Metropolitan Museum of Art

‘Heracles Driving Off the Cattle of Geryon’; engraving by Giulio Bonasone, sixteenth century

Middle Earth in a book title would suggest to most readers that it was about J.R.R. Tolkien and The Lord of the Rings, and it is likely that the publisher of Graham Robb’s book had an eye on this market. Robb has taken the phrase over to describe the European realm of the ancient Celts, the group of tribal societies that arrived in France from the Upper Danube beginning around 600 BC. But the Celts had no Middle Earth. The name, which was Germanic in origin, had reached Tolkien through his work on Old and Middle English, and its diffusion to a broad public in modern times was entirely the result of his own capacious imagination.

What the Celts did have, as Robb emphasizes, were dozens of cities called Mediolanum, a name best known as the Latin for Milan in Italy but widely spread across the Celtic territories of Gaul that lay to the north. Both the popularity and the etymology of Mediolanum are shrouded in mystery. The first half of the name, understood as Latin, means “middle” or perhaps “half,” whereas the second part, on the face of it, would appear to mean “wool.” But Middle Wool or even Half Wool is certainly not Middle Earth. There is reason to think that the Celts may have understood this name as meaning “middle enclosure,” but again nothing about Middle Earth. Yet all these Mediolana, carefully positioned in the Celtic territories of Roman antiquity, seem to conjure up a secret and sacred geography. Graham Robb has now found a new and unexpected way to discover what might have been going on.

Hitherto Robb has enjoyed a solid reputation for his work on French literature (Victor Hugo, Honoré de Balzac, and Arthur Rimbaud) and on the culture and history of France. Although Celtic Gaul was certainly a part of the evolution of France from antiquity onward, Robb’s new book is a bold departure from what he has done before. As he explains in his introduction, it all arose from plans for a bicycle trip to explore what was traditionally known as the Heraclean Road, the Via Heraclea. This was a legendary route that Heracles took from the southwestern part of the Iberian peninsula into continental Europe and across the Alps into Italy. One might well wonder how anyone could plan a cycling expedition over a mythical route that lacks any substantial literary or archaeological documentation, but Robb believes that he has been able to do exactly that. His book is a dense and breathless account of what he has discovered about his hero’s mythical itinerary. The thrill of his discovery sometimes overwhelms the clarity of his exposition, but author and reader alike share an exhilarating exploration of what is commonly called sacred geography.

In Greek mythology Heracles was very special. His mother was a mortal, Alcmene, whom Zeus had visited on one of his extramarital escapades, and so he began his tumultuous life as a hero rather than a god. Amphitryon, Alcmene’s husband, seemed not to mind, but Hera, who was not only Zeus’ wife but the deity whose fame (kleos) appeared to be enshrined in Heracles’ very name, conceived a fierce antipathy to her husband’s child. She sent large snakes to kill him, whereupon the baby Heracles promptly demonstrated the physical prowess for which he would later become famous by strangling the snakes as they coiled around him.

This was only the beginning of his legendary exertions. He was subsequently ordered to undertake twelve great labors, including clearing out the Augean Stables, subduing a lion, boar, and bull, and finding the apples of the Hesperides. He finally succumbed to a fiery death on a funeral pyre atop Mount Oeta in central Greece north of Parnassus. Once dead, Heracles became a god, after which his reputation continued to grow steadily. He became the chief rival to Dionysus in the Roman Empire. Both gods maintained their competitive edge even as the empire grew increasingly Christian.

Heracles, whom the Romans knew as Hercules, served to inspire both emperors and nations. The emperor Commodus, who reigned from 180 to 192 AD, represented himself as a reincarnation of Heracles, complete with lionskin and club, and Heracles was believed to be the founder of the Celtic nation through a union with a certain Celtinê somewhere in France. The grandeur and diversity of this hero-god are reflected in the traditions of many places throughout the ancient Mediterranean world. He traveled widely to carry out his twelve labors. For the tenth, which was one of his most remote geographically, he went to the Straits of Gibraltar, known henceforth as the Pillars of Hercules. His task was to kill the three-headed giant Geryon, who lived on an island called Red (Erytheia), in the vicinity of ancient Gades (the modern port of Cádiz in southwest Spain). After slaying Geryon, Heracles was charged with driving Geryon’s cattle all the way across Europe into Italy. That meant an overland journey out of Spain, over the Pyrenees, through France, across the Alps, and down to Rome.


