10 March, 2009

Chester in the seventh century: the fortress defences

Roman Chester was founded in around 74 AD as the legionary fortress of Deva, later called Deva Victrix from the name of its garrison, the Twentieth Legion Valeria Victrix. As a legionary fortress, it would obviously have been provided with a defensive wall. What were the defences like, and were they still standing in the seventh century?


The Roman defences

The initial Roman fort built in 74 AD or thereabouts had a turf and timber rampart. This may have been intended as a temporary structure from the start, and was soon replaced by a stone curtain wall. The date of completion of the stone defences is unknown, but usually placed somewhere around the turn of the first and second centuries (Mason 2001). On the south and west sides of the fortress, the Roman defences were demolished in the medieval period and replaced by defences extending to the river, so only the foundations and first few courses of the Roman walls survive to be recognised in archaeological excavations. On the east and north sides, some stretches of the Roman wall still stand to a considerable height with the later medieval walls on top.

Chester’s curtain wall was unusual, as it was entirely built using the monumental construction technique called opus quadratum, usually reserved for prestige structures such as gates. Large stone blocks up to 6 feet (1.8 m) long and 3.5 feet (1 m) wide were laid in courses 10–15 inches (25–36 cm) high, without using mortar. The wall was 4.5 feet (1.35 m) thick, reducing to 3.5 feet (1 m) at the top, and had an elaborate moulded cornice below the parapet (Mason 2001). This flashy form of construction is consistent with Chester having a higher status than other legionary fortresses in Britain (more on this in a later post).

The wall was built immediately in front of the original rampart, and the space between the wall and the front face of the old rampart was filled with rubble mixed with clay (mixed with mortar in the section between the south and east gates. This may indicate different work parties using different techniques in different sectors, or different phases of construction (Mason 2001).

How long did the defences stand?

A section of the north wall was taken down and rebuilt in the late nineteenth century, when it was found to contain re-used Roman tombstones and pieces of architectural sculpture. This could hardly belong to the original stone wall, as the fortress had only been in existence some 20 years at the time, surely not long enough to have generated large quantities of tombstones to be requisitioned for building work. Some of the tombstones commemorated serving legionaries who were married and thus probably third century. One commemorated Gabinius Felix, a soldier of legion II Augusta, and gave his legion the title Antoniniana, which was current in 213–222. The tombstone was very weathered, suggesting it was re-used in the wall no earlier than the late third century.

Further investigations have suggested two distinct phases of rebuilding of Chester’s curtain wall, both using large quantities of recycled Roman masonry and tombstones. In one phase the replacement wall was about 10 feet thick, roughly twice the thickness of the original wall, and this has been identified on the north wall immediately west of the north-east corner, and on the west wall north of the west gate.

In the other phase, the replacement wall was about 5 feet thick, roughly the same as the original wall. This rebuild has been identified on the north, east and west walls of the fortress. A section of this rebuild south of the east gate (at the Old Public Library on St John’s Street) allowed some of the sequence to be reconstructed. This showed that the ditch in front of the original Roman fortress wall had silted up completely to ground level, and on top of the soft fill lay a mass of heavy rubble including some damaged facing stones from the fortress wall. This is consistent with the wall having collapsed forwards and outwards (the wall leans outwards in the sections that are still standing further north along the east wall). This collapse might have been spontaneous or might have been deliberate demolition to make the site safe before beginning the rebuild. The rubble had been covered with a layer of sandstone brash up to 1 foot thick, and the wall rebuilt from the fourth course upwards, to the original width of about 5 feet, backed by clay instead of the original mortar, and containing at least one moulded block re-used from somewhere else. This rebuilt wall had itself collapsed, forwards and outwards again, onto the layer of sandstone brash. The rubble from this second collapse had eventually combined with the collapsed and weathered rampart behind to form a low mound. David Mason says there was evidence for a subsequent refortification, before the existing medieval wall and massive ditch was built in its current alignment probably in the late twelfth century. (Unfortunately, he doesn’t specify what the evidence for the post-Roman pre-medieval refortification consisted of).

