17
Exchanges with the wider world

That the British Isles were in contact with Continental Europe throughout the Iron Age is evident from many aspects of the material culture and from occasional anecdotes recorded in the classical literature. At times there may have been organized trade with well-defined procedures and recognized rates of exchange, but there were many other ways in which exotic material and ideas found their way into Britain, ranging from folk movements to systems of gift exchange between élites. Archaeology can seldom distinguish precisely between the various possible systems. In the discussion to follow, the scope of the evidence is laid out, roughly in chronological order, and, where it is thought to be appropriate, some of the wider implications of the material are discussed.

Britain, by virtue of its geographical position, has always been receptive of influences from two directions: from central and northern Europe across the North Sea (though with little direct Scandinavian influence (Thrane 1975, 233–5)); and from the Mediterranean and Iberia along the Atlantic coastal routes to south-west Britain and the Irish Sea province (Cunliffe 2001). Both spheres of contact were maintained throughout the Iron Age, but superimposed on these broad patterns were a variety of processes which influenced the intensity of contact and thus the volume of material which passed through the exchange systems. The reasons for these variations are difficult to isolate: they may have involved political policies which deliberately encouraged or hindered relationships, or social factors such as the development of conserving or inward- looking attitudes in society resulting in a lack of motivation to engage in commerce or social intercourse. Changing economic pressures among Continental communities would also have contributed to the complex situation. Occasionally these separate factors can be isolated with some degree of certainty but frequently they remain ill-defined and confused.

Viewed against the longue durée the beginning of the Iron Age in the eighth century BC marked a point of significant reorientation in the relationship between Britain and Europe. Until this time the Atlantic seaways had been the dominant axis of contact linking western areas of the British Isles to the coastal regions of western France and Iberia in a complex network of interactions fuelled, most probably, by exchange in metals – copper, tin and gold – with which the ocean-facing zone was well provided. Although the Atlantic seaways continued to be active in the following centuries, from the eighth century BC onwards the networks which linked Britain to middle and northern Europe seem to have become the dominant interface. Why this change took place is not easy to say. No doubt many factors were involved. Among them would have been the establishment of the Phoenician port-of-trade at Cadiz around 800 BC and the Phoenician take-over of the southern part of the Atlantic system. Another factor is the growth in power of the Hallstatt élite in west central Europe at about the same time and the new demands which this may have put on long-established exchange networks. And bound up with these new dynamics was the rapid development in the use of iron as a more effective material for making tools and weapons. The increasing use of iron rendered less essential consistent access to the copper and tin supplies of the west.

It will be convenient, in the discussion to follow, to deal first with the newly developing contacts with middle and northern Europe before considering the effects of these changes on the Atlantic system.

Contacts with middle Europe: eighth and seventh centuries (Figure 17.1)

The beginning of what can reasonably be called the Iron Age in Europe is linked to the development of the Hallstatt culture, characterized by the appearance of a horse-riding aristocracy using a long slashing sword and frequently burying their dead in timber-built tombs beneath barrows. Some of the burials were very rich, containing trappings belonging to the chieftain’s horses as well as a cart upon which the body may have been transported to the grave, along with weapons and selected items of personal equipment. Rich burials of this type cluster in the Czech Republic and southern Germany but are known sporadically across most of central Europe, spreading into central and southern Belgium. What their sudden appearance means is not immediately clear. Some writers proposed an invasion of an eastern aristocracy, ultimately from the Steppes, spreading west and gaining overlordship of the Late Urnfield communities. In support of such a view, close parallels are drawn between the harness-fittings and burial rites in the two areas; moreover, the rapid spread of the burials over Europe would be consistent with the concept of a mobile warrior aristocracy. Against this view, however, it can be argued that the characteristic long sword of the Hallstatt warriors owes nothing to eastern Cimmerian influence (Cowen 1967) and is best seen as the invention of the Late Urnfield bronze-smiths in response to cavalry warfare. Similarly, there is nothing new about rich burial rites involving the use of vehicles. Earlier examples are known in central and northern Europe but only incompletely, since the cremation ritual has destroyed much of the evidence. Extending these arguments, a more convincing model can be constructed which supposes that the Hallstatt aristocracy is simply an indigenous élite adopting the rite of inhumation and making an increased use of the horse as a cavalry animal. The arguments for and against these two views cannot be paraded in full here, but the majority of scholars now favour the latter view.

In Europe the chronology of the period is now tolerably well known. In the central region of southern Germany archaeologically imposed terminology recognizes the change from the Late Urnfield period, Hallstatt B2/3, to Hallstatt C as the transition to the Iron Age. Detailed typological studies, together with precise dates derived from dendrochronology, now place this change at around 800 BC and distinguish two phases in the earlier Hallstatt period. The first (Ha C1a), dating to 800–730/20 BC, is characterized by wagon burials and a distinctive bronze sword named after the site of Gündlingen, the form of which is rooted in indigenous traditions. The second phase (Ha C1b), dating to 730/20–620 BC, is characterized by the iron Mindelheim sword. The later Ha C2 phase (Pare 1998) is usually taken as lasting up to about 600 BC. In terms of the British bronze traditions (above, pp. 75–82) Ha C1a roughly corresponds with the period of overlap between the Ewart Park assemblage and the Llyn Fawr assemblage in the period 800–750. Since there is no sound reason to argue for there being any significant time lag between the first appearance of artefacts in west central Europe and their arrival in Britain we can allow that Gündlingen bronze swords and horse gear may have begun to reach Britain in the first half of the eighth century BC.

Figure 17.1 Distribution of selected types of Hallstatt metalwork (sources: various including Hawkes and Smith 1957 with additions).

The Gündlingen sword and its characteristic winged chapes are found widely distributed in north-western Europe including the British Isles and Ireland. In a detailed study, Cowen (1967) has distinguished four broad classes of Gündlingen sword on the basis of their hilt form (Figure 17.2). Of his classes a and b, both thought to be Continental products, twelve are known from Britain – eight of them from the Thames area; but with the possible exception of the sword from Ebberston, Yorks., none is from a closed archaeological context. Once established in Britain, the new weapon-type was closely copied by the local swordsmiths, giving rise to Cowen’s type c and d, of which twenty-seven recognizable examples are known in Britain and two from the Continent. Cowen has argued that in parallel with this development, various composite and hybrid forms were manufactured locally as native inventiveness took over, culminating in the evolution of the Thames type of sword, which represents the eventual dominance of the native traditions over the foreign. The type proved to be popular not only in southern Britain but also in north-western Europe, where some fifteen exported examples have been recognized. Although it has been suggested that the Thames type, rather than being derived from the Gündlingen prototype, was in fact its predecessor (Schauer 1972), the British evidence argues in favour of Cohen’s view.

The long sword was kept in a sheath of wood or leather terminating in a metal chape, the sides of which were splayed out into wings so formed, it is suggested, to enable a mounted warrior to keep the end of the sheath steady with his left foot while drawing the sword with his right hand. The chapes, like the swords, are characteristic of Continental burials and occur sporadically in Britain; there are eight known from England and ten from Ireland (Figure 17.3).

The occurrence of the swords and chapes in Britain (Figure 17.4) requires explanation. Clearly a number of weapons must have been brought in from the Continent, but conversely, once British manufacturing centres had developed, home-produced varieties were exported. Without special pleading, simple exchange mechanisms could explain the entire pattern. If, as was once argued, the swords had accompanied an immigrant warrior élite one would have expected them to be buried with the dead in the Continental manner, but this is not so; the majority were found in rivers where, like their Late Bronze Age predecessors, they had been deliberately deposited. No certain burial find is known in Britain – with the possible exception of the ill-recorded discovery at Ebberston in Yorkshire, where two swords and a chape were found together with a quantity of human bones. Even if this is accepted as a typical Hallstatt C burial in Continental style, it need not necessarily represent a foreign warrior aristocrat. Hallstatt swords are therefore most likely to have been worn by British warriors and disposed of in a manner consistent with traditions which had been in operation in Britain for centuries. The appearance of these swords is best explained in terms of a continuation of the existing exchange systems, linking Britain to northern Europe, which had been in operation for centuries.

