Frome in Wessex
Long Barrows and Ditches, Hill Forts and Wetlands: Prehistoric Frome
Abstract
Contents Updated: Thursday, 26 October 2006
- Bronze and Iron Age Forts
- Magnetometry
- Wetlands
- Wetlands Archaeology
- Wetlands Tracks
- A Wet Track to Glastonbury
- Back to Articles on Frome
Bronze and Iron Age Forts
The modern history of Frome began with Aldhelm and his monks in the seventh century but the area centred on Frome with a radius of about 30 miles is rich in prehistoric remains.
Overlooking Frome just a few miles to the east is Cley Hill looking rather like a very large long barrow. In fact, two bronze age chiefs did have their shining white barrows here but they have long been dug out and hardly remain. Eunice Overend tells us, the name does not mean “clay” but derives from a word meaning steep, the hill actually being of chalk—whence the whiteness of the tombs—because it is geologically really an isolated bit of Salisbury Plain to the east.
In the Iron Age the solitary hill dominating the adjacent landscape was ideal as a fortress and so it was fortified. The people living there had to walk down to Whitbourne Springs (which provides beautifully soft water from the greensand underlying the chalk) for their water supply and then take it all the way back. The springs arise because the greensands rest on a bed of impermeable clay. The Iron Age ramparts are now damaged by the needs of the Lords Bath in maintaining the colossal fifteenth century heap of feudal indulgence, Longleat House, which is only five miles from Frome. To provide mortar for maintenance and lime to sweeten the land of the estates a large chalkpit has been dug into the side of the hill damaging the ancient fortifications.
Anyone interested in a really magnificent Iron Age fort need only travel forty miles due south of Frome to near Dorchester to find Maiden Castle, the best Iron Age fort in Britain with enormous earthworks, excavated by the celebrated Sir Mortimer Wheeler in the 1930s. The ramparts and entrances are complicated but were not good enough to resist the Romans of Claudius in 43 AD.
Another great earthwork with huge banks and a ditch can be seen only thirty miles from Frome, near Salisbury, another famous old town with its spectacular 400 feet high spired cathedral. Old Sarum was built by iron age people about 500 BC but was used by each succeeding wave of invaders, the Romans, the Saxons and the Normans. William the Conqueror paid off his army here in 1070 AD and brought his nobles to order in 1085. The Normans made it into a major city with a royal castle and a cathedral, but it had a poor water supply and Salisbury, founded in the 13th century took away its raison d’etre and Old Sarum faded in favour of Sarum (Salisbury). It persisted as a “rotten borough,” able to send MPs to Parliament, even though it had a negligible population, until the reform act of 1832.
Magnetometry
More than 400 hill forts sit on top of the hills of Wessex. Barry Cunliffe, a well known archaeologist in recent work has clarified much of their mystery. Cunliffe has used magnetometers, one of the new nondestructive ways of testing the ground—looking at what is beneath the surface without having to dig trenches. Curiously, the magnetometer operators have found that some hill forts are devoid of anything while others are chock-a-block.
The value of geophysical methods is that they can show up details in sites that have been almost destroyed by ploughing for cereals. Norsebury Ring in Hampshire and Castle Ditches in Wiltshire reveal interesting features underground although almost all surface features have gone. Norsebury shows two “villages” within the earthworks and a thirty metre (100 ft) ring, too big to be a hut unless it was the Iron Age equivalent of Buckingham Palace—the home of a Celtic king. More likely perhaps is that it was a henge, like stonehenge—a shrine of some sort.
At Castle Ditches, ploughed over the years into a somewhat mounded field, magnetometry revealed more than 50 rings—the outlines of Iron Age huts but across them were cut more recent ditches, apparently to protect large barns used for storage from the wet. The circular huts were made of rings of hazel sticks inserted into the ground then woven into wattle and daubed or covered in turves. Oldbury Castle in Witshire was occupied over such a long period that the magnetometer results show successive phases of building. They also revealed a stretch of defenses that had completely disappeared to the naked eye.
