Nearly everyone is taught the names of rocks. Granite, limestone, slate — a list to memorise, the way you might memorise the capitals of countries you will never visit. It is a poor way to learn stone, because a name is only a label somebody hung on the outside. It tells you nothing about what happened.
This course does the opposite. We will not begin with names, and we will not begin with a shovel. We begin by asking what a rock actually is — and the answer, once you have it, quietly explains why some ground is worth breaking, why most ground is not, and how a person with no instruments and no maps can walk a hillside and know which way to go.
Pick up a stone from a path. You are holding, most people would say, a rock. Fair enough. But look closer — properly closer, with the stone tilted to the light — and you will see it is not one substance at all. It is a crowd.
Nearly every rock is an assembly of smaller pieces packed together, and those pieces have names of their own. Each kind of piece is a mineral: a substance with one fixed recipe, whose atoms stack in one fixed, repeating pattern. Quartz is a mineral. Salt is a mineral. Gold is a mineral. There are thousands of them, but you will meet the same dozen over and over, the way you meet the same dozen trees.
A rock, then, is simply minerals gathered together. Granite is not a substance. Granite is a crowd — pale blocky feldspar, glassy grey quartz, black flakes of mica — locked into one another. Hold a piece up close and you can pick the three of them out by eye, and once you have done that once, you can never quite un-see it.
And that is the important word. Cooked. Because a recipe that was actually cooked carries the marks of the kitchen it was cooked in — how hot, how long, how roughly it was handled, whether there was water about. That is where this whole course is going, and we can now say it plainly.
Here is the sentence the rest of this course rests on. It is not difficult, and almost nobody is told it before they are handed a list of rock names to learn:
Nothing about a stone is arbitrary. Not its grain, not its layers, not the size of its crystals, not its colour, not which minerals turned up in it and which did not. Every one of those is testimony — evidence left behind by heat, pressure, water and time doing particular things in a particular order. The stone is not decorated. It is written on.
This is a stranger idea than it first sounds, so sit with it. A rock cannot lie about its past, because its past is the only thing it is made of. A crystal that grew slowly is physically a different size from one that grew quickly — not symbolically, actually. A grain that travelled fifty miles down a river is physically rounder than one that fell off a cliff last winter. The evidence is not a clue the rock happens to be carrying. The evidence is the rock.
So the skill of the miner and the prospector is not memory. It is reading. And once you can read, the ground stops being a wall of grey and starts being a page that tells you two things:
Everything from here on is that one fact, unfolded. Nothing new gets smuggled in.
If a rock is a record, then something wrote it. There are only four hands on the pen, and almost everything you will ever see in stone is one of them, or two of them arguing.
Heat. Heat is what lets atoms move. Cold rock is rigid because its atoms are locked where they stand. Warm it and they can shuffle; warm it a great deal more and they can swim, and the rock is a liquid — molten rock, which we call magma while it is still underground and lava once it reaches the surface. Everything between rigid and liquid is a range, not a switch, and most of the interesting things in geology happen inside that range.
Pressure. Go down into the earth and the weight of everything above presses in. It rises fast: a kilometre down, every square inch is carrying the load of a kilometre-high column of rock. Pressure squeezes empty space out and pushes atoms into tighter arrangements — and, this matters later, when it squeezes harder from one direction than another, it lines things up.
Water. The quiet one, and the one that makes people rich. Water carries things. It dissolves substances, moves them somewhere else, and puts them down again — and it does this far more eagerly when it is hot and under pressure. Almost every concentrated deposit of metal on this planet is somewhere that water dropped a load it had been carrying.
Time. Not a force, but a multiplier of the other three. Processes far too slow to notice — a fraction of a millimetre a year — become mountains and canyons when you let them run for a hundred million years. Human patience is the wrong instrument here. The ground is not slow; you are brief.
If a rock records its conditions, then rocks fall into groups not by how they look but by what happened to them. And there turn out to be exactly three broad histories a piece of the earth can have had. Here is the first.