The eighth century saw the importation of a wide range of harness- and cart-fittings from northern and central Europe and their eventual deposition here in hoards, together with locally produced tools and weapons and sometimes imports from western Europe (Figure 17.1). Hoards of this kind are therefore of considerable significance in establishing synchronisms between local industry and datable foreign products. While it must be admitted that close dating is extremely difficult, the hoards can be divided into a loose sequence stretching throughout the eighth and seventh centuries. At the beginning of the sequence must be placed the famous collection of material found beneath a layer of stalagmite on the floor of the cave at Heathery Burn, Co. Durham. In addition to socketed spears, socketed axes, knives, gouges, awls, a bifid razor and a range of bone and stone artefacts, a group of bronze cart- and harness-fittings was recovered (Figure 17.5). The cart-fittings consisted of eight bronze bands, presumably for binding the ends of the axles of a wheeled vehicle. Similar objects are recorded in Late Urnfield contexts in central Europe dating to the ninth to eighth centuries. The harnesses are now represented by a bronze strap-distributor (i.e. a disc with perforated vertical sides to allow leather straps to cross at right angles), a ribbed disc with attachment loops, two large disc-shaped mountings and a group of bronze rings of various diameters. In addition, two cheek-pieces of bone were found. The bronzes are of the Ewart Park phase, and further dating evidence is provided by a bronze bucket (Figure 17.6) of a type in use in the eighth century – a date which corresponds well with the Ultimate Urnfield character of the material.

Figure 17.2 Hallstatt swords: 1 Brentford, Middx. (type a); 2 Henley, Oxon. (type b); 3 Newcastle upon Tyne (type c); 4 Cambridge (type d) (source: Cowen 1967).

Figure 17.3 Hallstatt C chapes (sources: Ebberston, Burgess 1969b; Brentford, Wheeler 1929; Thames, R.A. Smith 1925).

The strap-distributor and the looped disc can be closely paralleled at two other hoards from Welby, Leics. (Figure 17.5), and Parc y Meirch, Denbigh. (Figure 17.7), which together with the Horsehope hoard from Peeblesshire (Figure 17.5) constitute a closely similar group of horse equipment hoards broadly datable to the eighth century. At Welby five strap-distributors were found, together with a perforated looped disc and a double-looped harness-fitting identical to those from the other two hoards. Other central European types at Welby include two T-shaped handle attachments for a bowl and possibly the small carinated bronze cup with furrowed decoration on the shoulder (Figure 5.2), of a type which is frequently reproduced in pottery in southern England. The local bronze industry was represented by socketed axes, a spear and a sword.

The Parc y Meirch hoard is altogether more substantial, containing some ninety individual pieces, among which is a group of harness-trappings including double-looped fittings and strap- distributors as well as rattle-pendants composed of plaques of bronze joined loosely together with bronze rings (Figure 17.7). This type of bridle decoration is represented in Scandinavia in the eighth and seventh centuries, and a slightly different form of the same idea appears in one of the south Belgian burials at Court-St-Etienne in the early part of the seventh century. In origin the type is probably French (Savory 1976a, 44). In the Horsehope hoard a range of harness-rings occurs, including the characteristic double-looped ring, but in addition C.M. Piggott (1955) has recognized several cart-fittings which include a dish-shaped mounting and ribbed discs, the latter possibly serving as decorative attachments for the axle caps of a small (or model) vehicle. The closest parallels for these pieces are among the Hallstatt cart burials of the Czech Republic, dating to the eighth century.

Figure 17.4 Distribution of Hallstatt swords and chapes, (source: Cowen 1967 with additions).

Figure 17.5 Selection of Hallstatt horse-trappings and cart-fittings: 1 Newark-on-Trent, Notts.; 2, 9, 10 Llyn Fawr, Glam.; 3, 4, 8, 12 Heathery Burn, Co. Durham; 5, 7 Welby, Leics.; 6, 11 Horsehope, Peebles. (sources: 1, 3, 4, 8, 12 Inv. Arch. GB; 2, 9, 10 Grimes 1951; 5, 7 Powell 1950; 6, 11 S. Piggott 1955).

Figure 17.6 Bronze buckets (source: Hawkes and Smith 1957).

Of the later, seventh-century, hoards, by far the most significant is the collection of material recovered from the bottom of an ancient lake at Llyn Fawr in Glamorganshire. Local products are well represented by seven socketed bronze axes, four of them of a distinctive Welsh type, three socketed sickles, two of bronze and one of iron, and three socketed chisels. It is possible that the socketed iron spearhead recovered was also of local manufacture, but the type is simple and could equally well have been imported from the Continent, where identical forms were in use. Clearly, the presence of a local type of sickle made in iron implies that iron extraction was by now under way, if only on a small scale. The contribution of the Atlantic networks is represented by two bronze cauldrons of class B1 (Figure 17.9) – a type which first came into use in the first half of the seventh century or a little before. North European Hallstatt C types (Figure 17.5) include an iron sword with hilt plates of bone, two bronze cheek-pieces, three bronze discs or phalerae from harness decorations, an openwork harness-mount, a belt hook and a crescentic razor. All the types can be very closely paralleled in early Hallstatt C burials in southern Germany and Belgium, particularly at Court-St-Etienne. So close are the parallels that direct importation from the Continent is the only reasonable explanation for the presence of these objects in western Britain. Precise dating for the deposition of the hoard is, of course, difficult since there is no way of assessing how long the individual items had been hoarded, nor indeed is there any certainty that all were deposited together, but the fact that all appear to be broadly contemporary and identical to Continental types hints at a date not far removed from the middle of the seventh century. By this time, then, iron production was established.

Figure 17.7 Part of the hoard from Parc y Meirch, Denbigh. (photograph: National Museum of Wales).

A second hoard to show an association between Late Bronze Age types and iron comes from Sompting, Sussex, where the remains of a class B2 cauldron (Figure 17.8) were recovered, together with seventeen axes and a Hallstatt phalera of hollow conical form. To one of the axes adhered a mass of corroded iron. On the basis of the cauldron, a seventh-century date would seem to be most likely, but a later date in the sixth century cannot be completely ruled out.

Finally, dredging of the river Avon at Melksham, Wilts., produced two iron spearheads, three bronze socketed spearheads, the blade of a dirk and three bronze phalera (Gingell 1979; Osgood 1995). It is possible that the items were once part of a single deposit since all belong to the seventh century.

Several other but less closely datable discoveries of harness- and cart-fittings have come to light. Phalerae of Hallstatt type have been recovered from the Thames at Brentford, Middx., with masses of bronze work of Late Bronze Age type (O’Connor 1975). A phalera of similar date with a central perforation came from a hoard at Newark-on-Trent, Notts. (Figure 17.5), together with socketed axes and spears, and a hoard at Cardiff, Glam., produced an axle cap and two Hallstatt C razors as well as various implements in the local Bronze Age tradition.

This brief survey of the more significant hoards leaves very little doubt that throughout the eighth and seventh centuries, harness- and cart-fittings of Continental type were finding their way into this country in some quantity and not infrequently ending up in hoards along with local types. The mechanisms by which this material arrived are impossible now to recover. Conceivably horses and their tack were being traded together, perhaps with carts, in a system of élite exchange but we cannot completely rule out the possibility that a few people were now making their way into the country, bringing with them the outward and visible signs of their aristocracy.