The first hill forts were built about 800 BC, though fort is the wrong word to use because they were scarcely defensive constructions. They had banks and ditches but these were simply to define a space, being quite inadequate for serious defence. Barry Cunliffe thinks they were used for seasonal festivals. These are the empty ones. They were reserved for the seasonal gatherings and no one built on these spaces, so the physicists find nothing special. Balkesbury, near Danebury in Wiltshire, and Walbury in Berkshire are in this category.
The defensive castles are usually much smaller and began to appear in a rash about 600 BC. Maiden Castle is a particularly fine example but others are Danebury and Barbury castle in Wiltshire and Bury Hill and Woolbury in Hampshire. Besides having elaborate systems of ditches and ramparts, these were evidently occupied for a long time. The environment had obviously become hostile and the Celts were indulging in frequent fights against each other. the wars were not long but were evidently severe. Skulls found at Danebury have obvious injuries of swords and spears, and some were scalped. In a pit at the eastern end of Danebury were 11,000 pebbles used as sling stones.
Stanton Drew Standing Stones, an older and bigger stone age temple than Stonehenge. The stones make up three circles, the great circle, the south west circle and the north east circle. It was found in 1997 to be the site of a neolithic temple twice the size of stonehenge and 500 years older. English Heritage did a geophysical survey which turned up nine concentric circles constituting a wooden temple. The stones are on private land but can be visited for a small charge.
About twelve miles north of Frome, a little east of Bath is Bathampton where a huge Iron Age fort covers almost 80 acres. It was occupied for at least half a millennium and had a large population. Unfortunately, little of it remains although the Celtic fields can still be traced and it offers an excellent view over the Avon Valley. The Avon Valley in those days must have been a formidable boundary and several pre-Roman castles are built on the high lands overlooking it.
Next to Frome are the Mendip Hills upon which there are many traces of pre-Roman Celtic activity. Maesbury castle has two ramparts and ditches enclosing about seven acres. The outer rampart and ditch have essentially eroded into the contour of the hill from which they were originally dug but the inner rampart can still be clearly seen and is between six and twenty feet high. An interesting feature is that part of the ditch to the north is actually flooded forming a moat. Iron Age moats are uncommon especially as high as this one is.
Dolebury Castle is a few miles north of Cheddar and offers spectacular views in all directions except back towards the Mendips themselves. What is impressive about Dolebury as an Iron Age Castle is that the ramparts were faced with great blocks of stone which are still visible. The site encloses about 17 acres and obviously had a considerable population. Of course, to get to Dolebury from Frome, the journey is almost 30 miles to the Western scarp of the Mendips. Go thirty mile to the east of Frome and you get to Stonehenge.
The population of Danebury, a typical small hill fort investigated by Cunliffe, was about 300. Signs of religious ritual were that dead animals or birds were found buried and often human heads were found disembodied, the bodies apparently being left after the fashion of the Persians to be picked by animals and birds before being buried later.
The surveys found 4,500 deep conical pits, 3 m deep dug into the chalk. The pits were supposedly to keep grain. The theory is that the outside layers of stored grain would rot from the dampness of the chalk but, provided that the narrow top was kept sealed, the decay of the outer layers would use up oxygen and prevent further rot. It does not sound too healthy or convincing. Was the reason they were all trying to kill each other because they were all insane with ergotism—St Vitus’ dance?
Most of these hill fort villages were abandoned over a century before Claudius arrived. Cunliffe believes the effects of the growing Roman republic destabilised British society even before the Roman emperors finished it off. Perhaps they had started to buy decent sized pots from the Continent for storing their grain free of fungus and had stopped suffering from St Vitus’ Dance. At any rate the hill forts were used in the last stands of the British against the legions and then mainly became warrens for rabbits. And so they remained until today.
