m .1 , .

A CONTRIBUTION

TO THE STUDY OF

ANGLO-NORMAN INFLUENCE

ON

ENGLISH PLACE-NAMES

BY

R. E. ZACHRISSON

LUND 1909

FEINTED BY HAKAN OHLSSON

LUNDS UNIVERSITETS ARSSKRIFT. N. F. AFD. 1. Bd 4. Nr 3.

A CONTRIBUTION

TO THE STUDY OF

j

ANGLO-NORMAN INFLUENCE

ON

ENGLISH PLACE-NAMES

BY

R. E. ZACHRISSON x<

LUND 1909

PRINTED BY HAKAN OHLSSON

Preface.

The object of this treatise is to elucidate some questions concerning the influence. which French or rather Anglo-Norman

- the usual name for the French dialect spoken in England - has exercised on English place-names.

A detailed account of the various kinds of this influence would fill the pages of several volumes. In the present work I have only considered a number of sound-changes in English place-names which cannot be explained otherwise than as due to Anglo-Norman influence.

Here, again, I have mainly confined myself to dealing with such cases where the phonetic criteria do not admit of any doubt as to the cause of the change, and where, moreover, some of the forms due to Anglo-Norman influence have remained to the present day.

I do not presume to have definitely solved all the more or less intricate problems involved in these investigations; I hope, however, that my work may contribute to arouse an increased interest in the (sad to say) much neglected study of English place-names in general and in particular of French influence upon them.

It is with pleasure that I take this opportunity of expres- sing my sincere thanks to my teacher in English philology^ Professor Eilert Ekwall of the University of Lund, who has not

IV

only suggested this work, but has helped me with much kind advice and valuable information throughout the progress of it. My thanks are also due to Dr. G. E. Fuhrken of Gothenburg, who has revised my treatise from the point of view of the language.

Karlskrona, February 1909.

jR. E. Zachrisson.

Table of contents.

Preface IH

Table of contents V

Abbreviations.

Abbreviations of works consulted in ascertaining the early forms

of place-names VII

Abbreviations of names of counties XII

Other abbreviations XII

Introduction 1

I. Changes in English place-names due to Anglo-Norman sound-substitution.

§ 1. Anglo-Norman (ts) is substituted for English (t/) 18

Appendix A. The symbols c and ch in early A.N. and M.E. texts 32 Appendix B. The transition of (ts) to s in O.Fr. The date of

the D.B. MSS 37

§ 2. Anglo-Norman (t) is substituted for English initial ({>) 39

Appendix. Orthographic interchange between t and th for etymo- logical (t) 47

§ 3. Some difficult initial consonant-combinations are altered in various

ways to suit Anglo-Norman pronunciation (kn>ken, sl>l etc.) 49

§ 4. Anglo-Norman (dj) is substituted for English yiitial (j) 57

II. Changes in English place-names due to Anglo-Norman sound-development.

§5. s is lost before c, p, t 67

Appendix. Some cases of analogical transformation on the pattern

of forms in -cester, -Chester 78

§ 6. Loss of the interdental spirant.

A. — F. In place-names containing O.E. mupa, sup, wippe etc. 82

A.N. (t), (d) is substituted for M.E. final (p) and intervocalic (d) 95

VI

G. In place-names containing O.E. cepel 100

Appendix A. Various ways of rendering (p), (d) in early English

and Anglo-Norman MSS 115

Appendix B. Orthographic confusion between the symbols d and

i, g in the diphthong ei, ai etc 116

Appendix C. Loose s between two place-name compounds 118

§ 7. Change or interchange of the liquids (I, n,r}.

A. Interchange of I, n, r 120

B. Dissimilation of n to I, r by a following m 133

C. Dissimilatory loss of r 136

D. Interchange of -ling, -ning, -ring 138

E. Interchange of unstressed le (el), ne (en), re (er) 141

§8. I before a consonant is vocalized or lost 146

§ 9. a before nasal groups becomes au 153

§ 10. Change of (t/) to (/) 156

Index to names 161

Addenda et corrigenda.

Abbreviations.

Abbreviations of works consulted in ascertaining the early forms of place-names *.

A.S. Chr. = Two of the Saxon Chronicles parallel, ed. C. Plummer & J. Earle. Oxford 1892.

Anc. Ch. = Ancient Charters, royal and private prior to 1200 (1095 — 1200), ed. J. H. Round. Pipe Roll Society. London 1888.

Annales Mon. = Annales monastic! (mainly of 13 c.), ed. H. R. Luard. Chronicles and Memorials of Great Britain and Ireland 36. London 1846 — 49.

Bath Ch. = Two chartularies of the priory of St. Peter at Bath (most of the charters fall within the period 1066—1377), ed. W. Hunt. Som. Record Soc. London 1893.

Benoit = Benoit de Sainte Maure, Chronique des dues de Normandie, ed. F. Michel. Paris 1836—44.

Birch = W. de Gray Birch, Cartularium Saxonicum. London 1885 — 93. - The figures refer to the number of the charters.

Bruton and Montacute Ch. = Two cartularies of the priories of Bruton and Montacute (12 c.— 14 c.). Som. Record Society. London 1894.

C.R. = Calendar of the Close Rolls. - - The earliest (1204—27) are published by the Record Commissioners: Rotuli litterarum clausarum, ed. T. H. Hardy. London 1833—44. The others are being published in the Rolls Series. London 1892 ff. They consist of contemporary enrolments of Royal Letters Close. See the C.R. of Edward II., 1307—13, Introduction.

1 Our statements as to the date of the MSS. are generally derived from the Introductions to the various printed editions. We are indebted to Mr. H. Hall, F.S.A., Master of the Rolls, for the date of a few important MSS. For a detailed account of the import and contents of the various records, see S. R. SOAKGILL-BIED, A guide to the various classes of documents preserved in the Public Record Office (London 1908), or C. GBOSS, Sources and Literature of English History. London etc. 1900.

VIII

Capgrave = John Capgrave, The Chronicle of England, ed. F. C. Hingeston. Chronicles and Memorials 1. London 1858.

Cat. A.D. = Descriptive Catalogue of Ancient Deeds I— V. Rolls Series. London 1890 ff.

Ch.R. = A Calendar of the Charter Rolls I, II. Rolls Series. Lon- don 1903, 1906. -- Vol. I contains charters of 1226—1257, vol. II, charters of 1257—1300.

Cumbwell Ch. = Charters of Cumbwell priory (1160—1270). Kent Archseol. Soc. Archseologia Cantiana V, VI, VIII.

D.B. = Domesday book seu Liber Censualis Wilhelmi Primi regis Anglise, vols. I, II (= Exchequer D.B.), ed. A. Farley, London 1783; vols. Ill, IV, ed. H. Ellis. London 1816. — Vol. I contains all the counties of which a survey was taken, with the exception of Essex, Norfolk, and Suffolk, con- tained in vol. II; vol. Ill contains indices and a general introduction, vol. IV, Additamenta: Exon Domesday (= a survey of the five south-western shires), Inquisitio Eliensis, Liber Winton, Boldon Book. Hardy, Descriptive Cata- logue (London 1865) II, pp. 34, 36, assigns to the Exchequer and Exon MSS. so early a date as 1086. The evidence of the language points, however, to a considerably later date. Cf. below, p. 37 ff . and p. 72 note *.

Dv. D.B. = The Devonshire Domesday and geld inquest, extensions, translations, and indices, ed. J. B. Rowe. Devon Assoc. for Advancement of Science. Plymouth 1884—92.

Earle = J. Earle, A hand-book to the land-charters and other Saxonic documents. Oxford 1888.

F.A. = Inquisitions and Assessments relating to Feudal Aids etc. (1284—1431). Rolls Series. London 1899 ff. — Contains Kirby's Quest (1284), Nomina Villarum (1316), the Book of Aids (= Aid of 20 Edward III. and 3 Henry IV., MSS. early 15th cent.) the Subsidy Rolls of 6 Henry VI. and 9 Henry VI. Four volumes have appeared. This work is quite indispensable for an investigation into the history of English place-names. Kirby's Quest mainly exists in copies of the 15th and 16th cent. Of the Nomina Villarum there are original returns only for a small number of counties. In the In- troduction to the F.A., vol. I, p. VHI, we are expressly told, however, that »the readings of the original returns have been preferred to those of the books compiled from them>, and that »the spellings of names have been amended with the assistance of contemporary documents not always those from which the books were compiled but at any rate relating to the same transactions. >

Facsimile of Ch. = Facsimiles of ancient charters in the Brit. Mus., ed. E. A. Bond. London 1873—78.

Fantosme = Chronique de Fantosme. In vol. Ill of Benoit, Chronique des dues de Normandie, ed. F. Michel.

1 Dr. E. Walberg of Lund, who has examined a facsimile of the Exchequer D.B., is of opinion that, as far as the handwriting is concerned this MS. may very well be placed to the middle of the 12th century.

IX

Fr.Ch. = Calendar of documents preserved in France, illustrative of the history of Great Britain and Ireland (918—1206), ed. J. H. Round. Rolls Series. London 1899.

Frideswide Ch. = Chartulary of the monastery of St. Frideswide (1004—1537), ed. S. R. Wigram, Oxford Hist. Soc. 1895—96.

Gaimar = The A.N. Chronicle of G. Gaimar, ed. T. Wright. Lon- don 1850.

Geld. Inq. = The geld inquests of Np. and the five south-western counties; printed in Feudal England by J. H. Round (London 1892), and in the Viet. Hist, of the respective counties. — They record two assessments of Danegeld made between 1066 and 1084.

Gl. Recs. = Calendar of records of the corporation of Gloucester (1155 —1672), ed. W. H. Stevenson. Gloucester 1893.

Guill. le Mar. = Histoire de Guillaume le Marshal, ed. P. Meyer Paris 1891—1901.

H.R. = Rotuli Hundredorum (i.e. the »Hundred Rolls*), tempore Hen. III. et Edw. I. Record Comm. London 1812—18. — The Inquisitions of 1274—75 and 1279 are printed from contemporary MSS., not so the Inquisition of 1255.

I.C.C. = Inquisitio comitatus Cantabrigiensis (MS. c. 1180), subjicitur Inquisitio Eliensis, ed. N. E. S. A. Hamilton. London 1878.

I(nq.)El. = Inquisitio Eliensis, see I.C.C. — The MSS. of the I.E1. be- long to the end of the 12th cent.

I.P.M. = Calendar of Inquisitions Post Mortem. Rolls Series. London 1898 ff. — They consist of contemporary transcripts of the Inquisitions taken before the Es cheaters.

Index = Index to the Charters and Rolls in the British Museum, ed. H. Ellis and F. B. Bickley. London 1900.

Jesus Coll. MS. = Jesus College (Oxford) Manuscript 29, leaf 267; printed in An O.E. Miscellany, ed. R. Morris. Early English Text Soc. 49.

Ke. P.F. = Pedes Finium for Kent. Kent Archseol. Soc. Archseologia Cantiana I— VI, XI— XX.

Kemble = J. M. Kemble, Codex diplomaticus sevi Saxonici. London 1839 — 48. — Vol. VI contains an Index to place-names. The figures refer to the number of the charters.

La. P.F. = Final concords of the county of Lancaster (1196—1307), ed. W. Farrer. Record Soc. for La. and Ch. London 1899.

La. Pi.R. = the Pipe Rolls for Lancashire. Record Soc. for La. and Ch.

La. Recs. = Lancashire and Cheshire Records, ed. W. Selby. Record Soc. for La. and Ch. London 1882—83.

Langtoft = The Chronicle of Pierre de Langtoft, ed. T. Wright. Chron. and Mem. 47. London 1866—68.

La3amon = La3amon's Brut, ed. F. Madden. London 1847.

Le. Recs. = Records of the borough of Leicester, ed. M. Bateson. Cam- bridge 1899.

Le.S. = The Leicestershire Survey 1124—29; printed in Feudal Eng- land by J. H. Round.

Li.S. = The Lincolnshire Survey, temp. Hen. I., ed. J. Greenstreet. London 1884. Translation of the same survey by R. E. C. Waters. Lincoln 1883. — The MS. of the Li.S. is of the year 1119 or somewhat earlier.

Mdf. = H. Middendorff, Altenglisches Flurnamenbuch. Halle 1902.

Mi. F.F. = A Calendar of the feet of fines for London and Middlesex (Richard I. — 12 Elizabeth), ed. W. J. Hardy & W. Page. London 1896.

Mon. Brit. = Monumenta historica Britannica, ed. H. Petrie & J. Sharpe. London 1848.

Monks Horton Ch. = Charters of Monks Horton *priory (1140—1311), «d. J. R. Scott. Kent Archseol. Soc. Archteologia Cantiana X.

Mousket = Chronique riin^e par P. Mousket, ed. Reiffenberg. Briissel 1836 ff.

Muchelney and Athelney Ch. = Cartularies of the Abbeys of Muchel- ney and Athelney, ed. E. H. Bates. Som. Rec. Soc.

Napier-Stevenson = A. S. Napier & W. H. Stevenson, The Crawford collection of early charters and documents. Oxford 1895.

Nb. Pi.R. = The Pipe Rolls for Northumberland (1273—84), ed. W. Dickson. Newcastle 1854—60.

Nb. Rolls = Three early assize Rolls for Northumberland, ed. W. Page. Surtees Society 1891.

Np.S. = The Northampton Survey, Hen. I. — Hen. II. ; printed in Viet. Hist, for Np.

Nt. Recs. = Records of the borough of Nottingham (1155 — 1625), ed W. H. Stevenson. London 1882-89.

P.R. = Calendar of the Patent Rolls. — The earliest (1201—16) are published by the Record Commissioners: Rotuli litterarum patentium, ed. T. D. Hardy. London 1835. The others are being published in the Rolls Series. London 1891 ff. They consist of contemporary enrolments of Royal Letters Patent. See the P.R. of Edward III., 1327—30, Introduction.

Pi.R. = Pipe Rolls: Magnum Rotulum 31 Hen. I. (1130—31), ed. J. Hunter, 1833; The Great Rolls of the Pipe 2, 3, 4 Hen. II. (1155—58), ed. J. Hunter, 1844; Great Roll of the Pipe 1 Richard I. (1189— 90), ed. J. Hun- ter, 1844 etc. — Nearly all the Pi.R. are printed from contemporary MSS.

Pontefract Ch. = Chartulary of St. John of Pontefract, ed. R. Holmes. Yo. Archseol. Soc. Record Series XXV. Leeds 1899.

