The place-names Ystrad Alun, Yr Wyddgrug and Moldsdale

By Hywel Wyn Owen

This article considers the derivation and meaning of Ystrad, Alun, Yr Wyddgrug, Mold and Moldsdale. In discussing Yr Wyddgrug and Mold, I advance alternative interpretations to those hitherto fashionable.[1]


The meaning of Welsh strad or ystrad [2] is ‘valley bottom, wide valley, river-valley, river-meadow.’ The word is a distant relative of the Latin strata and the English street, all three denoting ‘flat’. In the Celtic languages it is found as Irish strath ‘river-meadow’, Breton strad ‘bottom, low place’, Cornish stras ‘flat valley, shallow valley, wide valley’ and Scottish strath ‘valley’- In fact, Scotland’s Allan Water in Perthshire was Alun and Strathalun in the twelfth-century, and is Strathallen today.[3] It was even adopted into English place-names, as strat as in Straddle in Herefordshire [4] (vallis Stradelei 1086, vallis de Strada 1169); there, dale, the second element, is a repetition of strat.

In Welsh, ystrad has a plural ystradau or stradau (recognisable as the Llanelli rugby-ground Strade or Stradey). The plural has the same connotation as the English valley ‘bottoms’. Therefore, ystrad referred not so much to the configuration of the valley (which in Ystrad Alun is very shallow) but rather to the lowlying land along the valley bottom, the wide meadows bordering the river. Several ystrad names contain river-names such as Ystradfyrnwy (Stradyvyrnwy 1578, the river Efyrnwy), Ystrad Fflur or Strata Florida (Stradflur 1181-2, Strata Florida. 1279-80, the nver Fflur ) and Ystradgynlais (Stradgenles 1372, the river Cynlais ) . However, ystrad is not exclusively associated with a river-name. Personal names also appear, indicating the owners or tenants of rich, prime land. Examples are Ystradelfedan near Welshpool (Stradelvedan 1232-3, the personal name Elfedan), Ystradfellte in Breconshire (Stratmelthin 1230, the personal nameMellte) and Ystradmynach in Glamorganshire (YstradManach 1636, mynach ‘monk’).

In Ystrad Alun the earlier strad is seen in Stradalun 1327-8, Stredalen 1536-9, and the later ystrad in Ystrad Alun 1435-70, Yslralalun c.1450and Ystrad Alyn c.1700. One poet, Tudur Penllyn( 1450-90), refers to it as dyffryn Alun (where dyffryn is ‘valley’).


River-names are amongst the oldest names we have. One reason is that frequently rivers were boundaries, and their names therefore well-established in usage and, fortunately, in documents. Another reason for their antiquity is that rivers were commonly accorded divine attributes and bear the names of gods and goddesses.

One of the old Celtic names was Alaunos, the name of a deity (possibly similar to the Roman Mercury) and a river-name. The meaning is unclear. Some scholars[5] believe it may contain the root al ‘wandering’ found also in the river-name Aled. A derivative of Alaunos was Alauna, the form in which various river-names appear on the Continent and in Britain. Modern forms of Alauna are Allonnes and Alaun (France), Allan (Scotland), Aln (Northumberland), Ellen (Cumberland), Allen (Cornwall), the Alun near St David’s (Pembrokeshire) and our Alun.[6]

Early records have Alun quite consistently from 1337 onwards with occasional variants such as Alen 1539, Alin 1672 and Allen 1695. The form Alyn appeared first (to my knowledge) c. 1700 but had no real currency until OS cartographers started using it between 1840 and 1871. Interestingly enough, the 1961 1/25,000 and the 1966 2 ” SJ 35 have River Alun, while the 1981 Pathfinder has River Alun (Afon Alun), and the 1983 Landranger has R.Alyn (Afon Alun). There can be no doubt that the ‘correct’ spelling is Alun.[7] The later emergence of Alyn can be ascribed to analogy with other Flintshire place-names ending in – yn (such as Mostyn, Prestatyn and Sychdyn). Alyn has also been perpetuated by the creation of the civic name Alyn and Deeside and the blend Delyn (influenced by telyn ‘harp’), and possibly by the perceived need to distinguish between Alyn the river and Alun the personal name (as in the bardic name assumed by John Blackwell 1797-1840, commemorated in the name of the school in Mold).

