by Decoding the Heavens- Solving the Mystery of the World's First Computer (retail) (epub)
That’s why the bronze statues brought up from the Antikythera wreck were quite well preserved – once they were cleaned, the original form of the ancient figures was revealed. But the chemicals formed by the corrosion of bronze can turn nasty. Copper chloride is a stable compound in water, but not in air. When objects that have been corroded in this way are removed from the sea, the copper chloride reacts with oxygen and moisture in the air to form hydrochloric acid. The acid attacks the uncorroded metal beneath to form more copper chloride, which again reacts with air to form more acid, and the cycle continues. If the reaction isn’t stopped, the object slowly and inexorably self-destructs.
For many months before it was discovered the mysterious Antikythera mechanism sat in a crate in the open courtyard of the National Archaeological Museum in Athens, unnoticed, untreated and literally eating itself away. By the time an unnamed museum worker noticed the significance of the decaying, fractured lump and brought it to the attention of the museum director Valerios Staïs, the outer layers of bronze had been completely destroyed.
Shrivelled fragments of wood clung to the bronze pieces, suggesting that the object had once been housed in a wooden box about the size and shape of a squat dictionary. Perhaps as the water evaporated out of the wood, the force of shrinking had literally pulled the contents apart. Perhaps a museum employee, eager to see what was inside, had hit the lump with a hammer. Either way, it was now in four crumbling bits.
An ugly layer of limestone – mostly calcium carbonate deposited as sea creatures feeding on the wreck died – covered much of the outer surfaces. But where the lump had cracked open patches of colour revealed the army of reactions devouring the bronze. The whitish green and bright blue-green of different forms of copper chloride dominated, but snaking through the green Staïs saw streaks of brownish-red copper oxide, the brown-black and whitish grey of various forms of tin oxide, and even the yellow and blue-black of tin and copper sulphide. Although a small core of metal remained in the centre, the surface of the fragments was made up of a powdery material that fell away at his touch.
Staïs was enjoying an impressive career. He was originally from the rugged island of Kythera, just north of Antikythera. Like his Uncle Spyridon, the Education Minister who first received news of the Antikythera shipwreck from Captain Kontos, Valerios had travelled to mainland Greece as an ambitious young man. He studied medicine, then archaeology, and he became director of the prestigious National Archaeological Museum of Athens aged only 30, just in time for the completion of the museum’s first permanent home in 1889. Since then the new buildings had been filled with ancient Greek statues, tools, weapons, pots, jewellery, and not least the fabulous finds from Antikythera, which, over the past few months – incredible, chaotic, wonderful months – had gained international fame for him and his museum. But with all the precious artefacts that had come through the museum’s doors, Staïs had never seen anything like this.
It was clockwork. Ancient clockwork. The largest piece of the strange object was about as wide as a page in a book, and almost as tall. One corner might once have been square, but the other sides were rough and eroded. Limestone formed an uneven layer over much of the front, although through it the outline of long-buried yet modern-looking gearwheels could still be made out. The overall effect was eerie and otherworldly, like finding a steam engine on the ancient, pitted surface of the Moon.
The clearest structure was a large wheel, almost as wide as the fragment itself, with a square hole in the centre where an axle might once have sat. Triangles had been cut out of the middle of the wheel, so that four unequally broad spokes formed the shape of a cross. Around the edge were about 200 tiny jagged teeth, carefully cut into triangles by some ancient hand and so small that they could only be counted with the help of a magnifying glass. A second, smaller, toothed wheel on the same side looked as if it might have engaged with the first, and there were hints, harder to make out, of other much smaller wheels or circles.
On the other side of this largest fragment, several more cogwheels were visible, with yet more little teeth – freshly revealed where the object had broken open and stunningly sharp and precise. Two medium-sized wheels lay one on top of the other, the upper one slightly offset from the lower, and several much smaller wheels were also visible, as well as a square peg. A thin, flat sheet of bronze appeared to be stuck over the bottom right-hand corner and it carried the remains of a Greek inscription. Strings of capital letters in a miniature, precise hand were so worn as to be practically unreadable, but they tantalisingly filled line after line without a single gap, as if the message had been too urgent to afford any pause between words.