The imagined route that Heracles took for this prodigious cattle drive became known later as the Via Heraclea, Via Herculia, or Via Herculis, and no one has been able to say precisely where it went. The name of this route first appears in a treatise probably from the later fourth century BC and ascribed erroneously to Aristotle. Traces of the Italian route have been identified in the Apennine Mountains—which run from the north to the south of Italy—and where it is now duly recorded in the Barrington Atlas, which is the standard modern atlas of the classical world, but the European route has for the most part been enveloped in mythical mist.

In the sixth century BC a Greek poet, Stesichorus, wrote an epic work called Gêryonêis on Heracles’ confrontation with Geryon and the abduction of his cattle, but unfortunately only fragments survive. Nonetheless, there were enough of them to inspire the brilliant poet Anne Carson, with her professional knowledge of the classics, to create a superb verse novel, Autobiography of Red, over a decade ago. In the surviving fragments of a lost tragedy, Aeschylus had brought on the titan Prometheus to advise Heracles before his departure for Europe and provide helpful predictions of what the hero would find on the way.

Graham Robb’s new book equally owes its inspiration to the accounts of Heracles’ mission to the West. Although it is not a work of poetry, it is very much a work of imagination, tempered by astronomical and mathematical calculations as well as computer technology. Whether or not Robb has proven his point, and many will be skeptical, the exploration he describes is consistently exciting, leaving the reader poised between wonder and disbelief.

Robb outlines a geographical pattern of settlement and urbanization in ancient Gaul and Britain that he ascribes to the Celts, and the Druids in particular. It is distributed along a grid that is based on solstice lines—the lines of the sun’s rays at sunrise on the longest and shortest days in the year, which are known as the summer and winter solstice. No one doubts that the Celts were capable of plotting solstice lines, just as no one doubts that their predecessors at Stonehenge could do the same thing. But the detail that Robb brings to his exposition is unusual in reflecting his use of computerized mapping to generate hitherto unrecognized patterns.

Without computers but with comparable learning and enthusiasm William Stukeley had done something similar when he published in 1740 his Stone- henge, a Temple Restor’d to the British Druids. Stukeley’s date for Stonehenge was wrong by several millennia—it is now dated between 3000 and 2000 BC, when there were no Druids. Still, as a scholar has written recently, he “became obsessive about the role of Druids at Stonehenge,” while his views nonetheless seem “almost credible when compared with some of the really fanciful interpretations of the site.”1

Robb is, in many ways, the Stukeley of our age. He is no less obsessed with the Druids and their mathematical skills, and fortunately, unlike Stukeley, he is writing about a time when there really were Druids in Europe. What he proposes seems almost credible at times, and his concentration on the Via Heraclea as the key to his argument is not unlike Stukeley’s concentration on Stonehenge. Sacred geography is definitely not a mirage; nor are solstice lines. The problem lies in relating these lines to points that are fixed by either historical data or archaeological remains.

Moving from myth to documented history on the ground is a treacherous business, although stories about the irrecoverable past undoubtedly had, and continue to have, an impact on the historical past and present. The Via Heraclea might have been worth trying out for a cycling expedition, but Robb’s assumption that the original path indicated in ancient texts could be recovered “if the surviving sections are projected in both directions” is highly questionable. We simply do not have the necessary sections of his imagined itinerary, and detecting an original path for a hero whose mythical exploits antedated the Druids in Europe is well-nigh impossible.

As recently as the 1960s Heracles’ mythic route was generally believed, on the basis of the text in pseudo-Aristotle, to have passed along the south coast of France between Marseille and Monaco, whence the feisty hero would have driven Geryon’s stolen cattle down into Italy. Prometheus’ speech to Heracles in Aeschylus’ lost play seemed to bear this out. He foretold that the hero would pass through the land of the Ligurians, who were Celts in southern France, and that he would overcome their forces through a hail of stones sent by Zeus to help him. This seemed to most readers a clear allusion to the rocky plain of La Crau just north of the coastal road in present-day Provence that Heracles was thought to have been taking.