Wetlands
Most of north Somerset was low lying and subject to floods so ancient people built lake villages on wooden piles driven into the lake-bottom ooze. The remains of these have been excavated at Glastonbury, an ancient and mystic town only twenty miles from Frome. Beyond the mound upon which the town was built were peat bogs extending to the Bristol Channel—the Somerset Levels. Lazy streams cut through swamps, pools and reeds punctuated here and there with thickets of willow, alder and oak.
The Somerset levels have been wet since the ice age and only began to be drained by the Romans, about the same time as they started to drain the fenlands. But the earlier marsh and lake dwellers led normal lives, making tools and weapons, clothes and ornaments like their contemporaries that did not have the water to contend with. Wetlands also retain signs of transport and communictions. The Somerset Levels have many tracks.
From 1892 to 1907 an area of slightly raised land near Glastonbury was cut into for excavation and discovered to be the site of an iron age village dated by sherds of pottery to about 60 BC. Two other sites at Meare were investigated from 1908. Work has continued intermittently since.
The structure of a hut of the lake villages was a base of timber and lathes held in place on the peat bottom of the lake by wooden piles. With a layer of clay placed upon it the floor of a circular hut was made. The centre of it was a hearth of baked clay. A structure like this gradually sank into the peat and for continued use had to be raised by the addition of more packing and a new clay floor from time to time. Some of the huts had ten consecutive floors. A similar construction was used to build causeways over the bog topped with timber and clay to provide dry access.
At Meare the villages seemed less substantial because the peat must have been less waterlogged and some of the huts were rectangular, apparently an earlier design. The village at Glastonbury must have been of some importance because it had a defensive wooden palisade surrounding it but those at Meare did not have any defences. The iron age tribes must have been warlike because the defended village was destroyed before the Roman invasion possibly with the massacre of the inhabitants. The sites at Meare lasted longer and apparently were re-occupied in the fourth century.
The relics of spinning and weaving equipment at the Somerset wetlands settlements shows that even in those days cloth was an important manufacture of the area. The acidic peat had preserved many wooden items which normally would have rotted including dug out canoes, and implements of iron and ornaments and personal decorations of bronze were quite common. Imported goods showed that trade routes were extensive. Tin for bronze ware came from Cornwall, shale bracelets from Dorset, jet came from Whitby in Yorkshire and even amber from the Baltic Sea coast was found. A broken axle proved that they had carts. The people lived on wheat, barley, peas and beans and cattle, sheep, horses, pigs and dogs were kept.
The Importance of Wetlands Archaeology
People lived in the north east Somerset area long before these lake folk however. Presumably pressure of population drove people into the swamplands but beforehand they preferred to remain on the hill tops where it was drier and healthier.
Water protects organic matter from decay by excluding oxygen, reducing the activity of bacteria and fungi. It also prevents insects like woodworm from invading the remains. Acidity is also helpful as long as the pH is not too low when bone and soft parts are dissolved. Wood left to the elements goes through the seasonal cycle of hot and cold and wet and dry, expanding and contracting all the while. This cracks it and weakens its structure allowing fungi and insects to penetrate and finish off the work. Underwater, the wetness is a constant that prevents the expansion and contraction cycle. The initial swelling might cause some damage but thereafter, the climate is uniform. The preservation of wood allows dendrochronology to be used to date the settlements.
The degree of preservation in wetland is remarkable. The bog bodies like Lindow Man are famous, but even the fish stew in a 6000 year old pot has been found in a Danish wetland site. Blood can still be detected on flint blades. The bodies of animals and molluscs can be preserved for centuries. In wetlands even basketwork has been preserved for over 3000 years.
Many wetlands contain archaeological evidence of past eras, and with modern techniques surprising information can be had. Analysis of lake sediments and peats yields up the microflora and microfauna of different times in history allowing the scientists to reconstruct ancient environments. The sequence of clearing, farming, abandoning and reclaiming can be seen.
Pollen remains give the history of local vegetation. In the muds below the settlement, the pollens reveal the state of the original land. Occasional appearances of cereal pollen shows that patches of brush were being cleared for temporary cultivation. Often these cleared and cultivated patches were actually on nearby drier patches, but they usually herald the utilsation of the watery areas, when population gets pressing. Settlement is indicated by a drop in native tree pollens accompanied by a spurt in cultivated plant pollens. The tree pollens that survive are often useful and quick growing types like hazel and alder.
