THE CLAUDIAN INVASION CAMPAIGN RECONSIDERED

Transcription

1 D.G. BIRD Summary. This paper attempts to build on Hind s hypothesis (1989) that the Roman invasion of AD 43 took place in Sussex, by examining the implications for the conquest campaign. It is suggested that the usual tactics of the Roman army and the size of the invading army are not consistent with the standard invasion campaign theory, in particular the events leading up to the so-called battle of the Medway. Other aspects of the evidence available to us are also inconsistent with the standard theory but can be explained in terms of a Sussex landing. The result is to offer further support for Hind s hypothesis. In a recent article J.G.F. Hind (1989) has challenged what has become the established version of events in AD 43 by arguing that the Plautian invasion force landed in the harbours behind the Isle of Wight, and avoided Kent altogether. Although this hypothesis has not yet been fully accepted it offers the opportunity to reconsider the Plautian conquest of southern Britain and perhaps make better sense of some of the anomalies, such as the rôle of the Dobunni, the curious absence of Caratacus and the early favouring of Verulamium. It also allows enough room for a very large Roman army to carry out the kind of campaign we should expect from it. 1 Hind has gathered and presented most of the available evidence (including a translation of Dio), and there is no need to repeat it here. Because of the need for very precise dating, archaeology can offer little help, although the increasing evidence for early Roman activity over the area between Fishbourne and Chichester is some support for the Hind hypothesis (e.g. Cunliffe 1998, 31). We remain largely reliant on the historical sources, and although they obviously include a degree of bias, especially towards the Flavian dynasty, there is no particular reason to suppose that the facts recorded are not presented more or less accurately, so far as was within the powers of the authors. The main source is of course the narrative of Cassius Dio, which seems to be a rapid but reasonably full summary in chronological order, based on earlier works which have not survived, presumably including the relevant part of Tacitus Annals. 1 A first draft of this paper was prepared in Since then I have been grateful for comments and assistance from a number of people, including Joanna Bird, Professor Barry Cunliffe, Geoffrey Dannell, Martin Henig, John Manley, John Mills and Mark Taylor. They would certainly not all agree with everything or perhaps anything of what I have written. There is, however, a growing body of opinion, especially among Sussex archaeologists, favouring the Hind hypothesis: see e.g. Rudling 1998, 98; Cunliffe 1998, 21; Henig 1998, 6. 19(1) ß Blackwell Publishers Ltd. 2000, 108 Cowley Road, Oxford OX4 1JF, UK and 350 Main Street, Malden, MA 02148, USA. 91

2 As Hind notes, there is not a mention of Kent in our sources and there is no need to assume that any part of the invasion force landed there, at least in the first part of the campaign. There is some evidence to suggest that the invasion fleet is more likely to have aimed for the harbours sheltered by the Isle of Wight, close to Verica s former kingdom. Hind (1989, 14) argues that this was a restoration, immediately providing the prospect of a friendly base and good harbours, well-known from cross-channel trade. He offers several parallels for the longer sea crossing, to which might be added the expeditions along the North Sea coast of Germany in the early years of the first century, involving considerable troop movements by sea over distances similar to those involved in an invasion headed for the area of the Isle of Wight (see e.g. Saddington 1990, 224 5). A Sussex landing would also offer the immediate prospect of a good corn supply: Plautius had a very large army to feed. The Sussex coast exported corn to the Continent in the medieval period (Campbell et al. 1993, 86, cf ), and its evident riches in the Roman and Iron Age periods suggest that the good soils were being exploited to the full earlier on. No doubt this was one of the areas which acted as a source for the corn exports noted by Strabo (cf. Cunliffe 1994, 80, who also notes earlier invasions in the Solent area). The supply of Roman armies on campaign is little understood (Goldsworthy 1996, 287), but there can be little doubt that locally available food sources played an important part. North-east Kent was also a major corn-producing area but as it was apparently in a potentially hostile area it might have been felt that there was a risk of facing the scorched earth policy tried against Caesar by Cassivellaunus (Dio notes the Roman concern about the possibility of an opposed landing). These factors may help to explain the site at Richborough. An island on the edge of a marsh (Hawkes 1968, 225) seems an odd place to choose for an invasion. The well-known early fortifications are described as follows: Although occupation was of a temporary nature, the double ditches and the well-built gate indicate that the camp was intended to be semi-permanent (Cunliffe 1968, 233). This semi-permanence is curious; does it in fact imply that the fort was in potentially hostile territory for a while, and was not a beachhead fortification so much as a fort placed to control the Wantsum channel? The granaries of the supply base are later than the fortifications, and may indicate exploitation of the corn land of Thanet for outward shipment (these arguments are developed more fully in Bird 1999). The monumental arch is often said to be a commemoration of the point at which Britain was first invaded, but there is no evidence that this was the case. In fact Strong (1968, 72) concluded that there is little doubt that its chief purpose was to symbolise the accessus Britanniae (i.e. to mark the entrance to (or exit from) the province). The commanders of the Spanish Armada considered the use of the Solent for an invasion and therefore thought of capturing the Isle of Wight (Mattingly 1959, 263). Vespasian of course is noted as having done so. It is difficult to see why it should have rated his attention sufficiently to have entered the records if it did not form part of the invasion strategy. He is also credited with overcoming two very strong tribes, fighting some thirty actions and capturing twenty oppida (Hind 1989, 3). These events of course are usually held to have happened after the capture of Camulodunum, but it is interesting that Suetonius tells us that they occurred partly under the leadership of Aulus Plautius and partly under the overall command of Claudius himself. The order in which this is put, with the emperor second, may suggest that most of the events had happened by the time of the capture of Camulodunum. This is perfectly possible if 92 ß Blackwell Publishers Ltd. 2000

