How to Identify an Unknown Gemstone

The property-based workflow used by GIA-trained gemologists.

Most gemstone identification errors come from relying on color or visual appearance. A red stone could be ruby, spinel, pyrope garnet, almandine, or red tourmaline — all can look nearly identical under the same light source. Photo-matching apps face this problem head-on and lose. Property-based identification sidesteps it entirely: the refractive index of ruby is always 1.762–1.770, regardless of what it looks like.

Property-based identification uses physical constants — RI, SG, optic character — that are determined by atomic structure and crystal chemistry. These values don't shift with lighting, treatment history, or cutting style. A well-calibrated refractometer gives you a number that maps directly to species. This guide walks through the standard workflow in order, from initial observation through final identification.

Before You Start — What You Need

You don't need every instrument on this list to start identifying gemstones, but the more you have, the fewer candidates remain unresolved. The required items handle most identifications. The optional items handle edge cases and high-value confirmations.

Required
Refractometer
With sodium light source or calibrated LED equivalent. Contact liquid (RI ~1.81) included. Reads RI to ±0.002.
Required
Longwave UV Lamp (365nm)
For fluorescence testing. Darkened room or UV box needed. Do not substitute a blacklight flashlight — output must be 365nm.
Required
10x Loupe or Microscope
For inclusion examination. A standard 10x triplet loupe is adequate for most observations. Darkfield illumination preferred for inclusions.
Recommended
Hydrostatic Weighing Setup
Precision scale, beaker, bridge or cage, and distilled water. Measures SG to ±0.01 on stones above 0.5ct.
Optional
Shortwave UV Lamp (254nm)
Adds a second fluorescence data point. Particularly useful for separating natural from synthetic diamonds and detecting certain treatments.
Optional
Chelsea Filter
A transmission filter that isolates red and green wavelengths. Most useful for emeralds, rubies, tsavorites, and chrome tourmalines.
Optional
Polariscope
Confirms singly vs. doubly refractive, and determines uniaxial vs. biaxial on DR stones. Adds optic sign as a data point.
Optional
Spectroscope
Absorption spectrum can distinguish chrome from vanadium coloration in emerald and identify rare gem species by characteristic absorption bands.

GemID app: The GemID app accepts measurements from all of the instruments above and ranks remaining candidates in real time. It also surfaces the next most useful test based on what candidates are still in contention — so you spend instrument time efficiently.

The Identification Workflow

Work through these steps in order. Each step eliminates candidates. By step 4, most unknowns are resolved to a single species or a short list requiring one confirmatory test.

Observe basic characteristics

Note transparency (transparent, translucent, or opaque), color and distribution, luster (vitreous, resinous, adamantine, waxy, or silky), and crystal habit if the stone is rough. For faceted stones, observe cutting style and proportions — some simulants are cut in ways that differ from natural stones. These initial observations narrow the field before any instrument is used and can flag anomalies worth investigating. An "emerald" with vitreous luster and no inclusions warrants skepticism before the refractometer is even touched.

Measure refractive index (RI)

Place a clean, flat facet against the refractometer hemicylinder with a small drop of contact liquid. The flat facet must be large enough to cover the hemisphere — a small culet or pavilion facet is ideal. Turn on the light source and look through the eyepiece: you will see a scale with a dark shadow edge. Read the position of the shadow edge to the nearest 0.002. For doubly refractive stones (most colored gems), slowly rotate the stone 360 degrees while watching the shadow edge. Two edges will appear and move. Record both: the lower reading is RI min; the higher is RI max. The difference (RI max minus RI min) is birefringence. Record all three values.

Determine optic character

While rotating the stone on the refractometer, observe whether one shadow edge stays fixed and one moves (doubly refractive, DR) or whether a single edge stays fixed throughout the rotation (singly refractive, SR). Singly refractive gems include cubic crystals (garnets, spinel, diamond) and amorphous materials (glass, opal). Doubly refractive gems include all uniaxial (hexagonal, trigonal, tetragonal) and biaxial (orthorhombic, monoclinic, triclinic) crystals. If you have a polariscope, use it on the DR stones to determine whether the stone is uniaxial or biaxial and whether it is optically positive or negative. This narrows the species list further and helps resolve ambiguous cases where multiple species share an RI range.

Measure specific gravity (SG)

Specific gravity is the ratio of a stone's weight to the weight of an equal volume of water, and it is the most powerful complement to RI. Weigh the stone dry in air using a precision scale — record this as Wair. Then suspend the stone in distilled water using a cage, hair, or purpose-built SG frame so it does not touch the sides or bottom of the beaker. Record this as Wwater. SG = Wair ÷ (Wair − Wwater). Inclusions, surface-reaching fractures filled with flux or oil, and coatings can affect the reading, so note any obvious treatments. Even a rough SG reading eliminates large groups: if SG is above 4.0, the stone is almost certainly zircon, cassiterite, or corundum. If SG is below 2.7, it is likely feldspar, opal, or a glass simulant.

