The Heroic Age
This paper argues that Ambrosius Aurelianus was a more important figure than Arthur in fifth/sixth-century Britain. The life of Ambrosius elucidates the continuity of Roman Britain beyond the formal end of the Roman period in 410 and the rise of kingship in the former diocese during the fifth-century.
This paper  reviews the historical evidence for Ambrosius Aurelianus and his family, and considers whether this evidence reflects the general situation in sub-Roman Britain. It is not my primary intention to develop a narrative history of sub-Roman Britain with a precise chronology.
The information provided by Gildas, principally in his Ruin of Britain, will provide the basis of this review. Although Gildas was probably writing in the first part of the sixth century and was therefore almost contemporary with the fifth century events he describes, Gildas has tended to be disregarded as a historical source because the details he gave of events were vague and inconsistent; furthermore, he did not provide a precise chronology for these events. However, more recent reviews have demonstrated that although Gildas was primarily concerned with societal matters of his own time, he is nevertheless an invaluable source of information on fifth and early sixth century Britain. The other source that refers to Ambrosius Aurelianus is the British History, attributed to Nennius. The British History is apparently based on earlier sources but was compiled in the early ninth century. It therefore post-dates Ambrosius Aurelianus by some 300 years, and so in this paper its evidence is regarded as secondary to the information provided by Gildas.
The period following the end of formal Roman imperial involvement in Britain has been perceived as a 'dark age' in terms of the lack of historical information available. In reality, there are numerous British and continental sources, which together provide considerable insight into sub-Roman Britain. With the rehabilitation of Gildas as a historical source, fifth/sixth century Britain is arguably better documented than much of the formal Roman period. This is illustrated by the evidence for Ambrosius Aurelianus and his family, an account that can be emulated for few personalities involved in Roman Britain, particularly during the later Roman period.
Gildas describes Ambrosius Aurelianus as follows:
[The British] leader was Ambrosius Aurelianus, a gentleman who, perhaps alone of the Romans, had survived the shock of this notable storm: certainly his parents, who had worn the purple, were slain in it. His descendants in our day have become greatly inferior to their grandfather's excellence.
The reference to his parents wearing the purple has been interpreted as meaning that Ambrosius Aurelianus's parents were in the upper echelons of the political hierarchy in Britain. Ambrosius Aurelianus's father (hereafter Ambrosius senior) is described in the British History as 'one of the consuls of the Roman people'.
The chronographic section of the British History includes the following computation: 'From the reign of Vortigern to the quarrel between Vitalinus and Ambrosius are twelve years, that is Guollopum, the battle of Guoloph.' The chronograph in the British History also dates the beginning of Vortigern's reign to c.425. This would give a date of c.437 for the battle of Guoloph. Moreover, the British History also states that: 'Vortigern ruled in Britain, and during his rule in Britain, he was under pressure, from fear of the Picts and Irish, and of a Roman invasion, and not least, from dread of Ambrosius.'
These references identify an Ambrosius who was a contemporary, indeed a rival of Vortigern; and Gildas refers to Ambrosius senior, who was killed during the Saxon revolt. It is very likely that Ambrosius senior and the Ambrosius who was a rival of Vortigern were one and the same person. Ambrosius senior, who probably fought at Guoloph c.437, who died during the Saxon revolt, and who was a contemporary of Vortigern, can therefore be given a floruit of the second quarter of the fifth century.
A more precise date for the death of Ambrosius senior would be forthcoming if the Saxon revolt could be more securely dated. The Gallic Chronicle of 452 states that around the year 441: 'The British provinces, which up to this time had suffered various defeats and catastrophes, were reduced to Saxon rule.' Although this entry could refer to the Saxon revolt, Alcock has suggested that this statement was an exaggeration; perhaps only a limited area, which had close links with the continent, had changed hands. Moreover, it may be that the statement in the Gallic Chronicle does not refer to the Saxon revolt at all. The phrase indicionem Saxonum rediguntur suggests a somewhat passive change of authority, perhaps by political means rather than by force of arms. Gildas states that the first two contingents of Saxones had been invited by the British authorities. The technical terms he uses indicate that these contingents were foederati, which implies the British were merely following the established Roman policy of inviting foreign troops to strengthen their military.