3 D.G. BIRD the main point of Vespasian s actions was to secure the army s left flank against the Durotriges and the anti-roman part of the Dobunni in the early invasion period. 2 Dio says that the Romans achieved surprise, and that the Britons had not mustered their forces. It will therefore have taken Caratacus and Togodumnus some time to gather their armies. We are told that once he had successfully landed, Plautius had difficulty coming to grips with the enemy because they melted into the marshes and forests (quite likely the Wealden forest and the marshy areas fringing it: cf. Hind 1989, 17). Roman armies did not fight to control territory; wars were always fought against a political entity, such as a tribe or state. A decision was reached when the enemy s will to fight was broken (Goldsworthy 1996, ). In fact the Roman army seems to have acted much like the British navy used to: in all circumstances take the offensive (Goldsworthy 1996, 114, 141 2). Plautius had a very large army; it was, for instance, about five times the size of William the Conqueror s force (Hind 1989, 14 n. 66). If he had landed in Kent, he would very soon have set off towards his main enemies, and it is unlikely that they would have had time to meet him in strength before he had crossed the Medway. He would have had to leave forces to guard his coastal base and communications even in theoretically friendly country: he could hardly clear the population, friendly or not. To bring the main enemy forces to battle he need only strike towards some major objective for which they would have to fight (cf. Goldsworthy 1996, 102). Eventually Plautius defeated first Caratacus and then Togodumnus and put them to flight (translation Hind 1989, 7). It is difficult to read Dio as meaning anything other than that these were not guerilla actions, and that the two leaders were not operating together. Yet this makes little sense if the Roman army was still in Kent, on the eastern side of the Medway. There is relatively little campaigning room in Kent north of the Downs (and the Weald) and we are specifically told that it was some time before Plautius was able to bring the Britons to battle, so they clearly had the opportunity to band together. Even if Caratacus offered battle first, why did not he and his remaining forces fall back on Togodumnus, who must have been somewhere nearby? Modern authors frequently write as though Togodumnus and Caratacus jointly led the resistance to the invasion, but at no point do our sources suggest that they acted in concert. 3 Cassius Dio mentions them together only when he says that they dominate a people he calls the Bodunni. There is archaeological evidence that this people had at least two separate groups (e.g. Frere 1987, 54), and we are specifically told that Plautius secured the alliance of only a part of them; it could be, therefore, that one group was paying tribute to Togodumnus and another to Caratacus. We are told that Caratacus was defeated first. This makes good sense if the invasion was in southern central Britain, because his power base was probably nearer; what little 2 Suetonius says that Vespasian was rewarded with triumphal ornaments and, very soon after, two priesthoods, and the consulship besides, which he held for the final two months of the year (translation Hind 1998, 3). Eichholz argument (1972) that he could not have gone to Rome to take part in Claudius triumph depends largely on the established version of events: Vespasian would not have had time to capture the Isle of Wight, etc, if he had had to start after the capture of Camulodunum. This difficulty is removed if Vespasian had already completed much of his campaign before Claudius arrived, and allows for the natural reading of Suetonius words to stand. We might note also that if the established story is correct, then Vespasian and Legio II must have been closely involved in the restoration of Verica or the installation of Cogidubnus/Togidubnus. Why then does this event not figure in the stories about him, when other events such as the conquest of the Isle of Wight do? 3 Noting the separate defeats of Caratacus and Togodumnus, Salway (1981, 83) suggests that there may have been landings both in Kent and Sussex. He still, however, prefers a main landing in Kent. ß Blackwell Publishers Ltd