Test UV fluorescence

In a darkened room or UV examination box, illuminate the stone with a longwave UV lamp (365nm). Note the fluorescence color (blue, orange, green, red, chalky white, yellow, or none) and intensity (none, weak, moderate, or strong). Then test with shortwave UV (254nm) if available. Record both readings separately — many species respond differently to longwave vs. shortwave, and the difference is diagnostic. Key patterns: natural rubies from Burma (Mogok) show strong red fluorescence under LW UV; heat-treated rubies from other origins may show weaker or no fluorescence due to iron content; Verneuil synthetic rubies often show stronger and more uniform red fluorescence than naturals; blue sapphires typically show no fluorescence or weak orange under LW UV; blue glass simulants often inert or chalky blue. Natural diamonds show blue fluorescence in about 25–30% of cases; synthetic HPHT diamonds may show yellow-green.

Chelsea filter observation

View the stone through the Chelsea filter under daylight-balanced or incandescent light. The filter transmits only red (around 690nm) and yellow-green (around 570nm) wavelengths. A stone colored by chromium — natural emerald, ruby, or chrome tourmaline — will appear red or pink because chromium has a characteristic red transmission band. Colombian emerald: strong red. Synthetic emerald: strong red. Tsavorite garnet: orange to red. Aquamarine: green (no chromium; iron coloration). Peridot: green. Paste (glass) emerald simulant: typically green, no red reaction. Demantoid garnet colored by chromium: orange-red. The Chelsea filter is useful but not definitive — always combine with RI and SG. Vanadium-colored emeralds may not show a red reaction despite being genuine emeralds.

Enter all measurements in GemID

Open GemID and enter the values recorded in steps 1 through 6: RI min, RI max, optic character (SR/DR/uniaxial/biaxial), SG, LW and SW fluorescence, Chelsea filter reaction, and any inclusion observations. GemID cross-references your input against its database of 130 gem species and ranks remaining candidates by how closely your measurements match the expected ranges. It also flags inconsistencies — for example, if your RI is consistent with sapphire but your SG reading is 2.65 (which would be inconsistent with corundum at 3.95–4.10), GemID will flag that discrepancy. Follow the "Next Test" recommendation to distinguish between remaining candidates until you reach a single consistent identification.

Reading the Results

GemID displays remaining candidates ranked by property match — not by visual similarity. The ranking reflects how well your measurements fall within each species' documented ranges for RI, SG, fluorescence, and optic character.

How GemID flags results

Consistent: Your measurement falls within the documented range for this species. The property supports the identification.
Inconsistent: Your measurement falls outside the expected range. Either the stone is not this species, or there is a measurement error worth reviewing.
Not entered: You have not yet measured this property. GemID's "Next Test" recommendation tells you which measurement would be most useful given what candidates remain.

When two or fewer candidates remain and all entered measurements are marked consistent, you have the information needed to record an identification. When one measurement is inconsistent, re-verify that measurement before concluding — calibration errors and wet facets on the refractometer are common sources of incorrect readings.

On reporting language: GemID shows results as "consistent with" a species — not "this stone is X." A field identification is a strong indicator; a certified laboratory report (GIA, Gübelin, SSEF) is a definitive determination. For stones with significant commercial value, always follow up with a reputable laboratory.

Common Identification Scenarios

These worked examples show how the workflow eliminates candidates in sequence. In each case, the identification follows from measurement — not from appearance.

Stone Key measurements Elimination logic Result
Unknown blue stone RI 1.762–1.770, SG 4.00, DR uniaxial neg., strong blue LW fluorescence (none) Tanzanite eliminated: RI 1.692–1.700. Aquamarine eliminated: SG 2.68. Blue topaz eliminated: RI 1.610. Blue tourmaline eliminated: SG 3.06. Measurements consistent with sapphire on all entered properties. Sapphire (consistent)
Unknown red stone RI 1.762–1.770, SG 4.00, DR, strong red LW fluorescence Spinel eliminated: SR (cubic), SG 3.52–3.57. Pyrope garnet eliminated: SR, SG 3.65–3.87. Red tourmaline eliminated: RI 1.624–1.644. Strong red LW fluorescence consistent with Burmese/Mozambique origin. Measurements consistent with ruby. Ruby (consistent)
Unknown green stone RI 1.577–1.583, SG 2.72, DR biaxial neg., Chelsea: red Peridot eliminated: RI 1.654–1.690. Tsavorite garnet eliminated: SR, SG 3.57–3.73. Demantoid eliminated: SR (cubic). RI and SG consistent with emerald; Chelsea red consistent with chromium coloration. Emerald (consistent)
Unknown colorless stone RI over 1.81 (off scale), SG 4.69, DR with strong birefringence visible to naked eye, inert UV Diamond eliminated: SR, SG 3.52. CZ eliminated: SR, SG 5.65–5.95. Moissanite eliminated: DR but SG 3.22. High RI + SG 4.69 + visible doubling: consistent with high zircon (SG 4.60–4.73, RI 1.925–1.984). Zircon (consistent)
Unknown purple stone RI 1.544–1.553, SG 2.65, DR uniaxial neg. Tanzanite eliminated: RI 1.692. Iolite eliminated: SG 2.57–2.66 (borderline), RI 1.542–1.551 (close — note biaxial vs. uniaxial). Uniaxial neg. + SG 2.65 + RI range: consistent with amethyst (quartz). Verify with conoscope interference figure if needed. Amethyst (consistent)

