mail: Bill Thayer	Help	Up	Home
previous: Introduction This webpage reproduces a chapter of Celtic Britain by Nora K. Chadwick published by Frederick A. Praeger New York 1963 The text and engravings are in the public domain. next: Chapter II

 p19  Chapter I

The End of Roman Britain

The peoples speaking a Celtic language are the earliest Britons of whom we have written records. Strictly speaking the term 'Celtic' is a linguistic one, and refers to a branch of the Indo-European languages. There is no 'Celtic' race or group of tribes, or any 'Celtic' area. In general speech, however, the term has come to be used in a wider sense of the people speaking one or other of the various branches of the Celtic languages, and then, by a further extension, of their countries. This extension of meaning is a convenient one and will be adopted throughout this book.

The Celtic languages have probably been spoken in the British Isles for more than three thousand years; they were brought by newcomers, perhaps already in the Bronze Age before 1000 B.C., and reinforced in the Iron Age after 500 B.C. by a later wave of Celts, whose language in their Continental home had undergone in the interval certain sound-changes shared by Gaulish also. We generally refer to the earlier groups as 'Goedelic', the later as 'Brythonic'. The Goedelic branch of Celtic has survived down to our own time in Ireland, the Highland and Islands of Scotland, and the Isle of Man. Brythonic has survived in Wales, Cornwall and Brittany. Scholars of the last generation believed that the Goedelic languages, entering from the Continent, were first spoken in Britain and were pushed westwards by the later Brythonic speakers, likewise entering from the Continent. Today we are asking if the Goedelic branch of Celtic did not by-pass Britain and go to Ireland direct by an early sea-route from the Continent. In any case the Irish language is believed to have been  p20 introduced into the Scottish Highlands and western Britain direct from Ireland in the early years of our era, and to have become what today we call Scottish Gaelic.1

In the second century A.D. the Greek geographer Ptolemy recorded the names and positions of three peoples of Britain.​a The Cornovii occupied the far north of Scotland, Sutherland and Caithness, and the Welsh Border in what later became Powys, with its capital at Wroxeter — the Viroconium of the Romans. The name is apparently preserved in the later British name of Cornwall, and Breton Cornouaille.

The Dumnonii were in southern Scotland along the Grampians, and also in Devonshire and the Cornish peninsula and part of Somerset. Their capital was Isca Dumnoniorum (Exeter). They are also identified with an early people of the west of Ireland known in early Irish tradition as the Fir Domnann ('The Dumnonii People').

The third group were the Coritani, whose native name was Qritani, later Cruithni, and who were known by fourth-century Roman and medieval writers as Picti ('Picts'). In Roman times the Picts were a powerful political force in the whole of Scotland north of the Antonine Wall, and as the Coritani they still occupied the north-eastern English Midlands. Ptolemy gives their towns as Leicester and Lincoln. Their language is unknown except in proper names and a number of undeciphered inscriptions, though in Scotland the royal Pictish families seem to have spoken Pictish as late as the ninth century. The Cornovii and the Dumnonii are believed to have spoken some form of Celtic; the Picts, a composite language of a form of Brythonic and an earlier indigenous language.2

At the time of the Roman Conquest Britain herself was divided into a large number of independent Celtic kingdoms, each ruled by its own royal house. In Scotland the kingdom or kingdoms of the Picts survived north of the Forth-Clyde isthmus; the Dumnonii in the south-west, and the kingdom or  p22 'tribe' of the Votadini in the south-east. Most of the north of England belonged to the largest tribe of all, the Brigantes, but in eastern Yorkshire were the Parisii. East Anglia was in the hands of the Iceni, the Coritani still held the Midlands, the Cornovii the north Welsh Border. The remaining tribes occupied positions indicated on the map overleaf.

Fig. 1. Political map of Roman Britain. The tribal boundaries are largely hypothetical (after Rivet).