4 evidence we have places him on the western side of the Catuvellaunian/Trinovantian empire (Cheesman 1994, 54, 60). He is said to flee, and is not heard of again in the invasion campaign. The next we know of him is when Ostorius Scapula has replaced Plautius and Caratacus is mentioned as operating with the Silures and the Ordovices in AD 47. Interestingly, he does not figure with the revolt of the Iceni, when Ostorius disarms them; he is also described as having considerable prestige because of his many undefeated battles and even many victories. Tacitus tells us that his name was known even at Rome (Annals XII.32 36). With such a high profile he would surely have been mentioned if he had taken further part in the invasion campaign. It may be suggested therefore that he fled first to the Dobunnic area, and, failing to get support (and perhaps triggering the embassy to the Romans?) had no choice but to continue into Wales, where he was kept out of events by a combination of a Roman garrison and unfriendly tribes between him and Camulodunum (for a consideration of his failure to raise the Verulamium area see below). According to Dio, Togodumnus was beaten next, and he too fled. For the separate defeats to be logical it makes sense for Caratacus to be beaten somewhere south of Silchester and driven away north-west before Togodumnus, coming from the east, was defeated somewhere in the general area of Silchester. It is reasonable to assume that there were wellestablished routes between Selsey/Chichester and Silchester and Silchester and Camulodunum and that these will have been used by the armies on both sides (see also below). Cassius Dio specifically links the flight of Caratacus and Togodumnus to the establishment of a Roman alliance with a group of the Bodunni. It is generally accepted that this means the Dobunni. This makes little sense unless the Roman army was within reasonable striking distance, which is strongly implied by Dio s words: Plautius left a garrison there and moved forward (Hind 1989, 7). Some have been able to accept the established version and send a flying column out to the Dobunni, but this seems unlikely. Those who have the Romans stuck in Kent have to allow the flying column to cross the Medway before the main army has to fight for two days to do so, and those who prefer a subsidiary landing in Hampshire/Sussex have to allow the Roman command to accept a division of forces and a garrison way out on a limb (an action which we might note would contrast strangely with the supine behaviour attributed to the main army after it had beaten both the main enemy leaders). 4 According to Cassius Dio, following the placing of a garrison with the Dobunni, Plautius moved forward. In the established version of events, Plautius, having defeated both Caratacus and Togodumnus sufficiently well to drive them to flight, failed to take the elementary step of pursuing them to the extent that the Britons were able to stand again behind the first major obstacle. Contrast this, for example, with the reaction after the battle on the river. If, however, we accept that Plautius was operating in the area around Silchester, and that he 4 Frere (1987, 50) avoids these problems by reading Cassius Dio to mean that the Romans did not go to the Dobunni but that the garrison was left at the crossing of the Stour at Canterbury. But the force of Dio s narrative seems clear enough, and why should this garrison be the only one mentioned in the entire campaign unless it was special? The supposed military occupation at Canterbury is rather less well attested than is sometimes implied (see Bennett et al. 1982, (e.g. a possible military occupation [28], not necessarily pre-boudica [30]). By contrast, military occupation at Chichester, Silchester and Cirencester, early enough to be in the invasion year, is accepted (Down 1998, 6 16; Wacher 1995, 272 3, 302). Those who worry about the Cirencester garrison as being a long way from the other forces, should remember the way in which Caesar operated with single legions in friendly territory and the accepted single legion groups splitting up to tackle different parts of Britain after AD ß Blackwell Publishers Ltd. 2000

