EPISODE 3

The Soldiers

THE RECITAL SEEMED INTERMINABLE, ITS prolongation assured by an enraptured audience and Nero’s eagerness to bask in its rapture. Nor was there likelihood of escape from the cycle of applause and encore till the emperor tired; and being scarcely beyond his teens and endlessly hungry for adulation, that was unlikely to be soon. Of all places which appreciated artistry, recognition in Greece, the land of artists, was doubly sweet. So it was that, as Nero struck the lyre yet again, a pin could have been heard to drop; while he sang they sat spellbound; and when he stopped they exploded, with shouts of ‘Blessed are they that hear thee!’, ‘Apollo, thou art with us!’ and ‘Surely it is Phoebus himself who sings!’ And yet in truth the playing was plain, the voice thin, the theatricality forced and the whole occasion acutely embarrassing.

Irksome as this was for the audience of Greek notables, it was doubly so for the emperor’s Italian entourage, who had no choice but to endure these unendurably boring exhibitions at each stop on the long itinerary. Performances had been known to last from early morning till late evening and some Greeks had hit on the idea of swooning with ecstasy, so they could be carried out as if dead: the only way of escape. For those in the imperial suite, on the other hand, endurance was perhaps a price worth paying for an otherwise pleasant and leisurely progress around the hospitable cities of Hellas, a country regarded by Romans with an affection similar to the Englishman’s view of Italy in the age of the Grand Tour. Besides, Greece flattered the emperor and moderated his moods, which made things easier as well as safer for his travelling companions. The catastrophe was for the host country, owing to the endless attainders and confiscations which Nero was currently devising to pay for his Corinth Canal project. It was said that the roads were busy with messengers carrying news of condemnations or confirmation of murders performed.¹

Among the Italians present on that fateful evening, quite distinct from the usual run of officials and hangers-on, was a senator, already in his forties, on whom the mantle of pretended pleasure sat uneasily. T. Flavius Vespasianus was not only a soldier among senators, he was even a rough diamond among soldiers, being of bourgeois rather than aristocratic background and rustic rather than metropolitan origin. Not that social handicaps mattered greatly. The awkwardness was in the man, here emphasized by the extreme contrast between caesar and soldier: Nero, last of the lines of Augustus and Livia, rouged, ringed and ringleted; and Vespasian, inelegant, gruff and practical. A portrait bust in Naples² shows him as bald, with vestiges of coarse and crinkly hair, the features clenched and determined, the expression searching, the mouth stingy but redeemed by an ironic smile. Though the eyes are blank marble, the sculptor has contrived to suggest a twinkle. Homely in looks and rough of tongue, short on social graces and long on common sense: such was the man designated by a jest of fortune to be Nero’s successor.

Why should the emperor invite so unlikely a companion on a fine arts tour of Greece? The answer can only have been that Nero liked him. He was also a national hero. When scarcely in his mid-twenties, he had done singularly well against the Britons. It seems Nero distrusted the commander of that campaign, Aulus Plautius, whose murder he would in due course arrange. Perhaps it was to spite Plautius that Nero now favoured his former lieutenant. Whatever the reason, Vespasian’s star was rising. He had come far since the ridicule to which he had been subjected as the youthful official in charge of the Roman street sweepers, when (if the expression may be excused) he had fallen foul of the then emperor: ‘Caligula, spotting a pile of mud in an alleyway, ordered it to be thrown onto Vespasian’s toga; he being at that time the official responsible for street cleaning.’³ Vespasian was obliged to offer up his white toga, with its senator’s broad purple stripe, much as a girl might hold her apron in gathering flowers, while guardsmen trowelled the ordure, doubtless with full measure of donkey droppings, into his lap. Suetonius enjoys the irony: ‘Accordingly the soldiers shovelled the dirt into a fold of his senatorial gown, filling it to capacity. This was later seen as an omen, that Vespasian would one day take the soil of Italy into his care.’⁴

Under Claudius had come a brighter turn. A contact in high places brought command of a legion. It was a job that would fit like a glove. ‘Vespasian was a born soldier: marching at the head of his men,⁵ choosing where they should camp,⁶ harrying the enemy day and night by his leadership and where necessary by personal combat; content with whatever was going in the way of food and dressing much like a private soldier.⁷’

His reward for a brilliantly fought campaign in Britain was the governorship of Africa. Straight as a die, he declined all bribes and in consequence was ruined by the grievous expense of this post. ‘He governed Africa with great justice – and with dignity – save once when the people of Hadrumentum⁸ pelted him with turnips. That he came back no richer than he went out is proven by his having to mortgage the family land and invest in the mule-breeding business, from which he got the nickname “muleteer”.’⁹

After Claudius came Nero and a return to better days. Perhaps the boyish caesar found Vespasian’s directness and jocularity refreshing. At any rate, it was highly unlikely that a ‘muleteer’ would ever grow too big for his boots. So Vespasian’s reward, both for his achievement and his modesty, was the honour of accompanying the emperor on this pleasantest of tours, plus a front seat at a long succession of imperial lyre recitals. We may assume that he was sitting near the front since Nero, in mid song, now spotted something which came close to costing Vespasian his life: Vespasian had fallen asleep!

How could a man of such sense allow himself so dangerous a lapse? We must suppose that disinclination toward the artistic, distaste for the pseudo-artistic, the excruciating tedium of the occasion, combined with the balminess of the evening, allowed his attention to wander further than usual, so that reverie slid into sleep, causing his head to nod under circumstances which, had he been less popular, would certainly have cost him it. We next hear of him dismissed from court, fleeing Greece and hiding in the Italian countryside; self-exiled to a life of total and indefinite obscurity. However, before pursuing this strange story, it may be rewarding to look back at those events in Britain which had brought Vespasian to prominence, indeed to consider that remote island generally.

Augustus’ advice to his successors, not to expand the empire further, had stood for almost thirty years until Claudius repudiated it by the British invasion of AD 43. This was a misguided enterprise, the decision to embark upon it taken against the weight of evidence. Certainly Rome’s northern frontier could hardly have been more clearly defined or better protected than by the Channel. ‘What wall’, Josephus asks, ‘could be stronger than the sea which is Britain’s bulwark?’¹⁰ Equally it was Rome’s. This Augustus recognized; and though he called himself Caesar’s heir, he had rejected the option of invasion:

Though Rome could have taken Britain she declined to do so. In the first place the Britons are no threat, having insufficient strength to cross over and attack us. In the second, there would be little to gain. It seems that we presently get more out of them in duty on their exports than we would by direct taxation, especially if the costs of an occupation army and of tax collecting be discounted. The same goes even more for the islands round about Britain.¹¹

For Claudius the temptation of Britain lay not in strategic or fiscal advantage but in newsworthiness and the stir which would be created by extending the empire across ‘outer ocean’, seen by Romans as a symbolic barrier. But though Britain possessed the potential for a quick propaganda coup, it was less easy to predict events in the long term, to foresee that the problem would not be crossing perilous water or landing on a hostile shore, but deciding how far to go and where to stop. This search for a stopping place would be long and vexatious, with no comfort in the knowledge that the Channel had been best in the first place. It would, alas, be characteristic of policy toward this island that practical arguments were overruled by emotional. Fame awaited the conqueror of Britannia. Conversely, infamy awaited whoever might relinquish her. So, once started, the line of action would be difficult to retract. Indeed, within ten years Nero was already wishing he could ditch the enterprise but found it politically prudent to stay, since ‘not to do so would have belittled the glory won by Claudius’.¹²

Glory was what Claudius most needed. Claudius the family fool; the coward who, during the desperate hours following Caligula’s assassination, had been found cringing behind a curtain in the palace. Claudius the insecure, who had bribed his way to the throne with payments of 3,750 denarii to each Praetorian guardsman, equivalent to sixteen years’ pay for a legionary private. Now he would show them. Subjugation of the ‘Celtic world’ had taken centuries and was still incomplete. Claudius would strike the culminating blow.

Ninety-seven years earlier Caesar had charted a comparable course. He too had recognized Britain’s power to stir the imagination. His visits may be connected with sensation-seeking and rivalry with Pompey, rather than with practical benefit. In the military sense they were reckless in the extreme. Gaul was not yet fully subdued and he had no business leaving her in his rear. But as the lion-tamer crowns his act by turning his back upon the lions, so it was Caesar’s pièce de résistance to leave the Gauls lightly guarded while turning with seeming nonchalance to the Britons.

In 55 BC Caesar had set sail with the bulk of his force in order to test the problems of crossing and the British reaction, anticipating a more ambitious operation in the summer following. He landed in East Kent, traditionally on the shingle at Walmer or, if his estimate of seven miles beyond the end of the cliffs is taken strictly, at Deal.

They embarked around midnight and Caesar, with the leading ships, reached the British Coast at about nine the following morning. He could see the enemy armed and in large numbers on all the hills. Hereabouts are steep cliffs, so close to the sea that missiles may be thrown from cliff top to beach. This seemed no place to land, so they pressed on for about seven miles to where the ships could be beached on a level and open shore.¹³

The sensation created by Caesar’s despatches, as well as the Roman view of the hazards of crossing, may be guessed from the Senate’s vote of a twenty-day thanksgiving on the army’s safe return.

On his second visit Caesar crossed by the same route. His army forded a river, doubtless the Stour, and took a hillfort, probably Bigbury near Canterbury. They then marched west, perhaps along the North Downs, and swung north to wade neck-deep across the Thames in the vicinity of future London, despite sharpened stakes hidden below the water’s surface. From here they penetrated the forest¹⁴ ringing north London, of which Epping is a remnant. After defeating a tribal confederation, said to have included 4,000 war chariots, they stormed and occupied the stronghold of Cassivelaunus, the British leader and paramount chief of south-eastern England,¹⁵ probably at Wheathampstead, north of present-day St Albans. But Caesar was soon obliged to retire to the coast, where his base camp was under attack; finally re-embarking before the autumn gales jeopardized safe passage.

This profitless adventure instigated a deception which could not fail to influence those who followed. It inferred that by seizing a major oppidum in tamest Hertfordshire the keys to Britain had somehow been secured. Geography seems to vindicate this view. Britain smiles toward the Continent, concealing troubles which will be revealed only gradually. In terms of terrain, climate and fertility, barely a third of the island lives up to its promise. What appears from the south to be an extension of France would, if seen from the north, be more like an extension of Norway. Even so, we must not leave the impression that antiquity was entirely ignorant of Britain’s physical and human geography. Evidence existed, some of it from Caesar’s own pen, on which a more realistic assessment might have been based.

In the passage following, Caesar refers to the ‘coastal’ and the ‘interior’ regions. By the former he meant the south-east, whose areas of recent Belgic settlement included Kent, the counties north of London and the Sussex coast. Here the wealthiest tribes lived. By the ‘interior’ he especially meant upland Britain; in modern terms the North, Scotland, Wales and the South-West. He contrasts the two ways of life, considering the interior peoples to be less advanced. Caesar also thought Britain’s axis lay north-west-south-east instead of north-south, so that the west coast faced toward Spain and the east coast in a more northerly direction.

The coastal regions are inhabited by Belgic invaders who came for plunder and stayed. The Belgians are numerous, with many cattle and farms, similar to those in Gaul. Like Gaul, too, there is ample timber, though without beech and fir. The climate is more temperate, with winter’s less severe than Gaul.

The island is triangular. One side, some 500 miles long, faces Gaul. A second faces westward, toward Spain. In this direction lies Ireland, thought to be half Britain’s size. Between Ireland and Britain is the Isle of Man; and it is believed there are also a number of smaller islands where some writers have described a total winter darkness lasting thirty days. This western side, according to the natives, is 700 miles long. The third faces north. There is no land opposite this, though its eastern corner faces Germany. Its length is reckoned as 800 miles.

By far the most civilized of the Britons live in Kent,¹⁶ where life resembles that of Gaul. Most interior tribes do not grow cereals but live on meat and milk and wear skins. All Britons dye themselves with woad, giving a blue colour and a savage appearance in battle. Hair is worn long; the entire body shaven except the moustache. Wives are held in common among groups of ten or twelve men, usually the males of the same family.¹⁷

This island had also fascinated Claudius’ predecessor, Caligula. Suetonius described the farcical events of AD 40, when the legions mutinied rather than embark. In a bizarre ceremony on the beach near Boulogne, Caligula declared Britain annexed, as it were, in absentia. The campaign would be sold to the Roman public as a victory against Ocean.