Wetlands Tracks
The Sweet Track is a neolithic pathway across the marshes 6000 years old. It is built of hazel and ash probably accidentally coppiced from some cleared trees regrowing from their stumps. By 5000 years ago the woven wooden hurdles they made showed that these people were deliberately coppicing and pollarding for a lathe and hurdle industry. Thereafter there was probably an unbroken tradition of coppicing in the Somerset Levels from prehistory until the Domesday Book when lots of coppiced woodland was recorded in Somerset. The tradition contined until the nineteenth century but lathe and wattle then lost its attractions.
The Walton Heath track, from a slightly more recent period, consisted of hazel rods 2.5 metres (8 feet) long and 15 mm across. To make 60 metres of track, 5000 of these rods were used in hurdles buried in the soft wet peat. 85 percent of the rods were of hazel. The Sweet Track had planks from massive trees of oak, ash and lime, 400 years old and a metre across, felled with stone axes and split with stone wedges then laid over the hazel hurdles used as its base. The planks were 5 cm thick, 40 cm wide and three metres long.
A Wet Track to Glastonbury
Bryony Coles (New Scientist)
Water protects remains from wind, rain and drying winds. Archaeologists can examine ancient brains or fish stew in a pot—if they were waterlogged
Bryony Coles lectures in prehistoric archaeology at the University of Exeter. With J M Coles she wrote Sweet Track to Glastonbury.
“Now I’ve dug posts instead of postholes, I could never go back to a dryland site.” This remark, typical of converts to wetland archaeology, encapsulates the immediacy of the evidence: solid wood, three dimensional structures and the unweathered debris of human activity sealed and preserved for thousands of years in waterlogged deposits. As wetlands shrink the world over, their archaeology has become an important area of research from southern Chile to Labrador, Japan to Ireland.
Throughout the world, there is time-consuming, expensive and immensely rewarding work to be carried out before desic cation destroys the organic record of past human activity.
Many wetlands contain archaeological evidence. Those that do fall into two main categories. First, the real wetlands, wet when there was a human presence in antiquity and wet ever since. Reed-swamp, marsh, fen and bog, as in the Somerset Levels, and lake-edge, as in the circum-Alpine zone, are often rich in freshwater sites. Estuaries, deltas and coastal areas may also be fruitful for archaeologists, as in the Rhine Delta or among the Keys of southwest florida.
The second category of wetlands consists of land that has become wet, holding evidence of the human activity that took place when the land was relatively dry, just before a nearby sea or lake rose to flood the area, or before a mudslide engulfed it. Such changes in water level have preserved many parts of the Danish coast. Some Alpine lake sites suffered a similar fate. This land is waterlogged, as opposed to wetland, although often the one shades into the other. There are also isolated, artificial wet areas, such as wells, pits and ditches. These occur mostly in towns or cities. They may preserve valuable evidence of the past environment and organic artefacts like those from Jorvik, Viking York, where archaeologists traced what people ate from remains in ancient latrine pits and found deer antlers carved into pins, combs and needles. But in such areas you see only part of the evidence of how people lived. Natural wetlands can reveal a fuller range of human activity.
The marsh and lake dwellers of ancient times lived normal lives. They were not restricted by their watery environment, possessing tools and weapons, clothes and ornaments just like their counterparts on dry land. The evidence from many prehistoric wetland villages suggests that they relied mostly on food from dry areas rather than food from rivers and the sea. The neolithic settlement of Charavines, on Lake Paladru in eastern France, which Aime Bocquet and his colleagues from National Centre for Underwater Archaeology in France, have excavated, demonstrates this well. They found collapsed structures built mainly of silver fir, charcoal from beechwood fuel, tool-handles from maple, ash and holly, and boxwood combs and spoons. Food remains include barley, peas, poppy and flax, three species of wheat and a host of wild plants including apples, hazelnuts, beechnuts and sloes, some probably collected as medicines or dyes. There were animal bones: these were poorly preserved, but they included cattle and deer. Along with the reeds for thatch, a few fish bones provide the only direct evidence for the exploitation of aquatic resources despite the fact that the settlements were on the side of a lake.
