How do wells become holy?


WELLS IN DEPTH –  Tristan Gray Hulse


How do wells become holy?   The problem of determining the status or category of any particular ‘holy’ well – boundary well, saint’s well, healing well, whatever – is neatly exemplified in the interesting group of eight Welsh wells named in some way for ‘Arthur’ noticed by Francis Jones. That this group is relatively large is in itself interesting; Geoffrey Ashe was able to trace references to only a further three in the whole of Britain – two Arthur’s Wells in England, and an Arthur’s Fountain in Scotland: Ashe 1997, pp.52, 42. He appears to know nothing of any of these eight Welsh wells. There are/were five Ffynhonnau Arthur, one each in the counties of Denbigh, Glamorgan, Monmouth, Montgomery, and Pembroke (Jones 1954, pp.177, 187, 196, 202, 213). There is a Ffynnon Cegin Arthur at Llanddeiniolen, Caernarfon (Ibid., p.154, 3, 107); a Ffynnon Graig Arthur (‘the Well of/at Arthur’s Rock’) at Trelawnydd, Flintshire (Ibid., p.180); and a Ffynnon Penarthur, in St Davids’ parish, Pembrokeshire (Ibid., pp.211, 5, 17). The abundance of Arthur-related toponyms and localised Arthurian traditions in Wales – cf. Ashe 1997, passim – suggests that there may well be other Arthur Wells still to be identified. Of these eight, Ffynhonnau Arthur in Glamorgan and Monmouth are known only from Medieval records, both as boundary marks. Two more, Ffynnon Graig Arthur, and Ffynnon Arthur, at Llangollen, Denbighshire, are simply mentioned from notes in the Parochialia of Edward Lhuyd. However, the remaining four – two of which certainly were noted as boundary marks – exhibit features which clearly suggest that they functioned as holy wells in the conventional sense of the term. The Pembrokeshire Ffynnon Arthur was on the land of Henfedde farm, at Clydey, where two fairs were held annually at Henfedde. Ffynnon Arthur at Llanfihangel-yng-Ngwynfa, in Montgomery, functioned as a boundary mark at least from 1309, and its site was still identifiable in 1910. Its one-time importance is shown by the fact that it gave its name to a township. Destroyed in the late-nineteenth century, it was stone-built, 4 feet deep and 12 inches square, and surrounded by a wall, which shows that it was used for bathing, and thus, presumably, for cures. Ffynnon Penarthur (the name of which means, not the Well of Arthur’s Head – pen – but the Well at the end – pen – of Arthur’s land: Jones 1954, p.5) was a manorial boundary mark, but its otherwise sacred status is suggested by the fact that formerly a highly-decorated dark age cross-stone marked with the personal name Gurmarc (now in St Davids cathedral), along with two other decorated stones, stood by the well. Local etymologising has made of Ffynnon Cegin Arthur, the ‘Well of Arthur’s Kitchin’ (cegin – oily deposits in the water ‘support’ this supposition, in that romantic writers of a bygone age said that the locals believed these deposits to come from the cooking fats used in the kitchen of King Arthur, floated from the greasy vessels during the washing-up!), but more prosaically, its location in the landscape determines its original meaning as the ‘Well of Arthur’s Ridge’ (cegin). It was long used for bathing for cures, of animals as well as humans, and for a short period functioned as a modest spa. It was even the subject of a book, King Arthur’s Well, by A. Wynn Williams (Ibid. p.107, note 78).

These few examples clearly demonstrate that at least some boundary wells were ‘holy wells’ in the conventional or accepted sense; and with these, conversely, it is useful to consider a large group of wells whose names suggest such status, yet for the majority of which available substantiating evidence is completely lacking. Jones (Ibid., p.46) found reference to 76 wells dedicated in some way to our Lady (Ffynnon Fair, St Mary’s Well, Lady Well) [10], there being at least one in each old Welsh county: Pembrokeshire alone could boast fifteen (Ibid., p.207). Of these, less than one third (23 out of 76) exhibit evidence beyond the simple nameform which seems to substantiate some form of cultic use [11]; and even then it is doubtful in certain instances that the Mary churches or chapels mentioned in association with them in some way actually had any specific demonstrable connections with the wells themselves. Wales, no less than any other Catholic country, had numbers of Marian sanctuaries in the middle ages which were the goals of pilgrims, and some of these appear to have had Mary wells associated with them in some way [12]; but it is instructive to note that in these cases the Ffynhonnau Fair were only one part or element of the pilgrimage complex, and that in only a single instance did a well of our Lady become a major focus of pilgrimage in its own right 13]. Jones’ distribution map of Mary wells (Ibid., p.220, map 4) reveals that the heaviest densities of wells are in those areas most thoroughly settled from England from the Anglo-Norman period onwards: Flintshire and the eastern half of Denbighshire, western Meirionydd (i.e. around Harlech, with its castle), Pembrokeshire, and Glamorgan.

