Did Greek gears move the planets?

It was just a corroded lump of metal when divers brought it up from a Roman grain ship that sank off the Greek island of Antikythera around 70 AD. But it turned out to be an extremely sophisticated mechanism, a pocket planetarium that showed the dates of future eclipses and probably the orbits of the planets. The Antikythera mechanism is the subject of this year’s Bernoulli Lecture on 13 April.

The mechanism has intrigued Professor of Astronomy Rien van de Weijgaert for many years. He has studied the corroded gears, pins and inscriptions, and is currently supervising a PhD project that is trying to determine how complicated the full mechanism really was.

When it was found at 40 meters depth, the mechanism wasn’t intact. Pinholes and a very intriguing gear with 63 teeth but no connections to other parts suggest that there may have been more to it. Parts may still lie on the seabed, if they haven’t been eaten away by seawater.

‘Investigators went back to the site of the wreck last year and did another survey. Their results suggest the ship may have been twice as big as thought. But the wreck is near a cliff in the seabed and part of it may have slid down.’

In 1901, when the Antikythera mechanism was found, it was just one of many objects divers brought to the surface. ‘The National Archaeological Museum in Athens has an exhibit showing what was found: loads of amphoras, a great many marble statues, some bronzes... The inventory of the ship’s kitchen was also found. Interestingly, among the human remains was the skull of a young woman, and some very valuable earrings were found near her.’

The ship, which as it now appears may have been a grain vessel (the bulk carriers of the Roman Empire and the biggest ships of their time), must have been stuffed with goods. It is very likely that it was a used as a cargo ship to take treasures from the Greek world to Rome. ‘And judging by the earrings, we can safely say there were some rich passengers on board.’

But the biggest mystery is the Antikythera mechanism. Over the past few decades, it has become clear that it was a sophisticated astronomical computer. It is certain that it showed the orbits of the Sun and the Moon. ‘Also, inscriptions on the spirally shaped bottom dial point to the years in which solar or lunar eclipses occur.’

Judging by the sequence of eclipses, it could very well describe a range starting in 205 BC, only seven years after the death of Archimedes. ‘It has often been speculated that the idea to represent the motion of planets on a kind of planetarium originated with Archimedes, and that he may thus have established this “tradition” of mechanical models of the heavens.’

The Sun and Moon may not have been the only celestial bodies represented in the mechanism. ‘There are various artefacts on the mechanism (pins, holes, etc.) that seem to suggest there may have been more to it.’ That’s why Van de Weijgaert’s PhD student Niels Bos has set out to investigate whether the mechanism contained the orbits of the five planets known at the time. ‘The names of these planets are inscribed on the mechanism, so it does seem likely.’ For Bos, it meant trying to come up with a set of gears that would fit inside the mechanism, include a 63 teeth ‘orphan’ gear that does not seem to connect to anything at present and show the proper orbits with the same degree of accuracy as the lunar orbit. ‘That is the tricky one, because the Moon is so close. A small error in the gears would quickly result in a big deviation in the Moon’s position. But the gears are incredibly accurate.’

At the Bernoulli Lecture on 13 April, Rien van de Weijgaert will bring the audience up to speed with the latest results in Antikythera research. Bos’s calculations suggest that the mechanism probably did show the orbits of the five planets. But why talk about planets in a lecture series on mathematics?

‘It was a mathematician, Christopher Zeeman, who was one of the first to analyse the gears of the mechanism and suggest its purpose’, says Van de Weijgaert. ‘And of course, the gears that drive the orbits are a mathematical series. Niels Bos had to do some very complicated mathematical modelling to see if the mechanism could have held gears to drive the planets.’

Perhaps most important is that the Antikythera mechanism is the ultimate translation – application if you want – of mathematical objects into hardware. ‘The maker or makers realized mathematical operations were equivalent to a mechanical process, like division is to the transfer of the rotation of a gear to another gear. In other words, this is the first real step towards the definition of a computer, computation by means of a real physical process.’

The 26^th Bernoulli Lecture will take place in the Aula of the Academy Building on 13 April at 19:30. Entrance is free and the lecture will be in English.