Digital TownsCrystalsShop crystals
  1. Crystals & Minerals
  2. How they form
  3. Agate Banding

How Agate Bands Form, and What Geologists Still Cannot Explain

By Digital Towns Crystals · Last reviewed October 8, 2026

Short answer: most agate forms inside old gas bubbles in basalt and andesite, where silica carried by water is deposited and slowly turns into fibrous chalcedony and quartz. Isotope and inclusion studies put the temperature below 200 °C and most likely below 100 °C. Bands either copy the cavity wall (fortification) or lie flat under gravity. A 2020 review of 250 years of research found temperature is the only settled question: where the silica comes from, and what the first deposit looked like, remain unproven.

What is a band, physically?

A band is a layer of silica that differs from its neighbors in crystal texture, porosity, water content or trace impurities. Agate is the banded form of chalcedony, which Minerals.net describes as quartz in microcrystalline form that does not occur in visible crystals, and in a thin section under a microscope the layers resolve into zones of fibrous chalcedony alternating with zones of coarser granular quartz. Pigment, usually iron oxide, sits in some layers more than others, which is why bands read as color stripes to the eye. Geology.com calls the banding within a cavity a record of water chemistry change.

The silica is not pure quartz either. In their review Agate Genesis: A Continuing Enigma, Terry Moxon and Galina Palyanova describe how Heaney and Post examined more than 150 samples of agate, chalcedony, chert and flint and found moganite, a second silica mineral, ranging from 23 to 2 wt% between samples. Moganite was only accepted as a separate mineral in 1999. Its share falls as agates age, a detail that turns out to be useful for dating them.

Bands also differ in porosity, and that is something a buyer can see. Geology.com notes that agate is porous and readily accepts dye, and that most of the spectacularly colored agate in the trade has been dyed. Dye soaks into the more porous bands and skips the tight ones, so a dyed stone exaggerates the natural banding instead of hiding it.

How does a banded agate form, step by step?

The sequence below is the version most geologists would sign. Each step is supported by evidence; the parts still argued over are flagged.

  1. A lava flow traps gas. As basalt cools, dissolved water vapor and carbon dioxide come out of solution and are frozen in place as bubbles. The Minnesota Department of Natural Resources describes this for Lake Superior agate, whose host lavas erupted more than a billion years ago when the continent began to split apart. The DNR adds that most of those gas pockets were small, about the size of a pea, which is why most Lake Superior agates are small too.
  2. Water moves through the cooled rock. Groundwater or hydrothermal water percolates through cracks and altered rock and reaches the empty cavities. Often a thin clay or mineral skin coats the bubble wall first.
  3. Silica arrives a little at a time. Silica is only slightly soluble in water. Moxon and Palyanova cite measurements of 100 to 140 ppm at 25 °C and 300 to 380 ppm at 90 °C, with little change between pH 1 and 9. Filling a cavity with solid silica therefore needs a great many pore volumes of water, which is one reason the supply question is hard. The chemistry of that water also drifts over time; Dexter Perkins notes in his mineralogy text that the concentric layers in a geode have slightly different compositions because the water composition changed a bit as crystallization occurred.
  4. The deposit reorganizes. The first silica is thought to be poorly ordered. The review sets out a transformation sequence, first argued for agate by Landmesser and documented in chert and sinter: amorphous silica, then cristobalite or tridymite, then chalcedony with moganite, then granular quartz. Each step is denser than the last, so the material shrinks as it matures.
  5. Quartz crystals may finish the job. Many agates end in a hollow lined with clear quartz points, or a solid core of crystalline quartz. Heaney's model, summarized in the review, explains this: chalcedony grows while the solution is rich in weakly polymerized silica, and once it is dilute enough to be undersaturated for amorphous silica but still saturated for quartz, prismatic quartz takes over.
  6. Gravity writes some bands flat. Geology.com summarizes the two geometries: silica deposits in concentric layers around the cavity walls, or in horizontal layers building up from the bottom. Where silica settled in a pool at the bottom of the cavity, the layers built up horizontally. The Minnesota DNR attributes these perfectly straight, parallel bands to solutions crystallizing inside the cavity under very low fluid pressure. In thin sections of Brazilian agates, Moxon and Palyanova show that wall-lining and horizontal bands in one stone can belong to different generations, so they did not form in a single episode.

