Antikythera Mechanism

from Wikipedia, the free encyclopedia

The Antikythera mechanism is an ancient device comparable to a later astronomical clock . With the help of many gears and dials, it showed considerably more astronomical and calendar relationships than the corresponding clocks that were to exist in the late Middle Ages and early modern times .

The mechanism was discovered in 1900 by sponge divers along with other finds in a shipwreck off the Greek island of Antikythera , between the Peloponnese and Crete . Coins from Pergamon on board could be traced back to the years between 86 and 67 BC. Dated to the years between 70 and 62 BC , coins from Ephesus . Therefore, the ship is believed to be between 70 and 60 BC. To have sunk; the find thus comes from the late Hellenism .

The device is incomplete and therefore no longer functional. The 82 preserved fragments are now in the National Archaeological Museum in Athens ; the three largest parts are on public display in the bronze objects department.

The discovery of the Antikythera mechanism was surprising in that such a technically sophisticated device like this and the technology and manufacturing method it contained was not previously known from ancient times.

The extensive, partly still ongoing reconstruction of the mechanism showed that it served as a model for the movements of the sun and moon observable from the earth with the help of displays on round scales. The displays, which are mostly scaled as a calendar , were changed synchronously with an adjustment aid. There were three large and three small displays, four of which were the most important:

Larger periods of time in the lunar and eclipses calendar were displayed on two other small scales.

Fragments B, A, and C (from left) exhibited in the National Archaeological Museum (Athens )
The front of fragment A with a four-spoke main drive wheel
The back of fragment A.

Find and initial situation for the reconstruction

Athens - National Archeological Museum - Paris (or Perseus) statue - 20060930.jpg
Bronze statue
of Paris
(or Perseus ?)
Philosopher bust bronze.jpg
Bronze head of a philosopher

The shipwreck lay at a depth of about 42 meters. Most of the extensive cargo had been lifted by autumn 1901 and taken to the National Archaeological Museum in Athens.

The archaeologists first focused on the obvious treasures among the finds,

"An extraordinarily beautiful statue of Paris (or Perseus ) made of bronze, a bronze head of a philosopher, three Ephebe (youths), a Kore (virgin), two statues of Aphrodite , two statues and a head of Hermes , two statues of Heracles , four of Apollo , one of Zeus , one of Philoctetes , two of Odysseus , one of Achilles and the four horses of a quadriga and other fragments "

all of which can be seen in the National Archaeological Museum in Athens. Original bronze statues from ancient times are extremely rare.

In 1902, the museum director and archaeologist Valerios Stais became aware of the inconspicuous, later so-called Antikythera mechanism, and recognized its importance.

The mechanism had been recovered as a lump of corroded pieces of metal. Many external parts that are likely to be present may have been lost during the salvage process or may have fallen off during the approximately 2000 years that the mechanism was in the sea. Some gears have not survived as whole parts. None of the scales are complete and there is only a single pointer in the fragments. When the intensive X-ray examination began 75 years after the recovery, the mechanism had already broken into many pieces. The reconstruction had to be based on material remains, on weakly recognizable traces of components (shaft remnants, pointer remnants) and form elements (fastening points for shafts, spacer bolts for supporting a bearing plate, hemispherical cavity for a display ball) and on indications in inscriptions only preserved in remnants.

Comparison of Antikythera's Mechanism to Other Devices


The two Greek scientists Ioannis N. Svoronos (archeology) and Perikles Rediadis (geodesy and hydrography) were the first to report on the mechanism in 1903, where they came to the conclusion that it must be a kind of astrolabe . From that moment on, the Antikythera Mechanism was long associated with the term astrolabe, before Price called it a computer.

The mechanism has in common with an astrolabe

  • that with both the movement of the sun can be modeled,
  • that it is adjusted or moved by hand.

The differences are:

  • The movement of the moon is usually not reproduced with an astrolabe.
  • The Antikythera mechanism does not reproduce the movement of the fixed stars.
  • An astrolabe usually does not contain any gears.

Analog calculator or computer

Price examined the mechanism extensively for the first time and called it a computer in his report.

Some authors referring to Price try to point out that it is more precisely an analog calculator or computer. It is taken into account here that the continuous change, for example the continuous shifting on the slide rule , is characteristic for analog computing and that the present mechanism was driven with the help of a hand crank.

The analogous way of working is applicable. However, unlike in arithmetic (regardless of whether by hand or with a calculating machine (including a computer)), two or more independent variables are offset against each other to form a third or additional variable. In the Antikythera mechanism, only fixed relationships between several periodically changing variables are represented. For a given value of one variable for a point in time, the mechanism automatically displays the value of the other variables for that point in time.

Astronomical art clock

The Antikythera mechanism generally has in common with a watch that it is equipped with cogwheels and hands on round dials. It has many hands and dials, such as an astronomical art clock built around the 16th century AD. It is also linked to the loss of clarity in comparison to the larger astronomical clock built from the 14th century AD, which like the astrolabe has several hands on just one dial and the position of the sun and moon not only relative to one another and in front of the fixed stars, but also relative to the horizon.

Conclusion from the comparisons

The mechanism has little in common with an astrolabe, because its construction without gears can only simulate the movement of the sun.

The comparison with a calculator or computer is, since there is no calculation, inappropriate (but often found):

The Antikythera mechanism, like a later simple astronomical clock or an astronomical art clock, was principally used to abstractly indicate relative positions between the sun and moon. In contrast to the astronomical clocks, it had no drive. The positions were set by hand from the outside: as with the astrolabe. A special feature is that it had two expanding, but at the same time also limiting, spiral scales with sliders, so it could not be turned continuously.

Historical classification

The Antikythera mechanism dates from a time when the existence of a complex gear mechanism was not only unknown, but from a time that was long believed to have preceded the start of technical development. In modern times, Greek science, which was only interpreted as a purely philosophical activity, had indeed expanded to include the successfully operated branches of mathematics and physics - here above all astronomy - but had no "practical use" as a result, as one thought. The knowledge of the Greeks would therefore only have entered apparatus and processes after their rediscovery ( renaissance ) about 1,500 years later and would only now have established the culture of technology . However, we now know that the Hellenistic Age , at the end of which the Antikythera Mechanism emerged, was marked by considerable technical creativity, although inventors like Archimedes or Heron of Alexandria never actually constructed many of their inventions. A more recent opinion is therefore, "... that the technology of the 18th century was rooted in the Hellenistic works".

Some well-known technical objects belong to the period of Hellenism. Essentially, these are descriptions of devices and automatons of Ktesibios as well as the screw pump and the war machines of Archimedes . It is likely that these things, although not in large numbers, were made in several copies. The detailed instructions for use affixed to the Antikythera mechanism are an indication that it was not a single piece for a single person and existed in several copies. Objects from Ktesibios were playful to use. In a certain sense, the Antikythera mechanism was also a play device that showed its user correct relationships with "rotating" cogwheels, but did not explain them to him (parallel in modern times: the art clocks). An astronomer could quickly determine pairs of values ​​with the help of his knowledge or prepare calendar tables that were comparable once and for all, containing the same number of columns as the Antikythera mechanism had displays. He did not have to rely on such a complex and therefore expensive machine.

