Illustration of the Battle of Mons Graupius
Long Reads

Rome Comes To Scotland – The Lost Battle Of Mons Graupius

Illustration of the Battle of Mons Graupius

It is surprising where a walk through the countryside can lead you.

A trip I once took as a child during my only ever visit to Scotland led up a large, heather-covered and slightly undulating hillside. 

I specifically remember there being lots of heather because I spent a good amount of the trip diving in and out of it. This was the only way to have avoided imaginary bullets in a game of war with my siblings, cousins and their friends.

The hill, it turns out, was called Bennachie, and it is a distinctive part of the Aberdeenshire landscape. 

Part of its distinctiveness is created by the sharp contrast between it and the relatively flat surrounding countryside. Mostly though, it sticks out because it has a number of peaks - some sources claiming four and others five. These are Craigshannoch, Bruntwood Tap, Watch Craig, Oxen Craig (the tallest, at 528 metres) and Mither Tap, the most prominent.

Mither Tap one of Bennachie's summits
Mither Tap, one of the peaks of Bennachie, seen from the south (image: Iain Millar on

Although I was not aware of it while fighting my imaginary war there, some historians believe Bennachie was the site of an enormous real-life battle almost 2,000 years ago.

The battle was so large that if I had somehow managed to visit during the late summer of 83 or 84 AD, I might have peered down at the base of the hill and seen up to 60,000 warriors assembling below. 

One side was made up of Celtic Caledonians and their allies, which is to say Scottish tribal forces. The other was a Roman army marching up to the hill to engage them.

To be fair, there is not complete agreement that Bennachie definitely was the location of Mon Graupius, the name the Roman historian Tacitus gave the battle.

Though in ‘Mons Graupius AD 83’, Duncan B Campbell explains that there are good reasons to suppose Professor Kenneth St Joseph was correct when he first proposed Bennachie was the site in 1978.

Although archaeological evidence is missing, it has never been properly sought there, and is also absent in significant amounts from other possible sites, Campbell says. 

What Bennachie has instead is the remains of a nearby camp from which the Roman forces may have been deployed.

Campbell also credits St Joseph with having described how the geography of the hill fits Tacitus’ description, and would have turned it into a kind of amphitheatre if indeed the battle did occur there. This of course suggests a landscape utterly transformed by two vast hordes of men, their pre-battle noises echoing around the hill and the surrounding countryside. 

There would have been tightly packed Roman legionaries and their auxiliary European allies visible, though unlike their opponents, they probably would have been relatively quiet and disciplined.

On the other hand, the Caledonians and their allies from other tribes would have been heard taunting their opponents as they flooded down from the heights and spread themselves out. One images there might also have been the thumping of metal weapons on shields as these men pumped themselves up for the fight.

Besides the man-made sounds, the clump of horses hooves would also have been audible. Roman cavalry had to get into position on the flanks of their comrades in the infantry, and the Caledonians used horses too - as their own cavalry and to haul their Celtic chariots. These vehicles might have been the most noticeable aspect of the battle about to unfold, possibly visible wheeling back and forth in the space between the two armies. They were, by this point, anachronistic in the rest of Britain, but they may have still fascinated the Romans opposite.

There also would have been Roman battle standards visible – eagle symbols held above each legion – and long horns protruding from the Caledonian side. This Celtic instrument was called a cernyx and it was shaped into an animal head at one end. This had the effect of making the listener think of its low hum as emanating from the mouth of the animal (see the video just below.)

The Caledonians and the other tribesman on their side were possibly more numerous. The Roman historian Tacitus’, whose account of the battle is the only source from around the time, put their numbers at more than 30,000. 

He describes them as having been large limbed and red haired, though Campbell points out he may have simply been overusing a stereotypical image of indigenous people.  

Tacitus also described them as having been naked, but here again a qualification is needed, since ‘naked’ very likely meant ‘without armour’. They may well, though, have been bare chested, with their scary-looking blue woad tattoos on full display.

The Romans, meanwhile, would have looked comparatively uniform and better organised. The main building block of the Roman army was the legion, within which there were around 5,000 men - or, to be precise, 5,280 men by this point in Roman History, as John Sadler explains in ‘Scottish Battles’.

