The Roman fort at Felixstowe in Suffolk stood on a sandy cliff. It went into the sea between 1700 and 1750, and there are still remains of it on the sea bed, a few metres off-shore. I collected some old drawings of the fort here, known as Walton Castle.
Today I saw on twitter here an aerial photograph of its sister fort in Norfolk, Burgh Castle. This has partly fallen into the river, and reminded me extremely strongly of those old drawings.
Here’s the aerial photo:
And here is the old plan of the vanished Roman fort at Walton Castle:
The walls of Burgh Castle still stand, and look like this:
And the drawing of the walls of Walton is here:
Note that this is taken from the landward side. The similarities are obvious.
If I had a drone, and could take an aerial photograph of the area where Walton Castle stood, it might be possible to “drop in” the image of Burgh, and recreate the appearance of the vanished fort. Maybe some day I will.
Regular readers will know that I have been attempting to locate a diving report from 1969 – the “Errington Report” -, which surveyed the ruins on the sea bed. I have been trying to locate this for a year. It is supposed to be at Suffolk Record Office. Sadly I have been unable to examine the relevant files so far.
Lidar is a technique for displaying the shape of the ground using pulses of laser light. The results have been widely used to discover Roman monuments, as they can process them to omit modern buildings, trees, etc. I have been interested in this ever since I discovered some Lidar images of the seabed showing the submerged ruins of the Roman fort of Walton Castle at Felixstowe.
Most of the United Kingdom has been surveyed using Lidar, and the resulting datasets are now freely available for download on the Environment Agency / DEFRA website. If you can download them, then you can pull them into a tool like QGIS, and turn the data into images. But this website is not well organised. I have never been able to work out how to download anything!
Partly this is because I used my Android mobile much of the time. Just don’t. You won’t be able to get it to work. Instead go to your trusty PC and open your browser. I did all this in Chrome.
[Update: 19 March 2024: The website has changed, so I have updated these instructions.]
1. Go to the Defra Home page, and search for LIDAR
Go to https://environment.data.gov.uk/, ignore all of it and put LIDAR in the search box. The next page gives you more options to search:
Change the search term to “LIDAR Composite” (without quotes)
Change “Organisation” to “Environment Agency”
Change “Sort by” to “Alphabetical A-Z”
Change “Results per page” to 100.
You currently (March 2024) get 14 results, which look like this. (Click on the image for a larger image) (You can get more results if you omit “composite” but none looked useful).
If you scroll down, you will find “LIDAR Composite … DSM – 1m” and “LIDAR Composite … DSM – 2m”, and LIDAR Cmposite … DTM – 1 m” etc.
The LIDAR Composite DTM and DSM materials are what you want, taken at various resolutions. DSM is the raw data. DTM removes surface objects like trees and houses.
Apparently “WMS” is is a way to get the datasets from the GIS tool directly. Not what I have in mind here at the moment.
The one you want is the “Download the Survey Data” which I have highlighted in red.
3. Click on “Open The Link”
This takes you to https://environment.data.gov.uk/survey, which after a pause builds the following inscrutable screen:
Ignore the “Download” box for now!
The bit you will need is the square. But… NOT YET! It is, in fact, a tool to draw an area on the map. We’ll use it in a moment.
4. First, zoom into the area that you want to look at
This bit is fairly obvious. Use the “+” icon to zoom, and drag the map around. Once you get far enough in, a grid will appear with references on it. If you know the reference, you can enter it in the search box, although I notice this sometimes does not work.
I’m using the area off Felixstowe, so I get to this.
Until you are zoomed in, you can’t do anything. You can only download datasets for small areas, you see. But this is probably enough.
5. Draw a polygon on the map of the area for which you want Lidar
Now at last you can click on the “Draw polygon” button! So long as you stay over the map, it is surrounded by a grey area with a yellow border. If you move off to another window, it cancels this, tho! But now you can draw. (This frankly can be pretty tricky too.)
Hover over the map at one corner of wherever you want to draw, and click. A small white circle should appear. If it doesn’t, look back at the “Draw polygon” icon – it’s probably lost focus.
Now move the mouse. Click again, and you’ll get another white circle, and a blue line back to the last corner. Here’s a screen grab, (although the cursor went everywhere!)
Repeat until your area is right, and then click back on your first circle to close the area. At that point the “Get available tiles” button on the popup will return to a normal green colour, and you can use it.
It will now look like this:
Note my polygon on the map.
Now, at last, you have something you can download. So click on the “Get Available Tiles”. You will now get a new menu:
What this lists is the various different types of dataset. In fact it lists the lot, of all sizes and resolutions. Whatever you choose, you get the link in blue at the bottom.
But there is more. Notice how the scroll bar to the right shows that the window is scrollable? If you scroll down you get this!
The link is to a zip file. In Chrome, just click on it to download to the Downloads folder; in IE, right-click and choose “Save target as” in the usual way. Either way you will end up with a lidar_composite_dtm-2022-1-TM33nw.zip file on your PC.
In my case, clicking “Download all” had the same effect.
[Note: I have not updated the following notes since 2019, so I don’t know if these work. I believe the bit above is what causes people trouble!]
6. Unzip the dataset
How you use the dataset is a different question, but I will give you what I found out.
First, you need to unzip the dataset. I use 7Zip on my PC, and right-click, 7-Zip, and extract to folder. So…
That created a folder Bathy-Coastal-Multibeam-2013-TM33nw in that directory.
I’m more interested in the DTM 1 meter stuff, so I get a download of LIDAR-DTM-1M-TM33nw.zip, and unpack it to a folder LIDAR-DTM-1M-TM33nw.
