Fortress of Louisbourg

In a very rare display of joint effort the British North Americans managed to get this establishment into their possession in 1745.

Plan of Louisbourg, published at the conclusion of the French & Indian War, from Bellin’s Petit Atlas Maritime. The map shows major fortifications and includes a key locating 14 important points of interest.

Louisbourg was originally settled in 1713, and initially called Havre à l’Anglois. Subsequently, the fishing port grew to become a major commercial port and a strongly defended fortress. The fortifications eventually surrounded the town. The walls were constructed mainly between 1720 and 1740.

By the mid-1740s Louisbourg, named for Louis XIV of France, was one of the most extensive (and expensive) European fortifications constructed in North America. It was supported by two smaller garrisons on Île Royale located at present-day St. Peter’s and Englishtown. The Fortress of Louisbourg suffered key weaknesses, since it was erected on low-lying ground commanded by nearby hills and its design was directed mainly toward sea-based assaults, leaving the land-facing defences relatively weak. A third weakness was that it was a long way from France or Quebec, from which reinforcements might be sent.

Louisbourg was first captured by New England based British colonists in 1745, and was a major bargaining chip in the negotiations leading to the 1748 treaty ending the War of the Austrian Succession. It was returned to the French in exchange for border towns in what is today Belgium. It was captured again in 1758 by British forces in the Seven Years’ War, after which its fortifications were systematically destroyed by British engineers.

Even counting in the St Lawrence settlements, the people of French Canada amounted to seventy thousand or less at the time of the Seven Years War, which put them at a numerical disadvantage of something like twenty to one compared with the British Americans to the south. Until almost the very end, however, the Canadians maintained a clear superiority in mobility and military prowess over the British – seemingly incredible assets which they owed to a greater centralisation of control (despite notorious corruption in high places), their skill at managing the canoe and the musket, the facility of water transport, and their generally good relations with the Indians.

In contrast, the open but far more thickly-settled British colonies of the eastern seaboard grew at the slow pace of self-sufficient agricultural communities. They were boxed into the north by the nations of the Iroquois confederation and their French associates, and to the west by the Appalachians. There was little sign of common purpose among the British colonies. Indeed, out of all the expeditions mounted by the British in the earlier wars the only ones which bore lasting fruits were the enterprises which wrested New Amsterdam (New York) from the Dutch in 1664, and gained Port Royal (Annapolis Royal) and the mastery of Acadia in 1710. Louis XIV had to renounce Acadia (a lightly settled coastal province) and the great island of Newfoundland at the Peace of Utrecht in 1713.

Building a Fortress

The French appreciated that they would have to take fresh measures to safeguard the seaward approaches to the St Lawrence. Already in 1706 an anonymous memorandum had urged the government to set up a fortified colony on lIe Royale (Cape Breton Island), which formed the southern shore of the entrance to the Gulf of St Lawrence:

The proposed establishment will concentrate all the fisheries in the hands of the French and deny them to the English altogether; it will defend the colonies of Canada, Newfoundland and Acadia against all the enterprises of the English … and ruin their colony at Boston by excluding them from this great tract of land; it will give refuge to our crippled vessels … it will promote Canadian trade and facilitate the export of its grain and other produce; it will furnish the royal arsenals with masts, yards, timbers and planks. (McLennan, 1957, 30-1)

The Conseil Royal decided in favour of the thing in 1715. The first small band of settlers came ashore in the following year, and the chief engineer of Canada, Jean-Francois du Verger de Verville began a series of lengthy reconnaissances. In 1721 work finally began on the new fortress of Louisbourg. The chosen site was on the east coast of the island at Havre al’Anglais, a roadstead capable of sheltering an entire French fleet, which might then bottle up any British ships that sailed into the St Lawrence. Verville planted the town on a peninsula, and closed off the neck by a perimeter of two full bastions, three curtains, and two half bastions – one on each of the seaward flanks. A highly original feature of the design was the way the full bastion to the right, looking from the town (Bastion du Roi) was formed into a miniature citadel with gorge wall, barracks, governor’s lodging and chapel. The one factor which the French left out of their calculations was the absence ofanything which could be termed a ‘building season’. The fog and rain prevented the mortar from drying out during the summer, and the imprisoned water froze every winter, with devastating results to the masonry. Verville disliked the Canadian climate intensely, and the Canadians still more, and he spent every winter in the comfort of France. Thus Louisbourg absorbed immense sums of money, without ever being in good repair, and Louis XV complained that he almost expected to see the ramparts of this costly ‘Dunkirk of America’ rising above the horizon of France.

In a very rare display of joint effort the British North Americans managed to get this establishment into their possession in 1745. The canny and popular merchant William Pepperell gathered 4,000 troops from the New England colonies (which was a considerable achievement in its own right) and sailed to Cape Breton Island in the company of Commodore Warren and 1,000 marines. The many seamen and backwoodsmen proved to be an immense Lhelp in building the siege batteries, though somebody complained that the force was ‘in great want of good gunners that have a disposition to be sober in the daytime’ (ibid., 152). There were no engineers with the expedition at all (until two officers arrived from Annapolis on 5 June), and the French were perplexed by the very irregularity and unpredictability of the conduct of the siege. Louisbourg fell on 17 June after six weeks of attack.

The new governor, Commodore Charles Knowles, had no very high opinion of any kind of fortress as a prize: ‘Neither the coast of Acadia nor any of the harbours in Newfoundland (except St Johns and Placentia) are fortified, and these but triflingly, and yet we always be masters of the cod fisheries for that year whether there be a Louisbourg or not’ (ibid., 175). Indeed, the British government was not disinclined to listen to the instances of the French, who at the peace conference at Aix in 1748 were determined to regain Louisbourg at almost any price. The Comte de Maurepas, the minister of marine, viewed the place as the guardian of both New France and the Grand Banks fisheries, which latter were of great economic importance and a nursery of seamen. Out of these considerations the French sacrificed Madras in far-off India and the brilliant conquests of de Saxe in the Netherlands.

The British accordingly gave up Louisbourg. They partially made up for the loss in 1749 when they built four forts and a barricade at Halifax on the adjacent peninsula of Nova Scotia (Acadia). Within three years Halifax had a population of four thousand, and the potential to become one of the most important avenues of entry for British power to North America.


Fort Douaumont is modernized

The task of modernizing the Verdun forts began in 1887 with Fort Douaumont, whose prime position protected the approaches to the city from the north and east and, in particular, from Metz. It was a gigantic task. The earth covering was first removed and the masonry was strengthened with pillars and concrete supports before being covered by a buffer of sand approximately one metre thick. Then, using for the first time a ‘continuous pour’ process, a thick layer of special concrete was poured on top of the sand. The eastern side of the barracks, some of the artillery bunkers and the tunnel into the barracks from the main entrance were covered with one and a half metres of concrete. The western side of the barracks and the remaining artillery bunkers, which were intended to form a strong ‘keep’ or redoubt for last ditch defence, received a covering two and a half metres thick. After completion, the whole of the concrete carapace was covered with a layer of earth between one and four metres deep. In 1888 a thick layer of special concrete was also applied to the open southern façade of the barracks. All in all, between April 1887 and November 1888 the work of strengthening Fort Douaumont required approximately 28,000 cubic metres of concrete and a team of some sixty construction workers.

To protect Fort Douaumont further, earth was banked up around the lower floor of the barrack block, covering it completely and effectively burying half the fort. The upper floor was now at ground level. To make the ditch less vulnerable to bombardment, the scarp wall was replaced by a sloping earth bank. To defend the ditch, strong, concrete galleries were constructed in the counterscarp, facing the fort itself. Armed with revolver guns and light cannon and later with searchlights, these galleries – single at the northern corners but double at the apex of the fort – were designed to sweep with enfilading fire any enemy who managed to penetrate into the ditch. Connected to the barracks by long underground tunnels, the counterscarp galleries could be reinforced regardless of enemy fire. The original gateway was scrapped and a new entrance – an independent blockhouse protected by double flanking galleries and a drawbridge – was constructed in the gorge (south) side of the fort. From the blockhouse a tunnel ran under the rampart to an entrance on the lower floor of the fort.

The armament of the fort

The original plan for the armament of Fort Douaumont had provided for twenty guns mounted on the parapet but the revolution in high explosives and artillery in the 1880s meant that henceforward guns had to be protected if they were to remain operational at all times. As a first step, ten of the guns were dispersed in batteries outside the fort but the introduction of steel-reinforced concrete in 1897 made it possible to construct shell proof gun positions in the fort itself.

The first protected guns at Fort Douaumont were installed in 1902-1903 in a new type of strong concrete bunker known, from the experimental range on which it was first tested, as a Bourges Casemate (Casemate de Bourges). Embedded in the southwest corner of the superstructure and shielded from direct fire by a long wall forming a protective wing, this bunker was strengthened with a layer of concrete almost two metres thick. It was armed with two quick-firing 75mm field guns installed in two chambers placed in echelon, whose embrasures allowed for fire in one direction only. The fixed guns, which had a range of 5,500 metres, were sited so as to cover the southwestern approach to the fort and to cover with flanking fire the defensive works situated along the ridge between Fort Douaumont and the Ouvrage de Froideterre. An observation post and magazines completed the installation.

The construction of the Bourges Casemate marked the beginning of the period that turned Fort Douaumont into a modern, armoured fort of enormous strength. Between 1902 and 1913, further armament was provided in the form of guns housed in retractable steel turrets of very advanced design which, by rotating through 360°, covered all the approaches to the fort. The turrets were activated by a vertical movement that raised them into the firing position and lowered them again once the gun had ceased firing. Raising the turret exposed the gun embrasures and allowed the guns inside to fire. When retracted, the guns were hidden from view and entirely protected by a steel dome which, in the case of the bigger turrets, was strong enough to withstand even direct hits by the heaviest shells. The turret was set in a reinforced concrete unit that also housed the activating machinery, magazine, replacement guns and range-finding equipment. Each one was coupled with an observation post protected by a dome of steel twenty-five centimetres thick and connected to the gun by speaking tube or telephone.

Four such turrets were installed in Fort Douaumont and linked to the barracks by underground tunnels. Two lighter models housed twin eight millimetre Hotchkiss machine guns that were mounted one above the other and fired alternately to avoid overheating. Intended for the close defence of the fort, the machine guns were installed at the northeast and northwest corners of the superstructure where the visibility was best. On the eastern side of the fort a short-barrelled 155mm gun capable of firing three rounds a minute over a range of 7,500 metres protected the vital north and northeast fronts. An armoured observation post at the entrance to the turret communicated with it through a speaking tube while another some distance away communicated with the gun crew by telephone. Twin short-barrelled 75mm guns housed in a similar turret in the escarpment to the north completed the armament. The 75mm guns, which together were capable of firing more than twenty two rounds a minute over a range of 5,500 metres, were intended to sweep the intervals between the forts. Their considerable firepower more than compensated for the dispersion of the remainder of the fort’s artillery in external batteries. By 1913, all the gun turrets, observation posts and the Bourges Casemate had been linked to the barrack block by strong underground passages so that they could be accessed at all times without going outside.

Mighty though it was, Fort Douaumont formed only one element in a strong and extensive ‘centre of resistance’ which also included Douaumont village, six ouvrages, five combat shelters, six concrete batteries, an underground shelter for reserves, two ammunition depots and a whole series of concrete infantry entrenchments. A revolving turret for twin 155s on the slopes to the south of the fort and another for a 75mm gun on the ridge to the east should have completed the defence but neither emplacement was complete when war broke out. A machine gun was mounted in the observation post on the ridge, while the incomplete 75mm gun turret became a shelter.

Access to the fort

Fort Douaumont was an immense structure, measuring almost 300 metres from north to south and 400 metres from east to west. It was protected on all sides by an open glacis offering wide fields of fire in every direction and surrounded by a belt of wire thirty metres deep, which was attached to metal picket posts set in concrete. At the top of the glacis, a line of stout spiked railings two and a half metres high ran along the counterscarp. On the floor of the ditch – approximately six metres below the top of the counterscarp – a further line of railings was set at an angle along the base of the scarp. The outer wall of the ditch was strengthened by a facing of masonry on three sides of the fort, but on the south side, where the inner wall was strengthened and provided with flanking blockhouses, it consisted of only a bank of earth.

