By Hywel Wyn Owen

The second article on place-names in Ystrad Alun deals with the derivation and meaning of the names of the townships [1 ] in Ystrad Alun [ 2 ] . The entry for each township consists of some historical forms ( in italics ) with dates [3] selected to demonstrate the major phonological features, the elements ( in bold italics) which comprise the etymology, and a discussion of the linguistic and topographical significance [4]. Other place-names are in bold.

Ardinwenl 1363, Arddunwent c.1500, 1545-6, Arddynwenl 1516, Arthunwenl 1544-5, Arlhinwenl 1547 The earlier forms have the Welsh voiced -dd- ( as in ‘farthing’ ) with ample evidence of the spelling substitution of -th-.
Ifor Williams [5] suggested gwent ‘maes’, but gwent does not appear as a noun in GPC or in GBGG (except as Gwent ). However, its existence in compounds such as cad went ‘battle’ ( from cad + gwent ) is evidence of its survival. But it is by no means common, and its late appearance is unhelpful. Currently no other element offers itself in the face of unequivocal and consistent spellings.

Ifor Williams and Melville Richards were certain that the first element is the feminine personal name Arddun [6] Ken Lloyd Gruffydd [7], D.Machreth Ellis [8] and Melville Richards [9] compare with Bryn Arddun (Caernarfonshire) and with D&l Arddun (Montgomeryshire). The name phrases Bryn Arddun and DSl Arddun pose no challenges; this is the expected word order for the personal name. Arddunwent, on the other hand, is not.
The precise significance of the origin of gwent need not concern us unduly here [10]. The argument hinges on the fact that several Roman settlements contained the element venta ( such as Venta Silurum later Gwent ( Chepstow ) , and Venta Belgarum later Winchester), while gwent also survived as a word in Welsh ( as in cadwent ‘battle’) and llin went ‘field for flax growing’ [1 ] ] . Its existence in cadwent from the 13th century suggests gwent was indeed a living element in the Middle Ages and that its meaning was something
like ‘field’ [12].

There is no evidence of arddun ‘fair, beautiful’ as an independent adjective before the late 18th century. So ‘fair field’, while feasible, is unlikely. The personal name Arddun, derived from it, is a safer candidate [13].

Argoyt 1477, Argoed 1539
The meaning of A r g o e d is ‘edge or border o f a forest’, or even ‘circle or enclosure of trees’. The name occurs several times in Wales [14] which is evidence of argoed being a place-name element in its own right [15]. Several of the literary contexts in which argoed appears ( from the 13th century onwards ) refer to being ‘ yn argoed ‘ ( ‘ in argoed’ ); one describes adog ‘mewn argoed ir ‘ ( ‘in a verdant argoed‘) [16J. Therefore argoed
seems to have signified a woodland, or a glade on the edge of a wood or forest. The settlement at Argoed was therefore neither opposite a forest, looking towards it, nor within the forest itself. The earliest record is fairly late (1477) but the wood or forest was old, probably the extensive forest called at various times Swerdwood [17] and later Ewloe Wood or Hawarden Wood. Buckley was one o f the first places to be cleared for agriculture or hunting.

Brakhaynyld 1477, Braclianyld 1477, Brachinillt 1580, Brachanillt 1622
Bruchanelt 1567, Briclunallt 1605, Brychanylll 1618
Brechamlt 1593, Brechanilll 1658
Byrchennall 1553, Byrchanylhd c.1700, Berchanillt 1683

This place-name has undergone several phonological changes, camouflaging its origin in the hillside now known by the name Bryn Coch. That land, we can conjecture, was referred to by the Welsh word brecan or brycan [18] ‘blanket, canopy’ possibly a reference to a canopy of trees. To this was added the English word hill [19]. The phrase *brycan-hill became *brycanill in which the stress shifted to the middle syllable [20] and the last syllable changed to – illt [21]. The varieties of spelling in the unaccented first syllable reflect the subsequent variants brecan and brycan with the additional variant o f -a-. They also show the influence of another (related) word brych with its variant brech [22] ‘speckled, variegated brown, freckled’. The unaccented first syllable also underwent what is called metathesis whereby -r- changes position, accounting for the spellings in Byrch– and Berch– and the modern Birchenald [23]. The final syllable, in the process of cymricizing English hill to – illt ( as in Gwersyllt ) , shows several intervening stages of -ild, -ilt, -elt, -aid, -alld and, predictably, the influence of the Welsh word allt ‘ hill ‘.
The whole explanation, while phonologically tenable, is still tentative because all the sound changes described above must be assumed to have occurred before the first record of the name in 1477 [24].

