Is staurolite a garnet?
No. Staurolite, Fe2+2Al9Si4O23(OH), is monoclinic and not one of the 32 garnet-supergroup species. It grows with almandine, at the same hardness, 7 to 7.5.
No. Staurolite, Fe2+2Al9Si4O23(OH) on the IMA list, is monoclinic and is not one of the 32 garnet-supergroup species in the IMA garnet report. Per 12 anions it has 1 large cation, 4.5 aluminium and 2 silicon; a garnet needs 3, 2 and 3. It grows with almandine at the same hardness, 7 to 7.5, but reads 14.65 to 14.78 g in water for 20.00 g against almandine’s 15.37 g.
In short
- The verdict. Grew et al., American Mineralogist 98, 785–811 (2013) define the garnet supergroup as all minerals isostructural with garnet: 32 approved species, general formula {X3}[Y2](Z3)φ12, most cubic in space group Ia3d. A script finds the word staurolite nowhere in the report’s text. The Handbook of Mineralogy staurolite sheet gives monoclinic, pseudo-orthorhombic symmetry.
- The three-two-three count. Per 12 anions a garnet carries 3 X, 2 Y and 3 Z cations. Staurolite’s IMA formula, Fe2+2Al9Si4O23(OH), has 24 anions; halved, that is 1 iron, 4.5 aluminium and 2 silicon. The sheet’s Black Mountain, New Hampshire analysis gives 0.91, 4.36 and 2.00. almandine, Fe2+3Al2(SiO4)3, is exactly 3, 2 and 3.
- Where the count fails, documented. Passing it proves nothing. Grew et al. exclude the tetragonal almandine-pyrope phase, which has garnet’s stoichiometry but not its structure. Vesuvianite’s IMA formula works out at 2.92, 2.00 and 2.77, and it is not a garnet either. The count can rule a mineral out; only the structure rules it in.
- Why they are confused. The staurolite sheet’s association line opens with almandine, the almandine sheet’s includes staurolite, and the International Gem Society staurolite article records garnet among staurolite’s inclusions. Both are 7 to 7.5 in hardness, both run from reddish brown to black on their sheets, both streak white or nearly so, and both are island silicates.
- Choose the test by the specimen. On a crystal, shape: prisms and 60° or 90° twins against dodecahedra and trapezohedra. On a loose clean piece, the balance: 14.65 to 14.78 g in water for 20.00 g against almandine’s 15.37 g and spessartine’s 15.23 g. On a cut stone, the polariscope: staurolite is biaxial, garnets isotropic.
| Test | Staurolite | Almandine | Other garnet sheets | Does it choose? | Wrong for (documented failure) |
|---|---|---|---|---|---|
| Three-two-three count (per 12 anions) | 1, 4.5, 2 (IMA); 0.91, 4.36, 2.00 (Black Mountain) | 3, 2, 3 | 3, 2, 3 for the six sheets’ ideal formulas | Yes, rules out | Passing proves nothing: vesuvianite 2.92, 2.00, 2.77; the tetragonal almandine-pyrope phase |
| On the 32-species list (Grew et al.) | No; the word does not occur | Yes | Yes: pyrope, spessartine, grossular, andradite, uvarovite | Yes, by definition | Needs the name to be right |
| Crystal system and shape | Monoclinic, pseudo-orthorhombic prisms to 12 cm; 60° twins common, 90° crosses less common | Cubic; dodecahedra or trapezohedra to 1 m | Cubic | Yes, on sharp crystals | Rounded grains and detrital material; two tetragonal garnets (Grew et al.) |
| Optics | Biaxial (+); α 1.736 to 1.747, γ 1.748 to 1.761 | Isotropic, anomalously biaxial; n 1.830 | n 1.714 (pyrope) to 1.887 (andradite) | Yes, with a polariscope | Anomalously biaxial garnets; grossular 1.734 sits by staurolite’s α |
| 20.00 g piece in water | 14.65 to 14.78 g (D 3.74 to 3.83) | 15.37 g (D 4.318) | Spessartine 15.23 g; pyrope 14.42 g; grossular 14.44 g | Against almandine and spessartine | Andradite 14.74 to 14.87 g and uvarovite 14.69 to 14.75 g overlap |
| Hardness | 7 to 7.5 | 7 to 7.5 | 6.5 to 7.5 | No | Every specimen |
| Streak and colour | White to greyish; dark reddish brown to blackish brown | White; deep red, brownish red, black | Varied | No | Brown and black specimens of both |
| Structure class | Orthosilicate (sheet’s reference); nesosilicate (Tulane) | Orthosilicate; nesosilicate | Orthosilicates | No | Every specimen: both are island silicates |
Is staurolite a garnet? The definition says no
A garnet is a mineral with the garnet structure, and staurolite does not have it. Grew et al., American Mineralogist 98, 785–811 (2013), the IMA garnet report, say the garnet supergroup includes all minerals isostructural with garnet regardless of which elements occupy the sites, so it spans several chemical classes: 32 approved species, with the general formula {X3}[Y2](Z3)φ12, where X, Y and Z are dodecahedral, octahedral and tetrahedral sites. Most are cubic, space group Ia3d; two OH-bearing species, henritermierite and holtstamite, are tetragonal.