The fort was accessed by means of a wagon road that led up the glacis on the south side of the fort, passed a guard house and came down in the ditch close to the main entrance or ‘peacetime gate’. The road then ran across a drawbridge and entered a tunnel under the rampart. This led to the ‘wartime gate’, which allowed direct entry to the lower floor of the barracks. At the end of the tunnel, two ramps provided access to the upper floor of the barracks and to the covered wagon roads that passed through the barrack block at each end. Emerging from the barracks on the north side, the wagon roads became the Rue du Rempart, which served the adjoining artillery shelters and ammunition depots. Access to the fort on foot was also possible by means of steps cut into the rampart, which led to a footbridge spanning the gap between the tunnel and the wartime gate. On the top of the rampart, two light steel domes allowed for observation over the south side of the fort

The barrack block was a two storey building. The lower floor, known as the ‘wartime barracks’, comprised the fort’s administrative services, siege headquarters for the commandant and his staff, depots, magazines and two groups of cisterns each holding 520 cubic metres of water. The upper floor, or ‘peacetime barracks’, provided accommodation for the garrison of 850, workshops, magazines and kitchens as well as a bakery and an infirmary. Staircases or metal ladders linked the two floors and on each level the barrack rooms opened onto a principal corridor. On the upper floor the corridor ran throughout the whole length of the barracks and joined the covered wagon roads at either end of the building.

One year before the outbreak of the war, Fort Douaumont was complete. The strongest and most modern of all the forts around Verdun, it was the cornerstone of the whole defensive system. Its construction, modernization and armament had required a total of twenty eight years and had cost 6,100,000 gold francs.


As early as September 1914, Fort Douaumont’s 155mm gun was in action against German positions to the north of the sector. The Germans soon replied with a barrage of medium and heavy calibre shells that caused some slight damage but left the fort’s vital organs unscathed. The operation was observed by a prominent guest, Crown Prince Wilhelm of Germany, who had been invited by the commander of V Reserve Corps, General Erich von Gündell, to view the shooting from his newly built observation tower, the ‘Gündell-Turm’. The French were not impressed and the tower soon became a favourite target of French gunners.

In December 1914 the 155mm gun was again in action, this time against the Jumelles d’Ornes, two hills that formed an important German observation post to the northeast of the sector. That brought retaliation from the Germans in February 1915 in the form of a ‘shooting match’ (Wettschiessen) between two of their biggest guns – a 420mm Krupp mortar and a 380mm ‘Long Max’ naval gun – during which thirty four huge projectiles were hurled against the fort and its immediate surroundings. Despite a great column of smoke which rose above the glacis and at first led the Germans to believe that the fort had been put out of action, only limited damage was done. On the eastern side of the barracks where the concrete carapace was only one and a half metres thick, three shells falling close together brought down the roof of the bakery and a nearby corridor, while the blast from a fourth fissured the floor and walls of the gallery leading to the 75mm turret. The guns, however, were unharmed. One 420mm shell striking the reinforced concrete collar of the 155mm turret left a deep hole but only slightly affected the turret mechanism and repairs were carried out within a day or two. Another fell without exploding close to the Rue du Rempart, where it was defused and sent to Paris for exhibition.

The fact that such a shell had hit the concrete covering of the fort and failed to explode probably encouraged the French High Command in its comfortable belief that the most powerful of the forts around Verdun was impregnable. Indeed, had the whole mighty system not been disarmed in the second year of the war, that belief might well have proved correct.

The World beyond Rome I

The Roman Empire was involved in networks of trade, diplomacy, and influence that, at their greatest extent, spanned Europe, Africa, and Asia. In the north, a Roman glass cup was buried in a fourth-century grave mound in Føre, Norway, above the Arctic Circle. In the east, a Roman glass bowl was buried in a fifth-century tomb in the Nara Prefecture in Japan. In the south, four Roman beads made of glass, silver, and gold were deposited in a third-century context at a trading site at Mkukutu in Tanzania. While these finds trace the outer edges of the reach of Roman trade goods, these regions were too far from the empire to play much role in frontier society. It is doubtful whether the nobles and merchants of Norway, Japan, and Tanzania who received these objects had any conception of the Roman Empire or knew where the luxury goods in their possession had been made.

Some knowledge of Rome reached China, where the Roman Empire was called “Great Qin.” Chinese sources reflect some eclectic but not inaccurate knowledge of Roman geography, government, and law. Romans had a similarly vague knowledge of the Chinese, whom they called “Seres,” being aware that their land was the source of silk and lay to the east beyond Parthia and India, but contacts were neither direct nor regular enough to leave much trace on the frontiers. The peoples, networks, and power centers that had a stronger impact on the frontier were found closer to the territory that the Romans had claimed as their own.

In North Africa, Roman administration covered the coastal agricultural regions, but in the broad zone of marginal lands between the coast and the Sahara desert there were numerous peoples, known to the Romans by such names as Mauri, Gaetuli, and Garamantes, who lived partly in and partly beyond the frontier region. Some of these peoples were dry-zone farmers who managed large-scale irrigation works. Others lived as nomadic pastoralists. There has been a long debate in the scholarship whether the settled and nomadic peoples of Rome’s desert frontiers, in Africa and elsewhere, lived in a state of cooperation or competition; the answer may well be both, depending on local circumstances and the fortunes of their farms and herds.

South of Egypt, on the middle reaches of the Nile, was the kingdom of Kush. In the aftermath of Octavian’s victory over Antony and Cleopatra and the incorporation of Egypt into the empire, Roman and Kushite forces clashed over control of the borderlands. After brief hostilities, Queen Amanirenas of Kush sent ambassadors to make a treaty with Augustus, and the peace held for most of the next few centuries. Occasional diplomatic missions helped keep the peace. One of these, likely from the third century, appears to be documented by a Latin inscription at Musawwarat es-Sufra in which one Acutus from Rome formally presents his good wishes to an unnamed queen. Evidence for the study of Greek in Kush may represent local officials keeping up the necessary language skills to send their own ambassadors in return. Kush also participated in the trade routes that connected the Mediterranean with the Indian Ocean and central Africa. Concern for the security of trade may have encouraged both states to keep relations stable.

The Arabian frontier, like North Africa, presented a mix of settled kingdoms and nomadic peoples. The trade routes that passed through the region brought in substantial wealth but also further complicated the relationships between these societies. The Nabataean kingdom was a Roman client state for the better part of two centuries. Its capital at Petra was adorned with rock-cut temples in ornate Hellenistic style, and its kings were important regional leaders. Trajan annexed the territory in 107 as the province of Arabia Petraea, or “Rocky Arabia.” Other kingdoms and tribal alliances competed for power and control of trade routes, sometimes allying with Rome and sometimes raiding the frontier.

The largest and most powerful of Rome’s neighbors was the Parthian Empire. The Parthian state, though a match for Rome in its ability to muster forces for campaigning, was decentralized, prone to divisive court intrigue, and contained numerous semi-autonomous subkingdoms. The administration of this unruly empire was as unwieldy a task as the administration of the Roman Empire with its restless provincials and ambitious generals. It is no wonder that, in the first century CE, the two empires mostly contrived to leave one another alone. Nevertheless, Parthia loomed large in the Roman imagination. It remained the big prize, the enemy against whom flattering writers and propagandistic artists could always imagine emperors leading the good fight. Rome was equally significant to Parthian policy. The Parthian kings positioned themselves as heirs to the Achaemenid dynasty and champions of the Iranian peoples against western aggression.

The period of relative stability was broken by Trajan, who invaded Mesopotamia and Armenia in 113. Although Trajan’s conquests were quickly reversed by his successor Hadrian, Roman-Parthian relations remained unsettled for the following century. Several emperors initiated or contemplated military action against Parthia, and several Parthian kings pursued more aggressive policies on their western frontier. No substantial changes to the border were lasting, however, and diplomatic relations continued in between bursts of conflict. The historian Herodian even reports that the emperor Caracalla, in the early third century, proposed marrying a Parthian princess, and that Caracalla’s successor, Macrinus, celebrated a peace treaty and hailed the Parthian king Artabanus V as a loyal friend.

On the Black Sea steppes, a variety of nomadic and seminomadic peoples continued to live in traditional ways while some peoples of the region also developed settled kingdoms. Romans tended to describe the region in vague terms that drew as much on the literary tradition going back to Herodotus’ Scythians as they did on contemporary knowledge, but we should not assume that life on the steppe was static. Literary sources name various peoples in this region, including Sarmatians and Alans. In some cases, these names seem to correspond to identifiable ethnic and political groups, but they can also be unreliable, as the complexities of steppe identities were sometimes lost on writers from sedentary cultures.

In the late second century, there is evidence of cultural changes around the northern shores of the Black Sea and the lower Danube that may reflect the arrival of migrating warrior bands from somewhere to the north and west. These new peoples are reflected in a distinct archaeological pattern of settlement types, pottery styles, and burial practices. These features are the earliest evidence for a cultural pattern that would become more pronounced in the third and fourth centuries CE, which modern archaeologists have termed the Chernyakhov culture. It is generally believed that the Chernyakhov culture is related to the people known as “Goths” in the literary sources, but how consistent the Chernyakhov-Goth connection is and how early we can speak of a Gothic presence in the region are matters of debate.

The Romans referred to the peoples who lived along the middle to upper Danube and Rhine as “Germans” (barring a few exceptions, such as the Dacians and Iazyges), but it is unlikely that the tribes and kingdoms of this region felt any kind of shared identity. Many individual tribal names are also known, but, as elsewhere, we cannot be confident that the Roman authors who recorded those names were applying them accurately. Many cultures existed in this region with different kinds of social and political organization. Some, such as the Dacians and Marcomanni, appear to have reached an early stage of state development, with power centralized in well-established royal families. Other peoples, such as the Frisians, lived in small, egalitarian communities with little in the way of formal power structures.

Farther to the north, away from the frontier zone but in close contact with the Roman world, another major power was rising. At Himlingøje in Denmark, a group of lavish burials filled with Roman luxuries marks the center of a commercial and political network that established itself in the late second century and spanned the Baltic Sea and southern Scandinavia. The warrior nobles of Himlingøje fought as auxiliaries in the Roman army and maintained strong trade and diplomatic connections to Rome after they returned home. Through these connections they acquired Roman goods, which they then used as prestigious gifts to expand their network of influence in the North. The numerous ritual deposits they made in Danish bogs of the weapons and armor of their defeated enemies show that they expanded their power in more aggressive ways as well. While many of the peoples who lived closer to the Roman frontier had unsettled histories with Rome, the rulers of Himlingøje appear to have remained on good terms with the Romans throughout their history.

Rome also had staunch allies in Scotland with the Votadini whose power center, a fortified hilltop site at Traprain Law, has yielded an extraordinary wealth of Roman imports ranging from gold brooches to iron door hinges. The precise boundaries of Votadinian power are uncertain, but other peoples certainly lived beyond the British frontier, both in Scotland and Ireland. Some of these peoples had large, settled societies, but others were small and mobile.

The peoples who lived in and beyond the Roman frontier zone varied widely in their ways of life, social organization, and political structures. While some maintained long-term diplomatic ties with Rome, others had volatile relations with the empire. This wide variety of frontier peoples challenged Rome’s limited capacity for maintaining foreign relations and managing the frontiers.

Emperors and Frontiers

The frontier was always an area of special concern to the emperors, even those with little direct experience of it. Imperial power depended on the support of two groups: the army, which was mostly stationed on the frontiers, and the people of Rome, who approved of victories over barbarians. Although imperial activity on the frontier could be haphazard and inconsistent, few emperors could afford to ignore the frontier entirely.

After the defeat of Varus, Augustus soured on expansion. He initiated no more conquests, and his final advice to his successor Tiberius was to keep the empire within its boundaries. The meaning of this counsel has long been debated. It is unlikely he meant that the empire should never expand again. The conquering ideal remained fixed in Roman ideology, and Augustus was not shy of bragging about the conquests accomplished under his authority. More likely it was personal advice to his successor not to embark on a new series of foreign campaigns for political purposes.