Biscopestreu 1086, Bissopestred 1093, 1151, Bistre 1533, Bistree 1593, Bystry 1603-4
Kares Esgob c.1550, Croesesgob 1602, Bistree Brit. Kroes esgob c. 1700 The two elements are the Old English personal name Biscop and treow ‘tree’. It is fairly common to find this type of combination [25] and one of the characteristics of such place-names is that they are boundaries, with the tree as a boundary marker with the named person living close to, or somehow associated with, the strategic location [26].

An alternative biscop ‘bishop’, although perfectly possible, has hitherto been considered less likely simply due to lack of corroborative evidence. Further research may throw a different light on ecclesiastical association [27]. However, trees in Old and Middle English place-names could be interpreted in the sense of a ‘Christian cross’ [28] and a boundary marker was associated with a place for meeting or for preaching. A well documented example is OswestryOswald’s tree’. Oswald was a common Old English personal name but when Oswestry’s church was dedicated St Oswald was an obvious choice, reinforcing the idea of ‘tree’ as a ‘Christian cross’ [29]. Bistre presents a very close parallel with Oswestry/Croesoswallt. In the Welsh form the personal name Biscop was interpreted as the ecclesiastical bishop, esgob, and ‘tree’ was replaced by croes ‘cross’. However, in
Oswestry, the two forms co-existed from very early on ( Oswaklestr’ c. 1180, Croesoswaki 1254). In Bistre, the Welsh interpretation does not appear until c.1550 by which time the elements in Bistre would be unrecognizable. We may conclude that Croesesgob was not an antiquarian coinage but was translated when ‘bishop’s tree’ were recognizable elements, probably in the 11th or 12th centuries. We can further assume that the Welsh form was not recorded in official documents in the interim, but was preserved in oral tradition.
However, it did not survive in modern usage. The phonological change from Bisceopestreu to Bistre can be explained by the process observed in cymricized place-names resulting in the loss of the middle syllable (called syncope) [30]. The Welsh influence has also
modified the spelling of Bistre to reflect tre ‘farm’, whereas the pronunciation of Bistre conforms to the regular development seen in Braintree , Elstree, Coventry and Oswestry.

Bruncol 1086, 1281, Brunkeed 1593
Fronkoed 1435, Trefronkoed c. 1500, Broncoil 1539, Broncoed 1600
Bryncoed 1675, Brynkoed c. 1700
The 1086 form appears to be the Old English brun ‘brown, dark’ [31] and cot ‘cottage, shelter’ [32]. Alternatively, the first element could possibly be the Old Norse brun ‘brow, the edge of a hill ‘ [33] since briln and brun are frequently indistinguishable [34]. However, the likelihood of this being an Old Norse name here in Moldsdale is very remote indeed. It must also be conceded that ‘shelter on the edge of a hill ‘ is not topographically appropriate. ‘Brown shelter’ challenges the imagination, especially since it does not occur
elsewhere in England or Wales; perhaps it was ‘muddy’ or ‘dark, shaded’ . Given the notoriously unreliable Domesday Book spellings and the possibility that it was Welsh all along ( as bryn ‘ hill ‘ + coed ‘trees’ ) [35], we must concede that Bruncol defies a wholly satisfactory etymological explanation.
No wonder, then, that with the passage of time Bruncol underwent several changes, reflecting Welsh influence. There are several other words with Old English -u- which became Welsh -u- or -y- [36] so this Brun- fell under the influence of bryn ‘hill’ ( despite it being topographically inappropriate ) . Perversely, the evidence also points to an earlier development to bron ‘hillside’ which has survived.

Gwernaffield / Y Waun
Gwernaffeld 1477, Gwemaffild 1486, Gwerneaffeyld 1526, Gwernafelde 1539, Gwernaffield 1598, yrwyen 1600, K VKKM c.1700, FAe Wfayn 1719, Wain 1726, Waen 1794
The meaning of gwern is ‘alder- tree’, ‘alder-grove’, ‘damp meadow’, ‘wet ground’. There are numerous gwern field-names in the near vicinity, with Gwernymynydd and Gwernaffield surviving as the names of settlements. Gwernymynydd was the ‘mountain gwern’; Gwernaffield was the ‘field gwern‘. The use of the English word field [37] is very difficult to explain, especially as there is no record of field being borrowed into Welsh as * ffild , and the numerous gwern names in the area have a Welsh rather than an English element following gwern . We can only draw a parallel with other English agricultural terms which characterised the spread of farming practice ( sometimes under the direction of an English-run estate ) such as croft/grofft , acre/ acer , close/ clos, stank/stang, sling/Sling , plot/Sblot. It would appear that in this one instance English field > Welsh ffild, although what 15th century agricultural practice distinguished this field from every other cae or maes in the area we cannot tell. Perhaps it was fairly early enclosure for arable purposes possibly encouraged by the Rhual or Gwysaney estates. Whatever the reason, the English field was a feature of the landscape and the adjacent gwern was known as ‘Gwern-y-field’ which was cymricized further to ‘Gwernyfield’ with the stress moving to the penultimate syllable despite it being the insignificant definite article [38]. This medial stress resulted in -y- becoming -a-. The element field itself became even more of a hybrid with the Welsh -ff- followed by an English -ie-.

