Sciences: 6. The biological behavior that motivated the mineral’s selection across the entire vertebrate lineage, independently, across millions of years — the behavior that is, as this post has argued, the first evidence of halloysite’s functional chemistry — generates less than 0.3% of the total literature.

Run a parallel search on Google Trends. Over 22 years of US search history, “halloysite” has never exceeded a relative search index of 3 — its all-time maximum, reached twice, in October 2009 and February 2026. “Healing clay” peaked at 100 in January 2018 and has averaged 22 since 2020, while halloysite has averaged 0.5 over the same period. That is a ratio of roughly 45 to one. The public searching for the mineral’s biological properties — the gut chemistry, the adsorption behavior, the thing this post is about — is not finding mineralogy. It is finding wellness culture. The discipline ceded the vocabulary and never reclaimed it.

The audience exists. The animals worked out the mechanism. The oncologists are publishing the results. And as of late 2025, the U.S. Department of Defense has a fourth reason to care. In December 2025, a Utah-based company confirmed a major halloysite deposit in the Lake Mountains and announced it is deriving nano-silicon directly from the nanotube structure — the same hollow cylinder described in this post — to produce EV battery anodes that charge to 80% capacity in under five minutes, compared to 40 minutes or more for conventional graphite batteries. The tube is the reason that works. On February 2, 2026, the U.S. government announced Project Vault, a $10 billion strategic reserve for critical minerals, with halloysite-hosted deposits in the conversation. A preliminary economic assessment is still pending, and the commercial claims deserve scrutiny before a feasibility study validates them. But the geometry that a gorilla selected empirically in Rwanda is now the subject of a federal industrial policy announcement.

Mineralogy — the discipline that named this mineral, that characterized its structure, that maintains the reference collections — is present in the literature, but peripheral to all four conversations it should be leading. The oncologists, the materials scientists, the wellness community, and now the defense supply chain strategists have all arrived at halloysite independently. None of them came through mineralogy to get there.

The Close

The mango quartz is still on my desk.

I keep moving it to make room for other things and then moving it back. I tell myself it’s because it’s beautiful, which is true. But I think the real reason is that I’m not done looking at it yet. Not because I don’t know what it is — I know considerably more than I did when I bought it — but because knowing what it is has not made it smaller. It has made it stranger.

The color at the tips is still the color of a mango in the last hour before you have to eat it. The halloysite is still there inside the quartz, invisible at any scale I can access without an electron microscope, present only as aggregate signal — the warm yellow-orange that stopped me at the dealer’s table. The tubes are still there too, doing what they have always done: holding geometry that living systems have been finding useful for longer than our genus has existed.

A gorilla in the dry season found it useful. A macaw at a river bank found it useful. A pregnant woman in Zanzibar, selecting clay from a specific deposit and carrying it home, found it useful. A materials scientist loading doxorubicin into a nanotube for targeted delivery to a breast cancer cell is finding it useful right now, in a lab somewhere, having arrived at the same architecture by a completely different road.

The mineral didn’t change. It was always this.

I bought it for the color. I kept it for everything else.

Share

References

Fossey, D. (1983). Gorillas in the Mist. Houghton Mifflin.

Gilardi, J. D., Duffey, S. S., Munn, C. A., & Tell, L. A. (1999). “Biochemical functions of geophagy in parrots: Detoxification of dietary toxins and cytoprotective effects.” Journal of Chemical Ecology, 25(4), 897–922. https://doi.org/10.1023/A:1020857120217

Krishnamani, R., & Mahaney, W. C. (2000). “Geophagy among primates: Adaptive significance and ecological consequences.” Animal Behaviour.

Mahaney, W. C., et al. (1990). “Geophagia by mountain gorillas (Gorilla gorilla beringei) in the Virunga Mountains, Rwanda.” Primates.

Mahaney, W. C., et al. (1995). “Mountain gorilla geophagy: A possible seasonal behavior for dealing with the effects of dietary changes.” International Journal of Primatology.

Young, S. L., Wilson, M. J., Hillier, S., Delbos, E., Ali, S. M., & Stoltzfus, R. J. (2010). “Differences and commonalities in physical, chemical and mineralogical properties of Zanzibari geophagic soils.” Journal of Chemical Ecology, 36(1), 129–140. https://doi.org/10.1007/s10886-009-9729-y

Young, S. L., & Miller, J. D. (2019). “Medicine beneath your feet: A biocultural examination of the risks and benefits of geophagy.” Clays and Clay Minerals, 67(1), 81–90. https://doi.org/10.1007/s42860-018-0004-6

Gray-Wannell, N., Cubillas, P., Aslam, Z., Holliman, P. J., Greenwell, H. C., Brydson, R., Delbos, E., Strachan, L.-J., Fuller, M., & Hillier, S. (2023). "Morphological features of halloysite nanotubes as revealed by various microscopies." Clay Minerals, 58, 395–407. https://doi.org/10.1180/clm.2023.37 — SEM, AFM, and cross-sectional TEM study of nine halloysite samples; establishes mean lumen diameter of 12 nm and demonstrates that the standard "carpet roll" model is an oversimplification. Open access.

Davies, T. C. (2023). “Current status of research and gaps in knowledge of geophagic practices in Africa.” Frontiers in Nutrition, 9. https://doi.org/10.3389/fnut.2022.1084589

Macaw photo: https://commons.wikimedia.org/wiki/File:Parrots_at_a_clay_lick_-Tambopata_National_Reserve,_Peru-8c.jpg

There is a specimen of apatite in my office at the Natural History Museum of Los Angeles County that I have been thinking about for over a year. It sits in a drawer with six others — same mineral, different histories — and I keep coming back to it because it is, in a precise and literal sense, the same mineral that is in your bones right now. Same formula. Different record.

I spent a week at Caltech in January 2024 arguing about what life is. Not philosophically. Practically. There were a little over thirty of us, funded by the Keck Institute for Space Studies, trying to answer a question that sounds simple until you actually try to answer it: how do you build an instrument to find life on another planet when you’re not entirely sure what life looks like?

I was the only mineralogist in the room. Nobody thought that was strange. I did.

The workshop was built around a question that has haunted planetary science for fifty years. In 1976, NASA landed two robotic spacecraft on the surface of Mars — the Viking landers, the most sophisticated machines humanity had ever sent to another planet. Their primary mission was not geology or weather or photography, though they did all of those things. Their primary mission was to find out whether anything was alive.

They ran the first — and still only — direct life-detection experiments ever conducted on another planet. The instruments worked flawlessly. The data came back clean. And then scientists spent the next fifty years arguing about what it meant.

That ambiguity is not a failure. It is the most honest scientific outcome possible. We built the best instruments we had, asked exactly the right questions, and the universe handed us a result we didn’t yet have the framework to interpret. Fifty years of better science followed directly from that unresolved result. That is what a good experiment does.

What Viking revealed, with extraordinary precision, was the shape of the problem. The Martian soil produced chemical signals that looked, in some ways, like the signals living things produce. And in other ways, didn’t. Was it life? Was it chemistry? We still don’t have a definitive answer.

I kept thinking about this in that room at Caltech. Not because of the biology. Because of the rocks.

I was in that argument. I helped shape what each category required. And the whole time, I kept drawing.

Not notes, exactly. More like a diagram I couldn’t stop refining — three overlapping circles on a iPad, with percentages at each intersection and a phrase scrawled underneath: Depth ≈ Time.

The people in that room were thinking about organisms. I was thinking about the archive those organisms leave behind. By the end of the week, I was certain those were the same problem — and that nobody in the room fully understood why.

Act 2: What Minerals Remember

Here is something most people don’t know about rocks: they remember.

Not the way organisms remember — not through neurons or behavior or instinct. But through structure. The way a crystal grows, the trace elements it incorporates, the defects locked into its lattice, the fluid trapped inside it when it formed — all of that is a record. A mineralogist reading a crystal is reading a document. The question is always: a document of what?

This is what I wrote under Metabolism on my diagram: Depth ≈ Time.

In biology, time runs through genealogy. You understand the past by tracing who descended from whom, how long the evolutionary clock has been running. In mineralogy, time is stratigraphy. Depth is time. The further down you drill into the earth, the further back you read. Each layer is a page. Each mineral in that layer is a sentence.

This is not a metaphor. It is how the geological record actually works. One of the concrete outcomes of that Caltech workshop was a mission concept built around exactly this idea: drill into the Martian subsurface, retrieve a vertical record of rock and salt and ancient brine, and read it layer by layer. When we drill a core sample from the Martian subsurface — something that mission concept is designed to do — we are not just collecting rock. We are collecting a physical archive of everything that happened in that place, at that depth, at that moment in planetary history. Temperature. Chemistry. Whether liquid water was present. Whether something was consuming that water and leaving waste behind.

A biologist looks at that core and asks: is there life here? A mineralogist looks at the same core and asks: what has this rock been through, what did it record, and can we trust what we’re seeing? Those are different questions. You need both.

Mineralogists have been part of Mars missions before — reading the mineralogy of the surface, identifying what the rocks are made of, characterizing the chemistry. That work is essential and it has been done well. But the specific question of life detection — designing the experiments, setting the burden of proof, deciding what counts as a biosignature and what doesn’t — has been largely a conversation between biologists, chemists, and engineers. The person trained to recognize when a mineral is mimicking life, or to know what a mineral record can and cannot preserve, has not always had a seat at that particular table. That Caltech workshop was one of the first times I felt the gap closing.

This matters because of what minerals actually do. Not just record. Preserve. Transform. And sometimes, deceive.

But the archive doesn't start at the moment life appears. The mineral record predates biology entirely — and the more you read it, the more you realize it didn't just witness life's beginning. It built the room.

Act 3: The Answer

Every serious scientific hypothesis about how life began on Earth involves minerals.

Not as background. Not as scenery. As the mechanism.

The leading hypothesis for life’s origin centers on hydrothermal vents on the ancient ocean floor — places where hot, mineral-rich water poured through cracks in the rock, creating chemical gradients across iron-sulfur mineral surfaces. Those gradients are thought to have driven the first proto-metabolic reactions. The chemistry of life didn’t happen in open water. It happened at the interface between water and rock.

Another hypothesis centers on clay mineral surfaces — thin, charged sheets of aluminosilicate that could have provided the ordered template that early RNA-like molecules needed to copy themselves. Without the clay, no copying. Without copying, no heredity. Without heredity, no evolution. Without evolution, no life.

A third hypothesis centers on evaporite minerals — salts that form when water evaporates. As ancient shorelines dried and rewet in cycles, organic molecules concentrated on evaporite surfaces, polymerized, and eventually crossed some threshold we don’t fully understand into something we would recognize as alive.

In all of these, minerals are not the stage. They are the actor. Life didn’t emerge despite the mineral world. It emerged through it.