Legions at Mons Graupius were likely understrength, however. It occurred at the end of a long campaigning season and multi-year effort to push into Northern Britain, and the Romans also had other concerns elsewhere in the Empire (more on this below.) 

Wikipedia estimates Roman numbers as being somewhere between 17,000 and 30,000 strong in total, and Campbell assumes a figure of just over 20,000.

Whatever the true amount, a significant portion of their numbers, and all of their front-line infantry, were auxiliaries. These were additional troops recruited from parts of the empire on the European mainland, in this case Germany and the Ardennes region. 

Campbell extrapolates from Tacitus’ account that some of these troops, the Batavians, were probably used to spearhead an assault on Anglesey (‘Mona’), off the Welsh coast, earlier in the campaign. Having originated from an island in the Rhine, they would have had practice crossing water with their weapons, making them ideal amphibious troops. 

It would be incorrect to think of these additional troops as legionnaires, however, since they were organised instead into individual cohorts.

Effectively, a legion could be thought of as an ancient brigade or, better yet, a small division, since it was set up to conduct administration as well as warfare. In other words, it was meant to be self-sufficient – a war making, then governing body controlled by a legate. 

Legions were divided internally into several cohorts, and it was this level of organisation that the auxiliary troops were placed in.

Apart from these separate cohorts for auxiliary troops, each legion contained ten cohorts of its own, the first of which was over-sized and the remaining nine being a standard size of 480 men each. 

We can perhaps think of these as small battalions, since battalions have varied historically, ranging in size from roughly 1,000 men at the larger end of the scale to 500 at the smaller end.  

Similarly, it is perhaps useful to think of foreign auxiliary units as being like independently-recruited battalions or regiments from particular areas taken by and friendly to the Romans. A good analogy might be the way Gurkhas have been recruited and fed into their own battalions, regiments and brigades within the British Army.

Cohorts were also subdivided into six centuries of 80 men, or three maniples of 160. Whereas cohorts were led by tribunes, who could be thought of along with legates as being officers, centuries were led by centurions, the Roman equivalent of senior NCOs. Centuries originally had 100 men, but were reduced in size in the first century BC. 

They in turn had 10 decurians serving under them who led subunits of eight men called contubernia (contubernium was the singular form, and centuria the Latinised singular of century). 

The contubernium was therefore like a modern infantry section, which in the British Army during the Second World War had 10 men each, often with only seven in combat and three in reserve. Centurions and decurions were thus like sergeants and corporals today. 

Bennachie shown on a Google map of Scotland
Bennachie shown under the arrow within Scotland (image: Google)

This all begs the question of just what had prompted the well-oiled Roman military machine to march this far north in the first place. 

The classicist Mary Beard has likened Britain to Rome’s Afghanistan, with Hadrian’s Wall becoming the northern frontier of their empire. Hadrian’s Wall, though, was not constructed until 122 AD, and at a location far to the south of Bennachie.

The truth is that Britain as a whole had always held a certain appeal for the Romans. Simon Schama opens his landmark series ‘A History of Britain’ with the following words:

“From its earliest days, Britain was an object of desire.”

In the accompanying book to the series, Schama spells out just why: Britain was, quite literally, a gold mine. It also had plentiful silver and pearls, although apparently, “they were grey like the overcast rain-heavy skies and … the natives only bothered to collect them when they were cast upon the shore”.

Map of Antonine and Hadrian's walls
Map showing Hadrian’s Wall and the Antonine Wall further to the north (image: Norman Einstein)

The first Roman foray into Britain had come in 55 BC, when Julius Caesar crossed over the channel from Gaul*.

(*This was an area that contained modern France, Belgium, Luxembourg, portions of northern Italy, Holland and Germany, and much of Switzerland). 

Though it was not until 43 AD, under the Emperor Claudius, that a permanent Roman presence was established, with a massive, 40,000-man army. 

This was a force that Schama says was beyond conception in the British Iron Age, the historical period that ran from 800 BC and was gradually phased out as Roman influence spread.

The fact that Claudius’ show of force was successful is evidenced by 11 local kings immediately pledging allegiance to him. Campbell indicates that one of them, Cogidubnus, seems to have effectively got a pat on the head in turn. He was declared “a friend of Rome” and given the inscription “rex magnus Britanniae” – ‘great king of Britain’. 

