Treating stones is at least three thousand years old and the recipes were written down before 400 AD. The gem trade answered with a set of code symbols, and asking which one applies is fair.
In 1832 the Swedish Academy received a metal box containing fourteen numbered papyrus sheets and one unrelated fragment, all in early Greek handwriting, sent from Alexandria by the Swedish-Norwegian vice-consul, a man who collected early Egyptian documents. The middle and longest section, in Kurt Nassau’s account of it for GIA’s own journal, “contains 73 recipes which deal with the falsification of pearls and gemstones”. Otto Lagercrantz worked out that the sheets were a copy made by a scribe about 400 AD in Greek-speaking Egypt.
They read like what they are, which is a working file. One entry, in full: “Bleaching Crystals: Dissolve rice in water, put the crystal in, and together with it, boil again the solution.”
Somebody in fourth-century Egypt kept careful notes on making one stone pass for another, and somebody else copied them out well enough that they lasted sixteen hundred years. Between that box and a crystal shop a great deal has changed, and almost none of it is the treating.
Pliny spends his last pages on how to catch a fake
A little over three centuries before that scribe sat down, Pliny the Elder was finishing the last book of his Natural History, and he spends its closing pages on the same subject from the other side.
He records the practice without much heat. “In general, all gems are rendered more colourful by being boiled thoroughly in honey, particularly if it is Corsican honey, which is unsuitable for any other purpose owing to its acidity.” He notes cutters removing parts of variegated stones to make novelties, and then gives their trick a name. “In order that these selfsame stones may not bear their usual name, they call them ‘physis,’ or ‘works of nature,’ and offer them for sale as natural curiosities.”
Works of nature. A Roman workshop cut a stone into something it had not been, then reached for a phrase meaning made by nature in order to sell it. The same move is on shelves today with the Greek swapped out for English.
Then he turns to the deliberate fakes. A sardonyx, he says, can be manufactured so convincingly by sticking three gems together, a black stone from one species, a white from another and a vermilion from a third, “that the artifice cannot be detected”. And there are manuals: “treatises by authorities, whom I at least shall not deign to mention by name, describing how by means of dyestuffs emeralds and other transparent coloured gems are made from rock-crystal, or a sardonyx from a sard, and similarly all other gemstones from one stone or another.” His verdict on the business is short. “There is no other trickery that is practised against society with greater profit.”
What he does next is the part worth having. “I, on the other hand, am prepared to explain the methods of detecting false gems, since it is only fitting that even luxury should be protected against deception.”
And he does, at length. Weight first, “because genuine stones are heavier”. Then coolness, “since genuine stones also feel colder in the mouth”. Then structure, because artificial stones “show globules deep below the surface, rough patches on the surface itself, filaments, an inconsistent lustre and a brightness that fails to strike the eye”. Test them early in the morning, he says, or up to ten o’clock, but on no account later. The most effective test is to knock a piece off and bake it on an iron plate, “but dealers in precious stones not unnaturally object to this, and likewise to testing with a file.” He offers a gentler one instead. “Flakes of obsidian will not scratch a genuine stone, but on a false stone every scratch leaves a white mark.”
Two thousand years ago a buyer had a published test list and a seller who did not want a file taken to the merchandise. Both of those people are still in the shop.
Two crafts, and how far back each one goes
The practice runs back a good deal further than the writing. Nassau records heated carnelian found in Tutankhamun’s tomb, “dating to at least 1300 B.C.”, which puts deliberate colour alteration more than a thousand years before anyone wrote down a recipe for it that survived. Foils go back further still: the use of shiny metal foils behind a stone to make it brighter or shift its colour reaches “at least to Minoan times (2000-1600 B.C.)” on the authority Nassau cites for it.
His summary of Pliny should be taken straight, because it is a gemmologist’s assessment rather than a historian’s flourish. Pliny “discusses many gemstone-enhancement techniques that are still in use today, almost 2,000 years later, including foils, oiling, dyeing, and composite stones.”
All four of those are on a modern trade’s list. The composite is the sharpest of them, because Pliny’s three-part stone, a black and a white and a vermilion glued into one convincing object, is precisely what the coloured stone trade now calls an assembled stone and gives a symbol to. The technique did not evolve into anything. It was described in the first century, it is made the same way, and what has been added since is a label.
