Cantre’r Gwaelod

The drowned kingdom of Cantre’r Gwaelod is a story associated with the inundation of land off the western coast of Wales between Ynys Enlli (Bardsey Island) in the north and either Cardigan or Ramsey Island in the south, in what is now Cardigan Bay. The earliest written source for the story of Cantre’r Gwaelod is a poem from the thirteenth-century manuscript Llyfr Du Caerfyrddin (The Black Book of Carmarthen) called ‘Boddi Maes Gwyddno’ – the drowning of Gwyddno’s plain.⁵ In the story, Seithenyn is the king of Cantre’r Gwaelod, a rich and fertile land that was flooded in the sixth century after the well-maiden, Mererid, neglected her duties and let in the waters.⁶ There is also a parallel inundation narrative of Helig ap Glannawg’s kingdom, Tyno Helig, in Conwy Bay. The stories have influenced one another over time and may represent the same original narrative that has been localised around two traditional sixth-century individuals.⁷ In addition, the inundation narrative of Kêr Is, off the coast of Finistère in western Brittany, shares many similarities with the Welsh stories of submergence.⁸

In the sixteenth century a different version of the Cantre’r Gwaelod tale emerged, in which King Gwyddno Garanhir ruled over sixteen cities that were protected from the sea by an embankment and a system of gates, or sluices. In this version, Seithenyn is relegated to the role of the dyke-keeper, who in his drunkenness forgets his vital duties, and consequently the land is flooded and everyone but the king is drowned.⁹ This version was popularised in the nineteenth century by works such as T. J. Ll. Pritchard’s poem ‘The Land Beneath the Sea’ (1823) and T. L. Peacock’s novel The Misfortunes of Elphin (1829).¹⁰ The Cantre’r Gwaelod story also gave rise to the popular song ‘Clychau Aberdyfi’ (‘The Bells of Aberdovey’), an eighteenth-century composition that was included in M. J. Williams’s prize-winning collection of unpublished Welsh airs presented at the 1838 Abergavenny Eisteddfod.¹¹

Later in the century, Celticists such as John Rhŷs commented upon the different versions of the tale, but it was the second narrative that captured the attention of the wider public.¹² In guidebooks such as Ward Lock’s North Wales (1912) and the Blue Guides: Wales (1922) it was the careless Seithenyn who neglected his duty in ‘a fit of intoxication’,¹³ and in A Book of North Wales (1903) Sabine Baring-Gould provided an excerpt of the poem attributed to Gwyddno, which included the lines:

Accursed be the sea-warden, who, after his carousal, let loose

the destroying fountain of the raging deep.

Accursed be the watcher, who, after drunken revelry, let loose

the fountain of the desolating sea.¹⁴

Beyond the guidebooks, Cantre’r Gwaelod was prominent in the artistic and intellectual lives of the Welsh people. For instance, the story provided subject matter for papers presented to the Cardiganshire Antiquarian Society in 1912, and for prize-winning songs at the eisteddfodau held in Merthyr in 1881 and Treorchy in 1928.¹⁵ David Owen, from Denbigh, won the Chair Prize at the 1908 Eisteddfod held in Barmouth for a poem based on the Cantre’r Gwaelod inundation, and a journalist from Aberystwyth called Dewi Morgan repeated the feat at Pwllheli in 1925.¹⁶ The tale of Cantre’r Gwaelod was adapted by Ernest Rhys and the Anglo-Irish poet Alfred Perceval Graves for the Harlech pageant in August 1920, when 200 actors performed the opening episode, ‘The Great Sea Flood’, in front of 5,000 spectators gathered in the castle.¹⁷ The performance was reprised in 1922, when the Western Mail praised the way the ‘little maidens clad in green and silver’ entered the stage with a ‘dainty dance’ to suggest the overwhelming sea, ‘driving before them the terror-stricken women of the countryside’.¹⁸ More generally, a reference to Cantre’r Gwaelod could denote a metaphorical drowning, for instance, when a letter written in 1902 to the South Wales Daily News suggested that the Welsh language was being submerged beneath a wave of anglicisation, ‘as was Cantre’r Gwaelod by the Atlantic’.¹⁹

For travellers and readers alike, the principal interest in Cantre’r Gwaelod lay in contemporary debates about the story’s veracity. These debates tantalised the traveller with the possibility that the authenticity of the story was visible in the physical geography of the area. Along the west coast of Wales there are three sarns, or causeways, that project out into Cardigan Bay. In the north, Sarn Badrig (St Patrick’s Causeway) runs about thirteen miles out to sea from the coast at Pensarn, between Harlech and Barmouth. The central Sarn y Bwch runs between four and five miles seaward from Tywyn, while the southern Sarn Cynfelyn projects eight miles out to sea from Y Borth, just north of Aberystwyth, culminating in a group of rocks called Caer Wyddno.²⁰

For many observers the remarkably straight form of the sarns indicated human design, which meant that the causeways might represent tangible traces of the system of embankments and sluices that had protected Cantre’r Gwaelod from the sea. Capitalising on the physical evidence, newspaper articles enticed tourists to the seaside towns of western Wales with the promise that they might find ‘traces of the legendary twelve walled cities’ of Cantre’r Gwaelod for themselves.²¹ Travel writers were also intrigued by this prospect, and when A. G. Bradley looked down on Cardigan Bay from the coastal road he drew the reader’s attention to the line of surf caused by Sarn Badrig just under the surface, one of Cantre’r Gwaelod’s ‘artificial banks’. At certain times, explained Bradley, ‘the wall itself’ was visible, ‘showing high and dry above the waves’ and ‘stretching for miles into the bay’, while further south ‘another mysterious wall of a similar kind’ confirmed Bradley’s assertion that ‘in the reign of Gwyddno Garanhir, a fair land covered with villages spread out here’.²² The writer Maud Stawell seemed to undermine any earnest consideration of the Cantre’r Gwaelod story by beginning her retelling with a variation of the archetypal fairy tale opening: ‘There was once a time very long ago, it is said, when all the bay that lies upon our right was a fertile plain, the Plain of Gwaelod’. However, the storyteller’s lightness of touch was also grounded in the geography of Cardigan Bay, and Stawell proceeded to draw her reader’s eyes seawards where, at low tide, ‘you may sometimes see the long line of the broken dyke’ that protected the ‘fertile plain’.²³

Not everybody accepted that the sarns represented the remains of embankments and sluice gates. Several guidebooks reproduced the geologist A. C. Ramsay’s conclusion that Sarn y Bwch was a ‘sunken reef’, which ‘may indicate the continuation of the Cambrian rocks’ out to sea.²⁴ Accordingly, Black’s Guide to North Wales (1897) remarked that the sarns ‘we would ascribe to nature rather than to man’, and the Gossiping Guide to Wales (1902) reassured its readers that, despite the imaginative associations put forward by Welsh antiquaries, Sarn Badrig ‘is now known to be a natural formation’.²⁵ But these dismissive comments were rare examples of certainty on the matter. It was more common for writers to draw attention to curious details, as the Canadian journalist Beckles Willson did in his book Lost England (1902). Willson pointed out that Sarn Badrig, when visible at low water, shelves to the south while it is steep to the north – exactly in the manner of a defensive sea wall.²⁶ Taking up a nuanced position, the geographer H. J. Fleure provided a geological explanation for the formation of the sarns, but concluded with the teasing line:

he would be a bold man who would definitely rule out the possibility of human work having contributed something to the form of these ‘sarns’ in the days when the lowland was still inhabited.²⁷

Likewise, Eiluned and Peter Lewis revelled in the lack of definitive answer when it came to the sarns in their evocative passage:

The sensible Saxon may stop his ears and refuse to hear the echo of ghostly bells on stormy nights, but even he cannot rule out the possibility of some drowned human effort when the spring tides uncover great ‘Sarns,’ or causeways, far out in Cardigan Bay.²⁸

Geological and archaeological research added fuel to speculation about the deep history of Cardigan Bay, and strengthened the sense that the Cantre’r Gwaelod story might be built on plausible foundations. In Britain and Ireland an understanding of ice-age conditions developed in the middle decades of the nineteenth century, when the Swiss-American geologist Louis Agassiz introduced his glacial theory to William Buckland, Charles Lyell and the Geological Society of London.²⁹ Over the next century, geologists remained fascinated by the changing conditions around the British-Irish Isles from the end of the Pleistocene, when the ice began to retreat. As J. G. D. Clark explained in The Mesolithic Settlement of Northern Europe (1936), geologists had to contend with the interacting forces of eustatic sea-level rise, as water was released from the ice sheet, and isostatic rebound, where formerly depressed areas of the earth’s crust recovered after the glacial weight was lifted.³⁰ Geologists generally agreed that significant and localised changes in the relative sea level had taken place since the end of the Pleistocene, but their understanding of the seabed and sunken lands remained limited. Moreover, the questions of precisely when marine transgression took place, its extent, and whether it was gradual or catastrophic in nature, remained open.