This explanation is plausible but could do with corroborative evidence. No other ffild exists in a Welsh placename, to my knowledge. No other stressed medial -y- changes to the consistent -a- found in Gwemaffield’s historical spellings. There is no evidence of an agricultural field project in the area. If a personal name such as Aphylt or * Affield came to light, it would provide a far more comfortable etymology [39].
Fortunately, the Welsh form Waun (from gwaun ‘moor or meadow’ preceded by the definite article y ) is less contentious, and waen is a frequent spelling variant of waun in Welsh. However, the Welsh form as an alternative to Gwernaffield appears to be fairly modern usage, possibly adopted from Waen Farm and the chapel ( and today the school) [40].

Quisnan 1086, Wissanei 1317, Wyssaney 1330, Gwysaney 1363,c. 1700 Gwysane 1450, Gwisane 1465, Gwesane 1498, Wisanna 1544, Gwizene 1555, Gwysaneu 1567, Guissany 1653, Gwissaney 1656
To understand the etymology of Gwysaney it is essential to start with the name of a brook which flows through the Sychdyn ( eastern ) side of the Gwysaney estate. Today, this brook is called Afon Ddu [41] or Black Brook but I propose that over ten centuries ago this was called *Gwys ‘a sow; a pig’ [42]. Names to describe rivers which burrow through the land can be paralleled elsewhere [43].

The second element in Gwysaney is the personal Aneu [44]. The identity of Aneu cannot be ascertained; but we may postulate the river Gwys being linked with an Aneu who resided nearby, possibly in a fairly dominant position, probably the site of the present Gwysaney Hall. Perhaps Gwysaney, ‘the Gwys of Aneu‘, [45] became the name of this section of the Gwys brook. In time, Gwysaney came to be associated with the house [46], and the brook’s name Gwysaney was gradually displaced by Afon Ddu.

Hartsheath / Hersedd
Hartsheth c. 1500, Hartesheath 1601-2, Hartsheath 1653
Herteshethe 1578-9, Hertsheath 1.586
Heartesheathe 1612-13, Heartshealh 1619
Herseth 1546, 1557-8, 1653, Hersedh c.1700

The meaning is ‘hart’s heath’ ( Old English heorot ‘hart’ and hxd ‘heath’ ). The subsequent phonological development is the result of two separate ‘Welsh’ and ‘English’ developments.
The English development is fairly regular, provided we remember that the Old English heorot was still written and pronounced hert in the late Middle Ages [47]. This explains some of the Herte– spellings above, and the later spellings in Heart- . We must also remember that heath was pronounced and written heth [48] which explains the spellings in -hetli(e) . The 15th and 16th century English pronunciation ‘Hertsheth’ was consistent
with the ‘Welsh’ ( or cymricized ) form which is a regular adaptation as far as the vowels are concerned. For some reason the consonantal cluster o f -rts- resulted in the loss of the medial -t- [49] producing the Herseth forms above. The change from Herseth to Hersedd [50] ( which Lhuyd notes as Hersedh c.1700 ) is probably by analogy with other Welsh words ending in -edd ( usually a suffix denoting abstract nouns such as mawredd
‘greatness’ or plurals such as gwragedd ‘women’).

Hendre Biffa
Hendre Bifau 1086, Hendrebiffa 1331, Hendrebyffa 1544-5, hendrebipliau 1681
In Welsh place-names hendre(f) means ‘winter dwelling located in the valley’, a permanent residence as opposed to the hafod ,’the summer dwelling’ [51]. There are several in the area such as Hendre Faelor ( in Bodfari ), Hendre Gaerwys ( in Caerwys ) , Hendre Mynach ( or Hendre Bach ) near Cwm [52]. Some are, not unexpectedly, combined with personal names such as Hendre Figillt ( the personal name Bugil ) now more commonly Hendre ( on the Denbigh road ). Hendre Biffa falls into the latter category, although the identity of Piffa is unknown.