Already c. 110‑100 B.C. a series of movements of mixed tribes known as Belgae from France and Belgium began to enter south-eastern England, and extended their overlord­ship to include Hertfordshire and Essex, where the tribe of the Catuvellauni were the dominant element. From this last area the Catuvellauni extended their supremacy under their king Cunobelin in the first half of the first century A.D. over the greater part of south-eastern Britain. The Belgae now formed the most advanced element in the population, and it was they who, towards the end of the second century B.C., introduced the first native coinage into Britain on the model of the coins of Belgic Gaul.​3 This movement of the Belgae into Britain was more in the nature of an expansion than a migration or a conquest. The in‑comers entered as tribal units, retaining ties with their Continental relatives, and by the time of the Roman Conquest trade relations between Britain and Gaul had been active for a long time. All the main Belgic areas had already become in part Romanised.

The Roman Conquest did not change the population or the Celtic tribal units, nor the language of the people as a whole. The history of the Occupation and final withdrawal suggests that the Celtic way of life, the economy of the country, continued more or less unbroken apart from a sharp increase of civilisation in the part of the country fa­cing the Continent in the south-east. Here towns and villas, in which Latin was spoken to some extent, were added to our map. True architecture, monumental sculpture, fresco painting, mosaic work were new arts brought from Rome, and amenities such as plumbing  p23 with a steady water supply were introduced. The official religion was Roman, first the Roman pantheon, then Christianity. But the native Celtic religion lingered on unhindered, and was even encouraged by the Romans, and fused with Roman cults; to the last, Britain remained a remote province of the Roman Empire, habitually referred to as 'a country of the setting sun, remote from our world'.

The Roman Conquest had been more like a flood than an upheaval. Caesar's two successive landings in 55 and 54 B.C. had been the first hint of the in-coming tide. The full flood tide set in in A.D. 43 with the conquest planned by Claudius. After A.D. 70 Wales had become virtually an armed outpost of the empire. The town of Caerwent, 8 miles from Caerleon, was the only major Roman civil settlement in Wales.

The revolt of the Iceni under Boudicca (Boadicea) in A.D. 61 did not stem the flood. In A.D. 71 began a northern campaign which was to extend the Roman Province over the Lowlands of Scotland. The legionary fortress of Inchtuthil, built in Perthshire under Agricola, lasted for a few years, and other Roman forts were built as far north as Kintore on the River Don; but from now onwards the waters began to ebb. Soon after A.D. 100 Roman forward aggression in Scotland came to an end. In Trajan's reign the advanced garrisons in Scotland withdrew. Scotland as a whole remained an unconquered country.

A powerful rebellion in Scotland and northern England brought the Emperor Hadrian to Britain in 122; Scotland was temporarily abandoned and the southern frontier was secured by the building of the Wall from Wallsend-on-Tyne to Bowness on the Solway. This frontier was advanced forty years later by the building of the turf wall from Forth to Clyde by Hadrian's successor, Antoninus Pius. Nevertheless, general Celtic risings in the Scottish Lowlands and the north of England in 155 and 181 destroyed the walls and almost all the forts, and in 196 a great part of Britain was overrun. The  p24 situation was recovered and the Wall and forts repaired, and in 208 Septimius Severus subdued the Caledonii of the Perthshire Highlands and the Picts of Strathmore and Strathearn in the Lowlands; but the Antonine Wall was no longer held as the northern frontier. The British tribes between the two walls were left to guard themselves and the Province against further incursions of the northern tribes, and it is a measure of their success that in the third century and especially during the fourth, civil Roman Britain enjoyed her most prosperous period the period of the country villas and flourishing, if changed, town-life (see ch. II).

But while Britain was enjoying the halcyon peace secured by the measures of Severus and by our northern British tribes against the Picts, the Romans of the Continent were passing through a heavy ordeal. The barbarian attacks on the northern frontiers were in full swing, and Gaul was in rebellion. How long would the Romans be able to spare troops to retain her remote western province? What would be the future of the Celtic peoples of Britain if the Roman forces withdrew?