The geology and geography of Cardigan Bay attracted so much attention because the low tide exposed the roots and trunks of submerged forests. In ‘Geology of Aberystwyth’ (1878), Walter Keeping noted the two sets of red deer antlers discovered in the submarine clay bed near Borth, where there were also trunks and branches of oak, pine and beech indicating the old land surface.³¹ The press drew readers’ attention to the visible remains of the ancient forests, too, and summarised lectures on the once ‘fertile plain’ that spanned Cardigan Bay.³² The sense of mystery lingered around sunken lands, so that by 1935 T. A. Glenn made essentially similar arguments to Keeping in an article for Archaeologia Cambrensis. Glenn drew attention to the submerged forests visible at low ebb from the Mersey to Cardiganshire and the seventy or more recent finds of prehistoric bronze, stone, bone and shell, but admitted that the dates for the various periods of uplift and submergence remained ‘yet to be worked out’.³³

As a result, when geologists and other captivated intellectuals turned their critical faculties to the legend of Cantre’r Gwaelod they could not simply dismiss it as mere fantasy. For the geologist William Boyd Dawkins, it was enough to sprinkle enticing references to Cantre’r Gwaelod in his books and speeches. In Early Man in Britain (1880), for instance, Dawkins analysed the geological evidence of land depression and suggested that it ‘could not be less than from thirty to forty feet’. This would, of course, add a considerable amount of land to the surface of Britain, especially in Cardigan Bay, ‘where the Welsh peasant still tells the story of the land swallowed by the sea’.³⁴ Others, however, subjected stories of sunken lands to more sustained analysis. In 1921, O. T. Jones, Professor of Geology at the University of Manchester, brought geological and archaeological evidence ‘to bear upon these legends’, to ascertain

whether there is anything in them beyond evidence of the fertile imagination and inventive ability of the Welsh and related Celtic peoples.³⁵

Jones explained that the geological record attests to ‘a general sinking’ of the southern parts of Britain and Ireland, ‘proved especially by the submerged forests’ at Borth, Barmouth and Tywyn. Various excavations during the construction of bridges and docks at Barry, Swansea and Barmouth revealed peat and trees up to sixty feet below the contemporary high-tide line, and the recovery of polished stone, bone and flint chips suggested that ‘Neolithic man’ – the ‘ancestors of the present population of Wales’ – once occupied the sunken lands.³⁶ Bringing the evidence together, Jones concluded:

It appears that the so-called fairy tales or folk tales of Wales deserve to be taken seriously, as they seem to have preserved for us by oral tradition past events in the history of the country and its people which would otherwise have been unrecorded.³⁷

Jones’s position was shared by D. J. Davies, who regarded the ‘glowing description’ of Cantre’r Gwaelod as the ‘poetic license’ exercised by Welsh bards, but admitted that ‘it is possible’ the legend derived from Neolithic people who witnessed the submergence of southern Britain after the glacial epoch.³⁸ Discussions of Cantre’r Gwaelod were characterised by this sort of equivocal language, which left the legend open to interpretation. This position was articulated by the University of Cambridge botanist Harry Godwin, who remained unsure whether the legend was ‘a genuine, if modified, folk memory of an actual catastrophe’, or a tale inspired by the sight of a submerged landscape.³⁹ Either way, the inability of such self-consciously rational and scientific minds to wholly dismiss the legend of Cantre’r Gwaelod produced a novel sense of curiosity, delight and wonder.

The earnest engagement with the story of Cantre’r Gwaelod and its landscape setting filtered into travel writing, popular history and the press. Writers of all kinds replicated the geologists’ distinction between the imaginative details of the story and the deeper truths that it communicated. This approach is encapsulated in an exchange of letters in the Manchester Times published in 1893. In the ‘Notes and Queries’ section, W. A. Part asked whether the ‘marvellous yarns’ about sunken land told by his skipper during a voyage across Cardigan Bay had any substance to them, and a reply the following week stated that the existence of Sarn Badrig, supposedly ‘a ruined sea wall’, suggests that the yarns ‘have no doubt some foundation in fact’.⁴⁰ The ‘foundation in fact’ explanation was repeated by Fleure in his introductory article in The Blue Guide: Wales (1922),⁴¹ and an 1888 article in the Aberystwyth Observer referred poetically to ‘sermons in stones’ that suggested the tradition ‘is not entirely to be disregarded’.⁴² Similarly, in A History of Wales (1911), J. E. Lloyd suggested that the story might represent memories ‘handed down through many generations’ of a subsidence ‘attested by geology’,⁴³ and in the same year the Encyclopaedia Britannica cited the ‘indications of submerged dwellings and roads’, which showed that the tradition was ‘evidently based on fact’.⁴⁴ In this case, even if the sarns were natural rock formations and there was no episode of drunken negligence in the Lowland Hundred, the story retained a foundation of truth that attested to historical human experience. This realisation was enchanting in its own right, and it relied on critical geological and archaeological inquiry.

Geological inquiry and travel writing came together in the pages of William Ashton’s Evolution of a Coast-Line (1920), in which the author recalled his visit to Cardigan Bay in September 1910, when Sarn Badrig was revealed at the lowest tidal ebb (see Figure 3.1). Ashton was struck by the ‘strong flavour of romance about the past of Cardigan Bay’ but maintained that he was a ‘critical enquirer’ who resolved to ‘proceed deductively’ and carry out ‘direct observation’ to ‘disentangle romance from fact’.⁴⁵ Marshalling the evidence of the traditional accounts, antiquaries, geologists, reports from local inhabitants and his own observations, Ashton advanced a series of arguments. First, combining geological knowledge of post-glacial subsidence in southern Britain with evidence provided by ‘experienced observers’ along the coast, Ashton argued that the subsidence was ‘still in progress’ at a rate of around three feet per century. This would account for forty-two feet of subsidence across fourteen centuries which, if reversed, would bring an area of land above the water corresponding ‘fairly well with that assigned to the lost Cantref Gwaelod’.⁴⁶

Second, Ashton accepted the conclusions of Professors Ramsay, J. R. Ainsworth and others that the sarns were naturally occurring at their foundation, and that tidal currents had covered them in stones, boulder-clay deposits and shingle.⁴⁷ However, after considering the historical precedent of embankments along the north German coast as well as Roman sea defences built in southern Britain and in Holland, Ashton surmised that ‘similar protection would be made on the subsiding Cardigan Bay coast’, and that ‘advantage would be taken of any lines of natural elevation upon which to erect such embankments’. This led Ashton to the conclusion that ‘embankments were once constructed on the natural rocky ridges of Cardigan Bay’.⁴⁸

Third, Ashton heard testimony given by Mr R. G. Humphreys from Porthmadog, who saw ‘a layer of 6 or 8 large oblong stones, as if part of a wall, near the south side of Sarn Badrig’, and Mr David Jones from Barmouth, who said that he found ‘stones with mortar adhering to them’ when dredging in six fathoms of water around six or seven miles from land opposite Dyffryn.⁴⁹ Bringing all of the evidence together, Ashton admitted that ‘we can hardly dismiss as unsupported legendary lore the tradition that the chief of the sixteen lowland towns, Caer Gwyddno, did once stand near the end of Sarn Cynfelin’.⁵⁰ So in the end, and despite Ashton’s efforts, ‘romance’ and ‘fact’ remained seductively intertwined.

The legend of Cantre’r Gwaelod entered the realm of public debate as contemporaries found new resonances in the account of lost land while witnessing destruction and flooding along their storm-battered and eroding coastlines. In north and west Wales the main railway lines were particularly vulnerable to storms and gales because they followed the coast for long stretches. In 1883 the railway embankment was breached at Barmouth, and along the Chester to Holyhead line Thomas Telford’s Stanley Embankment was partially washed away in 1887.⁵¹ In Pwllheli, a tramway used by tourists was damaged by a storm in 1896 and the same storm washed away the railway at Rhyl.⁵² There was a tragic incident in 1899 when the railway embankment collapsed between Conwy and Penmaenmawr, leaving the rails and sleepers intact but completely unsupported. An unsuspecting express goods train, travelling at night, crossed this section of track and tumbled into the sea, killing the driver and fireman.⁵³ Further destruction occurred at Rhyl in 1905 and again between Criccieth and Porthmadog in 1938, where a bridge collapsed shortly after a Great Western Railway train carrying sixty schoolchildren had crossed it.⁵⁴

The problem was partly caused by the infrastructure itself. Railway embankments caused coastal squeeze by preventing the natural migration of vegetation across inter-tidal land. In consequence, saltmarshes became mudflats, and as the vegetation declined so did its baffling effect on the energy produced by waves.⁵⁵ The embankments therefore created the conditions for their own destruction, and the combined threats of storms and erosion endangered the railway lines, roads, promenades, accommodation, tramways and pavilions in rapidly expanding seaside towns which relied economically on the summer tourism traffic.