The British History also recounts the arrival of two successive groups of Germanic troops. Furthermore, it states that Vortigern ceded Kent to the Saxons following the arrival of the second contingent, which comprised sixteen boatloads of warriors: 'So he [Vortigern] granted it [Kent] to them, although Gwyrangon was ruling in Kent, and did not know that his kingdom was being handed over to the heathens, and that he was himself given secretly into their power on his own.'
It is, therefore, possible that the Gallic Chronicle entry of c.441 refers not to the Saxon revolt, but to the settlement of the second contingent of Germanic troops in Kent. From Gaul, this event could have appeared to be an occupation of the entire country.
If the entry of c.441 in the Chronicle of 452 refers to a second major immigration of Germanic federates, it would follow that the Saxon revolt occurred after that date. In fact, Gildas says that the British supported the federates 'for a long time'; apparently, a noticeable period of time had passed between the second immigration and the revolt, in which case a date of c.450 or later is suggested for the commencement of the revolt. The British History also indicates that the Saxon revolt was a prolonged affair, with Vortigern's supporters (principally his son, Vortimer) attempting to contain the Saxons. Gildas concentrated on the after-effects, although his reference to 'repeated batterings' suggests that the revolt was more than a short-lived episode.
Incidentally, one reason why Gildas has been doubted as a historical source is his placing of a British appeal to the imperial authorities for help against raiders (the 'groans of the British'), apparently dating to 446 or later, before the Germanic immigrations. However, it has been argued that Gildas was simply identifying the recipient of the appeal as Aetius, military commander in Gaul, who was active from 425 to 454. If so, the appeal could have been made in the face of Pictish and Scottish raiding before the arrival of the Germanic foederati, without disrupting Gildas's relative chronology. Indeed, it is possible that the appeal to the military was associated with the documented appeal from the British that prompted the visit by St. Germanus in 429; the appeals mentioned by Gildas and in the Life of St. Germanus were both followed by military victories against the raiders.
According to the British History, after a campaign of fluctuating fortunes, the revolt ended disastrously for the British; Vortigern, still ruler of the British, was humiliated. Gildas remarks that as a consequence:
So a number of the wretched survivors were caught in the mountains and butchered wholesale. Others, their spirit broken by hunger, went to surrender to the enemy; they were fated to be slaves forever, if indeed they were not killed straight away, the highest boon. Others made for lands beyond the sea; beneath the swelling sails they loudly wailed, singing a psalm that took the place of a shanty: "You have given us like sheep for eating and scattered us among the heathen."
The disastrous (for the British) conclusion to the revolt evidently resulted in the emigration of a significant proportion of the population. Documentary evidence points to Gaul as the destination of many of these fugitives. For example, the military commander Riothamus, having reached Gaul 'by way of the Ocean', was made an ally of the Emperor Anthemius in 469; and Mansuetus, bishop of the Britons, attended the Council of Tours in 461. This evidence suggests that the Saxon revolt had ended by 460. However, the emigrations could have commenced during the revolt, in which case the Saxon revolt might have ended after 460.
If the Saxon revolt ended c.460, then the death of Ambrosius senior is accorded a similar date. Meanwhile, Gildas specifically states that Ambrosius Aurelianus was the `grandfather' of some of Gildas' contemporaries, at the time of his writing the De Excidio. The documentary evidence therefore suggests the following genealogy for the Ambrosian family, based on generations at thirty-year intervals with an average life-expectancy of sixty years and a floruit at around twenty-five to fifty years of age:
Ambrosius senior c.415-c.460 Ambrosius Aurelianus c.445-c.505 an unrecorded generation c.475-c.535 the grandchildren c.505-c.565
This genealogy requires long life spans and the presence of Ambrosius senior on the battlefield at Guoloph at a relatively early age. However, these elements are not wholly unlikely. Anyway, it is not necessarily the case that the opponent of Vitalinus, the rival of Vortigern and the father of Ambrosius Aurelianus were all the same person. Consequently, an alternative genealogy, based on twenty-five year generations and an average life-expectancy of fifty years, is possible:
Ambrosius 'I' (rival of Vitalinus and Vortigern) c.405-c.455 Ambrosius 'II' (killed in the Saxon revolt) c.430-c.460 Ambrosius Aurelianus c.455-c.505 an unrecorded generation c.480-c.530 the grandchildren c.505-c.555
There are three points worthy of note from this alternative genealogy. First, the floruit of Ambrosius Aurelianus occurs slightly later in this version of the genealogy than in the simpler version. Second, the relationship between Ambrosius I and Ambrosius II is unknown; it can only be assumed that Ambrosius II was the son of Ambrosius I (the documented ancestry of Ambrosius Aurelianus is referred to as 'Ambrosius senior' in this article, even though it may have involved two generations). Third, as Gildas was writing during the floruit of the grandchildren, the above genealogy points to a date of c.530x560 for the Ruin of Britain. This is consistent with the conventional dating for this work, based on its chronological context and the writing style.
