Glastonbury Tor: What the Archaeology Actually Found

The hill, the tower and what is actually up there

The Tor rises about 158 metres above a flat, wet plain, and that is the whole of its drama: the Somerset Levels were marsh and seasonal flood for most of their history, so a conical hill with a tower on it reads from miles off like an island. The later identification with the Isle of Avalon leans on that silhouette. The island is real. The rest is argument.

What stands on the summit is not a pagan temple but the west tower of a Christian church. An earlier church there was destroyed by the earthquake of 11 September 1275, one of the few medieval British earthquakes with a firm date. A stone church of St Michael replaced it in the fourteenth century, and when that in turn fell out of use the tower was left standing. Its north-east corner was rebuilt in 1804; the rest is medieval fabric.

The dedication is worth a moment. Michael is the archangel of high places and of the fight against what lies below — the standard patron for hilltop churches across Europe. A tower to St Michael on a steep hill is not an anomaly requiring an esoteric explanation. It is the most conventional thing that could be there.

What Philip Rahtz found on the Tor

Between 1964 and 1966 the archaeologist Philip Rahtz excavated the summit, and his results are still the backbone of everything defensible that can be said about the hill. Apart from a scatter of prehistoric and Roman material, the first real occupation dates to the sixth century AD.

He found post-holes from timber buildings, two hearths including a metalworker's forge, and a great deal of animal bone — somebody up there was eating well. He also found sherds of imported amphorae from the eastern Mediterranean, the pottery that in south-west Britain signals long-distance trade in wine and oil around AD 450–550, plus a small hollow bronze head and, from later centuries, a wheel-headed cross of the tenth or eleventh. Two burials were oriented north–south, which makes a Christian reading of them unlikely.

A small monastic settlement followed in the late Saxon or early medieval period, with cells cut into the rock and eventually the stone church. So: a high-status or monastic site with real imported goods, in the right century for the setting of the Arthur stories, on a hill visible for twenty miles. A genuinely exciting result — and precisely not evidence of Arthur. Rahtz said so more bluntly than anyone, calling the use of archaeology to chase mythical figures historically misleading and trivial.

The terraces: three explanations, no verdict

Seven broad, roughly symmetrical terraces wrap around the sides of the Tor. They are the feature visitors ask about most, and they are honestly unresolved.

The mainstream reading is that they are strip lynchets: the stepped edges that build up over centuries of ploughing a slope, common across medieval England. The difficulty is that some run across gradients steep enough to make cultivation miserable, and some sit on the north face, the worst ground on the hill.

The second reading, proposed by Geoffrey Russell in 1968, is that they form a three-dimensional labyrinth — a Cretan-style maze wrapped around a cone, walked as a ritual ascent. The third, raised by the historian Ronald Hutton, is more modest and rather persuasive: a spiral processional path so pilgrims could reach the summit church, of a kind known elsewhere in medieval England.

The archaeology done on the terraces is slight, and none of it settles the question — a more interesting answer than either camp's certainty. You can walk them as a labyrinth without needing them to have been built as one.

The grave of Arthur, found twice

In 1184 fire destroyed Glastonbury Abbey. Rebuilding on that scale needed money, and money needed pilgrims. The Lady Chapel went up on the site of the old church between 1184 and 1189, and two years after it was finished the monks announced that they had found the grave of King Arthur and Queen Guinevere in their cemetery.

The earliest account comes from Gerald of Wales, writing in 1193. The monks, he reports, dug down to a hollowed log holding a tall man and a woman, and above it a lead cross naming the burial: here lies the renowned King Arthur, with Guinevere his second wife, in the Isle of Avalon. That last phrase is load-bearing. It is the moment Glastonbury becomes Avalon in writing. The bones went into the church, ending in a black marble tomb before the high altar.

The story has a twentieth-century sequel. Ralegh Radford excavated at the abbey from 1951 to 1964 and went looking for the grave on purpose, using the medieval descriptions to pick his spot near the Lady Chapel. In the early 1960s he announced that he had found it, dating the pit by chippings of Doulting stone in its fill, which he believed arrived only with the rebuilding of 1184–89. Both halves of that argument have failed. Doulting stone is now recognised as the abbey's principal building material in every phase, so the chippings date nothing — and beneath the pit lies a cist grave, a stone-lined burial, of the tenth or eleventh century. A sixth-century king cannot be buried on top of a tenth-century one.