The frequency of such destructive events meant that sea defence and coastal erosion loomed large in the public consciousness, and the story of Cantre’r Gwaelod became a reference point for people across the country who raised concerns and demanded solutions. In 1896, a short article in the South Wales Daily News responded to the coastal erosion near Criccieth by asking ‘[a]re we going to have a second edition of [the] Cantre’r Gwaelod disaster?’⁵⁶ The legend even carried weight for those on the other side of the country. In the same year, after coastal erosion forced the abandonment of a coastguard station near Sheerness in Kent, a report in the Dover Express recognised that legends of dwellings ‘under the water of Cardigan Bay’ need not be accepted ‘in every particular’ for people to acknowledge that they contained an ‘element of truth’ and conveyed the strong belief held by past societies ‘in the overwhelming power of the ocean’ – a power that was still ‘going on before our eyes’.⁵⁷

In July 1906 a Royal Commission on Coast Erosion and Afforestation was appointed, chaired by Ivor Guest, 1st Viscount Wimborne, which brought together landowners, geologists, lawyers, and experts on sea defence with representatives of various government departments to investigate coastal erosion as a matter of national importance. This was welcome news to the people of Morfa Nefyn, a seaside town on the Llŷn Peninsula where the golf course and camping facilities were threatened by the encroaching sea. Moreover, according to a report in the North Wales Weekly News later that year, plans to extend the Colwyn Bay promenade were as much about improving the sea defences as about ‘visitors’ happiness’, and it was said that Colwyn Bay Hotel had been much further inland in the past. The article explained that the threat of the sea ‘is very real to us in Wales’, and, after considering the warning contained in the legend of Cantre’r Gwaelod, wondered ‘whether it is not possible that science has made sufficient progress by now’ to ‘rescue us from the inundation with which we are now threatened’.⁵⁸

The Royal Commission published its third and final report in 1911, which revealed the complexity of the problem without offering clear solutions. The report recommended that the Board of Trade be made responsible for sea defence and be given the power to control the removal of materials from the shore, but, as the civil engineer J. S. Owens pointed out, it also revealed differences of opinion among members of the commission as well as the ‘paucity’ of knowledge regarding coastal erosion.⁵⁹ The report itself admitted that systematic observation of the coastlines was required because ‘the information at present is scanty and vague’, and concluded that erosion was a localised problem and not a national service.⁶⁰ It was clear that much remained to be done, and frustration at the lack of action in the following decades was expressed in a letter to the Western Mail in 1929, when T. Wynne Thomas of Aberdyfi warned that Cantre’r Gwaelod was lost ‘owing to the laxity of the then Government officials’.⁶¹ Contemporaries recognised Cantre’r Gwaelod as a legend conveying the general and pressing truth that coastlines are elastic, the forces of nature have destructive potential, and human habitation is precarious.

Lyonesse

The lost land of Lyonesse is a legendary country that is conventionally situated between Land’s End and the Scilly Isles. Various iterations of the inundation story describe how many villages and 140 churches in this fertile tract of land were submerged:

when the sea made a clean sweep of the country and rushed, with stupendous speed, across the flat wooded lands until it was brought to a halt by the massive cliffs of what is now the Land’s End peninsula.⁶²

In its modern form, the legend of Lyonesse is a combination of several separate flood traditions, information taken from manuscript sources, and modern literary adaptations, which together formed the popular and thoroughly commercialised tale that enchanted travellers and writers in Cornwall and the Scilly Isles.

Several strands of the Lyonesse legend are French in origin, including the late twelfth-century Roman d’Aquin, which contains an inundation story within its fictitious account of a campaign waged by Charlemagne against pagans in Brittany. Another early strand is the twelfth-century Roman du Mont Saint-Michel, which is a history of the titular abbey translated (with additions) into Norman dialect from Latin documents that contains an account of a submerged forest. Alongside the inundation narratives, the earliest appearance of the name ‘Lyonesse’ occurs in Geffrei Gaimar’s twelfth-century chronicle Estoire des Engleis (written in the 1130s), in which ‘Loeneis’ refers to Lothian, in Scotland. Variations of Loeneis appear in the twelfth- and thirteenth-century manuscripts of Arthurian romance as the home of Tristram, and the name continued to change over time as the stories became localised around Cornish and Breton places and entwined with local traditions. By the fifteenth century, Thomas Malory used the more recognisable forms ‘Lyones’ and ‘Lyonesse’ for the name of Tristram’s country in Le Morte d’Arthur (1470).⁶³

Independent of Arthurian romance, medieval English chronicles described great inundations of the sea in an unspecified location. In the entry for November 1099, the Anglo-Saxon Chronicle described how ‘the sea-tide ran up so very high and did so much to harm as no one remembered that it ever before did’,⁶⁴ and the Chronicon ex chronicis recorded that, in the same month, ‘the sea overflowed the shore, destroying towns, and drowning many persons, and innumerable oxen and sheep’.⁶⁵ The first account of a specifically Cornish inundation appears in the Itinerary of William of Worcester (1415–82), who based his account of the submergence of land around St Michael’s Mount on a manuscript that, Bivar suggests, was influenced by the Roman du Mont Saint-Michel. John Leland (c.1503–52) also described land between Penzance and Mousehole that had been ‘devourid by the sea’ in his Itineraries (c.1538–43) and noted the ‘roots of trees in dyvers places’ near the low water mark in Collectanea (1533–6).⁶⁶

These threads were woven together in the sixteenth century, when the name Lyonesse became attached to the tradition of a Cornish inundation in two influential chorographical works: Richard Carew’s Survey of Cornwall (1602) and William Camden’s Britannia (1586). Carew conflated the tradition of submerged land in Cornwall with the name ‘Lionnesse’, which he situated in the thirty miles between Land’s End and the Scilly Isles.⁶⁷ Carew’s Survey was published in the early seventeenth century but it was circulated in manuscript form beforehand, and likely influenced Camden’s corresponding remarks about a land ‘farther to the West’ than Land’s End, called ‘Lionesse’.⁶⁸ This was a key moment in the development of the legend because the status of Carew and Camden ensured that the location of Lyonesse between Land’s End and the Scilly Isles was repeated uncritically by subsequent writers.⁶⁹

From the mid-nineteenth century, Lyonesse was popularised in literature that reworked Arthurian romance, such as Matthew Arnold’s poem ‘Tristram and Iseult’ (1852) and especially Alfred, Lord Tennyson’s Idylls of the King (1859–85), in which Lyonesse is the setting for the final battle between Arthur and Mordred, a land that has risen from the sea and is ‘to sink into the abyss again’.⁷⁰ The connection between Lyonesse and the south-west was strengthened after the publication of Walter Besant’s Armorel of Lyonesse (1890), a novel that placed the legendary country in the Scilly Isles. In the late nineteenth and early twentieth centuries Lyonesse was a useful setting that helped to convey themes of remoteness and tragedy in the poetry, plays and novels of several writers including Algernon Charles Swinburne, Thomas Hardy, Arthur Quiller-Couch, Walter de la Mare and Mary Ellen Chase.⁷¹ Lyonesse recurs as part of Hardy’s Wessex in several of his works, including in the poem ‘When I set out for Lyonnesse’ (1870) where the speaker returns ‘With magic in my eyes’, and again in the novel A Pair of Blue Eyes (1873) and the short story ‘A Mere Interlude’ (1885). Finally, a few years before his death, Hardy retold the story of Tristram and Iseult in the play The Famous Tragedy of the Queen of Cornwall in Tintagel (1923), a story of misfortune ‘well suited to his genius’, as one contemporary review put it, in which even the sea ‘sighs mournfully at the foot of the castle wall’.⁷² The play was reworked the following year by Hardy and Rutland Boughton as a musical drama presented at the Glastonbury festival in 1924, and in 1936 Gerald Finzi set ‘When I set out for Lyonesse’ to music as part of his Earth and Air and Rain song cycle.