As Gildas was denouncing the failings of his contemporaries, particularly their impiousness, Ambrosius was presumably a Christian; otherwise, Gildas would not have discussed him in such detail, and might even have omitted to mention him at all.
The documentary sources provide further information regarding Ambrosius Aurelianus himself. For example, the British History includes a 'Tale of Emrys', which relates how Vortigern tries to build a fortress in Wales, and eventually gives the fortress to Ambrosius. This tale was clearly intended to explain the derivation of the name Dinas Emrys, the fortress in question. In addition, it contains details that range from the inaccurate to the fantastic - not least alleging Ambrosius' immaculate conception. On the other hand, some incidental elements of the story could be accurate; indeed, facts might have been deliberately incorporated in order to lend the tale an air of credibility to what might otherwise have been readily dismissed as a total fabrication by its audience. Interestingly, the tale brings together Vortigern, approaching the end of his life, with Ambrosius as a boy; perhaps this is an indication of their age difference.
In addition, the 'Tale of Emrys' ends thus:
Then the king asked the lad "What is your name?". He replied "I am called Ambrosius," that is, he was shown to be Emrys the overlord. The king asked "What family do you come from?" and he answered "My father is one of the consuls of the Roman people." So the king gave him the fortress, with all the kingdoms of the western part of Britain.
This text supports Gildas's assertion that Ambrosius Aurelianus had a father of very high rank. In addition, it assigns Ambrosius a title, Guletic, which can be variously translated as `prince' or `overlord', perhaps implying a role different to that of a king. Gildas referred to Ambrosius as a dux, which was the title of a military commander in later Roman times. The British History even describes him as 'the great king among all the kings of the British nation'. It is difficult to know what to make of this range of titles, although as Gildas uses dux this was more likely to have been the term applied to Ambrosius during his lifetime. Overall, these titles suggest that Ambrosius was not just one of numerous local rulers - he was the pre-eminent leader. Furthermore, the 'Tale of Emrys' ends with Vortigern giving Ambrosius all the kingdoms of the western part of Britain; this can arguably be seen as another statement introduced to enhance the tale's plausibility. In fact, the British History also notes that one of Vortigern's sons, Pascent, 'ruled in the two countries called Builth and Gwerthrynion after his father's death, by kind permission of Ambrosius'. Taken as a whole, these references to Ambrosius suggest that he eventually succeeded Vortigern as the leader of the British in the former diocese.
Gildas recounts how Ambrosius Aurelianus attained such an elevated position. Following the catastrophic defeat of the British at the end of the Saxon revolt - but even then, only 'after a time' - Ambrosius emerged as the focus of a Romano-British revival:
Under him our people regained their strength, and challenged the victors to battle. The Lord assented, and the battle went their way. From then on victory went now to our countrymen, now to their enemies; so that in this people the Lord could make trial (as he tends to) of his latter-day Israel to see whether it loves him or not. This lasted right up to the year of the siege of Badon hill, pretty well the last defeat of the villains, and certainly not the least.