5 D.G. BIRD took time after the defeat of Caratacus and Togodumnus to help part of the Dobunni (and incidentally to split the opposing forces for good, driving Caratacus away towards Wales), then there is a logical break during which it would be possible for Togodumnus to recover and gather fresh forces. This version of events fits more naturally with Cassius Dio s story than the more forced established version. Where then did Plautius go next, and where was the battle on the river? I cannot follow Hind in having the Roman army go back to Sussex and strike out across the Weald towards London (Hind 1989, 17). His reasoning is that it is necessary to find a route to the Thames with only one river, but Cassius Dio does not say this. All that is needed is a river somewhere on the route which the Britons (wrongly) thought the Romans could not cross without a bridge. A crossing of the Weald would have been out of the question with an army, because it would have been necessary at some point to cross the Weald Clay; indeed, on the Stane Street route, as suggested by Hind, it would be necessary to cross it at its widest extent (see Fig. 1). A quotation from Cobbett s Rural Rides should be sufficient to illustrate the problem... the real Weald of Surrey... From Ewhurst the first three miles was the deepest clay that I ever saw, to the best of my recollection. I was warned of the difficulty of getting along; but I was not to be frightened at the sound of clay. Wagons, too, had been dragged along the lanes by some means or another; and where a wagon-horse could go, my horse could go. It took me, however, a good hour and a half to get along these three miles. Now, mind, this is the real weald, where the clay is bottomless; where there is no stone of any sort underneath... (Woodcock 1967, 163). Following this route with an army would be madness: had Plautius taken his army across the Weald he would have invited a disaster the like of which had not been seen since Varus: and to what purpose? 5 If, as suggested, Plautius had in fact reached the area around Silchester, it would form a very useful base at this stage. His newly freed allies were behind him, helping to guard and provide his supplies and his communications back to the Continent; his right flank was protected by the largely uninhabited Weald and his left by Vespasian s conquests; he was very well-placed to strike in any direction, and to make diplomatic contact with tribes to the north. It is interesting how many Roman roads join at Silchester: it is as well served as London (see also note 4). Plautius main objective, however, was still to the east, and there can be little doubt that it was always an aim of the invasion to take Camulodunum and make it into the new Roman capital. The Surrey area was probably former Atrebatic territory, at least in part, and therefore potentially friendly, but beyond that Kent was likely to be a difficulty. There can be little doubt that it had been effectively allied to the tribes north of the Thames for a considerable time; a good reason incidentally for avoiding it with the initial invasion. Plautius, 5 Ewhurst is close to the Stane Street line. I am not suggesting that the Weald was impassable. I am sure that its resources were used and that there were well-known paths, just as people knew the tracks through the marshes at the Thames crossing, and the Britons drove their animals into the woods away from Caesar. But this does not make it suitable terrain for an army. Magilton s suggestion (1996, 32) that Stane Street was laid out at an early date but originally directed to the Pulborough area strengthens rather than weakens this proposal. A direct route between Chichester and the London area would take a different line. If this was not possible at first then a line north from Chichester, turning to the east once north of the Weald Clay, might be expected. In this respect it is interesting that a recent study notes that there is no reason why a more direct route between Chichester and Silchester could not have been used (i.e. than the one taken by the known Roman road) (Turner 1997, 43). It may be that there is a road to be found between Iping and the Staines area (cf. Bird 1999). ß Blackwell Publishers Ltd

6 96 ß Blackwell Publishers Ltd Figure 1 Map of south-eastern England to illustrate places mentioned in the text. B = Braughing; Ca = Canterbury; Ch = Chichester; Ci = Cirencester; C = Colchester; Fi = Fingringhoe; F = Fishbourne; I = Iping; P = Pulborough; Re = Reculver; R = Richborough; S = Silchester; St = Staines; V = Verulamium; W = Winchester. Principal rivers are shown, and the line of the Wantsum channel between Richborough and Reculver is marked. The extent of the Weald Clay is shown by heavy lines. Some Roman roads are shown. The roads that later served London, from Canterbury, Chichester, Braughing and Verulamium are shown by dashed lines. Thic is not, however, intended to imply that the other roads shown were necessarily in existence as Roman roads in the earliest years of the conquest. (Drawing by the author.)

