Antikythera mechanism

Antikythera device / Antikythera computer
2026-10-07 add · 580 words

A bronze geared device recovered in 1901 from a shipwreck off the Greek island of Antikythera. It is dated to between the late 2nd century BC and the early 1st century BC^※1, and is thought to have been an astronomical calculator that computed and displayed the movements of the sun, moon, and planets, and eclipses. The next gear mechanism of comparable complexity does not appear until the astronomical clocks of medieval Europe, a gap of more than a thousand years. It is an artifact that rewrote the accepted picture of the level of ancient technology, and research and debate continue to this day.

Discovery and structure

Sponge divers found the wreck, and among the objects raised from it was a corroded lump of bronze. It turned out to be a mechanism combining more than thirty gears, and for over a hundred years since, analysis using X-rays and tomography has been carried out repeatedly.

The device was driven by a hand crank, and combinations of gears with differing numbers of teeth reproduced the periods of the heavenly bodies. Of particular note is the use of epicyclic gearing, which corrects, as a mechanism, for the fact that the apparent motion of the heavenly bodies is not uniform in speed, that is, for the irregularity of their orbits. The technical core of the device is that it embeds not merely gear ratios but the non-uniformity of the motion itself in the machine.

What it computed

It is widely accepted that it displayed the positions of the sun and moon, the phases of the moon, and predictions of solar and lunar eclipses (using the Saros cycle). It is also said to have handled the movements of the planets known at the time, and research at University College London (UCL) reports that, from evidence in Fragments A and D, a mechanism corresponding to a 462-year period relation for Venus was identified^※2.

As for the calendar dial on the front, the view that it corresponded to a 354-day lunar calendar rather than a 365-day solar calendar has gained ground in recent years. A reconstruction by Tony Freeth and others also presents a model that follows faithfully the specification in the inscriptions, in which the movements of the planets and the moon are shown by beads moving on rings.

Continuing debate

One recent point of contention is whether the device actually worked. In 2025, researchers at the National University of Mar del Plata in Argentina ran a simulation incorporating the triangular tooth shape and manufacturing errors, and announced that after about four months of turning, the gears would jam or disengage. This was widely reported under headlines asking whether it was in fact a prototype, or a toy.

The researchers themselves, however, state that the present measurements of the gears, after two thousand years of corrosion, are more likely to be what is off, and no conclusion has been reached. Another study has shown that the display for Mars could be off by up to 38 degrees, but the cause is interpreted as a limit of the astronomy of the time rather than a manufacturing defect. The work of separating problems of machining precision, deformation from corrosion, and limits of the underlying theory is still under way.

To verify

The maker and place of manufacture (there is a theory naming Rhodes, among others, but this is unconfirmed)
The assessment of the 2025 study claiming it did not work has not settled

References

Mentioned in

none

Similar