And this relationship didn’t end at the origin. It continued. When organisms die, minerals preserve them. When microbial communities alter their environment, minerals record it — in isotopic signatures, in oxidation states, in the specific crystal habits that only form under biological influence. The entire fossil record is, at its core, a mineral record. Everything we know about the history of life on Earth we know because minerals remembered it.

This is why minerals are the answer to the life detection problem. The biosignatures we’re looking for on Mars — the chemical traces that would tell us something once lived there, or still does — will be preserved in minerals. Evaporites. Fluid inclusions. Clays. The archive is mineral. The evidence, if it exists, is mineral. The medium through which we must read the past is mineral.

Which brings me to the problem.

Act 4: The Obfuscation

Here is the part that took me years to say out loud: the same properties that make minerals the answer also make them the noise. A mineral that preserves biological chemistry does so because it forms stable structures from the same elements, under the same conditions, that life requires. That's not a coincidence. It means that every mineral capable of holding the evidence of life is also capable of producing something that looks exactly like it. The archive and the counterfeit are made of the same stuff.

In 1996, a team of NASA scientists announced that they had found evidence of ancient life in a Martian meteorite.

The meteorite was called ALH84001. It had been sitting in a collection in Antarctica for years, a chunk of Mars that had been blasted off the surface by an ancient impact and eventually fallen to Earth. When scientists looked closely at the rock, they found several things that, taken together, seemed to point to biology: unusual carbonate minerals, organic compounds, and — most controversially — tiny structures that looked, under an electron microscope, like fossilized bacteria.

The announcement made the front page of every newspaper on Earth. President Clinton made a statement. NASA held a press conference. For a few weeks, it felt like the question had been answered.

It hadn’t. Over the following years, researchers showed that each piece of evidence could be explained by chemistry rather than biology. The carbonate minerals could form abiotically at high temperatures. The organic compounds were likely contamination from Earth. The tiny structures were almost certainly mineral artifacts — features that form naturally in rock and happen to resemble cells at nanometer scales.

ALH84001 is the most famous example of the problem, but it is not an isolated one. Minerals routinely produce structures, patterns, and chemical signatures that look biological and aren’t. Dendrites — the branching crystal formations you sometimes see in rock — look uncannily like fossil plants or coral. Certain iron oxide minerals form rounded, layered structures called spherulites that resemble microbial colonies. Silica deposits around hydrothermal vents create stalks and filaments that, without careful analysis, are indistinguishable from fossilized microbes. In the early Earth rock record, distinguishing genuine biosignatures from mineral mimics is one of the hardest problems in geology. On Mars, where we cannot bring the rocks back to a full laboratory for analysis, it may be the hardest problem we face.

Here is the duality, stated plainly: minerals preserve the evidence of life, and minerals produce false evidence of life. The same medium that holds the answer also generates the noise. Every signal we’re looking for on Mars will be embedded in — and potentially mimicked by — the mineral matrix that surrounds it.

This is not an argument against looking. It is an argument for understanding what you’re looking at. And understanding what you’re looking at, in a mineral, requires a mineralogist.

The frameworks the astrobiology community has built for life detection — careful, rigorous, genuinely impressive work — acknowledge this problem. The best current framework, developed in part from the work done at that Caltech workshop, explicitly targets evaporitic minerals, fluid inclusions, and crystal habits as the primary sample targets for Mars life detection. Mineralogical characterization is built into the baseline measurements before any biological experiment begins.

This is not a fringe concern. Charles Cockell, one of the scientists who co-led the Caltech workshop, has argued publicly that “life” is ultimately a human definition — that the difficulty of defining it precisely is what makes finding it on other planets so hard. The problem isn’t the instruments. It’s the definition the instruments are built to detect.

But building mineralogy into the framework is not the same as having mineralogists in the room when the framework is designed. One produces a checklist. The other produces judgment.

Act 5: The Diagram

Somewhere in the first week of that Caltech workshop, while the group was working through what would become their life detection framework, I drew a diagram.

Three overlapping circles: Patterns of Chemical Structures, Morphology, and Metabolism. Each circle alone gets you to 95% confidence that you’re looking at something biological. Any two overlapping: 99.7%. All three converging on the same sample: 100%.

The probability logic was simple. What stopped a few people mid-conversation when they saw it was the phrase underneath: Depth ≈ Time.

The biologists and chemists in that room were designing experiments for a single moment — you drill, you sample, you measure. I was thinking about the vertical record. What the rock looked like at ten meters versus fifty meters versus a hundred meters. How the mineralogy would shift with depth, how the chemistry would change, what that gradient would tell you about the history of the environment. In stratigraphy, depth is not just space. It is time made physical.

But there is another timescale problem that the mineral record forces you to confront. Mars is cold. Most of the time, extraordinarily cold. Life on Earth, when it exists in cold and extreme environments, doesn’t stop — it slows down. Metabolic rates drop to near zero. Cellular processes that take minutes at room temperature can take years, decades, or centuries at Martian temperatures. A single measurement, even a sophisticated one, taken over the course of an hour or a day, may show nothing — not because there is nothing there, but because whatever is there is moving on a timescale the instrument wasn’t designed to detect. The mineral record doesn’t have this problem. A crystal that took ten thousand years to grow still holds that growth in its structure, readable long after the process ended. Biology leaves traces in minerals precisely because minerals are patient in a way that instruments are not.

The mineralogist’s question is never just is there life here. It is what has this place been through, and does the mineral record support or complicate what the biological instruments are telling us. Those are different questions. The second one requires someone trained to read rock.

I showed the diagram around. The conversation shifted.

The second week of the workshop opened with a new banner image — the official Keck promotional image for Part II of the workshop. Earth on the left. Mars on the right. And between them, bridging the two planets: photographs of mineral samples. The life detection icons traveled in both directions along that mineral bridge.

The workshop’s own banner had made the argument. I had an iPad that said the same thing in three overlapping circles and a phrase about depth.

My diagram didn’t become the final framework verbatim. But the framework that emerged — and the mission design built from it — centers evaporitic minerals, fluid inclusions, and crystal habits as the primary targets. The mineralogist’s contribution was in the architecture, not just the instrument list.

That is what it looks like when a mineralogist is in the room. Not a checklist item. A different way of reading the same evidence.

Act 6: The Specimen Drawer

Back in Los Angeles, in my office at the Natural History Museum, there is a drawer of apatite specimens that I keep coming back to.

Same mineral in every box. Same formula: Ca₅(PO₄)₃(OH,F,Cl). Same crystal system. Completely different histories.

One formed on an ancient seafloor in Germany, built up from phosphate-rich sediment over millions of years. One crystallized deep inside a cooling magma body in Brazil. One grew from hot fluids moving through fractured rock in Portugal — a collector’s holy grail. One formed where magma met limestone in Quebec, the chemistry of both transforming into something neither would have produced alone. One grew alongside emeralds in the Andes. One is a geological hybrid from Utah, carrying two different formation stories in a single crystal.

And then there is the one from Durango, Mexico. Arguably the most famous apatite in the world — not because it is the rarest or the most beautiful, but because it is the most chemically consistent. Laboratories on every continent use Durango apatite as the global calibration standard for dating rocks. When scientists need to verify that their instruments are reading correctly, they run Durango apatite first. It is the zero point. The reference against which everything else is measured.

Every one of these formed through a different process, in a different environment. Every one of them is the same mineral. And every one of them is a record — of the conditions that made it, the fluids it grew from, the geological events it survived.

That is what minerals do. They remember.

Now consider what a life detection instrument would see if it analyzed these specimens without knowing where they came from. The German specimen has rounded, layered growth that looks almost biological — the kind of morphology that would stop a Mars instrument in its tracks. Several of the others carry trace element chemistries that overlap with biologically influenced mineralization. Some formed in exactly the kinds of iron-rich hydrothermal environments that Mars orbiters have flagged as biosignature targets.

None of this is evidence of life. Not one of these specimens has anything to do with biology. But a detector on Mars, encountering chemistry and morphology like this in the subsurface, would flag every one of them for follow-up. The archive and the noise look identical until someone who knows the difference is in the room.

This is not a hypothetical concern. It is the central analytical problem of the next fifty years of planetary exploration. The Durango apatite sits in my drawer as a calibration standard — the reference point scientists trust before they trust anything else. What we need for life detection on Mars is something equivalent. Not just better instruments. A better understanding of what the mineral record can and cannot tell us, built by people who have spent their careers learning to read it.

The crystals hold the answer and the noise. Learning to tell the difference — that’s the work.

But here is the question I keep coming back to — the one I’ll be sitting with on that panel stage in a few weeks. We set out to find life on Mars. We built the instruments, designed the experiments, and sent them across the solar system. And what the search has given back so far isn’t an answer. It’s a mirror. Every framework we’ve built to detect life elsewhere has forced us to confront how poorly we understand life here. What do we actually mean by it? Where does it begin? How do we recognize it when it doesn’t look like us?

My co-panelist has spent her career reconstructing ancient genes — reading the evolutionary record backward to understand what life looked like at its earliest moments on Earth. I’ve spent mine reading minerals — the physical archive that recorded those same moments from the other side of the biology. Two different archives. The same question. What does the search for life beyond Earth reveal about life here at home?

I don’t think that question has a clean answer yet. But I think we’re finally asking it in the right room.

References & Further Reading

The KISS Workshop and REVEAL Mission Perl, S.M., Cockell, C.S., Fischer, W.W., and colleagues. “Biological Validation and Agnostic Experiments for Extinct and Extant Microbial Life within the Martian Subsurface.” Astrobiology, Viking Special Issue (in revision, 2026). [Aaron Celestian is a co-author.]

Perl, S.M., Cockell, C.S., Fischer, W.W., and colleagues. “The Biology of Biosignature Detection: Rationale and Experimental Frameworks for Biological Validation.” Report prepared for the W.M. Keck Institute for Space Studies, California Institute of Technology, 2024. DOI: 10.26206/57j7-pk96 [Aaron Celestian is a co-author.]

The Ladder of Life Detection Neveu, M., Hays, L.E., Voytek, M.A., New, M.H., and Schulte, M.D. “The Ladder of Life Detection.” Astrobiology18(11), 1375–1402, 2018.

Mineral Evolution Hazen, R.M., and colleagues. Mineral evolution framework — the progression from ~60 minerals at Earth’s formation to 5,900+ today, driven in large part by biological processes. See Hazen et al., American Mineralogist, ongoing series.

ALH84001 and the Life on Mars Debate McKay, D.S., Gibson, E.K., Thomas-Keprta, K.L., and colleagues. “Search for Past Life on Mars: Possible Relic Biogenic Activity in Martian Meteorite ALH84001.” Science 273(5277), 924–930, 1996.