Inside the new folder are a bunch of .asc files. These together make up the dataset.
First, I installed the latest version of QGIS from the download site, which for me was 3.4.5. Look for the “long term stable release” stuff, and ignore the rest. This installed fine, and created a folder on my desktop, labelled QGIS 3.4, and an icon, “GRASS GIS 7.6.0”. Now … do NOT try to start that icon. Instead drag it into the folder, and forget about it.
Next open the folder, and double-click on the QGIS Desktop icon, again ignoring the GRASS thing. This will open something you can work with.
Next, create a project by Project -> New. Then do Project -> Save, and choose a name for your QGZ file – I used my own name.
Next, you need to import the dataset. Raster -> Miscellaneous -> Merge brings up a daft dialog box headed “Merge”.
Click on the “…”, and you get another daft dialog box headed “Multiple Selection”. Click on “Add”, and browse into the folder LIDAR-DTM-1M-TM33nw.
Select all the files in the folder, and hit “Open”. They will all appear in the “Multiple Selection” box.
Now hit “OK”. You’ll be back at the Merge dialog box.
You’ll want to save the resulting .tif file, so under “Merged” there is “Save to temporary file” – hit the “…” next to that and choose “Save to file”, and then pick a name.
Your “Merge” dialog will now look like this:
Don’t twiddle anything else.
Now hit Run, and go and make a cup of coffee. It takes a while.
When it finishes, it will pop up “Algorithm ‘Merge’ finished”, and look like this:
Hit “close” to get rid of the dialog box. You now have some results.
You can use the mouse to drag it around, and zoom in. The results are likely to look… disappointing.
On the left side is a box “Layers”. If you right-click on “Merged”, and choose “Properties”, you get stuff that you can play with. Select “Symbology”, and you can change the “Render type”. You can change it to “Hillshade” (whatever that is), and hit “Apply” and you get more details:
But that’s as far as I could get.
However, it IS more than I knew before.
Bibliography
Some links that I found useful:
http://apps.environment-agency.gov.uk/wiyby/151365.aspx – overview of the datasets
https://www.gislounge.com/what-is-a-shapefile/ – the datasets are “shape files”
https://www.gislounge.com/shapefile-viewers/ – possible viewers
https://houseprices.io/lab/lidar/map – Easily the best way to view Lidar. Only works on Chrome tho. Based on the 1m DSM data. Actually better than anything I got from this!
Walton Castle is the local name for the remains of a Roman fort, now submerged beneath the waves offshore at Felixstowe in Suffolk, Britain. Resources for study of this monument are limited, and I have discussed them in other posts.
One interesting article appeared in, of all things, a popular magazine. Such an item is, of course, not usually held by research libraries so can be very hard to locate. The item is Julian K. Hagar, “A New Plan for Walton Castle”, Popular Archaeology Today, vol. 8, no. 1 (February 1987), p.22-25. Fortunately today I was able to find a copy. In view of the difficulty of locating it, I will make a scan of the article available here:
Hagar’s analysis of the sketchy evidence is deeply convincing. He points out that by 1623, the earliest date at which we have knowledge of the site, the eastern side of the ruins must already be long gone. The cliff did not erode quickly, and indeed has only eroded another 50 metres in the last 200 years.
A sketch from 1623, reproduced in the Victoria County History of Suffolk, shows the west side of the fort more or less complete. The wall was perhaps 180 metres long, if we assume the measurements given in 1754 were copied from an earlier, now lost, account. Hagar suggests that this view is taken from the land-ward side, that the sea can be seen through the “gate”, and that we should disregard “cliffs” in the foreground as artistic license.
On this basis, he proposes this map for the castle in 1623:
By 1722 much of the remains had fallen into the sea, and the length of the west wall was only about 100 metres.
By 1750 it was all gone.
The highest point remaining, in the water, today, is still the ruin of the south-west corner tower and a fragment of each of the adjacent walls. This tends to favour Hagar’s theory.
Another point in favour of Hagar’s reconstruction is the north alignment that he gives the ruins, which does indeed seem to be the case. It is unfortunate that I was unable to locate the manuscript of the 1969 survey, which would have clarified this point.
But on the other hand, there are two obvious problems with Hagar’s plans.
The first is that the beach does not today run NW-SE as he shows it. It probably never did. It might have run N-S, but today it certainly runs SW-NE. The northern end of the ruins is therefore closer today to the beach, not the southern end.
The only feature of the landscape that would permit more erosion at the N end is the presence of “the Dip”, the shallow valley carved by the stream that runs into the sea at that end, which is still there today. But this seems doubtful.
The second point leads us in the same direction. If we agree with Hagar that the 1623 drawing was made from the landward side – for how else could it be drawn? – then the drawing shows quite a bit of masonry on the north side. This is equally obvious in what is probably the original of the drawing:
The map at the foot of the drawing is of the same period. Hagar is probably right to treat the east wall as an artist’s guess; and to suppose that the “corner towers” drawn at that end were in reality bastion towers, mid-way along the wall. The fort, then, was square, just like other Roman forts of the Saxon Shore.
But in Hagar’s favour, we can see that within the map are some ruins at the SW corner, possibly the remains of a demolished medieval castle keep mentioned in the sources. These must have been visible in 1623 or they wouldn’t be noted.
I would therefore suggest that perhaps the beach in 1623 did indeed run North-South, and preserved more of the North wall than Hagar allows. Perhaps the north side eroded faster – who can tell? But the hypothesis that half of the wall fell first does indeed seem to be correct.