He deployed the army in line of battle facing the sea, with the artillery in readiness. Everyone was wondering what it all meant when suddenly Caligula shouted ‘Gather sea-shells!’ By this he meant ‘spoils of ocean’, as an offering at the Roman Capitol. So the soldiers had to fill their helmets and kilts with shells. Then he promised a bounty of four gold pieces per man. The triremes used in the Channel on this occasion were transported to Rome, largely overland; and he wrote ahead, ordering a Triumph more spectacular than any yet seen.¹⁸

There can be no doubt that seasoned soldiers quailed at the sight of open sea. It was not that the Channel is rougher than the Mediterranean or Black Seas. The fear arose from a world picture consisting of three continents moated by a dark and savage deep, within which were sea monsters and beyond which was nothing. There were also superstitions regarding Britain herself: for example, the eerie story that she was the abode of the dead and that souls were rowed across in unmanned boats, which left the coast of Gaul at nightfall and returned before dawn. Thanet,¹⁹ the name of Kent’s north-eastern extremity, may originate in this legend.

It is therefore not surprising that the Claudian version of the same project should have started in much the same way. Again there was near mutiny when the soldiers realized they were to go ‘beyond the known world’.²⁰ On this occasion there was present one Narcissus, an imperial civil servant and former slave of the punctilious type Claudius liked to employ as his secretaries. This man, who we may suspect was empowered to offer an inducement, now mounted the general’s rostrum and attempted to calm the lads. Some wag then shouted ‘Io Saturnalia!’ (hurrah for the feast of Saturn) a greeting roughly equivalent to ‘merry Christmas’. For just as it is some armies’ custom that on Christmas Day the officers wait on the men, so for the holiday period beginning on 17 December, Roman masters and slaves traditionally swapped roles. The shout was thus a sarcastic reference to the indignity that an ex-slave should address the army in the emperor’s place. Taken up by others the whole parade was soon convulsed with laughter, easing the tension to such an extent that the soldiers forgot their fear and obeyed the order to embark.

For the campaign which followed, our principal source is Cassius Dio, writing 130 years later. By contrast Suetonius is brief and dismissive: ‘Claudius’ only campaign was of little importance. The Senate had already voted him an honorary Triumph but this he refused, lighting on Britain as the place to seek the real thing. Her conquest was a project which had lain dormant since Caesar’s day.’²¹

Four legions were mustered, one from the Danube and three from the Rhine, which had fallen unexpectedly quiet since the Varian Disaster. Aulus Plautius, ‘a distinguished senator’²² was appointed to command, with the youthful Vespasian leading the legion IIAugusta. In essence the thrust would follow that of Caesar. Dio specifies embarkation in three groups, at least one of which sailed westwards from Boulogne,²³ raising the possibility that a diversionary force beached in Chichester Harbour. Nevertheless, it remains likely that the main force landed unopposed at Richborough, Kent. Plautius then advanced to a river, too wide to be bridged, presumably the lower Medway. Here the crossing was strongly contested. In an action, probably near the future Rochester, Vespasian sprang to prominence, both because of his spirited attack and as the first senior officer across. The Britons fell back on the Thames and Plautius advanced to a position facing one of the fords: at Southwark, Westminster, or Battersea. Here, in temporary camps somewhere under today’s South London, the army settled to a long wait while a prearranged plan was set in motion. Claudius sailed from the Tiber to Marseilles, then probably up the Rhone and overland to the Channel, arriving in Britain with a large entourage of VIPs, guardsmen and elephants. All this involved a delay of six to eight weeks. The weather had been bad and on three occasions the imperial flotilla was almost shipwrecked.

At last Claudius joined the legions waiting by the Thames. Assuming command, he crossed the river, ‘defeated the barbarians’ and took the capital of the late king, Cynobellinus (Shakespeare’s Cymbeline) at Colchester. However, the real leader, Prince Caratacus, a stout fighter and clever tactician, fled to south Wales, where he would live to lead again. It seems unlikely that Claudius met much resistance. Suetonius states bluntly that ‘he fought no battles and took no casualties’.²⁴ Nothing is heard of the elephants. Doubtless their main contribution was in Claudius’ victory parade where their effect would be stunning, lumbering through the streets of a Colchester only days removed from prehistory.

Elsewhere elephants had proved a dubious asset. Their trial against the Celtiberians at Numantia²⁵ had been a débâcle, when the drivers lost control and the trumpeting animals ran amok. It is strange that while we spend hours in zoos or make expensive trips to Africa to look at elephants, the ancients tended to regard them with revulsion. In warfare there was a real risk that they would frighten the enemy less than the side which was using them. Thus Ammian: ‘With them came a striking sight: ponderous lines of soldier-laden elephants, bodies repulsively wrinkled; the most hideous of all forms of horror, as I have always maintained.’²⁶

Having cracked champagne against the hull of the new province, the emperor returned the command to Plautius with instructions ‘to subjugate the remaining districts’, and retired to Rome. In all he had spent sixteen days in Britain. So much for the victory. What of the territory? What to take and what reject, where to terminate the conquest and how to round it off? These were quite other matters and Claudius may have been content to leave them to his generals.

We do not know at what point Vespasian’s legion was detached from the main force and sent on the separate mission which would win him fame as perhaps the greatest gunnery officer the ancient world produced.²⁷ Suetonius mentions that officers decorated in Britain marched in the Triumph, suggesting Vespasian returned to Rome for this purpose and the conquest was not resumed till the following spring. Then, as the ponderous Plautius moved north, Vespasian swung west. For his part in the war one must be content with a single sentence in Suetonius, though a good one: ‘He went to Britain, where he fought thirty battles, subjugated two tribes and took more than twenty oppida, the Isle of Wight besides.’²⁸ In fact Vespasian overran the south and south-west of England at least as far as Exeter, ending perhaps with a sweep to the lower Severn. His offensive resembles that of 1944 when Patton, pivoting on the slower-moving Montgomery, sped across France. Vespasian does not lose from the comparison, for here was a commander who seized three quarters of the ground with a quarter of the army. What is more, he faced, in Wessex,²⁹ one of the most formidable concentrations of forts in the ancient world. His answer to these citadels of soil was artillery.

Roman guns were based on the idea of twisting a rope until tension was created, then releasing it; which is why they are sometimes known as ‘torsion artillery’. Hemp, horse, even women’s hair was used, but animal sinew had the greatest elasticity, though there was a problem in keeping it dry.³⁰ Guns were of two basic models. The scorpio (whose firing arm resembled the upreared tail of the scorpion) was based on a grounded chassis of heavy timber, across which was strung a thick skein of sinew into which the slinging arm was rooted. This was pulled back and the ropes wound by handles until the required tension was created. When released and the arm had reached the near-vertical, its rush was stopped by a padded beam, creating a jerk which projected the missile. The scorpiowas later known by other names, illustrating the inventiveness of army slang: ‘The machine is called tormentum (the rack) as the tension is created by twisting; and scorpio because of the raised sting. More recently it has been called onager (jackass) since, when wild donkeys flee from pursuers they kick stones backwards, so cracking skulls or shattering ribcages.’³¹ A crew of five is given for this machine (four winders and a loader-firer), a missile weighing fifty pounds and a range of 450 yards,³² though 700 yards is also recorded. Unlike modern pieces, designed so that parts will interchange and with calibre matched precisely to ammunition, the scorpio could be made to almost any specification; so it is not surprising that various missile sizes, ranges, mounting and haulage arrangements are claimed.

The ballista had a different appearance. It was in effect a large, stand-mounted crossbow, except that the bow consisted of two halves, each embedded in a sinew coil. This was a highly effective weapon at a hundred yards. Again, larger versions were available. A 4th-century source tells of a ballista able to shoot across the Danube.³³

Both scorpiones and ballistae could be adapted to stone shot or iron bolt. Ammian describes fire-darts, with reeds bound around a hollow, wooden centre into which glowing embers were placed.³⁴ All had devices for sighting, tilting and traversing, so that a target could be pounded once its range had been found. Accuracy called for standard missiles, though random rocks were best for anti-personnel bombardment, since the sound of a jagged object in flight is more terrifying. Various grades of ammunition must therefore have been carried on campaign. Added to this was the weight of the guns themselves. A full-scale scorpio model has been found to weigh over two tons.

It cannot be argued that artillery was a decisive arm in warfare generally. It was cumbersome, weather dependent, of little value in rough country and useless in forest. None the less it was highly effective in sieges. Stone walls could be shaken loose. Defenders could be driven from palisades, opening the way for infantry attack. Though ineffectual against the mighty hillforts, stones and firedarts could be lobbed across their outer mounds and ditches to fall among the thatched huts within. From wooden towers erected outside the defensive ring, observers could guide the shotfall onto selected targets; subjecting the defenders to an ordeal of whirring missiles crashing among them from guns they could not see.

Vespasian was quick to grasp artillery’s strengths and Wessex’s weaknesses and to see that these were complementary. Mighty earthworks were being used to protect flimsy, fire-prone villages, without internal shelters or warproofing of any kind. Furthermore the forts had become overblown. As with nuclear stockpiling, the rivalries which promoted their proliferation had become obsessional and scale had outrun ability to defend.

It is difficult to guess the number of artillery pieces allocated to Vespasian. Vegetius, a late-period writer, describes fifty-five ballistae and ten onagri per legion, drawn by mules or oxen and having crews of seven. As well as ammunition, Vespasian would be moving with equipment of other kinds: prefabricated towers and battering rams, as well as boats and planks for river crossing. Naval squadrons must have been operating in support up rivers like the Test, Wiltshire Avon and Frome, as well as in the assault on the Isle of Wight. With logistics like these we cannot assume lightning war. Indeed Vespasian’s temperament inclined him to the more deliberate school of generalship which, though for a time eclipsed by the showier styles of Caesar and Pompey, was in fact the Roman norm. Nevertheless the siege operations, once begun, were probably concluded with a speed which paralysed the enemy. The sudden surrender of ‘impregnable’ positions can have grave consequences for morale and when they crashed so quickly in the face of this unprecedented weapon it must have seemed, to tribes which awaited their turn, like the knock of doom. The remains of several Roman artillery pieces have been found and reconstructed versions may be seen at the Saalburg Museum, ten miles north-west of Frankfurt, and a half-scale scorpio at the Lunt Fort, Baginton, near Coventry, England.

Before pursuing Vespasian’s campaign, some account should be given of late Iron Age Britain as distinct from ‘Celtica’ generally. Terms like the latter are seldom used lest they imply unity in the Celtic camp. Of this there was little. Nor was there ever a Celtic empire. Separatism is, it seems, the enduring characteristic of a cultural group at loggerheads from that day to this; from whose differences the English would so frequently profit. In Britain’s case, the accepted picture of envelopment in a wider Celtic world must be qualified by a stark (and, to many, an unpalatable) fact. Except in the extreme south-east (and an enclave to the north of the Humber) archaeology has failed to reveal changes of sufficient magnitude to demonstrate migration into the British Isles.³⁵ This puts a rock into the river of prehistoric studies around which emotional currents are certain to swirl; for it implies that most of Britain was not peopled by incomers at all, but by a miscellany of native tribes surviving from the Bronze Age. Hence it would follow that British Celticism is a fraud; indeed that a pan-Celtic ancestry, knitting Europe’s Atlantic fringes into a cultural whole, is a modern idea, invented as a counterweight to the dominance of the English language and Anglo-Saxon institutions. It is of course true that no classical author used the word ‘Celt’ in a British context. Nor does ‘Celtic’ become familiar in this sense till the 18th century. However, a solitary word is not the sole issue. Opponents of a Celtic Britain must answer major questions posed by languages like Gaelic, Irish, Welsh, Cornish and Manx, related both to each other and to what we know of ancient Celtic. For instance, in his monumental The Celtic Place-names of Scotland (Edinburgh and London, 1926) W. J. Watson offers some 50,000 examples from that quarter alone. More broadly there is a common background of names for places and natural features, of which everyone will be aware: from Boulogne to Bologna, Trent to Trento, Severn to Seine, Ouse to Oise, Shannon to Saône, Mersey to Meuse, Don to Danube, Arun to Arno and Irun. From the Pennines of Northumbria to the Appennines of north Umbria, Western Europe is bound by eloquent strains of remembrance. Of what do they tell? Of Celtic invasions or merely of Celtic influences? Might this onomastic luggage have travelled without the passengers? It is a likelihood many will question. On the other hand, indigenous building methods, pottery and, to some extent, artistic styles, are among the evidence which continues to deny invasion. The controversy is complex and unresolved. It may nevertheless be accepted that the description ‘Celtic’ – albeit with modified meaning – retains at least partial validity and will continue to be employed by students of ancient Britain, if only because no untarnished alternative presents itself. Pro-Celtic propaganda (if such it is) has handed a resplendent past to northwestern Europe’s peripheral peoples. They will be reluctant to hand it back.