What is Life? (Video) Cockell, C.S. “Are Viruses Alive?” Astrobiology lecture series, University of Edinburgh, 2020.

Viking Missions Klein, H.P. “The Viking Biological Experiments on Mars.” Icarus 34(3), 666–674, 1978.

Biosignature Preservation in Evaporites Perl, S.M., Celestian, A.J., Seuylemezian, A., Tasoff, P., Baxter, B.K., Vaishampayan, P.A., and Corsetti, F.A. “Evaporitic Preservation of Modern Carotenoid Biomarkers and Halophilic Life in Martian Analogue Hypersaline Environments.” Astrobiology, 2025. https://doi.org/10.1177/15311074251392173

In the summer of 2023, a colleague introduced me by email to an artist named Melonie Ancheta. The introduction was two sentences. Melonie had been researching a particular mineral for twenty-five years. She thought we should talk.

I wrote back the same day.

Melonie is a Northwest Coast Native pigment specialist — her work centers on the traditional paints of the Haida and Tlingit peoples, what those paints were actually made of, and what happens to them over time on the objects they were applied to centuries ago. The mineral connecting us was vivianite. She was trying to stop it from changing color on Tlingit ceremonial objects held in museum collections across North America. I was trying to stop it from changing color in our gem and mineral hall. We had arrived at the same problem from completely different directions, separated by geology on one side and art history on the other, and neither of us had solved it.

Nobody has.

That’s what happens with vivianite. Everyone who gets close to it eventually ends up with the same unanswered question. And the more you understand about what this mineral actually is — where it comes from, how it forms, what it’s made of — the less the unanswered question feels like a failure, and the more it starts to feel like the point.

The Specimen in the Dark

We have a vivianite specimen at NHMLAC that I cannot put on display.

Vivianite is an iron phosphate mineral — one of the most visually arresting in any collection, and almost entirely unknown outside mineralogy. When you encounter it, the blue stops you.

If you’ve seen the Instagram post, you know what it looks like: a stack of dark metallic plates peeling apart from each other, with flashes of deep blue-green caught in the cracks between them, like someone has been forcing a book open from the spine and light got in between the pages. The outer surfaces are nearly black. The interior surfaces — the ones only recently exposed — are the color of a deep ocean in a painting you can’t quite believe is real.

The specimen has been doing this for decades, probably longer. It was doing it before I became curator. It will keep doing it after I’m gone. We keep it in the dark because light accelerates the process, and in the dark the process merely continues rather than races. The museum holds things in perpetuity — that is the entire institutional commitment — and so we hold this one in a drawer, and visit it occasionally, and note that it has progressed.

What is it doing? It is, in the most literal chemical sense, becoming something else.

Fe²⁺ and the Eight Waters

Vivianite’s formula is Fe₃(PO₄)₂·8H₂O. Iron, phosphate, and eight water molecules held together in a structure so water-rich it feels almost biological — which turns out to be entirely appropriate.

In a fresh crystal, all three iron atoms are in their reduced state: Fe²⁺, ferrous iron. This is iron that has held onto its electrons. It is the same oxidation state as the iron in your hemoglobin right now, carrying oxygen through your blood. In this state, the crystal is completely colorless. Not pale. Not faintly blue. Colorless — transparent as glass, with no color to speak of.

Then light arrives.

A visible photon strikes the crystal and knocks a proton loose from one of those eight water molecules. The displaced charge has to go somewhere, so a nearby Fe²⁺ gives up an electron and becomes Fe³⁺ — oxidized iron, iron that has lost something. Now two iron atoms sit side by side in the same lattice: one with the electron, one without. And the electron begins to move.

It hops. Back and forth, Fe²⁺ to Fe³⁺, Fe³⁺ to Fe²⁺, driven by the energy of incoming light. Physicists call this intervalence charge transfer, but what it looks like from the outside is blue — because the hopping absorbs the orange and red end of the spectrum and reflects what remains. The blue you see in vivianite is not a stable property of the mineral. It is the color of instability itself. It is what electron theft looks like, made visible.

And it cannot hold. The oxidation continues. More Fe²⁺ becomes Fe³⁺. The mixed-valence sweet spot that produces blue gradually disappears — there is less and less Fe²⁺ left to complete the pair, less hopping, less blue. The color deepens toward dark green, then black, then — when the iron is fully oxidized, every electron surrendered — it flips entirely. Fully oxidized iron phosphate absorbs blue and reflects yellow-brown. The end product is likely santabarbaraite: Fe³⁺₃(PO₄)₂(OH)₃·5H₂O. Same phosphate backbone, same basic architecture, every electron gone.

The full sequence: colorless, then blue, then dark, then yellow-brown. The same atoms the entire time, the same crystal shape, just electrons redistributed — a completely different mineral at each stage, wearing the same external form. The blue was never the destination. It was always the middle.

Somewhere in that middle, the formula is Fe²⁺Fe³⁺₂(PO₄)₂(OH)₂·6H₂O: metavivianite. A ghost of the original structure.

The specimen in our drawer is somewhere in this sequence. The black outer surfaces are late-stage. The blue-green caught in the cracks is mid-stage, exposed recently enough that the hopping still has somewhere to go. The crystal is not one thing — it is a gradient, a reaction front moving inward from every surface that has ever seen light. In the most honest description available to mineralogy, it is a record of every photon that has reached it.

This is why I cannot put it on display. The lights would finish what has already started.

The body does not stop being geochemistry when it stops being alive.

What Makes It, and Where

To understand why vivianite keeps appearing everywhere — in geological specimens, in burial sites, in the mud under peat bogs, in the sediment of alpine lakes — you need to understand what it requires. Three things: iron, phosphate, and very little oxygen. Reducing conditions, in the chemical vocabulary. That combination sounds specific. It isn’t.

Peat bogs. Lake sediments. Waterlogged soils. The reducing mud at the bottom of a marsh. The interior of a decomposing organism. All of these qualify. The organic matter provides phosphate as it dissolves. If iron is nearby — in the surrounding sediment, in buried metal objects, in the rocks of a glacier — the two meet in solution and nucleate. Crystals begin to grow. Colorless at first. Then the light finds them.

The places vivianite forms most readily are also, not coincidentally, places humans have been depositing their dead for thousands of years.

In 1996, a torso was recovered from a bay of Lake Brienz in Switzerland, exposed by an underwater landslide triggered by an earthquake. The body — nicknamed Brienzi by the forensic team, in the tradition of Ötzi — had been in the water since approximately the 1770s. Parts of it were blue. Vivianite had grown on the preserved adipose tissue, fed by iron from the lake sediment and phosphate from the dissolving bone. The forensic investigators used the mineral’s presence to reconstruct the burial environment and estimate the time since death. The vivianite had been keeping records the entire time, in the dark, at the bottom of the lake, waiting for an earthquake to bring it into the light.

Ötzi the Iceman — 5,300 years old, recovered from the Ötztaler Alps — carries blue spots on his skin where he was in contact with iron-bearing rocks. The glacier provided the iron. His own body provided the phosphate. The chemistry did not require anyone’s intent or awareness.

And US airmen missing in action in Vietnam, recovered decades after a 1963 crash, were found with blue encrustations on their skeletal remains. The iron came from the corroding aircraft. The vivianite told investigators the men had been buried in waterlogged soil — information recoverable no other way. A mineral that forms in the space between iron and phosphate and darkness had written a forensic record that outlasted every other witness.

Then there is John White, a railway engineer who died in 1861 and was buried in a cast iron coffin fitted with a small glass window — a Victorian custom, so mourners could see the face of the deceased when the lid was closed. At some point after burial the glass broke. Groundwater seeped in, met the iron of the coffin and the phosphate of the dissolving body, and over the following century assembled itself into vivianite crystals. When the coffin was exhumed during an archaeological rescue excavation more than a hundred years later, it was full of blue. John White had become, in part, a mineral. Nobody planned it. The chemistry simply proceeded.

Vivianite is not a rare mineral. It forms wherever the conditions align. And the conditions align, with remarkable frequency, wherever something organic has ended in the presence of iron and water.

What Your Body Is

Your body contains approximately 4–5 grams of iron. Most of it is in hemoglobin — the protein in red blood cells that carries oxygen — where the iron sits at the center of a ring-shaped molecule in its ferrous, Fe²⁺ form. Reduced iron. The same oxidation state as colorless, fresh vivianite.

Your body also contains roughly 700 grams of phosphorus, most of it locked into the mineral structure of your bones and teeth as hydroxyapatite: Ca₁₀(PO₄)₆(OH)₂. The phosphate groups in that formula are the same phosphate groups in vivianite’s formula. Right now they are doing a different job. They are making your skeleton rigid.

When decomposition begins, hemoglobin breaks down and releases its iron into the surrounding environment. Bone apatite dissolves and releases its phosphate. In a waterlogged, reducing burial environment — low oxygen, iron present — those two populations meet in solution for the first time. And they precipitate.

You are carrying the precursor minerals right now. The only thing preventing their reorganization into blue-green crystals is you.

The Painters and the Bog

In the 12th century, a painter decorating the Holy Sepulchre Chapel at Winchester Cathedral reached for blue. Not lapis lazuli — too expensive, too rare, too far from Hampshire. Instead, vivianite: the earthy blue-green mineral found in the peat and alluvial deposits of the local countryside, at Fordingbridge and the Isle of Wight, within traveling distance of the cathedral. The Getty Conservation Institute’s technical analysis of the chapel paintings established that the vair lining of Nicodemus’s cloak — the heraldic pattern of alternating pale blue and white squirrel fur — was rendered in vivianite. The conservators noted it was chosen not as an economical substitute but for its distinctive deep indigo quality, set against natural ultramarine for contrast. It was a deliberate aesthetic decision about a specific color.

They were not painting with a stable color. They were painting with a process.

The paintings at Winchester look green now. Nobody painted them green. Some of the vivianite particles have oxidized all the way through the blue-black stage and kept going — past metavivianite, into the fully oxidized iron phosphate territory where the color flips to yellow-brown. The yellow particles and the remaining blue particles, mixed together at the pigment grain scale, average out to green. The Winchester green is not a color anyone chose. It is the arithmetic of eight centuries of oxidation, caught partway through, different grains at different stages of the same sequence. The original intent is now recoverable only through laboratory analysis.

Four centuries later, in Dordrecht, Aelbert Cuyp was painting the Dutch countryside in warm golden light. The peat bogs around Dordrecht — the flat, waterlogged geography Cuyp knew as home — are exactly the kind of iron-phosphate-rich, anaerobic environment where vivianite forms as amorphous blue-green clay, easy to collect, easy to grind. Conservation scientists examining Cuyp’s paintings have identified vivianite in multiple works: in a milkmaid’s skirt in The Large Dort, in atmospheric green passages, in the cool distance of his landscapes. That blue-grey wash in a Cuyp painting is, in part, bog iron phosphate — the same chemistry, from the same kind of waterlogged organic ground, that produces vivianite in decomposing matter. Cuyp was painting the landscape partly with itself.