The same mechanism works in cracks rather than bubbles, giving vein agate, and in some limestones and fossil wood, though the review stresses there is no reason agates in different settings must form the same way.

How hot was it, and how do scientists know?

Temperature is the best constrained number in agate science, and it comes from several independent methods.

Method Agates studied Formation temperature reported
Oxygen and hydrogen isotopes (Fallick et al.) Scottish Devonian and Tertiary 50 °C
Oxygen isotopes (Harris) Namibia 120 °C
Harris data re-examined (Saunders) Namibia 39 to 85 °C
Aluminum content of central quartz Various 50 to 200 °C
Fluid inclusions in central quartz Northeast Russia, andesite-basalts about 100 to 170 °C

All figures are as summarized by Moxon and Palyanova. Their conclusion is that agate in these basaltic hosts formed below 200 °C and most likely below 100 °C, and that any proposal above 200 °C has to be discounted. The physical evidence for the ceiling is that heating agate drives off water irreversibly: loosely bound water goes below 200 °C, tightly bound water and silanol groups up to 800 °C, with maximum mass loss at 850 °C. Agate that had formed hot would not still hold that water.

That also means the agate formed well after the lava itself had cooled. The bubbles were already cold, empty holes when the silica arrived.

Is an agate as old as the rock around it?

Usually close to it, but not always, and the evidence is one of the more elegant parts of the field. Moxon's X-ray work tracked the size of quartz crystallites in agates from 17 regions with host ages from a few million to over a billion years. Crystallites grow steadily in agates from hosts up to about 60 million years old, barely change over the next 210 million years, grow again for about 30 million years, and stop after roughly 300 million years. Moganite content drops in step during the first 60 million years.

Because those properties track age, they can expose an agate that is younger than its host. The review's SEM work on agate from Rio Grande do Sul, Brazil, where the host basalt is about 135 million years old, found crystallite size and other properties pointing to a formation age of about 26 million years, with cristobalite still present. Most agates, the authors conclude, formed at about the same time as their host rock, which supports late hydrothermal water or hot springs as the silica carriers. The Brazilian case shows the exceptions are real.

What is still unknown about agate banding?

Moxon and Palyanova frame any full explanation as three questions: the formation temperature, the source of the silica, and the nature of the first deposit. They state that temperature is the only one that has been answered.

Where does the silica come from?

Leaching of the host rock is the most popular answer. Weathered basalt releases silica and the iron that colors the bands. But the review notes that some Scottish agates from heavily altered hosts lack the expected iron color, while others from similar rock are vividly red and yellow. Late hydrothermal solutions and hot springs are the other candidate, and the near-contemporaneous ages favor them for many deposits. Which source dominates, and whether it differs between deposits, is open.

Was the first deposit a gel?

This is the most repeated claim in popular accounts, and it is unproven. By 1915 the chemist Liesegang had produced convincing agate-like rings by diffusing metal salts through silica gel, and a 1994 hypothesis by Pabian and Zarins invoked an oscillating chemical reaction in a gel. The review lists the objections: gels crack and dehydrate to powder within weeks, a gel cannot migrate through solid host rock, and natural groundwater does not carry the high concentrations of silica, iron or aluminum needed to make one inside the cavity. The alternative is repeated deposition of amorphous silica that later crystallizes. Neither has been demonstrated end to end.

How do the bands get in?

Some agates show an infiltration canal, a channel that seems to breach the outer layers. In 1849 the geologist Noeggerath pointed to these canals as the entry route for later silica, asking how solutions could cross the first impermeable layers otherwise. Liesegang challenged the idea in 1915, and the canals are still read by some authors as entry points and by others as exits for water. Self-organization models, such as one published by Wang and Merino in 1990, try to explain rhythmic banding without repeated pulses of new fluid. No laboratory has yet grown true agate banding under natural conditions, so all of these remain hypotheses.