Yr Wyddgrug

The Welsh name Yr Wyddgrug appeared first (to my knowledge) in 1097 as Yr Wydgritc, and in its modern guise by the fourteenth century. There seems to have been a colloquial version (still heard locally) without the dd (as in wyrgrig c. 1460 and 1480). Edward Lhuyd notes this variant as Yr wrgryg (corrupie pro gwydhgrg) and yr wyrgryg (c. 1700), as well as the variant Gwydhgryg. [8]

There is another Gwyddgrug in Carmarthenshire; in Montgomeryshire are Cefn Gwyddgrug, Eskair Wyddgrig (1617) and castell yr Wydgruc (1263) now Nantcriba.[9] In our Yr Wyddgrug yr is the Welsh definite article (‘the’). After y or yr initial g w becomes w as in gwaun becoming waun, gwern becoming  wern, and so Gwyddgrug preceded by the definite article becomes Yr Wyddgrug. Why some place-names should have been intentionally distinguished by ‘the’ is not always clear, but it does seem to indicate a feature of local significance, such as Y Waun (Chirk) ‘the moor’ as opposed to Gwaun-y-terfyn (Acton) ‘moor at the border’.

The second element of Gwyddgrug and Yr Wyddgrug is crug ‘mound, knoll, hill; cairn, tumulus’. Elsewhere in Wales it appears in Crug Mawr (the prominent hill in Ceredigion), Bryncrug (Meirionethshire), Crucadarn (Breconshire), Crug-y-bar (Carmarthenshire) and in C r u g y w e l or Crickhowell (Breconshire) and Cricieth (Caernarfonshire). Its antiquity as a place-name element is evident from its appearance in England as Crich (Derbyshire), Creech (Dorset and Somerset), Crutch (Worcestershire) and Crook (Devon and Dorset).

The first element is a little more problematic, at least in its meaning. There can be little doubt that it is gwydd. The view hitherto held by all scholars is that gwydd here means ‘grave, burial mound, tumulus’. The same element, it is argued, can be seen in Yr Wyddfa (Snowdon) ‘the place of burial’ ( yr + gwydd + ma ); there is independent evidence from literature that gwyddfa means ‘grave’.[10] However, gwydd has an earlier meaning of ‘presence, sight, face’; the modern yng ngwydd ‘in the presence of’ still carries that meaning. [11]

I propose here that, in Yr Wyddgrug, gwydd referred, not to a grave, but to the open, prominent, highly visible hill which identifies Yr Wyddgrug with the Bailey Hill. There are four reasons for my proposal. One is that in GPC gwydd ‘presence, sight’ is recorded (independently of the place-name) from the twelfth century, much earlier than gwydd ‘grave’ recorded from the fourteenth century, and therefore accords much more closely with the first record of Yr Wyddgrug in 1097. The second reason is that it avoids the slightly odd repetition of ‘grave’ which is behind both gwydd ‘grave’ and crug ‘hill, cairn, tumulus’. The third reason is that the Norman-French Mont-hault ‘high hill’ (which later became Mold) can thus be interpreted simply as a loose translation of Yr Wyddgrug ‘the prominent hill or cairn’. The fourth reason is a negative one. It has been quite properly argued [12] that Bryn-yr-ellyllon in Mold has a far better recorded pedigree than Bailey Hill as a memorial mound or tumulus. Assigning the meaning ‘prominent’ to gwydd as opposed to ‘grave’ neatly side-steps the issue. In any case, building the castle ( kastell yr wydgrtic 14th cent.) may have obliterated any trace of a memorial mound or barrow, especially if it consisted of no more than a cairn of stones. The identification with Bailey Hill is the view held by most historians, local and national, from Edward Lhuyd onwards : ‘Mould is call’d y Wydhgrig i.e. Tumulus conspicuus….Ye first syllable in Wydhgrig signifies alius conspicuus.’ [13] It is significant that Lhuyd, too, gave that interpretation to gwydd.[14]

In summary, then, Yr Wyddgrug probably means ‘the prominent hill, cairn or tumulus’ and refers to today’s Bailey Hill, the dominant feature of the local landscape.


Many years ago, Ellis Davies [15] advanced the suggestion that the title de Monte Alto could have been a name which had been transferred to Mold from Normandy (as happened with Montgomery). [16] It is certainly an attractive possibility which cannot be discounted, but currently we lack corroborative evidence of a direct link with a Norman castle, town or village.[17] We must therefore accept for the time being that the place-name Mold refers to the same prominent topographic feature as the place-name Yr Wyddgrug, namely the Bailey Hill upon which the Norman castle was built.

A confusing characteristic of the early forms of Mold is that several appear in Latin or legal documents and so preserve a Latinized, grammatical, form (such as de Montealto 1151-81, Monte Alto 1240, Monlem Altum 1278).[18] We must assume that such grammatical forms had no existence outside the written documents, although, due to the conservative nature of legal documents, Latin forms did survive surprisingly well (such as Mons Altus 15th cent, Mounte Alto 1467-8, Mount Altye 1577 and Mont alto 1746).

