When an unknown genius sat down more than 2,000 years ago to design and build an astronomical instrument, chances are good that he or she didn’t think that entire academic institutions devoted to solving its mysteries would one day be established. But such is the enduring nature of the Antikythera mechanism, the gift from antiquity that keeps on giving long after being dredged up from a shipwreck in the Aegean Sea.
And now, new research on the ancient mechanism reveals that like other mechanical calendars, the Antikythera mechanism has a “day zero,” or a minimum possible date that it can display. The analysis by a team led by [Aristeidis Voulgaris] gets deep into the weeds of astronomical cycles, which the mechanism was designed to simulate using up to 37 separate gears, 30 of which have been found. The cycle of concern is the saros, a 223 lunar month cycle of alignments between the Earth, Sun, and Moon. The saros can be used to predict eclipses, astronomical events of immense importance in antiquity, particularly annular eclipses, which occur when the Moon is at apogee and therefore eclipses less of the Sun’s surface.
The researchers looked at historical annular eclipse data and found that saros cycle 58 had a particularly long annular eclipse, on 23 December 178 BCE. The eclipse would have been visible at sunrise in the eastern Mediterranean, and coupled with other astronomical goodies, like the proximity to the winter solstice, the Sun entering Capricorn, and the Moon being new and at apogee, was probably so culturally significant to the builder that it could serve as the initial date for calibrating all the mechanisms pointers and dials.
Others differ with that take, of course, saying that the evidence points even further back, to a start date in the summer of 204 BCE. In any case, if like us you can’t get enough Antikythera, be sure to check out our overview of the mechanism, plus [Clickspring]’s exploration of methods perhaps used to build it.
In all likelihood identical copies of the Antikythera mechanism were expensive but relatively commonplace. It demonstrates refined construction techniques and you don’t get those on a one-off prototype. You get it from building a lot of things based on the same technology.
This.
The significance of this artefact is that it suggests a continuity of sophistication: it’s almost certainly not a one-off as it would have required tooling to make that would have been used to make others like it. At the very least the sophistication strongly suggests undiscovered predecessors.
It demonstrates a level of manufacturing skill which was lost in Europe until it suddenly re-appeared among 14th century (AD) clock-makers. It suggests the skill was not lost at all, but kept alive among Eastern civilizations for over a thousand years while Europe went dark, and then “re-discovered” by Europeans trading with those Eastern civs.
As an astronomical instrument, it’s not all that useful, not because of mechanical or manufacturing issues, but because it assumes a geocentric solar system instead of a heliocentric one, so the accuracy is poor.
Worth noting Aristarchus of Samos proposed a heliocentric model 1300 years before Copernicus and Kepler (and ~50 yrs before the Antikythera mechanism) but nobody listened…
Thales of Miletus predicted a solar eclipse during his life born 636 died 545 bc.
If that were the case then don’t you think we would have found a few of them?
actually, if they only made say, less than 20, we were lucky to have found even one.
I disagree with both of you.
The first one is always bespoke. You don’t go from 0 to mass production. Mass production is less refined than the one-off watching-every-detail-to-make-sure-it-works version.
Just because there was a lot of something does not mean you will find a proportional or inversely-proportional number of them now.
15 million Model T’s were sold, how many are still around a century later?
If an artificer make a couple hundred of these mechanisms, I think the odds of finding several of them after 2000 years is pretty slim.
^This.
The fact that we’ve even found one of them in modern times is nothing short of miraculous.
Depending on originality or faithfulness to manufacturing specs: couple thousand perhaps. Modified:many more. Typically about 1% more or less.
Means you need a thousand of these made to have more surviving examples which may be improbable
We’ve lost over 99% of all literature written during the middle ages. We’ve lost over 99.9999% of all household goods used during the same period.
Even if you limit yourself to durable metallic weapons, far more than 99% of those made before the 20th century are gone.
Your point is valid, but the odds you cite are far too good.
There’s still an estimated 100,000 Model T Fords in the world, or about one in 150 ever built.
Given the same odds for the Antikythera mechanism, and 20 centuries of time, the number of devices originally built would be (1/150)^-20 which implies that there were originally 3 x 10^43 units built for every one that survives today.
If instead we insist that there were every only a couple hundred ever made, then finding just one after 2,000 years suggests the odds are about 1:1.3 over a century, which implies that we should have read reports of the existence of several such devices from medieval literature.
Another interpretation of the result is that applying simple odds to the case isn’t very relevant.
LOL. Love your math, but yea.. lot has happened in 2000 years.. So many comments are based on modern reproduction techniques and global population. Reminder that the population has doubled in 40 years. So there were WAY WAY fewer people around. But, the people who were around were making really spectacular things, because craftsmen were much more prevalent.