Staurolite on the sheet and the list. The September 2026 IMA list gives staurolite as Fe2+2Al9Si4O23(OH), G, 1792, country unknown, and two more staurolite-root rows, magnesiostaurolite and zincostaurolite. The Handbook of Mineralogy staurolite sheet writes (Fe2+,Mg)2Al9(Si,Al)4O20(O,OH)4, monoclinic and pseudo-orthorhombic, space group C2/m. A script over the saved text of the garnet report finds the word staurolite 0 times.
Same class, different structure. The Tulane notes list garnets, the Al2SiO5 minerals and staurolite among the common nesosilicates, and both the staurolite sheet and the six garnet sheets used here cite Deer, Howie and Zussman’s volume 1A, orthosilicates. Being an island silicate is what staurolite and garnet share; it does not make staurolite a garnet.
The three-two-three count, worked
The count: per 12 anions a garnet carries three X cations, two Y and three Z. That is Grew et al.’s general formula, {X3}[Y2](Z3)φ12. Almandine, Fe2+3Al2(SiO4)3 on the IMA list, fits exactly: 3 iron, 2 aluminium, 3 silicon to 12 oxygen. Two hydrous garnets on the IMA list keep the 3 and the 2 and change only the Z site: katoite, Ca3Al2(OH)12, and holtstamite, Ca3Al2(SiO4)2(OH)4.
Staurolite fails it twice. Its IMA formula, Fe2+2Al9Si4O23(OH), has 24 anions. Halved to 12, it holds 1 iron, 4.5 aluminium and 2 silicon: a third of the large cations a garnet needs, more than twice the aluminium, two-thirds of the silicon. The sheet’s Black Mountain, Grafton Co., New Hampshire microprobe analysis, recast as printed to 24 oxygens, gives 1.81 divalent cations and lithium, 8.71 aluminium-site cations and 4.00 silicon-site cations: per 12, 0.91, 4.36 and 2.00.
Where the count fails. It can only rule out. Grew et al. set aside the so-called tetragonal almandine-pyrope phase because it has garnet’s stoichiometry but not its structure, with edges shared between tetrahedra and octahedra. Vesuvianite’s IMA formula, (Ca,Na)19(Al,Mg,Fe)13(SiO4)10(Si2O7)4(OH,F,O)10, works out at 2.92, 2.00 and 2.77 per 12 anions, close to a garnet, and it is not one; that case is on is vesuvianite a garnet.
Why staurolite and garnet are confused
They grow in the same rocks and look alike. The staurolite sheet’s association line opens with almandine, and the almandine sheet’s lists staurolite with kyanite, andalusite and sillimanite; the staurolite sheet calls it common in pelitic schists and gneisses, the almandine sheet typical of mica schists and gneisses. The International Gem Society staurolite article lists garnet among the crystal inclusions recorded in staurolite.
In the hand. Staurolite is dark reddish brown to blackish brown, subvitreous to resinous, hardness 7 to 7.5, streak white to greyish. Almandine is deep red, brownish red, red-violet or black, vitreous to resinous, hardness 7 to 7.5, streak white. Hardness and streak choose nothing; method on how to test mineral hardness properly.
How do you tell staurolite from almandine?
On crystals, the shape. Staurolite forms rough prisms to 12 cm, and its sheet records 60° twins as common and 90° cruciform twins as less common; twin laws are explained on mineral twinning explained, and the fake crosses on staurolite. Almandine forms dodecahedra or trapezohedra to 1 m, and also rounded grains and massive material, where shape fails.
On a loose clean piece, the balance. Worked as 20 − 20/D, 20.00 g reads 14.65 to 14.78 g in water as staurolite, 15.37 g as almandine and 15.23 g as spessartine: gaps of 0.59 g and 0.45 g at the closest. But andradite, 14.74 to 14.87 g, and uvarovite, 14.69 to 14.75 g, overlap staurolite; pyrope, 14.42 g, and grossular, 14.44 g, sit 0.23 g and 0.21 g below it. Method on how to measure specific gravity at home.
On a cut stone, the polariscope. Staurolite is biaxial (+) with three indices, α 1.736 to 1.747, β 1.742 to 1.753, γ 1.748 to 1.761. The garnet sheets print one index, n 1.714 (pyrope) to 1.887 (andradite), almandine 1.830. A doubly refracting stone is not a garnet, with the caveat that almandine’s sheet says anomalously biaxial and grossular’s single index, 1.734, lies close to staurolite’s α. The refractometer rule for garnet is on how to tell if garnet is real.
Who each test is wrong for
Each test fits some specimens. Shape is right for a crystal and wrong for grains and worn detrital pieces. The balance is right against almandine and spessartine and wrong against andradite and uvarovite, and for anything in schist. The polariscope is right for a transparent stone and weakened by anomalous garnets. The three-two-three count needs an analysis and only rules out. Hardness, streak and structure class are wrong for every specimen.
Other non-garnets. Rhodonite and vesuvianite are on is rhodonite a garnet and is vesuvianite a garnet; an andradite that is a garnet is worked site by site on is andradite a garnet. The kyanite on staurolite’s association line is the subject of is kyanite a feldspar. The garnet species pages are almandine, pyrope, spessartine, grossular, andradite and uvarovite, and more answers are in the collecting guides. We hold no catalogue and nothing here is offered for sale; the wanted list is the only commercial route.