---
title: "How your pieces are tested, and what XRF can and cannot settle"
description: "How gold and silver are checked: marks, weight, magnet and density, what XRF reads and where it misleads. Nothing is cut before you accept an offer."
canonical: "https://assaypost.co.uk/testing-and-xrf"
modified: "2026-10-01"
language: "en-GB"
publisher: "Assaypost"
---

# How your pieces are tested, and what XRF can and cannot settle

Updated 1 October 2026.

Every piece is weighed and checked for its metal before an offer is made; the methods used depend on the piece and are listed on the offer. X-ray fluorescence reads what the surface of a piece is made of without marking it, but it cannot see beneath that surface. No analyser is in use yet.

**In short:** Every piece is identified by its marks first, then weighed and checked with a magnet; density is measured on unmarked solid pieces and bars, and wherever those checks disagree, if the shape allows. X-ray fluorescence reads surface composition only; no analyser is in use yet. Your offer says item by item which checks were used and what they could not settle.

- **What XRF reads**: A thin surface layer, not the core
- **What misleads it**: Plating, surface enrichment, solder, curves, dirt, filled and mixed construction
- **Checks used now**: Marks, weight, magnet and, where the shape allows, density
- **Anything that marks a piece**: Only with your written consent, item by item
- **Analyser**: None in use; your offer lists the checks used on each piece

## What the analyser measures

An X-ray fluorescence analyser directs X-rays at a small area of a piece. Atoms in that area give off X-rays of their own, at energies characteristic of each element, and the detector counts them. The result is a list of percentages, such as gold, silver, copper, zinc, nickel and palladium. Nothing is removed and nothing is marked.

The limit is depth. The signal comes from a thin surface layer only: in a gold alloy it is a matter of microns, and the exact depth varies with the metal, the element being measured and the instrument. The reading describes that layer and says nothing about what lies beneath it. No single test proves what a piece is made of, and none is described here as though it did.

## Where a surface reading misleads

These are the situations in which a precise-looking reading can be wrong, and each shapes how a piece is checked.

1. **Plating and vermeil**: A thick coat of fine gold over silver or brass can fill the reading with gold, so a plated ring may read high on the outside. Its weight against its size often shows the difference on brass, though on a small or thin piece even that can be inconclusive, and gilded silver comes close to 9ct gold in density.
2. **Surface enrichment**: Polishing, the maker's chemical cleaning (pickling) and years of wear can leave the outside of an old piece richer in gold than the metal underneath. Reading a worn shank and a protected inner face can help bracket the true alloy, but cannot confirm it.
3. **Solder at the joints**: Chain links, clasps and ring sizings are often joined with a lower-carat solder. A reading taken across a joint measures the solder as much as the piece, so a careful reading avoids joints or reads them separately.
4. **Filled and weighted construction**: A loaded candlestick is a thin silver shell over pitch or plaster, and gold-filled work is a bonded sheet of gold alloy over brass. The surface reads as the real metal because it is the real metal. Weight against size, and what the piece is known to be made like, is usually what shows how little of it there is.
5. **Curves, edges and tiny pieces**: The measurement assumes a flat, clean surface filling the window. A thin band, a wire earring or a domed cufflink returns less signal and a less reliable number, and dirt or residue on the surface skews it, so those pieces need readings at more positions and other checks.
6. **One object, several alloys**: A watch case, a bracelet with a replacement clasp, a brooch with a steel pin: each part can be a different metal. The parts are checked separately and recorded as such, never averaged into one answer.

## What goes beside the reading

The hallmark comes first, under magnification, because a full UK mark names the [sponsor](https://assaypost.co.uk/glossary#sponsors-mark), the standard, the assay office and often the year. Struck and laser-applied marks look different under a magnifier, and a mark added to a piece later often sits wrongly. A mark is strong evidence but not proof, so the metal is checked as well.

Then weight. Each piece is weighed on its own, and the weight recorded goes on your list. A piece much lighter than it looks is a warning sign before any other check is made.

A magnet finds steel cores, spring bars and clasp mechanisms that are not the metal you are selling. Density is the next check: weigh a piece in air, weigh it again suspended in water, and the difference gives its density ([specific gravity](https://assaypost.co.uk/glossary#specific-gravity)), which a thin plating barely changes. A hollow chain, a filled bangle or a gold-plated brass ring is usually lighter for its size than solid metal. Density is valid on solid pieces but not decisive on its own: stones, hollows, trapped air and small pieces all blur the result, and 9ct white gold and palladium can overlap.

An acid touchstone test, which rubs a small streak of metal onto a stone and tests it with acid, can help where the shape defeats the other checks. It reads only the surface unless a notch is filed, and it leaves a small mark, so it is proposed only when it would change the answer, and only with your written consent for that piece ([the terms, clause 4.2](https://assaypost.co.uk/terms#clause-4-2)). Please do not try it at home.

> **Anything that alters a piece is asked for first**
>
> Where a mark is too worn or a surface too dirty to read, a usable result can mean cleaning, degreasing or gently rubbing a tiny patch clean. That alters the piece, so it is never done without your written consent for that item, with a photograph of the place. Refusing is free, and you choose: the piece is assessed without it, with the gap stated on your offer, or it comes back unassessed at our cost ([the terms, clauses 4.2 and 4.3](https://assaypost.co.uk/terms#clause-4-3)).