The protracted Occupation of Britain and the forty years of peace, combined with the gradual weakening of Roman power, had brought about a great change in the relations of the Celtic peoples of Britain and their Roman rulers. The British chiefs of southern Scotland were fully prepared to co‑operate with the Romans to protect the northern frontiers against their mutual enemies the Picts. Also, in the late third century the relations between Roman and Britain in Wales changed radically.

About 275, serious Irish raids began in south-west Wales. An Irish dynasty and an Irish aristocracy settled in Pembroke and an Irish settlement from Leinster occupied western Caernarvonshire. Under this western pressure the Romans and the Britons of Wales found themselves no longer at enmity with one another but fa­cing a common foe. The chief Roman defences in eastern Wales had been the legionary fortresses of  p25 Caerleon and Chester, directed against the Welsh; part of the second legion was now transferred from Caerleon to Richborough in Kent, and a new fort was built at Cardiff against the Irish. In North Wales a fort at Holyhead, known as Caergybi, was built at this period as a protection against Irish raiders, and at Segontium (Caernarvon) at the mouth of the river a new fort was built repla­cing the earlier one; it was set on lower ground by the shore, in order to protect the trade of the valuable Anglesey copper mines from the same danger.​4 Nor was Morecambe Bay exempt from the raiders. At Lancaster a fort has recently been excavated, which had been built against danger from this quarter, and large forts at Piercebridge and Elslack were built to protect the area east of the Pennines from penetration by raiders from the rear.5

Simultaneously with the Irish encroachments on the west, serious Pictish raids were also taking place in the north. The usurper Carausius had reconditioned the fleet,​6 and his successor Constantius (293‑306) had built a new fleet and carried out defences in the north and west, restoring Hadrian's Wall, and building massive new foundations at High Rochester and the great multi-angular tower at York. The great castle-like forts round the coast from Brancaster in Norfolk to Portchester in Hampshire, each garrisoned by an auxiliary regiment, and commonly known as the 'Forts of the Saxon Shore', are now thought to have been built perhaps by Carausius and Allectus to protect themselves, not, primarily, against the Saxons, but against attack from the sea by Roman authorities during the usurpation.​7 The corresponding forts against the Saxons along the French coast and on some of the islands, e.g. Alderney, were much less massive. But the great British forts were undoubtedly used later as part of the system of defences against the Saxons, when their raids in Britain began. The great activity and expert training in massive stone building occasioned by all this defence work would account for the fact that Constantius  p26 seems to have supplied Gaul with masons from Britain to repair Gaulish ruined buildings, as we learn from the brilliant declamation of Eumenius of Autun in 298 on the subject of the restoration of the University of Autun.8

Fig. 2. The British Saxon shore forts (after White). The Notitia Dignitatum lists nine castella as being under the command of the Count of the Saxon Shore. They are: Branodunum (Brancaster), Gariannonum (Burgh Castle), Othona (Bradwell), Regulbium (Reculver), Rutupiae (Richborough), Dubris (Dover), Lemanis (Lympne), Anderida (Pevensey), Portus Adurni (Portchester)

Constantius was the idol of the Gaulish panegyric poets, not only in his own day but in that of his son and successor Constantine — as well he might be. For the danger to Britain was Gaul's danger also, and on sea and land he had won for both a respite from the marauding pirates; from the powerful Pictish fleet of the far north, no less than from the daring Saxon and Frisian fleets rapidly supplanting it in southern waters. Both Vegetius in the fourth century and Gildas in the sixth speak of sea raids from the north. In an anonymous prose panegyric (formerly erroneously attributed to Eumenius), probably delivered at Trier in 297 before Constantius himself on his victorious return from his British expedition, the speaker contrasts the easy victory that Julius Caesar had when 'this people were still primitive and accustomed only to fight the Picts and the half-naked Irish'.9