What can you read in a specimen?

Once you know the process, a polished face becomes a record.

  • Fortification bands that run parallel to the outer skin show where the cavity wall was. Minerals.net's agate page explains that each agate takes its pattern from its unique cavity shape.
  • Flat, parallel bands across one side show which way was down in the lava flow when they formed. Tilt the stone so those bands are horizontal and you are holding it the way it grew.
  • A clear quartz center marks the late, dilute stage when prismatic quartz took over from chalcedony. Minerals.net notes that a cavity is often only partly filled, leaving a hollow lined with crystalline quartz, which is how agate ends up as the outer wall of many geodes; see our geode guide.
  • A channel through the outer bands may be an infiltration canal.
  • Bands that glow differently under UV. Minerals.net notes agate is commonly fluorescent, usually green or white, and that some bands can fluoresce more strongly than others, which shows chemistry varied from layer to layer.

Two misconceptions are worth dropping. Bands are often described as yearly layers like tree rings, but none of the dating methods above counts bands, and bands in a single stone can come from separate episodes. And bright, saturated blue, pink or purple bands are rarely natural. Minerals.net says bright neon colors are rarely natural and that Brazilian agate is especially often dyed. Our guide to dyed agate covers the checks.

Choosing banded agate to see the formation story

For learning the process, choose pieces where the bands are sharp and the structure is complete: a full fortification pattern, a visible horizontal set, or a quartz center. Thin, backlit slices show translucency differences between bands that a solid piece hides. Natural earthy colors (gray, white, brown, red, ochre) are what basalt agates actually produce; treat electric colors as dyed until shown otherwise. For more on agate's varieties, see the agate guide and the Botswana agate page for one of the finest-banded localities.

  • Agate slice tea light holder: thick natural slices of 438 to 482 g, about 5 inches across. Light from behind makes the translucent chalcedony bands stand out against the denser ones.
  • Banded Agate Crystal Pyramid Polished Natural Stone: a 248 g pyramid with a 76 mm base in earthy tones; flat cut faces cross the bands at different angles, so you can follow one layer around the stone.
  • Agate cube with cut corner, 416 g: a 55 mm cube with concentric banding; the extra corner face gives a fourth cross section through the same layers.
  • Agate Geode, Druzy: a 6.25 x 4 x 4.25 inch geode with an open center, showing the agate wall giving way to drusy quartz, step 5 above.
  • Botswana banded agate tumble: an extra-small tumbled stone at an entry price, with the fine parallel banding Botswana is known for.

Digital Towns Market

Banded agate in stock

See all crystals and minerals

Frequently asked questions

How long does it take for agate to form?

No one has measured it directly, because agate forms underground and has never been grown with natural banding in a lab. Related silica sinters take around 50,000 years to complete their conversion to quartz, and agate crystallinity keeps changing for tens of millions of years after formation. A nodule is best thought of as forming over long geological time, not in a season.

What temperature does agate form at?

Isotope studies of Scottish agates gave about 50 °C, and a re-examination of Namibian data gave 39 to 85 °C. Reviews conclude agate in basalt forms below 200 °C and most likely below 100 °C. Heating agate above about 200 °C drives off water irreversibly, which rules out a hot origin.

Why are some agate bands straight and others curved?

Curved, concentric bands follow the cavity wall and are called fortification or wall-lining banding. Straight bands formed where silica collected in a pool at the bottom of the cavity and built up horizontally under gravity. Both can occur in the same agate, and thin sections show they can belong to different episodes of growth.

Does agate form from a silica gel?

That is a popular explanation but not an established one. Gels can mimic agate patterns in the lab, but researchers point out that gels dehydrate within weeks, cannot travel through solid rock, and need silica concentrations natural water rarely carries. Whether the first deposit is a gel or amorphous silica is one of the main open questions.

Is agate the same age as the volcanic rock it is found in?

Often roughly, but not always. Studies of crystallite size and moganite content suggest most agates formed close to the time of their host. Some Brazilian agates in basalt about 135 million years old show properties indicating formation around 26 million years ago, so silica can arrive long after the lava cooled.

Sources