He was not the only one. Vivianite has been identified in works from the workshops of Rembrandt and Vermeer. It turns up in the grey-blue of a carpet in Vermeer’s The Procuress. The broader pattern concentrates heavily in 17th-century Dutch painting, almost certainly because the Dutch peat bog was local, free, and abundant. The great masters of Northern European light were painting partly in the chemistry of decomposition.

And the paintings have been changing ever since. The vivianite in The Large Dort has degraded. The blue in the milkmaid’s skirt is no longer quite what Cuyp applied. Same reaction, same direction, same inexorable sequence — just on a longer timescale than the specimen in our drawer, slowed but not stopped by climate-controlled gallery conditions. The Large Dort is mid-transformation, hanging in the National Gallery.

This is the thing about vivianite as a pigment: the painters who used it were, knowingly or not, applying a material that was going to keep reacting. The blue was transitional when they brushed it on. They were not painting with a stable color. They were painting with a process.

The People Who Already Knew

Melonie Ancheta’s 2019 paper in the American Indian Culture and Research Journal established something that two hundred years of Northwest Coast scholarship had gotten wrong. The blue paint on Haida and Tlingit ceremonial objects — masks, war helmets, shaman’s regalia, bentwood chests — was not, as had been consistently assumed and repeated without a single laboratory test, a copper oxide. It was vivianite. Iron phosphate. The mineral that forms in bones and decomposing organic matter, pulled from peat bog deposits along the coast.

The error correction came from SEM/EDX analysis. The scholarly literature had simply repeated the first wrong guess for two hundred years. The objects themselves knew the truth the entire time. Nobody thought to ask them until Melonie did.

What makes the identification remarkable is not just the correction. It is what the identification means. The Haida and Tlingit did not use vivianite merely because it was the available blue. Other blues became available over time — azurite, Prussian blue in the early 1700s, then cobalt and cerulean — and artists like Charles Edenshaw were still choosing vivianite into the 1900s for specific categories of objects. The new pigments did not replace it for those uses. They could not.

The reason is that vivianite was not functioning primarily as a color. It was functioning as a material with properties no synthetic blue could replicate. Among the Haida, blue was a liminal color — a portal color, associated with the border between the living world and the spirit world. In a survey of three hundred masks, Ancheta found blue paint in tertiary fields 75% more often on shamanic objects than on non-shamanic ones. Shamans painted their regalia with it to move between realms. Warriors painted their helmets blue so their faces resembled the dead. And the mineral chosen to make that blue was specifically the one that forms in bones — the one that grows in the space between the living system and the mineral system, the one that is always in the middle of becoming something else.

The chromism was not a problem to be solved. It was part of the material’s power. A paint that transforms — that starts one color and becomes another over time — corresponded directly to a cosmology in which transformation was fundamental. The Haida believed bones were the essence of life, symbols of death and regeneration. A mineral that forms in bone, that changes color when light reaches it, that moves between states and cannot be fixed in place: this was not a pigment with an unfortunate instability. This was exactly the right material for objects used at the border of life and death.

It is also worth remembering that none of these objects were made for electric light. They were made for firelight. In shifting flames and moving shadow, the matte blue surface of vivianite paint — absorbing light rather than reflecting it, unlike every other pigment in the palette — would have animated differently from the black and red around it. The chromism was not a static property to be preserved. It was a behavior, visible in the right conditions, in the right context, to people who understood what they were looking at.

The Haida and Tlingit understood something that academic mineralogy spent two centuries not examining: where a material comes from is part of what it is. Vivianite formed in bone carries the meaning of bone. That is not metaphor. It is a material epistemology, and it is more rigorous than the scholarship that ignored it for two hundred years.

Chromism in Three Institutions

Winchester. Dordrecht. Los Angeles.

The bentwood chests attributed to Albert Edward Edenshaw at the University of British Columbia’s Museum of Anthropology are displayed in a glass room with constant UV exposure. The vivianite in their tertiary fields — once a deep, matte, velvety blue applied with water, no binder — has darkened toward black. The blue is nearly gone from entire panels. The object is still in the collection. The color it was made with is finishing its transformation under gallery lighting.

In the National Gallery in London, The Large Dort hangs with its milkmaid’s skirt altered — the vivianite Cuyp applied in the mid-17th century has degraded, shifted away from what he mixed. The conservators who cleaned and restored it documented the change. They could not reverse it.

In storage in Los Angeles, a geological crystal — not a pigment, not a painting, just a specimen from the earth — is doing the same thing. Pages peeling. Black where the light reached. Blue-green in the fresh cracks. The same Fe²⁺ oxidizing to Fe³⁺, the same water molecules being lost, the same slow drift from vivianite toward metavivianite, one photon at a time.

The mineral does not distinguish between a 12th-century English cathedral wall, a 17th-century Dutch canvas, a 19th-century Haida ceremonial chest, and a geological specimen drawer in a natural history museum. It is running the same reaction in all of them. Melonie cannot stop it on the Edenshaw chests. The National Gallery cannot stop it in the Cuyp. I cannot stop it in the drawer. We can slow it — keep things in the dark, control the temperature, filter the UV — but we cannot stop it, because the chemistry does not require our participation to proceed.

Melonie once described vivianite as a diva material — unpredictable, with behaviors nobody fully understands. She has been studying it for twenty-five years and still cannot fully predict when or why it darkens on one object and not another. I have been trying to stabilize it under museum lighting conditions for most of my tenure. We found each other because we were both losing the same argument from opposite directions.

What stays with me is not that vivianite changes color. It is that the spectacular, irreproducible blue is not the beginning or the end but the middle. That is what the Haida painted on the faces of shamans at the border of life and death. That is what Cuyp brushed into the shadow of a milkmaid’s skirt on the river at Dordrecht. That is what the Winchester painter reached for when rendering the squirrel-fur lining of a cloak. That is what is in the drawer. Everyone, across all these contexts and centuries, has been trying to hold onto the middle of a transformation that was already underway before they found it.

The Close

The specimen in our drawer is not entirely vivianite anymore. It is somewhere between vivianite and metavivianite — the same external crystal form, a different internal mineral. The museum holds it because that is the commitment: perpetual care for the object, whatever state the object is currently in.

There is a third object in the drawer. A few years ago, Bob Hazen — mineralogist at the Carnegie Institution and one of the architects of mineral evolution theory, the framework that tracks how Earth’s mineral diversity expanded alongside biological complexity — donated a sample of vivianite-bearing clay to NHMLAC. It came from Maryland, collected from biomass and concrete deposits along a riverbank. Evidence of anthropogenic mineralization: vivianite forming right now, in the present, from modern organic waste along an industrial waterway. Same iron-phosphate-reducing chemistry, same three requirements — iron, phosphate, very little oxygen — but the organic matter is not ancient peat or a centuries-old burial. It is recent. It is us.

The clay is grey-blue, partially oxidized already — the same earthy, amorphous form that the Haida collected from coastal deposits and ground into pigment. Not the spectacular crystal. The working material. It sits in the drawer next to the splitting geological specimen, and together they show you the two faces of the same mineral: the dramatic and the humble, the ancient and the contemporary, the geological and the accidental.

The sequence does not end at metavivianite. Given enough time and light, the iron continues its work, the structure continues to shift, and eventually the blue is entirely gone — the mineral arrives at santabarbaraite, yellow-brown, fully oxidized, stable at last. The transformation completes. The Winchester paintings got there centuries ago. The specimen in the drawer is still en route.

You are upstream of all of this. Your hemoglobin iron and your apatite phosphate are precursors in a sequence that vivianite sits in the middle of. The mineral world is not waiting for you. It has been running this reaction for as long as there has been iron and phosphate and water on this planet. You are a temporary organization of materials that have cycled through this process before and will do so again.

The Haida painted it on the faces of shamans because they understood that the border between living and not-living is not a wall. It is a gradient, and vivianite is one of the places the gradient is visible. The reaction is happening now, downstream from a waste site in Maryland, in a painting in the National Gallery, in a ceremonial chest in Vancouver, in a drawer in Los Angeles. And in the iron in your blood, patient and reduced, waiting for conditions that will not arrive while you are still organizing them otherwise.

I am still trying to stop the specimen from changing color. Melonie is still trying to stop the paint on the Edenshaw chests from going dark. Neither of us is going to fully succeed, and the mineral, in its patient way, is not particularly concerned.

It has been in the middle of becoming something else for longer than there have been people to notice.

References

Ancheta, Melonie. “Revealing Blue on the Northern Northwest Coast.” American Indian Culture and Research Journal43:1 (2019). Available at nativepaintrevealed.com.

Howard, Helen. "Pigments of English Medieval Wall Painting." In Historical Painting Techniques, Materials, and Studio Practice, edited by Arie Wallert, Erma Hermens, and Marja Peek. Getty Conservation Institute, 1995. Available open access at: https://www.getty.edu/conservation/publications_resources/pdf_publications/pdf/historical_paintings.pdf

Spring, Marika. “Pigments and Colour Change in the Paintings of Aelbert Cuyp.” In Aelbert Cuyp, exhibition catalog, National Gallery of Art, Washington, October 7, 2001–January 13, 2002. Edited by Arthur K. Wheelock Jr. Amsterdam/London, 2001–2002, pp. 65–73.

Spring, Marika, and Larry Keith. “Aelbert Cuyp’s ‘Large Dort’: Colour Change and Conservation.” National Gallery Technical Bulletin (London: National Gallery, 2009). Available at nationalgallery.org.uk.

The collector said it casually, already moving on to the next shelf.

We should get you a better one.

He wasn’t wrong. Our rhodonite from the Morro da Mina mine in Minas Gerais, Brazil — deep crimson bladed crystals set against a matrix of glassy brown cummingtonite — is a perfectly respectable specimen. But he’d scanned the Container the way experienced collectors do, eyes moving fast and efficiently, cataloguing decades of accumulated taste in seconds. The rhodonite didn’t make the cut. He said so kindly, the way someone who genuinely loves minerals tells a small truth, and then he kept moving.

I kept thinking about it.

Not because he was wrong — he wasn’t. A collector’s eye is trained on real things: crystal habit, color saturation, locality, the invisible calculus of what makes one specimen exceptional and another merely present. That’s a legitimate way to look at minerals. It’s probably the right way, if what you’re doing is building a collection worth looking at.

But I’m a curator. And curators, at least this one, have a bad habit of looking through the specimen rather than at it. The rhodonite on that shelf isn’t exceptional. What’s inside it is.

