by Graham Robb
He sailed through the Pillars of Hercules – or, if the Carthaginian blockade was in force, took the overland route along the rivers of Aquitanian Gaul – to the Atlantic coast. He passed through Corbilo at the mouth of the Loire. The mud of the estuary was already swallowing the port. Long before the Romans arrived, it disappeared, and nothing now remains of it, except perhaps a sandbank called the Banc de Bilho. From Corbilo, he sailed along the granite coast of a peninsula called Ouexisame in the Celtic tongue; its inhabitants were the Osismi – ‘the People at the End of the World’. The name has survived in ‘Ouessant’ (Ushant), the island off Finistère which marks the western extremity of Gaul. In February 1959, a few miles along the coast at Lampaul-Ploudalmézeau, a man who had been collecting seaweed for his vegetable patch noticed a gold coin glinting among his lettuces. It had been minted in the Mediterranean city of Cyrene in about 320 BC – a sign of how far the Greek trading empire extended, unless it came from the treasure chest of Pytheas himself. (Cyrenean gold coins would have made better bartering tokens than the drab, bronze coins of Massalia.)
North of Ouexisame, there was no sight of land for a day or two, until the stormy promontories of Belerion (Land’s End and the Lizard). This was the southernmost tip of the semi-mythical island or islands called Prettanike. Pytheas sailed along the busy coast among the coracles and canoes and larger vessels from the Atlantic lanes, to the headland at the other end of southern Britain: Kantion (Kent). According to one source, he then walked all the way through Britain. The natives lived in houses made of reed or logs, and threshed their wheat indoors because of the rainy climate. They made a beverage from honey and grain, which they drank with a dismal potage of millet, roots and herbs, having very little meat or fruit.
As he went, Pytheas calculated his latitude from the elevation of the mid-winter sun. In one place, it rose four peches (about eight degrees) above the horizon, which implies a latitude between the Peak District and the Yorkshire Dales. The next reading showed a solar elevation of only three peches (somewhere near the Moray Firth in Sutherland). Here, in the middle of a fourth-century BC winter, the history of Britain begins – not with Caesar’s summer raiding parties three hundred years later, but with the first identifiable visitor, a scientific traveller with a name, a place of birth and geographical coordinates, muddying his Mediterranean shoes on the soil of an island whose very existence was in doubt.
He reached the northernmost part of Prettanike at a place called Orka – possibly Duncansby Head, which looks over to the Orkneys. He had now gone far beyond even the imaginings of Homer. He set sail again, perhaps in a native boat. Six days out from Orka was an island called Thoule (the Faroes or Iceland), and a place where the sun kept watch all night in summer, barely rising from its bed. Further still, he came to a region that was neither land nor sea but a mixture of all the elements ‘on which one can neither walk nor sail’. In the fog banks and pack ice of the Arctic, he saw the earth in its troubled infancy or its confused old age.
He returned through the amber-rich Baltic and probably followed the river Dnieper to the Black Sea. In effect, Pytheas circumnavigated Europe. In the warmth and bustle of Massalia, or perhaps in a villa on Cap Croisette, he wrote up his journal. It was known by the title Peri tou okeanou (‘On the Ocean’), and it became one of the most famous books of the ancient world. No copy has ever been found – though some palimpsestic remnant may yet be hiding in a monastery – and one of the greatest voyages of discovery ever made is known only from brief, generally hostile references in a handful of Greek and Latin texts. The main source is the Geography of Strabo (7 BC), the geographer whose name means ‘one who can’t see straight’. The fantastic tales of that impudent Greek from southern Gaul made him burn with envy. How could Pytheas have talked to people who lived six days north of Britannia and thus beyond the limits of the habitable world? It was well known in 7 BC that nothing north of Ierne (Ireland) supported human life; Ierne itself was the home of ‘complete savages [who] lead a miserable existence because of the cold’. The exotic place names – Orka, Thoule, the Bed of the Sun – were obviously invented by Pytheas to give his incredible account an air of truth . . .
But even Strabo grudgingly admitted that there might be something to Pytheas’s scientific observations: ‘If judged by the science of the celestial phenomena and by mathematical theory, he might possibly seem to have made adequate use of the facts.’ This was, after all, the great contribution of Pytheas: whatever the official reason for his expedition, he had been collecting the priceless gems of evidence that would make it possible to construct a map of the world.
18. The Voyage of Pytheas
A scientifically produced map of the world in the fourth century BC seems as unlikely as the Antikythera Mechanism. If the warped, amoebic continents of medieval maps represent the sum of geographical wisdom, what formless fictions must have lived in the minds of the ancients? And yet, though precise measurements were the rarest of rare commodities, the principles were well established. Eighty years after the voyage of Pytheas, the chief librarian of the Library of Alexandria, Eratosthenes of Cyrene, made a momentous calculation. He had been told that at noon on the longest day of the year the sun at Syene (Aswan) shone directly down a well as though pouring its light carefully into the shaft without spilling a drop. A stick planted upright in the earth at Syene cast no shadow, whereas at Alexandria, eight hundred and fifty kilometres to the north, at exactly the same time, there was a very noticeable shadow. Given the distance between Syene and Alexandria (five thousand stadia) and the angle of the shadow to the sun’s rays (about one-fiftieth of a circle or just over seven degrees), Eratosthenes concluded that the circumference of the earth must be 250,000 stadia (5000 x 50).