Another group of items which might be regarded as representative of élite status are high- shouldered bronze buckets or situlae, of which twenty or so have been found in Britain and Ireland (Hawkes and Smith 1957; Briggs 1987; Turnbull 1995). While they are evidently inspired by Continental types of the eighth century, Briggs has argued that the extant British and Irish examples are all likely to have been of local manufacture. None the less their relative popularity and wide geographical extent would imply the importation of Continental prototypes (perhaps in leather as Briggs suggests) of which no example has yet been found. Such grand items may originally have served in displays of hospitality or feasts among the élite but a number were subsequently deposited in rivers or bogs presumably as offerings to the gods.

Figure 17.8 Bronze cauldrons (sources: Sompting, E.C. Curwen 1948; Colchester and London, Hawkes and Smith 1957).

Figure 17.9 Bronze cauldron from Llyn Fawr, Glam. (photograph: National Museum of Wales).

Superimposed upon this trade in aristocratic equipment was the importation of minor objects such as razors (Figure 17.10) (Piggott 1946), swan’s-neck pins (Figure 17.11) (Dunning 1934) and brooches (Figure 17.12) (Harden 1950). Razors from the Llyn Fawr and Cardiff hoards have already been mentioned; others have come from the settlements at Staple Howe, Yorks., and Potterne, Wilts., from a hoard at Danebury, Hants, unstratified from early occupation sites on Ham Hill and South Cadbury, and as isolated finds from half a dozen or more sites. Razors are frequently found in Hallstatt burials on the Continent, where they must represent a standard piece of a warrior’s equipment (Meyer 1986, 73–4). The British finds do not, however, come from burials and many of them may have been manufactured locally.

Swan’s-neck pins occur sporadically in Britain. In all about a dozen are known, mainly from the south, but apart from the examples at the settlement site of All Cannings Cross, Wilts., they are not generally found in closely stratified contexts, and since the type is likely to have continued in use for some time, it is not a very useful chronological indicator. Even less certainty attaches to the large number of Hallstatt fibulae from Britain (Figure 17.12). Not one has been recovered in an Iron Age context, but the number is so great that some at least must have found their way into Britain in the eighth to sixth centuries, on the dresses of visiting womenfolk or as traded trinkets.

In north-eastern Scotland a more substantial group of Continental-inspired metal types has been found, constituting what is called the Covesea phase of the local metal industry (Coles 1962b). Dominant among the types which define the phase are simple penannular armlets with their terminals expanded outwards, related to a type found in Late Urnfield contexts in northwestern Europe, whence it is suggested they were imported to Scotland. In two important hoards, Wester Ord, Ross, and Braes of Gight, Aberdeen, Covesea armlets were found associated with penannular necklets with free-swinging loops and pendants. These too have close north European parallels dated to c. 700 BC. In a third hoard of some significance, from Bal- mashanner, Angus, three Covesea armlets were found with a bronze bowl and an iron ring, demonstrating beyond doubt the overlap between the Covesea phase and the use of iron. Other associations showed that Covesea types were used in parallel to an already vigorous indigenous bronze industry, the various phases of which are named after the hoards at Adabrock, Ballimore and Tarves (Coles 1962b).

Figure 17.10 Hallstatt razors: 1, 6 Cardiff, Glam.; 2 Llyn Fawr, Glam.; 3 River Thames at Richmond, Surrey; 4, 5, 8 Staple Howe, Yorks.; 7 Midlothian (sources: 1, 6 Nash-Williams 1933b; 2 Grimes 1951; 3, 7 C.M. Piggott 1946; 4, 5, 8 Brewster 1963).

That Scotland appears to have maintained direct contact with northern Europe at this time is also shown by the distribution of distinctive sunflower swan’s-neck pins (Figure 17.16): thirteen from Scotland, one from Ireland and one from Fengate, Northants. Coles (1959) believed that the genesis of the type – the fusing of the Hallstatt swan’s-neck pins and the Irish-Scandinavian sunflower pins – probably took place in north Germany in the sixth and fifth centuries. However, Eogan (1974a, 58) argued for an earlier origin. This is supported by the fact that three Scottish hoards, St Andrews, Grosvenor Crescent, Edinburgh, and Tarves, Aberdeenshire, show that these pins were found in direct association with Ewart Park metalwork of the ninth to eighth centuries.

While a north European origin can be demonstrated for the Covesea types and the sunflower pins, the nature of the contact is far less clear. It has been suggested that the distinctive flat-rimmed pottery found at Covesea itself, Balmashanner and a number of other sites is of Late Urnfield derivation (Coles 1962b, 44), hinting at the possibility of a limited folk movement, but the pottery is so generalized that firm assumptions cannot be based on it. Seventh- century radiocarbon dates for the timber-laced forts of north-east Scotland have also been quoted to add support to the idea of an immigration from northern Europe (MacKie 1969a, 1976) but none of these observations is a compelling argument for folk movement: at best it shows a period of contact and innovation.

Figure 17.11 Bronze pins: 1 Hammersmith, London; 2 Brighton, Sussex; 3 Portslade, Sussex; 4, 6 All Cannings Cross, Wilts.; 5 Totternhoe, Beds.; 7, 8 Heathery Burn, Co. Durham (sources: various).

Atlantic trade: eighth to sixth centuries

Central to the discussion of Atlantic contacts in this period are the large sheet-bronze cauldrons of which about fifty have been found in Britain and Ireland and a further twenty in western France and Iberia. Two types have been distinguished, class A and class B. It was originally believed that the class A cauldrons were derived from a Mediterranean prototype of the eighth century (Hawkes and Smith 1957) but further research has shown the situation to be more complicated. Eogan (1974a, 322 n. 18) suggested that the idea was introduced from northern Europe, an idea later supported by Gerloff (1987) who argued that the earliest British examples, from Shipton and Colchester, probably date to the late second millennium. She further argues that cauldron production was taken up in Ireland, whence examples were traded along the Atlantic sea routes to England, Scotland, western France and Denmark. The tradition, once established in Ireland, continued with the development of the class B cauldron in the seventh and sixth centuries BC. Associations between cauldrons and Hallstatt C types are known at Sompting and Llyn Fawr.

Figure 17.12 Hallstatt fibulae: 1 Hod Hill, Dorset; 2 York; 3 North Wraxall, Wilts.; 4 Box, Wilts. (sources: 1,4 R.A. Smith 1925; 2 Antiq. Journ. 12 (1932), 454; 3 Cunnington and Goddard 1934).

The exchange networks along the Atlantic coast of Iberia, France and Britain, which had intensified in the Late Bronze Age, continued in the eighth century, by which time the tools and weapons of the Carp’s-Tongue Sword Complex were being widely distributed throughout northwestern France and south-eastern Britain (above, Figure 4.4). That the system continued into the seventh century in parallel with the arrival of Hallstatt C influences is shown by the occurrence of a Hallstatt C razor with Carp’s-Tongue Sword material in a hoard on Ile Guenoc, Finistère (Briard 1957, figure 28) and a radiocarbon date of 559±130 BC for a context at Saint- Guganen, Loudéac, Côtes-d’Armor.

On both sides of the Channel there appeared at about this time large numbers of small nonfunctional axes with rectangular sockets and unprepared cutting edges, known as Breton or Armorican axes. Altogether in north-western France, it is estimated that 315 hoards contained this type exclusively, producing some 38,000 individual specimens (Briard 1995, 164; Giot et al. 1995, 77–93), and the type is found widely in south-east Britain. One view is that they formed some kind of currency and continued to be used as such into the fifth century or even later. At Saint-Martin-des-Champs, Finistère, Breton axes were found in association with an iron ingot, demonstrating a late survival. The axes, then, are the last manifestation of the Atlantic bronze exchange systems, out of which were to grow the sophisticated trading contacts first mentioned by classical writers in the fourth century, and to which, a few centuries later, Caesar refers in some detail.