7 D.G. BIRD therefore, probably now took the bulk of his army along the Thames corridor, and it is somewhere along in this direction that we must look for the site of the river battle. That site is likely to have been reasonably near the Thames tidal lake as the crossing there seems to have been undertaken while the Romans were actively pursuing the Britons as they retreated after the battle. It must also accommodate the possibility of chariots operating in the area on and beyond the far bank. If as seems most likely Plautius struck along reasonably close to the south bank of the Thames, then one possibility for the river crossing would be the Wey, near Weybridge. Indeed this river s confluence with the Thames is not far from the confluence with the Mole; a contested crossing of one followed immediately by a contested crossing of the other might explain the need for a two-day battle, not properly understood by Dio or his source. But other rivers are also possible, even north of the Thames. 6 Following the battle, the defeated Britons were pursued to a crossing of the Thames at the tidal pool, presumably near the later site of London. An unforced reading of Dio suggests that the position of the bridge used by part of the army to cross further upstream is to be sought nearby, perhaps somewhere near Putney. (See Grimes 1968, 44 for a suggested prehistoric crossing point here. It is easy to imagine that the bridge will have looked something like the early eighteenth century structure that survived at Putney until 1886, illustrated in Vine 1994, 18.) Cassius Dio tells us that Togodumnus had died about this time, but the Britons, far from yielding, joined together all the more firmly to avenge his death. Because of this, and because of the losses encountered at the Thames, Plautius took alarm and advanced no further. Instead he proceeded to guard what had already been won and sent for Claudius; this is what he had been ordered to do, if there was any particularly stubborn resistance. Indeed extensive preparations had already been made in advance by way of gathering together various types of equipment, including elephants, to back up the invasion force (Hind 1989, 7). This is an interesting passage, but few would accept it as it stands. It surely originates in the official version of events, designed to stress Claudius part in the proceedings, and we are entitled to read between the lines. First of all we should note that Togodumnus has died, but there is no word of Caratacus now leading the resistance. In view of his fame in Roman eyes, can we doubt that he would have been mentioned if he was involved? Secondly, is it possible to believe that Plautius would have stopped at this point because of the resistance? He had already beaten both Caratacus and Togodumnus separately and put them to flight, fought and won an apparently difficult two-day battle and had now achieved a situation where the enemy had no leader of sufficient quality for his name to have entered the record at its fall, Camulodunum is described as the place which had been the 6 Ernest Black has recently suggested (1998) that the river battle may have been on the Thames itself. He argues that Dio was confused by his sources into believing that there had been two river crossing battles (on an unnamed river and the Thames) when in fact there was only one. This idea certainly has points in its favour but it is perhaps ultimately unconvincing. Although there are undoubtedly similarities between the engagements there are also some differences that seem difficult to explain even if two very different sources are postulated. For example in one event the Britons are camped on the opposite bank, in the other they are retreating across the river closely followed by the Roman army. In the first battle there is scope for chariot warfare while in the second the setting does not seem appropriate for chariots to be used. Again, the first battle is described as a sound defeat for the Britons, while the second ends inconclusively. In the first Plautius is described as in control, in the second he is not mentioned and the episode reads like an incautious pursuit. ß Blackwell Publishers Ltd