But in 1962 a French historian and geographer, Roger Dion, argued that the Via Heraclea belonged farther north and crossed the Alps precisely at the pass of Mont Genèvre on what is now the Franco-Italian border in the French Alps.2 His argument was not based on any archaeological evidence but solely on the historical testimony for Hannibal’s invasion of Italy. Although Dion convinced some historians but by no means the majority, his work is fundamental for Robb’s mapping. Hannibal had apparently seen himself as a new Heracles, not least because of the Carthaginians’ identification of the Punic god Melqart of Tyre with the Greek hero-god, as commemorated in cults at the Phoenician settlement at Gades (Cádiz) as well as on the Sacred Promontory at Cape St. Vincent (now in southern Portugal).


Graham Robb/Norton

A map from The Discovery of Middle Earth showing the route that Graham Robb ­proposes for Heracles’ mythic journey from southwestern Spain across the Alps to Italy. Cape St. Vincent, referred to in the review, is at the Sacred Promontory, and Mont Genèvre is at the Matrona Pass.

In 218 BC Hannibal took his army north from New Carthage—today’s Cartagena in southeastern Spain—toward the Pyrenees. After crossing the mountains into Gaul, he was on his way toward the Alps, with his army and elephants, when, according to the stories told by Celts in the region of the Rhône, a hero appeared to him and showed him where to go. This epiphany has been understood to refer to Heracles, although our source, the historian Polybius, does not name him. As a pro-Roman Greek, Polybius would certainly have been reluctant to identify a famous Greek hero as Hannibal’s patron deity, even if the story had its roots in Hannibal’s attachment to Heracles’ Punic avatar, Melqart. Robb wants Heracles to begin his journey from the Sacred Promontory at Cape St. Vincent and proceed in a straight line to Mont Genèvre before descending into Italy (see the map from Robb’s book above). This is what he declares to be the “original, mythic incarnation” of the road. Unfortunately this will not work.


Berthold Steinhilber/Laif/Redux

The Bibracte basin, excavated in the Burgundy region of France in 1987. ‘The oval shape of the basin is truly extraordinary,’ G.W. Bowersock writes. ‘But whether it is religious remains to be demonstrated.’

According to the myths available to us, Heracles found Geryon tending his animals on the island of Erytheia, in or near Cádiz, though we have only a few details of his movements after he killed the three-headed giant and stole his cattle. But the details we have are incompatible with Robb’s hypothesis that Heracles followed a Celtic solstice line from Cape St. Vincent to Mont Genèvre. He is described as setting forth from the area of Cádiz, and from several sources we can monitor what is said to have happened after that. A large fragment of an inelegant but instructive geographical poem in Latin iambics by Avienus in the fourth century AD, the Ora Maritima (Maritime Shore), describes the coast between Marseille and the Gulf of Cádiz, and Avienus clearly had access to earlier accounts of the topography and history of this coastline.

Avienus has much to say about the Pillars of Hercules at Gibraltar. He traces what he explicitly identifies as the Via Herculis from its beginning in Cádiz and from there over to the Sacred Promontory at Cape St. Vincent. He leaves no doubt that the route passed from one place to the other across a great causeway, which he says the Greeks called a herma, or rocky reef. He reports that this was believed to be Hercules’ road, “because Hercules is said to have paved the sea [stravisse quippe maria fertur Hercules] to open up an easy route for his captive flock.” An earlier ancient geographer had already mentioned the reef as lying just beneath the surface of the water. It was therefore by means of this rocky causeway that Heracles was supposed to have reached the Sacred Promontory at Cape St. Vincent. Nothing further can be said with certainty about the path to Mont Genèvre, if that is where he went, but the route out of Cádiz was an integral part of the Via Heraclea. Starting it at the Sacred Promontory suits Robb’s design, but at the expense of the only testimony we possess.