Hadrian’s Wall photographed at Greenhead Lough (image: Velella)
Hadrian’s Wall photographed at Greenhead Lough (image: Velella)

That, of course, is great king of Britain and not king of Great Britain, for tribalism had and would continue to be the norm. This is why Roman pacification of the whole island was both made easier (because indigenous peoples never united as one in resistance), and never happened completely (i.e. because lack of unity meant there was always some stubborn resistance to Rome somewhere.) 

Naturally, those in what is modern-day Scotland held out the longest. Though by the late summer of 83 or 84 AD, the Roman governor in Britain, Agricola, had simply continued the process of pacification up into the very north of Britain.  

Tacitus’ account of the battle was named ‘The Agricola’ after the governor, his father-in-law, and it was a hybrid of history and biography. 

The battle takes up a disproportionately large amount of the text precisely because the Roman victory is portrayed as an example of Agricola’s greatness. He would die a decade after the event, and Tacitus’ account was written five years on from that. 

Counter-intuitively, the Roman historian was not a fan of the Empire, at least not the way it was being run at that point. 

While he believed in the process of continued Roman expansion, he disapproved of the decadence of the era and disliked the emperor, Domitian. In the introduction to his translation of the work, Harold Mattingly explains that a complex back story lay behind this.

Tacitus and Agricola statues
A statue of Tacitus in Vienna (left, image: Pe-Jo), and, right, one of Agricola in Bath (image: Ostrich)

Tacitus’ friend Pliny had apparently known several high-level Romans who had been put to death and banished by Domitian. Their crimes appear to have been defending two senators who opposed the emperors Nero and Vespasian, and who were themselves put to death for it.

Rome had been a republic up until civil war broke out following Caesar’s crossing of the Rubicon in 49 BC. When the dust had finally settled from this, Caesar’s nephew Octavian took power as the Emperor Augustus, making Rome a full empire (as in, with an emperor and not just a republic that had acquired ever more territory.)

The resistance to the emperors Nero and Vespasian by the aforementioned senators was presumably a case of old republican habits dying hard - literally, it would seem.

Tacitus was most troubled by the climate of fear under Domitian and the stifling of independent thought that it created. Equally, though, he felt it was pointless to rebel overtly and risk one’s own execution or exile. 

This belief appears to have informed the way he wrote The Agricola, portraying its main protagonist, his father-in-law, as a model for these difficult times: someone who embodied honourable qualities like supreme ability as well as modesty, and who in his own way fought the decadence and corruption by doing the best he could under the circumstances.  

Tacitus also tells us Agricola was grieving the recent death of his infant son, and that he channelled his emotions into war as a means of dealing with the loss.

The general disapproval of what the Roman Empire had become also helps explain why Tacitus portrayed Agricola’s opponent in a positive light. This was the leader of the Caledonian tribe, Calgacus. 

Like Tacitis’ description of the enemy as red haired and large limbed ‘barbarians’, The Agricola has very little on the tribes of the north. The Romans referred to the entire area as Caledonia, even though the Caledonians were not the only group of people there. 

Map of tribes in Britain during Roman occupation
Northern portion of a map showing the tribes within Britain during the Roman occupation (image from ‘Mons Graupius AD 83’ by Duncan B Campbell © Osprey Publishing, part of Bloomsbury Publishing)

It is also possible that their leader Calgacus was an invention of Tacitus, a way to embody broad resistance to the Roman incursion in one man.

In any case, it is the stirring pre-battle speech that Tacitus attributes to Calgacus that is of most interest. Like the inspiring speech delivered by Mel Gibson’s William Wallace in the movie ‘Braveheart’, or that of Shakespeare’s Henry V that was its inspiration, Calgacus rouses his men by reminding them of just what they have to lose. Referring to the Caledonians having up until this point resisted incorporation into the Roman Empire, Tacitus has Calgacus say:

“You have mustered to a man, and to a man you are free.”

Bennachie viewed from the north
Bennachie from the north ground close up
Two views of Bennachie from the north, the side on which the battle took place (images: Google)

Noting the presence of the nearby Roman fleet, he speaks of only waves and rocks and Romans beyond them, and how, having exhausted the supply of land, the Romans now take the sea around Britain as well. (The Romans would, as part of this expedition, encircle the whole of Great Britain, including the Orkney islands).

