A loupe answers the question no measuring instrument can: what is going on inside this stone? Inclusions — the crystals, liquids, needles and growth features a gem carries — are as close as gemology gets to a signature. A refractometer reading is a number many species share; a patch of rutile silk intersecting at 120° belongs to a very short list of stones.

The numbers bear that out. In Loupewise's own engine, a single loupe inclusion observation narrows 130 species to a median of 6 candidates and settles 35 species outright — the strongest single observation in our entire dataset, on the cheapest instrument in it. For comparison, a refractometer reading alone leaves a median of 18, and a polariscope reading 74.

The honest caveat: these figures simulate a correct, confident reading of a diagnostic pattern on every species. Reading inclusions is a skill with a long tail — the numbers are what the observation is worth once you can make it, not a promise about your first week. That is also the argument for starting here: nothing else you can practise pays off as much.

Which Loupe

10× triplet. A triplet cements three lenses together to correct the colour fringing and edge distortion of a single lens. 10× is the working standard — it is the magnification diamond clarity grading is defined at — and it gives you about 2.5 cm of working distance with a usable field of view.

Not a 20× first. Doubling the magnification halves the working distance and shrinks the field and depth of focus, so you inspect the stone through a keyhole and miss the big picture — and the inclusion. Higher powers earn their place after a full pass at 10×, never instead of one.

Matte black housing. Bright metal casings reflect into the field of view and can tint what you see. A black-cased triplet avoids both.

Tweezers and a lint-free cloth. Fingers put grease on the stone; grease looks like a film of inclusions. Expect to drop stones while you learn — everyone does.

The Hold

Clean the stone

Dust, lint and fingerprints magnified ten times look like inclusions, and the back of a mounted stone is usually the dirtiest surface in the room. Clean thoroughly before you conclude anything — a wipe with a lint-free cloth for a loose stone; for a ring, get behind the setting.

Bring the loupe to your eye, not the stone

Hold the lens just in front of your eye — or just in front of your glasses, which can stay on — and rest the back of that thumb against the side of your nose. The loupe now moves with your head, and one of the three things that must hold still is anchored.

Bridge your hands

Hold the stone (or its tweezers) in the other hand and rest that thumb against the middle finger of the loupe hand. The two hands are now one braced unit, and the stone sits at the lens's working distance — roughly 2.5 cm for a 10×. Bring the stone into focus by rocking it, not by hunting with your head.

Keep both eyes open

Let the unused eye stare blankly. Squinting one eye for minutes at a time is what makes loupe work tiring; with both open, the brain suppresses the unused image on its own.

Light the stone from the side, against a dark background

The classic mistake is holding a stone up to a lamp and looking straight through it — glare in, nothing out. Inclusions show best lit obliquely, with no direct light entering your eye, so they appear bright against a dark field. A desk lamp beside the stone and a dark surface behind it approximates the dark-field illumination professional microscopes are built around.

Scan systematically, then change orientation

Wiggle the stone slowly through the focal plane — focus is a slice, and the interior has depth. Work crown, pavilion, and side views. Several diagnostic features, growth lines above all, are only visible from particular directions, so a one-angle look proves very little.

The Patterns Worth Learning First

Our database records loupe-visible inclusions for all 130 species. A handful of patterns recur across many stones — learn these before anything exotic, because each one you can recognise does real eliminating work.

Two-phase inclusions

A sealed cavity holding liquid plus a gas bubble. The classic form is a flat, angular “negative crystal” with a bubble inside it.

In our data: recorded for 55 of 130 species — quartz family, beryls, topaz and many more. Common, and strong evidence of natural growth.

Included crystals

Solid mineral grains with edges and faces of their own — zircon, rutile, spinel, mica, or the host mineral itself. Edges separate them from bubbles, which are round or teardrop and have none.

In our data: the most widespread solid-inclusion class. A crystal inside a stone is a strong vote for natural origin.

Silk

Fine rutile needles in intersecting sets — in corundum, meeting at 120° — showing a sheen in reflected light. Dense silk is what makes star stones.

In our data: 11 species. Beware the counterfeit: swarms of tiny bubbles in synthetic material can imitate silk's sheen until examined closely.

Fingerprints and feathers

Tiny flat liquid-filled cavities assembled in one plane — flat or gently curved sheets that look like a fingerprint pressed into the stone.

In our data: 12 species. Synthetic emerald's version is lace-like wisps of flattened canals — a pattern with no natural counterpart.