Fig. 3. The Gallic Saxon shore forts (after White). The modern names of the sites marked on the map in their original Latin names are as follows: Blaye (Blabia), Nantes (Manatias? Namnetibus), Vannes (Venetis), Aleth or Alet at St Servan/St Malo (Aleto), Avranches (Abrincates), Coutances (Constantia), Port-en‑Bessin (?Grannona), Rouen (Rotomagus), and Boulogne (Bononia). Osismiis may be Carhaix or Brest, Marcis may be Marck, Marquise or Mardyck; the headquarters of the Praefectus classis Sambricae at Hornensi may have been at Etaples, Le Crotoy or Cap Hornu

The menace of the Saxon pirates, 'the most dangerous of all foes', is vividly brought home to us in the fifth century by a letter from Sidonius Apollinaris to his friend the Admiral Namatius, patrolling the Straits of Dover: 'If he pursues he catches . . . He has no fear of shipwreck, it merely exercises him . . . In the hope of making a surprise attack he cheerfully risks his life amid rough seas and sharp rocks.'10

All protective measures ultimately failed to save Roman Britain from the barbarian forces gathering strength around the borders. In a conspiratio barbarica​11 in 367 Irish raiders from the west, Picts from the north, and Saxons from the east simultaneously made a swift and devastating raid over a large part of the country penetrating far inland. The dux Britanniarum, the official in charge of the northern frontier defences, was over­powered, and the commander of the Saxon shore killed.

 p29  The results to the prosperity and organisation of the country were serious for property, agriculture and commerce; but its effects must not be exaggerated. It was possible for the elaborate and wealthy pilgrimage temple at Lydney to be built during the succeeding period. Many towns such as Carlisle, York, Chester, and Leicester survived, while Verulamium also continued. The villas were not all destroyed, and in some instances, as at Langton in East Yorkshire, the damage was repaired. At East Denton in Lincolnshire an entirely new villa was built after the raid. The villa in the open country near Great Casterton in Rutland was built with a mosaic pavement and occupied for the first time in the late fourth or early fifth century — surely a sign of confidence. It is now suspected that the ultimate cessation of the villa system was due less to the insecurity of the country districts than to the general deterioration of trade owing to the disorganisation of world commerce on which they had depended.

Yet the end was gradually closing in. A brief respite was gained and security recovered by Theodosius. Hadrian's Wall was repaired, but the advanced posts were abandoned. Signal stations were set up along the Yorkshire coasts to give a warning of approaching raiding parties, but they lasted for only a brief twenty-five years before they were again destroyed by Saxon raiders in 390 and the raids continued relentlessly.

The Roman position was still further weakened by the action of Magnus Maximus, a Roman official of Spanish origin stationed in Britain, possibly either a governor or a legionary officer. His service while in Britain must have been remarkable, for the Anonymous Gaulish Chronicle states (s.a. 382) 'Maximus strenuously overcame the Picts and Scots' (i.e. the Irish); but in 383 his troops acclaimed him Emperor and crossed with him to Gaul. The nature or number of the troops which went with him is unknown. He did not remove the troops guarding the Saxon shore forts or the signal stations,  p30 and there is no clear evidence that he removed the troops from the wall.​12 His palatini, or personal bodyguard, are doubtless identical with the Seguntienses, troops referred to in the Notitia Dignitatum as garrisoning places in the Balkans, near Aquileia where he was killed by his rival Gratian in 388.13

The great General, Stilicho, arrived in 395 and reorganised the defence system, but once more withdrew the northern border to a line further south, apparently now based on York, leaving the northern British still more responsible for the defences against the Picts. And again the gratitude of the Gaulish panegyrist is eloquent contemporary testimony to the success of Stilicho's work in Britain; for in 399 or possibly 400, the poet Claudian recited his praise of Stilicho's consul­ship, picturing the grateful nations coming in procession to Roma's temple to give thanks for their delivery from their foes, and among them came Britain, 'clad in the skin of some Caledonian beast'.14

And in his poem against Eutropius he again reminds his listeners of Stilicho's work in Britain.

