Rome's Greatest Defeat

by Adrian Murdoch

  In one sense, only the harshest critic would have no sympathy at all with Varus for rejecting Segestes’ warning. He and Arminius had been constant companions throughout the summer. The Cherusci were not only allies and had given no sign of resenting the Roman yoke, but Arminius himself was a well-respected and integrated leader. He was a friend. His brother was a ranking Roman officer. Even Segestes’ son was working for the Romans – a priest at the Altar of the Ubii in Cologne. It was not just that Segestes’ tale sounded implausible; to him the warning must have appeared a ploy to involve the Romans in some sort of an intra-tribal dispute. Little surprise that, much in the same way that he had refused to act as arbitrator between Antipater and his father, so, too, Varus refused to be drawn in and intervene here.

  In another sense, however, Varus can be damned for misinterpreting his intelligence. That Segestes came to the commander suggests that he had what can loosely be called a counter-intelligence network of his own. The Romans had no compunctions about using spies and they did so on a regular basis. Varus himself had used the intelligence services in quelling the revolt in Judaea, thirteen years previously. Painting Arminius’ father-in-law as ‘M’ may be stretching a point, but there is no getting away from the fact that Segestes, either on an official or unofficial basis, had collected what he knew to be important strategic data, analysed it, and come to the conclusion that a revolt was brewing. Varus’ failure here was not one of intelligence per se. He knew about the revolt. Where the commander fell short is in not understanding what he was told.

  While the burden for the tragedy that was about to play out must inevitably be shouldered by Varus, this distraction of, and narrow emphasis on, the rejection of Segestes’ intelligence report in the classical historians hides a much broader Roman blunder. Varus not only misunderstood the extent to which Germany was in any sense pro-Roman or even pacified, he completely failed to appreciate the potential for unity among the Germanic tribes. The degree to which Augustus was at fault here is rarely considered. Had the emperor and the Senate begun to believe their own propaganda that Germany was essentially a province? ‘It was a miscalculation of massive proportions, caused by incomplete intelligence on several levels,’ writes one modern military historian.4 The dinners, the close comradeship, indeed everything that was to take place indicates that the idea of betrayal never once crossed Varus’ mind.

  The trap was sprung with a finesse bordering on genius. Arminius had organised an uprising within the fledgling province to draw the Roman army out. This was a manoeuvre requiring knife-edge precision. It had to be a significant enough incident for Varus to feel he needed to lead the army personally to put it down, but not so serious a revolt as to awaken suspicions either that this was a trap or that it was a precursor to a more concerted national uprising. Arminius’ plan was to ambush the Romans en route when they were off their guard. He reckoned that they could be overpowered easily while marching through what they believed to be friendly territory.

  Frustratingly, Cassius Dio does not mention which tribe it was that served to draw out the Romans, though it is possible to suggest a plausible candidate. Some may be discounted straight off, as they never broke their oaths of loyalty to the empire. The Ampsivarii were one of the few that were to remain loyal to Rome throughout the revolt; indeed their chieftain was to end up being arrested by Arminius for refusing to join him. The Chauci, whose territory was the north-west German coast, were in a similar position. They had been soundly beaten by Tiberius. Thereafter, during the campaigns of Germanicus, they even supplied support troops to help him. The Batavi, a tribe which occupied the area at the mouth of the Rhine around the modern town of Leiden, had supplied the emperor’s personal guards since AD 5 and were used as a base for Germanicus in the campaigns of AD 16, so they can probably also be discounted. It is a similar story with the Frisians, who occupied the modern Dutch provinces of Friesland and Gronigen.