Neither of the rebuildings can be securely dated. Mason concludes that the narrower rebuild probably belongs to the first quarter, of the fourth century and may be associated with an overhaul of Britain’s infrastructure and defences after Constantine the Great came to power. The thicker rebuild is identified as later on the basis of constructional technique. A single coin of Constantius II Caesar (324–337) was found in the thicker rebuild east of the north gate. Mason says that it means little by itself but does not elaborate; my guess is that a single coin could have been picked up accidentally along with the recycled building material and could have come from anywhere. At one point (p.211) he says the thicker rebuild could belong to the overhaul of army installations conducted by the elder Theodosius in about 370 AD, after the Barbarian Conspiracy, and at another (p. 204) he says it could be post-Roman and perhaps as late as the tenth century.

Documentary evidence

Annales Cambriae

601 The synod of Urbs Legionis [Chester].
--Annales Cambriae

This is probably the same synod mentioned by Bede for 603 or 604 AD. The use of Chester as a site for a major synod indicates that the city was still important, but does not necessarily say anything about the state of the defences.

The Anglo-Saxon Chronicle mentions Chester twice at the end of the ninth century:

AD 894.
....they [the Danish army] marched on the stretch by day and night,
till they arrived at a western city in Wirheal that is called
Chester. There the army could not overtake them ere they arrived
within the work: they beset the work though, without, some two
days, took all the cattle that was thereabout, slew the men whom
they could overtake without the work, and all the corn they
either burned or consumed with their horses every evening.

AD 907
.... Chester was rebuilt.
--Anglo-Saxon Chronicle

The 894 entry suggests to me that the English army tried to overtake the Danes before they arrived “within the work” at Chester, and therefore that Chester still had defences that were sufficiently serviceable to be of military use. The Chronicle often refers to “a work” in a context that implies it meant defensive earthworks. It also suggests to me that Chester and/or its immediate surroundings had sufficient of a population to have cattle and corn. The 907 entry suggests to me that Chester’s defences were incomplete enough to require rebuilding.


Clearly, the parts of Chester’s Roman walls that are still standing now (north wall and the east wall north of the east gate) would have been standing in the seventh century. It is impossible to be sure how much of the west and south walls were intact before they were demolished to extend the defences.

The second collapse of the rebuilt east wall south of the east gate was apparently left as rubble long enough to combine with the remains of the collapsed rampart behind, implying a long period without repair. If David Mason is correct that there was a subsequent refortification which predated the medieval wall, this is consistent with a long period of disrepair during the early medieval period, followed by a refortification either when the Danes briefly took the city in 894 or when “Chester was rebuilt” in 907. Since the wall was rebuilt to its original width and on its original foundation, it’s a reasonable first approximation that the rebuild may have stood about as long as the original. If the original stone wall was finished in the early second century (say 110 AD), and was demolished immediately before the rebuild was carried out in the early fourth century (say 310 AD), the original wall stood for about 200 years. If the rebuilt wall managed the same, it would have collapsed in the early sixth century (say 510 AD). It’s unlikely to have collapsed while the Roman Army was still in residence (i.e. up to c. 400 AD) or one would expect it to have been repaired again, and is perhaps unlikely to have stood much beyond, say, the early seventh century given the implied long period of disrepair. So we could tentatively assign a date of c. 500 AD, plus or minus a century or so either way, for the collapse of the rebuilt wall south of the east gate.

It is possible that the thicker rebuild of the curtain wall using recycled stone was part of the 907 rebuilding. However, the “refortification” of the east wall south of the east gate mentioned by David Mason presumably did not conform to the same pattern as the thicker rebuild, or one would expect him to have said so. So either there was a reason why the same fortification used different techniques, or they occurred at different times. I favour the latter as a simpler explanation.

On balance, I would suggest the following approximate sequence:

  • AD 100 approx – stone curtain wall built.