Curved lines and bubbles

Growth lines fine as a gramophone record's, curved — the signature of flame-fusion synthetic corundum — often with round gas bubbles, which natural corundum never contains.

In our data: curved striae and flow lines on 14 species between them. Natural colour banding is rigidly straight and meets at 120°.

Facet-edge doubling

Not an inclusion — back facet edges appearing as double lines through the table. It is double refraction made visible, and at 10× it is obvious in zircon, sphene and peridot.

Log it as optic character, not structure. It vanishes when you look down an optic axis, so tilt before concluding it is absent.

Where It Goes Wrong

Dirt reads as inclusions. The most common beginner error is diagnosing a stone that just needed cleaning. Surface features move or wipe away; inclusions do not. When something interesting appears, check first whether it is on the surface — refocus across it and look for it in reflected light.

Absence of evidence, again. Curved growth lines are only visible in the right orientation under the right lighting — in a mounted stone they can be nearly impossible to find even when present. Not seeing them is never proof a stone is natural; it is proof you have not seen them.

Bubbles imitating better things. Bubble swarms can pass for silk, and chains or sheets of bubbles in glass imitate feathers — until closely examined. Bubbles are round or teardrop-shaped with smooth outlines; a pointed teardrop is a shape peculiar to glass. Crystals have edges.

Reaching for more power. A 20× or 30× loupe halves the working distance, shrinks the field, and turns inspection into peering through a keyhole. Survey at 10× first, always; escalate only to inspect something already found.

Deciding species from one inclusion. An inclusion narrows the field; it rarely finishes the job on its own, and treatments and unusual specimens exist. Confirm with a measurement — RI, SG, optic character — before concluding.

Loupewise takes what you see — silk, two-phase inclusions, fingerprints, curved striae and more — and shows which of 130 species remain consistent with it, plus which test narrows fastest next.

Open Loupewise

Studying for the FGA or GIA practical? Loupewise drills inclusion recognition alongside instrument readings, and runs timed mock exams built around the real exam formats.

Practice exams →

Where This Comes From

The technique and the pattern descriptions are drawn from published gemological literature; the elimination figures are computed from Loupewise's own database and engine.

B. W. Anderson, Gem Testing, 6th ed., Heywood, 1958 — the pocket-lens doctrine, dark-field lighting principle, cleaning mounted stones, silk, feathers, two-phase inclusions, natural-vs-synthetic growth lines, and the bubble shapes peculiar to glass.
The Gemology Project wiki, 10x Loupe — the triplet construction, working-distance arithmetic, the 10×-before-20× rule, and the two-handed brace. Teaching-material grade, consistent with Anderson.
Loupewise engine measurements — median candidates and species counts computed across all 130 species; guarded by tests so they update when the data does.

Frequently Asked Questions

How do I use a jeweler's loupe for the first time?

Anchor it: lens just in front of your eye, thumb against your nose, both eyes open. Bring the object to the lens with your other hand and rest that thumb against the loupe hand's middle finger so the two hands brace each other. Practise on anything — print, fabric, a coin — before gemstones, and expect the working distance (about 2.5 cm at 10×) to feel uncomfortably close at first.

Can I use a loupe with glasses on?

Yes — hold the loupe just in front of your spectacles instead of your bare eye. The technique is otherwise identical, and keeping your glasses on is usually better than working with uncorrected vision.

Is 10x or 20x better for gemstones?

Start at 10×, every time. It is the standard diamond clarity is graded at, and its 2.5 cm working distance and wider field let you survey the whole stone. A 20× halves that distance and shrinks the field and depth of focus, so it is easy to miss an inclusion entirely. Use higher powers to inspect something the 10× already found.

How do I tell an inclusion from dirt on the surface?

Clean the stone first, then check depth: rock the stone slowly through focus. A surface feature sits in the same focal slice as the facet edges and often shows in reflected light; an inclusion sits below them. If in doubt, wipe and look again — dirt moves, inclusions do not.

Do inclusions mean a stone is natural?

Some do, some mean the opposite. Included mineral crystals and angular two-phase cavities are strong natural indicators. Round gas bubbles point to glass or flame-fusion synthetics; curved growth lines are a synthetic-corundum signature; and synthetic emerald grows lace-like feathers of its own. The pattern matters more than the mere presence of something inside — and either way, confirm with a measurement before concluding.

How to Use a Jeweler's Loupe for Gemstone Identification