Somewhere below you, rock got hot enough to melt. That melt is buoyant — liquid rock is lighter than the solid rock around it — so it rises, gathers in a chamber, and eventually stops rising because it has cooled too far to stay liquid. Then it freezes, in the exact sense that water freezes: a liquid becoming a solid. Rock born this way is called igneous, from the Latin for fire.
There are only two ways for the story to end, and they leave completely different stone behind.
One more thing to notice, because it becomes useful later. Molten rock does not sit politely in its chamber. It shoulders into cracks, drives sheets up through the layers above it, and bakes whatever it touches. Wherever a body of molten rock met colder rock there is a boundary — and boundaries are where the interesting minerals live.
The second history begins with destruction. Everything exposed at the surface is being taken apart: frost splitting it, rain dissolving it, roots levering it, wind sanding it. That is weathering — breaking rock where it stands. Then gravity, rivers, ice and wind carry the fragments away, which is erosion.
The pieces travel until whatever is carrying them slows down. And here is the key: moving water can only carry what it has the energy to carry. Slow it, and it must set down its heaviest and largest load first. Slow it further and the finer stuff drops too. Stop it altogether and even the finest mud settles out.
So sediment lands in layers, sorted by the energy of the day that delivered them. A storm drops coarse sand; the calm months afterwards lay fine mud on top; another storm, another sand bed. Bury that pile under the next few hundred metres of the same, squeeze the water out, and let dissolved minerals glue the grains together, and you have sedimentary rock — rock that was laid down.
These rocks are named by grain size, which is only another way of saying they are named by how much energy dropped them:
| The rock | Made of | Therefore it records |
|---|---|---|
| Conglomerate | Rounded pebbles set in finer material | Fast, powerful water — a river in flood, a beach in surf |
| Sandstone | Sand grains | Steady moving water or wind — a river bed, a dune, a shore |
| Shale, or mudstone | Mud and clay, too fine to see | Still water — a deep sea floor, a lake bottom, a quiet lagoon |
| Limestone | Calcium carbonate, much of it from shells and reefs | A warm shallow sea, biologically busy — this rock is largely made of corpses |
| Coal | Compressed plant matter | A swamp that buried its own vegetation faster than it could rot |
Read that table again and notice what it really is. It is not a list of rocks. It is a list of places and days, still legible, hundreds of millions of years after the water went home.
The third history takes a rock that already existed — from either of the first two families — and changes it without ever melting it. Bury it deep. Heat it. Squeeze it. Do that for long enough and the minerals inside become unstable in their old arrangement and rebuild themselves into new ones that suit the new conditions. The result is metamorphic rock, from the Greek for "changed form".
Nothing is added and nothing much is taken away. The same atoms are still there. They have simply been re-sorted, the way a room's furniture can be entirely rearranged without a single item leaving the house.
Now the part that turns this from a fact into a tool. If the squeeze comes equally from all sides, the new crystals grow in whatever direction they like and you get an even, sugary rock — limestone becomes marble, sandstone becomes quartzite. But if the squeeze is stronger from one direction than another, as it is when continents collide, then every flat, platy mineral in the rock ends up lying across the direction of the squeeze — rotated into that position, or grown there in the first place — because that is the attitude in which it is pressed least. Millions of tiny flakes all lying the same way produce a fabric.
That fabric is called foliation, from the Latin for leaf, and it is why slate splits into roofing tiles. The splitting planes are not the original sedimentary layers at all. They are a brand-new grain, imposed by the direction the mountains were built from — which means a slate roof is, quite literally, tiled along the line a continent once pushed.
The same starting mud, cooked and squeezed harder and harder, walks up a ladder:
Three histories, then. Molten and frozen; laid down in layers; cooked and squeezed. But they are not three destinations. They are three stations, and the same atoms go round and round between them for as long as the planet stays warm inside.
Granite that froze deep gets lifted by mountain-building, exposed by erosion, weathered into sand, washed to a coast and laid down as sandstone. Bury that sandstone deep enough and it becomes quartzite. Push it deeper still and it melts, and rises, and freezes into granite again. There is no first rock and no last one — only material moving between three sets of conditions, wearing whatever form the conditions demand at the time.