R.B. = The Red Book of the Exchequer, ed. H. Hall. Chronicles and Memorials 99. London 1896. — The MSS. are of the 13th cent, and later.

R.Br. = The Story of England by Robert Mannyng of Brunne, ed. F. J. Furnivall. Chron. and Mem. 87. London 1887.

R.G1. = The metrical Chronicle of Rob. of Gloucester, ed. W. A. Wright. Chron. and Mem. 86. London 1887.

Ramsey Ch. = Cartulariurn monasterii de Rameseia, ed. W. H. Hart <fc P. A. Lyons. Chron. and Mem. 79. London 1884—94.

Rievaulx Ch. = Cartulariurn abbathiae de Rievalle (1132—1539), ed. J. C. Atkinson. Surtees Society, 1889.

(Roman de) Rou = Maistre Wace's Roman de Rou, ed. H. Andresen. Heilbronn 1877—79.

XI

Searle = W. G. Searle, Onomasticon Anglo-Saxonicum. Cambrigde 1897 1.

Selborne Ch. = Calendar of charters and documents relating to Sel- borne and its priory, ed. W. D. Macray. Hampshire Record Soc. London 1891-94.

Sempringham Ch. = Charters relating to the priory of Sempringham, ed. E. M. Poynton. Genealogist, new series XV, XVI etc. London 1899 — 1900.

Sf. P.F. = A Calendar of the Feet of Fines for Suffolk, ed. W. Eye. Ipswich 1900.

Sf. S.E. = Suffolk in 1327, being a Subsidy Return from the Suffolk Green Book, ed. S. H. A. H. Woodbridge 1906.

Sm. K.Q. = Kirby's Quest for Somerset, Nomina Villarum for Som. etc., ed. F. H. Dickinson. Somerset Record Soc. London 1889.

Sm. N.V. = Nomina Villarum for Somerset, see Sm, K.Q.

Sm. P.F. = Pedes Finium for the county of Somerset (1196— 1346), ed. E. Green. Somerset Record Soc. 1892—98.

Sm. S.R. = Subsidy Rolls for Somerset, see Sm. K.Q.

Ss. S.R. — Subsidy Rolls for the county of Sussex. Sussex Archseol. Soc. Collections II, X.

Sweet, O.E.T, = H. Sweet, The Oldest English Texts. Early English Text Society, 83.

T.N. = Testa de Nevill sive Liber Feodorum, temp. Hen. III. et Edw. I. Record Comm. London 1807. -- »The bulk of the material in the printed work belongs to the first half of Henry III:s reign (1216 — 72). The compila- tion in its present form may have been made late in the reign of Edward I.», Gross, The sources and literature of Engl. history, p. 378.

Tax. Eccl. = Taxatio ecclesiastica Anglise et Wallise auctoritate Ni- colai IV. (1291). Record Comm. London 1802.

Thorpe = B. Thorpe, Diplomatarium Anglicum sevi Saxonici. Lon- don 1865.

V.E. = Valor Ecclesiasticus 26 Hen. VIII., ed. J. Caley & J. Hunter. Record Comm. London 1825 — 34.

Viet. Hist. = Victoria History of j the Counties of England, ed. A. Constable & Co. London 1900 ff. - - Contains excellent accounts of the various D.B. portions, with careful identifications of place-names etc.

Wace — Roman de Brut par Wace, ed. Le Roux de Lincy. Rouen 1836.

Yo. I.P.M, = Yorkshire Inquisitions (1241—95), ed. W. Brown. Yo. Archseol. and Topogr. Association. Record Series XII, XXIII, 1892—98.

Yo. K.F. = Knights' Fees in the Survey of York etc., ed. R. H. Skaife. Surtees Society, 1867.

Yo. K.Q. == Kirby's Inquest for Yo, see Yo. K.F.

Yo. N.V. = Nomina Villarum for Yo., see Yo. K.F.

Yo. P.F. = Pedes Finium Ebor, ed. W. Brown. Surtees Society, 1897, and Yo. Archseol. and Topogr. Assoc. Journal XI ff.

Yo. S.R. = Yorkshire lay subsidy 25 Edw. I. (1297), 30 Edw. I. (1301), ed. W. Brown. Yo. Archaeol. Society. Record Series XVI, XXI.

1 Searle always renders O.E. J5, # by th, and we have conformed to this habit when quoting instances from his Onomasticon.

XII

Abbreviations of Names of Counties.

Bd. = Bedford. Le. = Leicester.

Be. = Berkshire. Li. = Lincoln.

Bu. = Buckingham. Mi = Middlesex.

Cb. = Cambridge. Mo. = Monmouth.

Ch. = Cheshire. Nb. = Northumberland.

Co. = Cornwall. Nf. = Norfolk. .

Cu. = Cumberland. Np. = Northampton.

Db. = Derby. Nt. = Nottingham.

Do. = Dorset. Ox. = Oxford.

Du. = Durham. Sf. = Suffolk.

Dv. = Devon. Sh. = Shropshire.

Es. = Essex. Sm. = Somerset.

Gl. = Gloucester. Sr. = Surrey.

Ha. = Hampshire. Ss. = Sussex.

He. = Hereford. St. = Stafford.

Ht. = Hertford. Wa. = Warwick.

Hu. = Huntingdon. Wl. = Wiltshire.

Ke. = Kent. Wo. = Worcester.

La. = Lancashire. Yo. = Yorkshire.

Other abbreviations.

a. = anno.

A.N. = Anglo-Norman.

Archiv = Archiv flir das Studium der neueren Sprachen.

Bardsley's Die. — C. W. Bardsley, A Dictionary of English and Welsh surnames. London 1901.

Barth(olomew). = J. G. Bartholomew, The Survey Gazetter of the Bri- tish Isles. London 1904.

Behrens = D. Behrens, Beitrage zur Geschichte der f ranzOsischen Sprache in England. Franzosische Studien V, 2.

Berberich = H. Berberich, Herbarium Apuleii. Anglistische Forschungen 5. Heidelberg 1902.

Blach, Paulsschule = S. Blach, Die Schriftsprache in der Londoner Paulsschule. Halberstadt 1905.

Busch = E. Busch, Laut und Formenlehre der Anglonormannischen Sprache des XIV. Jahrhunderts. Diss. Greifswald 1887.

Btilbring = K. D. Biilbring, Altenglisches Elementarbuch I. Heidel- berg 1902.

c. — 1. century, centuries ; 2. circa, about.

Camden = W. Camden, Britannia, sec. ed. by E. Gibson. London 1722.

cent. = century, centuries.

cf. = confer, i.e. compare.

XIII

Dibelius = W. Dibelius, John Capgrave und die englische Schriftsprache Anglia XXIII.

Diez = F. Diez, Grammatik der romanischen Sprachen I. Vierte Auflage. Bonn 1876.

Duignan, St. Pl.Ns. = W. H. Duignan, Notes on Staffordshire Place Names. London 1902.

Duignan, Wo. Pl.Ns. = W. H. Duignan, Worcestershire Place Names. London 1905.

E. = East, Eastern.

e. = early.

E.D.D. = The English Dialect Dictionary by J. Wright. Oxford 1896 ff.

E.D.Gr. = The English Dialect Grammar by J. Wright. Oxford 1905.

E.E.T.S. = Early English Text Society.

E.E.T.S. E.S. = Early English Text Society. Extra Series.

Ekwall, Jones's Phonography = E. Ekwall, Dr. J. Jones's Practical Phonography. Neudrucke Friihneuenglischer Grammatiken 2. Halle 1907.

Ellis = A. J. Ellis, On Early English Pronunciation V.

Engl. Stud. = Englische Studien.

Forbiger = A. Forbiger, Handbuch der alten Geographic. Hamburg 1877'

Franzmeyer = F. Franzmeyer, Studien iiber den Konsonantismus und Vokalismus der neuenglischen Dialekte. Diss. Strassburg 1906.

Germ. = Germanic.

Grammont = M. Gramrnont, La dissimilation consonantique. Dijon 1895.

Griming = B. Griming, Schwund und Zusatz von Konsonanten in den neuenglischen Dialekten. Diss. Strassburg 1904.

Hildebrand = F. Hildebrand, Uber das franzosische Sprachelement im Liber Censualis Wilhelms I. von England. Zeitschrift fiir romanische Philo- logie VIII.

Hoevelmann = K. Hoavelmann, Zum Konsonantismus der altfranzo- sischen Lehnworter in der mittelenglischen Dichtung des 14. und 15. Jahr- hunderts. Diss. Kiel 1903.

Hope = R. C. Hope, A Glossary of Dialectal Place-nomenclature. Lon- don 1883.

Horn, Gramm. = W. Horn, Historische neuenglische Grammatik I. Strassburg 1908.

Horn, Gutturallaute = W. Horn, Beitrage zur Geschichte der englischen Gutturallaute. Berlin 1901.

Horn, Untersuchungen = W. Horn, Untersuchungen zur neuenglischen Lautgeschichte. Quellen und Forschungen 98. Strassburg 1905.

Joret = Ch. Joret, Des caracteres et de 1'extension du Patois Normand. Paris 1888.

Kluge, Paul's Gr. = F. Kluge, Geschichte der englischen Sprache in Grundriss der germanischen Philologie herausgeg. von H. Paul. Zweite Auflage. Strassburg 1899.

Kceppel, Spelling.-pron. = E. Kceppel, Spelling-pronunciations. Quellen und Forschungen 89. Strassburg 1901.

XIV

Kornmesser = E. Kornmesser, Die franzOsischen Ortsnamen germanischer Abkunft. Dies. Strassburg 1888.

Kruisinga = E. Kruisinga, A Grammar of the dialect of West Somerset. Bonner Beitriige zur Anglistik 18. Bonn 1905.

1. = late.

Lat. = Latin.

Ltbl. = Literaturblatt fur germanische und romanische Philologie.

Lucht — P. Lucht, Lautlehre der alteren La3amorfhandschrift. Palae- stra 49. Berlin 1905.

Luhmann = A. Luhmann, Die Uberlieferung von La3amons Brut. Stu- dien zur englischen Philologie herausgeg. von Morsbach, 22. Halle 1906.

M.Br. = Magna Britannia by Th. Cox. London 1726—31.

M.H.G. = Middle High German.

MS., MSS. = Manuscript, Manuscripts.

Mackel = E. Mackel, Die germanischen Elemente in der Franzosischen und Provenzalischen Sprache. Franzosische Studien VI, 1.

Mdl. = Midlands.

Menger = L. E. Menger, The Anglo-Norman dialect. New York 1904.

Meyer-Liibke, Gramm. = W. Meyer- Liibke, Grammaire des Langues Romanes I. Paris 1890.

Meynier = G. Meynier, Les noms de lieu romans en France. Extraits des Memoires de la societ^ d'emulation du Doubs. Besangon 1901.

Mod. = Modern.

Mod. E. = Modern English.

Morsbach, Gramm. = L. Morsbach, Mittelenglische Grammatik, Halle 1896.

Morsbach, Urkunden = L. Morsbach, Uber den Ursprung der neueng- lischen Schriftsprache. Heilbronn 1888.

Miiller = R. Miiller, Uber die Namen der nordhumbrischen Liber Vitse. Palaestra 9. Berlin 1900.

N. = North.

N.E. = New English.

N.E.D. = A New English Dictionary.

Neumann = G. Neumann, Die Orthographic der Paston Letters von 1422 — 61. Marburger Studien zur Englischen Philologie 7. Marburg 1904.

Nyrop, Gramm. = K. Nyrop, Grammaire Historique de la Langue Francaise J. Copenhague 1899.

O.E. = Old English.

O.F(r). = Old French.

O.H.G. = 01d High German.

Pogatscher, Lehnworte = A. Pogatscher, Zur Lautlehre der griechischen, lateinischen und romanischen Lehnworte im Altenglischen. Quellen und Forschungen 64. Strassburg 1888.

Quicherat = J. Quicherat, De la Formation francaise des Anciens Noms de Lieu. Paris 1867.

Rathmann = F. Rathmann, Die lautliche Gestaltung englischer Per- sonennamen in Geffrei Gaimars Reimchronik »L'Estoire des Engles*. Diss Kiel 1906.

XV

Rippmann = W. Rippmann, The Sounds of Spoken English. Lon- don 1906.

Scand. = Scandinavian.

Schroer, Regula = A. Schroer, Die Winteney-Version der Regula S, Benedicti. Halle 1888.

Sievers = E. Sievers, Angelsachsische Grammatik. Dritte Auflage. Halle 1898.

Skeat, Cb. Pl.-Ns. == W. Skeat, The Place-Names of Cambridgeshire. Cambridge 1901.

Skeat, Ht. Pl.-Ns. = W. Skeat, The Place-Names of Hertfordshire. Hertford 1904.

Stimming = A. Stimming, .Der anglonormannische Boeve de Haumtone. Bibliotheca Normannica VII. Halle 1899.

Stolze = W. Stolze, Zur Lautlehre der altenglischen Ortsnamen im Domesday-Book. Berlin 1902.

Sweet, H.E.S. = H. Sweet, History of English Sounds. Oxford 1888.

Sweet, N.E. Gr. = H. Sweet, A New English Grammar I. Oxford 1892.

Tanger = G. Tanger, Englisches Namen-Lexikon. Berlin 1888.

Taylor, Names = I, Taylor, Names and their Histories. London 1898.

W. = West.

Waltemath = W. Waltemath, Die frankischen Elemente in der fran- zosischen Sprache. Diss. Strassbnrg 1885.

Westphal = J. Westphal, Englische Ortsnamen im Altfranzosischen. Diss. Strassburg 1891.

Williams = C. A. Williams, Die franzosischen Ortsnamen keltischer Abkunft. Diss. Strassburg 1891.

Introduction.

1. Method of treating the subject of place-names. M. E. and O. Fr. sources of old forms used for our pur- poses. — Obviously the only right way to treat the subject of place-names is, in the first place, to collect a sufficient number of early spellings; secondly, to establish the etymologies on the evidence of the spellings; thirdly, to ascertain the modern pro- nunciation. Owing to the limited material upon which we can draw, this ideal treatment of place-names cannot yet be fully carried out. Nevertheless, by taking full advantage of the ma- terial at our disposal, we have tried to pay due regard to all these important points.