Leeswood / Coed-llai
Liegge 1086, Leygis Wode 1314, Leys Wode 1368, Lesewaod \417, Lysewood 1547, Liesewood 1566, Leeswood 1595

Koetyllai 1337, Coet y llai 1397, Coedllai 1490, KoedyLlai c.l500, Coed v Llay 1567′, Coed Llai 1773 The Domesday Book (1086). Legge [53] is the Old English leage ‘at the clearing*, from the Old English leak ‘clearing’, a very common element [54] in the place-names of England ( usually as Lea, Lee or Leigh or as ley at the end of place-names ). The description Legge usually denotes a’settlement at the clearing’, but recent research [55] indicates that leah and its derivative leage can refer to a settlement at or near to pasture or a meadow, especially near a river ( as the Denbighshire and Flintshire leah names suggest ) [56]. The wood referred to as Leygis Wode ( later Leeswood and Coed – llai ) was perhaps the original wood surrounding the cleared area adjacent to the river AIun.

The phonological development [57] of Legge to Welsh ( Coed – ) llai can be traced fairly easily if we remember that the Ley spelling above were cymricized to *llei and then to Llai . The cymricization was reinforced by the erroneous perception that the llai in Coed – llai was the adjective llai ‘less’, mirroring the English belief that Leeswood, too, was ‘Less-wood’ ( as in Lessewood 1566 ).

Llwynegrvn 1323

The second element is the personal name Egryn . His identity is lost to us but the same name appears as the name of the saint in Llanegryn ( Merioneth ) [58],
The first element llwyn can mean bush or hedge but was also used of a small wood, copse or grove. Llwyn is fairly common in place-names [59] and is frequently combined with a personal name [60]. The conclusion to be drawn is that llwyn in association with a named individual signifies a ( possibly privately owned ) grove that was managed for a particular purpose, probably coppicing [61].

Mold / Yr Wyddgrug
Both place-names have been discussed in detail in Ystrad Alun 1 (2000), 9-11.

Nerchgwys 1291, Nerchgoys 1398, Nercliois 1477
Nerthgoys 1397-8, Nerth Gwys 1490, Nerthkwys 1492
Nercwys 1492, Nerquis 1533, Nerquis als Nerthkwys 1634

Melville Richards thought the elements might be nerth ‘strength’ and cwys ‘furrow’ [62] but conceded that it made little sense, unless it referred to land which was difficult to plough. The grammatical formation ( of two nouns ) also sits uneasy. A possible explanation, however, is that Nercwys comprises the three elements yn ‘ in ‘ + erch ‘speckled’ (as in Nannerch )+ cwys. A hypothetical * Erchgwys ‘dark furrow’could have been the name of one of the tributaries of the Terrig which rises on Nercwys mountain. It is even possible that * Erchgwys was the name of that stretch of the river Terng [63] which is very close to Plas Nercwys. If * Erchgwys then came to be the name of a location ( perhaps Plas Nercwys or Nercwys itself ) , then the first element yn poses far less of a problem. It is not unknown for the preposition yn ‘ i n ‘ to become attached to the following placename ( as happened in ‘ yn Arberth’ becoming Narberth [64] ). Hence, yn *Erchgwys ( ‘in Erchgwys’ ) became Nerchgwys and eventually Nercwys.

Huall 1493, Yr Hual 1527, 1550
Y Rhyall 1621, Yr Rhual c.1900, Yr Rhyal c.1700
Riiall 1539, Rual 1644
Rhyallt 1597

The historical forms quite clearly point to two elements, namely yr ‘the’ and hual ‘fetter, shackle’, (figuratively) ‘restraint, hinderance, impediment’. The phonological process of metanalysis explains the linking, of yr with the following word ( as in yr hewl > Rhewl, yr hill > Rhyl , yr achub > Rachub ). There is evidence that yr was retained quite late ( as in Y Rhyall , Y Rhual ). The difficulty, however, is in interpreting hual . What was the fetter? Was there a perceived ‘ hinderance’ here? There are at least two other examples of hual in ( Rhos) Bodrual and ( the plural ) Bodruala ( both in Caernarfonshire ) but scholars are none the wiser. All three are near rivers. Was hual a dam of some kind ( as in the weir west of Rhual on the river Alun ) and the weir adjacent to Bodrual on Afon Saint near Caernarfon? Did hual signify sheepfold or pound?
The meaning was also evidently lost to earlier locals as suggested by the spellings of Rhyallt 1597-8 and Rhuallt 1689, influenced, probably, by the hill (allt ) down to Mold or the steep hill down to the river Alun, and also perhaps by Rhual It near St Asaph.