This is why "what kind of rock is this?" is a shallow question and "what has this been through?" is a deep one. The first has a one-word answer. The second tells you where to dig.
Sediment settles through water. Water lies flat. Therefore sediment lands flat, in sheets that spread out until something stops them. This sounds too obvious to be worth saying, and it is one of the most powerful observations anyone has ever made about the ground.
Because if layers start flat, then every tilted, folded or shattered layer you will ever see is telling you about something that happened afterwards. The tilt is not part of the original recipe. The tilt is a second chapter.
One honest qualification, since we are about to lean hard on this. A few things do build their own slopes as they are laid down — the downwind face of a sand dune, the front of a growing reef, a delta pushing out into a lake — so you will meet beds that were slanted the day they formed. They give themselves away: those slanting beds sit inside a flat-lying package, sliced off top and bottom, instead of the whole stack leaning together. The rule is about stacks, and the exceptions announce themselves.
Three rules follow, and all three are just the one fact applied honestly:
And there is a fourth, subtler reading. Suppose layers were laid flat, then tilted, then worn off level by erosion, and then new flat layers were laid down on top of the sawn-off stumps. What you see is flat beds resting on tilted beds with a clean line between them. That line is called an unconformity, and it is a gap in the record — a stretch of time with no rock to show for it, usually a very long one. It is a page torn out, and the tear itself is the evidence.
A century later, in 1788, the Scottish farmer and geologist James Hutton sailed along the Berwickshire coast with two friends and landed at Siccar Point, where near-vertical beds lie under gently sloping red sandstone — a clean, undeniable unconformity. One of his companions, John Playfair, later wrote that looking at it made the mind grow giddy at the depth of time involved. He was right to be giddy. The gap at that spot is something like sixty-five million years.
Back to the molten family, with the tool we now have. If a rock records its conditions, then the size of the crystals in a once-molten rock must record something. It records how long the melt had — and here is the mechanism, which is worth having properly, because it is not obvious.
Crystals do not appear whole. They start as a seed — a handful of atoms that happen to settle into the right pattern and stick — and then they grow outward, atom by atom, until they run into their neighbours and stop. So the final size depends on a race between two things: how many seeds get started, and how long each one gets to grow.
Cool the melt slowly and few seeds form, because the liquid stays close to the temperature at which crystals are only just stable — and each seed that does form then has an enormous stretch of time to feed. Few seeds, long growth, large crystals. Cool it fast and the melt plunges far below that temperature while still liquid, seeds appear everywhere at once, and each one is boxed in by its neighbours within hours. Many seeds, no time, tiny crystals. Cool it faster than the atoms can arrange themselves at all and you get glass — a liquid caught in the act, frozen mid-shuffle, with no pattern in it anywhere.
The fourth option there is the one that shows off the method. Some rocks contain a few big crystals floating in a mass of tiny ones. There is no way to get that from a single steady cooling. It has to be two coolings: a long, quiet spell at depth in which a few crystals grew large, interrupted by an eruption that threw the whole soup to the surface and froze the remaining liquid in days. That rock is a two-part diary, and you can read the interruption in it.
This matters for mining more than it sounds. Large bodies of slowly cooled granite are exactly where the last dregs of a melt — rich in water and in all the awkward elements that would not fit into the ordinary crystals — get concentrated and driven off into the surrounding cracks. Tin, tungsten, lithium and a great many gemstones are found around the edges of slow-cooled granite for that reason, and not around fast-cooled lava.
Now take the same reading skill to a handful of sand, because a loose grain is a record too. Two questions answer almost everything.
Is it angular or round? A freshly broken grain has sharp corners. Every mile it travels those corners get knocked off, because corners stick out and take the hits. So angular grains have not gone far — the source is close, probably a cliff or a slope of loose broken rock above you. Well-rounded grains have been in transport a long time, which means the source could be an entire county away.
Are the grains all one size, or mixed? Moving water sorts by weight. Give it enough time and steady conditions and it will deliver grains of one size to one place, because everything heavier dropped earlier and everything lighter stayed in suspension. So a bed of uniform grains means patient, steady sorting — a beach, a dune, a river bed. A jumble of every size mixed together means something dumped its whole load at once without sorting: a flash flood, a landslide, or the melt-out from a glacier, which sorts nothing at all because ice does not care about weight.
