How to identify a mineral specimen
Six tests you can do on a desk, in the order that eliminates the most possibilities fastest — and the point at which to stop guessing.
Identify a mineral specimen by habit and lustre first, then test hardness against the Mohs scale of 10 minerals: calcite is hardness 3, fluorite 4, apatite 5, quartz 7. Take a streak on unglazed porcelain, count the cleavage directions, judge the heft against quartz, and check fluorescence. Locality narrows the answer faster than any single test.
In short
- Habit and lustre first. Crystal shape and how the surface handles light rule out more species in ten seconds than any tool.
- Hardness next, on an inconspicuous point: fingernail ≈2.5, copper ≈3.5, steel blade and window glass ≈5.5. Fluorite is 4, quartz 7.
- Streak on unglazed porcelain: hematite streaks red-brown however black the crystal looks; pyrite streaks greenish-black, not gold.
- Count cleavage directions: galena 3 at right angles, calcite 3 not at right angles, fluorite 4, quartz none at all.
- Heft is a real test. Galena at density 7.6 feels absurd for its size next to quartz at 2.65.
| Hardness | Reference mineral | Everyday equivalent |
|---|---|---|
| 1 | Talc | Marked by a fingernail with no effort |
| 2 | Gypsum | Fingernail ≈ 2.5 |
| 3 | Calcite | Copper ≈ 3.5 |
| 4 | Fluorite | Scratched by a steel blade |
| 5 | Apatite | Steel blade and window glass ≈ 5.5 |
| 6 | Orthoclase | Hardened steel file ≈ 6.5 |
| 7 | Quartz | Scratches glass and steel readily |
| 8 | Topaz | Scratches quartz |
| 9 | Corundum | Scratched only by diamond |
| 10 | Diamond | Scratches everything |
Do the non-destructive tests first
Every identification test that involves scratching costs the specimen something, so work through the free ones first. Habit — the shape the crystals take — is the strongest single clue: cubes, hexagonal prisms, rhombs, blades, radiating sprays, botryoidal crusts. Then lustre: metallic, adamantine, vitreous, resinous, silky, earthy. Then colour, which is the weakest clue of all and the one beginners lean on hardest.
Colour deserves the warning. Fluorite occurs purple, green, amber, blue, pink and colourless; quartz does much the same. Neither can be identified by colour, and a great many misidentifications begin with somebody deciding a purple crystal must be amethyst.
Use a hand lens throughout. A 10x loupe shows crystal faces, cleavage traces, glue lines and coatings that are invisible at arm's length, and it costs less than a specimen.
Hardness, done without wrecking the specimen
Hardness is the most useful destructive test and the one most often done badly. Two rules: scratch the specimen with the reference, not the other way round, and do it on a back edge or an inconspicuous point that will never face outwards. Never test a display face. Never test a crystal you would be upset to mark.
Then check what you have actually done. A soft reference dragged across a hard mineral leaves a metallic smear that wipes off; that is not a scratch. A real scratch is a groove you can feel with a fingernail and still see after cleaning.
The everyday references are enough for most work: a fingernail at about 2.5, copper at about 3.5, a steel blade and window glass at about 5.5, and a hardened file around 6.5. Modern coins are usually plated steel rather than copper, so use a known piece of copper wire instead if you want the 3.5 point to mean anything.
Streak, cleavage and heft
Streak is the colour of the mineral powdered, taken by drawing it across unglazed porcelain, and it is far more constant than the colour of the crystal. It is decisive for the metallics: hematite gives red-brown no matter how black and mirror-bright the crystal is, pyrite gives greenish-black rather than gold, and galena gives lead-grey. Minerals harder than about 6.5 will not streak — they scratch the plate instead, which is itself information.
Cleavage is where a mineral parts along flat planes, and counting the directions is close to diagnostic. Galena cleaves in three directions at right angles and breaks into little cubes. Calcite cleaves in three directions not at right angles, giving rhombs. Fluorite cleaves in four directions, and the octahedral scars are unmistakable. Quartz has no cleavage at all and fractures conchoidally, in curved shell-like surfaces.
Heft is the test nobody writes down and everybody uses. Density separates lookalikes fast: galena at about 7.6 and barite at about 4.5 both feel far heavier than they look next to quartz at 2.65. Pick up known specimens until your hands are calibrated.
Fluorescence, magnetism and acid
A shortwave and longwave UV lamp answers questions nothing else will. Fluorescence is caused by trace activators rather than by the species itself, so it is not diagnostic on its own — but combined with a locality it often is, and it is the fastest way to sort some suites in a drawer.
A magnet separates the strongly magnetic minerals in seconds; magnetite is the one most often mistaken for something more exciting. A drop of dilute hydrochloric acid on an inconspicuous spot distinguishes the carbonates: calcite fizzes briskly, dolomite only reluctantly and mostly when powdered. Rinse and dry afterwards, and never do it on a display face.
When to stop and get it analysed
A large number of species simply cannot be separated by eye, and pretending otherwise is how wrong labels enter collections and stay there for decades. The lead members of the apatite supergroup — pyromorphite, mimetite and vanadinite — overlap in colour, habit and association. Roselite and wendwilsonite from Bou Azzer are visually the same mineral. Many of the copper arsenates need analysis.
Where the identification changes what the specimen is worth or what it means, the honest options are analysis or a hedged label. "Pyromorphite group" is a better label than a confident guess, and a collection full of hedged labels is more useful than one full of confident errors.
Local geological societies, mineral clubs and university geology departments are the usual route to someone with a microscope and an opinion. Most will look at something interesting.