Wheelhouses

The aisled wheelhouse at Middlequarter, Sollas, North Uist was inserted into a sand dune. It may have been partially roofed over, with an opening above the hearth. The door could be sealed by a draw-bar. There were cupboards in the house walls. The structure is unusual in having an additional side chamber on the north-east. (R.J.C.Atkinson)

One final group of buildings remains to be described, found mainly in the Outer Hebrides and, to a lesser extent, in the Northern Isles. Known as wheelhouses, these structures are usually circular and are divided up into compartments by internal stone walls arranged like the spokes of a wheel, with an open central area, often occupied by a stone hearth. In some examples, known as aisled wheelhouses the radial walls are separated from the house wall by a narrow gap.

Wheelhouses can be free-standing structures but more often they are sunk into a sand dune, the inner wall acting as a revetment against unstable sand. The radial walls must have helped support the roof which may well have been strengthened with rafters of whale bone, there being no local wood.

At Middlequarter, Sollas on North Uist the refuse from the house was piled in a midden over its roof, creating a virtually subterranean structure, entered from a long, funnel-shaped passage. The central room, about 9 m. (9.8 yd) in diameter, was divided into 12 bays, each wide enough to sleep two or three persons, or to act as store rooms. On one side a door led into an extramural oval chamber. Beside the central hearth, a stone-lined pit was used for storing water or shellfish. A stone quern indicated that grain had been grown and ground. The most bizarre feature of the Sollas wheelhouse was the discovery of some 200 sheep buried beneath the floor, mostly head downwards in small, conical pits. On the island of South Uist at the wheelhouse of A Cheardach Bheag some 20 deer jawbones had been set in an arc around the central hearth, and at the neighbouring site of A Cheardach Mhor 32 ox teeth were found beside one of the piers.

Wheelhouses usually occur close to fertile machair land and are not defended in any way. They seem to have been the Iron Age equivalent of the modern island croft. There is no clear dating for the beginning of wheelhouses although there are good reasons for thinking that they follow the demise of the brochs, their open, unprotected sites perhaps indicating less troublesome times. They seem to have flourished in the later Scottish Iron Age of the second to fifth centuries AD.

LINK

LINK

Crannogs

Reconstruction of the crannog on Milton Loch excavated in 1953. (Tracey Croft)

All over Scotland, but particularly in the west, are small, artificially constructed islands in lochs, marshes, rivers and estuaries known as crannogs. They were built of timber, brushwood or stones and often utilized natural features on the lake bed to give them greater stability. Many began as wooden platforms which were later reinforced with boulders and it is these which now appear as stone mounds protruding above the surface of a loch, or in reclaimed fields close to the water’s edge.

On the islands stood wooden buildings, usually circular huts, although rectangular ones are known. It is assumed that these were permanent dwellings which would also have served as refuges in times of emergency, providing safe store huts for food and livestock away from predators and vermin and also acting as fishing and fowling bases. Some were connected to the shore by stone or wooden causeways and it is likely that good farming land existed close by. From time to time log boats have been found offering an alternative means of access. Although excavated sites such as Milton Loch (Kirkcudbright) give dates in the fifth century BC, and Lochmaben (Dumfries) in the first century BC, it is very clear that crannogs were built over an excessively long period from early in the first millennium BC until at least the sixteenth century AD.

Mrs C.M.Piggot’s excavation of one of the crannogs in Milton Loch in 1953 is one of the best documented. An island constructed of timbers over clay lay some 35 m. (38 yd) from the present shoreline, to which it was connected by a wooden causeway. It had its own small wooden jetty and harbour leading out into deep water, which was partially supported by a rock outcrop. The island was mainly occupied by a round house with a conical reed-thatched roof. It was 12.8 m. (14 yd) in diameter, and had been partitioned into a series of rooms, one with a central hearth. Around the edge was a gangway 1.5m. (1.6 yd) wide supported on a framework of stout wooden piles.

The plough head and stilt of a wooden ard was uncovered beneath the foundations and was dated to 460–500 BC; an enamelled bronze dressfastener from within the house dated from the second century AD. These divergent dates suggest an intermittent use of the crannog over some seven centuries, and any attempt to interpret the excavation should bear in mind that all the features uncovered were not necessarily contemporary. Fragments of a quern and a spindle-whorl, together with the plough ard, suggest a simple economy of mixed farming on the loch shore, supplemented by fishing and fowling, during the late Iron Age.

LINK

LINK

The duns

Dun an Sticir (Dùn an Sticer)

This circular and galleried fort was built on a islet in Loch an Sticir, North Uist (Uibhist a Tuath). There are two causeways connecting it to another islet, Eilean na Mi-chomhairle, and from there to the north shore of the loch.

The remains of the fort are 3.6m high and about 18m in diameter. Its entrance faces the causeway and there are traces of a guard-cell and a gallery within the 3.5m thick wall.

There is a Medieval house built inside the fort. Tradition says that around 1600 Hugh MacDonald, who failed to overthrow his cousin Donald Gorm as 8th clan chief, hid there for a year. Then he was captured and left to die horribly in the dungeon at Duntulm Castle in Skye with a plate of salt beef and an empty jug.

Characteristic of western Scotland from Galloway to Lewis is a dense concentration of small stone-walled forts called duns. Circular or oval in plan, they have exceptionally thick walls enclosing an area of up to 375 sq.m. (448 sq. yd) which contains timber buildings. The layout of the dun was often determined by its location on a rocky knoll or promontory, an island in a loch or occasionally on flat ground (plate 50). The walls were usually solid and stood about 3 m. (3.3 yd) high, often with an inward sloping batter on the outside. Sometimes they had timber lacing to give them extra stability. The only break was an entrance passage, often with door-checks and bar-holes. There might be a cell built into the thickness of the wall which acted as a guard chamber.

On top of a long ridge on the island of Luing (Argyll), near Leccamore farm, stands a well preserved dun measuring some 20 m. by 13 m. (22 yd×14.2 yd), enclosed by a wall 5 m. (5.5 yd) thick and still standing 3 m. (3.3 yd) high. It is slightly unusual in having two opposing entrances, that on the south-west with well-preserved door jambs and a bar-hole and slot for securing a heavy wooden door. On either side of the north-east entrance are guard cells. The western one is carefully corbelled and contains a stair that allows access to the wall-head. To give the dun extra strength it is protected by an outer wall and two rock-cut ditches.

Dun excavations have not been particularly helpful in revealing details of interior structures but there is some evidence for timber buildings placed against the inner stone walls. A ledge or scarcement 1.5 m. (1.6 yd) above the floor of a dun at Ardifuar (Argyll) is seen as evidence for the anchoring of such constructions. Although we know next to nothing of their economy it seems almost certain that duns were fortified homesteads eking a fragile living from a difficult environment, and each jealously guarding its own immediate territory. The earliest duns may have originated in the seventh or sixth centuries BC but their type persisted until at least the third century AD. It should be pointed out that this part of Scotland was not occupied by the Romans and so an Iron Age way of life persisted into the first millennium AD.