Trefihyn 1275, Treffdyn 1372, Trefdyn 1382, trefddyn 1478-9
Trewethwyn , Trewenwauns 1286
Trythyn 1539, Treythyn 1565, Treyddyn 1590, Treuddyn 1664, Tredhyn, Treudhyn c.1700

The two elements in Treuddyn are tref and ddyn(n) . The meaning of tref has evolved in Welsh from ‘farm’ to ‘extended farm’ to ‘hamlet’ to today’s ‘town’. The earliest historical forms record tref ‘farm’ quite distinctly. The meaning of dyn(n) is less precise, partly because it is limited to a few words such as murddyn originally ‘defence’ and later ‘ruin’ and to such place-names as Creuddyn (where both elements in the name seem to mean
‘defence’ )[65]. The generally held view is that there are two loosely associated meanings for dyn(n) , namely ‘ hill , height’ and ‘fortification’ [66] although the fortification was probably no more than a fence or thorn-hedge as a protection from animals [67]. Certainly, ‘homestead’ would be a satisfactory meaning for Treuddyn [68]. The earliest forms are attempts to render ‘Trefddyn Fawr’ and ‘Trefddyn Fychan’ since the township was divided into two, a larger and a smaller part ( as in Lhuyd’s Y Treuddyn mav/r and Y Treuddyn bach c. 1700 ). The modern anglicized pronunciation of ‘Treiddyn’ can be ascribed to Welsh -eu- > -ei- ; some of the historical forms suggest that this parallel pronunciation has been around for several centuries [69].


