Sextant: A Voyage Guided by the Stars and the Men Who Mapped the World’s Oceans

Navigation posed many problems in the days before celestial navigation was perfected and the cause of a wreck can therefore seldom be attributed to any single factor. The notorious loss of the Dutch East India Company ship Batavia in 1629 is a case in point.

She was wrecked on a reef off the west coast of Australia after crossing the Indian Ocean on her way to Java. The reef in question was part of an extensive group of low islands discovered by a Dutch sailor called Frederik de Houtman in 1619, and the Batavia drove on to it under full sail. The lookout had in fact spotted white water ahead but the master, convinced that it was merely a reflection of the light of the moon, refused to alter course or shorten sail. A faulty DR estimate of the Batavia’s longitude might well have encouraged this disastrous misjudgement: the master apparently thought he was 600 miles from land.

Such an error would be quite understandable after a long ocean passage, especially since the prevailing westerly winds in the southern Indian Ocean generate an east-going current that would have been hard to detect. To confirm this, however, it would – at the very least – be necessary to know whether the master was using a chart on which the position of the reef was correctly recorded, and if so where he believed his ship was in relation to it. It is equally difficult to assess the reasons for the loss of the British warship Ramillies, which occurred as late as 1760. Having entered the English Channel, she was caught in a gale off the south coast of Devon and embayed. She tried to anchor but was driven ashore with the loss of all but twenty-seven of her crew of 800. Assuming that her commander knew the latitude correctly, was he mistaken in his longitude? Was the chart he was using in error? Or did poor visibility mean that those on watch failed to notice the approaching coast before it was too late? We have no way of knowing, but the impossibility of accurately determining their longitude put offshore navigators under a heavy handicap.

Commodore George Anson’s circumnavigation of the world provides the clearest illustrations of the navigational difficulties that sailors faced before the longitude problem had been solved. England being at war with Spain, Anson (1697–1762) was dispatched in 1740 with a squadron of ships and more than 1,900 sailors and soldiers

to harass the enemy’s colonial settlements in the Pacific. Long delayed by contrary winds, the squadron reached Madeira – their first port of call – in October, having already been at sea for forty days. The island’s longitude was laid down on contemporary charts – as it happens, quite accurately – in 17° West ‘of London’, but Anson and his men placed it somewhere between 18° 30' and 19° 30'.

Their DR was out by at least 75 miles. Five months later, in early March 1741, having passed through Le Maire Strait, north-east of Cape Horn, Anson’s ships struggled round into the Pacific, facing a succession of terrific gales: men were injured or lost overboard, some lost fingers or toes to frostbite, the ships began to leak heavily and both sails and rigging were frequently damaged. Typhus and dysentery had already weakened the squadron on the voyage south from Madeira, but scurvy too now began to take its toll.

On Anson’s flagship, the Centurion, the crew were so much weakened that they were unable to throw overboard the bodies of their dead shipmates. One old soldier, who had fought at the Battle of the Boyne in 1690, discovered as he lay dying that his long-healed, fifty-year-old wounds were reopening.

On 3 April the Centurion ran into a particularly severe storm:

In its first onset we received a furious shock from a sea which broke upon our larboard quarter, where it stove in the quarter gallery, and rushed into the ship like a deluge … to ease the stress upon the masts and shrouds we lowered both our main and fore-yards, and furled all our sails, and in this posture we lay to for three days …

Despite the appalling conditions Anson and his men were convinced that they had not only rounded Cape Horn but had also made good westerly progress out into the Pacific. Judging that it was now safe to head north, they were in for a nasty surprise. On the night of 13–14 April, when they thought they were hundreds of miles out to sea, the leading vessel caught sight of rocky cliffs – probably the western end of Noir Island off the south-west coast of Tierra del Fuego. She fired a gun and showed lights to warn the ships astern of her of the impending danger:

[the land] being but two miles distant, we were all under the most dreadful apprehensions of running on shore; which, had either the wind blown from its usual quarter [south-west] with its wonted vigour, or had not the moon suddenly shone out, not a ship amongst us could possibly have avoided …

One officer recalled seeing the cliffs rearing up ‘like two black Towers of an extraordinary height’, but every ship managed to get clear.