Assaulting Maiden Castle AD 43+ [45]

Legion IInd Augusta, commanded by the future Emperor Vespasian; according to Suetonius they fought 30 battles, conquered two tribes (almost certainly the Dobunni and Durotriges), and captured 20 towns and the Isle of Wight. Excavations at Maiden Castle and Hod Hill forts in Dorset have unearthed dramatic evidence of their storming under cover of barrages of catapult bolts.

#

The fort was conquered after a protracted assault on the E gateway by Vespasian and Legio II sometime after AD43; over 50,000 slingstones from nearby Chesil beach that were found in this area during excavations bear mute testimony to the stiff resistance the defenders offered. (The downhill range of a slingstone from the ramparts here would be c140m). In a British war cemetery located just outside the East gate were buried 30 warriors killed during the Roman assault. This, we know, because one of the skeletons exhumed had a Roman ballista bolt lodged in its spine. In addition, each warrior was accompanied by his tankard and a joint of meat, but no weapons, presumably as they had all been confiscated by the conquering Romans.

Archaeological evidence is not clear whether the Romans actually occupied the fort after the fighting. There is some indication that, having broken the back of the local tribes who made their stand there, Vespasian moved on without occupying the castle, as his aim of destroying the armed warriors was more important than detaching a garrison at every turn.

By AD70 the local civil population had relocated to the town of Durnovaria (Dorchester), and the fort remained deserted until late in the 4th century, when a 12m square Romano-British temple was built in the North-East area of the old fort, having a square cella at the centre surrounded by a verandah, and a 2-roomed priests house just to the North.

#

THE SIEGE OF MAIDEN CASTLE, ENGLAND. BY MORTIMER WHEELER.

Sir Mortimer Wheeler was one of the great excavators of the twentieth century. Colorful, military-like in his organization and attitude to excavation, he refined Pitt-Rivers’ methods to a high pitch in the 1920s and 1930s. He excavated mostly Roman sites, with the dig at the great Iron Age hillfort at Maiden Castle in southern England being the climax of his British work. Maiden Castle was attacked and invested by a Roman legion in A.D. 45. Here is Wheeler’s brilliant reconstruction of this long-forgotten event.

The Early Roman Period (c. A.D. 43–70)

And so we reach the Roman invasion of A. D. 43. That part of the army of conquest wherewith we are concerned in Dorset had as its nucleus the Second Augustan Legion, whose commander, at any rate in the earlier campaigns, was the future Emperor Vespasian Precisely how soon the invaders reached Maiden Castle can only be guessed, but by A. D. 47 the Roman arms had reached the Severn, and Dorset must already have been overrun. Suetonius affirms that Vespasian reduced “two very formidable tribes and over twenty towns (oppida), together with the Isle of Wight,” and it cannot be doubted that, whether or no the Durotriges (as is likely enough) were one of the tribes in question, the conquest of the Wessex hill-fort system is implied in the general statement. Nor is it improbable that, with the hints provided by the mention of the Isle of Wight and by the archaeological evidence for the subsequent presence of the Second Legion near Seaton in eastern Devon, a main line of advance lay through Dorset roughly along the route subsequently followed by the Roman road to Exeter. From that road today the traveller regards the terraced ramparts of the western entrance of Maiden Castle; and it requires no great effort of the imagination to conjure up the ghost of Vespasian himself, here confronted with the greatest of his “twenty towns.” Indeed, something less than imagination is now required to reconstruct the main sequence of events at the storming of Maiden Castle, for the excavation of the eastern entrance has yielded tangible evidence of it. With only a little amplification it may be reconstructed as follows.

Approaching from the direction of the Isle of Wight, Vespasian’s legion may be supposed to have crossed the River Frome at the only easy crossing hereabouts-where Roman and modern Dorchester were subsequently to come into being. Before the advancing troops, some 2 miles away, the sevenfold ramparts of the western gates of Dunium towered above the cornfields which probably swept, like their modern successors, up to the fringe of the defenses. Whether any sort of assault was attempted upon these gates we do not at present know; their excessive strength makes it more likely that, leaving a guard upon them, Vespasian moved his main attack to the somewhat less formidable eastern end. What happened there is plain to read. First, the regiment of artillery, which normally accompanied a legion on campaign, was ordered into action, and put down a barrage of iron-shod ballista-arrows over the eastern part of the site. Following this barrage, the infantry advanced up the slope, cutting its way from rampart to rampart, tower to tower. In the innermost bay of the entrance, close outside the actual gates, a number of huts had recently been built; these were now set alight, and under the rising clouds of smoke the gates were stormed and the position carried. But resistance had been obstinate and the fury of the attackers was roused. For a space, confusion and massacre dominated the scene. Men and women, young and old, were savagely cut down, before the legionaries were called to heel and the work of systematic destruction began. That work included the uprooting of some at least of the timbers which revetted the fighting-platform on the summit of the main rampart; but above all it consisted of the demolition of the gates and the overthrow of the high stone walls which flanked the two portals. The walls were now reduced to the lowly and ruinous state in which they were discovered by the excavator nearly nineteen centuries later.

That night, when the fires of the legion shone out (we may imagine) in orderly lines across the valley, the survivors crept forth from their broken stronghold and, in the darkness, buried their dead as nearly as might be outside their tumbled gates, in that place where the ashes of their burned huts lay warm and thick upon the ground. The task was carried out anxiously and hastily and without order, but, even so, from few graves were omitted those tributes of food and drink which were the proper and traditional perquisites of the dead. At daylight on the morrow, the legion moved westward to fresh conquest, doubtless taking with it the usual levy of hostages from the vanquished.

Thereafter, salving what they could of their crops and herds, the disarmed townsfolk made shift to put their house in order. Forbidden to refortify their gates, they built new roadways across the sprawling ruins, between gateless ramparts that were already fast assuming the blunted profiles that are theirs today. And so, for some two decades, a demilitarized Maiden Castle retained its inhabitants, or at least a nucleus of them. Just so long did it take the Roman authorities to adjust the old order to the new, to prepare new towns for old. And then finally, on some day towards the close of the sixties of the century, the town was ceremonially abandoned, its remaining walls were formally “slighted,” and Maiden Castle lapsed into the landscape among the farm-lands of Roman Dorchester.

So much for the story; now for its basis. First, scattered over the eastern end of Maiden Castle, mostly in and about the eastern entrance and always at the same Romano-Belgic level, were found upwards of a dozen iron arrowheads of two types: a type with a pyramidal point, and the simple flat- bladed type with turn-over socket. Arrowheads occurred at no other Iron Age level, but both types are common on Roman military sites where ballistae but not hand-bows are to be inferred. There, then, in the relatively small area uncovered, are the vestiges of the bombardment.

Secondly, the half-moon bay which represents the Iron Age B adaptation of the Iron Age A barbican, close outside the portals of the eastern entrance, was covered with a thick layer of ash associated with the postholes of three or more circular or roundish huts. In and immediately below this ash were quantities of late Belgic or “Belgicizing” pottery. In the surface of the ash was similar pottery with scraps of pre-Flavian Samian. There are the burned Belgic huts, covered by the trodden vestiges of the continued post-conquest occupation for which more tangible evidence will be offered shortly.