It is very unfortunate that the survey undertaken by Jeff Errington and his divers in 1969 cannot be located. Jeff told me that his divers did indeed locate the gateway area on the sea-bed. I have written again to the Ipswich Museum team to see if anything can be done to locate the survey manuscript. At present the material is unclassified and therefore researchers are not permitted to access it.
A kind correspondent, David Blocker, has looked at the Lidar images that I posted, of the ruins of “Walton Castle”, the Saxon Shore Roman fort lying submerged near Felixstowe in Suffolk, and annotated them. The results are fascinating:
Then with annotation:
As a reminder, the rough sketch map drawn by the diver Jeff Errington:
He adds:
The lidar image of the Felixstowe area reveals the ruins of the Roman fort.
The lidar image roughly corresponds to the sketch made by the scuba club in the 1970’s. There appears to be a shallow underwater channel between the beach and the offshore mound with the ruins, the groynes do not appear to be effective at retaining sand.
The lidar image hints at an under water structure between the Walton fort ruin and the scour channel. It is undoubtedly something that was built before the 1700s when the area was inundated. If it did not appear on maps or drawings of that time it was probably already buried under sand, it might be a road or walls from the Roman or Norman period. Where it extends closest to shore is approximately where the scuba club found a pile of red bricks. What I have labelled as “scour channel” might be the Roman era streambed. It would not be unreasonable to assume that there were out-buildings or a village near the fort walls.
Fascinating! I wish that I knew more about the Lidar world; undoubtedly the datasets created for flood planning contained very detailed aerial images, if one could but extract them.
I was able to sit at my computer this evening for the first time and work a little on the translation of chapter 11 of the Vita of St George. So I am clearly improving. But I still can’t really walk, or leave the house, and I must keep my foot elevated most of the time. So it will be a while yet. Another chapter (12) of the vita has come in, in very rough draft, so I will have to look at that some time.
I received an email yesterday from Suffolk Record Office, suggesting strongly that the report on the sub-aqua survey of Felixstowe / Walton Roman fort has been lost. It looks as if the archivist only looked at a catalogue, however, so there is still hope that it may just be mislaid and might be found on examination. This will have to wait until I am mobile again, however.
A rather large number of items have arrived in the last week or so which I have placed in my “things to blog about” folder. One day perhaps I will get to them!
I have now discovered why I was unable to locate the 1969 survey report by Jeff Errington, reporting on the dives to the submerged Roman fort at Felixstowe. The article from 2000 said that it was at Ipswich Museum. But an email from one of the article authors, Tom Plunkett, reveals that a mass of files were transferred away in 1999, out of Ipswich Museum and into the Ipswich Record Office. This included the ‘Parish’ files, referred to in the article.
The Ipswich Museum ‘Parish’ files are or were simply a collection of manilla envelopes, used as a filing system. There was one envelope for each geographical parish (hence the name). As stray material appeared, it was filed by location in the appropriate envelope. The contents of the files were never catalogued. The Errington report should be in the “Felixstowe” or “Walton” envelopes.
The transfer took place at a time of cutbacks, when the Museum dispensed with the services of an archaeologist, and was undertaken to ensure that the material was not simply thrown away.
This morning I travelled down to Ipswich Record Office, and registered there as a user, not without inexplicable difficulty. I was advised to write to the archivist, however, Louise Kennedy, which I have done. No doubt there will be a large box full of this stuff, which I shall have to sort through, but that will be a nice way to spend a few hours. I do have the time.
One impediment, however, is that I shall be going into hospital on Monday 1st April to have a trivial but annoying problem with my foot surgically attended to. It’s being done under general anaesthetic, so there is a small risk. The surgeon told me that I will be out of action afterwards for three weeks. Let us hope that I shall be able to use my PC!
The old Victoria History of the County of Suffolk, on the landscape near Felixstowe Roman fort, refers to a close known as “Great Long Dole”, which apparently bore that name in 1907. This gave no results in Google. Fortunately the old Ordinance Survey maps are online (although for some peculiar reason the new ones are not). This provides illuminating information, including the location of said “Long Dole”.
Great Long Dole is no longer marked – the name must have become obsolete – but Brackenbury Barracks have appeared, and the Cliff Road. Note also the stream, running onto the beach, near the castle ruins. This must be the source for the “Dip”, where the stream cuts through the soft sand to reach the sea.
Via GeoHack I find other maps – I’m not used to the world of maps, of course, so I am getting whatever I can – and especially this from Defra:
The barracks have vanished – this seems to be a contemporary map – and the whole area is now covered with houses. But the outfall of the spring is still marked. It’s now in a pipe running under the road. But this spring is probably the water source for the garrison. The location of the Roman fort need not be attended to, tho. I believe that in fact they are about 30 metres offshore from that outfall!
I’ve also wondered whether aerial photography might show the ruins. Surely it might!
Likewise… what about Lidar? There are downloadable datasets, I know; although, mysteriously, I could find no online browser to see the data. The nearest that I got was this, where the resolution is rubbish. Yet Lidar datasets are available down to less than 1 metre, this I know.
Is there something, just in front of the sea-front buildings? Maybe there is; but at that resolution, where a house is just a block, who can tell?
Lidar is beyond my knowledge. Googling is not producing anything very useful, although I did find this. I might have a go at this a bit later.
The ruins of the Roman fort of the Saxon shore at Felixstowe, known as “Walton Castle”, were examined in 1969 by a team of divers from the Ipswich branch of the British Sub-Aqua club, led by Jeff Errington.[1] Ipswich museum liason was Elizabeth Owles, although I have yet to locate the survey report filed with the museum.