What is the difference between the terms Celtic and Gallic? Names like Gaul, Galle,³⁶ Galatia, Galicia, Gaelic, Galway, Galloway, Donegal, Portugal and so on, remind us how the Celtic peoples, then as now, described themselves. As we have said, the two words are largely interchangeable,³⁷ except that in Latin, Gaul became associated with what is now France and posterity tends to honour the distinction. All were, however, one loose grouping which by late prehistory had colonized, absorbed or suffused western Europe from the Atlantic to Germany and from the north of Scotland to northern Italy, with an offshoot through the Alps into the Balkans and even an outlying pocket in Asia Minor.³⁸ Nevertheless, varying degrees of Celtic and pre-Celtic mixture, as well as centuries in various terrains and climates, created a wide range of development.

How did Celtic attainment compare with Roman? The traditional criterion of literacy as the difference between historic and prehistoric societies applies in this case. Though the arrival of Mediterranean influences had recently begun to provoke writing, using Greek and Roman alphabets, an illiterate majority may still be assumed and nothing resembling a Celtic literature had yet appeared. More recently the test of comparative technology has found favour; and in this sense achievement was close to Roman. Even so, common sense requires other insights; for Rome, after all, prevailed. At least one prehistorian suggests the real differences lay in social structure, civil order and organization: ‘between stability and the complex conduct of affairs of state, on the one hand, and the impermanence and emotion-charged atmosphere of the clan or tribe on the other’.³⁹ This seems closer to the truth, owing partly to developmental level but also to the ‘Celtic temperament’. In references to the lost account of Posidonius, in other classical authors, in the surviving Irish epics, even in echoes from the 18th-century Scottish Highlands, one has an impression of touchy pride, feud, argumentativeness and bombast, plus a life dominated by hunting, feasting and war. The sobriquet of the Irish hero, Conn of the Hundred Battles,⁴⁰ suggests this mood. Many such battles were doubtless mere cattle rustling. Others were over land and water claims, booty, revenge, or to repay some slight. Strabo went so far as to assert that ‘the whole Celtic world is war-mad’. This emphasis on warriorship would have tragic consequences in the prolonged clash with Rome, when honour would oblige the Gallic people to stand and fight where harrying tactics, on German lines, would often have served better. Celtic thinking on matters like peace and war, law and order, taxation and absorption into an alien regime was incompatible with Roman. Most of all, an aristocracy based on privilege and military prowess felt compelled to answer a challenge to either. These entrenched differences meant that incorporation into the empire would be painfully accomplished.

The pain was not entirely one-sided. We have spoken of the Romans as boreaphobic or fearful of the north, especially in connection with the Cimbric and Teutonic migrations of 100 BC. A far deeper scar had been left by the sack, almost two centuries earlier, of Rome herself. This followed an overspill from the Celtic movement into the Balkans. Seventy thousand rampaging Gauls looted and burned all Rome except the Capitol; but after a seven-month siege this too was taken. The Romans were obliged to buy off the raiders with a humiliating ransom. A result of this shock was the building of the Servian wall, so-called because its 19th-century discoverers assumed it to be that of King Servius Tullus.⁴¹ This was in huge blocks of dressed tufa, to a width of ten-and-a-half feet and a circuit of seven miles, the impressive relics of which may be seen outside Rome’s terminal railway station. A second result was the Celtic settlement of northern Italy which, despite Roman annexation and some colonization, was still in essence Gaulish. So Virgil, from Mantua, was perhaps an ethnic Gaul. Romans would not feel secure until the land beyond the Po – and ultimately most of Celtica – was under total control.

This would take time. Meanwhile, as with other eager consumers on Rome’s rim, Greek and Roman merchants were hustling the Gauls toward that ‘prestige goods dependence’, in which chieftainship is synonymous with showmanship. It was the same pattern as the later trinkets-and-gunpowder trade on the Gold and Ivory Coasts: luxuries, slaves to pay for them and war to get the slaves. We know the huge size of wine shipments but can only guess at the ills which this prolonged inflow of luxuries and outflow of slaves was inflicting upon the Celtic world.

What of the Britons? Et penitus toto divisos orbe Britannos⁴² (and the Britons, wholly sundered from the world) as Virgil put it. This was an isolation more romantic than real. Despite the mystery and metaphor, the less-than-poetic activities of the Roman merchant had already gained a firm hold in southern Britain. The main port appears to have been Hengistbury,⁴³ but commercial control soon passed to the south-east, where the adoption of the wheat symbol on pre-Roman coins⁴⁴ suggests a grain-based export drive, probably to supply Rome’s Rhine armies.

On the question of food production: opinion continues to revise this upwards. Strabo said of Gaul that ‘nowhere is untilled except where swamp and forest prevail’.⁴⁵ It is not of course known how much swamp and forest there was; but relative to today perhaps more swamp than forest. Mastery of iron had increased the tempo of man’s war on woods, with a far tamer terrain than that of prehistoric Germany as the probable result. Agriculture was in a state of revolution. In Britain productive wheat strains such as emmet allowed yields more abundant than was previously imagined.⁴⁶ Excavation of Roman forts in bleak, northern locations has in some instances revealed ploughmarks beneath their earliest levels. Though the north and west were still largely pastoral, these final decades of prehistory were a time of transition between grazing and planting. In the Celtic world generally this tendency was reducing the instability feared by Rome, though there was still the occasional migration. Another result of sedentary agriculture was abandonment of the native forts and the realization that wealth could better be pursued through activity in valleys than passivity on hilltops.

Europe’s face is as pocked by defensive works as the moon’s by craters. Most impressive of all are the earthen citadels (frequently though not invariably sited on dominant summits); Celtica’s quintessential landmark. Collectively they represent one of the biggest constructional feats in the human story, though their earlier origins and their presence in other parts of the world denies that they were Celtic only. Even within the Celtic lands there were major differences of style and scale. Near the Mediterranean, mimicking the town walls of Greek colonies, the forts had developed as dry-stone citadels, their walls ambitiously provided with towers, usually square; defending paved streets and stone houses. Prominent examples include Entremont, near Aix-en-Provence, where the bastioned stone walls enclose nine acres; Puig de Sant Andreu (near Empuries,⁴⁷ at the northern end of the Costa Brava); also in Catalonia, Castellet de Banyoles (Tivissa, twenty-eight miles west of Tarragona), with mudbrick towers on a stone base; and Citania (Portugal).

In Gaul proper another technique, which Caesar called murus Gallicus⁴⁸ (Gaulish walling) had emerged in reply to the approach of Rome. The defenders grasped that perpendicular walls make stiffer obstacles than sloping mounds. They experimented with dry-stonework but soon found that it was brittle to the battering ram and breachable under bombardment. Accordingly they devised double masonry facings, up to forty feet apart, the gap bridged by joists and cross-braced. This framework of nailed timber packed with rubble and backed by an earthern ramp, was resistant to shot and ram; but not to the slumping and sliding which follow woodrot. Though great metalworkers, good carpenters and moderate masons, the Celts never mastered the simple process of firing chalk or limestone to produce quicklime. Its full exploitation, though not its invention, was the work of Rome and crucial to almost all architectural advance during this period. Without lime there could be no mortar or concrete; and without mortar to grip or concrete to back its stone facings, the murus Gallicus remains a cul-de-sac in defensive building. None the less impressive results were produced. Excavation of the Bibracte oppidum (near Autun) and that of Manching (Bavaria) revealed three and four-mile circuits of Gaulish walling respectively.

However, these advances were never to reach the Britons, who were in all respects more conservative. In the south and west of England traditional construction continued, with billions of basketfuls of soil, hand-dumped in ant-like operations. Size and imposing position are characteristic, but the forts are also numerous. Herefordshire and Shropshire are their heartland, the seventy miles from Ross-on-Wye north to Oswestry containing some fifty major specimens, sometimes only a mile or two apart. Wessex too is rich in spectacular examples, including Maiden Castle (Dorset) of forty-five acres, capital of the Durotriges, a tribe opposed to Rome. Though less frequent, there are big hillforts in most parts of Britain. An impressive tribal capital is Eildon (near Melrose, Roxburghshire) of forty acres⁴⁹ and containing 300 roundhouses. North Wales, where rock is often more available than soil, developed its own style. Tre’r Ceiri, in a rugged and inaccessible situation near the north coast of the Lleyn Peninsula, is among Britain’s most romantic, with a stone wall surviving to rampart walk and internal buildings visible. There are some 200 British hillforts of more than fifteen acres. They grew considerably during the last four centuries BC, doubling and trebling their rampart rings and developing elaborate gateways with interned entrances and flanking guard chambers. Within were villages of 300–500 souls; but occasionally much larger, up to the 5,000 size.

Smaller works are almost uncountable. Fortified farms and defended villages are typical of the North and West, with clusters in central-south Scotland and Pembrokeshire densest of all. In general there was a wide variety in size, from capital ‘cities’ to single dwellings; and in function, from tribal strongholds, local refuges and communal animal pens, to markets, industrial centres and chieftains’ castles; sometimes all of these combined. In essence Iron Age forts were prototype towns and the Romans called them such. The very existence of this scale of structure suggests power, dictatorially held. Strabo states that Gaulish government was ‘normally autocratic’.⁵⁰ He adds that in Gaul ‘a single leader was chosen annually. In emergencies too, one man was elected by popular acclaim.’ These were late developments, in imitation of Roman practice and unlikely to be found in traditionalist Wessex and western Britain, where reoccupation of hillforts suggests increasing lawlessness. Smaller forts were being abandoned and the larger made stronger, perhaps in response to slave trafficking.

By contrast, what is now called south-eastern England was a separate world. Belgic invaders, arriving only a half century before Caesar, had brought new fashions and more advanced ideas. These did not, however, include non-violence. Their takeover of all southern England was probably forestalled only by the arrival of Claudius. Mighty men, painted and tattooed, with copious moustaches; glorious in war, with dashing chariots and flashing helmets of gilt and bronze; they carried shields of superb artistry, wore hauberks of iron rings⁵¹ and wielded long sword and spear. The Belgic settlers also led the descent to the plains. Though other areas were slow to follow, this process was now almost complete in south-eastern Britain, where topography was in any case less suited to hillfort building. Their oppida were nevertheless strong and guarded, presumably against chariot assault, by long dyke-and-ditch systems. These did not always form continuous circuits, but were often in straight lines, having some resemblance to a noughts-and-crosses layout, with settlement in the central square. Sometimes there are gaps, presumably where patches of woodland, thickened with thorn and felled tree trunks, made earthwork unnecessary. Caesar said that ‘the Britons call any fortified place in the thick forest which they have defended with dykes and ditches an oppidum (town) and use it as a refuge in time of trouble’.⁵²

Among these were the biggest British strongholds and the most sophisticated settlements. Wheathampstead (Herts), close to the Lea, encloses twice the space of Maiden Castle. Camelodunum (Colchester), scene of Claudius’ durbar, was biggest of all: twelve square miles, including pasture and fields. At its heart was a thriving urban centre and seat of kings, with a mint, plus commercial and industrial facilities. Though in time superseded by London (which did not yet exist) this would be Roman Britain’s first capital. Thus the process by which people congregate for safety, leading by degrees to formation of towns and so to social and cultural consequences unforeseen, was well under way by the eve of conquest. As an index of development, at least seventeen native coin types are known. These were adapted versions of classical originals, some with portraits of British kings and their names in Roman characters. They suggest an openness to Mediterranean influence and also the widening of a rudimentary reading ability, previously confined to the Druids. There were roads, unmetalled but serviceable, especially in summer. More broadly, here was a market economy able to produce food surpluses and to collect, transport and store them: factors which increased the risk of invasion, since Britain’s potential for feeding and moving armies can hardly have escaped the notice of Roman intelligence. By contrast there is no sign of native coinage and no trace of pre-conquest Roman money or trade goods north and west of a line from Bristol to the Humber.

