Something must be said of the towns. They were not numerous or specially important. It must be remembered that Britain as a province was comparatively poor and unsettled as compared with Spain or Asia Minor. Its relative prosperity at this period was thrown into relief by the fact that Gaul had long been suffering from barbarian invasions; it is not improbable that there had been immigration from the Continent. To return to the towns: Camulodunum seems never to have recovered from its destruction by Boudicca. When it was rebuilt, its walls enclosed an area less than that of several other places. Glevum, and probably Lindum, were colonia; Verulam had been a municipium since the commencement of the Roman epoch. Eboracum was undoubtedly the principal place north of the Trent; Corinium, Viroconium, Calleva, Isca Silurum, and other places, had considerable local importance. Aquæ Sulis was much frequented as a health resort; but the most important city of the province from the time of Carausius onwards was undoubtedly London, which about 340 received the title of Augusta, with, in all probability, exceptional privileges. Its name was far too ancient to be ousted by a mere honorary appellation, but Londinium Augusta was certainly the largest and most important, if not the finest, of Romano-British towns. Its walls enclosed an area of some 380 acres. The average of population per acre in modern London is about sixty, but in Berlin it is one hundred. Sanitation was ill-understood in those days, and in a commercial centre like London the crowding was probably often dense. The normal population may have been about 50,000. The area of Verulam was about 203 acres; its population, perhaps, 20,000. Such importance as it still retained may have been largely due to the fact that it was a pleasant place of resort from busy and overcrowded London. Viroconium, with an area of 170 acres, and Calleva with 102, can hardly be credited with more than 10,000 and 5,000 inhabitants respectively. Population does not flock into places of no special commercial importance. Glevum, Lindum, Eboracum, and Corinium were all, perhaps, as large as Verulamium. Many of the ports on the south coast must have attained considerable size. But Britain, unlike Gaul, was not a province of great cities. Apart from resorts of merchants, such as London, the real centres of social life seem to have been the numerous villas.
Nor, apart from prosperity at certain periods, can Britain ever have been a paying province. The number of troops permanently maintained there must have approached 40,000; while Severus was in the island there were probably 100,000; in 400 the Notitia shows over 50,000. Besides the expense of the army, there was that of the vast and ever-increasing bureaucracy, and of the maintenance of roads and military works.
When the chief towns were walled we do not know. There is some reason to think that the fortification of London was begun by Severus I., but it is probable that most of the wall dates from a later period. The opinion of the authors, based upon careful examination of the quality of the work, is that it was constructed in great haste. The wall of Verulam is especially strong and massive, but whether it is so early as it is claimed to be is somewhat doubtful. Inscriptions are plentiful on the Wall of Severus, but not on those of the towns; and since the practice of commemorative inscriptions tended to die out in the fourth century, we are, perhaps, justified in supposing that the fortification of the towns was undertaken comparatively late. Everything is doubtful at present. We only know that the ‘departure of the Eagles’ left most of the British towns walled. Since at Silchester the wall crosses the ends of streets diagonally, the inference is that it was built at a late date to enclose only the closely inhabited area.
About 343 the Picts, hitherto more or less quiescent, again took to the war-path. Apparently the defences of the North had been somewhat neglected; the Wall of Severus was pierced, Corstopitum burned. The Emperor Constans came in haste from Gaul in the winter to face the danger, drove back the raiders, and appears to have received some sort of homage from them, for Julius Firmicus speaks of the Emperor as having ‘extended the Empire.’ His coercion was at any rate severe enough to impose peace for seventeen years. But in 360 troubles again broke out. The Picts renewed their raids, and for the first time we hear of a new enemy to Britain—the Scots. They were destined to give their name to the northern part of Britain in the far future, but for the present they were probably neither more nor less than adventurers from Ireland. The name may signify a ‘broken’ or landless man, though the Scots as a whole appear to have had their home in north-east Ireland. Perhaps they were a confederation of broken clans and war-bands. They crossed to Caledonia, and established a settlement in the modern Argyll. At the same time some of the Picts had effected a small settlement in Ireland and in Galloway. The two peoples were thus in close communication, and a united attack from them was what might have been expected.
About 360, then, Picts and Scots began to direct raids upon the north and west of Roman Britain. After a while they were joined by the Attacotti, who seem to have been a confederation of the Britons beyond the Roman Wall—i.e., between Tyne and Forth. For a time these raids produced only slight effect, and in 360 Britain was exporting quantities of grain to Gaul for the relief of suffering provincials there. But by 364 the invaders were growing more daring. There is reason to think that other Irish tribes were assisting them, and now the ‘Saxons’—that is Angles, Frisians, Jutes, as well as Saxons—appear once more on the scene. Ammianus Marcellinus says that they were ‘in conspiracy’—that is, were acting in unison—and this is very probable. In 367 they made a combined attack and broke up the defence by two almost simultaneous victories. The Roman Army of the North was defeated; its commander, Fullofaudes, was slain, and the force broken up and dispersed; while Nectarides, Count of the Saxon Shore, was defeated and slain by the Saxons. The results were very serious. Probably here and there detachments of troops held out behind the walls of the larger towns and fortresses; but the invaders seem to have overrun a great part of the country north of the Thames. Ammianus says that they dispersed over the country in small marauding bands.