European contacts: sixth to fifth century

In central Europe the change to the Hallstatt D stage of culture took place a little before 600 BC. By this time the central areas of development, typified by the rich chieftains’ burials, had shifted west into the middle Rhine and Moselle region and into Burgundy – areas which now benefited from the new trade routes along the Rhône that had opened up following the foundation of the trading port of Massilia (Marseilles). Trading systems were by this time well established between central Europe and the Mediterranean world, and luxury goods belonging to wine-drinking and feasting rituals were being imported from Greek and Etruscan workshops, eventually finding their way into the tombs of the aristocracy (Cunliffe 1988a, 24–35). Trading contacts between Britain and the Continent appear to have been less intensive in the sixth and early fifth centuries than they had been in the preceding 100 years, but nevertheless a number of imported metal objects of Hallstatt D type are known.

The Thames was evidently an important point of entry, to judge by the number of new types found in the London area. From the Thames came a short iron sword with an antennae-shaped hilt (Figure 17.13) (Smith 1925; Stead 1984b) characteristic of sixth-century Continental types, and from the same general provenance was recovered a simple hemispherical bronze cauldron of Hallstatt D type (Hawkes and Smith 1957) which can be paralleled on many west central European sites, more commonly in the second half of the century (Figure 17.8).

The long-established school of swordsmiths in the Thames region, who by now were manufacturing their own varieties of the Hallstatt C imports, soon absorbed the idea of the Hallstatt D short dagger – a type in use in the early sixth century in southern Germany – and began to manufacture their own elegant versions with an improved method of suspension using twin loops at the back of the sheath to strengthen the attachment to the belt (Figures 17.13 and 17.14). Only one certain import is known, found in the bed of the Thames at Mortlake (Figure 17.13, no. 3) and its sheath was probably repaired in Britain (Jope 1983). Diplomatic gifts of this kind inspired a vigorous British tradition. The industry continued to flourish until about 300 BC, influenced in its later stages by new La Tène ideas coming in after about 450 (Jope 1961a).

A comparable example, showing the same process of adoption and rapid modification of Hallstatt D ideas, is exhibited by the representative of a class of fibulae found in the Thames near Hammersmith (Hodson 1971). Evidently there must have been schools of craftsmen working in the area, always receptive of new Continental ideas.

More widely flung contacts with northern Italy are emphasized by the ribbed pail, dug up at Weybridge, Surrey, made of strips of corrugated bronze riveted together along a single vertical seam and provided with two movable handles (Stead 1984a, 43–4). The type is well known in western Europe and was probably made in a European workshop some time in the sixth century whence, directly or indirectly, it reached Britain (Stjernquist 1967, 70–4).

Several other Italian Mediterranean bronzes have been found in Britain, but all in dubious circumstances (Harbison and Laing 1974). Among the more significant should be mentioned a fragment of ‘Rhodian’ flagon of late seventh- to early sixth-century date found at Minster in Kent, an Etruscan bronze oenochoe of the late sixth to early fifth century from Northampton, a trefoil-mouthed flagon of the early fifth century from the river Crouch, Essex, and two bronze jugs of fourth-century date, one from Tewkesbury, the other from Bath. In no case is it possible to be sure that the vessel was a contemporary Iron Age import, and the possibility must therefore be allowed of the more recent importation.

Other west European or Mediterranean objects found in Britain include about eighty examples of Continental fibulae dating to the period of the eighth to fourth centuries (Harden 1950). It is, of course, impossible to be sure when they were imported. Some of them may have arrived during the Roman period or even later in the collections of travellers, but many must have been imported into Britain during the time when they were current on the Continent. Other personal ornaments of Hallstatt type, including neck- and arm-rings of bronze with ‘nut’ moulded decoration, occur less frequently. One was found close to an early occupation site at Scarborough, Yorks., one from the hillfort of Castle Rings, Wilts. (Tomlinson 1986). Several came from the Iron Age port site at Mount Batten (Cunliffe 1988b), and another unassociated example is recorded from Clynnog, Caerns. (Figure 17.15). The general type does not seem to have been widely copied by local craftsmen.

Figure 17.13 Hallstatt D weapons: 1 River Thames; 2 Battersea, London; 3 Mortlake, Surrey (sources: 1 R.A. Smith 1925; 2, 3 Jope 1961a).

Another class of imported material is pottery. Several finds of Mediterranean vessels have been made in Britain, but again the date of their importation is, without exception, unknown since no examples have yet been found on stratified occupation sites. Several vessels have been dredged from rivers, including a Greek Black Figure kylix from the Thames near Reading, which is reliably dated on stylistic grounds to the late sixth century (Boon 1954, 178). From the same river, at Barnes, an Italic handled cup of the seventh century was dug out of the river bank, and from Barking Creek came a Greek hydriskos of the sixth century (Harden 1950, 321–2). A squat lekythos of fourth-century date was also recovered from Billingsgate. Finally, mention should be made of a group of three Greek vessels, one of them an Attic drinking cup of the late fourth century, which were found together in an artificial cave at Teignmouth, Devon, during the last century. It must be stressed that none of these finds need represent Iron Age importation, but there is nothing inherently difficult in supposing that some Mediterranean vessels found their way into southern Britain during the course of normal maritime exchanges.

Figure 17.14 Hallstatt D dagger and sheath from the River Thames at Hammersmith, London (photograph: Museum of London).

It can be seen from the above paragraphs that imported material of Hallstatt D date is sparse in Britain, but sufficient survives to imply continuous contact with the Continent throughout the sixth and early fifth centuries. Nothing in the surviving archaeological record needs, however, to be explained in terms other than of casual exchanges.

Figure 17.15 Hallstatt bracelets and neck-rings: 1 Scarborough, Yorks.; 2 Cold Kitchen Hill, Wilts.; 3 Clynnog, Caerns. (sources: 1 R.A. Smith 1934; 2 DMC; 3 Grimes 1951).

Contacts between Britain and the Continent: fifth to second century

From the fifth to the second century, contacts between Britain and adjacent parts of the Continent were maintained. Metalwork in the new La Tène style, which developed in Europe towards the beginning of the fifth century, found its way into the country largely through the exchange networks. The full corpus of material has been comprehensively reviewed and catalogued by Martyn Jope (2000). Many of the incoming types were rapidly copied by local craftsmen, giving rise to distinctive British varieties. In Yorkshire, however, a more extensive cultural assemblage appears to have been introduced, including not only metal types but alien burial customs best paralleled among the La Tène cultures of northern France. This group – referred to as the Arras culture – has been considered in some detail above (pp. 84–6). Elsewhere in Britain northern French decorative metalwork is rare, but a fine openwork disc, probably a horse-harness decoration, was found at the hillfort of Danebury, Hants (Figure 17.17). It is closely comparable to ornaments from the late fifth-century chieftains’ burials of the Marne region and is either an import or an assured local copy of one. A broadly similar type from the Thames at Hammersmith (Smith 1905, figure 122) is almost certainly a British-made version.

The sixth-century production of a distinctive series of daggers and dagger sheaths in the Thames region, noted above, was evidently carried out by local craftsmen influenced at first by Hallstatt D prototypes but soon evolving their own improved techniques and style. Production continued throughout the fifth and fourth centuries, absorbing, after about 450, La Tène improvements which were being developed in parallel on the Continent. The earliest of these La Tène-inspired sheaths, one discovered at Chelsea, is so close to northern French types of the mid- fifth century, while evidently itself of British manufacture, that the British craftsmen must have responded directly to new Continental ideas. Close contact appears only at this time; thereafter the British and French traditions diverged, hinting at the beginning of some degree of cultural isolation. The twenty or more daggers of La Tène I type found in Britain, mainly in the Thames area, can be arranged in a typological sequence lasting until the late fourth century. Among the latest to be made was the scabbard from Wisbech, decorated in a free-hand curvilinear style with S-shaped scrolls arranged in lyre and palmette patterns about a central ridge. This must be regarded as a provincial version of an art style current in Europe during the fourth century, from which the early beginnings of British Celtic art eventually developed.