The Lyonesse story also circulated in stories of church bells being heard beneath the waves, and in the tale of the lone survivor – a man who escaped the inundation on his horse. This tradition existed in two versions, each associated with an ancient Cornish family. In the Trevelyan version, the survivor came ashore on a white horse at Perranuthnoe, which is why the family crest displays a white horse emerging from the sea. In the Vyvyan version, the survivor rode ashore somewhere near St Buryan, and for this reason the Vyvyan family crest also depicts a horse. In 1921, the leader of the Cornish language revival Henry Jenner claimed that ‘until very recently’, according to Sir Vyell Vyvyan, the family kept a horse ‘ready saddled and bridled’ in the stable at Trelowarren ‘in case it [an inundation] should ever happen again’.⁷³

Unsurprisingly, the widespread cultural presence of Lyonesse shaped the traveller’s experience at the Cornish coast. Travel writing combined and repackaged these sources so that the traveller gazing westward from Land’s End would, like Sidney Heath, ‘think instinctively of the lost land of Lyonesse’.⁷⁴ W. Fordyce Clark also thought of Lyonesse ‘instinctively’ on his approach to St Michael’s Mount, where he felt ‘conscious of a sense of enchantment which is not easily shaken off’.⁷⁵ It was the ‘raw edge’ of Land’s End that arrested the imagination of the novelist Geraldine Mitton, where she and many other writers quoted Tennyson’s description of Arthur’s fight against Mordred:

So all day long the noise of battle roll’d

Among the mountains by the winter sea;

Until King Arthur’s Table, man by man,

Had fall’n in Lyonnesse about their Lord.⁷⁶

Alongside its literary purchase, the Cornish tourism industry was quick to seize the Lyonesse legend and make it a distinctly Cornish tale. In 1889, the passenger steamer Lyonesse began plying between Penzance and the Scilly Isles, crossing daily during the busy summer season. The steamer proved very popular with Scillonians and tourists alike, and frequently took over a hundred, and sometimes several hundred, excursionists on day trips from Penzance.⁷⁷ The association was reinforced by guidebook titles such as Faire Lyonesse: A Guide to the Isles of Scilly (1897) and Lyonesse: A Handbook for the Isles of Scilly (1900), and Lyonesse was also used for the name of a hotel on St Mary’s, as well as for the island’s cricket team. This was part of a broader claim over Arthurian legend, which also had strong connections with places in the English–Scottish borderlands and in Wales, but was most thoroughly exploited by the Cornish tourism industry.⁷⁸ As discussed in Chapter 1, the Southern and Great Western Railways made Lyonesse part of their marketing strategies, and Cornwall is drenched in Arthurian associations in J. P. Sayer’s Great Western Railway poster-map ‘Land of Legend, History and Romance’ (1934). The viewer’s eye is caught by storied locations such as Tintagel (‘Birthplace of Arthur’), Dozmary Pool (‘into which “Excalibur” is said to have been cast’), St Kew (‘Trysting place of Tristram & Iseult’), and Goss Moor (‘where King Arthur hunted’), culminating in the steamer route from Penzance which disappears off the western edge of the map, where ‘the sea covers the fabled land of Lyonesse’. The poster, which takes the form of an open book, is crammed with symbols, including the Cornish and the Great Western Railway coat of arms, and the bright colours and cartoon illustrations reflect the jovial and playful tone of the poster, in which Arthurian romance and the legend of Lyonesse rub shoulders with information about bathing beaches, golf courses and the region’s pilchard exports, as well as brief notes on historical events such as the Civil Wars and the failed invasion attempt of the Spanish Armada.⁷⁹ The Great Western Railway also designated the west Legend Land in a series of leaflets, in which George Basil Barham – under the pseudonym ‘Lyonesse’ – reworked ‘old tales’ told in the western parts of Britain served by the railway. ‘The Lost Land of Lyonesse’ was included in the second volume, which makes an appeal to the traveller in search of ‘truth mingled with the old legends’ as well as those looking for the ‘perfect place for a lazy holiday’ just ‘six-and-a-half hours from Paddington’.⁸⁰ The leaflets were well received, and they combined illustrations and descriptions of legendary lore with practical tips for visitors, while the map on the back page connected the stories’ locations with the Great Western Railway network.⁸¹

One response to the cultivation of an unashamedly commercialised Lyonesse was sober scepticism such as that adopted by J. C. Tonkin and Prescott Row, who signposted their transition from a passage on Lyonesse to a section on the Phoenicians, Romans and Anglo-Saxons as a move away from ‘charming stories of romance’ and towards ‘the facts of cold history’.⁸² Similarly, J. Henry Harris refused to be taken in by the Lyonesse story when he visited Land’s End, insisting that, whatever the traveller reads or hears, ‘[t]he fact remains that from Land’s End to Scilly is blue water’, and everyone who stands on the cliffs of Land’s End ‘turns his face seawards, and says, “There’s nothing between me and America”’.⁸³ Even the Great Western Railway saw no difficulty in refashioning Lyonesse into a marketing tool while simultaneously insisting to readers that the legend was ‘groundless’.⁸⁴ The Cornish writer Walter H. Tregellas was far more aggressive in his rejection of the Lyonesse tradition, and wished to mention the story only ‘to consign it to oblivion’.⁸⁵ A similarly sceptical argument was made in an article about Lundy Island in the Exeter Flying Post, which claimed the idea that the island might be another relic of Lyonesse was confined to geologists whose ‘poetical instincts are more vivid than their scientific consciousness’.⁸⁶

Another mode of engaging with the Lyonesse legend was much more akin to the travel writer’s use of storytelling conventions when indulging playfully in the mythical land Tír na nÓg. One point of similarity was the use of qualifying remarks in an introductory sentence, which served as a signal for the reader to consciously indulge in the Lyonesse legend. Arthur Norway, for instance, explained that at Land’s End ‘[t]his cliff, if there be any truth in tradition, did not always border the seashore’.⁸⁷ Likewise, the Ordnance Survey Archaeology Officer O. G. S. Crawford began his article on the Scilly Isles with the sentence : ‘Once upon a time (so tradition says) a region of extreme fertility lay between the Scilly Islands and Cornwall’.⁸⁸ Norway’s commas and Crawford’s parentheses are interruptions that serve as a knowing nod to the reader, prompting their scepticism at the same time as their imaginative faculties are being activated. Perhaps in an attempt to satisfy both the imaginative dreamers and the hard-headed sceptics, the journalist J. W. Lambert offered a choice to his readers at the edge of the land, where ‘you may if you wish marshal your visions of Lyonesse’.⁸⁹ In the same vein, Arthur Salmon mentioned that the visitor could, ‘if we choose … dream of some old Atlantis’ when looking out across the shallow western seas.⁹⁰ The effect of all these qualifying remarks is to encourage a suspension of disbelief; a knowing indulgence in the imagination, which can share in the wonders of Lyonesse without being beguiled by them. For Saler, Landy and the scholars of re-enchantment, this is a distinctly modern form of enchantment which ‘delights’ but does not ‘delude’.⁹¹

Indeed, some imaginative effort on the part of the traveller or reader was required for the full experience to be attained. To reject any sort of engagement with the story for want of definitive proof was to miss the enjoyment that could be found in the murky combination of Arthurian tradition, the Cornish landscape, and the distant past. Cornwall and its legends did not shine brightly but instead, as F. A. Ginever put it in the Cornishman, they ‘shimmer with romance’ in a region where ‘one walks on enchanted ground’.⁹² Arthur Norway emphasised the power of vagueness by asking his readers directly: ‘Is there any truth in this old story? Was there ever such a land stretching westwards from these cliffs?’⁹³ The clarity-seeking visitor was a figure of ridicule in Black’s Guide to Cornwall (1919), which poked fun at an ‘American fellow-tourist’ at the Tintagel ruins who was disappointed not to see the Round Table itself, and the guidebook teased that they were ‘even half expecting the Holy Grail to be on view’. When it came to Arthurian romance, to demand clarity was to misunderstand that ‘fragments and traces’ and ‘[p]uzzling hints’ held the key to enchantment, and when it came to the legend of Lyonesse, the guidebook concluded: ‘Let us, by all means, give ourselves up to such poetic dreams as have grown to haunt the Cornish headland’.⁹⁴

Travel writers also reported their encounters with local people, who served the dual purpose of either intensifying or mischievously undermining the legend of Lyonesse. After raising several questions with the reader, Arthur Norway turned his attention to ‘those who haunt the cliffside cottages’, among whom:

many a burly giant will drop his voice to an awed whisper as he tells you how he has seen towers and houses far down in the depths of the transparent water, and long oarweed waving round the steeples which once summoned the folk of some wide parish to prayer.⁹⁵

In this passage, Norway implied that the traveller and perhaps even the reader could indulge in the ‘rich Keltic fancy’ of the local people, whose thoughts, he suggested, ‘run quickly into the land of pure imagination’.⁹⁶ In contrast, when C. Lewis Hind met with two coastguards while tramping around Cape Cornwall he was initially met with scepticism when he suggested that between them and the Scilly Isles – ‘fairy isles floating in a fairy sea’ – lay the lost land of Lyonesse:

The coastguards were dubious. ‘There’s deep sea there,’ said the bearded one, ‘though the Seven Stones between here and Scilly be land sure enough’.

After Hind persisted by explaining in more detail the supposed location of the drowned land, one of the coastguards admitted: ‘ A chap told me that when the sea is quiet you can hear church bells ringing down below. I dunno’.⁹⁷ Later in the narrative, at Perranuthnoe near St Michael’s Mount, Hind tried his luck with his driver, to whom he mentioned the tradition that they were near the spot where an ancestor of the Trevelyan family survived the inundation on his horse:

The driver of the dogcart had not heard of the miracle of Perran Uthnoe. No Cornishman that I ever met was familiar with one of the legends; it is the guide-books that perpetuate them. ‘They wouldn’t know what you mean by Perran Uthnoe about here,’ he said smiling. ‘We call it Perran’.⁹⁸

For many travellers, then, the principal method of approaching the Lyonesse story was in a playful and knowing manner. Travellers frequently displayed a willingness to indulge in the story, to be entertained by it, and to give themselves up to ‘poetic dreams’ or even find enjoyment in encounters with local people who undermined them. However, there was also a strand of interest that treated the legend with the same type of geological curiosity as Hy Brasil and Cantre’r Gwaelod. Submerged forests in Mount’s Bay and submarine stone structures in the Scilly Isles brought geology, archaeology and tradition together in lively discussions about the truth behind the Lyonesse legend, in which the traveller might hear echoes of a historical inundation of the land.