It is not the use that makes the Antikythera mechanism so unique, but its existence as a device built from cogwheels that could model with high accuracy the relative movements between the sun, moon and earth, which were already well known at the time. Cogwheels, at least a collection of so many and small cogs, did not seem to exist in the Hellenistic age. It is surprising how many astronomical discoveries were common knowledge to the extent that a craftsman could incorporate them into a product like the Antikythera mechanism and a user of this product could playfully call it up. In particular, the technical treatment of the lunar anomalies by means of a planetary gear and the use of an eclipse calendar based on the Saros cycle are surprising . Gears for modeling the geocentrically observable movements of the planets are missing in the mass of the find. The fact that the Antikythera mechanism was also a planetarium cannot be directly recognized, but cannot be ruled out because of the information in the inscriptions on the mechanism. The necessary gears ( epicyclic gears and crank slides ) were already state of the art at the time, as the discovery made at Antikythera has proven.

At the beginning of the 1st century BC The geocentric worldview represented by Aristotle was generally accepted; Eratosthenes had calculated the circumference of the earth with a deviation of less than ten percent from the actual value. The intricate movements of the planets in the sky were explained with the help of the epicyclic theory , which Apollonius is said to have formulated about a century earlier. In addition to epicycles, Hipparchus used the eccentricity of the circle of deferents . The slider crank mechanism found in the Antikythera mechanism realizes the equant , a further aid for the construction of orbits of the celestial bodies , which was previously attributed to Ptolemy's later living .

At first ( Price and others) it was believed that the Antikythera Mechanism originated from the island of Rhodes . This speaks for:

  • The ship sunk near Antikythera made a stopover in Rhodes.
  • A few decades earlier, Hipparchus worked there, whose knowledge is essentially contained in the Antikythera Mechanism.
  • The parapegma on the Antikythera Mechanism is similar to that written by Geminos . Geminos lived until 70 BC. In Rhodes.
  • At the presumed time of the voyage, the respected philosopher and polymath Poseidonios also worked there , in which the young Cicero saw an instrument “whose individual rotations cause the same thing in the sun, moon and the five planets that happens in the real sky in the individual days and nights . "

Today, however, because of the Corinthian characters and terms used, it is mostly assumed ( Freeth and others) that the design for the Antikythera mechanism came from the environment of Archimedes who worked in Syracuse . There is also a supportive message for this from Cicero: At that time there was still an instrument built by Archimedes in Rome, which Cicero called a sphere that showed the movements of the sun, moon and planets around the earth. This instrument was around 150 years older than the Antikythera Mechanism and would point to a tradition that had lasted for more than a century.

No more is known than this historical mention of the two instruments associated with the Antikythera Mechanism. The mention, however, contradicts the earlier opinion of most historians that the technology for the construction of mechanisms like that of Antikythera was completely lacking in antiquity.


Only a few outer gears could be seen on the mechanism, which had corroded into a lump and soon broke into several parts after recovery. Nevertheless, it was assumed as early as 1903 that it must be a kind of astrolabe. This assumption was supported by the discovery of the Greek word for graduated scale in the inscriptions, which was also used for the zodiacal scale on an astrolabe. The inscriptions on the mechanism date from the 2nd century BC. BC and AD and interpreted as operating instructions.

In 1905 and 1906 the German philologist Albert Rehm examined the now cleaned find in Athens. He discovered the name of the month Pachon on the obverse and rejected the view that it could have been an astrolabe. In the 1930s, the Greek admiral Ioannes Theophanidis found a piece of graduated ring scale on the front. He shared the opinion of Rehm that the mechanism indicated the relative positions of the sun, moon and planets, but could not break away from the idea of ​​an astrolabe.

More intensive research began after World War II . In the first post-war years, they focused on the age and origin of the shipwreck and the salvaged cargo. Jacques-Yves Cousteau contributed to this in 1953 with the lifting of wreckage and some objects found afterwards. The wood for the ship was therefore already in the 3rd or 2nd century BC. As the radiocarbon dating showed. The cargo came from Asia Minor or the islands in the east of the Aegean Sea, from where the ship came from in the early 1st century BC. Must have left. The mechanism was initially unexamined. The wooden box surrounding it (or its wooden case) had already been lost due to rotting in the first few years after the salvage, so that the new radiocarbon dating could not be applied to it.

Cousteau re-examined the wreck at Antikythera in 1976. Coins were found that came from Pergamon and Ephesus. Their minting dates supported the previous knowledge that the ship came from Asia Minor, and limited the period of its sinking to 70 to 60 BC. A.

Investigations by Derek de Solla Price

Derek de Solla Price 1982
with a model of his reconstruction of the Antikythera mechanism,
rear displays simulated with small white discs

The science historian Derek de Solla Price saw the fragments of the mechanism for the first time in 1958, when he discovered further fragments of inscriptions on the obverse. From this he reconstructed a double-scaled ring scale: on the outside as a date scale, on the inside as a zodiac scale. The list below with the rising and setting times of selected stars (part of a parapegma ) that change over the course of the year is most similar to that of the ancient astronomer Geminos, which led Price to assume the island of Rhodes (birthplace of Geminus) as the origin of the mechanism. He realized that on the back of the mechanism there must have been at least two other larger displays over round dials and concluded that it must have been about "time in the most fundamental sense, measured by the movements of celestial bodies across the sky". He considered the mechanism to be so important that “a complete rethinking of the history of technology” was necessary.

Price succeeded in having the mechanism examined in 1972 at its storage location, the National Archaeological Museum in Athens, using X-rays and gamma rays in order to be able to identify and assess the internal gears. Although this recording method could not distinguish which of the multiple superimposed gears is arranged in front or in the back, Price succeeded in reconstructing the displays for the first time.

He recognized that a pointer moved from the outside by means of an adjusting element had to be set to the date of the year (sun pointer) above the front ring scale. Over three of the gear stages found, a pointer that was moved synchronously with the first pointer and arranged coaxially to the first pointer simulated the course of the sidereal months of the moon (about 27 ⅓ days) (sidereal pointer, about 13.37 times faster than the sun pointer) and on the zodiac division the circle scale shows the position of the moon in the zodiac (just like the sun hand in addition to its date display).

He was only able to partially and incorrectly reconstruct the displays on the back of the mechanism. He wrongly assumed the existence of a summing gear ( epicyclic gear in three-shaft operation ) in the drive to a pointer for the synodic lunar month (1 rotation / about 29 ½ days) . He placed this display below where, according to later investigations, the display of eclipse dates was. In order for its synodic moon display to work, it had to partly deviate from the number of teeth that had been found by means of the radiographs.

Research by Michael Wright and Allan Bromley

Allan Bromley was the first to have a model of the Antikythera Mechanism built.
Back: scales still with concentric rings instead of spirals, no drive from the side
Michael Wright: planet gears and crank loops mounted on the main drive
wheel to simulate the planetary movements

The physicist, historian and curator at the Science Museum in London Michael Wright was an early critic of Price's reconstruction, although his manipulation of the number of teeth first irritated him. He considered a summing gear to be too complicated a solution for the task at hand. He considered the fact that seven gears were used on the back just to drive a four-year hand was a “ridiculously simple idea”.

Wright's intense preoccupation with the mechanism did not begin until after Price's death (1983). In his opinion, the large, so-called main drive wheel, was predestined to carry several rotating gears in order to simulate the movements of the planets. Corresponding wheels could have been placed in the relatively large empty space between this wheel and the front dial, and there were references to planets in the inscriptions on the mechanism. Corresponding gears were not found.

Another critic of Price's results was the Australian computer historian Allan Bromley, who had made a model himself (see illustration) and who often visited Wright to discuss things with him. Wright was doing everyday work at the London Science Museum. He could only deal with the Antikythera Mechanism in his spare time and on vacation. Bromley obtained permission in 1989 to examine the fragments in the Athens Museum. Wright took unpaid leave and went with him. They looked at the fragments (including two newly emerged ones) carefully without making any significant new discoveries, except that Price did not properly position all of the fragments against one another. Renewed x-rays turned out to be worse than those made for Price because of errors in the museum's development laboratory.