The antecedent of Mold was the Norman-French mont-hault ‘high hill’. But from the beginning there were forms without the medial t (as in Mohault 1277 and Monhalt 1358) and without the final L (as in Muhaut 1241 and Mohaut 1297). The final t frequently appears as d (as in Midland 1241) which probably caused the L to reappear before d (as in de Moaldis 1220 and Moald 1284). There is ample evidence, too, that Mold was a two syllable place-name for a considerable period at time (as in the previous examples and in Mauliaud 1241, Moald 1310. Mohawt 1439, Moliaut 1498). Some earlier documents seem to have a single syllable (as in Mold 1297 and 1337, Molde 1341) but the general pattern reveals the modern form emerging with some consistency in the fifteenth and sixteenth centuries (as in Molde 1474, Moylde 1507-53, The Mold 1577, Mould 1581), but the variant spelling Mould continued quite late (as in Mould 1667 and 1811) no doubt influenced by the word mould. There were variant pronunciations current within the same period (as is revealed in Mohaut….otherwise Molde 1595). Such variants should not surprise us. We have two Norman-French words in everyday usage over eight centuries by English speakers and Welsh speakers, each subjecting the original name to the phonological changes which characterized English and Welsh.[19]

One of the difficulties in explaining Mold can be resolved by realising that we are not necessarily talking of a chronological progression from Mont-Hault to Mold. My view is that we can discern parallel forms existing side by side. From the examples cited above we can parallel de Moaldis 1220 and Mohault 1277, Moald 1241 and Muhaud 1241, Mold 1297 and Mohaut 1241. For me, the key evidence is Mohaut….otherwise Molde 1595. Two independent forms clearly existed as late as the sixteenth century. I suggest in this article that there were two ‘traditions’ : Moald/Mold which reflects a very rapid local naturalization of the Norman-French name, and Mouhall/Mohawt which represents a more conservative formal, literary, usage, perhaps confined to legal documents. There is really no need to explain the convergence of two conventions, because the Mouhalt/Mohawt pronunciation became obsolete. [20]


Dale is a common enough place-name element in the North of England, at least in those areas which were subject to Scandinavian influence. In Old English dsel meant ‘pit’ while the Old Norse word dalr meant ‘valley’. It is dalr which has influenced the meaning and distribution of dale ‘valley, vale’.[21] Place-name distribution of dale in England is north of the Mersey and Humber although some are found in Northamptonshire. Dale can be combined with a description of the character of the valley (as in Langdale ‘the long dale’); Professor Cameron cites examples with river-names (such as Wharfedale and Airedale) and buildings (such as Kirkdale) and Scandinavian personal names (as in Skelmersdale). An interesting comparison with Moldsdale is Borrowdale which contains a word for fortification.[22]

The forms for Moldsdale are similar to those for Mold. We find comparable Latin documentary forms (as in valle Montis Alti 1299, Valle montis altis 1331, vallis montis alti 1527). The same phonological pattern with loss of n and I (22] can be traced (as in Mohautisdale 1275 and Mohawtesdale 1439); some have d (as Mohaudesdale 1438) and a fair number of forms substitute n for l possibly under the influence of the wordmount (as Mohantisdale 1340, Mounisdale c. 1465, Mohuntsdale 1499, Mo’untzdale 1528 and Moliauniesdale 1533). There are several examples of the emerging Mold (as in Molysdale 1474, Mollesdale 1491, Moulsdale 1550, Mowlesdale 1547-8) and several which anticipate the modern form (as in Moldesdale 1314 and 1506, Moldysdaile 1509-53) but still seem to have three syllables. The modem form is first seen in Moldsdale 1515. The bilingual spirit is captured in Molesdale yn Welsch caullid Stredalen 1536-9.