Thirdly, into this ash a series of graves had been roughly cut, with no regularity either of outline or of orientation, and into them had been thrown, in all manner of attitudes — crouched, extended, on the back, on the side, on the face, even sitting up — thirty-eight skeletons of men and women, young and old; sometimes two persons were huddled together in the same grave. In ten cases extensive cuts were present on the skull, some on the top, some on the front, some on the back. In another case, one of the arrowheads already described was found actually embedded in the vertebra, having entered the body from the front below the heart. The victim had been finished off with a cut on the head. Yet another skull had been pierced by an implement of square section, probably a ballista bolt. The last two and some of the sword-cuts were doubtless battle wounds; but one skull, which had received no less than nine savage cuts, suggests the fury of massacre rather than the tumult of battle — a man does not stay to kill his enemy eight or nine times in the melee; and the neck of another skeleton had been dislocated, probably by hanging. Nevertheless, the dead had been buried by their friends, for most of them were accompanied by bowls or, in one case, a mug for the traditional food and drink. More notable, in two cases the dead held joints of lamb in their hands joints chosen carefully as young and succulent. Many of the dead still wore their gear: armlets of iron or shale, an iron finger-ring, and in three cases bronze toe- rings, representing a custom not previously, it seems, observed in prehistoric Britain but reminiscent of the Moslem habit of wearing toe-rings as ornaments or as preventives or cures of disease. One man lay in a double grave with an iron battle-axe, a knife and, strangely, a bronze ear-pick across his chest. The whole war cemetery as it lay exposed before us was eloquent of mingled piety and distraction; of weariness, of dread, of darkness, but yet not of complete forgetfulness.

The date of the cemetery was indicated by a variety of evidence. Most obvious is the Roman arrowhead embedded in the vertebra, but other associated relics point to the same conclusion. The seventeen pots put into the graves at the time of burial are all of that Wessex “Romano-Belgic overlap” class which has long been recognized at Jordan Hill, Weymouth, and elsewhere. The gear with one of the skeletons included, as has been remarked above, a Roman ear-scoop,” the use of which may or may not have been understood more clearly by its Belgic possessor than by the modern antiquary; at least it implies Roman contacts which, in Wessex, appear not long to have anticipated the Roman Conquest. One grave, moreover, contained a late British coin, and though it was impossible to say safely whether the coin was inserted at the interment or was incorporated in the loose ash into which the grave was cut, at least it was dropped within a very short time of the event. And finally, the materials included in the strata which “bracket” the cemetery are themselves, as noted above, sufficient to indicate a date at the end of the pre-Conquest period.

There, then, is the climax of the more human side of the story of conquest. But on the structural side the evidence for that event and for its sequel is no less vivid. On the topmost Belgic road-metal, in both portals of the eastern entrance but particularly in the southern, excavation revealed the tumbled stones from the massive walls that had formerly flanked the entrances. Here and there the fallen stones lay overlapping, like a collapsed pack of cards, in the sequence in which they had formerly stood as a vertical wall. With them was no cascade of rampart-earth such as might have implied a fall through subsidence, even could one presuppose the coincidence of the simultaneous fall of every part of the structure; the walls had been deliberately pulled down and no attempt had been made to replace them. But that was not all. Over the debris in each portal a new road had been built, metalled like the Belgic roads now buried beneath them. The new roads partially covered the surviving bases of the flanking walls, showing that the condition of these today is identical with their condition at the time of the road-building and confirming the permanence of the structural ruin. No provision of any kind was made in the new scheme for a gate; not a single post-hole was associated with the new road, and indeed the mutilated rampart-ends would have provided a poor setting for a fixed barrier. The implications of all this are evident. The entrance had been systematically “slighted’ and its military value reduced permanently to a minimum; but traffic through it did not cease, no interval occurred in the continuity of the occupation.

The picture is now complete in outline. Disarmed at the Roman Conquest, Maiden Castle remained in use for about a quarter of a century after the invasion, a pre-Roman city still in all essentials, partaking only a little of the cultural equipment of its conquerors. The picture is a reasonable and convincing one. The first generation of Roman rule was preoccupied with the subjugation of the difficult hill-countries of the north and west, with the development of mining areas, the planning of arterial roads, the founding or development of those few towns which had an immediate military or commercial function. Dorset offered, it is true, iron ore on a modest scale; but between Sussex and the Mendips there was little mineral wealth to attract the Roman prospector in the first flush of conquest. Wessex could wait. There was no urgent need to upset the traditional economic basis of the urbanized peasantry which crowded the downlands. To do so would have been to court added political difficulties at a time when difficulties were already manifold. It was better that, under surveillance, the Wessex farmers should for a time (and doubtless in return for the periodical payment of just or unjust dues) be allowed to maintain themselves in the fashion which they knew. The removal or, alternatively, the ennoblement of their rulers would rob them of independent leadership. A few police patrols would do the rest.

Celtic Farm - Butser Ancient Farm

Butser Ancient Farm, near Petersfield in Hampshire, England, is a working replica of an Iron Age farmstead where long-term experiments in prehistoric and Roman agriculture, animal husbandry and manufacturing are held to test ideas posited by archaeologists. The period of interest is that from 400 BC to 400 AD.

Butser Ancient Farm was founded in 1972 by experimental archaeologist Peter J Reynolds (6 November 1939 - 26 September 2001), and named after its original site at Butser Hill, a few kilometres from Petersfield. In 1976 a second site was opened at Hillhampton Down about a kilometre away, and in 1989 the original site at Butser Hill was closed down. In 1991 the project moved to Bascomb Copse at Chalton, Hampshire, about 5 km from the original site.

Buildings at the farm include simulated pre-Roman roundhouses and a simulated Roman villa. The Pimperne House was the first full-sized roundhouse to be built at the site, and at the time the largest in western Europe.

An episode of the 2005 BBC Television documentary series What the Ancients Did for Us examining the ideas and inventions of the Ancient Britons was filmed here.

The farm is open to the public.


LINK

CAUSEWAYED ENCLOSURES

An aerial view of Whitesheet Hill causewayed enclosure on the edge of the chalk escarpment in Wiltshire. A Bronze Age barrow and cross-ridge dyke also appear in the photograph. (J.E.Hancock)

Ground plans of some typical causewayed enclosures. 1 Windmill Hill, Wilts; 2 Robin Hood’s Ball, Wilts; 3 Whitesheet Hill, Wilts; 4 Briar Hill, Northants; 5 Whitehawk, Sussex. Ditch sections outlined. (Sources: various)

A human skull on the floor of a causewayed ditch at Hambledon Hill, Dorset. (Robin Holgate)

A section across the outer ditch of Windmill Hill, Wiltshire, excavated in 1988. (Alastair Whittle)

These enclosures, of which about 50 are known, are usually roughly circular in plan and consist of one, two or three concentric rings of ditches, dug as a series of irregular pits, probably by gang labour, and separated by undug causeways of soil, rather like a string of sausages. The material from the ditches was thrown up into an internal bank, sometimes revetted with posts or turf, or crowned by a stockade. Many of these banks have now disappeared entirely. It was originally thought that the gaps in the ditches were entrances for droving cattle, and that gaps in the banks corresponded with them. However, it is clear that the banks were much more continuous, with only a few breaks. If entrances existed they cannot always be positively identified, but groups of post holes at Hembury, Crickley, Whitehawk and Hambledon have been interpreted as the remains of wooden gates. Little sense has been made of the interior of the enclosures, which seem to contain a mass of pits, post holes and gullies, perhaps indicating at least temporary settlement.