Wetlands hold evidence for transport and communica tions, witness the many trackways of the Somerset Levels and of Lower Saxony-where archaeologists also found wooden wheels and parts of waggons or carts. They have excavated dugouts from river beds, like the recent discovery from Humberside. Enthusiastic amateurs from sub-aqua clubs have also found the wrecks of larger ships.
There is also evidence for the ritual side of life, and death. The discovery of bodies shows how well bogs can preserve them. Many of the prehistoric examples from northern Europe show signs of a violent end, similar to that of Lindow Man, who was hit over the head, stabbed and garrotted. In florida, archaeologists have discovered pools used for burial. Fort Center is a late prehistoric example, rich in zoomorphic wooden figurines, as well as human skeletons, which suggest elaborate burial rites. At Windover, also in florida, a pool held complete human brains dating from about 5000 to 6000 bc. (Radiocarbon dates are shown by bc).
The circumstances which brought Windover to public notice a few years ago are common to many wetland sites. They underline the dual threats of drainage and devel opment. Windover lies within a few kilometres of Cape Canaveral, in an area of rapid residential devel opment. Developers are converting every slough and marsh in sight to building plots. One developer—exceptionally, stopped work and called in archaeologists when land drainage exposed human bones-and the pool of human brains. Public interest in the discovery brought in enough money to fund the dig. But archaeologists realise that, as the florida wetlands disappear, other sites are being lost to the builders. Without federal or state money to back them, they cannot even monitor the loss to archaeology, let alone fund work on new discoveries.
Switzerland, on the other hand, does support research into its wetlands. The Swiss love their prehistoric lakeside villages~hildren learn about them in schools, much as Engiish children work on projects about the Tower of London. As a result of public pressure, archaeologists have the funds to carry Out all the necessary surveys and exca vations before a motorway route is started, for example. Swit zerland passed a law in 1961 that forced developers to add to their budgets the cost of archaeology from initial survey to publication of results if they come across remains during a development. The amount that Switzerland is prepared to pay for this work reveals how much we should be prepared to spend elsewhere. A budget of 12 million Swiss francs (£4.5 million) pays for a team of around 55 archae ologists to work on the bay of Hauterive, near Neuchatel, for seven years in an area of ancient lakeside villages. Motor way construction cannot proceed in that period. Time and money allow for a full investigation of the archaeological record. Other countries have similar needs for time and funds to realise the full potential of wetiand sites, but legis lation rarely backs the archaeologist.
The threat to wetlands is an old one. Archaeologists can trace the drainage of the Somerset Levels to the Roman period. It was probably underway at the same time in the Fens and, by the Middle Ages, farmers were “reclaiming” many British wetlands. In Britain and on the Continent, agricultural devel opments in the 19th century put a premium on drained organic soils, as in the Po valley of northern Italy. Here, farmers carted away ancient wetland settlements, the “terramare”, as fertiliser. But it was in Switzerland, where the lowering and control of lake levels exposed ancient settlements in the dry winter of 1854 that wetland archae ology began. Johann Aeppli, a local schoolteacher, recognised the sigmficance of the remains. He drew the attention of Ferdinand Keller, president of the Antiquities Association of Zurich, to his new discovery. Wetland archaeology had arrived, and the following years saw the investigation of sites in the Alpine lakes that were to become classics of continental archaeology-in Italy, France and Germany, as well as in Switzerland. Keller’s publications influenced work abroad, from that of Arthur Bulleid at Glastonbury in Somerset to Frank Cushing’s explorations at Key Marco in Florida.