On the whole, most of Augustus’ successors followed his advice. On the grand scale, the frontier was mostly stable. There were only a few large additions to the empire in the following centuries: the southern half of Great Britain, parts of North Africa, Dacia, parts of Arabia, Armenia, and Mesopotamia. The conquests of Mesopotamia and Armenia were brief accomplishments of Trajan’s and Septimius Severus’ wars against Parthia and did not long endure. Some of the expansions in Africa and Arabia came from incorporating client kingdoms rather than conquering new lands. On the small scale, however, the frontier was turbulent. Almost every emperor from Augustus to Severus Alexander fought frontier campaigns or faced unrest in frontier provinces. Most of these campaigns added little, if any, new territory to the empire, but few emperors actually treated the frontier as a limit not to be crossed.

Emperors who felt insecure in their position used foreign wars to prove their worth in the traditional expansionist mode. Claudius, who came to power unexpectedly, initiated the conquest of Britain, which the unloved Nero continued. Domitian, another surprise emperor, began his reign with a campaign in Germany that even his fellow Romans criticized as unwarranted. Trajan, though he grew to be one of the most beloved emperors, came to power through obscure political machinations, which may help explain his ambitious program of conquests in Dacia and Mesopotamia. Septimius Severus, the victor of a civil war, spent much of his reign fighting in Mesopotamia, North Africa, and Scotland. These campaigns not only showered military glory on the emperors but also enriched the empire with plunder and slaves while keeping potentially restless soldiers occupied.

Restless soldiers were no trifle. Revolt by troops who felt ignored by the emperors was a recurrent threat to imperial stability. Sometimes this discontent could be softened by letting the soldiers pillage across the frontier. On other occasions, successful frontier generals could harness their soldiers’ dissatisfaction in a bid for the throne. Vespasian and Septimius Severus both came to power in this way, and many more attempted the feat unsuccessfully or managed it only to be quickly ousted by a rival general.

While the Romans pushed the frontier, the frontier pushed back. There were few major incursions on Roman territory in the first centuries of the empire, but some threats demanded the emperor’s attention. Relations with the Parthian Empire remained unresolved as both empires pressed for greater influence along their mutual border, but neither could secure a lasting victory over the other. Trajan, Severus, and Caracalla all led major campaigns against Parthia, but their gains did not last. The Parthians backed Pescennius Niger, a general in Syria who competed with Severus for power, but Pescennius’ bid for the throne failed.

Away from the Parthian front, the most serious threat to the Roman frontier in this period developed along the Danube in the late second century. Termed the Marcomannic Wars by modern scholarship, this diffuse and protracted series of conflicts involved many of the peoples of the region, chiefly the Marcomanni, Quadi, and Iazyges, and kept the emperor Marcus Aurelius occupied from the early 160s to 180. Smaller-scale troubles rarely claimed the attention of the emperors, but raiding, local resistance, and discontent among the soldiers were constant nuisances in the frontier zone that could flare up into more serious trouble if not kept in check.

Emperors undertook a variety of different policies toward the frontier. In the early empire, rulers such as Augustus and Nero were content to govern from a distance and entrust even major campaigns to subordinates, but the rise of frontier generals as claimants to the throne demonstrated that it was dangerous for an emperor to leave the frontier in anyone else’s hands. There were those, such as Trajan and Severus, who threw themselves into aggressive frontier campaigning. Others, notably Hadrian and Marcus Aurelius, were led, either by temperament or circumstance, to focus on consolidating and defending the territory claimed by their predecessors. Only a few emperors such as Antoninus and Elagabalus largely ignored frontier problems, being either fortunate enough to rule in a period of relative calm or else too busy with their own concerns.

Because of the practicalities of governing a continent-spanning state in an age when messages could take weeks and armies months, if not years, to reach the frontier, an emperor’s ability to effectively manage the frontier was limited. At the same time, as proven by generals such as Vespasian and Severus, delegation of too much power was risky. Wars against barbarians or restless provincials were potent propaganda tools, and emperors were wary of letting anyone else get their hands on them. It was a conventional charge against bad emperors that they did not trust their subordinates, but even the most popular emperors understood the importance of preserving personal control over frontier policy. After the Julio-Claudian age, most emperors learned to keep frontier generals on a short leash.

The World beyond Rome II

Fort at Vindolanda, AD 105. The fort housed the First Tungrian cohort and a Batavian cohort.

The effects of imperial neglect can be seen on the British frontier. The archaeological evidence from Scotland shows a lively cross-frontier exchange in the first and early second centuries. Roman goods found their way into native hands, from fine enameled brooches and sets of bronze tableware to hinges and horseshoes. While the Votadini enjoyed a profitable alliance with the Romans, deposits of mixed Roman and non-Roman scrap metal at several sites indicate that local smiths were also doing jobs for the Roman soldiers stationed on the frontier. Even some modest farmsteads had access to Roman goods. During this period of strong cross-border ties, many emperors devoted at least some of their energies to Britain, and the frontier was briefly advanced into Scotland in the mid-second century. Starting around 160, however, the Marcomannic Wars took imperial attention away from Britain for several decades. Despite some frontier shakeups under Commodus, it was not until 208 that another emperor, Severus, took an active interest in the province. Roman artifacts in Scotland show a corresponding decline after 160. Even casual exchanges, such as Scottish crafters working for frontier soldiers, seem to have dried up. While we might have expected provincial commanders to take up the slack and maintain regional ties when an emperor was busy elsewhere, the Scottish evidence suggests that they did not—or, more to the point, they were not permitted to.

The Roman emperors’ relationship to the frontier was contradictory. They could have enormous effects on frontier societies, whether by leading their soldiers out on campaign or by pulling them back and assigning them to border control. When an emperor turned his attention to a frontier area, it must have been akin to an earthquake or flood: an unpredictable, irresistible event that could change local conditions for generations, but whose aftereffects were mostly left to the locals to deal with. When they turned their attention elsewhere, their subordinates were limited in what they could do to compensate for their neglect. Most of the empire’s frontiers, most of the time, were left to themselves, shaped largely by the actions of the peoples who lived along them.

The Army on the Frontier

The most stable Roman presence on the frontier was the army. While some frontiers were more fully militarized than others, all were marked with fortresses and outposts where Roman soldiers were stationed to maintain security and control. In regions with urbanized societies, such as Egypt and Syria, the army’s influence was mostly limited to the hinterland zones. In other areas, where local societies functioned on a smaller scale, such as Britain and Arabia, the army’s effect on social and economic conditions was more widespread. Across the Roman world, the peoples who lived at the fringes of Roman power mostly knew Rome through its army, whose presence could be both beneficial and disruptive.

Soldiers were usually well paid, since the emperors depended on their loyalty. The regular provision of wages and supplies brought a steady flow of cash and merchants into regions that in many cases had previously been economically underdeveloped. The frontier army was a market for goods and services from both inside and outside the empire. In the West, the pottery and bronze industries of Gaul were stimulated by demand in the frontier regions. The economic effect was less visible in the more developed East, but in outlying regions such as the Egyptian oases, Roman forts provided a new market for local goods. The reach of the frontier market extended well outside the range of Roman authority. Peoples as far away as Himlingøje and Mecca increased their leather and textile production to meet Roman demand.

The Roman army also offered employment to soldiers recruited in and beyond the frontier zone. Barbarian auxiliaries were a vital part of the Roman army for the same reasons that Greek mercenaries had been employed by Egyptians and Persians: economically underdeveloped regions make prime recruiting grounds for troops. After the revolt of Batavian soldiers serving near their homeland in 69 CE, the Roman army began to station auxiliary units away from the regions where they were recruited, so that future rebels would not have the benefit of being surrounded by their own people. Once stationed in their new locations, these units tended to recruit locally and lose their original ethnic character over time, but troops were also relocated from one part of the empire to another as military needs dictated. Because of this reshuffling of personnel, we find, for example, a Pannonian soldier commemorated with a funerary stela at Gordium in central Anatolia and offerings to Syrian gods in the forts of Hadrian’s Wall in northern Britain. Some of these soldiers married local women and started families, creating new communities with ties to both the army and the local peoples. Their sons were often recruited into the Roman army a generation later. Other auxiliary veterans returned home across the frontier and played a role in mediating trade and diplomatic connections between Romans and non-Romans. Recruitment from beyond the frontier fostered the growth of a distinct military society that was neither entirely Roman nor native to the lands in which it developed.

The Roman army could also be disruptive. The militarization of the frontier interfered with traditional trade routes and seasonal movements of laborers and pastoralists. Tacitus noted that unimpeded border crossing was a privilege reserved for few, such as the friendly Hermunduri tribe:

For them alone among the Germans is there trade not only on the [Danube] riverbank but even deep in the most magnificent colony of the province of Raetia. They cross here and there without guards and while to other people we show only our arms and forts, to them we have opened our homes and estates.

The portoria, a customs duty of 25 percent, was collected on all goods entering the empire’s eastern provinces. On other frontiers the rates may have been lower, but there were still fees. The eastern trade routes could be highly profitable: the record of a loan contract from Egypt documents a cargo of perfumes, ivory, fabrics, and other luxuries from India in the second century CE valued at more than 9 million sestertii. (For comparison’s sake, by the late second century, a fortune of 20 million sestertii could put one in the lower echelons of the imperial aristocracy.) High customs fees and valuable cargoes encouraged smuggling. The Romans began to station customs enforcers in client kingdoms beyond the frontier to help monitor the traffic.

Simply knowing what was going on along the frontier was a challenge in itself. Surveillance posts and patrols were obtrusive shows of force, but more subtle forms of spying are hinted at by the historian Ammianus Marcellinus’ mention of the arcani, or “hidden ones”: “Their duty was, by hastening far and near, to keep our generals informed of disturbances among nearby tribes.” A fragmentary tablet from Vindolanda, a Roman fort in northern Britain, with the text miles arcanus (“hidden soldier”) may relate to these same spies, and another Vindolanda text possibly records a scrap of an intelligence report on the locals’ fighting capabilities.

All this surveillance can only have been an aggravation to those who lived along the frontier. Tacitus described a Germanic tribe complaining that the Romans would not allow them to meet with their fellow Germans who lived within the borders, “or else charge us a fee to meet unarmed, practically naked, and under guard, which is even more insulting to men born to arms.” The authority of frontier soldiers to stop, search, and tax travelers was ripe for abuse. A merchant’s letter of complaint found at Vindolanda suggests some of the misconduct soldiers indulged in. The beginning of the letter is damaged, so the details are unclear, but it seems both the merchant and his goods were threatened with violence, perhaps as part of a shakedown:

he beat me further until I would either declare my goods worthless or else pour them away. . . . I beg your mercy not to allow me, an innocent man from abroad, about whose honesty you may inquire, to have been bloodied with rods like a criminal.

The letter further details how the mistreated merchant had appealed up the chain of command as far as the provincial governor with no luck.

If a merchant who could write good Latin and knew how to work the system got so little satisfaction for his grievances, the ordinary people who lived in the outer shadow of Rome’s frontier cannot have fared much better. With no effective recourse against exploitation, peoples of the frontier zone resorted to raiding and revolt, such as the Frisians, who were required to pay a tribute of oxhides to Rome, even though they lived beyond the Rhine. In 28 CE the Roman centurion assigned to oversee the tribe demanded hides of higher quality than the Frisians could supply. When their appeals for relief brought no results, the Frisians revolted, killing more than a thousand Roman troops before they were subdued.

Acting both as agents of imperial power and on their own motivations, Roman soldiers made up one of the main forces at work on frontier society, but Rome was not the only force along the frontier. Many other peoples, cultures, and political forces, both those local to the frontier zone and those farther away, interacted with Rome, pursuing their own agendas and putting their own pressures on those who lived at the edges of Roman power.

Between Rome and a Hard Place

A series of inscriptions from Volubilis in the foothills of the Atlas Mountains on the Atlantic coast of North Africa records eleven occasions over the first and second centuries CE when Roman officials held negotiations with the Baquates, a collection of seminomadic tribes. To judge from the inscriptions, the negotiations seem to have come to a satisfactory end on each occasion. These inscriptions testify to the possibility of peaceful coexistence among those who lived at the fringes of the Roman world, but the fact that these negotiations had to be repeated over and over again also indicates that, in the long term, frontier relations remained unstable.

What was true at Volubilis was true of the frontier as a whole. While a tranquil coexistence was sometimes possible, and large-scale hostilities were relatively rare in the empire’s first two and a half centuries, the frontier was never quite settled. The disquiet of the frontier arose partly from the nature of the societies along it, but also from the way it was caught between worlds. The society of the frontier was constantly being pushed and pulled by many different forces, both Roman and non-Roman. These tensions were felt both inside and outside the demarcated boundaries of Roman control. The conflict between different forces with different agendas destabilized local societies.

