Why do mineral names end in -ite?
We machine-read the September 2026 IMA list: 5,988 of 6,195 names end in -ite, about 96.7%. The rules governing mineral naming do not require the ending at all.
We machine-read the September 2026 IMA-CNMNC master list and extracted 6,195 of its 6,239 valid mineral species names. Of those, 5,988 end in -ite, about 96.7%. The published rules do not require it: a mineral may be named for a locality, a person or a property. Quartz, spinel and realgar were named before anyone described a crystal system.
In short
- It is a convention, not a rule. The IMA's published Procedures and Guidelines on Mineral Nomenclature set out what a mineral may be named after — a locality, its discoverer, a person prominent in mineralogy, or a property of the mineral — and say nothing whatever about how the name should end.
- 96.7% is the actual figure. We counted it rather than repeating it: 5,988 of the 6,195 names we could machine-read from the September 2026 master list end in -ite.
- 91 valid species end in nothing standard at all, and they are the famous ones: quartz, corundum, spinel, zircon, rutile, realgar, turquoise, schorl, grossular, pyrope, chrysocolla, analcime. The oldest names are the exceptions, because they predate the convention.
- -lite is not a separate ending. 531 names end in -lite, but that is -ite attached to a root ending in l, usually from the Greek lithos, stone. Natrolite is soda-stone-ite.
- The rules that do exist are about avoiding confusion, not about the suffix — no two names close enough to be confused, no commercial sponsors, no capital letters or spaces inside a name, and permission required before naming a mineral after a living person.
| Ending | Count | Share | Examples | What it usually signals |
|---|---|---|---|---|
| -ite | 5,988 | 96.7% | Fluorite, bournonite, goethite, brochantite | The default. No information beyond 'this is a mineral' |
| -ine | 36 | 0.6% | Aegirine, almandine, celestine, cuspidine, devilline | Mostly older names, many Greek or French in origin |
| -ane | 17 | 0.3% | Glaucophane, hedyphane, allophane, stilpnomelane, rhabdophane-(Ce) | Greek -phanes, appearing or shining, in most of them |
| -ide | 17 | 0.3% | Auricupride, cuproauride, palladoarsenide, diopside | Mostly modern alloy and intermetallic names, built like chemistry |
| -ore | 15 | 0.2% | The pyrochlore group, clinochlore, baileychlore, diaspore | Almost entirely -chlore, from the Greek for green |
| -ium | 15 | 0.2% | Aluminium, cadmium, iridium, osmium, palladium, tellurium | Native elements, carrying the element name itself |
| -ase | 8 | 0.1% | Anatase, euclase, dioptase, orthoclase, clinoclase, periclase | Greek klasis, breaking — most of them describe cleavage |
| -yte | 4 | 0.1% | Baryte, eudialyte, feroxyhyte | A spelling variant, and why British usage writes baryte |
| None of these | 91 | 1.5% | Quartz, corundum, spinel, zircon, rutile, realgar, turquoise | Names that entered mineralogy before the convention existed |
What the nomenclature rules actually say
The governing document is the International Mineralogical Association’s Procedures and Guidelines on Mineral Nomenclature, 1998, published in The Canadian Mineralogist. It is the authority every new mineral proposal is judged against, and it is freely available.
Its section headed “Choice of a new mineral name” says this: a mineral is commonly named for the geographical locality of its occurrence, for the discoverer of the mineral — although not if he or she is the author — for a person prominent in the field of mineralogy, or for a particular property of the mineral.
That is the whole of it on what a name is based on, and there is no mention of -ite anywhere in the section. The rules that follow are all about avoiding problems rather than about form: names must be sufficiently different from existing ones to prevent confusion, and the document cites celadonite and caledonite, mallardite and malladrite, rhodesite and rhodizite and rhodusite as the confusions already in the literature. Naming minerals after commercial organisations that have made no contribution to mineralogy is to be discouraged. A name taken from a person must lose its spaces and internal capitals — mcnearite, not mcNearite; joesmithite, not joe smithite. Permission is required before naming a mineral after a living person.
So -ite is what mineralogists do, not what mineralogists are told to do. It survives because it is unambiguous: an unfamiliar word ending in -ite is recognisably a mineral name, in any language that uses the Latin alphabet, which is exactly the property a global nomenclature needs.