Roberta Gilchrist, who led the ten-year project that reanalysed all thirty-six excavation seasons, put it plainly: merely a pit in a cemetery, datable only to somewhere between the eleventh and fifteenth centuries. Radford, she noted, may have overstated his evidence. It is tempting to file the whole business as fraud. The Reading team make a fairer point: medieval people did not draw the hard line between fact and story that we do. Filling a gap with what you believed ought to be true was not, to them, lying. It was restoration.

The director who dug by automatic writing

Which brings us to the most remarkable fact about Glastonbury archaeology, and the one most guidebooks skip. The first director of excavations at the abbey, appointed in 1908, was Frederick Bligh Bond (1864–1945) — a respected ecclesiastical architect from Bristol, co-author of a standard work on rood screens, and a Freemason, Theosophist and member of both the Society for Psychical Research and the Societas Rosicruciana in Anglia. He dug at Glastonbury for thirteen years. He also decided where to dig by asking the dead.

With a retired naval captain, John Allan Bartlett, writing as John Alleyne, Bond held sittings in which Bartlett's hand produced automatic script in a patchwork of low Latin and archaic English. It was signed by monks — Gulielmus Monachus, William the Monk, was the first — and by a collective calling themselves the Watchers. They described buildings that no longer stood, among them a great chapel east of the choir: the Edgar Chapel, with an apsidal east end.

Bond dug, and found the Edgar Chapel. He kept quiet about the method for a decade, then published the record in 1918 as The Gate of Remembrance. His employers in the Church of England, who disapproved of spiritualism, did not take it well: the Edgar Chapel was argued over in Parliament, and in 1921 Bishop Armitage Robinson dismissed him.

His fingerprints are still all over the town. The wrought-iron cover of the Chalice Well — the interlocking-circle vesica piscis that has become Glastonbury's unofficial emblem, reproduced on every second shopfront — was his design, given as a gift in 1919. The man the Church sacked for consulting spirits is the reason modern Glastonbury has a logo.

What Bond's own book admits

It is worth reading The Gate of Remembrance rather than reading about it, because the book undercuts its own legend twice, in Bond's own words. The first is the failure rate: in his preface Bond concedes that some of the script contained, as he puts it, very obvious misstatements — and names the worst himself. Asked for the Edgar Chapel's measurements, the script gave thirty virgae (yards) long and twenty wide; asked again, it said fifty; asked a third time, it returned to thirty. Bond ruled out the fifty as an obvious mistake and the twenty-yard width as, in his word, absurd. When a source must be asked three times and is wrong twice, the hits need a better explanation than the misses.

The second is the year 1907. Bond records that in the twelve months before the sittings began he and Bartlett worked systematically through essentially every surviving account of the abbey — William of Malmesbury, Adam of Domerham, John of Glastonbury, Leland, Dugdale, Stukeley and the later antiquaries. The sceptic Joseph McCabe wrote in 1920 that the two had steeped themselves for a year in the literature and lived for months in the medieval atmosphere. He meant it as a demolition. Bond had already written the same thing, as a credential.

That is the crux, and it requires nobody to have lied. An ecclesiastical architect who has spent a year absorbing every plan and description of a building, and who then lets his attention go slack, is well placed to guess where its foundations lie — without recognising the guess as his own. The Reverend H. J. Wilkins, reviewing the scripts in 1922, found absolutely nothing supermundane in the whole of the script; later writers point to the ideomotor effect, the mechanism behind the ouija board. The twist is that Bond half agreed. He never claimed to be chatting with ghosts — he proposed that the script drew on a vast store of subconscious memory reached by something like telepathy, and called his book the story of a psychological experiment. The sceptics' mechanism was not far from his own.

Avalon by design: 1929, 1935, 1969

Modern sacred Glastonbury was assembled in living memory, and the dates are known. In the 1920s the artist Katharine Maltwood was mapping the journeys in The High History of the Holy Grail when she became convinced that the beasts the knights fought were physical effigies in the Somerset landscape, outlined by rivers, roads and earthworks across a circle roughly ten miles wide. Her map appeared as a supplement to the 1929 edition of that text; her full account of a prehistoric Temple of the Stars followed in 1935, backed by commissioned aerial photography. No archaeologist has supported it. The figures combine features of wildly different dates, many of them modern field boundaries, and the same method applied to any English county produces a zodiac too.

