FREQUENTLY ASKED QUESTIONS
Q: How can we be sure the Mechanism is not a fake? A: Since the wreck was only found in 1901, the remains of the Mechanism cannot have been dropped into the water before then. The history of the Mechanism within the Antikythera collection since 1901 is well attested in newspapers, early photographs, academic research. etc. Hence there are only two possibilities: Either it could be a forgery made c.1900 and deliberately inserted into the Antikythera collection shortly after that date, or It is a piece of antique clockwork that has accidentally made its way into that collection. The first possibility is ruled out by the fact that the depth and density of marine encrustation on the exposed surfaces means that it must have lain under the sea for at least several centuries. The second possibility requires the maker to have had the skills and knowledge necessary to produce the instrument. The Mechanism is densely inscribed in Greek letters characteristic of the 2nd century BC. These inscriptions include technical terminology and astronomical/calendrical data which was only previously known in a very small number of texts from quite disparate disciplines. For example, the distribution of skipped days on the Metonic dial; terminology such as the ancient Greek words for “bearing”, “pointer”, “disk”, and the names of regional athletic games. It would have taken a tremendous effort of scholarship in the 17th or 18th century to create authentic inscriptions which stand up to close scrutiny. But this second possibility can be completely ruled out by the fact that the Mechanism incorporates calendar data which was not known until the early 1970’s. Until 2008 the only known Corinthian calendars were for Epirus and Taormina. Even as late as 1972, only eight names of the months in the calendar for Epirus were known, and the full calendar was only finally resolved in 2003. Although known since 1833, only seven names in the Taormina calendar match those on the Mechanism. So the Mechanism incorporates data and terminology which has been confirmed only recently to be authentic but was not known before 2000. (These arguments are documented fully in the recent paper "The Epoch Dates of the Antikythera Mechanism" written by Alexander Jones).
Q: How reliable is the reconstruction? A: Correctly built, it's pretty reliable. The first machines were delivered in 2019 with no issues reported. Since then we’ve improved the design, adopting harder-wearing nylon for the gears and metal plates for the front and back of the planetarium gearbox. The instrument comes with a 2-year guarantee but used with care it should last indefinitely.
Q: How accurate is the reconstruction? A: As accurate as possible, given the stage of development of the Greek astronomical theories at that time. The calendar calculator is exact to within a tiny fraction of a day over the 76-year Callippic cycle. The position of the Moon is accurate to within a day of its motion. In the case of eclipse predictions, a well-known characteristic of the Saros is that it “drifts” slowly, so the instrument should give accurate forecasts for the first two cycles, i.e. 37 years, but the accuracy of the timings will then start to deteriorate. The accuracy of the calculation of a planet’s position then depends on many factors including the eccentricity of its orbit, perturbations caused by the gravitational pull from other planets when they align, etc. The kinematic theories of the Greek astronomers do not take these factors into account, but they give a reasonable first approximation of the position of most of the planets, most of the time - sufficient to enable an observer to find the planet in the night sky. The Babylonian and Greek astronomers were aware of variations of a planet's position within its synodic cycle, which is particularly evident for Mercury and Mars, but it was only satisfactorily incorporated into the Greek theories by Ptolemy, some 300 years later. Our goal is that the reconstruction should be accurate to within +/-5° over 20 years, but at times the position of Mars may sometimes deviate from true by as much as 30° because of these synodic anomalies.
Q: What is the design of the planetarium gears based on? A: The remains of the planetarium gears have never been found at the wreck site, but several different researchers have proposed possible solutions for the gear ratios, based on the remains discovered, and the text inscribed on the machine. Until more evidence is found no-one can be sure which design is right. We have taken the best elements of the two most recent proposals, one by Michael Wright for the inner planets and one by Tony Freeth for the outer planets and combined them to give the most reliable and accurate configuration overall.
Q: Why is the key or handle in a different position to the original? A: In the ancient Greek design gear B1 is driven by A, a large crown gear on the side to which a handle seems to have been connected. This introduces significant stress on the gears, particularly in the B:C:D:E gear chain which drives the lunar pointer. There is a high step-up ratio to this fastest moving part of the whole mechanism and the pressure on the smallest gears is a major source of unreliability. To make our reconstruction as reliable as possible we avoid this issue by driving the machine directly from axle E, a solution first proposed by one of the researchers, Professor Allan Bromley, in 1993.
Q: How do we know it was not the first machine of its kind to be built? A: Firstly, we can identify technical solutions to issues which could not have been anticipated in an entirely original design. There are features of the Mechanism which have clearly been implemented to improve reliability, even though they may lead to other operational issues such as increased friction. Some of these could be workarounds added after the Mechanism was tested, but others could only have been built into the design to overcome defects found in earlier machines. Secondly, there are aspects of the design which could not have been invented from scratch without having been trialled in an earlier machine. For example, the design of the clips holding the pointers in place at the centre of the spiral dials for the Metonic calendar and the Saros eclipse predictor. It is a difficult engineering challenge to get these clips to perform correctly.