  1. The townships have not always remained consistent in number or extent, I am grateful to Ken Lloyd Gruffydd for the map which locates the townships and for his invaluable comments on the draft of the article.
  2. Several of the township names have been discussed by me in PNDA.
  3. The historical forms are extracted from research by Ken Lloyd Gruffydd, Mary Read, B.G.Charles, Ellis Davies and from my own researches. Special mention must be made of the Melville Richards Archive (AMR) in the Place-name Research Centre at the University of Wales Bangor. To have included references to the documentary sources of the historical forms was thought to be too cumbersome for the purposes o f this publication. The reader is welcome to contact me for references ( at the Place-name Research Centre ).
  4. A fuller explanation of the methodology of place-name research is to be found in Ystrad Alun 1(2000), 12.
  5. III. 12.
  6. ibid, and in AMR.
  7. In a personal note.
  8. In his annotated copy of ELI (in my possession).
  9. AMR.
  10. They are summarised in Enwau, 7-8.
  11. ELI, 42.
  12. Although T.S.O’Maille has argued that its original meaning could have been ‘river’ ( in ‘Venta, Gwenta, Finn, Guen’ in Nomina xi (1987), 145-51.
  13. See n.6.
  14. Argoed in Bedwellte (Gwent), in Nevern (Pembrokeshire), in Caron Isclawdd (Cardiganshire), and in Overton (Flintshire). In England, there is an Argoed in Kinnerley (Shropshire), an Argoedd in Cuddington (Cheshire) and an Argoed in the ‘Old North’ of Strathclyde.
  15. The element argoed is made up of the prefix ar – ‘adjacent, opposite’ (as in Arfon ‘opposite Mon’) and coed ‘trees’.
  16. GPC s.v. argoed
  17. Swerdewod 1278.
  18. GPC s.v. brycan , brecan .
  19. This hybrid type ( of Welsh and English ) is to be found elsewhere, as in Gwernaffield
  20. Stress-shift in place-names has been investigated and described by me in WSP. This demonstrates that the attraction of stress to the Welsh penultimate syllable transforms the phonological character of an English place-name. Thus, apparently ‘Welsh’ place-names such as Erddig, Gwesbyr, Bagillt, Golftyn and Gwersyllt have been shown to be English place-names in origin. So brycan-hill, accented on hill, can be compared with presta-tun resulting in Prestatyn; a similar process occurred in Brogyntyn and Sellatyn (both in Shropshire) and Carlatton (Cumbria).
  21. Compare Coleshill/ Cwnsyllt (from the Old English personal name Col1(1) + hill) with Kwnsalll 1543 showing the influence of Welsh allt ‘ hill ‘, and Gwersyllt( from the genitive of an Old English personal name to give Wersiges + hill ) as Wersuld 1315, Wersult 1315 and Gwersild 1393.
  22. GPC s.v. brych , brech (2). Compare Brechfa (Carmarthenshire) which is ascribed to the varied nature of the soil or land ( £ 7 G , 140). The same characterstic is seen in Brithdir (Merioneth, Glamorgan and Montgomeryshire) ‘speckled land’ and Brithdir Bach / Mawr on the eastern slopes of Moel Fama. In Mold it is worth noting the modern Bryncoch ‘red hill ‘.
  23. Compare Burntwood and Brentwood .
  24. It is tempting to see the entire place-name as being English as if it were broken-hill as in the fieldname le Brokynhulle 1425 in Surrey meaning ‘undulating hill ‘ ( VEPN , 41..) and the field-name Brokenhyl 1293 in Huntingdonshire (ibid. ) However, the varieties of -o-, -a-, -y- and -e- in the first syllable would be far more difficult to explain, as would the predominant -a- in the second syllable.
  25. Margaret Gelling states that 38 examples of this type are documented as well as ‘an uncounted number in the boundary clauses of Anglo-Saxon charters’ (PNSa i, 231 ).
  26. For a detailed discussion of the historical Bistre , see Ken Lloyd Gruffydd, ‘Bistre in the Domesday Book: some observations’, Buckley 19 (Spring 1995), 3-12.
  27. For example, D.Pratt has suggested that there may have been a Norman abbey on the site of the nearby Spon Chapel ( in ‘One of our abbeys is missing’, Clwyd Historian XXVI (1991), 30-4 and in ‘The Augustinian Priory of St Thomas the Martyr, Spon’ Buckley, 17 (Spring 1993), 1-12). That in turn might have been the location of some bishop-associated land before 1086.
  28. loc. cit. Of course, crosses (usually wooden) continued to be associated with boundary markers, frequently at cross-roads and with personal names attached, e.g. ‘Kroes wion ydyw’r tervyn rh. y Plwy ymma a Chaerwys’ (‘Kroes Wion is the boundary between this parish {Bodfari} and Caerwys’ (c.Paroch. i, 69.) In Flintshire the other cross whose antiquity can be compared to Oswestry and Bistre is Atiscross/ Croes-Ati ( Atiscros 1086 ) in Flint.
  29. See PNSa i, 231
  30. Compare Kelsterton / Kelstryn , and Wolfynton 1283 becoming Golftyn . The process of ‘naturalization’ of place-names of English origin which I term ‘cymricization’ is discussed in WSP .
  31. Compare a field in Littleton (Cheshire) called Brunlawejuriang ‘furlong at the brown hillock’ PNCh, 4, 115.
  32. Compare Mancot ( Maiiecote 1284, PNEF, 85 ).
  33. Compare Burn Toft in Durham ( Bruntoft 1188-96, VEPN, 49 ).
  34. The difficulties of distinguishing between brun and brun are noted in PNCh 5 i , 119., and in VEPN , 48-9.
  35. The present explanation supersedes my previous comments (PNDA , 15-16.) which favoured bryn + coed for Bruncot.
  36. Compare Old English tun ‘farm’ becoming Welsh -tyn ( in M o s t y n , Sychdyn, etc., cited in WSP , 103), and Old English pund ‘pound’ becoming Welsh punt and Old English slir ‘sour’ becoming Welsh sur ( EEW, 30 ).
  37. The meaning from the 10th century seems to have been ‘communally-cultivated arable’ land as part of the open-field system (LPN , 270-1 ).
  38. This is a well documented phonological process by which pen-y-berth > Penyberth , tref-y-clawdd Trefyclo ( Knighton ) , llan-y-cil > Llanycil, and pen-y-goes > Penegoes.