Many of these discussions centred on the submerged forest in Mount’s Bay, which had provoked interest at least as early as 1757, when the Cornish naturalist William Borlase encountered roots, trunks and branches of oak, willow and hazel up to ‘300 yards below the full-sea-mark’, where they lay below ‘at least 12 feet of water’ at high tide.⁹⁹ In Submerged Forests (1913), the geologist Clement Reid summarised the state of contemporary knowledge when he explained that in Cornwall ‘old land-surfaces can be found far below the sea-level’, and he drew attention to Mount’s Bay, where the submerged traces of stumps and roots suggested that the notion that St Michael’s Mount was once surrounded by a wood ‘is probably quite accurate’.¹⁰⁰ Reid suggested that the change in sea level amounted to ‘80 feet, or perhaps rather more’, and, based on the known rate of ongoing encroachment, he estimated that ‘the sea cannot have reached the Mount till long after the Roman period’.¹⁰¹ In the book’s conclusion, Reid wondered whether stories of deluge had their origin in the inundations evidenced by the geological record. Rather than pursuing the matter further, Reid merely acknowledged that the submergence of land probably began around 3000 BC. This was not a period of ‘great geological antiquity’, but a time when ‘geology, archaeology, and history meet and overlap’.¹⁰²

If Reid was unwilling to pursue the matter, travel writers certainly were, and many of their accounts drew on the antiquarian, chemist and photographer Robert Hunt, who wrote about Mount’s Bay in Popular Romances of the West of England (1865). In his chapter on ‘Romances of Lost Cities’, Hunt aimed to ‘show that those dim traditions point to some buried truth’.¹⁰³ One of these traditions was preserved in the name of an anchoring place on the western side of Mount’s Bay called Gwavas Lake. ‘It is not a little curious that any part of the ocean should be called a lake’, wrote Hunt, and according to tradition the lake was overwhelmed along with churches, houses and people during the ‘great flood’ that separated the Scilly Isles from England. Furthermore, evidence in support of this tradition could be seen on ‘a fine summer day, when the tide is low and the waters clear’, in the form of a submerged forest.¹⁰⁴ In a footnote Hunt recalled, as a schoolboy, setting out past the large tree trunks cast upon the shoreline and walking far out into the bay after a ‘violent equinoctial gale’ where there were ‘scores of trees embedded in the sands’.¹⁰⁵ This evocative memory made an impression on C. Lewis Hind, who quoted a sizeable chunk of Hunt’s text when, as he looked down on Mount’s Bay, his thoughts turned to the ‘primeval forest’ and ‘the tempest and flood that overwhelmed Lyonesse’.¹⁰⁶ Charles G. Harper also quoted Hunt’s childhood memory, before confirming that ‘I, too, have found, cast upon the shore, traces of this submarine forest’.¹⁰⁷

Geological research had established the idea that, as an article in the Cornishman put it in 1896, ‘a great part of the British Isles had sunk below the sea, and risen again, time after time’.¹⁰⁸ In other words, ‘this Kingdom was and is, physically speaking, in a fluid state’.¹⁰⁹ In this way, geology carved out new space for the imagination when it came to stories of sunken lands. In a lecture on the ‘History of Devonshire scenery’ in 1901, the principal of the Royal Albert Memorial College, Arthur W. Clayden, took this to the extreme when he declared that the ‘fabled land of Lyonesse there can be no doubt was once a reality’,¹¹⁰ and Thomas Cornish had made a similarly bold claim when he told the Penzance Natural History and Antiquarian Society that ‘science would, before long, establish the fact that the Land’s End and the Scilly Islands were at one time united’ and that Lyonesse ‘is a reality’.¹¹¹

Most travel writers did not stretch the imagination so far, and instead remained enchanted within firm limitations, sticking to what might be deduced from the physical evidence of the submerged forest in Mount’s Bay. Arthur Salmon claimed that tales such as Lyonesse were generally ‘inaccurate in detail’ but were ‘rarely baseless’, and the ‘fact that the shallow seas extend far westward cannot be ignored’. When it came to Lyonesse, wrote Salmon, ‘we are not dealing with absurdities’, and this was ‘an alluring vision on which we can linger without the sense of being actually unhistoric’.¹¹² This was typical of the way travel writers tempered their imaginations. In Sidney Heath’s words, the denudation was ‘of course within the bounds of geological possibility’, and ‘there is probably a certain rough truth in the old legends’.¹¹³ For Arthur Norway the alluring and enchanting element of the legend was ‘a kernel of truth’ that, however ‘twisted or concealed by careless repetition’, could not be ignored.¹¹⁴

At the other side of the submerged realm, the Scilly Isles were equally compelling to naturalists and geologists. After some initial observations in the eighteenth-century publications of Robert Heath (1750) and William Borlase (1753, 1756), a more systematic treatment of the islands’ geology emerged in Henry De la Beche’s Report on the Geology of Cornwall, Devon and West Somerset (1839) and Joseph Carne’s ‘On the Geology of the Islands of Scilly’ (1850).¹¹⁵ By the early twentieth century, it had long been understood that a submerged inland valley existed between the islands of Samson, Tresco and Bryher, which, as George Barrow explained in Geology of the Isles of Scilly (1906), was formed through a combination of denudation and subsidence.¹¹⁶ As in Cornwall, geological knowledge of subsidence prompted discussions of the Lyonesse legend, but inspiration travelled in the other direction, too. For instance, in an article published in 1912, Samuel Hazzledine Warren presented evidence of a Mesolithic and Neolithic land off the Essex coast and proposed that it should be named the ‘Lyonesse Surface’.¹¹⁷

The interplay between research and tradition continued in the interwar period. In a journey that paralleled William Ashton’s visit to Sarn Badrig in Cardigan Bay, in 1927 the aforementioned O. G. S. Crawford published an account of travelling to the Scilly Isles in the first issue of Antiquity.¹¹⁸ Crawford’s article, ‘Lyonesse’, is heterogenous and entertaining. As a field archaeologist writing a journal article, he duly set out to collect and assess ‘evidence, both archaeological and historical’, and situated his observations in relation to previous research by Borlase, Barrow and Reid. Yet, as quoted above, the paper began with the traditional fairy tale opening ‘Once upon a time’ followed by a summary of the Lyonesse legend. In the manner of a travel writer, Crawford asked whether the story had ‘any real basis in fact, or is it merely an invention of the “dreamy Celt”?’¹¹⁹

To answer this question, Crawford set out on a ‘submarine walk’ across the sandflats between Samson, Tresco and Bryher, taking advantage of the lowest spring tides in March 1926. Accompanied by his photographer, Alexander Gibson, Crawford wished to photograph and examine at close quarters a long straight line of stones he had seen from the high ground on Samson. As they walked out over the sands, it became clear that ‘the line of stones was undoubtedly the remains of a wall of human construction’:

It was one of those thrilling moments which occasionally occur in the life of an archaeologist. Here before us was tangible proof that the land had sunk since prehistoric times; for no one makes walls like this below high water mark.¹²⁰

Gibson’s wide-angle photograph illustrates the straight trajectory of the ‘boulder-hedge’ across the tidal wetland from a distance, and this is accompanied by a detailed image of the stones with Crawford in the background, surveying the scene in contemplation. Gibson’s photographs also draw comparison with contemporary field walls on St Mary’s, which ‘are made in exactly the same way’.¹²¹ Crawford’s role as a pioneer in non-military aerial photography is indicated by his article’s use of aerial photographs of the flats, taken by the coastal reconnaissance unit No. 480 Flight based in RAF Calshot.¹²² These observations led Crawford to the belief that the ‘whole archipelago’ was at one time ‘one or more large islands’, and the contemporary sandflat was once a fertile level plain with deep soil and ‘sheltered on all sides by higher ground’. This is where the ‘prehistoric builders of the submerged walls’ grazed their flocks and herds until the inland plain was submerged by the sea.¹²³ Crawford suggested that, if Lyonesse were associated with the Scilly Isles rather than with the region between them and Land’s End, then there were ‘good reasons for believing that the substance of the legend is true’.¹²⁴

