Founding Patrons
Les Bulley Mabel Hudson Dorothy Meek Edith Thomas Colin Titcombe Marjorie Morgan
Some thoughts about the above – an ongoing collection for
Caerwent Historic Trust
01291 420745 john@rose-cottage.freeserve.co.uk
2. Introduction – suggestions re Navigation
3. The Whirly Holes
8. Llyn Liwan, Liuan or Livan
14. The Wondrous Plank
Caerwent Historic Trust
to work towards re-establishing a museum or Visitor Centre in Caerwent.
Roman Navigation – from Steve Farrar 8.1.2004
The Romans opened up the Neddern for ships, or if not then for fair sized boats, not quite as far as Caerwent. They established docks, including one very large one, possibly 1000 x 200 yards, downstream from Caerwent’s Brook. This occurred to me over the last 5 years or so: linear earthworks are noticeable in that area and their form is more easily seen as a whole in conditions of moonlight. The dock system is also discernible in high summer when differences in ground colour show up – an aerial photo would be useful.
Lower down, nearer Caldicot Pill, I think the water level was controlled by a system of locks. I would like to see an archaeological investigation, ideally a trench across a key part of the valley, but boreholes would be a cheap alternative to this.
I would also imagine that the stretch of the valley overlooked by Mount Ballan had a strategic role for the Romans in repelling ship-borne raiders – their boats could be trapped by fireships being sent down at them but also launched downstream from where Caldicot Castle now is. . They could then be a sitting target for a defending force on the banks.
The present landscape is difficult to interpret because of the culverting of stretches of the Neddern, both above and below Caerwent’s Brook, and the re-routing of its outfall, which used to be at Sudbrook, from Deepweir to the Estuary (Caldicot Pill).
reported to JN by phone 8.1.2004
Denis is planning a walk for CHT to embrace Crick and the Neddern. he regards the mound of the Motte at “The Berries” to be much more ancient than Norman. A cursus of pudding stones (conglomerate rock) links it with the Crick Burial Chamber. It also lies on an alignment of Caldicot Church with the tumulus just above Crick at Bentley Green.
Denis has access to a computer programme which can plot the contour lines. This has shown many interesting details about the evolution of the coastline and may help identify dock sites etc. He feels doubtful that ocean going ships will have gone as far as Steve thinks, but goods will have been ferried by lighter vessels, maybe flat bottomed.
The ancient Neddern was full of fascination, largely wiped out (I believe) by T A Walker in the hope of solving his problem of draining the Severn Tunnel. The Whirly Holes just below Caerwent’s Brook, listed by Nennius among the 13 Wonders of Wales, were bunged up with clay!
to JN 9 and 12.1.2004
Steve has lots more theories including a possible chariot racing circuit around the lane Brockwells – Shirefield. According to Steve the Romans will have built their villas on the high ground overlooking the Neddern. This is borne out by the discovery by Colin Titcombe of numerous coins and other fragments of Roman origin during his period of residence at Brockwells Farm.
The Whirly Holes were brought to our attention by Denis Manning, Monmouthshire C C Countryside Warden, Caldicot Castle Country Park, in January 2004:
The ancient Neddern was full of
fascination, largely wiped out by T A Walker in the hope of solving his problem
of draining the Severn Tunnel. The Whirly Holes just below Caerwent’s Brook,
listed by Girallt Cymru among the 13
Wonders of Wales, were bunged up with clay! The bed of the stream was culverted
all the way from Llanmelin to Caerwent Brook and presumably its diversion away
from Sudbrook pill made irreversible at that time.
We haven’t unearthed a description of them yet. It seems there were 2 major sites, as there are 2 “whirlpools” shown on a 1775 map, and two Whirly Pool fields identified from the Caerwent Tithe map.
Tithe Map adapted by Glyn Parkhouse shows Whirly Pool Mead and Whirly Pool Meadow.
Mentions their existence but nothing more!
They were points of resurgence and disappearance of the Cas Troggy (Neddern) Brook. Nowadays the stream disappears for the summer, at the mouth of the Cwm and only has running water again from around its junction with the Crick Brook. The stream must have carried more water past Caerwent in the 1930s as older people refer to bathing in it etc. though this must have been thanks to the culverting, whose benefit has not lasted due to the formation of new swallow holes moving upstream in the Cwm.