It has been calculated above that the end of the Saxon revolt occurred c.460, and 'after a time' suggests that there was a significant interval (probably at least a decade) before the recovery, under Ambrosius, began. Consequently Ambrosius's floruit probably began c.475. It should also be noted that Gildas appears to regard the military campaign of Ambrosius as continuing up to the 'siege of Badon hill'. Gildas' narrative covering the later fifth century can be subdivided into the aftermath of the Saxon revolt (25.1), the British recovery led by Ambrosius (25.2-26.1), and the ensuing period of peace (26.2-4). There is no indication in the text that the battle of Badon was separated in time from the campaign of Ambrosius. Indeed, it seems that the initial victory marked the beginning of, and the battle of Badon concluded, a single campaign that was led throughout by Ambrosius.
The battle of Badon has attracted much attention partly because Gildas highlights it as marking the (successful) end of the British recovery, and partly due to references to a battle of Badon in the British History (the campaigns of Arthur), and in the Welsh Annals (for the year 516). This British victory is usually credited to Arthur on the basis of the latter two references. However, doubt has been cast over such annalistic entries, which were compiled long after the event and were relatively susceptible to amendment. As a virtually contemporary, detailed narrative account, the evidence of the Ruin of Britain ought to carry much greater weight. Gildas not only fails to acknowledge Arthur as a champion of the British cause, he does not mention Arthur in the Ruin of Britain at all.
Gildas's reference to Badon is somewhat laboured; 'pretty well the last ... and certainly not the least [British victory],' fought in the year that Gildas was born. There are a number of possible reasons why Gildas mentions Badon, and in such a manner. Firstly, it was not an exceptional battle, but it was singled out because it was fought in the year that Gildas was born. Secondly, it was one of a number of British victories, singled out because it was fought by Ambrosius Aurelianus. Finally, it was the decisive British victory, which just happened to be fought in the year that Gildas was born.
The first possibility is the least likely. The fact that the battle was fought in the year of Gildas' birth would not have elevated its importance in the minds of the audience to whom the Ruin of Britain was directed; it would not have furthered Gildas's argument. The second possibility is plausible, if a number of British kings were carrying out their own, localised military campaigns. However, Gildas identifies Ambrosius Aurelianus as the leader of the British, which suggests that there was one, co-ordinated campaign rather than a series of localised conflicts. Gildas' convoluted reference to Badon is typical of his complex rhetorical writing style, but his meaning is clear; the battle was not the final defeat of the Saxons, but it was the greatest. Consequently, the likeliest reason for Gildas mentioning Badon appears to be the third option - it was the most important battle, which just happened to be fought in the year that Gildas was born. Gildas felt it unnecessary to name Ambrosius as the victor in that particular battle, because he had already identified him as leader of the whole campaign.
As with most of the events that occurred in sub-Roman Britain, the date of the battle of Badon is uncertain. The only absolute date offered by the documentary sources is that of 516 in the Welsh Annals; but this is an annalistic entry, and cannot be relied upon in isolation. Ian Wood has suggested that Ambrosius Aurelianus's victory occurred forty-four years before the writing of De Excidio, and the battle of Badon took place only a month prior to Gildas writing his text. However, Gildas refers to Badon as `nearly the last [battle]' and declares that 'external wars' were over - statements which could only have been made with the benefit of considerable hindsight. Wood also argues that 'Gildas tells us that a previous generation witnessed the double miracle of the Saxon revolt and Ambrosius Aurelianus' victory,' implying that Badon was somewhat removed in time from the two earlier events. But Gildas actually states that the previous generation witnessed the revolt and the 'unlooked for recovery.' According to Gildas, Ambrosius' first victory did not in itself constitute a recovery, merely the beginning of one. Ambrosius' campaign was one of fluctuating fortunes. In addition, the recovery was not completed until at least one victory (and perhaps one or more defeats) after Badon. Gildas refers to his own time as 'an age ... that is ignorant of that storm and has experience only of the calm of the present'. It is difficult to believe, as Wood's hypothesis requires, that this calm had involved at least two battles in the preceding month. After all, the form of Gildas's text suggests that there was a significant interval of time not between Ambrosius' initial victory and Badon, but rather between Ambrosius' campaign (including Badon) and the relative calm of Gildas' time. The most plausible interpretation of Gildas' reference to Badon is that the battle did take place forty-four years before Gildas wrote the Ruin of Britain.