What is new, then, is not the threat but the scale of the threat. Ireland’s peat, which constitutes the bulk of Ireland’s wetlands, will disappear by the end of the century if the present rate of extraction continues. Drainage in Florida threatens even the Everglades. In north Germany only a 3 few scraps of relic bog remain of the original vast peatlands.
Also new is the level of information we can extract from wetland evidence. We can analyse lake sediments and peats, identily their microscopic and macroscopic animals and plants, and use our findings to reconstruct the varied environments, dry as well as wet, that the local inhabitants could exploit. We can trace how people affected their environment, and plot the relative sequences of events, such as the classic cycle of clearance, farming, abandonment and regeneration. We can date all these events precisely by dendrochronology, that is, by measuring the growth of rings on wood, particularly oak, and matching them to a sequence of rings of known date. These precise dates represent a completely new development in archaeology, and are likely to revolutionise the study of prehistoric Europe more profoundly than radiocarbon dating has done.
Once we know what the environment was like, we can recover an immensely rich record of human activity from a wetland settlement. Immediately apparent is the large amount of objects made from organic materials, such as maplewood bowls, ashwood axe-handles, fine basketwork and linen textiles. Many of these need expensive conser vation treatment to survive exposure to air. Even the inor ganic objects, such as flint and pottery, may hold more infor mation than on a dryland site: archaeologists can detect blood on a flint blade and food in a cooking pot, such as the fish stew in a vessel 6000 years old from the coastal site of Tybrind Vig in Denmark.
Wet sites also provide evidence for the exploitation of woodlands, and the skills of the woodworker. To study the ways in which ancient people used the woodland around them, an archaeologist has first to reconstruct the composition of the forest. What species of tree and shrub were growing, and in what quantity? Wet sites preserve two relevant categories of evidence, pollen and wood, and close cooperation between archaeology, palynology-the study of ancient pollen~and dendrochronology make the most of information from these sources. Pollen remains provide an outline history of local vegetation. Analysis of the wood gives direct evidence of human exploitation. Studies of the neolithic settlement at Charavines show how we can reconstruct a detailed and dynamic picture of local forests and their exploitation.
Analysis of pollen from the sediments below the settlement reveals, which trees covered the catchment area over six or seven centuries. Silver fir was the main species, followed by beech. Periodic dips in the percentage of tree pollen, some times accompanied by the presence of small quantities of flax or cereal pollen, suggested that people were clearing small patched of forests to grow crops. Interestingly, archaeologists found no tools or human remains. Perhaps the farmers lived on dry land where their settlements were soon obliterated, like countless others. Pollen at the level of the earliest settlement of Charavines shows a marked decline in the amount of silver fir in the forest, from about 50 per cent to a low of 5 per cent. Surviving wood from the contemporary settlement, distinguished from the mass of posts by dendro chronology, consists mainly of silver fir: the farmers were felling selectively to build their houses.
A short while later, they felled most of the nearby forest, reducing the amount of tree pollen. The presence of cereal pollen at the same level suggests farmers grew crops in the clearings. Analysis of three rings indicates that the inhabitants rebuilt their houses several times, and that within three decades of its foundation the people had abandoned the small lakeside settlement.
In the forest that grew over the vacant fields, beech became as common as silver fir. Alder and hazel were more abundant than before. For two or three decades the trees grew appar ently undisturbed. At the lake edge the ruins of the first settlement silted over. Then people came once more to Char avines and built a large house from silver fir, and various other structures of elm, ash, willow and alder. The mixture of species-identified by archaeologists from the posts of the second settlement-reflects an overall fall in tree pollen and a renewed rise in cereal pollen. The only major discrepancy occurs with beech. The pollen count reveals a decline in the number of these trees slightly greater than for silver fir, even though people apparently did not use beech to build houses. On the other hand, identification of charcoal from contem porary sediments indicates that people favoured beechwood as a fuel. In this instance, analysis of macroscopic remains gives detail to the general picture of how people exploited the forest, which pollen analysis reveals.
