Species · Micas

Muscovite

Muscovite is a mica with crystals recorded to 4.5 m and 77 tonnes. The Handbook gives it two hardness values — 2.5 along the cleavage and 4 across it.

Formula KAl2(Si3Al)O10(OH,F)2Hardness 2.5 along {001}, 4 acrossDensity 2.77 to 2.88Largest crystals 4.5 m and 77 tonnesCleavage {001}, perfectLaminae Flexible and elasticFound with Quartz, tourmaline, topaz, feldspar

Muscovite is the common potassium mica of granite pegmatites, where it grows with quartz, tourmaline and topaz, density 2.77 to 2.88, with perfect {001} cleavage splitting into flexible, elastic sheets. The Handbook of Mineralogy records crystals to 4.5 m and 77 tonnes, and gives two hardness values for one species: 2.5 along the cleavage and 4 across it.

In short

  • The Handbook of Mineralogy gives muscovite two hardness numbers — 2.5 parallel to {001} and 4 perpendicular to it. A scratch test on a mica returns a different answer depending on which way you push, and both answers are correct.
  • Crystals reach 4.5 metres and 77 tonnes, which makes muscovite one of the largest-crystallising minerals known. Phlogopite, the magnesium mica, is recorded larger still at 10 m and 270 tonnes. Both come from granite pegmatites.
  • Elastic, not merely flexible. A bent muscovite lamina springs back. Molybdenite, graphite and chlorite all cleave into flexible sheets that stay bent. That single test separates the micas from every other sheet mineral in a collection.
  • Lepidolite is not a valid species. It is absent from the September 2026 IMA list; the valid lithium micas are trilithionite and polylithionite, both settled by a 1998 special procedure. This is the fourth such case this site has recorded.
  • Twinning forms six-pointed stars, on the composition plane {001} with [310] as twin axis. Euhedral crystals are uncommon, so most collection specimens are books or plates on matrix, often as an accessory on tourmaline, beryl or topaz rather than as sharp individuals of their own.
The micas a collection actually holds, and the one name that is not a species
NameFormulaHardnessDensityIMA status, September 2026
MuscoviteKAl2(Si3Al)O10(OH,F)22.5 along, 4 across2.77 to 2.88Valid; approved 1998 s.p.
PhlogopiteKMg3Si3AlO10(F,OH)22 to 32.78 to 2.85Valid; forms a series with biotite
LepidoliteK(Li,Al)3(Si,Al)4O10(F,OH)22.5 to 42.8 to 2.9Not on the list — a series name
TrilithioniteKLi1.5Al1.5(Si3Al)O10F2Valid; redefined 1998 s.p., Sweden
PolylithioniteKLi2AlSi4O10F2Valid; approved 1998 s.p.
BiotiteA series name, not a species

One species, two hardness numbers, and why that is not an error

Read the Handbook of Mineralogy sheet for muscovite and the physical properties line says: Hardness = 2.5 ∥ [001]; 4 ⊥ [001]. Two values, for two directions, in one mineral.

The structure explains it. Muscovite is built of silicate sheets held together across the stack by potassium. Drag a point along the sheets and you are separating layers held by a weak bond; drive it across them and you are cutting the sheets themselves. The first is a hardness-2.5 operation and the second a hardness-4 one.

This is the second documented case of directional hardness on this site. The first is kyanite, covered in our note on testing hardness properly, where the difference runs from about 4.5 to 7 along different axes of the same crystal. Kyanite is the famous example and muscovite is the common one — a collection is far more likely to contain a mica than a kyanite, so this is the case a scratch test is more likely to meet.

The working rule: a hardness test on a platy mineral is only meaningful if you record which direction you tested, and on a mica the useful direction is across the plates, because along them you are testing the cleavage instead.

Flexible and elastic — the one test that sorts every sheet mineral

Several minerals in a collection cleave into thin sheets: the micas, molybdenite, graphite, chlorite, and gypsum in its foliated form. They look alike in the hand and the confusion is routine.

The Handbook distinguishes them with one word. Muscovite's laminae are recorded as flexible and elastic; tough, and phlogopite's as flexible and elastic. Molybdenite's are flexible, not elastic, and graphite's flexible but not elastic; sectile.