Celtic Christians were as devoted to the Virgin as any other Christians. As noted in the Introduction, the Celtic Churches received the cultus of the saints already fully formed, absorbing it with the rest of the new faith at the time of their conversion. Devotion to Mary was flourishing by the end of the seventh century, as is revealed in a continuous series of texts from Ireland, beginning circa 700 (cf. O’Dwyer 1976 for a detailed study of these sources). No comparable documentation survives from Wales (literary survival from Wales for this period is in any case virtually nil overall), but the continuous interaction of the Welsh and Irish Churches over the previous two centuries dictates that Welsh patterns of devotion would have closely paralleled those in Ireland. Texts such as the circa 800 Martyrology of Oengus Céli Dé (Stokes 1905) demonstrate that numbers of feasts in her honour were being observed, and carvings on crosses in Ireland and Scotland and manuscript illuminations reveal a familiarity with her iconography which suggests the presence of her icons in the churches. Relics of her hair were claimed at Armagh (Ibid., p.221). But Welsh Christians differed from their continental brethren in one respect, in that their churches always bore the names of native saints, never those of biblical or other saints, as became the custom in Europe from the sixth century at latest. The single exception to this rule was the Archangel St Michael (cf. note 6), who had churches dedicated to him in the pre-Norman period, as is witnessed in the Book of Llandaf. With a single possible exception (a c. 910 grant in Liber Landavensis mentions Lann Meiripenn Ros, glossed, possibly at a later period, as ecclesia Sanctae Mariae – see Davies 1979, p.122), even Mary had to wait until the Anglo-Norman invasion of Wales to have churches dedicated in her honour. What is commonly accepted as being a reference to the first Welsh church erected in her honour is an annalistic entry, its importance being stressed by the fact that the note is common to numbers of medieval Welsh chronicles. Thus, under 1155 (recte 1156), Brenhinedd y Saesson reads:

‘Ac y kyssegrwitt eglwys Veir yn Meivot (And the church of Mary at Meifod was consecrated)’.(Jones 1971, p.157).

