A device that did not fit
In 1900 and 1901, an unusual object was recovered from the sea near the Greek island of Antikythera. It was a corroded lump of bronze and wood with visible gear wheels and inscriptions. Over the following century and more, careful study showed it to be a hand-cranked astronomical device, with many interlocking gears, far more complex than anything else known from the Greek or Roman world. The “mystery” in the title is mostly about its makers and its place in ancient technology, because much of how it worked has now been worked out.
The wreck and its discovery
According to the accounts, sponge divers from the island of Symi were sheltering from a storm near Antikythera in 1900 when one of them found a wreck on the seabed. The Greek authorities organised recovery operations with the help of the navy, and marble and bronze statues, pottery, glass and coins were brought up, along with the lumps from which the mechanism later emerged. The recovery took place in 1900 and 1901.
The ship is generally dated to the first century BCE, with a date of around 70 to 60 BCE often given on the basis of pottery and coins, though these estimates are approximate. It was apparently carrying luxury cargo, possibly heading west toward Italy. The mechanism itself was noticed after the fragments were in the National Archaeological Museum in Athens, when a piece of the corroded mass broke and revealed gearing. Early studies were published in the first decades of the twentieth century, but some scholars at the time thought it could not be ancient.
What survives
The remains consist of a number of fragments, commonly counted at around eighty, from a wooden case roughly the size of a shoebox. Today the pieces are held in Athens. About thirty gears survive in total, made of bronze, with teeth that appear to be triangular and cut by hand. Inscriptions on the surfaces form a kind of user manual, with text that researchers have been able to read in part through imaging.
Imaging and what it revealed
In the 1970s, the historian Derek de Solla Price and a Greek colleague used gamma radiography to study the fragments, and Price proposed a first gear reconstruction. A far more detailed study began in the early 2000s with the Antikythera Mechanism Research Project. Using three-dimensional X-ray computed tomography and a technique of reflectance imaging of surfaces, researchers could see inside the fragments and read much more of the inscriptions. Results were published in Nature in 2006, with further work reported in 2008.
From this work, the following picture is generally accepted:
- Front dial. A pointer shows the position of the Sun and Moon against a scale of the zodiac, alongside a calendar scale related to the Egyptian year. Parapegma-style inscriptions listed star risings and settings.
- Upper back dial. A spiral dial follows the Metonic cycle, the observation that 235 lunar months correspond to about 19 solar years, which allowed a lunar calendar to be matched to the Sun.
- Lower back dial. A second spiral follows the Saros cycle, a period of 223 lunar months over which eclipses tend to repeat in similar patterns, with smaller dials for related cycles.
- Games dial. A small dial records a four-year cycle naming major festivals, including the Olympic Games, as identified in 2008.
- Lunar motion. A pair of gears with a pin in a slot appears to model the Moon’s changing speed across the sky, an effect known to ancient astronomers.
The Metonic and Saros cycles show a debt to Babylonian astronomy, whose long records of eclipses are the likely source for such periods, though the mechanism’s design is Greek in its language and in its geometric approach.
What is debated
Several issues remain open. The planetary display is a good example: the lost front section probably showed the planets, but how, and which ones, is uncertain. Proposals, including a published reconstruction in 2021 by a research team in London, offer plausible gear arrangements, but they are reconstructions, not recovered parts.
The date of construction is also discussed. Estimates generally fall in the second or first century BCE, and some arguments about the calendar details point to particular places or periods. Where it was made is not settled. Candidates in the ancient world include Rhodes and other centres of astronomy, but the evidence does not allow a firm answer. Later writers such as Cicero mention devices that modelled the heavens, which suggests the mechanism was not unique in kind, although no similar object survives from antiquity.
Why it matters
The Antikythera Mechanism shows that Greek craftsmanship in metal and astronomy could reach a high level of mechanical complexity, though it does not by itself prove that such devices were widespread. The idea of an “ancient computer” is sometimes used, but a more careful description is a special-purpose analogue calculator for calendar and astronomical cycles. Studying the wreck also illustrates underwater archaeology methods, with later campaigns by Greek and international teams returning to the site since the 1970s and again in recent years, and recovering more objects. Alongside other ancient shipwreck finds like the Uluburun shipwreck, it demonstrates how sea trade carried not only goods but also knowledge and technology across the ancient Mediterranean.