Then Calgacus eloquently sums up the greed and brutality of Rome in this oft-quoted section of the speech: 

“Neither East nor West has served to glut their maw. Only they, of all on earth, long for the poor with as keen a desire as they do the rich. Robbery, butchery, rapine, these the liars call ‘empire’: they create desolation and call it peace.”

The Caledonians had previously avoided a direct fight with the Romans, though Agricola had managed to force one by advancing along the flat ground around the east coast of Scotland. He created a corridor for his army by flinging up camps and forts along the way to block attacks coming down from the Highlands, and by keeping a fleet at sea alongside him so that his forces were fully supplied and supported.

As they crept further north, the Romans began to approach the Caledonian granary, the fertile land alongside the Moray Firth in the far north where they grew their crops. This forced the Caledonians to rally with other tribes at Bennachie to pre-empt this move and put a stop to the ongoing advance. It was almost certainly their last chance, though it was also not their first attempt. 

Map of Roman Scotland in AD 82 - 83
A map showing Roman progress up the Scottish coast, with the Moray Forth and Bennachie labelled near the top (image from ‘Mons Graupius AD 83’ by Duncan B Campbell © Osprey Publishing, part of Bloomsbury Publishing)

In what was probably the year 82 AD, Caledonians had ambushed the IX Legion (sometimes written as the VIIII Legion) in their camp during the night. But the legionaries had both fought back and received reinforcements urgently dispatched by Agricola, and the Caledonians had been driven off.

Now, on a late summer day the following year, the whole of Agricola’s force faced all those under Calgacus.

Tacitus speaks of the ‘barbarians’ cheering with wild enthusiasm at their leader’s speech. Once more like Braveheart and Henry V, he has Calgacus finish by reminding them to go into battle thinking of those who had gone before and of those in the future.

Agricola, meanwhile, urges his men to seek honour in battle rather than shame in the safety, or perceived safety, of retreat. Reminding his troops how this is the culmination of seven years’ worth of campaigning to finish conquering all of Britain, he speaks of them having threaded through various woods and crossed estuaries in their march north. And he reminds them that if they are to lose and run for safety, they would be pursued by an enemy who knows the terrain far better than they do. 

Then he says:

“If we must perish, it would be no mean glory to fall where land and nature end.”

This fits perfectly with the subtitle of Campbell’s book on the battle: ‘Rome’s battle at the edge of the world’.

Agricola then goes on to remind his troops that the enemy are at heart cowardly, as evidenced by their night attack on the IX Legion that turned into a rout the moment they were resisted. 

He ends by urging his men to end fifty years’ worth of Roman expansion** in Britain with a glorious victory, and that it should be remembered that rebellions and wars there have not been the fault of Roman soldiers.

(**Historians actually put Claudius’ arrival in Britain as having happened in 43 AD, so 40 years is a closer approximation than 50, even if Caesar’s earlier excursions to Britain are included).

Campbell assumes that the Roman force now assembled consisted of the 5,000 cavalry on the flanks and in reserve, the 8,000 auxiliary infantry in the centre, and a comparable number of Roman legionaries in reserve. Thus, 21,000 in total. 

He also deduces that the easiest way for Agricola to have addressed his troops was not as they were spread out on the battlefield, like a horse-borne Mel Gibson in Braveheart. Rather, he assumes that the Roman governor probably spoke to his troops as they came out of their marching camp at Durno, which is northeast of Bennachie, and before they crossed the river Urie to go and fight beneath the hill’s north face. 

Durno near Bennachie Logie Elphinstone Community Action
The site of the Roman camp at Durno, which now has a rapeseed field in the middle of it, lies to the northeast of Bennachie, which can be seen in the distance (image: Dave Pert)

There is a section in The Agricola that attributes the placing of the auxiliary troops in front as a means by which the governor intended to prevent the shedding of Roman blood. Campbell argues Tacitus’ interpretation is likely incorrect in this case, since auxiliary troops were highly valued members of Rome’s force.

He suggests there is another reason the auxiliary troops might have been placed in front, and it has to do with open versus closed-order combat.

These two styles of fighting essentially existed on a scale, with the Caledonians on one end and the legionaries on the other. They were almost certainly on display at the battle played out.

It started with both sides flinging missiles at each other, though the Caledonians skilfully parried the Roman javelins, or pilums and lancea, using their broad swords or small shields.