Back in London, Wright became aware of tomography . He even managed to build a suitable recording device at home with a commercially available X-ray source. In the following four winters - again and again as a companion of Bromley - he took about 700 pictures in different depths of the mechanism with which it could be reconstructed spatially. As a senior university professor (Wright was a curator at the time), Bromley took the recordings with him to Sydney , but wanted to send Wright copies after scanning. The recordings were neither scanned nor evaluated by Bromley. As a result, Wright initially concentrated on the construction of the supposed additional planetary functions of the mechanism (see figure opposite) and ultimately spent at least as long on it as on the reconstruction of what the illuminated fragments directly produced. He could only begin with the latter when the recordings were given to him after Bromley's serious illness in 2000 (part) and after his death in 2003 (the remainder that can still be found).

With the help of his layer recordings, Wright recognized further gears (31 instead of the previous 27) and some essential connections that Price had not yet recognized:

  • There was a spiral scale at the top of the back. He added three fictitious cogwheels with which a lunar hand indicated 235 synodic lunar months in 19 solar years on the five spiral orbits and another pointer simultaneously indicated a quarter turn in a quarter of 76 solar years ( Calippian period ) on a smaller scale.
  • There was an additional idler gear in the gear train that used to move the front moon hand that Price hadn't seen. It was used to reverse the direction, so that the large additional wheel, which Price assumed under the front panel, was also moved in the opposite direction by the crown wheel of the side drive, was not necessary.
  • A newly discovered small cogwheel protruding from the front and a hemispherical cavity led to the conclusion that a two-colored small ball rotated on the front moon hand to indicate the phases of the moon .

Wright saw some details that made Price's unsuccessful reconstruction of the lower rear display now satisfactorily possible. He had not yet worked them out when a group of impatient prospects began to speed up the reconstruction and the answering of open questions through their own research. In a hastily given, ultimately unsatisfactory answer, he said that the rear lower hand rotated once per draconian month (about 27 1/5 days). Between the drive from the sidereal to the draconian moon pointer, a gear ratio of around 1.004 is necessary, for which he considered the use of the existing planetary gear - now with only one input ( two-shaft operation ) - still to be excessive, but justified. He neglected to accept a lost wheel on a stub shaft he had seen and to allow it to mesh with the existing large wheel with 223 teeth. He could have found the much more conclusive display of 223 synodic months in 18.03 solar years (Saros period) on the spiral scale with four revolutions. Instead, he assumed that an old used gear with coincidentally 223 teeth had been reused as the spider gear of the epicyclic gear.

Wright had already seen the connection (revolving crank loop ) in the revolving wheel pair in the epicyclic gear train, which was not rigid but was made with a pin and slot . Since he did not recognize any purpose at that point (he used several such pairings in his reconstruction relating to the planets in the outer part of the mechanism, see figure above), he came to the conclusion that they were also recycled, in the mechanism on the same axis firmly connected wheels acted.

A short time later (2005/2006) the research group mentioned above came up with the more satisfactory and more likely form of the rear lower display and the purpose of this revolving crank loop. Wright subsequently adopted both partial answers in the model of the mechanism he built.

Research in the Antikythera Mechanism Research Project

An international group of researchers and helpers brought together by two Englishmen since around 2002 - the astronomer Mike Edmunds and the mathematician and documentary filmmaker Tony Freeth - later organized in the so-called Antikythera Mechanism Research Project. Organization, evaluation and publication of the work of this group is mainly in the hands of Tony Freeth.

After his experiences with Bromley, Wright understandably did not want to give up his layer photographs, which is why the group tried to be able to examine the fragments in the Athens Museum again themselves. After more than ten years, the newer computed tomography promised better X-ray images than those produced by Wright. And the group believed that deciphering additional characters could bring advances in the reconstruction and interpretation of the mechanism. A new process developed by Hewlett-Packard for three-dimensional and detailed surface images seemed suitable for this. In autumn 2005 the group was allowed to work in the museum with their modern heavy equipment for about a month.

Although the images were sharper than any previously made, they essentially confirmed the number of teeth and wheel sequences that Wright was already familiar with (which Price also knew for the most part, but did not always follow). Decisive advances resulted from the interpretation of the characters. A fragment that had been forgotten in the museum and had not yet been examined had only recently emerged and additionally helped to explain the lower rear display to which it belongs. The four-spiral scale that was supplemented with this now showed detailed, labeled subdivisions and indicated the display of 223 synodic months in 18.03 solar years. This period is the Saros period. All of the solar and lunar eclipses contained therein and clearly marked on the scale repeat themselves regularly after this time. The correct rotation of the corresponding eclipse hand resulted from turning the large wheel with 223 teeth by means of the conceivable wheel with 27 teeth that Freeth inserted.

Freeth recognized the purpose of the circulating crank loop in the planetary gear. The drive to the front moon hand led via a detour first to this crank loop and only then to the front. This superimposed the small fluctuation caused by the non-circular lunar orbit on the rotation of the pointer. Since the wheel with 223 teeth revolved around it once in about 9 solar years, the slow rotation of the elliptical lunar orbit in space was also simulated.

Wright had assumed that a small ball, which could move radially on the hands, would engage the spiral grooves of the rear displays to mark the way in which the display was currently being displayed. In the new sharper X-ray images, a slide with a nose engaging in the groove was discovered instead on the upper display. Wright's display of the fourfold Meton period with the help of a smaller pointer at the top at the back was moved to a different position because the position he had provided turned out to be an indication of the four-year period Olympiad.

The investigations of the group from the Antikythera Mechanism Research Project brought to light many other text fragments, in addition to new labels on the dials, which seem to come from an extensive operating manual. The group believes that the Antikythera Mechanism was not an instrument for astronomers, but a luxury item for a wealthy, non-astronomical client, and possibly not a one-off. The labels found have not yet been finally evaluated. They further supported Wright's initially vague assumption for a planetarium part.

In 2008, the reverse of the construction was also examined and interpreted more closely by the Antikythera Mechanism Research Project. On the large upper display at the back, the names of the months could be identified and it could be determined that these names are of Corinthian origin. Like Price, the group had previously been of the opinion that the mechanism had originated on Rhodes because the astronomer Hipparchus and the polymath Poseidonius worked there. It is now believed that the mechanism came from Corinth or a Corinthian city foundation such as Syracuse , Archimedes' hometown.

In 2012, Freeth reported on the inclusion of the movements of the planets in his virtual computer model. There was also a second sun pointer, which imitates the not quite constant speed of movement of the sun through the zodiac. The previous sun hand was retained as a calendar hand. Apart from other interpreted written references from the inscriptions, it is based on several spacer bolts recognized on the main drive wheel. In his considerations, these together with the main drive wheel and an assumed plate parallel to this form a rotating cage in which the numerous lost planetary gears were stored. Thus, the Antikythera Mechanism Research Project now also assumes that these additional displays exist. However, since no other gear wheel required for this was found apart from the main drive wheel (with its cage marks), the technical solution could not be reconstructed. The otherwise reconstructed virtual model was extended for the planetary part with the results of a current designer.



82 individual parts of the Antikythera mechanism have been preserved, seven large fragments (fragments A – G) and 75 smaller fragments (fragments 1–75). How many parts it originally consisted of is unknown, as the mechanism has not been fully preserved. The largest fragment obtained (fragment A) has a size of 18 cm × 15 cm. The entire mechanism should have been about 31.5 cm × 19 cm × 10 cm in size.


The mechanism's gears, hands, indicators, and suspected cover plates are made of bronze in an alloy of 95% copper and 5% tin. All parts had been cut out of a 1 to 2 mm thick bronze sheet.