Anson’s own log gave the Centurion’s longitude on 13 April as 87° 51' W, while another surviving log gives a longitude of 84° 12' W just before land was sighted.

The difference between these estimates is itself an indication of how difficult it was to determine longitude reliably by DR. In fact the longitude of Noir Island is about 73 degrees West, which means that Anson’s estimate was out by nearly 14 degrees – a distance of almost 500 nautical miles in this latitude. The authorized account of the voyage plausibly places the blame for these very large errors on the unexpected strength of the ocean currents in this locality:

It was indeed most wonderful that the currents should have driven us to the eastward with such strength; for the whole squadron esteemed themselves upwards of ten degrees more westerly than this land, so that in running down, by our account, about nineteen degrees of longitude, we had not really advanced above half that distance.

The squadron – from which two ships had already separated – faced further battering by storms as it struggled northwards. By the end of April, as the death toll from scurvy rapidly mounted, each of the surviving vessels found itself alone. Having barely survived a hurricane at the end of May off the island of Chiloé,

the Centurion headed for a planned rendezvous at the island of Juan Fernández [now Robinson Crusoe Island] off the coast of Chile where desperately needed fresh provisions could be found. However, to save vital time, Anson ‘resolved, if possible, to hit the island upon a meridian [of longitude]’.

In other words, rather than heading north up the coast of Chile and then running down the island’s latitude in the time-honoured fashion, he took the chance of sailing directly for it. On 28 May they had nearly reached the latitude in which the island was laid down and ‘had great expectations of seeing it: But not finding it in the position in which the charts had taught [them] to expect it’ they were afraid that they might have gone too far to the west.

Though Anson himself was ‘strongly persuaded’ that he had glimpsed the island, his fellow officers were unconvinced and, following ‘a consultation’, it was agreed that they should head back to the east. They sighted the distant, snow-capped peaks of the Chilean cordillera on 30 May:

Though by this view of the land we ascertained our position, yet it gave us great uneasiness to find that we had so needlessly altered our course when we were, in all probability, just upon the point of making the island; for the mortality amongst us was now increased in a most dreadful degree, and those who remained alive were utterly dispirited by this new disappointment …

It took nine days to regain the ground they had lost and it was not until 10 June that the Centurion, whose crew at full strength would have numbered between four and five hundred men, at last reached Juan Fernández ‘with not above ten foremast men in a watch capable of doing duty’. This single navigational error had cost the lives of ‘between seventy and eighty of our men, whom we should doubtless have saved had we made the island [on 28 May], which, had we kept on our course for a few hours longer, we could not have failed to have done’.

Once the remnants of his squadron had gathered at Juan Fernández and the surviving members of the ships’ crews had recovered their strength, Anson carried out a raid on a small Spanish coastal settlement in Peru, and also captured a few merchant vessels. He next tried to intercept a Spanish treasure ship that was expected to sail from Acapulco, but news of his presence had reached the Spanish authorities and the ship remained safely in port. Disappointed but undaunted, in May 1742 he set out across the Pacific with his two surviving ships.

In the course of an agonizingly slow voyage, again greatly complicated by unreliable charts and uncertainties about their longitude, one ship had to be abandoned and the Centurion, with eight or ten men dying every day ‘like rotten sheep’, was barely afloat when she reached Tinian in the Marianas Islands in August.

Anson’s determination now, at last, paid off. He was able to reach Canton (modern Guangzhou) where the Centurion was repaired and then succeeded in capturing the treasure ship, Nuestra Señora de Covadonga, off the Philippines. The Covadonga was carrying more than 1.3 million pieces of eight and 35,682 ounces of silver. The exact value of the loot Anson finally brought home in 1744 is uncertain, but as a ship’s captain as well as commander-in-chief, he would have received three-eighths. His share of the treasure from the Covadonga alone may have amounted to the vast sum of £91,000. For comparison the pay due to him in the course of the whole voyage fell just short of £720.

The captured treasure was paraded through the streets of London to national jubilation, and little attention was paid to the fact that 1,400 men had failed to return home. Propelled by this success Anson was later ennobled and rose to the very pinnacle of the Royal Navy, serving twice as First Lord of the Admiralty.

