This book belongs to
THE CAMPBELL COLLECTION
purchased with the aid of
The MacDonald-Stewart Foundation
and
The Canada Council
i /'
,; e of M
Li n n A ^ v '*
IbKAKY
CAMPBELL
COLLECTION
LUNDS UNIVERSITETS ARSSKRIFT. N. F. Avd. 1. Bd 12. Nr 6.
CONTRIBUTIONS TO THE HISTORY OF
OLD ENGLISH DIALECTS
BY
EILERT EKWALL
LEIPZIG
C. W. K. GLKERUP OTTO HARRA9SO\VITZ
MAY 1 * 197
LUND 1917
PRINTED BY HA KAN OHLSSON
Abbreviations of Sources of Early Forms.
Anc. D. = A Descriptive Catalogue of Ancient Deeds, preserved in
the Public Record Office I VI. London 18901915.
Ass. R. = A Calendar of the Lancashire Assize Rolls, preserved in
the Public Record Office. Record Society for Lancashire and Cheshire 47.
Birch = Cartularium Saxonicum ed. W. de Gray Birch.
Cl. R. = Rotuli Litterarum Clausarum in Turri Londinensi asservati,
London 1833.
Colch. Ch. = Cartularium Monasterii Sancti Johannis Baptiste de Cole-
cestria I, II. London 1897 (Roxburghe Club). A compilation made from
the original documents in the time of Henry III. Some entries are later.
Cockers. Ch. = The Chartulary of Cockersand Abbey, ed W. Farrer.
Chetham Society Manchester 1898 etc. The chartulary was compiled by
Brother Robert de Lachford 126768.
DB Domesday Book.
Dugdale = Monasticum Anglicanum, originally published by Sir Wil-
liam Dugdale. New edition London 1846.
Exon DB Exon Domesday in Domesday Book IV.
FA = Inquisitions and Assessments relating to Feudal Aids A.D.
1284-1431. London 18991908.
Glouc. Ch. = Historia et Cartularium Monasterii S. Petri Gloucestriae
I III, ed. W. H. Hart. Chronicles and Memorials 18631867. The MSS
date from the 15th century.
Godst. N. = The English Register of Godstow Nunnery, near Oxford,
written about 1450, edited by Andrew Clark. Early English Text Society
129, 130 (1905-6).
HR = Rotuli Hundredorum, temp. Hen. III. et Edw. I. Record Com-
mission 181218.
IN = Nonarum Inquisitiones . . temp. Regis Edwardi III. Record
Commission 1807.
Inq. El. = Inquisitio Eliensis in Domesday Book IV.
I. p. m. = Calendarium Inquisitionum post Mortem, sive Escaetarum ;
ed. J. Caley. (Record Series) Vol. I.
Johnston = Johnston, The Place-Names of England and Wales. 1915.
Kemble = Codex Diplomaticus Aevi Saxonici ed. Kemble. London
183948.
4 Eilert Ekwall
La. Ch. = Materials for the History of the Church of Lancaster ed.
W. O. Roper. Chetham Society 1892 ff. The chartulary was compiled in the
15th cent. The charters mostly date from the twelfth and thirteenth centuries.
Lacy Comp. = Compoti of Henry de Lacy, Earl of Lincoln. Chetham
Society 112.
La. Inq. = Lancashire Inquests, Extents, and Feudal Aids ed. W.
Farrer. Record Society for Lancashire and Cheshire 48.
Lay Subs. = Exchequer Lay Subsidy Rolls 1327, 1332. Record Society
for Lancashire and Cheshire 31.
L.F. = Final Concords of the County of Lancaster ed. W. Farrer. Re-
cord Society for Lancashire and Cheshire 39, 46.
Lib. Gust. = Liber Custumarum in Munimenta Gildhallse Londoniensis,
ed. H. T. Riley. Chronicles and Memorials 12, 1859-1862.
Malmesbury Reg. = Registrum Malmesburiense, ed. J. S. Brewer (Chron-
icles and Memorials). MS. from about 1300.
Middendorff = Middendorff, Altenglisches Flurnamenbuch. Halle 1902.
Or. R. = Rotulorum Originalium in Curia Scaccarii Abbreviatio, temp.
Hen. Ill Edw. III. Record Commission 1805, 10.
Pat. R. Rotuli Litterarum Patentium in Turn Londinensi asservati.
London 1835.
Percy Ch. = The Percy Chartulary (Surtees Soc. 1911).
Pipe R. = The Great Roll of the Pipe. Publications of the Pipe Roll
Society 1884 ff.
Plac. A. = Placitorum in Domo Capitulari Westmonasteriensi aaser-
vatorum Abbreviatio. Record Commission 1811.
Ramsey Ch. = Cartularium Monasterii de Rameseia. Chronicles and
Memorials 79. London 1884 93. The MS dates from about the middle of
the fourteenth century.
RB = The Red Book of the Exchequer ed. H. Hall. Chronicles and
Memorials 99. London 1896.
Reg. Osen. = The English Register of Oseney Abbey by Oxford, writ-
ten about 1460, edited by Andrew Clark, Early English Text Society 133, 1913.
Rot. Ch. = Rotuli Chartarum in Turri Londinensi aseervati. London 1837.
Scarisbrick Ch. = Ancient Charters preserved at Scarisbrick Hall.
Historic Society of Lancashire and Cheshire 1896 7.
Whalley Ch. = The Coucher Book of Whalley Abbey ed. W. A. Hul-
ton. Chetham Society 10 etc. The chartulary was compiled in the four-
teenth century.
I. West Germanic a before 1 followed by
a consonant.
It is generally held (cf. e. g. Sweet, History of English Sounds
427, 436, Sievers, Angelsachsische Grammatik 80, 158, Bill-
bring, Altenglisches Elementarbuch, 134, Luick, Historische Gram-
matik 138, 146), that West Germanic a before I followed by a
consonant developed differently in different dialects, so that it
became ea (was broken* to ea) in Saxon and Kentish, but re-
mained as (or became , which later returned to) a in Anglian.
The development of W. G. a in this position, consequently, is.
considered to be an important dialect criterion. Yet no one, so
far as I know, has ever really tried to prove that breaking took
place universally in Kentish and Saxon dialects. Moreover, the
correctness of this view has been called into question. Thus two
pupils of Professor Morsbach's (Dolle, Zur Sprache Londons vor
Chaucer, p. 83, and Schlemilch, Zur Sprache und Orthographic
spatae. Sprachdenkmaler, p. 27) have lately advanced the opinion
(in reality that of Professor Morsbach) that breaking only took
place in the South Eastern dialects (Kent-Hampshire). Under
these circumstances an investigation into the question, which brings
fresh material under discussion, will not be looked upon as un-
necessary.
A fresh examination of the OE material would further us
very little, if at all. Few OE texts can be located with any
amount of certainty. Moreover, West Saxon began to dominate
at an early period as a written language, and, at least in the case
of later Old English, it is difficult to find any South English text
that is not more or less influenced by it. Also the Middle Eng-
lish literature (at least literature proper) would yield very meagre
material. Again, it is difficult to find texts that can be assigned
with certainty to definite dialect districts; at least many (or rather
6 Eilert Ekwall
most) dialects are not represented in ME literature. Further,
OE ea at an early period became a. In some cases, it is true,
an OE ea would leave traces in ME: when it was preceded by
a c or 3, for these consonants would be palatalised and appear
as ME ch, y; and when it stood before Id, for here ea would be
lengthened and become ME e (or the like). In dialects where
breaking took place we expect forms such as chalf, ~$ld as against
calf, (>ld in dialects where a remained as . But it is a well-known
fact that forms like chalf, eld are rare in ME except in Kentish,
even in dialects where we should expect breaking to have taken
place. We must assume that to some extent the broken forms
(or, as I shall call them, the ea-forms) have been supplanted by
the unbroken (the #-forms). If we could find no other material
than that alluded too, we should indeed have little hope of being
able to solve the question of the OE breaking. Fortunately there
is one kind of material that has so far been used only to a very
small extent, viz. the place-names. If the breaking took place in
the dialect of a given district, we should expect to find some
traces of ea-forms (such as chalf or ~$ld) in its place-names, more
especially in the earliest forms of these. The following study is
based exclusively on place-name material. .
The words that have to be taken into consideration here
are chiefly the following:
a. c(e)alc 'chalk', c(e)ald 'cold', c(e)alf (c(e)alfor) 'calf (prob-
ably also Cealf p. n.), g(e)alga 'gallows'.
b. (e)ald 'old' (Ealda p. n. and names in Eald-), h(e)ald '
'hold', w(e)ald 'wold'.
The word gealga, however, is rare in place-names. No ea-
forms of it have been met with in place-names. Personal names
containing Beald-, Eald- are of little use as material, because
in these the diphthong would as a rule be shortened (or perhaps
rather not lengthened). For the same reason names containing
eald, weald are not always conclusive.
As regards weald it should also be remembered that it is
not always conclusive when forming the latter part of compounds.
Sievers points out (Indogerm. Forschungen 14, 36 if.) that break-
ing did not always take place in weak-stressed syllables, and that
1 Of heald adj. 'sloping' I have found no conclusive examples.
Contributions to the History of OE Dialects 7
ven before the breaking a in such syllables sometimes became
o, as in -bold for -beald. As a matter of fact forms in -wold are
found as early as DB : Sudwoldam = Southwold Suff., Prestewolde
Prestwold Leic. So a ME -wold is not a certain proof of
breaking not having taken place in the district where it occurs.
But very likely breaking generally took place in a syllable with
medium stress (e. g. in trisyllabic forms). Moreover many com-
pounds in -wold are obviously late formations.
Also the trustworthiness of OE cealc may be called into
question, as in the standard language and many dialects it has
developed differently from c(e)alf, c(e)ald ( : chalk, as against calf,
cold); cf. especially Ritter, Die Verteilung der ch- und /c-Formen im
Mittelenglischen, p. 20 ff. It has also been suggested (cf. Ritter
and literature quoted by him) that in OE cealc ea is not due to
breaking but to the influence of c in an earlier OE *ccelc (< Lat.
calceni). But it is not to be doubted that the word is an early
loanword, as it occurs in several early Germanic languages (OHG
kalch, kalk, MDu kalk etc.). OE *C(dc would have to be explained
as due to readoption of the Latin word. It is quite unnecessary
to make such an assumption, as the form chalk in the standard
language can be easily explained as a Southern form and the
spread of the form in dialects is accounted for by the influence
of the standard language. To some extent, direct influence of the
Southern dialects on those more to the north may be assumed.
This view is supported by the fact that chalk is particularly char-
acteristic of Southern, especially South-eastern England. Just
because the form chalk may to some extent have spread from the
Southern dialects, names containing this element are perhaps not
quite so certain examples as the others. However, it should be
noted that the form chalk has been found in place-names only
in such districts where we have every reason to assume breaking
to have taken place. For my own part, I believe examples con-
taining this element are equally valuable with the others.
The material on which this study is based consists chiefly
-of names, forms of which have been found in early records.
To some extent I have been able to use monographs on the place-
names of particular counties or other works dealing with place-
names, before all Johnston The Place Names of England and
Wales. But the material found in such sources was rarely suf-
8 Eilert Ekwall
ficient for my purpose it may be mentioned here that I did
not find one single example in Skeat's Berkshire Place Names -
and I have had to supplement it by collections of my own. Some
books on place-names have not been accessible to me x . Heuser's
Alt-London contains valuable material for some counties.
Of the records consulted, some of the earliest, as the Domes-
day Book, must be used with caution, because their forms are
strongly influenced by Norman pronunciation and spelling tradition.
As has been shown especially by Zachrisson, Anglo-Norman In-
fluence, p. 32 ff., the early Norman scribes used the symbols c,
ch in accordance with principles quite different from later ME
ones. In DB and other early documents * ch is used to denote
both | k and tf . The symbol c in DB is obviously used now .
and then even before a to denote the sound going back to OE
palatalized c, i. e. | tf or a Norman substitution for this, as in
Calmimdelei, now Cholmondeley (OE Ceolmund-) or Calmonleuge now
Chulmleigh; cf. also Chalvington Suss, and Chalfield Berks infra.
That c is often used in DB and other early texts before e, i to
denote the sound later written ch (or a Norman sound substituted
for it) is a well-known fact.
Spellings with ea in late OE charters have not. been taken
into consideration; ea may be due to influence from the West
Saxon standard language. Early ME spellings with e (e. g. in
the DB) must be used with caution. It is possible that e denotes
or a diphthong ea or a sound intermediate between that and
later a, but we cannot be sure that such is the case. In DB we
find e. g. Seltwella (OE salt-, sealt-) in Norfolk, where we may
take it for granted that breaking did not take place. How this
e is to be explained, it is not easy to say. Partly e may be due
to a slip of the pen. DB sometimes has a instead of e; inversely
e may have been used for a. Possibly late OE ea-spellings may
have played some part 8 . Anyhow such forms in DB as Eldeham
1 Not e. g. Duignan's monograph on Staffordshire, Skeat's on Cam-
bridgeshire, Hertfordshire, and Huntingdonshire names.
2 Examples of ch = (k) in the Pipe Rolls are: Chardif (= Cardiff]
11756, Chocham (= Cookham) 1183-4, Chunegescliue 11845.
1 I consider it quite likely that the Domesday compilers sometimes
used Saxon documents in drawing up their returns. Domesday forms may
therefore to some extent show influence from the late West Saxon standard
language. An example of such influence I see in the quite common occur-
Contributions to the History of OE Dialects 9
Herts and Eldewincle Nhp, which are not supported by later
forms, I do not consider as sufficiently conclusive l .
Quite isolated forms in early records must of course be used
with caution, as they may be mere slips or due to some special
circumstance which it is impossible to judge of.
The majority of the texts used are original records, and
these naturally form the most valuable part of the material. I
have also used a number of records preserved in somewhat later
copies, especially chartularies of abbeys and similar institutions.
These latter sources are valuable, because in many cases they
contain a comparatively large number of examples, names of
fields, and the like. These sources must of course be used with
some caution. It should especially be noted that the dates ap-
pended to forms taken from such sources are those of the original
records, not those of the transcript. In many, perhaps most, cases
the copyists were no doubt careful to render their originals as
carefully as they could, but of course the forms familiar to them
would be apt to displace the old forms they were copying. Ob-
viously such forms as may be suspected to be late forms inserted
by the copyist (or translator) must be used with caution. But
when in these records we find such forms as eld for old or chalf
for calf, which by the time the copy was made, must be as-
sumed to have begun to go out of use, we may be sure that the
scribe copied his original carefulty or else that he used such forms
himself. We may say that a-forrns are often suspicious in such
records, whereas ea-forms may generally be trusted.
The collection of material does not claim completeness. The
number of records available has been restricted owing to outward
rence of L instead of I, J in DB. Zachrisson, who points out such cases
(Anglo-Norman Influence p. 66), as Loletorp, Lucufled (= Yawtlwrp, Tokefleet)
thinks these forms are merely due to orthographic confusion between 1, J,
L, which are extremely alike in the MSS of DB. But it is difficult to be-
lieve that they are not, to some extent at least, due to misapprehension on
the part of Norman scribes of the AS type for (r, which is often like an L.
1 But in spite of the fact that the early forms have thus often com-
paratively little value for our immediate purpose, I am careful always to
quote them, and for two reasons. In the first place they often give us valu-
able hints us regards the etymology of names. In the second place they
tell us that the name in question is one of old standing, which is often of
importance to know.
10 Eilert Ekwall
circumstances. Yet I think the material is full enough for my
purpose. No doubt an examination of an additional number of
sources would add valuable material, but I doubt whether the
results would be materially altered. In one direction fresh ma-
terial would doubtless be of considerable value; it would enable
us to form a somewhat more definite opinion as to the time when
the ea-forms began to be and were finally displaced by a-forms.
This is a question which I do not claim to solve in this study.
Material.
I deal first with the counties south of the Thames, the Avon,
and the Bristol Channel.
Kent.
In Kent ea-forms are well represented even in ME literature.
They are common also in place-names.
CHALK: Celca DB, Chelke FA 1284, Chalke ib. 1316, Chealke
ib. 1346 (:OE cealc).
Chaldegate Rot. Ch. 1208, Cheldegate HR I. 224 (1275).
CHALLOCK: Cealfloca 835 (Johnston), Challok Auc. 1). VI
(1429).
YALDHAM: Ealdeham FA 1346.
YALDING: Elding 1 HR I. 224 (1275), Ealdinge FA 1316, Eal-
dynge ib. 1346 etc. (probably an OE patronymic Ealdinyas).
Ealdehalle FA. 1346.
WEALD, THE WEALD: Wald' HR I. 218 (1275), hundr. de
Waldis ib. I. 206. Cf. Weldslade FA 1284, Waldeslade ib. 1346.
RINGWOULD: Ridlingwalde HR I. 206 (1275), Eidelyngwealde
FA 1316, Eynewolde Or. R. II. 103 (1337).
WOMENSWOULD: Wymelingweld FA 1316, Wymelyngeswold Or.
R. II. 268 (1362).
SIBERTSWOLD: Siberteswald DB, Syberteswold FA 1346.
The early forms Wald, -wald are doubtless due to early
shortening of ea in accordance with Kentish soundlaws (Morsbach
ME Gram. 58, an in. 2).
CALDECOTE: Caldecote FA 1346.
Caldham Auc. D. I (1426) is stated to be in Kent. I can-
Contributions to the History of OE Dialects 11
not find such a place, but the correctness of the statement is
perhaps corroborated by the family name de Caldehum FA 1346
(Kent).
WOULDHAM has OE u (Wnldaham c. 970, Johnston).