Many of the peoples who lived in and around the Roman frontiers are conventionally described as “tribes.” This vague word is applied to various kinds of small-scale societies with no formal government that are held together by networks of extended family ties and personal relationships. Where Roman authors such as Caesar and Tacitus imagined stable ethnic groups with names and defining traits, we should instead see most of the Roman frontier zone inhabited by loose and changeable conglomerations of people who were ready to form, dissolve, and re-form alliances as their interests shifted. Trying to cope with these unstable groups was a challenge for the limited resources of Roman foreign policy. The brutality in many of Rome’s interactions with these peoples only sowed further disruption.

There were other societies at the edges of the Roman world that were larger, more stable, and better able to deal with Rome on an equal footing, including Kush, Parthia, and Himlingøje. For much of the first few centuries of the Roman Empire, these peoples enjoyed relatively peaceful relations with Rome. Their stability and organization made it easier for them to pursue consistent long-term policies toward Rome and to rebuff Roman efforts to meddle in their spheres of influence, but the existence of smaller, less well organized states and peoples in between these major players also helped stabilize relations. Kush had ongoing conflicts with the same desert raiders that harassed the Roman southern frontier. Rome and Parthia managed to keep the peace for more than a century in part because they were able to limit their conflicts mostly to competition over influence in Armenia. Relations in the North were helped because, during the Marcomannic Wars, the rulers of Himlingøje were at war with the same peoples the Romans were fighting.

Caught in between these larger forces, the “tribal” peoples of the frontier did what was necessary to survive. Sometimes they were able to make a profitable peace with Rome and their other powerful neighbors. Sometimes they were pushed into open war. Much of the time, they got by in a state of uneasy cooperation, taking chances to profit from trade or military service when they could get them, indulging in petty raiding and customs evasion when they could get away with it, and suffering the abuses of bored soldiers when they had to.

Good fences may make good neighbors, but what is good for the neighbors is not always good for the fence. Earlier conceptions of the Roman frontier often imagined the peoples just beyond the Roman borders as an outer wall of client states, held in place by Roman diplomacy and intimidation as a bulwark against uncertain threats from the unknown lands of the far distance. When significant new threats to the security of Roman military and political authority arose in the third century, however, they did not come from the far-off reaches of Scandinavia or central Asia but from the frontier zone itself. The peoples that Rome had been bribing, intimidating, patrolling, and generally meddling with for centuries finally began to push back in more effective ways. In the third century, peoples all around the edges of the Roman world—in Scotland, Germany, the Black Sea steppes, Arabia, and North Africa—began to succeed at what Arminius had attempted in the first decade CE: to create large, stable alliances that could stand up to Roman power.

Mesopotamian linear barriers


1. Muriq Tidnim (conjectural)

Babylonian Line 1

2. Nebuchadnezzar’s Babylon to Kish Wall (conjectural)

Babylonian Line 2

3. Habl es-Sakhar (Nebuchadnezzar’s Sippar to Opis Wall, Median Wall)

Line 3 (uncertain)

4. El-Mutabbaq

5. Sadd Nimrud (also called El-Jalu)

6. Umm Raus Wall (site of Macepracta Wall(?), Artaxerxes’ Trench(?))



Mesopotamia and the Rivers Tigris and the Euphrates

Egypt shows that the conjunction of irrigated lands and nomads produced linear barriers – even if the evidence might seem elusive and inconclusive. Therefore, might also then Mesopotamia, with the similarly intensely irrigated Tigris and Euphrates Rivers, produce evidence of walls in the presence of nomads?

In Mesopotamia, the area of irrigated lands runs along the flood plains of the Tigris and the Euphrates up to, and somewhat beyond, the convergence points of the rivers between ancient Babylon and modern Baghdad. Above that point the alluvial plain peters out and the land becomes too hilly to allow for intense irrigation. From the north-east flows the Diyala River which passes through the Zagros Mountains to join the Tigris, linking the high Persian plateau to Mesopotamia. Around the river was especially valued irrigated land. The area of convergence of the Tigris and Euphrates constituted a constricted land corridor. Local nomads and semi-nomads would have been expected to press particularly hard on the rich and productive irrigated lands of Mesopotamia.

As in Egypt, civilisation, sustained by the irrigated lands of Mesopotamia, came early, in the fourth millennia BC, with the Sumerians. Again, as with Egypt, there is evidence of climate change. In the last century of the third millennium BC the stream flow of the Euphrates and the Tigris was very low, according to analysis of sediments in the Persian Gulf. The end of the Akkadian era, due to defeat by the hated Gutian peoples from the mountainous east in the twenty-second century BC, coincided with a few decades of intense drought which was followed by two to three centuries of dry weather. Ur revived and under Ur-Nammu defeated the Gutians and established the third dynasty of Ur, commonly abbreviated to Ur III, in 2112 BC. The Sumerians initiated a short period of cultural renaissance in a time of constant conflict with the semi-nomadic Martu – more familiar as the biblical Amorites.

Indeed, Ur III may have faced two reasonably distinct threats. From the north-west there was the Martu whose aim may in part have been to gain sustenance for their herds in times of drought. The direction of the threat that they posed would have been through the relatively flat lands between and to both sides of the convergence point of the Euphrates and the Tigris. To the north-east were the Elamites and Shimashki confederation in the highlands to the east of the Tigris. Their lines of attack would have been more focused down river valleys – perhaps the Diyala River flowing through the Zagros Mountains to the Tigris.

Mesopotamian linear barriers

In this early period there is only textual evidence for linear barriers, based on letters that remarkably survive from the third dynasty of Ur. These writings between Sumerian kings and their often disobedient generals and officials, are called the Royal Correspondence of Ur (abbreviated to the RCU). Much of the correspondence in the twenty-two or so surviving letters was about defence against the Martu. There was also information about linear barriers in the year names of Sumerian king lists (Mesopotamian kings named each year of their reigns after some major event).

The Sumerian kings Shulgi (2094–2047 BC), Shu-Sin (2037–2029 BC) and Ibbi-Sin (2028–2004 BC) were mentioned in the context of three walls:

bad-mada/Wall of the Land – The Wall of the Land is known only from one reference in the king lists: ‘Year 37: Nanna (the god) and Shulgi the king built the Wall of the land.’10 Shulgi was on the throne for forty-seven years so the wall belongs to the last quarter of his long reign. This was a time of increasing pressure on central and southern Mesopotamia from the Martu.

bad-igi-hur-sag-ga/Wall Facing the Highlands – In the RCU there are several references to bad-igi-hur-sag-ga – both during Shulgi’s reign and that of his successors, Shu-Sin and Ibbi-Sin. The bad-igi-hur-sag-ga has been variously translated as the Wall, Fortress, or the Fortification facing the highlands or mountains – making it uncertain whether this was a continuous linear barrier. If, however, Shulgi really did build a long wall then he has the distinction of being the first known builder of such a barrier. This obstacle possibly faced a threat coming down the Diyala River as it faced the Highlands, presumably the Zagros Mountains.

Muriq Tidnim/Fender off of the Tidnim – There are three references to Muriq Tidnim, or fender (off) of the Tidnim and Shu-Sin. First, the king lists of his fourth regnal year said: ‘Shu-Sin the king of Ur built the amurru (Amorite) wall (called) ‘Muriq Tidnim/holding back the Tidanum’’ Second, there is an inscription in a temple built for the god Shara: ‘For Shara Shu-Sin built the Eshagepada, his (Shara’s) beloved temple, for his (Shu-Sin’s) life when he built the Martu wall Muriq Tidnim (and) turned back the paths of the Martu to their land.’ Third, the most informative reference to the Muriq Tidnim is in a letter from Sharrum-bani, an official of Shu-Sin. ‘You sent me a message ordering me to work on the construction of the great fortification Muriq Tidnim … announcing: “The Martu have invaded the land.” You instructed me to build the fortification, so as to cut off their route; also, that no breaches of the Tigris or the Euphrates should cover the fields with water … from the bank of the Ab-gal watercourse to the province of Zimudar. When I was constructing this fortification to the length of 26 danna, and had reached the area between the two mountain ranges, I was informed of the Martu camping within the mountain ranges because of my building work.’

In this letter, the construction is described as ‘great’. Whatever the uncertainties about Shulgi’s earlier edifices, it is difficult not to interpret this passage as describing a major continuous linear barrier. In the west the Ab-gal canal is associated with an earlier western course of the Euphrates and to the east the province of Zimudar is identified as being on the east side of Tigris in the region of the Diyala river. A danna is about two hours march so 26 dannas may be over 150 kilometres. Therefore, the edifice appeared to extend from the Euphrates to the other side of the Tigris because its length was much greater than the distance between the two rivers. The instructions to build the walls specifically cite stopping the semi-nomadic Martu from overwhelming the fields by a breach between the Tigris and the Euphrates, showing that irrigated land was perceived as particularly vulnerable.

Analysis – Ur III

In the hillier east controlling access down the Diyala river area there may have been a single fortification, the bad-igi-hur-sag-ga or the Wall/Fortress facing the Highlands, first built by Shulgi, which might or might not have been part of another system bad-mada (the Wall of the Land) built in the flatter west. During the reign of Shu-Sin it seems more likely that a linear barrier called Muriq Tidnim was built from new, or it consisted of earlier lines that were linked and much reinforced including Shulgi’s Wall of the land. This is all speculation but there is good if circumstantial literary evidence that Ur III’s strategy for defence against the Martu involved the construction of what would be the first recorded long continuous non-aquatic linear barriers.

There does seem to be a fairly general academic acceptance that under Shulgi and Shu-Sin long walls were built and their purpose was to keep out nomads. For example: ‘Even as early as year 35 of Shulgi, the (nomad) problem was becoming so grave that Shulgi constructed a wall to keep them (pastoral and semi-nomadic Amorites) out, and Shu-Sin built another barrier, called “fender off of Tidnim,” 200 kilometres long, stretching between the Tigris and the Euphrates across the northern edge of the alluvial plain.’ Also: ‘Yet despite Shulgi’s talents, within a few years of his death in 2047 BC his Empire, too, imploded. In the 2030s raiding became such a problem that Ur built a hundred-mile wall to keep the Amorites out.’

Later Mesopotamia

Looking at later Mesopotamia, after the fall of Ur III, how did it defend itself in times of necessity? What emerges is three intense periods of barrier building: firstly, that already discussed, during the short lived Ur III period; secondly, in the neo-Babylonian period associated with Nebuchadnezzar in the sixth century BC; and thirdly, later in the fourth century, aquatic linear barriers were built by the Sasanians. There are also a number of major but little studied walls, discussed below, north of the Tigris and Euphrates convergence point, which are not clearly dated.

After Ur III fell to the Elamites and the Shimaskhi confederation, the so-called Amorite dynasty of Isin completed its breakaway. Given that lower Mesopotamia had fallen to peoples from outside the region there was no reason for a barrier between the north and southern Mesopotamia. Also, Martu or Amorite semi-nomads were becoming increasingly sedentarised. Subsequently, the Babylonians of the era of Hammurabi were able to project their power well to the north of Babylon. The Assyrians, coming from the north, had no need for walls around 700 BC to defend Babylon in this region as they controlled the regions to its north and south.

The neo-Babylonians recovered control of their city in the sixth century BC and made it the capital of the region. The second period of major barrier building materialised in this later Babylonian period, associated with Nebuchadnezzar and textually with Queens Semiramis and Nitocris. Nebuchadnezzar II ruled for forty-three years from 604 to 562 BC. The Medes’ conquest of Lydia made Nebuchadnezzar suspicious of their intentions and this led him to strengthen his northern border. Behind the Medes loomed the Persians. This was clearly seen, rightly as it turned out, as a real, unpredictable threat – and one that prompted the construction of a comprehensive linear barrier system. Notwithstanding this attempt, in 539 BC Cyrus the Great led the Medes and the Persians into Babylonia which was absorbed into the Achaemenid Empire.

Linear barriers – survey

There were three lines of barriers at and above Babylon looked at here, starting in the south and going to the north.

