Duncan Steel starts his book by playing a little trick on the reader. Which Eclipse, he asks, has exerted the greatest influence over our affairs? Aha! Anybody notice that capital "E"? This Eclipse is not a celestial phenomenon but an eighteenth-century racehorse; never once beaten in a race. After his retirement he spent 20 years at stud, as a result of which every thoroughbred nowadays carries a few of his genes. An annual race at Sandown Park and an annual set of horse awards in the US bear his name.
This is not just trivial. In Britain, Ireland and France, horse racing and its associated gambling make an important contribution to the economy, while some countries see it as among their largest industries. In the same way, total solar eclipses also play their part in the world economy as dedicated eclipse-chasers pursue them to every part of the globe, often giving a much-needed boost to the tourist industry in poor and backward countries.
Lunar eclipses occur at a rate of about fifteen per decade, a little less than half of which are total, and are, of course, visible to everybody on the night side of the Earth. Steel persists in the idea, popular with journalists, of the eclipsed Moon turning to "the colour of blood". This is something I can honestly say I've never seen. Copper-coloured, yes. The colour of old brass, yes. Dark grey, even. But blood-red no way.
For total solar eclipses, the average number in a decade is seven or eight, but since the Moon's shadow as it tracks across the Earth's surface is only around 60 to 100 miles (100-160km) in diameter, you have to be in the right place at the right time to see them. On the other hand, the penumbra is over 4000 miles (6400km) in diameter, and covers a large fraction of the dayside, giving a fairly wide area from which a partial eclipse can be seen.
The most famous solar eclipse recorded in the Bible, and dated by modern methods to September 30, 1131 BC, is the one where Joshua described the Sun as standing still and even moving backwards. This has long puzzled scholars. The same claim has been made for several other eclipses, including one in Ireland in the fifteenth century; but the phenomenon seems to be merely a visual illusion caused by the Moon overtaking the Sun so that the Sun appears to slip backwards.
Eclipses, of course, have always been taken as omens, arousing fear and dread, and there is even a recorded case of a king being scared to death by one. Charlemagne, emperor of the Holy Roman Empire, died in 814, but between 807 and 810 a peculiar set of solar and lunar eclipses had been visible from his kingdom. Although he himself understood that they were natural phenomena, his son and successor Louis got it into his head that the eclipses had been evil portents of his father's death. When, on May 5 840, a total solar eclipse took place, he believed that his days in turn must be numbered, and never recovered from the fright it gave him, dying a month later.
Until a couple of decades ago, computer programmes were generally punched on to 80-byte cards dating back to Herman Hollerith, who introduced a machine in the late nineteenth century to process the information resulting from a population census of the USA. The basic idea of coded cards came earlier, dating back to a loom capable of producing woven patterns, the invention of Joseph-Marie Jacquard (1752-1834). The Jacquard loom operated with flat tablets of wood with a series of holes cut into them, each hole causing a particular thread to be raised, whereas unpunched wood had the effect of making another thread drop.
When Charles Babbage was working on his "analytical engine" in the 1830s, his intention was to use a card system copied from the Jacquard system in his machine, but the project was never completed. This is connected with eclipses in two ways. The first is that Babbage's earlier proposed calculating machine, the "difference engine" also never completed was intended for the computation of mathematical and astronomical tables such as might be used to predict eclipses. It was also intended to correct mistakes in the nautical almanac, which was prepared by human "computers" rather than an error-free machine.
The second connection is that a Jacquard weave provides a parallel to the pattern of eclipse recurrences. Steel's "eclipse tapestries" for both solar and lunar eclipses show the result of plotting total, partial and annular solar eclipses, and total and partial lunar eclipses, against the days of the year, 1 to 365, and the years from 1900 to 2300. Patterns emerge revealing the 19-year metonic cycle, the saros sequences, and another shorter cycle of 3.8 years (47 synodic months) after which eclipses are repeated. A further downward-sloping line is produced by eclipses coming earlier from year to year by an amount representing the difference between the solar and the lunar year 10.88 days.
This was where the penny dropped on something odd that had been vaguely bothering me about the book. What it says is not 10.88, as above, but I0.88. In fact, all through the text, including the page numbers, a capital "I" regularly appears in place of the figure "1". This may be a quirk of the typeface used although it's hard to imagine a font lacking a figure 1 but the effect is quite weird, making the number I0, for instance, look like the name of one of Jupiter's satellites.