But his confidence was grievously misplaced. We are on the eve, not only of the end of Roman Britain, but of a world crisis, and the fate of Britain is only the outer ripple of the tidal wave. In 406 a mixed body of barbarian peoples crossed the frozen Rhine, annihilating the frontier troops, and poured over Gaul. In 409 or 410 Rome, 'the mistress of the world', fell to the Goth, Alaric, and was sacked. The contemporary Anonymous Gaulish Chronicle records a specially heavy raid on Britain in the same year.

In 407 a soldier, Constantine, was acclaimed Emperor by the army in Britain, possibly against his will, and apparently by a friendly arrangement with Honorius. First taking such  p31 measures for the defence of Britain as he was able, he crossed to Gaul as Emperor, taking British troops with him to the Rhine to help the Continental army. He was subsequently defeated in battle and slain, and Procopius, our informant, adds, 'Notwithstanding this the Romans were never able to recover Britain which henceforth continued to be ruled by usurpers.'​16 Zosimus states that at this point the Britons and some of the Gauls (doubtless including the people of Armorica) seceded from Rome, took up arms, and struggling bravely on their own behalf freed themselves from the onslaughts of the barbarians.17

During this closing phase of the Occupation some new officials appear in the records of affairs in Britain, probably connected with the defensive measures taken on her behalf by the usurper Constantine on the eve of his departure. Here our chief guide is the Notitia Dignitatum. This document enables us to watch a process of devolution at work in Britain analogous in many respects to that which had already taken place on the Continent since the reform measures of the Roman Army by Diocletian (286‑305) and Constantine (305‑307). Briefly stated, this process entails the withdrawal and supplanting of the Roman sedentary troops massed on the frontier, known as limitanei (L. limes, a 'frontier'), by a local militia, consisting of foederati or federate native troops, and the simultaneous transformation of the limitanei into comitatenses, the mobile and more privileged reserves behind the limitanei. Naturally, transference from the ranks of the limitanei to those of the comitatenses was much coveted, and as the length of our land frontiers caused a high number of limitanei to be employed in Britain, an ambitious general could always offer to take them overseas and transform them into comitatenses.

The official in charge of the limitanei was the dux, and in Britain his was an old office functioning in the north. From the Notitia Dignitatum it would seem that he had sent in no recent 'returns' to the Chancellery, and this suggests that there was no  p32 longer a dux on the Wall, and is perhaps a further indication of the substitution of native British foederati for Roman troops on the northern frontiers, following the system already adopted on the Continent.

A second Roman official of some importance who appears about this time is the comes littoris Saxonici, 'The Count of the Saxon shore', who is believed to have been responsible for the manning and provisioning of the Saxon shore forts, each of which was garrisoned by an auxiliary regiment. The title first occurs in the Notitia Dignitatum (Occidentalis XXVIII). We have no reference to any official as comes, 'Count', in Britain before the mention by Ammianus Marcellinus (XXVII.VIII.1) of a comes maritimi tractus in connection with the great raid of 367; but title appears later among a small group created by Constantine (407‑411) as part of his defence measures for Britain before he left.18

If the 'Saxon Shore' defences were in existence at an earlier date they must have been under another command, which may have been wider, and included the forts and fleets on both sides of the Channel. Carausius seems to have held a command on the Gaulish side of the coast before his usurpation. At some later period the Gaulish forts may have been transferred to a wider command extending along the whole coast; but the history of the Gaulish fleet at this time is very obscure, and excavation of the forts is badly needed. For practical purposes the Saxon Shore troops would count as limitanei. They might thus come gradually under native command; but no doubt this would be at a late stage.

