Lava that cools and hardens becomes igneous rock. The specific type depends on cooling speed and chemistry: rapid chilling at the surface produces fine-grained basalt, while slow cooling underground produces coarse-grained granite. When lava cools and hardens, the result is extrusive igneous rock—most commonly basalt, which makes up roughly 90 percent of all volcanic rock on Earth (USGS).

How cooling rate locks in texture

The single biggest control on what rock forms is how fast heat escapes. Lava erupted into air or water loses heat quickly, and crystals get almost no time to grow. The result is a fine-grained or glassy rock where individual minerals are too small to see without magnification.

The physics is straightforward. Crystal growth requires atoms to migrate through melt and attach to a growing mineral face. At cooling rates above roughly 1 to 10 degrees Celsius per minute, migration can't keep pace, so you get many tiny crystals instead of fewer large ones. Quench lava in seawater and you can freeze it into volcanic glass—obsidian—with no crystals at all.

Cooling speed also controls where the crystals end up. Slow cooling lets early-formed minerals sink or settle, producing layered or patchy textures. Fast cooling traps everything in place, giving a uniform, fine-grained fabric.

Here is the plain version: fast cooling equals small crystals, slow cooling equals big crystals, and the chemistry of the melt decides which minerals appear. Both the speed and the chemistry are baked into the final rock.

Cooling times and crystal sizes, quantified

Cooling time spans an enormous range, and texture tracks it closely. The table below pairs typical settings with approximate cooling timescales and the textures and rock types they produce.

Setting Approximate cooling time Crystal size Typical rock
Submarine eruption Seconds to minutes Glassy, no crystals Obsidian, pillow basalt
Thin lava flow on land Days to weeks Under 1 mm Fine basalt
Thick lava flow interior Months to years 1–5 mm Basalt with coarser core
Shallow intrusion (dike, sill) Years to decades 1–10 mm Diabase (dolerite)
Deep magma chamber 10,000–1,000,000 years 1–10 cm or more Granite, gabbro

Crystal size roughly scales with the logarithm of cooling time, so a thousandfold increase in cooling time doesn't give a thousandfold larger crystal—it gives perhaps a tenfold jump. This is why deep plutons capped at a few centimeters rather than meters (USGS). The relationship is real but sublinear, which surprises people who expect giant crystals from very slow cooling.

Temperature thresholds matter too. Basaltic melt sits around 1,100 to 1,250 degrees Celsius at eruption. It stays fully liquid until it drops to roughly 1,000 degrees Celsius, then a burst of crystal nucleation begins. Below about 900 degrees Celsius the melt is mostly solid. The window where crystals can grow is narrow in fast-cooling settings and wide in slow ones.

Why chemistry changes the answer

Cooling rate explains texture, but chemistry explains which minerals and which rock name you get. Silica content, in particular, sets the melt's viscosity and its mineral menu.

High-silica melts (about 65 to 75 percent silicon dioxide) are thick and polymerized. They resist crystal growth and tend to erupt explosively, producing rhyolite at the surface and granite underground. Low-silica melts (about 45 to 52 percent silicon dioxide) are fluid and runny, producing basalt at the surface and gabbro underground.

Two rocks can share the same texture but carry different names because of chemistry. A fine-grained dark rock and a coarse-grained dark rock are both low-silica, but one is basalt and the other is gabbro. Text and chemistry together determine the classification.

Water content adds a wrinkle. Dissolved water lowers the melting point and makes melt less viscous, which can actually speed crystal growth. When water escapes during eruption, the melt stiffens abruptly and can shatter into glass or pumice. The counterintuitive detail: adding water to a melt can make the resulting rock coarser, not finer, because it lowers viscosity and speeds atom migration.

Edge cases that break the simple rule

The fast-equals-fine rule has exceptions that trip up students and field geologists alike.