8 Figure 2 98 ß Blackwell Publishers Ltd. 2000

9 D.G. BIRD ß Blackwell Publishers Ltd

10 Figure 2 (On pages 98 to 99) A possible reconstruction of the invasion campaign. The Weald Clay is marked by vertical lines and the line of the Thames is shown. The Wantsum channel is marked but no other attempt is made to show the Roman period coastline. The dots mark the areas seen as occupied by peoples allied to or submitting to the Roman army at each stage. Abbreviations for places are the same as for Figure 1. 1 Landings at and near Fishbourne. Development of bases and the securing of the Sussex corn supply. Vespasian leads the subjugation of the Isle of Wight and then campaigns against the eastern Durotriges. The main army drives north and defeats first Caratacus and then Togodumnus. 2 Some of the Dobunni seek help and Plautius drives Caratacus and his surviving followers into Wales, placing a garrison at Cirencester. Verulamium provides assistance in some way at about this time, perhaps refusing aid to Caratacus and offering support and supplies to the Roman army. The eastern Britons regather their strength. The fleet is active around the coasts and perhaps now a fort at Richborough is established. 3 The Cirencester garrison and the Veralamium Catuvellauni, probably aided by other groups further north, prevent Caratacus from gathering forces to aid Togodumnus. This would also be a logical time for the forts along the Durotrigian frontier. Plautius advances east and defeats the Britons at the river battle. Incautious pursuit to the Thames near the tidal pool leads to losses. Plautius sends for Claudius and stops further advance north of the river. While waiting for the emperor he gains the submission of tribal groups in Kent and Surrey and prepares for the advance to Camulodunum and the meeting there between tribal leaders and Claudius. 4 Claudius sails to Fishbourne and instals Togidubnus (the elephants perhaps travel via Richborough). He then journeys north and joins the army by the Thames for the final advance on Camulodunum, enjoying a largely token victory en route. After receiving the submission of various tribal leaders he returns to the continent via the Wantsum channel and Richborough. (Drawing by the author). capital of Cunobelinus. The general impression then is of an unfocused and fairly hopeless resistance. It is significant, too, that the preparations supposedly to provide further support for Plautius included special flourishes, such as elephants. There can be little doubt that these preparations were designed as special accompaniments to Claudius visit. He himself had to stay in Rome until he could be sure of a success, but the entire trip was evidently planned well in advance. If Claudius really felt the need for a personal military success to cement his hold on power, then he cannot have been too happy about being well away from Rome and particularly on the other side of the Channel. He will not have wanted to spend more than the 16 days he is said to have spent in Britain. He could not therefore afford the prospect of serious resistance when he arrived: even minor reverses would have interfered with his programme and he could hardly leave again until a proper success had been achieved. We may accordingly be sure that Plautius was under orders to stop before the final victory, but that he was also under orders to ensure that the victory was in the bag. Plautius would not have sat still to wait for Claudius, as some seem to think. From the military point of view this could have been disastrous: if the Britons were really joining together all the more firmly to avenge Togodumnus, then to sit and let them regather their strength would have been almost criminal in a commander. Under any circumstances, to stop and wait would have been to invite renewed opposition, and it is unlikely that the emperor would have thanked him for that. It may therefore be suggested that Plautius did indeed stop any further advance on Camulodunum, but that he pursued the overall objective elsewhere. In particular now would be the time to deal with any opposition in Kent, perhaps working from both directions, with the support of the fleet (hence the base at Richborough and other related evidence). 100 ß Blackwell Publishers Ltd. 2000

11 D.G. BIRD This would also have been the time for diplomatic effort to prepare the ground for the submission of tribal leaders to Claudius and complete the isolation of Camulodunum. It is interesting to consider the Verulamium factor. The Catuvellauni are supposed to have been the Romans main enemy, and in due course Verulamium became the chief town in their tribal area in the Roman period. Yet far from suffering as a result the town appears to have had specially favoured status as a municipium from early on, and the area around it soon appears prosperous (see e.g. Millett 1990, 100, table 4.3, and Wilkes 1996, 28, who describes it as a Romanised British settlement full of rich collaborators already [sc. at the time of Boudica] enjoying some of the luxuries of a Roman lifestyle ). Even if it was not a municipium this early, as some would argue, it was still clearly of considerable importance by AD 60. A useful parallel might be the case of Volubilis: this town helped Rome during the war in Mauretania, shortly before the invasion of Britain. In return it received Roman citizenship, municipal status and exemption from taxation for ten years (Scullard 1963, 308). Claudius is well-known for his pro-provincial policies. If Verulamium was really a municipium by AD 60 it can only have received this status at the hands of Claudius or Nero. Salway (1981, 751) discusses the possibility that Claudius wanted to advance some of the Britons. Could it be that Verulamium and its territory sided with Plautius and received its due reward? Reasons might include pique at the transfer of the main seat of power to Camulodunum and a previous pro-roman feeling. Perhaps the leader of a tribal sub-group opposed Caratacus and Togodumnus or had been overpowered by them (like their brother Adminius, driven to take refuge with the Romans a few years before the invasion, and possibly even, therefore, travelling back with them [cf. Wacher 1995, ]). This suggestion has the merit that it also helps to explain why Caratacus was apparently unable to assist with the final defence of Camulodunum, and why there is so little sign of any need for campaigning in Catuvellaunian territory (the evidence for previously claimed forts is flimsy; see e.g. Bird 1996, 228 n. 14). It still seems to be the usual assumption that, having stopped his advance, Plautius set up a base at or near London. Yet the results of recent archaeological activity suggest that no such base existed. London itself appears to be a deliberate foundation around AD 50 55, and there is nothing to suggest that the roads that served it, which required considerable engineering across the marshes, were built before this (Bird 1999). It is important to remember that the crossing at the tidal pool mentioned by Dio was made in pursuit of a beaten army: it was not a crossing place chosen by the Romans and proved to be difficult to manage and dangerous to them once they were across. It was also obviously not the crossing place for a normal route but one used by the retreating Britons precisely because of its considerable difficulties, in an attempt to shake off pursuit. There is accordingly no reason to suppose that the Roman army will have chosen to cross here again during the campaign, and it is not surprising therefore that no Plautian base has been found at or near London (discussed more fully in Bird 1994). Another reason for the suggested London base is because Dio tells us that when Claudius arrived he joined the army which was waiting at the Thames. The usual interpretation is that the troops were waiting at London, but this is just because some of Plautius army was last mentioned hereabouts. We need not assume it and it is not a likely crossing point for the reasons set out above. The army need not have been waiting long at the Thames; they could have been ordered up to a start line to coincide more or less with Claudius arrival. ß Blackwell Publishers Ltd