A second slightly smaller fragment also had a flat sheet stuck on to one face, engraved with another inscription. On the back of it had been cut a series of concentric circles, which looked as though they might have served as guides for a rotating pointer. Rocky deposits completely covered the front of the third fragment, but on the back of it was part of an illegible inscription, as well as a raised ring that intersected with another raised, curved edge. An inscribed letter ‘Τ’ was just discernible inside the ring, and something that looked like a movable pointer projected from the centre. The surface of the fourth fragment was completely eroded, but from the size and shape it looked as though it might contain a lonely cog.
The number of gearwheels and the precision with which they were cut, along with the presence of various scales, pointers and inscriptions, perhaps instructions, immediately suggested to Staïs that this was a mechanical device for making accurate measurements or calculations.
But it couldn’t be. The pieces crumbling in his hands had to be 2,000 years old and nothing like that had ever been found from antiquity. The ancient Greeks (or anyone else around at the time) weren’t supposed to have had complex scientific instruments, or even, according to many scholars, any proper science at all. And clockwork wasn’t supposed to have been invented until the appearance of, well, clocks, in Medieval Europe more than a thousand years later.
It’s hard to overestimate the uniqueness of the find. Before the Antikythera mechanism, not one single gearwheel had ever been found from antiquity, nor indeed any example of an accurate pointer or scale. Apart from the Antikythera mechanism, they still haven’t.
Ancient texts reveal slightly more, although with written descriptions it’s harder to tell how the objects being described actually worked, or whether they were ever made. Also you often have to rely on texts in which the writer is describing something long after the event, or texts that have been copied and recopied many times and therefore could have been corrupted. But there are a few scattered mentions of gearing. The earliest may be a treatise on mechanics dating from around 330 BC, dubiously attributed to the revered philosopher Aristotle. It discusses circles that roll in contact, pushing each other round in opposite directions. The author might be talking about gears, but there’s no mention of cogs or teeth, so it’s hard to know for sure.
The first Greeks we know of to use working gears were the two most famous inventors of the third century BC: Ctesibius and Archimedes. Ctesibius, the son of a barber, became the greatest engineer of the time that we know about – after the legendary Archimedes – and he worked in Alexandria – in fact, he was probably the first director of the famous museum there. None of Ctesibius’ writings survive, but we hear a lot about him from later authors, such as the Roman architect Vitruvius, writing a couple of hundred years later. Vitruvius said that Ctesibius built a water clock in which a float that rose with the water level moved an hour pointer by means of a ‘rack-and-pinion’ gear. This is a set-up in which a single gearwheel engages with a flat, toothed rack, and it’s used to convert linear motion into rotational motion or vice versa.
Archimedes lived in the rich city state of Syracuse on the island of Sicily, although during his youth he almost certainly worked in Alexandria with Ctesibius. Among many other things he is credited with the invention of the wonderfully named ‘endless screw’ – in which
a threaded screw is used to engage a toothed wheel with a much larger gear ratio. One full turn of the screw only turns the wheel through one tooth’s worth of rotation – meaning that a lot of gentle winding turns the wheel only a small distance, but with a much stronger force than that originally applied by the winder. According to the ancient historian Plutarch, such a device allowed Archimedes to impress Syracuse’s king by single-handedly dragging a ship over the ground, ‘as smoothly and evenly as if she had been in the sea’.
Another, slightly more complicated device described by Vitruvius was a distance-meter or odometer, based on the principle that chariot wheels with a diameter of about four feet would turn 400 times in one Roman mile. For each revolution a pin on the wheel’s axle engaged a 400-tooth cogwheel, moving it around the equivalent of one tooth, so that the cogwheel made one complete turn for every mile. This wheel engaged another gear with holes along its circumference that held pebbles, so that as the gear turned the pebbles dropped one by one into a box. Counting the pebbles therefore gave the distance travelled, in miles.