The Emperor Valentinian I. sent to cope with this most dangerous irruption a gallant Spanish officer, Theodosius, entrusting to him large reinforcements of Teutonic mercenaries and two regiments of the Imperial Guard. Theodosius’s first care was to clear the Midlands, a task involving much rapid marching and hard fighting, but successfully carried out. He wisely did not threaten disbanded troops with military punishment, and thus was able to rally the Army of Britain on the corps which accompanied him, and to completely reorganize it. In 369 he cleared the north, and, so Claudian tells us, pursued the enemy oversea to their refuges—presumably the Irish coast and the Hebrides. Whatever exaggeration may be behind the poet’s eulogy, there is no doubt that Theodosius gained great and, for the time, decisive successes. Although the theory of a new ‘province of Valentia’ between the Walls of Hadrian and Antoninus is due to a misunderstanding of the words of Ammianus, there is some reason to think that Theodosius did in a sense advance the border. He abolished the ‘Arcani,’ a sort of frontier intelligence corps composed of border Britons, and this may imply supersession of them by advanced detachments of regulars. Secondly, we find Attacotti soon after serving in considerable numbers in the Roman army, a circumstance which seems to imply complete defeat, if not political subjection. Thirdly, while Ammianus says that Theodosius restored all the frontier forts, excavation appears to indicate that the line of the Wall with its mile-castles was not repaired. The evidence is not quite conclusive, for the remains of the last occupation lying nearest to the surface would be the first to perish. The frontier was guarded in force till forty years later. This has been contested by Mommsen, who suggested that the roll of the Army of Britain in the Notitia Dignitatum was copied from an earlier list in order to hide the chasm caused by the destruction of corps in the Picto-Scottish wars. Professor Oman has satisfactorily rebutted this theory. He points out that, though there is a remarkable survival of old regiments, yet intermingled with them are many with unquestionable fourth-century titles such as ‘The Thundering Moors,’ ‘The Senior Lions,’ and ‘The Bears of Valentinian.’ Claudian distinctly states that to meet Alaric, Stilicho withdrew troops from the North of Britain. Coins of Maximus (383-388) have been found on the Wall, proving an occupation until almost the end of the fourth century; and lacking, as there is, information of any great disaster, it cannot be asserted that the Roman hold on Britain was not effective until the last. The case of the Attacotti is suggestive, and there is evidence (some of it certainly late) that the British tribes between the walls were practically adjuncts of the province and co-operating in its defence.
The arrangements of Theodosius sufficed to ensure the safety of the province for some fourteen years. That much damage had been inflicted is certain, and perhaps Britain never entirely recovered from the effects of the invasions of 364-368. It has been suggested that Deva, Viroconium, and other towns in the west, had been destroyed by the Scots; but this seems very doubtful. A fact to be noted is that there was already a considerable Teutonic element in the island in the shape of many numeri formed out of prisoners taken in the chronic wars on the Rhine. In 371 the Emperor Valentinian sent over to Britain a whole Alemannic sub-tribe.
In 383 the Army of Britain revolted against Gratianus, the successor of Valentinian I., and proclaimed as Emperor an able Spanish general, Magnus Clemens Maximus, who held high command in the island, but had been passed over by Gratian’s ministers for promotion. The Picts and Scots seized the opportunity to renew their raids, but were repelled by Maximus, who then, however, crossed to Gaul to expel Gratian. The troops joined him, Gratian was murdered by one of his officers, and Maximus became supreme over Gaul and Spain. Gratian’s brother, Valentinian II., retained Italy for a while, but in 387 Maximus expelled him. He hoped, perhaps, to repeat the deeds of Constantine I., who had conquered the whole empire from the West, but fate decreed otherwise. In 388 Theodosius I., Emperor of the East, son of Count Theodosius, came up against him, and he was defeated at Aquileia, captured, and executed.
Gildas says that, to defend himself against Theodosius, Maximus stripped Britain of her warriors, and so paved the way for the ruin that was to come. This, however, is very doubtful, and Gildas cannot be relied upon except for his own times. But Claudian may be believed when he says that Britain suffered from Pictish and Scottish raids, though he no doubt paints his picture in the darkest colours. The ‘Historia Brittonum’ says that about 385-390 the Scots were in possession of North Wales, but that they were driven out by an army led by Cunedda and his eight sons from the land of the Otadini—i.e., the Lothians (Manau Gododin). If this statement—and it is very precise—may be taken as historical fact, it can only mean that the Otadini now formed part of the province, and that an auxiliary force led by one of their chiefs was employed to clear North Wales of the Scots. Cunedda was clearly a Romanized Briton; his father Æternus and his grandfather Paternus bear just the quaint names that were common among Romans in the fourth century. It is possible that Cunedda’s campaign was initiated by Maximus. ‘Maxim Gwledig’ (= Maximus Imperator) bulks largely in British legend, and it is permissible to suppose that there was some solid reason for the respect paid to his memory. The theory that Wales, and possibly Damnonia, were defended by their own local levies accounts satisfactorily for the fact that in the Notitia we find no regular troops stationed in those regions. It also explains the early formation of monarchical states among them, which is a feature of the next century. Finally, if a large part of the warriors of the Otadini went to Wales, we might expect to find the defence of the north weakened, and, if Claudian may be trusted, this is what did happen.
When Theodosius the Great died in 395 the Roman Empire was already sorely pressed, but for more than ten years ruin was staved off by the great Vandal Stilicho, guardian of the weak young Emperor Honorius, and Commander-in-Chief in the West. Amongst other things, he reorganized the defences of Britain. The General in the North was called the ‘Duke of the Britains’ (Dux Britanniarum). From Brancaster, in Norfolk, to Southampton Water extended the district of the Count of the Saxon Shore (Comes Littoris Saxonici). Both were under the supreme command of the Count of the Britains (Comes Britanniarum), who controlled a reserve force which could be used at need to strengthen either north or east. The VIth Legion was still at York; the IInd was now at Rutupiæ; and there were besides thirty-seven auxiliary regiments of infantry and sixteen of cavalry—nearly 60,000 men in all. There were also two naval squadrons—one stationed on the ‘Saxon Shore,’ the other off the Lancashire coast.