Regarding population, the once accepted figure of two million may now be trebled. For example, the capacity of the Welsh border hillforts points to about sixteen persons per square mile, suggesting that the more populous areas of the late Iron Age were similar to the quieter regions of Britain today. Though farm had cut deep into forest, landscapes were far from treeless. The natural climax-vegetation of southern England is oak, elm, alder, lime and hazel, with holly, yew and bramble underbrush; mix and density depending on local conditions. Large forests persisted. The Fenlands and many smaller areas were still waterlogged. The notion that Britain’s moors and mountains remain unspoilt, while the lowlands have been transformed beyond recognition is, alas, a hiker’s view of history. On the contrary, while pockets of near-native woodland survive in the South, almost the entirety of upland Britain has, in relatively recent times, been laid bare by felling and kept bare by sheep; with poisonous or inedible plants like heather, bracken and gorse unnaturally dominant. Wales and the North were once cloaked in oak, elm, alder and Scots pine (pinus sylvestris). The glens of northern Scotland are thought to have been thicketed in stunted oak, thinning to birch on the lower slopes, with pine up to 2,000 feet. Ancient sources refer to northern woodland, including the great Caledonian Forest,⁵³ probably covering the entire central Highlands.

The Celts were among antiquity’s greatest artists. Content was primarily abstract, consisting in the main of curving and twirling lines. These have been traced to Greek and other sources, selectively imitated in a peculiar way. In effect the Celts copied detail such as decorative surrounds, of incidental importance to Mediterranean art, ignoring the representational part which was its true point. Marginal devices, like flowers, fronds, vine stems, abstract patterns and flourishes, were lifted so to speak from the edges of the originals and placed at the centre of Celtic expression. It is as if one were excited about the mouldings on a frame and ignored the picture, or bought a house because of its wallpaper. But the La Tène⁵⁴ artist was not a student of trivia. The borrowed bits-and-bobs were stylized and developed till the source was forgotten and the original excelled. Soon this school, fusing art with finest craftsmanship and metallurgy, developed a fluid beauty and a curious mystery of its own; simpler than later abstractions, like Arabesque, but bolder and more haunting also. Alongside these masterpieces, Roman provincial work, especially in the western provinces (where it lacked the hand of Greece) seems lustreless. The art of the Roman army too is generally childish without childhood’s charm. Unfortunately the imported would tend to replace the native, for La Tène modes of expression withered at the touch of Rome, not through deliberate discouragement but because they hung upon an aristocratic patronage which did not survive the conquest. On the other hand we must not be too hard on the Roman provincials, who make amends for miserliness in the fine by prodigality in the useful arts. Many have admired the profusion of instruments and practical objects to be seen in the most modest Roman museum. The Celts had style; but it was doubtless less unpleasant to have a dental abscess or a splinter in the eye on the Roman side of the frontier.

Artistic flair appears also to distinguish Gaul from German, an opinion to which Frenchmen still subscribe. Nevertheless Roman authors showed scant interest in Celtic creativity or capability, confining their comment to foibles. Caesar speaks of the Gaulish temperament as inconstant and untrustworthy.⁵⁵ Tacitus uses the expression inertia Gallorum⁵⁶ (the good-for-nothingness of the Gauls). Dio wrote of the Britons that ‘their boldness is rashness’.⁵⁷ As we have noted, recklessness, fecklessness and untrustworthiness were standard Roman views of the northern European. Tacitus adds that the British tribes ‘were once ruled by kings. Now the ambitions of petty chieftains pull them apart.’⁵⁸ This is substantially untrue. Kingly power was growing. Nevertheless in Britain, as in Gaul and Germany, there would be no coherent reply to the Roman threat. If a tendency can be recognized it is that the weaker tribes sided with Rome against the stronger. In Britain the east coast peoples, less confident of the terrain’s ability to protect them, generally accepted the invader more readily.

Hints from ancient authors suggest that Romans, who had viewed pre-conquest Britain with awe, tended to speak slightingly or mockingly thereafter. Hence Appian (2nd century): ‘The Romans already have the best half of Britain and do not need the rest, for even the part they have profits them nothing.’⁵⁹ There is also Dio’s tongue-in-cheek reference to an officer who had been severely disciplined: ‘Lucius Verus did not put him to death, but merely sent him to Britain!’⁶⁰

What was the Celtic view of Rome? Though some chose to resist and others to acquiesce, it is not difficult to imagine that all feared and hated her. What was imperium (Roman rule) but greed? Despite lack of nationhood there were doubtless pan-Gallic sympathies, fed by rumour and refugees. Even the remotest Britons must have been aware of Rome’s long progress through Spain and Gaul toward their shores, generating deep dread and resentment against a seeming conspiracy to expunge Celtic independence.

Turning to Celtic and in particular to British society: what we have said about style and swagger applies only to the upper class. That this was an aristocratic culture can be seen from social differences in burial. The majority was poor, ill accoutred and with little occasion for feasting and bombast. This underclass would not have much to lose and perhaps something to gain from a Roman occupation. A third estate was the Druids, a name which may have meant ‘brotherhood of the oak’:⁶¹ a priestly class in which religious knowledge, learning and literacy resided. Their presence is attested in Gaul and Britain only, with Britain as their heart and Mona (Anglesey) at the heart of Druid Britain. Anglesey, a fair island beyond wildest Wales, its legacy of magic stretching back to the Stone Age, was southern Britain’s ultimate refuge. This was doubtless a place apart, a northern Mount Athos, though we must not press that parallel, since human sacrifice was conducted there and observances included the burning of gargantuan effigies of straw and wood, with live humans or sacrificial animals inside. Because of these practices and because the Romans saw it as a focus of resistance, Druidism had been outlawed since earliest contact. This was a tactical error for it would give a desperate edge to resistance in places where the cult was most entrenched.

The Druids worshipped natural spirits, including trees, and believed in reincarnation. They were an intellectual élite, dedicated to philosophic enquiry and the pursuit of nature’s secrets, ancestors perhaps of the medieval alchemists: ‘In intellect lofty and united in brotherly societies, they scorned matters mortal and preached the eternity of the soul; prophesying the eventual ending of the world in a deluge of water and fire.’⁶² They were also society’s arbiters and peacemakers: doves made hawks by the knowledge that Roman victory would mean their dissolution. No major religious ceremony was possible without them. Though the Britons, like the Germans, worshipped in woods, a number of pre-Roman temples have been found: usually small and rectangular, like the one under Heathrow Airport. A custom which preserved some of their best metalwork was the throwing of votive objects into rivers and lakes. In addition there were sacred springs and wells containing human and animal bones, the deepest of which, at 140 feet, was found in Bavaria. There was also a cult of the severed head. Skulls, real or carved, adorned lintels of houses as well as gates and stockades of forts, though whether to glorify defenders or horrify attackers is not known.

Round houses, based on wooden uprights, leaning inwards, were the prehistoric norm, since this is the easiest way to make a shelter. The roof structure of an oblong or square building is more difficult, though once mastered there are advantages when it comes to division into rooms or fitting into rows to make streets. Rectangular buildings had long been standard in the Mediterranean. The Celts were now in transition between the two styles, though in Britain circular huts were dominant still: thatched and either all-timber or timber on low, dry-stone walls.

Another symptom of British conservatism was the use of war chariots, already obsolete on the Continent and long abandoned by the Romans (though still used for ceremony and racing). The idea of driving them through city streets, or of chariot-mounted generals leading the legions, is cinematic licence. The British chariots fascinated Caesar and he gave a detailed description of their tactics and the drivers’ acrobatics.⁶³ Obviously they were worthless against forts and other substantial defences. Furthermore, they represented a fighting method based on individual prowess and difficult to subordinate to a battle plan. Their real value had been as a shock weapon, but by now they had become something of an heroic cliché; except in south-eastern Britain, where echoes of the Celtic dream-world still resounded and chariots were an aspect of its bravura.

Strabo, geographer of the Augustan period, draws together several impressions in the following passage:

In Rome I have myself seen British youths, six inches taller than the city’s tallest, though bowlegged and of a displeasing appearance generally. In habits they are not unlike Gauls, but simpler and more barbaric. In warfare they use chariots. Their cities are the forest, for they enclose large areas of it by felling trees, building huts and keeping cattle within. The climate is rainy rather than snowy. Sometimes, on days otherwise fine, the fog hangs about so long that the sun comes through only for three or four hours around noon. The deified Caesar visited the island twice but did not stay long and accomplished little.⁶⁴

Such was southern Britain in the last years of her prehistory: defiant in isolation; in some ways advanced yet stranded psychologically in the heroic age; an apple ripe apparently for biting, though one which would prove unexpectedly sour.

Let us return to the young Colonel Vespasian, now in the heart of Thomas Hardy country, where resistance was most resolute, tradition longest and forts strongest. There is, of course, a limit to what we may make of one sentence in Suetonius. But archaeology assists with the exciting concordance between ancient pen and modern spade which emerged from Mortimer Wheeler’s 1930s excavation at Maiden Castle. Twenty-eight war graves were discovered near the east gate, containing skeletons with severe wounds of sword-thrust type, one with the head of a Roman ballista bolt lodged in the spine. The gateways had been slighted, the site evacuated and the inhabitants moved, perhaps to become the first residents of nearby Dorchester. Further, in the 1950s, Ian Richmond’s excavation inside the fort of Hod Hill (near Blandford Forum) found fifteen bolts in a cluster round the largest house. The rampart’s height, plus the hilltop situation, mean that fire must have been directed from a tall observation tower outside. After surrender the circumvallation was partially lowered and a Roman fort built into the north-west corner. A third hillfort thought to have received the artilleryman’s attention is Cadbury Castle, Somerset.

Two of Vespasian’s methods are now clear. The first was to drive the defenders away from an entrance with artillery fire, climaxing in an infantry rush. The timber gates would then yield to burning. Alternatively bombardment could be conclusive on its own, especially if observation of shot-fall were feasible and where defended areas were small enough to be within the field of fire, as in the case of Hod Hill. The oppidum would soon become a place of panic, with people and animals rushing about; and wherever they ran the guns could be directed to follow. So, of the ‘more than twenty oppida’ said to have fallen to Vespasian, we have evidence for two and the likelihood of three; a fair score when one considers the odds against finding a few pieces of roundshot or boltheads buried in those vast and numerous earthworks. It is also likely that Vespasian directed a sizeable resettlement operation: demolishing the stockades, with which the forts were crowned, reducing the ramparts and allocating lowland sites. It has been estimated that some 30,000 hilltop dwellers were in due course displaced in Hereford and Shropshire alone.⁶⁵

It is probable that Vespasian found, in Exeter, a suitable site for a legionary base and cornerpost for his flank of the advance; as had Plautius, in Lincoln, for its eastern end. Joining them was the Fosse Way, later name for a prehistoric track following a discontinuous limestone ridge for over 200 miles from Devon to Lincolnshire. This would be the approximate halt-line of the Claudian invasion. It is not clear whether the location was preselected or the offensive ran out of steam; whether Claudius thought he could hold southern Britain and ignore the rest; or whether he had no diagnostic insight and simply left his field commanders to amputate as best they could.

Here we leave the army of Britain, about to deploy along its overstretched and not overstrong frontier, while Vespasian returned to Rome, a hero’s welcome and the governorship of Africa. Claudius died in AD 54, reputedly poisoned by his wife. No matter, Vespasian continued to enjoy favour under Nero; that is until the calamitous lyre recital, when a career of promise was shattered in the winking of an eye. Suetonius develops the story:

In Greece, as a member of the imperial entourage, he committed the most appalling faux pas, falling asleep during one of Nero’s recitals. The upshot was total disfavour and dismissal from court. He then fled to some obscure country town, where he went into hiding for fear of his life. But in the end he was offered a province and an army command.⁶⁶

The province was Judaea. The command to crush the Jewish rebellion of AD 67. ‘In the end’ meant at the age of fifty-eight, twenty-three years after he had last seen service in Britain. Why this sudden recall; this amnesty, so seldom offered by the spiteful Nero; this reinstatement of which Ovid had vainly dreamed? It was because the Jews had gone on the rampage, murdered their procurator and destroyed a legion. Now, bold behind strong city walls, they defied Nero and all his works. Artillery was needed; a general who understood guns and knew how to use them against great fortresses. Then someone thought of that fellow, long rusticated, who bred mules; the one who had done so well against the Britons.