Crawford’s flair for storytelling recurs at several points in the article, including his brief description of getting caught at the turn of the tide while walking between Samson and Tresco. Furthermore, alongside geology, Crawford was equally interested in folkloric explanations of how the boulder-hedges came about, and he retold stories of giants taken from Hunt’s Popular Romances and drew on the work of the folklorist Charlotte Sophia Burne when making a comparison with tales that surround submerged structures in Shropshire.¹²⁵ Bringing his wide-ranging interests together, Crawford concluded that the fishermen who told tales of a sunken land were correct, although this process was gradual rather than cataclysmic. Combining enchantment with laconic humour in his closing line, Crawford remarked that the legend of Lyonesse ‘undoubtedly contains a vestige of antiquity’ but, given the ‘infinite slowness’ of the inundation, Trevelyan ‘hasted in vain’.¹²⁶

There was a wide range of responses to the Lyonesse legend. Some writers preferred to reject it outright whereas others looked for corroborating facts in medieval texts and in the archaeological and geological record. The Great Western Railway marketed the legend as an attraction for tourist consumption, and their playful engagement with the story set the tone for many travel writers. There remained, however, an irresistible ambiguity surrounding the origin of the tale that, as Jessie Mothersole put it, ‘sober, unromantic people spend their efforts and waste their breath in trying to disprove’.¹²⁷ Catherine A. Dawson Scott called in vain for a wealthy philanthropist to dredge the seabed between Land’s End and the Scilly Isles to ‘settle the vexed question for good and all!’¹²⁸ But the question could not be settled, and beneath the tale of Arthur’s last battle and Merlin’s magic the existence of some underlying truth was widely accepted. An article in the Globe in 1906 described this as a ‘qualified acceptance of tradition’ that was aided by science. Those who ‘mock at the tale of Lost Lyonesse’, the article went on, request ‘dates of this and that’ or ‘documents in support’, but these were the ‘subterfuges of the sceptic’ and the work of ‘mental miscreants’ who lacked ‘reverence for the ancestral ages’.¹²⁹ It was also a futile endeavour because Lyonesse exercised a power over the imagination that was enhanced by ambiguity, which in turn opened a space in which scepticism, rationalisation and imagination existed together in frustrating, tantalising and enchanting combination.

Conclusion

Mythical islands and sunken lands remain alluring today. At the 2023 Venice Architecture Biennale, called ‘The Laboratory of the Future’, the Irish Pavilion presented ‘In Search of Hy Brasil’. The exhibition was based on fieldwork undertaken on Ireland’s remote islands, showcasing local materials and raising awareness of sustainable living by promoting renewable energy, ethical food production and biodiversity.¹³⁰ Sunken lands periodically crop up in news stories when new evidence comes to light, such as the spectacular skull and antlers discovered at Borth in 2016 which, initially believed to be between 4,000 and 6,000 years old, are now known to have belonged to a deer that lived and died in the Bronze Age, between 1200 and 1000 BC.¹³¹ Moreover, as concerns about rising sea levels and climate change mount, the last decade or so has seen a series of major research projects in the sciences, demonstrating that the subject continues to grasp the academic imagination. Between 2009 and 2013, The Lyonesse Project carried out research into the coastal and marine environment of the Scilly Isles, concluding that a rapid submergence did indeed take place around 7000 BC, and the large area of saltmarsh between the islands would have been used for grazing animals before it eroded around the seventh century.¹³² Two years later another project, Europe’s Lost Frontiers, began mapping landscapes that were submerged at the end of the last ice age, focusing mainly on Doggerland, but also on areas such as Cardigan Bay.¹³³ From 2023, the Portalis Project has undertaken drilled core sampling, excavation and laboratory analysis to examine coastal communities in eastern Ireland and western Wales, exploring historical connections going back 10,000 years, and the pilot project revealed evidence of Mesolithic communities abandoning submerged terrain and settling on higher ground.¹³⁴

In an article called ‘Semantics of the Sea’, geo-archaeologists Erin Kavanagh and Martin Bates attempt to bring together ‘flood-related folk tales’ and geological evidence, and, in their research on the post-glacial coastal evolution of Cardigan Bay the physical geographer Simon Haslett and the historical linguist David Willis found parallels between geological and bathymetric evidence, tales of sunken lands, and historical sources such as the Gough Map.¹³⁵ The attempt to bring together geology, history and folklore remains appealing but, as Kavanagh and Bates admitted, the precise relationship between these ‘disparate’ bodies of evidence remains ‘opaque’.¹³⁶ The two chapters in this section have shown how we might understand the significance of this relationship, at least in the late nineteenth and early twentieth centuries, when the authority of geological and archaeological research provided a key ingredient in the new fascination with imaginative stories of elusive islands and sunken lands. This fascination was sustained by the power of ambiguity and doubt, and it took a variety of forms, from ‘double consciousness’ and playful indulgence to sincere and affecting encounters with the Atlantic edge. It also influenced the way people expressed their alarm at the precarity of coastal habitation and lack of control over ongoing erosion and rising sea levels, which reminds us that the sea was not solely associated with imperial expansion and the lyrics of ‘Rule, Britannia’. As one commentator put it in 1904, evidence of coastal erosion and tales such as Lyonesse and Cantre’r Gwaelod made it clear that ‘the sea that has made can also mar. The useful servant may become a tyrannical master’.¹³⁷ More broadly, the cultural developments explored in these two chapters were driven by the expansion of infrastructure, a mass print industry, and by self-consciously sceptical readings of the stories, which, in the end, could not be dismissed as pure fantasy. If these are some of the hallmarks of British and Irish modernity, then western landscapes, seascapes, elusive islands, and sunken lands were especially potent places of modern enchantment.

Notes

1. 1. K. E. Kavanagh and M. R. Bates, ‘Semantics of the Sea: Stories and Science along the Celtic Seaboard’, Internet Archaeology vol. 53 (2019), accessed 1 February 2025, https://doi.org/10.11141/ia.53.8; see W. B. Dawkins, Cave Hunting: Researches on the Evidence of Caves Respecting the Early Inhabitants of Europe (London, 1874); C. Reid, Submerged Forests (Cambridge, 1913); J. G. D. Clark, The Mesolithic Settlement of Northern Europe (New York, 1970 [1936]).

2. 2. An exception is the reference to the North Sea as ‘a wide and level country’ in H. G. Wells, ‘A Story of the Stone Age’ (1897), published in H. G. Wells, Tales of Space and Time (London, 1900), 61–163, at 61.

3. 3. W. Pooley, ‘Doubt and the Dislocation of Magic: France, 1790–1940’, Past and Present 261, no. 1 (2024), 133–67, quotations at 135, 138, 164.

4. 4. Pooley, ‘Doubt and the Dislocation of Magic’, 142.

5. 5. ‘Flood Legends’, in Celtic Culture, vol. 2, ed. J. Koch (Santa Barbara, CA, 2006), 754–5.

6. 6. ‘Flood Legends’, 754–5; R. Bromwich, ‘Cantre’r Gwaelod and Kêr-Is’, in The Early Cultures of North-West Europe, ed. C. Fox and B. Dickins (Cambridge, 1950), 215–41, at 221–2; ‘Cantre’r Gwaelod, Cantref-y-Gwaelod’, in A Dictionary of Celtic Mythology, ed. J. MacKillop (Oxford, 2004).

7. 7. ‘Flood Legends’, 754–5; Bromwich, ‘Cantre’r Gwaelod and Kêr-Is’, 229–31.

8. 8. Common spelling variations include the City of Ys or Is (English), Ville d’Ys (French), Ker Is and Kêr Iz (Breton).

9. 9. ‘Cantre’r Gwaelod, Cantref-y-Gwaelod’, in Dictionary of Celtic Mythology. The two identities of Seithenyn seem to merge in some accounts. For Seithenyn as the ‘Drunkard, Son of the King of South Wales’, see J. C. Davies, Folk-Lore of West and Mid-Wales (Aberystwyth, 1911), 321.

10. 10. Pritchard’s poem, and the wider movement of popularising legends attached to Welsh landscapes (and seascapes), is discussed in Morgan, ‘From Death to a View’, 86.

11. 11. ‘The Bells of Aberdovey’, The Cardiff Times, 25 July 1908, 10.

12. 12. John Rhŷs commented on the different versions of the tale in ‘Chwedloniaeth Gymraeg’, Y Traethodydd 35, no. 4 (1880), 476–84, specifically at 479–81.

13. 13. Quotation from North Wales (Southern) (1912), 100–101; Blue Guides: Wales (1922), 140. This version of the story was also summarised in ‘Cardiganshire’, in Chrisholm Hugh ed., Encyclopaedia Britannica, 11th edn (Cambridge, 1911), vol. 5, 320.

14. 14. S. Baring-Gould, A Book of North Wales (London, 1903), 239–40.

15. 15. ‘Cardiganshire Antiquarian Society’, Cambrian News, 20 September 1912, 6; ‘The London Welsh Choir and the Merthyr Eisteddfod’, South Wales Daily News, 7 January 1881, 4; ‘Second Day’s Awards’, Western Mail, 8 August 1928, 10.

16. 16. ‘Denbigh’, North Wales Times, 19 December 1908, 4; ‘National Eisteddfod Crown’, Western Mail, 7 August 1925, 9.