Perhaps Roman chariot drivers charged by the mile. We don’t know for sure that the devices were ever built, but the general idea seems sound enough and they may have been around much earlier than Vitruvius’ time. Alexander the Great was accompanied on his campaign in Asia in the fourth century BC by ‘bematists’, who had what must have been one of the most boring jobs in the ancient world – counting their steps to measure distances. Their accuracy even over journeys of hundreds of miles (they were often less than 1 per cent out) has led to suggestions that they must have had mechanical odometers to help them.
The height of invention in ancient Greek gears was supposedly reached by the instrument-maker Hero, another follower of Ctesibius and a lecturer at the Alexandria museum some time later in the first century AD. Hero wrote about the principle that Archimedes had started to develop of using gearwheels of different sizes to change the strength of an applied force.
In particular, he talked about a weight-lifting machine called a baroulkos. He drew a picture of it, showing how a series of gearwheels of increasing size would allow a relatively small force to lift a heavy weight. There’s no evidence that it was anything but an armchair invention – indeed many scholars have argued that the teeth wouldn’t have been strong enough for the device to work in practice – but the description alone proves that the principle of intermeshing gearwheels was understood. Another device described in detail by Hero was an elaborate dioptra or sighting instrument, which used an endless screw and cogged semicircle to allow it to be aligned accurately.
So we know that the Greeks used toothed gearwheels in simple mechanical devices from around 300 BC onwards. But most of these devices involved just one or two wheels that engaged with a screw or rack and they didn’t need to be particularly precise, they just needed to apply force or lift a weight. Even so, Hero has been seen as an aberration in the history of technology: a genius who did not represent his age but described mad, impossible devices far beyond the comprehension of his peers. One eminent publication on the subject from the 1950s describes Hero’s dioptra as ‘unique, without past and without future: a fine but premature invention whose complexity exceeded the technical resources of its time’.
But compared to the baroulkos and dioptra – which were supposedly so far ahead of their time – the gearing in the Antikythera mechanism was undoubtedly real, and its complexity was breathtaking. These were precisely cut bronze gearwheels, clearly meant for some mathematical purpose. It was hard to count the gears embedded in the battered fragments, but to Staïs and his colleagues at least 15 wheels were visible on the eroded surfaces alone. They seemed to have interacted to make certain numerical calculations, the answers to which would have been displayed via pointers on a scale.
Rather than anything the ancient Greeks were supposed to have built, the sophistication of this mechanism made it look more like a clock or calculator. But if so it had to be nearly 2,000 years ahead of its time. Mechanical clocks of such a small size required delicate springs and regulators and didn’t appear in Europe until the fifteenth century, and the first mechanical calculators – complex contraptions that used metal gearwheels to add, subtract, multiply and divide – weren’t devised until some 200 years after that.
Today we’re so used to electronic computers and calculators that the idea of a calculation using metal gearwheels might seem bizarre. Imagine, for example, that you have a gearwheel with 20 teeth that engages with a gearwheel with 10 teeth. Each time you turn the first gearwheel through one complete revolution, the second one will turn twice. In other words, your input has been multiplied by two (actually the second wheel turns in the opposite direction to the first, so you could argue that the input has been multiplied by minus two, but you get the picture). This is part of what clocks do – converting the seconds that tick by into the minutes and then the hours of the passing day. The more gearwheels you have, in series or in parallel, the more complex the calculation you can make.
The idea that an ancient clock or calculator might have been found caused excitement and some consternation at the Athens museum. Realising that interpreting it was beyond his expertise, Staïs quickly called in two experts. The first was John Svoronos, director of the National Numismatic Museum in Athens – the keeper of the nation’s ancient coins and an expert in reading ancient lettering. Svoronos was one of the most senior archaeologists in the country and hugely knowledgable; unfortunately, he was also prone to coming up with eccentric theories about which few dared to disagree. The second expert was Adolph Wilhelm, a young and brilliant Austrian expert in inscriptions, who was stationed in Athens at the time.