The element is manganese. And the offhand comment about finding us a better specimen sent me somewhere I didn’t expect to go.

But first — the Container.

The vault door is heavy enough that you feel it in your shoulders when you pull it open. It doesn’t swing — it concedes, slowly, like something that has decided to let you in. On the other side is a room about the size of a generous home office, maybe twelve by eighteen feet, and everything in it is grey. The walls are grey. The floor is grey. The safes and locked metal cabinets lining every surface are grey, unlabeled, giving nothing away. The lights run cool and bright — 4000 Kelvin, the color of overcast daylight — and the room echoes when you move. There is no softness in here to absorb sound.

This was intentional. When we designed the Container, we made a deliberate choice to drain it of competition. The room would not have a personality. It would not have warmth or atmosphere or visual interest of its own. It would simply hold things and get out of the way. Because what it holds doesn’t need any help.

Well, most of what it holds does not need any help.

ACT 2: THE PINK STONE AND THE BLACK VEINS

Rhodonite is a manganese inosilicate, (Mn,Ca)SiO₃, crystallizing in the triclinic system with a characteristic rose-red to pink color caused by Mn²⁺ within the structure. It’s genuinely beautiful when it’s beautiful — gem-quality crystals from the Ural Mountains or Broken Hill, Australia can rival any pink gemstone in intensity. The Morro da Mina material has its own character: those deep crimson blades rising out of a glassy brown cummingtonite matrix, the contrast between the two sharp enough to stop you for a moment.

Many rhodonite specimens — particularly those from Franklin, New Jersey — show black veining that cuts through the pink matrix like ink strokes through watercolor. Those veins are not an impurity in the conventional sense. They’re manganese oxide dendrites, formed when the same manganese that gives the pink its color oxidizes at fracture surfaces and grain boundaries, changing oxidation state from Mn²⁺ to Mn⁴⁺ and precipitating as MnO₂.

The mineral is quite literally painting its own portrait in two oxidation states simultaneously.

This is worth pausing on, because oxidation state is the real story of this element across all of human history. The chart below shows the primary manganese minerals and where rhodonite sits in terms of Mn content. Yes, it’s at the low end. No, that’s not the point.

The high-Mn ore minerals — pyrolusite, hausmannite, braunite — are all black. All of them. The mineral only turns pink when it’s sitting quietly in the crystal between two chains of silica, surrounded by enough silicon and oxygen to keep it reduced and beautiful. The moment it oxidizes, it goes dark. That transformation, it turns out, is the engine behind almost everything manganese has ever done for us.

It has been going unrecognized for a long time.

In ancient Magnesia — in what is now Greece or western Turkey — two black minerals were found in the same region and given the same name: magnes, after their place of origin. One attracted iron. The other didn’t, but was used to decolorize glass. To distinguish them, later writers assigned them a sex. The lodestone was the male magnes — active, attracting, doing something visible and dramatic. The pyrolusite was the female magnes — quieter, its action invisible, its chemistry not understood.

From magnesia came manganesum. From manganesum came manganese. And from the same ancient root, by a different path, came magnesium — two different elements sharing an etymology, one of them accidentally linked to magnetism despite being entirely non-magnetic.

Manganese is the least magnetic of the transition metals. The female magnes, it turns out, was named for attraction and has none.

The collector would have felt right at home in ancient Magnesia.

ACT 3: THE FIRST MARK

Before there was writing, before there was agriculture, before there was anything we would recognize as civilization, there was manganese.

Specifically, there was pyrolusite — MnO₂, manganese dioxide — ground to a fine black powder, mixed with animal fat, and pressed by a human hand against a cave wall. The paintings at Lascaux. At Altamira. At Chauvet. The bison and horses and aurochs that have survived thirty thousand years in the dark were put there with manganese (with charcoal and soot) — manganese, chosen specifically because its chemical stability meant the pigment would not fade, would not migrate, would not be undone by time.

We do not know if the people who made those paintings understood why the black held so well. They knew it worked. They came back to it, generation after generation, extracting pyrolusite from the same deposits, carrying it into the deep chambers where the art was made. One of the oldest, if not the oldest, sustained materials science programs in human history, and the active ingredient was the same element sitting in the pink silicate on my shelf.

Pyrolusite — black, heavy, unremarkable to look at — is Mn⁴⁺. Manganese in its highest common oxidation state, fully oxidized, stripped of electrons. It is not pink. It will never be pink. That requires the element to calm down considerably, to sit in a silicate structure as Mn²⁺ with a full complement of electrons, surrounded by oxygen and silicon in a structure that coaxes out the rose color we find beautiful enough to carve into imperial tombs.

The cave painters weren’t thinking about oxidation states. But they were, without knowing it, exploiting the same chemical flexibility that makes this element so remarkable — its ability to exist in multiple oxidation states, each one behaving like an almost entirely different material. Black and permanent for pigment. Pink and beautiful for jewelry. And as we’ll see, other oxidation states entirely for the things that came next.

The darkness of Lascaux and the pink of a Brazilian mine are the same atom, differently charged.

That’s worth sitting with for a moment.

ACT 4: THE GLASSMAKERS’ SOAP

Here is a question worth asking: when did human beings first make something truly transparent?

Not translucent — not the thin-scraped animal hide stretched over a window opening, not the oiled parchment that let in a grey approximation of daylight. Truly transparent. Something you could hold up and see the world through clearly, something that separated you from the weather without separating you from the light.

The answer is glass. And the answer to how they made it clear is manganese.

The problem with early glass is iron. Sand — the raw material of glass — is almost never pure silica. It contains iron impurities, and iron in glass is not neutral. Depending on its oxidation state, ferrous iron (Fe²⁺) tints glass blue-green; ferric iron (Fe³⁺) pulls it toward yellow. The result, without intervention, is glass the color of a shallow harbor — attractive in its way, but not transparent. Not useful for seeing through.

Egyptian glassmakers at Amarna were working with this problem as early as 1500 BCE. Roman glassmakers systematized the solution by around 100 CE, and it spread through the empire from workshops in Alexandria. The secret ingredient was pyrolusite — the same black manganese dioxide the cave painters had been grinding for pigment for twenty-eight thousand years before anyone thought to put it in molten glass.

What pyrolusite does in a glass melt is chemically elegant. Mn⁴⁺ oxidizes the Fe²⁺ to Fe³⁺ — the stronger blue-green absorber becomes the weaker yellow absorber — while the Mn⁴⁺ itself is reduced to Mn²⁺, which in small quantities is nearly colorless. Two problems cancel each other out. The glassmakers called it their soap — sabon de verre in later French usage — because it cleaned the color from glass the way soap cleans dirt from cloth. They didn’t know they were running a redox reaction. They knew it worked.

Medieval Venetian glassmakers inherited this knowledge and refined it. By the 14th century, the workshops of Murano were producing glass of extraordinary clarity using manganese dioxide sourced from deposits across Europe. The transparency of Venetian glass — that quality that made it so coveted, so expensive, so widely imitated — was built on manganese chemistry that traced an unbroken line back to Alexandria, and before that to Egypt, and before that to the same pyrolusite deposits that stocked a cave painter’s kit.

Then something unexpected happened. Centuries later, in the older houses of New England, people began noticing that their clear window panes were turning purple.

Not all of them. Not quickly. But glass that had been installed in the 18th and early 19th centuries — glass made with manganese dioxide as a decolorizer, in the tradition that stretched back to Rome — was slowly, visibly, changing color in the sunlight. Developing a lilac, then an amethyst hue, deepening over decades.

A New England glass manufacturer named Thomas Gaffield investigated in the 1820s and found the culprit: the manganese. The Mn²⁺ that had been doing its quiet decolorizing work was being photo-oxidized by sunlight — the same photochemical process, driven by the same light energy, reverting it back toward its original oxidized state. The glass was, in a sense, remembering what the manganese had been before the iron reduction happened. It was returning, molecule by molecule, to its cave-painting self.

American glass collectors now prize solarized amethyst glass as a mark of authenticity — a signature of age, of a particular era of manufacture. What they’re looking at, held up to the light, is manganese chemistry running slowly in reverse. An unintentional clock. The glass is purple because the manganese is oxidizing, because the sun is doing what sunlight does, because the element that cleaned the glass two hundred years ago is now, patiently, undoing its own work.

There is a word for this in mineralogy. We call it alteration.

ACT 5: THE ELEMENT THAT WON WARS — AND NOBODY KNEW IT

There is a persistent mystery in ancient military history.

The Spartans were not the largest army in the ancient world. They were not always the best supplied, the most numerous, or the most strategically sophisticated. But their steel had a reputation that preceded them — weapons that held an edge longer, armor that absorbed impact differently, metal that behaved under stress in ways that their enemies’ metal did not. Ancient writers noted it. Modern historians have puzzled over it. The explanation, when it finally came, had nothing to do with Spartan smithing technique or secret forging knowledge.

It was the ore.

The iron deposits available to Spartan metalworkers happened to contain manganese — not by design, not by understanding, but by geology. When that ore was smelted, trace amounts of manganese entered the iron, and the resulting metal was measurably harder, tougher, and more resistant to deformation than iron produced from cleaner deposits elsewhere. The Spartans had a materials advantage they could neither explain nor reproduce intentionally. They just knew their metal was better. They were right. The reason was sitting in the ore the whole time, invisible, doing what manganese does.

This would remain unexplained for roughly two thousand years.

The formal understanding of manganese’s role in steel began to emerge only in the early 19th century. In 1816, a German researcher documented that iron alloyed with manganese was harder without becoming more brittle. Patents followed in Britain. Industrial production of ferromanganese began. And then, in 1856, a British steelmaker named Robert Forester Mushet solved a problem that had been threatening to derail the entire Industrial Revolution.

Henry Bessemer had invented his converter — the process that would make mass-produced steel possible, that would build the railroads and the bridges and the structural skeletons of modern cities. But the Bessemer process had a flaw: it left excess oxygen and sulfur in the steel, making it brittle and unpredictable. Unusable at scale. Mushet’s solution was to add spiegeleisen after the blow — a pig iron rich in manganese and carbon. The manganese scavenged the oxygen and sulfur, the carbon restored the carbon content, and the steel that came out the other end was consistent, workable, and strong.

Mushet saved the Bessemer process. The Bessemer process built the modern world. And the active ingredient in Mushet’s fix was manganese.

Then Robert Abbott Hadfield took it further. In Sheffield in 1882, Hadfield produced an austenitic manganese steel containing roughly 12% manganese — and discovered something that seemed almost physically impossible.

The more you hit it, the harder it got.

If that sounds familiar, it should — Marvel’s writers were working from real materials science when they dreamed up vibranium. The Black Panther suit doesn’t just resist impact; it absorbs and stores it. Hadfield got there first, in Sheffield, in 1882.