The detection craft cannot be dated as far back, and it would be easy to overstate. What can be said is that the two run together for as long as there is writing about either. The earliest surviving recipe collection for faking stones and the earliest surviving instructions for catching it are three centuries apart, which on this timescale is close enough to call them contemporaries. Neither craft is the intruder.
Is any of this new?
Most crystal buying advice will tell you to look for untreated stones, and it is not wrong that what was done to a stone is worth knowing. Where it goes wrong is the story wrapped around the instruction, which is that treatment is something modern creeping into a trade that used to be straight. That second half has no history behind it at all.
The framing does real damage to a beginner, which is why it needs taking apart rather than just correcting. If treatment is a modern corruption, then an untreated stone is the pure article and everything else is a compromise, so the sensible response is to hunt for purity and to feel cheated on failing to find it. It is a framing that sells worry, and then sells rough stone at a premium.
Take the framing away and the picture is calmer. Heat, dye, oil, wax and assembly are ancient crafts. Several of them are the reason a given stone is wearable at all, and some are stable enough that a piece will look the same in thirty years, though which ones and how stable depends on the material and is exactly the kind of thing a chart is for. A treated stone is a real object with a real history, and nothing in your collection is diminished by having had something done to it.
None of that makes undisclosed treatment acceptable. It relocates the problem. The problem was never that somebody heated a stone. It is that you were not told, and the telling is where the last two thousand years made their difference.
Fifteen symbols, one shelf over
The American Gem Trade Association publishes a Gemstone Information Manual, revised into a seventeenth edition in August 2026 and submitted to the Federal Trade Commission for informational purposes, and near the front of it sits a list of symbols for what has been done to a stone.
There are fifteen, and fourteen of them are treatments. N is the odd one out: natural, meaning “no modification (or currently has no known modification process)”. ASBL is for assembled stones, “doublets, triplets or assembled materials of all types”, which is Pliny’s three-part stone. B is bleaching. C is coating, covering “lacquering, enameling, inking, foiling or sputtering of films”, and foiling is on that list because it never left. D is dyeing, F is filling, H is heating, HP is heating and pressure, HR is heating with residue, I is impregnation, L is lasering, O is oiling and resin infusion, R is irradiation, U is diffusion, W is waxing.
Fifteen symbols, and one of them means nothing was done. Pause on that, because the crystal trade has no agreed way of saying it.
The letters are only part of the grammar. The same manual runs three more scales alongside them. Frequency, from rarely through occasionally, commonly and usually to always, with unknown available when nobody is sure. Stability, from excellent down through very good, good and fair to poor, with variable for materials that will not sit still. And care, in three steps: normal, special and extra special. A single row therefore tells a dealer what was done, how often it is done to that material, whether it will last, and what the owner has to do differently.
The manual also fixes what permanence means, and it does it by behaviour rather than by chemistry. An enhancement counts as permanent “as long as the effect of the enhancement does not change under normal wear, cutting, repair, cleaning or display conditions”, and where it is not permanent, “appropriate disclosure is required”. That is a definition a buyer can use. It does not ask whether a colour sits in the lattice or on the surface. What it asks is whether the stone will still look like this after you have owned it a while and cleaned it a few times, and that is the question you had anyway.
There is a companion rule for stones that have had more than one thing done to them. “Multiple enhancement techniques are sometimes applied to the same material. All treatments must be listed”, and the manual’s own example is a diamond marked LF for lasering and filling. Nothing gets to hide behind the more respectable of two letters.
And the manual is candid about the softest part of its own chart. One sentence there we would put in front of every crystal writer: “Designations are based on a consensus of opinion rather than any available documentation.” Its frequency column, it says, “represents a reasonable estimate of how commonly a particular enhancement process is utilized in the trade, based on a consensus of opinion.” A trade body publishing the industry’s reference chart tells you which of its columns are measurements and which are the collective impression of experienced people. That is a standard of care worth borrowing. Any confident figure about how often crystals are treated should be traced back far enough to see whether it was ever more than that.