And there is the practical payoff, which prospectors have used for as long as there have been prospectors. If you find fragments of something interesting lying loose in the ground — miners call such loose fragments float — then their angularity tells you roughly how far you are from the source, and gravity tells you which way it came. Float travels downhill and downstream. Walk up the slope, testing as you go, until the float stops. Somewhere just above that last piece is where it is coming from.
With one exception, and it is the same fact behaving consistently rather than an escape clause. Ice is not water and does not obey gravity in the same way: a glacier pushes material along its own line of travel, uphill included, and drops it wherever it melts. So in country that has been glaciated — the northern United States and most of Canada — the trail does not lead up the slope. It leads back along the direction the ice came from, which you can read off the scratches and the shape of the ground the ice left behind. Same rule, different carrier.
A mineral has one fixed recipe and one fixed atomic pattern. Therefore its properties are fixed too, and therefore they can be tested — with a pocket knife, a fingernail and a piece of unglazed tile. This is the oldest laboratory there is, and it still works.
Hardness. Which scratches which. In 1812 a German mineralogist named Friedrich Mohs simply lined ten minerals up in order, from talc, which your fingernail dents, to diamond, which nothing else touches, and numbered them one to ten. It is not a measurement of anything else — it is purely an order of scratching. What makes it useful in the field is the household equivalents: your fingernail sits at about two and a half, a steel knife blade or a pane of window glass at about five and a half, a steel file at about six and a half. Quartz, at seven, scratches glass easily. Feldspar, at six, scratches glass with effort. Calcite, at three, is scratched by a copper coin.
Streak. Drag the mineral across the back of an unglazed tile and look at the powder it leaves. Colour on the outside of a mineral is the least trustworthy property there is — quartz comes in every colour of the rainbow depending on trace impurities — but the colour of the powder is far more constant. Haematite can look silver-grey, black or red, yet its streak is always a rusty red-brown, and that is what names it.
Heft. Pick it up, and pick up an ordinary stone of the same size in the other hand. Most rock weighs about two and a half times what the same volume of water would, and after a week of handling stone your hands learn that as "normal". Galena — the mineral nearly all the world's lead comes out of — is about seven and a half times water, and feels shockingly, almost comically heavy for its size. Gold is more than nineteen times. Heft is a cruder test than the others, and it is very often the first one that tells you something is worth a closer look.
How it breaks. Some minerals split along flat planes, because their atomic pattern is weakly bonded in certain directions — that is cleavage. Mica peels into sheets. Galena and rock salt break into little cubes. Calcite breaks into leaning boxes. Quartz has no cleavage at all and instead breaks in smooth curved shells, like a struck bottle. How a mineral breaks is the atomic pattern showing through at a size you can hold in your hand.
Two quick chemical tells. A drop of vinegar on limestone or marble fizzes faintly, because the acid is attacking the calcium carbonate those rocks are built from and letting its gas go; on granite or sandstone it does nothing at all. And a magnet will jump to magnetite from a centimetre away, which is one reason the black sand in a stream bed is always worth a second look.
If a rock is a record of its conditions, then every mineral in it formed under those same conditions. Therefore the minerals in a rock are not a random draw. They are the guest list of one particular event — and guest lists repeat.
Certain minerals need the same heat, the same pressure, the same chemistry, the same water. Where those conditions occurred, all of them formed together; where the conditions did not occur, none of them did. So finding one member of a group is genuine evidence that the others are somewhere nearby. That group is called an assemblage, and it is the most useful idea in prospecting.
The friends that are easy to see and reliably keep company with something valuable have a name: indicator minerals, sometimes called pathfinders. They are not the prize. They are the signpost, and the whole art is knowing which signpost points at what.
The same logic runs in reverse, and that saves you far more time than it costs. If you know the conditions a mineral requires, and the ground in front of you plainly never experienced them, then it is not there. No amount of hopeful digging will change that. Ruling ground out is not defeat — it is most of the job, and it is why a person who can read stone covers in an afternoon what a person who cannot will spend a summer on.