The entire apparatus was originally packed in a wooden frame or in a wooden box. At the time of the recovery, there were still wood residues, but over time they crumbled and got lost due to drying out in the air. Radiocarbon dating of the mechanism is therefore no longer possible.


The front and back of the mechanism were probably protected with a metal cover each when not in use. These covers are usually referred to as doors in literature, but it has by no means been proven that they were hinged covers like a book, as no hinges have been preserved. More obvious, as they are more practical to use, are removable lids. They have been left out of all replicas.

Some inscriptions on the lids were found and identified as imprints on the remains of the dials at the front and back.


Seven pointers were required for the six indicators of the mechanism that were relatively reliably recognized. Only:

  • a relatively complete pointer for the lunar calendar (back of fragment C; 55.0 mm long, 4.2 mm wide and 2.2 mm thick),
  • the hub of the pointer for the smaller eclipse display.
Wheels scheme of the mechanism, front view, as of 08/2012, wheels for displaying the moon phases have been omitted, wheels marked in red are hypothetical
axis 1 2 3 4th 5 6th comment
a 48 drive
b 224 Sun
b 64 32 moon
c 38 48
d 24 127
e 32 32 223 188 50 50
f 53 30th
G 54 20th Saros
H 60 15th
i 60 Exeligmos
k 50 50
l 38 53
m 96 15th 27
n 53 15th 57 Meton
O 60 Olympics
p 60 12
q 60 Callippus

It doesn't matter that most of the pointers have been lost. It can be assumed without further ado that the scales and pointer axes found also included a pointer.

For the only assumed additional displays of the planetary movements (Mercury, Venus, Mars, Jupiter and Saturn), five more pointers would have been necessary on the front of the mechanism (there previously two: for the sun and moon).

Gear train


There are 30 gears preserved, of which 27 are in fragment A and 1 each in fragments B, C and D. Fragment D also contains 1 further gear, which probably belonged to a planet display that has not been preserved and for which further missing gear would have been necessary. For the seven reconstructed displays, 8 not found gears were necessary (gears including the number of teeth are hypothetical).

Except for two crown gears , all gears were spur gears (with teeth perpendicular to the axis of rotation of the wheel). A crown wheel meshes with the large main gear, and its axis points out to the side. It is believed that it was associated with a side-mounted adjuster that could drive the entire mechanism. A second crown wheel was on the front moon hand. Almost all gears were closed disks. Only the main drive wheel had four wide spokes, one of which had already been repaired.

The teeth of all spur gears have the shape of an isosceles triangle (60 degree angle at the tip and the tooth base) and are of the same height (about 1.5 mm) so that each gear could mesh with every other gear. It is a primitive tooth system, which on a small scale does not have a completely uniform translation from tooth to tooth or does not meet the first law of toothing .


The gears were combined into several gears , with which the input rotation made by hand was transmitted in suitable gear ratios to the pointers on the front and back.

Adjustment element

By means of a side-mounted rotary knob or a crank, all gears and thus all displays of the mechanism could be set in motion via an angular gear.


The mechanism was repaired at least once when one of the four spokes on the main drive wheel was replaced. Furthermore, a tooth of a gear is said to have been replaced. This shows that the device was used often.

Wheels scheme of the mechanism, sectional drawing, status 08/2012, gears marked in italics are hypothetical


Astronomical periods

The purpose of the mechanism was to simulate the simultaneous periodic celestial movements of the sun and moon and to show the mutual relationships of their period durations .

Basic periods are:

  • the sidereal period of the sun (about 365 days, the solar year),
  • the sidereal period of the moon (about 27 days for a 360 ° rotation of the moon on its orbit),
  • the synodic period of the moon (about 29 ½ days between the same moon phase).

The mechanism also contained calendar displays, because the solar year and the lunar month are calendar units.

Indirect periods were also to be shown, which are characterized by certain simple integer relationships between basic periods:

  • the Meton period, which is both 19 solar years and 235 lunar months,
  • the Saros period, which is both 223 lunar months and 242 Draconite lunar periods (about 27 1/5 days) long (about 18 solar years).

The Meton period is the basic relationship for binding a lunar calendar to a solar calendar. The mechanism made this bond clear.

The Saros period is the most important eclipse period , after which all eclipses occurring within this period are repeated for about 1000 years. There was a pointer on the mechanism that indicated the lunar month in which an eclipse took place.

The presumable addition to the representation of the planetary movements supports the conclusion that the Antikythera Mechanism was primarily a model for the movement of the heavenly bodies. Its function as an adjustable mechanical calendar was secondary and limited. It did not contain any jumps in movement that are required in calendars for intercalations (leap days and months).

Front displays

The large display on the front of the mechanism was made up of two ring-shaped scales, an inner zodiac and an outer date scale (solar calendar).

  • The inner ring scale was divided into 12 sections for the 12  signs of the zodiac .
  • The outer ring scale was divided into 365 sections for the 365 days of the year according to the Egyptian calendar : 12 months of 30 days each and 5 additional days  . It could probably be rotated by a day unit in order to be able to manipulate it briefly every four years because of the inserted leap day .

Two coaxial, mutually synchronized pointers could be set using the scales:

  • Sidereal sun pointer:
    On the outer scale it approximately indicated the date in the solar calendar year and on the inner scale the position of the sun in the zodiac, which changes over the year.
    The sun pointer was fastened together with the main wheel b1 on the same shaft  b (hollow shaft); a special gear stage was omitted.
  • Sidereal moon hand:
    One full turn of this hand corresponds to the period of the sidereal month (approx. 27 1/3 days). Its position above the inner scale indicated the position of the moon in the zodiac. 19 solar years contain 254 sidereal months. The moon hand turned accordingly more often:
    b2 / c1 × c2 / d1 × d2 / e2 = 64/38 × 48/24 × 127/32 = 13.37 = 254/19 .
    From wheel e2, the movement is guided via the detour e5 - k1 - k2 - e6 - e1 to wheel b3 and thus to the moon hand. The gear ratio is 1 in each of these stages . But within one revolution there is a periodic fluctuation of this ratio, which simulates the uneven course of the moon on its elliptical orbit. The unevenness is produced by a slight mutual axial offset of the two wheels k. k1 drives k2 via a pin which can slide in a slot in k2 (rotating crank loop). The wheels k are mounted on the large wheel e3, so they revolve with it, which also simulates the rotation of the lunar orbit in space ( apsis rotation: once every 8.9 years). Apsid
    rotation: b2 / l1 × l2 / m1 × m3 / e3 = 64/38 × 53/96 × 27/223 = 0.1126 = 8.88 . (hypothetical: m3 )
  • Moon phase ball:
    ball placed on the moon hand is driven by an angular gear with a gear ratio of 1 due to the difference in speed between the moon and sun hands. The gear wheel parallel to the dial rotates with the sun hand. Thus the ball rotates once completely between two meetings of the moon with the sun (about every 29 ½ days, synodic month). When they meet, there is a new moon. Between two encounters, the ball shows the different light shapes of the moon.
  • Planet
    : For the display of the positions of the planets (Mercury, Venus, Mars, Jupiter and Saturn) on the zodiac scale, there would have been five additional coaxial hands in addition to those for the sun and moon. The planetary movements, like those of the moon, are not used in a calendar in a sidereal month. The display of the sidereal month and the non-reconstructable displays of the planetary movements are an indication that the front displays on the Antikythera mechanism were primarily a mechanical model for the movements of all movable celestial bodies and only secondarily an adjustable mechanical calendar.