The importance of all this for an understanding of the problem of Mary Wells in Wales – eighty or more wells dedicated in her honour, less than one third of which exhibit any evidence of any form of cultus whatsoever – becomes clearer when it is considered in the context of the sacred landscape of the country. Very briefly, in the Celtic period llannau (sacred, or monastic, enclosures – today’s churchyards) were founded, in which churches were erected, then or subsequently. The llannau were named after their founders – the Celtic saints, as we think of them today – or named for great monastic founders by their disciples. Because these enclosures, their churches, and the lands around the llannau (the ancestors of the modern parishes which still bear the same names) were once occupied by their founders, a variety of landscape features around each – rocks, caves, mountains, even trees, etc. – very commonly bore their names, because they had owned or used them in some way. Principal among such landscape features, of course, were springs, for a source of clean water would have been essential for such ecclesiastical communities, for drinking, washing, and cooking, and possibly also for baptisms. (It is noteworthy that the founders, being monks, almost always chose remote and uninhabited sites for their foundations – the villages and towns which today take their names from the llannau grew up around them at a later period) [14]. After their deaths, the founders came to be venerated as saints, with all the outward expressions of such cult – holy tombs, and other relics, pilgrimages, feastdays, etc. – and the local landscape of the founder developed into the sacred landscape of the saint. Imitating older saints’ cults, their wells developed functions other than the purely utilitarian, such as drinking or bathing in the water for healing; or as a locus for penance and pilgrimage. Charms and divination could be included in such rituals, for, as scholars are increasingly able to recognise, such practices were thought of as morally or religiously neutral, rather than as pagan, until the developed theology of the high middle ages began to demand their extirpation (cf. e.g., Jolly 1996, on England in the late Anglo-Saxon period); and even then it was possible for ritual magic to spring up and flourish in an exclusively Christian context, with a detailed rationale and defence of such practices, despite the opposition of theologians such as Aquinas (see for example, the articles by Robert Mathiesen, Nicholas Watson, Claire Fanger, and Richard Kieckhefer, in Fanger 1998, pp.143-265) – such items of practice were not of necessity ‘pagan’ in origin, as is so often assumed. A similar sacrality was perceived to inhere in all the local landscape features which had associations with the saint, and on individual pilgrimages or more formally in processions on feastdays, were each visited in turn, each round of the sites delineating the sacred landscape of the saint. To the present day, such ritual perambulations of the saint’s landscape (occasionally in legend referred to as literally following in the footsteps of the saint, who was said to make the rounds yearly, or even daily) are a regular feature of individual or collective pilgrimages in honour of certain Celtic saints in Brittany and Ireland (for numbers of pertinent Irish examples, see Logan 1980, passim). These countries, of course, have retained their Catholicism, and thus their direct links with their Celtic pasts, and there is sufficient evidence to show that such sacred journeys were once made in Wales. But the post-Reformation break-up of the cults of the saints in Wales has led to a de-sacralisation of the landscape, and few sacred landscapes can now be reconstructed with certainty. Many individual sites on these itineraries can no longer be identified, but as springs cannot easily be stopped-up, and as in any case they retained their utilitarian functions – drinking, bathing for cures – well into the modern age, most of the old holy wells are still remembered. But their place in the once-sacred landscape has been forgotten.   But at the time of the Anglo-Norman invasion, this was the ubiquitous practice of Welsh Christianity. The Normans introduced the newer patterns of church dedication, and in the twelfth century embarked on an ambitious programme of church building. In line with prevailing devotional patterns, large numbers of these new churches were dedicated to our Lady, such as that at Meifod. Devotion to her in Wales received a new impetus, but it is clear from innumerable placenames across Wales that this, along with the new churches dedicated to her, was swiftly incorporated within traditional Welsh devotional structures. Thus, in imitation of the older churches, many of these new foundations were called Llanfair, even though the eponym was not herself the actual founder of her llan, as had been the case with the earlier founders. Almost certainly it is to this period that we must look for the origins of most, if not all, wells bearing the name of the Virgin. Just as was the case with the older churches, dedicated to Celtic saints, these new churches observed the feastdays of their patron; and they observed them in the ways traditional to the country. But there were no topographical features in the local landscape to be visited in memory of a saint who had never set foot there, so the Welsh set about creating them. Some of these new landscapes could be surprisingly detailed; hand- and footprints were discovered on rocks, and a few locations developed legends of Mary herself having visited Wales (Jones 1954, pp.46-7 – it might be the case that numbers of these traditions retain the memory of some visionary experience, garbled by time?), but the most common element in the new Marian landscape was the well (the vision traditions just mentioned are usually associated with wells). In fact, so closely was a visit to a well sacred to a particular saint associated with the observance of her or his feastday, or with individual acts of devotion to the saint, that Mary Wells were created in parishes not specifically dedicated to her. Wells did not need to be immemoriably sacred, or associated in some particular way with their patron in historical time. They could simply be blessed in honour of the saint, a practice which has been continued in both Ireland and Brittany into modern times, when a newly-introduced devotion to a particular saint into a parish has been felt to require expression in this time-honoured way.

But this creation of new holy wells was not spontaneous, as had been the case with the gradually-perceived sacrality of the older wells of the Celtic saints. Created specifically in connection with the newly-awakened cult of the Virgin in the twelfth century, most of them appear to have remained little more than a formal adjunct of formal church ritual, which is why some two-thirds of the Mary Wells noticed by Jones now exhibit no evidence of cult beyond their mere names. As an example, the well at Llannefydd, in Denbighshire can be cited. Nefydd has never been satisfactorily identified [15], but he was clearly thought of as the founder-saint, as the toponym reveals. His well survives in a thicket some 300 yards North of the church (Jones 1954, p.175), and a little poem was written in its honour by a seventeenth-century rector. But at some time in the middle ages, possibly no earlier than the fifteenth century when the church was rebuilt, its dedication was changed from Nefydd to Mary. And closer to the church than Ffynnon Nefydd is her well, now a neglected spring in Gwyndy farmyard. No tradition of any form survives, and so far forgotten has it been that it remained unnoticed in print or manuscript until 1997 (see note 10). It is clear that any cult at this well had been formal and perfunctory, while the parish remained attached to its first holy well.

