TEie Celtic world from the sixth century BC inherited a very long and sophisticated tradition of woodworking, in continental Europe and Britain going back to at least the beginning of the fourth millennium. The recovery of this fundamental episode in the history of technology has been a product of the exploitation over the last three decades of the potentialities of what has become known as wetland archaeology, where waterlogged conditions provide a medium for the preservation of wood, otherwise a fugitive natural substance save in exceptional circumstances such as extreme aridity or permafrost. An early example of wetland investigation was of course that of the Swiss lakeside sites first revealed m the 1850s and increasingly with the lowering of water levels in 1870-5; the great iron age site of La Tene was first discovered in 1858. In Britain the Glastonbury and Meare wetland iron age sites were excavated from 1892 (Coles and Coles 1986), but the co-ordinated work of archaeologists and natural scientists on the Somerset peat bogs and their contained artefacts, which has given us such astonishing new information, dates from the early 1960s, supplemented by similar projects in the Cambridgeshire Fens. We can now talk with some confidence about the salient facts of prehistoric European woodworking shared by the Celts before the Roman period.
All woodworking demands one or other of two types of timber, heavy or light (Coles et al. 1978; Piggott 1983: 16-21, 28 with refs). Traditionally, from medieval Latin to modern English, ‘timber’ denotes heavy stuff for posts and beams, or split into planks of appropriate thicknesses; ‘wood’ small-diameter poles which can themselves be split for hurdling or basketry. Already by the early fourth millennium BC in neolithic Britain there is evidence of conscious woodland management both in the selection of trees (usually oaks) for heavy timber posts up to a metre in diameter, log-boats from halved hollowed stems, and planks split or riven ‘on the chord’ and along the natural medullary rays for, among other things, single-piece or composite disc wheels and plank-built boats. More surprising is the neolithic achievement of the techniques of coppicing and pollarding for the controlled production of pole wood and its use when split for hurdling, the wattle element in wattle-and-daub walling, and basket-making. All this was the product of stone tools, to be replaced first by a bronze and in the Celtic world by an iron toolkit of axe and adze, chisels, gouges and knives. The saw was a late-comer to European prehistory, and from the Late
Bronze Age and the iron-working Celtic world usually small joiners’ tools, but the planks of the burial chamber of the Hohmichele, of sixth-century Hallstatt D date, over 6 m long and 0.35 m wide, appear to have been not split, but sawn by a large timber-yard saw. The other technological innovation, also sixth-century in the Celtic world of west Europe, was the adoption of the lathe for wood-turning, to which we will return later.
In the meantime, heavy timber posts and beams were extensively used in the Celtic world, especially in fortifications (Figure 5.3). The problems of felling and handling massive tree stems had been faced and mastered by neolithic societies building in Britain the ceremonial and burial monuments of henges and long barrows by the third millennium BC, when oaks of up to a metre in diameter were worked into posts up to an estimated 3.25 m in height and a weight of over 2,000 kg. Such timber could be split and hollowed into log-boats (Ellmens 1969; McGrail and Switsur 1975) as detailed in Chapter 15 (Figure 15.6), but Chapter 5 describes the vast amount of timber used in the facing and internal lacing of so many types of defensive ramparts of forts and the later oppida of the Continent, where timber-laced walls with lengths of 5 or 6 km in circuit must have drawn heavily on the natural resources of woodland. All types of wood, from heavy timber to brushwood, and including discarded worked pieces, were used to make up the artifical islands for settlement in lakes or marshes, as at Glastonbury and the crannogs of Ireland and Scotland, as well as piles which could be massive posts driven into the lake bed (Morrison 1985). In the Milton Loch crannog in Scotland the woodwork included the plough-head and stilt of an ard which gave a radiocarbon date coincident with a structural pile of c.450-500 BC (Guido 1974). An ard beam from 45 km away is of the first century BC and the type goes back to the second millennium in Italy and south Russia. Heavy timber again was needed for the single-piece or composite wheels of ox-wagons, as we shall see on pp. 324-6, on the craft of the Geltic wheelwright, but before this we may consider two minor forms of woodworking, figural sculpture and wooden containers, which may involve raw material of varying dimensions.
