When the dragon swallows the sun

Eclipses have fascinated mankind throughout the ages – and even changed the way we view the world

March 18, 2015

A solar eclipse is an impressive sight to behold. Even thousands of years ago, people were spellbound by it and devised all kinds of myths to account for this phenomenon. It is even said that wars were won or lost when the new moon slowly slid across the sun’s disk. In the 18th and 19th centuries, researchers became increasingly interested in studying total eclipses. And in 1919, observations of such a cosmic game of shadows even reinforced Einstein’s General Theory of Relativity.

Text: Helmut Hornung

Cosmic game of shadows: a solar eclipse occurs when the sun, new moon and Earth are aligned, causing the moon to cast a shadow onto Earth's surface.


Natural phenomena move people, yet none more so than a solar eclipse. Even the ancient Babylonians, who observed the celestial bodies as they travelled across the night sky around 2000 BC, noticed a cycle lasting about 18 years after witnessing that similar eclipses kept recurring. Yet their priest astrologists could not explain what caused these events, nor were they able to predict them. After all, solar eclipses are a visible result of the perpetually ticking gears of the heavens, the workings of which astronomers did not begin to understand until the 16th century.

The actual race of the sun, moon and stars probably also remained a mystery to the Greek mathematician and philosopher Thales of Miletus. Nevertheless, it is said that he prophesied a total solar eclipse in the year 585 BC. According to the historian Herodotus, this eclipse over Asia Minor even determined the outcome of a war between the Lydians and the Medes. On the very same day that the decisive battle was being waged, the sun was obscured by the moon. Thales had warned the Lydians in advance, while the Medes had no knowledge of the upcoming darkness. Full of fear, they fled the battlefield and made peace with their enemies.

However, this cosmic game of shadows did not only instil fear and terror in the hearts of man in ancient times. When the moon blocked the sun over Kenia on 16 February 1980, many of the country’s citizens fled into their huts or used whatever they could find to make deafening noise in the hopes of banishing the Demon of Darkness. The Chinese, in the earlier centuries, witnessing such events first-hand probably reacted in a similar manner when, to their horror, a dragon with glinting eyes swallowed the sun. (Thankfully the mythical beast always decided to spit it out again afterwards.) In astrology, as the interpretation of stars is formally known, a solar eclipse is traditionally considered an evil omen.

Yet what actually causes this natural phenomenon? While the scientific answer to this question is far less exciting than the mythical tales of dragons and demons, it is based on what could be regarded as a magical coincidence. After all, the moon – which is almost 3500 kilometres in diameter – fits pretty much exactly over the sun, which has a diameter of 1.4 million kilometres. That is because the daystar is 400 times larger than the moon – and 400 times further away. As seen from Earth, both heavenly bodies therefore appear to have approximately the same diameter of about half a degree of arc.

In and of itself, however, this whim of nature is not enough to produce an eclipse. The moon’s trajectory has to pass right over the sun’s glowing disk. That can only happen at new moon. Only then – when observed from far away in outer space – is the moon aligned directly between Earth and sun. This constellation occurs every month, or every 29 days, 12 hours and 44 minutes, to be exact. Yet total solar eclipses are not very common. That is because the moon’s orbital plane is tilted at five degrees to the plane on which Earth orbits the sun once a year; this plane is known as the ecliptic. In most cases, the new moon misses the sun’s disk, passing by above or below it unnoticed.

Every now and then, however, the new moon is positioned very close to or directly in one of the two nodes, as astronomers call the points at which the moon’s orbit and the ecliptic intersect. Even today, these two orbital nodes are known as the Dragon’s Head and the Dragon’s Tail – in reminiscence of the mythical beast with an appetite for suns featured in ancient Chinese folklore. The degree of totality of each individual eclipse depends on the distances between moon and Earth and also between Earth and sun, among other factors. The most favourable constellation is when the new moon is close to Earth, which in turn is far away from the sun. Yet even then, the umbral cone covers an area on Earth’s surface that is a maximum of 300 kilometres wide, because only the cone’s tip ever touches our planet.

Since the umbral cone travels East at 2000 kilometres per hour, a total solar eclipse is a very fleeting affair in any geographical point located within this narrow corridor. Totality can last a maximum of 7 minutes and 31 seconds. The eclipse on March 20th will last 2 minutes and 47 seconds at most.

The umbra often does not reach Earth. If you are lucky enough to be standing in the cone’s extension, you will see an annular (ring-shaped) solar eclipse: all that remains visible of the sun is a thin border around the moon. Eclipses can start out being annular and then become total in the umbra’s corridor before they become ring-shaped again when they end.

Brain power: Theodor von Oppolzer and his associates catalogued 8000 solar- and 5200 lunar eclipses, calculating each of them by hand using only paper, pencils and slide rules. Their compilation Canon of Eclipses was published in Vienna in 1887, one year after Oppolzer's death.


And finally there is a third possible scenario: a partial solar eclipse, which is visible outside of the umbra wherever the 7000-kilometre wide penumbra brushes Earth. This natural phenomenon draws larger crowds because it is more common in certain locations than a total eclipse. A partial eclipse also occurs when the umbral cone does not reach the Earth, i.e. when a total eclipse is not visible anywhere on our planet.

In the modern day and age, computers and specialised programmes can be used to precisely predict eclipses right down to the second. When Theodor von Oppolzer and his associates set out to compile all (solar and lunar) eclipses that occurred or would occur between 10 November 1208 BC and 12 October 2163 AD, computers had not yet been invented. In the 1880s, the only tools they had at their disposal were paper, pencils and slide rules. It took the scientists years of painstaking effort to catalogue 8000 solar and 5200 lunar eclipses – each of them calculated in meticulous detail.