Chapter 6 (#ulink_57200407-90fd-5a21-a7a9-c52d6c2d3c49)

The Marine Chronometer (#ulink_57200407-90fd-5a21-a7a9-c52d6c2d3c49)

Day 7: Stayed in my bunk ’til 0745 and then sat in the sun reading Slocum while watching clouds building up – the barometer is falling and the weather is starting to break. The wind veered to W and increased to F 5 so we took down the main and ran on under genoa at 6–7 knots. Colin was still not comfortable though, so we went right down to the pocket handkerchief No. 2 stays’l which cut our speed to 4 knots.

After lunch – the usual sandwiches though the bread is getting mouldy round the edges – I went to sleep or tried to for two hours. Much rolling and rattling. It’s amazing how much the weather affects one’s mood out here. All the same, we’ve made good progress so far and today we’ve covered 144 miles, noon to noon. Our course is now 105° magnetic. Helped Colin work out our noon position: 42° 34' N, 46° 16' W.

Celestial navigation would be easy if the sun and all the other heavenly bodies stood motionless in the sky – as Polaris does, almost.

It would then be possible to fix your position by sextant sights without the need to know the time or even the date. The cosmos, however, is not that obliging. Not only does the earth rotate completely once every day, but its axis of rotation – currently inclined at roughly 23.5 degrees to the plane of its orbit around the sun – also changes gradually over a cycle of about 25,800 years.

To complicate matters further, the earth’s orbit around the sun is elliptical rather than circular, with the result that the interval between one passage of the sun over the observer’s meridian and the next is not quite constant. So not only do the heavens appear to be in motion, but that motion itself is also changeful. This is most obviously revealed by the variations in the path of the sun across the sky – which is measured by its declination to the north or south of the equator – the phenomenon that gives rise to the seasons. The behaviour of the planets and the moon is yet more complex.

The ancient Greeks and Romans, who leaned heavily on earlier Babylonian learning, had a well-developed understanding of the paths that the various heavenly bodies described, as did the Arab astronomers who followed them. They clung, however, to the view – associated with the astronomer Ptolemy (c.90–168 CE) – that the earth was at the centre of the universe, and this theory prevailed until the time of Copernicus (1473–1543).

Though Ptolemaic orthodoxy may have been misguided, it did not prevent astronomers producing accurate solar declination tables as far back as the end of the fifteenth century. These enabled mariners for the first time to adjust their observations of the sun to allow for the seasonal changes in its meridian altitude. Now they could determine their latitude at noon as the sun crossed their meridian, as well as after dark (from the height of Polaris), subject to the limitations of the instruments then at their disposal. Moreover, they could continue to find their latitude when south of the equator – when Polaris had disappeared below the northern horizon. This breakthrough helped the Portuguese to open up an enormously valuable trade route into the Indian Ocean round the Cape of Good Hope. Early in the sixteenth century the Portuguese also devised a rule for determining latitude by reference to the stars of the Southern Cross – which lie some distance from the south celestial pole.

While latitude could be determined quite easily, the earth’s motions meant that the measurement of longitude was a much more difficult challenge. Early in the sixteenth century the astronomer Gemma Frisius (1508–55) realized that a promising approach to solving the longitude problem would be to find a way of measuring time accurately – whether on land or sea. An observer equipped with an accurate enough clock set to the time at a reference meridian could in principle compare the time of an event (such as sunrise or sunset) with the predicted time of the same event at a reference meridian – such as Greenwich.

The observer’s longitude could then be derived by converting the time difference in hours and minutes into a spatial displacement measured in degrees and minutes east or west – one hour being equal to 15 degrees of longitude (360 divided by 24).

It was not until the early seventeenth century that Copernican theory was firmly established on the basis of the observations of Tycho Brahe (1546–1601), Galileo Galilei (1564–1642) and Johannes Kepler (1571–1630). Galileo’s momentous discovery of the moons of Jupiter and, soon afterwards, of the changing phases of the planet Venus not only provided overwhelming evidence that the earth was not the centre of the universe but also opened the way to a proper understanding of planetary motion.