Illustration of Celtic light infantry types, 1st century BC or 1st century AD
Celtic warriors from the period – the man in front has a small shield like those in Campbell’s book, though his sword is thinner than those depicted therein (image from ‘Rome’s Enemies (2)’ by Peter Wilcox © Osprey Publishing, part of Bloomsbury Publishin

As John Sadler explains though, the Roman javelin tips were made to be soft enough to bend, so that the shaft would fold over and make the enemy shield it became embedded in useless (i.e. because it ended up with a bent, unwieldy pole on the front catching on things and generally getting in the way.)

Having said that, even those Caledonians who retained their shields were at a disadvantage when it came to closing with Roman legionaries. 

The textbook manoeuvre for a Roman legion was a close-order assault. While Caledonian broad swords and small shields were well suited to open order combat, they became less advantageous when the Romans advanced on them in close order. 

Protected behind their tightly-packed formations, the Romans could advance into the enemy behind their walls of large, curved shields. Furthermore, their short sword, the gladius, was thin and sharp-tipped, the oppose of the broader, more rounded Caledonian sword.

Roman tortoise by Dun.can
The testudo, or 'tortoise', an example of a close-order Roman formation (image: Dun.can)

This meant the Romans could stab their short and sharp weapons through the openings in their shield formations when the opportunity presented itself.

When it did not, smashing the enemy with the metal boss in the middle of their shields was also effective. It must also have provoked their enemies into taking a swing with their swords, an attack that would land on the Roman shields and leave them open for being stabbed. Penetration of only a few centimetres was probably all that was required to kill or incapacitate the Romans’ enemies before they moved forwards to repeat the process.

I once ran across some Roman legionary re-enactors in the US. One of them explained to me that they had worked out very quickly just why it was so advantageous for the Romans to have relatively short swords, and why they carried them on their right hips. Packed tightly behind a wall of shields, it would have been difficult or impossible to then reach across the body and unsheathe a sword that was on the left hip. Taking an overhand grip on the relatively short gladius on the right hip, however, and then unsheathing it while the shield was kept in place with the left hand was perfectly doable. 

Sadler also explains how this all must have worked like clockwork - again, almost literally. 

By this point, a Roman soldier’s military career was usually 25 years long, running from the ages of 18 to 43, with the first five years spent training for 10 hours a day. 

In combat, the same adherence to routine ensured a ready supply of fresh troops as units were broken down into ten waves and periodically rotated to the front. The Romans had some ability to keep track of time through the use of water clocks, something that is explained in more detail in an article entitled ‘Timekeeping In The Roman Army’ on In this way, Roman legionaries were often able to triumph over less well-organised enemies.

Illustration of Agricola's Army is drawn up at Mons Graupius
View from the position of the Roman auxiliary foot soldiers and the cavalry next to them (image from ‘Mons Graupius AD 83’ by Duncan B Campbell © Osprey Publishing, part of Bloomsbury Publishing)

Though this textbook scenario is not really what happened at Mons Graupius. 

The fighting style of auxiliary troops lay somewhere between the two extremes of the close-order legionaries and the looser formations of the Caledonians. In other words, they were well-drilled and organised, but they were accustomed to fighting in a more open-order formation.

And it is this, Campbell suggests, that may have been the main reason they were placed in front. Because once the Caledonian chieftains began zipping back and forth in their chariots, breaking formation was almost certainly how the Romans defeated them.

Chariot wheels had scythes that protruded from them, and these could be absolutely lethal if run along lines of enemy troops.

Since the time of Alexander the Great, however, those facing chariots had begun to learn that the answer in this case was to break formation and let them sail though. They could then be surrounded and dispatched in the rear, while the line reformed in the front. Similar, in a way, to how closed infantry ranks with spears or muskets with bayonets were best for dealing with cavalry charges, and more dispersed formations for artillery attacks during, say, the Napoleonic era.

The auxiliaries carried different shields to the legionaries, which were oval shaped rather than the iconic curled, rectangular shape. They were still a lot larger than the Caledonian shields though, and one illustration in Campbell’s book shows how they could and probably did form a kind of shield wall that blocked the Caledonian charge.

The illustration also depicts the Roman marching camp at Durno in the distance, northeast of Bennachie, as noted in the picture caption above. It was from here that Campbell says the Romans first deployed. Agricola stayed behind the battle, ready to direct formations wherever they might be needed.