  That leaves us with the Chatti, the Bructeri and the Angrivarii. The Chatti, whose homeland pretty much corresponds to the modern state of Hesse, certainly had no love of Rome. They had been perennial thorns in Drusus’ side and there is no doubt that they did take part in the rebellion. In fact, they were the first of the tribes to suffer Germanicus’ wrath in AD 15, but they do not appear to have been the Germans alluded to here. Their territory was simply too far south for Varus to risk marching on them.5

  The last two tribes are plausible suspects. Like the Chatti, the Bructeri certainly took part in the uprising and had been consistently staunch opponents of Rome. Based in the general area of the town of Münster, between the Ems and the Lippe, the tribe was in the ascendant at the moment. Although it had been nominally defeated in a naval battle by Drusus, and then again by Tiberius a decade previously, it was a staunch ally of Arminius and played a leading role at Teutoburg. Even Germanicus’ campaigns were not enough to quell the tribe’s revolutionary tendencies. It was a prime instigator in the uprisings that shook Germany in AD 69–70. At the end of the first century, Pliny the Younger was still able to refer to the Bructeri as ‘savage people’.6 Again, however, it remains unlikely that they caused this diversionary uprising. Kalkriese is too far north from their territory and a geographically implausible march for Varus to have made.

  The Angrivarii remain the best guess. It is credible for Varus to have marched on their territory, which was predominantly between the Weser and the Elbe, just north of the modern town of Hannover. The tribe’s identity is confirmed if Arminius’ tactical imperatives are considered. As the Angrivarii were to the north of the Cherusci, Arminius’ troops did not have to travel far and so there was little risk that the Romans could have suspected what they were plotting. The final corroboration is that the Angrivarii were one of the major objects of Germanicus’ wrath.7

  As the Roman legions marched out, heading north-west, they had no inkling of what was to come. Varus was later to be criticised for not maintaining tactical integrity as he marched, for not ordering the army to march in a state of full war-readiness. The train was scattered with civilians, men, women and children from the camp. As with so much that was written at the time, it is difficult not to see this as wisdom after the fact. It was a march through friendly, not hostile, territory.

  Nonetheless, the sight of the Roman army in full marching order, stretching out over some 8 to 10km, must have been an impressive sight. Lightly armed auxiliaries and archers marched first. Their role was to act as scouts, to identify any potential trouble spots and to neutralise them. It is likely that Arminius was part of this vanguard. Certainly it made sense to have those who knew the landscape and the routes go first, to check that all was clear before the army came through.

  After some heavy infantry and cavalry, which would act as support troops for the archers and auxiliaries should the van be attacked, came the engineers and a number from every legion, carrying flags. The engineers would work to make sure that the army could proceed as easily as possible, removing logs, cutting down trees, building or rebuilding bridges over rivers and marshland. The army was marching along country paths, not paved, cambered Roman roads. The flag-bearing legionaries were those who would mark out the camp in the afternoon. They had to make sure that the camp had access to firewood for legionaries, fodder for the horses and water for both. They had to take topography into consideration (so that a freak storm did not result in a river flowing through the camp) and strategy was a factor too.

  Then came Varus himself, along with his senior commanders, those like Numonius Vala, his deputy, and the camp commanders, Lucius Eggius and Ceionius. To protect what was the heart of the army, the senior officers were shielded by select infantry and cavalry, the best of the best. Traditionally, the senior officers were followed by mules dragging the artillery, as they would have been here. The presence of catapults is attested by a number of bolts that have been found on the field.

  Next came the bulk of the army, led by the ensigns bearing the standards. These eagles were the spirits of t
he legions: ‘The king, and the strongest of all birds, which seems to be a signal of dominion for the Romans and an omen that they shall conquer all against whom they march,’ in the words of one Roman contemporary historian.8 The Roman legionaries marched six abreast behind the trumpeters, many wearing the new, segmented armour called lorica segmentata. Familiar from recent Roman motion pictures like Gladiator, this was a technological innovation which allowed the soldiers both more flexibility and protection. It was also noticeably lighter than the ring armour worn up to now; 9kg, compared to 16kg. Their centurions, among whom we know the names of Marcus Caelius and Fabricius, marched behind them to keep an eye on their men.

  The bulk of the army had now passed. An eyewitness who waited around would have seen slaves leading the mules that carried the soldiers’ baggage and then more infantry and cavalry units protecting the rear, ready to give warning against any ambush from behind.

