London in the 5th century


English Heritage


Year published





Comment on this article
Double click anywhere on the text to add an annotation in-line

Citation Show another format:

'Introduction: London in the 5th century', An Inventory of the Historical Monuments in London, Volume 3: Roman London (1928), pp. 62-68. URL: Date accessed: 18 April 2014. Add to my bookshelf


(Min 3 characters)

8. London in the 5th Century.

In the year 410, as recorded by the 5th-century historian Zosimus, the emperor Honorius, hard pressed by the Visigoths in Italy, formally delegated to "the cities in Britain" the initiative in home-defence. There is no clear evidence that this responsibility was ever resumed by the imperial administration, and the weight of both the literary and the archæological material is strongly against the probability of any effective link between the province and the central authority after that date. (fn. 1) It is sufficiently clear that Romanized Britain was left to drift, with only such impetus as remained to it from more than three and a half centuries of imperial direction. At first the changed circumstance may scarcely have been perceptible in the normal life of the province, if only because the unrest and bloodshed which had characterized the last half-century of the weakened Roman régime could hardly have been aggravated in the years immediately following its final lapse. There is at least no good reason for ending the story of Roman London with the rescript of Honorius, and it is necessary therefore to include a brief discussion of the problems relating to the city in the uncertain period which heralded the birth of Anglo-Saxon Britain.

Historically, the evidence is of the slightest. The only direct mention of London which need concern us is that of the Anglo-Saxon Chronicle under the year 457, when the Britons are said to have sought refuge in the city after their defeat by Hengist at Crecganford (possibly Crayford in Kent). This might be regarded as witness for a London still effective after the middle of the 5th century, but that the whole account belongs to a 9th-century compilation which is thus too late to stand by itself. Of more use is the less explicit evidence of the 5th-century life of St. Germanus, who visited Britain twice, in 429 and 447, and at least on the former occasion found municipal governance still in force in the south-eastern part of the island. (fn. 2) The historical record is perhaps just strong enough to suggest that towns such as London and Verulam were able to maintain a real civic status something like half a century after the rupture with Rome. (fn. 3) The Picts, Scots and Saxons, however, were an ever-present peril. The Scots or Irish had been raiding and settling upon the western coasts of Britain before the end of the 3rd century, and in the same period the imperial authorities had found it necessary to fortify a "Saxon Shore" against Teutonic invaders on both sides of the English Channel. By the second half of the 4th century, Roman Britain was struggling between the scissor-blades of simultaneous Irish and German attack. Thus in a single year (367) both the officer in command of the defences of the Saxon Shore and his colleague in command of the northern British frontier were captured or slain in onslaughts from the two directions. Small wonder that the contemporary historian, Prosper of Aquitaine, was able to write that in 409, on the eve of the break, "the strength of Britain was desperately attenuated by the weakness of the Romans." By building or rebuilding forts along the eastern and western coasts of the island, the Roman authorities had made a last despairing effort to hold the province against its double foe. Under the chaotic conditions of the time, the effort was foredoomed to failure. And twenty years afterwards, when Germanus came to denounce the Pelagian heresy, the saint went from the synod at Verulam to lead the British forces to victory against a joint invasion of Picts and Saxons at some unknown spot in the midlands or the north.

Prosper, writing in Gaul and therefore presumably with some real if fragmentary knowledge of contemporary events across the channel, states that in 442 "Britain up to this time, harassed by various disasters and vicissitudes, was brought under the domination of the Saxons." (fn. 4) Four years later, according to the less substantial authority of the 6th-century Jeremiads of Gildas, the famous letter bore the "groans of the Britons" in vain to Rome. Nevertheless, in the year following this appeal Germanus was able to repeat his visit to this country and seems to have found some part of it still sufficiently Romanized and free from military distraction to listen once more to his denunciations of Pelagianism. Nor is there on this occasion any hint of a martial interlude such as that which had made him the hero of the "Hallelujah Victory" in 429.

These familiar passages from the earlier historians suggest that in the 5th century, as in the 4th, the island was swept from the north-west and the south-east by waves of invasion which were sometimes beaten back (as in 429) and sometimes carried all before them (as in 367 and 442). Every now and then, like rocks amongst the breakers, the old Romanized cities of the south-east seem to rear their heads above the flood with something of the obstinacy of the contemporary cities of invaded Gaul. How long they survived complete submersion, if such was ever their fate, the historical evidence does not tell us. Gildas, writing in the middle of the 6th century, says that "even our cities are not now inhabited as they were of yore"; but he clearly knows little of Britain east of the Severn, and in any case his statements are fraught with an almost hysterical exaggeration and more often than not suggest vague and violent generalizations based merely on particular instances. His evidence scarcely affects the problem of south-eastern Britain. Historically, there is no valid obstacle to the postulation of an enduring if attenuated London throughout the 5th and 6th centuries.