This morning I met Jeff Errington (now aged 75) at his business, Dive Line, in Ipswich. He lent me two clippings from local newspapers. I attach images as a PDF below, but I thought that I would transcribe these here.
Usefully the articles confirm that “the Dip” is in fact a ravine cut in the soft sand by a freshwater stream, and now brought under the cliff road by a pipe discharging on the beach. The angle of the ruins is correct – the walls run even closer to the beach than might be supposed.
The first, shorter article, was in the East Anglian Daily Times on December 15, 1969; the other in more detail was in the Mercury, on December 19, 1969, page 8. The latter included a sketch map of the site, based upon a drawing by Jeff Errington.
* * * * * * *
From the East Anglian Daily Times:
Submerged Roman fort yields up some of its secrets
By Don Black
Long submerged in the sea off Felixstowe, a Roman castle has yielded up some of its secrets to a team of amateur divers.
Yesterday, when a low tide exposed great chunks of a corner bastion, they concluded their first season of survey dives.
The divers have found that at least two walls still extend unimagined distances from the tumbled bastion—after 17 centuries of assault by armed men, stone robbers, cliff falls and the pluck and knock of powerful waves.
Yesterday’s high wind, gusting to gale force, whipped the sea against the masonry with such strength that further serious study is having to be put off until next year.
Walton Castle, as the place was known, will be explored as far as the ramparts that stood, on its seaward side, the first, to fall to erosion of the soft cliff on which it stood.
Fifteen dives
Ipswich branch of the British Sub-Aqua Club has made 15 fully recorded dives to the ruins since May, – its members working in an average depth of 12ft., with visibility ranging from eight foot to nil.
light under the sea may still perhaps be found the ghosts of men and shadows of the sunken stronghold.
‘‘Seaward of Felixstowe Ferry, the Roman castle of Walton once stood guardian against the raiding ships of the Saxon. These walls, bound
with mortar no less hard than the stone itself, were proof against the storms of a thousand years.
“In Walton castle, in 1338, Edward III lodged royally just before he sailed for France in the campaign of Crecy. Above the scarred walls, even then ancient, the banners floated, and under shelter of the castle lay its own harbour, Wadgate haven, crowded no doubt with the barrelled masts of medieval shipping.
“The sea has overwhelmed everything. Not a trace remains; Walton has vanished.” If the research done at Walton castle is anything to go by, the Ipswich team can produce results of scientific value from the sport they enjoy most.
* * * * * * *
From the Mercury:
Roman castle fortifications discovered by divers
Long submerged in the sea off Felixstowe, a Roman castle has yielded up some of its secrets to a team of amateur divers.
On Sunday, when a low tide exposed great chunks of a corner bastion, they concluded their first season of survey dives.
The divers have found that at least two walls still extend unimagined distances from the tumbled bastion—after 17 centuries of assault by armed men, stone robbers, cliff falls and the pluck and knock of powerful waves.
Sunday’s high winds, gust-ing to gale force, whipped the sea against the masonry of Walton Castle with such strength that further serious study is having to be put off until next year.
Ipswich branch . of the British Sub-Aqua Club has made 15 fully recorded dives to the ruins since May, its members working in an average depth of 12ft. with visibility ranging from eight feet to nil.
They have discovered that one wall runs for about 90 yards almost parallel with the shore, and that the other wall extends more than 40 yards out to sea.
The divers plotted their finds on a large sketch map. When this was showed to Walton’s historian, 89-year-old Mr. Samuel Wall, he declared: “This is wonderful. I was under the impression that the greater part of the masonry, septaria, was dredged away in the 1860’s to make cement.
Miss Elizabeth Owles, archaeological assistant at Ipswich Museum, has been advising the club on what to look for and how to record any finds.
“Underwater archaeology is incredibly difficult in this country,” she said. “Virtually nothing reliable is known about Walton Castle and any hard facts from the divers are most welcome.
“Old books and prints on the subject are open to suspicion. Perhaps the divers will find the position of the main gate . . .”
Miss Owles believes the map shows the walls running a little too tidily.
To help the divers identify Roman masonry, as opposed to natural formations, Ipswich Museum told them to look for layers of red brick. After removing a large amount of marine growth their search was successful.
“We have many more measurements to take and we want to bring out loose pieces of wall, if possible a good cross-section of bricks and stone” said the club’s diving officer, Mr. Geoff Errington (26).
They may do that by attaching five-gallon cans to the masonry and letting the incoming tide provide the lift.
When they made their first dive, they found that the wall running seawards appeared to be about nine feet thick and four feet to five feet high. Later, several points were seen where the brick runs in two bands of three layers.
Weed clings to the higher parts and there are crabs and the occasional lobster living among the ruins.
Many pieces of masonry lie throughout the area. These have been far too numerous to plot and are evidently scattered far out to sea and to the north of the main ruins.
Walton was built in the 3rd century AD, with Burgh Castle, two in a chain of forts from the Wash to the Solent that were intended to protect Roman Britain from Saxon invaders.
Both Suffolk strongholds were later used by the Normans, who built keeps inside the walls. But these additions have disappeared, that at Walton having been dismantled by King Henry II in about 1174 when he put down a barons’ rebellion.
The advancing sea completed his destructive work in the early part of the 18th century.
While firm evidence of the fort’s layout and history is limited, romantic writing on the subject abounds almost as much as that about Dunwich or Lyonesse.