This is the earliest record of a rational mind embracing the whole earth. Inevitably, Eratosthenes’ calculations were the approximate caresses of a lover who was unable to express the precision of his desire. He assumed that the earth was a perfect sphere, that Alexandria lay due north of Syene, and that Syene lay directly on the tropic (where the summer solstice sun is directly overhead at noon). It is impossible to say how close he came to the mathematical truth: the exact value of the stadion he used is unknown; the error was somewhere between two and thirty per cent of the actual figure. But the essential point was the application of theory to the cosmos: equipped with nothing but a stick, a distance measurement and simple geometry, Eratosthenes had established the basis of a world map.
19. Eratosthenes’ experiment
Ancient accounts usually attributed discoveries to heroic individuals when in fact they probably dawned in different places and at different times. When Eratosthenes scratched the cosmic truth into the wax of a writing tablet, several vast realities of terrestrial existence were already widely known or suspected: the world was not flat, and since the sun was very large and very distant, it was mechanically absurd to assume that it orbited the earth. In the 320s BC, Pytheas knew that latitude could be calculated from the shadow of a gnomon, from the length of the longest day, or from the rising and setting of a star. When he walked through Britain, he felt the curve of the earth beneath his feet, and when he sailed on the ocean, he knew that the swelling tides were in some manner caused by the moon.
The information was available, and so were the means of fashioning it into imaginable forms. By modern standards, it was a motley collection of data. In the Library of Alexandria, Eratosthenes would have found a few precise readings from gnomons and water-clocks, astronomical observations made by Babylonian astrologers, measurements of distance provided by armies, camel-trains, the mensores (land-surveyors) of Alexander the Great, and the betamists (professional walkers) and ‘rope-stretchers’ who re-surveyed Egypt to re-establish land boundaries after each major flooding of the Nile. Ships putting in at Alexandria were searched and any scrolls found on board were confiscated; copies were made and the originals placed in the library, marked ‘from the ships’. This wily acquisitions policy must have produced
a fine collection of sailors’ periploi, which listed river mouths and headlands, and gave distances in days of sailing.
Most of this geographical knowledge survives only as a muddle of rumour and misreporting in later Roman texts, but the early descriptions of the oikoumene (the inhabited earth) show that certain key coordinates had been identified – Rhodes, the Pillars of Hercules, Byzantion, Borysthenes, etc. When Pytheas sailed from Massalia, these points of reference were already being used to organize a conception of the earth that had once seemed a prerogative of the gods. Accurate depictions of the coastlines and continents were still many centuries away, but a semblance of the oikoumene could now be held within a human mind, thanks to one of the great inventions of the ancient world: the division of the terrestrial sphere into zones of latitude called klimata.
20. The oikoumene
The numbers on the left (hours and minutes) give the length of the longest day in 300 BC at that latitude. (The decimal degrees are those of the modern coordinate system.) The commonest intervals between klimata were half an hour; smaller divisions were also made. The latitude lines are derived primarily from Dicaearchus, Eratosthenes and Timosthenes. See also heren.
Determining latitude is fairly simple. Schoolchildren supplied with sticks and protractors regularly perform the exercise on sunny days. Determining longitude is far trickier and sometimes thought to have been beyond the capability of the ancient Greeks.
The erratic path of an ancient meridian or line of longitude can be seen towards the right of the map. The six places from Borysthenes (the mouth of the Dnieper) to Meroe (in the Sudan) were considered by Eratosthenes, for the sake of argument, to lie on the same north–south line. In reality, they occupy a time zone that covers almost six degrees of longitude: noon occurs at Rhodes twenty-two minutes later than at Meroe. For a sailor, the error would be catastrophic. The usual solution was ‘latitude sailing’: head north or south until the latitude of the destination is reached, then sail west or east until land is sighted. The same form of navigation was used by Columbus and Vasco da Gama almost two thousand years later. Once it became possible to calculate longitude as well as latitude, a ship could steer a direct, diagonal course along a rhumb line,19 holding to the same bearing throughout the voyage, instead of laboriously following two sides of the triangle. But with no landmarks, a surface that never stood still and no chance of repeating the experiment, it was impossible to determine longitude at sea with any degree of accuracy until the invention of a reliable marine chronometer in the late eighteenth century.
The inconveniences of ocean sailing greatly magnify the problem of longitude, which is not quite as intractable as it seems. On land, various forms of triangulation can be used to create a network of coordinates. The gridlines of Greek colonies on the Gaulish coast and parts of the Celtic system of Mediolana were, in a sense, accurate maps of miniature worlds. On a national or global scale, something more manageable was required, but according to the written record, not even a theoretical solution existed until about 150 BC. It was then that the Greek astronomer Hipparchos suggested that meridians east and west of a zero longitude (Rhodes) could be established by recording the local times at which a lunar eclipse was observed.