The causewayed enclosures crown rounded hills in the chalk lands like The Trundle in Sussex and Knap Hill in Wiltshire, but they are also found in low-lying valleys like Abingdon in Oxfordshire and Staines in Middlesex, and on saddles and ridges as at Combe Hill and Whitehawk, both in Sussex. They are almost exclusive to lowland Britain and stretch from Hembury in Devon north to Alrewas in Staffordshire and Barholm in Lincolnshire, with a possible addition at South Kirby in West Yorkshire.

There is as yet little evidence that the majority of enclosures were built for defensive purposes, and the earthworks sometimes slope down the side of a hill across the contours as though deliberately displaying their interiors to the outside world. However, some East Anglian sites seem to have massive stockades inside the ditches, which might be interpreted as a form of defence. Isobel Smith suggests that they follow a ‘predetermined plan carried out regardless of topography’. Size varies considerably from less than a hectare (3 acres) at Rybury in Wiltshire to over 8 ha. (20 acres) at Windmill Hill and Hambledon Hill.

Most excavation of causewayed enclosures has concentrated on the ditches which are usually some 3 m. (10 ft) wide and seldom less than 1.5 m. (5 ft) deep. Where more than one ring of ditches occurs, the outer ring is usually the deepest as at Windmill Hill and The Trundle. It is not clear if each circuit of ditch is contemporary with its neighbour, and some enclosures may have increased their size as their importance grew. At Etton, Francis Pryor has suggested that only a few segments of ditch were dug at any one time. Although the ditches were basically the quarries for bank material, it is clear that they had a part to play in the activities at the site, since they were cleaned out on a number of occasions. Even so excavations show that they frequently contain large quantities of domestic rubbish. This usually consists of layers of animal bones, (especially cattle, sheep, pigs and deer), mixed with fragments of pottery, vegetable refuse and charcoal, broken flints, the occasional dead dog and human bones. Often this rubbish was carefully covered with soil as though to reduce the smell of rotting garbage.

Most of the pottery found in the ditches belongs to round-bottomed, baggy-shaped vessels. Finer quality carinated bowls of Grimston type seem to have been of special significance. Numerous axes of non-local stone and pottery tempered with grit from Cornwall and found as far east as Gloucestershire and Sussex suggest that these objects had been brought to the enclosures from long distances. The animal bones show the cut marks of flint knives, perhaps indicating on-the- spot butchering, and some ox skulls show signs of pole-axing with a sharp flint point over the left eye.

Deserving particular attention are the human remains found in the ditches. A large number of human skulls are recorded, often lacking their lower jaws. Some of these, at Hambledon Hill for example, seem to have been deliberately positioned on the ditch floor, as though to ward off evil spirits. Others are more casually scattered, and mixed with other human bones, suggesting that they may have been swept into the ditch in a cleaning operation. If this is the case then it is likely that they originated in the centre of the enclosure where corpses may have been decay. Roger Mercer has written of the central enclosure at Hambledon Hill, describing it as ‘a vast, reeking open cemetry, its silence broken only by the din of crows and ravens’. This implies that causewayed enclosures played a far greater part in the funeray ritual of neolithic Britain than has been realized. In Celtic Iron Age times human heads were collected as trophies and stored as prize possessions. It is perhaps worth wondering if a similar cult existed in the neolithic period.

Child burials seem to have had a special place in this ritual. At Windmill Hill the deliberate burials of two young children were found in a ditch, together with the skulls of three more. At Whitehawk the excavator found the ashes of a hearth containing fragments of five skulls, all of them of young people between 6 and 20 years. In the same ditch section was the complete skeleton of a young mother with her new-born child. Child burials also occurred in the ditches at Hambledon Hill, where they accounted for 60 per cent of the burials. Some lay crouched in the ditch bottom with cairns of flints above them. At the same site the lower trunk of a 15-year-old boy had been dragged into a ditch, perhaps by animals, whilst the flesh was still upon it. Complete adult skeletons have also been found at Offham Hill (Sussex), Abingdon (Oxon) and Staines (Middx).

Probably the best-known example of a causewayed enclosure, and first to be investigated, is Windmill Hill, 2.5 km. (1 1/2 miles) north-west of Avebury in Wiltshire. Three roughly concentric rings of causewayed ditches circle a low hilltop. The outer ditch has a diameter of 365 m. (1,200 ft). The mean diameter of the middle ditch is 200 m. (660 ft), whilst the inner measures about 85 m. (280 ft). The ditches do not follow the contours of the hill; instead they hang lop-sidedly down the steeper northern hillslope. They may not all be contemporary and recent excavations by Alastair Whittle suggest that the outer ditch may have been added later (plate 8). All the ditch sections are very irregular and vary considerably in size. Excavation by Alexander Keiller between 1925 and 1938 showed them to be flat bottomed, and deepest in the outer circle and shallowest in the inner ring. Only at the eastern side of the outer circle can any trace of the bank now be seen, though the excavations showed that it was present inside all the ditches. We shall probably never know if it was topped by a stockade, thus making it defensive.

Many fragments of early neolithic pottery were found in the enclosure. Nearly one-third of it had been made from Jurassic clays found some 30 km (20 miles) away around Frome and Bath. How such fragile material was carried to Windmill Hill remains a mystery. Also deliberately buried in the ditches were domestic objects such as flint scrapers, stone axes and animal bones suggesting some form of settlement, either temporary or permanent, as well as the skeletons of the two children mentioned above.

A variety of interpretations of causewayed enclosures have been offered over the years. Settlement sites and defended camps were first suggested 50 years ago, and this was possibly closer to the truth than has more recently been supposed. The large quantities of domestic rubbish in the ditches at some of these sites would support such explanations. The fact that some of the rubbish had come from some distance away led to the idea that the enclosures were centres for periodic fairs and tribal gatherings. The wealth of animal bones was used to suggest that the sites were corrals where cattle were annually rounded up for branding, gelding or culling. Following the work at Hambledon Hill the idea of the central area of the enclosures being used as a vast mortuary for the exposure of corpses has proved most popular. There is no clear answer. It is probably wrong to try and see each site as serving the same function but the various similarities between them perhaps indicate that they were most likely ritual cult centres, where people met at certain times of the year to mourn their dead and celebrate the well-being and fertility of their crops, their animals and themselves.

Amongst the causewayed enclosures are a small group of strongly defended settlements with causewayed and continuous ditches, which tend to be sited on hill spurs. The best known examples are Hembury in Devon, Crickley Hill in Gloucestershire, Hambledon Hill in Dorset and Carn Brea in Cornwall. These sites were defended by steep natural hill slopes as well as man-made defences. At Carn Brea a massive enclosing wall, built of boulders and 2 m. (6.6 ft) wide at the base, surrounded an area of 0.8 ha. (2 acres). The local geology precluded the digging of a causewayed ditch. Post holes indicated a number of probably domestic buildings in the enclosure, some of which had been destroyed by fire. Outside, larger enclosures were used for agriculture. More than 800 leaf-shaped arrowheads, some broken, and others amongst the boulders of the rampart, strongly suggested that Carn Brea was attacked on more than one occasion.