Given this relative date for Badon, a more precise date for the battle would be forthcoming if it was possible to determine when the Ruin of Britain was written. As discussed above, this work was probably written during the second quarter of the sixth century, and the proposed Ambrosian genealogy dates Gildas' contemporaries to the same period. Placing Badon forty-four years before the writing of the Ruin of Britain would date the battle to around the last decade of the fifth century. This would have been well within the floruit of Ambrosius Aurelianus, and so it is quite possible that he was the British leader at Badon as Gildas implies.
In order to understand how and why members of the family of Ambrosius Aurelianus could have been so influential in the affairs of sub-Roman Britain over perhaps five generations, their activities have to be considered within the context of diocesan political authority. As the nature of political authority is not a central theme of the Ruin of Britain, the information that Gildas provides on this subject may have been largely free from selectivity. The first words in the Complaint section of the Ruin of Britain are that 'Britain has kings, but they are tyrants.' This suggests that although Gildas questioned the quality of kingship displayed by certain individuals of his time, he accepted the existence of kings at the highest level of political authority in Britain. Yet little more than a century earlier, the diocese had a centralised government that was subject to the Roman emperor. How did the change from imperial, diocesan government to kingships come about?
The Notitia Dignitatum indicates that in the late fourth century Britain was controlled by a vicarius. Under him, two of the five provinces were governed by consulares, of senatorial rank, and the rest were administered by praesides, of equestrian rank. There were also three military commands: a dux and two comites. The title of consul appears to have been in use in Britain during the early fifth century, and it is likely that these posts were filled until the upheavals of 406-10.
According to Zosimus, upon the collapse of Constantine III's government in 409, Britain established its own constitution. As the inhabitants of the diocese asked for guidance from Honorius, they apparently expected the empire to resume its control of Britain. Zosimus states that the civitates led this appeal. In the absence of diocesan and provincial officials such as the vicarius, who formed the government of Constantine III and were presumably overthrown, the next level of civic government would have been the local government of the civitates. Indeed Gratian, one of the short-lived British emperors of 406, is described by Orosius as municeps tyrannus. If this means that Gratian was a town official who had (from Orosius's imperial perspective) usurped power, perhaps the authority of the imperial administration was being contested by local officials prior to 409.
For the civitates to act together on behalf of the diocese, some form of unifying mechanism would have been necessary, and this may been provided by a diocesan council comprising representatives of the civitates. Concilia provinciae had acted for the provinces on a limited scale during the Roman period. The empire did not resume control over Britain after the fall of Constantine III, and it seems the diocesan council remained in control for a time. Gildas mentions a council and its members acting for Britain prior to the Saxon revolt, and in the British History a 'whole council of the British' was involved in Germanus's second visit to Britain c.435.
However, the absence of imperial authority following the Rescript of Honorius left a power vacuum in the diocese. The diocesan administration had been answerable to the Praetorian Prefect of the Gauls, but at least the authority of the diocesan government had been supported and legitimised by the empire. The post-409 British government was answerable to no one, but this also meant that internal problems could not be referred to a higher authority. There may have been occasional imperial interventions, including the visits of St. Germanus in the 420s-30s, but with time it would have been apparent that the empire would not normally become involved in British affairs. It was, perhaps, inevitable that the effectiveness of the council would have been compromised due to internal tensions. Tribal, religious and political divisions may have been involved, and personal differences were also a likely factor. There may also have been difficulties in keeping the military commanders subordinate to the council.
Consequently, a new form of authority appears around the second quarter of the fifth century - that of kingship. Kings probably originated at the tribal level, as the dominant family of each tribal aristocracy steadily consolidated its local political influence. It is likely that sub-kings controlled smaller areas, for example 'city-states' centred on walled towns in the more urbanised south-east part of the diocese. In the British History, St. Germanus is said to have encountered a king within his defended city during his first visit to Britain in 429 ; and Gildas refers to kings ruling in Britain before the Saxon revolt. The title tyrannus appears as well as rex, but it seems that the former term is used to describe a bad king.