17. 17. ‘Pageantry at Harlech’, Western Mail, 24 August 1920, 8; A. Bartie, L. Fleming, M. Freeman, T. Hulme, A. Hutton and P. Readman, ‘The Pageant of Harlech Castle’, The Redress of the Past, accessed 1 February 2025, http://www.historicalpageants.ac.uk/pageants/1287/.

18. 18. ‘Harlech Castle Pomp’, Western Mail, 22 August 1922, 4.

19. 19. A. P. Gomer to the editor, ‘The Preservation of the Welsh Language. To the Editor’, South Wales Daily News, 31 May 1902, 3.

20. 20. Bromwich, ‘Cantre’r Gwaelod and Kêr-Is’, 225–7; W. Ashton, The Evolution of a Coast-Line (London, 1920), 268, 270. Sarn Dewi and Sarn Cadwgan are similar formations, though less commented upon in the source material.

21. 21. ‘Whitsun and Early Holidays’, Sheffield Independent, 5 June 1935, 3.

22. 22. Bradley, Highways and Byways in North Wales (1898), 388–9.

23. 23. Mrs R. Stawell, Motor Tours in Wales and the Border Counties (Boston, MA, 1909), 123–4.

24. 24. A. C. Ramsay, The Geology of North Wales, Memoirs of the Geological Survey of Great Britain and of the Museum of Practical Geology, vol. 3 (London, 1866), 26; Ramsay’s phrase ‘sunken reef’ is repeated in ‘Our Reviewer’s Table’, Barmouth and County Advertiser, 5 March 1903, 7.

25. 25. Black’s Guide to North Wales, 21st edn (London, 1897), 188–9; Askew Roberts and Edward Woodall, Gossiping Guide to Wales (North Wales and Aberystwyth) (London, 1902), 93. Black’s North Wales (1897) was similarly sceptical of those who ‘still fancy they can trace the ruins of Helig’s pride’ along the northern coastline between Llandudno and Penmaenmawr – see 38–9.

26. 26. B. Willson, Lost England: The Story of Our Submerged Coasts (London, 1902), p. 36.

27. 27. Muirhead, Blue Guides: Wales (1922), xix.

28. 28. Lewis and Lewis, Land of Wales (1937), 1–2.

29. 29. L. Agassiz, ‘On Glaciers, and the Evidence of Their Having Once Existed in Scotland, Ireland, and England’, Proceedings of the Geological Society of London 3, pt. 2, no. 72 (1840), 327–32; Agassiz had previously introduced his glacial theory in Études sur les Glaciers (Neuchâtel, 1840).

30. 30. Clark, Mesolithic Settlement (1970 [1936]), 7–8. In Scotland, the rebound outpaced the rise in sea level.

31. 31. W. Keeping, ‘Geology of Aberystwyth’, Geological Magazine, 5 (1878), 532–47, at 542–3.

32. 32. ‘Buried Welsh Cities. Cardigan Bay Once a Fertile Plain’, Liverpool Daily Post, 23 November 1932, 4.

33. 33. T. A. Glenn, ‘Distribution of the Graig Lwyd Axe and Its Associated Cultures’, Archaeologia Cambrensis, 90 (1935), 189–218 at 206–7. Quotation at 206 n5.

34. 34. W. B. Dawkins, Early Man in Britain (London, 1880), 253. Dawkins made a similar comment in his ‘Address on the Place of a University in the History of Wales’ (Bangor, 1900), quoted in John Rhŷs, Celtic Folklore (London, 1909), 388–9.

35. 35. O. T. Jones, ‘The Origin of Welsh Legends’, The Welsh Outlook 8 (1921), 309–12, at 309.

36. 36. Jones, ‘Origin of Welsh Legends’, 310–11.

37. 37. Jones, ‘Origin of Welsh Legends’, 312.

38. 38. D. J. Davies, ‘Cantref y Gwaelod’, Transactions and Archaeological Record, Cardiganshire Antiquarian Society 5 (1927), 21–33, at 26, 29.

39. 39. H. Godwin, ‘Coastal Peat Beds of the British Isles and North Sea: Presidential Address to the British Ecological Society 1943’, Journal of Ecology 31, no. 2 (1943), 199–247, at 225

40. 40. ‘Notes and Queries’, Manchester Times, 15 September 1893, 5; ‘Answers and Comments’, Manchester Times, 22 September 1893, 5.

41. 41. Fleure, ‘Land of Wales’ (1922), xviii. Also see H. J. Fleure, Some Aspects of British Civilisation (Oxford, 1948).

42. 42. ‘Places Worth Visiting. Aberystwyth’, Aberystwyth Observer, 10 November 1888, 8.

43. 43. J. E. Lloyd, A History of Wales: From Earliest Times to the Edwardian Conquest, vol. 1 (London, 1911), 5.

44. 44. Encyclopaedia Britannica (1911) vol. 5, 320.

45. 45. Ashton, Evolution of a Coast-Line (1920), 265.

46. 46. Ashton, The Evolution of a Coast-Line (1920), 272. Ashton’s claim that subsidence was ongoing was disputed by other geologists, who generally insisted that it took place in the Neolithic period. O. T. Jones disagreed with Ashton’s theory explicitly in ‘Origin of Welsh Legends’ (1921), 311.

47. 47. Ashton, Evolution of a Coast-Line (1920), 271.

48. 48. Ashton, The Evolution of a Coast-Line (1920), 273–4.

49. 49. Ashton, The Evolution of a Coast-Line (1920), 274–5.

50. 50. Ashton, The Evolution of a Coast-Line (1920), 265–75.

51. 51. ‘The Storm and High Tides’, The Morning Post, 20 October 1883, 5; R. Duck, On the Edge: Coastlines of Britain (Edinburgh, 2015), 69.

52. 52. Duck, On the Edge, 57–8, 70. Despite being re-laid further inland, the Pwllheli tramline was destroyed again in 1927.

53. 53. Duck, On the Edge, 70–71.

54. 54. ‘Havoc on the West Coast’, Western Chronicle, 1 December 1905, 7; ‘Gale and Flood’, The Times, 24 November 1938, 14.

55. 55. Duck, On the Edge, 200–201.

56. 56. ‘Welsh Gossip’, South Wales Daily News, 6 February 1896, 4.

57. 57. ‘Our Wasting Shores’, Dover Express, 15 May 1896, 3.

58. 58. ‘Greedy Neptune’, North Wales Weekly News, 14 September 1906, 7.

59. 59. J. S. Owens, ‘Royal Commission on Coast Erosion’, The Geographical Journal 38, no. 6 (December 1911), 598–601.

60. 60. Quoted in Owens, ‘Royal Commission on Coast Erosion’ (1911), 600.

61. 61. ‘Coast Erosion’, Western Mail, 22 March 1929, 6.

62. 62. G. B. Barham, Legend Land: Being a Collection of Some of the Old Tales in those Western Parts of Britain Served by the Great Western Railway, Now Retold by Lyonesse, vol. 2 (London, 1922), 29.

63. 63. A. D. H. Bivar, ‘Lyonesse: The Evolution of a Fable’, Modern Philology 50, no. 5 (1953), 162–70, at 162–4, 167–9.

64. 64. The Anglo-Saxon Chronicle, trans. E. E. C. Gomme (London, 1909), 215.

65. 65. The Chronicle of Florence of Worcester, ed. and trans. T. Forester (London, 1854), 206. This chronicle is now attributed to another monk, John of Worcester.

66. 66. Bivar, ‘Lyonesse’, 165–6. Worcester’s itinerary bears the influence of the twelfth-century French source Roman de Mont Saint-Michel, which in turn drew on the earlier, twelfth-century Roman d’Aquin. See Bivar, ‘Lyonesse’.

67. 67. R. Carew, The Survey of Cornwall. Written by Richard Carew of Antonie, Esquire (London, 1602), 3. EEBO, accessed 1 February 2025, http://name.umdl.umich.edu/A17958.0001.001.

68. 68. For an English translation of Camden’s comments, see W. Camden, Camden’s Britannia Newly Translated into English, With Large Additions and Improvements, trans. E. Gibson (London, 1695 [1586]), see the section on ‘Cornwall’.

69. 69. Bivar, ‘Lyonesse’, 170.

70. 70. A., Lord Tennyson, Idylls of the King (London, 1904), 402. The quotation is from ‘The Passing of Arthur’, first published in 1869.

71. 71. Examples include A. C. Swinburne, ‘Tristram of Lyonesse’ (1882); W. Besant, Armorel of Lyonesse (London, 1890); T. Hardy, ‘When I Set Out for Lyonnesse’ (1870), A Pair of Blue Eyes (1873), ‘A Mere Interlude’ (1885), and The Famous Tragedy of the Queen of Cornwall in Tintagel (1923); W. de la Mare, ‘Sunk Lyonesse’ (1921); M. E. Chase, Dawn in Lyonesse (1938).

72. 72. ‘ “The Famous Tragedy”. Dorset Players Perform in London’, Western Morning News, 22 February 1924, 4.