In 1969 John Michell published The View Over Atlantis, which took Alfred Watkins' straight trackways, fused them with sacred geometry and a lost universal science, and put Glastonbury at the centre of the resulting map — including the St Michael line, the alignment of Michael dedications across southern England. More than any other book, it made the town the capital of the British New Age. The statistical objection is straightforward: a landscape as dense with churches, mounds and crossroads as England's generates alignments by chance, as Williamson and Bellamy set out in Ley Lines in Question (1983). We looked at the same problem in our piece on Wilhelm Teudt and the Externsteine, and in our guide to geomancy, ley lines and power places.

None of this makes the town a fake. It makes it a place where the layers are unusually legible: a real sixth-century site, a medieval abbey that marketed itself brilliantly, an Edwardian architect with a planchette, an artist with an ordnance survey map, a countercultural writer with a thesis. Most sacred landscapes have the same layers. Glastonbury kept the receipts.

Going anyway: how to visit without fooling yourself

The two springs at the foot of the hill are the best argument for going. The Chalice Well is chalybeate — genuinely iron-rich — and iron oxide stains its channels a deep rust red, which is why it has been read for centuries as blood, as the Grail, as the nails of the crucifixion. A few steps away the White Spring runs over calcium carbonate and leaves the opposite colour: two springs of different mineral content, side by side, doing something visibly strange. You need believe nothing for that to be worth standing in front of.

On Wearyall Hill there is the Holy Thorn, or its descendants: hawthorns that flower twice a year, in spring and again around midwinter — unusual enough to have carried a legend since the medieval period, the staff of Joseph of Arimathea thrust into the ground and taking root. A budded branch from the thorn in St John's churchyard is still sent to the reigning monarch at Christmas.

The practice that fits here fits anywhere. Walk up rather than driving as close as you can. Give it longer than you planned. Notice what the wind is doing at the top, because on that hill it always is. Do not arrive demanding an experience; a place is not obliged to perform, and the visitors who report the most are rarely the ones auditioning it. Our guide to the meaning of Archangel Michael covers the figure whose name is on the tower.

Frequently asked questions

Is King Arthur buried at Glastonbury?

There is no evidence for it. The claim dates from 1191, seven years after a fire left the abbey needing funds, and the earliest account was written two years later by Gerald of Wales. Reassessment of the excavation records shows the pit identified as the grave lies above a stone-lined burial of the tenth or eleventh century, so it cannot hold a sixth-century king.

What did archaeologists actually find on the Tor?

Rahtz's excavations of 1964–66 found sixth-century occupation: timber buildings, two hearths including a metalworker's forge, abundant animal bone, and sherds of amphorae imported from the eastern Mediterranean. Later phases include a small monastic settlement with rock-cut cells and the medieval church whose tower still stands.

Are the terraces on Glastonbury Tor a labyrinth?

Unproven. The most widely accepted explanation is that they are medieval strip lynchets formed by ploughing, though their steepness and the terraces on the north face complicate that. Geoffrey Russell proposed a three-dimensional labyrinth in 1968, and a medieval spiral processional path for pilgrims has also been suggested. The limited archaeology on the terraces does not settle it.

Did an archaeologist really excavate Glastonbury Abbey using automatic writing?

Yes. Frederick Bligh Bond directed the excavations from 1908 and used automatic script produced by John Allan Bartlett to choose where to dig, publishing the record in 1918 as The Gate of Remembrance. He was dismissed in 1921. Critics point out that Bond spent the whole of 1907 studying the abbey's documentary sources — a fact he reports himself — which offers an ordinary explanation for the hits.

Is the Glastonbury Festival held at Glastonbury Tor?

No. The festival is held at Worthy Farm near the village of Pilton, several miles east, and has no connection to the Tor, the abbey or the springs. Visitors looking for the sacred sites want the town of Glastonbury itself — a separate trip, in a much quieter week.