OSPREY Fortress PUBLISHING Hadrian's Wall AD 1 2 2 - 4 10 N Fields • Illustrated by D Spedaliere & S Sulemsohn Spedaliere First published in Great Britain in 2003 by Osprey Publishing, Elms Court. Editor's note Chapel Way, Botley. Oxford OX2 9LP United Kingdom, Email: [email protected] When classical authors are referred to throughout the text the standard form of reference has been adopted.The formula used is © 2003 Osprey Publishing Ltd. 'author','title' (if the author wrote more than one work) followed All rights reserved. Apart from any fair dealing for the purpose of private study, by either a two- or three-figure reference, indicating "book, research, criticism or review, as permitted under the Copyright. Designs and Patents chapter" and 'paragraph'.Thus 'Polybius (6.33-7)' refers to Ace 1988. no part of this publication may be reproduced, stored in a retrieval system, chapters 33 to 37 of the sixth book of the only surviving work by or transmitted in any form or by any means, electronic, electrical, chemical, mechanical, Polybius. whilst 'Tacitus Annales 13.35' refers to the 35th chapter optical, photocopying, recording or otherwise, without the prior written permission of the 13th book of the Annales by Tacitus. of the copyright owner. Enquiries should be addressed to the Publishers. Artist's note ISBN I 84174 430 2 Editorial: llios Publishing. Oxford, UK (www.iliospublishing.com) Our sincere thanks to all who have helped in the preparation of Design: Ken Vail Graphic Design, Cambridge, UK this book, especially to Marcus Cowper and Nikolai Bogdanovic Index by Alison Worthington who believed in us and gave us the wonderful opportunity to Originated by Grasmere Digital Imaging, Leeds, UK illustrate the book, and Dr. Nic Fields, who enabled us to go deep Printed and bound by L-Rex Printing Company Ltd. into the subject with his great knowledge.This book is dedicated to our dearest daughter Alina, and to our parents, Misu, Edith, 03 04 05 06 07 10 9 8 7 6 5 4 3 2 I Antonio and Maria. A CIP catalogue record for this book is available from the British Library. Readers may care to note that the original paintings from which the colour plates in this book were prepared are available for FOR A CATALOGUE OF ALL BOOKS PUBLISHED BY OSPREY MIUTARY AND AVIATION private sale.All reproduction copyright whatsoever is retained by PLEASE CONTACT: the Publishers. AII enquiries should be addressed to: Osprey Direct USA, c/o MBI Publishing, PO Box I. 729 Prospect Ave. Osceola. WI 54020, USA. Sarah Sulemsohn Email: [email protected] Tel-Fax: 00-39-0575-692210 Osprey Direct UK.PO Box 140, Wellingborough, [email protected] Northancs. NN8 2FA, United Kingdom. [email protected] Email: [email protected] www.alinaillustrazioni.com www.ospreypublishing.com The Publishers regret that they can enter into no correspondence upon this matter. Abbreviations AE L'Annee Epigraphique, Paris, 1888- BAR British Archaeological Reports, Oxford BMC III Coins of the Roman Empire in the British Museum II London, 1936 CIL Corpus Inscriptionum Latinarum, Berlin, 1862- CPL Corpus Papyrorum Latinarum,Wiesbaden, 1956-8 ILS Inscriptiones Latinoe Selectae, 2nd edition, Berlin, 1954 P. Mich Papyri in the University of Michigan Collection, Ann Arbor, 1931- P. Oxy. The Oxyrhynchus Papyri, London, 1898- RIB Roman Inscriptions of Britain I, 2nd edition, Stroud, 1995 RMR Roman Military Records on Papyrus, New Haven, 1971 SB Sammelbuch griechischer Urkunden aus Agypten, Strassburg, Leipzig, Heidelberg,Wiesbaden, 1913- SHA Scriptores Historiae Augustae SP Select Papyri: non-literary Papyri, Cambridge. Mass., 1923-4 Tab. Vindol. II The Vindolanda Writing Tablets II, London, 1994 Contents Introduction 4 Roman fortifications 5 Fortresses • Forts • Fortlets »Watchtowers Marching-camps • Practice-camps The origins of Hadrian's Wall 9 Chronology The anatomy of Hadrian's Wall 12 The Wall • Ditch • Military Way •Vallum • Milecastles "Turrets • Forts Bridges • Northern outposts • West coast defences * Phases of construction The construction of Hadrian's Wall 26 Chronology • Materials • Builders The function of Hadrian's Wall 35 The garrison of Hadrian's Wall 40 Life on Hadrian's Wall 48 Duties • Diet • Alcohol • Bathing • Gaming Sports • Women • Leave * Soldier and civilian The sites today 60 Useful contact information Further reading and research 61 Glossary 62 Index 64 Introduction As the first century AD matured, the boundaries of the Roman Empire become increasingly fixed, and what were once temporary stop lines become firm frontiers. Significantly, from the Latin for