As Jones indicates, approximately one-third of Welsh Mary Wells exhibit some evidence of cult (not all of it satisfactory or conclusive). Some few were attached to Marian sanctuaries (the most famous is at Penrhys – see note 12), but at none of these was the well the principal object of devotion. Only a single well of Our Lady achieved status as an object of more than local veneration, in a sanctuary where it was the principal focus of devotion. This was the well-shrine known as Capel Ffynnon Fair, the Chapel of Mary’s Well, at Cefn Meiriadog, in Denbighshire. Even here, the evidence is frugal, and in the main can only be deduced from the ruins of the shrine itself. Not a single reference to it has been traced from the pre-Reformation period, and the first mention is the simple notation ‘Cap: Funhown vaier’ on Christopher Saxton’s 1579 map of Flintshire and Denbighshire. A rhyming calendar in Welsh, dated 1609, under 15 August, conveys the fact that the principal date for pilgrimages was the feast of the Assumption, and demonstrates that pilgrimages were still in full swing there, 70 years after the Reformation.

‘Gwyl fair gynta or kynhaia
ysbort a gair yn ffynhownfair'(quoted in Thomas 1908, p.395, note 1)

which may be translated as ‘[On] the first feast of Mary in harvest [the common Welsh designation of the Assumption feast]
“fun and games” [lit, sport and word] at Ffynnon Fair.’    It is noticed in the antiquarian literature fairly frequently from the time of Edward Lhuyd onwards (Lhuyd 1909, p.47), when it is said to be ‘quite ruinous’. Browne Willis’ description of 1720 has already been quoted (note 3).   It is the architecture of the shrine which tells of its one-time importance. It consisted of a well immediately outside the North wall of a chapel. The water was conducted underground to a bath in a corner of the chapel. The well-chamber was reconstructed in the late-fifteenth century, in a star shape closely resembling St Winifred’s Well at Holywell; and at the same time the chapel was doubled in size by the addition of a new chancel, to the South. The architectural splendour and sophistication of this work argues for a popularity and prosperity of the pilgrimages which took the well as their goal. But the original chapel, immediately beside the well, was a much simpler building, with cyclopean stonework around the doors, and arguably it dates to the twelfth century. This was the time of the consecration of Meifod, and the explosion of Mary churches and Mary wells across Wales. At that time, the well was in the parish of St Asaph. This little town housed the cathedral of the diocese, and it is possible that the consecration of the well to the Virgin and the building of a chapel to her represent the parochial or even the diocesan response to the upsurge of devotion to Mary in the twelfth century. Certainly the chapel had no function in the ordinary parochial structure, and has never been anything else than a purpose-built place of pilgrimage. There is but a single tenuous clue to its ultimate origins. Margaret Williams spent her girlhood, before World War I, at Dolbelidr, a house just across the river Elwy from Capel Ffynnon Fair, and remembered the place well. I interviewed her some years ago, shortly before her death, when she told me she remembered that people still bathed in Ffynnon Fair in the pre-War period, for the relief of rheumatism. But she told my friend Judy Corbett, whom she knew well, something more – she remembered a local tradition to the effect that sometime long ago there had been a vision of Our Lady at the well, and that the chapel had been built as a consequence of that vision. This tradition has never been recorded before, and I am grateful to Judy for passing it on to me, for record here.

This convoluted paper may be conveniently terminated at this point. Hopefully it will have shown, not just the available documentation for a single North Welsh holy well, now sadly no more; but also how a consideration of such documentation can usefully be used in a critique of a major work of scholarship, Francis Jones’ book The Holy Wells of Wales. It has not been my intent to demean this seminal work, but to begin the task of revalidating it for the future, by making readers aware of its limitations, and by taking a little further its considerations of such perennial themes as the survival of ‘paganism’ into Christianity, the problems presented by boundary wells, the point of origin of the significant group of Marian wells, and the perplexing subject of what exactly it is that makes a holy well holy – while also permitting myself to ride a number of favourite hobby-horses along the way!


Ffynnon Ddeier: legend

Ffynnon Ddeier: cultus



How do wells become holy?



Page designed by Rich Pederick (© Living Spring Journal, MM)
Written & maintained by Katy Jordan & Rich Pederick
Created  May 29, MM

%d bloggers like this: