How to tell if a mineral specimen is repaired
Repair is normal in the mineral trade and undisclosed repair is not. A six-step inspection order, what raking light and UV each reveal, and the effect on value.
Hold the specimen 30 cm from a bare lamp and rake the light across it: adhesive shows as a change in reflectance before it shows as a change in colour. Then check it under longwave UV in the dark, because most epoxies and cyanoacrylates fluoresce differently from the mineral and the matrix around them.
In short
- Repair is not fraud. Undisclosed repair is. A clean break in a specimen's own material, rejoined and declared, is trade-normal on old collection pieces. A crystal glued on from somewhere else is not a repair at all.
- Work in a fixed order. Most people start by squinting at the crystal; most repairs are on the back, the base, or a matrix contact, because that is where specimens break and where a repairer expects nobody to look.
- Raking light is the highest-yield tool and it costs nothing. A glue line differs in gloss from the mineral around it even when it matches in colour.
- Longwave UV is the confirmation, not the search. It will not find a repair you have not already suspected, but it will settle one you have.
- The value question is not whether a piece is repaired but what was repaired and whether you were told. Rejoined matrix costs little; a rejoined crystal termination costs a great deal.
| What was done | What it means | Trade view | Effect on value |
|---|---|---|---|
| Cleaning — acid, ultrasonic, mechanical | Removal of coating, not addition of material | Normal, usually not itemised | None, if the species survived it undamaged |
| Matrix trimming | Excess rock removed to improve balance | Normal, usually not itemised | Usually positive |
| Repair: the specimen's own broken piece, rejoined, disclosed | A clean break put back | Accepted, and universal on 19th-century material | Reduces, mildly on matrix, sharply on a crystal |
| Repair, undisclosed | The same act, without the sentence | Not accepted | Material, and it damages the seller more than the piece |
| Restoration: added filler, tinted resin, rebuilt termination | New material that was never part of the specimen | Must be disclosed explicitly | Substantial |
| Composite: parts of two specimens married into one | Presented as a single natural piece | Not accepted | Destroys collector value |
| Fabrication: crystals set into a base with adhesive and dusted matrix | Manufactured object | Fraud | No collector value at all |
The six-step check, in order
Order matters, because each step is cheaper than the one after it and rules out more.
- Turn it over. Look at the back and the base before the display face. Specimens break where they were struck loose or where they were dropped, and both are usually behind or underneath.
- Rake the light. Hold the piece about 30 cm from a bare bulb or a phone torch, at a low angle so the light skims the surface, and rotate it slowly. You are watching for a line where the gloss changes — adhesive is almost always either shinier or duller than the mineral, even when a repairer has matched the colour perfectly.
- Follow every crystal-to-matrix junction. Nature grows crystals out of matrix, so the contact is intergrown and irregular. Adhesive sits on matrix, so the contact is a seam. A crystal whose base meets the rock along a clean flat line did not grow there.
- Look for filler. Restoration commonly uses adhesive loaded with crushed matrix or mica to bridge a gap. It matches in colour and fails in texture: it is matte where the rock around it has micro-glitter, and it has no grain.
- Longwave UV, in the dark, after your eyes adjust. Most epoxies and cyanoacrylates fluoresce — typically a dull bluish white — against a mineral and matrix that do something else or nothing at all. Interpret with care: plenty of minerals fluoresce in their own right, so a glowing area is a question, not an answer.
- Ask, and ask for it in writing. The reply is the most informative test on this list. A seller who says yes, there is a repair through the base, here is a photograph of it has told you both about the specimen and about themselves.
What raking light shows that a photograph never will
Almost every online listing is lit flat and frontally, which is exactly the lighting that hides a repair. That is not usually deliberate — flat light is also the lighting that makes colour read correctly — but the effect is the same. A glue line is a specular feature: it announces itself in the reflection, and a photograph optimised for colour has removed the reflections.
So when buying at distance, ask for one extra photograph taken with a single light source low and to one side, showing the back and the base. It is a small request, it is instantly recognisable to any experienced seller as a condition question, and a refusal is itself informative. On old British material this is close to routine: much of it was collected in the 1800s, has changed hands many times, and has been repaired at some point by somebody. We say the same thing about condition on our page about buying old collection material, and it applies with particular force to Cornish copper species, which are both fragile and old.
Repair, restoration and fabrication are three different things
The vocabulary matters, because the words are used loosely and the difference in value is enormous.
Repair puts a specimen's own broken piece back where it came from. Nothing is added but adhesive. On a 150-year-old Cornish specimen this is close to expected, and a well-executed, disclosed repair through a matrix is a minor matter.
Restoration adds material that was never part of the specimen: filler in a void, tinted resin rebuilding a chipped termination, a sculpted replacement for a missing crystal tip. This must be disclosed explicitly and described, because the buyer is being asked to pay for a surface that is not mineral.
Fabrication is manufacture. Loose crystals stood upright in a bed of adhesive with crushed mica dusted over it to hide the join is the pattern documented most often, and the giveaway is usually statistical rather than physical: real vugs do not produce a dozen equally sized crystals all standing at the same angle. This is not a condition issue, it is a fake, and we treat it as one on how to tell if a crystal is fake or treated.
A fourth category sits awkwardly between them: the composite, where genuine pieces of two different specimens are married into a better-looking one. Everything on it is real. The object is not.
What a repair does to what a specimen is worth
We do not publish price guidance and there is no market rate to quote, but the direction and rough proportion of the effect are consistent enough to state:
- A repair in the matrix, away from the crystals, is the mildest case. The specimen still shows what it is meant to show.
- A repair through a crystal is much more serious, and it is worse the closer it is to the termination, because the termination is the part that carries the aesthetic and the part that is hardest to fake convincingly.
- A repair on a species that is common matters more than the same repair on a species that is rare, because the buyer has alternatives. A repaired specimen of something that comes to market twice a decade is still the only one available.
- Disclosure changes the transaction, not the object. A repaired specimen bought knowingly at a fair price is a good purchase; the identical specimen bought unknowingly is a bad one. That is why the sixth step of the check is asking.
Doing this yourself is entirely realistic, and worth saying plainly: a torch, a dark room, a cheap longwave lamp and the discipline to always turn the piece over will catch most of what is out there, and it costs an evening to learn. What it will not do is scale. Checking one specimen properly takes ten minutes; checking a whole cabinet coming out of a deceased estate takes days, and the pieces you most want are the ones you will see photographs of rather than hold. Where that becomes the constraint, tell us what you are after on the wanted list — the value is in somebody having had it in their hands under a raking lamp before you commit, not in us holding anything, because we hold no stock.