Only a small fragment in the form of fragment C has survived from the large front display. There the inner ring shows the constellations Virgo (Virgo) and Libra (Libra) , which the Greeks called Parthenos (Παρθένος, Virgo) and Chelai Skorpiou (Χηλαί Σκορπίου, claws of the scorpion). The outer ring shows the Egyptian months of Pachon ('March') and Payni ('April'). You can also see a pointer.

Rear upper displays

Back upper display (Meton period)

235 synodic months (each about 29 ½ days) are contained in 19 solar years or in a so-called Meton year (great year). Synchronized with the sun hand, another hand could be used to display the lunar date corresponding to an annual date on a 235 synodic month long lunar calendar scale. The display was rough because the scale was only scaled in months. The latter was attached in a spiral (five whole whorls) on the back of the mechanism.

  • synodic moon pointer:
    The corresponding pointer could turn five times more often than on a scale with only one contact.
    b2 / l1 × l2 / m1 × m2 / n1 = 64/38 × 53/96 × 15/53 = 0.2632 = 5/19 . (hypothetical: n1 )
    A slide mounted on the pointer engaged in a spiral groove. This marked the handling on which the ad was currently being made.

This display was a bound lunar calendar that counts the months from 1 to 12 and bundles them into calendar years in packages of 12 or 13 months (leap as the 13th month). The 235 months were noted on the spiral scale with names derived from a Corinthian calendar.

Small left display at top rear (Callippian period)

The finding of the number 76 in the inscriptions led to the assumption that it meant the Callippian Period , which was 76 solar years long , and that an indication existed for it. Freeth's group put this ad to the left of the Olympics ad and used the gear q1 they found for it.

  • Callippus pointer:
    The corresponding pointer turned
    1/4 times when the Meton pointer turned five times (5 revolutions for 19 years).
    n2 / p1 x p2 / q1 = 15/260 × 12/60 = 0.05 = 1/20 = 1/5 x 1/4 . (hypothetical: n2, p1 and p2 )

The Meton period had been equated by the astronomers at Meton's time with 6,940 days except for 19 solar years. At the time of Callippus , it was believed that this number was 1/4 day too long. In whole days this meant that 76 solar years (four Meton periods) had to be shortened by one day. The Callippian Period was equated with 76 solar years and 27,759 days ((4 × 6940) - 1). With the corresponding application of the Antikythera mechanism, the day display of the front sun hand, which had rotated 76 times, was currently error-free with regard to a year length assumed to be 365 1/4 days.

Small right display at the top back (four-year period / Olympics)

The Olympiad calendar was important for the Greek calendar, which took place in Olympiads , i.e. four-year periods. So the dates of historical events were given as in year 1, 2, 3 or 4 of a certain Olympiad.

  • Olympiad hand:
    The corresponding hand rotated once when the front sun hand rotated four times.
    b2 / l1 x l2 / m1 x m2 / n1 x n3 / o1 = 64/38 × 53/96 × 15/53 × 57/60 = 0.25 = 1/4 . (hypothetical: n1 , n3 , o1 )

On this quartered circle Panhellenic competition games were noted at six competition venues. Five of the entries could be identified. These are the Olympic Games in Olympia , the Pythian Games in Delphi , the Isthmic Games in Corinth , the Nemean Games in Nemea and the Naaic Games in Dodona . Of these, the Isthmian and Nemean Games took place in year 2 and 4, the rest in year 4. According to tradition, the first Olympic Games took place in 776 BC. Chr.

Rear lower indicators

Back lower display (Saros period)

223 synodic months (29.53 days each) form the so-called Saros period (18.03 solar years). Since all eclipses repeat themselves with a similar course after this time, an indication of eclipses taking place does not need to extend over a longer period of time. Within this period, the timing of the eclipses that occur is constant over a long period of time. Usually six (less often five) months pass between two solar or two lunar eclipses. The rear lower display was also spiral, it had four revolutions. All solar and lunar eclipses were plotted on it according to the scheme valid at that time.

  • synodic eclipse
    (Saros): The corresponding pointer could turn four times more often than on a scale with only one contact.
    b2 / L1 × L2 / m1 ×  m3 / e3 x e4 / f1 × f2 / g1 = 64/38 53/96 × ×  27 /223 × 188/53 × = 0.2219 = 30/54 4 / 18.03 ( hypothetical: m3 ).
    A slide mounted on the pointer also engaged in the spiral groove. This marked the handling on which the ad was currently being made.

The Antikythera mechanism thus also contained a table with times for solar and lunar eclipses.

The months with solar and / or lunar eclipses were provided with abbreviations (glyphs). A total of 18 monthly labels of 51 (38 for lunar eclipses and 27 for solar eclipses) have been preserved. The labels consist of the following information:

  • Σ  (ΣΕΛΗΝΗ, Selene, Greek for 'moon') for lunar eclipses,
  • Η  (ΗΛΙΟΣ, Helios, Greek for 'sun') for solar eclipses,
  • Η \ Μ  (ΗΜΕΡΑΣ, Hemeras, Greek for 'day') for taking place during the day,
  • Ν \ Υ  (ΝΥΚΤΟΣ, Nyktos, Greek for 'night') for taking place during the night and
  • ω \ ρ  (ωρα, ora, Gr. for 'hour') for the respective hour of the day or night.

This means that for an eclipse that occurs during the day with a full moon or with a new moon, the night or day hour was also specified.

Small display in the lower back (Exeligmos period)

The small display inside the large display was used to expand the eclipse calendar to three times the value of the Saros period. The latter is around 6,585 1/3 days long at 18.03 years. This means that after three Saros periods (19,756 days) an eclipse occurs again at about the same time of day. This longer duration is the Exeligmos period at around 54 years. For this purpose, the display was divided into three sectors, which indicated whether the eclipse took place either in the 0th, 8th or 16th hour.

  • synodic eclipse
    (Exeligmos): The corresponding pointer turned 1/3 times if the eclipse pointer turned four times (4 revolutions for 18.03 years).
    g2 / h1 x h2 / i1 = 20/60 × 15/60 = 0.083 = 1/12 = 1/4 x 1/3 .

From this ad the two numbers 8 and 16 were found.

Advertisement overview

location display function
front central Big circular display Solar year
  with outer ring scale 12 Egyptian months + 5 leap days
  with inner ring scale 12 Babylonian Zodiac Signs
  with sun pointer Annual calendar date and position indicator of the sun in the zodiac
  with moon hand Position indicator of the moon in the zodiac
  with moon phase ball Moon phase display (ball on moon hand)
  ?? several planet hands ?? ?? Position display of the planets in the zodiac ??
back top central Large spiral display Meton period (19 solar years)
  with 5-part spiral scale 235 synodic months, summarized in 19 lunisolar calendar years
  with pointer Display of the synodic month 1, 2, ... or 235
and the lunisolar calendar year 1, 2, ... or 19 in the Meton period
back top central left Small circular display Callippus Period (76 solar years)
  with pointer Display of the 19-year sub-period 1, 2, 3 or 4 in the Callippus period
back top center right Small circular display Olympic period (4 calendar years)
  with pointer Display of year 1, 2, 3 or 4 in the Olympics period
rear bottom central Large spiral display Saros period (≈ 18 solar years)
  with 4-part spiral scale 223 synodic months, additionally scaled with solar or lunar eclipses
  with pointer Display of the synodic month with a solar or lunar eclipse
rear lower central right Small circular display Exeligmos period (≈ 54 solar years)
  with pointer Display of the 18-year sub-period 1, 2, or 3 in the Saros period

Inscriptions and scale markings

The mechanism had four large areas of inscription, which were located on the inside of the front and rear cover facing the apparatus and on the open areas above and below the displays on the front and rear. In addition, all scales were labeled, with their scale values ​​predominantly being given in words - not just numbers or symbols.