frontier, limes (pi. limites), we gain our word 'limit'. Consequently, the army's role predominately became one of policing the frontier tribes, preventing livestock rustling and tax evasion, mounting punitive raids, and showing the flag to friendly tribes outside the empire. The frontiers of the empire took many forms. Some of them were completely- open with scarcely any boundary definition, while a military road marked others. The Wall at Walltown Crags, Some of them followed the lines of rivers, while others were closed off with looking south-west from turret manmade barriers. The latter were not of uniform design, except that most were 45a (Walltown), showing the accompanied by one or more ditches. Hadrian's Wall was extremely elaborate, facing-stones and rubble-core composed of three separate defensive features, a ditch to the north, then the used in its construction. This is a wide stone curtain-wall with turrets, milecastles and forts strung out along it, and section of the Narrow Wall. (Author's Collection) finally a larger earthwork to the south. Running some 75 miles from sea to sea, it has been justifiably described as over the top. Other frontiers were less complex. In Germania, Hadrian built a palisade fronted by a ditch, replaced at a later date by a bank of earth. In Britannia Hadrian's Wall was replaced for a short time by the Antonine Wall, 45 miles to the north, built not in stone, but turf-blocks. In Raetia, approximately the area of modern Switzerland and Austria, a stone curtain-wall was constructed, but not so wide as Hadrian's Wall. In Africa stretches of dry-stone walling have been found marking sections of this very long frontier, other sections of which were left open, but not necessarily unguarded as is evident from a number of blockhouses. In most frontier provinces legionary fortresses were situated in the interior, some distance behind the borders. On parts of the Rhine and Danube, particularly where the frontiers were marked by the rivers themselves, the legions were stationed at strategic points close to the river banks, sometimes so close that the fortresses were washed away and had to be rebuilt further back from the rivers. Auxiliary troops were generally stationed in forts on the line of the frontier itself, actually attached to them as on Hadrian's Wall and the Antonine Wall, or some short distance (c. 1 mile) behind them, as in Germania and Raetia. Most frontiers were equipped with smaller fortlets like the milecastles attached to Hadrian's Wall, or the freestanding Kleinkastellen along the Rhine. 4 In between these were usually watchtowers. Roman fortifications Most of our knowledge concerning the layout and terminology of Roman military installations derives from two literary sources. The earliest surviving description of a marching-camp is that given by Polybius (6.33-7), who was writing in the middle of the second century BC. In the reign of Trajan (AD 98-117) a surveyor commonly known as Hyginus Gromaticus wrote a theoretical surveying manual (De munitionibus castrorum), which was intended to provide the appropriate accommodation for every type of army unit the student was likely to encounter. Despite being written nearly three hundred years apart both accounts are still broadly comparable, with divergences due mainly to the differing needs of an army in permanent garrison as opposed to a temporary rest camp. Archaeology and aerial reconnaissance, especially in Britain, have demonstrated that the basic principles laid down by these two writers were incorporated into the planning of fortifications from the late Republic until well into the third century AD. When the army was on campaign it constructed marching-camps to provide security at night, and, once an area was conquered, a network of turf and timber forts roughly a day's march apart. In Britannia this phase lasted until the mid-80s AD. Additionally, before the legions had established their permanent bases in Britannia, they constructed fortresses either to provide part of a legion with a summer campaign base (costal aestiva) or winter quarters (castra hiberna). Once the army was no longer poised to continue the expansion of the empire these fortresses and forts became permanent, their plan and design preserving the main defensive features of the maiching-camp from which they had evolved. The shallow ditch and palisade of the latter were, however, replaced by more substantial earthworks in permanent fortifications, often with two or more V-shaped ditches and an earth or turf rampart surmounted by a timber parapet. The four gateways