Oppolzer died one year before his book Canon of Eclipses (German: Canon der Finsternisse) was published in Vienna in 1887. He had long ago found an explanation for the 18-year saros cycle, which had already been observed by the ancient Babylonians. The sun passes a certain node in the moon’s orbit every 346.62 mean solar days; this period of time is known as an eclipse year and is around 20 days shorter than our common calendar year. Nineteen eclipse years are equivalent to 6585.78 solar days.

A synodic month – the interval between two new moons – lasts 29.5306 days. Coincidentally, 223 synodic months are almost exactly as long as 19 eclipse years or 18 calendar years, i.e. 6585.32 days. This implicates that after each of these 18-year periods, the celestial repertoire begins anew, due to the fact that virtually identical eclipse conditions are in place. The aforementioned Thales of Miletus probably predicted his solar eclipse with the help of this saros cycle.

An eclipse begins when a new moon first comes into contact with the sun’s eastern rim. This indentation gradually grows in size. When the moon completely occults the sun, i.e. when totality is achieved, the sun is surrounded by a shimmering, blue and white aura. The spikes that flare out into the sky from this aura can be up to twice as long as the diameter of the sun’s disk. At that precise moment, observers are looking at the corona, the sun’s hot outer atmosphere that reaches temperatures of approx. one million degrees Celsius. In times of minimal solar activity, it appears asymmetrical with short “plumes” at the poles and long rays in the equatorial regions. During the maximum of the solar cycle, the sun’s activity appears more uniform.

For modern day science, solar eclipses – whether partial or total – no longer yield large amounts of valuable information. In the 18th and 19th centuries, on the other hand, these moments of darkness were illuminating for many a scientist. Especially the hopes of observing the corona and solar prominences – large gaseous features that appear as filigree arcs on the sun’s rim during totality – led researchers to set off on expeditions to remote corners of the globe.

On 18 August 1868, for example, Jules Janssen travelled to India, where he separated the light of the prominences into a spectrum and discovered that these tongues of fire mainly consist of hydrogen. Back then, scientists already knew that the sun is a gigantic ball of gas that has no solid surface. The corona appeared to be the surface extending into the atmosphere. Charles A. Young and William Harkness were determined to study it in more detail during the eclipse that occurred on 7 August 1869. They did so by separating the pale light of this corona into a spectrum. This spectrum showed a bright line shining in the green range. They were not able to attribute it to any element known on Earth.

Young and Harkness had apparently discovered a new element – one that only exists in the sun’s corona, which is why it was named coronium. It was not until 1942 that the Swedish scientist Bengt Edlén identified this mysterious coronal emission line: it is in fact caused by iron atoms that each lost half of their 26 electrons. This is only possible in simultaneous states of very low dilution and extremely high gas temperatures. The corona had to have a temperature of one million degrees. Its approximately 300-kilometre thin photosphere – the visible gaseous envelope around the sun –, on the other hand, only reaches temperatures of 5500 degrees Celsius.

Tongues of fire: during a total eclipse, bright eruptions of hydrogen gas become visible close to the sun's rim, mostly in the form of filigree arcs. In the past, astronomers were particularly interested in these prominences. This photo was taken during the solar eclipse that occurred on 28 May 1900.


Total solar eclipses not only helped us gain a better understanding of the daystar that illuminates our planet. The difference between the predicted and the actual times of contact indicate that the moon’s orbit is disrupted and Earth’s rotation is irregular. Discrepancies were discovered in ancient and Arabic records. Scientists thus conclude that the moon moves away from Earth by four centimetres each year. Furthermore, our planet’s rotation seems to be slowing down; the day’s length increases by 0.0016 seconds per century.

The perhaps most significant discovery during a solar eclipse was made on 29 May 1919. It confirmed a new physical construct. In the year 1907, Albert Einstein pondered the question of how gravity influences the path of light – the same question the astronomer Johann Georg von Soldner had studied over a century earlier. If light is made up of particles, he mused, then it must obey gravity like a rock that is thrown up into the air and plummets back down to earth.

Einstein calculated that a ray of light brushing the sun should theoretically be diverted by 0.875 arc seconds due to the sun’s gravity. This hypothesis needed to be tested during a total solar eclipse, because only then would it be possible to observe the sun and the stars at the same time. The scientists had to measure the position of the stars near the sun’s rim and compare them to the data recorded in their catalogues to determine the discrepancy.

In the early 20th century, this type of observation required highly precise instruments. 0.875 arc seconds is a tiny unit, equivalent to approx. two thousandths of the full moon’s diameter. Yet in the year 1915, Einstein doubled this value, as dictated by his General Theory of Relativity, which postulates that mass practically bends space – just like a sleeping person forms a dent in the mattress. This curvature of spacetime should therefore divert the light’s trajectory and shift the location of a star close to the sun’s rim by 1.75 arc seconds.

On 8 March 1919, two groups of astronomers set off from England to go on expeditions. One group travelled to the island of Príncipe off the coast of Spanish Guinea, while the other headed for the city of Sobral in northern Brazil. The aim of the researchers was to observe the total eclipse that would take place on 29 May 1919. On Príncipe, Sir Arthur Stanley Eddington was able to record the phenomenon in 16 photographs, of which only two were viable. They each showed five stars – and the effect predicted by Einstein. Andrew Grommelin’s mission in northern Brazil was successful as well. His eight photographic plates depicted stars that had been displaced by roughly the predicted value.

Today we know that these observations were largely caused by measurement errors. Nevertheless, Einstein’s General Theory of Relativity had passed its first test. “Revolution in Science”, the headline of the London Times exclaimed on 7 November 1919. Once again, astronomical observations had changed the way we view the world.

Go to Editor View