Sussex.
Cf. Roberts, The Place Names of Sussex 1914.
Chalkmere MR II. 211 (1275).
CHALDER: Chaldore HR II. 211 (1275). Former element
probably OE ceald.
CHALVINGTON: Calvintone DB, Chalvynton and the like 1278
etc. (: Cealfan tim Roberts).
OLDCOURT (now lost): Veillecourt 1289, Ealdecourt 1315, El-
decourte 1324, Oldconrt 1318 (Roberts). Cf. also Eldeton (campus)
Custumals of Battle Abbey (Camden Soc.), 18 (1307).
Early shortening has taken place in WALDRON : Waldrene
DB, Waudern 1268, Wulderne 1278 etc. (Wealda + earn Roberts;
more likely perhaps weald -\- fern). An early a- form is:
OLD SHOREHAM: olde Shoreham Godst. N. 589 (1218, late
copy!). COLD WALTHAM I do not find in early sources.
Surrey.
CHALDON: Cealfadune Kemble 812 (1062), Chaddon FA 1316,
Chalvedoun, -don ib. 1428 (: OE cealf). Cf. also Chalnemer Auc.
D. I (1336), presumably for Chaluemer, and Chalveteghe, Chalfetheghe
Custumals of Battle Abbey (Camd. Soc.) p. 138 f. (1312); -teghe
< OE teag 'enclosure, pasture'.
ELDEBURY (or ELDERBURY) : Eldeb&'ie DB (: eald or Ealdan?).
Cf. Eld-, Eldehawe Custumals of Battle Abbey 138 f. (1312).
An early example with OE a is Calkote Auc. D. I. (1362);
not found in other sources.
Hampshire.
CHARLTON: Schalston FA 1316, Chalghton ib. 1346, Auc. D.
II (1335), Chaulton FA 1428, Challeton ib. 1431, Chaloughton Anc.
D. I (1505). Former element OE. cealc. It is uncertain whether
CHALTON has the same origin, as certain early forms are wanting
(Johnston identifies OE Cealhtune 722 with it).
Chaldecote Or. R. II. 185 (1347).
12 Eilert Ekwall
STOKE CHARITY (formerly OLD STOKE) : Eledestoke FA 1346,
Eldestolce ib. 1428, Oldestoke ib. 1431.
WIELD: Walde DB, Welde FA 1316, IN 1341, ADC. D. V.
(1554), Wilde FA 1346, Weelde ib. 1428, Wildemore HR II. 222
(1274). The vacillation between e and i is curious; may influence
from the adjective wild be assumed ?
CHOLDERTON (Cerewartone DB, Chelewarton FA 1316) does not
belong here (: OE Ceolweard-).
From the Isle of Wight one example may be adduced:
Chalfham Anc. D. II (t. Hen. Ill); latter element no doubt
hamm 'enclosure'.
CALBOUENE (: Gawd-, Caulbourne FA) does not belong here.
Berkshire.
No examples in point are given by Skeat, The Place Names
of Berkshire 1911, but the following two names containing OE
cealf clearly belong here:
Chahrofte Godst. N. 52 (13823), 60 (1442), 61 (1453).
Litell Chaleueye ib. 56 (1404), litett Chalueye 57 (1410).
Doubtful examples are Chaldesworth IN 1341 (now lost?)
and Eldeslei DB. CHALLOW appears as Ceawan-hlacwe Birch II.
601 (Skeat).
There is one certain a-form:
CALCOTT: Caldecote FA 1428. I have noticed no early forms
of COLD ASH.
Wiltshire.
CHALK, BOWER CHALK: Chelche DB; Chalke, Burchalke FA
1316; Chalk, Borchalk ib. 1428. Cf. also Chalkebourn Anc. D.
Ill (1513).
CHALFIELD: Caldefelle DB, Chaldefeld Cl. R. 1216, FA 1428,
Anc. D. Ill (1405), Chadefelde FA 1316 (: OE ceald). Cf. also
Luttlechaldeforlong (High worth) Godst. N. 625 (c. 1270), but
litilcoldeforlonge ib. 620 (1270).
Chalvecrofte (Kemble) Malmesbury Reg. II. 408 (1309).
OE. eald in: Eldelotide (Highworth) Godst. N. 625 (c. 1270).
But aid in: Oldedich, Oldelonde (Corston) Malmesbury Reg. II.
109, Oldedone (Colerne) ib. 97.
Coldecote Auc. D. I (1442) presupposes OE cold.
Contributions to the History of OE Dialects 13
Early forms are not available for OLDBOROUOH. CALSTONE
(: Calestone DB, FA 1316) and CHOLDEETON (: Celdrintone DB,
Chaldrynton FA 1316; former element a name in Ceol-?) do not
belong here.
Dorsetshire.
CHALDWELL (now HOLY WELL): Chaldeivelle FA 1316.
CHALDON, East and West: Celnedime, Calnedone DB, Chal-
vedon Cl. R. 1224, HR I. 103 (1275), FA 1285, 1346, Chahedon
Boys FA 1346 etc. (: OE cealf).
The personal name Chaldecote occurs several times in the
Dorset Feudal Aids: W. de Chaldecote p. 25 (1285), E. Chaldecote
p. 117 (1431) etc.
CALLOW WESTON (: Calweston FA 1431) and COLWAY (: Calwe-
hcgh FA 1346, 1428) no doubt contain the OE adjective calu 'bald'
or a derivative of it.
Somersetshire:
. CHOLWELL HofosE: cf. W. de Cheldewall FA IV, p. 287 (1284)
under Somerset.
CHELVEY: Caluica Exon DB, Calviche DB, Chalvy FA 1284,
Chehy ib. 1303. 1316, 1346, Or. R. I. 228 (1316), Chelvey FA
1428, ? Chalvehulle Anc. D. IV (1313, identified with Chelvey by
the editor). The first element is no doubt OE cealf with late
W. Sax. change of ea into e (Biilbring 314). The latter element
seems to be OE wic (cf. gatawic etc. Middendorff, p. 149), which
has lost its w after v and its final consonant (cf. ME -li < -llch etc.).
CHILTON-UPON-POLDEN may belong here : Cahalton (no doubt
for Chalton) FA 1284 (R. de Chelton ib.), Chauton FA 13031428.
First element OE cealc or ceald or cealf? The modern form seems
to be due to the influence of three other places of the name of
Chilton in Somerset (: Cildetone, Cilletone DB).
Eldebreche (Low Ham) HR II. 122 (1274). OE eald 'old'
and bryce 'a clearing'.
Caldecote DB is quite isolated. KELSTON (: Calveston FA 1303.
1346, 1428, Kilveston ib. 1316, E. de Calveston ib. 1316) is oty-
mologically obscure. If Kilveston 1316 is simply a bad spelling,
the former element may be OE C(e)alf and we have an OE a-
form before us. It is to be noticed that Kelstou is situated near
14 Eilert Ekwall
the Gloucestershire border. ALDWICK appears as Aldvic DB, Hoi-
deivyk FA 1303 etc. (probably OE old 'old'). Also this place is
in the N.E. not far from the Gloucester border.
Devonshire.
CHALLACOMBE: Celdecome DB, Chaudecumb FA 1284, Chal-
decumb ib. 1316, Choldecomb ib. 1346, Cholcombe ib. 1428. Cole-
comb ib. 1303 seems to be due to confusion with CULLJCOMBE
(earlier Colecumb etc.). The former element seems to be OE ceaid;
forms such as (E. de) Chaluecumba Pipe R. 1167 8, (David) Chal-
vecombe FA 1428 (p. 466) are no doubt due to mistake.
CHOLD ASH: Choldasshe FA 1346, Cholde Assh ib. 1428.
CHAWLEIGH: Calvdie DB, Chalvelegh FA 1284, 1316, 1346,
Chalvary ib. 1428 (: OE cealf}. Cf. also CHELFHAM, early forms
of which have not been found.
Certain early examples of the -form have not been noted.
Caldebir' Plac. A. 72 (c. 1200) is stated to be a place in Devon,
but I have found no other traces of it. Of COLD EAST no early
forms have been noted. COLDRIDGE appeal's as Colrugg(e) FA
1303, 1316.
CAI/VERLEY appeas as Kaletvode-, Calwodelegh in FA. It ob-
viously does not belong here.
Cornwall.
The only example found is:
Choldecote Anc. D. II (1375). Cf. the personal name Chole-
co/e FA I, 232 (1428) under Cornwall. The personal name Chole-
wyll ib. 236 (1428) may be a case in point, but a place of this
name has not been found.
The development ceuld >> ceald > ME chold is worth noticing.
Besides in Somerset, Devon, and Cornwall it has been exemplified
in a London place-name (Choldherberwe infra).
We pass on to the counties north of the Thames etc. where
place-names exhibiting OE ea have been found.
Essex.
Cf. Heuser, Alt-London p. 36.
The material is fairly extensive, thanks especially to the
Colchester Chartulary. I give fairly numerous examples from this
Contributions to the History of OE Dialects 1J>
collection notwithstanding the fact that Heuser quotes several.
Heuser omits the dates and also some examples important for
our purpose.
CHADWELL: Celdeuuella DB, Chaudewell Rot. Ch. 1205, Chal-
dewell Cl. R. 121213, Anc. D. I (1348), FA 1428 etc. But Cal-
dewell Plac. A. 225 (1291). Caldeuuelle Colch. Ch. 649 is a place
in Hunts.
Cheldelerecroft (Bergholt) Colch. Ch. 289. But Cheldebenes-
mad ib. 433 is stated to have belonged to a person named Chel-
debeue; this name is obscure.
OE cealf forms the first element of Chaluedona Colch. Ch.
62 (1179), 68 (1202), -done 96 (1237), also Cealuedune ib. 143, Clial-
vedon HR I. 154 (a place in Pitsea); Chalfpyghtel (printed Chats-)
Colch. Ch. 501 (1364), also Chalfpitel 610 f., -pichtel 667, -pytel
614 etc. (all in late, probably 14th cent, charters), in Colchester.
An obscure case is CHAFFORD HD.: Ceffeorda, -urda H. DB r
Cliafford RB, FA 1303 etc., but Chalfewurd' (Chaff ewurd') hund-
redum Pipe R. 11823 (: OE cealf?).
OLD HEATH: Ealdehethe, Eldehethe Colch. Ch. 307 f., but
the Old-Hyth Dugdale, IV. 603 (1405), Oldhythe ib. 612 (1544).
Latter element OE hyp 'landing-place'. OE eald also appears in:
Eldenaptune (Colchester) Colch. Ch. 329 f., 441, ^Eldeneptune ib.
540. Chenewoldes ealdecote (Rayne) ib. 233, but Chenewoldcs aide-
cote ib. 232. Cf. also Eldefelde Plac. A. 216 (1288).
A doubtful case is YELDHAM : Geld(e)liam DB, Geldham Colch.
Ch. 96 (1237) etc., Gelham ib. 565 etc., FA 1303 etc. According
to Heuser <C Eald-.
But OE aid in the late forms Oldegatyslond (Colchester) Colch.
Ch. 501 (1364), Oldhalle (Manningtree) Anc. D. II (1475).
WEALD (SOUTH W., NORTH W. or W. BASSET): Welda, Walda
DB, Weld FA 1303, 1428, Welde ib. 1346, Northtvelde Anc. D. I
(1280, 1324), IN 1341 etc., -weld Anc. D. I (1300), Walde HR.
I. 153 (1275), de Waude (Walde) Colch. Ch. 433. The occasional
form Walde is no doubt due to shortening of ea. OE weald is
perhaps the former element of WALTHAM: Weald ham, Waltham
Colch. Ch. 22 (110018), Waldham ib. XLII.
An interesting case is Cestreswold (Bergholt) Colch. Ch. 425.
The word does not look like an old compound, otherwise an OE.
16 Eilert Ekwall
change a > o in a weak-stressed syllable might be thought of.
The place is in the extreme N., on the Suffolk border.
KELVEDON does not belong here. It appears as Cynlaue
dyne 998 (Johnston). Calewecroft (Fordham) Colch. Ch. 469 probably
contains OE. calu 'bald' or a derivative of it.
Suffolk.
Cf. Skeat, Place Names of Suffolk 1913.
In this county ea-forms are rare :
OLD NEWTON: Eldeneuton Or. R. II. 171 (1345), Eldventon (sicl)
ib. 102 (1336), but Olde Neuton FA 1346, IN 1341. Cf. also El-
delonde Anc. D. I.
Otherwise only a-forms have been found:
Caldecota DB, Caldecotes HR II. 162 (1275).
Caldewalle Anc. D. I (1296), Caldewdl ib. (1296).
OULTON: Ouliun TN etc. (Skeat), Oltun HR II. 170 (1275),
Oldton IN 1341.
SOUTHWOLD: Sudwoldam DB, Suthwald, -ivold HR etc. (cf.
Skeat). As shown by the DB form this is not a conclusive example.
It should be noted that Old Newton lies about the centre
of the county, whereas Oulton is in the N.
Middlesex (London).
Cf. Heuser, Alt-London.
CHELSEY: Cdchyth Birch I. 355(789), ChekhedDE, Chelchehuth
FA 1316 etc.
ChoUherlerwe Calendar of Wills 12581349 (Heuser).
OE eald frequently appears as eld(e) in names of buildings,
streets etc., as Eldefishstrete Lib. Cust. 276 (1324), 410 (1321), but
Oldefishstrete ib. 404, 409 (1321), Oldefi(s)shstret Auc. D. I (1386,
1398:, Oldefahestretelane Or. R. II. 177 (1346); Eldeforde Auc. D. I
(1323), but Oldeforde Plac. A. 316 (1313); the Eldynne Gesta Ab-
batum Mon. Sancti Albani (Rolls Ser.) III. 514 (1408); Elde Marie
church Anc. D. I (t. Hen. III.), Eldemariecherche ib. (1287), -chirche
Lib. Cust. 229 (1303); Eldestrate HR I. 432 (1275).
Heuser p. 23 gives quotations for various names of this
kind from Calendar of Wills I (12581349), as Elde Jewrie, Elde-
chaunge etc., without appending the exact dates. He quotes 32
cases of Elde- as against 7 of Olde- in this collection.
Contributions to the History of OE Dialects 17
Hertfordshire.
CHALDEANS: Celgdene DB, Chaldene FA 1303, Chaldenes ib.
1402, Chalden ib. 1428. The DB form points to OE cealc as
former element, but if the following early Pipe Roll forms refer
to this place, OE cealf is perhaps more likely : Chaluedon Pipe
Rolls 11656, -dona 11667, -dene 11689, 11701, Calueden
1169 70 etc. (stated to be in Herts or Essex).
Chalkdelle Gesta Abbatum Mon. Sancti Albani (Rolls Ser.)
II. 262 (1380).
CALDECOTT: Caldecota DB, Chaldecote FA 1303 (bis), Caldi-
cote ib. 1428, Caldecot HR 1. 194 (1275), CaldeJcote IN 1341.
The FA spelling of 1303 clearly represents the earlier form of
the name.
OE cealf forms the first element of Chalcrofte, Chalncroft
(Saudon) S:t Paul's Doomsday (Camden Soc.) 13 f. (1222).
Eldehallegrene Anc. D. Ill (1423).
The DB spelling Eldeham (probably ALDENHAM) is not con-
clusive.
Buckinghamshire.
CHALVEY: Chalveye HR I. 34 (1255), FA 1302, 1316, Chalf-
heye ib. 1346, Chalveye Anc. D. I. (1344). Former element clearly
OE cealf, as in Chalveh HR I. 46 (1276), which seems to be a
different place.
The rest of the names point to OE a:
CALDECOTE (Newport Pagnell) and CALDECOTE (Bow Brick-
hill): Caldecote DB, HR I. 40 f., FA 1284 etc., Plac. A. 261 (1307).
It is not always easy to keep the two apart, but both are given
in FA and HR.
CALVEBTON: Calvretone DB, Calvertone FA 1284, 1316, 1346 etc.
Somewhat doubtful is the personal name atte Wolde FA 1316,
de Walda, de la Welde 1346. The last-mentioned form may be
a scribal error. No place Wold is known to me in Bucks.
CHALFONT does not belong here. It appears as Chalfhunte
etc. FA, but as Ceadeles funtan in OE (Johnston).
It should be noticed that Chalvey is situated in S. Bucks,
Caldecote and Calverton in N. Bucks, Calverton close to the bor-
der of Northants.
18 Eilert Ekwall
Bedfordshire.
Of. Skeat, Place Names of Bedfordshire 1906.
CHALQEAVE : Cealhgrcefan 926 OE charter, Celgrave DB, Choi-
grave HR etc. (Skeat), FA 1284.
CHALTON: Cealhtun OE charter (Skeat).
CHAWSTON: Chauelestorne, Calnestorne DB, Calvesterne RB,
Chalvesterne, Chalsterne FA etc. (Skeat). Seems to be OE cealf
or Cealf and porn, ]>yrne. The early spellings with C- may be
disregarded.
OE a-forms must be assumed for:
CALDECOTE: Caldecote I. p. rn. and CAULCOTT: Calcote I. p. m.
(Skeat). Cf. Caldecote Anc. D. II (1351), IH. (1388) etc.
CALDWELL (an old abbey): Caldewell Godst. N. 90 (1235),
HR. II. 333 (1279), Caudewelle ib. II. 325, Caldewelle FA. 1302.
HARROLD: Hareuuelle DB, Harewold FA etc. (Skeat) is not
a quite certain example; cf. p. 6.
Chawston is in the northernmost part of the county.
Huntingdonshire.
In this county there is much vacillation between ea- and a-
forms. The latter are in the majority. Of ea-forms I have found
the following:
Chelfrebecche (Sawtrey) Ramsey Ch. I. 161 (114653), Chelwre-
mare, Cheluremere, -beche ib. I. 160, 162 (1146 53), Chalu(e)re-
mere ib. I. 165 (1146 53). The former element is clearly OE
cealf (cealf or}.
Chalpithanedyn HR II. 630 (1281; in Ellington) seems to
contain as its former element OE cealcpyt. The latter I cannot
explain.
OE eald in Waltone Eldlode Ramsey Ch. II. 244 (131642),
Waltone eeldelode ib. I. 187 (1342).
WEALD: Welde Ramsey Ch. I. 195 (1244), 212 (1341), Weld'
HR II. 671 (1281); once Wolde Ramsey Ch. I. 211 (1341), probably
miswritten.
Of a-forms the following deserve mentioning:
CALDECOT: Caldecote DB, FA 13031428, Ramsey Ch. I.
305 (1251), IN 1341.
Caldeuuelle (Hamerton) Colch. Ch. 649.
Contributions to the History of OE Dialects 19
Caldemowehact Kemble 1364 (late charter 1 ), Caldemowehat,
Kaldemowehath (Warboys) Ramsey Ch. I. 308 (1251). The word
is not quite clear, but the first element is OE cold 'cold', the
last OE hd]) 'heath'.
Calfo (Spaldwick) HR II. 615 f. (1281). Probably OE calf
+ hoh.
Calurecroft (Abbot's Ripton) Ramsey Ch. I. 322 (1252).
Calver(e)wode^ate (Walton) Ramsey Ch. I. 187 (1342), II.
243 (1316-42).
Holdedole, Holdheyre (Broughton) Ramsey Ch. I. 331 (1252):
OE aid 'old'.
OE wold is found in OLD HURST Waldhirst Ramsey Ch. II.
163 (11857), Woldehyrst ib. I. 284 (1251), Woldhirst HR II.
603 ff. (1281), Woldhurst Pl&c. A. 331 (1319), and in OLD WESTON:
Weston de Wald" (Wold) HR. I. 197 (1276), Weston(e) de Waldis
FA 1285, Wolde Westone ib. 1316, Woldwestone Ramsey Ch. II.
300 (1301) etc. Cf. also Longewolde (Warboys) Ramsey Ch. I.
307 (1251), Waterwolde (Broughton) ib. I. 331 (1252).
The ea- and a-forms are distributed as follows.
Weald is in the narrow southern strip of the county which
extends along the boundary of Beds. Ellington (with Chalpit-
hanedyri) is in the S.; Sawtrey (with Chelurevnere etc.) lies to the
N.W.; Walton (with Eldlode and Holdedole etc.) S.E. of Sawtrey,
slightly N. of the centre of the county.
Caldecot is in the N.W.; Old Weston and Spaldwick (with
Calfo) in the W. part of the county; Abbot's Ripton and Broughton,
S.E. of Sawtrey and about the centre of the county; Old Hurst
and Warboys (with Calurecroft etc.) in the E.; Hamerton (with
Caldeuuelle) near Sawtrey. No certain a-forms have been found
in the S. half of the county.
Northamptonshire.
All the examples of ea except the first two are more or less
open to doubt.
Chaldenhul (Hulcote S. Nhp.) Anc. D. II. p. 381 (1372).
Eldebroc (Hulcote) Anc. D. III. p. 280 (no date).
CHELVESTOSTE : Celvertone, Celuestone DB, Chelveston FA 1284.
1 Note such forms as Cnottis rode, le banue, Asshebech, Hertyngts in
this charter.
20 Eilert Ekwall
The former element may be, and most likely is, OE Ceolf for
Ceolwulf; cf. Ceolf estun (Somerset) Kemble 816 (1065). The DB
form Celuertone is probably a mistake for Celuestone ; if r is ori-
ginal, OE Ceolferp may be thought of. But it is also possible
that the former element is OE Cealf; cf. Chawson Wore., Causton
Warw. The regular e in the first syllable would be somewhat
remarkable, but cf. the names in Chelvre- Hunts. Chelveston is
in the E. of the county near the border of Hunts.
Celvrecot DB possibly contains OE. c(e)alfra, but the form
is 1 not conclusive. It is difficult to locate the place exactly, but
it seems to have been in the S. or S.W. part of the county.
Elddvere (Whiston) Ramsey Ch. I. 57 seems to contain as
its former element OE eald 'old', but the latter part of the word
appears to be corrupt. The place must have been situated in the
E. not far from the Bedford border.
The DB form Eldewincle (also Aide-} for ALDWINKLE is not
conclusive proof of OE Eald- (cf. p. 8). The other early forms
found have a : Aldwingel 1137, Aldewincle 1166, Audewynde 1298
(Johnston), Aldewinda Liber Niger 1125 8 (in Chronicon Petro-
burgense, Camden Soc., p. 166), Aldewinde RB 1166, Aldwinde
Plac. A. 325 (1317). Also this place is in the E. of the county,
near the border of Hunts.
An apparent case is CHALCOMBE (CHACOMBE): Chaucumbe
HR II. 6 (1276), FA 1284, Chacombe FA. 1316, 1428 (: OE
Ceawa or Ccec).
The rest of the forms point to OE a:
CALDECOT (Chelveston cum C., E. Nhants): Caldecote DB,
Chaldecote Pipe R. 1184 5, -cota ib. 1185 6. There is another
Caldecote in DB (seemingly in the N.W. of the county). A third,
in the S. of the county, is mentioned by Bartholomew. It is un-
likely that Ch- in the Pipe R. examples denotes the palatalised
sound.
COLD ASHBT: Coldesby FA 1316, Colde As&eby ib. 1428,
Anc. D. I. (1452), Coldassheby Or. R. II. 256 (1361).
Caluhay (printed Calnhay) Or. R. II. 74 (1333, in forest of
Clif, probably King's Cliff, N. Nhants).
WOLD (OLD): Waldo, DB, Waude FA 1284, Wolde Or. R. I.
6 (1244), FA 1316, 1428, Anc. D. III. (1405).
Cold Ashby and Wold are in the northern part of the county
Contributions to the History of OE Dialects 21
Oxfordshire.
Cf. Alexander, Place Names of Oxfordshire 1912.
CHALFORD: Celford DB, Calk-, Chalcford 12749, ChalTce-
ford 1316 (:OE cealc; Alexander).
CHALGROVE: Cealhgrcefan etc. 926, Celgrave DB, Chalgrave
12749 etc. (Alexander). Cf. Chalkeforlonge (Shillingford) Godst.
N. 542, 544 (1322 f.).
OE ceald in Chaldfeld (Wolvercote) Godst. N. 571 (c. 1140)
and Chaldewelle (Sandford) HR II. 723 (1279).
OE eald in: Eldedieh (Tew) Godst. N. 551 (c. 1240), Eldee
(Oseney) Reg. Osen. 52 (c. 1180), 57 (1225), Eldhe HR II. 788
(1279); the heldegore (Tew) Godst. N. 551 (c. 1240), but oldegore,
old gore ib. 552 (1266), 558 (c. 1270).
Middisladeheld (Tew) Godst. N. 552 (1256) but -hold ib.
(1266), Myddelsladesholde ib. 557 (1270); cf. The Hold piscaria
(Tackley) HR II. 859 (1279). The meaning of -held, -hold is
obscure, but it presumably goes back to OE h(e)ald sb. 'hold'.
WEALD: Welde DB, Walde 1229, Wealde 1253, Welde 1269
etc. (Alexander), Welde 01. R. 252 (1216), Estwelde FA 1316,
Westwelde Anc. D. I (1316), but Westwdd FA 1316; cf. Nete-
wolde FA 1316 (=Nuffield?).
The following names have only a-forrns:
CALCUTT, CAULCOTT: Caldecof HR II. 827 (1279). It is doubt-
ful which of the two places is meant.
Caldewell (Steeple Barton) Reg. Osen. 122 (c. 12,00), Kalde-
well Plac. A. 49 (1206).
COLD NORTON: Caldenortone RB 799.
Oldebarton' (Headington) HR II. 711 (1279). Cf. Oldegore
(Hooknorton) Reg. Osen. 176 (c. 1270), Oldefeld (Cassington) Godst.
N. 290 (c. 1270).
Of the places mentioned, Chalford, Chalgrove, Weald, Sand-
ford (with Chaldewelle) are in the S. part of the county, Oseney
(with Eldee} about the centre, and Tew (with Eldedieh etc.) in the
N. part of the county.
Cassington (with Oldefeld), Steeple Barton (with Caldewell),
and Headington (with Oldebarton) are situated about the centre,
Calcutt and Caulcott in the N. of the county.
22 Eilert Ekwall
Gloucestershire.
Cf. Baddeley, Place Names of Gloucestershire 1913.
ea-forms are rare:
CHALPORD: Chalkforde 1297 etc. (Baddeley).
CHALKWELLS l : Cealcweallas 743 (Baddeley).
Eldeberge (Alderton) Glouc. Ch. I. 167 is not quite certain.
It probably means 'old barrow', but the former element might
be OE died 'fire'. Eldbertone DB does not belong here; it appears
as Aylberton, Aylbryghton in FA.
The rest of the early forms point to OE a:
COLD ASTON: Colde Astone c. 1224 (Baddeley), Colidaston
FA 1303, Coldaston ib. 1346. Cf. Coldeassheton Or. R. II. 184 (1347)
and Colde Newijnton Or. R. II. 168 (1345).
CAUDLE GREEN: Kalvelegh, Calveleesfeld Anc. D. IV (1310).
Cf. also the following names, all containing OE calf: Galvescrofte
(Wheatenhurst) Glouc. Ch. Ill, 227, 230 (1343?), 230 (1358); Calf hey,
Calfheyesputte ib. 230 f. (1358), Calverecroft (Standish) ib. II. 260
(126384), Calvecroft (Standish) ib. III. 44 f. (1267-8). Baddeley
also gives CALF- WAY, found as a personal name from 1271.
OLDBURY UPON SEVERN: Oldebiri c. 1200. OLDBURY ON THE
HILL: Aldeberie DB (Baddeley). Cf. Oldebury FA 1303, 1346,
Oldebur' IN 1341.
Cf. also Oldebreche (Eastleach) Glouc. Ch. I. 271 (probably
13th c.), Aldelande (Alderton) Glouc. Ch. I. 167, Oldelonde ib. I.
168, II. 189.
OE wald in: de la Wolde (pers. n.) Glouc. Ch. I. 195(1263
84), Newynton on le Wolde Anc. D. Ill (1379), further in STOW
ON THE WOLD and:
COTS WOLD Hills: Coteswold 1231, Coteswalde 1300, Cottasowlde
a. 1500 etc. (Johnston), Coteswolde 1318, Cotteswolde 1360 etc.
(Baddeley).
WIG WOLD: Wyggewold 1358 etc. (Baddeley), Wygewold, -wade
FA 1284.
Celflede Hd DB (quoted by Baddeley) does not mean 'calf-
1 I give this name on Mr. Baddeley'a authority, though I look upon
it as suspicious. Baddeley'a only form is taken from Kemble, who gives
Chalkivells as an identification of OE Cealcweallas. Kemble italicizes the form,
thereby marking that he does not know if there is a place of this name now,
but that this is the form it would be expected to have.
Contributions to the History of OE Dialects 23
way'. DB also has forms like Ceolflede H. (obviously OE Ceolfled
p. n.).
Chalford and Chalkwells are in the S. E., Aldertou (with Elde-
berge) in the N. part of the county. The places whose names
go back to a-forms are pretty evenly distributed over the whole of
the county ; thus Wigwold and Cold Aston are in the E., Stow on-
the Wold in the N., Wheatenhurst (with Calfhey etc.) in the W.
Worcestershire.
Cf. Duignan, Place Names of Worcestershire 1905.
There is only one name which points to OE ea, viz.
CHAUSON: Celvestune DB, Chalvestone 1108, 1275, 1327 (Duignan).
All the rest point to OE a:
GOLDICOTE: Caldicote 1275 (Duignan).
CALDWELL (Kidderminster): Caldewell 1275. CALDWELL (Per-
shoie): Caldwelle 1275 (Duignan). Cf. Caldewelle Anc. D. I (1391),
-hull ib., Caldewell ib. HE (1382 f.).
Calveham (Lindridge) Registrum Wigorniense, Camden Soc.
91, p. 11 (1240), Calvesteil ib. p. 116 (1240). Both are names of
small pieces of land. Calveham is probably OE calfa g. pi. -j-
hamm 'enclosure'; Calvesteil, perhaps 'calf's tail'. Fields and the
like sometimes had curious names of a similar kind.
OLDBURY : Ealdanbyri 972 (Duignan). The OE spelling need
not prove that the name had the eo-form, as it may be due to
the influence of the West Saxon standard language.
OLDENHALL: Aldenhulle, Oldenhull 13 cent. et6. (Duignau).
OLDINGTON : Aldintone DB (Duignan).
OLD SWINFORD: Oldeswynesford IN. 1341.
Chauson is situated in the centre of the county; most of
the others in the N., Goldicote, Caldwell (Pe.) in the S. E.
Shropshire.
There is one apparently conclusive example of OE ea:
ELCOTE: Eldecocote (obviously for Eldecote) Anc. D. Ill (1395),
Eldekot ib. VI (1529), Eldecote (1553), Elcote (1558). Former
element OE eald or Ealda. Another, though doubtful, example is:
OLDBDRY: Eldebur Cl. R. 260 (1216), but Aldeberie DB,
Aldebyr' HR II. 82, Hadeburi FA 1284, Oldebury ib. 13161431,
Oldebur' IN 1341.
24 Ellert Ekwall
EYTON ON THE WEALD MOORS appears as Eyton super Wilde-
more FA 1346, 1431 (: OE wilde-}.
Otherwise all the forms point to OE a:
Caldebrok HR II. 110 (1274). Cnldeford ib. II. 99 (1274).
COLD HATTON: Colde Ration FA 1346, 1428.
CAUSTON: Caluestone DB.
CALVERHALL: Cavrahalle DB, Calfrehall HR II. 59 (1255),
de Calverhal(e) FA 1346 (IV. 236). Cf. Calvecrofte Or. R. II. 98
(1336), Calverleye ib. 291 (1367).
OLDPIELDS (Moreton Say): Oldefeld FA 1284. Cf. also le
Oldeford Or. R. II. 166 (1345).
Elcote is (or was) situated in the S.E. part of the county on
the Worcester border. The isolated form Eldebur' of 1216 is
hardly worth much attention; it is worth noting, however, that
the place is in the S.E., though not so near the Worcester border
as Elcote. Of the other places mentioned, Causton is in the S.W.,
Calverhall, Cold Hatton and Oldfields in the N. of the county.
In the rest of the counties certain examples have been found
only of OE a.
Norfolk.
OE calc forms the former part of Kalkmellis HR I. 530
(1275). Cf. R. de Calchill, A. de Calkhyll FA 1302 (Norfolk,
p. 422).
CALDECOTE: Calda(n}chota DB, Caldecote Inq. El., HR I. 460
(1275), Caldecotis FA 1428.
*CALVELEY (no modern form found): Celueleia Inq. EL, Chalue-
lega, -lea Pipe R. 116572, 11834, 11845, Kalvele Plac. A.
243 (1301), CaWleye Or. R. II. 266 (1362), J. de Calveleye FA
1302. On the early spelling Ch- see p. 8.
CAWSTON: Caluestune DB, Caustone, Caulstone RB p. 788 f.,
Causton FA 1316.
Calvebeche (Walsoken) Ramsey Ch. III. 289 (c. 1174?).
Galgou, Galhou DB, Gal-, Galehowe Hd. HR I. 454 ff. (1275).
Former element OE galga or ON galgi. If the latter element is
ON haugr, an ON origin of the former element is of course
probable.
METHWOLD: Methelwalde, Matelwalde DB, Methelwalde, -wolde
Contributions to the History of OE Dialects 25
HR I. 438 (1275), Methwolde FA 1316, Methelwolde ib. 1346, Plac.
A. 253 (1305).
NORTHWOLD: Nortwalde DB, Northwalde, -wolde HR I. 438
(1275), Northwold Plac. A. 196 (1279), Norwolde FA 1302, North-
wolde ib. 1316, 1346.
HOCKWOLD: Hocuuella DB, Okewold HR I. 438 (1275), HoTce-
wold(e) FA 1302, 1316, Houkewold ib. 1346.
The three names in -wold are clearly late formations : Hock-
wold and Methwold have a Scandinavian former element (Hauler
p. n. and medal adj. 'middle'). The three places are situated in
the S. W. corner of Norfolk close to the Suffolk border.
Cambridgeshire.
CALDECOTE: Caldecote DB, HR II. 517 ff. (1279), Plac. A.
208 (1285), FA 1316-1428, IN 1341, Cnudecote HR. I. 54.
COLDHAM: Coldham FA 1302, Ramsey Ch. III. 122 ff. (1329).
Caldewell HR. II. 436 (1279).
Calwyscroft HR II. 498, Calvenebreg' ib. 507 (1279).
Le Oldlode Plac. A. 326 (1317), Holdelode (Chatteris) Ramsey
Ch. I. 430, Holdebreche ib. I. 434.
WOLD: Wolde FA 1302, 1346, 1428.
An apparent exception is SHELFORD: Esceldford, Sceldford
c 1080, Escelforde DB, Selford 1210 (Johnston), Shelf orde FA
1346 1428. Identical with this is SHELFORD Notts: Scelforde
DB, Schelford 1278, 1284 (Mutschmann). For Shelford Cambr.
derivation from OE sceald- would not be impossible; the place is
situated in the S. of the county. But for Shelford Notts the
assumption of OE ea is not well admissible. Clearly the same
explanation must be sought for both names. Either the former
element is not connected with OE sceald but some other word,
e. g. OE sceld 'shield'; or, what is more likely, we have to assume
a derivative of OE sc(e)ald or an i-mutated form of it. Mutsch-
inaun proposes OE *scielde by the side of OE sceald; OE *sceldu
'shallow place' might also be thought of.
Rutlandshire.
The only example met with is:
CALDECOTT: Caldecote FA 1316. Challecote Anc. D. Ill is
hesitatingly referred to this place. If so, Ch- is doubtless only
an early spelling for C-. The document is undated.
26 Eilert Ekwall
Leicestershire.
COLD NEWTON: Coldenewton FA 1428.
COLD OVERTON: Caldoverton Plac. A. 44 (1204), Coldovertone
RB 121012, Coldoverton FA 1428.
HORNINGHOLD : Horniwole DB, Hornyngwold FA 1428.
PRESTWOLD: Presteuuald, -wolde.D'B, Prestwold FA 1316, 1428.
WIMESWOULD: Wimundeswald DB, Wymondeswold ib. 1428.
Cf. WOLD HILLS, the name of an elevated tract in the N. of
Leicester.
A doubtful case is CHADWELL (in the N. part of the county).
In the Index to FA Caldwell 1316 is given under Chadwell; if
that is correct we must refer to it also: Caldcuuelle DB, Caldewella
Pipe R. 11767, ChaldewelV ib. 11823, -wella ib. 11845,
11856, -well Plac. A. 65 (1210), Caldewett HR I. 237 (1276).
The early spellings with Ch- need not be regarded ; the early
forms point to an a-form (Caldewelle). The relation between this
form and the present Chadwell may be as follows. There are in
Leicester a parish Wycombe and Chadwell and a hamlet Caldwell,
which must be situated close to each other (Caldwell in Wycombe-
and-Ch.?). The names Caldwell and Chadwell may have been con-
fused with each other. Anyhow the only certain early form found
is Caldwell or the like.
Warwickshire.
Cf. Duignan, Place Names of Warwickshire 1912.
CALDICOTE: Caldecote DB, HR II. 225 (1276), Anc. D. V
(1330), I (1358), FA 1428. There is also a hamlet CAULCUTT
(CALDICOTE Duignan), to which some examples may belong.
Caldefordemulne Auc. D. I (1325), Caldeford ib. (13 c.).
CAWSTON: Calvestone DB, Causton 12 c. (Duiguan). Cf. Calves-
croft (Bidford) HR II. 226.
OLDBURY: Aldebcrie, Aldberi 12 c. (Duignan), probably Holde-
borwrth HR. II. 228 (1276).
OLTON HALL: Oldenhale FA 1316. Cf. le Oldefeld Anc. D, II [.
Wolde FA 1316 (hamlet Snitterfield).
The places mentioned, so far as they can be located, are
mostly situated in the N. and E. parts, Calvescroft and Wolde,
however, in the S.W. part of the county.
Contributions to the History of OE Dialects 27
Herefordshire.
COLDBOROUGH : Colcheberge DB, Calbarwe FA 1303, -barewe
1346, Caldebarewe ib. 1428 (: probably OE calc).
CALDECOT: Kaldekote Anc. D. I (1340), ?Caldecot, Chaldecot
01. R. 1220 (Ch- no doubt a Norman spelling for C-). Cf. COLD-
WELL (no early forms found).
OLDCASTLE: Oldecastell Anc. D. Ill (1408).
Monmouthshire.
CALDICOT: Caldecot(e) DB, Caldecote Or. R. I. 185 (1312), II.
290 (1367).
Staffordshire.
CALDMORE : cf. Caldemoretona Pipe R. 1175 6.
COLDECOTE: Coldecote FA 1284. I know no early forms of
COLD NORTON.
CAULDON: Celfdun 1004 (Johnston), Caldone DB, Caldon FA
1316.
CALF HEATH: Calfr-e heie 994 (Jolmston). Cf. Calvermedowe
Or. R. I. 179 (1311). Early forms are wanting for CALWICH
(Calf-wie?).
It is uncertain whether la Coldewall and Oldefeld Or.R. I.
112 (1300) are in Stafford or Shropshire.
The remaining counties may be dealt with more briefly.
Lincolnshire.
CAWKWELL: Calchewelle DB, CalJcewell Plac. A. 68 (c. 1200),
FA 1316, 1401. Cf. Kelkefeld HR I. 340, 374 (1275 f.).
CALCOT: Caldecot FA 1284, Calcot ib. 1401.
COLD HANWORTH: Coldhameworth Plac. A. 276 (1284).
Galwecroft HR. I. 336 (1275).
THE WOLDS (chalk hill tract); cf. the following names.
WOLD NEWTON: Walde Neuton FA 1316, Wolde Nettton Or.
R. I. Ill (1300).
CDXWOLD: Cucvalt DB, Cokewald HR I. 294 ff. (1275), FA 1316.
Coketveld Plac. A. 229 (1292) is obviously corrupt.
STIXWODLD: Stigeswalde, -uuald, -wait DB, Stykeswold FA
1303, StiJceswald ib. 1316.
28 Eilert Ekwall
Nottinghamshire.
Cf. Mutschmann, Place Names of Nottinghamshire 1913.
CALVERTON: Calvretone DB, Calverton 1284 (Mutschmann).
WOLDS: Wolde 1252, Waldos 1363 (Mutschmann), le Wold
IN 1341, Wylughby super Wolde Or. R. II. 133 (1340).
On SHELFORD see under Cambridgeshire. The DB forms of
AWSWORTH Eldesvorde, Eldeurde need not be taken into considera-
tion; cf. p. 8.
Derbyshire.
CALKE (CAULK): Calc Dugdale VI. 598, Calk ib. 429 (1297).
CALLOW: Caldelauue DB.
CALDWELL (CAULDWELL): Caldewelle DB, Caldewell HR I. 58
(1276).
Caldecotes DB, Caldecote Or. R. I. 280 (1324).
CALVER: Caluoure DB, Calvore FA 1431 (: OE calf).
Cheshire.
CALDY: Caldeahundredo Pipe R 1185 6, Calday Anc. D.
Ill (1392), Caldey ib. I (1516).
CALDECOTT: Caldecote DB.
CALVELET: ? Cavelea DB.
Yorkshire.
KELK : Chelche, Chelchis 1 DB.
KELFIELD: Chelchefeld DB, Kelkfeld HR I. 122 (1276), Kelke-
felde Plac. A. 210 (1285).
Caldecotes DB, HR I. 106 (1276).
CALDEWELL: Caldeuuelle DB, Caldewell Plac. A. 333 (1319).
CALVERLEY: Calverlege Plac. A. 94 (c. 1200).
Galleghaye HR 1. 116 (1276).
COXWOLD: Cucualt DB, Cuckewalde Plac. A. 240 (1298), Cok-
wold IN 1341.
EASINGWOLD: Esingwald Plac. A. 70 (c. 1200), Esyngwald IN
1341. Cf. THE WOLDS (a chalk hill tract in East Riding).
1 Kelk seems to go back to an i-mutated form of OE calc. The same
form appears to enter into Chelsey (Middlesex). I have no explanation to
offer for this form.
Contributions to the History of OE Dialects 29
Lancashire.
COLDCOATS: Caldecotis 1258, Coldecotes 1298 etc. (Wyld and
Hirst, Place Names of Lancashire).
Northumberland.
CALDCOTES: Caldcotes FA 1346.
COLDWELL: Caldwell FA 1346, Or. R. II. 256 (1360).
CALLEETON: Calverdon FA 1346 etc.
For Cumberland, Westmoreland, Durham and Scotland I
have collected no early forms.
Results.
As will have been seen, the material is fairly extensive,
and we are no doubt entitled to draw rather definite conclusions
from it.
We can unhesitatingly conclude that in OE or early ME
time broken forms (ea-forins) were used in the dialects of the
districts where names containing such forms have been found,
for we may assume that borrowing of place-names from one
district into another has taken place only under quite special
circumstances. But it does not necessarily follow that breaking
had taken place in the district in question, for the dialect of this
district may have been influenced by that of a neighbouring
district where breaking had taken place. Especially in districts
where a- and go-forms are both well represented in place-names
or where a-forms are in the majority, the a-forms might well be
the genuine forms of the dialect, the ea forms borrowings from
other districts. However, there can be no doubt but that the
general tendency is for the a-forms to spread southwards and for
the m-forms to be superseded. Such is the case in the spoken
language in general, and of course this tendency has extended
itself also to place-names. For this reason I believe that upon
the whole <?a-forms are indigenous in the districts where they are
found, and that even isolated names (ea-forms of which are well
evidenced) may as a rule be taken to prove that the district in
question belonged to the a-area, whereas a-forms occurring by
30 Eilert Ekwall
the side of ea-forms are mostly due to influence from the dialects
of other districts. Of course I do not deny that occasionally ea-
forms have spread beyond their original domain. We must especially
reckon with the well-known fact of overlapping of dialects in bor-
der districts.
The correctness in the main of this view is corroborated by
the following considerations. The material shows that there are
very few counties where only <?a-forms are found in place-names.
Even in Kent there are isolated a-forms, as Caldecote, found from
1346, and -wold begins to crop up by the side of -weld, -wald in
the former half of the fourteenth century. That e. g. in the
South East a-forms are due to the influence of more northerly
dialects, hardly any one will deny, especially as these a-forms
consist of words or elements which must have been etymologically
absolutely clear. But an examination of the material from the
other counties (Oxfordshire etc.) will show that here too the a-
forms are mostly etymologically perfectly clear or rather must
have been perfectly clear to people in the Middle Ages, whereas
the ea-forins to a great extent consist of words whose etymology
is less evident. A form such as Chalveye (Bucks) would not be
easily associated with chalf 'calf, and would therefore retain Ch-,
even if chalf was supplanted by calf in the dialect. Names con-
taining the personal name Cealf (as Chauson Wore.) would of course
not be liable to any associative influence. Among a-forms we
find particularly distinctive additions such as Cold, Old, placed
before a name (COLD WALTHAM etc.). Many of these were probably
added at a comparatively late period. Frequently the a-forms
occur in names of smaller pieces of land, enclosures, fields or
the like. These are not far removed from common nouns and
would be apt to exchange an earlier ea-torm for the a-form adopted
by the dialect; Calfo, Calurecroft Hunts are cases in point. The
-form has perfectly ousted the ^a-form in the case of CALDECOT
(CALCOTT etc.); even in early sources ea-forms are rare. But this
name is so common - - there are at least 30 different places of
this name in England that one is inclined to believe that it
is in reality an old common noun. Like Cold Harbour it may
have meant 'a place of shelter from the weather for wayfarers,
constructed by the wayside'. Taylor, Words and Places, assumes
a sense like this for the word. If this is right, it is not sur-
Contributions to the History of OE Dialects 31
prising that the earlier Chaldecofe was completely supplanted. But
even if it is not, the connection between this name and the adjec-
tive c(e)ald must have been so strongly felt that the place-name
would be liable to exchange the m-form for the a-form when the
form cald (cold) supplanted chald, cheald in the dialect.
It is also of importance to note that in some cases the ea-
form is found in the earliest sources, the a-form in later sources.
Examples have been given e. g. under OLDCOURT Sussex, STOKE
CHARITY Hants, OLD HEATH Essex, OLD FISHSTREET Middlesex,
CALDECOTT Herts, OLD NEWTON Suffolk and particularly under
Oxfordshire.
The same considerations render it unlikely that a-forms in
such districts should be due to breaking not having taken place
in certain subdialects l .
1 A few notes on the geographical distribution of the names dealt
with in the material, may aptly be appended here. Only counties in which
a-forms prevail or preponderate are considered; for the others see remarks
under Material and infra.
Kent. The ea-forms are evenly distributed: Chalk, Yalding, Yaldham
Weald are in the W. or N.W.; Challock, Ringwould, Womenswould, in the E
- Caldecote is in the W.; Sibertswold, in the E.
Sussex. Chalder is in the S.W.; Chalvington, in the S.E.
Hants. Charlton is in the S.; Stoke Charity and Wield, about the
centre.
Wilts. Chalk is in the 8.; Chalfield. in the W.; Chalvecroft, Highworth
(with Eldelonde\ in the N.
Dorset. Chaldwell is in the W.; Chaldon, in the S.E.
Somerset. Eldebreche is in the S.; Chilton, in the W.; Cholwell House
and Chelvey, in the N.; Kelston and Aldwick, likewise in the N.
Devon. Challacombe is in the extreme N.; Chold Ash and Chawleigh,
in the N. half.
Essex. The places dealt with are evenly distributed. Weald and Pitsea
(with Chaluedone) are in the S.; Chadwell, in the S.W.; Rayne (with Chene-
woldes ealdecote), in the N.W.; Bergholt (with Cheldeberecroft and Cestresivold],
Colchester (with Chalfpitel etc.), Old Heath, in the N.E.
Herts. Chaldeans is in the E.; Caldecote (earlier Chaldecote), in the
N.E.; Sandon (with Chalcroff), in the S.E.
Beds. Chalton and Chalgrave are in the S.W.; Chawston, in the N.E;
Caldecote, in the E.; Caulcott, in the N.; Harrold (doubtful), in the N.W.
As regards the counties of Devon, Somerset, Wilts, Hants, Dorset (and
Sussex, Berks, Oxford) the distribution of forms going back to OE ie (< i-
mutated ea) should also be noticed (see Ch. II). The ea-forms taken together
with the examples of OE ie prove that breaking had taken place in most
32 Eilert Ekwall
We will now proceed to draw our conclusions more in detail.
We may unhesitatingly affirm that all the counties south
of the line Thames Avon Bristol Channel belong to the ea-area
{: Kent, Surrey, Sussex, Hants with the Isle of Wight, Berks,
Wilts, Dorset, Somerset, Devon, Cornwall). The place-names do
not give the slightest support to the view that the breaking was
restricted to the South East. It is true that the material is some-
what scanty in the case of Somerset, but even if we had no
certain examples from this county, we might be sure that it
belonged to the ea-area; in Devon breaking is very well evi-
denced, and Devon and Somerset must belong together linguistic-
ally. It looks, however, as if a-forms had come into use early
in the N.E. part, clearly owing to influence from Gloucestershire.
The following counties north of the line Thames Avon-
Bristol Channel obviously belong to the a-area: Essex, Middle-
sex, Herts, Beds, Oxford. Some counti.es call for a more detailed
discussion.
The northern half of Suffolk belongs to the a-area; in the
southern half isolated ea-forms occur. It would be rash to con-
clude that the boundary line between a and ea is to be drawn
across Suffolk. The ea-forrns consist only of two or three examples
of Eld- for Old-. Obviously these forms may be due to influence
from the Essex dialect. If we are at all entitled to assume ea-
forms to have spread beyond their original domain, we are par-
ticularly justified in doing so in the case of the Eastern counties,
where the broken forms were longer preserved than in any other
part of England. I believe, in spite of the ea-forms found, that
all Suffolk belonged to the a-area. I am strengthened in this
belief by the following consideration. There is in N. Essex a place
Cestreswold Colch. Ch. (13 c.), situated close to the Suffolk border.
It is impossible to doubt that we have an a-form before us, and
the only explanation of its occurrence in Essex at such an early
date, so far as I can see, is influence of the Suffolk dialect. Cestres-
wold is a case of early overlapping of a in a border district.
or practically all the dialects of some counties, as Devon or Dorset. In the
case of counties in which i-mutated ea appears as e and names containing
a-forms are rare, it is of course impossible to demonstrate by the help of
place-names the extent to which breaking had taken place. Cf. p. 38 infra.
Contributions to the History of OE Dialects 33
In Hunts a-forms are in the majority, but ea-forms are also
very well represented. In the S. of the county there is the im-
portant example Weald, which renders it practically certain that
this part of the county belongs to the ea-area. In the northern
half numerous ea-forrns are recorded in the 12th century round
Sawtry, and even in the 14th cent. Eldlode occurs in about the
same neighbourhood. It seems impossible to account for all these
forms by assuming influence from the dialect of Bedfordshire -
the neighbouring counties of Cambridge and Northants are no
doubt rt-districts. In Bedford ea-forms are well evidenced, but
these do not seem to have had a very firm position; a-forms are
found early. This is not to be wondered at in view of the vi-
cinity of Cambridgeshire and Northants. So far as 1 can see,
the place-names show Hunts (at least the greater part of it) to
belong to the en-area. It is of course possible that the parts of
it adjoining Cambridgeshire and Northants are to be reckoned to
the a area, but the a-forms found in the county are all easily
accounted for as due to the influence of the dialects of these
counties. It seems very probable that the ea forms began at an
early period to be supplanted by a-forms, a very reasonable sup-
position in view of the situation of the county. It is worth no-
tice, however, that with the exception of Caldecote DB (where C
may denote the palatalised sound), no a-form has been evidenced
before the 13th century. Waldhirst 11857 may have short a
<C fa. The example quoted from Kemble occurs in an obviously
late copy.
Hulcote (with Chaldenh'ul and Eldebroc) is in S. Northampton,
near the border of Bucks. The two forms are probably cases of
overlapping of the Bucks ea. The other ea-forms quoted from
Northants are all extremely dubious. However, it is possible fur-
ther material will show one or the other to be really examples
of OE ea. If so, they are accounted for without much difficulty.