In Judaea, Vespasian’s first task was to reduce the Galilean stronghold of Jotapata.⁶⁷ Here the Romans faced skilful and desperate defenders, led by the historian Josephus, one of the few eyewitnesses of the Roman army in action, whose descriptions throw light on the effects of artillery as a terror weapon:

Placing his 160 pieces in a ring facing the wall’s defenders, Vespasian commanded the bombardment to begin. A salvo followed, the scorpions shooting bolts, the ballistae hurling stones weighing over 100 lbs, plus flaming torches and a hail of arrows. All this not only cleared our men off the wall but also from a broad zone where the missiles were landing within; for the Arab archers, the javelin throwers, the slingers and the guns were all firing in unison.⁶⁸ Such was the artillery’s power that a single bolt transfixed a row of men. The stones removed battlements, even dislodging the corners of towers. The formation does not exist that can stand against rocks like this which carve through rank after rank. A man beside me on the wall was decapitated, his head rolling three furlongs. Most frightening of all was the rushing sound as the missiles flew through the air and the sickening thump of impact. Added to this was the thud of bodies as they fell from the wall. Soon the sentry-walk could be reached simply by scrambling up the corpses. Inside was the wailing of women, outside the groans of the dying.⁶⁹

The master had not lost his touch. Of the siege of Jerusalem, at which Vespasian’s elder son Titus later commanded, Josephus described how lookouts, seeing the ballista-shot on its way, shouted warnings in time for the defenders to take cover. In response, the gunners blackened the stones, making their flight almost invisible. Reminiscent of Hod Hill, though using larger ordnance, catapult balls up to eighteen inches across have been found beneath the convent of the Sisters of Zion and at other Jerusalem locations.

The siege of Jotapata ended with the city’s fall and Josephus’ capture. Hearing his prisoner had the gift of prophecy, Vespasian requested that his own fortune be told. Anticipating Macbeth’s witches, the Jew assured the Roman that he would be emperor thereafter. The superstitious Vespasian was greatly impressed. Josephus was held under gentle duress and persuaded to act as interpreter and mediator, in which role he witnessed the six-month agony of Jerusalem from the Roman side. He would later become Vespasian’s friend, living as a pensioner in Rome and writing his history of the war.

Now Vespasian’s fortunes were about to take another extraordinary turn. The year after Jotapata, events in Rome caused the rebellion to be shelved, forgotten almost, and the rebels at least temporarily reprieved. Nero had been toppled. The Delphic oracle is said to have told Nero to beware the age of seventy-three. Only thirty-one, the emperor congratulated himself on having forty-two years to live. But already Galba, governor of Spain, aged seventy-three, was marching on Rome.⁷⁰ Deserted by the Praetorians, disowned by the army, reviled by the Senate: time was running out for the Julio-Claudian dynasty. Hiding in a servant’s house on the city’s edge, Nero stabbed himself in the throat and expired, crying, ‘Jupiter, what an artist dies in me!’ This was June, AD 68. The following year, known as the Year of the Four Emperors, was one of Rome’s worst. ‘The vacancy of the throne’, as Gibbon commented darkly, ‘is a moment big with danger and mischief.’⁷¹

So it was that Vespasian, a highly successful and popular general, whose rocklike character had immense appeal after the caprices of Nero and Caligula, happened, at that perilous moment, to be far from the dangers of Rome and close to the protection of armies. The prefect of Egypt broke the ice by declaring in Vespasian’s favour. Emerging one morning from his tent, soldiers began to greet him as emperor, soon followed by the entire Judaean expeditionary force. Finally Syria, jewel of the eastern provinces, came out on his side. Now he could wait while the western claimants killed each other. When the time was right he would intervene as restitutor orbis, putter-to-rights of a Roman world gone wrong.

Toga muddied by Caligula, pelted with turnips in Africa, caught napping by Nero, hiding in hick towns, breeding mules: such was the improbable path to power of T. Flavius Vespasianus who, in the ten years remaining him, would do more than any emperor until that time to caulk Rome’s leaky frontiers and bring Rome’s insulation from the outside world a step closer to completion.

Four characters now enter the picture. First Vespasian’s sons, Titus and Domitian. Mindful of the Year of the Four Emperors, Vespasian insisted that the Senate accept them as his heirs. The elder, Titus, who completed his father’s work in Judaea, succeeded him in AD 79. His reign is memorable for the eruption of Vesuvius and the inauguration of the Colosseum. He was dashing, generous and popular. However, at the age of forty-two his health gave way and he died after little more than two years on the throne.

He was succeeded by Domitian, the shy and quiet younger brother, soon to reveal himself as a tyrant of Stalinesque suspicion, whose jealous eye lighted on all successful people. During his sixteen-year misrule, especially its sinister second half, the Roman state would be paralysed, its servants daring neither to fail nor succeed.

Then there is our principal recorder of events: the historian, Cornelius Tacitus, whose tortured personality may be understood as a product of terror, for his boyhood was passed under Nero and his manhood, from the ages of twenty-six to forty-one, under Domitian. He published nothing. Not only was he in danger of writing something which could give offence, it might be fatal just to write well.

Finally there is Julius Agricola, second soldier of this Episode, who served three times in Britain and whose name is forever associated with the country now called Scotland. Vespasian had been a genial, approachable and sound emperor, whose policies were of peaceful consolidation and thrift. Britain was the exception. There he had won his spurs and it pained him to see the enterprise languish. Accordingly he dispatched a series of Rome’s best men. With Agricola, history is helped by a remarkable coincidence in which Tacitus is linked to the battle for Britain. The connection was the youthful historian’s marriage to Agricola’s daughter. In due course the son-in-law would be the father-in-law’s biographer. Roman historical writing, fascinated by events at the centre, rarely mentions the margins, let alone the barbarian lands; and we are grateful to scavenge a sentence or two. In Tacitus’ Life of Julius Agricola we have an entire book about the empire’s edge. It is also a book about personalities and problems, in which a general’s duty is played against an emperor’s envy. Its subject is a province without a frontier, unable to find a territorial balance or to strike a durable bargain with the outsider; its context an empire no longer decisive about whether to take territory or to leave it. Though not an eyewitness account with the intense involvement of Ovid, it is the next best thing; casting a powerful beam into Europe’s farthest corner, where we would otherwise grope by archaeological candlelight.

Agricola was born in Forum Julii (Fréjus, Côte d’Azur) and educated in Marseilles. When Claudius invaded Britain he was three. At eighteen his first posting was as tribune (second lieutenant) attached to the staff of Britain’s fifth governor, G. Suetonius Paulinus.⁷² By now the Fosse Way had failed at its western end, largely because of the Welsh wasps’ nest, prodded by the fugitive prince, Caratacus, and buzzing still.

The second governor, P. Ostorius Scapula, had advanced his left to Gloucester and probably then to Chester. This was a crucial position, commanding the approach to North Wales and severing a possible alliance between Welsh and Pennine tribes. According to Tacitus, Scapula died of stress and exhaustion. Nevertheless he defeated and captured Caratacus, though the latter lived to ask his famous question of imperialism, whose artlessness disguised a sarcastic comment on Roman greed: ‘And what do you, who have so much, want with our wretched tents?’⁷³

Moving on to Paulinus (with the young Agricola on his staff): here was a general who had earned himself a reputation for mountain warfare in Morocco as first to lead Roman soldiers across the Atlas. He now spent two years on the reduction of North Wales, finally isolating Anglesey, heart of nationalist hopes and Druidical dreams. Its capture would be an achievement to equal that of Nero’s general, Domitius Corbulo, in Armenia:

Britain’s new governor was Suetonius Paulinus, Corbulo’s rival both as a strategist and for public esteem. Could he produce victories to match the retaking of Armenia? He now decided on the capture of Mona (Anglesey) which had been a refuge for so many. Flat-bottomed boats were built to take the infantry across the treacherous shallows. The cavalry used fords, some troopers swimming beside their mounts. The armed enemy crowded the opposite shore. Among them were women, robed in black, hair wild like Furies, waving flaming firebrands. Nearby the Druid priests, with hands raised, called down terrible curses from heaven. This awful spectacle brought our soldiers up short. They stood as if frozen, until the general broke the spell by shouting how shameful it would be if they were halted by a gang of lunatic women. So the eagles surged forward, hemming in the enemy who was burned by his own torches. Paulinus occupied the island, felling the sacred groves dedicated to Anglesey’s vile rites; for it was among their beliefs that altars should run with captives’ blood and that prophecies should be made by examining human entrails.⁷⁴

But Britain, a reluctant yielder of laurels, was not yet ready to let Paulinus win his. At this juncture there came news of rebellion, 250 miles to the rear. It involved the Icenians of Norfolk and the Trinovantians of Essex and Suffolk who had, for more than a decade, been simmering with resentment at the granting of their lands to Roman veterans. The late Icenian King, in the hope of saving some family influence, had willed his territory in part to his daughters and in part to Rome. It was like asking a pig to leave half the trough. When his widow Boudicca (Boadicea) protested, she is reported to have been flogged by Roman officials and her daughters raped. Such was the flashpoint for the last great attempt to reverse the conquest of the Celtic world. Dio describes the queen as:

Boudouika [sic] a British woman of royal blood, with more brain than women usually have; tall, terrifying, with flashing eyes, menacing voice and a wild mass of yellowish hair falling to her waist; wearing a great, golden neck-torque, a many-coloured dress and thick cloak, fastened with a clasp. Spear in hand she harangued a gathering of armed men 120,000 strong.⁷⁵

One should interpolate that what leaders said to armies, when no Roman was in earshot, was largely conjecture. Indeed it was an ancient convention to put speeches into the mouths of commanders; following Thucydides, who confessed that, since verbatim reporting was seldom possible, he would write what he thought the occasion demanded! Nevertheless such speeches rarely lack information. Here Dio makes the interesting logistical comment through Boudicca (already quoted): ‘While we are able to subsist on wild plants and water, they depend on bread, wine and olive oil; and if one of these should fail them, they are finished.’⁷⁶ Grain was indeed essential; and while it was an exaggeration to claim that the eagle would not fly without wine and oil, sour wine in which to dip the bread and oil as the universal cooking medium were the soldiers’ normal expectation.

Much of Boudicca’s tirade is of course predictable: northern liberty contrasted with oriental servitude, British hardihood versus Roman decadence:

I am queen not of toiling Egyptians or money-grovelling Assyrians.⁷⁷ I beseech heaven for victory against these insolent and insatiable men – if those who take warm baths, eat sweetmeats, drink wine unwatered, smear themselves with scent and lie with boys on soft couches, deserve the name of men! They who are lackeys to a lyre player – and a bloody awful one at that!⁷⁸

Boudicca now struck swiftly at the lightly guarded towns of an already Romanized South-East. ‘Avoiding fort and strong-point, the rebels headed for where the booty was richest and protection poorest; itching to chop heads, stretch necks, burn and crucify; as if taking revenge now for the retribution which would come later. The Roman provincial dead were put at 70,000.’⁷⁹ Colchester, St Albans and London were destroyed and their Roman or pro-Roman populations butchered before Boudicca was defeated, perhaps in Northampton-shire or elsewhere in the West Midlands. The queen and her family took poison. Such were Agricola’s experiences at twenty years of age and his first taste of service in Britain.

In due course there followed the Year of the Four Emperors, with the heavy drafting of British troops to fight in the civil war. Those who stayed held steady; perhaps too busy with Wales to become involved in continental politics. None the less, reduction of the army’s numbers would have dangerous repercussions in the North. The Brigantians were, according to Tacitus, the biggest British tribe, whose territory covered almost the entirety of what is now northern England, from Peak to Solway and from coast to coast. They were shepherds and stock rearers, ruled by a rich nobility. Their forts were few but impressive, like that on the summit of Ingleborough; or the sprawling, lowland complex of Stanwick, in North Yorkshire, seemingly enlarged in panic at Rome’s advent to 730 acres! From here, or perhaps from Almondbury, near Halifax, Queen Cartimandua ruled in friendly alliance with Claudius and Nero. However, Brigantian politics included the usual anti-Roman faction and AD 69, Rome’s hour of weakness, saw a successfulcoup, led by Venutius, the queen’s ex-husband. Again one senses the stress set up by Rome’s proximity; sundering husband and wife as it had Armin and his brother. Venutius now thought to assume the mantle of Boudicca as leader of British resistance. He might have been less eager had he reflected that this same year, which brought him to power, also called to the purple the master-gunner, whose Wessex campaign of a generation earlier had consigned so many British strongholds to oblivion. Vespasian did not return to Britain in person but would field his strongest side in order to quicken a conquest begun twenty-six years earlier and whose termination was exasperatingly overdue. Accordingly he appointed three successive soldier-governors of unusual capability: Cerialis, Frontinus and Agricola, who would respectively take in hand the North of England, Wales and most of Scotland.