The second settlement at Charavines was as short lived as the first. Within 30 years, forest and lake had reclaimed the areas used briefly by a small group of people. The lake buried pollen and wood alike and sealed 4hem in a waterlogged environment until their discovery in the early 20th century.
We do not know the age of prehistoric Charavines precisely. We can work out which wood remains were the earliest on site and which came later but we do not have a dated sequence of silver fir tree rings against which we can match the pieces of fif found on the site. Radiocarbon dating places the first settlement at around 5000 years ago. The floating silver fir chronology does indicate that the inhab itants abandoned the second settlement within a century of founding the first.
At about the same time, the people living around the Somerset Levels were perhaps deliberately managing their woodlands. At Charavines, and indeed earlier in the levels, people selected particular species and types of growth from the forest. But they do not seem to have deliberately manipulated the growth of trees. For example, the long straight poles of ash and hazel from the Sweet Track, a neolithic pathway built across the marshes of the Somerset Levels about 6000 years ago, may have come by chance, from secondary growth on tree stumps in abandoned clearings. But in the Levels, from about 5000 years ago, there is a remarkable series of woven panels that show increasing mastery of cop picing and hurdle-making.
Soon after this, people were harvesting thousands of long straight rods of hazel, up to 2.5 metres long and 14 to 16 millimetres, in diameter, to make the panels of the Walton Heath track. More than 5000 rods went into a stretch of track 60 metres long. The hurdles were bedded into soft wet peat which preserved many hundreds of rods in excellent condition. There are abundant signs of growth from a coppiced stool-the stump of the mature tree trunk, cut down to encourage young growth. There are heels at the base of a rod where it was cut or pulled from the stool, long and narrow stems without side branches, shooting straight up in the competition for light. They grow strongly because an established stool has a disproportionately large root system to feed the young shoots. But it is concentration on one species, hazel, that points most strongly to deliberate management of woodlands. Less than 15 per cent of the wood we identified from the Walton Heath trackway came from species other than hazel.
Once established in the Somerset Levels, coppicing persis ted. There may have been an unbroken tradition from its prehistoric beginning to the coppiced woodlands recorded in the Domesday Book and during later centuries. Other wetland sites throughout prehistoric Europe show that coppicing was common, and that early farmers exploited and controlled the growth of woodlands as did their medieval descendants.
Analysing the age of wooden remains has proved fruitful elsewhere, as in the neolithic settlement of Hornstadt-Hornle, which lies on the shores of the Bodensee in south Germany. Excavation by Helmut Schliterle and his colleagues from the Bodensee-Oberschwaben project has in recent years yielded a wealth of archaeological data, with exciting results from the analyses of tree rings by Andre’ Billamboz from the same group. This makes it possible to build up a detailed picture of the exploitation of oak from the local woods.
Oak was one of the kinds of trees felled for house-building, and south Germany has an oak master-chronology stretching back several millennia. So archaeologists could establish absolute dates for the second neolithic settlement at Hornstadt-Hornle from the many surviving oak uprights. The settlement was founded in 3586 BC and lasted for a century. Analysis of the uprights enabled Billamboz to isolate individual houses and to show the age and size of tree that people used during each building episode. During the first period, from 3586 to 3582 BC, housebuilders used oak 80 to 90 years old, and 30 to 35 centimetres in diameter. Builders split each felled trunk into six or more hefty radial planks to build a structure about 8 ~ 5 metres, in ground area. From 3570 to 3562 BC, people rebuilt the structure, using generally younger and slighter trees, 50 to 60 years old and under 30 centimetres in diameter.