He had already thought to place some of his cavalry on the flanks, but as the vast melee erupted between the auxiliary foot soldiers and the Caledonians at the base of the hill, enemy reinforcements were dispatched. 

Caledonian cavalry had by this point already been driven off, according to Campbell. The mass numbers of Caledonian foot soldiers still on the heights, however, descended and attempted to outflank the Romans. 

Battle of Mons Graupius 3D map
The Caledonians in blue, shortly before they tried to use their superior numbers of infantry to outflank the Romans (in red) at the base of the hill (image from ‘Mons Graupius AD 83’ by Duncan B Campbell © Osprey Publishing, part of Bloomsbury Publishing)

They were first blocked by reserve cavalry dispatched by Agricola, then they were themselves surrounded and attacked from the rear. Another illustration in Campbell’s book depicts a scene of utter chaos for the Caledonians as they stumbled through marsh and over fields, pursued and cut down by the Roman horsemen. 

The auxiliary Roman infantry, meanwhile, had continued to march up Bennachie, pressing into the ranks of the Caledonian warriors, cutting them down as they went. By keeping this murderous routine up, the battle soon tipped in Rome’s favour.

Campbell quotes Tacitus describing the scene afterwards this way:

“Everywhere, there were weapons, bodies, mangled limbs, and blood-soaked earth.”

Many more were pursued by the Romans, some being captured and killed, others escaping into woodland. 

Tacitus put the number of Roman deaths during the battle as having been 360, against 10,000 for the Caledonians. Given that his is the only source, and that The Agricola was written as a glowing eulogy of its protagonist, it is no surprise historians question the veracity of these figures.

Still, assuming there even was a battle of Mons Graupius, it seems likely it was an overwhelming Roman victory. A similarly crushing defeat had been inflicted during the Boudica revolt just over 20 years beforehand, after all. 

And yet, the Romans did not stay in Caledonia. 

While their fleet continued on its path up and around the tip of Britain, Roman troops were withdrawn again, and Agricola recalled to Rome. 

Mattingly points out that historians see this as being perfectly justified on Domitian’s part, and not necessarily motived by jealousy of Agricola. Problems elsewhere in the empire are a more likely explanation for a pull back from Caledonia, reduction in the number of troops in Britain and the recall of its governor. Again, these issues had already caused Roman numbers in Britain to become somewhat thinned out, and were probably a large part of the reason legions in Caledonia were not at full strength. 

The Romans would make an attempt to advance their sphere of control northward again, with the construction of the Antonine Wall. At first used in conjunction with Hadrian’s Wall, it was later abandoned in favour of the latter as the only barrier between Rome and the Caledonians.

As for the Battle of Mon Graupius, once again, no one can say for certain that it was really the ‘Battle of Bennachie’, given the lack of archaeological evidence. 

And yet, Campbell points out that the location and distinctive nature of the hill would have made it an obvious landmark around which the Caledonians and other tribes could have assembled. 

So perhaps one day, some other country walkers about to trek up Bennachie will stumble on some of this missing archaeological evidence. Then we might find out for certain if it really was the place where Agricola’s and Calgacus’ armies met all those years ago.

For more on the Battle of Mons Graupius, read John Sadler’s ‘Scottish Battles: From Mons Graupius to Culloden’, Ken Mattingly’s translation of ‘The Agricola’, ‘Mons Graupius AD 83: Rome’s Battle at the Edge of the World’ by Duncan B Campbell and ‘Rome’s Enemies (2): Gallic and British Celts’ by Peter Wilcox. Visit Osprey Publishing for more illustrated military history.

Picture of Mither Tap from Iain Millar on, and map of Antonine and Hadrian’s Wall by Norman Einstein. Map of Roman Empire just below by Steerpike and Andrei nacu.

Map of the Roman_Empire in around 69 AD
Map of the Roman Empire as it was in 68 – 69 AD, eight years before Agricola become governor in Britannia and expanded the Empire into Wales and then continued into Scotland

Join Our Newsletter


How is Estonia dealing with heightened Russian threat to its security?

RAF airman John Nichol's life-saving decision to eject from burning Tornado jet

Tough three-day course BEFORE starting Royal Marine Commando training