73. 73. H. Jenner, ‘Lost Lyonesse’, Western Morning News, 21 February 1921, 4.

74. 74. Heath, Cornish Riviera (1911), 38. For other examples of travel writers turning their thoughts to Lyonesse at Land’s End, see A. H. Norway, Highways and Byways in Devon and Cornwall (London, 1897), 302–3; Bates, Gretna Green to Land’s End (1908), 377–8; and Hind, Days in Cornwall (1907), 350.

75. 75. W. F. Clark, ‘The Guarded Mount. A Cornish Beauty Spot’, The Scotsman, 12 September 1927, 8.

76. 76. G. E. Mitton, Cornwall, painted by G. F. Nicholls (London, 1915), 55–6. For other examples of Tennyson’s lines quoted in guidebooks and travelogues, see A. R. H. Moncrieff, ed., Black’s Guide to Cornwall, 22nd edn (London, 1919), 175; J. Mothersole, The Isles of Scilly. Their Story, Their Folk and Their Flowers, 3rd edn (London, 1919 [1910]), 62–3.

77. 77. ‘The Lyonesse at Scilly’, Cornishman, 25 April 1889, 7; ‘About the Islands of Scilly’, Derbyshire Advertiser and Journal, 6 June 1890, 5; ‘Penzance’, Cornishman, 18 August 1898, 4.

78. 78. S. Barczewski, Myth and National Identity in Nineteenth-Century Britain: The Legends of King Arthur and Robin Hood (Oxford, 2000), 201–2.

79. 79. J. P. Sayer, ‘Cornwall, Land of Legend, History and Romance’, c.1934 [Great Western Railway poster].

80. 80. Barham, Legend Land, vol. 2 (1922), 28–31.

81. 81. For a review of the leaflet series, see ‘Lure of Legend Land. More Great Western Travel Pamphlets’, Bath Chronicle and Weekly Gazette, 9 February 1924, 25.

82. 82. J. C. Tonkin and P. Row, Lyonesse: A Handbook for the Isles of Scilly, 2nd edn (St Mary’s, 1900 [1898]), 32.

83. 83. J. H. Harris, Cornish Saints and Sinners (London, 1906), 30.

84. 84. GWR, Cornish Riviera (1905), 69.

85. 85. W. H. Tregellas, Tourist’s Guide to Cornwall and the Scilly Isles (London, 1878), 2.

86. 86. ‘Lundy Island’, Exeter Flying Post, 11 April 1883, 6.

87. 87. Norway, Devon and Cornwall (1897), 303.

88. 88. O. G. S. Crawford, ‘Lyonesse’, Antiquity 1, no. 1 (1927), 5–14, at 5.

89. 89. J. W. Lambert, The Penguin Guides: Cornwall (London, 1939), 96.

90. 90. A. L. Salmon, The Cornwall Coast (London, 1910), 15.

91. 91. Landy and Saler, ‘Varieties of Modern Enchantment’, 2.

92. 92. F. A. Ginever, ‘ “One and All” Notes’, Cornishman, 17 August 1905, 4.

93. 93. Norway, Devon and Cornwall (1897), 303.

94. 94. Black’s Cornwall (1919), 27–8.

95. 95. Norway, Devon and Cornwall (1897), 303–5.

96. 96. Norway, Devon and Cornwall (1897), 303–5.

97. 97. Hind, Days in Cornwall (1907), 163–4.

98. 98. Hind, Days in Cornwall (1907), 198.

99. 99. W. Borlase, ‘An Account of Some Trees Discovered Under-ground on the Shore at Mount’s-Bay in Cornwall’, Philosophical Transactions (1683–1775), vol. 50 (1757–8), 51–3.

100. 100. Reid, Submerged Forests (1913), 100–101.

101. 101. Reid, Submerged Forests (1913), 100–101.

102. 102. Reid, Submerged Forests (1913), 120.

103. 103. R. Hunt, Popular Romances of the West of England (London, 1865), 207–8.

104. 104. Hunt, Popular Romances of the West of England (1865), 219.

105. 105. Hunt, Popular Romances of the West of England (1865), 219–20n.

106. 106. Hind, Days in Cornwall (1907), 191–2.

107. 107. C. G. Harper, The Cornish Coast (South) and the Isles of Scilly (London, 1910), 206.

108. 108. ‘Occasional Notes’, Cornishman, 28 May 1896, 4.

109. 109. ‘How England Grows by Winning Land from the Sea’, Falkirk Herald, 30 July 1902, 8.

110. 110. ‘Devonshire Scenery. Mr Clayden’s Lecture Concluded Yesterday’, Western Times, 13 June 1901, 4.

111. 111. ‘Occasional Notes’, Cornishman, 15 March 1888, 4.

112. 112. Salmon, Cornwall Coast (1910), 14–15.

113. 113. Heath, Cornish Riviera (1911), 38–40.

114. 114. Norway, Devon and Cornwall (1897), 303–5.

115. 115. R. Heath, A Natural and Historical Account of the Islands of Scilly (London, 1750); W. Borlase, ‘Of the Great Alterations Which the Islands of Sylley Have Undergone Since the Time of the Ancients’, Philosophical Transactions 48 (1753), 55–69; W. Borlase, Observations on the Ancient and Present State of the Islands of Scilly (Oxford, 1756); H. De la Beche, Report on the Geology of Cornwall, Devon and West Somerset (London, 1839); J. Carne, ‘On the Geology of the Islands of Scilly’, Transactions of the Royal Geological Society of Cornwall 7 (1850), 140–54.

116. 116. G. Barrow, Geology of the Isles of Scilly (London, 1906), 1–3.

117. 117. S. H. Warren, ‘The Classification of the Prehistoric Remains of Eastern Essex’, The Journal of the Royal Anthropological Institute of Great Britain and Ireland 42 (1912), 91–127, at 104. This is mentioned in Crawford, ‘Lyonesse’ (1927), 10.

118. 118. Crawford founded the journal and was its editor until he died in 1957.

119. 119. Crawford, ‘Lyonesse’ (1927), 5.

120. 120. Crawford, ‘Lyonesse’ (1927), 5.

121. 121. Crawford, ‘Lyonesse’ (1927), plate I facing, page 5; plate II, facing page 6; quotation at 7.

122. 122. Crawford, ‘Lyonesse’ (1927), facing 14.

123. 123. Crawford, ‘Lyonesse’ (1927), 9.

124. 124. Crawford, ‘Lyonesse’ (1927), 5.

125. 125. Crawford, ‘Lyonesse’ (1927), 8, 13.

126. 126. Crawford, ‘Lyonesse’ (1927), 14.

127. 127. J. Mothersole, The Isles of Scilly. Their Story, Their Folk and Their Flowers (London, 1910), 76.

128. 128. C. A. D. Scott, Nooks and Corners of Cornwall (London, 1911), 101.

129. 129. ‘Lost Lyonesse’, Globe, 10 August 1906, 1–2.

130. 130. ‘Hy-Brasil – Ireland’s representation at the 2023 Venice Architecture Biennale’, Irish Government, Department of Tourism, Culture, Arts, Gaeltacht, Sport and Media, 4 May 2023, accessed 1 February 2025, https://www.gov.ie/en/press-release/04a32-hy-brasil-irelands-representation-at-the-2023-venice-architecture-biennale/.

131. 131. ‘4,000-year-old Red Deer Skull and Antlers Found in Borth’, BBC News, 28 April 2016, accessed 1 February 2025, https://www.bbc.co.uk/news/uk-wales-mid-wales-36157302; D. Davies, ‘Ancient Antlers Shed Light on Borth Coastline’, Cambrian News, 15 August 2022, accessed 1 February 2025, https://www.cambrian-news.co.uk/news/ancient-antlers-shed-light-on-borth-coastline-559308.

132. 132. D. Charman, C. Johns, K. Camidge, P. Marshall, S. Mills, J. Mulville and H. Roberts, The Lyonesse Project: A Study of the Coastal and Marine Environment of the Isles of Scilly (York, 2014).

133. 133. V. Gaffney and S. Fitch, eds, Europe’s Lost Frontiers, vol. 1, Context and Methodology (Oxford, 2022).

134. 134. I. Dunne, ‘New Evidence of Mesolithic Settlers Discovered in Ireland and Wales’, Portalis Project, 18 July 2023, accessed 1 February 2025, https://portalisproject.eu/new-evidence-of-mesolithic-settlers-discovered-in-ireland-and-wales/.

135. 135. Kavanagh and Bates, ‘Semantics of the Sea’; S. K. Haslett and D. Willis, ‘The “Lost” Islands of Cardigan Bay, Wales, UK: Insights into the Post-glacial Evolution of Some Celtic Coasts of Northwest Europe’, Atlantic Geoscience 58 (2022), 131–46.

136. 136. Kavanagh and Bates, ‘Semantics of the Sea’, see ‘Conclusion’.

137. 137. ‘Diminishing England’, Lincolnshire Chronicle, 20 May 1904, 6.