All the places, except Cdvrecot, which may be disregarded, are
situated near the border of Hunts or Beds, and an ea-form may
be due to influence from the dialects of these counties. If it
should be possible to prove the occurrence of occasional ea-forms
in the parts of Northants adjoining Hunts, this would strongly
support the theory that the latter county belongs to the ea-area.
3
34: Eilert Ekwall
The Buckinghamshire material is unfortunately very scanty;
with the exception of the two Caldecotes in N. Bucks, there is only
one certain ea-form in the S. and one a-form (Calverton) in the
extreme N. of the county. The latter form shows that the ex-
treme N. of the county must at least have been influenced early by
a-dialects (no doubt Northants dialects). The situation of the
county between Oxfordshire, Beds, Herts, Middlesex and Berks
renders it fairly certain that it must belong to the ea-area. This
is corroborated by the occurrence of ea- forms in S. Northants 1 .
In Gloucestershire and Worcestershire there is a peculiar
state pi things as regards a- and ea-forms. In both counties a-
forms greatly preponderate. In Gloucester there are of ea-forms
only Chalford and Chalkwells in the S. and the somewhat doubtful
Eldeberge in the N. In Worcester there is only one ea-form, viz.
Chawson, which must be looked upon as a certain example of
breaking. The form cannot be due to influence from neighbouring
dialects; neither the Gloucester nor the Warwick dialects could
have exercised such influence. But if Worcester belongs to the
m-area, Gloucester must, in my opinion, do so too. The claims
of Worcester to be an old e-district are further strengthened by
the occurrence of an isolated ea-form in S. Shropshire (Elcote).
It cannot be doubted that the main part of the latter county be-
longs to the a-area. The isolated case (or cases, if the nonce-
form Eldebur' 1210 deserves credit) of ea near the Worcester
border must be due to overlapping of Worcester ea.
But the preponderance of a-forms in Gloucester and Wor-
cester proves that these began at a very early period to supplant
the ea-forms in the dialects of these counties. It is quite possible
that this process began even in OE time.
To the a-area obviously belong : the greater part at least of
Suffolk, Norfolk, Northants (at least the greater part), Rutland,
Leicester, Warwick, Stafford, Shropshire (except perhaps the ex-
treme S.), Hereford and all the counties north of these.
1 We have here some interesting examples of blending of dialects in
border districts. Calverton N. Bucks seems to be due to influence from
Northants dialects; inversely Chaldenhull, Eldebroc in S. Northants seem to be
due to influence from Bucks dialects. Similar are the relations between forms
in Eld- in S. Suffolk and Gestreswold in N. Essex.
Contributions to the History of OE Dialects 35
It is generally held that breaking of a before /-groups is a
distinctly Saxon-Kentish phenomenon, whereas a remained intact
in Anglian dialects. It remains to inquire whether this view holds
true, in other words whether there is reason to believe that the
large ea-area we have established may be looked upon as probable
Saxon-Kentish domain from a linguistic point of view. Fortunately,
this is a comparatively easy thing to do thanks to Professor Brandl's
admirable book Zur Geographic der altenglischen Dialekte 1915.
1 here make use only of the first two chapters of the book, in
which Professor Brandl tries to trace the boundary line between
Saxons and Angles by the help of direct statements in early hi-
storical works and similar sources, and of the important indications
offered by the early ecclesiastical division of England.
Professor Brandl holds that on the whole the early dioceses
correspond to the early kingdoms; sometimes, of course, one king-
dom was divided up into two or more dioceses. Man hat durch
das 7. und 8. Jahrhundert niemals Angeln und Sachsen, iiberhaupt
Teile verschiedener Hauptstamme, zu einer Diozese zusanimenge-
legt. The results arrived at by Professor Brandl may be summed
up as follows, in so far as they are of importance for the present
investigation.
The kingdom of Essex, according to OE sources, compre-
hended Essex, Middlesex and the E. part of Herts. These districts
made up the diocese of London. The northern boundary of Essex
and Herts is also the old boundary between Angles and Saxons.
The kingdom of Wessex from an early period comprehended
Hants, Wilts, Dorset and also Surrey and Berks. Out of these
the two bishoprics of Winchester and Sherborne were formed later
on. In the 6th century the West Saxons extended their dominion
over Bedford, Bucks, Oxfordshire and the W. part of Herts, which
later - - with the exception perhaps of Herts formed the diocese
of Dorchester, the northern boundary of which, in Professor Brandl's
opinion, corresponded to the northern boundary of the said coun-
ties. Here we have the old boundary between Saxon and Anglian
country. Western Herts at a later date appears as part of the
archdeaconry of Huntingdon. In the 8th century the whole of
this conquered district was lost to the Mercians, but its inhabitants
were and remained Saxons.
36 Eilert Ekwall
Likewise in the 6th century the West Saxons brought under
their dominion and colonised the country of the Hwiccas, i. e.
Gloucester, Worcester and Warwick, but lost the whole of this
area to the Mercians after about half a century. The northern
part of it seems to have been recolonised by Mercians, whereas
the southern part remained Saxon. The bishopric of Worcester
embraced the main part of Gloucester, Worcester and the S.W. part
of Warwick; it presumably corresponds to the part of the country
of the Hwiccas that remained Saxon. The greater part of War-
wick belonged to the diocese of Lichfield, which also embraced
Stafford etc.
On the kingdoms and bishoprics N. of those discussed we
know much less with certainty. The West Angles held Hereford;
the later bishopric of Hereford embraced besides Hereford a small
part of Gloucester (The Forest of Dene) and S. Shropshire. The
domains of the other Mercian kingdoms are very obscure, and the
southern boundary of the diocese of Leicester cannot be traced
with certainty. Oman (quoted by Brandl) includes in this diocese
Leicester, Northants with Rutland, Hunts, and Cambridge. The
correctness of this view may be doubted as regards details. East
Anglia embraced Norfolk, Suffolk, Cambridge, and probably parts
of adjoining counties, as Northants and Hunts. The earliest East
Anglian diocese was that of Dunwich.
Professor Brandl's book contains an ecclesiastical map of
England in 780, made up by Professor Oman. In this the boundary
between the Saxon and Anglian dioceses is drawn. If we compare
this boundary line with the a/fa line which we can draw by
the help of place-names, we can hardly fail to be struck by a
remarkable correspondence between the two. There are some dis-
crepancies, however, and these are partly not less remarkable.
The chief are as follows:
1. Hunts must be considered to belong at least partly to
the ea-area. This is the only m-district that lies outside the Ken-
tish-Saxon border as drawn by Brandl (and Oman). It would be
very remarkable if we could establish breaking on old Anglian
territory. We must inquire whether it can be looked upon as cer-
tain that Hunts is not an old Saxon district. The earlier history
of this county is obscure. On its population in OE time nothing
is known ; Brandl supposes the eastern part of it to have belonged
Contributions to the History of OE Dialects 37
tc the region of the Gyrwas and East Anglia, but this is hardly
more than a hypothesis '. Oman and Brandl suppose Hunts to
have belonged to an Anglian diocese. The northern boundary of
the Saxon diocese of Dorchester is considered to correspond to that
of Oxford, Buckingham and Bedfordshire. But the original extent
of this old diocese is extremely doubtful. It was combined at
an early period (in the ninth century) with the Mercian dioceses of
Leicester and Lindsey to the diocese of Lincoln. The boundary-
line drawn by Oman and Brandl is therefore founded on the in-
dications offered by the mediaeval archdeaconries, as determined in
the Taxatio Ecclesiastica of 1291. These generally corresponded
to counties. The archdeaconries of Oxford, Buckingham, and Bed-
ford, which we know to have been colonised by West Saxons,
correspond exactly to the counties of these names. This, in Pro-
fessor Brandl's opinion, proves the northern boundary of the said
counties to have been the boundary of the old diocese of Dorchester
and of the early West Saxon territory.
But there is nothing in the ecclesiastical geography to preclude
the possibility or probability of Hunts having been Saxon territory
as well. Rather we might say that what little we know of its
early ecclesiastical position speaks in favour of such .a theory.
It is a very curious fact that Hunts with Western Herts forms
an archdeaconry in the diocese of Lincoln. If we are entitled
to believe that Hunts was an old Saxon district, then the com-
bination with Western Herts would become much less remarkable.
Oman's reasons for reckoning Hunts as part of the old diocese
of Leicester are not conclusive, as hinted at by Professor Brandl.
We know very little about this diocese, and the territory assigned
to it by Oman must be looked upon as to a great extent hypo-
thetical.
My conclusion is that nothing prevents us from assuming
Hunts to have been an old (West) Saxon colony. The fact 4hat
the county belongs to the ra-area tells strongly in favour of the
theory that it really is one.
2. No ea-form has been found in Warwick. This is not
to be wondered at, for as it seems only a small part of the territory
that forms this county remained Saxon. Moreover, there is every
1 The charter of 656 where Huntingdon is mentioned is considered to
be a forgery of the time of Edgar (Earle-Plumraer, Sax. Chr. II, 25).
38 Eilert Ekwall
reason to believe that in Warwick, as in Worcester and Glou-
cester, the m-forms were displaced at an early period by a-forms.
This circumstance is easily accounted for by the political position
of these counties, as briefly indicated above.
The rest of the discrepancies between Oman- Bra ndl's boun-
dary line between Angles and Saxons and the a/ea line are in-
significant. We found isolated m-forms in S. Suffolk, in S. and
possibly E. Northants and in S. Shropshire. These are all easily
accounted for as cases of overlapping of ea in border districts.
The tt-forms found south of the boundary line are no doubt chiefly
to be looked upon as due to comparatively late influence from
Midland dialects. In some cnses (as Cestreswold Essex and Cal-
verton Bucks) early overlapping of a in border districts is to be
reckoned with.
We may state, then, that with the exception of Hunts our
ea-area corresponds remarkably closely to the Saxon-Kentish terri-
tory established by Brandl and Oman. 1 do not hesitate to draw
the following conclusions from the previous discussion:
1. Breaking of a before I -f- cons, is a universal
feature of Saxon and Kentish dialects.
2. Breaking took place only in Saxon and Ken-
tish dialects. Huntingdonshire is therefore to be
looked upon as a Saxon colony.
In early OE texts, both Kentish and Saxon, a is frequently
written before I -f- cons. Some scholars (e. g. Luick Hist. Gr.
146) are inclined to believe that breaking did not take place
at least in all West Saxon subdialects. As already pointed out
(p. 31, foot-note) it is impossible to prove by means of place-
names, that breaking took place in all the subdialects of the
counties referred .to the m-area. On the other hand nothing in
the place-name material supports the theory that it did not.
Breaking is a very early phenomenon; it must have taken place
already before the Anglo-Saxon migration to England. It would
be a very reasonable supposition that it had not taken place in
the dialects of certain Saxon tribes. But in that case we should
expect it to be lacking in the dialects of some comparatively
large Saxon district, say one corresponding to a later county.
We should not expect to find it lacking in isolated subdialects.
But, as the material shows, breaking is evidenced in all the coun-
Contributions to the History of OE dialects 39
ties that are to be looked upon as originally Kentish and Saxon.
I believe the early Kentish and Saxon a is due to influence from
Mercian and Northumbrian orthographical tradition.
A few more remarks may be worth adding.
The results arrived at mostly refer to OE (in some cases
even early OE) time. The place-names in many cases tell us
that in certain dialects breaking had taken place, but they do not
prove that the broken forms (or forms such as chalf, eld) were ac-
tually used say in early ME time. Conclusions as to how long
ea-forms were preserved in originally Saxon and Kentish dialects
must be drawn with much caution from place-name forms. Such
forms as Eldlode (Hunts) c. 1342, Eldehawe Surr. 1312, Eldestole
Hants 1428 may have been used traditionally long after the usual
dialect form had come to be old. On the other hand place-name
forms give important information as to the time when a-forms
began to displace ea-forms. It should be remembered, however,
that the earliest sources must be used with caution; Caldecote in
DB e. g. is not absolutely conclusive. We have found a-forms
very early in some Saxon districts, particularly, of course, in those
near the Anglian border, as in Hunts, Gloucester, Worcester etc.
A tabulation of a-forms will hardly be of much use.
It follows from what has been said that our results must be
used with great caution in trying to determine the dialect of ME
literary documents. An illustrative example is offered by the Owl
and the Nightingale. There can hardly be the leasj doubt but
that this poem was written in Dorsetshire early in the 13th cen-
tury (Breier, Morsbach's Studien 39, p. 157), that is to say in a
district where we should decidedly expect ea-forrns at such an early
period. Yet with the exception of the rhyme iweld : culceweld and
belde 1715 all the forms point to OE a (.cold, folde etc.). There
are even the rhymes iholde : wolde, wolde : iholde, which tell against
the supposition that the poet may have written ihelde and that, the
o-forms mav be due to a scribe.
II. On i-mutation of West Germanic a before
1 followed by a consonant.
A discussion of the history of a before I -\- consonant when
i- mutated forms a natural complement to the preceding study and
also to some extent corroborates the results obtained in it. But
on the whole the place-names containing z-mutated a tell us little
as regards the boundary between Saxon and Anglian territory.
Their chief value is that they throw some light on the develop-
ment in different dialects of OE ie (< i-mutated ea) and that they
help us to determine the area of ce, as the result of -mutation of
W. Germ, a before Z-groups. The opinion generally held is that
ea when ^-mutated became West Saxon ie, Kentish e, whereas
Anglian a before Z-groups became ce, rarely e. Yet e occasionally
occurs also in West Saxon (Biilbring, Ae. El. 175, 179 f., Luick,
Hist. Gr. 188 f., 194 etc.).
The place-name elements that may be taken into considera-
tion here are chiefly the following:
1. OJD fielging, fudging 'fallow' (a rare word).
2. OE wiellj -a, -e (wcdl etc., well etc.) 'well', wielm, wtrlm
'flowing' etc. The latter element is rare.
3. Personal names in JElf- (as Alfred] and JElla, jElle
(probably pet forms of names in .^EV/-).
[4. Patronymics in -ing derived from names containing a
before I -f- cons.].
The elements under 3. and 4. are of little value. Patronymics,
indeed, have been totally disregarded, because in these the mutated
vowel was often supplanted by the non-mutated one. It is often
impossible to decide whether we have to start from a mutated or
a non-mutated form. Names in JElf- have been almost altogether
disregarded. If, as seems probable, this element goes back to Prim-
itive Germ. *alfii- we should expect the regular West Saxon form
Contributions to the History of OE Dialects 41
to have been lelf-, Ilf- etc. No such -forms have been found in
OE, though no doubt there were such, as shown by one of the
names discussed in the material. The West Saxon form jElf-
(JElfred, jElfric etc.) cannot be genuinely West Saxon; it must
have been borrowed from Anglian '. Names of this kind tell us
nothing about the West Saxon development of i-mutated a before
/-groups. But this must render it very doubtful whether names
of this kind can be implicitly trusted when occurring in Anglian
place-names. I have found it safest not to include them in the
material. Another reason for not doing so is the fact that it is
often very difficult or even impossible to distinguish names in
sElf- from names in Eald-, Ealh- and even JEpcl -. Names in jElf-
may even have been confused with and influenced by these. On
this point I refer especially to Zachrisson, Anglo-Norman Influence
p. 107 ff.
Also the names sEUa, JKlle are doubtful. Again we should
expect West Saxon forms in ie, y etc.; no such forms occur, so
far as I know. Yet these would not be so liable to confusion
with other names as those in sEIf-, and I have occasionally quoted
examples containing them, without, however, drawing any con-
clusions from them.
It follows that this chapter is based almost exclusively on
names containing OE wiell etc. This word is an extremely fre
quent element in place-names. However, even this word has its
drawbacks (or at least apparent drawbacks), which must be briefly
discussed.
In the first place it is sometimes assumed that Old English
had an ablaut form welle (< ''tvdlon-) by the side of widla etc.
I do not think there are sufficient reasons for assuming such a
form; yet the possibility must be counted with. Secondly, as -w(i)ell
chiefly occurs as the latter element of compounds, its vowel may
have been weakened in an unstressed position ; -wylle e. g. may
have been weakened to -irelle. This has no doubt often been the
case, but on the other hand the element in question may be as-
sumed frequently to have had a strong secondary stress, and
moreover, the tendency to weakening would be counteracted by
the analogy of the simplex well. The material seems to show that
1 Cf. Bjorkman, Anglia Beiblatt 1914, p. 358.
42 Eilert Ekwall
weakening has not taken place to the extent that might have been
expected. Thirdly, the frequent form -walle may be due to con-
fusion with wall OE weall) or to the influence of the verb
wallen 'to well' (OE weallari). No doubt occasional instances of
-walle may be due to one of these circumstances. Also confusion
with French -mile, common in place and personal names, is to be
reckoned with. As a matter of fact the spelling -wile (or -vile)
occurs occasionally instead of -welle etc. Fourthly, as -wille, -welle
etc. were no doubt associated with the simplex well, names con-
taining this element would be liable to be influenced by the form
of the simplex. A scribe to whom welle was the familiar form
would be apt to substitute -welle for a -wille or -walle heard by
him or found in the original he was copying. If, owing to in-
fluence from some other dialect or the standard language, the
genuine dialect form wille or walle was supplanted by welle } the
place-names in -wille etc. would necessarily come to be influenced
ff
by the new dialect form. This is the chief drawback of the ele-
ment we are discussing and it is a serious drawback, as we shall
see later on.
Besides from monographs on place-names the material of
this chapter has been taken chiefly from the Domesday Book, the
Hundred Rolls, the Feudal Aids and the Nonaruin Inquisitiones.
Occasionally forms (especially forms of particular interest) are quo-
ted from the Ancient Deeds or other sources. In the case of some
counties, as Wilts and Lancashire, I have made use of special
sources, in order to find sufficient material.