Odder still, the same substitution also appears throughout another book, J.P.McEvoy's Eclipse. Although the books are from different publishers, both were printed in Cornwall, so maybe this is a peculiarity of Cornish printers, or else it's a result of the race to get eclipse books out in time for August 1999 and could also account for small mistakes like Duncan Steel's "anthropomorphic" greenhouse effect.
Modern astronomers predict eclipses through their knowledge of the very complicated movements of the Sun and Moon relative to the Earth; ancient predictors had to rely on historical records, from which they worked out cycles of regular recurrences. The importance placed by emperors and monarchs on eclipses as omens made this a high-profile activity, and getting it wrong cost many predictors their lives.
A significant eclipse, the first known to have been observed in England, took place, according to records in English monasteries, on May 3, 664AD. This was also the year of the Synod of Whitby, at which King Oswy of Northumberland decided to abandon the Celtic Church in favour of the Roman mainly on the strength of his belief that the Roman Church had more expertise in the movements of the Sun and Moon and hence the calculation of the all-important date of Easter. In reality, the Roman tables contained an error in the date of the new Moon and therefore also of the solar eclipse, which the Church craftily covered up by falsifying the records and pretending the eclipse had happened on May 3, although every-where else in continental Europe the correct date of May 1 was recorded.
The first real predictor of eclipses was Edmond Halley (1656-1742), who discovered the saros, or rather rediscovered it, since it had been known to the Babylonians at least as early as the third century BC. Halley went beyond the Babylonians, though, in recognising that eclipses repeated on the cycle also repeated the same characteristics of duration, etc.
During medieval times, short-term predictions of eclipses could be made using the metonic cycle (19 solar years equals 235 lunations). In 1472 the first astronomical almanac was printed by Johannes Mόller (Regiomontanus, a latinized form of the city of Kφnigsburg where he was born). His tables of lunar eclipses were used in the 1580s when Sir Walter Raleigh was exploring the possibility of setting up an English colony in North America. It was known from the tables that a total lunar eclipse would occur on November 17-18, 1584, so two astronomers and their assistants were landed on Roanoke Island, where they timed the start of the eclipse. Astronomers back in England did the same, and the difference in time gave the difference in longitude or, in effect, the width of the Atlantic Ocean.
Christopher Columbus also used the tables to scare the natives of Jamaica into submission with a total lunar eclipse on February 29, 1504. This is supposed to be a true story, but fictional eclipses abound as well, and Steel mentions a solar eclipse described in Rider Haggard's King Solomon's Mines, in which, according to Haggard's account, totality apparently lasted for upwards of an hour. (This is also mentioned in the Guinness Book of Astronomy, which says that in the first edition of the novel, Haggard wrote of a full Moon one night being followed by a total eclipse of the Sun next day; but when his mistake was pointed out, he changed the eclipse to a lunar one in the second edition.)
Sir Arthur Eddington took advantage of the eclipse of May 29 1919 to verify Einstein's contention that rays of light passing the Sun were bent twice much as was predicted by Newton's theory. This was an ideal eclipse, with a totality of 6 minutes 51 seconds and the Sun passing through the Hyades, giving a background of bright stars to measure. In spite of sporadic cloud cover, photographs were secured. The astronomers then had to remain on site for some months so that they could photograph the same star fields in the absence of the Sun, for comparison. It was November before the results of the investigation were revealed as confirming Einstein's theory, and they made headlines around the world, securing instant fame for Einstein.
The effect of the deflection of light beams around the Sun is to make it act as a gravitational lens. Steel mentions the idea of placing a satellite at the focus of this lens, resulting in a "telescope" with an aperture of 865,000 miles, giving unheard-of resolving power; but since the satellite would have to be at a distance of 500 AU, about twelve times as far away as Pluto, it seems hardly practical at the moment. There is, though, the possibility that the Earth might happen to be near the focus of a gravitational lens produced by some other star, and Einstein published a paper on this subject, but regarded the idea as being only of theoretical interest. A year later another astronomer suggested that distant galaxies might produce the same effect, but it was forty years before the first example was discovered. In the 1990s the discovery of lensing galaxies became commonplace, and an illustration in the book shows a Hubble photograph of a massive galaxy surrounded by four gravitationally lensed images of a more distant quasar, an effect known as the Einstein Cross. But can this really be regarded as an eclipse, asks Steel anxiously, not wanting us to think he's straying from his subject and padding out the book. Yes, by a litttle bit of lateral thinking, it can because the galaxy is blocking our direct view of the quasar!