Two words first, because the table leans on both and neither has been earned yet.
Sulfide is the family name for a mineral in which a metal is joined to sulfur. Galena is lead and sulfur. Pyrite — fool's gold — is iron and sulfur. Lead, zinc, copper, silver and nickel nearly all travel in this form, because sulfur is what they readily bond with in hot underground water. The important habit of sulfides is that they are unstable in air and rainwater: at the surface they rot, and they leave rust and stains behind them. That is why the top of a rich body of rock so often looks nothing like the rock below it. Silica is the other word, and it is simpler — silicon joined to oxygen, the stuff quartz is made of, and the commonest material in the crust.
Now the classic assemblages, each one an illustration of the same principle. Do not memorise the table. Read down the right-hand column and notice that every single entry is a statement about conditions.
| If you find | Look for | Because |
|---|---|---|
| Milky quartz veins with rusty pitting | Gold | Hot water carried silica and gold together and dropped them in the same crack; the pits are where iron sulfide has since rotted away |
| Arsenic minerals | Gold | Arsenic travels with gold in the same hot fluids and is far more abundant — a wide target that points at a narrow one |
| Galena, the heavy lead mineral | Zinc, and usually silver | Lead and zinc sulfides form under near-identical conditions, and silver atoms sit directly inside the lead crystal. Most of the world's silver is not mined for its own sake at all — it comes out as a by-product of other ores, and lead and zinc are the largest single source of it |
| Green or blue stains on a rock face | Copper | Copper sulfides weather at the surface into bright green malachite and deep blue azurite. Humans have followed that green stain for at least six thousand years |
| A rusty, spongy, honeycombed cap of iron oxide | A sulfide body underneath | Rainwater rots sulfide minerals from the top down and leaves the iron behind as rust. Miners call that cap a gossan, and it is the loudest signal in prospecting |
| Purple-red garnet and bright green chrome diopside in loose gravel | Diamond | Those minerals come up from the same great depth as diamond, in the same rare volcanic pipes. They survive weathering, they are far more numerous, and they are much easier to see |
| Limestone hard against a body of granite | Copper, tungsten, iron, garnet | The granite baked and chemically attacked the limestone at the contact, and that reaction zone is a natural trap for metals |
Look at the diamond row again, because it is the whole argument in miniature. Nobody sensible walks a glacial gravel plain looking for diamonds. They sieve for garnets, which are hundreds of times more common, brightly coloured and easy to spot — then follow the trail of garnets upstream, or back along the direction the ice came from, until it narrows to a point. The diamond gets found by not looking for it.
Here is the problem with metals, and it is worth stating baldly. The earth's crust does contain gold, copper, lead and the rest — but spread so thinly through ordinary rock that mining ordinary rock would cost more than the metal is worth, for ever. Gold in average crust runs at a few parts per billion — crush a whole tonne of ordinary rock and the gold in it amounts to a few thousandths of a gram, a speck you would need a lens to find.
So a mine is never a place where metal exists. A mine is a place where metal has been concentrated — gathered out of an enormous volume of poor rock into a small volume of rich rock. Rock that has been enriched far enough to be worth digging, in a form the metal can actually be got out of, is called ore. Note that this is an economic word as much as a geological one: the same rock stops being ore the day the price falls. Something had to do the gathering. Almost always, that something was water.
Water buried in rock and heated by a body of molten rock below becomes a different substance from the water in your kettle. Hot, under crushing pressure, and usually salty, it will dissolve metals it could not touch cold. It then rises, because hot water is buoyant, and it rises by the only routes available — cracks and faults, which act as plumbing.
And as it climbs, everything that let it carry the metal is taken away. It cools. The pressure drops. It may boil. And a fluid that cannot hold its load must set that load down — on the walls of the crack, crystal by crystal, out of the flow. Do that for long enough and the crack fills with quartz and with metal minerals, growing inward from both walls. That filled crack is a vein, and it is what hard-rock mining follows.
Most of what fills a vein is worthless — quartz, usually, or calcite. Miners have a word for the worthless filling: gangue, said "gang". It is worth knowing because it is most of what you dig, most of what you carry, and most of what you crush, and the metal is the minority passenger the whole way.