The most absurd mistakes may arise when attempts at etymologizing or explanations of modern pronunciations are not based on the evidence of early forms. By disregarding this canon some eminent scholars have more than once arrived at erroneous conclusions. In various essays on English philology (HOLTHAUSEN, Indogerm. Forsch. XV, p. 275, E. RUDOLF, Die engl. Orthographic, Diss. Marburg 1904, p. 39) we are told that Bristol < Bristou is due to dialectal addition of I, just as Tyn- dale has rainboll = rainbow. The addition presupposes loss of final 1. This loss is mainly confined to some N. and Mdl. dialects (Yo., La., Ch., Nb.f Cu., Db. etc.; cf. E. D. Gr. § 255), and does not seem to have been evidenced until late M.E. or early N.E. times. I in Bristol is different both as to the place and the time of its appearence. Somerset does not belong to the dialects for which the loss has been evidenced, and we have found such forms as Bristollum, Index, BristolL H.R., T.N.

1

2 R. E. Zachrisson

as early as the 12th and 13th centuries l. - - KOEPPEL, Spelling- pron., p. 55, has the following remark on the English place-name Coventry: »Afz. cuvent > me. ktwent, covent ne. heute noch (kevant) in den Ortsnamen Covent Garden etc. Hingegen steht fiir den Stadtenamen Coventry neben der alten (u > -B) Lautung jetzt auch eine Schriftaussprache (kov9ntri)». But Coventry has nothing whatever to do with O.Fr. cuvent. The etymology is O.E. Cofantreo, A.S.Chr. = »the cave tree* < O.E. cofa, N.E. cove. Consequently, the pron. (kovantri) represents the phonetic development. - JESPERSEN, Engl. Stud. 31, p. 241, suggests that Birmingham might be explained from Brummagem with insertion of n. LOGEMAN, Engl. Stud. 34, p. 250, adopts this explanation without making any restrictions. BITTER, Archiv CXIII, p. 34, is of opinion that Brummagem has been altered to Birmingham by association with the numerous English place- names which contain -ing. All these theories are, however, flatly contradicted by the fact that, as was correctly pointed out by Jespersen, the forms in -ing are by far the oldest: Bermingeham, D.B., Burmingham, T.N. etc. Forms without n do not appear until the 14th and 15th cent.: Birmycham (1489), Brymmycham (1519), Index. Cf. also the material given under Brummagem. in N.E.D.

These and similar errors render it necessary for us to use the current literature on English place-names with the greatest discrimination.

We have endeavoured to gather a copious and trustworthy material of early forms. As the French 2 influence, naturally, does not commence until the time of the Conquest, we have mainly gone through charters and feudal records from the 12th to the loth cent. England may well be proud of her long, almost unbroken series of such records, which are excellent sources for historical and philological researches.

Most of these documents are written in Latin, it being a well-known fact that literary monuments in English are of very

1 To account for the form Bristol I have formed a theory of my own,, which I hope soon to publish in one of thie philological periodicals.

1 For the sake of convenience the term French is often used through- out this work in the sense of Norman or Anglo-Norman. When no other- wise stated, 0. Fr. implies both Continental French and Anglo-Norman.

Anglo-Norman influence on English place-names 3

rare occurrence even as late as the beginning of the 13th c. A comparison indicates, however, that the forms of a certain place- name are, on the whole, the same in a Latin record as in an English one of the same date, and, consequently, the forms in Latin documents are to be considered as reliable material for philological investigations. Such cases of Latinizing as occur, are generally confined to the addition of a Latin ending, -a, ~ia, -urn, e.g. Cestria, Index = Chester, Wika, T.N. = Wyke Regis (Do.), Bristollum, Index, Bristola, V.E. = Bristol etc. It is even a characteristical feature of the Exon D.B. that almost all place-names there exhibit the Latin ending -a. More seldom we meet with translations, like Album Monasterium (1244), Index = Whitchurch (Sh.), Villa S. Botulphi, Index = Boston (Li.), Nova terra (1275), Subs. Roll = Newland (Wo.), Novus lurgus, New- borough (1356) = Nobury (Wo.); cf. Duignan, Wo. Pl.Ns., p. 115, p. 116.

Numerous forms of interest have also been noted from M.E. chronicles and other purely English works. We have moreover gone through the most important O.Fr. and A.N. chronicles in which English place-names occur. Such purely French forms are generally not given in the lists preceding the various paragraphs. We have thought it a better plan to give them in the text itself, with a view to proving that some changes in which they also have participated, are really due to French influence. On the other hand forms from the interesting collec- tion, Documents preserved in France = Fr.Ch., are always given in the lists. In this place we will once for all draw atten- tion to the importance of the various spellings found there. Nearly all the changes which we have ascribed to French in- fluence, occur also here, a fact which seems eminently to corro- borate our conclusions.

Most of the works we have consulted, are to be found in the three well-known series: (1) Calendars of State Papers, (2) Eerum Britannicarum medii cevi scriptores or Chronicles and Memorials of Great Britain and Ireland, (3) Publications of the Record Commissioners.

It is much to be regretted that we", have nothing in Eng- lish corresponding to the excellent series of French Topographical Dictionaries, where early forms are recorded of most place-names

4 R. E. Zachrisson

in a certain district. The only works we know of containing a fairly large collection of early spellings from various times and for various counties, are the Index and the Feudal Aids. It has, indeed, been a toilsome and laborious task to go through scores of books merely to ascertain a sufficient number of early forms for the places of one single county. Nearly all the works we have used, however, are provided with indices to the place- names, arranged alphabetically or according to counties. In some instances only the old forms are given, but in the majority of cases they are identified with, their modern equivalents. When this has not been the case we have spared no pains to supply the identification ourselves; moreover, in most dubious cases old identifications have been verified. For such purposes the Feudal Aids and some old geographical works, like CAMDEN'S Britannia, sec. ed. E. GIBSON, 1722, and THOMAS Cox's Magna Britannia, c. 1750, in which the old division in hundreds is still kept, have proved very useful. KELLY'S Directories, where the place-names of all countries are given alphabetically in their modern forms, have also been of great help to us.

As a rule it has not been considered necessary to note forms later than the 15th cent. The few changes which may appear in the spellings of place-names after that time, can hardly be put down to French influence. On the other hand, as this influence is particularly strong in the 12th cent, and the begin- ning of the 13th when contemporary records are comparatively scarce, we have tried to bridge the gulf by using as many printed sources as possible from that period. Thus numerous charters, most of which have been published separately by learned Societies, have been consulted.

The modern forms are generally taken from the Survey Gazetteer, ed. J. G. BARTHOLOMEW, 1904 (= Barth.) A number of small places which are not given in Barth., wo have quoted from Kelly or the Index.

2. Etymologies. O.E. sources. — To give full ety- mologies of all the place-names we have to deal with, would not accord with the general plan of our treatise. We have often only to ascertain that owing to A.N. influence a sound or sound- group has been changed in a certain direction. Hence it will

Anglo-Norman influence on English place-names 5

sometimes suffice to adduce the modern form and a few M.E. references exhibiting the regular development. Nevertheless, as it is always desirable and sometimes necessary to establish the O.E. ground-forms, we have consulted for this purpose several collections of O.E. charters and similar documents in which place-names abundantly occur. Here the identification offers the greatest difficulties. Out of many hundreds of O.E, place- names only a small number have been identified, and the identi- fications are not always correct. When it has been impossible to find any references to the O.E. ground-form, we have often tried to establish the etymology on the evidence of our earliest and most trustworthy M.E. forms, mainly with the aid of SEARLE'S Onomasticon and MIDDENDORFF'S Flurnamenbuch. O.E. words which have not found a place there, are quoted from the Anglo-Saxon Dictionaries of SWEET and Bos WORTH-TOLLER.

3. Pronunciation of place-names. Influence of the spelling. — As to the modern pronunciation of English place- names there is often a marked difference between the pron. of natives and that of persons unacquainted with the place. The former mostly reflects the phonetic development, the latter is often altogether dependent on the spelling. It is obvious that spelling-pronunciation must be very usual ia a language like English, which is composed of so many heterogeneous elements and so conservative in its orthography, that in spite of the most thorough-going phonetic changes the modern spelling in many respects is still the same as in Chaucer's time. Cf. on this question E. KOEPPEL, Spelling-pronunciations.

Naturally the names of large and well-known places are pronounced in about the same way by all Englishmen, but where small places are concerned, which are hardly known outside the parish or the county, there often exists by the side of the local phonetic pronunciation an artificial one due to the spelling.

Indeed, the spelling plays a very important part in the history of place-names. Various forms which at the first glance might be taken to represent a sound-change, are merely due to careless orthography. The scribes who had to copy various deeds, did not always acquit themselves of their task in the most conscientious way. Nothing is less consistent than early

6 R. E. Zachrisson

M.E. or early A.N. orthography. Various symbols were used to denote the same sound, some of them representing the historical spelling, others being attempts to render new sounds phoneti- cally. When owing to continual development within the respec- tive dialects two letters which originally had distinct values, came to denote the same sound, one was frequently put for the other, and this use might be extended even to positions where the old values were still kept.

To make things worse, some letters were often confused, partly owing to outward resemblance, partly from reasons which have not as yet been cleared up. Some of these confusions are really astonishing. A place in Devon called Instow, Innestow, F.A., is mentioned in both our D.B. Mss. as Johannesto, -stou. Ins- or bines- of the original has evidently been taken for a shor- tening of Johannes.

Such erroneous forms due to false orthographical represen- tation are sometimes adopted into the spoken language, and fre- quently used by the side of the original ones. Thus Thystleworth (1428), Index, Thistleworth, Camden, I p. 368, seems to have been a usual variant of Isleworth (Mi.) < Gistelesworde, D.B. Here th is obviously due to confusion between f, th and 3, y. Such forms may even ultimately supplant the etymological ones. It is very likely that Jesrnond (Nb.) has originated from Jesmonth, * Jesmonde, n being an error for u. Cf. § 4.

It has been pointed out quite correctly that in the majority of cases spelling-pronunciations have not originated until com- paratively late times when books and the reading of books were coming into more general use. In our opinion the English place- names must to a certain extent be considered as exceptions to this rule. We must keep in mind that a great number of the early documents in which these place-names occur, were drawn up or copied by Normans, who often may have had a very imperfect knowledge of the English language. Under such cir- cumstances the written forms may, particularly in the case of small places, have been invested with more importance than those used by a small number of people living in the place itself. To support this opinion of ours we will adduce some in- stances where f orms due to false orthography seem to have been

Anglo-Norman influence on English place-names 7

in frequent use at a comparatively early date. . Nearly all the old spellings of Dyseworth (Le.), Dichesword (12 c.), Le. S., Dicheswrth, Ch.R.II, Dychesworth, H.K., T.N., point to O.E. die as the ground-form. Such early forms as Digacteswrff, Digtfeswrthia (1180), DyMswort.il (1246), Index, Divvort (with loss of J5), D.B., render it very probable, however, that ch is due to confusion between ch and th. The modern form may be accounted for both ways. Of particular interest is Sawbridgeworth (Cb.),

Sabrixteworde, D.B., Sabrihteswrthe (1166), R.B. < O.E. SceberU, Searle. Cf. SKEAT, Pl.-Ns. of Ht., p. 57. We account for the mo- dem form in the following way: t being lost before s gives origin to such forms as Sabrycheswrfh, H.R., Sdbrichesworth, Cat. A.D; ch, originally a spelling for the palatal spirant in O.E. beorht, is now taken to denote (t/), and Sdbriche- is changed into Sabrige- owing to weak stress. Hence Sabrigeworth (1428), F.A. The phonetic development is reflected by the local pronun- ciation (ssepso). Some more instances are noted under § 10, where the alternative spellings of etymological -ic, -isc are discussed.

Only a thorough detailed investigation will enable us to realize what a vast and important part spelling has played in the history of English place-names.

Useful references to works in which confusion of letters and similar phenomena have been considered, are given by MORSBACH, Mittelengl. Gramm., p. 28. In addition the following may be mentioned: BERBERICH, Das Herbarium Apuleii (Ang- listische Forschungen, 5), p. 7 ff., LUHMANN, La3amons Brut (Studien zur Engl. Phil, herausg. v. MORSBACH, XXII), p. 42 ff., KELLER, Angelsachsische Palaographie (Palaestra XLIII 1, 2); cf. SCHROER, Anglia, Beiblatt XVIII, p. 96 f.

So very little being known about early M.E. spelling, we have often had to make rather long digressions in order to establish what value is to be assigned to the various orthogra- phical symbols. It has often been difficult to find a suitable place for such discussions of spellings. We have tried to relegate them to foot-notes or to appendices after the paragraphs con- cerned.

It is a great drawback that we have no satisfactory pro- nouncing-dictionary of English place-names. For want of better we have used: TANGER, Englisches Namen-Lexikon, RIPPMANN,

8 R. E. Zachrisson

The Sounds of Spoken English, p. 111—119, HOPE, A Glossary of dialectal place-nomenclature (the absence of a phonetic nota- tion considerably lessens the value of this book). A few local pronunciations are also given by A. J. ELLIS in vol. V of Early English Pronunciation.

The system of phonetic notation we have used is, on the whole, the same as that in SWEET'S A Primer of .Spoken English (Oxford 1900). The only alterations we have made is that the symbol (B) has been adopted for the sound in bwt, the symbols (o) and (oo), respectively, for the short and long sounds in not and all. The palatal or guttural spirant in O.E. cniht, brohte has been transcribed with (/). All phonetic notations are placed between round brackets. Only a few transcriptions borrowed from Ellis' phonetic alphabet have been placed inside square brackets.

4. Influence of the dialects. — There is one more question of the greatest importance for place-name study, viz. the influence of the dialects. Here many sound-changes have taken place which have no correspondence whatever in literary English, and such changes may also appear in the names of places. In treating this subject one of our greatest difficulties has been to distinguish properly between the changes which are due to French influence, and those which have originated through dialectal sound-development. Sometimes the chronology may serve as a criterion. If a certain phenomenon nowadays com- mon to various dialects but not evidenced in e.M.E., appears in the place-names of Anglo-Latin records of the 12th and 13th cent., we seem justified in pointing to French influence, suppo- sing the changes in question admit of such an explanation. — Another criterion is the geographical distribution of the forms. Thus, if certain peculiarities confined to a few English dialects have been noted by us in the names of places situated in the most different parts of England, French influence would seem admissible.

More than once we have had to enter into a rather de- tailed discussion of these questions. Besides numerous M.E. texts quoted separately we have used for this purpose works oa e.N.E. and Mod. E. dialects by ELLIS, WEIGHT, HORN, KEUISINGA, FEANZ-

Anglo-Norman influence on English place-names 9

MEYER, GRUNING etc. Of. the list of grammars and philological works quoted in this treatise.

