The first of these insular Emperors after the war of Theodosius was Clemens Maximus, an Iberian or Spaniard by birth, who had served under Theodosius in Britain, and was now, twelve years later, in command of the Roman army there. Taking advantage of the unpopularity of the Emperor Gratian with the army owing to favour shown to the Alans, and jealous of the elevation of the younger Theodosius to a share in the Empire, he excited the army in Britain to revolt, and was proclaimed Emperor in the year 383. In the following year he repressed the incursions of the Picts and Scots,[96] and forced the hostile nations to yield to his power. He then crossed over to Gaul with the army of Britain, slew the Emperor Gratian, and after maintaining himself in Gaul for four years, he entered Italy, and was finally defeated and slain by the Emperor Theodosius at Aquileia, in the year 388.
The withdrawal of the Roman troops from Britain by Maximus left the province exposed to the incursions of their old enemies, and the two nations of the Picts and Scots—the one from the north, where the regions beyond the Forth and Clyde formed their seat—the other from the west, where lay the island of Ierne, whence they proceeded—continued to harass the provincial Britons for many years with their piratical incursions, which they were the less able to resist as the usurper Maximus had drained the province of the young and active men who could be trained as soldiers, as well as withdrawn the army.
The Britons at length applied to Stilicho, the minister of the young Emperor Honorius, and a legion was sent to Britain, which, for the time, drove back the invading tribes, and garrisoned the wall between the Forth and the Clyde. The recovery of the territory at the northern frontier was on this occasion, as well as when Theodosius repelled the invaders from it, followed by a part of the nation of the Attacotts being enrolled in the Roman army, where they bore the name of Honoriani in honour of the Emperor Honorius. The Roman historians affording us but little information regarding these renewed incursions of the Picts and Scots, their place is now supplied by the British historians Gildas and Nennius; while the allusions to these events in the poems of Claudian enable us to assign the somewhat vague and undated accounts of the British historians to their true period. They tell us of this irruption of the Picts and Scots, and of the arrival of the legion to the assistance of the Britons. The poet Claudian connects this with the name of Stilicho. He alludes to the legion which bridled the Scot, or the Saxon. He describes it as guarding the frontier of Britain, as bridling the Scot, and examining, on the body of the dying Pict, the figures punctured with iron. He depicts Britain as saying that Stilicho had fortified her by a wall against the neighbouring nations, and that she neither feared the Scots crossing from Ierne, nor the Pict nor the Saxon ravaging her coasts.[97] This fixes the date of the expulsion of the Barbarians and arrival of the legion at the year 400, and Stilicho appears on this occasion to have also enrolled bodies of Attacotts in the Roman army.[98]
Four years later the legion was recalled from Britain in consequence of the Gothic war and the attacks of Alaric, and left the island, having, as we are informed by Nennius, appointed a leader to command the Britons. They had no sooner gone, however, than the old enemies of the provincial Britons—the Picts and Scots—again broke into the province and renewed their ravages.
After three years, Stilicho sent assistance to them. He appears to have feared the total loss of Britain to the Romans, and, apparently desirous to make a great effort for its permanent recovery on this occasion, he restored the army of Britain to its usual strength, consisting of three legions—the second, the sixth, and the twentieth—by whom the province was effectually freed from the invaders and garrisoned by Roman troops.[99] As long as this army remained in Britain, the province was protected in its full extent to its frontier at the Firths of Forth and Clyde; but the position of the army, as indicated in the Notitia Imperii, sufficiently shows the imminence of the danger which now threatened the province in Britain, and the quarter from whence it was dreaded. The three legions which now protected the frontiers of this distant portion of the Empire, in the last notice which we have regarding the Roman troops in Britain, are found stationed in greatest force along the wall which extended from the Tyne to the Solway, and in the garrisons between that barrier and the Humber, and likewise in those that protected what was now termed the Saxon shore, extending from the Wash to Portsmouth. The ‘Comes Britanniarum’ guarded the western frontier of the two Britains, where the new province of Valentia had probably been formed, with troops which may have been stationed at Caerleon and Chester, the old headquarters of the second and twentieth legions, and the interior of the country is comparatively ungarrisoned.
The doom of this great Empire was now, however, rapidly approaching, and the withdrawal of the troops from the remote frontiers to protect the seat of power precipitated the fate of the frontier provinces. The great invasion of the Vandals with the Alani and Suevi, which took place in the year 406, and was the first of those fatal inroads of the Barbarians into the very heart of the Empire which led to its final ruin, alarmed the troops which remained of the army in Britain, who, on the irruption of the Barbarians into Gaul, found that they would be cut off from the other forces of the Empire and exposed alone in their insular position to the attacks of the enemy, and led them to resort to the step which had now become the habitual tendency of a Roman army so placed—to proclaim an Emperor. Accordingly they terminated their four years’ residence in Britain by revolting, and selected Marcus as Emperor. He was soon slain by Gratianus, who assumed the imperial authority, and after a four months’ enjoyment of it, was in his turn slain by the soldiers.