Crickley Hill, in its final phase, was defended by a continuous ditch and a strong stone-built internal wall topped by a stockade. Four hundred arrowheads and signs of burning again indicated a dramatic end for the enclosure. A violent end to Hambledon Hill is also seen as likely from the signs of the burning down of the stockade on top of the bank, and the skeleton of a young man with an arrowhead in his chest found there, apparently killed whilst rescuing a child from the burning enclosure.

Roger Mercer has observed that after these defended enclosures went out of use, no further defensive sites are known until the appearance of hillforts a thousand years later. It is unlikely that the causewayed enclosures were dug by the very first farmers who arrived in Britain. At first concentrated communal effort was required to establish the farming way of life with its forest clearance, house building, crop growing and animal husbandry. Only after a generation or two would there be time for large numbers of people to gather at the slack times of the year when the farming calender was not too busy, to engage in the construction of communal earthworks.

HILLFORTS IN SCOTLAND AND BEYOND

Professor Ian Ralston, Professor of Archaeology, University of Edinburgh

It is impossible in a written report to convey the dramatic visual impact of the large number of excellent colour slides used by Professor Ian Ralston to illustrate his lecture to the Inverness Field Club.

Professor Ralston drew from his long experience culminating in the Professorship of Archaeology at University of Edinburgh to demonstrate the many and varied links between the structure of hill forts in Scotland and those in the rest of Great Britain and different parts of mainland Europe. Latest estimates are that there are about 20,000 hill fort sites in Europe of which some 1,000 are in Scotland. Maiden Castle in Dorset is a typical upland site with a very imposing set of banks and ditches designed in a warrior context to help both with attack and defence. These are now grassed over but when first constructed the gleaming white chalk would have presented a formidable appearance. In Ireland certain hill forts are traditionally associated with resistance to the arrivals of peoples speaking Celtic languages. In a later period the conflict between Iron Age people and Roman invaders in the first century BC is demonstrated not only by the archaeological record but also by textual references such as the account by Julius Caesar of the wars in Gaul. At Alesia Roman siege works surrounded the Iron Age fort prior to the decisive battle, which led to the Roman conquest of Gaul. At that time hill forts were very widely distributed throughout Europe and were not confined to Celtic areas. Brochs were the only structures that were unique to Scotland. There is evidence going back to the bronze shields of the Bronze Age, prior to 1,000 BC, that warfare was very much an aristocratic pursuit with warriors having fine parade equipment. By the end of the period the archaeological record shows that individual named people such as Vercingetorix can be linked to specific sites.

A basic text is Hillforts of Britain, by A.H.A. Hogg, 1975. Hogg identifies four types of hill fort: Contour forts, promontory forts, cliff forts and ridge forts. This classification is not very precise. Contour forts where banks and ditches are moulded to the shape of the hill are the dominant ones in Scotland. Promontory forts, such as the largest one in Scotland at the Mull of Galloway, are usually a coastal feature. As their name implies, cliff forts are on the top of cliffs. Ridge forts often have a route way through them. There is a large variation in the size of these hill forts from the enormous to the minute. Some large upland sites in south west Germany and France extend to 1,000 hectares (2,400 acres), i.e. larger than the site of classical Rome. The largest sites were defendable blocks of upland country surrounded by cliffs. Sites at a low altitude often had a dense population within their perimeter. By contrast a few sites are at an extremely high altitude. In county Sutherland it is well worth ascending to the remarkable remains in a very exposed situation on top of Ben Griam Beg 2,000 feet above sea level. There is even evidence of some external settlement outside the fort. In Ireland, Caherconree fort in county Kerry is at an altitude of 2,700 feet but it is unlikely to have witnessed any permanent settlement. The date range of hill forts extended over 2,000 years. Early examples in Germany date from the late Bronze Age about 2,000 BC whereas forts from the Slav part of Europe were constructed as late as the 9th Century AD. Nearby the fort at Burghead dates from the first millennium AD but it has subsequently been badly depleted. In some areas burrowing by rabbits has caused more destruction than human activities!

Different types of defensive style occur throughout the Iron Age period. In later years they were a response to Roman siege warfare. There are different combinations in the use of earth, stone or timber. Timber is frequently infilled with stone or other materials. In Europe the timber is often arranged vertically but in Scotland horizontal timbers predominated. The use of timber increased the risk of damage by fire. There is evidence in Europe that styles were copied from classical Greece although the purpose of certain military features such as vertical stone walls with bastions was not fully understood. Some styles appeared formidable but they were not truly defensive because access to the forts remained open at the sides or rear. This was the case with certain forts in Shetland.

In Scotland certain sites are difficult to dig because they continue to be in use. Obvious examples are the castles at Edinburgh and Stirling. At Bourges in France the Iron Age remains are now five metres below the surface. Many rural sites are still easily available for study from actual remains or from crop mark records. Quarry working has destroyed certain sites such as Broxmouth in East Lothian. At Broxmouth there was evidence that stone-built houses replaced timber houses about 200 BC and that the hill fort settlement was replaced by an unenclosed settlement. Traded Roman coins are evidence of a Celtic reoccupation of upland forts in Roman times. Traprain Law is the largest surviving hill fort in Scotland although it is only 16 hectares in extent. It goes back to the Bronze Age and it was there that the Traprain Law silver treasure dating from Roman times was found. Other large hill forts in Scotland were Dunagoil in Bute, Eildon Hill North dating from the Bronze Age and the Tap O'Noth. Elsewhere in Europe hill forts were abandoned and many people moved to live in oppida or towns established under Roman rule.

One intriguing subject is the origin of vitrified forts. They are forts where heat has transformed the surface of the stone to glass. Vitrification is witness to the systematic deliberate destruction by fire of conquered sites in the early part of the first millennium AD. To the local people it would have been a spectacular intimation of power. Good local examples are Craig Phadraig above Inverness and Knockfarrel near Strathpeffer. A series of dramatic slides illustrated experiments in 1980 to reproduce the process of vitrification. A wall constructed with stone and horizontal timers was set alight and kept burning vigorously throughout the night. By morning the wall had become unsafe and some vitrification had taken place.

In conclusion there were significant differences between Scotland and the rest of Europe. In Scotland many of the early sites were very large as at Eildon Hill North and Tap O'Noth. Their construction represented a large social effort. They date from the late Bronze Age and are separated by a gap from later occupation coinciding with the Roman period. In Scotland it is the early sites that are big in contrast with Europe where the larger sites are later. In Europe a money economy developed, the leaders began to live in large houses and the style of living became more luxurious with features such as the increasing use of wine. In Europe it was a civilised society that the Romans took over and enhanced. By contrast Scotland was too difficult for the Romans to take over and absorb.

THE NORTHERN IRON AGE

Castle Law Hillfort - this photo shows the earth banks and terraces of the hill fort.

Castle Law fort, Abernethy (Perth), showing the slots for timber lacing on the outer face of the inner wall. (R.C.H.M. Scotland)

In most parts of Scotland after the climatic deterioration at the beginning of the first millennium BC, there was a gradual expansion of the population. Bronze Age traditions survived and there was cultural continuity through into the Iron Age. Regionally the country is very diverse, with sparse settlement along the edges of the Atlantic coast and in the north-east, and greater density between the Tyne and the Forth, and along the Solway and Clyde estuaries.