Such human representations as survive seem to fall into two classes (Coles 1990). The first comprises seven dated figures from Britain and Ireland, all naked and some ithyphallic, and ranging from the early second millennium to the fourth century BC. Those from the sixth-century could include the large Ballachulish female nearly 1.5 m high of 728-524 BC, the contemporary little figures holding round shields and standing in a boat from Roos Carr, and the 34 cm ithyphallic figure from Kingsteignston, of 426-352 BC. Later prehistoric figures from the Nordic world beyond the Celts provide vague parallels. The second class is formed by an enormous number of Gaulish or Gallo-Roman wooden votive figures from two sanctuary sites at sacred springs, 300 or so from Sources-de-la-Seine near Dijon, and up to about 5,000 at the twin springs at Chamalieres in the Massif Central. Some of the wood carvings are ex votos of human organs, but the complete human figures are clothed, and at Sources-de-la-Seine less romanized than at Chamalieres (Megaw and Megaw 1989: 172-3, with references). Animal sculpture survives from one site, a waterlogged well or shaft within a ritual enclosure of the Viereckschanze type at Fellbach-Schmiden near Stuttgart with a tree-felling date of 123 BC, where a spirited stag protome and an antithetical pair of rams or goats originally flanking a human figure show us how much Celtic art in fugitive substances we must have lost (Megaw and Megaw 1989: 162-3).
Small pole-wood obtained by coppicing was extensively used for building trackways across bogs and marshes from the early third millennium BC, and in later prehistory in the Somerset Levels and the Cambridgeshire fens and the Dutch bogs, and in late prehistory in north Germany, sometimes flanked by highly stylized human figures cut from planks. Planks were used for shields of long oval type from the third century BC at Hjortspring on the Danish island of Als and at La Tene (Rosenberg 1937; Vouga 1923; pi. xvi-xvill) and of course for arrows and spear-shafts, bows, and the handles of domestic tools from Neolithic times onwards. Woodwork was also freely used for vessels and containers of various types and it is here we must consider the early use of the lathe in prehistoric western Europe.
The earliest evidence for lathe-turning is in the ancient Orient and Aegean, and a lathe is technologically the application of the principle of the long-established bow-drill to a horizontal spindle driven into intermittent rotation by a wrapped-round cord or strap. This motion needed an assistant strap-puller or the return of the slack obtained by a springy upright in the pole lathe, common to most European peasant contexts, and in Britain surviving until modern times among, for instance, the chair-bodgers of the Chiltern beech woods. In prehistoric Europe the lathe must derive from Mediterranean sources, either Aegean or Etruscan. There is inferential evidence for the lathe finishing of bronze bowls in late bronze age central Europe, but general agreement puts the first real use of the wood-turner’s lathe to sixth-century Hallstatt craftsmen in south Germany (Piggott 1983; 162), and a rapid adoption of the pole lathe or its equivalent throughout the Celtic world. The true lathe with continuous rotation had to await the invention of the crankshaft around AD 1400 (White 1978: 305).
Despite their importance in the Celtic world, wooden containers for liquids or dry stuffs have scarcely been discussed by archaeologists and the only technical study seems to be that made on the Glastonbury material (Earwood 1988; 1993). Latheturning may play a part, but vessels hollowed from the solid by axe and adze, chisel gouge and knife are part of the neolithic ancestry of woodcraft: at the lakeside settlement at Chalain in the Jura, for instance, hand-carved wooden bowls copy the contemporary early third-millennium neolithic pottery styles. In late prehistory finds from Northern Ireland are remarkable feats of carving from the solid wood block: one such cauldron is over half a metre in diameter (Coles et al. 1978) and another, from Altatarte, Co. Monaghan, and 30 cm across, is decorated with incised La Tene ornament and has free ring-drop handles (e. g. Raftery 1931: fig. 256). Some of the bowls and platters from La Tene itself were certainly turned (Vouga 1923: pi. xxix) and in the third-century Hjortspring find were elegant little turned boxes with knobbed lids (Rosenberg 1937: fig. 34). From Glastonbury came the two well-known shallow flat-bottomed bowls with incised La Tene ornament, turned from solid blocks some 35-40 cm across. Decorative lathe-turned bases were incorporated in some British stave-built vessels, a class of construction of widespread employment in the Celtic world (Earwood 1993).