Figure 17.16 Sunflower swan’s-neck pins: 1 Taves, Aberdeen; 2 Loch Broom, Ross; 3 Campbeltown, Argyll. (source: Coles 1959).

While the later British daggers continued to evolve, throughout the beginning of the La Tène period a range of swords of La Tène I type was being manufactured. At least seventeen have been identified from the rivers Thames and Witham, but of these only two or three have any claim to being actual imports. One came from Wandsworth and had a blade engraved or chased with close-set parallel lines in a style known as laddering, best represented on Swiss examples; another, from the Thames at Hammersmith, was found in a scabbard decorated at the hilt end with facing dragons in a style more common in Hungary (Stead 1984a, 47–50) (Figure 17.18). A third, possible, import was a finely decorated version from the river Trent at Sutton, Notts., which is closely similar to examples belonging to the Swiss sword style. Even the scabbard from Standlake, Oxon. (Figure 18.26), which has close connections with Continental technological tradition, is best regarded as a British product standing at the head of a long development series lasting throughout the rest of the pre-Roman period. What is perhaps most impressive about the British swords is their relative isolation from Continental development. This fact, together with the general lack of imported varieties, might suggest that Britain now lacked the close contacts with the Continent which it had previously maintained during the seventh and sixth centuries.

Figure 17.17 Openwork disc of early La Tène date from Danebury, Hants, actual size (photograph: Institute of Archaeology, Oxford. Danebury Trust).

Personal ornaments are found scattered over much of southern Britain and of these the commonest category is brooches. The earliest of the La Tène brooches found in Britain – those of La Tène I type (c. 450–300 BC) – are the most numerous: over 100 have come to light, mostly from the south-east. Characteristically they have an arching bow, bilateral spring and a catch plate made by bending the foot back to touch the bow (Figures 17.19 and 17.20). The bow is often decorated with simple grooved, curvilinear or stamped decoration and the foot is usually expanded into a disc shape decorated with stamped or incised motifs, or very occasionally inlaid with coral. The first serious attempt to classify these brooches was put forward by Fox (1927). There have been several subsequent additions to the debate, but the publication of a full corpus (Hull and Hawkes 1987) put the subject on a firm basis and more recently Martyn Jope has thoroughly reviewed all the evidence (Jope 2000, 39–50). From the point of view of the present discussion it is sufficient to note that of the large number of La Tène I brooches known from Britain very few can be identified as actual imports. However the number alone, and the Continental style and technology which they adopt, show that Britain remained in contact with Continental developments. The earliest brooches are of the Marzabotto type found commonly in Europe north of the Alps and in smaller numbers in northern Italy (Stead 1984b, 50–2). They were in use from the late fifth to the early fourth century. At the Swiss cemetery of Münsingen, where a comprehensive sequence has been established, they are known as type Ia (Hodson 1968, 35). Of the British examples fewer than a dozen are possible imports. The type was particularly popular in Wessex, where a distinctive local version was produced (p. 501). Of the next group, the Dux fibulae (Münsingen Ib, c. 350–280 BC), no contemporary imports have been recorded, if the Wallingford examples are now accepted as recent imports (Stead 1984b, 53–4), but there are a few native copies. Thereafter imported La Tène types and their copies become far less common. There is a bent silver ring characteristic of the Münsingen Ic phase (c. 280–200 BC) from Park Brow, Sussex (Figure 17.19, no. 8) and a few British brooches of Ic type, but the paucity of material is in marked contrast to the number of earlier La Tène types found. The same may be said of the succeeding La Tène II period (c. 200–100 BC). Thus, if the number of brooches found in Britain can be taken to be a reflection of the intensity of importation, then the period from 450 to 350 was a time of much interaction, after which, until about 100 BC, the intensity of contact had dramatically declined. The figures cannot, however, be taken too literally since a high percentage of the locally manufactured Ia types were probably being made and used in the third and even the second century. Even so the relative lack of inspiration from Continental types of the period 350 and 100 is best explained by a sharp diminution in imports.

Figure 17.18 Imported La Tène sword from the Thames at Hammersmith (source: Stead 1984a).

Figure 17.19 La Tène style fibulae and a bent silver ring: 1 Box, Wilts.; 2 Blaise Castle Hill, Avon; 3 Merthyr Mawr Warren, Glam.; 4 Findon Park, Sussex; 5 Water Eaton, Oxon.; 6 Blaise Castle Hill, Avon; 7 Merthyr Mawr Warren, Glam.; 8 Park Brow, Sussex (sources: 1, 5 R.A. Smith 1925; 2, 6 Rahtz and Brown 1959; 3, 7 Grimes 1951; 4 Wolseley, Smith and Hawley 1927; 8 R.A. Smith 1927).

Figure 17.20 La Tène I style fibula from Danebury, Hants (actual length: 46 mm) (photograph: Nick Bradford. Danebury Trust).

Bracelets of La Tène types are not common in Britain outside the territory of the Yorkshire Arras culture, where two basic types are found: the relatively plain version with a ‘tongue-in- glove’ fastening and the more elaborate knobbed and ribbed type with a thick and often heavy body formed into adjacent bosses or ribs (Figure 17.21). The plain types occur once at Cowlam and four times at Arras, both Yorkshire graves. The only other examples to be found are the pair of bracelets with circular bezels from Coygan Camp, Carm.; but here the fastening is simple and not specially tongued. The knobbed and ribbed varieties are rather more widely spread but are difficult to distinguish from Hallstatt types. It is impossible to be certain how many of these bracelets are direct imports and how many are locally made, but most of the specimens can be paralleled among La Tène I and early La Tène II contexts in eastern France and Switzerland, whence it is likely they or the types which inspired them were derived.

Trading contact between the west of Britain and the Atlantic coasts of Gaul and Iberia is well attested both in the classical literature and in the archaeological record. The fourth-century Roman poet Avienus, in his poem Ora Maritima, uses scraps of information taken from a very early account of the Atlantic sea-ways known as the Massaliote Periplus, which purports to describe the journeyings of the Tartessians and the Carthaginians from southern Iberia to Brittany, Ireland and Britain (Albion) for purposes of trade. After Tartessos had become a Phoenician monopoly in the early fifth century, the Greek city-states which had previously been supplied with tin via Iberia began to look further afield for their vital supplies. One Greek merchant, Pytheas, sailed to Brittany along the Atlantic route between 330 and 325 and recorded his experiences, which survive only in the writings of later classical authors such as Strabo (Hawkes 1978; Cunliffe 2001b). Nevertheless, sufficient appears in these secondary sources to show that the tin-producing areas of the south-west were in constant contact with the Mediterranean world from the fifth century onwards, no doubt following upon trading traditions already established in ancient times.

Figure 17.21 La Tène bracelets: 1 Coygan Camp, Carm.; 2, 4 Arras, Yorks.; 3, 6 Cowlam, Yorks.; 5 South Ferriby, Lincs. (sources: 1 Wainwright 1967a; 2–6 Stead 1965).

The texts relating to the later period are somewhat obscure and have led to extensive debate (Maxwell 1972; Cunliffe 1983; Mitchell 1983; Hawkes 1984). The question hinges upon the location of Ictis mentioned by Diodorus Siculus, in a famous descriptive passage based upon earlier sources, probably Pytheas. He describes the peninsula of Belerion where tin is to be had. After mentioning the process of extraction, leading to the preparation of knuckle-bone-sized ingots of metal, he goes on to say that they are carried

to an island which lies off Britain and is called Ictis, for at the time of ebb-tide the space between this island and the mainland becomes dry and they can take the tin in large quantities over to the island on their wagons… On the island of Ictis the merchants purchase the tin of the natives… whence it is taken to Gaul and overland to the Mediterranean.