The most intensive research has taken place in the Tyne and Forth area and this might slightly distort the overall view. Tiny hillforts, often less than a hectare in extent, are thick on the ground there. Hownham Rings (Roxburgh), developing from palisaded enclosure to defended hillfort, has already been quoted, and this sequential pattern seems to have been common, though many palisaded homesteads existed side by side with hillforts. The excavation of Broxmouth, on a low hill near Dunbar (East Lothian) revealed a very complex hillfort. An unenclosed homestead containing a single round house was succeeded by a fort with a single rampart and ditch. This defence was later doubled, and then reduced to a univallate form with a variety of entrances throughout. Five wooden round houses each about 11 m. (12 yd) in diameter, were detected in the fort, as well as circular stone-built examples which were later than the defences. The latter had timber posts to support the roof and beaten earth floors. Later the floors were paved and the roof posts must have stood on stone supports. A deposit of ox skulls beneath a house wall may have been a dedication burial.

Most hillforts in central and eastern Scotland were timber-laced. This means that horizontal timbers were laid within the core of both stone walls and earthen ramparts, rather than being set vertically as in southern Britain. Sometimes when the fort accidentally caught fire (or was deliberately fired) the timbers burnt at temperatures over 900°C, causing the stones to fuse together producing vitrification. This is seen clearly at Carradale (Argyll) and Dunagoil (Bute). There has been much debate as to whether vitrified forts were the result of a deliberate attempt to strengthen the defences, but the general consensus seems to be against the suggestion. Although some very early radiocarbon dates have been obtained for timberlaced forts, a beginning perhaps in the eighth century BC seems most acceptable. Finavon (Angus) is the best known of the timber-laced forts which suffered extensive vitrification. Although only enclosing 0.4 ha. it is defended by a massive stone wall some 6 m. thick, surviving to a height of 4.9 m. externally. Excavation by Gordon Childe in 1933–4 showed that only the top of the wall had vitrified, perhaps because the upper section had contained more timber to strengthen it. The excavator found traces of timber houses against the fort wall, and hearths, a possible oven, domestic refuse and crucibles for metalworking. Childe also uncovered a rock-cut cistern or well to a depth of 6.3 m. Such cisterns occur in other north-eastern forts, such as Castle Law (Abernethy, Midlothian). This latter site, also excavated by Childe, passed through at least three phases consisting of palisade, single timber-laced rampart and multivallate fort. In the fort ditch an underground storage place or souterrain had been constructed, similar to the Cornish fogous. About 200 examples of souterrains are known from northern Scotland, though most are dated between the last century BC and the third AD. Those in Orkney and Shetland tend to be entirely underground, but in eastern Scotland they are only partially subterranean. As in Cornwall their function is uncertain though storage seems to be most probable.

LINK

CELTIC COMMUNITIES: FORT AND FARMSTEAD

Exterior and interior views of the Iron Age 'Pimperne house' (named after the site of an important archaeological find) which was carefully reconstructed at Butser Hill near Petersfield, Hampshire as part of an experimental recreation of a working farmstead of the Celtic period. These photographs remind us that the phrase 'thatched hut' can be misleading: this is a large, solidly-constructed dwelling of sturdy appearance. We have no idea what the interior arrangements or furnishings were like, since archaeologists have little more than post holes and the traces of hearths to go by. Experience suggests that it is probably a mistake to assume primitive squalor. (Richard Muir)

The oppidum of Entremont, seen from the east. It occupies a dominating position on the edge of the Puyricard plateau between two valleys. Detail is excavations.

Since the Celts left no written record, our only knowledge of the arrangement of their lives and their communities comes from the brief, and perhaps unreliable accounts left by Roman writers, and from the evidence of the spade. There are few clues to any detailed understanding of their society. We know that they were a 'tribal' people; we do not know exactly what their tribal structure was. We are told that they were a society divided by caste into a warrior 'aristocracy', a priestly class, and an underclass of peasants. We know that they practised slavery.

As for their pattern of building, the modern academic view is that a fairly highly organised society of scattered farms and farming hamlets looked towards local 'hill forts’ as the focus of their lives. These 'forts’ present a bewildering range of size, local density, and apparent purpose. Some are only an acre or two in extent, with a simple rampart-and-ditch defence, and traces of a handful of huts. Others enclose within huge multiple rampart systems scores or even hundreds of acres, and traces of up to several hundred huts. There are examples which fall at every point along this range of size. Some may have been villages; some were almost certainly simply refuges for people and their beasts in time of war; and the largest and most densely built-up can only be described as 'towns'— perhaps even as local 'capitals’.

We simply do not know how Celtic 'political’ society worked; so we cannot make intelligent guesses about the comparative frequency of purely military 'forts’, fortified refuges, permanent fortified villages, or massively defended 'royal capitals’. One safe assumption is that the time-smoothed banks and faint traces of post-holes, which today crown almost every skyline in some parts of Britain and continental Europe, give an altogether too primitive impression. The archaeological evidence shows a wide range of construction techniques, some extremely sophisticated. Ramparts which even today survive to a height of 90 feet would then have been much more sharply sloped and sculpted. Some were built up by means of timber lacing, rubble in-fill, and vertical facing walls of dry stone blocks. Some had defended gateways with indirect approaches and outworks which are reminiscent in their sound design of 18th century forts. We find evidence for massive timber gates surmounted by patrol-walks; for multiple stone-faced ramparts, quite possibly spaced according to the effective range of the available missile weapons; for a hierarchy wielding enough authority to stockpile 50,000 large sling-stones in handy positions on the ramparts of a fort, after gathering them from a beach some miles away. Whatever our ignorance of these people, one thing is sure: their chieftains had real authority, and wielded it over a social system wealthy and organised enough to put considerable manpower at their disposal for sustained tasks.

Julius Caesar describes encountering in central and northern Gaul a type of solid defensive wall which he terms murus Gallicus. This can best be described as a skeletal grid of timber beams placed crossways and nailed together, built up in layers, with earth and rubble rammed down into the spaces between the beams at every level. A dry stone wall faced this construction front and back— Sometimes covering the ends of the lateral beams, sometimes leaving them exposed. The core of the wall thus gave good resistance to battering rams, even when the facing had been breached; and the facing and in-fill protected the timber skeleton from fire. This 'Gallic wall' is known to have been at least 12 ft high in some cases.

The Gallic Celts came in contact with Greek settlers in southern France, and it is tantalising to wonder how much this contact affected Celtic defensive engineering. In this area several strongholds have been identified which boasted ramparts of stone construction, rather than of stone-faced earth. The best-known is Entremont, which the Romans described as an oppidum—'town'. Overlooking Aix-en-Provence, this triangular fortress, captured by the Romans in 123 BC, had walls of rough-cut stone blocks defended at intervals of about every 20 yards by towers with solid rubble-packed bases; the walls probably boasted battlements or parapets originally. Britain has not produced evidence of comparable sophistication. There are signs that some British forts were given improved defences at several periods; in about the 3rd century BC there was a general deepening of ditches and heightening of ramparts, and on some southern British sites the 1st century BC saw the raising of additional belts of ramparts and ditches and the construction of sophisticated indirect entranceways.