Babylon to Kish – Line 1

Two walls of Nebuchadnezzar (604–562 BC) are known from a clay cylinder, dated to 590 BC when relations between the Babylonians and the Medes had deteriorated. (These compose Line 1 and Line 2 in this and the next section.)

Nebuchadnezzar’s Wall from near Babylon to Kish – This cylinder is inscribed: ‘In the district of Babylon from the chau(s)sée on the Euphrates bank to Kish, 4 2/3 bēru long, I heaped up on the level of the ground an earth-wall and surrounded the City with mighty waters. That no crack should appear in it, I plastered its slope with asphalt and bricks.’ A bēru is the distance which could be travelled in two hours so is variable according to terrain. At five kilometres an hour this barrier would be about 47 kilometres long. The problem is that this is considerably longer than the distance between Babylon and Kish – which is little more than 10 kilometres – unless the barrier followed a particularly circuitous route. Also, it would seem a fairly pointless military exercise building a barrier from Babylon to Kish leaving the flood plain open to the east from Kish to the Tigris. Using up the surplus kilometres would take the wall further east to Kar-Nargal, near an earlier channel of the Tigris, hence blocking the land corridor between the Euphrates and the Tigris. No physical evidence of this wall has been identified.

Opis to Sippar – Line 2

The second line ran between the cities of Sippar, above Babylon on the Euphrates, and Opis on the Tigris, the precise position of which has been lost. A number of walls are associated with this location in texts and there is a surviving wall called Habl-es-Sakhar.

Nebuchadnezzar’s Wall from Sippar to Opis – Nebuchadnezzar’s inscribed cylinder described the second wall as follows: ‘To strengthen the fortification of Babylon, I continued, and from Opis upstream to the middle of Sippar, from Tigris bank to Euphrates bank, 5 bēru, I heaped up a mighty earth-wall and surrounded the city for 20 bēru like the fullness of the sea. That the pressure of the water should not harm the dike, I plastered its slope with asphalt and bricks.’ This Opis to Sippar wall would have been about 50 kilometres long. Both the Babylon to Kish and the Opis to Sippar walls were water-proofed by asphalt so they must have been built in proximity to water – possibly water-courses like canals or in flatlands prone to flooding or swamping.

Wall of Semiramis – The geographer Strabo, citing Eratosthenes, when describing Mesopotamia, said the Tigris, ‘goes to Opis, and to the wall of Semiramis, as it is called.’ Therefore, this wall was in the region of the Tigris and the Euphrates’ convergence point. (Herodotus mentioned Semiramis’ works but did not specify a wall. Rather he described levees which controlled flooding.)

Wall of Nitrocris – Herodotus also described a Babylonian queen called Nitocris – possibly the daughter of Nebuchadnezzar and the mother of the Book of Daniel’s King Belshazzar brought down by Cyrus – whose constructions in Babylon were mainly connected with diverting the Euphrates. Nitocris built works in the entrance of the country (which is clearly a description of a land corridor) against the threat of the Medes. ‘Nitocris … observing the great power and restless enterprise of the Medes, … and expecting to be attacked in her turn, made all possible exertions to increase the defences of her empire.’

Wall of Media – In the Anabasis, Xenophon described how he led the 10,000 Greeks back from Mesopotamia. In it he encountered the Wall of Media twice. Here what is described is the second occasion when Xenophon actually crossed the wall itself following the battle of Cunaxa in 401 BC. ‘They reached the so called Wall of Media and passed within it. It was built of baked bricks, laid in asphalt, and was twenty feet wide and a hundred feet high; its length was said to be twenty parasangs, and it is not far distant from Babylon.’ Assuming that a parasang is the same as a bēru or a danna, that is a two hours march, then the wall was about 100 kilometres long.

Habl-es-Sakhar – There is a surviving wall in the vicinity of Sippar. In 1867 one Captain Bewsher described the ruins of a wall then called Habl-es-Sakhar – which translates from the Arabic as a line of stones or bricks. ‘The ruins of this wall may now be traced for about 10½ miles and are about 6 feet above the level of the soil. It was irregularly built, the longest side running E.S.E. for 5½ miles; it then turns to N.N.E. for another mile and a half. An extensive swamp to the northward has done much towards reducing the wall.… There is a considerable quantity of bitumen scattered about, and it was probably made of bricks set in bitumen. I can see nothing in Xenophon which would show this was not the wall the Greeks passed, for what he says of its length was merely what was told him.’ The description of the ‘baked bricks laid upon bitumen’ is like Nebuchadnezzar’s description of his wall between Opis and Sippar: plastered with asphalt and bricks.

In 1983 a joint team of Belgian and British Archaeological Expeditions to Iraq investigated the ruins of Habl-es-Sakhar. This confirmed that Habl-es-Sakhar was built by Nebuchadnezzar, for bricks marked with his name were found during its excavation. The team reported that Habl-es-Sakhar is the name of ‘a levee 30 metres wide and 1 metre high which could be followed for about 15 kilometres. A trench across the levee to the north of the site of Sippar revealed baked brick walls (largely robbed) on either side of an earth embankment. The earth core was about 3.2m wide and the brick walls about 1.75m in width. Between the brick courses was a skin of bitumen. On the bottom of each brick was a stamp of Nebuchadnezzar. If the wall extended to the ancient line of the Tigris it would have been nearly 40k long.’

The wall stood astride the northern approaches to Babylon itself. The wall’s function appeared primarily to have been military as it was not well situated to protect land against the flooding of the Euphrates which lay to the south. It is ‘beyond reasonable doubt’ that Habl-es-Sakhar is Nebuchadnezzar’s wall and Xenophon’s Wall of Media due to the location north of Sippar, the details of the construction, and the stamped bricks set in bitumen. This is rather satisfying because a surviving wall has been matched up with literary text.

Umm Raus to Samarra – Line 3

A third line of walls runs from Samarra on the Tigris to Ramadi on the Euphrates which delineated the upper limits of the alluvial plain where intense irrigation was possible. Here the fertile plain is not continuous between the Tigris and the Euphrates but the regions close to the rivers fit the description of valued irrigated land. As the rivers have diverged already significantly in the area of the third uppermost line, compared to the lower two lines, a wall that extended the whole distance would have had to have been much longer. Central sections might also have been purposeless as there was little valued, highly irrigated, land to protect and attackers would not have wanted to stray too far into less fertile land. This area is the site of two walls described in ancient texts and three surviving linear barriers.

Trench of Artaxerxes – In the Anabasis Xenophon described the march along the Euphrates, at the point where canals began, thereby indicating intense irrigation: ‘Cyrus … expected the king to give battle the same day, for in the middle of this day’s march a deep sunk trench was reached, thirty feet broad, and eighteen feet deep.… The trench itself had been constructed by the great king upon hearing of Cyrus’s approach, to serve as a line of defence.’ The trench does not appear to have survived but the site might have been reused to build later walls – the first being the Wall of Macepracta, discussed next, and second the surviving wall at Umm Raus.

Wall of Macepracta – Ammianus Marcellinus, describing the assault in AD 362 by the apostate Emperor Julian on the Sasasian Empire of Shapur II, wrote: ‘our soldiers came to the village of Macepracta, where the half-destroyed traces of walls were seen; these in early times had a wide extent, it was said, and protected Assyria from hostile inroads.’

There is a surviving belt of linear barriers which extends – with long gaps – between the Euphrates and the Tigris. The three walls mark the line where the fertile Babylonian plain peters out. There is the rampart starting at Umm-Raus which extends east from the Euphrates; El-Mutabbaq is a burnt brick wall with towers running west from the Tigris; and between them is a dyke named Sadd Nimrud (also called El-Jalu). Their dating is very uncertain.

Wall at Umm Raus – The wall, running east from the Euphrates, has been described: ‘From Umm Raus we see the wall running inland for a distance of about 7 miles, with rounded bastions at intervals for 2½ miles.… The wall appeared to be about 35–45 ft broad, with bastions projecting about 20ft. to 25ft., set at a distance of about 190 feet axis to axis. At its highest point the mound made by the wall is about 7 to 8 feet high. From the air it can be seen that there are about forty buttresses in all.’

The line may follow that of Artaxerxes’ trench. It is not a brick wall but an earth rampart. It was ‘never defensible, perhaps never finished’. Also: ‘This wall must have been designed … to protect the suddenly broadening area of fertile irrigated land to its south from raids and infiltration; large armies entering Iraq by the Euphrates would not have found it a serious obstacle.’

Again, there is the explicit mention of defending irrigated land. The Umm Raus rampart must date between 401 BC, as it is not mentioned by Xenophon, and AD 363, when a ruined wall was described at Macepracta by Ammianus Marcellinus.

El-Mutabbaq – The modern name, El-Mutabbaq, means built in layers or courses of bricks. This is a massive rampart lying at the boundary of the irrigatel alluvium of the widening Tigris valley south of Samarra and the desert to the north-west. It is about forty kilometres long and ‘has traces of turrets and moat on the north-west side and follows … the natural contours of the land. The rampart was four to six metres high, thirty metres wide at the bottom.’ It is, ‘a mud-brick wall three and a half bricks wide behind which is 10.5m of gravel-packing held in by a small mud-wall. The gravel packing was compartmented by mud-brick cross walls. There are projecting towers at regular intervals and a ditch about 20 to 30m. wide which is now about 2m. deep.’

The following description shows El-Mutabbaq as being designed to protect valued land against a nomad threat: ‘Herzfeld (a German explorer and historian) attributed construction to the threat of the Bedouin invading the fertile area along the Tigris by the river Dujail.’ These walls were seen as intended to stop nomads thereby affirming their ineffectiveness against great armies: ‘Cross-country walls of this type are notoriously inefficient at stopping great armies; this particular example could be outflanked without any difficulty at all. A stronger objection to any theory that it was designed to stop a great army is that it blocks the one route into southern Mesopotamia which, because of natural obstacles north of Samarra, invading armies have preferred never to use.’ The walls were intended to defend irrigated land: ‘El-Mutabbaq was more probably intended to help protect the irrigated land from unwanted settlers and raiding parties coming from the desert.’ There is no consensus as to the builder although they are described as Sasanian. Basically, these linear barriers do not seem to have been examined since the 1960s and remain effectively undated.

Sadd Nimrud – A dyke called Sadd Nimrud or El-Jalu, which is about forty kilometres long, that lies to the west of El-Muttabaq. This linear barrier does not extend the full distance between El-Muttabaq and the Wall at Umm Raus: ‘The fortification in the central area peters out in the direction of Falluja – perhaps as a considerable gap did not need to be defended – as armies could not advance far into the desert away from water.’ No date, other than this possibly being pre-Islamic, has been suggested.

Analysis – three lines at the Euphrates and Tigris convergence point

These three barriers between the Tigris and the Euphrates present a very baffling picture. They follow roughly the line where intense irrigation ceases. Rather than being a single response, however, they seem to be three discrete AD hoc reactions to separate threats to irrigated lands near the rivers Tigris and Euphrates. They can lay claim to being among the longest and oldest walls outside China, excepting certain Roman and Sasanian walls, yet there appears to have been no very detailed study of them. The attribution is generally vague – with comparisons made to features on Sumerian to Sasanian walls, in other words millennia apart. Generally commentators do regard them as forming part of a local response to the need to protect valued irrigated land in the immediate vicinity, rather than as having any strategic purpose to block routes into central and southern Mesopotamia.

Sasanian aquatic barriers

In the early fourth century AD a semi-nomadic people, the Lakhmid Arabs, who were originally from the Yemen, emerged as a serious threat to Sasanian Mesopotamia.

Khandaq-i-Shapur – Arab tradition associates Shapur II (AD 309–379) with a defensive dyke that reputedly ran west of the Euphrates, from Hit to Basra. This barrier is looked at again later when Sasanian barriers are discussed. It is clear however that the linear barrier was built to hinder the nomadic Arab people from the desert. Although this Khandaq is much later than the Egyptian Walls of the Ruler, it throws an interesting perspective on it. Firstly, there is neat symmetry. In the face of a threat from nomadic Asiatics, the response to both the east and the west of the Arabian Desert was to build a moat or canal. Secondly, the historian Yāqūt, writing later in the Islamic period, said that Anushirvan (531–579) who rebuilt Shapur’s earlier work, ‘built on it (the moat) towers and pavilions and he joined it together with fortified points.’ Therefore, this was a continuous fortified aquatic linear barrier. The fact that such a barrier was constructed by the Sasanians perhaps meant that Egypt’s early Walls of the Ruler were also a continuous aquatic barrier, strengthened by forts.