  39. J.E.Lloyd, for example, suggested an unattested *Hatfield ( in ‘Flintshire Notes : Hint and Mold’, Archaeologia Cambrensis (1940), 60 ) but the historical forms provide no evidence to support this.
  40. The two standard reference gazetteers of Welsh place-names, REL (1957) and WATU (1969), do not record Y Waun ; neither does FPN (1959). However, in AMR there exists a handwritten letter dated 20 October 1972 that testifies to Y Waen being the name given by ‘the Authorities to a small road of bungalows in Gwernaffield’.
  41. yravon dduy 1563
  42. Independent evidence for the existence of such a name is Edward del Gwysse 1390, and Trevays y Gwys 14th Century. I am grateful to Ken Lloyd Gruffydd for these references.
  43. GBGG s.v. gwys states that gwys originally meant ‘sow’ as do discussions in BBCS iv,127andii, GPC s.v. gwys also confirms ‘sow’ as the original meaning. Interestingly enough, Gwysaney, in its form Quisnan 1086, antedates these dictionaries which have no record of gwys before the 12- 13th centuries. According to Ellis Davies (FPN, 29), Gwys was the former name of the river Galchog to the east of Caerwys. There is a Gwys ( near Ystalafera) which flows into the river Twrch ‘boar’. Other examples are Hwch ‘sow’ (Caernarfonshire) and Banw ‘piglet’ ( Montgomeryshire) where Banw and another Twrch flow into each other. We can compare Gwystre (Radnorshire) probably gwys + tref ‘pig farm’ ( RadPN, 60 ).
  44. This rare name is also found i n Coedana in Anglesey ( coet aneu , a reference to the battle of 1194), now the name of the parish. There is also some doubt as to whether Aneu was ‘a female saint’ (ELM , 114.) or a man (ETG , 238). Interestingly, Dorothy Sylvester (RLWB , 467) notes that the Domesday Book mentions a priest at Quisnan.
  45. The -ey of the final syllable in most of the forms shows the similar development of -e / -au found in Plasau / Plasey, Parciau / Parkey, Ponciau / Ponkey found near Wrexham (CELW , 1) where it is shown to be an English development of the dialectal Welsh plural -e ( rather than -au ) . In Gwysaney the spellings support -aneu > -ane > -aney.
  46. Ellis Davies proposed gwas ‘abode, mansion’ as the first element on the basis of a personal name Gwassane in Colwyn ( FPN s.v. Gwysaney ). This suggestion, published in 1959, was taken up by GPC part l x v published in 1973 s.v. gwas (2), possibly on the basis of Ellis Davies’s view that ‘gwas Aneu’ might possibly be the origin. However, the historical forms cannot support such a derivation
  47. OED s.v. hart ‘slag’ has liert(e) as a spelling until at least the 16lh century. In place-names we may compare with Hartford (Cheshire) which was Hertford alias Herford 1581 ( PNCh 3, 188).
  48. OED s.v. heath has hethfe) as a spelling between the 15th and 16th centuries. In place-names compare Horseheath (Cambridgeshire) which was Horseth 1245 (CDEPN ), Heath (Shropshire) which was Hethe 1237 (ibid.), Heath (Cheshire) which was le hethe 1412 ( PNCh 4, 243 ), and Cheshire’s unidentified Blakheth 1478 ( PNCh 3, 188) and Walsnemansheth 1507 ( PNCh 2, 179 ).
  49. The same consonantal modification can be seen in Horsley (Bank) (Cheshire) where Hertislip 1413 (‘the hart’s leap’) becomes Horslyppe 1470 and Horseleys 1583. It is interesting to note that one of the forms for our Hartsheath is Horsfieath 1577.
  50. The Welsh form is scarcely documented after c. 1700 (according to Ken Lloyd Gruffydd).
  51. GPC s.v. hendref which denves it from hen ‘old’ and tref ‘farm’. The Domesday Book form Hendre Bifau 1086 precedes GPC s first record in the 12th century.
  52. See f T N , 86 lor further information on these hendref place-names.
  53. I am grateful to Ken Lloyd Gruffydd for first drawing my attention to the identification of Legge 1086 with Leeswood.
  54. EPN has 147 examples in Derbyshire alone, 42 of them as parish names.
  55. LPN , 237-9 and CELW , 2-6.
  56. Compare Llai/ Llay (Wrexham). In CELW it is argued that llai was a living word in this area for a ‘(settlement in a) river-pasture’ but that it did not survive and did not find its way into GPC. On later attempts to make sense of an obsolete word, see PNDA , 23.
  57. Detailed analysis and documentation is provided in CELW , 2-6 and related footnotes.
  58. LBS , i i , 415.
  59. REL has 16 llwyn place-names, and WATU (which includes some place-names no longer current) has 24 such as Llwyneliddon (Glamorgan), Llwyndafydd (Ceredigion), Llwyngwril and Llwyngriff’ri (Merioneth), Llwyndynwal (Caernarfonshire).
  60. In FPN Ellis Davies lists Llwynerddyn (Ysgeifiog) probably Erfyn , Llwyn Ifor (Whitford) and Llwyn Owen (Caergwrle); PNEF. 271 has llwyn mab hova 1372 (Rhanberfedd) ‘grove of the son of Hofa’.
  61. This interpretation is in accord with the interpretation of grove in English place-names. Margaret Gelling and Ann Cole (following Peter Kitson) have shown that a grove was a fairly limited woodland, managed probably as a coppice, and surrounded by a boundary ditch (LPN , 226-30). In England, too, grove is frequently found with a personal name.
  62. EIG , 248-9 and a note in AMR. He was perhaps influenced by a similar Nerthgwys in Llanddyfnan, Anglesey ( Nerthgwys 1413). The number of spellings with -th- suggests that nerth did make some sense.
  63. Terrig is first recorded in 1492 as terrygebrooke .
  64. Other examples of metanalysis are Sain Tathan alongside St Athan, and yr hill becoming RhyI, and yr hual becoming Rhual . In England, attenash ‘at the ash tree’ became Nash.
  65. So ETG , 89, whereas EIJ ,78 believes creu can mean ‘pig sty’.
  66. GPC s.v. dyn(n)
  67. ETC , 88. ETC , 88.
  68. This is the meaning offered by Melville Richards (AMR) and Ifor Williams (ELI , 78 ).
  69. Such as Trythyn 1538, Trythin 1746.