The invention of the first pendulum clock in the 1650s by Christiaan Huygens (1629–95) also marked a great advance. It was now possible for astronomers to measure time with sufficient precision to predict with great accuracy the positions of all the major heavenly bodies day by day – though, as we shall see, there was one troublesome exception: the moon. The establishment of the two great Royal Observatories in Paris (1667) and Greenwich (1675) contributed notably to this process. These technical developments, coupled with major theoretical advances – of which the publication in 1687 of Newton’s laws of motion was the most significant – were crucial steps on the path to the eventual solution of the longitude problem.

By the end of the seventeenth century, the laborious observations of astronomers had yielded the first accurate ephemeris tables.

An observer on dry land supplied with a pendulum clock could now regulate it by reference to the predicted events and thereby establish his or her longitude. French scientists were the first to apply the new technology to the making of accurate terrestrial maps and the results were sometimes surprising. In 1693 a new map of the coast of France based on an elaborate survey supervised by the astronomers Jean Picard (1620–82) and Philippe de La Hire (1640–1718) revealed that the kingdom had shrunk. The port of Brest, for example, had moved 50 miles to the east of its position on the best existing map. King Louis XIV is reputed to have complained that he had lost more territory to his astronomers than to his enemies.

The French undertook much basic research – including heroic efforts to determine the precise shape of the planet, a knowledge of which was essential if maps were accurately to reflect reality. Scientists were sent all over the world in an attempt to decide whether Newton’s prediction that the earth bulged slightly around the equator was correct. If it did, the geographical length of a degree of latitude would increase as one moved away from the equator towards the poles. While one such expedition went to Finland and another to South Africa, a third, led by Louis Godin, set out in 1735 for the Andes to try to measure a degree of latitude at the equator. Godin and his team endured extraordinary hardships, first struggling through tropical jungles and then working at heights of over 16,000 feet on the freezing mountaintops, as they carefully measured baselines and extended a network of triangles along the mountain chain over a distance of some 200 miles.

Their efforts, combined with the work of the other expeditions, confirmed Newton’s prediction.

The British were initially slow to learn from the French map-makers, but a growing awareness of the great military and commercial advantages conferred by good maps and charts prompted action. Murdoch Mackenzie Senior (1712–97) led the way with his pioneering marine survey of the Orkney Islands in the 1740s, based on a rigid system of triangulation using precisely measured baselines, the first of which was laid out on the frozen surface of a lake.

Mackenzie’s were the first accurate British charts, and he also invented the system of symbolic abbreviations that survives to this day. His Treatise on Maritim Surveying (published in 1774) was to set the pattern for every hydrographic survey conducted over the next hundred years, and in it he listed the quadrant and sextant as essential items of the marine surveyor’s equipment. Of the sextant he had this to say:

This instrument may be used with great Advantage in Maritim surveys, on most Occasions; being more portable, more readily applied to the taking of Angles, and generally more accurately and minutely divided than Theodolites are: an Observer is less liable to make mistakes with it; and, which is a very material advantage, he can take angles with it at sea, as well as on Land.

Mackenzie is also credited with the invention of the ‘station pointer’ – an invaluable instrument that enables the coastal navigator quickly to fix his position by taking horizontal sextant angles between three or more fixed points.

Not until the 1790s did the newly established Ordnance Survey follow Mackenzie’s example and begin mapping Britain by triangulation. In 1797 the intricate network of triangles was extended from Land’s End to the Scilly Isles and, to general consternation, it emerged that the position of the islands shown on contemporary charts was out by the astonishingly wide margin of 20 nautical miles.

Cold comfort for poor Shovell.

*

For all the progress that was being made in mapping the land, accurate position-fixing at sea still remained an impossible dream in the early eighteenth century. In fact it was no closer to reality than it had been 150 years earlier when the Spanish, conscious of the vital commercial importance of their overseas colonies and the difficulties surrounding accurate and therefore safe navigation, began to seek a shipboard solution to the longitude problem. In 1567 King Philip II offered the first cash prize to anyone who could crack it, and in 1598 his successor, Philip III, raised the stakes: the winner would receive a one-off payment of 6,000 ducats together with an annual pension of 2,000 ducats. So important was the goal that this princely annuity was promised even to the heirs of the eventual winner.