were retained, but towers now defended them, and further towers were added at the four angles and at intervals between. It should be emphasised that there is no such thing as a typical Roman fortress, fort or marching-camp. The basic layout of a fortress, for instance, was A turf and timber fort, as depicted on Trajan's Column (Scene LI), showing detail of the installation's fortified gateway and angle-towers. The two buildings just inside the fort appear to be granaries. (Reproduced from Lepper, F. and Frere, S. S., Trajan's Column: A New Edition of the Chicorius Plates, Sutton, Stroud, 1988) 5 supposed to be standardised, but closer examination shows that there were considerable differences in detail between individual fortress plans, and between the same types of building at different sites. What follows, therefore, is an outline illustrating the different categories of Roman military installations. Fortresses Prior to Domitian (r. AD 81-96), fortresses were permanent bases accommodating two legions. During Domitian's reign, however, fortresses were reduced in size (c, 20-25ha) and housed only one legion, or were smaller still (c. 10-15ha) and housed either a full legion or several of its cohorts, sometimes with auxiliary troops, for a campaign. The term 'vexillation fortress' was coined by archaeologists for the latter type of site. Forts The framework of Roman occupation and control was firmly based on the fort (c. l-5ha), a permanent base accommodating an auxiliary unit. The layout of the auxiliary fort was essentially a miniature of the legionary fortress plan. A fort of the period AD 80-125 was protected by an earth rampart - encased with either timber or turf and founded upon a bed of logs or a stone base - surmounted by a spilt-timber breastwork or wattle hurdles and fronted by one or more V-shaped ditches. The rampart was pierced by four gateways, each with a timber tower above the gate passage itself or towers to either side. Further towers, set within the body of the rampart, stood at the angles as well as being spaced at regular intervals around the perimeter. The forts along Hadrian's Wall, however, were built with curtain-walls, towers and gateways of stone. Tacitus (b. c. AD 55) rightly calls the fort the 'soldiers' hearth and home' (Historiae 2.80), the objective being to provide a permanent and tolerably comfortable quarter for its garrison. As such, it compared well with the fortress of the legions. It also had to be secure against the possibility of surprise attack. However, a fort was not designed as an impregnable stronghold. On the contrary it was a jumping-off point, a base for wide-ranging activities. In wartime the enemy was engaged at close-quarters in the field, while at other times the garrison would have patrolled well beyond the frontier, either to support allied tribes or to conduct punitive campaigns. Fortlets A very much smaller installation was the fortlet (c. 1ha). Placed at intermediate points along a frontier system, along a road, or at a river crossing, these usually accommodated a century or more of an auxiliary cohort. A fortlet, unlike a fort, only had a single gate through the rampart, with a timber tower above, and one or two ditches beyond. With their towered gateways fortlets would have fulfilled a 'see and be seen' role. Watchtowers Commodus (r. AD 180-192), so as to safeguard the provincials of Mauretania Caesariensis, modern Algeria, 'built new watchtowers and repaired the old ones by the work of his soldiers' (ILS 396). Invariably only three or four metres square at the base and at least two storeys high, the term is often used indiscriminately, but is usually taken to cover small sites without significant barrack accommodation. Probably manned by a contiibernium of eight men, these installations were usually made of timber, set within a low rampart and a single or double ditch, and spaced out along a road or river to observe traffic and population movement. Marching-camps Josephus (b. AD 37) says that whenever the Romans entered hostile territory, 6 they would 'first construct their camp' (Bellum Judaicum 3.76). Marching-camps, LEFT A turf and timber fortlec on the Danube frontier, as depicted on Trajan's Column (Scene I), showing detail of the installation's fortified gateway and timber palisade. (Reproduced from Lepper, F. and Frere, S. S., Trajan's Column: A New Edition of the Chicorius Plates, Sutton. Stroud. 