All this means that there are only two or three identifiable points on Heracles’ alleged route: Cádiz, where he started his cattle drive after killing Geryon; Cape St. Vincent, to which Heracles passed with his cattle on a mythical causeway he built; and possibly the Col de Mont Genèvre, if Hannibal actually crossed the Alps by that pass. No other points along the road can be confirmed in the mythic literature, either traditionally or historically. Cults of Heracles appear in widely scattered places, but they do not necessarily point to stations along the road. If they did, the cult of Heracles at Sagunto would bring Heracles all the way over to the east coast of Spain, even though there is not the slightest reason to believe he went there. The cult of Heracles at Nîmes, ancient Nemausus, may, but need not, indicate a stop on the hero’s path.

Robb’s desire to begin the Via Heraclea at Cape St. Vincent, from which he could draw a diagonal line straight across to Mont Genèvre, allows him to construct a Celtic network of solstitial lines with reference to it. But if the Celts chose to ignore Cádiz as the start of the route, they could only have done so by also ignoring the earliest reports of Heracles’ journey. There would be no basis for considering it “original.” Furthermore, we cannot be sure where Hannibal crossed the Alps, and so we are even less sure about where Heracles supposedly did.

The dilemma that confronts anyone trying to make sense of this route emerges repeatedly in one of the few recent studies devoted to this subject. Robert Knapp observes that the archaeology of the Via Heraclea is even more difficult than the analysis of its literary sources: “The Via Heraclea is a concept, an abstraction, not a real route that was laid out and well made…. Archaeological remains of the route itself are not known anywhere. Equally other references to the via are scarce.”3 Under these circumstances it is extremely risky to undertake computerized mapping. Robb’s encounter with the mythological tradition sounds rather like an ecstatic vision: “I began to see, as though in some miraculously preserved document, the ancient birth of modern Europe…. The geography of the Western world had been organized into a grid of ‘solstice lines,’ based on the original Via Heraklea.”

If there was any original road, we can say nothing about it. The Celts may have heard reports that Heracles had seduced Pyrene, the mythical figure whose name was given to the Pyrenees, or that he fathered many children who gave their names to regions in Europe. Timagenes of Alexandria, in the late first century BC, claimed to have learned about Heracles’ paternity from the Druids. Not long before, the Romans had heard about those learned Celtic priests from Julius Caesar, who had written at length about them after encountering them during his Gallic wars. But in the Greco-Roman world before that time the Celts had been known chiefly as invaders from the north. They descended into Rome in 390 BC and into Delphi in 279. Their movements about that time brought them as far as the territory of modern Turkey, where they settled under the name of Galatians.

It was in Gaul that the Druids, the members of the priestly class in Celtic Europe, acquired their reputation for being erudite, a reputation famously maintained later in Britain. The solstice grids that Robb identifies as having been conceived in both countries may well reflect a sophisticated knowledge of mathematics. If Robb’s chapters “The Druidic Syllabus, I: Elementary,” and “II: Advanced” exaggerate in implying that a modern academic course on Druidic science cannot be far away, one of his prime exhibits is incontestably spectacular. It is a large oval water basin in pink granite, some thirty-four by twelve feet—excavated in 1987 at Bibracte on Mont Beuvray, west of Autun in Burgundy. This object, oriented on the solstice line at Bibracte, is too shallow to have been a cistern. There was no adjacent source of running water, and Robb postulates, a little incautiously, some kind of sacral character.

But the oval shape of the basin is truly extraordinary. It was generated by two circles overlapping at one fifth of their diameter with their intersection points along the solstice line. This means that lines drawn from the oval’s top and bottom to the center of each circle formed, together with the solstice line running through the oval, two Pythagorean triangles, one on each side of that line. The excavators who published this basin recognized the mathematical precision of their bizarre discovery.

This moved Robb to hyperbolic enthusiasm about the oval’s esoteric use in iconography for the mandorla—the almond-shaped area of light surrounding the resurrected Christ and other sacred figures—as well as the vulva. “Though it hardly bears comparison with the Acropolis or St. Peter’s Basilica,” as if anyone would wish to make such a comparison, “the pink-granite basin of Bibracte is one of the most remarkable religious monuments of Europe.” It is undeniably remarkable, but whether it is religious remains to be demonstrated.