The stream certainly seems to have had a reputation for appearing and disappearing quite suddenly. The Whirly Holes were swallow holes, a well known feature of limestone country, and I guess these had a more or less permanent existence as ponds. The most frightening event seems to have been a sudden loss of water, accompanied by a severe whirlpool effect. We can suggest this occurred due to variations in the surface water in the stream and the underlying water table. The effect may have been exaggerated by a siphonic arrangement of chambers under the ground. Maybe there was a link with the tides - presumably the ebb tide lowered the water table so created a faster drainage down the hole, hence the whirlpool effect.
The site of the upper Whirly Hole is a very obvious hollow in a meadow that is divided by the defunct upstream bed. Downstream the defunct bed follows the field boundary. The site of the lower hole must be tucked into the bank where there is a sudden rise in ground level to the main fields of Slough Farm – I haven’t quite found it yet!
from Mabel Hudson 2005
October 2005 (aged 94) her uncle worked for the farmer of the day, a Mr Knight, whose son had been lost down a Whirly Hole (whirlpool) in the Neddern and never seen again.
Mabel and her sisters were warned to keep right away from the whirlpool area on account of this, though it seems the upper one and probably the lower one too had been filled in as part of the efforts made by T A Walker to cure the problem of the Severn Tunnel. The children used to walk across from Crick across Mr Edmunds’ land, keeping to the footpath, to go ice skating – maybe more in the area which has been fairly constantly flooded in winter. She is speaking of whirlpools from her childhood, 1910 onwards – much smaller than the original Whirlies, which were filled in by then.
Colin Titcombe also reports small whirlpool/swallow holes from 1950 onwards.
Very frustrating – doesn’t show any Whirly Holes at all! Though the field names refer to them (see front page). Presumably the fairly recent course of the stream was counted to be the boundary, hence they weren’t in Caerwent.
Caldicott Tithe Map 1842 – area of Caerwent’s Brook
Seems to show the upper Whirly hole as double. Also a pond by the bend in the road linked by a channel to the Whirly Hole.
[Puzzles – this pond seems to have become a grass verge now. Also there was a there was an old bridge, only fell down in the 1970s, at about the place were the pond joins with the channel. I had assumed this to be over the original route of the Neddern, and to carry the Caerwent-Caldicot road.]
Ignore differences in shading.
1775 Map (Caldicot Lordship Sale) – two “Whirlpools”
Map of 1775 for Caldicott Lordship Sale
It seems there are two Whirly Holes, both of them double. There seem to have been many changes in the course of the stream. The parish boundary (dotted) probably follows its course about 700AD when Brockwael transferred the ownership of the Southern part of Caerwent to the Church, whence all the land to the south of the Neddern ended up as “Caldicot Parish”.
edited JN 23.2.2006
for Caerwent Historic Trust
Dear John,
I have read with great interest your work on the Whirly holes. You are absolutely correct. I had come across an erroneous description of the Nedern/Crick confluence somewhere but can’t remember where... Hando?
I contacted the Cambrian Cave Registry regarding the Whirly Holes resurgence and indeed they have it recorded as such. They have very kindly sent me a copy of the entries for that area. The Nedern sink has interestingly been traced to the Great Spring in the Severn Tunnel by the University of Bristol Spelaeological Society. I will give you a copy at the next meeting but little else is of interest.
The Cambrian Cave Registry would very much like (and benefit) from a summary of your work on this matter if you would be willing to share it with the record. I hope to look into it further also.
Regards, Mark Lewis 11.1,2006
1887 6inch(?) OS map
Shows various ponds of which I think the residual Whirly Holes are at the extremes of the extract below:
The label whirlpools has been transferred upstream. I think the main holes were filled in by this time and various subsidiary holes were evident all along the valley – Colin Titcombe remembers some from about 1950.
