Turquoise
Turquoise is a copper aluminium phosphate, hardness 5 to 6, density 2.86, crystals to only 3 mm. Howlite at hardness 3.5 is the substitute a coin will find.
Turquoise is a hydrated copper aluminium phosphate, hardness 5 to 6, measured density 2.86, triclinic. Crystals are rare and reach only about 3 mm; the material collectors and jewellers handle is cryptocrystalline, nodular or massive. It forms two series, one with chalcosiderite and one with planerite.
In short
- Crystallised turquoise reaches about 3 mm. That single number disposes of a great deal of misdescription: any blue mineral offered as a turquoise crystal over a centimetre across is something else. The crystallised locality of record is the Bishop mine at Lynch Station, Virginia.
- Hardness 5 to 6 against howlite at 3.5 is the working test. A steel point or even a copper coin will mark howlite and will not mark turquoise. Dyed howlite is the commonest substitute in the trade and this separates it in a second.
- Density does very little. Turquoise 2.86, howlite 2.53 to 2.59, variscite 2.57 to 2.61, faustite 2.92. The whole group sits between 2.5 and 2.9, so hydrostatic weighing cannot be relied on alone.
- Faustite is the zinc analogue and it is apple-green, which is why green material sold as turquoise may be correctly named something else. Variscite, an aluminium phosphate with no copper, is also green and is a separate species entirely.
- There is a Cornish occurrence. The Handbook of Mineralogy names the Bunny mine at St Austell and other Cornish localities, and chalcosiderite — turquoise's iron end-member — has its type locality in Cornwall.
| Mineral | Formula | Hardness | Density | Colour | The separator |
|---|---|---|---|---|---|
| Turquoise | CuAl6(PO4)4(OH)8(H2O)4 | 5 to 6 | 2.86 | Sky-blue, bluish green, apple-green | — |
| Howlite | Ca2SiB5O9(OH)5 | 3.5 | 2.53 to 2.59 | White when natural; dyed blue to imitate turquoise | Hardness. A coin marks it, turquoise resists a steel point |
| Variscite | Al(PO4)(H2O)2 | 4.5 | 2.57 to 2.61 | Pale green to emerald-green; rarely blue | Colour and the absence of copper |
| Faustite | ZnAl6(PO4)4(OH)8(H2O)4 | 5.5 | 2.92 | Apple-green | Composition — it is the zinc analogue, not a fake |
| Chrysocolla | (Cu2-xAlx)H2-xSi2O5(OH)4 · nH2O | about 2 to 4 | 1.93 to 2.4 | Blue, blue-green, green | Density, which is far lower, and a waxy to earthy break |
| Chalcosiderite | CuFe6(PO4)4(OH)8(H2O)4 | 4.5 | 3.22 | Green | The iron end-member of the same series; Cornish type locality |
The howlite test, and why hardness beats every other check
Dyed howlite is the substitute you will actually meet, in beads, cabochons and tumbled material, and it is separated by a single property. Howlite is hardness 3.5. Turquoise is 5 to 6. That gap is enormous in practice: a steel knife point at about 5.5 will bite into howlite immediately and will skate on good turquoise.
Test on the back, on a drill hole, or on an area that will not show, and press lightly — the point of the test is to see whether the point bites, not to gouge the piece. Our note on testing mineral hardness properly covers the technique and the ways it goes wrong, and how to tell if a crystal is dyed covers the dye itself, which on howlite tends to concentrate in the fractures and in the dark veining.
The complication, and it is a real one: much commercial turquoise is stabilised. Porous low-grade material is impregnated with resin to make it workable, and a resin-filled surface can resist a point more convincingly than the mineral underneath deserves. Stabilisation is a legitimate and near-universal treatment; failing to disclose it is not. If a piece is being sold as natural, ask, and ask in writing.
Why the density column is nearly useless here
Hydrostatic weighing is the standard fallback when hardness is inconclusive, and on this group it barely works. Read the column: 2.86 for turquoise, 2.53 to 2.59 for howlite, 2.57 to 2.61 for variscite, 2.92 for faustite. The entire set fits inside a range of about 0.4, and the measurement error on a small irregular specimen weighed at home is a meaningful fraction of that.
The one place density does pull its weight is chrysocolla, at 1.93 to 2.4. That is a genuinely different number — a chrysocolla nodule is noticeably light in the hand for its size, where turquoise feels ordinary. Chrysocolla is also much softer, at about 2 to 4, and breaks with a waxy or earthy surface rather than a smooth conchoidal one.
The general point is worth stating because it generalises beyond this group. A test is only useful when the values it returns are further apart than the error in measuring them. On turquoise and its look-alikes, hardness clears that bar and density does not.
Crystallised turquoise, and the 3 mm ceiling
The Handbook of Mineralogy describes turquoise crystals as rare, steep pinacoidal, with {010}, {110} and {001}, to 3 mm. Everything else is fine granular to cryptocrystalline, nodular, in globular crusts, veinlets or massive.
Three millimetres is the fact worth carrying, because it settles a whole category of claim on its own. The crystallised material of record comes from the Bishop mine at Lynch Station, Campbell County, Virginia, and there are good crystals from Katonto, north of Kolwezi in Katanga, Congo. Large crystals are reported from the Itatiaiuçu iron mine southwest of Belo Horizonte in Minas Gerais, Brazil, and crystals from Narooma in New South Wales.
The great commercial sources are different again and are almost all massive: Ma'dan, 45 km northwest of Neyshabur in Iran, the historical source and the one that gave the mineral its name — from the French Turquie, because Iranian material reached Europe through Turkey; the Cerrillos and Burro Mountains districts in New Mexico; Mineral Park, Morenci and the Globe-Miami district in Arizona; Lander County in Nevada; and Hubei and Shaanxi in China.
The Cornish connection, through chalcosiderite
Turquoise is not a mineral anyone associates with Britain, but the Handbook names the Bunny mine at St Austell, and elsewhere in Cornwall, in its distribution list.
The stronger British link runs through the other end of the series. Turquoise forms a series with chalcosiderite, CuFe6(PO4)4(OH)8(H2O)4, which is the same structure with iron in place of aluminium — and chalcosiderite's country of first description on the IMA list is the United Kingdom, with 1814 as the year. Our chalcosiderite page covers the Cornish material.
So a British collector building a phosphate suite has a legitimate route into turquoise that does not involve buying Iranian or Arizonan cabochons: the Cornish end of the series, in small, unglamorous, entirely genuine specimens. That is the sort of thing a wanted list is for, since it is defined by locality and species rather than by appearance, and it does not come onto the market often enough to wait for it to appear in front of you.