5. Works bearing on the English place-name study in general. - • The best known English works of this kind are: EDMUNDS' Traces of history in the names of places, TAYLOR'S Words and Places, and Names and their histories. These books are useful on account of the old forms there recorded. From a philological point of view Edmunds' work is antiquated and of no value. Taylor is more conscientious, though also here too many hap-hazard etymologies and uncritical, often erroneous remarks are to be found.

Monographies exist only for a few counties. Among these we have to note in the first place three works by Prof. SKEAT (The Place-Names of Cb., Ht., and Bd.), which may well be said to form a new era in the history of English place-name study. DUJGNAN'S Notes on Staff. Place Names, and Wore. Place Names offer excellent collections of early forms, though we have some doubt as to the correctness of several etymologies given there. A work by STREATFEILD called Line, and the Danes is to be used with great caution for philological purposes.

To some extent we have availed ourselves of the material brought together in the books referred to above. So far as we have been able to ascertain, it has proved quite trustworthy.

6. Works bearing on the subject of A.N. influence on English place-names. - - Hitherto the influence which A.N. may have exercised on English place-names, has not been made the subject of any serious philological investigations. Only some special question in connection with it has occasionally received some attention. Editors of several early records containing place- names may have pointed out some » curiously corrupted forms», but it has not occurred to many that this corruption may reflect an A.N. adaptation, may be an attempt to express phonetically the various changes which many English place-names necessarily underwent in the speech of the Normans.

In his above-mentioned works on place-names Prof. Skeat has paid due regard to the possibility of French influence. In our opinion he considerably overestimates this influence, and

10 R. E. Zachrisson

above all does not seem to make any clear distinction between the changes which are due to dialectal sound-development, and those caused by French influence.

The English place-names in D.B. have been made the subject of a special investigation by STOLZE, Zur Lautlehre der altengl. Ortsn. im D.B. This in many respects very meritorious work reveals a considerable defect. The authgr has evidently not had recourse to the modern D.B. literature, and, consequently, the identifications are not always correct. Sometimes two forms are pointed to as alternative spellings, whereas, in reality, they stand for two distinct places. Here is a typical instance to illustrate these deficiencies. On page 32, Anm., the following forms are adduced to serve as a proof that French is was substituted for English Jcs: Aisse, Aisselie, -ford, -minstre, Esseminstre (Dv.); Aissa, Aixa, Aissebrige, -Jord, -cote, Aisecome (Sm.). None of all these examples are conclusive. They are stated to be derived from O.E. Exan-, Exe-, Ax-, but, in reality, Aisse, Aisselie, -ford (Dv.) correspond to the present places Ash, Ashley, Ashford; Aissa , Aixa, Aissecote, Aisecome (Sm.) to Ash, Ashcombe, and Ashcot all of them < O.E. cesc, Mdf., p. 9. Only Aisse-, Esseminstre, and Aisseford are really due to O.E. Exe-. As the spellings ai, ei often indicate e in D.B., we are, however, only justified to con- clude that Norman s was substituted for English Jcs. Cf. Essesse (= Essex): cesse, G. le Mar., v. 19007, Ozenefort (< O.E. Oxena- ford), Wace, v. 12376. Other similar mistakes will be pointed out further on. We may add that such minor shortcomings do not essentially lessen the value of the book, as in most instances at least some of the numerous instances brought together to exemplify various phonetic phenomena are quite to the point.

The purely French word-material in D.B. has been treated by HILDEBRAND in Zeitschrift fur Rom. Phil, VIII.

We only know of two more special investigations which have some bearing on our subject, viz. RATHMANN, Englische Personennamen in Gaimar, and WESTPHAL, Englische Ortsn. im Altfranz. To the latter work there is appended a useful, though by no means exhaustive, list of English place-names from various O. Fr. sources.

Anglo-Norman influence on English place-names 11

7. Origin and various kinds of the A.N. influence in question 1. • Immediately after the Conquest a great number of Frenchmen settled in England, and for more than 200 years there was a steady immigration of fresh crowds attracted by prospective advantages. Most secular posts of any importance were bestowed by William and his immediate successors on Frenchmen, and the whole clergy, of higher and lower rank mainly consisted of Frenchmen. Merchants and artisans of various trades also immigrated, and settled down in different parts of England; cf. VISING, I p. 6 f. In 1154 Norman noblemen were in possession of 1115 fortified English castles. The monasteries and religious houses were 'filled with French monks of various orders; cf. VISING, I p. 8.

Numerous French-speaking centres having thus originated, it goes without saying that the language of the conquerors must have left some traces in English place-nomenclature. The proud Normans, with all their aversion to the English language, had to pronounce the names of the places they had occupied.

Only a few instances are recorded of an English place- name having been replaced by a new French one. Sheen has been superseded by Richmond (TAYLOR, Names, p. 235), Tates- hale, D.B., by Pontefract (TAYLOR, p. 225), Lutgaresburg by Mont- acute (FREEMAN, V p. 572) etc. »On the whole really new names were confined to really new foundations. A good many castles and abbeys reared by Romance-speaking founders not unnaturally received Romance names »; FREEMAN, ibid.

It is much more usual that a French element is added to an English one, which gives origin to hybrid formations, like Old- castle (Mon.), Newcastle (Nb.), Newe castelle, R.Br., v. 250, Nof- chastel sur Tyne, Langtoft, II p. 192, Castlecombe (WL), Cumbe (John), Index, Droitwich (Wo.), Saltwic, Birch 138, Wich, D.B. Cf. DUIGNAN, Wo. Pl.Ns., p. 53.

Often this element is the name of a Norman family, which mostly is put after the place-name as a separate word, e. g. Stoke Mandeville, Bury Pommeroy, Ashby de la Zouch. More

1 In working out this section and the following one we have mainly depended on the following works: FREEMAN, The Norman Conquest I— V J. VISING, Franska spraket i England I, II, III (Gothenburg 1900), BEHRENS, Zur Geschichte der Franz. Sprache in England (Paul's Gr.).

12 R. E. Zachrisson

seldom the English word and the Norman name have coalesced, as in Hurstmonceaux (Ss.) < O.E. hyrst + O.Fr. monceaus (cf. Willelmus de Moncels, D.B., HILDEBRAND, p. 339), Hurstpier point (Ss.) < O.E. hyrst -f- O.Fr. Pierpoint (still used as a surname, see BARDSLEY'S Die.), Amport (Ha.), Anne, D.B., Anne Port (1306), Anne de Port (1325), Index (cf. Hugo de Port, D.B., HILDEBRAND, p. 341), Bridgewater (Sm.)1, Bruges Walteri (1^27), Ch.R., Bru- geswalteri, -water, H.R., Briggewauter (1346—49), C.R., Briggewater, (1401—05), P.R., Gentleshaw (St.), Gentylshawe (1505); according to DUIGNAN, St. Pl.Ns., p. 67, in 1341 a person called John Gentyl is mentioned as living in the samrne parish.

Not unfrequently the A.N. saint, which soon took the place of O.E. sanct, has been added to a place-name, mostly in connection with the name of a saint. For examples see BAR- THOLOMEW'S Gazetteer.

Such forms as Wicmaret, Wivelescomet, Estochet, D.B., Stoket (1284), F.A. = Wickmere (Nf.), Wiveliscornbe (Sm.), Stoke (Sm.), seem to instance the addition of the French diminutive suffix-e£. Cf. Montamys (1484), Montamiset (1498) = Montamy (Calvados), Die. Topogr. du Calvados.

The preposition de d' frequently used to connect a per- sonal name and a pjace-narne, is sometimes found prefixed to the latter: Disenhurst (c. 1270), Selborne Ch. = Ysenhurst (Ha.), W. de Deverwik (3 X), Ramsey Ch. = York. Cf. Aniblainvilla and Demblenvilla (1128) = Damblainville (Calv.), Avevilla (1060), Dyau- villa (14 c.) = Deauville (Calv.), Die. Topogr. du Calvados.

Of particular interest are the cases in which the French definite article appears either prefixed to the place-name : Laierda, D.B. = Yard (Dv.), Lalege, D.B. = Leigh (Gl.), Lacherne, D.B. = Kerne (Ha.), Ladone, D.B. = The Down (Ha.), La Hide, G. le Mar., v. 15904 = Hyde Abbey (Ha.), La Kenape, G. le Mar., v. 15886 = Knepp Castle (Ss.), La Forde (1275). Subs. Roll = Lifford (Wo.), Pole (1342), La Pole (1454) = Lappal (Wo.); or inserted between the two components: Barton le Street (Yo.), Thorpe le Soken (Es.), Sutton le Marsh (Li.) etc. 2.

1 B. was originally called Brugge or Burgh Walter from the Norman baron to whom the manor was given at the Conquest; cf. EARTH.

1 This question is best treated in connection with the corresponding use of the article in French place-names. We have gathered some material

Anglo-Norman influence on English place-names 13

As the places of Normandy often contain elements from various Germanic dialects, English place-names are sometimes remodelled after the pattern of such or other well-known forms. Norman -tot and- ware corresponding to O.E. -toft and -mere, we some- times find the former substituted for the latter: Wicmara, D.B. = Wickmere (Nf.), Slidemare, Ledemare, D.B. = Sledmere (Yo.), Hotot, -toft, D.B. = Huttoft (Li.), Langetot, Stouwdangetot (13 c., 14 c.), Index = Stowlangtoft (Sf.), Wilgetot, D.B. = Willitoft (Yo.). Such forms are explained in a different manner by JELLINGHAUS, Anglia XX, p. 325 (tot a dialectal variant of loft) and STOLZE, p. 16 (mare for mere owing to Scand. influence). We believe our explanation to be the correct one. The irregular forms of the above-mentioned places hardly appear except in D.B. All later records we have consulted, have -toft and -mere. The only place for which the forms in -tot have been kept somewhat longer, is Stowlangtoft, which may have been the seat of an old Norman family, there being a manor-house in its immediate neighbour- hood. — Westminster appears in Wace and Fantosme as Westmoslier, -mustier] cf. WESTPHAL, Index. To O.E. mynster corresponds O.Fr. moustier. — Fleshy (Es.), formerly Plesincho, D.B., Plesingho, T.N., F.A., seems to have been transformed on the analogy of numerous French place-names in ei < etum. Cf. such Latinized forms as Plesseto, H.R., Plessetum (1408), Index 1.

From the above-mentioned cases we have to keep apart such arbitrary alterations as Beauxfield for Whitfield (Ke.) Beaus- feld (Hen. III.) Index, or Belleau (Li.) for Elgeh, D.B. Here the first English element has been replaced by the very usual French place-name component beau, bel.

A few times attempts have been made at translating English place-names into French: Chef de Bede, Gaimar, v.

for a special investigation into the subject. POGATSCHEE, Anglia, Beiblatt XVIII, p. 105, is of opinion that le, when inserted between two Engl. place- name components, is a remainder of^the French preposition lez. His theory does not seem te be substantiated by the early spellings. We must keep in mind that, though lez sometimes may serve to connect two French place- name components, a noun preceded by the article le, la is also often added to a place-name, as it would seem, to distinguish the place in question from others with the same name. Cf. such early forms as Fresneium la Mere (12 c.) = Fresn4-la-Mere (Calv.), Die. Top. du Calvados.

1 Remains of an aucient castle witness to Norman settlement; BAETH.

14 R. E. Zachrisson

1416 < O.E. Badanheafod, A.S. Chr., Grandham, D.B. = Greet- ham (Li.)

We will, however, not enter into a detailed discussion of all these interesting phenomena, our main object being to ac- count only for such changes in English place-names as are due to differences between English and Anglo-Norman phonology.

7. Victory of A.N. forms. The causes. For a period of more than 300 years French was the official language in England, understood and spoken not only by the Normans and the French, but also by the more refined among the Eng- lish; cf. VISING, I p. 19. We are justified in assuming that at least for the whole of this time double forms existed of many English place-names, one of them purely English, the other exhibiting A.N. peculiarities. From being used originally only by the Normans, the latter might in course of time easily have been adopted by the English themselves.

Such forms may also have exercised a mutual influence on each other. Indeed, in one or two cases it seems as if an A.N. form has given origin to a corresponding English one. Cf. § 5 Appendix.

Some of the French forms were merely transitory, as Nicole for Lincoln, others, like Londres for London, were only used by the French themselves, but not unfrequently the English forms have been replaced by the French.

No doubt, the supremacy of French as the language of the conquerors has been the principal cause of this victory. Whereas English was looked down upon as an inferior language (cf. the often quoted passage from R.G1., PAUL'S Gr., p. 951), French was highly appreciated. This does not surprise us when we consider that it was spoken by the men of the highest rank in the country.

With reference to place-names it has already been pointed out that, may-be, a faulty pronunciation used by people in a higher position and on a higher cultural level, may easily be adopted by the inhabitants of the place itself. Zinal and Za (Wallis), in which z signifies a local pronunciation ts, are natu- rally pronounced with voiced s by the French and, being situa- ted on French territory, also by many German tourists, and this

Anglo-Norman influence on English place-names 15

pronunciation is now corning into use even among the natives. See MEYER-LUBKE, Einfiihrung in das Studium der Roman. Sprachwissenscbaft, p. 193.

It is principally where large places are concerned, which have played an important ecclesiastical or political part, and in the population of which the A.N. element certainly was very well represented, that the French forms have heen victorious. When the same has been the case with smaller places, this may often be traced back to the fact that a Norman castle, monastery, or manor-house was situated in the immediate neighbourhood. There were a great many religious houses in England, and as late as 1252 French and Latin were often the only languages known by the monks. Of. VISING, I p. 7, 9, II p. 8. Thus our theories are sometimes supported by historical evidence.

Lastly, we will draw attention to one more circumstance which certainly has contributed not a little to the victory of the Anglo-Norman forms. »At the time of the Norman invasion of England the practice of hereditary surnames seems still to have been a novelty, but a novelty which was fast taking root. The mem- bers of the great Norman houses already bore surnames, some- times territorial, sometimes patronymic, of which the former class became hereditary »; FREEMAN, V p. 565. This custom the Normans brought over with them to England, where it soon spread rapidly; Vising, I p. 20. Many Norman noble families were named after the place where they had their landed pro- perty. Of. Gislebert de Cliva, D.B., HILDEBRAND, p. 346. = Mod. E. Clive < O.E. clif. It. is evident that such place-names used as surnames by the Normans were developed according to French sound-laws, and in some cases the pronunciation of the name may also have been transferred to that of the place.

9. Plan of the treatise. — We have thought it suitable to divide our treatise into two parts, according to the different nature of the phenomena we have to deal with.