The stave-built hooped cask or barrel (Figure 8.3) to hold liquids was a Celtic invention, a northern barbarian counterpart to the Mediterranean amphora. Such a barrel is shown on the well-known Gallo-Roman relief from Langres, a huge cask, perhaps a couple of metres long, mounted on a mule drawn-wagon and usually taken to indicate the bulk transport of wine (Eydoux 1962: fig. 232). Such casks have been found reused as well linings in Roman Gaul and Britain (e. g. at Silchester). A series of late iron age stave-built vessels, often with decorative bronze bindings, occur in the Celtic world of Britain and the Continent. The great straight-sided stave-built vessel from Marlborough in Wiltshire, 60 cm in diameter and height (Fox called it a ‘vat’) has elaborate repousse bronze decorative hoops and panels, and may be a Gaulish import, but unquestionably insular are stave-built tankards, banded or sheathed in bronze (e. g. Trawsfynydd and Shapwick), with decorative lathe-turned bases (Fox 1958: 68-70, 108-10). A distinctive series of stave-built buckets with hoop handles and decorative bronze mountings, in which three or four staves are prolonged into feet, are known from Belgic Britain (e. g. Aylesford and Baldock) and the Continent (Fuxembourg) (Megaw and Megaw 1989: 184-7). Fragments of undecorated vessels of this type come from Glastonbury. From this site too come fragments of the walls of boxes made from very thin (3-5 mm) split and adzed strips of wood bent into a cylinder up to 15 cm in diameter and with incised geometric ornament. The technique was known in bronze age Denmark.
An important branch of wood technology in the Celtic world was that of the wheelwright and builder of four-wheeled wagons and two-wheeled carts or their specialized versions, chariots for parade and war. The writer surveyed the subject on a European basis twelve years ago and, where not otherwise documented, detailed references will be found in this source (Piggott 1983). It should be noted at the outset that our knowledge of Celtic vehicles from the sixth century BC to Roman times is very unequal and largely consists of wheels alone. The four-wheeled carriages from the Hallstatt D princely graves in the sixth century can be reconstructed in some detail (Pare 1992) from their elaborate bronze or iron sheathing and ornamentation, which allowed full-sized replicas to be built (Barth 1987), whereas our factual knowledge of the Celtic chariot, despite recent excavations of chariot graves in Belgium (Piggott 1983: 205) and Yorkshire (Stead 1991) depends on soil stains of the basic plans, and assumptions based on two or three representations on coins or sculpture (Piggott 1983: figs 127-9).
The earliest wheeled vehicles, from 3000 BC, were ox-drawn wagons and carts with simple disc wheels, either one-piece or of tripartite plank construction. Both demand adze and chisel work of heavy timber split on the chord: a single-piece, 90 cm diameter wheel, as in the Dutch Neolithic, would need a plank of 3 cubic metres weighing 322 kg. The wheel of three dowelled planks, the central twice the width of the lateral members, was early achieved and lasted not only throughout prehistory but until the 1950s in rural Ireland (Piggott 1983: 19-26).
In the Celtic world tripartite disc wheels presumably associated with heavy ox-drawn vehicles occur, as in the Nordic world beyond the Celts (Piggott 1983: 197-9; Schvosbo 1987), as early as 490-430 BC at Doogarymore in Ireland, 530 BC in Nordic Dystrup. Celtic examples continue on the Continent (200 BC at Mechanich-Antweiler, second-first century BC at Ezinge in Holland).
