The Antikythera Mechanism: The 2,000-Year-Old Computer

Discovered in a 1900 shipwreck, the Antikythera Mechanism is an ancient Greek bronze device revealing a 2,000-year-old astronomical computer with intricate gears.

In the spring of nineteen hundred, a boatload of sponge divers from the island of Symi was blown off course in the Aegean Sea. They took shelter near Antikythera, a small, rocky island between Crete and the Peloponnese. When one diver came back up from the seabed, he was shaken. He said he had seen dead bodies below.

They were not bodies. They were bronze and marble statues from an ancient shipwreck, lying about forty-five meters, or one hundred and fifty feet, underwater. The Greek authorities soon organized a salvage operation, and throughout nineteen hundred and one, divers worked the wreck in heavy canvas suits, brass helmets, and lead boots, with air pumped down from the boat above.

It was dangerous, exhausting work. One diver died, and others were injured by decompression sickness, then poorly understood. But the finds were extraordinary: statues, glassware, jewelry, amphorae, and fragments of fine furniture from a Roman-era cargo ship that had probably gone down in the first century before Christ.

Among the treasures was a lump of corroded bronze and rotted wood, green as old pennies and easy to overlook. At first, it looked like scrap. Then, after months in the museum, it cracked open. Inside were thin bronze plates, tiny writing, and the unmistakable teeth of gears.

That was the moment the story became strange. Greek statues made sense. Luxury cargo made sense. But precision gearing from an ancient shipwreck seemed almost impossible, as if someone had found the inside of a clock in a world that should not yet have had clocks.

The fragments went to the National Archaeological Museum in Athens. Scholars could see wheels and inscriptions, but corrosion had fused everything into a puzzle box. Was it a navigation instrument? A calendar? A planetarium? For decades, no one could prove the answer.

The first major breakthrough came from Derek de Solla Price, a British-born historian of science who worked in the United States. In the nineteen fifties and sixties, he studied the fragments with the Greek physicist Charalambos Karakalos. Using X-ray imaging, they saw that the object was not a decorative fitting or a single gear. It was a compact machine with many interlocking wheels.

Price published his great study, “Gears from the Greeks,” in nineteen seventy-four. He argued that the device was an ancient astronomical calculator. Turn a handle, and it modeled cycles in the sky. At the time, that claim sounded daring. Later research has shown that Price was right about the main idea, even if many details have since been revised.

The next leap came in the two thousands, when the Antikythera Mechanism Research Project used high-resolution X-ray computed tomography and advanced surface imaging. Think of a hospital CT scan, but aimed at a fragile block of ancient metal. The scans revealed hidden gear teeth and thousands of letters, some less than two millimeters high.

Today, about thirty bronze gears are known to survive, though the original machine probably had more. It was roughly the size of a shoebox, set in a wooden case, with dials on the front and back. A user turned a handle. Pointers moved over scales. In plain terms, it was a mechanical model of time.

On the front, it showed the calendar year and the positions of the Sun and Moon against the zodiac. One remarkable gear arrangement reproduced the Moon’s uneven motion, because the Moon does not appear to move through the sky at a perfectly steady speed. The mechanism used a clever pin-and-slot system to imitate that irregular rhythm.

On the back were spiral dials for longer cycles. One followed the Metonic cycle, a nineteen-year pattern that helps align lunar months with solar years. Another followed the Saros cycle, about eighteen years long, used to predict eclipses. The eclipse inscriptions could indicate whether an eclipse would be solar or lunar and give information about its timing.

This was not magic, and not fortune-telling. It was mathematical astronomy translated into bronze. Another small dial tracked the cycle of major athletic festivals, including the Olympic Games. That detail makes the machine feel practical as well as brilliant: a portable calendar for a Greek world where religion, politics, travel, sport, and the heavens were all tied to time.

Who made it? We do not know. The honest answer is that no maker signed it. The inscriptions are Greek, and the science behind it belongs to Greek astronomical tradition. Some scholars cautiously connect it with Rhodes, an important center of astronomy and mechanical skill. Hipparchus, one of antiquity’s greatest astronomers, worked there in the second century before Christ, and his lunar theory may be reflected in the design. But that remains a hypothesis, not a proven fact.

Ancient writers show that such machines were not unimaginable. Cicero described mechanical models of the heavens associated with Archimedes and Posidonius. Those passages do not describe the Antikythera device itself, but they tell us that educated Greeks and Romans knew of geared celestial models.

So why do we not find machines like this again for more than a thousand years? Here, too, we are dealing in hypotheses. Bronze was valuable and often melted down. Workshops vanished. Patronage changed. Technical manuals on papyrus or parchment decayed. Perhaps other devices existed, but were recycled, broken, or lost.

Complex astronomical gearing appears later in the medieval Islamic world, in Byzantine traditions, and in European astronomical clocks. Yet the Antikythera Mechanism remains older by a startling margin. It reminds us that technological history is not a neat staircase. Knowledge can flare brightly, travel in small circles, and disappear.