E N D – N O T E S
  1. Within the discussion of each name you will find certain words in italics followed by dates. The convention in discussing place-names is to cite early forms of the place-name (in italics) and give the dates of these forms. Place-names change, sometimes quite dramatically, over a period of time and the further back we go the closer we get to the original meaning. Occasionally however, there comes a point beyond which we cannot go simply because the documentary evidence does not exist, or is ambiguous or is seemingly meaningless. Then we have to be judicious in our interpretation or frequently rely on similar place-names elsewhere in Britain. Forms and dates provide the evidence on which to base the explanation for a place-name. You will also find words in bold italics type; they are the elements which make up place-names. Place-names are in bold. Counties are referred to by their pre-1974 designations. I am grateful for having been granted access to the Melville Richards Archive at the University of Wales Bangor. Many of the historical forms are derived from this unpublished source material. Others are derived from publications by B G Charles, Ellis Davies, Richard Morgan and Ken Lloyd Gruffydd as well as my own research material. The latter has, over the years, been most generous in sharing with me placename data which he has gathered in the course of his own researches.
  1. The older form was strad but by the twelfth century in both native Welsh words and words adopted from English and Latin, y developed before an initial s followed by a consonant. Examples are ystrad, ysgol ‘school’ and ysbryd ‘spirit’; in these examples the additional first syllable was stressed. The Liber Landavensis c. 1150 has Strat, Istrat and Estrat. The best discussion of the development is in \VG,26 and 49. Suffice to say that in our Ystrad Alun we have early evidence of the older strad and the emerging ystrad, but the latter soon became the dominant and eventually the exclusive form.
  1. .SPN 186
  2. DEPN s.v. Straddle
  3. Such as Sir Ifor Williams ELI 47
  4. PNRB 243-7
  5. Scholars are agreed on that point, e.g. Sir lfor Williams ELI 47, Professor Kenneth Jackson LHEB 306, 309, Professor Melville Richards WAR/ 6, 206 and ETG 32, 253, Ellis Davies FPN 4, and the Board of Celtic Studies Gazetteer 3
  6. Paroch.i 78, 88, 89, 98
  7. castell yr Wydgruc was Gytne Gnug 1281; hitherto incorrectly identified with Yr Wyddgrug/Mold according to Richard Morgan in a personal note.
  1. Cited by Sir Ifor Williams EL/14
  2. GPC distinguishes between gwydd ‘presence’ and gwydd ?grave but suggests that the latter probably developed from the former.
  3. By Ken Lloyd Gruffydd in personal correspondence, and by Ellis Davies FPN 180.
  4. Cited by Nesta Lloyd in ‘The correspondence of Edward Lhuyd and Richard Mostyn’ FHS 25(1971-2),43. I am grateful to Ken Lloyd Gruffydd for drawing my attention to the reference.
  5. In a forthcoming publication on Montgomeryshire place-names, the same conclusions is being proposed independently by Richard Morgan.
  6. FPN 113
  7. Roger de Montgomery named his castle Montgomery after his castle in Montgommery in Normandy.
  8. Despite lengthy correspondence over the years with several personal-name experts such the late Cecily Clark, Dr George Redmonds, Peter McClure and Dr Keats-Rohan. All, however, admit the possibility. There is, for example, a Monthault in Ille-et-Vilaine in Normandy.
  9. Compare the Norman-French beau maris Tine sea-marsh’ becoming Beaumaris and Biwmares but appearing in Latin documents as de Bella Marisco in 1284.
  10. It is interesting to see the development elsewhere of the same Norman-French place-name. In Scotland for example we find Monlealt becoming Monaut and then the modern place-name Mowat. In Riddlesden, Yorkshire, the twelfth-century family name de Moliaut survives as Maude while in the Bradford-Leeds area we find Monalie alias Mawde 1542, MontealdaliasMawde 1555, and MawhautaliasMawde 1585. We cannot link the Yorkshire family with the Flintshire Mont-liault. I am grateful to Ronald Black for the information on Mowat and to Dr George Redmonds for the note on the Yorkshire surnames. The influence of Welsh speech on English place-names in the area has been exemplified thoroughly in Hywel Wyn Owen, “English place-names and Welsh stress-patterns\Nomina XI (1987), 99-1.14,
  1. A similar example is to be found in G o l f t y n ‘Wulfwine’s farm’ which in 1585 was Gol/ton alias 12 Woo/fingion.
  2. For a further discussion see EPN 189.
  3. The n occurs too consistently for us to assume u miscopied as n.

in the order in which the references appear in the end-notes.

WG … J Morris Jones, A Welsh Grammar (Oxford 193 1)

SPN … WFH Nicolaisen, Scottish Place-Naines (London 1976)

DEPN … E Ekwall, The Concise Oxford Dictionary of English Place-Names (Oxford 1960)

ELI …   I Williams, EnwauUeoedd (Liverpool 1945, 1969)

PNRB … A L Rivet and C Smilh,772e Place-Names of Roman Britain (London 1979)

IJTEB …   K H Jackson, Language and History in Early Britain (Edingurgh 1953)

WATU … M Richards, Welsh Administrative and Territorial Units (Cardiff 1969)

ETG …   M Richards, Enwau Tir a Gwlad, ed. B L Jones (Caernarfon 1998)

FPN   …   Ellis Davies, Flintshire Place-names (Cardiff 1959)

Gazetteer … E Davies (ed.), A Gazetteer of Welsh Place-names (Cardiff 1967)

Paroch …   Edward Lhuyd c. 1700, Parochialia’m Arcliaeologia Cainbrensis (London 1909, 1910, 1911)

GPC … Geiriadur Prifysgol Cymru (Caerdydd 1950- )

FHS   …   Journal/Publication of the Flintshire Historical Society (1911 – )

Nomina   … Journal of the Society of Name Studies in Britain and Ireland 1977- )

EPN … K Cameron, English Place-names (London 1996)



Copyright o f articles published in Ystrad Alun lies with the

Mold Civic Society and individual contributors.

Contents and opinions expressed therein remains the

responsibility of individual authors.