Impact didn’t damage Hadfield steel. It strengthened it. The technical term is work-hardening: manganese stabilizes the austenite phase of iron so that deformation induces martensitic transformation at the surface while the bulk remains tough and ductile underneath. In practical terms, you could build things that needed to be simultaneously hard on the outside and resilient within — railway track switches that grew stronger under every passing train, rock crushers, and eventually, tank treads and armor plate. The violence, counterintuitively, was doing the maintenance.

During the Second World War, the Ural Mountains — the same geological formation that had been producing rhodonite for imperial Russian lapidary work, the same mountains the Tsarina’s sarcophagus came from — were supplying roughly 70% of the Soviet Union’s manganese. The decorative stone in the Peter and Paul Cathedral and the armor on the T-34 tank share a mineralogical address. The pink and the black. The beautiful and the brutal. Coming out of the same mountains, in the same decade, for completely different purposes.

No one was thinking about oxidation states (I don’t think). No one was thinking about rhodonite. They were thinking about steel.

But the element doesn’t care what you’re thinking. It just performs.

ACT 6: THE PATIENT ELEMENT

In 1866, a French engineer named Georges Leclanché sat down and invented the battery that would eventually power the modern world — not immediately, not dramatically, but with the quiet persistence that seems to characterize everything manganese touches.

The Leclanché cell was simple by the standards of what came before it. A zinc anode. An electrolyte of ammonium chloride solution. And a cathode made of manganese dioxide packed around a carbon rod. Pyrolusite again. The same black mineral that colored cave walls and cleaned glass and silently improved Spartan weapons was now sitting at the heart of an electrochemical cell, doing something entirely new: accepting electrons during discharge, reducing from Mn⁴⁺ toward Mn³⁺, storing and releasing energy.

The cell worked. It was cheap. It didn’t require a liquid electrolyte that would spill — later refinements produced the dry cell, which made portable electrical devices possible for the first time. The flashlight. The portable radio. The transistor radio your grandfather carried. Every alkaline battery you have ever put in a television remote or a child’s toy contains manganese dioxide as the cathode material, running the same basic electrochemical reaction that Leclanché worked out in 1866.

One hundred and sixty years of batteries. The same element. The same oxidation state chemistry.

What makes manganese so useful in batteries is the same thing that made it useful everywhere else: its willingness to change. Mn⁴⁺ accepts electrons readily, becoming Mn³⁺, then Mn²⁺. It does this reversibly, or nearly so, which is the fundamental requirement of a rechargeable system. It is abundant — the 12th most common element in the Earth’s crust, present in accessible deposits on every continent and in extraordinary concentrations on the seafloor in polymetallic nodules that may eventually represent one of the largest untapped mineral resources on the planet. It is relatively non-toxic compared to the cobalt and nickel that dominate current lithium-ion battery chemistry. And it is cheap — orders of magnitude cheaper than the materials it might replace.

Battery engineers have known all of this for decades. The challenge has been getting manganese to behave well enough in rechargeable systems at the voltages and energy densities that modern applications require.

Lithium manganese oxide — LMO, the spinel structure, LiMn₂O₄ — was one of the first serious answers. Insert lithium ions into the manganese oxide framework during charging; extract them during discharge. The manganese cycles between oxidation states to compensate the charge, the spinel structure provides channels for the lithium to move through, and the result is a cathode material that is safer, cheaper, and more thermally stable than lithium cobalt oxide. LMO has been in commercial lithium-ion batteries since the 1990s. It is in your power tools. It is in some electric vehicles. It is, right now, the subject of active research into how to improve its cycle life, reduce its capacity fade, and extend its useful temperature range.

It is also, in a different form, central to my current research.

I work on lithium extraction from brine — the recovery of lithium from the saline waters that saturate the sediments beneath salt flats, the same brines that are now central to the global scramble for battery materials as electric vehicle demand accelerates. The challenge is selectivity: brines contain not just lithium but sodium, potassium, magnesium, calcium, and a periodic table’s worth of other ions, all competing for the same extraction pathways. One of the most promising approaches uses manganese oxide frameworks — synthetic analogs of the natural mineral structure — as ion-selective sieves. The framework’s tunnel geometry and the specific coordination chemistry of the Mn sites preferentially accommodate Li⁺ over larger competing ions.

It is, in its way, the same trick the cave painters were running. Find a manganese oxide mineral that works the way you want it to, and exploit what it naturally does.

The rhodonite on the shelf in the Container started as manganese settling out of a Palaeoproterozoic ocean — somewhere between 2.07 and 1.86 billion years ago, when that ocean was still negotiating its relationship with oxygen. The manganese precipitated as oxide particles in shallow, oxic water, then dissolved and redeposited as it crossed into the euxinic depths below. Eventually it was buried, metamorphosed at temperatures exceeding 600°C, folded, sheared, and exhumed over the better part of two billion years. What came out the other end was queluzite — a manganese silicate-carbonate rock — at the Morro da Mina mine in Minas Gerais, Brazil. The rhodonite is part of that rock, those deep crimson crystals the visible evidence of manganese that survived conditions that would have destroyed almost anything else.

The manganese oxide framework I work with in the lab was synthesized last year in an oven, by my colleague and retired Lawerence Livermore National Laboratory scientist. The geological and the synthetic materials are separated by nearly two billion years but just a few drawers away from each other.

That proximity is not an accident. It’s the reason I’m here. California is rich in minerals and gems, but also rich in amazing scientists to work with.

I came to manganese through the research — through lithium extraction, through brine chemistry, through the problem of getting one specific ion out of water that contains everything else. But I stayed because of what manganese kept turning out to be. Every time I thought I understood it, it opened another door. Cave paintings. Glass. Steel. Batteries. And a lab inside a natural history museum, a synthetic version of its crystal structure being asked to help power the energy transition.

The work is ongoing. The mechanism is not fully understood. My colleagues and I published our most recent findings in 2025 — if you want the crystallography, it’s linked below. But the short version is this: manganese is still surprising us. After thirty thousand years of human use, we are still learning what it does and why.

That is, I think, the most honest thing I can say about why I love this element. Not because we understand it. Because we don’t, quite, yet.

EPILOGUE: BACK TO THE CONTAINER

The collector was generous with his time that day. We talked for a while about rhodonite — the Brazilian deposits, what makes a specimen exceptional, the collector’s market for high-quality material. He knew his stuff. He knows rhodonite the way people who love minerals know things: deeply, specifically, with real feeling for the object.

He was evaluating the specimen. I kept drifting back to the element.

A specimen is a window. What you look through it at depends on what questions you’re carrying when you pick it up. He was carrying aesthetic questions — color saturation, crystal habit, locality, the invisible calculus of what makes one rhodonite better than another. Those are legitimate questions. They are not the only questions.

The collector offered to find us something better. I told him I’d love that. I meant it — a beautiful rhodonite would be a genuine gift, and the right specimen in the right light is its own argument for why these minerals deserve our attention.

But I’ve also become rather fond of the one we have.

It painted the first human mark on a wall. It made Roman windows clear. It gave the Spartans an advantage nobody could explain for two thousand years. It is, right now, on a shelf in a grey container, waiting for the next person who picks it up and asks what it does rather than what it’s worth.

Not much to look at. Performs impressively.

If you found this interesting, Pocketful of Xtals goes out regularly to readers who think minerals are more than pretty rocks. Subscribe below — it’s free, and I promise the Curator’s Notes will keep getting longer.

Subscribe now

CURATOR’S NOTES

1. On the durability of manganese pigment in cave environments

The ongoing conservation crisis at Lascaux is frequently misunderstood as a threat to the pigments themselves. It isn’t — not primarily. The danger is to the limestone substrate beneath them. Elevated CO₂ from visitors’ breath dissolves into condensation water on the walls, forming carbonic acid that attacks the calcium carbonate bedrock. When the wall destabilizes, the pigment comes with it. The iron-based ochres — the reds and yellows — are additionally vulnerable to microenvironmental redox changes, and any organic binders used to fix pigments are highly susceptible to biological colonization.

The manganese black, by contrast, is chemically among the most stable pigments in the cave. MnO₂ is already fully oxidized — there is nowhere left for it to go. It doesn’t react meaningfully with carbonic acid under the conditions present. Thirty thousand years of geological time: survived. Forty years of uncontrolled human visitation: the wall begins to fail. Lascaux was closed to the public in 1963 and a replica built nearby. The manganese is still there, in the dark, perfectly intact, waiting. The most durable thing the painters left behind is also the simplest chemically — an element that had already given up everything it had to give.

2. On Hadfield steel and the work-hardening mechanism

The reason Hadfield steel hardens under impact is a function of its unusual phase stability. At roughly 12% manganese, the austenite phase — normally only stable at high temperatures in plain carbon steel — is stabilized at room temperature. When the surface is struck, the mechanical energy induces a localized transformation: austenite converts to martensite at the point of impact, creating an extremely hard surface layer. The bulk of the material remains austenitic and therefore tough and ductile. Repeated impact keeps driving this transformation outward from the surface, so the material literally gets harder the more it is used.

Railway track switches made from Hadfield steel last significantly longer than those made from conventional steel precisely because train wheel impacts are, counterintuitively, maintaining them. The violence is doing the maintenance. This is also why Hadfield steel is nearly impossible to machine — any cutting tool impact hardens the surface faster than it can be cut. It has to be cast to near-net shape or ground, not machined.

3. On the lithium extraction research

The brine extraction work uses a spinel-structured lithium manganese oxide framework as an ion-selective sieve for lithium recovery. The full mechanism — including what we found about how the structure fails under maximum loading conditions, and what that means for practical cycle life — is described in our 2025 paper in the Journal of Raman Spectroscopy. https://analyticalsciencejournals.onlinelibrary.wiley.com/doi/10.1002/jrs.70013

This work was supported by the U.S. Department of Energy’s Office of Energy Efficiency and Renewable Energy under the Advanced Manufacturing Office, in collaboration with colleagues at Oak Ridge National Laboratory and Mineral Selective Technologies.

She wasn’t supposed to stop there.

Most people walk past galena. It sits in its case looking like something an architect might keep on a desk — a stack of perfect silver-gray cubes, edges so sharp they seem machined, surfaces that catch light like a fragment of mirror dropped into stone. It doesn’t have the fire of a ruby or the depth of a sapphire. It doesn’t glow. It doesn’t pulse. It just sits there, impossibly perfect, being a cube.

But she stopped.

I won’t pretend I don’t remember who it was. Angelina Jolie was making her way through the mineral hall at the Natural History Museum of Los Angeles County, and something made her stop at the galena case. She tilted her head — the way people do when geometry surprises them — and asked why it was so perfect.