The DB forms, however, should be used with caution,
because OE <E, y, i frequently appear as DB e; cf. Stolze, Laut-
, lehre der ae. Ortsnamen im Domesday Book 4, 9, 15.
Material.
Traces of OE ie, i, y are found with certainty in some Saxon
counties, viz. Devon, Somerset, Dorset, Wilts, Hants, Oxfordshire
Sussex, and possibly Berks.
*
Devonshire.
ILFRACOMBE: Alfreincome DB, Affride-, Aufredecumbe Close
Rolls 1233, Alfridecumbe ib. 1234 (Johnston), Hfredecomb FA 1303,
Contributions to the History of OE Dialects 43
llferdecumb ib. 1316, Hfordecombe Or. R. I. 297 (1326), Tlfrycomb
FA 1346, Ilfardy combe Anc. D. IV (1356). Obviously the former
element is OE JElfred or (perhaps rather) a patronymic derived
from it, not with Johnston OE Ealh frith. The variation between
initial A and / is remarkable. So far as I can see we must
assume that Ilfrede- 1303 etc. represent the local pronunciation,
while Alfride- and the like are due to influence from the usual
form of the name, OE Alfred, ME Alfred l .
Wille DB, Wille FA 1284, Welle ib. 1303 (: OE wieUa).
As the latter element of compound names, -wille (-welle) is
extremely common in Devon. I give a fairly full collection of
examples.
ABBOTSKERSWELL: Karswill Abbatis FA 1284, Abbodescarswill
ib. 1316.
BLAKE WELL: Blachewelle DB, Blakewill FA 1303, -wyll 1346.
BRADWELL: Bradewelle DB, -well FA 1303, -wille 1316, -will
1346.
COPPINSWELL: Wille Coffin FA 1284, Coffynys Wille ib. 1316.
COLWELL: Colewille DB, -well FA 1284, 1303, Colwill ib. 1346.
DUNKESWELL: Doducheswelle DB, Dunkewell HR I. 76 (1275),
Dunkeswill FA 1284, Donkyswille ib. 1316.
EDGINSWELL: Wille Eygelf FA 1284, Eggereswill ib. 1303,
EggeryswiU ib. 1316.
GABWELL: Gabbewill FA 1303, Gabwill ib. 1346, Gawyll ib.
1428.
GOOSEWELL: Gosewelle DB, Goseiull (for -uill) FA 1284, -will
ib. 1303 etc.
HALSWILL: ? Hasswell HR I. 91 (1276), Hahwill FA 1303,
1346, Hahewyll ib. 1428, Hahwyll Anc. D. 1, II. (1497, 1502).
HALWELL (Brixton) : ? Hagewile, Halgewelle DB, Halghewell
FA 1284, -wylle ib. 1346, llalghwill ib. 1428.
HALWILL: Halgawill FA 1284, Halgh(e}wiH ib. 1303, 1428.
HOLEWILL: Holeivill FA 1303, 1346, -wyll ib. 1428.
HORSEWELL: Horswill FA 1284, 1303, -tt^/Z ib. 1346, Anc.
D. 1 '(1466).
KERSWELL: Carswill FA 1303, 1346. KERSWILL: Carswille FA
1 This is the only certain example of OK lelf- found. It is just possible
that IL.SINGTON Devon (Ylstyngton FA 1284, Ilstyngton 1303 etc.) might be
another (Aelfstaning- ?). Earlier forms than 1284 not found.
44 Eilert Ekwall
1316. Cf. Carsewelle, Carsvelle DB, Car swill HR I. 71, 89 (1275
f.), -wille I. 72 (1275), -well I. 89 (1276). Of. COAESWELL: Kars-
will FA 1284.
KINGSKERSWELL : Karswill Regis FA 1284, Carswell Eegis ib.
1346, Carssewill Eegis ib. 1428.
LODDISWELL: Lodeswelle HR I. 79, -well ib. 91 (1275 f.),
Lodeswill FA 1284, 1316, Lodeswell Anc. D. I (1397).
MEDWELL: Medweye FA 1303, Medewylle ib. 1346, -will
ib. 1428.
NUTWELL: Noteswille DB, Notcwille, -will HR I. 66, 92 (1275
f.), Notewell FA 1284, -w7Z ib. 1303, wille ib. 1316.
OFFWELL: Offewille DB, OffewellFA 1284, Uffewtllib. 1303.
OGWELL: Wogewille, Wogwel DB. EAST 0.: Estwoghwil, Est
Woggewill, Est Woggewell HR I 846 (1275), .Es* Wogliwill FA
1284, j&tf Wogwille ib. 1316 etc. WEST O.: Wcs< Wogw'dl, West
Woggewill HR. I. 84, 86 (1275), West Wogwill FA 1284, West-
woggewill ib. 1303 etc. O. PEYTEVIN : Wogewill Peytevyn FA. 1303,
Wogwyll P. ib. 1346 etc.
PICK WELL: Wedicheswelle DB, Paddikkeswell FA 1284, Pydeke-
will ib. 1303, 1316, Pydekuill ib. 1346.
Pis WELL: Pissewelle DB, Piseweyll FA 1284, Pysewill ib. 1303,
Pysewill ib. 1346.
RINGWELL: Ryngeswelle HR I. 86 (1275), Eyngeswille FA
1284, -wi'M ib. 1303, Etjngiswyll ib. 1346.
SHIRWELL: Sirewelle DB, Syrewell HR 78 f. (1275), Scirewill,
Schirewell FA 1284, Schyrewill ib. 1303.
SPARKWELL: Sperkwille FA 1284, Sparkewell ib. 1303, -M;^
ib. 1346.
TOTTEWELL: Tottewitt FA 1303, Totewill ib. 1346.
WILLAND(: We/towd DB, La Willelaund FA 1284, Wilklond
ib. 1346) very likely belongs here, but OE WY/Ja may also be
thought of.
Only with e:
WELCOME: TFie/fecome DB, Wellecomb FA 1303, PTe/cwmft
ib. 1316.
In one case there seems to have been a change of the vowel
into u , but there are no early instances of this vowel:
WOOLH ANGER (N.E. Devon) : Welhanger FA 1284, Welhangre
ib. 1303.
Contributions to the History of OE Dialects 45
The material will have shown beyond -doubt that the regular
representative of OE ieea before / -\- cons.) in Devonshire is
i. Besides i we find e, but this is much less common and no
doubt to be explained as due chiefly to scribes who used the form
well(e). Later -ivell (and Welcome) is probably due to influence
from standard pronunciation or from other dialects. The places
in question are distributed all over the county.
There is one apparent example of a, viz. PARTRIDGEWALL
(in the centre of the county): Pertrikes-, Pertrycheswall FA 1303,
1346. I have no doubt the latter element is OE weall 'wall'.
Somersetshire.
The usual form of well is in this county -ivell(e\ but there
are occasional examples also of -wille or the like. The type -walle
is rare and doubtful. Also the type wulle seems to occur. I give
first the names exhibiting forms in -welle, -wille.
BACKWELL: Bacoile DB, Baggewoil (sic) RB 549 (121012),
Bacwell Plac. A. (1203), HR II. 130 (1276), FA 1284, Bacc-
welle ib. 1346.
BANWELL: Ban welle DB, Banewell FA 1316, 1428 etc.
HASWELL: Haseivelle DB, Halswille FA 1284, Halleswelle ib.
1303, Haleswdle ib. 1346 etc.
Milewardiswill (Kingston) Anc. D. I (1357).
RODWELL: Radewelle FA 1303, Radewyll FA 1346, Radewell
ib. 1428, 1431.
Roughwylle (Kingston) Anc. D. \ (1389).
STAWELL: Stowell FA 1284, Stawell ib. 1316 etc.
STOWELL: ? Stawelle DB, Staville, Stawell HR II. 118, 133
(1274, 1276), Stauwelle FA 1284, Stawell ib. 1303.
THRUBWELL: Trubwell FA 1316, Trollevill ib. 1428.
TUXWELL: Tochestvelle DB, Tokeswell FA 1431.
de Wybbewyll FA 1284.
WELLS: Wyllon Kemble 708 (1002-14), Wyllan ib. 716
(9961006), Welle DB, Welles HR II. 118 (1274) etc.
Of the following names there are forms in -walle:
EAST WELLS: Bstwalle FA 1303, Eat Wallye ib. 1346, Est-
well ib. 1428.
HOLWELL: Holewale RB 549 (121012), FA 1431, Holetvalle
ib. 1316. Cf. also CHOLWELL p. 13.
46 Eilert Ekwall
There is one example of a change into | u
WOOLLEY (N.E. Som.): Wilege DB, Wolleye FA 1303, Wellelegh
ib. 1316, Wulleye ib. 1346, Willtty ib. 1428. At least it seems
probable that the former element is OE wiella 'well'.
The places whose names exhibit forms in -wille or the like
are fairly evenly distributed over the county. There is e. g.
Backwell in the N.E., Haswell about the centre, Kingston (with
Milewardisivill and RoughwyVe] in the S.E.
The three cases of -walle are doubtful. Cholwell (N.E. Som.)
is found only once (Cheldewall pers. n.). East Wells seems to be
part of Wells (N.E. Som .). I am not convinced that the latter
element is really well; Estwalle 1303 might mean 'east wall' and
later Eastwell be due to association with Wells and names in
-well. Holwell (S.E. Som.) may also contain the word wall; cf.
Thirlwall (Northumberland). If -walle really stands for -'well in
these words, the a may either be due to influence from the verb
wallen or be borrowed from Gloucestershire dialects. It is difficult
to believe that it could be the regular form of Somerset dialects.
Dorsetshire.
The typical Dorset form of the word well is wolle (wulle),
but in the records welle is somewhat more common ; -wille is rare.
ARMSWELL: Ermyngyswelle FA 1303, Ermyngeswoll 1346,
Ermyngitwill ib. 1428.
ASKERSWELL : Oscherwille DB, Eskereswdle FA 1285, Asters-
will ib. 1346, Askereswell ib. 1428, 1431, IN 1341; de Oskereswill
(Eskereswiil) FA 1285.
BLACKEN WALL: Blachenewell FA 1285, Blachyngwell ib. 1316.
ELWELL: ? Halegrwelle DB, Hallewolle FA 1285, Halewolle
ib. 1428, Hal(e}well ib. 1303, 1346, 1431.
GORWELL: Gorewull FA 1285.
HOLWELL: ? Hellewell HR. I. 101 (1275), Hellewolle FA 1285.
There is another Ho/well, which appears as Holewale FA 1428.
LEWELL: Lcwelle, Liwelle DB, Lywolle FA 1285.
POXWELL: Pochesivdle DB, Pokeswell FA 1285, 1346, -e ib.
1303, Pokeswull IN 1341. *
WARM WELL: Warmewelle, Warmemoille DB, Wermewell FA
1285, Wermwolle, -welle ib. 1303, Warmwull IN 1341.
Contributions to the History of OE Dialects 47
WOOL: Wille DB, Well HR I. 103 (1275), Welle, Wolle
FA 1285.
It seems very likely that the word well forms the first ele-
ment of the following three names:
WOOLBRIDGE: Celebrigg (sic!), Welesbrigg FA 1303, Wolleberg
ib. 1316, Wolbnjgge ib. 1428, Wellesbrigg ib. 1346.
WOOLCOMBE (Melbury Bubb): Wolecumbe, Welecombe FA 1303,
Wellecombe ib. 1316. - - WOOLCOMBE (Bingham): Wollecumbe FA
1285. Cf. Welle-, Wileeome DB.
At least one example of -wolle -wulle has been found for most
names. No doubt this is a genuine Dorset form.
The form -wille is rare except in Askerswell. This place lies
in the W., not far from the Devon border. The other places
mentioned are distributed over the county; Armswell is in the
W., Elwell, Holwell, Lewell, Poxwell, Warrawell in the middle
8., Wool in the S.E. etc.
The late and isolated Holewale may be a bad spelling. -
BESTWALL (Beastewelle DB, Byesteivall FA 1316, Beestwall ib.
1431) no doubt has OE weall for its latter element (perhaps be
eastan wealte 'east of the wall').
Wiltshire.
Names in -well are common. Forms in -well(e) preponderate,
but there are a good many examples of -wulle, -wile.
Boterewelle (near Crudwell) Malmesbury Reg. II. 274.
BrodewuUe (Brokenborough) Malraesbury Reg. I. 315. Chegge-
hemwllesbroke ib. 314. Crudehamwlleslake ib. Cutelwlle ib. 315.
Merewlle ib. 314. All five forms occur in -a ME copy of an
OE charter 1 with modernised forms.
Kadcwolle Plac. A. 12 (1195).
CRUDWELL: Credvelle DB, Credewalle Plac. A, 14, 11 95 (perhaps
a mistake for -wulle), Credewell FA 1316, Cruddewelle IN 1341.
Extremely often mentioned in the Malmesbury Reg., only in the
form -well(e).
CUMBER WELL: Cumbrewelle DB, Comerwell FA 1316, de Cum-
berivell Malmesbury Reg. II. 107, de Covnerwlle ib. 395.
1 Another version is given by Kemble (N:o 460, Eadwig 956, but doubtful
or spurious). The forms are here: BrodewuUe, Chegghemwlles broke, Crudeham
wiles lake, Cutelwlle, Morewelle.
48 Eilert Ekwall
Fulewelle (Sherston) Malmesbury Reg. II. 132. Holewelle ib.
GROUNDWELL: GrendeweUe DB, Grundewell FA 1428.
QUOBWELL: Cuabwelle Malmesbury Reg. II. 83, Knabwell
FA 1428.
WHORWELSDOWN: Wervollesdone HR II. 277 f. (1275), Where-
milestone FA 1316, Whereweldoune ib. 1402. Cf. Th. de Were-
wulle Plac. A. 17 (1195).
There is one example of OE wielm:
EWEN: Ewlma Malmesbury Reg. I. 356 (1156), 383, Ewlme
I. 140 (1284), II. 55; Ewelma I. 349 (12. c.), I. 360 (1191), Ewelme
II. 282 etc., FA 1428.
Cf. also Bokewelle Malmesbury Reg. II. 297, Bradewelle ib.
1. 445, II. 221, Hore(s)welle ib. II. 273, Thickewelle ib. II. 132.
Most of the places whose names exhibit forms in -wulle etc.
are situated in N. Wilts, as those in Broken borough and Ewen; in
W. Wilts are Cumberwell, Whorwelsdown. Also Crudwell, Quob-
well, Ground well and Sherston (with Fulewelle} are in N. and
N.W. Wilts. Certain examples from S. Wilts have not been found.
Oxfordshire.
Here -well is extremely common in place-names. Alexander,
Oxfordshire Place-Names, gives 21 such names with very full
and valuable material. The form of the element is regularly
-well(e] except in OE forms and in Kad(e}w<dle 1196 (CADWELL
S.E. Oxf ), which is presumably due to some special circumstance
(influence from wallen vb or the like). Of EWELME he gives the
forms OE ^wylm 931 and Ewolm* HR II. 31 (1276), which point
to -wylm, otherwise only -welm(e). I have myself found the following
noteworthy forms: Ewelme occurs as Ewnline HR II. 30 (clearly
misread or miswritten for Ewulme}. BRADWELL (Bamptou) is
Bradewll Plac. A. 117 (1241). CHERWELL (River) is Charwylle
HR II. 789, but ChareweW ib. II. 35 etc. These cases will be
discussed further on.
Hampshire.
The regular form of well is well(e). I have found only one
quite certain exception (Wools). For isolated forms like colwullan,
bradeivlle in OE charters, see Williams, Vokale der Tonsilben im
Codex Wintouieusis p. 84. Possibly these are in reality ME forms.
Contributions to the History of OE Dialects 49
WOOLS (Romsey, W. Hants): Welles FA 131 6. (the only early
form found).
WALLOP is doubtful : Wallop(e] DB, FA 1428 etc., but WeUtop
1. R. 1217 (Johnston), Wellop Plac. A. 186 (1272) ; Overe-, Nuther-
>coUop FA 1316, Wollop, Nether Wollop ib. 1346, Wolhop IN 1341.
The former element is possibly OE tviell, exchanged for wall-
owing to some special influence.
MAPLEDERWELL: Mapledrewelle DB, Mapeldurewelle Anc. D.
II. (1223), Mapelderwll ib. IV (1262), Mapeldrewelle HR II. 221
f..(1274), Mapelederewell FA 1316. The form -wll may be due to
clerical error.
BURWELL: Burwelle FA 1316, Borewdl ib. 1346, 1428.
GREWELL: Greywelle Plac. A. 1275, FA 1316.
ITCHINGSWELL : EchetiesweUe FA 1316.
SHORWELL: SoreweUe DB, North- , Suthshorewelle FA 1316,
.Shorewelle IN 1341.
WHERWELL: Wherewelle FA 1316, Wherewell IN 1341.
Cf. also HOLYWELL, NUNWELL, WATCHINGWELL, WHITWELL l .
Sussex.
Roberts, Place-Names of Sussex, gives only forms in -well(e)
of DUDDLESWELL, GRAYLiNGWELL and MiswELL (except meoswille,
-ivylle 772). SHOYSWELL is ShoweW HR II. 216, Shoeswell FA
1316. I have found two instances of -wull, viz. Berewull Percy
Oh. 390, 397 (probably 13 c.), but Berewell 402, 410. The place
seems to have been in W. Sussex.
Berkshire.
Skeat, Berks Place-Names, has only -well(e) except in OE
forms. The element is fairly common. Cf. BRIGHTWELL, COXWELL,
SDNNINGWELL etc. In Chronicon Monasterii de Abiugdon (Chron-
icles and Memorials) a few forms occur which may belong here.