I. Changes in English place-names due to Anglo- Norman sound-substitution. An English sound or group of sounds was, generally or in a certain position, absolutely un- known to the Normans, and was, consequently, altered in one

16 R. E. Zachrisson

way or another to suit A.N. pronunciation. Changes of this kind took place, of course, immediately after the Norman in- vasion.

II. Changes in English place-names due to Anglo- Norman sound-development. The A.N. dialect was, natu- rally, in a state of continual development. At> the time of the Conquest sound-laws were acting and continued to act within it, and others made their appearance during the period of more than three hundred years when A.N. was spoken in England. It is evident that numerous English place-names used by the Normans in the written and spoken language should be subjec- ted to changes caused by those sound-laws.

As to the treatment of the various paragraphs we have not thought it advisable to lay down any hard and fixed rules al- ways to be observed. The phenomena we have to deal with involving manifold and dissimilar problems, we have striven to obtain the best results by adopting what in every special case seems to be the best method of treating the various matters in hand.

Note. It was our original idea to devote a special chapter to some changes in the pronunciation of English place-names which are due to A.N. orthography. Considerations of time and space have made us desist from this intention of ours. We will content ourselves with giving a few examples here. In rendering English sounds the Normans sometimes used symbols which easily might give rise to misinterpretations. Thus O.E. u is often rendered by o. Of. SWEET, N.E.Gr., § 775. If o was kept in orthography, this might lead to a spelling pronunciation (o) for («, u). To the examples adduced by Koeppel, Spelling-pron., the following may be added: Lolworth (Cb.\ pron. (loula). I am indebted for the local pronunciation to the vicar of the parish. Old forms are Lolesuuorde, D.B., Lulleivorth (1284), F.A.; cf. O.E. Lulleswyrp, Kemble 714. Rotherhithe (Sr.), pron. (rodarip), Tanger. On the etymology cf. Koeppel, p. 56. Compton (Gl. etc.) pron. (komtan), Tanger; cf. O.E. Cumbtun, Cumtun, Thorpe. Combermere (Wo.), pron. (kombamia), Tanger; cf. Comberton (Wo.) < O. E. Cumbrincgtitn, Birch 1282. Cromwell (Nt), Crunwelle D.B., Crumbwell, Crombewdl, F.A. At least the surname is nowadays pronounced with (o) l. The Cavaliers' toast: — >Wash this crumb

1 According to Hope, Cromwell (Nt.) is dialectally pronounced »Crum- mil>. Mr. Fuhrken has kindly pointed out to me that the pers. name Crom- well is still sometimes pron. with (B), and that the current pron. of Com- bermere is (kBmbamia) not (kombamia), as stated by Tanger.

Anglo-Norman influence on English place-names 17

well down> (cf, BAEDSLEY, Die., p. 218), enables us to conclude that the present pronunciation had not yet arisen at the time of the famous Eng- lish statesman. In Bodley, Bodleian, pron. (bodli, bodlian), an instance is offered of spelling-pron. having been introduced into the personal name, whereas the corresponding locality, Budleigh (Dv.) exhibits the historically correct spelling and pron.

10. Arrangement of the material. - - The material of the old forms we have collected has been brought together in lists at the head of the different paragraphs. In drawing up these lists we have proceeded in the following way. The various records we have used are arranged chronologically 1. Records of one and the same century are given in alphabetical order. Old forms of place-names taken from the same document are ar- ranged alphabetically according to counties. The figures at the top refer to the Mod. and O.E. forms, which always head the lists.

When from various reasons exemples of place-names are adduced in the text itself the forms meant to illustrate a certain sound-change are generally given first, then the present name of the place, and lastly a few references to early regular forms. All the early spellings both of place-names and of ordinary words have been printed in italics.

1 As it would have been impossible to ascertain the exact date of the MSS. of all the works we have consulted, we have had to base the chrono- logy on the date of the originals. The age of many MSS. is, however, stated in the list of works consulted preceding the Introduction.

I. Changes in English place-names due to Anglo-Norman sound-substitution.

§ 1. Anglo-Norman (ts) is substituted for English (t/).

A. In initial position.

1. The Anglo-Norman forms have prevailed:

a. In place-names containing O.E. cester, ceaste, caster"1. 1. Exeter (Dv.) < O.E. Exanceaster, A.S. Chr. 2. Frocester with F. Court (Gl.). 3. Ci- rencester (Gl.) < O.E. Cirenceaster, A.S. Chr. 4. Gloucester (Gl.) < O.E. Gloive-, Gleawcester, A.S. Chr. 5. Leicester (Le.) < O.E. Ligeraceaster, A.S. Chr. 6. Craster (Nb.). 7. Towcester (Np.) < O.E. Tofeceaster, A.S. Chr. 8. Bicester (Ox.). 9. Wroxeter (Sh.). 1O. Rocester (St.). 11. Alcester (Wa.). 12; Mancetter or Mancester (Wa.). 13. Worcester (Wo.) < O.E. Wigraceaster , A.S. Chr.

D.B. Execestre1, Frowcestre*, Cirecestres, Glouuecestre*, Ledecestre*,

Bernecestre*, Rochecestreg, Roivecestre10, Wirecestrels. 12th c. — Index. Circestria* (1149), Gloecestria* (1130). Pi.R. Legrecestre5

(1131), Leicester'0 (1155). 13th c. — C.B. Leicester6, Towcester'1 , Alencesire11 (31—34). Ch.B. Exeter1

(1242). F.A. Leycestrie*. Proclamation of Henry III. 1258

1 Nearly all M.E. forms of O.E. cester, ceaster, ccester exhibit e in the root-syllable. This e was in O.E. confined to Mercian and 1. West-Saxon. Cf. BULBRING § 314. M.E. e instead of a may sometimes be due to weak stress (thus MORSBACH, Urkunden, p. 52 f.) or even to the analogy of the numerous forms where e is the regular development. La3amon has several instances of ce, a as well, but it is very difficult to decide which was the reading of the original, as in the text ce, a not unfrequently occur for e < O.E. e. Chastre, R.G1. (= Chester), Chastrefeld, Chaster/eld, Index (= Chesterfield Db., St.,) may go back to O.E. ceaster, ccester, though, on the other hand, asso- ciation with O.Fr. chastre < Lat. castra is not excluded; confusion with O.Fr. chastel is indisputable in Chastelfield (= Chesterfield, Db.), R.G1.

Anglo-Norman influence on English place-names 19

(in Kluge, Mittelengl. Lesebuch, p. 25). Gloivchestre* , Leirchestre*, Wirechestre13. MS. Jesus Coll. Oxford (O.E. Miscellany, Early E. Text Soc. 49). Execestre1, Gloucestreschire*, Leycestreschire5, Wirecestre13. La3amon. JExechcestre1 A, Excestre1 B, v. 9755, Excestre1 A, v. 31112, Axcetre1 A, v. 30865, Cher inches tre3 A, v! 29112, Chirenchestre3 A, v. 29122, Chirchestre* A, v. 29205, Gloichestre* A, Gloucestre* B, v. 9629, Gleoch&stre* A, Gloucestre* B, v. 10429, Gloucester* A, v. 27206, Leirchestre5 A, Leycetre* B, v. 2915, Leirchcestre* A, Leycestre5 B, v. 24346, Leycestre* A, v! 27272. P.B. Cyrencestre3, Gloucester* (32—47), Leircestre* (16—25), Touecestre1 (01—16), Worcester18 (32—47). B.G1. Excestre1, Exes- tre1, Excetre1, Gloucestre*, Gloucetre4, Leycester*, Leicetre5, Wor- cestre13, Wircestre13, Wircetre1* (for references to line and page se the Index to Wright's ed.). T.N. Crancestre*, Toucestre1, Bernecestre*, Alencestre11, Manecestre1'2. Tax.Eccl. Excestre*, Cirencestre3, Leicestre*, Toivecestre1 , Wrokcestre9, Alicestre11 Man- cetre1*.

14th c. — C.R. Exeter1, Cirencester*, Gloucester4, Leicester'0, (07-13), Wrox- cestre9 (46—49), Worcester™ (07—13). P.A. Frowcester* (1316), Leicester6 (1316), Crauucestre6 (1346), Wroccestre9 (1316). Index. Glousetre* (1374). P.B. Excestre1, Cirencester*, Gloucester*, Lei- cester5 (91—96), Wroxcestre9 (46—49), Wircestre13 (91—96).

15th c. — Capgrave. Excetre1, Excetir\ p. 122, p. 129, Cicetir3, p. 275, Glouceter*, pp. 136, 147, Leicester5, p. 100, Leycetir*, pp. 118, 157, Wyscetir13, p. 281, Wycetir13, p. 292. F.A. Leycestrie5 (1428), Craucestre* (1428). I.P.M. Exeter1 (Hen. VII.), Bursester* (Hen. VI.), Burencestres (1485), Bysseter* (1495).

16th c. — V.B. Frowceter*, Cirincester, -cetre3, Gloucester, -ceter*, Glociter4, Glocitershire4, Leicestre, -ceter*, Biscyter*, Wroxcetor, -cetour*,. Wroxtor®, Rocettre10, Mancettur12, Worcestre, -ceter13.

b. In other cases. 1. Cippenham (Bu.). Cf. Chippenham and Chip- ping, where, according to Tanger, ch is pronnounced (if). 2. Cerne Abbas, Nether and Upper Cerne (Do.). 3. Cirencester (Gl.) < O.E. Ciren-, Cyren-,. Cyrnceaster, A.8. Chr. 4- North and South Cerney, Cerney Wick (Gl.) '.

1 Cerne, Cirencester, and Cerney may all of them contain O.E. ciren, which seems to have been a not uncommon name of rivers in the West of England. There are three rivers the names of which may be derived from ciren: (1) Churne < O.E. Cirnea, Kemble, in Glouc, on which are situated Cirencester, N. and S. Cerney, and Cerney Wick; all these places have probably been named after the river; (2) Cerne in Dors., on which we note Cerne Abbas, Nether and Upper Cerne; (3) Churnet and Churnet Valley, Chirnete (1284) (see DUIGNAN, St. Pl.Ns., p. 40) in Chesh. — The ultimate source of O.E. ciren is obscure. Like the names of so many other English rivers it may be of Celtic origin. POGATSCHEE, Lateinische Lehnworter, p. 125, derives the first element of Cirenceaster from Lat. Corinium. This-

20 R. E. Zachrisson

D.B. Cerne*, CcrneP, Cerneli*, Obcerne*, Cirecestre*, Cernei*. 12th c. — A.S. Chr. Cernel* (1114), MS.H. Bruton and Mont. Ch. (tres)

Cernels*. Index. Circestria*. Pi.B. Cerna* (1131), Cerne* (1189). 1 3th c. — H.B. Chippenham1, Cirincestre*. Lasamon. Cernele* A, v. 29674,

Cherinchestre9 A, v. 29112, Chirchestre* A, v. 29205. P.B. dp*

penham1 (16—25). T.N. Cern?, Cernay*. Tax. Eccl. Ccrne1,

Cirencester*. 14th c. — O.B. Cippenham1 (46—49), Cirencester* (07—13). F.A. Sowth-

serney* (1303), Southcerney4' (1316, 46), Northcerney4* (1303, 16, 46).

Index. Upcerne* (1379). P.R. Sippenham1 (48—50). 15th c. — Capgrave. Cicetir3, p. 275. I.P.M. Cerne* (1487). Index.

Syppenham1 (1429).

2. Jfo Anglo-Norman forms have not prevailed.

a. In place-names containing O.E. cester, ceaster, ccester. 1. Chesham (Bu.). 2. Chesterton (Cb.) 3. Chester (Ch.) < O.E. Ceaster, A.S. Chr. 4. Chesterfield (Db.). 5. Dorchester (Do.) < O.E. dornuuarana ceaster, Kent. Chart., Sweet, O.E.T., p. 434. 6. Lanchester (Du.). 7. Colchester (Es.) < O.E. Colneceaster, A.S. Chr. 8. Woodchester (GL). 9. Winchester (Ha.) < O.E. Wintanceaster, A.S. Chr. 1 0. Silchester (Ha.). 1 1 . Porchester (Ha.). 12. Cheshunt (Ht.). 13. Godmanchester (Hu.). 14. Rochester (Ke.) < O.E. Hrofesceaster, A.S. Chr. 15. Manchester (La.) < O.E. Mameceaster, A.S. Chr. 16. Ribchester (La.). 17. Irchester (Np.). 18. Dorchester (Ox.) < O.E. Dorceceaster, A.S. Chr. 19. II Chester (Sm.). 2O. Chichester (Ss.) < O.E. Cisseccaster, A.S. Chr.

D.B. Cestreham1, Cestretone* , Centre*, Dorecestre'0, Colecestra1 , Widecestre*, UdecestreB, Silcestre10, Portcestre11, Cestrehunt1*, Rove- cestre1*, Mamecestre™, Ribelcastre™, Dorchecestre™, Cicestre™. 12th c. — Index. Chestreunt1*, Cestrehont™, Cestrehunt1*, Gutmuncetre1*

(46—54), Gumcestria1* (1194).

13th c. — O.B. Cestrefeld*, Cilcestre™, Porcestre^, Mamecestre™, Cycestre™ (31—34). Ch.B. Cestresham1 (1257), Langecestre" (1292), Cole- cestria1 (1255). F.A. Irencestre" (1284). H.B. Cestresham1, Dor- cestre5, Colcestre"1, Colcetre"1, Cestrehonte1*, Gomcestre*9, Irencestre1^ , Dorkccestre1*, Cycestre™. La3amon. Winchestre9 A, B, v. 2819, Winchcestre* A, v. 9431, Portchcestre11 A, Portcastre11 B, v. 9234. P.B. Colecestre1 , Winchester9, Porcestre, -chestre11 (32—47), Gome- cestre1*, Gurmechestre18 (81—92), Mamecestre™ (25—32), Maume- cestre" (72—81), Rybecestre1*, Ivelcestre™, Cycestre™ (32-47). B.B.

etymology is by no means certain. Cf. FOKBIGER, Handbuch der alten Geogr. Ill, p. 289. Corinium would, besides, have given O.E. cyren with a gut- tural k. In Anglia XXXI, p. 272, POGATSCHKR accounts for s as due to spelling- pron. This explanation seems to us very unsatisfactory. The c/i-forms in La3amon indicate that in M.E. a pron. with (t/) existed by the side of the present one, where s, no doubt, is due to French influence.