In Halley's time, the corona was still thought to be of lunar rather than solar origin, a sort of lunar atmosphere which could be seen only when the Sun was illuminating the Moon from behind. It was not until 1890 that it was definitely known to belong to the Sun. The advent of the spectroscope revealed unknown lines in the coronal spectrum, suggesting that a new element not found on Earth was present. Sir Norman Lockyer named this element "helium". Then another new element, Coronium, was found in the solar corona, and yet another, called Nebulium, in distant galaxies. All three turned out to be familiar elements subjected to temperatures of millions of degrees, and therefore emitting lines unachievable in terrestrial laboratories.
The corona can be studied without an eclipse by means of a coronagraph, an instrument in which a disk artificially blocks out the bright body of the Sun, and this is best done outside the Earth's atmosphere with the use of satellites such as the Solar and Heliospheric Observatory (SOHO), which has allowed the corona to be studied out to thirty times the Sun's radius. Another method is the use of a special filter which admits only the red H line at 6563 . This cuts out much of the light from the photosphere while admitting light from the hot hydrogen in the chromosphere and corona, in effect "eclipsing" the solar disk.
Tidal friction slows down the Earth's rotation; modern high-tech methods have shown that the day in 1999 is about 0.17 milliseconds longer than it was in 1989. A small change, but it builds up over the centuries, as can be shown by solar eclipses recorded in ancient times. For instance, a chronicle for 181BC reports that a total solar eclipse was seen at Chang'an, then the capital of China. Because the track of a total eclipse is so narrow, astronomers were able to calculate backwards and pin-point its position on the globe. If the Earth's rate of rotation had been the same as it is now, Chang'an would never have seen the eclipse, which would have happened some distance west of it. But in fact the day then was marginally shorter, resulting in the Sun reaching the meridian maginally sooner each day than it does now. Over the centuries these tiny amounts add up and in this case they made a difference of 3 hours and 20 minutes, or 50 of longitude.
The tidal force that slows the Earth's spin rate also speeds up the Moon, causing it to recede from the Earth and preserving the total angular momentum of the system. In 1969-72 the Apollo astronauts left several reflectors on the Moon's surface. By bouncing short laser-light pulses off them, physicists can measure the round trip as taking slightly over 2.5 seconds, giving an accurate distance for the Moon. Over the past three decades, they have recorded an annual increase in the distance of an inch and a half. The lod (length of day) is calculated as increasing by 1.7 milliseconds per century; however, going by the Moon's measured rate of recession, the equivalent increase in the lod would be 2.3 milliseconds, a difference of 0.6 milliseconds.
The discrepancy is partly due to the rebound of continental crust since the removal of the huge weight of ice after the last Ice Age ended, about 10,000 years ago, and partly to the migration of the resulting melt-water. The effect is a marginal change in the shape of the Earth as a whole, reducing its oblateness and making it more spherical. Less of an equatorial bulge tends to reduce the tidal drag of the Moon and Sun and increase the planet's rate of spin by the amount of ... exactly 0.6 milliseconds per century!
The last time a total eclipse crossed London was in 1715, and this had been predicted by Edmond Halley, who published a chart showing the position of the track. After the eclipse he received a flood of information from numerous observers, mainly amateurs, which enabled him to fix the exact dimensions of the track. It proved to be considerably wider than his prediction, the reason being that precise distances and sizes for the Sun and Moon were not then available.
Halley's data later proved useful to twentieth-century scientists. Sunspot numbers are known to vary, and this is suspected to have some influence on the Earth's climate. Studies of stellar evolution also indicate that, since it first went nuclear, the Sun's energy output has increased by about 40%. Scientists talk of the "Early Faint Sun Paradox" the fact that, if the Sun really had been initially so much fainter, Earth would have been a ball of ice, raising the question of how the first primitive living things could manage to evolve and survive.
Any increase or decrease in the Sun's size would similarly be a factor in the Earth's development. In the 1980s, comparison between measurements of the apparent solar diameter made in the late 17th and early 18th centuries and those made in recent times seemed to show that three hundred years ago the Sun was 4 arcseconds wider than it is now, that is, about 1 part in 480 of the solar diameter. If these early measurements were accurate, the Sun must be shrinking, which might cause it to heat up and thus contribute to the greenhouse effect. Some check was needed to settle the matter, and it was found in 1988 when researchers studying Halley's report of the 1715 eclipse, realised that, if the Sun really had been 4 arcseconds larger, the path of totality would have been some 6.5 miles narrower.
It was perhaps a bit cheeky of Duncan Steel to introduce chapters on what he calls Eclipses of the Third and Fourth Kind actually meaning occultations and transits. True enough, a total eclipse of the Sun could just about be classed as an occultation one celestial body passing in front of another and obscuring it although it isn't usual; but calling a transit an eclipse is rather stretching things. However, our author can be forgiven, because these chapters contain some very interesting material.
The wave nature of electromagnetic radiation imposes a fundamental limit on the resolving power of even the best telescope, making it impossible to measure the disks of even the nearest and largest stars. But this can be achieved by means of occultations. If the light from a star is fed into a detector which gives a readout of its intensity as it changes every microsecond, the variation as the Moon passes across the disk allows its size to be deduced. The same technique can be used with close binaries, galaxies and quasars.
Useful information can also be gleaned from occultations by asteroids not about the occulted star this time, but about the asteroid itself. Unlike in the wonderful flyby pictures of such bodies as Gaspra and Ida, the average asteroid appears in the telescope as a mere speck of light. But as its minute shadow, cast by some distant star, races across the surface of the globe, just as the Moon's does in a solar eclipse, observers stationed in the shadow path and along its immediate fringes can gather data that will reveal the size and shape of the occulting asteroid, along the axis perpendicular to its apparent motion as well as along the shadow path itself.
In the case of comets, the revealing factor is non-occultation; stars are seen to shine undimmed through the insubstantial material of the coma. Similarly with Mars. In 1830 an amateur astronomer, Sir James South, observed the occultation of a bright star in Leo by the planet Mars, and noted that it did not suddenly vanish but that the light slowly flickered and faded. There could be only one answer: Mars had an atmosphere.
South's private observatory was situated among the (then) green fields of the borough of Kensington. The map still shows Observatory Gardens, now a road (location of the Observatory House Hotel where Harry and I used to stay on visits to London). The observatory stood for 40 years until South's death in 1867, and the blue plaque marking it (not there in our day, as far as I know) states that it housed the largest telescope in the world at the time not quite correctly, it seems, since its 12-inch refractor was outdone by some of William Herschel's reflecting telescopes. South was one of the founding members of the Astronomical Society of London and, through the influence of aristocratic visitors who flocked to look through his telescopes, played a vital part in securing the Society's royal patronage and Charter, granted in 1831.
In 1977 a good occultation by Uranus of a bright star was due, and a Nasa aircraft was used to follow it along the shadow path, the idea being to gather more information about the planet's atmosphere before Voyager 2 reached it. To the surprise of the observers, a series of dips in the light signal revealed that Uranus, like Saturn, possessed a set of rings, tipped sideways because of the planet's orientation. Voyager 2 was instructed to look for them, and they were later featured in pictures from the Hubble telescope
Similar occultation observations of Neptune suggested that it too seemed to possess rings, but of a rather puzzling and elusive nature. The mystery was solved when, in 1989, Voyager 2 sent back data revealing the planet's unique arc-rings. Even Saturn's well-known rings revealed a new complexity when the Voyagers recorded blips in the light of stars occulted by them as the spacecraft swept past. Instead of being simply flat, featureless bands, the rings were shown to consist of thousands of individual strands, twisting and twining and "shepherded" by tiny Moons.
By a gift of fortune to astronomers, the atmosphere of distant Pluto was also detected through an occultation. Pluto was near perihelion when in June 1988 it passed in front of a bright star, allowing the presence of an atmosphere to be confirmed by the dimming of the light. But the atmosphere, consisting of carbon monoxide, nitrogen and methane, is only temporary; as Pluto recedes from the Sun, the gases will freeze and the atmosphere vanish.
“Eclipses” of the Fourth Kind were (until now!) usually known simply as transits, and were associated with Venus, Mercury and, in its day, Vulcan, whose transits were regularly reported by enthusiasts until its non-existence was finally demonstrated. Transits of Mercury occur in May or November, at the nodes of its orbit, May being better because Mercury is then near aphelion and so presents a larger disk. On average thirteen transits of Mercury take place each century, at rather irregular intervals governed by a complicated series of cycles. The first transit to be observed was predicted by Kepler and seen from Paris in November 1631, by Pierre Gassendi and others. Twenty years later, in 1651, another transit of Mercury was calculated by an English amateur, Jeremiah Shakerley who, realising it would occur during the night in Britain, travelled all the way to India to observe it, earning him the title from Steel of the "first eclipse-chaser".