A novel type of vehicle, of light construction and drawing on small timber and the produce of coppicing, is that associated with paired horse draught, having four, but more often two, spoked wheels, when it takes the form of the chariot for parade, hunting or warfare. Such chariots were highly developed in the Near East from early in the second millennium BC, and their counterparts are known in Europe in the Urals and the Carpathians from the middle of the millennium. Immediately antecedent to our Celtic period are the Hallstatt C (seventh-century) graves of central Europe such as those in the Grosseibstat cemetery (both wagons and a chariot) or the Czechoslovakian wagon graves such as the Elradenin cemetery. It is, however, in Hallstatt D, from 600 BC, that the great series of wagon or carriage burials belong, to be replaced from the fifth century by chariot burials (Piggott 1983: 138-70; 1992 k
The Hallstatt D graves with four spoked-wheeled wagons or carriages (Figure 26.8) which have been excavated number nearly 250, distributed from the Alps to Alsace, from Burgundy to Bohemia (Pare 1992; Barth 1987); the timber burial chambers have often been plundered in antiquity, but occasionally have survived, as at Vix or the staggering intact burial at Hochdorf (Biel 1985). The vehicles, many with elaborate bronze or iron sheathing and decoration on the basic woodwork, have a surprising uniformity in their structure and even dimensions. The axle-trees have a separation (wheelbase) averaging about 1.8 m, and a wheel track or gauge of about
1.3 m. This average track goes back to the earliest ox-wagon of the third millennium BC in south Russia, continues in the La Tene chariots and survived to determine the ‘standard gauge’ of 4 ft 8'A in for English railways and their continental derivatives of the nineteenth century (Piggott 1992). The Hallstatt D axles were joined by a Y-shaped perch and there was an articulated draught pole for a pair of horses under a yoke in the manner of oxen, a system of harnessing replaced in western Europe by collar and shafts only in the early Middle Ages.
The Hallstatt undercarriage supported a shallow and narrow body, only about 10 cm deep, with bronze-decorated sides and something under a metre wide. The long (40-45 cm) wheel naves allowed for a pivoted front axle with a quarter-lock of 20° and a turning circle of about 8.5 m, but most examples seem to have had a pair of fixed axles, as in all four-wheeled vehicles up to the later fourteenth century AD.
The spoked wheels used on horse-drawn vehicles in the Celtic world demand a fuller discussion. As we saw, the spoked wheel as opposed to the single or tripartite disc is an invention of the second millennium BC in the ancient Orient and eastern Europe as a part of the creation of a lightly built vehicle suitable to paired equid traction at high speeds compared to the lumbering ox-cart. Already by Hallstatt C in the seventh century such wheels were made with nail-studded iron tyres on wooden felloes which could be of complex construction incorporating both bent wood and segmental elements with elaborate iron clasps. By 600 and the beginning of Hallstatt D, wheel types had become simpler, with a single-piece bent felloe around 80 cm diameter, made from a coppiced pole 6-7 cm in diameter, over 3 m long, heat-bent into a true circle with an overlapping scarf joint, a type of construction in the Celtic world until Roman times. Iron hoop-tyres became a constant feature, the nails diminishing in number until about the second century BC, when the shrunk-on tyre with no nails was achieved {Piggott 1983: 216).
All vehicles, from the earliest disc wheels through to spoked wheels of all types, had wheels rotating freely of their axles, and held in place by linchpins of bronze or iron often made into decorative features. The long cylindrical naves of the sixth-century Hallstatt wheels were frequently completely sheathed in bronze or (as at Hochdorf) iron. One vehicle, chat from Vix of just before 500 BC, has bronze-sheathed naves in an elegant double-curved classical moulding, the cyma recta, which, without sheathing, became standard in La Tene, and here we must look for ultimate origins in the western Greek world (Piggott 1983: 165, 214). All these naves imply lathe-turning on a massive scale, initial shaping with adze and chisel, as the unfinished nave from Glastonbury shows (Earwood 1988: 89; 1993).
The question of wooden horse-yokes for paired draught is a little obscure. In the seventh-century Hallstatt C there was a short-lived fashion for very elaborate bronze-studded, leather-covered wooden yokes, but in the following century they were unknown. The maplewood yoke from Hochdorf is a simple affair with bronze bands and a pair of little cast bronze horse figures. The finely carved yoke from La Tene and simpler versions from Ezinge and unassociated bog finds have all been classed by Fenton as ox-yokes but an equine use is not excluded for some, at least (Piggott 1983: 218).
The body of the ‘Celtic’ chariot above its axles is, as we have seen, wholly unknown except for inferential double-hooped screens of some material suggested by representations on the third-century BC Padua reliefs and Roman coins of 50-60 BC.
Figure i8.i Reconstruction model of a chariot, based on fittings from Lynn Cerrig Bach, Anglesey. By courtesy of the National Museum of Wales.
World History