The Siege of Antwerp

Parma’s bridge across the Scheldt at Antwerp

Over the next few years the Duke of Parma consolidated the line between the loyal south and the rebellious north, and set about reducing the northern strongholds by means of a long succession of sieges, a process that culminated in the thirteen-month-long Siege of Antwerp – one of the most fascinating operations of the Eighty Years War. Parma’s plans involved cutting the city off from the north by building a bridge across the Scheldt. To many this was the strategy of a lunatic. That a river half a mile wide could be bridged while there were so many rebels around to prevent its construction was one reason for the scepticism. The other reason was that some years previously, when Antwerp was still in Spanish hands, William the Silent had attempted to build a bridge, only to see his creation swept away with the coming of winter and the pounding of ice floes. Nevertheless, William remained one of the few people to take Parma’s threat seriously, and he proposed a drastic course of action to frustrate Parma’s plans.

William’s plan involved the almost total inundation of the area. Downstream from Antwerp, the Scheldt was confined within its banks by a complex system of dykes, the most important of which extended along its edges towards the sea in parallel lines. On the right bank this barrier became the mighty Blauwgaren dyke, which was met at right angles by the equally formidable Kowenstyn dyke. Not far from where they joined, the Dutch had a strong fortress called Lillo. If the Blauwgaren dyke was pierced, it would take the Kowenstyn dyke with it and would cause such an extensive flood that Antwerp would become a city with a harbour on the sea. It would then be almost impossible to starve out.

Had William the Silent’s orders been carried out immediately, then Antwerp might indeed have been safe, but a fateful and time-wasting debate took place, and just a few weeks later William was assassinated. The idea of a massive flood was certainly not well received. In an echo of Alkmaar, it was pointed out that twelve thousand head of cattle grazed upon the fields protected by the two dykes. If Parma was intent upon starving Antwerp’s citizens, then surely there was no better way of helping him than by the Dutch destroying such a huge food supply.

The tiny village of Kallo, which lay about nine miles from Antwerp, became the construction site for Parma’s bridge, but the scheme was such a huge undertaking that by the autumn of 1584 little seemed to have been achieved. Antwerp continued to be supplied by flotillas of craft, which exchanged fire with Parma’s forts as they boldly made their way upstream. The Antwerp authorities then made an astounding blunder. It transpired that grain bought in Holland could be sold for four times its original price in beleaguered Antwerp, a mark-up that was attractive enough to make Spanish cannon fire an acceptable hazard. But the city fathers then set a fixed price for supplies brought in, and simultaneously regulated the accumulation of grain in private warehouses. Seeing their profit wiped out, the ships’ captains stopped the traffic stone dead. Even Parma could not have created such an effective blockade!

At the same time, the inundation urged by William the Silent had actually begun, albeit in a much-reduced fashion. Yet, ironically, the opening of the sluices on the Flanders side actually made Parma’s communications that much easier, because the flooded countryside now enabled him to give Antwerp a wide berth. By the time it was finally decreed that the dykes of Blauwagaren and Kowenstyn should be cut there were strong Spanish garrisons in place to prevent this happening. The Kowenstyn in particular now resembled a long, bastioned city wall bristling with cannon and pikes.

Meanwhile, the bridge grew slowly. On the Flemish side a fort called Santa Maria was erected, while on the Brabant side opposite developed one named in honour of King Philip II of Spain. From each of these two points a framework of heavy timbers spread slowly towards the middle of the river. The roadway was twelve feet wide, defended by solid blockhouses. Numerous skirmishers attacked the workmen in order to prevent the two halves meeting, but skirmishes is all that these attacks were. In spite of entreaties from Antwerp the vacuum of power since the death of William the Silent prevented any concerted attack from occurring.

Parma was also suffering from a lack of money. His army had not been paid for two years, and he was not yet in a position to promise early payment from loot. A botched attempt by the rebels to capture s’Hertogenbosch, Parma’s main supply centre for the siege, served only to increase the commander’s determination to complete his bridge, against which the wintry weather was now providing the only real challenge. The ocean tides drove blocks of ice against the piers, which stood firm, but in the centre portion of the construction the current was too strong to allow pile-driving, so here the bridge had to be carried on the top of boats. There were thirty-two of them altogether, anchored and bound firmly to each other and armed with cannon.

Parma’s bridge was completed on 25 February 1585. It was twice as long as Julius Caesar’s celebrated Rhine bridge, and had been built under the most adverse weather conditions. As an added precaution, on each side of the bridge there was anchored a long heavy raft floating upon empty barrels, the constituent timbers lashed together and supported by ships’ masts, and protected with iron spikes that made the construction look like the front rank of a pike square. An entire army could both sit on the bridge and walk across it, and, to impress the citizens of Antwerp, Parma’s soldiers proceeded to do both.

So that they should be under no illusions as to the strength and size of the edifice, a captured Dutch spy, who expected to be hanged, was instead given a guided tour of the bridge and sent safely back to relate in wide-eyed wonder what he had seen. `Tell them further’, said Parma to the astonished secret agent, `that the siege will never be abandoned, and that this bridge will be my sepulchre or my pathway into Antwerp.’

The Dutch ship Fin de la Guerre (“End of War”) during the Siege of Antwerp in 1585.

The first marine application of mine warfare occurred in 1585 at the city of Antwerp. Fighting for their independence from Spain, the Dutch were under siege by Spanish forces, who had built a fortified bridge across the Scheldt River to prevent supplies from entering the city. Frederigo Gianibelli sent a small ship loaded with gunpowder down the river, with a time fuse. The ship detonated directly beneath the bridge, destroying it and the Spanish soldiers guarding it.

The Diabolical Machine

The besieged citizens of Antwerp, however, still possessed one possible winning card. In their city lived a sympathetic Italian engineer by the name of Frederigo Gianibelli, and in a similar display of enthusiasm to that with which Parma had built his bridge, so did this Gianibelli determine to destroy it using exploding ships. His proposal to the city authorities involved the construction of a fleet, but by the time his project was approved the parsimonious city fathers had reduced the fleet to two ships, which disgusted Gianibelli, even though each of the vessels, to be optimistically named Hope and Fortune, was enormous. The two ships were nothing less than artificial volcanoes. In the hold of each was a chamber of marble, along their entire length, built upon a brick foundation. This chamber was filled with gunpowder under a stone roof, on top of which was a `cone’ – also of marble – packed with millstones, cannonballs, lumps of stone, chain-shot, iron hooks, ploughshares and anything else that could be requisitioned in Antwerp to cause injury when blown up. On top of all of this were piles of wood that gave the vessels the appearance of conventional fireships. The one difference between the two ships lay in the means of ignition of the volcanoes within. On the Fortune this was to be done by means of a slow match. On the Hope the business would be done by clockwork and flint, rather like an enormous wheel-lock pistol. The progress of these infernal floating mines was to be preceded on the ebb tide by thirty-two smaller vessels laden with combustible materials, which would keep the defenders of the bridge busy until the two great ships reached Parma’s masterpiece and utterly destroyed it.

The date for the attack was to be dusk on 5 April 1585, and the enterprise was placed into the hands of Admiral Jacob Jacobzoon. He began badly, sending all the thirty-two vanguard ships down the Scheldt almost all at once rather than in the steady progression previously agreed upon. On each bank, and from every dyke and fortress, the Spanish troops gathered in their thousands to gaze at the burning flotilla that was turning the night back into day with its ruddy glow. Some of the boats hit the forward barges of the bridge and stuck on the spikes, where they burned themselves out ineffectively. Others struck the banks or ran aground. Some simply sank into the river as their own fires consumed them.

To the guardians of the bridge the attack seemed to be having no effect, but behind these minor vessels there now loomed the two great ones. They meandered somewhat aimlessly with the tide and the current, because their pilots had long since abandoned them. There was a moment of concern for the Spanish when the Fortune swung towards the side of the river, completely missing the forward protective raft. It eventually ground itself while, unknown to the Spanish defenders, the slow match burned through. There was a small explosion, and some minor damage, but so slight was the effect that Parma sent a boarding party to examine the interior of the ship.

They did not stay long, because the Hope had now followed its sister downstream. Its precision in finding its target could not have been better if it had been guided until the very last moment, because it managed to hit the bridge next to the blockhouse where the middle pontoons began. However, as Parma had confidently expected, the bridge had been so strongly built that the impact alone caused it no damage. Expecting it to be another fireship, Spanish boarders leapt on to the deck, and with excited whoops of laughter promptly extinguished the decoy fire. With some sixth sense, an ensign rushed up to his commander and begged him to leave the scene. So earnest were the man’s pleas that Parma reluctantly withdrew to the Fort of Santa Maria. This saved his life, for at that very moment the Hope exploded.

Not only did the ship vanish, so did much of the bridge, the banks, the dykes, the fortresses, and for a brief moment even the waters of the Scheldt, as possibly the largest man-made explosion in history up to that date lit up the night sky. The facts and statistics of the act took months to establish, and still have the power to cause amazement. The entire centre section of the bridge disintegrated. More than a thousand Spanish soldiers died instantly, and their bodies were never found. Houses nearby collapsed as if hit by an earthquake, and the pressure wave blew people off their feet. From the sky there began to fall the cannonballs and stones that had been crammed into the ship, accompanied by the mortal remains of its immediate victims. Slabs of granite were later found buried deep in the ground having travelled six miles from the scene of the explosion.

The personal tales were also quite remarkable. One Marquis Richebourg, who had been in command on the bridge, simply disappeared. His body was located several days later, its progress through the air having been arrested by one of the chains Parma had strung across the river. Seigneur de Billy’s body was not located until months afterwards when his golden locket and an unpleasant stain on one of the surviving bridge supports provided identification. The fortunate Duke of Parma was merely knocked unconscious by a flying stake. One captain was blown out of one boat and landed safely in another. A certain Captain Tucci was blown vertically into the air in his full armour and dumped in the river, where he still retained the presence of mind to remove his cuirass and swim to safety. Another young officer was blown completely across the river and landed safely after a flight of half a mile.

The original plan was that immediately after the expected explosion Admiral Jacobzoon should launch a signal rocket that would send boatloads of armed Dutchmen pouring on to the scene. Instead, he was totally stupefied by the explosion and gave no order. No rocket was fired and no one advanced. During the hiatus Parma regained consciousness, and by displaying leadership skills of unbelievable quality he managed to marshal his men to begin to repair the damage. Even though the Dutch advance was expected at any moment, it never came. By daybreak, even Parma began to believe the unbelievable – that the Dutch rebels, having set off the largest explosion since the introduction of gunpowder to Europe and blown a hole in his bridge, were now going to let him mend it. Yet this is precisely what happened.

The battle for the Kowenstyn dyke

The Kowenstyn Dyke

With the initiative lost it took the defenders of Antwerp a full month to mount another attack on Parma’s besieging army. The new attack was not against his damaged bridge but on the mighty Kowenstyn dyke. As the target was an earthen dam explosives would not have been effective, so the goal of breaking the great barrier would be made by men capturing the dyke with pike and musket and then cutting it with pick and shovel. It was a low-tech solution, and it was likely to be a very bloody one.

Following a successful landing a fierce `push of pike’ began on top of the Kowenstyn dyke. The rebels could well have been shoved back into the water had it not been for the arrival of the other half of their army downstream from Antwerp. For once in this campaign a co-ordinated effort had actually worked, and three thousand men now occupied this small section of the dyke. Among them was an eighteen-year-old youth called Maurice of Nassau, the son and heir of William the Silent, who was experiencing his first real taste of combat in what was to become a renowned military career. While two walls of soldiers shot, cut and speared their enemies, the sappers began two very different but complementary operations: to reinforce the dyke with trenches and mounds, and also to cut a hole through it. At last a loud cheer went up as the salt water rushed in a torrent through the newly created gap. A few moments later a Zeeland barge sailed through.