  • AD 260 – 310 approx - Chester largely abandoned while the Twentieth Legion is fighting elsewhere in the Empire. The ditch silts up and parts of the curtain wall collapse due to lack of maintenance.

  • AD 310 approx – Damaged parts of Chester’s curtain wall repaired to original width.

  • AD 370 or later – Other damaged parts of Chester’s curtain wall, which had presumably not needed repair in 310, rebuilt to double width.

  • AD 500 approx (plus/minus a century) – Part of the wall rebuilt in c. 310 collapses south of the east gate and the breach is ignored, the rubble being left to combine with the collapsed rampart. Either the breach was still defensible with a sufficiently large and determined force, or Chester was not an important defensive structure at the time, or any repair/refortification has not left any evidence).

  • AD 907 – Chester refortified by Aethelflaed Lady of the Mercians to create a defensive burh.

Full-text sources available online are linked in the text.
Mason DJP. Roman Chester: city of the eagles. Tempus, 2001, ISBN 978- 0-7524-1922-0.


Anonymous said...

Lovely column Carla,always a pleasure to read your work

very interesting is the time when Chester actually fell to the Angles.
Yes Aethelfrith was victorious in 613 but we dont know if he was claiming territory or just asserting power.
The campaigns of Cadwallon,did they reclaim the area of Chester or at this point was it still in the hands of the Brythonic dynasties?
Upon the fall of Cadwallon who claimed the Chester area?

and which Brythonic Dynastic was ruling Chester?-Gwynedd? Powys ? Rheged? or another?
Certainly because of its very placing Chester would be of great value as a defensive/offensive base for any ruler.

Meghan said...

"A section of the north wall was taken down and rebuilt in the late nineteenth century, when it was found to contain re-used Roman tombstones and pieces of architectural sculpture."

This caught my attention because when I went to Athens last summer I got to see Themistocles' wall, which was hastily built against the returning Persian army (479 BCE) and contained tombstones and whatever was at hand. It sounds like the part of the wall in Chester was repaired in haste as well...?

I also love that so many ancient things still stand in England even today. Being from California it's hard for me to imagine tripping over Roman ruins on a daily basis!

Carla said...

Y Rhechen Hen - Hello and welcome. As usual, the answer is that we don't really know what happened to Chester between around 400 AD and 900 AD. In the Roman period it was in the territory of the Cornovii, so on that basis it's perhaps more likely to have been associated with Powys than Gwynedd, though if Gwynedd absorbed the territory that peviously belonged to the Decanglii (north-east Wales) that would have brought Gwynedd very close to Chester. I would guess it was disputed between the two. I think it's probably a bit far south for Rheged, although we don't really know where Rheged was (it's normally attached to the Carlisle area, but that's not much better than a guess). If Powys was a confederacy, as is sometimes suggested, Chester could have been semi-independent (as could Wroxeter to the south) and may have changed allegiance according to political circumstance. I would guess that Aethelferth was there to assert power and take tribute, rather than to claim territory in any modern sense. If Cynddylan of the Canu Heledd poetry ruled all of Powys (as is implied) as well as his own areaa of Tren/Pengwern, then given that he was a contemporary of Penda and fought at Maes Cogwy in 642, he would be a good candidate to be in control of the Chester area after Cadwallon's death in 633/4 (and possibly beforehand; we don't know whether Cynddylan joined Cadwallon and Penda's anti-Northumbrian alliance in 633, opposed it, or stayed neutral).

Meghan - Yes, that struck me too! The archaeologist didn't comment on it, but I'm sure it wasn't standard Roman practice to pinch tombstones to build walls. The two dates he suggests for the two rebuildings are both consistent with some sort of military emergency. He attributes the first rebuilding to Constantine the Great, and evidently the fortress was in a shocking state of repair beforehand, if the ditch had filled up to ground level and parts of the wall fallen on top of it. I imagine that a legion posted to a fortress whose walls were falling down, by an Emperor who'd just usurped the job and would presumably have to fight to keep it, would repair the defences in double-quick time with whatever they could grab! The second rebuilding he attributes variously to Theodosius in c. 370 or to some unknown post-Roman date. If it was post-Roman it could be anything, but if it was rebilt by Theodosius, he was retrieving Britain after the disaster of the Barbarian Conspiracy and that would also have come under the heading of a military emergency. I wondered if I was reading too much into it, since the archaeologist hadn't commented on it, so thank you very much for the example of Themistocles' wall!