The test takes a second and needs nothing. Bend a thin lamina at the edge of the specimen and let go. A mica springs back. Molybdenite and graphite stay bent. No other property separates that group so cleanly, and it costs one expendable flake rather than a scratch on a face.

Density is the confirmation if you want one, and the gap is enormous: muscovite 2.77 to 2.88 against molybdenite at 4.62 to 4.73. The method is in our note on measuring specific gravity at home.

Lepidolite is not a species, and that is now a pattern

Search the September 2026 IMA-CNMNC master list for lepidolite and there is no entry. The Handbook sheet says why, in four words at the end of its polymorphism note: a group name. The valid lithium micas are trilithionite, redefined by a 1998 special procedure from Sweden, and polylithionite, approved under the same procedure.

This is the fourth time this site has run into the same thing. Apophyllite resolved into fluorapophyllite-(K) and its relatives; tetrahedrite into a suffixed series; stilbite into stilbite-Ca and stilbite-Na; and now lepidolite into trilithionite and polylithionite. Biotite went the same way and is also absent as a species.

The consequence for a label is the same in every case and it is not that the label is wrong. An old specimen marked “lepidolite” has been determined to a series, not misidentified, and relabelling it to a suffixed species without analysis would be claiming a determination nobody made. Our note on labelling and cataloguing sets out how to record a name that has since been split, and the honest form is to keep the original and add the current position underneath.

The largest crystals of anything, and what a collection actually holds

Muscovite's crystal-size record — 4.5 m and 77 tonnes — is among the largest recorded for any mineral, and phlogopite's 10 m and 270 tonnes is larger still. Both come from granite pegmatites, where slow growth in an open cavity is possible on a scale nothing else allows.

What a collection contains is different in kind. The Handbook opens its distribution paragraph with the honest sentence Uncommon in euhedral crystals, and that is the practical fact. Most muscovite is a book, a plate on matrix or a rosette, not a terminated crystal, and the material worth a cabinet is usually muscovite as an accessory on something else — on tourmaline from Minas Gerais, on beryl and topaz from Pakistan, or the fine crystals the Handbook names from near Shelby in North Carolina and Pennsbury in Pennsylvania.

The exception collectors chase is the star: twinning on {001} about [310] produces six-pointed stellate groups, and a sharp one is a genuinely uncommon specimen. Our note on mineral twinning covers why these forms arise.

We hold no catalogue and nothing here is offered. If you have old pegmatite material with its original labels, the wanted list sets out what we look for, and our species notes cover the tourmaline, beryl and topaz muscovite arrives on.

Questions

What is the hardness of muscovite?
Two values, and both are correct. The Handbook of Mineralogy gives hardness 2.5 measured parallel to {001} — along the cleavage — and 4 measured perpendicular to it. A scratch test drawn along the plates separates layers; one driven across them cuts the silicate sheets. If you record a hardness for a mica, record the direction with it or the number means nothing.
Is lepidolite a real mineral?
Not as a species. It is absent from the September 2026 IMA-CNMNC master list, and the Handbook of Mineralogy calls it a group name. The valid lithium micas are trilithionite and polylithionite, both settled by a 1998 special procedure. A specimen labelled lepidolite is identified to a series, which is a real determination and not a mistake — it simply is not a species name.
How do I tell muscovite from molybdenite?
Bend a flake. Muscovite springs back because its laminae are elastic; molybdenite stays bent because its are flexible but not elastic. Colour and density confirm it: muscovite is colourless to brown and 2.77 to 2.88, molybdenite is lead-grey, metallic and 4.62 to 4.73 — more than half again as dense. Molybdenite also leaves a bluish-grey streak, and muscovite a white one.
Why does mica flake off my specimen?
Because the {001} cleavage is perfect and hardness along it is 2.5, so the plates separate under handling pressure that would not mark most species. Mica on matrix is best handled by the matrix, kept out of tissue and cotton wool that catch under the edges of plates, and never cleaned with a brush drawn across the plate edges. The damage is not recoverable, as a detached lamina cannot be reseated.