The rise of kingship was probably concomitant with the decline in the power of the diocesan council. If Britain continued to be governed as a diocese, it was inevitable that the tribal kings would attempt to gain supremacy at this political level. Perhaps the first king to succeed in this aim was Vortigern. Gildas claims that the superbus tyrannus, who was probably Vortigern, instituted the policy of settling Germanic foederati in Britain. This person was apparently a supreme king, who held sway over the council. According to the British History, Vortigern was the emperor of Britain, but was 'under pressure from fear ... of a Roman invasion, and, not least, from dread of Ambrosius.' Whether or not Vortigern had usurped control of Britain, he evidently had rivals and was threatened by the re-imposition of imperial control of the diocese.
Certain titles, if not offices, may have been inherited from the imperial administrative system. Although Vortigern might have been expected to have taken the role of vicarius, the title of emperor may have been preferred in order to emphasise his authority over the military. Ambrosius senior appears to have worn the purple and been a consul, indicating a senior role in government. One possibility is that the provinces continued as administrative units after 409, and Ambrosius senior had governed one of them. If the provincial structure did continue, one of the roles of the council may have been to elect the governors (and the vicarius/emperor), presumably from its own membership. However, even under this system there would be pressure for such offices of power to become hereditary, and so they would have ended up as kingdoms.
In this context, it is likely that Ambrosius senior was one of the kings who assumed power during the second quarter of the fifth century. Ambrosius senior's documented friction with both Vitalinus and Vortigern (above, notes 8-9) probably represents rivalry between kings. That this rivalry was deemed worthy of record, and that it involved the 'High King' Vortigern, point to more than just localised conflict. These individuals could have controlled large kingdoms from their heartlands by enforcing the allegiance of sub-kings, or may have been nominal leaders of political factions.
An Ambrosian kingdom is likely to have been located in one of two areas of the diocese. The first is eastern Catuvellaunia, which contains a concentration of Ambros- place-names, suggesting an association with a person or persons of that name. The same area has been tentatively identified as a British enclave, centred on St. Albans, which may not have fallen under Germanic control until the battle of Biedcanford, which occurred in 571 according to the Anglo-Saxon Chronicle. However, this means the kingdom ruled by Ambrosius Aurelianus and his descendants would have been located on the eastern side of the island, isolated from the rest of the British territories. It is difficult, but not impossible, to see how Ambrosius Aurelianus could have acted as the supreme British ruler in such circumstances.
The second possible location of an Ambrosian kingdom is the territory of the Durotriges. It has been suggested that one of the few British kings not criticised by Gildas was the ruler of the Durotriges , and Ambrosius Aurelianus's grandchildren also avoided incurring Gildas's wrath; perhaps they were one and the same. In addition Amesbury, one of the few Ambros- place-names outside Catuvellaunia, lies in Durotrigan territory. Other, less reliable traditions link Ambrosius with Amesbury (see below, notes 74-6). The place-name Amesbury probably means 'the fort of Ambrosius', and may have originally applied to the nearby hillfort now known as Vespasian's Camp. This hillfort could have been re-fortified during the later 5th century, as many hillforts in the south-west were, in order to provide Ambrosius Aurelianus with a stronghold. It lay south of the Wansdyke, and might have formed with it a barrier against incursions by the neighbouring Germanic territories to the north-east. Certainly, the Durotrigan kingdom would have been at the forefront of British resistance following the Saxon Revolt, making it an appropriate location for Ambrosius Aurelianus. Other possible interpretations of this evidence are that there were Ambrosian dynasties, possibly unrelated, in both Catuvellaunia and Durotrigia; or that Ambrosius senior was a king of the Catuvellauni, but Ambrosius Aurelianus came to rule the Durotriges, presumably by marriage.
That at least three generations of the Ambrosian family out of four (or four out of five according to the alternative genealogy suggested above) were politically active, commencing at a time when kings first appeared in sub-Roman Britain, strongly suggests that they formed a tribal dynasty. The basis of Ambrosius Aurelianus's power was that of a tribal king. His use of the military title dux (see note 31), along with his apparent supremacy over the other British kings, indicates that the pre-eminence of Ambrosius Aurelianus was due to him taking the leading role in a military campaign (against the Saxons) that involved other British kings.