If we turn now to the archæological evidence, the problem may be approached from two sides—the Roman and the Saxon. It is unfortunate that the numismatic material ascribed to London is generally too uncertain in origin to justify discussion (fn. 5) and we must look elsewhere for help from this type of evidence. The only excavated Roman site of possible significance in this context is Richborough in Kent, where excavations carried out over a period of six years within the area of the Saxon-Shore Fortress have yielded many thousands of Roman coins. Amongst them, attention has already been drawn to the very large preponderance of late types. (fn. 6) That these were rarely minted later than c. 395 is of little moment, since the issue of copper coinage in Gaul virtually ceased about that date. As Mr. Salisbury remarks, "the fact of the cessation of bronze coinage in the west would act like a dam behind which the money would accumulate at the geographical limit of circulation." How long and in what area did this late coinage thus accumulate?

The first question is difficult or impossible to answer. From the figures prepared by Mr. Hayter and Mr. Salisbury, the late coins found at Richborough between 1922 and 1926 are as follows:—

Coins minted between 320 and 364: 3,495.

Coins minted between 364 and 398: 12,569.

It cannot be assumed that the relative value of the coinage or the density of the population at Richborough remained anything like constant from the beginning to the end of the 4th century, and it is therefore unprofitable to attempt to calculate from these figures the exact chronological implication of the great accumulation of coins during the latter part of the century. But, as Mr. A. W. Clapham points out, the fact that the coins minted during the last 35 years of the century outnumber the total minted during the preceding 45 years by some 320 per cent. at least suggests the possibility that the former represent an occupation carried well on into the 5th century.

The second question is less difficult. Mr. Salisbury and others (fn. 7) have collected evidence to show that late silver coins (of Honorius and Arcadius) occur in hoards in Suffolk, Essex, Middlesex, Berkshire, Hampshire, Wiltshire, Dorset and Somerset, i.e. in the immediate hinterland of the Saxon Shore. (fn. 8) The copper coins of the same emperors are much more widely spread, although none occurs on Hadrian's Wall and only three or four scattered in Wales outside Monmouthshire. But apart from some twenty copper hoards, extending from Northumberland to the Severn (though mostly in the southern counties), (fn. 9) very few of these late coins are found on any single Romano-British site, and the complete contrast with the thousands found in the soil of some two acres of Richborough is striking in the last degree. It is clear that, as Mr. Salisbury infers, we may visualize the submergence of Roman Britain as proceding from north-west to south-east, i.e. as being due primarily to the Picts and Scots rather than to the Saxons. Some part of south-eastern Britain survived, at least for a while; but how far the numismatic evidence so liberally furnished by Richborough will be matched by other sites in this region remains to be seen. Verulam, which received Germanus in 429, may be expected to provide an analogy; and, if so, London may, on this ground alone, march securely between Kent and Hertfordshire into the 5th century.

The relative evidence from Saxon sites is slight but suggestive. The earliest Saxon remains in England are probably two burials found many years ago at Dorchester in Oxfordshire. (fn. 10) These were ascribed to a period as early as the 3rd century by Sir Arthur Evans and to the 4th century by Dr. Salin, but have more recently been assigned to "the early half of the 5th century at the latest" by Mr. Thurlow Leeds. In Berkshire, at Frilford, a Saxon cemetery immediately adjoined a late Roman one in such a way as to suggest something approaching continuity of use. The neighbourhood of the Lower Thames has produced early Saxon cemeteries (burnt burials) at Croydon, Mitcham, Beddington and elsewhere. There is certainly sufficient evidence to indicate that the Saxons were penetrating far into the Thames Valley within a very few decades of the rescript of Honorius, if indeed they had not already formed occasional settlements there actually before that date. It is permissible to imagine their small fleets passing up-river beneath the closed gates of a London which may have regarded them with the same wary indifference wherewith, four centuries later, Saxon London often enough regarded the roving long-ships of the Vikings. (fn. 11)