Anglo-Norman influence on English place-names 21

Cestretone* (1210), Cestrehunte™. B.G1. Chastre*, Chastelfield*, Golchester\ Winchestre9, Wince fre9, Winchetre9, Wincestre9, Cili- cestre10, Rouchester1*, Roucestre^, Roucetre1*, Chichestre™, Cicestre™, Cicetre™. T.N. Cestresham1, Colecestre\ Wodecestre*, Porcestre11, Mamecestre1*, Irecestre*1 , Dorkecestre±B. Tax. Eccl. Cestrefeud4, Colecestre1 , Colechestre1 , Porcestre11.

14th c. — C.B. Silchestre10, Cilcestre™ (46—49). P.A. Chesham* (1302), Wodecestre8 (1303, 46), Gurmencestre™ (1346), Roucestre, -chestre14 (1346), Irencestre11 (1316). Index. Cestre* (1331), Cestrefeld* (1306), Cheslrefeld* (1384), Rouceste1* (1326). P.B. Langecestre* (1399— 1401), Gummecestre^ (21—24), Gurmundchestre1* (43—45), Mani- chestre™, Mamcestre™ (91—96). B.Br. Chester*, p. 19, Chestre- feld\ p. 223, Winchester9, p. 19, Wincester9, p. 121.

1 5th c. — Capgrave. Chester*, p. 133, Winchester9, p. 109, Rochester1*, p. 214, Chichester™, p. 187. F.A. Chesthunte1* (1402), Rouches- tre1* (1404), Irencestre11 (1428). I.P.M. Dorchester18 (1486). P.B. Godmuncestre1* (22—29).

1 6th c. — V.E. Lanchestre6, Wodchester*, Winchestre9, Dorchester1*, (hundred

de) Dorcestre1*. b. In other cases. Any number of instances might he adduced. We

refer to the discussion on the proportion between the English and French

forms on p. 28 below.

B. In medial position.

1. The Anglo-Norman forms have prevailed.

1. Messing (Es.). Cf. O.E. (ge)mcecca, also used as a personal name, Mdf., p. 91. 2. Whissonsett (Nf.)? < O.E. wee, Mdf., p. 150. Cf. Whit- chingham (Nf.), Witcingeham, D.B., Wychingham, H.R., Witchingham, F.A.

D.B. Matcinga1, Witcingkeseta? . 13th c. — C.B. Messing1 (88—96). Ch.B. WissingeseP, Wilsingesete* II.

H.B. Wychingsete*.

14th c. — F.A. Messing*. P.B. Wyssingsete* (27—30). 1 5th c. — I.P.M. Wyssyngset* (Hen. VII.). P.B. Messynge1 (1401—05).

2. The Anglo-Norman forms have not prevailed.

1 . Matching (Es.). 2. Latchingdon (Es.). 3. Fletching (Ke.). 4. Blet- chingley (Sr.). Cf. O.E. Blcecca, Searle, Bleccingdenn, Thorpe = Blatching- don (Ke.).

D.B. Matcinga1, Lacen-, Lachenduna*, Flescingess, Blacking elei*. 1 3th c. — C.B. Machinge* (72—79), Flechyng* (72—79), Blechingeligh* (72—79),

Blessingleye* (88-96). Ch.B. Fletching6 (1252), Blescingeleye4- II.

H.B. Flessing3. Index. Blechynglie*. 1 4th c. — F.A. Massing1, Maching1, Lacindon* (1303). Index. Blechynglie*.

22 R. E. Zachrisson

C. In final position after i.

1. The Anglo-Norman forms have prevailed.

1. Chatteris (Ob.). It is hard to find any suitable etymology of this word. SKEAT, Pl.-Ns. of Cb., p. 69, suggests it was originally a river-name in -ic. 2. Hiz, a small river in Ht. Cf. hundret de Hiz in the D.B. of Ht. 1 Tbe genuine English form appears in Hichin and Hichin Priory, situated on the Hiz. Of these places we have noted some early Spelling, like Hiche (1210, 1303), R.B., Index, Hechyn (1541) Index. 3. Diss (Nf.)? < O.E. die.

D.B. Cetriz1, Dice9. 1 2th c. — I.C.C. Chateriz\ Chatriz1. I .El. Catriz1, Cetriz1, Cateriz1, Dicse*.

Index. Cateriz1 (1177).

1 3th c. — Kamsey Ch. Chaterik1, Cceateric1, Chateric1, Ceairice1. T.N. Discs*. 14th c. — Index. Chaterfz1.

2. The Anglo-Norman forms have not prevailed.

1 . Greenwich (Ke.) < O.E. wic. 2. Woolwich (Ke.). 3. Harwich (Nf.). 4. Norwich (Nf.) < O.E. Norpivic, A.S. Chr. 5. Launditch Hundred (Nf.) < O.E. die. 6. Ipswich (Ox.). 7. Combwich (Sm.). 8. Evercreech (Sm.). 9. Chadlewich (Wa.).

D.B. Grenviz1, Hulwiz*, Commie1, Eurecriz*.

\ 2th c. — Index. Chadeleswiz9 (John). Pi.R. Gippeswiz6 (1157). 1 3th c. — Ch.R. Gyppcivych* II. F.A. Wolwy** (1284). H.B. Laundiz*,

Gyppeiviz, -wice6. Index. Herewiz*, Norewiz*. \ 4th c. — F.A. Wolewych* (1346), Laundiz5 (1302), Landich* (1316), Laun-

dich* (1346).

At a very early period in the life of the English language (k) under certain circumstances was fronted, and; at least in the South of England, further developed to (t/j, about the same sound as in Mod. English church. Cf. SIEVEBS § 206; BULBKING § 491 ff. Although original (k) had developed into two different sounds, the same orthographical symbol c was mostly used for both of them, and, consequently, in initial, medial, and final position might have the value of (t/).

Latin (k) was also fronted in some parts of the Romance linguistic territory, at different times before different vowels. As

1 SKEAT, Pl.-Ns of Ht., p. 63, remarks that the present name of the river may be due to an artificial revival of the old Domesday form. Whether this is true or not can be ascertained only by an examination of a suffi- cient number of early spellings.

Anglo-Norman influence on English place-names 23

regards (k) before a, two districts are to be distinguished in Nor- mandy. In the North along the Channel (k) was kept, whereas in the South just as in Central French it became (if), and this stage had already been reached at the time of the Norman Con- quest. Cf. K. BEETZ, C und Ch vor lateinischem A in altfranz. Texten (Diss. Strassburg 1887), p. 67 1; MEYER-LUBKE, Gramm. I § 410 f.

Before e and i the fronting of (k) had begun at a much earlier date, but varies in its results. It appears from loan- words and other criteria that in the llth cent, (t/) was prevalent in the North-East of France, (ts) in other parts. Cf. MEYER- LtfBKE, Gramm. I §§ 406 f. That Normandy belonged to the district where (k) had become (ts), is clearly indicated not only by the orthography of e. A.N. MSS., but also by the phonology of the French loan-words in English. Most early instances exhi- bit ce, ci; the few with die, chi in Modern English are almost all of later introduction. Cf. BEHKENS, p. 185.

As to the various ways in which the sounds (k), (t/), and (ts) are orthographically represented in the English and French MSS. of the 12th cent, and later, cf. Appendix A.

The comparison we have made between the development af original (k) in O.E. and O.Fr., results in the fact that in the English of the llth cent, (k) and (t/) appear for original (k), whereas A.N. has three sounds, i. e. (k), (t/), (ts). Now, it seems very plausible that the Normans to whom (t/) was quite a well- known sound, should have kept it in all positions, but this has by no means been the case. There must, indeed, have been a great acoustic resemblance between (ts) and (t/), and, in fact, Eng- lish (t/) is rendered by the Normans sometimes with one, some- times with the other. English (t/) seems to have been kept only in positions where it occurred in Norman too; on the other hand, (ts) was substituted for it in such positions and before such vowels where (ts) only occurred, or was at any rate prevalent in the Norman dialect itself. This whole procedure of substitution may be characterized as a tendency to bring the phonology of the English words in accordance with the Norman vocabulary.

In the following discussion we have, consequently, to disting- uish between (t/) in initial, medial, and final position and, at the same

24 R. E. Zachrisson

time, take due regard to the vowel which follows immediately after (if).

A. (if) in initial position. 1. (if) is followed by se.

As to the A.N. way of rendering Englisn (t/se) < cea, c& there are two possibilities to be taken into account. O.E. CB, which was something between Norman short open e and cr, has either been taken for e, in which case (ts) was probably substi- tuted for (if), or it has been identified with a. If so, the same interchange of c = (k) and ch = (if) sets in which we have already observed in the Norman dialect itself, ca instead of cha in South English place names cannot be due to native development for - with the single exception of cea before I -\- consonant — the assi- bilation had taken place in the whole Southern part of England. Cf. HITTER, Die Verteilung der ch- und &-formen irn Mittelengl. (Diss. Marburg 1904), p. 21, p. 80.

We have chosen the following examples from D.B. to illus- trate the first possibility, A.N ce < O.E. cce: Cetriz = Chatteris (Cb.), Cedeslai = Chaddesley (Wo.) < O.E. Ceaddcsleage, Birch 357, Celdvic (with Id for dl) D.B. = Chadwich (Wo.), Chadeleswich, Chaddewyke (13 c.) (cf. DUIGNAN, Wo. Pl.Ns., p. 35) < O.E. dead- wealla Searle.

On the other hand the same interchange as in French Cartrai, Chartreia, Kartrai, D.B. l (cf. HILDEBRAND, p. 333), ap- pears in Kagefort, Exon. D.B., Chageford, Exch.D.B. = Chagford (Dv.), Chage-, Chagliford, F.A. < O.E. cea(c)ge, Mdf., p. 25, N.E, dial. chag; a branch of broom or gorse (E.D.D.), and Chadele- dona, Exon D.B., Cadeledone, Exch. D.B. = Cheldori (Dv.) <O.E. Ceadwealla, Searle. ca for cha, and vice-versa, is also to be found in later documents. Chatteris appears as Cairiz, Chateriz- in the LEI., Catareio and Chateriz in the I.C.C. The Index has Cateriz (1177). Catshill (Sr.)is Chatishille (1131) and Cateshella (1141), Index. In Wace we have noted Chatenois, v. 2365 = Caithness (Scot- land), ChantorUre, v. 4079 = Canterbury. The same interchange appears even before a (French alteration!) < Kentish e < y in

1 For other similar instances in Norman place-names cf. JOEET, p. 139.

Anglo-Norman influence on English place-names 25

Cameshing, Chamsesinges, R.B. = Kemsing (Ke.) < O.E. Cymesinc (a. 822), Index; after s in Escardeburg, Eschardeburg, Ch.R. I = Scarborough (Yo.). It is worthy of note that M.E. carl in Gaimar appears as carles, charles, v. 5486 etc., which latter form cannot be accounted for by O.E. ceorl, whose regular M.E. forms are cherle, churle. See the Dictionaries of MATZNER and STRATMANN- BRADLEY. Modern English place-names do not, as far as we have been able to ascertain, preserve any traces of this A.N. peculiarity.

2. (t/) is followed ~by e, i.

In this position the Normans substituted (ts) for (t/), as appears from the nearly exclusive spellings ce, ci in D.B. A great many English place-names are also still pronounced with s < (ts) < (t/). Such a development is from an English point of view quite without a parallel, but to the Normans the substitu- tion came quite natural, as their own language did not possess (t/i) except in one single word (chiche < Greek xtxxov), and (t/e) in a very small number of words in unstressed position (cheval etc.). These few exceptions could not have counted very much in comparison with the great number of words with (tse), (tsi). In fact, cester is the only spelling that occurs for numerous Eng- lish places in O.Fr. texts. Cf. the instances brought together by WESTPHAL, Index, to which may be added: Colecestre, I, p. 76, Cestre, 1, p. 306, Wincestre, I, p. 32, Cestrefeld, II, p. 148, Lang- toft; Cestre, v. 15235, Porecestre, v. 15101, Rouvecestre, v. 15072, Guill. le Mar. Cf. also Certesie, v. 15767, Guill. le Mar. = = Chertsey (Sr.), Certesore, v. 823, Gaimar = O.E. Cerdices ora, A.S. Chr.

There are, however, a few spellings che, chi for O.E. (t/e), (t/i), in D.B. (not noted by Stolze): Chipeham = Chippenham (Cb.), Celes-, Chelesham = Chelsham (Sr.), Ckercke, Exon D.B., Cerce, Excheq. D.B. (Devon) = Church. These spellings might be considered as at- tempts to render the English sound properly, but are on account of their rare occurrence perhaps best explained as errors. We

1 The regular French development of cestre, i.e. cetre, is still preser- ved in Mod. Fr. Bicetre: — »Lieu, pres Paris, ou e"tait un chateau. — Etym. Winchester, nom d'un personnage anglais, qui, dans le XIV6 siecle, poss^da ce chateau*; LITTKE, Die. francais I, p. 337.

26 R. E. Zachrisson

must keep in mind that the D.B. MSS. are copies of original returns, perhaps even second-hand copies, and a thoughtless scribe who was used to write c and ch indiscriminately in some positions, e. g. before a, might very well have transferred this interchange to c in other positions. Indeed, it is hardly probable that the c^-spellings belong to the originals. In the MS. of the LhS. (1119), which may be of an earlier date than those of D.B., ce ci occur with nearly perfect regularity for O.E. (t/e), (t/i).

B. (t/) in medial position.

Here the Normans had both (t/) and (ts) not only after a consonant, but also between two vowels, at least when e followed: manche < manicam, herce < herpicem, lecher < leccare, sace < sa- piam etc. This state of things seems to be reflected in such spellings as Lenche, Achelenz, Circelenz, D.B. = Lench, Atchlench, Church Lench (Wo.), Stanlim, D.B. = Stallenge (Dv.), Stanlyncli (1316), F.A., -Wynch, Weinz, F.A. = Winch (Nf.). Vowel + che is also much more usual in D.B. than initial che.

Medial s in Whissonsett and Messing (cf. list B, 1.) is the more likely to be due to A.N. influence, since the Normans had only the sound-combination vowel + (t/e), not that of vowel + (t/i).

C. (t/) in final position after i.

Final c after i developed in the South of England to (t/), whereas the stop was probably kept before a following vowel. Cf. BULBRING §§ 496, 497 Amn. This theory is supported by the phonology of M.E. and N.E. dialectal words. Cf. HITTER, p. 33 ff. In late O.E. we have, consequently, (ik) in the North of England, (it/"), or perhaps sometimes (ik) owing to the influence of the inflected forms, in the South.