Brittewilla occurs as a personal name II. 26 (1087 1100). The
charters are chiefly Old English, and contain forms such as Ceare-
wyllan (955), Tidewaldes wylle (955), Sunnigwellan (811) etc. But
the copy is late (13 c.), and Cealden wulle (955), cealdan wulle (952)
1 Early forms of names referred to in this way are generally to be
found in HR or FA.
50
Eilert Ekwall
are possibly to be looked upon as containing the early ME form.
For other examples cf. Kiigler, ie und seine Parallelformen im
Ags., p. 69 (with reference).
An a, obviously going back to OE <E (walls <C OE wcella etc.)
occurs in certain Western counties, Hereford, Shropshire, Stafford,
Derby, Cheshire, S. Lancashire, more rarely in Gloucester, Wor-
cester, Warwick, S.W. Yorkshire.
Gloucestershire.
The normal form is -welle; -walle is rare.
CARLS WALL: Crasowel DB, Cassewalle FA 1303, Car&wall r
Karswell ib. 1346. In Newent, close to the border of Hereford-
shire. Baddeley gives Carswell in Newent (evidently = Carls-
wall) and under Karswell a Carls wall in Newent with partially
different early forms. Presumably the same place is meant.
CLEARWELL: Glower-wall, Clowrewalle, Clewer-ivell (Baddeley
without dates). In Forest of Dene, the part of Gloucester be-
longing to Hereford diocese.
Sepeshuswatte (Eastleach, E. Glouc.) Glouc. Ch. I. 274(1263
-84).
SOILWELL: Solewalle 1281, Soilewell (Baddeley). In Forest
of Dene.
.Of names only found with e, the following may be mentioned :
BOXWELL: Boxewelle DB, Boxwella Glouc. Ch. I. 349 (1154
-89), III. 2 (1200), Boxwell FA 1303 etc.
BROADWELL: Bradewelle DB, BradeweU HR I. 178 (1276),
IN 1341 etc.
ROWELL: Eawelle DB, Eowett FA 1316, Rowelle IN 1341 etc.
Cf. CLIMPERWELL, DOWDESWELL, KARSWELL, PINSWELL, Sisr-
WELL, STOWELL, WANSWELL, W T ASHWELL in Baddeley.
Worcestershire.
Only one name exhibiting a has been found:
PEPWELL FARM: de Pepewalle 1275 (two examples Duignan).
In N.W. Worcestershire not far from the Shropshire border.
For forms in -well(e) cf. Duignan s. v. BLACKWELL, CALD-
WELL, HAZLEWELL, HEWELL, KERSEWELL, PERDISWELL. Cf. also
CoMeswett', Colewelle, Hnrjr. -iH R II.
CAMPBELL
COLLECTION
Contributions to the History of OE Dialects 51
Warwickshire.
Forms in a are rare:
Clarissewallefeld (Solihull, N.W. Warw.) Anc. D. Ill (1352).
SHOTSWELL (S.E. Warw.): Sotesivalle, -well 1123 (Duiguan,
who gives several examples of -welle), Scotestvell FA 1316, Shottes-
well, -woll (no doubt a clerical error) ib. 1428; several examples
of -well Anc. D. IV.
BERKSWELL (N. Warw.): Berchewelle DB, Berkeswelle FA 1316,
-well IN 1341 etc.
HOREWELL (centre of. county): Horewell 13 c. (Duignan).
WALMLEY is derived by Duignan from OE weelm, but his
only example is Warmleye a 1300 (no doubt OE Wearm-).
Herefordshire.
The typical form of the word well is watte.
Walleforlonge (Hope Mansell) Glouc. Ch. I. 338 (126384).
Former element perhaps OE wcella.
COLWALL: Colewelle DB, Colewille HR I. 186 (1275), Colewall
FA 1316, Colwall IN 1341, Colewalle FA 1428.
CRASSWALL: Crassewalle HR II. 69 (1255).
Dryebrookesivalle (Hope Mansell) Glouc. Ch. III. 236 (1338),
Dribrokeswalte ib. 239 (1339).
DEWSALL: Deiveswall IN 1341, FA 1428.
ECCLESWALL: Ecleswelle HR I. 186 (1275), Ekleswall FA 1303,
Ecdeswall ib. 1346, Anc. D. Ill (1394), Ekeswall FA 1428, Egglis-
wall ib. 1431.
There can hardly be any doubt that the rare forms -welle
are either early spellings with e for & or else introduced by a
scribe who used the form -welle himself. The isolated form -wille
is most probably a clerical error (for -welle).
Shropshire.
There are very few certain examples.
ASTOL: Estwalle HR II. 70 (1255), Astewall FA 1284, Ast-
walle Anc. D. II (1417).
CHATWALL: Chatewalle HR II. 71 (1255), Chatwalle, Chate-
walle Anc. D. II (1417).
WestwaUe HR II. 70 (1255).
52 Eilert Ekwall
WALL: Watte DB, HR II. 71 (1255). Doubtful. It may
be OE w(e)att.
ELLERDINE (Elleurdine DB) seems to contain OE Ella
(MXa). Le Blakewall, la Coldewall Or. R. I. 112 (1300) are in
Shropshire or Staffordshire.
Staffordshire.
FALLINGS HEATH: Olde Falinge a 1200 (Johnston. Not 'fel-
ling of trees', but OE f edging}.
BRADWELL: Bradewalle Anc. D. V (1310).
CAVERSWALL: Cavreswelle DB, Cavereswall FA 1316, Cares-
welle IN 1341; as a pers. n. Cavereswall FA 1284, 1316, Caverswell
ib. 1284.
Codeuuale DB.
CRESSWALL: Cressvale DB, Crestewall FA 1284, Crassewall
ib. 1316, Creswatt ib. 1428.
FRADSWELL: Frodeswelle DB, Frotheswelle, Frodeswall a 1300
(Johnston), Frodeswell FA 1284.
FAREWELL: Farwalle IN 1341, Farewall FA 1428.
HOPWELL: Hoppewell FA 1284.
Walschemonesfold Anc. D. I. p. 457 (1333). OE walisc.
But WALL (Le Wall FA) is OE weall 'wall'.
No doubt -watte is the genuine form ; -welle is to be judged
of in the same way as welle in Hereford.
Derbyshire.
Gf. Walker, The Place-Names of Derbyshire, who gives very
full and valuable material. The examples below are mostly taken
from my own collections.
BAKEWELL: Badequella DB, Bathekwalle 1254 (Walker), ' Bauke-
wett HR I. 58 (1276), Batltequall ib. II. 291 (1275), Bathewell ib.
297, Bakewall 1481 (Walker).
BLACKWELL: Blacheuuelle DB, Blacwell FA 1284 etc.; Bla(c}k-
wall 1339 etc. (Walker).
BRADWELL: Bradewelle DB, HR II. 287 (1275), Bradewalle
1339 etc. (Walker), Bradwall FA 1431, Anc. D. VI (1469).
CALDWELL: Caldewelle DB, Caldewell HR I. 58 (1276), Calde-
walla, Caldewalle etc. 11001486 (Walker), Caldewall FA 1431,
Anc. D. VI (1459).
Contributions to the History of OE Dialects 53
ETWALL: Etewelle DB, EtewalV HR II. 293 (1275), Etewelle
FA 1302 etc., Etwall ib. 1428, 1431.
GLAPWELL: Glapewelle DB, Glappwell HR I. 60(1276), Glap-
well FA 1284, 1346, 1481.
HOPWELL: Opeuuelle DB, Hopwell FA 1302, 1428, 1431.
TIDESWELL: Tidesuuelle DB, Tideswall 1216 etc. (Walker),
Ticks-, Tudiswelle HR I. 60 (1276), II. 287 (1275), Tyddezwalle
Anc. D. I. 1402, Tyddeswell FA 1431.
Cf. CRESWELL, HOLY WELL, WHATSTANDWELL, WHITWELL
(Walker) and Chillewall HR I. 58 (1276), Herdwyckewalle (Worm-
hill) Anc. 1). I. 1402.
ALSOP: Elleshope DB, Alsop FA 1431 etc. (OE JEUe). ELTON:
mtune DB, Elton FA 1431 etc. (OE &lld).
Blackwolle 1355 (Walker) is no doubt a scribal error. Clape-
vill 1282 (Walker) is due to influence from Fr. -ville. Cresseivyle
HR II. 289 is used as a personal name; the isolated spelling
may be disregarded.
There is much vacillation between -welle and -walle; -welle
is more common. Yet in some names -walle is very well evi-
denced, as Bradwell, Caldwell, Etwall, Tideswell. It is noteworthy
that all these places are in the W. part of Derby. No examples
of -walle are given by Walker (or have been found by me) in
the case of Glapwell, Hopwell; of Cresswell and Whitwell one
form in -walle has been found (Cresswall 1549!, Wytwalle 1272
1307). All four names denote places in E. Derby. There are
two places named Blackwell, one in E. Derby and one in W.
Derby; Walker does not keep the two apart. We must conclude
from the material that -walle is at least much less common in
E. than in W. Derby.
Cheshire.
The material is very scanty, because no returns from this
county are found either in the HR, the FA, or the IN.
HASSAL (W. Ches.): Haselwall Auc. D. I (1480 etc.).
HESWALL (W. Ches.): ? Eswette DB, Haselwall Anc. D. Ill
(1309), Hasilwall ib. VI (1325), Hasulwall ib. Ill (1430), Haselivelle
ib. (1317).
WIRSWALL: Wireswelle, -uelle DB.
54 Eilert Ekwall
No early forms are available for BRADWALL. But THELWALL
contains OE weall (: 'wall made of planks').
Lancashire.
For this county the material at my disposal is very full.
In Lane, south of the Ribble a preponderates ; north of that line
I have only found e (welle etc.). I give first examples from S.
Lane.
ALSTON (near the Ribble): Alston Pipe R. 1226, L.F. 1313,
1373 etc. (: probably JElles tun).
FALINGE (Salford H.): ffalenges etc. Whalley Ch.
FALLING (W. Derby H.): Falinge Cockersand Ch. 665 (1190
-1212).
ASPINALL (W. Derby H.): Aspinwalle Scarisbrick Ch. 1280
etc. (often), rarely -welle.
CHILDWALL (W. Derby H.): Cildeuuelle DB, Chillewelle Pipe
R. 1177, -well La. Inq. 1302 etc. Usually -walle, as Childewalle
La. Inq. 1212, Lay Subs. 1332, L.F. 1376, -watt IN 1341, Child-
wall L.F. 1423 etc.
HALLIWELL (Salford H.): Halitvdll(e) Cockersand Ch. 698 ff.
(1190 1220 etc.), -wall Lay Subs. 1327 etc.; sometimes -welle,
as Hall-well Ass. R. 1246, -welle Lay Subs. 1332.
WINEWALL (Blackburn H.): Wyneicelle Lacy Comp. 1294 6,
Wyntvell La. Inq. 1323 4; no early forms in -wall found, but
examples of the name are rare.
WISWALL (Blackburn H.): Wisewall La. luq. 1311, L.F. 1331,
FA 1346 etc.; -well is rare: Wisewell L.F. 1207 etc.
Only e is found in CROMWELLBOTTOM, but here weakening
of a has probably taken place in the weak syllable. Similarly
the absence of -wall in the early forms of IRWELL (: It-well Ass.
R. 1246, Urwil Cockersand Ch. 709 etc.) is doubtless due to in-
fluence from IRLAM, a village situated on the Irwell (: Irwelham
Pat. R. 1259, Irrewilham Ass. R. 1277, Urwilham Cockers. Ch.
719 f. (118490) etc.).
Examples from N. Lane, are not very numerous. 1 mention :
ELLEL: Ellhale DB, Elhale L.F. 1208, Ass. R. 1277 etc.
la).
COLLOWAT: Collingeswdle Cockersand Ch. 813(1199 1206).
HAWKSWELL: Hokes-, Haukeswelle La. Ch. 182, 215.
Contributions to the History of OE Dialects 56.
Keldbrekewelle Cockersand Ch. 106 (1206-46). Sledwelle
ib. 801 (11901220).
Cf. also Wyld and Hirst, Place Names of Lancashire, and
Sephton, Handbook of Lancashire Place Names.
On S.W. Yorkshire see infra, p. 58.
In the remaining counties practically only forms in e (welle
etc.) have been found. Forms with a (-ivalle etc.) are only oc-
casional (except in S.W. Yorkshire) and may be explained as
due to clerical error or some special circumstance. It is hardly
necessary to give very full material. I begin with the remaining
Saxon counties.
Buckinghamshire.
BRADWELL: Smdeuuelle DB, Bradewell HR I. 30 (1255), IN
1341 etc.
CRANWELL: Crendeivelle FA 1302, Crandewelle ib. 1316.
HARTWELL: Herdeuuelle DB, Hurttrelle FA 1284, Herttvelle
ib. 1302 etc.
Cf. BRITWELL, FULWELL. Exceptional is Dewalle Anc. D.
I (1342).
Bedfordshire.
Several examples are given by Skeat: HOLWELL, ICKWELL,
RADWELL etc. Only -well(e) found. Cf. CALDWELL p. 18.
Huntingdonshire.
HOLYWELL: Haliewelle DB, Haliwell HR II. 603 (1281), IN
1341 etc. Numerous examples in Ramsey Ch., as Flaxlandewelle
I. 307, Eedwelle II. 329, Toftwelle I. 307, Wysewelle I. 353. The
only exception found is Bricstaneswlle III. 283 (no doubt a clerical
error for -welle). Cf. also Werkwelle HR I. 621 (1281), Anc. D.
I (1352) etc.
Hertfordshire.
Names in -well are fairly numerous. Only -well(e) found.
AMWELL: Emmetvelle DB, Emwelle FA 1303, Amewell ib. 1402.
ASH WELL: Esce-, Asceuuelle DB, Asswell FA 1303 etc.
DIGSWELL: Dichelesuuelle DB, Digen-, Dygenes-well HR I.
188, 192 (1275), Digonestvell FA 1303, Digneswell ib. 1346.
56 Eilert Ekwall
Of. BRIGHTWELL, HORWELL, SOPWELL.
Several examples are found in Gesfca Abbatum Mon. S:i
Albani (Chronicles and Memorials) passim. Only -welle. Syentries-
walle III. 186 is explained as quodam antique muro terreo
(OE weall).
Surrey.
CAMBERWELL: Cambrewelle DB, Gamer-well FA 1316, 1428 etc.
STOCKWELL: Stokewelle FA 1316.
EWELL: Ewell FA 1428.
Middlesex.
CLERKENWELL: Clerkenwell FA 1428 etc.
HANWELL: Hanewelle DB, Ranwell FA 1428.
STANWELL : Stanwelle DB, Stanewell FA 1353.
Cf. also CHELSEY p. 16 and 28!
Essex.
Names in -well are fairly numerous.
BRADWELL: Bradewelle FA 1303, Bradwell ib. 1428 etc.
FRESHWELL: Frossewella DB, Frosshwell FA 1346.
PRITTLEWELL : Pritteuuella DB, Pritelewell FA 1303, P-ritel-
well ib. 1346 etc.
RUNWELL: Runeuuella DB, Ronewelle IN 1341 etc.
CHADWELL see p. 15. Cf. CHIGWELL, HAWKWELL, RIDGEWELL.
Johnston gives for Bradwell a form Bradwall a 1300 without
reference; I cannot judge of the form. Cheldewalle-lane Dugdale
IV, 603 (1405) is hardly trustworthy.
Kent.
BUCKWELL: Bokewell FA 1284, Bukwell ib. 1346.
EASTWELL: Estwelle DB, Estwell FA 1284, Eastwellib. 1846 etc.
EWELL: Ewette DB, HR I. 209 (1275), Ewell FA 1284 etc.
Cf. HAWKSWELL, WESTWELL.
An exceptional case is Welmwalle RB 751, perhaps merely
miswritten. WALMESTONE is Wyelmeston FA 1346, 1431; presumably
OE Wighelm.
We pass on to the remaining Anglian counties.
' *
Contributions to the History of OE Dialects 57
Northamptonshire.
There are numerous names in -well, e. g.
ASTWELL: Estwelle DB, AstweM HR II. 6 (1276) etc.
HARTWELL: Hertewelle DB, Hert(e}well HR II. 11 (1276),
Hertwell FA 1284 etc.
MAIDWELL: Medewelle DB, Maidewetl FA 1284 etc.
SYWELL: Sewelle DB, Siwell FA 1316 etc.
Only -welle found. Cf. also BARNWELL, CASWELL, HOLLO-
WELL, NOBWELT., ROTHWELL, SCALDWELL, YARWELL. An uncertain
example of OE ce is WANSPORD: Welmesford 657 (Johnston),
Walmesford FA 1284. Perhaps OE welm, wcelm. But if so, we
have rather to assume remodelling of Welmesford under the in-
fluence of .wallen vb.
Leicestershire.
Only -welle found; isolated -woll is no doubt a clerical error.
BARWELL: Barewelle DB, Barewell FA 1284 etc., Barewoll
Anc. D. VI (1290).
SHAWELL: Sawelle DB, Schathewell HR I. 239 (1276), Showell
FA 1284, Shathewell ib. 1428.
Cf. BlTTESWELL, CHADWELL, EA8TWELL, HOLWELL, PlCKWELL.
Nottinghamshire.
There are numerous names containing the word -well. The
early forms have regularly -well(e); cf. Mutschmann under BUL-
WELL, CHILWELL, CROMWELL, WELHAM etc. An apparent exception
is TROWELL: Trowalle, Torwalle DB etc. Mutschmann derives the
latter element from OE weall 'wall' or ON vpllr. Clearly the
name does not contain the word well.