12 In his 16 days in Britain, Claudius landed, made his way to the Thames, defeated the Britons and took Camulodunum, following which, according to Cassius Dio, he won over many peoples, some by diplomacy, some by force of arms, and he was saluted as imperator several times... He confiscated the weapons of these peoples and handed the tribes over to Plautius, and left him with orders to subdue the remaining regions (translation Hind 1989, 7). If there is any truth in this story then Claudius can hardly have had time to draw breath! Even if he sailed back directly from Camulodunum he could only have had nine or ten days there at most. 7 We must therefore see the visit as a show, designed to impress the British chiefs and their aristocracy and to allow Claudius to take part in the army s triumph, and then, in due course, to put on another show in Rome. It will have been Plautius job to ensure that all went smoothly for the visiting emperor, and with such a tight timetable everything must have been carefully planned. Claudius may have entered the country by the Kent route, but that area might still be rather dangerous. It would be more pleasing and ceremonious to go to the area of Chichester (from which the distance to Camulodunum is not a great deal further than from Richborough that is, if a London area crossing is used). Under these circumstances it is justifiable to imagine Claudius landing at Fishbourne, allowing a restoration of Verica (or installation of Cogidubnus/Togidubnus) in person. Nothing could be more calculated to demonstrate that a Roman alliance was now something to be reckoned with. Claudius could then join his army, ready to enjoy a token victory over the leaderless remnants of the Catuvellauni (and/or Trinovantes), and ceremoniously enter Camulodunum. Here, symbolically at the enemy capital, he will have held court, and accepted the homage of a number of peoples, previously arranged by Plautius. The elephants, and no doubt other suitably imperial trappings, will have added greatly to the scene (surely their main purpose). Our only evidence for the events of the invasion is Dio s text and the few extra references in other sources. The hypothesis set out here together with Hind s earlier work fits those sources at least as well as the established version, and in several cases provides a more natural interpretation. In particular it allows a very large Roman army to behave as we should expect. Archaeological evidence is simply too imprecise to demonstrate which theory is the more likely, but the established version of events would not explain why the Romans were so obviously totally in control when Claudius arrived. In particular, where was Caratacus? Why had no effort been made to free Verica s kingdom and why is it not mentioned as a task for Plautius once the emperor had gone? The restoration of Verica (or installation of Togidubnus) is indeed curiously absent from the texts, given that Dio and Suetonius both suggest that it is likely that the official version of events was that the invasion was undertaken to aid a Roman ally. The explanation could be that this is because it was taken for granted as being achieved as soon as the Romans landed. The hypothesis set out above provides a more logical answer to these problems and fits better with the apparent ease of the conquest of the whole of south-east and southern central Britain. It accords better with other Roman campaigns than the established version of events, which has a First World War feel about it. After all, we are told that Caratacus and 7 Barrett (1998) suggests, on the basis of a coin from Alexandria, that Claudius must have completed his British success by mid-august of AD 43 at the latest, and probably rather earlier. This tends to support the hypothesis presented here, in that it emphasizes the speed of the conquest of south-east England. I am grateful to Martin Henig for drawing this reference to my attention. 102 ß Blackwell Publishers Ltd. 2000