Three consequences, all of them practical, all of them just the picture above taken seriously:
The second thing water does needs no heat at all, and it is the reason the poorest person with the simplest equipment has always had a fair chance at gold.
Once a vein reaches the surface, weathering destroys it like anything else. The quartz breaks up, the sulfide minerals rot away, and whatever was durable and heavy — gold, platinum, tin ore, gemstones — is released as loose grains and washed into the nearest stream.
Now the sorting rule from Module 10 does the work for you. Moving water carries what it has the energy to carry, and gold is about seven times heavier than the sand travelling beside it. So the instant the water slows, the gold stops and the sand keeps going. It does not travel far. And it works its way downwards through the gravel with every flood, because heavy grains settle through shaken material, until it reaches something it cannot pass. A deposit gathered this way is called a placer, and it is nature having already done the crushing, the washing and most of the sorting.
Therefore you know exactly where to look, without anybody having to tell you:
One last piece of reasoning, and it is exactly the sort of thing this course exists to make possible. Rivers move. A stream that flowed here a million years ago may now be running half a mile away and eighty metres lower, having left its old gravel bed stranded up the valley side as a bench or terrace. That old bed still has its gold in it, sitting on its old bedrock floor, nowhere near any water at all. Nobody has to teach you to look there. It follows.
Everything so far assembles into a sequence. It is not a secret, and it has not changed much since Agricola wrote it down, because it is simply the one fact applied in a sensible order — widest question first, narrowest last.
Notice that six of those seven steps involve no digging at all. That is not caution, it is efficiency. Digging is the most expensive way there is to acquire information, and every step above exists to make sure that when you finally do dig, you are digging to confirm something you already have good reason to believe.
And one duty that is not geology at all but belongs here anyway. The minerals under a piece of land are owned by somebody — sometimes the surface landowner, often not, and the rules differ by country, state and county. Get permission, in writing where it matters, and find out who holds the mineral rights before you invest a season in a hillside. Old workings almost always belong to somebody, and so does the liability if you fall into one.
Reading rock tells you where value sits. The same reading, done honestly, tells you where the ground will kill you — and that is the more valuable half of the skill, because the ground does not negotiate and does not give its warnings twice.
Every hazard below is the one fact turned round: the rock records what happened to it, and some of what happened to it has left it unable to hold itself up, or has left the air inside it unbreathable.
If you followed all of that, you were not handed a list of rock names. You were handed one fact and its consequences, and here is the whole chain in a single breath.
A rock is a record of the conditions that made it. Therefore nothing in a stone is arbitrary, and everything in it is evidence. Therefore there are three families because there are three histories — once molten, laid down in layers, or cooked and squeezed. Therefore layers mean it was laid down flat, and anything tilted has been moved since, and anything cutting across is younger than what it cuts. Therefore big crystals mean slow cooling deep down and tiny ones mean fast cooling at the surface, and a rock with both means two coolings with an interruption between them. Therefore rounded grains have travelled and angular grains have not, and jumbled grains were dumped in a hurry. Therefore minerals that formed together turn up together, so one of them tells you to look for another, and the common visible one is how you find the rare hidden one. Therefore water is the great mover — dissolving metal at depth and dropping it where it cools, which is what a vein is, and sorting it by weight where it slows, which is what a placer is. Therefore reading rock is how you know where to dig, and, more importantly, where not to — because the same reading that says value is here also says this will not hold itself up and this air will not do.
That is what a first-principles education buys. Not more answers. Better questions, and the ability to work the answer out yourself on a hillside nobody has surveyed, with nothing in your hands but a hammer and your eyes.
"We find no vestige of a beginning — no prospect of an end."
— James Hutton, on the rock record, 1788
The Guild of the Deep · Course One of the Twelve Guild Courses · Iron Roots Supply
Back to the Guild Hall for the rest of the courses.
If this course sends you to one book, send yourself to Agricola. De Re Metallica, 1556 — the Hoover translation of 1912 is free to read online, and the woodcuts alone are worth the evening.