You don't trip over Roman ruins on a daily basis everywhere in Britain :-) - though there are quite a lot around if you look for them! I can't really imagine living somewhere that hasn't got historical structures stretching back into the remote past. One of the things that got me interested in history as a child was wanting to know who built such-and-such a castle or stately home or whatever.

Gabriele Campbell said...

Very interesting and informative post. I'll have to search my photos for the Chester town walls. Then I can be lazy and just link to your post for explanation. *grin*

Rick said...

I had to go to Google Maps to have the first clue about this discussion, namely where Chester is!

Now that I've done that, the natural segue from the last discussion is what was going on there, or not going on there, when the wall collapsed sometime between 400 and 600. Whoever was in control then, or for the next 300 years, either did not have the means to repair it or didn't need to.

The sort of people who fortified South Cadbury or built public works at Wroxeter presumably had the resources to repair Chester, but only if they controlled it and had use for it. You could imagine a situation in which the last thing someone wanted was Chester repaired, if it was not part of their own defense perimeter but close enough to be a threat if fortified and held by an enemy.

'Works' (usually in the plural) for fortifications is an expression I have seen many times, so it seems to have been English usage for a very long time!

Constance Brewer said...

Very cool! You make my little engineer heart go pitty-pat. Great information, thanks for sharing.

Gabriele Campbell said...

Plus the weir in the Dee river led to the old harbour and a bend in the river silting up - it's today the site of the racing course. I can imagine that also changed the way the ground reacted to downpours (already too soaked to take much, maybe) and thus the walls may have got shaky foundations over the years.

Carla, I've a little award waiting on my blog. :)

Carla said...

Gabriele - feel free! I have another post coming up on Chester in the near future too. Whether the silting up of the harbour would have altered the drainage enough to affect the wall foundations I'm not sure, but it's an interesting thought. And thank you for the award! I'll do something with it next week.

Constance - you're welcome.

Rick - should I routinely put a link to Google Maps or its equivalent, do you think? Anyway, so you now know where Chester is - at a strategic point on the Dee estuary, and in Roman times possessor of the biggest port on the north-west coast. I have another post or two in preparation on Chester and its area in the seventh century - the information wouldn't all fit in one post! - which might tie in with your questions. It's possible that the breach(es) in the Roman wall were repaired with some simple fortifications, like a palisade on top of the rubble, that might have deterred an enemy if effectively defended but that haven't left any recognisable trace. Mason's tantalising mention of "subsequent refortification" predating the medieval wall is also undated, so although it seems plausible to attach it to the burh of 907, it could be considerably earlier. Since it's behind the line of the Roman wall, I suppose the rubble could have mouldered gently away in front of it for a long time. I don't favour this, but in large part that's because I like the idea of being able to identify Aethelflaed's rebuilding :-)

Chester clearly wasn't simply abandoned as soon as the Roman army left, since the synod was held there in 601/603. (Of course it could have been abandoned after that, although I doubt it). As far as I know, none of the known early medieval kingdoms claims to be centred on Chester, so I wonder if it was a disputed border zone. Gwynedd and Powys might be expected to have an interest (see reply above), and Mercia, Rheged, Northumbria and/or Elmet might have had some territorial interest at various points in their history. It may not be an accident that Aethelferth of Northumbria fought a major battle against Powys (and perhaps also Gwynedd, and maybe others as well) at Chester in 613/616. One thing that isn't often discussed is the salt trade. Salt in Britain comes from sea water, or from a handful of inland salt springs, three of which are close to Chester. I do wonder if control of the salt works was fantastically profitable and worth fighting for, independent of any "strategic" considerations. Early medieval kings ate their way round their territories like medieval kings, so perhaps Chester was a place where the king (of whichever kingdom happened to be in control at the time) turned up once a year to collect his rent or tribute or tax. (If the inhabitants were really unlucky, perhaps three or four kings turned up in some years....). In this scenario, it's not worthwhile for any of the kings to put the effort into repairing the defences because they're not going to maintain a garrison so far from their heartlands.