After a gap of 21 years, the third building phase, from 3541 to 3531 BC, saw the use of predominantly 40-year-old trees, 20 centimetres or less in diameter. They usually split these slight trunks in half and no further. The inhabitants built at least six structures. All the trees used for these houses had begun to grow after the first house was built. Some had grown relatively quickly, reaching the same diameter as trees of the first and second phase in less than half the time, in 25 years as opposed to 75 years in one case. This implies coppiced growth, perhaps starting from the stools of trees felled for the earlier building. Woodcutters felled oak which had also started to grow after the first phase. In these extra years, it had reached a diameter of about 35 centimetres, stimulated perhaps by the thinning between 3541 and 3531 BC. The people had not exhausted the supply of mature oak, however. In 3508 and 3507 BC the builders felled oaks that were over a century old, which had been small trees when people built the first house some 80 years earlier. At felling, they were as big or bigger than the trees used in the following phases. It is not clear why the people changed their pattern of exploitation, switching from older to younger trees, but it was probably a matter of choice rather than coping with dwindling forest resources.
The Sweet Track in Somerset may be contem porary with Hornstadt Hornle. We have no precise date as yet for the wood used to build it, but the analyses of Ruth Morgan, from the University of Sheffield, of the wood that forms the tracks add greatly to our under standing of neolithic carpen try. Along with deliberately or fortuitously coppiced wood, massive trees provided planks for the track’s walk way. These trees were up to 400 years old and a metre in diameter. They were felled and split by people who had no bronze or iron tools, only stone axes and wooden wedges and mallets. Oak, ash and lime were the main species for planl:-oak was the most common. The woodworkers split them into planks which could be 40 centimetres wide, 3 metres long and less than 5 centimetres thick, a triumph of their art. They generally split large trunks into radial planks. We have found these as the most common type all along the 1800-metre track, except at its southern end where these are also transverse planks. The rings in the planks show that the transverse planks were split from relatively young trees, only about 120 years old. In order to get planks which were as wide as those from the 400-year-old trees out of these smaller trunks, the wood workers had to split them across the grain to use the full diameter of the trunk.
A number of the planks in the Sweet Track have a hole cut though one end. People axed across the grain at either end of where the hole was to be, and then wedged out the surplus wood in between. They worked the hole from both sides of the plank, so that it was hourglass-shaped in section. Holes in planks raise the possibility that the wood workers used mortise and tenon joinery, but no one has observed anything of this sort along the Sweet Track. Instead, it would seem that the holes were cut so that a person could tie a rope securely to the heavy planks to pull or tow them down to where they were needed in the reeds. However, Jean~Pierre Petrequin reports that he has found a rudimentary form of mortise and tenon joinery from Clairvaux, a late neolithic settlement near Charavines in eastern France. You cannot deduce any of these details from dryland evidence: a hole in a plank exists only by virtue of the solid wood that suffounds it.
Appreciation of the returns from wetland archaeology shows in the growing number of teams specialising in this field. In England, where the work has been largely funded by the Historic Buildings and Monuments Commission, the Somerset Levels Project started in 1973. The Fenland Research Committee followed in 1981, and is carrying out major surveys in several counties, as well as excavations that include the unique sites of the waterlogged neolithic long barrow at Haddenham, complete with timber mortuary chamber, and the large Bronze Age lake settlement at flag Fen, where rectangular wooden houses sat on an artificial island of timber. The Northwest Wetlands Project began preliminary work in environmental archaeology this year in Merseyside, Greater Manchester, Cumbria and Shropshire, as has the Severn Estuary Levels Committee in Wales. The Severn group will be hard at it in face of the projected Severn barrage. In Scotland, research into ancient lake or bog dwell ings, or crannogs, leads the wetland field, while in Ireland archaeologists are making brave attempts to record trackways from the vanishing bogs.
In 1986, a grant from the University of Exeter Research Fund, which raises money from conferences, helped to set up the Wetland Archaeology Research Project (WARP), which reflected the growing involvement of archaeologists in wetlands. WARP exists to help archaeologists specialising in wetlands to keep in touch with each other through a news letter, whether about simple matters, such as digging tools, or about the refinements of remote sensing. After a year, WARP already ha,s more than 200 associates throughout Europe, North and South America, Austraiasia and Japan. I hope the network will soon cover all threatened wetlands, for past experience suggests that every river, lake, estuary and bog in the habitable world was exploited by someone at some time in the past. In a wet site there is every chance that evidence of that exploitation survives.