13 D.G. BIRD Togodumnus were sufficiently heavily defeated to be put to flight. After his one victory at Hastings William the Conqueror could roam over all of south-east and southern England Kings Road Guildford Surrey GU1 4JW REFERENCES BARRETT, A.A. 1998: The date of Claudius British campaign. Classical Quarterly new series 48.2, BENNETT, P., FRERE, S.S. and STOW, S. 1982: The archaeology of Canterbury vol 1. Excavations at Canterbury Castle (Maidstone, Canterbury Archaeological Trust). BIRD, D.G. 1994: The origins of Roman London. London Archaeologist 7.10, BIRD, D.G. 1996: The London region in the Roman period. In Bird et al. 1996, BIRD, D.G. 1999: Early days at London and Richborough. London Archaeologist 8.12, BIRD, J., HASSALL, M. and SHELDON, H. (eds.) 1996: Interpreting Roman London. Papers in memory of Hugh Chapman (Oxford, Oxbow Monograph 58). BLACK, E.W. 1998: How many rivers to cross? Britannia 29, CAMPBELL, B.M.S., GALLOWAY, J.A., KEENE, D. and MURPHY, M. 1993: A medieval capital and its grain supply: agrarian production and distribution in the London region c 1300 (Historical Geography Research Series, 30). CHEESMAN, C. 1994: The coins. In O Connell, M. and Bird, J., The Roman temple at Wanborough, excavation Surrey Archaeological Collections 82, CUNLIFFE, B. (ed.) 1968: Fifth report on the Excavations of the Roman fort at Richborough, Kent (Report of the Research Committee of the Society of Antiquaries of London, 23). CUNLIFFE, B. 1994: After hillforts. Oxford Journal of Archaeology 13.1, CUNLIFFE, B. 1998: Fishbourne Roman palace (Stroud). DOWN, A. 1988: Roman Chichester (Chichester). EICHHOLZ, D.E. 1972: How long did Vespasian serve in Britain? Britannia 3, FRERE, S. 1987: Britannia, a history of Roman Britain (3rd ed. London). GOLDSWORTHY, A.K. 1996: The Roman army at war 100 BC AD 200 (Oxford). GRIMES, W.F. 1968: The excavation of Roman and medieval London (London). HAWKES, S.C. 1968: Richborough the physical geography. In Cunliffe 1968, HENIG, M. 1998: Togidubnus and the Roman liberation. British Archaeology 37, 8 9. HIND, J.G.F. 1989: The invasion of Britain in AD 43 an alternative strategy for Aulus Plautius. Britannia 20, MAGILTON, J. 1996: Roman roads in the Manhood Peninsula. In Southern Archaeology, The Archaeology of Chichester and District 1995 (Chichester, Chichester District Council), MATTINGLY, G. 1959: The defeat of the Spanish Armada (Oxford). 8 Obviously circumstances were different but after all Plautius had potential allies. He also had a far larger army: Stenton (1943, 584) says William probably had fewer than 7000 men (Hind [1989, 14 n. 66] gives 8000). Apart from occasional setbacks which caused units to require bringing back up to strength, the Roman size of army proved sufficient to take and hold most of Britain, often operating as groups based on only one legion or less. ß Blackwell Publishers Ltd

14 MILLETT, M. 1990: The Romanization of Britain. An essay in archaeological interpretation (Cambridge). RUDLING, D. 1998: The Roman villas of Sussex, with particular reference to the villas at Beddingham and Bignor. Proximus 2 (publication of the archaeological round-table Dieppe 17th and 18th September). SADDINGTON, D.B. 1990: The origin and nature of the German and British fleets. Britannia 21, SALWAY, P. 1981: Roman Britain (Oxford). SCULLARD, H.H. 1963: From the Gracchi to Nero. A history of Rome from 133 BC to AD 68 (2nd ed., London). STENTON, F.M. 1943: Anglo-Saxon England (Oxford). STRONG, D.E. 1968: The monument. In Cunliffe 1968, TURNER, D. 1997: The course of the Roman road north of Chichester. In Southern Archaeology, The Archaeology of Chichester and District 1996 (Chichester, Chichester District Council), VINE, P.A.L. 1994: London to Portsmouth Waterway (Midhurst). WACHER, J. 1995: The towns of Roman Britain (2nd ed., London). WILKES, J. 1996: The status of Londinium. In Bird et al. 1996, WOODCOCK, G. (ed.) 1967: William Cobbett. Rural rides (Harmondsworth). 104 ß Blackwell Publishers Ltd. 2000