To complete this picture, it may be recalled that by the end of the 4th century, whatever the state of the towns of Roman Britain, the countryside had to a large extent been shorn of the best of its Roman elements. No Romano-British country-house seems to have yielded any evidence of occupation after the third quarter of the 4th century, and the anarchy which accompanied and followed disasters such as those of the year 367 appears to have put an end to Romano-British country-life on any significant scale. Roman Britain was driven into its walled towns and its fortresses, and the open country with its half-civilized peasantry and bands of brigands was now easy prey to the groups of Saxon yeomen who came as settlers in the van of the main Teutonic immigrations. The newcomers were thus not, of necessity, brought into close contact with the remnants of Romano-British civilization, and under favourable conditions, particularly where (as in London) the commercial rather than the territorial factor was paramount, the urban life which was foreign to the German tradition may well have continued to exist amidst an increasingly Teutonized countryside. Without pressing too closely analogies from Merovingian Gaul, it may be pertinent to remark that there the Germanic "invasions" were "often far less subversive of the old social order than we might at first suppose," and, apart from intermittent outbursts of robbery and violence, "ordinary Gallo-Roman life probably went on as it had done for generations before the Visigoths appeared at Toulouse or the Franks at Cambrai." (fn. 12)

In summary, the evidence from all sources lends itself to the following interpretation. On historical grounds it may be affirmed that London existed until after the first visit of St. Germanus in 429, and probably until after the second visit in 447. With less assurance, the period may be extended to 457, the date ascribed by the Anglo-Saxon Chronicle to the rout of Crecganford. Generally consistent with this evidence—though not at present susceptible to certain interpretation—is that of the abundant late coinage recovered in recent years from Richborough, the only important fortified site yet excavated scientifically in south-eastern Britain. It is clear, therefore, that on certain sites in this part of the country—London amongst them—occupation was continued until well after the year 410.

This is the less surprising when it is remembered that the Saxon immigrants, though savage enough on occasion, were primarily not (like the Picts and Scots) destroyers, but armed yeomen in search of comfortable patches of riverside-land where they could resume their comparatively settled rural existence. (fn. 13) Upon them the rescript of Honorius in 410 conferred, in a sense, the freedom of south-eastern Britain. True, the older histories have usually exaggerated the importance of the rescript, and so have tended to obscure the essential continuity of the events of the half-centuries preceding and following it. But it did at least imply that such co-ordinating authority as then remained to the Imperial Command was withdrawn. "The cities" of Roman Britain were now left to fend for themselves—a situation which was probably not so much created as aggravated by the formal severance. Often enough during the 4th century the central power had been lacking in times of crisis; but there was now, after 410, even less guarantee of concerted action, and no Imperial reserve from which a Theodosius or a Stilicho could be drawn at need. The fenced cities remained, but the defensive screen between and around them was now of little substance. Through its gaps, up the Thames and Medway valleys, small groups of Saxon settlers were thus able to find their way, with little hindrance or none, into an open countryside which for a generation had lain almost desolate. Relics of them dating from the early years of the 5th century, have been found as far up-river as Oxfordshire.

The immediate result of the lapse of centralized control in Roman Britain may thus for a time have been actually to dimmish conflict with the Saxon immigrants. The Imperial defence, by damming the tide of invasion, had doubtless augmented its violence, and the partial breaking of the dam after 410 may be supposed to have reduced the pressure, at the expense merely of the flooding of a countryside that was already abandoned by the more civilized elements in the population. There is no substantial reason why the few surviving walled cities should not have remained as islands in the flood. To the Saxon farmer of the 5th century they must have counted even for less than did their successors to the Danish Vikings who in the 9th century sometimes took toll of them and sometimes passed them by. There is a danger, in approaching the problem of the Romano-British towns in the 5th century, of exaggerating their importance and of regarding the contemporary Britain from their standpoint. In reality, save as occasional links with the Latin world, their importance had almost vanished. The centre of interest was already shifting to the new population which had begun to dominate the country districts. In rural Gaul, the Frankish invaders found a Roman villa-system in full working order; they insinuated themselves into it, and it in turn gradually Romanized and even urbanized them. In Britain, as we have seen, the corresponding villa-system, in so far as it had developed at all, had been wiped out, largely by the Picts and Scots, during the last half-century of Roman rule. Here, therefore, there was no link between the essentially rural German and the essentially urban Roman civilizations. (fn. 14) The walled towns of the south-east, out of reach of the Picts and out of mind of the Saxons, may thus be thought to have lingered on almost as "reservations" for the secondary Romano-British population. The silence of history in regard to them is probably just; London in the year 500 can have mattered little to anyone save to a few decivilized sub-Roman Londoners. Only when, in the course of the 6th century, a settled and wealthy Germanic aristocracy (particularly amongst the comparatively civilized Jutes) began to emerge and to renew contact with the Latin world, did the cities begin once more to come into their own. The process was sealed by the reintroduction of the centralizing ministration of Roman Christianity, and it is no accident that the history of London is resumed with the advent of St. Augustine and the revival of the ancient bishopric in 604.