Green stone with Chelsea filter

The Chelsea filter is most useful as a second-step test after RI has already shortened the list. As a standalone screen for green stones: Chelsea red = likely emerald or chrome tourmaline; Chelsea green = likely peridot, tsavorite, or demantoid. But note that vanadium-colored emeralds (some from Zambia and Brazil) can show a Chelsea green reaction despite being genuine emeralds. Always confirm with RI and SG.

GemID handles the candidate matching and next-test logic automatically. Enter your measurements and let the app work through the elimination sequence.

Open GemID →

Frequently Asked Questions

Can I identify a gemstone without a refractometer?
Limited, but possible in some cases. Specific gravity combined with fluorescence and careful visual observation can narrow candidates to two or three species when the stone has distinctive SG values. For example, a colorless stone with SG 3.52 that is inert to UV and optically isotropic is almost certainly diamond or cubic zirconia (SG 5.65–5.95, easily distinguished by weight). But RI is the single most diagnostic measurement in gemology for most species pairs. Most gemology textbooks — GIA's Gem Reference Guide, Anderson and Jobbins' Gem Testing, O'Donoghue's Gems — rank RI as the primary instrument precisely because it is repeatable and highly species-specific. Without RI, many identifications remain provisional.
What if my stone reads off the scale on the refractometer?
Standard gemological refractometers top out at approximately 1.81. A reading that goes to the top of the scale or produces no readable shadow edge indicates a high-RI stone. Candidates include: demantoid garnet (RI 1.880–1.889, SG 3.82–3.85, SR, green, characteristic horsetail inclusions), zircon (RI 1.810–1.984, SG 4.60–4.73, DR with very high birefringence — doubling of facet edges visible under 10x), cassiterite (RI 1.997–2.093, SG 6.7–7.1, very dense), and synthetic YAG (RI 1.834, SG 4.55, SR, colorless or tinted). Specific gravity is the most useful discriminator: weigh the stone in hand first — cassiterite is noticeably heavier than demantoid. Chelsea filter can help: demantoid shows orange-red; YAG typically green.
How accurate does my RI reading need to be?
Within ±0.003 is adequate for most identifications. This is achievable with a well-calibrated gemological refractometer and good lighting. Many species have RI ranges that differ by 0.01 to 0.05, so ±0.003 is more than sufficient to discriminate between them. Where it matters most: distinguishing aquamarine (RI 1.577–1.583) from blue topaz (RI 1.610–1.620) — a 0.027 gap, well within instrument range — or blue sapphire (1.762–1.770) from blue spinel (singly refractive, single reading in range 1.712–1.762). Calibrate your refractometer regularly using the standard glass calibration block that should come with the instrument. If your block reads outside its certified value by more than 0.002, service or replace the instrument.
Can GemID identify synthetic gemstones?
GemID includes a dedicated Natural vs. Synthetic protocol for the 32 gem species that have commercially significant synthetic counterparts, including corundum, emerald, alexandrite, spinel, tourmaline, and opal. The protocol is a two-phase workflow: Phase 1 identifies the species by properties; Phase 2 guides through inclusion examination to look for synthetic growth features. Property measurements alone typically cannot distinguish natural from synthetic origin for most species — the crystal chemistry is identical, so RI and SG values overlap completely. Origin determination requires magnification: Verneuil synthetic corundum shows curved growth lines (chevron striations) and gas bubbles in rounded or teardrop shapes; hydrothermal synthetic emerald shows chevron growth patterns and two-phase inclusions; flux-grown synthetic ruby shows wisps of remaining flux and platinum platelets. For high-value stones, submit to a gemmological laboratory (GIA, Gübelin, SSEF, AGL) for definitive determination.