It's also worth bearing in mind that although Roman fortifications look obviously useful to us, they may not have been as useful as we think in the circumstances of post-Roman Britain. You need a LOT of men to defend something as big as a Roman legionary fortress, otherwise the enemy just sneaks in round the back with a few ladders and grappling irons. Unless a sizeable population lives in or near the fortress, and is able and willing to defend it, the walls themselves may not be much use, so no need to bother to repair them.

Rick said...

Map links would be great! Do other 'Murrican lovers of British history do as they should, and sit down with an atlas? Or are they like me, and most of the time content to have no earthly notion where the stuff they read about is going on. It is enough that it has wonderful and quaint names, and presumably looks more or less like the Shire.

Except for Wales, which is sort of like where Frodo had a tipple, took a wrong turn on the motorway, and somehow got onto Highway 1 headed for Big Sur.

Decaying fortifications don't necessarily say a thing about the town - it could result from a time of peace. Flip side, well kept fortifications don't prove that a town is thriving, only that it is strategic.

Good point also about the usefulness or otherwise of 'work' this big. I infer that Aethelflaed had a big army to call on, in the thousands - about what Chester was designed for in the first place, in fact.

And it isn't just army size. 'Petty kings' would be hard put to rebuild and garrison it, but wealthy and powerful rulers can also have a positive allergy to fortifications.

Anonymous said...

Sorry Carla,I have a great problem with Cynddylan and his largely inflated heroic status!
A huge modern propagated myth is the geography and pseudo Historization of the area of Pengwern.
According to our modern `experts` ,the Kingdom of Powys lost its Eastern territories(modern day Shropshire) to the Kings of a Gwynedd sub-Kingdom called Dogfeiling whose rulers held the territory until it was destroyed by Northumbrian Kings and annexed to Mercia in the mid seventh century.

Let us look at a clearer explanation of what has happened and the true identification of the area of Pengwern.
The Area of Dogfeiling is named after Dogfael ap Cunedda and roughly covers the area East of the River Clwyd it was bordered by Lal, which is roughly modern day Denbighshire. The area of Lal was the ancient seat of the Cadellings who ruled from the Lal area and eventually who ruled over all Powys until the 11th Century.

Our first mention of the pedigree of Dogfael is in the Jesus College MS 20

[M]euruc m elaed m elud m glas m elno m docuael m cuneda wledic.

However this pedigree has been mixed with the Harleian MS 3859 #14

I]udnerth map Morgen map Catgur map Catmor map Merguid map Moriutned map Morhen map Morcant map Botan map Morgen map Mormayl map Glast, unum sunt Glastenic qui uenerunt [per villam] que uocatur Loytcoyt

The two previous pedigrees have been merged with the Welsh poems of the Canu Heledd, thus the Glas of Harleain MS 3859 becomes the same as Glas of the Jesus college MS 20.We suddenly find our King of Dogfeiling also ruling the area of Glastenning and also due to misinterpretated readings in the Canu Heledd all of Shropshire also!
Lets have a look at the two poems `Canu Heledd` and `Marwnad Cynddlan`

Cynddlyan ap Cyndwrn ,King of Dogfeiling and oppressor of the Cadelling: we know from the poem the Marwnad Cynddylan that his wars are against the Cadellings of Powys who would of course be his border neighbours. Now forgive my geography but since when has modern day Clwyd`s borders ever been modern day Shropshire? Why would Pengwern be right the other side of Lal?
The misidentification has been from one line in the Canu Heledd