Robb ascribes to Druidic science the placement of the many cities called Mediolanum, and he found an ally in this interpretation in the French geographer Xavier de Planhol, who became convinced that each Mediolanum was situated equidistant from two others. If there are insufficient points along the route of Heracles to show where it went, the cities called Mediolanum are equally frustrating in a different way because there are so many. Robb gives their number as both sixty and thirty-six, and he explains this inconsistency by a problem that will be familiar to anyone who studies place-names (toponymy): “a dream in which familiar faces turn out to belong to strangers.” He is alluding to a process of extrapolation back from a modern toponym to a presumed ancient name, but as he says, despite Mediolanum as the name of Italian Milan, “most of the other places called Milan have an entirely different origin.”

Even so, the number of multiple Mediolana is impressive, and it is hard to resist the notion that they were part of some kind of solstice network. Caesar reports that the Druids met once a year at a sacred place in the territory of the Celtic people called the Carnutes, which was thought to be the middle (media) region of all Gaul. This is undoubtedly the area of modern Orléans, although it would scarcely be the middle of a network that covered the whole of France. If the Druids had been such accomplished mathematicians, they would have known this. There is no way of saying much about what the Druids did at their annual meeting, but Caesar reports that Celts came from far and wide to have their disputes heard and adjudicated by these wise men.

The Druids were a privileged and learned class, exempt from tax and military service, and schooled for up to twenty years. According to Caesar, they studied astronomy, theology, the natural sciences, and “the magnitude of the world and its lands.” It is reasonable to assume that they learned to observe solstices and lay out solstice lines, but whether they used this knowledge to construct complex grids is another matter. If they really believed that Orléans was in the middle of all Gaul, something was wrong. On the most indulgent interpretation, the sacrality of the place where the Druids held their annual meeting was more important than its centrality.

An embarrassing reputation of the Druids for human sacrifice ultimately overcame any Roman respect for their scientific prowess, and in the reign of Tiberius they were obliged to take their erudition and their sacrifices across the Channel to Britain. Robb says candidly that he had intended to end his book with that dramatic exodus, but picking up on Caesar’s hint that some Druids had already made their way to Britain previously, he opens a long exposition of presumed Celtic roads and grids in Britain. This naturally takes him far away from the bicycle trip with which he began, and Heracles quickly fades into the background.

But curiously, only a few generations before the Druids returned to Britain under Tiberius, Heracles had made a conspicuous comeback at Rome through the ambitions of its first emperor, Augustus. This would have made a fitting conclusion for Robb, who starts with the Via Heraclea. After all, Heracles’ legendary journey from the Iberian peninsula ended with his depositing Geryon’s cattle in Rome, where Augustus founded what we know as the Principate (or Empire) after the Roman civil wars. In his early years he launched a war against rebel holdouts in Spain, and this took him precisely into the northern part of the Iberian peninsula, where Heracles would have passed on his way to the Pyrenees. Augustus led a strenuous campaign in 26 BC against the people in Cantabria, in northern Spain. Despite an illness he supervised the continuation of the war through the following year, when he declared victory. That victory proved to be premature and had to be reclaimed six years later by another general on his behalf, but Augustus’ return from Spain in 25 BC was glorious. He was welcomed as a new Heracles. In an ode from that time the poet Horace hailed the emperor “like Heracles…he came back a victor from the Spanish shore.” In another ode Horace envisioned a pacified Spaniard, “mindful of the great Heracles,” cheerfully praying to the new emperor.

It is impossible to infer anything from Augustus’ operations in Spain about where Heracles might have gone. Besides, Augustus did not cross the Pyrenees, although he did build a road of his own in the south. But he boasted, in an autobiographical inscription, that he had pacified “the Gallic provinces and Germany, surrounded by Ocean, from Gades to the mouth of the Elbe.” This reference to Gades may well allude to the place from which Heracles began his legendary journey across Europe.

Yet the mythical path that Heracles took from there remains as mysterious and seductive as ever. It can hardly be said to lead us into Middle Earth, but it does take us into a bewildering world of Druids, solstice lines, and grids, with all the unexpected information that the Celts knew how to provide for foreigners. Julius Caesar even heard, and apparently believed, a report, worthy of Tolkien, about elks that slept while leaning against a tree because their legs lacked any joints. Robb may not have discovered Middle Earth, nor is his narrative as enchanting as The Lord of the Rings, but it is full of surprises that reflect the unflagging curiosity and fertile imagination of its author.