A soldier named Constantine was then chosen, owing his elevation mainly to his name being that of the celebrated Emperor; and this new Constantine no sooner assumed the purple, than, with the fatal policy of his predecessors, he resolved to strike a blow for the possession of Gaul, and Spain likewise. Before withdrawing the troops from Britain, however, he counselled the provincial Britons to abandon the districts between the walls, a territory now barely and with difficulty maintained by them, and to protect the remainder of the province by maintaining garrisons on the southern wall. At the same time the valleys on the north side of the Solway Firth appear to have been protected by an earthen rampart and fosse, which extends from the shore of the firth opposite the western termination of the wall across the upper part of the valleys till it terminates at Loch Ryan. On the south coast, where the province had been exposed to the piratical descents of the Saxons, and had hitherto been protected by the Roman vessels, he erected towers at stated intervals. Having thus taken the best measures in his power to enable the provincial Britons to protect the province, Constantine crossed over to Gaul with the army, and the Roman legions left Britain, never again to return. They had no sooner been withdrawn, than the old enemies of the province occupied the district as far as the southern wall to which Constantine had withdrawn the frontier; but although the Roman troops had left the island, the civil government of the Romans still remained in force, and the provinces of Britain continued to form an integral part of the Empire. The events, however, connected with the usurpation of Constantine speedily led to the termination of the Roman government in Britain, and its final separation from the Empire. Constantine had no sooner landed in Gaul than an engagement took place between the British army and the Barbarians who had entered Gaul by the passes of the Alps, in which the former were successful, and a great slaughter of the enemy took place. The Roman troops in Gaul submitted to Constantine, and he thus obtained possession of the whole of that country. In the meantime, intelligence having reached Rome of Constantine’s successful usurpation, and that the provinces of Gaul had become subject to him, Stilicho returned to Rome from Ravenna, and sent Sarus in command of an army against him. Justinian, one of Constantine’s generals, was encountered and slain. Neviogastes, another, was put to death by treachery; and Sarus proceeded to besiege Valentia, where Constantine then was. The usurper now appointed Edovinchus, and Gerontius a native of Britain, his generals; and Sarus, dreading their military reputation, retreated from Valentia, which he had invested for seven days. The new generals followed and attacked him, and it was with difficulty he reached the Alps and escaped into Italy, having had to bribe the ‘Bagaudæ,’ or armed peasantry, who were in possession of the passes, by giving up to them the whole of his booty to permit his army to pass through.
Constantine now placed garrisons in the passes of the Alps, and likewise secured the Rhine, in order to protect the territory he had acquired from invasion. Being now in undisturbed possession of Gaul, he created his eldest son Constans, who had been a monk, Cæsar, and sent him into Spain to wrest that country likewise from the government of Honorius. Constans proceeded accordingly to Spain, having Terentius as his general, and Apollinarius as prefect of the Prætorium, and was encountered by the relatives of Honorius who commanded there, and who surrendered to him after a battle in which Constans had the advantage, and an unsuccessful attempt to destroy him by arming the peasantry. Having thus become possessed of two of the relations of the Emperor—Verinianus and Didymus—Constantine sent messengers to Honorius entreating forgiveness for having allowed himself to accept the Empire, and stating that it had been forced upon him by the soldiery. The Emperor was in no position to contend with Constantine, and being afraid of the fate of his relations, acceded to his request and admitted him to a share in the imperial authority.
Constans in the meantime returned from Spain, bringing with him Verinianus and Didymus, having left there Gerontius, the Briton, as general, with the troops from Gaul, part of which consisted of the British nation of the Attacotts, who had been enlisted in the Roman army by Stilicho,[100] to guard the passes through the Pyrenees. The unfortunate relatives of Honorius were no sooner brought before Constantine than they were put to death, and an embassy was sent to Honorius in the person of Jovius, a distinguished orator, to excuse the death of his relatives, and to request that the peace might be confirmed. The plea was, that they had been put to death without his consent. Jovius prevailed with Honorius by pointing out to him that he was in no condition to act otherwise, and by promising him assistance from Constantine’s army in quelling commotions in Italy and Rome.
Termination of Roman Empire in Britain.
Constans had, in the meantime, been sent back to Spain, and took with him Justus as his general. This gave great offence to Gerontius the Briton, who probably only waited for a pretext to endeavour to overturn the government of Constantine; and, having gained over the soldiers in Spain, who, being principally Attacotts, were probably more accessible to the influence of their countrymen, he incited the Barbarians in Gaul to revolt, and invited those beyond the Rhine to enter the provinces. The latter ravaged them at pleasure, the main attack having been upon those of Britain. This took place in the year 409, and that part of the Barbarians who were thus invited and encouraged to attack the provinces of Britain were, we know from other sources, their old enemies, the Picts, Scots, and Saxons. The civil government of the Romans still continued in Britain, but Honorius, being unable to afford them assistance, wrote letters in the following year to the cities in Britain, urging them to look after their own safety. This was equivalent to an abandonment of the imperial authority over Britain; and the provincial Britons, who, no doubt in common with the inhabitants of the other provinces, groaned under the intolerable weight of the Roman civil government, rose against them, and having, by one unanimous and vigorous effort, freed their cities from the invading Barbarians, drove out the Roman prefects likewise, and shook off the Roman yoke.
In the following year Honorius, finding that the existence of the opposing tyrants, Constantine and Gerontius, had prevented him from opposing the Barbarians, and led to the defection of Britain and Armorica, resolved to make an effort for their destruction, and sent Constantius into Gaul with an army, who shut Constantine into the town of Arles, took it, and slew him. Gerontius, at the same time, no doubt aiming at the possession of Britain for himself, followed up his proceedings by slaying Constans at Vienne, and setting up Maximus, said by one author to have been his son, in his place. Gerontius was shortly after slain by his own soldiers, and Maximus, stripped of the purple, fled into exile among the Barbarians in Spain. The death of Gerontius thus prevented him from reaping the fruit of his designs, whatever his object in precipitating the Barbarians again upon the provinces of Britain may have been.
No attempt was made to recover Britain. It no longer formed a portion of the Roman Empire, and the Roman legions never returned to it. This great and momentous change in the political and social condition of the island took place in the year 410; and thus terminated the Roman dominion in that island, which, for good or for evil, had so long endured, and so powerfully influenced the fortunes of its inhabitants.
Such is the narrative of the Roman occupation, so far as it affected the northern portion of the island; such the knowledge the Romans had attained, and the record their historians have left us, compressed in few facts, and accompanied by meagre details of the position, character, and habits of the northern tribes occupying the barren regions of Caledonia, who, though often assailed, and sometimes with temporary success, preserved their independence, and remained in hostility to the Roman government throughout the whole period of their dominion in the island.[101]
55. The author has felt himself obliged to enter somewhat into detail regarding the Roman geography of Scotland, as the subject has been so much perverted by our best writers, owing to their unfortunate adoption of the spurious work of Richard of Cirencester. It is time that the credit of Ptolemy should be restored, and it is impossible for any one to compare his statements with the actual face of the country, without being struck with their general accuracy. Between the Solway and the Tay the country is distorted and the distances thrown out of proportion by the unfortunate mistake which turned the north of Scotland to the east. The effect is to increase some of the distances to a little more than double their proper length, and proportionally to diminish others. The whole country being placed in too northern a latitude does not affect the distances, and the smaller degree of longitude would be taken into account in laying down the positions; but it must be kept in mind that Ptolemy uses no smaller division than one-eighth of a degree, giving a possible variation to each place of seven miles in one direction and five in another. Taking all this into account, however, the distances between the leading features of the country, which it is impossible to mistake, are wonderfully correct.
The Latin editions of Ptolemy are the earliest, and are greatly to be preferred to the Greek. They were printed from a translation into Latin by Jacob Angelus, and consist of an edition at Bologna bearing the date 1462, but which is believed not to be the true date; one at Vincenza in 1475, which is really the earliest edition; that of Rome, 1478, the first with maps; Ulm, 1482 and 1486; Rome, 1490 and 1507; Venice, 1511; Strasburg, 1513, 1520, and 1522, in which the text is compared with an old Greek MS.; an edition in 1525, which bears to be a correction of Jacob Angelus’s translation by I. de Regiomonte, and in which the principal changes introduced into the later Greek editions first make their appearance; and the two editions by Servetus in 1535 and 1541. The principal Greek editions are those of Erasmus in 1533 and 1546, Montanus in 1605, and Bertius in 1619. A recent edition has appeared, by Dr. F. G. Wilberg, in 1838, from a collation of nine mss. with the editions of 1482, 1513, 1533, and 1535, and with a ms. at Milan, another at Vienna, and two Latin mss. collated by Mannert. The author has himself collated for this work the Latin editions of 1482, 1486, 1520, 1522, 1525, 1535, with the Greek editions of 1605 and 1619, and with Wilberg’s edition, and he agrees with Mannert in giving the preference among the early editions to the Ulm edition of 1482, and the Strasburg editions of 1520 and 1522. In the so-called corrected edition of 1525 he has no confidence. The variations occur both in the names and in the latitudes and longitudes. In cases where all the editions agree, there can be no doubt as to the genuine text used. When they differ, he has laid down the positions according to the variant readings, and selected the one which best corresponded with the appearance of the country. The agreement is mainly in the position of the towns, and the variations in the features of the coast, and are, therefore, more easily corrected.
56. The Vedra might more naturally be supposed to be the Tyne, but an altar found at Chester-le-Street, on the Wear, on which the name Vadri occurs, indicates the Wear as the river, to which indeed the name bears a greater resemblance. There is no variation in the position of these two places.
57. The early Latin editions have, instead of Ienae aestuarium, Fines aestus. It is possible that this may be the correct reading, and that Wigtown Bay may have marked the utmost limit to which the Roman troops penetrated in Agricola’s second campaign.
58. The early Latin editions all read Leva. The edition of 1525 first altered it to Deva, and is followed by the late editions, and also by Wilberg; but the distance both from the Firth of Tay in the south and from Kinnaird’s Head corresponds more exactly with the mouth of the North Esk than with that of the river Dee.
59. The editions of Ptolemy all vary as to the situation of Loxa. The Ulm editions place it after the Varar aestus at Lossiemouth; the Strasburg editions at the mouth of the Nairn; while Wilberg’s edition places it before the Varar, at the Dornoch Firth. The Ulm reading is here preferred from the resemblance of Loxa to Lossie. The reading which places it north of the Varar seems inadmissible, as it is described by Ptolemy as the mouth of a river, and not an estuary or a bay, such as the Dornoch or Cromarty Firths would be described.
60. This has generally been supposed to be Loch Fine, in the usual random way of selecting the first large loch near about that part of the coast, but the position corresponds much more nearly with that of Loch Long. Its distance from the promontory of Kintyre is too great, and its vicinity to the Clyde too marked, for Loch Fine. The name, moreover, has clearly reference to the neighbouring Lake of Lomond, and the district of Lennox, the old name of which was Leamhan.
61. In the same loose way the Linnhe Loch is usually supposed to be meant by the Longus Fluvius, but it is impossible to suppose that a great arm of the sea—the greatest on the west coast—could be expressed by the word Fluvius. The editions give two different readings of the position, but that of all the editions, except Wilberg’s, corresponds with the mouth of the river Add.
62. Epidium has generally been identified with the island of Isla, from the natural enough inference that its name connects it with the Epidium promontorium, and consequently historians have been much at a loss where to look for the two Ebudas, and have resorted to mere conjecture. The Epidii seem, however, to have occupied Lorn as well as Kintyre, and the name would be appropriate to any island on that coast. Ptolemy places the two Ebudas close together, and makes them the most westerly of the group, while Maleus or Mull is the most northerly, placing it between Engaricenna and Epidium, which latter is the most easterly; and a comparison between Ptolemy’s positions and those of the islands south of the point of Ardnamurchan seems to leave little doubt as to their identity.
Ptolemy’s five Ebudas with Monarina form the group of islands frequently mentioned in ancient Irish documents as ‘Ara, Ile, Rachra acus innsi orcheana,’ that is Arran, Isla, Rachra, and the other islands.
63. This appears from many circumstances. Pausanias implies it when he says that Antoninus, who advanced the frontier of the province from Hadrian’s wall to the Firths of Forth and Clyde, took land from the Brigantes (Paus. viii. 43). Tacitus mentions Venusius, King of the Brigantes, hostile to Rome, and that his frontiers were to the north of the province appears from the geographer of Ravenna placing the town of Venusio north of the stations at the wall. The early Latin editions of Ptolemy omit the Gadeni, and call the tribe north of the wall Otalini; but the edition of 1525, and the later editions, have ‘Gadeni, the more northern (western); Otadeni, the more southern (eastern);’ and the name Gadeni occurs in inscriptions. If this is the correct reading, however, it is obvious that Ptolemy considered them as substantially the same people, as he places the towns of Bremenium and Curia among them generally, without distinguishing to which tribe each belonged, and the terminations are the same. Inscriptions mentioning the god of the Gadeni have been found at Reesingham and at Old Penrith, within the territory of the Brigantes. On the other hand, an inscription to the goddess Brigantia has been found at Middlebie, within the territory of the Selgovæ.
64. This seems to be the town mentioned by Bede, Ec. Hist. B. i. c. 12, ‘Orientalis (sinus) habet in medio sui urbem Giudi.’
65. Trimontium has been identified with the Eildon hills in Roxburghshire, owing simply to the resemblance of the form of the hill with a station supposed to be called the Three Mountains; but it is more probable that the syllable Tri represents the Welsh Tre or Tref, and that it is a rendering of Trefmynydd, or the Town on the Mountain. To place it at the Eildon hills is to do great violence to Ptolemy’s text.
66. The first of our historians to make use of Ptolemy was Hector Boece, but he placed his names too far north. He puts the Brigantes in Galloway, and the Novantes in Kintyre, and hence their towns are placed in Argyll instead of Wigtown. The Ulm edition of 1486, which is a very inaccurate one, was apparently the edition used by Boece, and in it the name Rerigonium is misprinted Beregonium. Boece applied the name to the vitrified remains, the correct name of which was Dunmhicuisneachan, the fort of the sons of Uisneach, now corrupted into Dunmacsniochan, and thus arose one of the spurious traditions created by Boece’s History.
67. In some of the editions this name is Vanduara, and is considered by Chalmers to have been Paisley, and he has been followed by all subsequent writers. His reasons are very inconclusive, viz. that there are said to have been Roman remains at Paisley, and that Vanduara is probably derived from the Welsh Gwendwr or White water, and the river at Paisley is called the White Cart. But rivers do not change their names. If it had ever been called Gwendwr, it would have borne the name still; and to rest the identity of Vanduara with Paisley upon a mere conjectural etymology is the reverse of satisfactory. The best editions give Vandogara as the form of the name, which obviously connects it with Vindogara or the bay of Ayr; and Ptolemy’s position corresponds very closely with Loudon Hill on the river Irvine, where there is a Roman camp. What confirms this identity is, that the towns in the territory of the Damnonii appear afterwards to have been all connected with Roman roads, and there are the remains of a Roman road leading from this camp to Carstairs.
68. All editions agree in placing Devana in the interior of the country, at a distance of at least thirty miles from the coast. Its identity with the sea-port of Aberdeen rests upon the authority of Richard of Cirencester alone.
69. Mr. Burton, in stating his disbelief in the genuineness of Richard and its results, adds, among other things to be abandoned, ‘the celebrated Winged Camp; the Pteroton Stratopedon can no longer remain at Burghead in Moray, though a water tank there has become a Roman bath to help in its identification, and it must go back to Edinburgh or some other of its old sites.’—(Vol. i. p. 62.) He is, however, mistaken in supposing that its identification rests upon Richard. Ptolemy is in reality the authority for Alata Castra and its position on the shore of the Moray Firth.
It is of course absurd to recognise Roman remains there at that early period, but there can be no question that a native strength existed on that headland. See Proc. Ant. Soc. vol. iv. p. 321, for an account of the remains.
70. The only authorities for the events in the reign of Antoninus are two short passages. One, the passage of Pausanias, referred to in Note 63, and the other of Julius Capitolinus, who says (De Anton. Pio, 5), ‘Per legatos suos plurima bella gessit. Nam et Britannos per Lollium Urbicum legatum vicit, alio muro cespiticio submotis barbaris ducto.’ The expression ‘submotis barbaris’ proves that this wall now formed the boundary between the barbarian or independent tribes and the Roman province. It is analogous to the expression used by Aelius Spartianus of ‘qui barbaros Romanosque divideret,’ in stating the building of Hadrian’s wall. It does not necessarily imply an actual driving north of the people, but only the extension of the province, so that the part hostile to the Roman power came to be farther removed.
Chalmers has treated the Roman wars in Scotland very strangely. His narrative of the actions of Lollius Urbicus extends over seventy closely printed pages; while for all this the actual authority is comprised within exactly fourteen words of Julius Capitolinus. The campaigns of Severus, by far more important, occupy just six pages; and yet for these we have the detailed narrative of two independent historians.
71. The principal stations on the wall were at the following places—viz., West Kilpatrick, Duntocher, Castlehill, East Kilpatrick, Bemulie, Kirkintilloch, Auchindavy, Barhill, Westerwood, Castlecary, and Rough Castle; and as they are in general constructed partly of stone, and some of them connected with baths and more elaborate works, they are probably to be attributed to a later age. See a paper by David Milne Home, Esq., in the Trans. Roy. Soc. vol. xxvii. part i. p. 39, for the latest account of the wall.
72. Et adversus Britannos quidem Calphurnius Agricola missus est.—(Capitolin. Mar. Aur. 8.)
73. In Britannia, in Germania, et in Dacia imperium ejus recusantibus provincialibus, quae omnia ista per duces sedata sunt.—(Lamprid. Comm. c. 13. Conf. Dion. 72. 8.)
74. Dio, 75, 76, 77; Herodian, iii. 7; Capitolin. Clod. Alb. c. 9; Eutropius, viii. 18.
75. From Magh, a plain. The same word seems to enter into the name Vacomagi.
76. Colonel Gurwood’s Speeches of the Duke of Wellington, vol. ii. p. 729.
77. At Habitancum, a station on Watling Street, on the south bank of the Rede, inscriptions have been found showing that Severus restored the gate and repaired the walls of the station. See Bruce’s Roman Wall, p. 384.
78. These camps are as follows—viz., Wardykes, near Keithock; Raedykes, near Stonehaven; Normandykes, on the Dee; and Raedykes, on the Ythan.
79. The account of the campaigns of Severus, and of the state of the hostile nations at the time, is given at length in the two independent narratives of Dio (as abridged by Xiphiline) and Herodian, and therefore rests upon peculiarly firm ground. A great deal too much has been made of the Mæatæ by previous historians. It has been stated, as if it were a name in general use and applied to the tribes between the walls during the whole period of the Roman occupation of Britain; but the fact is that the Mæatæ are mentioned by Dio alone, and on this occasion only. We never hear of them before or after. Innes and Chalmers talk of the Mæatæ or Midland Britons (that fatal or of historians implying an identity assumed but not proved), as if there were some analogy between the names. There is none. The term Midland Britons nowhere occurs, and the root of the name Mæatæ is probably the word for a plain, nearly the same in Welsh and Gaelic—Maes, Magh. That both nations were in Caledonia is plain, independently of the position that the wall alluded to by Dio is the wall between the Forth and the Clyde, for Dio styles them both ‘the inhabitants of that part of Britain which is hostile to us,’ that is, extra-provincial. Moreover, Dio’s expression ‘advanced into Caledonia,’ is the equivalent of Herodian’s, ‘he passed beyond the rivers and fortresses that defended the Roman territory.’ That Severus constructed roads and built bridges is emphatically stated by both Dio and Herodian, and it is to him alone that the classical historians attribute such works in Britain.
80. The Horesti are mentioned in the inscription noticed in chap. i., Note 52. The other inscription is as follows—‘In H.D.D. Baioli et vexillarii Collegio Victoriensium signiferorum Genum de suo fecerunt viii. kal. Octobr. Presente et Albino Cos.’ which places it in 239.
81. That Severus built or had reconstructed a wall in Britain rests upon the direct authority of Aurelius Victor, Eutropius, Spartian, Orosius, and Eusebius. Spartian, who wrote in 280, says (c. 18), ‘Britanniam, quod maximum ejus imperii decus est, muro per transversam insulam ducto, utrimque ad finem oceani munivit.’ Unfortunately he does not give the length of the wall, which would have indicated its position; but he also says (c. 22), ‘Post murum aut vallum missum in Britannia, quum ad proximam mansionem rediret, non solum victor, sed etiam in aeternum pace fundata;’ which shows that it was after his expedition into Caledonia; and it is rather remarkable that at Cramond—the proxima mansio—behind the wall of Antoninus, was found a medal of Severus, having on the reverse the inscription, ‘fundator pacis.’ Aurelius Victor, who wrote 360, says, ‘His majora aggressus Britanniam quae ad ea utilis erat, pulsis hostibus, muro munivit, per transversam insulam ducto, utrimque ad finem oceani’ (De Caes. 20). And again: ‘Hic in Britannia vallum per triginta duo passuum millia a mari ad mare deduxit’ (Epit. 40). And Eutropius, who wrote at the same time, says, ‘Novissimum bellum in Britannia habuit: utque receptas provincias omni securitate muniret, vallum per 32 millia passuum a mari ad mare deduxit’ (viii. 19).
Both these writers place the construction of the vallum after the war, and if it was thirty-two Roman miles in length, it can only have extended across the peninsula between the Forth and the Clyde. Orosius, who wrote in 417, says, ‘Severus victor in Britannias defectu pene omnium sociorum trahitur. Ubi magnis gravibusque praeliis saepe gestis, receptam partem insulae a caeteris indomitis gentibus vallo distinguendam putavit. Itaque magnam fossam firmissimumque vallum, crebris insuper turribus communitum, per centum triginta et duo millia passuum a mari ad mare duxit.’ Eusebius, as reported by St. Jerome, says, ‘Severus in Britannos bellum transfert, ubi, ut receptas provincias ab incursione barbarica faceret securiores, vallum per 132 passuum millia a mari ad mare duxit.’
The length here given of 132 Roman miles is as inconsistent with the distance between the Tyne and the Solway, as it is with that between the Forth and the Clyde. Horsley, who considered that the earthen vallum between the Tyne and the Solway was the work of Hadrian, and the murus or wall which runs parallel to it, the work of Severus, supposed that in the original MS. of these writers the distance had been written LXXXII and that C had been written by mistake for L, which would reduce the distance to eighty-two miles; but no MS. supports this conjecture, and Mr. Bruce, in his work on the wall, clearly establishes that both are the work of Hadrian.
It is inconceivable that our best historians should have gone so entirely against the direct testimony of the older authorities. They have in this given too much weight to the opinion of Bede, who first declared the remains of the wall between the Tyne and the Solway to be those of Severus’s wall, for opinion it is only, and he was naturally biassed by the remains of the northern rampart being always before his eyes. Nennius gives the native tradition before his time when he quotes the passage from Eusebius, and adds, ‘et vocatur Britannico sermone Guaul a Penguaul quae villa Scotici Cenail, Anglice vero Peneltun dicitur, usque ad ostium fluminis Cluth et Cairpentaloch, quo murus ille finitur rustico opere;’ thus clearly placing the wall between the Forth and Clyde.
Moreover, placing Severus’s wall between the Tyne and Solway involves the manifest inconsistency, that, after penetrating almost to the end of the island, and making a peace, in which territory was ceded to him, he abandoned the whole of his conquests, and withdrew the frontier of the province to where it had been placed by Hadrian. Chalmers, who saw this difficulty, supposes that he built the wall before he commenced his conquests; but this is equally against the direct statement of the older authorities, that it was built after he had driven back his enemies and concluded peace. Mr. Bruce has the pertinent remark that ‘if Severus built the wall (between Tyne and Solway), we should expect to find frequent intimations of the fact in the stations and mile castles. The truth, however, is that from Wallsend to Bowness we do not meet with a single inscription belonging to the reign of Severus, while we meet with several belonging to that of Hadrian’ (p. 382).