  The subject of how many men Varus had with him on his fate-ful march has been endlessly debated. Velleius Paterculus states that ‘three legions, the same number of divisions of cavalry and six cohorts’ were involved.9 At face value, that would give a figure of around 18,000 legionaries, some 900 cavalry and a further 3,600 allied auxiliaries. That would account for the figure of 22,500 men which is commonly bandied about. It is of course, at this distance of history, impossible to know what the actual numbers were. Troop figures in the ancient world are notoriously unreliable.

  But it is extremely unlikely that all, if any, of the battalions were fighting at full strength. The interim strength report of a cohort which was found at Vindolanda camp on Hadrian’s Wall, dating to the end of the first century, gives food for thought. The document says that the first cohort of Tungrians from northern Gaul was 752 men strong, commanded by 6 centurions. Written in business-like Old Roman cursive script, it paints an illuminating picture of the disparity between the paper strength and the actual strength of a unit. Of those 752 men, 46 were on secondment to the governor of Britain’s guard; 337 men and 2 centurions were at Corbridge, another camp on the Wall; and 1 centurion was in London on business unknown. It is impossible to read where the others were, but in the end, there were only 296 men left in Vindolanda under 1 centurion – of whom 35 were unfit for duty, either ill or wounded. That leaves 35 per cent of the men on active service. This state of affairs was not exceptional and is corroborated by a daily report on a detachment of soldiers found on an ostracon in Bu Njem in Tripolitania in the third century. Of the fifty-seven men stationed at the fort, over half were away on exercises, sick, or seconded to other projects.10

  Certainly, logic dictates that this was the normal state of affairs and none of this seems especially extraordinary for any army before or since. If one considers that some of Varus’ soldiers were out on patrol or policing duty in other parts of the province, even if one adds in the non-combatants who tagged along – women, children and slaves – a total figure for those who set out from Minden of under 14,000 would be well within the realms of possibility.

  Towards the end of the day, Arminius and his auxiliaries, Varus’ vanguard, begged to be excused. As part of the advance team, they needed to get away, they said, to mobilise other tribal auxiliaries in support of the general and to clear the way for the Roman army. Again, while those hostile to Varus have seen his acquiescence as a negligent move on his part, it was procedurally common to allow the advance guard to go ahead to smooth the passage of the army. The general was perfectly well aware of the need for reconnaissance in territory like this and as his intelligence sources had failed to give him any advance warning about the tribal uprising, it was reasonable for him to expect Arminius to bring back more-detailed information about what was actually happening on the ground. The problem was that the person he chose to act as scout, to inform and to warn him was the person plotting to attack and kill him.

  As Arminius and his Cheruscan auxiliaries rode away from the Roman army, the moment came for the German to cross his Rubicon. The die was now cast. He had thrown off civilisation and his commanders. For the historian Edward Creasy, Arminius’ soul was burning at the thought that his tribe would be forced to yield to these ‘debased Italians’. That is probably a fairer reflection of anti-Catholic tendencies in nineteenth-century England than first-century Germany. But to what extent was Arminius pondering his destiny? Did he take the fate of the Gaulish revolutionary Vercingetorix into consideration, who surrendered to Julius Caesar and was then ignominiously paraded through Rome like an animal? Or of the numerous other Germans who had set themselves up against Rome?

  Arminius rendezvoused with his fellow conspirators who were waiting nearby. Although, by the end of the revolt, many, indeed most, of the tribes of Germany were involved in the uprising, at this point we can be confident of only three: the Cherusci, the Bructeri, mentioned above, and the Marsi, a much smaller tribe which lived south of the Lippe and east of the Rhine.

  With Arminius heading off in the late afternoon, it was time to build the camp, the first one seen by Germanicus’ men. ‘Varus’ first camp with its wide circumference and the measurements of its central space clearly indicated the handiwork of three legions,’ writes Tacitus.11 Still, as centurions posted sentries, as the cavalrymen saw to their horses and as the legionaries erected their tents and ate their rations, there is no indication that they were under attack or expected to be. It was the routine of a large army on campaign; it was the calm before the storm.