Gildas refers only to Ambrosius Aurelianus during his near-contemporary account of a Romano-British revival that lasted for over half a century, and does not mention Arthur at all. However, in the British History it is Arthur (not Ambrosius) who is named as the leader of the kings of the British in twelve battles, the last of which was Badon. In the Welsh Annals, Arthur is mentioned in the entry for 516: 'The battle of Badon, in which Arthur carried the cross of our lord Jesus Christ for three days and three nights on his shoulders and the Britons were victors.'
Another entry in the Welsh Annals, the battle of Camlann in 537, refers to Arthur . The 'wonders of Britain' in the British History also mentions Arthur ; he is twice referred to as 'the warrior' in examples of wonders - the stone that returns if moved, and the grave with fluctuating dimensions - that show how Arthur was developing as a legendary figure even at this early stage. References to Ambrosius Aurelianus in the British History are restricted to an aside during a discussion of Vortigern's family, and the 'Tale of Emrys'. The generally mythical character of these references, which were compiled over three centuries after the event, illustrates why the British History should be treated with caution. Barely another three hundred years later, the Arthur in Geoffrey of Monmouth's History of the Kings of Britain is a totally legendary figure. Chronologically and literally, the British History stands midway between reality and legend.
The British History has a strong northern British bias. The list of Arthur's battles is consistent with this, for most of the battles that can be located with any confidence are in northern Britain - Silva Celidonis (Caledonian forest); castello Guinnion (from Roman Vinovium = Binchester, County Durham); and Breguoin (from Roman Bremenium = High Rochester, Northumberland). Badon is one of a number of battle sites in the list that cannot be precisely located, although it is generally agreed that it was located in southern Britain. Moreover, Gildas appears to have been writing for a southern British audience, for the kings he criticised were all located in what is now Wales and south-west England. If Badon was fought in southern Britain, it seems out of place in a northern battle list. It so happens that in different versions of the British History the eleventh of the twelve battles is named as either Breguoin or Agned, or both. It has been suggested that Breguoin and Agned were originally the last two battles on the list, and that Badon was inserted into the list as the last battle later, with the original name of the last battle either omitted, or conflated with the eleventh battle. There is other evidence that places Arthur in northern Britain. The battle of Camlann, where Arthur was killed according to the Welsh Annals, has been identified with Camboglanna (Castlesteads, on Hadrian's Wall). The Gododdin, a British poem of c.600 telling of a raid against Northumbria, compares one of the heroes to Arthur. At such an early stage, it is likely that this is a reference to a local warrior, before the legend began to spread around Britain. Considering these northern links, and the absence of evidence that would place Arthur in southern Britain, it seems likely that Arthur was a British leader operating in northern Britain during the earlier sixth century; perhaps he was even an early ruler of the British kingdom of Rheged. However, this does not in itself explain why Arthur's fame increased, and Ambrosius' did not.
Although Geoffrey of Monmouth's History of the Kings of Britain cannot be employed directly as a source of information on 5th/6th-century history, it can be used to identify changes in the received wisdom on that history up to the 12th century. It might even be possible to examine the processes that cause such changes. For instance, in one episode of the History of the Kings of Britain Aurelius Ambrosius encounters Ambrosius Merlin - two forms of Ambrosius Aurelianus. Aurelius Ambrosius is possibly a combination of Ambrosius Aurelianus and his father. Ambrosius Merlin is an interesting character, a prophet who also performs amazing feats, including the building of Stonehenge. Ambrosius is also identified with Merlin by Geoffrey of Monmouth in his version of the 'Tale of Emrys': 'Merlin, who was also called Ambrosius.'