Petilius Cerialis was the emperor’s son-in-law, a dashing general in the Caesarian mould, whose success in AD 49 had beaten Vespasian a path to the throne. He was first sent to put down a revolt on the lower Rhine. Tacitus describes his raffish and unorthodox character, careless of the trappings of discipline and a brilliant improviser. There had been a whiff of scandal. The enemy attacked while Cerialis was out of camp spending the night with a local woman. However ‘luck always covered his lapses’. It was in a speech to the Germans that Tacitus attributed to him that crisp appraisal of the Roman system: ‘no peace without armies, no armies without pay, no pay without taxes.’⁸⁰

Cerialis took Agricola to Britain as one of his legionary commanders. It was his second tour and he was now thirty-one. We know little of the campaign, doubtless because Tacitus wished to underplay this part of the biography, keeping his best cards till its hero would himself be governor. We do, however, know that Cerialis sealed off Wales then headed north, defeating Venutius on some unrecorded field.⁸¹ On the dreary Stainmore Pass is a legion-size, twenty-acre marching camp, cut in two by the modern road (A66), whose square shape and eleven gateways suggests this advance. With two similar camps in the upper Eden and Petteril Valleys, these point straight toward Carlisle, where Cerialis’ drive is thought to have ended. The IXth Hispana legion now moved up from Lincoln to York, though there is no evidence for a widespread occupation of the North at this date.

Next came Roman Britain’s tenth governor, Julius Frontinus, author of two books on the art of war: De Rei Militari (Matters Military) and Strategematon (Strategems). The former and more important is lost. The latter, its supplement, gives examples of military ruses said to illustrate the principles of the earlier book. These are, in our view, largely nonsense. Nevertheless there was nothing nonsensical about his struggle in South Wales, where the resurgent Silurians had drawn the conclusion, from Vespasian’s campaign of vivid memory, that Rome was better fought by taking to the hills than by waiting in the hillforts. It was this adoption of guerrilla tactics, plus Rome’s preoccupation with civil war, which lengthened the Welsh involvement to thirty years and thirteen known offensives.

Pacification of Snowdonia, followed by the second invasion of Anglesey, was left until Agricola’s first year as governor. His Welsh conquest employed the network method, in which successive mountain blocks were isolated by building roads in the surrounding valleys and placing forts at key intersections.⁸² Agricola would also apply this to southern Scotland; which helps explain the large number of forts and immense road mileages with which he is credited. First, however, Wales and its borders received over 700 miles of road and thirty-eight forts or fortlets, plus strategic supervision from Legio XX Valeria Victrix at Chester and II Augusta at Caerleon-on-Usk (near Newport, Gwent); though all would be held in sketchy fashion during the subsequent offensive into north Britain.

Let us back-track for a moment, to include a happier interlude. While Frontinus was struggling in South Wales, Agricola enjoyed the governorship of Aquitania (south-western France) followed by the consulship. It was toward the end of this period that his daughter (whose name, like that of Ovid’s wife, is not mentioned in her husband’s work) married Tacitus:

After less than three years as Governor of Aquitania he was recalled with the consulship in view. It was while he was consul that I was betrothed to his daughter: a girl of exceptional promise – and I still in my youth. On the expiry of his year of office he placed her hand in mine just before he was given the governorship of Britain.⁸³

So Agricola, now thirty-eight, returned to Britain as eleventh governor and commanding general of her four legions, with clear instructions from Vespasian to finish off Wales and then proceed to the subjugation of Britain in toto. In practice it meant the conquest of what is now Scotland, which would occupy most of the seven-year term of governorship and is the main subject of Tacitus’ biography. This was written some eight years after Agricola’s death, a postponement caused by the long wait for Domitian to die in turn. Nevertheless, some time after AD 96, with Domitian replaced by a humane successor, Tacitus was able to affirm: ‘At last Nerva has united the two things we had ceased to believe compatible: rule by an emperor and freedom. Meanwhile fifteen years have been erased from our lives, during which time the young became middle aged and the elderly old, all without daring to say a word.’⁸⁴

As long as Domitian lived, Tacitus, like Ovid, could be described as a literary exile, though of an opposite kind: the one able to write but condemned to live in Tomis, the other able to live where he wished but condemned to silence. Now, past forty, he was free to begin and the Agricola would be his first effort, published just before the Germania. Both are preserved in their entirety.⁸⁵ The earliest known reference to an Agricola manuscript is from the monastery of Monte Casino, around the time of the Norman conquest of England. It then disappeared and was presumed lost. In 1431 the Florentine scholar Niccoli discovered a 9th-century version at the Hersfeld Monastery in southern Germany. Pope Nicholas V, an avid collector, ordered his agent Enoch of Ascoli, who scoured northern Europe for this purpose, to acquire it; and receipt was recorded by the Papal Secretary Decembrio in 1455. It was printed in Milan, c. 1475; and it is perhaps from this version that the spelling mistake Mons Grampius for Graupius first arose, beginning the tradition which saddled Scotland with the erroneous name Grampian to this day.

Why should Vespasian favour the total conquest of Britain, irrespective of the island’s value? Unlike Spain, whose promise revived as the gold-rich, north-west corner came closer, Britain’s dwindles as one moves in that direction, in all senses except the scenic. However, in Spain’s case the completion of conquest had brought savings as well as profit. Victory allowed defence cuts. With time these increased to the extent almost of demilitarizing the Iberian peninsula. This was a formula which could be applied to a Britain grossly over-garrisoned in relation to her worth. A supreme effort now could save indefinite occupation costs, inescapable as long as Rome sat in one part of the island with enemies at large in the other. So, 134 years after Caesar and thirty-six after Claudius disembarked in Kent, Rome’s eyes turned toward the far north at last.

Though we use it for convenience, the name Scotland is strictly speaking incorrect until the Scots’ arrival from Ireland in the post-Roman period. Pre-Roman Scotland (if we may so call it) was more populous than is often thought. There were some thousands of hillforts of less than one acre. In what are now known as Northumberland and southern Scotland almost every hill was utilized, the densest clusters being on the upper Tweed (near Peebles) and upper Teviot (near Hawick). Defended settlements in valley situations were also numerous: farms or small villages of round, timber-and-thatch huts, encircled by mound and ditch. In the Border region there were a few larger strongholds resembling those of southern England, such as the already-mentioned Eildon Hill (near Melrose), the Selgovian capital, an oppidum with perhaps 2,000 inhabitants. Even the Highlands were moderately peopled. Here were the small, stone-walled forts known as duns, most frequent in Argyll and on Skye. Further north still were the brochs: round, dry-stone towers as high as forty feet, with a staircase between double walls, especially common in Sutherland, the Western and Northern Isles. All tribes were pastoral with some cultivation. Tillage was more developed and forest clearance more extensive than once supposed. These amounts of settlement and development would have been less surprising to someone living only 150 years ago, for the emptying of the rural north is relatively recent.

The existence of so many separate communities, plus an obsession with defence, suggests Iron Age Scotland as a particularly peaceless place. It is clear from Roman sources that a Highland-Lowland distinction already existed and, as in later ages, these will doubtless have been at loggerheads. Pliny gives the expression Caledonia Silva (the Caledonian Forest) and Ptolemy places the Caledonian tribe right across the central Highlands, south of the Great Glen; a tract without a single large fort or town. He locates another, the Vacomagi, in north-eastern Scotland. By the 1st century AD these were probably one confederation. Tacitus speaks of a distinct Highland type, ‘whose red hair and long legs betray German origins’.⁸⁶ Red hair perhaps, but German origins are unlikely. Despite this and other hints of racial differences which today might be seen as evidence of the earlier-than-Celtic origins of the British peoples, Tacitus concludes that the generality of Britons came from Gaul, a view disputed by modern archaeology:

You will find there Gallic customs and beliefs. Nor is the language dissimilar. Temperamentally there is the same rashness: dashing into danger and, when it is met, dashing out of it with equal eagerness. Even so the Britons are more mettlesome than the Gauls, having been under Roman occupation for a shorter time. Those already conquered are moving toward what the Gauls are now; and those unconquered remain as the Gauls once were.⁸⁷

As ever the weakness was tribalism. Tacitus’ comment that ‘the ambitions of petty princes divide them’,⁸⁸ was evidently truer of Britain than of Gaul. Indeed quisling chieftains, like Cogidubnus of today’s West Sussex had proved a valuable tool of conquest, ‘using the time-honoured means of Roman diplomacy, by which kings are persuaded to serve as servitude’s instruments’.⁸⁹ The following sardonic passage on the sedation of the northern peoples – by following a brutal conquest with the introduction of Mediterranean pleasures – is equally a compliment to Rome on her handling of the newly conquered:

In order that a rude and scattered people might be coaxed toward peaceful paths through comfort, he [Agricola] encouraged the building of temples, markets and houses. The sons of chieftains were educated in the liberal arts; and those who not long before spurned Roman speech began to aspire to rhetoric and adopt the toga. So by slow degrees the Britons were seduced by pleasant pastimes, like strolling through colonnades, relaxing in the baths and attending polished entertainments, till finally the gullible natives came to call their slavery ‘culture’!⁹⁰

Returning to Scotland, and to geography, Tacitus describes the sudden widening of Britain upon ‘crossing’ into Caledonia (i.e. after leaving Central Scotland) when it becomes ‘a vast and irregular shape’; adding, in a flash of knowledge about the north-west Highlands, ‘Nowhere has the sea more mastery, more conflicting motions. Nowhere is land and water more intermingled; the sea among mountains, making them its own.’⁹¹ He also gives a fair summary of Britain’s climate:

The sky is cloaked in cloud, the rain incessant, though the cold not unduly severe. The length of day is difficult for us to grasp. The nights are clear and – in the extremity of Britain – brief, so that a short interval separates dusk and dawn. They even say that if no clouds obscure the sun it shines all night, tracking across the sky without either setting or rising. The soil (save for olives, wine and other warm-climate fruits) permits crops and favours cattle. Seeds sprout quickly in the damp earth but ripen slowly in the damp air.⁹²

As described, Agricola spent his first campaigning season, probably that of AD 78, in Wales. In 79 he consolidated Cerialis’ conquest of northern England. However, Vespasian’s death, in the late June of that year, put a spoke in Agricola’s wheel. All campaigning ceased. War was an imperial prerogative and each new emperor must confirm or veto its continuation. Though this meant delays, worse lay ahead. From a Vespasian unable to accept that the British venture should fail, power would pass to a Domitian unable to stomach its success. Nevertheless, Vespasian’s policies were for the moment safe in the hands of his elder son; and during the winter of 79–80 one may assume that Titus sent Agricola orders to proceed. The crucial phase of the British plan now began to unfold.

In the spring of 80 Agricola advanced into new territory. His eastern column pioneered the inland route (now the A66), choosing Cheviot rather than coast, since the latter was longer, with forest barring the lower Tweed, while the almost harbourless Northumbrian shore offered little likelihood of naval support. Southern Scotland was ‘networked’, but compared with Wales resistance proved slight. This remarkable season ended with Agricola’s vanguard at the Tay.

A question now arose regarding the Forth-Clyde isthmus. Here Agricola had found the wasp’s waist, a defensible line of only thirty-five miles, seven times shorter than the Fosse Way. Doubtless his despatches pointed to the advantages of utilizing this line as a frontier and contrasted it with the severity of the Highland alternative. Titus may have accepted such a view. At any rate the northward advance appears to have ceased and the 81 and 82 seasons were spent consolidating southern Scotland.

Even so, this seems an unduly long pause, best explained by Titus’ death on 1 September 81. Again Agricola was obliged to await new orders. His campaign had encountered the double-postponement of two emperors’ deaths in two years. All chance of surprising the Caledonians had been lost. To make matters worse, Domitian had decided to lead his own expedition across the Rhine. For this he would require detachments from all four of Agricola’s legions. The year 82 was therefore one of dismal delay and manpower reallocation in which, as the army of Britain was pared down, the chances of taking and holding northern Scotland dwindled.