The question is complicated however by a text of Pliny, quoting Timaeus (Nat. Hist. IV, 16, 104) in which he tells us that ‘there is an island named Mictis lying inwards, six days’ sail from Britain where tin is found and to which the Britons cross in boats of osier covered with stitched hides’. The two texts taken together provide a fertile ground for imaginative speculation! The simplest interpretation would be that Ictis and Mictis, despite similar sounding names, were two entirely different places, Ictis being just off the shore of south-western Britain with Mictis being considerably nearer, along the trade route, to Rome. Six days’ sail ‘inwards’ could place it anywhere off the Armorican coast as far south as the mouth of the Loire.

The location of the British trading port, Ictis, has been hotly debated, the most favoured contender being St Michael’s Mount, Cornwall, for which no shred of archaeological evidence has been presented. A better claim can be made for Mount Batten, a promontory jutting into Plymouth Sound (Cunliffe 1983). Before the construction of the Plymouth breakwater it frequently became an island at high tide. Moreover, it is admirably sited to provide a well-protected haven and to command the rivers Tamar, Tavy and Plym which lead to the metal-rich fringes of Dartmoor. There is ample archaeological evidence to show that it was actually used from the Late Bronze Age and throughout the period described by the classical writers (Cunliffe 1988b), but whether or not Mount Batten was Ictis will remain a matter of personal opinion.

An interesting discovery which may reflect on the early tin trade was made by divers exploring a reef in the estuary of the river Erme where it flows into Bigbury Bay on the south coast of Devon. The divers recovered a number of tin ingots evidently from an ancient shipwreck site. The ingots were undated and were thought to be of Dark Age date (Fox 1997) but the possibility remains that they are much earlier. Most are of bun shape but a few are quite distinctly the size and shape of a knuckle-bone recalling the description of Diodorus Siculus. That said, without firm dating evidence the intriguing possibility of this being an Iron Age wreck remains unproven.

Archaeological evidence for trade along the Atlantic sea-ways is not plentiful but contact with the south-west of Europe is demonstrated by a group of distinctive fibulae, two of bronze found at the port of Mount Batten (Figure 17.22) and two bronze and one iron from the cemetery at Harlyn Bay on the north coast of Cornwall. The brooches are characterized by a knob-ended cross piece for the pin to pivot on, a high bow and an upturned foot ornamented with a large disc head. They belong to a well-recognized class of fifth-century brooches found in Spain and Aquitania (south-western France), and while it has been argued that the British finds were locally made (Boudet 1988), the essentially Atlantic distribution of the type is a clear indication of maritime activity which presumably brought to Britain the prototypes which inspired them.

A more enigmatic object, frequently quoted as evidence of Atlantic trade, is a small bronze statuette found on the shore of Aust close to the site of the Severn Bridge (Dawson 1980) (Figure 17.23). The figure, a female, is provided with a crescentic head-dress and has eyes enlivened with inset glass beads. Traditionally it has been referred to as an Iberian import but Stead has cast doubt on this attribution, preferring a British origin (Stead 1984b, 60). However, not only can no British parallels be quoted, but the head-dress and, more significantly, the stance and simplification of the figure have distinct analogies among first-millennium BC Iberian figurines. The uncertainty well illustrates the problems posed in attempting to identify imports.

The documentary and archaeological evidence together emphasize the continuance of well- established patterns of trade along the Atlantic sea routes from the fifth to the second century. It may have been by these means that some, at least, of the large number of Greek and Carthaginian coins arrived in Britain. The distribution pattern now extends across much of southern and eastern Britain, but rarely have the coins come from undoubted Iron Age contexts (Laing 1968). While it remains a possibility that some of them may have been brought in during the Roman period or as later collectors’ items, many must have arrived during the pre-Roman period in the wake of Atlantic trade. Only one, a coin of Ptolemy V (204–181 BC) from Winchester, has so far been found stratified in an Iron Age level (Cunliffe 1964, 75; Collis 1975; Biddle 1975b).

Figure 17.22 Fibulae of ‘Iberian’ type from Mount Batten, Devon (source: H.O’N. Hencken 1932).

One final item remains to be considered – the famous ‘hanging bowl’ found in a stone cist at Cerrig y Drudion, Clwyd, in 1924 (Figure 17.24). A reinterpretation of the surviving fragments has suggested that more than one item may be represented and that the major pieces are probably part of an ornamented lid rather than a bowl (Stead 1982), the chains being for attachment, not for suspension as originally thought. The domed surface and the underside of the flange are decorated with an elaborate scheme of incised palmettes and acanthus half-palmettes thrown into greater prominence by a cross-hatched ‘basketry background’. The style of decoration has much in common with Celtic art styles in western France and in particular with the finely decorated pottery of Brittany, but the ‘basketry’ technique is generally considered to be a British development. Thus, one of the finest pieces of Celtic art found in Britain must remain of uncertain origin. Its importance, however, is that it demonstrates a lively exchange in artistic concepts in the western sphere of contact in the early fourth century, at a time when the British schools of craftsmen were beginning to develop their own distinct styles in the service of their artistic patrons (Cunliffe 1990b).

The archaeological evidence for contact along the Atlantic sea routes in the period from the fifth to the second century, though not extensive, usefully augments the tantalizing accounts provided by the classical texts. The impression given is that trade was organized and that certain locations had emerged as ports-of-trade where commodities could be exchanged in safety. Ictis is the only named place, but in all probability a number of coastal locations in the south-west had by now developed trading connections. Mount Batten is the best attested archaeologically but others suggest themselves. At Exeter, for example, a surprisingly large number of Mediterranean coins have been found which may reflect pre-Roman trade. More certainty attaches to Methyr Mawr Warren, at the mouth of the rivers Ogmore and Ewenny in South Wales, where a range of material including La Tène brooches and evidence for metalworking indicates a settlement of some importance. The site is well located to be a port articulating with the metal-producing areas of south Wales. Somewhere in the vicinity two bronze helmets were found in 1808 (now lost). Jope suggests that they may have been Italo-Celtic imports of the fourth century BC. Uncertainty remains but such finds would be consistent with Methyr Mawr Warren being a major entry port (Jope 2000, 228; Toft 2001).

Figure 17.23 Bronze figurine from Aust, Avon (actual height: 146 mm) (photograph: British Museum).

Sites such as these may well have taken on a more than local significance at times when the Atlantic route was being actively exploited.

Trade in the first century BC and early first century AD

After the beginning of the first century BC evidence for trade and exchange with the Continent increases dramatically. This is amply demonstrated by a wide range of imports including Gallo- Belgic and Armorican coins, north-western French pottery, Italian and Spanish amphorae, bronze and silver tableware from Italy and a number of other luxury commodities. In addition to actual imports, rapid style change in the pottery of the south-east and the appearance of large numbers of fibulae modelled on La Tène III types in circulation on the Continent show that Britain and the adjacent parts of Europe were now in close and continuous contact. What we are seeing is essentially the bow-wave effect in advance of the increasing Romanization of Gaul. To understand it something must be said of events on the Continent.

Figure 17.24 Bronze vessel from Cerrig y Drudion, Clwyd. The original drawing (reproduced here from Smith 1926) shows the find reconstructed as a hanging bowl. It is now thought to be parts of one or more lids. Actual maximum diameter of flange 26 cm.