I told her about the crystal structure. About the way lead and sulfur ions lock together in a face-centered cubic lattice, and how the mineral cleaves along those planes with almost no resistance, leaving faces so flat they’re essentially mirrors. The perfection isn’t polish. It isn’t craft. It’s the mineral expressing its own internal architecture at the scale of your hand.

Then I told her about the radio.

She looked at me the way people look at me when I say things like that.

A chunk of galena, I explained, can pull a radio signal out of the air with no batteries, no power source, no electricity whatsoever. Touch a fine wire — a “cat’s whisker” — to the right spot on the crystal surface, and the mineral rectifies the alternating current of a radio wave into the direct current your earphone needs to produce sound. It’s the world’s first semiconductor. It works because of quantum mechanical electron behavior at the crystal surface, which in 1900 nobody understood and nobody needed to. The galena just worked.

She said that sounded useful for a desert island. A MacGyver situation.

I said it only receives AM radio. So unless you’re really into conservative talk radio, you might be disappointed.

She laughed. We moved on.

But I’ve thought about that exchange more times than I can count. Because the joke — the punchline about AM talk radio — depends on galena being a novelty. A curiosity. A party trick for mineralogists.

And galena is so much more than that. It is, in ways that are genuinely hard to convey, one of the most consequential minerals in the history of human civilization. Not because it’s rare. Because it isn’t.

The Oldest Mascara

Let’s go back. As far back as we can.

Six thousand years ago, in the Nile Delta, Egyptian women — and men, and children — were grinding galena into a fine black powder and applying it to their eyelids. They called it mesdemet. We inherited the word kohl later, from Arabic — but the practice and the mineral are Egyptian, and they are very old.

The Egyptians used mesdemet for reasons that were cosmetic, religious, and possibly medical simultaneously. That deep black outline around the eye is immediately recognizable across three millennia of portraiture. But in 2010, French researchers reported something more unexpected: the lead compounds formed when galena was ground and prepared may have had genuine antimicrobial properties, triggering an immune response that reduced eye infections in a culture living alongside a river that was also an open sewer. The Egyptians almost certainly didn’t understand the chemistry. But they noticed, across generations, what worked — and they refined their preparations with striking sophistication, deliberately synthesizing lead compounds that don’t occur naturally in Egypt and would have required controlled wet chemistry to produce. The world’s first cosmetic industry was also, in some sense, its first pharmaceutical one.

Galena was also the primary ore from which lead was smelted, and here the story bifurcates. Lead — the element — has one of the most complicated relationships with human civilization of any material on Earth. It is extraordinarily useful: dense, malleable, resistant to corrosion, with a low melting point that makes it easy to work. The Romans built their water infrastructure from it. The Latin word for lead, plumbum, is why we call plumbers plumbers. They lined their wine vessels and cooking pots with lead, used lead acetate as a sweetener, and consumed it at levels we now recognize as neurologically catastrophic. Some historians have argued — controversially, but not without evidence — that chronic lead poisoning contributed to cognitive deterioration across the Roman ruling class.

The mineral itself was blameless. Galena forms in hydrothermal veins, crystallizing from hot metal-rich fluids moving through fractures in the crust. It has been doing this for billions of years. What humans decided to do with the lead they smelted from it is a different story entirely — one that runs from Roman aqueducts through Elizabethan face paint through Victorian plumbing through leaded gasoline through the Flint water crisis. The mineral sat in the earth, perfect and indifferent. The choices were ours.

The Cube Explains Itself

I want to stay with the cube for a moment, because it deserves more than a passing description.

Galena is lead sulfide: one lead atom for every sulfur atom, arranged in a face-centered cubic crystal structure that crystallographers call the rock salt structure — the same geometry as table salt, the same geometry as the mineral halite. The lead ions and sulfur ions alternate in three dimensions, each surrounded by six of the other type, the whole assembly held together by ionic bonds.

Galena’s perfect cubic cleavage — those razor-sharp right-angle faces — is the direct physical expression of that internal architecture. When you cleave galena, you’re not breaking random chemical bonds. You’re separating the crystal along the planes where the ionic bond density is lowest, which happen to be the planes parallel to the cube faces. The mineral isn’t being cut. It’s being asked to express a preference it already had.

The result is a surface that catches light like metal — not because it’s optically smooth (under a microscope, fresh cleavage faces are anything but), but because galena’s electrons are delocalized across the crystal in a way that more closely resembles a metal than a typical ionic mineral. It sits at the boundary between ionic and metallic bonding. That same electronic character — the loosely held electrons, the narrow band gap at the crystal surface — is precisely what gives galena its semiconductor properties and made it the foundation of the radio age.

The Cat’s Whisker

In 1874, a German physicist named Karl Ferdinand Braun was twenty-four years old and teaching in Leipzig. He had been studying electrolytes — the way metal salts conduct electricity when dissolved in water — and had started asking a stranger question: could solid metal sulfide crystals conduct electricity without being dissolved at all? He began pressing metal electrodes against crystal surfaces and measuring what happened.

What happened was strange. When he probed galena with the point of a slender silver wire, current flowed easily in one direction and poorly in the other. The crystal wasn’t behaving like a resistor, which treats current symmetrically. It was rectifying it — acting differently depending on which way the current was pushed. Braun published the finding in Annalen der Physik that November. He couldn’t explain it. Nobody could. He noted it as a curiosity and moved on.

This is the part of the story I find most honest about how science actually works.

The Higgs boson was predicted mathematically in 1964. When CERN finally confirmed it in 2012 after decades of effort and billions of dollars, physicists were relieved rather than surprised — it was the most expensive checkbox in the history of science. The public experienced it as a revelation. Most scientists experienced it as a confirmation of something they already believed. Predicted discovery is a different category of event than accidental discovery, and the two feel nothing alike from the inside.

Braun had no prediction. He had a twenty-four-year-old’s hunch about electrolytes and a slender wire. He found something that had no name, no theory, no framework, and no application. The conceptual tools to understand what rectification meant for electron behavior in solids wouldn’t exist for another half century. So he did what honest scientists do when they find something they can’t explain: he described it carefully, published it accurately, and filed it away.

Why silver wire specifically? Braun never explained his choice in the surviving record, but it wasn’t a random one. Silver has the highest electrical conductivity of any metal — better than copper, far better than iron or brass — which meant the probe itself would introduce almost no resistance, leaving Braun measuring the crystal’s behavior rather than the wire’s. It can also be drawn to an exceptionally fine point without hardening and snapping. Whether he chose it by calculation or by a good experimentalist’s instinct, it was the right tool. The name “cat’s whisker” came later, once people could see what the device actually looked like: a slender wire probing gently across a crystal surface, exactly like a cat testing something unfamiliar.

Twenty-six years passed. Then radio arrived, and everything changed.

In 1900, an American inventor named Greenleaf Whittier Pickard began experimenting with galena crystals as radio detectors. The physics of radio reception requires rectification — the conversion of the oscillating alternating current of a radio wave into the unidirectional direct current that drives an earphone. Coherers and electrolytic detectors had been doing this job, clumsily. Galena did it better.

Pickard’s crystal set was disarmingly simple: a long wire antenna to gather the radio signal, a tuning coil, a pea-sized crystal of galena, and a cat’s whisker that the operator touched to the crystal surface, moving it slowly until the signal sharpened into sound. No batteries. No vacuum tubes. No external power of any kind. The energy in the earphone came entirely from the radio wave, harvested by the antenna and rectified by the crystal.

By the early 1920s, when commercial radio broadcasting began, an estimated forty million crystal sets were in use worldwide. For millions of families who couldn’t afford the expensive battery-powered tube radios, a galena crystal and a coil of wire was the first radio they ever owned. Farmers in rural areas without electricity pressed a cat’s whisker to galena and heard weather forecasts. Children built crystal sets from kits and from scratch. The Boy Scouts used them as educational projects.

Galena — the most widely mined lead ore on Earth, the mineral people had been smelting for six thousand years — turned out to have been a semiconductor the entire time. It just took humanity until 1900 to notice.

What Soldiers Heard

The first military application came quickly.

In World War I, the U.S. Army fielded the SCR-54, a portable crystal radio receiver small enough to carry in a pack, used at artillery stations to receive messages from fire control aircraft. The galena detector was standard. It required no batteries, which meant no supply chain vulnerability, no dead cells at critical moments. The crystal worked or it didn’t.

By World War II, the technology had been nominally superseded by vacuum tube radios. But galena — and the improvised ingenuity it inspired — proved to have a second life that no military planner anticipated.

In the German-occupied Channel Islands, the Nazi command confiscated all civilian radio sets. The purpose was isolation: no BBC, no news from the outside, no signal that the war might be going differently than the occupiers claimed. In occupied Guernsey, to possess a radio was to risk imprisonment. In some occupied territories on the continent, it was to risk death.

The BBC broadcast instructions anyway, through a program hosted by a figure known only as “Colonel Britton.” The instructions were for building crystal sets from whatever materials were at hand.

On Guernsey, a man named W.A. Renier and a friend built more than fifty sets and distributed them in secret. The crystal detectors — the essential component — were homemade. The method: mix yellow sulfur and lead chips in equal parts, pack the mixture into a German rifle cartridge case, seal it, and bake it in a fire. The result, when cracked open, was crude lead sulfide — the same compound as galena, if not its crystalline equal — a glassy material with bright metallic spots that proved sufficient to rectify a radio signal.

They used the enemy’s ammunition casings to synthesize a semiconductor. They listened to the BBC. They typed up summaries of the news and passed them through the community. They did this until liberation in May 1945, knowing that discovery meant prison at minimum and possibly worse.

Galena. Made from German cartridges. Keeping occupied people connected to the truth.

I think about that whenever someone asks me why minerals matter. The answer is always longer than they expected.

After the Crystal

The germanium diode, developed during World War II, finally made the cat’s whisker detector obsolete. The era of galena radios ended. The mineral returned to its primary industrial role — lead ore, smelted in enormous quantities for batteries, ammunition, and radiation shielding.

But the physics that Braun noticed in 1874, that Pickard exploited in 1900, that occupied islanders weaponized in 1944 — that physics didn’t go away. The point-contact semiconductor junction at the galena crystal surface was the conceptual ancestor of the transistor, developed at Bell Labs in 1947. The transistor begat the integrated circuit. The integrated circuit begat everything that followed.

Braun did eventually win a Nobel Prize — in 1909, shared with Marconi, for contributions to wireless telegraphy. Not for semiconductors. The semiconductor work wouldn’t be recognized as foundational until transistors proved it, decades after his death. He found the most important thing he ever found and didn’t know it. That’s not a failure. That’s what the frontier actually looks like.

Every electronic device you own traces its conceptual lineage to a cat’s whisker touching galena.

