Glastonbury Tor Maze - click to see the book

The Glastonbury Tor Maze

Geoffrey Ashe

Author of Traveller's Guide to Arthurian Britain

Click here to order this book

The Theory

Visitors to Glastonbury often ask whether the Tor is artificial. The answer is that it is a natural hill, but one that shows signs of having been artificially shaped. Along its sides are a number of terraces, one above another. From the upper part of Well House Lane, several of them can be seen running along the north face. They are worn and weathered, but traceable over long stretches.

The Tor, with Chalice Hill and the town in front

According to a theory put forward by Geoffrey Russell, they are the principal remains of a maze: not in the sense of a puzzle, but in the sense of a long, twisting, devious approach to a centre - a labyrinth. Made in the remote past for ritual purposes, it spirals round the Tor seven times, and ends - or may be supposed to end - at the summit where the tower now stands.

It is argued further that the spiral is not a simple one, but a three-dimensional adaptation of a more complex pattern which is found in antiquity and in widely separated places. This may be called the Cretan spiral, because it appears on Cretan coins and has some connection with the Labyrinth legend.

But the same maze or its mirror-image also occurs on an Etruscan vase assigned to the seventh century B.C., on a pillar at Pompeii, and on rocks at Tintagel and in Co. Wicklow. It is known in Wales in the shape of rustic mazes called Caerdroia, 'Troy Town'. It is even found among the Hopi of Arizona, as the 'Mother Earth' symbol. Philip Rahtz, who excavated the top of the Tor in 1964-6, said of Russell's idea: 'The argument is complex, but it is worth consideration'. And elsewhere: 'If the maze theory were demonstrated to be true, it would clearly be of the greatest relevance to the origins of Glastonbury as a religious centre'.

Cretan Maze

The theory requires that there should once have been seven paths going completely round the Tor, all running along continuous terraces, with vertical connections between them. Weathering, trampling, and shiftings of soil and strata have made parts of this hypothetical scheme a matter of conjecture. Yet terraces can indeed be distinguished at seven different levels, and while they are not now continuous, they are more nearly so than a glance might suggest. Effects of light and shadow, variations at ground level, make it difficult to take in the whole system at any one time or from any one angle. Sometimes a path is hard to recognise when one is on it, yet easily visible from a distance. Sometimes a terrace is almost indiscernible from a distance, yet well defined when looked at from directly above or below.

The north end of the Tor

Scrutiny in a combination of ways shows how much of the system is, arguably, there. Furthermore, even where a presumed maze-path has vanished, there is usually at least a small trodden track preserving its course. The whole maze can be reconstructed in such a way that a person threading it is hardly ever simply improvising a route through featureless grass. There is nearly always a visible way of some sort. The few total gaps are either at the lowest levels where farming activity has altered the surface, or at the highest where the effects of steepness might be expected to do likewise. These facts do not prove the maze's reality, but they go some distance towards proving its credibility.

To account for the terraces without it, four ideas have been advanced. They may be merely tracks worn by cattle. They may be remnants of a spiral route of some sort which, however, was made for a more mundane purpose such as getting horse-drawn carts up the Ton They may be early agricultural workings - strip-lynchets, shelves for vineyards. They may even be natural: freak results of the erosion of strata of differing hardness.

The first suggestion seems unlikely. Known cattle-tracks on parts of the Tor itself, and on Chalice Hill opposite, are far narrower, closer together, and more numerous. The notion of a track for carts is disproved by the actual course of the terraces: the transitions from level to level are too steep for the imagined ascent. With the remaining suggestions, the only useful procedure is to see whether the system has features which the maze theory accounts for and the others do not.

For discussion, some terms need to be agreed on. The seven paths - sometimes with well-defined strips of terrace to run along, sometimes not - may be numbered as Paths I, II, III, IV, V, VI, VII, starting from the lowest and outermost which goes round the base, and ending with the one nearest the summit. They are at right angles to the two principal routes up the Tor - or to be more precise, they would be if the Tor were circular like the Cretan maze, but its elongation causes distortion. One of these routes is reached from Well House Lane near the Chilkwell Street end. After a steep stretch of surfaced path between hedges, a climber enters the Tor field and passes up through it, to a metal gate or stile in the Tor's boundary fence; then, just above, starts the ascent of the Tor proper near a bench. This may be called Ascent Route A.

The climb is long, but a good deal of it is gradual, along a ridge. The other upward route is at the far end of the Tor. Again the climber goes through a field and boundary gate, and then begins the ascent, which is shorter and steeper. This may be called Ascent Route B. Neither A nor B is a single, clear-cut, narrow way like a flight of steps. Both are fairly broad corridors within which the line of ascent can be varied.

Copyright Sig Lonegren

From Sig Lonegren's book Labyrinths

A walker who turns left off Ascent Route A, to circle the Tor clockwise by any of the seven paths, passes along the north face towards Ascent Route B. From Ascent Route B a similar progress can be made along the south face, back to Ascent Route A. It should be added that the Tor belongs to the National Trust, and that the boundary fence round the base, along most of its length, separates the Tor itself from fields and other properties to which the public has no access.

Who, When, Why?

Russell proposed to connect the Tor and its maze with early Welsh poetic allusions to Caer Sidi, the 'turning' or spiral castle. This was a place in pre-Christian mythology which housed a magic cauldron. Caer Sidi was a point of contact with Annwn, a Celtic Otherworld sometimes pictured as underground. Its wonder-working vessel may have been the same as a cauldron of inspiration that belonged to the goddess Ceridwen.

Certainly the legends of Glastonbury link up with these themes. The cauldron, in one guise or another, is a factor in the making of the Grail story, and a very early Welsh poem tells how Arthur and his men went in quest of it. A tale about a visiting saint, Collen, shows that the Tor was regarded as an entrance to Annwn. Russell suggests that the Quest of the Grail has pre-Christian roots in a Celtic ritual which involved threading the Tor maze to the summit and, presumably, attaining a real or symbolic sacred vessel of otherworldly character.

However, while British Celts of the pagan Iron Age doubtless had notions about the Tor, and might even have made some use of its maze, they are unlikely to have done the original work. According to Philip Rahtz a probable date would lie in the second or third millennium B.C. As a religious structure comparable to Silbury, the maze, if confirmed, could have a bearing on the debate as to how far Neolithic religion was centred on the cult of a Great Goddess or Earth Mother. It has been pointed out that the Glastonbury hill-profile, viewed from a certain angle, evokes a recumbent female figure with the Tor forming the left breast. Early images of goddesses do occasionally have lines circling and meandering on their bodies.

The special Cretan maze-pattern in its complexity and wide distribution is an unsolved problem, though a ritual origin is agreed to be likely. Another relevant theme is that of the holy mountain. Several early Asian mythologies refer to a mountain where the gods dwelt. Hindus called it Mem, placed it at the world's centre, and sometimes - interestingly - made it septenary, with seven sides or levels. In Babylon the great Ziggurat or temple had seven tiers, and may have been intended as a model of the mount of the gods. The motif passed into Islamic legend, also into the poetry of Dante, who portrays Purgatory as a seven-tiered mountain. The ingredients for a combination of holy mountain and septenary maze certainly existed from ancient times. How or why such a grandiose work might have come to be undertaken in Britain is a further question.

Should the maze be generally accepted, this will have one notable result even if the background remains obscure. For the first time it will become feasible to reconstruct and re-live a Neolithic ritual, at least to some extent. We do not know what was actually done at Avebury or Stonehenge. But if people went to immense trouble to make a maze on the Tor, then we can be sure that whatever else they did, they threaded it. So can we.

   Glastonbury Tor