Wansdyke and Offa's Dyke

In various parts of England and Wales there are to be seen by the wayfarer long travelling earthworks consisting of ditched embankments which pertain to long-bygone ages of our storied past. Of these works the greatest in southern England is one, stretching for eighty miles from Ingpen Beacon in Berkshire to the Severn estuary, which has been known from early times as Woden's Dyke or Wansdyke. Another, even more impressive, extends more or less continuously for 140 miles across the Welsh Marches between the Dee and Severn estuaries, and has been known from the days of Alfred the Great as Offa's Dyke. The problem of their origin, intention and use is of high interest, and because of certain features which they possess in common, it will be convenient and instructive to deal with these two dykes together.

Wansdyke was formed by digging a ditch and throwing up the material derived from it into an embankment on its southern verge. It traverses a region of oolitic limestone and chalk, with rolling downs of no great elevation. There are places, as at Shepherd's Shore, between Marlborough and Devizes - overshadowed by the great Post-Office aerials on Morgan's Hill - where there is a slant of 30 feet from parapet to silted ditch. But for long stretches the present contour is a hardly perceptible hump, and notwithstanding centuries of denudation by weathering and of destruction by the plough it would be hard to envisage in many places an original contour much more substantial. Throughout its length there were great camps at Maesknoll, Stantonbury, Bathampton, Chisbury and elsewhere. These may have helped to defend it, whether they were planned for that purpose or lay already upon the course marked out by the dyke-makers. In this respect they may be compared with the outlying defences of Hadrian's Wall, which also was a ditched embankment.

The bank and its northward-facing ditch passes to the south of Bristol and Bath, crosses the Avon twice, utilises for fourteen miles the unbending Roman road from Bathford to Morgan's Hill, where it swings southward, while the earlier road pursues its even way towards Silchester and London. It is lost in the thickets of Savernake Forest, and is often weak just where ancient forests are likely to have been. One end of it rests on Portbury Flats at the coast by Avonmouth, as Offa's Tump rests on the red cliffs near Chepstow. The inland end is traceable no farther east than Ingpen Beacon, which owed its protection against outflanking movements to the marshy region in which it once was set.

Offa's Dyke traverses sinuously portions of four Welsh and three English counties, from Flintshire to Gloucestershire. At its mightiest it is poised high up the western slopes of hills or along the crests of ridges, the ditch always towards the west. Through stretches of twenty miles and more it maintains a normal elevation of 1,300 feet above sea-level, and at Llanvair Hill climbs over a crest 1,408 feet high. O'er moor and mountain it pursues its relentless way, to descend cross valleys by the shortest course, only to climb again as steeply on the other side. Nowhere does it traverse flat plain for any distance, except in the Severn valley near Welshpool. The embankment seldom exceeds 12 feet, and the ditch is rarely deeper than 7 feet, except where streams have scoured it to a greater depth. The finest intact trace is at Baker's Hill, between Llanymynech and Selattyn, where the ditch is 18 feet wide and 8 feet deep.

The dyke being necessarily made of local materials, the varied rock structure of the Marches, embracing shales, grits, sandstones, limestones and glacier clays, accounts for the diversity of its substance. These materials were never wrought into coursed masonry, as those of Hadrian's Wall were ultimately wrought; but were embanked loosely, as thrown up by pick and shovel. At the river crossings a solitary house is often found, and here and there tiny chapels stand just within the dyke, as at Craignant near the Berwyn Hills, or Mainstone in Clun Forest. These may recall stations established when the watch and ward of the dyke was active. At Hergen, on Selattyn Hill, and elsewhere, the ditch is protected by an outer wall of earth, and in the north some sections are ditched on both sides. There is no chain of supporting camps comparable with those on Wansdyke.

No record remains of the original contour of the work, and since it attracted the attention of John Leland in the sixteenth century no scientific plans had been made until those now in progress. The weathering of centuries, the destructive action of ploughman and roadmaker, and the burrowing of rabbits are responsible for incessant and irretrievable wastage. The dyke serves here and there as a motor road, and then may fade into a farm-track, be converted into pond or watercourse, vanish amid reedy swamps, be overgrown with a tangle of vegetation, survive as a switchback across a highway, or be seen only as a faint shadow on a field in the evening light. For five or six miles, all told, it serves in three places as a national boundary between the England and Wales of to-day.

In the vicinity of both these massive relics of a distant past lie lesser works of like construction. Eastward of Offa's Dyke at its northern end lies an embankment known from an early time as Wat's Dyke. For a space of forty miles this is roughly parallel to Offa's, with a distance ranging from 500 yards to three miles between them. Wat's Dyke has often been confused with the greater work, in such wise as to be called Clawdd Offa in local tradition, not necessarily as though part of it but as having equal right to the name. So also Wansdyke is to be studied in conjunction with Bokerly Dyke, which is but four miles long, is in part the boundary between Wilts and Dorset, is fashioned in the same way, and, like it, presents its ditched face towards the north.