The mineral in the hall is still a cube. Still mirror-bright. Still stopping people in their tracks with the simple perfection of its geometry. It carries six thousand years of human history in its cleavage planes — Egyptian cosmetics, Roman plumbing, WWI field communications, WWII resistance networks, the entire semiconductor age.

It only gets AM radio.

But I’m not sure there’s a better argument for why you should take mineralogy seriously.

Aaron Celestian, PhD is Curator of Mineral Sciences at the Natural History Museum of Los Angeles County.

I pick it up without thinking. I slot it into my weapon like a battery — the game’s way of saying this gem now powers you. My fire damage increases. I move on.

It takes another twenty hours before the scientist in me finally wakes up and says: wait.

Where did that shaman get a ruby?

I mean this literally. The shaman was alive — animated, aggressive, casting spells at me. Before I killed it, it was a functioning creature moving through a cave. It had no pockets I could see. No mining equipment. No gemological supply chain. And yet, inside it, apparently, was a ruby. A corundum crystal formed under metamorphic pressure, somewhere between five and fifty kilometers underground, over a span of time measured in millions of years.

The shaman had been carrying it around.

I put down the controller and stared at the ceiling for a while. Then I went back to playing, because I needed more rubies.

This is the Monster Problem. Not that games do this — games abstract everything, and I’ll defend that shortly — but that we accept it so completely, so automatically, for minerals specifically, that it takes forty hours of active play before the question even surfaces. A ruby falls out of a goblin and it feels correct. That feeling is interesting. I want to understand it.

I work in a building full of rubies. As Curator of Mineral Sciences at the Natural History Museum of Los Angeles County, I spend my days explaining to visitors how gems actually form. A ruby is corundum — aluminum oxide — crystallized under the specific pressure and temperature conditions of regional metamorphism, its red color produced by chromium atoms substituting for aluminum in the crystal lattice. The whole process requires conditions that would kill anything living: depths measured in kilometers, temperatures above 500 degrees Celsius, timescales measured in tens of millions of years.

The formation history is the identity. That’s the thing I keep coming back to when I think about the shaman and his ruby. A ruby isn’t a red stone that happens to be hard and valuable. It’s a record — a physical inscription of specific pressure, chemistry, and time. You can read those conditions in the crystal structure. You can trace the tectonic history of the region in the inclusion patterns. The ruby is a document written by the earth over millions of years.

The shaman had been using it as a stat boost.

I have a database of 291 video games spanning four decades of game design. I built it to understand how games use minerals — what they include, what they ignore, and what their choices reveal about our collective relationship with the material world. And the first thing the data tells me is this: more games give you gems by killing enemies than by mining the ground.

172 games use enemy drops as a source of minerals. 141 use mining. And 122 games — 42% of all mineral-containing games in the database — give you gems exclusively from monsters, with no geological sourcing whatsoever. No mines. No ore veins. No geological context of any kind. The ruby comes from the shaman, full stop.

This has been normal game design for forty years. No one has asked why.

Before I ask why, I want to be fair about what this actually is. Games abstract constantly and gloriously. Your character carries fifty swords in an invisible backpack. You heal catastrophic injuries by eating a sandwich. You fall from a great height and walk it off. These aren’t failures of realism — they’re the necessary simplifications that make interactive entertainment work. The goblin carrying a ruby is in the same category as the inventory system that holds it. Game logic, not geological logic.

I’m not here to correct the goblin. I’m here to ask what it reveals that we never notice him.

The data has an answer, and it’s more interesting than “games are unrealistic.”

When I separate the 122 enemy-drop-only games from the 91 mined-only games and look at what minerals actually do in each group, a clear picture emerges. In games where gems come from enemies, their dominant function is power augmentation — they directly increase your combat statistics, boost your abilities, make you stronger. 59% of enemy-drop-only games use minerals this way.

In games where gems come from the ground, the dominant function is crafting and progression — you build things, upgrade equipment, advance economically. 70% of mined-only games use minerals this way. A further 58% use them as economic trade goods.

These aren’t just different game mechanics. They’re two different theories of what minerals fundamentally are.

When a gem comes from the earth, it’s a material — something extracted, processed, and transformed through labor and knowledge. The earth doesn’t yield a finished ruby; it yields corundum in a host rock, and understanding what that rock is, how it formed, and where to find it is the work. This is, more or less, how economic geology actually operates. A mining geologist reads the landscape for structural signatures, traces the metamorphic history of a region, and uses that knowledge to find what’s there. The games that model this well — and a handful do — quietly teach players to think like geologists without ever saying so.

When a gem comes from a monster, it’s concentrated power — something the creature held, something that passes to you when you defeat it. The monster was a vessel. The gem is the essence. You absorb it.

That second model is ancient. It precedes science by millennia. Every folk tradition that assigns healing, protective, or magical properties to minerals is operating on exactly this logic: power concentrates in objects, objects can transfer that power to their holder, and the holder’s virtue or strength determines how much power transfers. The healing crystal doesn’t care where it formed. The amulet doesn’t require a geological context to work. The power is in the object, independent of its origin.

The geological theory — where a mineral’s properties are caused by its formation conditions, where the ruby’s hardness comes from its crystal structure and its color from chromium substitution and both of those trace back to specific metamorphic conditions — is the scientific model. And in my database, it’s the minority position.

The clearest illustration of how this happened sits in five rows of my spreadsheet: Diablo (1996), Diablo II (2000), Diablo II: Resurrected (2021), Diablo III (2012), Diablo IV (2023). Five games, twenty-seven years. Every single one: gems from enemies, no mining, pure power augmentation. The Diablo franchise didn’t invent this design philosophy, but it codified it, refined it, and distributed it to hundreds of millions of players across three decades.

Compare that to Dragon Quest, which ran from 1986 to 1995 on a different model. In the original Dragon Quest games, gems don’t fall from enemies — they appear in treasure chests. That sounds like a small distinction. Mechanically, it barely matters. But semantically it’s enormous: a chest implies someone stored this. The gem had a prior life as a physical object that a person or entity put somewhere for a reason. It doesn’t live inside a creature. It exists in the world as a material thing.

Not every game made this trade. A small number of titles built their mineral systems around geological reality and are more compelling for it, not less. Vintage Story gives you cassiterite, galena, and sphalerite — real ore minerals with real names — and asks you to understand their properties to survive. Final Fantasy XIV includes twenty-three scientifically accurate mineral names, each tied to specific deposits in its world. Players of both games have written their own field guides. The knowledge stuck because the economy required it. These games are outliers in the database — but they’re the proof of concept that the formation story can be the economy, if someone chooses to build it that way.

The Combat Augmentation cluster in my database — the group of games most purely organized around combat power mechanics — contains 63 games. Every single one drops gems from enemies. Eight of them bother to mine.

Here’s what I actually find fascinating about all of this, and why I don’t think the goblin is the problem.

The gem drop works — works completely, works addictively, keeps you playing for forty hours before you notice — because the game built the economy before it handed you the gem. By the time that ruby drops from the shaman, you already know exactly what rubies do, why you need them, what you’re going to socket them into, how many more you need. The origin is irrelevant because you’re already inside a system where the ruby’s function is established and consequential.

That’s the mechanism. Not variable reward psychology, not dopamine hits from loot drops — those are real but they’re downstream of something more fundamental. The gem means something because an economy was running before you got it. Strip the economy and the gem is confetti.

Now consider how science presents the same object. Here is corundum. Here is chromium substitution. Here is the Mohs scale. Here are the metamorphic conditions under which corundum forms. The facts arrive correct and complete. And they don’t stick, for most people, because there’s no economy running. There’s no system you’re inside where knowing about chromium substitution changes your situation. The knowledge is true and inert.

The goblin’s ruby isn’t invisible because players are scientifically illiterate. It’s invisible because the economy that makes the ruby matter was built first, and that economy had no need for a formation story.

What the shaman was carrying, underneath the game logic, was something genuinely ancient: the intuition that power concentrates in living things, that defeating something releases what it held, that gems are the currency of that transfer. That intuition is geologically wrong and culturally universal. It’s the same logic behind every healing crystal, every protective amulet, every gemstone prescribed in ancient medical texts — the ones I study in our collection, the ones that predate the science of mineralogy by four thousand years.

Games didn’t invent this. They inherited it. And because it’s so deeply intuitive — because it maps onto something real about how we understand living things to carry value — it never needed to be questioned.

The formation story got left out because the economy didn’t require it. Twenty-five million years of metamorphic pressure, chromium substituting atom by atom into an aluminum oxide lattice, kilometers of overburden pressing down while tectonic plates shifted overhead — none of that needed to be in the game for the ruby to function as a reward.

The job isn’t to fix the goblin. It’s to build an economy where the formation story is the thing you need to survive. Where knowing how a ruby forms changes your situation in the world you’re navigating.

That economy isn’t hypothetical. It has been running for five thousand years. The Bronze Age wasn’t named for a feeling — it was named for an alloy of copper and tin, two minerals that happen to co-occur in specific geological settings, and the civilizations that figured out where those settings were didn’t just make better weapons. They built the ancient world’s first global trade networks: copper from Cyprus, tin from Cornwall and Afghanistan, meeting in furnaces from Mesopotamia to the Aegean. The people who understood the geology had the power. Everyone else bought from them.

The economy is still running. The phone in your pocket contains lithium from brine deposits in the Atacama, cobalt from the Democratic Republic of Congo, rare earth elements refined almost entirely in China — each one the product of specific geological conditions that took hundreds of millions of years to create and that exist in only a handful of places on Earth. The countries that control those deposits are renegotiating the global order right now. Knowing where spodumene forms, and why, is not an academic exercise. It is geopolitics.

In the collection I curate, we have specimens from most of those deposits. Rocks pulled from the ground before anyone knew what they would eventually power. They look, to most visitors, like rocks. Interesting rocks, perhaps — some of them are beautiful — but rocks. The economy they’re embedded in is invisible because no one built the tutorial. No one set up the crafting system before handing them over.

Meanwhile, enrollment in geology and mineralogy programs has declined for two decades. The wellness industry — selling the same minerals, the same crystals, the same objects from the same deposits — is worth tens of billions of dollars annually and growing. The sacred register migrated. The stakes stayed exactly where they always were. We just stopped making people feel them.

That’s the job. Not to fix the goblin. To build the scientific economy where the formation story is the resource — because it already is one. We just haven’t made it feel like one yet.

And don’t blame the games. They figured out what people actually want long before we did. I’m still playing Diablo II. I still need more rubies. I just really want to know where that damn shaman got his.

The data in this post comes from the Pocketful of Χtals games database, v3 — 292 games across four decades of game design, coded for mineral content, origin mechanics, role functions, and design philosophy. The database is ongoing. If you think I’m missing a game, I probably am.