The most interesting and the most obscure of these new officials is the comes Britanniarum, or Britanniae, who appears at this time in the Notitia Dignitatum (Occidentalis I.V and viii), and whose office is believed to be quite a new creation, perhaps the command of a mobile field army mainly of cavalry, but partly of infantry units. It may have been built up of such  p33 remnants of the limitanei of the dux of the north, and of the comes littoris Saxonici of east and south, as did not accompany Constantine to the Continent.19

In 410 came the famous rescript of Honorius, informing the cities in Britain that they may look after themselves.​20 This has been variously interpreted, but coinciding as the date does with the fall of Rome, there can be little doubt that it is simply an official rescind of the Roman law prohibiting barbarians from bearing arms. What, then, is the date of the end of Roman Britain? Perhaps the question is hardly a valid one, for we have been following, not a single crisis, nor even an event or series of events, but an ordered process. The Notitia Dignitatum enables us to watch the gradual transformation of Roman military organisation and of Roman frontier defences into native British organisation and defences. As to when the process was completed, opinions have varied by as much as fifty years. The late Martin Charlesworth​21 held that the military end came in 407 with the removal of the troops by Constantine, but that the civil administration went on till 418, the time of the council of the Seven Provinces at Arles. A date of c. 415 is favoured by most recent scholars, with possibly a brief re-occupation in a very limited area in the south-east.22

According to the doubtful testimony of Gildas (ch. 20), the Britons sent a letter to the Roman Consul Aetius in his third consul­ship, which fell in 466, expressing 'the groans of the Britons', and their sufferings under the depredations of the Picts and Scots. But no help came. By the middle of the fifth century the four centuries of the Roman Occupation had passed like a dream. From the time when the last Roman ferry weighed anchor, carrying the last of her troops to Gaul, till the foundation of the Saxon kingdoms in the sixth and seventh centuries and the spread of the Saxon conquests to the Highland line and the Welsh Border, to the Tay and the Severn, the whole of Britain was governed by her native Celtic princes.

 p34  It was in this period of freedom between the Roman and the Saxon Occupations that the deals and the literature took shape which still characterise the Celtic peoples wherever the Celtic languages are spoken.

The Author's Notes:

1 Jackson, P. B. A.

❦

2 Jackson in Wainwright, P. P., Appendix.

❦

3 D. F. Allen, FSA, 'The Origins of Coinage in Britain: A Reappraisal' in Problems of the Iron Age in Southern Britain, edited by S. S. Frere, published as Occasional Paper II (1960) U. L. I. A.

❦

4 Richmond, R. B., 155.

❦

5 Ibid., R. N., 113.

❦

6 Atkinson, C. B., 7.

❦

7 White, L. S., passim.

❦

8 Panégyriques Latins, ed. Galletier, Vol. I, 122.

❦

9 Ibid., 91.

❦

10 Epistola VIII, vi.

❦

11 Ammianus Marcellinus, Rerum Gestarum Libri XXVII, viii.1.

❦

12 See Charlesworth, L. P., 26; Collingwood and Myres, 287.

❦

13 The identification was made by Stevens, É.C. III (1938) 86.

❦

14 Poems of Claudian, Panegyric on Stilicho's Consulship II.247.

❦

15 Ibid., In Eutropium I.392.

❦

16 Vandal War III.II.

❦

17 Historia  VI.v, vi, x, Zosimus derived his information about events in Britain from a Byzantine historian, Olympiodorus, who dealt mainly with western events. He was an exact contemporary of the last days of the Roman Occupation of Britain, and is of a high degree of authenticity. See also Thomson, Antiquity XXX (1956), 163‑7.

❦

18 Eutropius IX, XXI.

❦

19 For the preceding suggestions relating to these late officials I am indebted to the invaluable article by Stevens on the Notitia Dignitatum (cf. p167 below).

❦

20 Zosimus, Historia VI.v.3 and 10.

❦

21 L. P., 34.

❦

22 See references cited by Chadwick, S. E. B. H., 11, note; and cf. more recently Richmond, R. B., 185.

Thayer's Note:

a Geography, II.2.

◂ previous						next ▸
Images with borders lead to more information. The thicker the border, the more information. (Details here.)
UP TO:		Chadwick's Celtic Britain	British History	Home
A page or image on this site is in the public domain ONLY if its URL has a total of one *asterisk. If the URL has two **asterisks, the item is copyright someone else, and used by permission or fair use. If the URL has none the item is © Bill Thayer. See my copyright page for details and contact information.

Accessibility