It is to the great credit of the Spanish commanders on the scene that they did not immediately panic; they stayed calm, even though their leader was some distance away. They were also sensible enough to realise that a breach sufficient to allow a Zeeland barge through was by no means sufficient to permit the passage of an entire fleet, and if the dyke could somehow be recaptured then the rupture might even be repaired. Five attacks followed along the dyke in a manner that demonstrated beyond all doubt why the Spanish were regarded as the finest infantry in Europe. The last assault was successful, and it was not long before intelligence arrived in Antwerp that the wild celebrations currently taking place were somewhat premature.

The failure plunged Antwerp into despair and forced its rulers back to the negotiating table. They sought three reassurances from Parma: that religious freedom would be granted, that troops would not be stationed in the city, and that the hated citadel would not be rebuilt. Knowing that King Philip II would accept none of these `exorbitant ideas’, as Parma termed them, he reminded the citizens of Antwerp of the stranglehold he still had on their city. But he had other cards to play, and drew their attentions to the role of Antwerp as the `great opulent and commercial city’ that it had been in the past and could be again. What cause, what real cause, did rich Antwerp have with the heretical Sea Beggars of Holland and Zeeland? Surely the loyal south was more to their liking?

Parma’s own fears lay with the winter that was fast approaching. It turned out to be so severe that Parma’s bridge would have been unlikely to survive, but by the time winter came a settlement had already been reached. A minor concession regarding the troops to be stationed in Antwerp proved sufficient for all parties to be satisfied, and Antwerp capitulated with honour on 17 August 1585 without a shot having been fired at the city itself. There was no massacre, no sack, no pillage and Parma’s soldiers were paid not by loot but in hard cash. The noble Duke of Parma had achieved his objectives, and, unknown to him at the time, he had actually achieved something quite remarkable. By detaching the fate of Antwerp and the lands to the south from the United Provinces of The Netherlands he had effectively created a recognisable and workable border. In 1648, as part of the Treaty of Westphalia, this border was to be given both recognition and reality, confirming that Alexander Farnese, Duke of Parma, had invented Belgium.

Trench Warfare in the American Civil War

The Peninsula Campaign and Trench Warfare

When General George McClellan persuaded Lincoln (against the latter’s judgement) to leave only 75 000 men guarding Washington from behind fortresses and land more than 100 000 men on the Yorktown Peninsula on 22nd March 1862 to strike at Richmond by sea, he sowed the seeds of failure by keeping secret from the President the fact that he was leaving only 50 000 to guard the capital. For when Lincoln discovered the deceit, he withheld 25 000 men from McClellan. By then the General was enmeshed in what amounted to almost constant and costly siege warfare against a series of well-entrenched lines of resistance, dug across his predicable line of advance through the ten-mile neck of the peninsula, but guarded by only 60 000 enemy troops under General Joseph Johnston.

The campaign developed a pattern hitherto unknown in warfare. Excepting sieges of fortified cities, combat in the past had been of short duration, major battles rarely lasting for more than a day. The Peninsula campaign, commencing with the Battle of Kernstown on 23rd March as part of General Jackson’s diversionary activities in the Shenandoah Valley, and ending in the withdrawal of Federal forces from the Peninsula in August, consisted of almost ceaseless fighting. Including the siege of Yorktown from 4th April to 4th May, there were no less than six major battles in the Valley and nine in the Peninsula, connected by continual skirmishing and one major raid by a cavalry division. Moreover, the Peninsula fighting coincided with a major campaign in the west, on either side of the Mississippi, where the struggle to control that jugular vein of the Confederacy culminated in the bloody Battle of Shiloh on the 6th and 7th April; the capture of New Orleans by a Federal fleet of 17 warships under Admiral David Farragut on 25th April; and the fall of Memphis to Federal river gunboats on 6th June. Losses were colossal – 14 000 Federal and 11 000 Confederate troops at Shiloh alone. Exhaustion became endemic, halting operations.

Although these heavy losses could partly be ascribed to errors of raw troops, as well as to poor staff work, the underlying reasons were improved technology which had redoubled firepower, and crippling deficiencies in communication which technology had not yet solved. When McClellan advanced from Yorktown in the direction of Richmond, his progress was slowed by an out-numbered Confederate rearguard which gave ground only grudgingly on a wide front. This was possible because no longer did men need to be packed into tight ranks in order to generate sufficient volume of fire to maintain their position against assault. Reciprocally, the thinning out of ranks made them less vulnerable to incoming fire. Such gams were ameliorated further when men took to lying down to shoot or, better still, made a point of firing from trenches or behind cover instead of standing up in the open, as so recently in the past decade. Not that either army was yet able to apply the full devastating potential of modern weapons. Many old, muzzle-loading rifles were still in service, but a sound of the future ripped forth at the Battle of Fair Oaks when on 31st May, within sight of Richmond, a battery of hand-operated Williams machine-guns chimed in to support the first Confederate counter¬ stroke – a battle which was to save their capital city though it failed, with losses of over 6000 men, to drive McClellan back.

As had been shown in the Crimea and at Solferino, head-on assaults against a well emplaced enemy of equal calibre were no longer profitable operations of war. Even less viable was cavalry against modern artillery and rifle fire. The only chance of making a worthwhile mounted contribution was to ride through gaps in the enemy lines, both for reconnaissance and for raids, into the wide-open spaces of the enemy rear. In a country the size of America, and with relatively small forces engaged, there would always be gaps and nobody was better at exploiting these than General JEB Stuart, as he demonstrated between 12th and 15th June when he rode right round McClellan’s army, creating havoc in the rear and returning to Lee (given command in the field after Fair Oaks) with invaluable information about Federal dispositions.

Trench Warfare – The Lines of Petersburg

When Grant deluded Lee as to his true intentions after the Wilderness Campaign, managing suddenly to appear in mid-June with massed forces at Petersburg instead of further north as expected, the thinly defended city lay at his mercy. But war weariness and a conditioned caution held back the Federal troops who now approached all entrenchments warily as a matter of course. One quick determined rush by a Corps on 15th June 1864 might well have broken through. Three days later an army of 65 000 was insufficient to overcome the 40 000 men Lee had rushed to the spot by rail. Faced at first by an improvised line, the initial Federal assault failed from lack of co-ordination. Detachments advanced independently, inadequately supported by artillery, and were pinned to the ground by fire of only moderate intensity. By the time a set-piece attack could be launched on the 18th, the volume of defensive fire was annihilating, compelling Grant to call a halt and commence probing the city’s southern flank with a view to isolating it. Keeping pace with each Federal sidestep to their left, the Confederates extended their entrenchment to their right, always in time to meet each assault while fiercely contesting Grant’s further attempts to cut the line to Richmond or the one running westward from Petersburg. Assault was usually of the battering-ram sort – a blasting of the selected point of attack by artillery and mortars (the latter, with their plunging fire, being particularly suitable for striking at the deeper enemy emplacements) followed by a massed infantry charge.

Historians accuse those in charge of a succession of failed set-piece attacks with bungling. To some extent they are right, although they tend to overlook that the dimensions and ferocity of modern war had produced a complex problem beyond the knowledge and technology of the day to solve. `In war’, said a Prussian officer called Hindenburg, many years later, `only simple plans work’. In 1864 simplicity could not be adopted. Even if every plan had been perfectly devised, staff work impeccable, communication arrangements fault¬ less and every order executed implicitly, the weather, or the enemy could be expected to disrupt them. But nothing could be perfect in this form of warfare, with masses of men and numerous weapon systems somehow to be coordinated. Humanity failed in all its unpredictability. That way chaos and slaughter were assured.

The attack on the Redoubt at Petersburg on 30th July demonstrated in utter confusion the inability of commanders to make human courage prevail over material factors and human frailty. As a powerful augmentation of the, by now, conventional artillery concentration of fire, a mine containing four tons of black powder was to be exploded beneath the redoubt and its defenders. Placed in a cross shaft at the end of a 511-foot tunnel which a regiment of coal miners secretly dug, it was blown at dawn without warning to the enemy. General Ambrose Burnside, whose four divisions of infantry were to exploit the explosion, seems to have relied too much upon the shock effect of the mine; beyond doubt the measures he took to ensure that the troops not only occupied the crater but pressed on rapidly beyond, were ambiguous and unambitious. As for the troops, so staggered were they by the enormity of the explosion, the air pressure of its blast and the scene of carnage which met their eyes when they poured into the crater, that they lost all sense of purpose and stayed there all morning, poking about among the grisly ruins of dismembered men and equipment. On the Federal side leadership came to naught while among the Confederates initial shock was gradually overcome and a counter-stroke launched in the afternoon. Artillery sealed off the flanks of the 500-yard breach in the defences, as infantry rushed to the lip of the crater where they fired volleys into the disorganized mob below. The Federals were flung back with the loss of 3793 men. That day the Confederates lost 1182, including those blown up.

For the rest of the summer and into the fall, Grant strove fitfully to break the deadlock in front of Richmond and Petersburg, creating for the logistic support of his troops a comprehensive conglomeration of base depots, camps and railway spurs leading to within artillery range of the enemy. Facing the Confederate capital the entrenched front was some 37 miles long, manned by 90 000 well-provisioned Federal troops on one side, and 60 000 deprived but fanatically determined Confederates on the other. Try as he would to smash through, Grant was defeated. Likewise, Lee was rebuffed when in March 1865, a last sortie took Fort Stedman but got no further than its ramparts before it was stopped by a Federal counterattack. In a four-hour battle, the attackers lost twice as many men as the defenders – 4000 to 2000.

Had Grant’s exploits comprised the sole Federal effort in 1864 they could well have led to his and President Lincoln’s downfall in an election year. The disgruntled General McClellan was campaigning for the Democratic candidacy with a call for an end to the war. He might have won if General William Sherman, taking advantage of the dram of Confederate strength to the east, had not struck the decisive blow in the west.

Hohentwiel Fortress

The fortress resisted five Imperial sieges in the Thirty Years’ War, under the command of Konrad Widerholt between 1634 and 1648. The effect was that Württemberg remained Protestant, while most of the surrounding areas returned to catholicism in the Counterreformation.

The mountaintop fortress of Hohentweil under attack in October 1641.

1643 illustration of Festung Hohentwiel (Hohentwiel Fortress), near modern Singen, Germany.

The only feasible way from Alsace for an army is around the southern end of the Black Forest where the route divided. One branch ran north east to the upper reaches of the Danube around the Württemberg enclave of Tuttlingen and thence to Bavaria. This route was overlooked by the duke’s impregnable castle of Hohentwiel perched on an extinct volcano 263 metres above the surrounding plain. The other branch ran east through the towns of Überlingen, Lindau and Radolfzell along the northern shore of Lake Constance to the Bregenzer Klause, the pass giving access to the Tirol and Valtellina. The area between the lake, the Danube and the Bavarian frontier was studded with walled imperial cities, notably Ravensburg, Kempten, Memmingen, Ulm and Augsburg. The emperor was rarely able to devote significant resources to defending these positions, despite a strategic importance that grew with French intervention in 1635. Defence was left largely to local militia, especially in Villingen and the imperial city of Rottweil which guarded the back door from Württemberg through the Black Forest to Breisach.

Sweden Loses Southern Germany

Against this the Habsburg loss of 2,000 seemed slight and enabled them to claim a major triumph. Following a long succession of defeats, Nördlingen appeared to be vindication for Wallenstein’s murder and cemented the influence of Gallas and Piccolomini by associating them with victory. As at Breitenfeld, the scale was magnified by the demoralization of the enemy army. News of the defeat reached Frankfurt on 12 September along with a flood of refugees. The remaining Heilbronn delegates fled the next day. Oxenstierna tried to improvise a new line of defence along the Main to contain the defeat to the south. No one cooperated. Johann Georg failed to launch the requested diversionary attack against Bohemia, while Duke Georg refused to move south to hold the middle section of the river. Wilhelm of Weimar abandoned Franconia and fell back with 4,000 men to his base at Erfurt, exposing the upper Main to Piccolomini and Isolano who approached with 13,000 troops from Nördlingen and north-west Bohemia. Piccolomini took Schweinfurt, while Isolano destroyed the Suhl arms workshops that had supplied most of the Swedes’ small arms and munitions since 1631. Isolano and 6,000 Croats then swept down the Main in November, rampaging into the Hessian possession of Hersfeld.