`I have gazed from Wroxeter`

and the misidentification of Caer Lwtgoed

`y rhag Caer Lwytgoed, neus dug Moriael`

The usual identification of which is Lichfield which does have a point as is mentioned in Nennius 9th Century `Historia Brittonum` as such .However Caer lwtgoed means `Fort in the Grey wood` and one must again wonder Geographically what a sub-King of Gwynedd was doing in modern day Staffordshire! Then the use of the name Caer lwtgoed has been further embroiled with the Harleain MS 3859 pedigree of Glastenning due to the line

Glastenic qui uenerunt [per villam] que uocatur Loytcoyt

So the whole History and Geography of Pengwern and glorious career of King Cynddylan have been built all on the shoulders of the modern historian.
The reality of the situation is that Cynddylan was nothing but a sub-King of Gwynedd who had some success against his Eastern Border, his prowess and death were recorded in poetic verse and magnified by modern historians. He never conquered Shropshire or Glastennin though he probably along with other British Kings was allied with the Anglo-Saxon King Penda of Mercia.
So where was Pengwern? Even on a modern map it continues to be there, in Clwyd is a small village not far from Rhuddlan called Pengwern its believed the hall of Cynddylan stands now within the grounds of Bodelwyddan Castle


Carla said...

Rick - Okay, I give up - you will have to explain your comment about Wales and Frodo on Highway 1 to me. I'm as lost as Frodo :-)

Alfred's burhs were set up for defence by a citizen militia. The size of each burh was determined by the population it served, with 1 hide (a hide was the amount of land required to support one family, roughly speaking, although as you would expect (!) there are disagreements about its definition) = 1 man = 4 yards of wall. The population served included the people living in the burh and those living in its immediate area (within about 20 miles or so). When danger threatened, the people living outside were supposed to move into the burh for their own protection and to contribute to its defence. So not an army as such, but a defence force of sufficient size - by definition - to defend the walls. Some burhs re-used existing fortifications, Roman or otherwise, others were constructed de novo, and many became thriving market towns. Google for "Burghal Hidage" if you want details of Alfred's system in Wessex. I imagine Aethelflaed (his daughter) applied the same system or something very similar in Mercia.

Y Rhechen Hen - there are a lot of interpretations of the history (if any) behind medieval Welsh poetry and genealogy. Topographical place names could be purely descriptive and could be used for several places independently, like all the "Ben More" (Big Hill) mountains in Scotland. Caer Luitcoet is usually assigned to the ex-Roman station of Letocetum (modern Wall by Lichfield), but there could have been more than one place called 'fort of the grey wood'. I think Geoffrey of Monmouth places Caer Luitcoet, or something like it, at Lincoln, and there are several Pengwerns on the modern map. Cynddylan (if he existed) might have had a court at any of them, or at another location that now has a different name.

Rick said...

Highway 1 is the California state highway running most directly along the coastline, and the portion through Big Sur in Monterey County is reputed to look rather like Wales. It is also old time California counterculture country, where some people would hardly raise an eyebrow if Frodo showed up.

I meant 'army' in the sense of troops available for muster, not just regulars. As militia armies go, I imagine the English one by this time had had a fair number of veterans, and was more like reservists than a mere armed crowd milling around.

No one in western Europe had any large number of 'regulars,' that I know of, between the end of the western empire and 15th century France.

Carla said...

Rick - thanks, all is made clear!
Quite right that nobody went in for large regular armies much. Whether burh defence would come under 'troops available for muster' I'm not quite sure. If I understand it correctly (and I haven't looked this up), there seem to have been two types of military service under Alfred, not counting the 'professional' warriors of the aristocracy. The burh system required one man per hide to defend the burh when needed, and one man per five hides had to serve with the king's army for part of the year. The difference being that if you were a burh defender you only had to serve in your local area (a sort of Home Guard, if you like), while if you were in the king's army you were expected to fight away from your local area if need be. The king's army service might be closer to modern reservists, though I would expect there was a lot of overlap, and no doubt people who had served in the king's army would also serve in their burh defence, either instead (perhaps when they got a bit older?) or as well as army service.