Abbreviations and Bibliography

AMR Archif Melville Richards: Melville Richards’s place-name material on permanent . loan at the Place-name Research Centre in the University of Wales Bangor.
BBCS Bulletin of the Board of Celtic Studies ( Cardiff 1921-)
Buckley Buckley [The Magazine of the Buckley Society] ( Buckley 1970- )
CDEPN E.Ekwall, The Concise Oxford Dictionary of English Place-names ( Oxford 1960) Cymreigio H.W.Owen, Cymreigio Enwau Lleoedd yng nghylch Wrecsam ( Bangor 2000 )
(Bangor Papers 1, ed. B.Jarvis, P.Lynch, University o f Wales Bangor)
EANC R.J.Thomas, Enwau Afonydda Nentydd Cymru (Caerdydd 1938)
EEW T.H.Parry-Williams, The English Element in Welsh ( London 1923 )
ELI [. Williams, Enwau Lleoedd ( Liverpool 1945)
ELM G.T.Jones, T.Roberts, Enwau Lleoedd Mon ( Llangefni 1996)
Enwdu B.L.Jones, Enwau { Llanrwst 1991 )
EPN K.Cameron, English Place Names (London 1996)
in’G M.Richards, Enwau TiraGwlad ( ed. B.L.Jones, Caernarfon 1998)
FPN EDavies, Flintshire Place-names ( Cardiff 1959)
GRCXi J.Lloyd Jones, Geiriadur Barddoniaeth Gynnar Gymraeg ( Cardiff 1931 )
GPC Geiriadur Prifysgol Cymru (Caerdydd 1950-)
LBS The Lives of the British Saints (ed. S.Baring-Gould, J.Fisher, London 1907-13)
LPN M.Gelling , A.Cole, The Landscape of Place-names ( Stamford 2000 )
Nomina Nomina [ Journal of the Society for Name Studies in Britain and Ireland ) ( Hull,
Cambridge 1977-)
OED The Oxford English Dictionary ( Oxford 1888-1933 ) E.Lhuyd (c. 1700) Parochialia in ArchaeologiaCambrensis ( London 1909-11)
PNCh J.McN.Dodgson, The Place-natnes ofCheshire (Nottingham 1970-97)
PNDA H.W.Owen, Place-names of Dee and Alun (Llanrwst 1995)
PNEF H.W.Owen, The Place-names of East Flintshire ( Cardiff 1994)
PNSa M.Gelling, The Place-names of Shropshire (Nottingham 1990)
RadPN R. Morgan, A Study of Radnorshire Place-Names ( Llanrwst 1998 )
REL E.Davies, Rhestr o Enwau Lleoedd : A Gazetteer of Welsh Place-names ( Cardiff . . 1975)
RLWB D.Sylvester, The Rural Landscape of the Welsh Borderland. A Study in Historical . . Geography { London 1969)
VEPN The Vocabulary of English Place-Names , eds. D.Parsons, T.Styles, C.Hough ., (. . .. Nottingham 1997-)
WSP H.W.Owen, ‘English Place-names and Welsh Stress Patterns’, Nomina xi (1987), 99- . 114.
WA’I’U M.Richards, Welsh Administrative and Territorial Units ( Cardiff 1969)
Ystrad Alun Ystrad Alun [ Journal of the Mold Civic Society ] ( Mold / Yr Wyddgrug 2000-)

Copyright of 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.