“Suffolk Sea Borders,” published in 1926, includes this account of the approaches to the Deben estuary:
“LOST ATLANTIS”
“If entering the haven from the southward, a ship must sail over the dim ruin of a lost Atlantis. In the purple twilight under the sea may still perhaps be found the ghosts of men and shadows of the sunken stronghold.
“Seaward of Felixstowe Ferry, the Roman castle of Walton once stood guardian against the raiding ships of the Saxon. These walls, bound with mortar no less hard than the stone itself, were proof against the storms of a thousand years.
“In Walton castle, in 1338, Edward III lodged royally just before he sailed for France in the campaign of Crecy. Above the scarred walls, even then ancient, the banners floated, and under shelter of the castle lay its own harbour, Wadgate Haven, crowded no doubt with the barrelled masts of medieval shipping.
“The sea has overwhelmed everything. Not a trace remains; Walton has vanished.”
The sea on Sunday was surprisingly warm, 40 degrees F, the same temperature as the divers find in water-filled gravel pits in summer.
But they were glad to leave the turbulent conditions at Felixstowe for their normal Sunday training session at Fore Street baths, Ipswich.
* * * * * * *
Jeff told me that the pictures of the divers were all posed; it was far too rough to dive that day. All but three of the divers are now dead. Also he added that the sea-level was lower than he had ever seen it, which explains the very visible remains. The rocks that break the surface are those in the corner of the fort, where the ruins stand 8 feet tall.
Here is the raw article images. I do have higher resolution scans, but these are entirely readable.
My sincere thanks to Jeff Errington for meeting with me, and his great kindness in loaning me the clippings and telling me about this fascinating episode!
We do possess a number of old drawings of the Roman “Saxon Shore” fort that once stood on the cliff at Felixstowe. These show what it looked like, before it went over the cliff into the sea, and then after. These were printed in 1907 in The Victoria history of the county of Suffolk, Vol. 1, between p.288 and 289,[1] and appear in the Wikipedia article.
The first was drawn in 1623:
It was accompanied by a plan:
The “ruins” may be the remains of a medieval baron’s castle slighted by Henry II.
Next the Victoria History prints a drawing from 1766 by Francis Grose of the ruins lying on the beach after the sea undermined the sandy cliff on which it stood.
The Victoria County History volume was published in 1907, at a time when the landscape around the area was rather different, for the port of Felixstowe did not exist, and much of the area now occupied by the town was just marshes!
The two rivers, the Stour and the Orwell, pour into the sea together to the south. The river Deben is to the north. The VH reads (p.288; I have added paragraphing and modernised spelling where needed):
… the united waters of the Stour and Orwell poured themselves into the sea by a passage running at the foot of high land called at its eastern end Bulls Cliff, at Felixstowe. Marshes at the foot of this high land, and the traces of a waterway in the marshes, in fact, seem to point out the line or which the united waters of the Stour and Orwell reached the sea in very early times.
The VH continues:
Between the Deben and these other estuaries lay a broad flat tract of land some miles in extent, a sort of peninsula. In the Roman period this tract projected farther into the North Sea than it does now. On a site on this land about a mile south of the mouth of the Deben, but possibly then at some distance from the sea-shore, stood the fortress whose history, scanty as it is, is still worth tracing.
Of the walls not a fragment remains above the waves, which undermined the last relics of them in the 18th century. Now and again when the tide is at its lowest two or three weed-covered masses may be seen, but that is all there is to show that a Roman station once stood on the spot.
Fortunately the memory of it has been preserved in letters and drawings which show clearly enough the character of the buildings. The place was known in the 17th century, and probably much earlier, by the name of Walton Castle, no doubt from some traces remaining within the inclosure of the Roman walls of a keep and other structures of the 12th century, built by the rebel earl Hugh Bigod, whose castle here was destroyed by the king (Henry II) after the suppression of the rebellion in which the earl had been engaged.[11]
11. On the Pipe Roll of 22 Hen. II, an. 1, 176, is the amount of the costs involved in its destruction. The Norman additions to the Roman fortress would find a parallel in those made within the Roman walls of Pevensey (Anderida), another of the ‘forts of the Saxon Shore.’
The intruder structure then is very like the situation at Porchester (not Pevensey), where a Norman keep stands in one corner of the Roman enclosure:
The VH has more to tell us, however:
However this may be, it is not until the 18th century that it is possible to speak with any certainty as to the remains called Walton Castle, and then the descriptions show them to be clearly Roman. The first in date is the following, which occurs in vol. i of the Minute Books of the Society of Antiquaries under date 28 November 1722, in a letter from Dr. Knight. It runs thus:
“Some distance east of this town (i.e. the neighbouring village of Walton) are the ruins of a Roman wall situate on the ridge of a cliff next the sea between Languard fort and Woodbridge River (the Deben) on Bawdsey haven. It is 100 yards long, five foot above ground, 12 broad at each end and turned with an angle. It is composed of pebble and Roman bricks in three courses, all round footsteps of buildings, and several large pieces of wall cast down upon the strand by the seas undermining the Cliff, all which have Roman brick. At low water mark very much of the like is visible at some distance in the Sea. There are two entire pillars with balls, the cliff is 100 foot high.”
In 1722, then, part of the fort had fallen into the sea, but the landward side still stood, as a “Roman wall” which was 100m long.