The settlements which were scattered right across the Celts' geographical range offer just as wide a variety of sizes and designs as the Torts', from isolated farmsteads perhaps supporting one extended family, to quite large villages of up to 40 acres or so—larger than most medieval and many modern villages. There have been several recent experiments in reconstructing, from archaeological evidence, working Iron Age farmsteads. A project on Butser Hill near Petersfield, Hampshire included several different types of living units based on post-hole measurements and surviving fragments of hut fabric. In fields cultivated by hand, or with primitive ploughs drawn by cattle, experimental crops of cereals thought to resemble contemporary grains have been raised. Crops such as spelt and emma were found to average some 1,600 lbs yield per acre even in poor conditions. Breeds of horse, cow, poultry and sheep which approximate ancient strains have been raised on these experimental farms—for instance the agile and hardy St Kilda sheep, a small goat-like creature raised for its wool. Weaving, potting, charcoal-burning and metal-smelting—all necessary to a Celtic community— have been practised on these sites using the reconstructed technology of the period. In the lower strata of Celtic society most men, women and children would have spent the bulk of their lives carrying out these labour-intensive tasks.

Courtyard houses

The interior of a courtyard house at Chysauster (Cornwall) with small rooms leading off from the central open area. (J.Dyer)

Plan of the completely excavated site of Walesland Rath, Pembroke, showing circular huts and rectangular structures, not all of which were contemporary. The excavator interpreted some of the buildings close to the bank as a continuous run of huts, but some of these could be interpreted as four-post structures as indicated. (After Wainwright, 1969)

As one moves from central southern England to the west and north, areas begin to display their own regional peculiarities. In western Cornwall and Scilly are about 60 courtyard houses, usually sited on fairly high ground. The best known examples have been excavated at Chysauster and Carn Euny where they occur in groups, but they are more frequently found as isolated farmsteads set amongst fields. A typical house was oval in plan and consisted of a central courtyard, sometimes paved, and almost certainly open to the sky. Around it were ranged four to seven rooms with thatched roofs, the largest opposite the courtyard entrance. This big circular room was often raised above the level of the yard and was most probably the living room. As well as a central hearth it often contained benches and drains. Other rooms may have been used for sleeping, working, storage and perhaps housing animals. The houses were very strongly built with carefully designed granite walls often 2–3 m. (2.2–3.3 yd) thick. A number of the houses had small paddocks or gardens attached, whilst beyond were extensive field systems. Most courtyard houses seem to date from around 200 BC and continue into the Roman period.

Another feature of western Cornwall are underground structures known locally as fogous. Most are found attached to settlements like Chysauster, Halligye and Carn Euny. They vary considerably in size but consist basically of a stone-lined and roofed underground passage. At Chysauster it is very short, but at Carn Euny the passage is 20 m. (22 yd) long with a corbelled stone chamber 4.5 m. (5 yd) in diameter on one side. It is almost certain that these structures were cellars or storage places, where food could be kept at a fairly even temperature. As hiding places they would have been death traps. A radiocarbon date from Carn Euny suggests that some may be as early as the fifth century BC thus predating the courtyard houses, but they continued in use throughout the later Iron Age.

Banked and ditched enclosures known as rounds are another feature of the Cornish and Devon countryside in the late first millennium BC. Essentially agricultural in economy they are difficult to define since they come in many shapes and sizes. Suffice to describe them both as farmsteads and hamlets, surrounded by a non-defensive bank and shallow ditch, and seldom more than a hectare in extent. Each contains a number of huts, usually sited against the inner face of the bank. They fit without difficulty into the pattern of enclosed farmsteads found in the rest of Britain.

Many similar sites also existed in southwest Wales where banks and ditches sometimes make it difficult to decide whether the site is not in fact a very small hillfort, since 50 per cent of Welsh forts enclose less than 0.5 ha. (1.2 acres). One such site, extensively excavated in 1967–8 was Walesland Rath in central Dyfed (fig. 60). Earliest occupation began in the third century BC with a settlement enclosed by a low clay bank and shallow external ditch. The western entrance had limestone walls leading up to a timber gateway. Over the south-east gate was a bridge or tower supported by six massive posts. The inner edge of the bank was lined with a continuous run of roughly rectangular wooden huts, with hard earth floors, occasional paving and indications that some had been used for animal housing. Evidence of grinding grain and metallurgy suggest that most of the buildings were occupied by people. Other circular houses with conical roofs stood inside the enclosures, together with at least one of rectangular shape.

The peripheral buildings at Walesland Rath have been compared with similar structures at Clickhimin broch in Shetland and various Scottish duns, and it is possible that they were a feature of contemporary sites along the Irish Sea coast. One word of caution; some of these buildings at Walesland may be interpreted as four-post granaries.

Glastonbury lake village

Reconstruction of an Iron Age house at the Peat Moors Centre. The house is based on one found at Glastonbury Lake Village.

A reconstruction of the Glastonbury lake village. (Tracey Croft)

The nearby village of Glastonbury was very different, partly because of the waterlogged nature of the site; it was excavated by Arthur Bulleid and Harold St George Gray more than 80 years ago and many otherwise perishable items have been preserved. It was built on an artificial island or crannog composed of tree trunks and brushwood and was completely surrounded by water, making it accessible only by boat. Alder logs, together with some of oak, ash and birch, had been felled on the shores of the lake and brought to the island, where they were laid in layers at right angles. The surface had then been levelled with bracken and peat, rubble and clay. A close-set wooden palisade around the edge of the site enclosed 1.4 ha. (3.5 acres). The posts, sharply pointed with axes, had been driven down into the peaty bed of the lake.

Within the enclosure were about 80 buildings, which were certainly not all contemporary, and only a few were major dwelling houses. A number of attempts have been made to reinterpret the old excavation report, and the writer is inclined to accept some of the views of the late Professor E.K.Tratman (1970), who considered that the village had passed through two distinct periods of occupation. At first square or rectangular houses were built, supported above the ground on oak piles. The builders were fine carpenters and all the wood was carefully jointed and the houses were skillfully constructed with walls of daubed hurdle work. After a period of abandonment, the artificial island was constructed around the ruins of the old houses, which were replaced by new circular buildings between 5.5 and 8.5 m. (6–9.3 yd) in diameter. They had walls of wattle and daub, floors of clay, some with central hearths and ovens, and reed-thatched roofs. The excavators observed the huts as low mounds, created by the build-up of as many as ten successive clay floors. Eight houses had wooden floorboards.

The most recent study of Glastonbury, by Bryony and John Coles, speculates that the village may have been occupied for 300 to 400 years into the first century AD. They suggest that as well as a series of dwelling houses, the village contained specialized areas for working wood and bone, bronze and iron objects and pottery manufacture. Most weaving and basketry probably took place in individual houses. Because the site was waterlogged numerous wooden objects have survived, including bowls and tubs turned on a pole lathe, containers and baskets, a ladder and a stout door a metre high. Turned axle boxes and wheel spokes belonged to chariots or carts pulled by pony-sized horses which wore iron snaffle bits and bronze terret rings, as well as old-fashioned bone and antler cheek pieces in a late Bronze Age style. The wooden handles of many iron tools such as sickles and saws, knives and billhooks can still be seen in Taunton Museum, together with weaving equipment and loom weights, local glass beads and fine pottery with beautiful flowing linear patterns.