Although much has been written about these great earthworks, it is only of late that their scientific study has been undertaken. Some years ago the late General Pitt-Rivers made actual excavations on Bokerly and at the Wiltshire end of Wansdyke. He was able to dismiss for ever the speculations of some of the older antiquaries, which had remained untested for generations. Over a period of years the late Albany Major traversed Wansdyke from end to end, and after the war made further surveys in conjunction with Mr. E. J. Burrow, which have been fully published. When these have been supplemented by indispensable field-work, especially in Somersetshire, some puzzling questions at present unsolved may be easier to answer.

In 1925 a campaign for the careful survey of Offa's Dyke was inaugurated by Dr. Cyril Fox, who had already excavated the Devil's Dyke in Cambridgeshire, once part of the boundary between East Anglia and Mercia. This survey is planned to cover several years, and provides for actual digging at suitable places in order to reveal the method of construction and, if possible, to recover contemporary objects. The results already attained enable some useful conclusions to be drawn.

What, then, of the origin and intention of these majestic monuments? In the case of the Welsh embankment tradition and recorded history supply answers not only definite in themselves but also supported by modern research. Offa, king of Mercia, is referred to in the story of Saxon England. His reign was marked by warfare with the western Britons, culminating in their subjugation about a.d. 779. Within a century of his death Asser, the Welsh biographer of Alfred, wrote of Offa, " this strenuous king... ordered to be made between Britain and Mercia a great dyke from sea to sea."

Attempts have been made to claim for that work, or parts of it, an origin dating back to the Roman occupation. Let it suffice to say that where it passes through a Roman settlement, as at the Ffrith, the Roman objects unearthed by the modern digger were clearly derived from old deposits scattered by the dyke-makers. Indeed, such a work would have served no useful purpose to the Roman legions, which depended for the strategic control of that region rather on the great road system between Caerleon, Wroxeter and Chester.

It may be reasonably held that the work belongs to one period of construction, and was designed to define such a boundary as history attributes to the Mercian king. Sometimes it utilised as sighting-marks structures which represented the abodes and sepulchres of older populations, as when it led up to an encircled Bronze-age barrow and emerged on the other side. The political boundary which it created was maintained by drastic sanctions. In the twelfth century a Welsh writer referred to the dyke in relation to his countrymen, "crossing or passing which they were made to atone and to mourn the loss of a foot."

Here and there it doubtless served at times as a military defence. Troops could be moved rapidly up the gentle ramp on the Mercian side, and the flattened crest of wall may in times of stress have had a timber palisade thrown up on its western edge. Wat's Dyke, perhaps enshrining the name of a Mercian general officer, may have been the first portion to be built, and may never have been supplemented by another along the gap of nineteen miles between Offa's extant terminal and the Flintshire coast.

The problem of Wansdyke is not so ready for solution. The early writers are silent. Its structural form does not bear the indispensable stamp of one master mind, nor does it seem to have been set up as a continuous political boundary. It suggests a primary purpose of defence, and may prove to be a chain of regional lines of varying periods, the western portions of which were strung together in a time of stress, and the eastern portions similarly, but not necessarily under the same conditions. The linkage of the two sections along the disused Roman road may mark a final stage. General Pitt-Rivers showed that the eastern section did not precede the advent of the Roman legions, and casual excavations farther west support these conclusions.

Working backwards, if we accept the eleventh century statement of William of Malmesbury - the earliest literary mention of the name, though not of the work - that a battle fought at Woddesbeorg in 592 took place at Woden's Dyke, we antedate Offa's by at least two centuries. Attempts to attribute the work to one of the early followers of Cerdic, who landed in Wessex a century earlier still, lack archaeological support, and may be met by an appeal to the -dyke's very name. Were the work before us of English workmanship it would probably bear a personal name, as Offa's and Wat's do, and not the name of a deity, which was more likely to be a folk-name attributing superhuman origin to a work whose very builders were forgotten.

Attention is thus directed to the period of disturbance when the Roman occupation was drawing to its close. It is tempting to recall the fact, as one antiquary has recently done, that when Caledonian hordes were breaking down the northern walls and ravaging central Britain down to the Thames, the Romano-British population, as the legions were withdrawing, set up defensive walls against them. Where are these walls, and was any part of Wansdyke one of them? All such speculations are best left for further excavation on the spot.

Happily the importance of these works is at length receiving recognition. Large sections of them have already been scheduled by the Office of Works under the Ancient Monuments Protection Acts. Ingpen at the eastern end of Wansdyke, the section between Chisbury Camp and the Bedwyn road, and Woodyates on Bokerly are now out of danger.

The scheduling of Offa's Dyke has already secured Ysceifiog Circle and its Bronze-age barrow, 530 yards at Coed Talon Banks, and 130 yards near Hope, in Flintshire, together with 370 yards at Brymbo Hill, 100 yards at Cae Llewelyn, the section at Ffrith Hall, 240 yards at Pen y Coed and 380 yards at Vron Farm in Denbighshire. The scheduling of other sections is proceeding, under the direction of the experienced archaeologists to whom is entrusted the responsible task of keeping a careful watch and ward upon our national heritage.