The Normans had no final (it/) — at least not until e was lost in such words as riche — and it was, therefore, quite natural that they substituted the well-known (its) for English (it/) 1. That we have here to do with a real pronunciation (its), later on (is), is

Examples from O.Fr. text are adduced by WESTPHAJL, p. 26. For instances of 4z instead of -ich in personal names see SKEAT, Pl.-Ns. of Cb., p. 69; RATHMANN, p. 45.

Anglo-Norman influence on English place-names $7

obvious from rhymes like Dunewiz: escriz, Fantosme, v. 872 = Dunwich (Sf.); Sanwis: recuittis, Gaimar, v. 2429 = O.E. Sandwic, A.S. Chr., Chenris: osciz, Gaimar, v 1789 = O.E. Keneric, Searle; Sanvis: envis, Norviz: escris, Guill. le Mar, v. 17187, v. 14120.

By the side of -«> we find in D.B. and later -ic, -ich. For instances cf. STOLZE, p. 44 f., RATHMANN p. 44 f., WESTPHAL, p. 23. It is very difficult to decide on the value of these spel- lings. In French words final c, ch stand for (k) in D.B. Cf. HILDEBRAND, p. 359. It is, however, not impossible that -ic, -ich are only spellings which were also pronounced (its). We can quote at least one rhyme in favour of this view, viz. Evrewic: Edelfriz, Gaimar, v.* 1005. This instance is of particular interest, as ~ic corresponds to Northern (ik). Cf. also the place-name Warwick, which in R.Br., p. 113, exhibits the regular form War- wik, but in Langtoft's chronicle I, p. 474, appears as Warwys: repris.

The spelling Norwece, Paston Letters (see NEUMANN, p. 72) may be a late M.E. instance of A.N. s for English (t/). Cf. Norwyce, Langtoft I, p. 474, corresponding to Norwiche, R.Br., p. 208.

From the examination above we seem justified in drawing the conclusion that A.N. (ts) was generally substituted for English (t/): (1) initially before e and i, (2) medially before i after a vowel, (3) finally after i. In positions where the Normans had (t/) as well as (ts), both these sounds seem to have been used in rendering English (if).

The French and English forms existed side by side for many centuries, as is amply proved by the numerous double forms in the lists.

1 Loss of c. after i we have noted in D.B. only in two rather dubious cases, viz. Porbi, Porbiche = Purbeck (Do.), Tori = Torridge (Dv.) < O.E. Toric, Earle. The forms without c are, no doubt, only spellings on the ana- logy of such double forms as Albert, Alberic, Baldri, Baldric, D.B. etc. See HILDEBKAND, p. 349; MACKEL, p. 146. — Torix, D.B. < Toric, may in the same way be compared to such pairs as Perdix, Perdriz, E.B., and is best explained as a sort of false latinizing. As for modern Beverley (Du.) < O.E. Beoferlic, A.S. Chr., Beverli: transi, Gaimar (cf. WESTPHAL, p. 23) we refer to § 8. The form Beoferlic in the A.S. Chr. is hardly etymologically correct; the c is either due to faulty orthography, or to A.N. sound- substitution.

<$ E. E. Zachrisson

We have examined several documents of a somewhat more recent date than D.B. in order to ascertain the frequency of the French forms with initial ce, ci in comparison with the English ones in che, chi. From this examination all places in -cester, -Chester are excepted, as also the few other cases in which the A.N. forms prevailed: Li.S., ce, ci, with great consistency; Inq. El. and I.C.C., ce, ci and che, chi, of which the former are a little more 'usual than the latter; Fr.Ch., ce, ci and che, chi. In the Index (12c.) we find by the side of numerous ch- forms a few in ce, ci. The Np.S. has only ch (8 X). The O.E. personal names in Gaimar appear generally with ce seldom with che. Cf. RATHMANN, p. 57. From the beginning of the loth cent, the French forms are in an absolute minority. We have gone through the first volume of the Ch.R., and only found the following in- stances: Cistelet = Chistlett (Ke.), Ceddre (1227) = Cheddar (Sm.), Certesey = Chertsey (Sr.), Cettre = Chitierne (Wl.). A few more may be added from the Index: Cercelle (Hen. III.) = Churchill (Ox.), Cerde (Edw. I.) = Chard (Sm.), Cisselberg (1250) = Chisel- borough (Sm.). In the T.N. and the H.R. c is, on the whole, very rare; we have noted only a small number of cases. In the 14th and 15th cent, ch is the rule. The exceptions are few. The Index has Cedene = Cheddon (Sm.); in the F.A. we find Seriton (1346) = Cheriton (Ke.), and a few more as late as 1428: Ceder = Cheddar (Sm.), Cynnok = Chinuock (Sm.), Cicelbere = Chisel- borough (Sm.). The French forms, which predominated in D.B. and the Li.S., have, with a few exceptions, quite disappeared, 300 years later.

We will now proceed to examine more closely the cases in which the A.N. forms have prevailed.

In place-names from O.E. ceaster the A.N. influence has been of a very thorough-going nature. To account for this we must bear in mind that a great many of these places were in the Middle Ages very important and flourishing towns, which played an important political or ecclesiastical part. That this latter circumstance was apt to favour the French influence, we have already pointed out. Cf. Introduction, p. 15. Indeed in a few cases where the English forms have gained the victory, the A.N.

Anglo-Norman influence on English place-names 29

forms are still in use — but only as episcopal signatures: Cicestr = Chichester, Cestr = Chester 1.

That many French forms were more in use and had, so to speak, their stronghold among scholars and book-learned men, who often may have known Latin and French better than Eng- lish, is sometimes clearly indicated by the very distribution of the forms. In a collection of La. Records, Ducatus Lancastrise, Inquisitiones post mortem (1485—1625), Manchester, the most important place of this county, appears both as Manchester and Mancestre in the Latin parts, but only as Manchester in the English parts of the book. — In Magna Britannia we find Lanchester as the regular form of the Northumbrian town of that name, whereas an abbey in the immediate neighbourhood of the town is spoken of as Langcester 2.

For some places the A.N. forms are in such a majority that we hardly meet with any English ones. This is the case with Gloucester, Leicester, and Worcester, which we have found in their native forms only in La3amon and in the Proclamation of King Henry. In La3amon's Brut the forms in -Chester seem — in spite of Wace's invariable -cestre — to have been the original ones. Only here we find JExechcestre and Cherinchestre = Exeter and Cirencester, and the few instances of -cester in the MS. A. are, no doubt, the work of the copyist. It does not at all astonish us that La3amon, the great patriot and sworn enemy of the French intruders, should prefer the native forms. In the MS.B. the copyist has made use of the cester- forms in all cases where they are still found in Mod. English.

In course of time the native forms have slowly gained ground, so that many French forms of frequent occurrence as late as the loth c. are now replaced by the corresponding Eng- lish ones. In R. of Gloucester's chronicle we find -cester as well as -Chester for nearly all modern chester-towns, while Capgrave about a hundred years later has nearly the same distribution as Mod. English.

1 Of. THE CHUKCHMAN'S POCKET BOOK, 1906, p. 179 ff., where we note some more archaic forms in the same use: Cantuar, Winton, Exon, Norvic, Oxon, Dunelm, Sarum etc.

2 Lancester used as a personal name is recorded as late as 1790. Of. BABDSLEY, Die., p. 466.

30 R. E. Zachrisson

There is, however, even in some works of the 18th c. not exactly the same proportion between French and native forms as there is now. A comparison between all places in -Chester and -cester here treated and their equivalents in Camden's Britan- nia shows that, at least in one case, Carnden prefers the French forms, viz. Silcester, I, p. 147 = Silchester (Ha.). Though Dor- chester (Ox.) keeps its regular form in the tex4 itself, it occurs as Dorcester on the appended map of Oxfordshire. - - Modern Outchester (Nb.), Ulecester, T.N., is still called Ulecester in M.Br. l.

That the A.N. forms have prevailed in Cerne (Do.) is by no means astonishing, for the place was already in O.E. times famous for its old abbey, which is supposed to have been founded by St. Austin, the apostle of the English. See CAMDEN I, p. 56 2.

Just as the prevalence of the cester- forms in many small places may be due to the influence of such great and import- ant towns as Exeter, Gloucester etc., Cerne may have contri- buted to the victory of the A.N. forms in Cerney, Cerney Wick etc. 3.

1 In this place we will call attention to a few places in -cester the names of which are now obliterated from the English map. In the D.B. of Essex we find Effecestre, and in the F.A. for the same county, Wibcester. More examples occur in CAMDEN: — »In this part of the county (Li.) stood formerly the city Sidnacester, once the seat of the Bishops of those Parts, who were call'd Bishops of the Lindifarii; now neither ruins nor name are in being*, (I, p. 571 f.); »In this place (Le.) stood once a very nourishing city call'd Cley-cester», (I, p. 531); » Whoever views the neighbouring heap of rub- bish, which is now called Colecester* (Nb., II, p. 1085).

1 The great veneration which the place enjoyed is piously expressed by the following curious etymology in La3amon, v. 29674 ff.:

pene stude he (St. Austin) cleopede Cernele.

cerno, cernis;

fat is Latin ful iwis.

cerno an Englisc leodc ;

ich iseo swa hit is iqueoen.

el is Ebreounsc;

pat is godd ful iwis.

pen? tun he cleopede Cernel;

ich iseo drihten iwildel.

' We may add that association with the etymon cerne, seme, cherne (Godefroy), not uncommon in French place-names may have contributed to- give credit to the A.N. forms. In the Topographical Dictionaries of Marne-

Anglo-Norman influence on English place-names 31

Besides Whissonsett and Messing, where medial (t/i) has been replaced by (tsi) > si, we have noted one case of s for me- dial (t/) before a following et viz. Ticehurst (Ss.), pron. (tais). Old forms are Tycheneherste (1180), Cumbwell Ch.; Tycheshurst (1295), Ch.R.; Ticehurst (1321— 30), Ticheshurst, Tyshurst(138l—Sb\ Tysherst (1399—1401), P.R.; Tycheshurst (1426), Tyseherst (1543), Index. The first element seems to contain O.E. ticcen, Mdf., p. 134. To account for s in this word it is not necessary to assume French influence; ch may have been dropped before s after the syncopation of unstressed e. Cf. Tisted (Ha.) < O.E. Ticcestede, Birch 786. A pronunciation (tis) would then have been the regular one. The occurrence of (ai) obviously points to French influence, unless it can be explained as due to spell- ing-pron. Cf. the discussion on Riding, Clive § 6.

Although several instances of French (ts) for English (t/) in final position have been noted, the French forms have not prevailed save in a few cases. Chatteris (Cb.), already in the Conqueror's time a famous abbey, appears still in a form of pronunciation similar to the one which was used by its French inmates. - - The little river Hiz, on which the old town of Hichin is situated, has probably derived its name from the town, and exhibits the French form. - - To judge from the form Dice, D.B. — where, however, c may stand for s - - Diss (Nf.) should be a third example of the French forms having prevailed.

Only few attempts have been made to elucidate the ques- tions we have treated in this paragraph. MORSBACH, Urkunden, p. 88, has a short note on Gloucester, Leicester, Worcester, the modern forms of which he rightly ascribes to French spelling and pronunciation without defining more closely the nature of the French influence.

WESTPHAL, p. 22, makes several objections to the theory of sound-substitution: »Die Stufe (t/) war dem Gemeinfran-

zosischen des 12. Jahrhunderts gelaufig vor e in Wortern wie cheval (= lat. caballus], dem Pik. in Wortern wie chest = ecce iste.» The force of these arguments is considerably reduced if we con- sider: (1) that e. A.N. is almost free from Picard influence; (2)

and Calvados we find several places called Cernay with old forms like Cerneium (1260), Cyrneium (1234), Cernacum (1303), Cerneium (1310), Sernayum, Charny (14 c.), Cernay (1411).

32 R. E. Zachrisson

that the number of such words as cheval was very inconsiderable in comparison with the many others which had ce; (3) that in the earliest A.N. texts cha, ca are in those words quite as usual as che, ce, and may have been still more so at the time of the Conquest.

A theory of MAYHEW, Academy 1240, p. 17 (cf. also TAYLOR, Names, p. 32 ff.), according to which the modern cester- forms are confined to the old Mercian Kingdom, where in O.E. times W.S. ceaster, Northumbrian caster appears as cester, is easily confuted, as a glance at the lists of forms heading this paragraph is sufficient to show that Chester and cester occur promiscuously in the most different parts of England, and, conse- quently, dialectal influence is left out of the question.

Appendix A. The symbols c and ch in early A.N. and M.E. texts. —

To judge with some certainty of the various forms in which the place-names treated in this paragraph appear, it will be necessary to establish what values are to be assigned to the symbols c, ch in A.N. and English texts from the time after the Conquest.

1. Value of c, ch in A.N. texts.

With the value of (k) ch was used already in Latin origi- nally to render Greek / in words borrowed from that language. Cf. MAOKEL, p. 130. Possibly in imitation of this Latin habit we find it in some very early (8th and 9th cent.) French or rather Frankish documents used by the side of c to render Teutonic (7) before consonants, and strongly aspirated (h) before vowels in proper names like Chlodovius, Charialdus, Sygichelme etc. <*hlupa, *harja, *helma. Cf. WALDEMATH, Die frankischeu Ele- mente in der franz. Sprache (Diss. Strassburg 1885), p. 40, p. 42, p. 44. In both cases the value of ch was probably (k) as in Latin. This (k) has been handed down to our time in French Clovis, the archaic form of the name of first Merovingian, whose modern correspondence is Louis, and perhaps in chemise, Vulg. Lat. camisia. Cf. MACKEL, p. 131.

Anglo-Norman influence on English place-names 33

In the earliest A.N. MSS. ch has certainly the double value of (k) and (if). In fact, h is put after c as a sort of diacritic sign, denoting that ch followed by a certain vowel is to have a pronunciation different from the one which is generally assigned to c before the same vowel.

The following short analysis of the use of ch and c before the different vowels is chiefly based on investigations of e. A.N. texts by BUHLE, Das C im Lambspringer Alexius etc. (Diss. -Greifswald 1881), VARNHAGEN, Das altnormannische C (Zeitschrift fur Rom. Phil. Ill), and BEETZ, C und Ch vor lat. A.

ch and c before a < Lat. ca in texts of the 12th c. are by Varnhagen and Buhle taken to denote the same sound probably (t/). BEHRENS, p. 204, has pointed out, however, that, as far as the French loan-words in English are concerned, c in this position always denotes (k), while ch stands for (t/). We are, consequently, obliged to assume with Beetz that in Norman and A.N. both pronunciations existed, (k) originally representing the •dialect of the North of Normandy, (t/) that of the South. In the course of the 12th cent, ch gets the upper hand, at least orthographically, probably on account of the growing influence of the Isle de France dialect. Cf. BEETZ, p. 66 f.