Meanwhile Agricola was pondering the possibility of invading Ireland. We glimpse him standing with his staff at the Rhinns of Galloway and looking across to the last outpost of free ‘Celtica’, only twenty-five miles distant: ‘Often I heard him say how Ireland could be taken and held with only one legion and auxiliaries; and that this would help regarding Britain too, for with Roman arms appearing everywhere, liberty would seem to be nowhere.’⁹³

The foremost source of ambiguity remained central Scotland. Should he fortify the isthmus as a frontier or not? Here Tacitus is unusually specific on geography but vague on policy, evidently reflecting the indecision of that year:

The fourth summer was spent securing what had been so swiftly gained. He could in fact have found a stop-line within Britain itself, had the army been less keen and Roman pride less easily satisfied. For here Clyde and Forth, driven inland by the tides of opposing seas, are sundered only by a neck of land. This was now secured with garrisons and the whole tract taken in hand, the enemy being pushed back into what is in effect another island.⁹⁴

The strategic view from central Scotland was quite different from that further south. In the first place, discovery of the Forth-Clyde isthmus had shown how easy it would be to seal off the Highlands. Secondly, proximity to the Highlands made it clear how difficult their invasion would be. Finally, revelation of the Irish Sea’s narrowness made a move in that direction more attractive than before. We may guess that, in view of his loss of manpower to the German expedition, Agricola favoured the last of these choices. Being less mountainous, Ireland would require a smaller force, leaving a reserve to man central Scotland. It is presumably with reference to an Irish invasion that Tacitus tells us he occupied the coast of south-west Scotland ‘hopefully rather than defensively’.⁹⁵ He had a substantial fleet and was ready to use it. Of the two courses an invasion across ocean would have been the more prestigious and in later life he spoke wistfully that it was not chosen. By contrast, the army, always shirking the sea, preferred to march north. ‘Penetrandam Caledoniam’,⁹⁶ they are heard shouting (‘let’s get stuck into Caledonia’).

The dilemma was resolved by a directive from Rome: Agricola was granted a second term as governor and ordered to proceed with the conquest of all Scotland. This is consistent with Domitian’s malice. Agricola was being manoeuvred into the most difficult of the options with a much reduced force. Meanwhile the emperor would lead an army into Germany with more achievable objectives and shorter supply lines.

So, unaware of being on a collision course with the imperial psyche, 83 and 84 see Agricola advancing through Strath Allan, Strathearn and out into the broad corridor of Strathmore, setting the pattern for all subsequent invasions by skirting the Highlands in a right-flanking movement toward Aberdeenshire. The navy’s role was crucial to this east coast strategy. By constant contact with its warships and supply vessels at the Tay, Montrose, Stonehaven and Aberdeen, the army could be fed and supported. Agricola’s handling of combined operations proved to be his strongest card, more successful in the light of others’ failure. Fleet support had been the weak link in Caesar’s British forays.⁹⁷ We will recall Germanicus, the drownings in the Waddensee and shipwrecks on the Frisian Islands. During the conquest of north-west Spain a comparable strategy seems to have lapsed owing to fear of the open Atlantic.

Agricola was the first to use the navy as part of his armoury; and it made a stirring sight as it brought up the rear. So the offensive was carried by sea and land simultaneously. Many a time the infantry, cavalry and sailors would meet over a shared meal, capping each others’ stories: the one about forest and hilltop, the other about wind and wave; conquest of the land on the one hand and of the ocean on the other. For their part the British were stunned by the fleet’s appearance. It was as if their seas and shores were being stripped of secrets and their ultimate bolt-holes barred.⁹⁸

Attempts have been made to trace the campaign through its marching camps. Indeed the terrain has yielded little else. This ubiquitous instrument of Roman mobility is worth describing in broad terms. Marching camps were usually square or rectangular. Internal layout reflected the order of march, with the commander at the centre, the vanguard in the front and the main force behind. Since everyone’s tent and each tent’s place were always the same, a standardized pattern was intended to soothe tired men and speed them to their rest. All soldiers bore its blueprint in their minds and could recreate it perhaps in a couple of hours, using the trenching tools which were standard equipment. An advance party under the praefectus castrorum (camp boss) selected and marked out the site. While the main body dug the ditches and threw up the spoil into perimeter mounds, the rearguard unloaded the pack animals and the leather tents were erected in precise pattern within. Every infantryman carried a pila muralia (wall spear), which was driven into the rampart so that it bristled like a porcupine.

However, camps were not wholly stereotyped. The praefecti had their quirks and latitude was allowed in external proportions or the design of gateways, with their various inturned or outturned extensions, intended to impede or deflect attack. For example, one of Agricola’s camp bosses favoured a peculiar sickle-and-hammer-shaped opening. In theory such variants allow us to plot a progression of camps, distinguishing them from other campaigns or periods.

Ancient sources tell of camp construction every evening and destruction every morning; the unit refilling the ditches before it marched away. If a camp were left intact the enemy could occupy it, perhaps to lie in wait for a tired Roman force expecting to use it again. But in practice, due to laxity or successive occupancy, demolition was not always carried out. Had it been, Britain would not have inherited so many fine examples. There are also instances of camps superimposed, or of small built into the corners of large.

A camp’s size gives an idea of numbers. Since wasted space meant defending a larger perimeter, sprawling layouts would be avoided. Occupation density sub pellibus⁹⁹ (under canvas) has been calculated as about 300 men per acre.¹⁰⁰ Spacing between camps tends to be about fifteen Roman miles, the standard day’s march for fully laden troops. This is a reasonable distance in view of the additional chores of making and striking camp, packing and unpacking, foddering animals, cooking one’s own meals and so on.

From the Firth of Forth to the Moray Firth some fourteen marching camps associated with Agricola’s campaign curve northwards in a 150-mile arc. Near their beginning is the great base of Ardoch,¹⁰¹ where six overlapping camps, three clearly visible, show it as the hub of northbound traffic. Of this long line Raedykes, behind Stonehaven, is probably the most rewarding to visit. The furthest known is at Bellie, near the Moray Firth. The furthest visible on the ground is Ythan Wells, Aberdeenshire, whose gorsey dyke signals with yellow flag the empire’s northernmost upstanding work: 1,300 miles as the crow flies from Rome, three months’ march at standard pace. The Moray Firth may be seen as the limit of a thousand-mile swath, cut by Rome into the Celtic and British worlds.

Here, then, Tacitus’ text, the marching camps, and the lie of the land, are the sum of evidence. None is overhelpful. Tacitus is unspecific on geographical detail, though he gives the useful hint that Agricola ‘divided the army into three’.¹⁰² The camps resolve into two main sizes. The larger, around 110 acres, shows the army marching as one. At 300 men to the acre, this is consistent with a total force of 32,000: four understrength legions of about 4,000 each and an equivalent number of auxiliaries. The smaller, around thirty acres, is roughly compatible with division into three. But there are many imponderables. Missing or undiscovered camps, multiple use, varied unit groupings and the likelihood that some camps were built during return journeys, make this a dubious guessing game. It may, however, be pessimistic to say that the story will never be told. Archaeology and aerial survey are a potent partnership; and the camera’s eye is sharpened by drought, which increases colour differences in herbage, caused by soil disturbance. Remarkable detail can be wrung from these inscrutable bivouacs, such as the location of rubbish pits, where dateable material might be found. Recent years have been exceptionally dry. Of the ten hottest summers on record, six were in the 1980s. ‘Global warming’ brings some recompense.

If details of Agricola’s march are obscure, his strategy is not. First he must avoid entanglement in guerrilla warfare, which meant avoiding the hills. Second, he must bring the enemy to him, forcing a battle on favourable ground. The accepted way to provoke confrontation was rampage: destroying crops, burning houses, butchering animals and even people. North-eastern Scotland, a tract of lowland in a highland setting, is well suited to such grim tactics. Its rich lands could be devastated while their agonized tenants watched from the hilly grandstand where they had taken refuge. By putting the following sentence into the supposed speech of Calgacus, the British commander, Tacitus allows himself a veiled comment on this sinister aspect of the pax Romana: ‘What they call “empire” is theft and butchery; and what they call “peace” is the silence of death.’¹⁰³ Barbarity was not, it seems, a barbarian monopoly.

Nevertheless, the policy worked: drawing the enemy toward a rendezvous at Mons Graupius, somewhere along the line where plain and mountain meet. Where is Graupius? Though it has given its name to Britain’s biggest massif, the battle site was probably at the foot of a single hill. This must have been big enough to hold a substantial British host, close to a Roman camp, with some flat ground adjacent and rising ground behind. Tacitus describes it briefly: ‘The Britons were drawn up in a striking and frightening way: their front rank level with ours while the rest, on rising ground, appeared to tower up behind them; the chariots rushing noisily to and fro across the fields between.’¹⁰⁴ These vague clues could apply to almost any hill near almost any marching camp along a Highland front of nearly 200 miles. Here was a veritable Teutoburg Forest of a puzzle, combining hints from a great historian with clues from a romantic landscape. What could excite scholarly Scotland more? The result was centuries of learned (and sometimes not-so-learned) debate, with everyone adding his piece, including Sir Walter Scott: ‘Our Scottish antiquaries have been greatly divided about the local situation of the final conflict between Agricola and the Caledonians. Some contend for Ardoch in Strathallan, some for Innerpeffry, some for Raedykes in the Mearns and some are for carrying the scene of action as far north as Blair in Athole.’¹⁰⁵

So, between about 1725 and 1975, the sharpest eyes of Romano-British scholarship ranged up and down the line where grass and heather meet, its prestigious pens producing theories neither provable nor disprovable. But at the latter date debate subsided. J. K. St Joseph (1912–94) longtime director of the Cambridge University Committee for Aerial Photography, whose work in Britain (from 1930) had revealed forty-three forts, twenty fortlets, fourteen signal stations and 235 camps, played a strong card.¹⁰⁶ A photographic sortie of 26 July 1975 revealed a new marching camp at Durno, six miles northwest of Inverurie and twenty-three north of Aberdeen. At 144 acres this is the largest known encampment beyond the Forth, big enough to accommodate Agricola’s entire force with room for prisoners of war besides. Might this imply that, although forward units had already reached the Moray Firth, all were now recalled and recombined to meet some exceptional event? Above Durno towers the bulk of Bennachie¹⁰⁷ (1,733 feet), a plateau four miles long and two wide, linked by a ridge to the body of the Highlands and commanding views southwards to Stonehaven and eastwards to the North Sea. Pending discovery of weapons, graves, or other finite evidence, this is presently the strongest candidate for the identity of Mons Graupius.

Before the battle, Agricola addressed his army, emphasizing the feat of exploration and the pride in trailblazing: ‘You have outshone previous armies and I previous governors. Britain’s bounds are no longer rumour and guesswork, but as factual as our forts and our strength. We have brought Britain both to light and to heel.’¹⁰⁸ The encounter itself was a tragic preview of Culloden: tribal motley facing trained troops, the unwieldy against the compact, wild passion versus professional calm, 10,000 clansmen slaughtered for the loss of 360 Roman auxiliary soldiers. The legions did not even engage. However, the Britons were so numerous that the majority could not come to grips. Twenty thousand were able to scramble back up the hillside or disappear into nearby forest, where nightfall covered their retreat. With so many left to fight again and the Highlands yet unpenetrated this could hardly be called a decisive victory. At most it was a promising start. Beyond lay an interior made for guerrilla warfare, unsuited to the formal encounter, hopeless for cavalry and demanding a dangerous dispersal of force. Wales, a mere fifty miles wide, had required thirteen offensives. The Highlands are over 120 miles and behind is an even wilder world: a jigsaw of sea and land; as if carved by the hand which had tried to make Greece and tossed it, as a reject, into Europe’s opposite corner.

So far, however, the Scottish campaign had gone well; perhaps too well. Before long rumours of glorious Graupius would reach a Roman public more interested in the glamour of battles than in their long-term inferences. Agricola now marched on to the Moray Firth and a rendezvous with his fleet. The Boresti are thought to have given their name to Forres, near Elgin.

For the victors the night was bright with celebration and looting. For the vanquished the weeping of men and women mingled as they dragged off the wounded and searched for survivors, abandoning their homes and even setting them alight in their rage. The morning revealed the aftermath of victory: everywhere the dismal silence, the hilly solitude, the smoking ruins. Since our scouts met no one it was apparent that the enemy was not regrouping. Accordingly, since the summer was too far advanced for the war to be enlarged, he marched the army down into Borestian territory. There he gave orders to his naval commander to circumnavigate Britain.¹⁰⁹

These orders doubtless included probing for a Highland back door, which exists in the shape of Loch Linnhe. The army’s obvious next step, the Great Glen, would require new thinking and a fresh start. Little could now be done until the April of 85. The battle had been fought at too late an hour on a day too late in the year.