Glastonbury also produced two iron currency bars; these were long iron bars resembling swords with rounded tips. Caesar referred to them in the mid-first century BC as one of three types of currency in use in Britain at that time, the others being coins of bronze and gold. Rotary handmills or querns were replacing the earlier saddle querns and were used to grind mixtures of wheat, barley and oats, some of which were combined with honey to make bread and small cakes or buns. Wild berries, acorns, parsnips, peas and dwarf broad beans were gathered. Meat in the form of mutton and lamb was most common, followed by beef and pork and perhaps horse flesh. This was supplemented with wild birds, including the pelican, and freshwater and sea fish. Perhaps indicating the inhabitants’ relaxations were finds of bone and antler dice and dice-boxes.

LINK

OPEN VILLAGES

Bronze Age House at Flag Fen, Cambridgeshire

Iron Age House at Flag Fen

1300BC, the people in the area built a massive kilometer-long row of posts which runs across the wetland and reaches dry land on either side. Towards the centre, the posts run across a man made timber platform of about a hectare in area. This site was the focus of many religious rites. It may also have served as a defensive barrier and route across the marsh. At Flag Fen, Cambridgeshire.

Scattered amongst the developed hillforts, but more often spreading over eastern England where forts are scarce, a series of settlements which we can best call open villages (some known since about 800 BC) now began to proliferate. An expanding population may have led to overcrowding in single farmsteads and the growth of nucleated clusters of houses in which the occupants could diversify their talents, for instance as metalsmiths, weavers, potters and carpenters, as well as farm and general labourers. Unenclosed round houses with streets and paddocks formed extensive groups which were usually located on the lower lying ground, and seemed to favour the river gravels.

Francis Pryor’s excavations of the Cat’s Water site on the extreme edge of the Fens at Fengate near Peterborough (Cambs) show that it dates from between 400 and 100 BC, and revealed more than 50 circular buildings. Clearly they were not all contemporary, nor were they all used for habitation. It was suggested that not more than a dozen would have been in use at any one time, perhaps occupied by a total of 25 to 30 people together with their animals. In general the houses were placed around the outside of the settlements, whilst the animals were penned in fenced and ditched enclosures in the centre. The site was divided up by numerous drainage ditches, made necessary by the high water-table so close to the Fens. Wild animals, fish and waterfowl supplemented the meat from cattle and sheep which the village produced. There was much evidence for pasture and meadow land around Fengate but cereal cultivation was poorly re-prepresented. Four post granaries were absent and the subsoil was too waterlogged for storage pits. A crucible with tin traces in it suggests that some metallurgy was practised at Cat’s Water.

The construction of a bypass at Little Waltham (Essex) revealed part of an open village settlement. Fifteen circular houses were uncovered dating from about 250 BC. Assuming that the unexcavated area was equally densely occupied, we can estimate a total of 30 to 35 houses. They stood on slightly raised ground above the river Chelmer, and each averaged 12.5 m. (13.7 yd) in diameter. The hut walls were made of wattle and daub panels supported by substantial timbers 0.20 to 0.35 m. (0.22–0.38 yd) thick which stood in deep wall trenches. An inner ring of posts to support the roofs was inferred from the great span of the rafters which would have needed internal support. There were no storage pits at Little Waltham, probably due to the poor soil conditions. Instead 8 four-post granaries were traced and more than 30 two-post drying racks. Bones of cattle, horses, pigs and sheep suggest that a mixed agricultural economy was practised. Pottery bowls of local manufacture were found in large quantities, together with everted-rim and footring bowls transported from the Mucking- Chadwell area of the Thames estuary.

At Mucking, also in Essex, the sites of more than 100 round houses have been uncovered in rescue excavations which are still largely unpublished. Little remained of the actual houses, but the circular drainage gullies that caught the water as it fell from the roofs were clearly visible. Some of the houses were in small garden-like compounds, but most were free-standing. Pits and post holes were common on the site, but were difficult to interpret as they ranged in date from the Bronze Age through to the Anglo-Saxon period.

On the south side of Bredon Hill beneath the hillforts of Bredon and Conderton Camp lies Beckford (Hereford and Worcs), an extensive village of houses and enclosures dating from 250–50 BC and covering several hectares. The enclosures seem to have been non-defensive, their ditches most probably for drainage. Each may have been individually owned. Inside were houses marked by rings of stakes and doorposts, cobbled yards, four-poster granaries and grain storage pits 0.5–1.0 m. (0.55–1.1 yd) deep and 1.0–2.0 m. (1–1–2.2 yd) in diameter. Examination of bones from the site shows that cattle and sheep were the main animals kept, but pigs, dogs and horses were also present. Under one of the cobbled yards a bundle of ten spit-shaped iron ‘currency bars’ was found.

On the north-eastern side of Winchester aerial photography revealed an enclosure on Winnall Down. Excavation has shown that it had a long history beginning in the neolithic period and continuing until medieval times. In the late Bronze Age four possible post-built round houses with west-facing porches were constructed. Later a Dshaped enclosure ditch surrounding half a dozen round houses, 24 storage pits and 19 four-post structures was dated by radiocarbon to the sixth century BC. By the second century BC the ditch had silted up and an open village of 10 round houses had been built over it, together with a large rectangular structure that may have been a sheep-fold, and a group of 80 storage pits to the east, along with some 16 four-posters. Eighteen burials were associated with the village, 12 of them children or infants. They had mostly been buried in crouched positions in quarry hollows and storage pits. The only adult male burial was also the only one with grave goods—a shale bracelet and a bronze thumb ring. The economy of the village seems to have been based like the others on mixed farming, which consisted of rearing sheep, cattle, pigs and horses, and growing sixrow hulled barley and wheat (sown in the autumn) together with beans and peas. The pottery used on the site consisted mainly of saucepan pots common throughout central southern England by the third century BC. Pieces of briquettage vessels, probably used for transporting salt, were also found.

As with the hillforts of the mid-first century BC there is little in the layout of the villages to suggest that they housed a hierarchy; every one seems to have been more or less equal. Currency bars and salt containers are seen by some as symbols of exchange. Perhaps villages could also be exchange centres? In Somerset the marsh villages of Glastonbury and Meare produced a range of semi-exotic goods. Meare Village West stood on the edge of a raised bog [less distance to throw witches]. Its houses had clay floors but very flimsy superstructures suggesting that they might only have been tents. The village was occupied only in the summer when its residents concentrated on the production of a wide range of specialist crafts, including glass beads, loom weights and antler combs. Among the many other materials used in the village were iron, tin and lead, shale, bone, wood and wool. Bryony Coles, one of the excavators of Meare, finds the situation of such a prosperous village in such a wet marginal location puzzling— but perhaps it was deliberate. She suggests that Meare may have been a market centre, a meeting-place, or seasonal fair on the periphery of a number of adjoining communities, sited in a neutral ‘no-man’s-land’ position, rather than a private hillfort: a clever expedient for keeping the peace. For a flourishing rural economy such exchange centres would be essential, and if she is right many more must remain to be identified.