## How readings become a fineness

No analyser is in use yet. When one is, readings will be interpreted against the standards a piece could plausibly be, not rounded to whatever is nearest. A 14ct alloy is at least 585 parts gold per thousand; readings within the analyser's stated uncertainty of that figure, agreeing across two positions and supported by density, would be recorded as consistent with 14ct. The example below is an illustration of what such readings look like, not a record of a real item.

### An unmarked yellow ring, as an illustration

1. The ring weighs 6.40 g and has no stones or other parts.
2. A reading on the outside of the shank gives gold 58.9%, silver 5.1% and copper 35.6%.
3. A second reading, on the inside of the shank away from any join, gives gold 58.4%, silver 5.4% and copper 35.9%.
4. Weighed in water, the ring is consistent with a 14ct yellow alloy, not with plate.

Recorded as consistent with 14ct (585), with both readings, their spread and the density result written beside it.

When the checks and the mark disagree, we do not guess. A shank marked 750 that reads well below it may be rhodium-plated white gold, have a lower-carat repair, or not be what the mark says. You are told before anything else happens: what was found, where, what it suggests, and the choice between a further test and having the piece back.

## What you are shown for each item

Your offer carries the list of everything that arrived ([the terms, clause 5.1](https://assaypost.co.uk/terms#clause-5-1)). For every piece it records:

- The item number, a description and the photographs taken on arrival.
- The weight recorded, and any weight left out of the assessment, with the reason.
- Every check used and what it showed, with the position where it was made.
- Any test you asked us not to carry out.
- The fineness indicated, or a plain statement that it could not be established, and what the checks could not settle.
- Whether the piece was set aside as worth more whole than melted, and why.

The offer is one amount for the lot, set out with that list, and accepting or declining it is up to you. [Offers, returns and payment](https://assaypost.co.uk/offers-returns-and-payment) explains what you can do with it.

## The equipment used

Not yet: No analyser is in use. When there is one, its make, model, reference standards and last verification date will be published here, so that the claim can be checked.

Not yet: Fire assay is the reference method for gold, and it consumes part of the sample. There is no arrangement with an assay laboratory, so it is not available for a disputed piece. If you are not satisfied with a result, you can have the piece back at our cost rather than accept a cautious price.

Parcel openings, weighing and testing are not filmed, so there is no footage; the photographic record described on [posting safely and cover](https://assaypost.co.uk/security-and-insurance) is what exists. If filming ever starts, this page will say what is captured, who can view it, how long it is kept and how to get a copy or have it erased.

## Questions sellers ask

### Will the X-ray damage my jewellery?

No. We do not use an X-ray analyser at present, and where one is used, a reading is aimed at a few millimetres of surface for a short time and removes nothing. What can alter a piece is preparing a surface so it can be read at all, such as rubbing a tiny patch of a worn mark clean, and that never happens without your written consent for that item.

### Can you tell plated from solid without cutting it?

Often, though not always. A plated brass piece is usually too light for its size when it is weighed in water, though small pieces and gilded silver can be inconclusive. Where the checks are still ambiguous we say so rather than treat the piece as solid. Our guide to [plated, rolled and gold-filled](https://assaypost.co.uk/guides/plated-rolled-gold-filled-and-vermeil) pieces explains more.

### My pieces have no hallmark. Is that a problem?

Not necessarily. Plenty of older, imported and lightweight jewellery was never marked, and below the exemption weights a mark is not required. The metal is checked, and what the checks show, including what they could not settle, goes on the record. The [hallmark guide](https://assaypost.co.uk/guides/uk-hallmarks-and-fineness-guide) explains the exemptions.

### What happens if two readings disagree?

More checks, then density, then a note on your record explaining the spread. If only a test that marks the piece would settle it, we ask your written consent first. Say no and we either price the piece on the lowest figure the checks support or leave it out of the offer, and tell you which ([the terms, clause 4.4](https://assaypost.co.uk/terms#clause-4-4)). Or you can have it back unassessed, at our cost.

### Do you ever cut or scrape an item?

Nothing is cut, filed, polished or melted, and no stone is taken out, before you have accepted an offer for it. Before then, the only things that can mark a piece are an acid touchstone test and preparing a small window on a worn mark, and each needs your written consent for that item.

Further reading: [the physics and the limits in full](https://assaypost.co.uk/guides/xrf-testing-capabilities-and-limits) · [UK hallmarks and fineness](https://assaypost.co.uk/guides/uk-hallmarks-and-fineness-guide) · [weighted and loaded silver](https://assaypost.co.uk/guides/weighted-and-loaded-silver)

## Start with photographs

- [Send your photos on WhatsApp](https://wa.me/447369296063?text=Hello.%20I've%20got%20a%20piece%20I%20can't%20identify%20and%20I%20wanted%20to%20ask%20how%20it%20would%20be%20tested.)
- [How it works](https://assaypost.co.uk/how-it-works)
- Telephone [07369 296063](tel:+447369296063), 6am to 10pm, seven days a week