Large crystals in a fine matrix: a porphyritic rock has big crystals floating in fine groundmass, meaning slow growth in a chamber followed by fast cooling after eruption. Glassy rocks that aren't obsidian: tachylyte forms when basalt quenches so fast that iron and silica stay disordered. Vesicular textures: gas bubbles trapped during rapid cooling create holes, as in pumice and scoria, with no bearing on crystal size. Devitrification: glass can slowly crystallize over millions of years, blurring the original cooling story. Hydrothermal alteration: hot fluids can recrystallize and overprint the original texture entirely. Anorthosite: plagioclase-rich intrusions can host crystals up to a meter, far beyond what cooling time alone predicts, because crystal settling concentrates them.

One professional caveat deserves emphasis. Texture records the last thermal event, not the whole history. A rock can cool slowly, then get reheated and quenched by a later intrusion, leaving no trace of the first episode. Geologists call this a reset. Reading rock texture as a simple thermometer is a mistake; it is a record of the final cooling path.

Myths vs. measured evidence

Several popular ideas about lava rock don't survive contact with the data.

The first myth is that all lava hardens into the same rock. In reality, basalt dominates because most eruptions are basaltic, but andesite, dacite, and rhyolite are common where subduction or continental crust is involved.

The second myth is that cooling always produces crystals. Glassy rocks exist whenever cooling outruns nucleation, which happens in water or air contact.

The third myth is that lava rock is always dark. Color tracks mineral content, not cooling rate. Rhyolite is often pale, and obsidian is black only because of its glassy structure and iron oxides.

The fourth myth is that underground cooling never produces fine rock. Thin dikes and sills can chill quickly against cold country rock, producing fine-grained margins and coarse interiors in the same body.

The fifth myth is that once lava hardens it stops changing. Weathering, hydrothermal fluids, and burial metamorphism all alter igneous rock over time.

Field evidence supports these corrections. Mapping shows chilled margins on every intrusion, glassy rims on every submarine flow, and porphyritic textures wherever magma stalled before erupting.

Reading rock in the field: practical steps

If you want to identify a lava-derived rock yourself, work through these steps in order.

Check grain size with a hand lens. If you can see individual mineral grains, it cooled slowly enough to grow crystals. Measure the largest crystals. Grains under 1 mm suggest days to weeks of cooling; grains over 1 cm suggest thousands of years or more. Test for glass. A glassy luster and conchoidal fracture point to quenching, not slow cooling. Note color and density. Dark, dense rock suggests low silica (basalt family); pale, low-density rock suggests high silica (rhyolite family). Look for vesicles. Abundant bubbles mean gas was trapped during rapid cooling. Estimate the cooling environment. Surface flows chill fast; intrusions chill slow; contact zones show chilled margins.

For safety, treat fresh lava and active flow fields as hazardous. Temperatures above 700 degrees Celsius persist in flow interiors for months, and surface crusts can be thin enough to break underfoot. Never walk on recent flows without local guidance. When collecting samples, check land access rules and park regulations first.

Limits of what texture can tell you

Texture narrows the cooling story but doesn't fully resolve it. You cannot read exact cooling rates from grain size alone, because composition, water content, and nucleation history all shift the result. Two rocks with identical textures can have cooled at very different rates if their melts differed.

Absolute ages require radiometric dating, not texture. A fine-grained basalt could be 10 years old or 10 million years old; only isotope measurements distinguish them. Radiometric dating using potassium-argon or argon-argon methods can pin eruption ages to within thousands of years on young flows and better on older ones.

There are open questions too. How nucleation actually begins in natural melts is still an active research area. Small mineral seeds, impurities, and pressure changes can all trigger crystallization, and geologists cannot always predict which one dominated in a given body. Laboratory experiments on melt cooling run for months, not millennia, so extrapolation carries uncertainty.

What is solid: lava that cools and hardens forms igneous rock, cooling rate sets texture, and chemistry sets mineral composition. Both together give you the rock name. Everything beyond that rests on careful field observation and measurement, not assumption.