The next notice that can be given is from Kirby’s Suffolk Traveller (1735), where it is stated under the head of Walton :—
“For the neighbouring Parish of Felixstowe on the Colnes side of Woodbridge (Bardsey) haven, still appear the ruins of a quadrangular castle, advantageously situated”;
and further and more fully in the second edition of his book, published in 1754, he says:
“He that would look for the site of this castle within the bounds of Walton strictly taken will never find it; but upon a high Cliff in Felixstowe, at the distance of about one mile from the mouth of Woodbridge River (the Deben) and two miles from Orwell Haven, part of the foundation of the west side of it is still to be seen: being now one hundred and eighty-seven yards in length, and nine feet thick; it is called by the country people, Stone-Works. How much larger it was we cannot judge. Part of the south end being washed away: and the sea, which is daily gaining upon this coast having swallowed up the ruins. Such was the condition of it about the year 1740 : but since then the sea hath washed away the remainder of the foundation. There can be no doubt but Walton Castle was a Roman fortification as appears from the great variety of Roman urns, rings, coins, etc., that have been found there, etc.”
A page further on contains the following paragraph, which seems to identify the place:
“In a survey of the Manor of Felixstowe Priory made in 1613 we find a close of arable land called Great Long-Dole, in which close are the ruins of Walton Castle.”
The close is still known by its ancient name, and its position may be seen on the Ordnance Survey maps. It lies close against the sea.
The landward side of the fort fell into the sea, then, between 1740 and 1754. Sadly a Google search on “Great Long-Dole” reveals that, whatever it was, it is no longer known by that name.
Grose, in his supplement to the Antiquities of England and Wales, published in 1787, gives a view of the wall in question, which by that time had fallen on the beach. He says:
“Its remains in 1766 when this view was drawn, were only visible at near low water, the sea having gained so considerably on this coast as to wash away the cliff on which it stood. A gentleman now living remembers the ruins of the castle to have stood at least fifty yards within the extremity of the cliff.”
Thus the only remaining fragment of Roman enclosure had fallen beneath the waves in 1766. The wall unquestionably formed part of a station of the same class as the one by the Waveney, though perhaps not covering quite so large an extent of ground. Kirby speaks of it as standing on a ridge. This was the southern edge of a wide depression, scarcely a valley, which, wide upon the shore, gradually narrowed as it ran westward to the ancient village of Felixstowe. Both slopes of this shallow valley appear to have been used as the cemetery of the station; a boarded well has been discovered ; and perhaps some few scattered houses, judging from the building material found, may have existed here. On the north side of the valley, where a fall of the cliff occurred in 1853, two skeletons were uncovered. These had bronze armlets, which were placed with the bones in the museum at Ipswich.[12]
12. For an account of other views representing the ruins of the Roman station called Walton Castle see Topographical Index under Felixstowe.
The “valley” is today known as “The Dip”, and any driver passing along the coast road will recognise the big dip in the road instantly.
From a report [13] in the Proceedings of the Society of Antiquaries (1885) we learn that ‘during the progress of works carried on in what is known as the Park, situated a short distance from Felixstowe Church (close to, if not actually part of, the field known as the Great Long-Dole), the men in their search for Coprolites came upon many most interesting relics of the Roman occupation.’ Amongst the list of objects turned up occur ‘coins of Severus, Gordianus, Gallienus, Victorinus, Constantinus, Arcadius’ &c. Of more consequence than the miscellaneous finds were the cinerary urns found, ‘containing bones and ashes and either closed with a cover or with a stone only.’ The discovery of these last showed the situation as part of the cemetery.
From what has preceded it is clear that there was on this spot of the Suffolk coast a Roman station, approximating in size and details of construction to those noted for the defence of the east coast. It occurs at a most important point for the protection of this coast, and its omission would have caused a dangerous gap in the line of defence. On the cliff 100 ft. above the sea it commanded a full view of both the Stour and the Orwell and of the Deben, the harbour at whose mouth, called Bawdsey Haven, scarcely a mile away, would have held the auxiliary ships of the fleet. It seems strange that the site should be so little known by antiquaries, perhaps owing to its disappearance early in the 18th century and the very little interest shown by them in the later Roman fortresses in comparison with those which are more obvious and earlier.
It is a curious fact that though eight of the ‘forts of the Saxon Shore’ have been identified with all probable correctness, the place of one of them has never been settled with general assent. This is the Portus Adurni. The station has been supposed to have been situated on the Adur, a stream flowing into Shoreham Harbour, in Sussex, the name being given to it from that of the stream ; but this has been shown to be an error, as the little river in question had no certain name till the 17th century, when that name was bestowed upon it.[14] Portchester has also been suggested for its site, as a large Roman station of the late class still exists there at the head of Portsmouth Harbour. There is, however, no river or stream here whose name might be associated with that of the port. From the most westerly of the identified ports of the Saxon Shore, Pevensey (Anderida), round to Brancaster on the Wash, in Norfolk (Branodunum), the sites of the stations have been found and named, yet here at Walton on the most important point in the whole line lies a station as important as Gariannonum, which station it is suggested was Portus Adurni. If we knew the Roman name of the Stour such a suggestion would be settled.
In the Notitia Dignitatum Portus Adurni is spoken of as garrisoned by a body oi exploratores ; that is, in modern phrase, mounted scouts. No better troops could be selected for such a fort as that at Walton. There seems not very much doubt that if all the facts connected with the site could be fairly arrayed the name given above would with little hesitation be bestowed upon it and the last gap in the list of the Notitia would be satisfactorily filled.
The “Topographical Index” is later in the volume, on p.300. Pp.305-7 contains yet more information. I will give it all, having dug it out, but it is rather dry. The possibility of more drawings, however, deserves attention.