Perhaps a better way to run thee odds is to look at how many of those ships survive in any sort of form, then work out how many might have had the luxury of such a mechanism (I’m assuming this was not a standard fit on most ships) and then how likely we are to stumble across even one of them.
A lot of the comments also ignore possibilities for damage, destruction, salvaging/re-purposing of parts, or just melting down for the metal content. If there’s one guy or one workshop on the planet making & repairing these things, and it breaks or is damaged, or the dude who knows how it works dies – the unit becomes an expensive ornament at best.
There’s also stuff throughout history like those marvellous religious folk who would consider this some sort of affront to God and try to have it destroyed / purged.
Comparing these things to Model T Fords feels a bit off, they clearly would not be “mass” produced – perhaps they’re more akin to early supercomputers. We’ve only got one Colossus and a few Bombes (even then, only after a massive rebuild effort) despite there being many hundreds built. How many Cray-1’s still survive ot of the many that were built?
I agree whole heartedly! Thx 🙏🏻, for spreading the knowledge. It keeps informed, & prepatred.
Maybe it was a prototype and they refined the techniques while building it.
The technology was later forgotten and didn’t return to that level of sophistication in Europe until the 14th century. I think that rules out the idea that it was most likely commonplace.
Maybe it was a prototype and they refined the techniques while building it.
The technology was later forgotten and didn’t return to that level of sophistication in Europe until the 14th century. I think that rules out the idea that it was most likely commonplace.
I misread the title and thought they had discovered an 0-day. (A brand new security exploit) Which would have been completely hilarious.
I had the exact same interpretation, and then laughed at myself for a bit.
“Hacker finds new antikytheria exploit, Greeks HATE him!”
The device had at least 30 gears, the largest of which had 223 teeth.
Those teeth could have all been cut and filed by hand, but the person doing it would have quickly realized they could have hobbed them on a machine with improved accuracy and less work.
Such a hobbing machine is far less complicated than the mechanism of the A.M., so on balance it is more likely they did it mechanically. And if they did it mechanically they could have made more.
That doesn’t mean they were mass produced. Likely a single workshop was making them, and they would have been very expensive – same as 14th century clocks. And while there are wealthy nerds in every age, they is a limited market for such a device: Syracuse (where it was probably made) in 200 BC had a population of maybe 30’000 people, only 20-30 who could afford one, and maybe 4-5 who actually wanted one. So, a market for 50 total in the entire Mediterranean world, and most potential clients would never had learned of the availability anyway. So maybe 10 or 20 were ever made…Yes, those figures are guesses.
My main point was that it is so complex that it definitely wasn’t the first device made by that workshop, and probably not the last. But the likelihood of ever finding one of these devices is basically zero. The one we have was found by chance at the bottom of the sea. Not impossible that we find another, but 22 centuries is an ocean of time for an artefact to cross to reach us.
Sic biscuitus disintegratum.
The difficulty is coming up with a machine that is repeatable enough. If you end up off by half a gear tooth by the time you’ve gone around the circle, you just made a very expensive error.
The accuracy of the cut teeth is less important than their accurate placement around the circle, so they would have spent a lot of effort measuring and marking first, and the job of actually filing the gear would have been the easier part given to some apprentice.
Hand filed seems far more plausible to me – its so easy to hand file and unless you are setting up to make hundreds of them the material and time investment to make any form of mechanical gear cutting aid, with all the adjustability to do all the various gears in the mechanism would be wasted effort, probably take you longer to make the the gear cutting machine work reliably than make a good number of geared mechanisms by hand…
When you look at just how much hand finishing and entirely hand crafted parts are in things like firearms and clocks well into the 1800’s (arguably even into the middle of the 20th century) despite a very developed industrial base I really can’t see such niche production from so very long ago being more mechanically mass produced… I have no doubt the idea of a gear cutting system would occur to folks back then, as they are clearly well developed in mechanical thinking, but actually building something so expensive to make another expensive device very very slightly cheaper and easier to make doesn’t seem likely.
I’m reading what it does all I could think of is a similar device in avatar the last air bender for predicting eclipses.
23 December 178 BCEFGHIJKLMNOPQRSTUVWXYZ!@#$%^&*()_+…….
23 December 178 B.C. <– much better
Not everyone is a theist and from that group not everyone is a Christ.
The IP was bought by the Swiss watchmakers and buried in the vaults of the Gnomes of Zurich….
So when is year zero?
Clickspring got a mention! Cool!
I traveled to Greece in 1972 . I read about it and I wanted to go see it. My wife and I flew to Amsterdam and found a bus to take us to Greece. $30. The device was in a glass case at a university. It was a lot of turmoil going on in Greece at that time. There is a military dictatorship that was not very popular. The students were rioting in the military ended up running a tank through a wall in front of the university. But I did get to see it in person. I’ve read about this all my life and thank you for all the information you guys have given me