The Scottish mathematician James Gregory proposed in 1663 that transits of the inner planets could be used to find the distance of the Sun, but it was Edmund Halley who, having seen a transit of Mercury in November 1677 from St Helena, developed the technique. However, he realised that only transits of Venus would be of use, although he could not hope to see one himself since the next were not due until 1761 and 1769.
Transits of Venus occur more regularly than Mercury's, at intervals of 8.0, 121.5, 8.0 and 105.5 years. Venus orbits the Sun thirteen times to our eight, and therefore after eight years returns approximately to the same alignment with the Earth at one of its nodes, in early June or early December. After that, because of the precession of the two orbits, the alignment moves beyond the ecliptic limits and no transit can take place.
Only five transits of Venus have ever been observed in December 1639, June 1761 and 1769, and December 1874 and 1882. The next pair are due in June 2004 and 2012. Kepler had predicted that in 1639 Venus would pass just below the Sun, but an amateur astronomer, the Reverend Jeremiah Horrocks, recalculated the figures, realised that the transit would take place, and managed to rush home from church in time to see it. The sighting was confirmed by his friend William Crabtree, the only other observer he managed to alert.
Kepler's laws of planetary motion give the relative distances between the bodies in the solar system, so an accurate measurement of one distance can give the scale of the whole. Halley's method depended on the fact that observers at two different locations on Earth will see Venus cross the Sun on two different paths (chords). If the distance between the observers is known, and the angular distance between the two chords can be accurately measured, the distance between the Earth and Venus can be calculated. According to Kepler's laws, the ratio between Venus's distance from the Earth at inferior conjunction and its distance from the Sun is 72:28, therefore from this the Earth-Sun distance can be found.
Preparations for the 1761 transit were put in hand by British, French and other teams. Finding the scale of the solar system was of prime importance at this time, since an accurate distance for the Moon was essential for the compilation of navigation tables. The main point was to observe the transit from situations as far apart in latitude as possible. In this case the point furthest north at which the whole transit was timed was near 66 , the furthest south was Calcutta at 22.5 N, giving a separation of only about 40 . The British had intended to send a team to a location in Sumatra, 4 south of the equator, but as they were then engaged in driving the French out of India, the expedition which set forth under Charles Mason and Jeremiah Dixon (of the Mason-Dixon line) was not surprisingly fired on by a hostile French frigate in the Channel and had to limp back to Portsmouth with a number of dead and wounded. The expedition finally made it to the Cape of Good Hope, from where only the egress could be observed. Unfortunately, the mathematician charged with analysing the results decided, mistakenly, that their timing was inaccurate, "corrected" it, and came up with a result for the Sun's distance that was more than 10% too low. The outcome was a bit of a shambles which no one was anxious to see repeated in 1769.
In 1768, Lieutenant James Cook was despatched in the Endeavour to Tahiti then known, to the British anyway, as King George's Island to observe the 1769 transit. Camp was set up at a location Cook called Fort Venus, because of the trouble they experienced in guarding their equipment and supplies from thieving natives. On the day, not a cloud was seen in the sky and observations were successfully completed. Cook then opened a sealed envelope he had been given at the beginning of his voyage and, in accordance with the instructions it contained, embarked on a search for the fabled southern continent, Terra Australis Incognita, eventually claiming New Zealand and the east coast of Australia for the British crown.
A transit of Mercury was also due in 1769, in November, and its timing had been calculated with some precision. Cook chose a suitable location in New Zealand, near modern-day Auckland, which he named Mercury Bay, set up his equipment and, by comparing the time the transit was visible in England with the local time according to the Sun, successfully determined the longitude of New Zealand.
In addition to Cook's Tahiti observations, timings were made at Murmansk, Hudson's Bay and locations in Norway. The date, June 3, was fortunate, since the Sun is above the horizon for most of the 24 hours in these latitudes. Again, though, the mathematician handling the results chose his data in a dubious way, selecting some measurements and ignoring others seemingly at random, and coming up with a distance for the Sun not greatly different from the 1761 result. On the whole, it could be said that the costly expeditions for both transits turned out to be dismal failures.