Assuming it was treated
The manual’s most useful idea is a default rather than a letter, and it reverses the posture a crystal shelf teaches.
“Since many enhancements are difficult or impractical to prove definitively,” it says, “the approach taken in this manual is, unless otherwise indicated, to assume that such traditional enhancement has been done to that particular gemstone. This assumption has been made in order to protect both the seller and the consumer.”
Assume it happened. That is the trade’s own working position, and note who it protects: the seller as much as the buyer, because a seller who cannot prove a negative should not be asked to.
Which turns untreated into a claim rather than a description, and the manual handles it as one. The N symbol appears on the chart only for materials not currently known to be enhanced. A seller may still use it for one specific stone, on two conditions. They must “provide a guarantee that there has been none”, and the stone “must be accompanied by a commercial document, such as an invoice, a memorandum and/or a laboratory report, to support the fact that the gemstone is not enhanced.” Nothing was done to this is a sentence somebody has to be willing to sign, and the paperwork is the signature.
The codes are also not what a customer is owed, and the manual is blunt about it. Its worked example is a ruby marked F, whose surface cavities have been filled with glass. Within the industry, it says, that stone must carry the letter. Then: the information “must also be provided to the consuming public in writing using plain language. Abbreviations and codes are not sufficient.” The letters are shorthand between dealers. The person at the counter gets words, in writing, and a seller who hands over a code instead has met the trade’s obligation to the trade and not the one to you.
Here, then, is what actually changed in two thousand years, and it is not a small thing. Pliny put the work on the buyer. He wrote a test list and expected you to use it, before ten in the morning, with a flake of obsidian to hand. The modern trade puts the work on the seller. It asks for a declaration in plain language, from whoever is holding the stone at the time of sale. That is a genuine improvement and it is most of a solution.
Where it thins out is at the crystal shelf. The manual binds the trade that publishes it, a crystal shop is not usually inside that trade, and a buyer standing in front of one is closer to Pliny’s position than to a gem dealer’s, with a worse test list than Pliny supplied.
Our standard
Here is the bar, then, and it is short enough to carry into a shop.
Assume the stone was treated, exactly as the trade assumes it. That is the calmest starting position available, and it takes away the least pleasant part of buying crystals, the feeling that you are supposed to be catching somebody out.
Ask for the sentence rather than the absence of one. Untreated is something a seller says, and a seller who means it will write it down. Wanting it written down is not an awkward request, because it is what the coloured stone trade already requires of itself.
Expect plain language and not a symbol. A seller who does use the letters is telling you something good about where they buy their stock, but the sentence you are owed is the one in ordinary words.
At Crystalance we write what is known about a material and what the trade does to it, and where a frequency is somebody’s estimate rather than a measurement, we say so and name whose estimate it is. We do not print a treatment claim we cannot source. The composition and hardness that decide what a treatment can even do to a given stone sit in the Crystalance Mineral Library. The wording sellers use around all this is taken apart in what “natural” actually means, and the separate question of which materials get faked outright rather than treated is handled on its own.
The scribe in Alexandria was writing down what worked. Pliny was writing down how to catch it. Both of them assumed a reader would need to know, and after three thousand years the only thing worth adding is that you should not have to work it out alone.
Sources
- Kurt Nassau, “The Early History of Gemstone Treatments”, Gems & Gemology, Spring 1984, vol. 20 no. 1, for the Stockholm Papyrus and its 73 recipes, Lagercrantz’s dating, the bleaching recipe, the heated carnelian in Tutankhamun’s tomb, the Minoan dating of metal foils, and the assessment of Pliny’s techniques still in use: gia.edu
- Pliny the Elder, Natural History, book 37, sections 195 and 197 to 200, Rackham, Jones and Eichholz translation, for the boiling in honey, the stones sold as “physis”, the three-part stone, the treatises on counterfeiting, and the tests for a false gem: attalus.org
- American Gem Trade Association, Gemstone Information Manual, 17th edition, August 2026, for the fifteen enhancement symbols, the frequency, stability and care designations, the definition of permanence, the rule that all treatments are listed, the assumption of traditional enhancement, the conditions on the N symbol, and the requirement of plain language: agta.org