82. Aurel. Victor. de Caes. 39; Eutrop. ix. 21; Orosius, vii. 25.
83. Eumenius, Paneg. Const. c. 12. Eutrop. ix. 22.
84. Eumen. Pan. Const. Caes. c. 6. Mamert. Pan. Max. Herc. c. 11, 12.
85. Comp. Eumenius, ‘prolixo crine rutilantia,’ with Tacitus, ‘rutilae Caledoniam habitantium comae.’
86. Non dico Caledonum aliorumque Pictorum silvas et paludes.—Eumen. c. 7.
87. Appian. Alex. Hist. Rom. Præf. 5. Eumen. Pan. Const. cc. 9-19.
88. Sunt in Gallia cum Aquitania et Britanniis decem et octo provinciae ... in Britannia, Maxima Cæsariensis, Flavia, Britannia Prima, Britannia Secunda.—Sextus Rufus Festus (360), Brev. 6.
89. ‘Consulatu vero Constantii decies, terque Juliani, in Britanniis cum Scotorum Pictorumque gentium ferarum excursus, rupta quiete condicta, loca limitibus vicina vastarent, et implicaret formido provincias præteritarum cladium congerie fessas.’—Am. Mar. B. xx. c. 1. The sentence which follows—‘Hyemem agens apud Parisios Cæsar distractusque in solicitudines varias, verebatur ire subsidio transmarinis; ut retulimus ante fecisse Constantem,’ etc.—implies that there had been a previous attack in 343, but as this part of Ammianus’s work is lost, it is impossible to found upon it. The peace said to have been broken probably followed it.
90. The early legends of Wales show that the seaboard of that district had been exposed at an early period to the attacks of the Scots. Nennius, in giving the early settlements of the Scots in Britain, says—‘Filii autem Liethan obtinuerunt in regione Demetorum’ (that is South Wales), ‘et in aliis regionibus, id est, Guir et Cetgueli, donec expulsi sunt a Cuneda et a filiis ejus ab omnibus Britannicis regionibus.’ And again—‘Scotti autem de occidente et Picti de aquilone.’ And again—‘Mailcunus magnus rex apud Brittones regnabat, id est, in regione Guenedotæ’ (that is, North Wales), ‘quia atavus illius, id est, Cunedag cum filiis suis, quorum numerus octo erat, venerat prius de parte sinistrali, id est, de regione quæ vocatur Manau Guotodin centum quadraginta sex annis antequam Mailcun regnaret et Scottos cum ingentissima clade expulerunt ab istis regionibus et nunquam reversi sunt iterum ad habitandum.’—Nennii Brit. Hist.
91. Hoc tempore (364) ... Picti Saxonesque et Scoti et Attacotti Britannos ærumnis vexavere continuis.—Ammian. Mar. xxvi. 4.
92. Illud tamen sufficiet dici, quod eo tempore Picti in duas gentes divisi, Dicalidonas et Vecturiones, itidemque Attacotti, bellicosa hominum natio, et Scotti, per diversa vagantes, multa populabantur.—Ammian. Mar. xxvii. 8, 9.
The ‘Caledonii’ of Dio we know were the most northerly of the two nations; and the ‘Dicalidonæ’ of Ammianus must have extended along the coast bounded by the Deucaledonian Sea of Ptolemy.
93.
It has generally been supposed that the province had at this time only extended to the wall between the Solway and the Tyne, and that Theodosius added the additional territory, which now for the first time became a province under the name of Valentia. But the words of the historian are directly opposed to this: ‘Recuperatamque provinciam, quæ in ditionem concesserat hostium, ita reddiderat statui pristino.’—Am. Mar. B. xxviii. c. 3.
94. The Notitia Imperii, compiled subsequently to this expedition, has the following bodies of Atecotti in the Roman army who were stationed in Gaul:—
St. Jerome says that he saw in Gaul the Atticotts, a British nation, which implies that they were inhabitants of Britain. He says (Adv. Her. ii.), ‘Quid loquar de cæteris nationibus, quum ipse adolescentulus in Gallia viderim Atticotos, gentem Britannicam, humanis vesci carnibus.’ As St. Jerome says that he was then ‘adolescentulus,’ and was born in the year 340, it is supposed that this could not have been later than 355; but this is a mistake arising from overlooking the lax sense in which Jerome uses the word ‘adolescentulus,’ which he stretches into very mature age. He uses the expressions of ‘puer’ and ‘adolescens’ for himself when he was at least thirty years old. St. Jerome was in Gaul at only one period of his life, and that we know from other circumstances must have been about the period of Theodosius’s conquest. That the Atecotti were inhabitants of the district between the walls appears from the fact that they only joined the invading tribes after the latter had been four years in possession of that territory; and that no sooner was it again wrested from the invaders by Theodosius, than we find them enlisted in the Roman army.
