The ancient Greeks were avid astronomers, despite not really understanding what they were observing. None knew that the visible planets orbited the Sun along with the Earth, or that the apparently stationary stars were incredibly distant suns themselves. Their ignorance didn't stop them from creating complex, empirical formulae to describe and predict the motion of celestial bodies. Astronomical cycles were serious business for philosophers and sailors alike, as the position and motion of the stars made accurate navigation possible.
The Greeks weren't particularly known for inventiveness, however. Unlike the Chinese with their handy-dandy abacus, all those brilliant Greek mathematicians and astronomers had to do their calculations by hand - or did they? Though no historical accounts from the period give any real evidence of Greek computers, the discovery of a remarkable machine in a first-century BCE shipwreck near the tiny island of Antikythera proves that they must've been onto something. A historical anomaly and a technological marvel, the so-called "Antikythera mechanism" used an elaborate system of gears to compute the positions of sun, moon and planets as a function of time.
In 1900 CE, divers searching for sponges off the coast of Antikythera in the Aegean Sea came across an ancient shipwreck loaded with artifacts, most of which soon found their way to the local market. A team of archaeologists eventually arrived at the scene and recovered what hadn't already been sold. Among statues, coins, jewels and pottery of every sort, an unidentifiable lump of bronze in a wooden frame was brought to the surface and sent to the local museum, almost as an afterthought.
Archaeologists and museum personnel were too busy with the obvious treasure hoard to pay much attention to the unidentified, but apparently prosaic piece. Eventually, someone got around to cleaning two millennia's worth of marine gunk off of it, and it became apparent that the bronze was a gearing mechanism, though Greek historians couldn't agree on just what it was. Some thought it an astrolabe made for navigational purposes, while others pointed out that the workings were far too complex for such a relatively simple tool. After half a century had passed, British physicist and historian Derek de Solla Price arrived at the museum in Athens to examine the mechanism and decipher its meaning.
With the help of a Greek who translated the inscriptions in the bronze, Price was able to determine the form and function of the mechanism. It consisted of thirty-one bronze gears, including a differential, mounted on a plate of the same material. That came as a great surprise, as differential gearing wasn't supposed to have been invented until more than a millennium after the date of the shipwreck, which was known to within about fifteen years. It had been operated by a hand crank and driving wheel, and turned three dials, one on the front of its case and two on the back.
Price originally published his findings in the June 1959 Scientific American, and expounded on them in the 1975 book Gears from the Greeks. He had discovered that the front dial traced the motion of the Sun through the zodiac, the upper back dial displayed the positions of the five known planets (Mercury, Venus, Mars, Jupiter and Saturn), and the lower back followed the Moon through the synodic year; that is, the year containing 12 lunar months. Only through use of a differential would this be possible in such a machine, since the length of the Moon's phase is a function of its position relative to both Earth and Sun, and longer than the sidereal month determined by a single lunar orbit. Far more complex than an ordinary orrery, the Antikythera mechanism was undoubtedly on the cutting edge of ancient technology.
The mechanism's purpose is no longer a mystery, but its origin is still in doubt. Price believed it to have been built in Rhodes. No records exist of its construction, but its inscriptions are very similar to passages in a book written by Rhodian astronomer Geminus, whose countryman and contemporary Posidonus made great advances in astronomy, possibly building a machine of his own upon which Geminus' was modeled. The finding of great numbers of Rhodian amphorae in the wreck strengthens Price's claim, as does Rhodes' reputation among the ancients as a great center for both natural science and invention. We may never know for sure.