The same spring, which had seen Agricola entering north-eastern Scotland, also saw Domitian’s army crossing the Rhine at Mainz. The emperor led in person. His nominal aim was to chastise the Chattans, a troublesome tribe. Perhaps he also intended to enlarge the Roman hold on Germany’s south-west corner. Without venturing deeply into barbarian territory, it was tempting to seek a short cut between the Rhine and Danube frontiers, reducing the long re-entrant via Basle. The improvement would in due course be made and Domitian’s campaign was a significant step toward it. This was, however, incidental to the emperor’s purpose. Domitian knew that fame would not be won by road improvements but by a victory, like that of his brother at Jerusalem (strikingly shown on the sculptural panels of the Arch of Titus, where they may still be seen). Such éclat could now be matched by a rousing success in Germany.

However, the Chattan expedition would merit no arch. The enemy simply retreated into his forest. Ground was gained, but memories of Quintilius Varus forbade a penetration of more than fifty miles beyond the Rhine. So, after two or three seasons of exhausting and frustrating effort, there were no notable battles, no striking successes and few captives or trophies to parade through the Roman streets. Domitian then dashed to the lower Danube, perhaps grateful for a diversion from his German stalemate and a second chance of glory. The ambitious and malicious Decebal, king of Dacia (today’s Romania) had used the imperial preoccupation to descend from the Carpathians, cross the river and savage Rome’s Balkan provinces; an attack which cost the life of Moesia’s governor, Oppius Sabinus. Domitian summoned the Praetorian Guard from its comfortable quarters on the Tiber to meet him on the Danube. An expedition into Dacia was hastily mounted, doubtless backed by inadequate staff work, in which the praetorian prefect, Cornelius Fuscus, was routed and killed, prompting Juvenal’s sneer:

Fuscus, who did his soldiering in banqueting halls,

Little dreamed himself a banquet for Dacia’s crows.¹¹⁰

Dio would be scathing about Domitian’s role in this campaign:

He took no part in the fighting, remaining in one of the Moesian cities and living it up, as usual; for he was lazy as well as cowardly; and dissolute toward women and boys besides. So he sent others to fight his battles, claiming any successes for himself and blaming someone else for the reverses. He even celebrated a Triumph, as if he had won a victory; in which, instead of captured spoils, he paraded ‘props’ drawn from the government furniture store!¹¹¹

A second attempt led to a Roman victory at Tapae, only thirty-five miles short of the Dacian capital. Success at last. But news of a mutiny on the Rhine eclipsed Domitian’s moment of glory and sent him scurrying north. This was the rebellion of Antoninus Saturninus, who blackmailed the legionaries of the Mainz garrison into supporting him by sequestering their life savings, deposited in the regimental strongroom. However, the rest of the Rhine army stayed loyal and the rising collapsed. For a second time Domitian marched against the Chattans, who had dared support Saturninus. At this unfavourable juncture the Danube again erupted with the Marcomanns or south Germans, probably prompted by Decebal, attacking from today’s Slovakia.

Domitian had now reigned nine years in an accelerating nightmare of reverses, rushing between Rhine and Danube in a manner prophetic of the later empire. Rich in promise, his wars had been poor in results, with no prizes to assuage his insecurity or soothe the envy which corroded his spirit. He had backed the wrong horse. He could, like Claudius, have chosen to visit Britain and reap the triumph of Agricola’s 600-mile advance, crowned by victory on the ancient world’s northernmost field. Instead the setbacks tipped him into darkness, creating, during his last six years, a time when Rome spoke in whispers, senators chose to potter on their estates and writers pretended their muse had forsaken them.¹¹² ‘He was suspicious of all mankind’ (commented Dio, from the safety of a later age).¹¹³ Dio also described¹¹⁴ how he invited senators and knights to banquets conducted in near-darkness, with the guests’ places at table written on imitation gravestones, served by black-painted slaves, with black tableware and the accoutrements of a funeral feast; the host droning on about topics related to death while his guests reclined in shivering silence. Informers were ubiquitous, show-trials frequent, murders and enforced suicides a daily event. Domitian was even said to have executed a man because he had a map of the world painted on his bedroom wall and could be accused of ‘dreaming dangerously’!¹¹⁵

What of Agricola, commander of the only successful front who, in six years, had doubled the area of Britain under Roman control? His successes had coincided with Domitian’s early disappointments, fortunately not quite as dangerous a time as these later years. Here was a dutiful but ingenuous soldier, nearing the end of a long and hazardous mission during which every congenial voice had fallen silent and every friendly door had closed.

These achievements, though played down in Agricola’s despatches, were received by Domitian with a mixture of pretended pleasure and disguised disquiet. He knew that his own recent Triumph over the Germans had been a fraud and the subject of ridicule. In place of captives he had purchased slaves who could be kitted out to look like prisoners of war. By contrast here was the genuine article: thousands of casualties inflicted and a decisive victory for all to see. The very thing the emperor most dreaded: that a private citizen’s name should outshine his own!¹¹⁶

It was time for Agricola to return, entering Italy on tiptoe lest his coming should be greeted with a warmth Domitian himself had been unable to inspire. Then, ‘so that his entry into the city might not excite attention, he avoided all friends, slipping into Rome by night and by night visiting the palace, as required. There, after a perfunctory peck and not a word spoken he quickly melted into the crowd of creeps round the throne.’¹¹⁷

Willy-nilly our general was a national hero. Triumphal ornaments were awarded and the regulation statue put in hand. Spite bided its time. Though only forty-five, Agricola would never work again. Prudently he withdrew to Fréjus and obscurity, dying there at fifty-four (three years before Domitian), probably poisoned by an imperial courier. ‘For the remainder of his life Agricola lived not only in disgrace but in actual need, just because the things he had accomplished were too great for a general. That is why Domitian finally had him murdered, despite giving him the triumphal ornaments.’¹¹⁸

Tacitus leaves his hero with a phrase famous for its bitterness: perdomita Britannia et statim omissa¹¹⁹ (Britain, no sooner grasped than let slip). His meaning is that Mons Graupius clinched the conquest of Britain and the jealousy of Domitian threw it away. But the majority of the enemy had escaped from that battle and unpropitious country lay ahead. History does not judge Mons Graupius a Waterloo and questions whether Agricola’s depleted force could have won one. On the other hand, there is general truth in Tacitus’ judgement, for it had required a painstaking pyramid of effort to put an army into northern Scotland and it would not easily be constructed again. Soon it would begin to topple and the will to total conquest pass, perhaps for ever.

Agricola had made one of Rome’s longest advances, extended knowledge, established sixty forts and built some 1,300 miles of road.¹²⁰ This may be compared with the 18th-century Highland road-building of Generals Wade and Caulfield: 860 miles in something over a million man/days. Agricola’s effort, nearly twice that mileage in a quarter of the time, has been estimated as 900,000 man/days,¹²¹ though allowance must be made for more modest specification.

All would be wasted. Agricola’s fate was not to be remembered in a Romanized north Britain, but in the dim earthworks of marching armies and on Tacitus’ bright page. At Richborough, Kent, marble splinters and cruciform foundations still recall the tetrapylon, or cross-arch, at Rome’s principal port of entry, thought to commemorate Agricola’s completion of a conquest begun from that spot by Claudius forty years earlier. But if there were a Richborough Monument its salute was unwarranted. Before long the frontier would be back on the Tyne-Solway line where Petilius Cerialis had left it a generation earlier.

The crumbling of Roman Scotland was not, however, instant. Tacitus’ adverb statim (straight away) is denied by archaeology, which shows Agricola’s successor guarding his gains with forts, though none is yet certain north of Stracathro.¹²² Agricola was recalled in 84 or 85. In about 87 came the abandonment of Perthshire and a shift in the army’s centre of gravity to southern Scotland. Around 100–105, in Trajan’s reign, this too would be abandoned. The step-by-step shrinkage of Roman North Britain echoed a larger shift in the continental fulcrum from Rhine to Danube. The fate of what we know as Scotland will be decided in the country now called Romania. Britain was a side-show, opened because Claudius needed a success and reopened because of Vespasian’s emotional attachment. Though Tacitus argues, ad hominem, of a constructive Agricola and a destructive Domitian, the reduction of the British legions may have been less a matter of Domitianic spite than of Rome’s weakness. She had declined from the strike-where-she-liked situation of the late Republic to one of rob-Peter-to-pay-Paul. The army was scattered around the frontiers and the emperors were too fearful of a military conspiracy to risk the establishment of a central pool or strategic reserve. To attack on one front meant borrowing from others. Domitian saw Scotland, as Hitler did North Africa, in terms of priorities. Indeed resemblances between Agricola and Rommel are too numerous to ignore: the steadfast soldier, the demonic master, the distant war-theatre, the stretched supply lines, the breathtaking advance, the surge of popularity at home, the prize almost within reach; then starvation of resources, frustrations of the start-stop kind; and finally recall, muted praise and death under suspicious circumstances. Both were perhaps happy to leave a darkening stage.¹²³

Compared with the inner provinces, Roman Britain has little to show in the way of prestigious buildings, impressive urban sites or engineering marvels. Nevertheless, Agricola’s mark on the northern landscape is part of a larger legacy of military remains – especially in relation to active campaigning – of which we should learn to be proud. Most remarkable is the number of temporary camps. Four hundred have so far been identified; three quarters of them from the air since 1950. They vary in size from one to 165 acres and include winter camps, siege camps, construction camps, as well as some fifty ‘practice camps’, built by soldiers in training. Most, however, were marching camps. Britain is fortunate that warfare and military occupation were largely in remote or upland regions, little disturbed by later ages. This wealth in Wales, England’s North and Scotland is in strong contrast to continental Europe, where no more than a handful of upstanding camps survive and relatively few are known, even from the air. The south of England is also poor. In Wessex, where Vespasian is said to have fought his thirty battles, few have been found. Heavy and prolonged ploughing is usually given as the reason, but we may ask why other faint markings, like Iron Age fields, are discernable; and why there is so little imprint on the chalk downs, where ploughing was slight and Vespasian active. We can only assume him to have been lax regarding the drill of nightly camp-making; or unusually thorough in their destruction. More probably he bivouacked within friendlyoppida and captured forts. Perhaps we should not be asking why lowland camps are few but why upland are many. Here we must not discount the obvious: the mood of moorland Britain, then largely cloaked in forest and even more sombre than now; to say nothing of the truculence of the remoter tribes.

In Rome of the mid 90s the boil of terror ached for the lancet. The empress Domitia, believing her own arrest imminent, finally found courage to do what all Rome had vainly hoped of her father, Domitius Corbulo. He it was who had conquered Armenia, only to die on the whim of Nero. Now his daughter would give tyranny their joint reply. At 5 a.m. on 18 September 96, despite his palace walls being clad in mirrors and the dagger beneath his pillow, Domitian was attacked in his sleep and succumbed to eight stab wounds, inflicted by a slave acting for the empress and her co-conspirators, who included the praetorian prefect. As the news broke that morning senators hurried to the Chamber, jostling to make jeering speeches and to push through a motion damnatio memoriae¹²⁴ (in condemnation of his memory). This execration carried with it the annulment of the late emperor’s laws, suppression of his titles, removal of his portraits or emblems, and erasure of his name from every inscription in the empire. The Younger Pliny described the destruction of his golden statues:

The pleasure in being present as those proudest of faces bit the dust, of hacking and chopping them with sword and axe, as if each blow were piling on the pain. All got a kick from seeing these likenesses mutilated and dismembered; that hateful, fearful face cast into the furnace and the thought that from this melting down of menace and terror something useful and enjoyable might be made.¹²⁵

Like the 1991 toppling of the statue of Felix Dzerzhinski, founder of the KGB, it was feeble recompense for so much suffered so meekly for so long.

Thus ended Domitian in the sixteenth year of his reign at the age of forty-four, whose father was the genial Vespasian and whose brother had been the darling of the Roman crowd. His memory was treasured solely by the soldiers, demonstrating the army’s disinterest in liberty and devotion to dynasties. Of these there had been two since Augustus established the Principate: the Julio-Claudian and the Flavian, the latter ending on Domitian’s death without issue.

For a final comment on the loss of Scotland one may perhaps defer to the authority of Edward Gibbon:

The masters of the fairest and most wealthy climate of the globe turned with contempt from gloomy hills assailed by the winter tempest, from lakes concealed in a blue mist and from cold and lonely heaths, over which the deer of the forest were chased by a troop of naked barbarians.¹²⁶