Of the estimated 15,000 letters, around 3,000 letters have been preserved (relation between the area of ​​the text received and the presumed area of ​​the originally existing text), whereby it must be taken into account that the Greek letters were used for numbers in the Greek numerals .

The font height ranges from 2.7 mm in the parapegm on the front to 1.2 mm in the lettering of the spiral displays on the back. The shape of the letters is typical of stone inscriptions. That means: there are no gaps between the individual words, but there are gaps in front of and behind the letters that served as digits. There are also gaps in front of new paragraphs.

The state of preservation of the individual texts is extremely poor. Some of the inscriptions on the lost lids have only survived because of their impressions in the encrustations. Only 11 lines of the text on the lid are reasonably legible.

The texts are in Koine , the Greek commonly used at the time. Exceptions are the 12 month names of the lunar calendar in the upper display on the back of the mechanism, which come from the Corinthian dialect and have Doric features. It can therefore be assumed that the designer of this mechanism spoke Corinthian Greek, i.e. in Corinth or in a Corinthian colony such as Syracuse, at home and, for example, Archimedes (* 287 BC; † 212 BC) or a student was from him.

Content of the labels

The areas on the inside of the two lids contained instructions for use for the mechanism in a relatively small font. Examples are:

  • On fragment E there is the only 2 mm high Greek text: “Spiral divided into 235 sections”, which is an indication of the Meton period with 235 lunar months.
  • Fragment 19 contains the Greek words “76 years” and “19 years”, which is an indication of the 19-year Meton period and the 76-year Callipean period on the back of the mechanism.

The areas above and below the large front display contained lists ( parapegmas ) with the morning and evening rising and setting of important stars and constellations .

Part identification letters

Each part and hole had an identification letter that was used to aid assembly.

Use of the mechanism

The Antikythera Mechanism was a movable model for the movements of the sun, moon and probably also the planets, observable from Earth, which were simulated with several hands that moved synchronously with one another.

Ads on the front

On the large front display area, the coaxial pointers rotated - one for each celestial body - whereby the different orbital speeds and directions of movement (temporary return of the planets) were illustrated.

In a snapshot or a specific setting, the current positions of the celestial bodies in the zodiac belonging to the calendar date of the year (indicated by the position of the sun) could be seen on the front double scale. The current year could not be read. If only the mutual positions of the sun and moon were considered or no planetary displays existed, it could be one of 19 possible years.

The date of the year was inaccurate up to about a day, because the orbit of the sun hand was always the same, while the calendar year could be 365 or 366 days long.

Ads on the back

The setting or turning by means of the annual calendar in front was synchronously transferred to two hands via spiral scales on the back of the mechanism. Both scales were divided with the synodic lunar month as a unit.

The spiral lunar calendar was scaled with the 13 month names of a bound lunar calendar used at the time. When the front sun hand was turned 19 times (19 solar calendar years), 235 scale units (235 synodic months) were passed through at the back. The astronomical connection known as the Meton cycle (19 sidereal solar periods are roughly equal to 235 synodic lunar periods) was represented in this way. On the other hand, it was possible to identify which of the 19 bound lunar calendar years with a leap month was to be extended to 13 lunar calendar months. The indication of a Meton period as a quarter of the Callippi period was an addition. The slider engaging in the spiral groove had to be set back to the beginning of the spiral so that three more Meton displays could be transferred to the Kallippian display. In order to be able to use the lunar calendar, it was necessary to know for which historical 19 years it was made, or which of the years 1 to 19 was the current one.

Knowledge of the currently valid four-year period was also required for the use of the attached Olympiad advertisement.

The spiral scale of the back eclipse calendar was 223 synodic lunar units (about 18 sidereal solar periods) long. After this time interval, the eclipses repeat themselves regularly over several centuries within a day. Therefore, a list of the different eclipses that took place in this so-called Saros period was sufficient. On this list, which is underlaid as an eclipse scale, the synodic months were marked in which a solar eclipse (with a new moon) and / or a lunar eclipse (with a full moon) took place. After three Saros periods (around 54 years of Exeligmos), the eclipses even take place at around the same time of day. The corresponding additional display with three Saros periods per revolution was used to indicate whether the time of day indicated on the main dial was valid or whether 1/3 or 2/3 days were to be added. The slider engaging in the spiral groove had to be reset to the beginning of the spiral so that two more Saros advertisements could be transferred to the Exeligmos advertisement.

The eclipse calendar could be used again after about 54 years. Since there is no reference to the 19-year Meton period, it had to be known for which historical 54 years the eclipse calendar was made, or which of the three times 223 synodic lunar periods was the current one, in order to be able to use it for the prediction of eclipses.


Reconstruction after Derek de Solla Price in the National Archaeological Museum (Athens)
side and back view

An early material partial reconstruction was made by Allan Bromley together with the watchmaker Frank Percival from Sydney, before he and Michael Wright were able to X-ray the Antikythera mechanism by tomography from 1990 to 1993. The necessity of a more precise X-ray analysis of the original mechanism resulted from its replica, which did not run smoothly enough despite relocating the drive to the faster large gear inside. But it ran much better than a replica of Derek de Solla Price.

Michael Wright tested his research using real models from the start. He also provided this with a part showing the movements of the planets, but no wheel fragments were found for this. They were always equipped with a side drive crank on the heavy front main drive wheel. His latest model also contains some findings from later researchers (particularly the Englishman Tony Freeth).

Replica of the Antikythera Mechanism by Mogi Vicentini, 2007

Tony Freeth (Antikythera Mechanism Research Project) only “built” his reconstruction virtually.

There are individual replicas that, on the one hand, aim as closely as possible to the original (see figure on the left), on the other hand they are just a mechanism that has to fulfill the same tasks as the original.

Several museums have a model of the Antikythera Mechanism:

The English orrery builder (manufacturer of planetary machines) John Gleave brought a well-functioning series of the Antikythera mechanism on the market. It roughly corresponded to Wright's first reconstruction without the spiral displays on the back.

The Geneva watch manufacturer Hublot presented a variant with a modern design as a homage to the Antikythera mechanism and for advertising purposes in October 2011 at the Belles Montres trade fair in Shanghai.

The Youtuber Clickspring has been working on a replica of the mechanism since the beginning of 2017. He uses modern tools and machines of course, but adds attempts to make and work with some antique tools.


  • Ioannis N. Svoronos : The Athens National Museum. Volume 1. Beck & Barth, Athens 1908, pp. 1–86 ("The Finds of Antikythera") (online)
  • Gladys Davidson Weinberg , Virginia R. Grace, et al. a .: The Antikythera shipwreck reconsidered. (= Transactions of the American Philosophical Society NS 55, 3) American Philosophical Society, Philadelphia 1965 (on the ship and its dating, to be dated to 80-50 BC after the finds).
  • Peter Cornelis Bol : The sculptures of the Antikythera ship find. Gebr. Mann, Berlin 1972, ISBN 3-7861-2191-5 (on the sculpture finds from the ship's find).
  • Derek de Solla Price Gears from the Greeks. The Antikythera mechanism - a calendar computer from approx. 80 BC (= Transactions of the American Philosophical Society NS 64, 7) American Philosophical Society, Philadelphia 1974, ISBN 0-87169-647-9 .
  • Heather Couper, Nigel Henbest: The History of Astronomy . Frederking & Thaler, Munich 2008, ISBN 978-3-89405-707-7 .
  • Jo Marchant: Decoding the Heavens - Solving the mystery of the world's first computer . William Heinemann, London 2008, ISBN 978-0-434-01835-2 .
  • Jo Marchant: Deciphering Heaven: The First Computer - Solving a 2000 Year Old Mystery . Rowohlt, Reinbek 2011, ISBN 978-3-498-04517-3 .
  • Nikolaos Kaltsas , Elena Vlachogianni, Polyxeni Bouyia (eds.): The Antikythera Shipwreck: The ship, the treasures, the mechanism. National Archaeological Museum, April 2012 - April 2013. Hellenic Ministry of Culture and Tourism; National Archaeological Museum, Athens 2012, ISBN 978-960-386-031-0 .
  • Jian-Liang Lin, et al .: Decoding the Mechanisms of Antikythera Astronomical Device. Springer Berlin 2016, ISBN 978-3-662-48445-6 .
  • Alexander Jones: A Portable Cosmos. Revealing the Antikythera Mechanism, Scientific Wonder of the Ancient World. Oxford University Press, Oxford 2017, ISBN 978-0-19-973934-9 .