The count, and how we did it
Rather than repeat a figure, we counted. Method, stated so it can be checked or disputed: we downloaded the September 2026 update of the IMA-CNMNC master list as published at cnmnc.units.it, converted the PDF to text, and extracted the name column from every row carrying an IMA status code and a year. That yielded 6,195 distinct names of the 6,239 species the document states it contains — the 44 not captured are rows whose layout defeated the extraction, not species we are disputing. Chemical and rare-earth suffixes such as -(Ce), -Ca and -(Fe2+) were stripped before classifying the ending, so that chabazite-Ca counts as an -ite name.
The result: 5,988 of 6,195, or 96.7%, end in -ite. Everything else put together is 207 names, and no other ending reaches 40.
Two details worth having. 531 names end in -lite, but that is not a separate suffix: it is -ite on a root ending in l, most often the Greek lithos, stone. And 30 end in -phyllite, from the Greek for leaf, which describes a habit — as in the apophyllite group, where the six species are separated by chemistry rather than by that shared root. Our page on apophyllite covers that group.
The 91 that end in nothing standard, and why they are the famous ones
The list of species whose names end in none of the recognisable mineral suffixes is short, and reading it is like reading a list of the minerals everyone can name.
Quartz. Corundum. Spinel. Zircon. Rutile. Realgar. Turquoise. Schorl. Grossular. Pyrope. Analcime. Chrysocolla. Chrysoberyl. Chrysotile. Diaspore. Epidote. Harmotome. Calomel. Massicot. Litharge. Natron. Trona. Lime. Cinnabar’s relatives metacinnabar and hypercinnabar. The native elements — antimony, arsenic, bismuth, nickel, platinum, silicon, tungsten.
The pattern is chronological. These names entered European mineralogy from Greek, Latin, Arabic and German mining dialect before a naming convention existed to regularise them, and the Handbook of Mineralogy records the derivations. Realgar is from the Arabic rahj al ghar, for powder of the mine. Schorl is from the Old German Schörl, of uncertain derivation, perhaps meaning impurities. Quartz is of obscure origin in the Middle Ages, first applied to gangue in Saxony, Germany. That last one is a good illustration of how old these words are: the name of the commonest mineral on the planet cannot be traced with confidence. The -ite convention firmed up in the eighteenth and nineteenth centuries alongside systematic mineralogy, and it was applied to what came afterwards.
There is a practical collector consequence. A name that does not end in -ite is usually an old name, and an old name usually means a mineral known and traded long before the modern species concept — which is exactly the group where varietal names, trade names and species names get confused. Amethyst is a variety of quartz and not a species at all; the guidelines say so explicitly, noting that existing varietal names such as amethyst and kunzite are not species and do not come under the Commission’s jurisdiction. Our species reference keeps to the species names.
What a name can and cannot tell you
A mineral name is a label, not a description, and almost nothing can be inferred from it. A name ending in -ite tells you it is a mineral and nothing else. This is worth internalising, because collectors routinely over-read names.
Where a name does carry information, it is one of four things. A locality, as in mendipite from the Mendip Hills or bassetite from Wheal Basset. A person, as in goethite, bournonite, brochantite and torbernite — four of the species we published this week, all named for individuals. A chemical element, as in the cupro-, ferro-, plumbo- and magnesio- prefixes, which do mean what they say. Or a property, as in crocoite from the Greek for saffron, or lepidocrocite from the Greek for scale and thread.
Where a name actively misleads: pseudomalachite is not malachite and is not even a carbonate, fool’s gold has no formal standing, and marcasite in a jewellers’ window is pyrite. The Handbook of Mineralogy records marcasite as a word of Arabic or Moorish origin early applied to pyrite and other substances, which is how that one happened. Our pages on pseudomalachite and marcasite deal with both.
And modern names increasingly carry a suffix that does matter. Fluorapophyllite-(K) and hydroxyapophyllite-(K) differ in an anion; tetrahedrite-(Fe) and tetrahedrite-(Zn) differ in a cation; and bare “tetrahedrite” is no longer a valid species at all — the September 2026 list carries seven suffixed tetrahedrites and no unsuffixed entry. An old label reading tetrahedrite is a group determination. The rest of the collecting guides cover what to do about labels like that, and our wanted list is where to tell us what you are looking for.