1887 map with present features superimposed © Ordnance Survey
Whirly Holes and Llyn Liwan Caerwent
from www.wondersofbritain.org by Andrew Evans accessed 20.5.2006
translated from Nennius, Latin, pre 900AD
“Another wonder is the mouth of Linn Liuan, the mouth of which river opens into the Severn, and when the tide flows into the Severn, the sea in the like manner flows into the mouth of the aforesaid river, and is received into a pool at its mouth, as into a gulf, and does not proceed higher up. And there is a beach near the river, and when the tide is in the Severn, that beach is not covered; and when the sea and the Severn recede, then the pool Liuan disgorges everything that is devoured from the sea, and that beach is covered, and it breaks and spews in one wave. And if the army of the whole country should be there, and should front the wave, the force of the wave would drag down the army, its clothing filled with water, and the horses would be dragged down. But should the army turn their backs towards the wave, it will not injure them. And when the sea has receded, then the whole beach which the wave had covered is left bare again, and the sea ebbs from it.” 12
This confusing account apparently describes a bay3 on the Severn, which swallows the tide as it rises, only to disgorge the water in a giant wave as the tide retreats4. Linn Liuan is probably a misspelling of the Welsh “Llyn [Lake] Lliwan”567 talked of in other works. However, this doesn’t help us much, as, despite being mentioned over many years8, no one now knows where this is either! One work that features it is a Welsh story “Culhwch and Olwen”, which gives it as near the Wye and downstream of Gloucester. On this, and other evidence,9 Celtic specialists Rachel Bromwich and Simon Evans* have suggested this is the mouth of the River Usk (Photo), though the mouth of the River Taff10 or Wye11 (Photo) also seem possible.
Lake Liuan is also described by Geoffrey of Monmouth in his “History of the Kings of Britain” (~1136 C.E.). He relates a conversation between Arthur and a companion1 about the lake. A Welsh version of this conversation gives...
“And he told him also that there was another lake in Wales near the Severn, which the men of that country called Llyn Llivan; and that lake, when the sea flowed, received water into it, and swallowed it as though it had been a mountain, until it overflowed its banks; and if it chanced that any stood with their faces towards the lake, and any of the spray of the water touched their clothes, it was hard for them to avoid being drawn into the lake; but if their backs were towards it, how near soever they might stand to its edge, it would have no effect upon them.”2
This might be a mistranslation of the version by Nennius, or it may suggests the water itself has magical powers, drawing in bystanders to their doom. While the lake isn’t noted for its special character in the tale “Culhwch and Olwen”, it does suggest the lake is also the home of the oldest animal in Britain3 - a giant Salmon4, who helps Arthur’s men rescue a prisoner from upstream of the lake at Gloucester as part of their quest for the giant boar Twrch Trwyth. This quest also features in the Wonder of Builth Cairn.
www.blaenau-gwent.gov.uk has similar in folklore section
Comment JN on behalf of Caerwent Historic Trust
These quotes have solved a question which has vexed us for some time, though we may not be the first to reach this conclusion, and Chris Barber pointed the way in his first book on King Arthur Journey to Avalon pp 197-8
Just east of the former Roman town of Caerwent there were two large swallow holes, blocked up 125 years ago, called the Whirly Holes. Swallow holes are a feature of limestone geology and often, as with icebergs, there’s more below the surface than can be seen above! We gather that the water of the stream made quite dramatic disappearances and resurgences at these holes, such that they were regarded as one of the Wonders.
We have been looking for an eye witness account of them and at last these have come to our knowledge, though we hardly expected that the most recent reports we could find would be 1000 years old!. We are inclined to believe that though the descriptions may be rather exaggerated, they did indeed refer to our place. Although the site is now 2 miles from the banks of the Severn Estuary we believe that the water at high tide rose to only a short distance from Caerwent, hence encouraged the Romans to build there due to ease of transport. If the ground water level were low, the stream would be dry, its underground passages empty and the incoming tide would pour down the swallow holes. If the holes, and the (nowadays) water meadow around them, were filled then residual water might drain very suddenly down the holes at a certain stage of the ebb tide.
It is only in the last 400 years that the sea wall and drainage schemes have established a reliable coastline here (the green line on map over the page). The drainage increased with the building of the Severn Tunnel and has increased again such that Llyn Lliwan, from being a brackish or tidal lake has shrunk to a wetland which is becoming less and less wet, drained by a frequently dry stream (the Neddern or Cas Troggy brook). However some winters Llyn Lliwan reasserts itself with a vengeance, attracting hundreds of migratory geese, and fortunately for motorists, the M48 is carried on a viaduct above it.