1988) BELOW A watchtower on the Danube frontier constructed of turf-blocks, as depicted on Trajan's Column (Scene I), although the first-floor balcony and palisade are of timber.A torch for signalling purposes projects from the upper window. (Reproduced from Lepper, F. and Frere, S. S., Trajan's Column:A New Edition of the Chicorius Plates, Sutton, Stroud, 1988) to which Josephus is referring, were overnight halts for armies or units on campaign. Each had a shallow ditch and low earth rampart constructed of material thrown up from the ditch, some 1.5 metres high, topped with a palisade made up of the two wooden stakes (pila muralia) carried by each soldier. Rather than having one end driven into the ground, it is now believed that the pila muralia were tied together in groups of three to form a kind of large 'caltrop'. There were no gates in the gateways of a marching-camp as sentries guarded them. The open gateway would normally receive additional protection from a mound with a ditch (titulus), which was built several metres to its front, or through an extension of the rampart (clavicula) that curved either outwards or inwards. These camps provided a simple measure of security for troops camped under canvas. Practice-camps Sextus Iulius Frontinus, governor of Britannia (AD 73/4-77/8) and engineer of note, wrote several technical treatises. In one he quotes with approval the maxim of Gnaeus Domitius Corbulo, a commander renowned for his realistic training methods: 'Domitius Corbulo used to say that the pick (dolabra) was the weapon with which to beat the enemy' (Strategemata 4.7.2, cf. Tacitus Annales 13.35). This can only be a reference to the proven ability of the Roman army to build camps for itself. Obviously, recruits would have to be instructed in these military techniques, whereas fully trained soldiers would have to be exercised at fairly frequent intervals so as to maintain standards. 7 A scene from Trajan's Column Britain easily provides the largest number of practice-camps in the empire, (Scene CX) showing a the most common size being around 30.5 metres square. Often a mile or two marching-camp. A number of away from the site of a fort and close to a Roman road, these sites are where leather cents, designed for a troops trained in constructing marching-camps and in particular the most cent-unit (eontubernium) of eight difficult sections of the camps, the corners and gateways. men, are clearly visible within. (Reproduced from Lepper, F. and Frere, S. S., Trajan's Column: A New Edition of the Chicorius Plates, Succon. Stroud, 1988) 8 The origins of Hadrian's Wall Chronology 55 BC and 54 BC Gaius Iulius Caesar leads two punitive expeditions to southern Britain. AD 43 Invasion and conquest of the southern part of Britain by the Emperor Claudius. AD 119 Quintus Pompeius Falco, governor of Britannia, puts down a revolt in the province. AD 122 Visit to Britannia by the Emperor Hadrian and soon afterwards work begins on the building of Hadrian's Wall. AD 142/3 Building of the Antonine Wall from the Forth to the Clyde. Hadrian's Wall is abandoned until AD 163/4. AD 208-11 The emperor Lucius Septimius Severus campaigns against the Maeatae and Caledonii. briefly reoccupying the Antonine Wall. AD 367 The so-called 'Barbarian Conspiracy' - the Picts may at this time have overrun Hadrian's Wall. c. AD 400 The Notitia Dignitatum catalogues the units commanded by the dux Britanniarum and includes a sub-section entitled 'also, along the line of the Wall' (in part/bus Occidents XL ). 32-53 AD 407-11 Gradual withdrawal of the bulk of the remaining garrison of Britannia and, according to St Gildas (De Excidio Britanniae 18), a letter was sent by the emperor Honorius urging the people to see to their own defence, c. AD 700 The Ravenna Cosmographia (107 ), which records the towns and rivers 24-30 of the Roman world, lists the Wall forts from east to west. AD 731 The Venerable Bede at Jarrow Monastery describes the Wall as 'eight feet in breadth, and twelve feet in height, in a straight line from east to west, as is clear to beholders to this day' (Historia Ecclesiastica 1.12). According to Tacitus, Britain's reduction to a province was only achieved 'gradually' (Agrkola 14.1). Indeed, some 80 years after the Claudian invasion, Roman Britain had no effective northern frontier that could be compared to the Rhine, Danube or Euphrates. Although the Stanegate, the Roman road The visible trace of the Stanegate connecting Corbridge with Carlisle, marked the northern limit of military that runs immediately south of the occupation in Britain by the reign of Trajan, it was not a frontier system. modern road leading to Chesterholm fort (Vindolanda), Since the reign of Claudius (AD 41-54), the