One further aspect of the problem may be noted. The view here taken that, in spite of occasional outbreaks of violence, the Saxon settlers of the 5th century probably continued—as they certainly began—to settle in the country districts of south-eastern Britain without completely obliterating the walled towns, has perhaps a secondary implication. The most consistently destructive enemies of Roman Britain had been the Picts and the Scots. With them the Saxon immigrants, or some of them, occasionally joined forces, as on the occasion of the victory of Germanus in 429. But more often, as their fundamentally different traditions and intentions rendered inevitable, the two groups of invaders were hostile to each other; and it may not be altogether unreasonable to suggest that, paradoxically enough, it was the early penetration of Saxon settlers into the south-eastern part of the island that incidentally saved some of the remaining Romano-British cities there from destruction. Having plundered their way as far south as the midlands, the mountaineers from the north found in front of them not merely some of the strongest Roman towns in the land, but, between them, a countryside that was already filling up with determined Saxon settlers who had as little desire for disturbance as the towns themselves. Regarded in this light, the Saxons of the early 5th century may actually have done more to preserve than to destroy the surviving elements of Romano-British civilization. It is not necessary to suppose that they went so far as their Frankish kinsmen in Gaul, who even found it politic to shoulder the defence of Gallo-Roman life against their more subversive rivals, the Huns and Visigoths. But it is easy to imagine that the contemporary circumstances in south-eastern Britain were not wholly dissimilar, and that the infiltration of the German warrior-farmers may there also have done something to grout the broken fabric of Romano-British civilization, and so have enabled it in its walled towns to escape that complete destruction which seems to have overwhelmed the more remote and exposed cities known to Gildas.

Fig 8.


1 For the well-known discussion on this point, see J. B. Bury, Journ. Rom. Studies, X, 131; R. G. Collingwood, Ib. XII, 74; F. S. Salisbury, Antiq. Journ., VII, 268, and Num. Chron., 1927, 108; also C. Oman, England before the Norman Conquest (1927), 174.
2 Germanus and Lupus refuted the Pelagian heresy in a synod held at Verulam. Mention is made of "a man of tribunician rank," and of wealthy magnates [see C. Oman, England before the Norman Conquest (1927), 195].
3 It is scarcely necessary to refer here to Geoffrey of Monmouth's Lot, a member of the royal house, who was established by King Uther and again by his son King Arthur in" the consulship of Londonesia and the other provinces belonging to him," and led the British army against the Saxons at the end of the 5th century, with indifferent success [Historia Britonum, VIII, 21, and IX, 9]. According to the same romancer, London, Winchester and other places remained more or less intact throughout the 5th and 6th centuries. Unfortunately the Historia is not evidence, although its general testimony happens to coincide with the view taken in the present section.
4 The precise significance of this statement is impossible to estimate. It may reflect some unusually violent but comparatively local wave of conquest such as that which then or shortly afterwards established the Jutes in Kent.
5 Not only are many of the coins, ascribed to London, of uncertain origin, but the upper deposits which should have contained evidence relating to the latest Roman occupation have long ago been removed in the course of building operations.
6 F. S. Salisbury, Antiq. Journ., VII, 268; and Num. Chron., 5th Ser., VII (1927).
7 Notably, Sir Arthur Evans, in Num. Chron., Ser. 4, XV, 498.
8 A remote hoard such as that which included coins of Honorius and Arcadius near Coleraine, Co. Deny, was obviously plunder and thus helps, if anything, to emphasise the restricted distribution noted above.
9 See lists collected by R. G. Collingwood, Journ. of Rom. Studies, XII, 90.
10 E. Thurlow Leeds, The Archæology of the Anglo-Saxon Settlements, 55.
11 As Mr. Leeds remarks, " The whole results of archæological research .... prove conclusively that the Saxon immigrants made full use of the Thames route, unhampered by London or its inhabitants." Op. cit., 54.
12 S. Dill, Roman Society in Gaul in the Merovingian Age, 12, 25, 38, etc.
13 So also, as noted above, the warrior-farmers of Frankish Gaul. "Their early kings were men of war. But when they came to settle on the fertile lands of the Scheldt and Moselle, their chief object must have been to secure a tranquil, prosperous life beside their Gallo-Roman neighbours. They were a keen and greedy race, fully appreciating the possession of lands and flocks, and herds, and eager for such wealth as these could yield " [S. Dill, discussing the evidence of the Salic Code, op. cit., 60; cf. 12, 25, 38, etc.].
14 See R. G. Collingwood in Antiquity, I, 117.