Over the next few days Wilhelm cautiously dated the writing on the mechanism to somewhere between the second century BC and the second century AD. Meanwhile, Svoronos and some of Greece’s top scholars exchanged rival and rather pompous articles in the national press, hotly debating what the bizarre instrument might be – their talk of cogs and scales appearing alongside reports of Cuba’s newly won independence from the United States and Britain’s violent takeover of South Africa. Then the initial excitement died down, as the experts each went away to write up their various theories for scholarly publication.
Svoronos got there first, in a 1903 report written with Pericles Rediadis, a professor of geodesy and hydrography (fields concerned with measuring the Earth and the sea). Rediadis, a senior member of the Archaeological Society of Athens, was also interested in naval history and he was well known for his studies of the site on which the famous Battle of Salamis (480 BC) was fought.
Svoronos pored over the cryptic inscriptions on the Antikythera mechanism with a magnifying glass. He was able to decipher 220 scattered Greek letters, though very few whole words, and he compared their style to those on the ancient coins that he knew so well. He overruled Wilhelm’s opinion on the mechanism’s age and announced instead that the writing dated from the first half of the third century AD, a turbulent time of civil war when the Roman empire, including Greece, was ruled by a succession of leaders who each briefly seized power before being brutally assassinated.
Meanwhile Rediadis provided a description of the Antikythera fragments – the first technical, if rather vague, account of what he called ‘this completely strange instrument’. He noted that the mechanism had been carried in a wooden box, as nautical instruments on ships still were in his own time, and deduced that the object was not part of the Antikythera ship’s cargo, but a navigational instrument used by the crew.
From the scraps of lettering deciphered by Svoronos and Wilhelm, Rediadis suggested that the inscriptions were operating instructions, and put great importance on one particular and very unusual Greek word: μοιρογνωμóνιον. This is a technical term referring to a graduated scale. The word was used to describe the zodiac scale in the earliest known account of the astrolabe, written in the sixth century AD. Svoronos and Rediadis concluded that the Antikythera mechanism must there
fore be a kind of astrolabe.
Astrolabes were among the cleverest instruments thought to have been around in antiquity, and they were calculators of a sort. They were used for solving problems relating to the time and the position of the Sun and stars in the sky, and they were popular until the seventeenth century or so, when increasingly accurate clocks and astronomical tables began to render them obsolete.
The essence of an astrolabe, however, was something that the new technologies could never replace. The name means ‘star catcher’ and it is apt: holding the engraved, metal circle of an astrolabe you have the whole of the heavens in the palm of your hand. From Aristotle’s time onwards, it was accepted (with just a few dissenters) that the Earth lay motionless at the centre of the universe, with the Sun circling around it and a sphere of fixed stars rotating behind that. An astrolabe is a flat disc in which one circular plate rotates over another to represent in two dimensions the spinning heavens as seen from Earth. The Sun, stars, horizon and even the very sky itself are represented by intricate patterns on its face. The inscriptions look complex and alien to us today, but they are the result of centuries of astronomical observations, and they elegantly encode our place in the visible universe.
The circular instrument’s base – called a mater (or ‘mother’) in Latin – had a central pin over which a flat metal plate fitted snugly, like a disc on a record player. Engraved on it was a bewildering yet beautiful set of intersecting curves, lines and circles. This was a map of the sky, imagined as a sphere and projected on to the flat disc with the North Pole in the middle – just as a map of the Earth represents the spherical surface of the planet on a flat piece of paper. Although the sky looks like a featureless expanse, you can actually draw lines that mark very specific locations on it. The plate was engraved with a straight vertical diameter to show north and south, for example, and a horizontal diameter to show east and west. A series of curves and circles depicted the celestial equator (the Earth’s equator as extended straight up into the sky), the Tropics of Cancer and Capricorn, and the horizon, as well as marking various altitudes above the horizon and degrees from north. The position of these lines is dependent on how far north or south of the equator you are, so most astrolabes came with a series of interchangeable plates, each engraved for a specific latitude.