security of the north had been looking east - note the evidence of founded on a treaty between Rome and the Brigantian queen, Cartimandua. In the pronounced camber and AD 69, however, her consort Venutius ousted Cartimandua and friendly relations drainage ditches to either side, between Rome and the Brigantes came to an abrupt end. In the cutting words of (Author's Collection) Tacitus, 'the kingdom was left to Venutius, the war to us' (Historiae 3.45). At a time of civil war in the empire, the governor, Marcus Vettius Bolanus (AD 69-71), was able to do little more than rescue the client-queen. There are hints of rather more military activity during his period of office than Tacitus reveals, but it seems highly improbable that Bolanus operated in Caledonia as the contemporary poet Statius implies (Silvae 142-9). The arrival of the new governor, Quintus Petillius Cerialis (AD 71-73/4), saw renewed activity in Brigantia. Tacitus (Agrkola 17.1) fleetingly refers to Cerialis winning bloody battles against the tribe after campaigning widely in their territory. Although he built upon the successes of his energetic predecessors, credit for the eventual 9 subjugation of northern Britain is rightly given to Gnaeus Iulius Agricola, the father-in-law of Tacitus and governor of Britannia for seven years (AD 77/8-83/4). The new governor was neither a stranger to the province nor unaware of its problems, having served there as a military tribune during the Boudican revolt (AD 60-61) and later as legate of legio XX Valeria Victrix. However, although the northern border may- have appeared to be secure Hadrian's Wall did not stand in isolation. It was built in reaction to something or someone. Hadrian (r. AD 117-138) had a general policy of defining fixed limits for the empire, but a fresh outbreak of conflict in northern Britain at the outset of his reign might have been the immediate reason why the Wall was built. A tombstone from Ferentinum in Italy names Titus Pontius Sabinus, erstwhile primus pilus of legio III Augusta, who commanded vexillationes of legiones VII Gemina, VIII Augusta and XXII Primigenia on the 'expeditone Brittanica' (ILS 2726, cf. 2735) - see the glossary on pages 62-63 for a full explanation of terms. A reference in the text to his decoration by the deified Trajan makes it probable that this expedition to Britannia occurred in the latter part of Trajan's reign or under Hadrian. For further evidence of a conflict in The principal military and civilian Britannia under Hadrian we can cite the tombstone inscription of Gaius Iulius sites of northern Roman Britain, Karus, prefect of the auxiliary unit cohors II Asturum, decorated bello Brittanico based on an original map by Guy de before his transfer to Egypt to serve as a tribune in legio III Cyrenaica (AE la Bedoyere. (© Copyright Osprey 1951.88). This revolt is also registered, albeit briefly, in a late fourth-century Publishing Limited) source as being one of the numerous troubles to afflict the new emperor, whereby 'the Britons could not be kept under Roman control' (SUA Hadrian 5.2). Further evidence is found in a letter written some 40 years after the event by Marcus Cornelius Fronto. Addressed to Marcus Aurelius (r. AD 161-180), his former pupil, Fronto consoles the emperor for the heavy losses his army had sustained in Parthia by recalling 'what a large number of soldiers were killed under your grandfather Hadrian by the Jews, what a number by the Britons' (22,2). A fragmentary tombstone from Chesterholm suggests that fighting took place precisely in the area where Hadrian was to have the Tyne-Solway system established: T • ANN[IVS...] CENTVR[IO...] TVNGR[ORVM...] ... INBELL[0... INTER]FECTVS... Titus Annius, a centurion of a legion serving as acting commander of the auxiliary unit cohors I Tungrorum at Chesterholm, may have been one of the casualties of this revolt (inbell[o ... inter]fectvs) that flared up at Hadrian's succession. As for reasons for the uprising in Britannia, a Chesterholm-Vindolanda writing-tablet offers stark evidence for what could have caused local resentment. The derisive attitude to the Brittunculi, a previously unattested word that means something like 'nasty little Brits', and the derogatory comments on the fighting qualities of the 'naked Britons' (nu[di] Brittones) contained in the letter, suggests no great sympathy for the subject population (Tab. Vindol. II 164). From the turn of the second century, this memorandum presumably refers to Britons who 10 had been recruited into the Roman army to form irregular units (numeri). The