In the latter part of the second century BC Rome was increasingly drawn into the affairs of southern France in response to pleas from the Greek cities of the Mediterranean coast for military support against the aggressive attentions of neighbouring hill tribes. By the 120s events had become so serious that the Roman authorities decided that the only solution was to station a legion at Aquae Sextiae (Aix-en-Provence). In this way the Romanization of Gaul formally began and Roman entrepreneurs poured in, bent on exploiting the new commercial opportunities which occupation provided. In 118 BC a colony was founded at Narbo Martius (Narbonne), greatly facilitating access to the new markets. For the next sixty years or so the province of Transalpina, as it was known, was thronged with merchants trading with the barbarian Gauls beyond the frontier. Enormous quantities of Italian wine, and no doubt other consumer goods, were exchanged for slaves and raw materials, particularly metals (Tchernia 1983; Cunliffe 1988a, 80–92). The sudden development of this market invigorated the traditional trading routes, particularly the Atlantic route along the western coast of France to southern Britain.

In 59 BC Roman involvement in Gaul entered a new phase when Julius Caesar began a war of conquest against the Gauls which was to last for almost a decade. By the time it was over the political geography of Europe had changed out of all recognition. Rome had now established a somewhat shaky control over a vast territory of what is now France, Belgium and parts of Holland up to the river Rhine. This new reality, and the fact that Caesar had campaigned in Britain in 55 and 54 BC, where he had negotiated treaties with the tribes of eastern Britain, brought about a reorientation of trade with the island. While the old Atlantic route still continued to function, trade was on a drastically reduced scale: it was the coastal tribes of eastern Britain who were now in the forefront of commerce.

By about 10 BC Gaul had become thoroughly stabilized. From this time onwards the Rhine frontier, and the route via the Rhône leading to it, became a ‘commercial corridor’ – a zone of production, consumption and transshipment linking the Thames estuary to the Mediterranean. The effects of all this on the tribes of south-eastern Britain can be clearly recognized as quantities of comparatively low-value goods, such as pottery, manufactured in the commercial corridor, began to pour in.

It is against this broader background that the pattern of imports enjoyed by Britain from about 100 BC until the Roman Conquest must be assessed.

The effects of the reinvigoration of the Atlantic route, following the foundation of the Roman province of Transalpina, can best be seen at Hengistbury Head, Dorset (Figure 17.25). The headland, a dominant feature of the Solent coast, protects the sheltered expanse of Christchurch harbour from the predominant south-west winds and currents, thus providing an attractive anchorage for shipping (Figures 8.5). Not only this, the site was also admirably located at the hub of a natural route network which linked the productive hinterland of Wessex (via the rivers Stour and Avon), to the land-hugging south coast routes and the favoured cross-Channel passage, via the Channel Islands, to the ports of the Côtes-d’Armor and beyond (McGrail 1983; Cunliffe and de Jersey 1998). Excavation has shown the existence of a substantial settlement along the protected northern shore, producing a wide range of imported goods including quantities of Italian Dressel 1A amphorae, lumps of raw purple and yellow glass, figs, and masses of pottery from Brittany including elegant black cordoned wares, graphite-coated wares and rilled vessels made in a micaceous fabric (Figures 17.26 and A:33). There can be little doubt that in the first half of the first century BC Hengistbury was receiving cargoes of merchandise made up of goods amassed at various points along the 1,600 km route which led from northern Italy, via Narbonne, through the Carcassonne Gap and the Garonne/Gironde, along the Atlantic coast of France and around Brittany to the Solent shore. In return the merchants at Hengistbury were collecting together an impressive range of exports. Local products included iron, Kimmeridge shale and salt; grain came from the Wessex chalk- land, while metals including gold, silver, copper, tin and lead were gathered as ore, ingots and scrap from the Mendips, the south-western peninsula and elsewhere for refinement, and perhaps manufacture, before export. All these commodities are attested in the archaeological record but many more were probably involved. When, a few decades later, Strabo listed the principal exports from Britain, he included ‘grain, cattle, gold, silver and iron… also hides, and slaves and dogs that are by nature suited to the purposes of the chase’ (Geog. 4.5.2). His list gives some idea of significant local products which are archaeologically invisible.

Figure 17.25 Hengistbury Head, Dorset. The excavation of the Late Iron Age port on the shore of the harbour (photograph: author).

Figure 17.26 Distribution of Late Iron Age pottery made in Brittany (source: Cunliffe and de Jersey 1997).

Although Hengistbury was clearly a major focus of trade, it may not have been the only one in the area. Poole Harbour, nearby, has produced a comparable range of imports, especially from Green Island and Hamworthy, from both of which have come Italian Dressel 1A amphorae and north-western French pottery. An impressive array of Mediterranean coinage has also been found in the vicinity (Cunliffe 1982a, figure 3). The discovery of two substantial jetties, dating to the second century BC, leading out into deep water from Green Island and the adjacent mainland shows that the local community was investing much effort in its maritime infrastructure. With ease of access to salt and Kimmeridge shale and with a convenient route, via the river Frome, deep into Durotrigan territory, Poole Harbour has obvious attractions as a trading focus. We might therefore regard the two harbours as integral elements in a single contact zone. Further to the west the Isle of Portland commanding Weymouth Bay may also have served as a centre for articulating local exchanges (Taylor 2001).

From this Solent contact zone imports were distributed to the hinterland. French pottery (and its contents?) barely got as far as 10 km from the ports, amphorae were found further inland up to 50–60 km, while coins of the Armorican tribes, particularly the Coriosolites but also the Baio- casses, Redones, Unelli and Osismii, reached much further (and may anyway have arrived by a variety of routes) (Cunliffe and de Jersey 1998, 72–103). Of the British commodities needed for export, metals from the south-west would have come by sea while the lead/silver alloy from the Mendips was probably brought via the Stour valley route (Figure 17.27).

The dense distribution of Dressel 1A amphorae in Armorica (Figure 17.28), the presence of Armorican pottery and coins in the British contact zone and the occurrence of Kimmeridge shale bracelets in Brittany are sufficient to show that the tribes of Armorica were intimately bound up in the trading system. Caesar discovered this, and writing of them says

the Veneti are by far the strongest. They have a great many ships and regularly sail to and from Britain. When it comes to knowledge and experience of navigation, they leave all the other tribes standing… They are able to extract tolls from almost all who regularly use those waters.

(BG iii, 8)

It is curious that in Britain Venetic coins are extremely rare while those of the Coriosolites, from the Côtes-d’Armor, are far more common. One possible explanation for this is that the Veneti were the receivers of the Atlantic cargoes coming from the south, and were thus well known to Roman traders, while the final leg of the journey from Brittany to Britain was under the control of the Coriosolites. If so it might imply a vigorous cross-peninsula trade from the south coast of Armorica to the north.

Exactly how the process of cross-Channel trade was articulated it is difficult to say but the exceptionally large quantity of Armorican pottery at the port of Hengistbury hints at the possibility of a significant presence of Armoricans. One possibility, therefore, is that Armoricans actually lived at Hengistbury during the summer months when the trading enterprises were under way and were visited there by Britons bringing products from various parts of central southern and south-western Britain. Once it had become known that traders would be present for a fixed period every year it is easy to see how a momentum would build up.

A trading network as complex as this could not have survived Caesar’s devastating attack on the Armorican tribes in 56 BC and the evidence from Hengistbury bears witness to this. The importation of French pottery virtually ceases and amphorae which can be dated with some certainty to the second half of the first century BC occur in negligible quantities. A few Dressel 1B types show that a little Italian wine was arriving and occasional examples of Dressel 1/Pascual 1 reflect the growing importance of the Tarragona region of Spain as a supplier, but even together the quantity represents only a fraction of that which had previously arrived in Dressel 1A amphorae. While the chronology of these vessels is not precise, the evidence from Hengistbury clearly shows a marked decline in wine importation after the mid-first century BC (Cunliffe 1987). In Poole Harbour, however, that some level of cross-Channel contact was maintained is shown by imported pottery of the late first century BC and early first century AD (Cox and Hearne 1991, 114–22).

