How to tell real gold

The first thing to look at is the hallmark, usually on the inside of a ring or at the clasp of a chain. In Germany there is no obligation to hallmark. If a fineness is stamped, the seller is liable for it (Section 7 of the German Fineness Act, Feingehaltsgesetz); there is no state assay, but false stamps carry a fine (Section 9 of that Act). Jewellery may be stamped in any fineness, stated in thousandths such as 333, 375, 585 or 750.

A hallmark is therefore a clue, not proof. Stamps can be forged, and the number alone does not tell you whether it sits on solid gold or on a gold-plated core. A maker's mark next to it makes the piece traceable, but not any more genuine.

Certain additions reveal a layer: GP or GEP stands for gold electroplated, HGE for a thicker electroplated layer, GF, RGP or gold overlay for a mechanically bonded gold layer, called Doublé in German. A fraction such as 1/20 or 1/40 before the carat number belongs here too. The carat in 14K GF or 18K GP describes only the layer, not the piece.

A calculation shows how little gold this is: with 1/20 14K GF, one twentieth of the metal weight is 14 carat gold, which is about 2.9 per cent fine gold. A 10 gram chain with this mark therefore contains about 0.29 grams of fine gold, a solid 585 chain of the same weight 5.85 grams.

Gold is not magnetic. If a strong magnet clearly attracts a piece, it is not gold all the way through but contains iron, steel or nickel. If it does not stick, that proves little: copper, brass, silver, lead and tungsten do not stick to a magnet either. A fridge magnet is usually too weak for the test.

Many gold chains have an exception: the spring inside the spring ring or lobster clasp is often steel and reacts to the magnet, even though the chain is gold. The German Fineness Act expressly leaves such springs out of the fineness mark.

If you rub a piece across the rough, unglazed underside of a tile, it leaves a streak. If the streak is gold-coloured, at least the surface is probably gold; it says nothing about the inside or the fineness. In the trade, a dark streak is read as a sign of base metal.

Because the test removes material and leaves scratches, we advise against it, especially with heirlooms. In any case it cannot tell gold plating from solid gold as long as the streak only touches the plating.

Gold is one of the densest metals, and that can be measured. You need an accurate scale, a glass of water and a thread: weigh the piece, put the glass of water on the scale and tare it, then hang the piece in the water on the thread without touching the bottom or sides. The display now shows the weight of the displaced water, and the weight of the piece divided by this value gives the density. The table shows where the usual finenesses lie.

The test only works for solid pieces with a smooth surface. Hollow jewellery, air bubbles and set stones distort the result. Against a tungsten core it does not help at all: at 19.25 g/cm³ tungsten is almost as dense as fine gold. With bars, professionals therefore also test with ultrasound or electrical conductivity.

The vinegar test proves nothing. Gold does not react with vinegar, but neither does gold plating; if a piece stays unchanged in vinegar, it may just as well be plated. The same goes for toothpaste, which only rubs the surface and reveals nothing about the material underneath.

Biting the coin does not work either. Alloyed gold such as 585, 750 or 22 carat is too hard for tooth marks, whereas soft lead does show them. The risk is borne by your teeth and the piece.

In the touchstone test the piece is rubbed across a black touchstone of siliceous slate, next to comparison streaks made with test needles of known alloys. Testing acids dissolve the base metals, and how the streaks react shows the fineness. Experienced testers can detect differences of ten to twenty thousandths this way.

The touchstone test, too, initially checks only the surface. If plating is suspected, a little material is filed off in a hidden spot and the fresh metal is tested. Testing acids are corrosive, so the test belongs in experienced hands.

An X-ray fluorescence device, XRF for short, determines in seconds which metals are present and in what proportion, without damaging the piece. However, it only measures the surface to a depth of about 15 to 20 micrometres. Ordinary gold plating is only 1 to 3 micrometres thin, so the device also picks up the metal underneath. A deliberately applied, thicker gold layer, on the other hand, can fake a higher carat, and it cannot see a tungsten core.

Professionals therefore do not rely on a single method. They combine the measurement with weight, dimensions and density and, if in doubt, file a hidden spot to measure again underneath.

In our shop in Nordhorn we test every piece free of charge and in front of you. First we look at the hallmark and maker's mark under the loupe and hold the magnet to it, then we measure with the X-ray fluorescence device. If we suspect thick plating, we add a density check or the touchstone test. Weighing is done on the calibrated scale.

You see every result yourself before money is discussed. If a piece turns out to be plated, we tell you straight away and you take it home again. If it is genuine, we tell you today's value and you decide in peace whether to sell.