Lincolnshire.
Names in -well numerous. Only -well(e) found. Examples:
CR AN WELL: Cranewelle DB, Cranewell FA 1303 etc.
HARPSWELL: Herpeswelle DB, Herpeswell FA 1303 etc.
HOWELL: Huuelle DB, Howell FA 1303 etc.
TATHWELL: Tadewelle DB, Tathewell FA 1303 etc.
Cf. further BRAUNCEWELL, BURWELL, CALKWELL, GOKEWELL,
GREETWELL, HEMSWELL, HOLYWELL, MAIDENWELL, ROTHWELL.
WALMSGATE: Walmesgare FA 1316 is of doubtful etymology
(perhaps OE Wealdhelm-}.
58 Eilert Ekwall
Cambridgeshire.
BUEWELL: Burewelle DB, BureweW HR I. 54 (1276), Bure-
welle FA 1302, Burwelle ib. 1316, Borewell IN 1341.
KNAPWELL: Chenepewelle DB, Knapwell HR I. 53 (1276) etc.
WHITWELL: Witeuuelle DB, Wyten-eW HR II. 563 (1279),
well FA 1316.
Cf. BARNWELL, ORWELL, OUTWELL, SNAILWELL (several forms
in Johnston) and Fulburnewelle HR II. 445 (1279).
Norfolk.
Names in -ivell are common. Examples:
BAWDSWELL: Baldereswella DB, Baldeswell FA 1316.
BEXWELL: Becheswella DB, Belceswelle FA 1302, 1316 etc.
OUTWELL: Utuuella DB, Utwelle FA 1316.
WELLS: Wella DB, Welles FA 1302, 1316 etc.
Cf. ASHWELL, BEECHAMWELL, BUNWELL, FELTWELL, TITCH-
WELL, UPWELL.
Suffolk.
Skeat gives several names in -well, as BARDWELL, BRADWELL,
BRIGHTWELL, all only with forms in -welle. I have myself noted
three isolated instances of -wall(e) : Brighteivall FA 1316, Calde-
walle Anc. D. I (1296), Kentewalle ib. II (but also -welle ib., -well
FA 1401). These examples are too few, in comparison with the
forms in -welle, to render OE tvcetta probable for Suffolk.
Yorkshire.
In S.W. Yorkshire there are a few examples of -wall(e); cf.
Goodall, Place Names of S.W. Yorkshire:
CHURWELL: Cherlewall 1226, Chor(e)lwell 1296 etc. (Goodall).
HELIWELL: Heliwall 1297, Heliwelle 1373 (Goodall).
CHISLEY: Chesewaldeley 1296, Chesewelley 1307, Cheswalleye
1308, Chesewalleye 1309. (Goodall).
Only -well(e) is recorded in HOLYWELL, OAKWELL, SPINK-
WELL. Moorman, W. Riding Flace-Names, has only forms in
-well of BRACEWELL, BRAITHWELL, KETTLEWELL, ROTHWELL, SHAD-
WELL, WOMBWELL. On BRAITHWELL cf. also Lindkvist, ME Place-
Names of Scand. Origin, p. 27, where numerous examples are
given. Cf. also Welle DB, Well IN 1341.
Contributions to the History of OE Dialects 59
From the East and North Ridings comparatively few examples
are available. A few forms are taken from The Visitations of
1409 ff. in Surtees Soc. Vol. 127 (Vis. 1409 etc.). Cf. also KELK,
KELEIELD (E. Rid.) p. 28.
HARSWELL (E. Rid.): HorsenweW Vis. 1409.
HAUXWELL (N. Rid.): Havocswelle DB, Haukesicell IN 1341,
HawkeswelV Vis. 1428.
HINDERWELL (N. Rid.): Hildencell Percy Ch. 150 (1325), Vis.
1410. Of. numerous examples in Lindkvist, op. cit. p. 10.
RIMSWELL (E. Rid.): Rimesvvelle DB, Rymeswell Anc. D. V
(1431). .
WHITWELL (N. Rid.): Witevella DB, Whytewell Anc. D. V
(1252), Whitewell Percy Ch. 7 (1260).
WELHAM (E. Rid.): Wellun DB, Wellom HR I. 120 (1276),
Percy Ch. 190 (13267).
Durham.
A few examples are taken from the Durham Assize Rolls
(Surtees Soc. 127). Durham is not represented in FA, HR, IN.
BLACKWELL: Blalcewelle Durh. A. R. 35 (1242), Blacwelle ib.
87 (12356).
CHOPWELL: Chepwelle Durh. A. R. 72 (1242).
WHITEWELL: Wytewell, Whitewelle Durh. A. R. 22 ff. (1242).
Cf. Hesseivelle, Essewelle Durh. A. R. 33 f. (1242).
Northumberland.
BYWELL: Bywell FA 1346, 1428, Or. R. II. 349 f. (1377).
TRANWELL: Tranewell HR II. 23 (1275), Trenwell FA 142H.
Cf. COLDWELL, CRESSWELL, HAWKWELL, HOLYWELL, and Ber
tewell Percy Ch. 263 etc.
BENWELL (Bynnewalle a. 1100 Johnston, Benwell FA 1346,
1428) is no doubt correctly explained by Johnston as meaning
'within the (Roman) wall'. Also THIRLWALL (ThirlewalV Or. R.
II. 12, 1327) no doubt contains OE w(e}all. The place is situated
near Hadrian's wall. An isolated example of -wall for -well is
Colewall Or. R. II. 12 (1327), but -well ib. 113 (1337), Colwell
FA 1346.
No early examples from Cumberland arid Westmoreland
are given by SedgeHeld, Place-Names of Cumberland and West-
moreland.
60 Eilert Ekwall
Results \
On old West Saxon territory we have found chiefly three
forms of the word well : -wille, -wulle (-wolle), and -ivelle. The
former two go back to OE wiella (wielle).
The form -wille has been found almost only in Devon and
Somerset. For the personal name Cholewyll in Cornwall see p.
14. In Devon and Somerset -wille is obviously the genuine re-
presentative of OE wiella etc. OE y ^-mutated u) appears in
Devon and Somerset place-names as u or i. OE ie (<C ^-mutated
ea before I -f- cons.) very rarely becomes u (Woolhanger Devon,
Woolley Sorn.), and this u is very likely due to late rounding of
i between w and I. We may therefore conclude that i in -wille
does not go back to OE y (< ie), but to OE i ie}. In the
dialects of Devon and Somerset OE ie must have developed into
an j-sound, obviously the sound denoted by the frequent OE
spelling i (fillan 'to fell' etc.). One explanation of these OE i-
spellings may be that they represent Somerset and Devon forms.
In the westernmost part of Dorset -wille was also used ;
otherwise the typical Dorset form is -wulle (-wolle), i. e. a form
with u . This form is common also in Wilts, and there are
traces of it in Hants. It must go back to OE wylla (wylle). In
the dialects of these counties OE ie became y at least after w.
The place-names tell us nothing about the development of y in
other positions.
In all these counties -welle occurs by the side of -wille and
-wulle. It is not likely that both -welle and -wille or -welle and
-wulle were genuine dialect forms of the same district. To some
extent welle may be due to scribes to whom this form was familiar;
this is particularly obvious in the case of Devon and probably
also of Dorset. On DB -welle cf. also p. 42. But this is not
sufficient to explain the frequent occurrence of -welle. It is a
fact that -welle is more common in later than in earlier sources.
Thus it is the regular Wilts form in later sources ; all the examples
of -wulle have been found in early sources. This indicates that
1 These are based only on ME forms. I have not quoted many OE
forms, but some are. given in the monographs referred to. Cf. also Kugler,
ie und seine Parallelformen im Ags., p. 66 ff. An OE spelling -wylle in a
charter need not prove that this was the genuine dialect form of the district
in question; -wylle may be due to influence from Standard West Saxon.
Contributions to the History of OE Dialects 61
iii districts where -welle occurs by the side of -wille or of -wulle
the latter is the earlier form. The form -welle may here be due
to weakening of the vowel (/, u) in an unstressed position, or it
may have been introduced from districts in which it was the
genuine dialect form (OE wella etc. < *wallian- etc.). I think
it is safest not to operate with an OE ablaut form (*welle <C
Weakening of the vowel (, M) in an unstressed syllable has
hardly played an important part. It is difficult to believe that i
was more liable to weakening in Somerset than in Devon, or u
more in Wilts than in Dorset. But that the spread of the form
welle is partly due to this circumstance, I readily admit. In
Wilts earlier Eivulme is supplanted by Ewelme. As ME welm
was a rare word and can hardly have caused the change, it is
most natural to explain Ewelme as a later development of Ewulme.
Similarly -wulle, -wille may occasionally have become -welle.
But chiefly the spread of -welle is, in my opinion, due to
influence from dialects in which it was the regular form. Just
as chalf, chald, eld were supplanted by Midland calf, cold, old,
the genuine forms wille, wulle were supplanted in Devon etc.
dialects by welle, and this form soon began to influence place-
names containing this element.
My conclusion is that wille is to be looked upon as the
genuine dialect form of Devon and Somerset, wulle as that of
Dorset, Wilts, and probably Hants.
Traces of OE ie have been found in Oxfordshire (Ewolin
etc., Charwylle). In view of the great number of names con-
taining the element well and the very full material available these
traces are extremely faint, and we must conclude either that a
genuine Oxford -wille (or -wulle} was supplanted early by a -welle
introduced from neighbouring dialects or that welle was the
genuine form of some Oxfordshire dialects.
The traces of OE ie found in Berks are still fainter, but
the situation of the county between Wilts, Hants and Oxford-
shire leads us to expect OE ie as i-mutation of ea in its dialects.
It is very interesting to find examples of -wull in W. Sussex;
they prove that a form wylla (wylle) was used in Sussex dialects.
Verv likelv the form was restricted to W. Sussex.
62 Eilert Ekwall
In the remaining West Saxon districts we have only found
-welle (and occasionally -walle, on which see infra).
The usual form in Gloucester and Worcester is -welle. One
is tempted to conclude that in the dialects of these districts e
was the regular OE representative of i- mutated a before / -(- cons.
But in certain ME texts which seem to belong to Gloucester (as
the Chronicle of Robert of Gloucester) forms such as fulle, uldere,
widest (for felle, elder, eldest) occur, which must make us cautious.
It is quite possible that wulle was the genuine Gloucester (and
Worcester) form, which was, however, displaced at an early period
by Midland welle.
In the Eastern West Saxon dialects (in Bucks, Herts, Beds,
Hunts, Surrey), in East Sussex, in the East Saxon dialects (in
Essex and Middlesex) the regular form is -ivelle. The results of
the preceding examination must of course make us cautious in
drawing conclusions from the material at hand, and I should not
be surprised if isolated instances of -wille or -wulle should turn
up in early sources not accessible to me. The following con-
sideration, however, makes me inclined to believe that the regular
OE representative of ^-mutated ea in the dialects of these districts
was e. The common occurrence of -welle in Hants and probably
Oxford and Berks is no doubt to a great extent due to the in-
flujence of the dialects of neighbouring counties-. Welle should
therefore have been the genuine dialect form in the counties
east of these; from Kentish dialects alone this influence cannot
have emanated. If we assume that e was the regular OE re-
presentative of a-mutated ea in e. g. E. Sussex, Surrey, Middlesex,
Beds, Herts, the early spread of -welle in Hants etc. is easily
accounted for. The not infrequent West Saxon e as ^'-mutation
of ea may be an Eastern West Saxon development.
In Kent -welle is of course the regular form.
In several West Midland counties, viz. Herefordshire, Shrop-
shire, Staffordshire, (probably) Cheshire, and S. Lancashire, the
typical representative of ^-mutated a before I -\- cons, is a, ob-
viously going back to OE ce (walle < OE walla etc.). The not
infrequent e-forms (-welle etc.) are no doubt due partly to scribes
to whom these were familiar, partly to influence from E. Midland
dialects and the standard language. In DB e doubtless partly
Contributions to the History of OE Dialects 63
denotes | se | . The form -ivall(e) is common also in W. Derby-
shire, very rare in the E. part of this county. It is possible
that the boundary line between welle and walle once ran right
through Derbyshire, but 1 consider it more likely that -walle is
the genuine dialect form of all Derbyshire, though it was at an
early period practically supplanted in the E. part by -u-elle.
We also found -icalle in the originally West Saxon counties
of Gloucester, Worcester, and Warwick. Two certain examples
of -walle were found in the Forest of Dene, the part of Gloucester
belonging to the diocese of Hereford; the dialect may here well
have been nearly related to that of Herefordshire. Another cer-
tain example was found in W. Gloucester, close to the Hereford
border (Carlswall). Further, one isolated spelling was found in
E. Gloucester. One Worcestershire name, denoting a place situated
not far from the Shropshire border, has forms in -walle. Also
two Warwickshire names exhibit isolated forms in -icalle.
Clearly walle is the typical form of the W. Midland counties
mentioned above. This form has spread southwards over War-
wick, Worcester and Gloucester, though welle is here more com-
mon. From Gloucester iralle may have spread also to Somerset.
The examples of -wall(e) in S.W. Yorkshire may be due to
comparatively late influence from Lancashire or Derby dialects;
but they may also point to early Mercian colonisation of this
district l .
In the rest of the Anglian (The East Midland and Northern)
counties, the regular representative of /-mutated a before I -f-
cous. is e (welle). Quite isolated spellings -walle may be dis-
regarded.
It is extremely interesting to note that the area in which
a is the regular representative of t- mutated a before Z-groups is
almost exactly conterminous with the two dioceses of Hereford
and Lichfield (cf. the ecclesiastical map in Brandl's Geographic
der ae. Dialekte); it is true the traces of -tvalle in Warwickshire
1 Cf. on this point Brandl's Geographie der ae. Dialekte, p. 14 f. where
the early history of the West Riding is discussed. Brandl points out that
at least part of this district was occupied by Mercians under Penda in the
seventh century and that quite possibly some Northumbrian border districts
(\\7,. the West Riding) were recolonised by Mercians. This theory receives
some support from the occurrence of forms in -icalle in S.W. Yorkshire.
64 Eilert Ekwall
are faint. Further the boundary between walle and welle tallies
with that between u and i <C OE y as regards the N. Midlands,
for according to Brandl, op. cit. p. 75, Nottingham belonged to
the *-area, Derbyshire to the w-area. In how far the boundary
between walle and welle may represent an old tribal boundary,
is a question which I have not sufficient material to enter into.
The results obtained as regards Anglian dialects, however,
are in conflict with those obtained by a study of OE texts and
with the rules given in OE grammars. In OE texts fmutated
a before I -\- cons, as a rule appears as ce in Anglian texts, not
only in those that may be looked upon as West Mercian, as the
West Mercian Psalter. On the other hand our results are cor-
roborated by ME dialects and the standard language, for in the
latter as well as in East Midland and Northern dialects the re-
gular forms are felle, elder, eldest etc.
It is not sufficient to point to the rare OE e that appears
by the side of ce in certain OE texts (cf. p. 34). If ce was prac-
tically the regular Anglian form, we should at least expect nu-
merous forms in a (as walle etc.) in place-names and ME texts
from East Midland and Northern districts. I see no other way
out of this difficulty than the following.
OE ce before / -f- cons, does not denote exactly the same
sound in West Mercian and in other Anglian texts. West Mercian
<K denotes a real | se , which became ME a. In other Anglian
texts & denotes an open e-souud, something like e i, which was
distinct from e in set and ce. in heett, yet sufficiently like the
latter to be usually written ce; this open e was later on merged
in the e-sound pronounced in OE bed, fell. Occasionally this
open e-sound was written e (OE wella, fellan etc.).
How this different development is to be explained, is un-
certain. It may be Primitive English a before I -\- cons, when
e- mutated became a closer sound in Northumbrian and East Mer-
cian than in West Mercian dialects. Or West Germanic a before
I -\- cons, may have remained as Primitive English a in West
Mercian also when an i (j) followed in the next syllable, and
become ce by i-mutation; whereas in other Anglian dialects, when
followed by i (j), it became Primitive English ce and through i-
mutatiou the open e-sound suggested above. Cf. the explanation
Contributions to the History of OE Dialects 65
proposed by Morsbach for occasional OE fellan etc. by the side
of fcellan (Biilbring, Ae. Elernentarbuch 175, anm.). But whatever
may be the explanation of the different development, it is neces-
sary to assume that this dialectal difference existed in OE time.
We may sum up the results of the present chapter
as follows.
1. Saxon ea (<^a before Z-f-cons.) when /-mutated
became ie in some West and South Saxon dialects.
This ie became:
a. OE i, ME i in Devon and Somerset;
b. After w OE ?/, ME u (o) in Dorset, Wilts, Hants,
partly at least in Oxford and W. Sussex, probably
also in Berks.
2. In Kentish, probably in East Saxon, in East
Sussex, perhaps also in the Eastern West Saxon dia-
lects en was i- mutated to e\
3. Anglian a before I -f- cons, when i-mutated
developed differently in West Mercian and other
Anglian dialects; it became:
a. In the former OE a, ME ;
b. In the latter OE a [s], ME e.
Addendum.
Page 8. On e for a (and a for e] in early records (DB etc.) see now
also Zachrisson, Studier i Modern Sprakvetenskap utg. af Nyfllologiska sall-
skapet i Stockholm VI. 9 (1917) p. 285 f., where a good many examples
are given.
Ilcvall, E.
Contributions to the history
of Old English Dialects
PE
287
,E38