Article (most of the articles are cited in the notes, see there)


  • Antikythera's miracle machine. Documentary, France, Greece, Great Britain, 2012, 74 min., Director: Mike Beckham, production: Images First, arte , ERT , NHK , summary by arte.
  • Harald Lesch: Unsolved riddle about an ancient machine. ZDF, TerraX series, 2018, documentation with the replica from the Astronomical-Physical Cabinet Kassel.

Web links

Commons : Antikythera mechanism  - collection of images, videos and audio files

Antikythera Mechanism Research Project of the Hellenic Ministry of Culture:

Soundless videos from the Antikythera Mechanism Research Project:

Video in German:

Video in English:


National Archaeological Museum in Athens:


  1. This remarkable solution is rare even today.
  2. The pointer made a complete rotation in one solar year (about 365¼ days). However, this was necessarily assigned a whole number of calendar days (365 days), which is why the display could only be a good approximation.
  3. A certain one of the many simultaneously existing Saros cycles repeats itself for more than a millennium before it disappears.

Individual evidence

  1. Jo Marchant: The Decipherment of Heaven . Rowohlt, 2011, ISBN 978-3-498-04517-3 , pp. 85-88.
  2. ^ The Antikythera Mechanism at the National Archaeological Museum. (No longer available online.) Archived from the original on February 21, 2017 ; accessed on October 26, 2012 (English).
  3. One conclusion is, for example: "... that the technology of the 18th century was rooted in the Hellenistic works ...". Cf. Lucio Russo : The forgotten revolution or the rebirth of ancient knowledge. Springer, 2005, ISBN 3-540-20938-7 , p. 156.
  4. a b c d e f g Tony Freeth: The decryption of an ancient computer in spectrum of science , May 2010, pp. 62–70.
  5. ^ Ioannis N. Svoronos : The Athens National Museum. 1903, p. 14 , accessed October 26, 2012 .
  6. ^ Attilio Mastrocinque: The Antikythera Shipwreck and Sinope's Culture during the Mithridatic Wars. In: Mithridates VI and the Pontic Kingdom. 2009, p. 2 , accessed on October 26, 2012 (English).
  7. Derek de Solla Price : Gears from the Greeks. The Antikythera Mechanism: A Calendar Computer from ca. 80 B. C. Science History Publications, 1975, p. 9.
  8. ^ A b Ioannis N. Svoronos : The Athens National Museum. 1903, p. 50 , accessed October 26, 2012 .
  9. a b Derek de Solla Price: An Ancient Greek Computer . In: Scientific American , 200 (6), 1959.
  10. a b Derek de Solla Price: Gears from the Greeks. The Antikythera Mechanism: A Calendar Computer from ca 80 BC Science History Publications, 1975.
  11. Although gears are equipped with a discrete number of elements (teeth), gear drives move steplessly or continuously. With the tooth shape in the Antikythera Mechanism, the only thing that is not quite even is the small movement from tooth to tooth.
  12. Siegfried Wetzel: Left and right turning sun and moon hands on astronomical clocks / 4. Sun and moon hands on separate dials, Fig. 4. In: Chronométrophilia (N o 55). 2003, pp. 73-76 , accessed October 26, 2012 .
  13. The Antikythera Mechanism Research Project: "The Antikythera Mechanism is now understood to be dedicated to astronomical phenomena and operates as a complex mechanical 'computer' which tracks the cycles of the Solar System. [1]
  14. Seaman, Bill; Rössler, Otto E. (January 1, 2011). Neosentience: The Benevolence Engine. Intellect Books. p. 111. ISBN 978-1-84150-404-9 . "Mike G. Edmunds and colleagues used imaging and high-resolution X-ray tomography to study fragments of the Antikythera Mechanism, a bronze mechanical analog computer thought to calculate astronomical positions" [2]
  15. Swedin, Eric G .; Ferro, David L. (October 24, 2007). Computers: The Life Story of a Technology. JHU Press. p. 1. ISBN 978-0-8018-8774-1 . "It was a mechanical computer for calculating lunar, solar, and stellar calendars." [3]
  16. Paphitis, Nicholas (30 November 2006). "Experts: Fragments an Ancient Computer". Washington Post. "Imagine tossing a top-notch laptop into the sea, leaving scientists from a foreign culture to scratch their heads over its corroded remains centuries later. A Roman shipmaster inadvertently did something just like it 2,000 years ago off southern Greece, experts said late Thursday. " [4]
  17. Lucio Russo : The forgotten revolution or the rebirth of ancient knowledge . Springer, 2005, ISBN 3-540-20938-7 , p. 149.
  18. Lucio Russo: The forgotten revolution or the rebirth of ancient knowledge. Springer, 2005, ISBN 3-540-20938-7 , p. 156.
  19. a b c Jo Marchant: The decryption of the sky. Rowohlt, 2011, ISBN 978-3-498-04517-3 , pp. 227-234.
  20. Norbert Froese: Eudoxos & Co. - The beginnings of scientific astronomy. (PDF; 543 kB) May 19, 2012, p. 24 (Fig. 10) , accessed on October 27, 2012 .
  21. a b In De Natura Deorum , Book II, 88. Cf. Jo Marchant: The decryption of the sky . Rowohlt, 2011, ISBN 978-3-498-04517-3 , pp. 257/258.
  22. a b In De Re Publica , Book I, chap. 21-22. Cf. Jo Marchant: The Decryption of Heaven . Rowohlt, 2011, ISBN 978-3-498-04517-3 , p. 263.
  23. Jo Marchant: The Decipherment of Heaven . Rowohlt, 2011, ISBN 978-3-498-04517-3 , p. 52.
  24. Rehm never published the results, but they became known through a lecture by Georg Karo in Athens in December 1906. Jo Marchant: Deciphering the Sky . Rowohlt, 2011, ISBN 978-3-498-04517-3 , p. 58.
  25. Jo Marchant: The Decipherment of Heaven . Rowohlt, 2011, ISBN 978-3-498-04517-3 , p. 60.
  26. Jo Marchant: The Decipherment of Heaven . Rowohlt, 2011, ISBN 978-3-498-04517-3 , p. 76.
  27. Jo Marchant: The Decipherment of Heaven . Rowohlt, 2011, ISBN 978-3-498-04517-3 , p. 82.
  28. Jo Marchant: The Decipherment of Heaven . Rowohlt, 2011, ISBN 978-3-498-04517-3 , pp. 85-88.
  29. Derek de Solla Price: Gears from the Greeks. The Antikythera Mechanism: A Calendar Computer from ca. 80 B. C. Science History Publications, 1975, pp. 111/112.
  30. Jo Marchant: The Decipherment of Heaven . Rowohlt, 2011, ISBN 978-3-498-04517-3 , p. 113.