This location fits in well with the Culhwch and Olwen story: Twrch Trwyth probably knew his escape routes - he will have been heading for the Severn crossing at Black Rock or the nearby one at Sudbrook. These are between the mouth of our stream and the mouth of the Wye. However unless he reached there at low tide, he is unlikely to have survived. Also the escape to Cornwall may not have happened because Kernow seems to have also been a name applied to S E Wales, surviving eg in Coed Cernew today (seems funny that he needed driving into the sea twice).
The phenomenon described by Nennius is nothing like the bores on the rivers because they show no disappearance, the water only gently subsides. The Usk and Wye have bores infrequently, and they are trifling except for the rarely seen “winter bore” on the Wye, which can rise to 4 feet and be quite scary. The Severn bore only rises to a violent 4 feet within 10 miles of Gloucester, so suggestions that the descriptions are derived from a bore phenomenon are far less tenable than our pointing out actual holes through which the water could have flowed.
And the Great Salmon - well, they were all over the place here - whenever the sea came in the salmon would too - an ideal salmon trap. I’ll check as to their recent history, but they are still harvested on the estuary shore by lave nets and putchers.
January 2004 based on Crown copyright.
The blue area of the map shows the area below the present 10m contour. We regard this as the probable Llyn Llyn Liwan. The square top left is the Roman walls of Caerwent and T A Walker in his book about Construction of the Severn Tunnel refers to iron rings set in the south wall – though he found it hard to believe, he had no alternative idea from the local tradition of boats having been moored to them.
Dennis’s map also shows a number of ancient sites on or near the border of the blue area whilst no sites from pre-1500AD have so far as I know been found within it. We must also compare the various maps with the record of the area kept by Colin Titcombe from about 40 years ago.
present coast green
10 metre or 30 foot(?)contour blue
A Caerwent Roman town 100AD
B the Berries motte & bailey 1100AD ditches & cursus 2000BC?
C Caldicot castle1100AD
D Crick round barrow 2000BC
E Crick Manor House or bailey behind it 1200AD
F Brockwells
G Sudbrook Promontory Fort 600BC
J Harold’s hunting lodge 1065AD
some dates speculative!
The existence of this lake which we have decided is the original Llyn Liwan has apparently been acknowledged by archaeologists following up the discovery of a prehistoric boat, 1980s I think, at Caldicot Castle. They just called it the “Caldicot Castle Lake”. I’m not clear to what extent they thought it may have been tidal.
It’s strange that so far as I know the lake has existed for over half the year most years since at least 1950, yet it is never shown on OS maps. Two winters ago the water was about 6 feet higher than this.
Looking W alongside the M48
Looking NW from under the M48.
Looking NW to Caerwent from the M48
Looking S to Caldicot Castle C across the M48 “The Berries” B on the bank to the left
The meadow inside the barbed wire is about 12 feet higher than present water level but is at the top of a quite steep bank. This is typical of the contour near the water’s edge, ie the lake could hold a much greater depth of water without encroaching much on useful land.
Looking SW from below the M48 – at the foot of the steeply wooded hillside opposite is a large area of water difficult to see because it is enclosed by greenery.
Canon E T Davies locates this at St Tewdric’s Well, Mathern, just 3 miles east of Caerwent – see extract over page.
If this location is correct then it indicates the tide to have risen to a height presently more than 15 metres whereas the height of the Neddern at Caerwent Brook remains below 10 metres and the 15 Metre contour comes well up the fields under the Roman Walls. There’s no doubt that Nennius is describing a closely similar tidal situation to that of the Llyn Liwan Wonder. The Mathern inlet is more or less continuous with the Caerwent one as we progress up the Severn coast and they presumably flooded in the same way.
I don’t think the actual site of St Tewdric’s Well matches the environment in which the plank’s weird behaviour would be possible. A small distance to the west there is a long narrow hollow, of similar height to Caerwent’s Brook. If you go to the edge of the Churchyard you will see that part of it is even today occupied by a pond. More likely the plank was bridging this narrow channel and was washed up and down it during extreme tides. Indeed there are two more channels between Mathern and St Pierre which obviously would have connected with the estuary in the past. Both of these also carry narrow ponds at present. It’s the middle one of the three which carries the main stream from Pwll Meyrick village, so I guess it may be the best bet for the plank!
by Canon E T Davies, Vicar 1943-48, extract from booklet 1950/1990