Here again, the sand in the hour-glass seems to flow in the wrong direction. If no means of determining longitude was devised before Hipparchos, how did Hannibal, for instance, manage to navigate his way along the Heraklean diagonal, and how was he able to rejoin the solar path after being forced to deviate from it? How could the pathways of the Celts have been anything but a wonderful mirage? A clue can be found, surprisingly, in Pliny the Elder’s credulous encyclopedia of wonders, Historia Naturalis (c. AD 78). In a chapter which is bizarre even by Pliny’s standards, he describes two ancient ‘experiments’, the purpose and result of which escaped him almost entirely. He knew only that they had something to do with the curvature of the earth, which ‘discovers and hides some things to some, and others to others’.
Someone had evidently tried to explain to Pliny the problem of longitude. The first example describes a fictitious phenomenon reminiscent of the puzzle with which brains befuddled by jetlag often have to wrestle: one of Alexander’s high-speed couriers could run 1200 stadia (about 200 kilometres) in nine hours, but only when he was heading west with the sun; the return journey took him six hours longer, despite the fact that it was downhill all the way. The other ‘experiment’ was just as confusing and equally impossible as described by Pliny. ‘High beacon-towers’ were erected by Hannibal in Africa and Spain. The beacons were lit at the sixth hour of day (noon). At the same moment, their light was seen in Asia at the third hour of night (9 pm).
This may be the only surviving indication of the kind of rapid, long-distance triangulations that enabled the Carthaginian general to follow the path of the sun. Hannibal and his astrologers performed this geodetic feat in 218 BC, perhaps with the aid of an early form of theodolite called the dioptra – a surveyor’s rod with sights at both ends through which the relative positions of the beacons could be calculated (see here). By then, the process may have been well established. A semi-instinctive form of rhumb-line sailing may have been adapted for use on land so that armies could navigate the continents like armadas.
There is even some unnoticed evidence of a longitude experiment further back in time, using astronomical rather than terrestrial measurements. In 331 BC, a lunar eclipse was observed at Arbela (Arbil, in Iraq), eleven days before the battle of Arbela at which Alexander defeated Darius III. The eclipse was recorded in at least three different places three thousand kilometres apart. All these places – Arbela, Syracuse and Carthage – lie in the same klima or latitude zone, as do two of the scientific capitals of the ancient world – Rhodes and Athens (fig. 20). Rhodes had recently become part of Alexander’s empire. Like Hannibal, Alexander had an urgent need for accurate maps and global positioning techniques, and perhaps it was at Rhodes that his geometers and surveyors coordinated the world’s first international scientific experiment.
Ultimately, the precision of these measurements depended on time-keeping. Most ancient records of eclipses give the times to the nearest hour or half hour. Since sunlight passes over the Mediterranean at about twenty-two kilometres a minute, the margin of error is enormous. Timing to the nearest minute was practically impossible, which is why ancient Greek and Latin have no word for ‘minute’. The material evidence is barely enough to fill a small cardboard box. Fragments of four water-clocks have survived from Egypt and Greece. They appear to have been capable of measuring time to within ten minutes in a twelve-hour period, though more refined readings were apparently attainable: some of the day lengths reported by Pliny include thirds, fifths and ninths of an hour, and even, in one case, a thirtieth (two minutes). This would still have produced a very erratic meridian. The great advance in time-keeping came much later, in eleventh-century Spain, when an Arab engineer invented a water-clock with epicyclic gearing (here). But historical chronologies are changing all the time. As we now know, the same technology was already performing its clockwork miracles in the second century BC. Someone on board a ship sailing west in the Aegean may have known exactly where he was before Poseidon reached up and confiscated the magical mechanism near the island of Antikythera.
The measuring and mapping of the world coincided with the first great age of European exploration. Discoverers of distant places were usually said to have been blown off course and to have been led by the gods to lands ruled by monsters or by women of unimaginable beauty and sexual appetite. But some of those supposedly accidental voyages were so long and successful that the adventurers must have been prepared, both materially and scientifically. When Eudoxus of Kyzikos set off to reach India by circumnavigating Africa in the late second century BC, he had a well-fed, motivated crew which included doctors, craftsmen and ‘flute girls’.
Not all those navigators arrived on uncultivated shores inhabited by Stone Age tribes. In the traditional European view of exp
loration, the bold adventurer is always more intelligent than the natives. While this was certainly true in the case of Hanno the Carthaginian, who reached equatorial West Africa in the early fifth century BC and captured three wild and hairy ‘women’ of the ‘Gorillai’ tribe, there is no reason to suppose that it was always the case. The people of Belerion on the south coast of England had been civilized by their contact with foreigners, and Pytheas was able to take his latitude readings throughout Britain without suffering the fate of some eighteenth-century French cartographers who were savagely attacked by suspicious natives. Pytheas’s measurements of the elevation of the mid-winter sun are revealing in more than one respect. In order to take his readings at locations several degrees of latitude apart, he would have had to remain in Britain for over a year, and quite possibly several years, if the modest British sun failed to show itself in late December. Perhaps he really did endure two British winters, or perhaps the land of astronomically aligned stone temples was able to provide him with the information.