In Geoffrey of Monmouth's account of the construction of Stonehenge, Ambrosius Merlin the prophet undertakes the work for Aurelius Ambrosius the king. This story is an attempt to explain Stonehenge as the burial site of British elders massacred at a parley by the Saxons, which may itself have been based on a tale in the British History that explained how the British were defeated during the Saxon revolt. The name given to the stone circles is 'Mount Ambrius,' which was also the location of 'a monastery of three hundred brethren'. Stonehenge is not on a hill, but less than two miles from Stonehenge is the town of Amesbury, and between the two is the hillfort of Vespasian's Camp - the 'fort of Ambrosius' discussed above (see note 64). Amesbury was the site of an abbey at least as early as 979, but probably not much earlier. Amesbury church is dedicated to Saint Mary and Saint Melor; the latter saint appears to have been an early (possibly 6th century) tradition from Brittany, a very unusual dedication. A possible explanation for this is that an early tradition linking Ambrosius Aurelianus with the hillfort later incorporated the adjacent sites of Stonehenge and Amesbury Abbey. Geoffrey of Monmouth's tale may have been an acknowledgement of this tradition.
Merlin, on the other hand, is an essentially northern British personality. The name is thought to be a derivation of Myrddin, the late sixth-century northern British poet. Merlin is mentioned in the Welsh Annals for the year 573:
'The battle of Arfderydd between the sons of Eliffer and Gwenddolau son of Ceidio; in which Gwenddolau fell; Merlin went mad.'
Arfderydd is identified with Arthuret in Cumbria. It may seem remarkable that Ambrosius Aurelianus could be conflated with such an enigmatic (northern) character as Merlin, but even in the ninth century Ambrosius was being credited with magical powers. The 'Tale of Emrys' in the British History credits Ambrosius with prophetic ability that enables him to confound Vortigern's magi; Merlin was the next step in the mythology.
Ambrosius Aurelianus was apparently the foremost Romano-British leader operating in southern Britain during the late 5th century. Unfortunately for Ambrosius, his successors did not rule long enough for an Ambrosian heroic tradition to develop. The southern part of the diocese, the most Romanised area, was lost during the later 6th century to renewed Germanic incursions. What native traditions did survive were from the less Romanized northern and western parts of Britain. Culturally and geographically, the historical Ambrosius was out of place in these traditions. It seems his military exploits, which seemed almost miraculous to Gildas, came to be expressed as magical feats. Meanwhile, the role of warrior/hero was gradually conferred on Arthur, a northern, British warrior. By the 12th century Arthur had become a legendary warrior. Ambrosius, on the other hand, was of secondary importance, and he had been conflated with Merlin, who was originally a secondary figure in British tradition. During the medieval period, the Arthurian cycles of chivalry and romance served to consolidate Arthur's central position in the 'Matter of Britain.' This trend has continued to the present day.
Archaeologists once deferred to the historians when considering the end of Roman Britain. As the rescript of Honorius in 410 was seen as a turning point in the fortunes of Roman Britain, the sub-Roman period was seen as being unconnected with the Roman period. In contrast with the imperial control economy, there was a tribal society with an economy limited to barter and gift exchange. Consequently, interpretation of the archaeological evidence for sub-Roman Britain was constrained by the supposition that the society and economy of the 5th century must have been essentially different from that of the formal Roman period. This view was first challenged by Richard Reece, who argued that the decline from the classical Roman model of society had begun as early as the 2nd century, so that the distinction between the archaeology of 4th and 5th century Britain was much less clear. It is now possible for the archaeology of sub-Roman Britain to be seen as part of a socio-economic continuum, albeit radically affected by (mainly) external events.
The documentary evidence also suggests that the society of sub-Roman Britain was strongly rooted in the Roman period. Indeed, many aspects of late Romano-British society continued to the middle of the fifth century, if not beyond. Gildas's 'Ruin of Britain,' and the fact that he was addressing a wide audience, are evidence of the continued provision of education in Latin to a high standard in Britain into the early 6th century. According to Gildas, no single event clearly marked the end of Roman imperial involvement in Britain. From his viewpoint, the Saxon revolt had by far the greatest impact on Romano-British society, but there was not a complete break with the old order even then.
The family of Ambrosius Aurelianus provides strong evidence in support of the model of socio-political continuity in the diocese of Britain beyond the end of the formal Roman period. The family's political influence had presumably been established during the Roman period if not the Iron Age, and it continued until at least the middle of the 6th century. The remarkable exploits of Ambrosius Aurelianus in particular serve to demonstrate the importance of the 'Roman' contribution to the character of 5th and 6th century Britain.