The main imperial army moved west, bypassing Ulm to enter Stuttgart on 19 September. Duke Eberhard III fled to Switzerland and the last Württemberg fort surrendered in November. Only the isolated Hohentwiel on the upper Danube held out. While the Imperialists made themselves at home, the Spanish continued westwards and the Bavarians, now under the command of Werth, captured Heidelberg on 19 November, though its castle remained defiant. Riding ahead, Werth’s cavalry harried the remnants of Bernhard’s army as it fled to Frankfurt. The commanders of Sweden’s Rhine army refused to join him, on the ground this would spread demoralization to their own troops. Birkenfeld abandoned Heilbronn and retreated to the Kehl bridgehead opposite Strasbourg. His hopes of replacing Bernhard were dashed by Oxenstierna, who felt there was no realistic alternative to the defeated general. The death of Count Salm-Kyrburg to plague on 16 October enabled Bernhard to incorporate the former Alsatian units into his command.

Leaving Werth and Duke Charles to complete the conquest of the Lower Palatinate, Fernando continued his march down the Rhine, crossing at Cologne on 16 October to reach Brussels nineteen days later. Meanwhile, Philipp Count Mansfeld collected the Westphalians at Andernach, allegedly accompanied by a hundred coach-loads of Catholic lords and clergy eager to recover their property. As Philipp marched south, it looked as if Bernhard would be crushed between his hammer and Gallas’s anvil.

The situation mirrored that of 1631, only this time Protestant areas were the ones affected as government collapsed in the wake of the headlong flight of Sweden’s German collaborators. Suffering was also more general because the plague hindered the harvest, causing widespread hardship. There were signs that Emperor Ferdinand had learned the lessons of 1629 as efforts were made to restrain over-zealous Catholics. He intervened to stop Bishop Hatzfeldt punishing the Franconian knights for collaborating with the Swedes and the Jesuits were refused permission to take over Württemberg’s university at Tübingen. Political considerations undoubtedly influenced this, since Vienna did not want to jeopardize promising negotiations with Saxony. Archduke Ferdinand’s presence was another moderating factor. However, it often proved impossible to stop officers and administrators exploiting the situation, either to enrich themselves or to find cash for the perennially underpaid imperial army. Catholic government resumed relatively quickly in Würzburg despite the Swedish garrison holding out on the Marienberg and in Königshofen until January and December 1635 respectively.

Oxenstierna worked feverishly to salvage what he could of the situation, reconvening the Heilbronn League congress at Worms on 2 December. Though some members were willing to fight on, most sought a way out through Saxon mediation. Saxon and Darmstadt envoys agreed draft peace terms, known as the Pirna Note, on 24 November. Oxenstierna tried to stem desertion by publishing what he could discover of the terms, notably the suggestion of 1627 as a new normative year that would secure many of the Catholic gains.

Partisan Leaders

The horrendous losses of the 1638 Rhine campaign were a major factor behind this decline. Bernhard of Weimar was determined to achieve the objective set the year before and establish a firm foothold for France east of the river. This time he prepared thoroughly. Since he had wintered in Mömpelgard and the bishopric of Basel, he was already close to the stretch of the Rhine along the Swiss frontier to Lake Constance. This route offered an alternative to the previous year’s attempt to punch directly across the Black Forest. Though he had been joined in person by Rohan, who had escaped from the Valtellina, he still had few troops. Savelli’s Imperialists held Rheinfelden with 500 men, with other garrisons in Waldshut, Freiburg and Philippsburg and Reinach’s regiment in Breisach. These posts would have to be taken if the river was to be secured. He also needed a base beyond the Black Forest to tap the richer resources of Württemberg and the Danube valley.

Fortunately, Bernhard had an excellent spy network and knew how weak his opponents were. He also had the services of Colonel Erlach, a veteran of Dutch service who had been wounded at White Mountain and subsequently served Mansfeld and Sweden until 1627. Since then he had commanded the militia of his homeland, the Protestant canton of Bern. He joined Bernhard’s army in September 1637, although he did not leave Bernese service until the following May. His contacts with the canton’s patriciate ensured a good flow of supplies to Bernhard’s army.

Erlach also opened negotiations with Major Widerhold, a Hessian who was the Württemberg militia’s drill instructor and commandant of the Hohentwiel, the duchy’s only fortress still holding out against the emperor. Though he has now faded from the local popular consciousness, Widerhold occupied a prominent place in Swabian patriotic folklore into the twentieth century. He exemplifies the partisan leaders who played an increasingly important role as the rapid escalation of the conflict left numerous isolated garrisons scattered across the Empire. These sustained themselves by raiding and acted as potential bases should friendly forces return to their area. The Swedes in Benfeld, Ruischenberg’s Imperialists in Wolfenbüttel and Ramsay’s Bernhardines in Hanau are three examples encountered already. Others included the Hessians in Lippstadt under the Huguenot refugee Baron St André and his subordinate, Jacques Mercier from Mömpelgard, known as Little Jacob, who rose through the ranks of Hungarian, Bohemian, Russian and Dutch service. Both were contemporary celebrities incorporated by Grimmelshausen into his novel. A counterpart in Habsburg service was the Swiss patrician Franz Peter König, ennobled in 1624 as von Mohr, who distinguished himself in skirmishes around Lake Constance in the early 1630s. As these background sketches indicate, such men generally came from relatively humble backgrounds and made reputations and fortunes through daring exploits. They never rose to command armies and were often difficult to control. König was dismissed after becoming embroiled in a feud with the highly disagreeable Wolfgang Rudolf von Ossa, Habsburg military commissioner for south-west Germany.

Widerhold acted nominally in the name of Duke Eberhard III of Württemberg, but pursued his own agenda. Mixing terror with benevolence, he spared the immediate vicinity of the Hohentwiel and concentrated on longer-range raids against Catholic communities, forcing 56 villages, monasteries and hamlets to provision his garrison that rose to 1,058 men and 61 guns by the end of 1638. He was well-supplied with intelligence from friendly villagers who often participated in his plundering expeditions. He returned the favour on his death, leaving a large endowment for the local poor. His exploits became legendary. Once he caught the bishop of Konstanz out hunting and stole his horse and silver, and later he netted 20,000 talers by capturing the local imperial war chest in Bahlingen.

Blockaded since Nördlingen, Widerhold agreed to remain neutral after February 1636 because of renewed talks to include Württemberg in the Prague amnesty. Ferdinand III made surrender a condition for restoring Eberhard III in 1637, but Widerhold ignored ducal orders to comply and declared for Bernhard in February 1638. He remained a constant thorn in the Habsburg side, not least by raiding the Tirolean enclaves, and did not submit to ducal authority until 1650.


By draining other regions of their troops, Ferdinand III managed to collect 44,000 men in Bohemia by January 1640. Of these, only 12,400 were available as a field army under Archduke Leopold Wilhelm, reinforced by 4,100 under Hatzfeldt who had wintered in Franconia. Piccolomini was down to 13,000 in Westphalia, while the Saxons mustered 6,648, or only a quarter of their strength five years earlier. The Brandenburgers had effectively been knocked out. The Bavarians still totalled about 17,000 men, most of whom were on the Upper Rhine where there were perhaps 10,000 in total, including a few Imperialists. The rest were in winter quarters around Donauwörth and Ingolstadt. As these figures suggest, it was now very difficult to launch major operations in more than one region at a time.

His enemies were in a similar position. Banér was reduced to 10,000 effectives, while the other Swedish commanders had only enough men to hold their current positions. Banér had little choice but to evacuate Bohemia in March and fall back the way he had come the previous year to join Königsmarck at Erfurt. The units left to hold Saxony were defeated at Plauen on 20 April 1640, forcing the garrison in Chemnitz to surrender while most of the others abandoned their positions.

The challenge over the coming two years was for France and Sweden to establish a viable framework for military and political cooperation that had to include the Hessians and Guelphs, while Ferdinand pinned his hopes on frustrating this with one last effort to rally all Germans behind the Prague settlement. The emperor’s preference for negotiation was cruelly exploited by the Guelphs and Hessians who had used the winter to gather their strength and now declared their hand in May 1640. Duke Georg did this openly by sending troops to Banér, counting on Swedish help to prevent an invasion of Hildesheim. He nominally mustered 20,000, but in fact had 6,000 at Göttingen and garrisons along the Weser, plus a field force of 4,500 under Klitzing. Amalie Elisabeth acknowledged her French alliance in March, but still promised to respect the truce in Westphalia. With French agreement, Melander moved the 4,000-strong Hessian field force east to the Eichsfeld in May to reinforce Banér. Richelieu summoned de Longueville from Italy, hoping that he possessed sufficient personal authority as a duke to master the 8,000-strong Bernhardine field army. This moved back down the Rhine to join the allied concentration.

The emperor was obliged to match these moves. He still hoped to win over the Hessians and so accepted Amalie Elisabeth’s assurances. Nonetheless, Wahl, the new Cologne commander, was authorized to recover the positions her troops had seized over the last two years in breach of the truce. Hessian garrisons also became bolder, now raiding Paderborn. Piccolomini followed Melander east and joined Leopold Wilhelm at Saalfeld, south of Erfurt, on 5 May. They entrenched to block the way into Franconia. After a two-week stand-off, Banér fell back north-west into Lower Saxony, alarming the Guelphs who feared he would abandon them. Once they had promised another 5,000 men, he marched south again to Göttingen and Kassel. Leopold Wilhelm shadowed him, moving through Hersfeld to entrench again at Fritzlar in August. It was cold all year, the summer was wet and miserable and food proved hard to find. Banér’s second wife died and de Longueville fell ill, relinquishing command to Guébriant again. The Bavarian field army arrived from Ingolstadt, bringing Leopold Wilhelm back up to 25,000 men. After another four-week stand-off, Banér withdrew, allowing the archduke to advance north down the Weser to join Wahl’s 4,000 field troops. Together, they took Höxter in October, but the men were exhausted and ill-disciplined. The weather grew windy and even colder. Leopold Wilhelm retreated south to winter at Ingolstadt. Banér left 7,000 to blockade Wolfenbüttel, while the rest of his army made themselves comfortable at the expense of the Guelphs’ villagers.

Seemingly uneventful, this campaign completely shifted the war’s focus to northern Germany, transplanting the ‘little war’ of outposts from Westphalia to the Upper Rhine instead. Under Erlach’s direction, the Bernhardine garrisons operated from Breisach and the Forest Towns in conjunction with Widerhold in the Hohentwiel. The Bavarians retaliated from Philippsburg, Heidelberg and Offenburg, while the Imperialists sortied from Konstanz and Villingen. Neither side managed to spare more than 3,000 men from their fortresses, severely restricting what they could achieve. Erlach helped disrupt plans to besiege the Hohentwiel in 1640 by sending cavalry to collect the Swabian harvest. Claudia scraped together another expedition against Widerhold in 1641, but heavy snow and lack of food forced this to be abandoned in January 1642. Erlach and Widerhold scored the only success, briefly combining the following January to take Überlingen by surprise.

The Freiburg Campaign

The situation appeared promising for the emperor at the beginning of 1644. Sweden’s decision to attack Denmark at the end of the previous year removed the threat to the Habsburg hereditary lands. The forces there were reduced to 11,000 men under the rehabilitated Field Marshal Götz, allowing Ferdinand to mass 21,500 under Gallas who marched down the Elbe to help the Danes. Buoyed by their success at Tuttlingen the previous year, Mercy’s Bavarians totalled 19,640, or twice the size of France’s Army of Germany despite Mazarin spending 2 million livres to rebuild it during the winter. This was the first time since 1637 that the emperor and his allies began the year with an army large enough to go on the offensive on the Upper Rhine.

General Turenne had been recalled to command in Alsace, but his forces were too weak to fulfil Mazarin’s expectations of conquering land beyond the Black Forest. Mercy attacked instead, retaking Überlingen on 10 May, eliminating the last French gain from 1643. Having been repulsed from the Hohentwiel, he left 1,000 men to blockade Widerhold’s garrison and crossed the Black Forest to recover the areas lost in the 1638 campaign. Turenne was forced to abandon his own advance through the Forest Towns, double back into Alsace and re-cross to save Breisach. Duke Charles broke off another of his periodic negotiations and renewed raiding into Lorraine. Mazarin was obliged to redirect d’Enghien from covering Champagne to retrieve the situation on the Rhine. Despite dashing 33km a day, d’Enghien arrived too late to save Freiburg, which had surrendered to the Bavarians after a prolonged bombardment on 29 July.