Why Do Rocks Come in Three Families?

I came to geology by way of earthworms. I was standing in the vegetable patch, wondering where all the soil had come from, and the answer turned out to be "rocks, mostly, and a very long time". Once you start asking that question you can't really stop. Every pebble in the path becomes a small mystery.

The good news is that nearly every rock you'll meet fits one of three life stories. Learn the plots, and you can start to read the clues.

Three life stories

Geologists sort rocks into three families, and the sorting is about how a rock got to be the way it is, not what it's made of.

• Igneous rocks cooled and hardened from molten rock.
• Sedimentary rocks are made of bits (grains of sand, mud, shell, and so on) that settled out, piled up, and were pressed and glued into stone.
• Metamorphic rocks were already rocks, then were changed by heat and pressure without melting.

That's the whole scheme. "Igneous" comes from a word for fire, "sediment" from a word for settling, and "metamorphic" from Greek for changing shape. So you can think of them as cooled, settled and changed.

(Three is partly a convenience. Rocks don't care about our filing system, and I'll come back to that. But it's a very useful filing system.)

Family one: cooled from melt

Start with something familiar: anything that goes from runny to solid when it cools. Honey in the cupboard on a cold morning. Wax dripping down a candle.

Molten rock does the same thing, except it's far hotter. Melt forms deep underground, and then one of two things happens. It may push up and erupt, cooling quickly on the surface as lava. Or it may stall underground and cool slowly over a very long time. Rocks from the first route are called extrusive (or volcanic). Rocks from the second are intrusive (or plutonic).

This matters because of a simple idea about crystals. As melt cools, atoms lock into orderly crystal patterns, and those crystals need time to grow.

Here's a thought experiment. Imagine a hall where people are being asked to find a seat in tidy rows. If you give them an hour, they'll settle into big, neat blocks. If you give them thirty seconds, they'll grab whatever seat is nearest and you'll get lots of tiny, scrappy clumps. Crystals are like that. Slow cooling gives big crystals. Fast cooling gives tiny ones.

So when you look at an igneous rock, the crystal size tells you something about where it cooled:

• Granite has crystals big enough to see without a lens, interlocked like a jigsaw. That's slow cooling, deep underground. You'll find it in places like Abel Tasman and Fiordland, where uplift and erosion have since uncovered rock that once sat far below the surface.
• Basalt is dark and fine-grained, with crystals too small to see easily. That's quick cooling at or near the surface. Auckland's volcanic cones are basaltic.
• Pumice is so full of gas bubbles that it can float. The bubbles are frozen-in froth from a gassy eruption.
• Obsidian is volcanic glass. It cooled so quickly that the atoms had no time to form crystals at all.

One common misunderstanding: people often assume "igneous" means "volcano". It doesn't. Granite never saw daylight while it was molten. It's igneous but not volcanic in the everyday sense.

Family two: settled and pressed

Now for a different plot, one I understood much better after I tried a trick with my garden soil.

Put a couple of handfuls of soil in a clear jar, fill it with water, shake it, and leave it overnight. By morning, you'll see layers. The coarse sand falls first and sits at the bottom, the finer silt settles on top of that, and the tiniest clay particles take longest, sometimes leaving the water cloudy for a day or more. (I did this expecting a muddy mess and got something rather beautiful.)

That's the beginning of a sedimentary rock. Over time, rivers, wind, waves and ice break rock into pieces and carry them off. When the water slows down, the pieces drop, coarse ones first. Layer after layer builds up, and the ones at the bottom get squashed under the weight of those above. Water seeping through leaves dissolved minerals behind, and these act like a cement. The geologists' name for this squashing and gluing is lithification, which just means "turning to stone".

The main types follow the kind of material:

• Sandstone is cemented sand grains.
• Mudstone is cemented mud and clay.
• Limestone is mostly calcium carbonate, often from shells and tiny sea creatures that piled up on the seafloor.
• Greywacke, which you'll see all over New Zealand, is a hard, grey, gritty rock made of sand and mud that was buried and compacted until it was tough. Many of our riverbeds are full of it.

The clues to look for are layers (called bedding), grains you can see or feel (rub a fingertip across a sandstone and it can feel like sandpaper), and sometimes fossils, which survive in sedimentary rocks far more readily than in the other two families because the rock is formed from gentle settling rather than heat.

Little trick for limestone: a drop of vinegar on a scratched or powdered patch will fizz gently, because the carbonate reacts with the acid. It's not a dramatic fizz, so look closely. I got this tip from a neighbour who'd been poking at rocks for years, and I thought she was pulling my leg until I tried it.

Family three: changed without melting

The third plot is my favourite, because it has a quiet twist. The rock changes while staying solid.

Think about a potter's kiln. A lump of soft clay goes in, is baked at high temperature, and comes out as hard ceramic. It never melted, but it's not clay any more. Its minerals have changed. Metamorphic rocks follow the same idea, with the earth as the kiln.