  The attack came the following afternoon. Varus could not know how widespread or perfectly choreographed the revolt was already. The detachments that had been sent to the various communities had already been massacred. According to Roman accounts, the terrain was difficult; more mountainous than Varus and his legions were used to and much more forested. Forward detachments had been engaged in clearing a path for the army, chopping down trees and building bridges. The weather had worsened as they headed towards the Wiehengebirge mountain chain. The lashing rain and wind was quickly becoming a storm. The march slowed to a walk, then almost to a standstill, the storm front off the North Sea extinguishing both light and hope. The ground was slippery, then became muddy, causing wagons to stick. Forward movement became even more difficult as the tops of trees snapped off in the wind, blocking the path.

  From The Tempest to Wuthering Heights, adverse landscape and weather have often been used as a literary device. Certainly it is not uncommon in the classical canon, especially in connection with a disastrous campaign at the edges of the known world. But while the difficulty of the terrain may well be exaggerated, as anyone who has visited northern Germany in the autumn knows, the account of the storms is very plausible.

  The Roman force suffered its first defeat east of Kalkriese. To attack the Romans in their camp would have been folly and Arminius had waited for the legions to march out and allowed nature to begin to wear them down. The first assault came when the soldiers, tired, wet and anything but alert were beginning to think of rest and their evening meals. An attack from all sides surrounded the army in the forest. An escape route, the way they had come, had been blocked off. The question of how large Arminius’ army was is complicated, but a guess can be attempted. With no reliable literary account as a starting point, it becomes a question of population density. Excavations on Danish islands suggests that 300 soldiers could be recruited from 1,500 people (some 15 villages of 10 homesteads, each with 10 inhabitants), the approximate population density for an area between 10 and 20km in diameter. Arminius could easily have massed an army 15,000-strong, drawing on a mere 750 settlements. While these figures are conservative, it is possible to detect a vague homogeneity in the figures in written sources that goes beyond literary topos. Julius Caesar mentions that Ariovistus’ army numbered 12,000 men in the 50s BC. In the 270s AD, the Emperor Probus drummed a force of 16,000 Germans into his army, and in AD 357, the future emperor Julian faced a German army of 35,000 men near Strasbourg under the command of Chondomar, who had melded together six tribes.12<
br />
  The sky, already black with storm clouds, darkened further as the Roman army was showered with German spears. With Roman cavalrymen protected by short-sleeved, hip-length mail armour, the obvious initial target for the Germans was the horses themselves. Arminius knew that in a confined space, cavalry is effectively useless. It is true that, when fighting at close quarters, cavalry have a height advantage, but with wounded animals slipping in the mud, rearing, throwing their riders, galloping uncontrolled towards their own infantry, the Cheruscan had realised what an advantage he could give his men. And any riders who survived would be a liability, their unwieldy swords and shields practically useless as they got in the way of their colleagues.

  The Germans also knew that many mounts that were not either killed as the arrows buried and tore their way into them or dispatched by the Roman legionaries in self-defence would die soon afterwards. To all intents and purposes, this was a variation of germ warfare. At a time before immunisation shots, tetanus was one of the most dangerous threats to horses. The disease can take up to three weeks to incubate under normal circumstances, but just over a week is more common, a period that can be hastened to a day or two in battlefield conditions. Some sense of the scale of the problem this would give to the Romans can be seen by the fact that the aftermath of the American Civil War Battle of Gettysburg in 1863 required the removal of 2,250 tonnes of horseflesh.

  In this time of mechanisation, it is easy to forget how important horses were. The cavalry units were probably mounted on German or Gaulish horses. These had been drafted into the Roman army for years. Although Julius Caesar dismisses them as ‘poor and misshapen’, that conclusion should be taken with a pinch of salt.13 Horse breeding formed a crucial part of Celtic culture and in general their animals were praised for their speed and endurance. Attacking horses was an effective way to immobilise and demoralise an army and cause huge logistical problems.