When the transit of 1874 came around, the British, French, Russians, Germans and Americans sent out numerous expeditions, the US alone sending three, to Siberia, Japan and China, to obtain northern sightings, and five to New Zealand and Australia for southern sightings. This was the first transit when photography was used, but many groups were unlucky with the weather, and again results were disappointing.
In 1882 there was less enthusiasm. The transit method depends on timing the exact moment of ingress and egress, which is difficult because of optical illusions caused by Venus's atmosphere, such as the "Black Drop" effect, making the method unreliable; and by this time new ways of finding the scale of the solar system were being developed. French astronomers had attempted to do this in 1751 by measuring the parallax of Mars from two locations 6,000 miles apart, but without great success. Then in 1898 the Earth-approaching asteroid Eros was discovered and its parallax measured during a close pass (15 million miles). Finally, radar pulses bounced off the planet Venus established the solar distance with an accuracy undreamt of by Halley, Cook or any other earlier observers.
"Eclipses" are of course regularly seen with Jupiter's satellites, either passing in front of their primary (transit) or vanishing into its shadow (geniune eclipse). The possibility of one of the planets occulting another also exists. In 1591 Kepler believed that he had seen Mars occult Jupiter, but modern back-computations show that it was actually a near miss. The only recorded Venus- Mercury occultation was witnessed in 1737 by one man, John Bevis of Greenwich, and even that was later shown not to have been total. For the future, the predictions are that Mercury may occult Neptune in 2067, Venus may occult Jupiter in 2123, and Mars may do the same a hundred years later, in 2223. Close planetary encounters, of whatever kind, are obviously very rare occurrences.
The first eclipsing binary to be recognised was Algol. Variations in its light were recorded in 1667, then in 1882 a systematic study was made by 18-year-old amateur John Goodricke, who noted that its brightness decreased and increased regularly over a cycle of 69 hours, and suggested that this might be due to some unseen body orbiting the star. A few years later, the sizes of the two stars and the distance between them was calculated. One star is brighter than the other, so when it passes in front, Algol's light falls off only slightly, but when the larger, dimmer star eclipses it, the drop is considerable. Binary asteroids have also been detected in this way, for instance, Dionysus, which consists of two lumps of rock orbiting each other.
Chapter 11 sorry, Chapter II gives us some advice on watching the August 1999 eclipse, the most important being not to go to Cornwall. Metz, Stuttgart, Bucharest, even Iran, Pakistan or India anywhere except Cornwall. Too right, as we know. As consolation, though, Steel adds that if it rains just a little, and the clouds allow a small glimpse of the chromosphere and corona, the viewer may be privileged to see a pink rainbow pink because the arc is dominated by the red of the chromosphere with little colour input from other parts of the spectrum.
An Appendix, useful for reference, sets out the complexities eclipse predictors have to cope with in the Earth-Moon-Sun system, from long-term ones such as variations in the eccentricity of the Earth's orbit and the tilt of its axis to the ecliptic, and relatively shorter ones like the precession of the equinoxes and the precession of the perihelion. The Moon's perigee also moves round its orbit, in the same direction as the Earth's rotation (eastward) in a period of 8.85 years, and the line of nodes retrogrades (in the opposite direction) in a period of 18.61 years, the node moving backwards to meet the Moon at the rate of 19.5 per year. There's more, lots of it, but we won't go into all that now.
While Earth's shadow stretches into space for 1,382,000km (850,000 miles), with a slight variation because of the ellipticity of the terrestrial orbit, the Moon's shadow on average is only 375,000km (232,500 miles), a little shorter than its mean distance from the Earth of 382,160km (236,900 miles). Under favourable circumstances, though, it can touch the Earth's surface (obviously, or else we'd never see a total eclipse), and even occasionally extend 29,300km (18,200 miles) beyond, giving the shadow its maximum possible cross-section of 269km (167 miles).
Finally, it seems that we inhabitants of the northern hemisphere come off better in the total solar eclipse stakes than our southern neighbours. More total eclipses occur between May and August than between November and February, because the Earth is then near aphelion (beginning of July), so that the Sun's apparent disk is smaller and therefore more likely to be covered by the Moon. At the same time, the northern hemisphere is inclined towards the Sun so that a larger area faces sunward than in the southern hemisphere; the overall effect being that, in the long run, the lion's share of total eclipses falls to the north.