Bury a rock deep enough and two things rise: temperature and pressure. Deep down, or when plates push against each other and crumple the crust, the original minerals are no longer stable. They slowly rearrange, and sometimes new minerals grow in their place, all in the solid state. (If it gets hot enough to melt completely, it's no longer metamorphic. Cool that melt, and you've got an igneous rock.)

The pressure is the clever part. If a rock is squeezed hard from one direction (as happens where plates collide), flat or needle-shaped minerals such as mica tend to grow pointing across the squeeze. Over a large volume of rock, that gives you aligned minerals, which look like bands or a layered sparkle. Geologists call this foliation, from the Latin for leaf. It's a nice name for something that often looks like a stack of thin pages.

You can follow a story with increasing heat and pressure. Start with mudstone. With modest change you get slate, which splits into smooth sheets (think old roof tiles). More change gives a silky, shiny rock called phyllite, then schist, where you can see glittering mica flakes. Push harder still and you may get gneiss (pronounced "nice"), with coarse, contrasting bands. A lot of the schist in Otago started life as sediment rather like greywacke, which I find rather wonderful. Same ingredients, different journey.

Not all metamorphic rocks are banded. If the original rock is made of one kind of mineral, there may be nothing to line up:

• Marble is limestone that has recrystallised. You can see it in Takaka Hill country, and it fizzes with vinegar, just like limestone.
• Quartzite is sandstone that has been fused into a tough, glassy mass.

A quick guide to guessing

Here's the method I use. It isn't foolproof, and I'll come to that. Take a rock you've found and ask three questions.

1. Do I see interlocking crystals, like a jigsaw? If so, it's probably igneous (or possibly metamorphic). Big crystals suggest slow cooling, tiny ones suggest fast cooling, and bubbles or a glassy look point to a volcanic origin.

2. Do I see layers, rounded grains, or fossils? That points to sedimentary. A hand lens helps (a cheap one is plenty). If the grains are rounded and held together by cement, you're probably looking at sedimentary rock. If you can rub off a bit of grit, even better evidence.

3. Do I see bands, a silky shine, or flaky sparkle all pointing the same way? That's the hallmark of metamorphic.

A few things can mislead you, which is why I'm still learning here:

• Layers don't always mean sedimentary. Volcanic ash falls and lava flows can stack up in layers, and metamorphic rocks can be banded too. The difference is often in the texture. Sedimentary layers show grains settling, while metamorphic bands show crystals growing.
• Crystals don't always mean igneous. Marble is made of interlocking crystals, and it's not igneous at all.
• Weathering hides things. The outside of a rock may be stained or crusted. If you can find a fresh broken face (safely, and ideally from a loose piece rather than a rock face), the inside is much easier to read.

Even the practised get stumped

I'll admit, my first guesses are still wrong a fair bit of the time. I once spent a whole walk convinced a layered grey rock was sedimentary, only for someone who knew better to point out the glinting mica that told me it was schist. I felt a bit silly, but she just smiled and said she gets caught out too.

That's not politeness, either. Professional geologists often slice a rock paper-thin and look at it under a microscope, because some rocks simply can't be identified by eye. Some rocks are in-between cases, such as a sedimentary rock that has just started to change. So if you're stumped, you're in good company. A rock labelled "not sure" is a perfectly good scientific answer.

A loop, not a ladder

Here's the idea that made the three families click for me. It's tempting to picture them as a ladder: igneous at the bottom, then sedimentary, then metamorphic at the top. But that's not what happens.

James Hutton, the eighteenth-century Scottish naturalist, was among the first to see that rocks are constantly recycled over vast spans of time. We now call this the rock cycle. It's not a conveyor belt with one direction. Think of it more like a road network with lots of junctions and shortcuts:

• Any rock can be weathered into sediment, which can become sedimentary rock.
• Any rock can be buried and changed into a metamorphic rock.
• Any rock can melt, and when the melt cools, it's igneous.
• Metamorphic rock can be uplifted, worn down, and turned into sediment again.
• An igneous rock can be uplifted and worn away without ever passing through the "middle" stages.

There's no fixed order. A bit of granite might become sand, then sandstone, then schist, then melt, then granite again, or it might skip half of those steps. The pace is slow: we're talking millions of years, driven by heat from inside the Earth and by wind, water and ice at the surface.

It's also why the families are a handy filing system and not a hard rule. Each name describes a chapter in a rock's story, not its whole biography.

Try it this week

You don't need much: a hand lens if you have one, a jar for the soil trick, a small bottle of vinegar and a bit of curiosity. Pick up a few pebbles from your garden, the beach, or a riverbank, and look for crystals, layers and bands. Don't worry about getting it right. Write down what you see before deciding what it is, and enjoy being wrong now and then. I certainly do.

Thank you for reading, and happy rock-watching.

And here's something I still wonder about. The pebble in your pocket has been through many chap