The Antikythera mechanism, a mysterious ancient Greek device that is often called the world鈥檚 first computer, may not have functioned at all, according to a simulation of its workings. But researchers say we can鈥檛 be sure of this since the machine is so badly damaged.
Since the mechanism was discovered in 1901, in a shipwreck thought to date to around 60 BC, researchers have struggled to work out exactly why it was built. X-ray scans and digital reconstructions show that it was originally a 30-centimetre box containing interlinked systems of bronze gears. These appear able to track the positions of the moon, sun and future eclipses on a large circular dial, as well as other features such as a calendar that includes the dates of Olympic competitions, which were related to astronomical events.
Much of the device has been corroded and damaged from centuries spent underwater, making it hard to measure exactly how the gear teeth may have functioned. In 2006, at Cardiff University, UK, and his colleagues used CT scans to try to estimate the accuracy of the device. and make its predictions fairly inaccurate 鈥� suggesting the Antikythera mechanism may have been for educational use or display rather than a working tool.
Advertisement
Now, and Gustavo Arenas at the National University of Mar del Plata in Argentina have developed a two-dimensional computational model that simulates how the triangular teeth of the gears meshed together, incorporating previous errors measured by Edmunds and others. If those earlier measurements are correct, the device wouldn鈥檛 have worked, the model suggests. 鈥淭he mechanism would not have even been able to move, because it would have jammed or also the teeth would have disengaged,鈥� says Szigety. 鈥淥ne tooth would rotate and the other wouldn鈥檛 rotate.鈥�
He thinks this conclusion is unlikely. 鈥淗ow could it be that someone invested so much time and so much effort for it to not work in the end?鈥�
The obvious alternative, says Szigety, is that the past errors were too pessimistic, and that the device in fact worked more accurately than previous research has suggested. If so, then the mechanism would have been able to predict eclipses with 鈥渧ery acceptable results鈥�, says Szigety, though it is still unclear how accurate it would have been for other uses, such as predicting the position of the moon.
However, Edmunds disagrees that smaller errors would mean the device could have been an accurate calculator. 鈥淓ven if you do come back down to smaller errors that allow it to work, then the major conclusion of my paper isn鈥檛 altered 鈥� that the lunar pointer on the front was not particularly accurate, and that indicates that it was probably a display or educational device, rather than an accurate calculator for really accurate astronomical calculations,鈥� he says.
Time spent underwater converted the bronze parts of the machine into a mineral called atacamite, which cracked and shrank when it was removed from the ocean, changing its dimensions, says at the Thessaloniki Directorate of Culture and Tourism in Greece. 鈥淭herefore, any attempt to apply precision measurements on the current condition of the gears, axes, includes the effect of the deformation. In this way, we cannot say that 鈥榓ccording to our precise measurements the mechanism never functioned鈥�.鈥�
arXiv