Felixstowe.—Here was a Roman station known under the name of Walton Castle, formerly existing on land near the village of Felixstowe, which has long been washed away by the encroachments of the sea (see p. 287). The site of the cemetery attached to it still remains. The following are the principal authorities for the former existence of the station :—
A communication from Dr. Knight to the Society of Antiquaries of London, which has been inserted in the manuscript Minute Book of that society [i, 71, 2]. The letter is dated 28 November 1722 [A Tour in the Whole Island of Britain, by a Gentleman (ed. 3), 39-40].
Mr. T. Martin, in his account of the remains, which is dated 16 September 1725 [Church Notes, i, 185], says, ‘ About half a mile from the town (of Felixstowe) are the ruins of a Roman fortification upon the brink of the cliff (great part being already fallen down, a few years are likely to put a period to the whole), ’twas built very substantially with rock-stone and Roman brick.’ . . . . . . . . . . . . . . . . The portion of this wall of the station, still standing in 1740, and the site generally, are described by J. Kirby [Suffolk Traveller (ed. 2), 89-91].
Camden mentions Walton Castle, but not as a Roman station.
Gough in his ‘Additions’ to his edition of Camden speaks of ‘a castle (at Walton), the foundations 187 yds. long and 9 ft. thick . . . plenty of Roman coins and other antiquities found there’ [Camden, Brit. (ed. Gough), ii, 85]. The mention of the length of wall remaining is probably taken from Kirby.
In Grose’s account accompanying his text a view apparently taken in 1766 is given showing the remains of the wall lying upon the beach [Grose, Antiq. of Engl. and Wales ; Suppl. ii].
Besides the view of the ruined wall in Grose’s account others may be mentioned.
There is a small sketch in Indian ink entitled ‘ Remains of Felixstow Castle 1780,’ the drawing being by Isaac Johnson, in vol. ii of Suff. Illus. (Fitch Coll.).
Another drawing in pen and ink, washed, was in the collection of the late Mr. Eyre of Ipswich. It purports to be a copy from a print in the possession of Mr. Oxburgh of Hartcliff in Kent, the print itself having been executed from a drawing made in 1700 by Thomas Bates of Ipswich. A second copy of this print is in the collection of Miss Nina Layard of Ipswich. Neither print nor original drawing is now traceable. These copies show the ruins of a circular bastion on the edge of the cliff, evidently that of the one at the south-west angle of the station, with a small portion of the south wall running seawards, and overhanging the precipice. Masses of the same wall lie upon the beach.
Mr. Eyre also had a tracing from a pen-and-ink outline drawing washed with colour, of much earlier date than the last mentioned. Upon the original drawing there appears to have been this inscription in writing of the 17th century: ‘E Pros (?) Walton Castel,’ and it was signed in one corner ‘John Sheppard 1623.’ The sea front of the station is represented entire, the edge of the cliff being in the foreground. A circular bastion is seen at each angle, and a wide break about the centre of the front marks a gateway. Beneath the view is a rough plan indicating that the walls of the station formed a parallelogram with a bastion at each angle. The drawing from which this tracing was made is not now to be found, and judging from the tracing it had the character of a sketch of the 18th rather than of the 17th century, although the writing on it seemed to be of the latter period. Unless something more could be ascertained respecting the original drawing from which the tracing in question was made, its value as an authority for the condition of the remains early in the 17th century must be considered doubtful.
Many objects of the Roman period have been found about the site. Beginning with coins, it may be mentioned that Davy [Suff. Coll. B.M. MSS. 19087, fol. 53-60] described a collection made at Felixstowe in the years 1742-3-4, in full detail. The coins range from Pompey the Great (ob. 48 B.C.) of whom there was one example, to Honorius (A.D. 395-423), and their total number amounted to 420. This collection was the property of the Rev. W. Brown of Saxmundham, and at the sale of his goods in 1827 it passed into the possession of the Rev. W. Layton of Ipswich.
Coins were found at a later date, of Victorinus (a.d. 265-8), Tetricus (a.d. 268-73), Urbs Roma, Valens (2) (a.d. 364-78), and Gratian (a.d. 375-83) [Journ. Brit.Arch. Assoc. xiv, 271] ; also a gem of oval form, possibly a cornelian, engraved with a figure wearing a petasus, and with a panther skin hanging from one arm. In the right hand it held a poppy head and wheat ear, and in the left a pedum [ibid. 339].
A find is recorded in 1749 of very small Roman coins and some pieces of metal, as if melted in the fire, near Bawdsey Haven [MS. Min. Soc. Antiq. v, 241].
During the process of digging for coprolites in the field known as the Park near Felixstowe Church, many objects of the Roman period were turned up. Such as were noted were as follows :—
Vase of so-called Samian ware, with hunting scenes and ornament of oak leaves and acorns (purchased by the South Kensington Museum), flue tiles, amphorae, lagenae, a small glass phial, bronze pins, tweezers, a speculum, several fibulae and gold rings, silver rings, some set with stones, a gold chain of twisted wire, and a bronze bracelet, a bronze disk enamelled (a circular fibula ?), a bronze enamelled tag ox fibula, and other objects in the same metal ; coins of gold, silver, and bronze, of Severus (a.d. 222-35), Gordian III (a.d. 238-44), Gallienus (a.d. 253-68), Victorinus (a.d. 265-8), Constantine (a.d. 306-37), and Arcadius (A.D. 383-408). There was a great quantity also of mussel, periwinkle, and cockle shells, and of snail shells. Many sepulchral urns were dug up containing bones and ashes, closed in some instances with a cover, in others only with a stone [Proc. Soc. Antiq. (Ser. 2), xi, 12, 14: ‘Communication from Mr. E. St. F. Moore of Woodbridge’].