  31. Derek de Solla Price: Gears from the Greeks. The Antikythera Mechanism: A Calendar Computer from ca. 80 BC, Fig. 33: Sectional diagram of complete garing system. (No longer available online.) In: Science History Publications. 1975, p. 43 , archived from the original on April 23, 2013 ; accessed on October 26, 2012 (English).
  32. American Mathematical Society: The Antikythera Mechanism I; 3. The Sun-Moon Assembly. Retrieved October 26, 2012 (English).
  33. Derek de Solla Price: Gears from the Greeks. The Antikythera Mechanism: A Calendar Computer from ca. 80 B. C. Science History Publications, 1975, p. 43, Fig. 33.
  34. Jo Marchant: The Decipherment of Heaven . Rowohlt, 2011, ISBN 978-3-498-04517-3 , p. 160.
  35. ^ Allan Bromley (historian) in the English language Wikipedia.
  36. Jo Marchant: Video: Michael Wright explains his model of the Antikythera Mechanism. Retrieved October 26, 2012 (English).
  37. a b c d Jo Marchant: The decryption of the sky . Rowohlt, 2011, ISBN 978-3-498-04517-3 , pp. 193-198.
  38. Jo Marchant: The Decipherment of Heaven . Rowohlt, 2011, ISBN 978-3-498-04517-3 , p. 173.
  39. Michael Wright: The Antikythera Mechanism reconsidered in Interdisciplinary science reviews, March 2006, pp. 27-43.
  40. a b Mogi Vincentini: Video: The model of the Antikythera mechanism built by Michael Wright in an exploded view. Retrieved October 26, 2012 (English).
  41. ^ The Antikythera Mechanism Research Project, project overview. Retrieved November 3, 2012 .
  42. Jo Marchant: The Decipherment of Heaven . Rowohlt, 2011, ISBN 978-3-498-04517-3 , p. 212.
  43. Jo Marchant: The Decipherment of Heaven . Rowohlt, 2011, ISBN 978-3-498-04517-3 , pp. 229/230.
  44. Picture in the English language Wikipedia.
  45. a b c d e T. Freeth, Y. Bitsakis, X. Moussas, JH Seiradakis, A. Tselikas, H. Mangou, M. Zafeiropoulou, R. Hadland, D. Bate, A. Ramsey, M. Allen, A Crawley, P. Hockley, T. Malzbender, D. Gelb, W. Ambrisco, MG Edmunds: Decoding the ancient Greek astronomical calculator known as the Antikythera Mechanism . In: Nature . tape 444 , no. 7119 , October 30, 2006, p. 587–591 , doi : 10.1038 / nature05357 ( Internet archive [PDF; accessed on April 15, 2020]).
  46. Ziggurathss: YouTube video with a rotating crank loop ( Memento from April 23, 2010 in the Internet Archive )
  47. Jo Marchant: The Decipherment of Heaven. Rowohlt, 2011, ISBN 978-3-498-04517-3 , p. 229.
  48. Tony Freeth et al. a .: Calendars with Olympiad display and eclipse prediction on the Antikythera mechanism. In: Nature , July 31, 2008, pp. 614-17 (quotation, pp. 616/617)
  49. ^ A b c d Tony Freeth, Alexander Jones: The Cosmos in the Antikythera Mechanism. In: ISAW Papers 4 Preprint 2012. Retrieved November 2, 2012 (English).
  50. Martin Allen: What was it made of? In: Antikythera Mechanism Research Project. July 4, 2007, accessed October 26, 2012 .
  51. a b c M. T. Wright: The Antikythera Mechanism reconsidered . In: Interdisciplinary Science Reviews . tape 32 , no. 1 , 2007, p. 27–43 ( [PDF; accessed October 26, 2012]).
  52. Wolfram M. Lippe: The Antikythera gear train. (PDF; 1.3 MB) In: History of calculating machines. 2011, p. 5 , archived from the original on March 3, 2014 ; Retrieved October 26, 2012 .
  53. Figure 24 in Tony Freeth, Alexander Jones: The Cosmos in the Antikythera Mechanism 2012 Preprint.
  54. Martin Allen: How many gears does it have? In: Antikythera Mechanism Research Project. May 28, 2007, accessed October 26, 2012 .
  55. a b c d Wolfram M. Lippe: The Antikythera gear train. (PDF; 1.3 MB) In: History of calculating machines. 2011, p. 9 , archived from the original on March 3, 2014 ; Retrieved October 26, 2012 .
  56. Wolfram M. Lippe: The Antikythera gear train. (PDF; 1.3 MB) In: History of calculating machines. 2011, pp. 3, 9, 10 , archived from the original on March 3, 2014 ; Retrieved October 26, 2012 .
  57. James Evans, Christián C. Carman, Alan S. Thorndike: Solar Anomaly and Planetary Displays in the Antikythera Mechanism. (PDF; 2.1 MB) In: Journal for the History of Astronomy, February 2010, Pages 1-39. P. 3 (Photo 2) , accessed on October 26, 2012 (English).
  58. Tony Freeth, Alexander Jones, John M. Steele, Yanis Bitsakis: Calendars with Olympiad display and eclipse prediction on the Antikythera Mechanism . In: Nature . tape 454 , no. 7204 , July 31, 2008, p. 614-617 ( [PDF; accessed on October 26, 2012] p. 615 (2), Fig. 2, above).
  59. a b c d e f Tony Freeth, Alexander Jones, John M. Steele, Yanis Bitsakis: Calendars with Olympiad display and eclipse prediction on the Antikythera Mechanism . In: Nature . tape 454 , no. 7204 , July 31, 2008, p. 614–617 ( [PDF; accessed October 26, 2012]).
  60. Tony Freeth, Alexander Jones, John M. Steele, Yanis Bitsakis: Calendars with Olympiad display and eclipse prediction on the Antikythera Mechanism . In: Nature . tape 454 , no. 7204 , July 31, 2008, p. 614-617 ( [PDF; accessed on October 26, 2012] p. 615 (2), Fig. 2, below).
  61. Tony Freeth: Decoding an Ancient Computer. (PDF; 1.5 MB) In: Scientific American, December 2009. p. 79 , accessed on October 26, 2012 (English).
  62. Tony Freeth, Alexander Jones: The Cosmos in the Antikythera Mechanism. In: ISAW Papers 4 Preprint 2012. Retrieved April 15, 2020 (English).
  63. Wolfram M. Lippe: The Antikythera gear train. (PDF; 1.3 MB) In: History of calculating machines. 2011, p. 6 , archived from the original on March 3, 2014 ; Retrieved October 26, 2011 .
  64. Jo Marchant: The Decipherment of Heaven . Rowohlt, 2011, ISBN 978-3-498-04517-3 , p. 168.
  65. Jo Marchand: Video: Michael Wright explains his model of the Antikythera Mechanism. Retrieved October 26, 2012 (English).
  66. Tony Freeth: Decoding an Ancient Computer: Greek Technology Tracked the Heavens. In: Scientific American. November 30, 2009, accessed October 26, 2012 .
  67. Andrew Carol: Ancient calculating machine built from Lego bricks. Retrieved October 26, 2012 (English).
  68. : gears
  69. - information with images of the Lucerne duplicate and further links (English)
  70. John Gleave: Antikythera. (No longer available online.) In: Grand Illusions. Archived from the original on January 4, 2013 ; accessed on October 26, 2012 (English).
  71. Shahendra Ohneswere: New Watch Alert - Hublot. October 18, 2011, accessed October 26, 2012 .
  72. YouTube. Retrieved February 10, 2019 .