Figure 17.27 Model for trade with Hengistbury (source: Cunliffe 1982a).

Direct archaeological evidence of Iron Age shipping around the shores of Britain is slight, but a complete iron anchor together with a length of chain was discovered in a hoard at Bulbury, Dorset, together with other material suggesting a date in the first half of the first century BC (Figure 17.29). Although the anchor may have come from an Armorican ship (Cunliffe 1972b) its form suggests that it is more likely to have been of Roman origin. More recently a lead anchor stock of Greco-Roman type was found by divers off the coast of north-west Wales, near Porth Felen (Figure 17.29). Typological considerations indicate a date in the second or first century BC (Boon 1976). These two anchors provide an interesting hint of the type of Mediterranean shipping with which the inhabitants of pre-Roman Britain may well have been familiar.

The Porth Felen anchor is a reminder that some foreign vessels were now penetrating along the Atlantic seaways. In this context we may note what is evidently a beach port at Meols on the Wirrall peninsula which apparently served as a landing place for trading vessels over a long period of time. Among the collection of unstratified material recovered from the site were three Carthaginian and two Coriosolites coins which suggest that the landing may have originated in the latter part of the Iron Age (Matthews 1999).

Figure 17.28 Distribution of Dressel 1A amphorae in Brittany and southern Britain (source: Cunliffe 1987 after Galliou 1982 and with additions).

In the last hundred years before the Claudian conquest, the tribes of south-eastern Britain seem to have developed a lively trade with Roman Gaul. Throughout the second half of the first century BC, Italian wine was imported in large quantities (as the widespread occurrence of Dressel type 1B amphorae shows) (Figure 17.30), and along with the wine came the tableware appropriate to its consumption. Bronze jugs (oenochoe) and bronze patellae, found in the burials at Aylesford and Welwyn, were imported from the Ornavasso region of northern Italy some time between 50 and 10 BC, while silver cups of Augustan date were brought in, eventually to be buried with dead chieftains at Welwyn and Welwyn Garden City. Bronze bowls and wine- strainers, probably of Gaulish or Italian manufacture, were also introduced. The forty years or so following Caesar’s conquest of Gaul saw the tribes of south-eastern Britain develop in parallel with their neighbours in Belgic Gaul who were now under Roman authority. Exchange was probably still manipulated through traditional mechanisms embedded in the social systems of the two areas (Cunliffe 1988a, 137–44), but with the more rigid control of Gaul and the development of the Rhine ‘corridor’ the old order gave way to a new commercialism.

Figure 17.29 Anchors from Britain: 1 iron anchor from Bulbury, Dorset; 2 lead anchor stock dredged up from near Porth Felen, Aberdaron (sources: 1 Cunliffe 1972b; 2 Boon 1977).

From about 10 BC until the Conquest of AD 43 the volume of trade seems to have increased. Fine Gallo-Belgic tablewares were being imported in quantity, together with smaller consignments of Arretine vessels made in northern Italy: later, samian ware from southern Gaul reached Britain. A limited number of glass vessels also made an appearance at this time. Wine too continued to be imported in bulk, as well as increasing quantities of fish sauce and olive oil from the Spanish province of Baetica, the two delicacies arriving in characteristic amphorae (Peacock 1971, 1984). In addition to food, wine and tableware, other luxury goods appeared, like the pairs of silver fibulae from Great Chesterford and elsewhere (Stead 1976a), the set of glass gaming pieces from the burial at Welwyn Garden City and the small medallion of the Emperor Augustus found in the tumulus burial at Lexden – the latter surely a diplomatic gift of some kind. In fact, in the few generations before the conquest, the wealthy members of south-eastern British society must have been able to enjoy much the same range of Roman consumer luxuries as their distant relations now living across the Channel in Roman Gaul.

While some of the imported commodities may have represented gifts used to establish and maintain diplomatic relationships there can be little doubt that regular trade was now under way on a large scale. Among the British exports corn and other raw materials would have featured large, but manpower, in the form of slaves, was always a desirable commodity in the Roman markets. In this context we might note the slave chains found at Bigbury, Lords Bridge and Llyn Cerrig Bach (Thompson 1994).

Figure 17.30 Distribution of Dressel 1B amphorae in eastern Britain (source: Fitzpatrick 1985a with additions).

Some insight into this flow of commodities is provided by a telling passage written by Strabo at the turn of the millennium, ‘Some of the chieftains’, he writes,

after procuring the friendship of Caesar Augustus… have managed to make the whole of the island virtually Roman property. Further, they submit so easily to heavy duties, both on the exports from there to Celtica and on the imports from Celtica (these latter are ivory chains and necklaces, and amber-gems and glass vessels and other pretty wares of that sort) that there is no need of garrisoning the island.

(Geog. 4.5.3)

A few decades after these words were written the Roman legions landed in Kent.

Summary of Continental contacts from the seventh century BC to the first century AD

Standing back from the mass of evidence briefly surveyed in the foregoing pages, it is possible to make a number of broad generalizations. From the eighth century until the middle of the fourth it would seem that the British Isles maintained vigorous trading contacts with the Continent, demonstrated by the appearance in these islands of a large number of metal types ranging from personal ornaments like pins and bracelets to swords and horse-harness fittings. Much of this exotic material is of central and north European origin, but the long-established trade routes along the Atlantic coasts of France and Iberia (leading ultimately to the Mediterranean) continued to be used, particularly by the traders distributing bronze weapons and tools manufactured in the local Atlantic Bronze Age styles. Imported types did not, however, materially alter the traditional range of weapon- and tool-types although many of the new ideas were adopted and rapidly modified by the British craftsmen.

The appearance in the eighth century of imported Hallstatt C swords together with harness- and cart-fittings was once thought to imply an incursion of a mobile, horse-riding aristocracy from Continental Europe, but no positive trace of a large-scale folk movement into Britain survives, and the virtual absence of the characteristic Hallstatt burial rite in Britain argues that no such movement took place. Indeed swords and horse-trappings were already being imported in the preceding period. The general picture to emerge is therefore of a broadly parallel development between Britain and the Continent, the two areas retaining a close contact, which encouraged a free flow of ideas and an exchange of goods, while indigenous traditions remained dominant.

This pattern was maintained throughout the fifth and fourth centuries, allowing the importation of luxury objects such as daggers and swords, by means of which elements of the contemporary Continental art styles first appeared in Britain. Along with the more valuable types came a scattering of trinkets including fibulae, bracelets and occasionally rings. Throughout this period the Atlantic trade routes to and from south-western Britain were maintained, the principal export being Cornish tin for the Mediterranean world. Occasional references to the tin trade in classical literature suggest that the system was well established and flourishing. No doubt equally complex trading patterns linked the south-east of the country to mainland Europe.

Some time just before 400 BC it seems probable that a group of people migrated from France to eastern Yorkshire, establishing a community which maintained its identity until, and even after, the Roman invasion. Apart from this one possible instance, there is no evidence to support the idea of a widespread invasion.

The material culture and settlement pattern of the third and second centuries suggest that Continental contact decreased during this period, allowing the British communities to assume an intensely regional aspect reflecting little of contemporary European developments. Contact was reintensified about 100 BC when the activities of the Roman entrepreneurs in southern Gaul impinged upon the Solent harbours and the south-west and bound these areas closely once more in an Atlantic trading network. But after Caesar’s conquest of Gaul this trade rapidly declined, to be replaced by even more intensive contacts between the Roman world and the pro-Roman tribes of eastern Britain. This system was subsumed when Rome invaded Britain in AD 43.