Somebody, somewhere, is being handed a garnet this January. The custom feels immemorial, and jewellers are generally happy to let it feel that way, but the list pinned up in most shop windows was agreed by a trade association meeting in Kansas City in 1912 and has been amended by committee several times since. The older story behind it is real, and considerably more tangled than the counter display suggests.

Twelve stones on a breastplate

The starting text is Exodus 28, which describes the hoshen, the ceremonial breastplate worn by Aaron as high priest: twelve stones set in four rows of three, each engraved with the name of one of the tribes of Israel. The passage is repeated almost verbatim in Exodus 39, and the stones are listed by name in both places.

The difficulty is that nobody knows for certain what those names mean. The Hebrew terms, odem and pitdah and bareketh and the rest, identified stones by colour and by source rather than by the crystal chemistry a modern gemmologist would recognise, and several appear nowhere else in the language. When the text was rendered into Greek for the Septuagint and later into Latin by Jerome, translators substituted whichever classical gem names seemed closest, and those carried their own confusions: ancient sapphirus almost certainly meant lapis lazuli, ancient topazion may well have been peridot, and jacinth, ligure and chalcedony have all shifted meaning since. Every English Bible offers a slightly different set. Any confident modern identification of the twelve breastplate stones is a reconstruction, and a disputed one.

The leap to the calendar

Nothing in Exodus links the stones to months or to birthdays. That connection was made, as far as the surviving record shows, by the Jewish historian Flavius Josephus in the first century CE. Writing about the priestly vestments, he observed that the twelve stones might be understood as corresponding either to the twelve months of the year or to the twelve signs of the zodiac. Jerome, some three centuries afterwards, repeated the observation approvingly, and from there it passed into the mainstream of Christian commentary.

It is worth noticing what sort of claim this was. Josephus was performing symbolic exegesis on a garment, in a period when educated readers fully expected sacred objects to encode the order of the cosmos. He was not describing anybody's habit of wearing stones, and he was certainly not suggesting that anyone acquire one. The entire edifice rests on a scholarly aside.

Owning twelve, wearing one

Through the medieval centuries gems were taken seriously as active substances rather than decoration. Lapidaries, treatises cataloguing the powers of stones, circulated very widely indeed; the most influential, by Marbode, Bishop of Rennes, was written around 1090, survives in hundreds of manuscripts and runs through some sixty stones and what each was good for. Amethyst against drunkenness, sapphire for cooling the passions, ruby for the blood. Crucially, these virtues attached to the stone, not to the owner's birthday.

The practice of owning all twelve and rotating through them, wearing each during the month when its power was thought strongest, appears in later accounts of medieval and early modern habits. The decisive modern move is the shift from that collection to a single stone marking your own birth month, and its origins are murkier than most retellings allow. A widely repeated version credits eighteenth-century Poland and the gem traders working there, but the documentation is thin and it is fairer to say the custom spread across Europe during the eighteenth and nineteenth centuries without a clean point of origin. What is not in doubt is that Victorian sentimental jewellery, with its mourning rings and acrostic brooches whose gems were chosen to spell out words, made personal gemstone symbolism enormously fashionable, and a birth-month stone slotted straight into that fashion.

A committee in Kansas City

By the early twentieth century American jewellers faced a practical nuisance. The lists in circulation contradicted one another, and a customer who had read one book and then walked into a shop stocked from a different one was an unhappy customer. In 1912 the American National Retail Jewelers' Association, meeting in Kansas City, adopted a single official list. That body is now Jewelers of America, and its 1912 list is the direct ancestor of the one used across the English-speaking world today.

It pays to be clear-eyed about what took place. This was a commercial standardisation, and its choices reflect what could be reliably supplied and profitably sold in 1912 at least as much as anything inherited. Encyclopaedia Britannica puts it flatly: the stones now associated with each month bear only slight relationship to the ancient beliefs, the list having been tempered by availability and cost. Sardonyx, long the stone for August, found itself sharing the month with peridot. Where the old lists favoured the obscure, the new one favoured the saleable.

Committees kept meeting. In 1952 the Jewelry Industry Council revised the list, adding alexandrite for June, citrine for November, tourmaline for October and zircon for December, in several cases giving a month two or three acceptable stones, which had the convenient effect of widening the range of price points on offer.

The most recent entries are recent indeed. Tanzanite, the blue-violet variety of the mineral zoisite, was found only in 1967 in the Merelani Hills of northern Tanzania and was named by Tiffany and Company; it became a December birthstone in 2002, thirty-five years after the mineral was first identified. Spinel joined August in 2016. A gemstone can therefore acquire an ancient-sounding calendrical meaning well within a single human lifetime, which tells you most of what you need to know about how this tradition actually operates.

Other lists, other logics

The Anglo-American list is not the only one going. Britain has historically used a variant of its own, differing on several months, and the traditional lists printed alongside the modern ones by the Gemological Institute of America preserve stones the 1912 committee set aside.

More interesting are the traditions that do not use months at all. In Indian jyotish the reference set is the navaratna, the nine gems, each tied to one of the nine celestial influences of Vedic astronomy: ruby for the sun, pearl for the moon, red coral for Mars, emerald for Mercury, yellow sapphire for Jupiter, diamond for Venus, blue sapphire for Saturn, and hessonite and cat's eye for the two lunar nodes. A stone is prescribed only after a practitioner has examined an individual birth chart, so two people born in the same month may be advised quite differently, and a stone judged unsuitable for a chart is thought to do active harm. Tibetan lists circulate as well, along with various so-called mystical sets whose provenance is uncertain and often surprisingly modern.

Then there is the plain mineralogy, which has a charm of its own. Ruby and sapphire are the same mineral, corundum, coloured differently by trace elements, so July and September are twins. Emerald and aquamarine are both beryl, making May and March another matched pair. Amethyst is quartz, the commonest mineral family on the planet, prized for a purple that comes from iron and irradiation. Twelve months do not represent twelve substances, and the collections at the Smithsonian's Hall of Geology, Gems, and Minerals make the family resemblances very plain.

Wearing one anyway

Knowing the list was fixed by a trade body obliges nobody to stop enjoying it. Most durable customs turn out to have a committee somewhere in their past, and a garnet given in January carries whatever the giver put into it, regardless of what was decided in Kansas City.

What the history offers is a better sense of proportion. Nobody in Exodus was thinking about birthdays. Josephus drew a symbolic comparison and moved on. Medieval readers wanted stones for their powers rather than their dates. The neat month-by-month scheme in a modern shop window is a nineteenth-century sentiment given a twentieth-century polish, and its newest entry came out of the ground in 1967. That is less a debunking than a lineage, and lineages, much like the stones themselves, are more interesting once you can see the inclusions.