In 1843 Prof. George Henslow and his father found ‘ a sort of kitchen midden in the London clay cliff near Felixstowe.’ Rough pottery, bones, cores of deer horns, a skull, rusty nails, and a so-called Samian vase, were among the objects discovered in it [Antiq. xlii, 283 (1906)]. The Roman objects of bronze in the collection of Lord Londesborough from this site are a bust of Mercury and several keys [Antiq. Etching Club Publ. iii, pi. 25].
In the Ipswich Museum there are three flue tiles, perfect, one of large size with reeded ornamentation ; vessels of ordinary ware, some probably cinerary urns ; a large thumb-pot of unusual size, and a harp-shaped fibula 2½ in. long. Here are preserved an arm and a finger bone from one of the skeletons found by Prof. Henslovy, together with bronze bracelets. The objects added to the museum in 1897 were fragments of vases of Durobrivian, Upchurch, and so-called Samian ware (all plain) ; pieces of a glass vessel, one fragment of window-glass, and small portions of roof and flue tiles. The animal remains consist of fragments of horns of red deer, &c., and there are some oyster shells.
In the Norwich Museum (Fitch Coll.) are preserved keys, tags of belts, tweezers, fibulae—two perfect, two in fragments—nails, rings, a small bronze column 3 in. high, a figure of a goat, a head of an animal, two small busts, one perhaps of Mercury, a leaf, a portion of a vase, and fragments of ligulae. All these bronzes appear to be water-worn.
In the Bury Museum are a small vase with black glaze, presented by the late Lord John Hervey, 1853; needles and bone pins (Acton Coll.); implement in bone, probably a mesh gauge ; jet pins ; a double comb in bone, one button in shale and one in jet.
In the British Museum are bronze studs found here in a leaden coffin in 1853; the central band of an enamelled buckle ; a bronze toilet implement ; a brooch ; a pin ; beads ; an armlet ; a necklace of glass beads and shell found in a glass urn with bone bracelets in 1853; glass beads, one being engraved ; a small rude flanged pan of dark drab ware, a pot of grey ware ornamented with groups of dots in slip, a small black vase with glazed bands ; a small pot of rough reddish grey ware ; and a reddish-grey vase with black bands. A vase of so-called Samian ware of oval form has the body ornamented with a wide band and simple marginal lines, this band having vine foliage, amongst which are interspersed figures of stags ; the animals appear to be of slip, and resemble those seen on Durobrivian ware. This vase was given by Sir A. W. Franks in 1881 [For site see O.S. 6-in. xc, NW. and SW.].
The Victorian County History is something with which I have hitherto had no acquaintance. It does contain a frankly impressive quantity of information! But I think that’s enough for now!
I’ve written a couple of posts about the remains of the Roman fort of the Saxon Shore, lying under the water offshore at Felixstowe. I’ve been trying to get hold of a survey report from 1969, done by members of the British Sub-Aqua club. This seems to be the last work done on the fort. The survey is known as the “Errington manuscript”, and is proving elusive indeed. It was lodged with Ipswich Museum, who have not been able to locate it following a move. No doubt it will turn up in time. But I learned from the local branch of the BSA that “Errington” was still alive, and the proprietor of a local diving business. So I emailed him a few days ago.
This afternoon, to my enormous surprise, my mobile rang and a voice introduced himself as Jeff Errington! He had received my email, and modestly apologised for being away on business in Scotland and that he had only got back last night. He was pleased to learn that his work had made it “into the history books”, as of course it has.
Sadly he himself does not have a copy of the detailed survey report that he did for Elizabeth Owls, the County archaeologist. But he did have cuttings from the newspaper from the time, showing the layout of the fort, and I have been told to pop along to his business and get copies. I hope that it will be possible to meet Mr Errington personally, although he is in fact a very busy man even in retirement.
I asked him about how he had come to dive on the ruins, and he told me a little about it.
In 1969, Mr. Errington was in the navy. After an accident overseas, he had come back to the UK to recover. The navy had posted him for six weeks to HMS Ganges, the now-defunct naval base on the Shotley peninsula. He was already a member of the British sub-aqua club, and eager to learn about ancient remains of any sort in the area in which he happened to be. On learning that there was a Roman fort off the shore at Felixstowe, he organised the local sub-aqua club to conduct the survey. He wasn’t able to get the navy to do the job (“of course”), but was able to get them to contribute some items of support or equipment.
The ruins of the Roman fort, he told me, are not parallel to the modern-day shoreline, because the coast has changed. But they are directly off the lowest point of “the Dip”. About 30 yards offshore there are black, weed-covered rocks which break the surface at a sufficiently low tide.
The highest section of the ruins is about 8 feet tall. The divers could stand up next to it, on the sea bed, and it was taller than they were. They found a section of wall, a corner, and a further section. They then found a gap in the wall, and then more wall. He told me that he had looked at the Norfolk Saxon Shore fort at Caistor, and the gap was about the width of the gateway there. So he presumed that this was the opening for the gate, facing inland.
The ruins are in very shallow water, which presented problems for the divers, who ended up “bobbing about”. A chunk of wall/brick was retrieved and handed to Elizabeth Owls, who sent it off for analysis.
This is fascinating, and I will go and look at the clippings, which include a map.
I think that I have seen the “rocks” myself, however. Last week I made a visit to “the Dip” at low water, and a group of rocks were clearly visible. Here are my photographs:
Sadly the camera on my phone is not of sufficient resolution to show much. But … it sounds from what Jeff said as if this is indeed it!