95. The three bodies of infantry were the Victores Juniores Britanniciani, the Primani Juniores, and Secundani Juniores. The six bodies of cavalry, the Equites Catafractarii Juniores, the Equites Scutarii Aureliaci, the Equites Honoriani Seniores, the Equites Stablesiani, the Equites Syri, and the Equites Taifali. The Equites Catafractarii were stationed at Morbium, supposed by Horsley to be Templeburgh on the south bank of the river Don. The provinces are twice given in the Notitia, and the order is the same in both—Maxima Cæsariensis, Valentia, Britannia Prima, Britannia Secunda, Flavia Cæsariensis. The position usually assigned to these provinces rests entirely upon the authority of the spurious Richard of Cirencester, and involves the supposition that when Constantine divided the provinces into four, he substituted the name of Maxima Cæsariensis for that of Lower Britain, and divided Upper Britain into three provinces, forming the district of Wales into a separate province called Britannia Secunda; but if the order in the Notitia is geographical, and proceeds from north to south, Maxima Cæsariensis is the most northerly, then Valentia and Britannia Prima extend across the island from west to east. Then south of them Britannia Secunda, and farther south Flavia Cæsariensis; and thus, before Valentia was formed, Maxima Cæsariensis and Britannia Prima would represent what had been Lower Britain, and the Dux Britanniæ would command the troops within it; Britannia Secunda and Flavia Cæsariensis what had been Upper Britain, and the Comes tractus maritimi, the troops within it. The new province would be formed in the west to meet the invasion in a new quarter from a new people, the Scots; and a new commander, the Comes Britanniarum, or Count of the two Britannias, would be placed there to protect the western frontier.
96. Incursantes Pictos et Scotos Maximus strenue superavit.—Prosper. Aquit. Gratian. iv.
97.
98. Of the four bodies of Attecotti in the Roman army, the first two were those probably enrolled by Theodosius, and seen by St. Jerome in Gaul. The two last, which are termed Honoriani, must, from their name, have been enrolled by Stilicho, the minister of Honorius. Orosius called the latter ‘Barbari ... qui quondam in fœdus recepti atque in militiam adlecti Honoriaci vocabantur’ (Oros. vii. 40). Thus, on the two occasions in which the territory between the walls was recovered, Attecotti were enrolled in the Roman army. They were Barbari who ravaged Britain, when the Barbarians occupied this part of the province. They were ‘in fœdus recepti et in militiam adlecti’ when the Romans recovered it—a combination only applicable to the half-provincial half-independent tribes between the walls; and they were probably the same people whom Ptolemy called the Ottedeni and Gadeni, who extended from the southern wall to the Firth of Forth. The same word seems to enter into the composition of the names Ottedeni and Attecotti.
99. The army is mentioned in Britain in 406. Stilicho was consul the preceding year. The Notitia Imperii refers to a state of matters after Theodosius, for the province of Valentia is mentioned, and the army there described must have been in Britain at this time.
100. Adversus hos Constantinus Constantem filium suum, proh dolor! ex monacho Cæsarem factum, cum barbaris quibusdam, qui quondam in fœdus recepti atque in militiam adlecti, Honoriaci vocabantur, in Hespanias misit.—Orosius, vii. 40.
101. This account of the usurpation of Constantine, and its consequences, is taken from Zosimus and Olympiodorus, two contemporary historians. The opinion generally entertained that the Roman troops returned to Britain after the year 410 rests upon no direct authority, and is opposed to the testimony of those contemporary historians. Mr. Bruce, in his Roman Wall, makes the pertinent remark (43): ‘The series of coins found in the stations of the north of England, and in the camps and Roman cities of the south, extends from the earlier reigns of the Empire down to the times of Arcadius and Honorius, and then ceases. Any legion coming later must have been destitute of treasure.’
The mistake has arisen from the false chronology of the invasions of the Scots and Picts, and of the assistance of the Romans in repelling them, applied to the narrative of Gildas. No dates are given in the work of Gildas; but if the mind is disabused of preconceived conceptions in this respect, it is impossible to compare Gildas’s narrative with the notices of the legion sent by Stilicho, and of the army which elected Constantine, the attack which followed, and the repelling of the invaders by the provincial Britons, without seeing the absolute identity of the events.
The following comparison will show this more clearly:—
| Roman and Greek Authors. | Narrative of Gildas. | |
| 383 | Maximus revolts. | Revolt of Maximus, who withdraws the army with the youth from Britain. |
| 387 | Withdraws Roman army from Britain. | |
| 396 | A legion sent by Stilicho, who drive back Picts and Scots, and garrison wall. | First devastation of Picts and Scots. Britons apply for assistance. A legion sent, who build northern wall. |
| 402 | Legion withdrawn. | Legion withdrawn. |
| Second devastation of Picts and Scots. | ||
| Britons again apply for assistance. | ||
| 406 | A Roman army in Britain — stationed ‘per lineam valli.’ | Roman troops sent, who fortify southern wall. |
| 407 | Constantine withdraws Roman army. | Roman troops withdrawn, ‘never to return.’ |
| Picts seize up to wall. | ||
| Break through wall and ravage. | ||
| 409 | Gerontius invites Barbarians. Honorius frees province. |
Provincials take courage and repel them. |
| Provincials raise and repel invaders. | Vortigern invites Saxons. |