Before i, ch denotes (k), c stands for (ts) < Lat. ci etc.

Before e, ch denotes (k), c (ts) < Lat. ce etc. In one case, however, ch may stand for (t/), viz. in words originally containing Latin ca in open syllables. In such forms as rief, ceval < Lat. *capum, caballum, c has been kept as the historic spelling, but may have the value of (t/).

Picard (t/e), (t/i) < Lat. ce, ci etc. is in A.N. rather unusual and of sporadic occurrence. For references cf. STIMMING p. 233. His earliest examples are from Les Eois (c. 1170) and Brandan (c. 1200). Those are, however, looked upon as mere mistakes by BUHLE, p. 24, and SCHLOSSER, p. 75. WAHLBERG, Le Bestiaire de Philippe de Thaiin (Lund 1900), p. XCIV, takes a similar view, when judging of a few similar instances from this text. The original had c = (ts); ch is probably a false spelling.

Before u, ch and u denote (k).

Before o < Lat. o, c and ch denote (k), but before o < Lat au, ch has, of course, the value of (t/).

Our results may be briefly summed up as follows:

3

34 R. E. Zachrisson

c(d) = (k) : camp < campum ; ch(a} = (if) : champ < campum.

c(i) = (ts) : cil < ecce ille ;chi=l. (k) : chi < qui, 2. (t/) spo- radically and in comparatively late texts: Jrunchist, Ipomedon, v. 2390. Cf. STIMMENG, p. 233.

c(e} = 1. (ts) : cent < centum, 2. (t/) : ceval < caballum.

ck(e) = 1. (k) : unches < unquam, 2. (t/) : cheval < caballum, jeche, Ipomedon, v. 875 < vicem.

c(u), ch(u) = (k) : cure < curam, auchun < *alicunum.

c(o) = (k): cor < comw; cfe(o) = 1. (t/): dtase < causam, 2. (k): eschor dement, Bestiary, v. 2776 < cordem.

Apart from position before a certain vowel, ch may always denote (k) in some learned words, especially personal names derived from Greek or Latin, e.g. Achanius, Wace, Acharies, Alexius, see BUHLE, p. 20, p. 21.

Contrary to what is supposed by HILDEBRAND, p. 359, we are of opinion that c and ch in D.B. are used in conformation to the rules given above. In purely French words ca alternates with cha, ce occurs for Latin ce : Salceid < Salicetum = Saussay ; we find c(i)e by the side of ch(i)e < Latin ca in open syllables : Cheure, Chieure, Cieure < capram etc. Cf. HILDEBRAND, p. 333, p. 342.

It may be of some interest to give a short account of the various ways in which the sound of (k) is expressed before e and i in some of the early documents we have consulted. D.B. has mostly ch, more seldom Jc, qu (cf. STOLZE, p. 42 f.); a few instances of c may either be explained as mere errors, or the forms have been more or less directly copied from O.E. originals in which c, of course, had the double value of (k) and (t/). The same explanation may be adopted for sporadic c by the side of ch, k in Gaimar: Chenret, v. 1566, Renret, v. 1599 < O.E. Cen- rcd, A.S. Chr., but Cenwalh, v. 1403, Cenwaille, v. 1356 < O.E. Cenwealh, A.S. Chr. The Lincolnshire Survey, this very old (1119) and as to its orthography much more consistent document than D.B., has ch almost exclusively. The LEI. and the I.C.C. (1180) have ch and /;, of which ch is the more usual; the R.B., ch and &; the Index, ch and 7c; the Fr.Ch., ch and k. In the short but very interesting Northhampton Survey (end of 12th c.) we only find & (5 X). In Westphal's examples from English chronicles it strikes us that /• is much more usual than ch. After the 12th cent, we have not

Anglo-Xorman influence on English place-names 35

been able to find one single instance of ch—'k before front vowels. On the other hand ch, now as well as in .earlier MSS., occasionally occurs with the value of (k) before back vowels and consonants: (Jhneija (1268), Ch.R. = Knaith (Li.); ( Walter of) Corf- gate, Choriet, Corefat (c. 1275), Gl. Records. Thus ch predomi- nates in the earliest MSS., but is in the course of the 12th cent, slowly superseded by k.

2. Value of c, ch in early English.

We will now try and set forth the most salient points in the use of ch as an orthographical symbol in English. As to literature on the subject cf. KLUGE, Paul's Gr., p. 990; RITTER, Die Verteilung der ch- und &-forrnen, p. 5, and above all the excellent account given by WYLD in the Transactions of the Philological Society, 1899—1902, p. 140 ff.

ch was very often used both in O.E. and M.E. to denote the guttural or palatal spirant (/) 1.

In early M.E. it seems, moreover, to have been adopted from French in the special use of distinguishing between (k) and (t/). At first h after c was, just as in French, only a diacritic sign; consequently, ch could stand both for (k) and (if). Towards the end of the 12th cent, ch occurs more frequently, and has now nearly always the value of (t/). In some texts from the beginning of the 13th c. (Ancren Riwle, La^amon etc.) final c may occasionally have the value of (t/); otherwise we may con- sider ch for (t/) by this time as an orthographical rule. The only exceptions to this rule are, of course, M.E. copies of O.E. origi- nals, in which the old spelling either prevails, or is at least to a great extent preserved by the side of the new one. Lastly, we must not forget that in English on the whole ch may always denote (k) in learned words, borrowed from Greek or Latin, as also in a great many foreign personal names. Thus, although Orrmulum with very great consistency writes ch for (t/), we find just in the personal names of this text some instances of ch =

A) In O.E. a few sporadic instances of ch = (k) have been noted. Cf. Miss LEA, Language of the North. Gloss (Anglia XVI), p. 128, BOUTEKWEK, North. Evang. CXXXVIII. Cf. also WYLD, p. 140.

36 fi. E. Zachrisson

(k). See REICHMANN, Die Eigennamen im Orrmulum, p. 81 (Stud. zur Engl. Phil., herausg. von Morsbach, XXV).

The following account of the use of ch in some of our earliest M.E. MSS. tends to prove the correctness of what has been said above. If not otherwise stated, c is the usual way of denoting both (k) and (t/) in the texts considered below.

Laws of Canute (MS. A, c. 1120), ch = (k):'fokhisne; ch = (t/): michel Cf. L. WROBLEWSKI (Diss. Berlin 1901), p. 52.

Textus Roffensis (MS. 1130—50), ch = (k): Falche-Jalchen- ham. Cf. W. GORNEMANN (Diss. Berlin 1901), p. 44.

A.S. Chr. (1137), ch = (k): rachenteges.

Herbarium Apuleii (MS. c. 1150), ch = (t/): drenche, wyrche. Cf. BERBERICH in Anglist. Forsch. 5, p. 29.

Kentish Gospels (MS. Hatton 38, c. 1150), ch = (tj) quite often; ch = (k) principally in foreign words. Cf. WYLD, p. 141.

Poema Morale (MS. Lambeth 487, c. 1170), ch for (t/) seems to be the rule.

Abingdon Chartulary (MSS. 1200, 1260, late copies of -O.E. charters), ch = (t/) quite often ; ch = (k) : Dorchecestere, Bochelande. Cf. F. LANGER (Berlin 1904), p. 71 £.

M.E. version of the Rule of Saint Benet (MS. c. 1200), ch = (t/) quite often. Cf. J. TACHAUER (Diss. Wiirzburg 1906), p. 9.

O.E. Homilies (MS. Vesp. A 22, c. 1200), ch = (tf) quite often; ch = (k): dierchin. Cf. WYLD, p. 141.

La3amon's Brut (MS. A, c. 1222, MS. B, c. 1275), ch = (if) is the rule; ch = (k}: awachede, A, v. 2685. Cf. LUCHT (Palaes- tra XLIX) p. 73.

Kentish Sermons (MS. Laud 471, c. 1200—50) ch is always used for (t/). Cf. WYLD, p. 142.

Consequently, in the beginning of the 13th cent, ch is nearly always put to denote (t/); ch = (k) in native words is after this time extremely rare. In the Paston Letters we find one instance, i.e. dronchyn = druncen; see NEUMANN, p. 73. For instances of ch = (k) in learned words see NEUMANN, p. 73, RU- DOLF, Die engl. Orthograhie, p. 40, KOEPPEL, Spelling-pron., p. 31.

Anglo-Norman influence on English place-names 37

Appendix B. The transition of (ts) to s in O.Fr. The date of the D.B. MSS. -

The sound-combination (ts) became s in both Continental French and A.N. Authors do not agree as to the time when this transition took place. To judge from rhymes in Computus and Gaimar final (ts) of various provenience seems to have become s already towards the beginning of the 12th cent. Cf, STIMMING, p. 230. Unfinal (ts) may have been kept somewhat longer. NYROP, Gramm. § 404, assumes for the transition so late a date as the 13th cent. MEYER-LUBKE, Gramm. 1 § 406, thinks it goes as far back as the 12th cent. Several early spellings with so and s for c < Lat. ce, ci etc. may be quoted in favour of this view. See STIMMING, p. 232 f. Cf. also Scessuns (1127), A.S. Chr. = Sois- sons <. Suessionis (QUICHERAT, p. 17); Scand laudan, ibid. = Saint Lo (Manche). The latter example is perhaps not quite so con- clusive; See often occurs as a shortening of Sancte, and, conse- quently, Scand may be an orthographic blending of Sancte and See.

That nearly all French loan-words in English exhibit s does not prove very much, as on the one hand the unusual initial sound- combination (ts) may have been replaced by 6-, on the other assimilation of (ts) to s is usual in English itself. Cf. KLUGE, Paul's Gr., p. 1012. In Fitz the old pronunciation has been kept, in spite of (ts) becoming s quite early in this posi- tion. We must, however, keep in mind that Fitz mostly occurred as part of a personal name, in which, contrary to the rules ordinary words are subjected to, an archaic pronunciation may very well have been preserved. As for Mod. E. assets < O.Fr. asez there is strong evidence for believing that the spelling has influenced the pron. For instances of early forms (16th c. and later) see N.E.D.

A very curious and hitherto unnoticed circumstance is that in D.B. s, sc, c are interchangeable in the spelling not only of the English, but also of some purely French words, viz. Corcel, Corscella, Coursella = Courcelles, Wissand, Wiscant = Wissant. Cf. HILDEBRAND, p. 334, p. 345.

The following instances of interchange in English words may also very well be worth mentioning.

38 R. E. Zachrisson

1. c for O,E. s: Nacefeld = Nesfield (Yo.), Nesse-, Nesefeld, Yo. K.Q., Yo. N.V. < O.E. nces, ness, Mdf., p. 97; Torgis, Exch. D.B. of Dv. against Torgic in the Exon version < O.E. Thurgils, Searle.

2. * for O.E. c = (t/): Ciltendis, -die = Childerditch (Es.), Chiltendich, H.R., T.N. < O.E. die; Belicolt, Berisout = Bircholt (Ke.), Biricholt (1204), Ke. P.F. < O.E. %rce, fync, Mdf., p. 23; Cecandene, Secendune = Checkendon (Ox.); Loteis = Lutwyche (Sh.), Lolwich, H.R. < O.E. wic; Hancese = Hanchurch (St.)

3. sc for O.E. c = (t/): Beriscolt = Bircholt (Ke.); Flescinges = Fletching (Ss.), Flechyng (14 c.), C.R., P.R.

But the graphic confusion goes further still. O.E. sc = (/) is in D.B. generally rendered by sch, sc, s (cf. STOLZE, p. 45), and c appears not unfrequently even for such an s, sc, sch.

4. c for O.E. sc = (/): Celdestone = Shalstone (Bu ) -- the Ch.R. have sc, the T.N. sh, the H.R. s — < O.E. sceoWe, Mdf., p. Ill; Celdefordam, Scaldefort = Shalford (Es.), Scaldeford, H.R., Shaldeford, F.A. < O.E. scealde; Ece = Ash (Ke.) < O.E. <esc, Mdf. p. 9 ; Keilmerse, Cailmarc (the latter form may also be due to confusion with marca, marc, very usual in Norman place- names, cf. QUICHEKAT, p. 121) = Kehnarsh (Np.); Celdeslai = Shelsley (Wo.), Sceldeslege (12 c.) (cf. DUIGNAN, Wo. Pl.Ns., p. 148); Caldeslai = Shelsley Walsh (Wo.), Sceldeley Walleys (1275) (cf. DUIGNAN, p. 148); Calvrestot, Celrestuis = Shelvestrode, manor in East Grinsted (Ss.), Shelvistrode (1411), Ss. S.R. < O.E. scylf, scelf, Mdf., p. 114; the second element of the D.B. form is probably due to association with such usual Norman place-names as Hottot, Thuit, see JORET, p. 53, p. 81.

This double use of c for O.E. c = (if) and O.E. sc = (/) will perhaps account for a few more very peculiar spellings in D.B. By the side of regular Tacesbroc we have noted the curious form Taschebroc = Tachbrook (Wa.). If sometimes c could stand not only for c = O.E. (t/), but also for s, sc, sch = O.E. (/), sch may here have been thoughtlessly written for c. Moreover, Chenet- jeldam by the side of Scenefeldam = Shenfield (Es.) is perhaps nothing but one more instance of c for s, ch being occasionally used in D.B. to render O.E. c = (t/) before e.

All these comparatively numerous cases of orthographic in- terchange between s, sc, c in D.B. seem to indicate that, at the

Anglo-Norman influence on English place-names 39

time when the present MSS. of Liber Censualis were written, the transition of (ts) > s had already commenced. On the other hand, as such spellings are rare even in texts from the middle of the 12th cent., we can hardly account for their appearance here without assuming that the Exon and Exchequer MSS., contrary to a general belief, are not originals but only 12th cent, copies of untraceable originals.

In documents of somewhat more recent date than D.B. we have noted a few instances of s for c, and vice-versa, in a posi- tion where c generally has the value of (k), and these, con- sequently, are to be considered as still more glaring mistakes by thoughtless scribes. They, of course, owe their origin to the circumstance that, as we have already seen, c in some positions - before e, i - - might have the value of s.

1. s for c: Salthorn, Yo. N.V. = Cawthorne (Yo.), Cal- torn, D.B.

2. c for s.

a. Before a vowel: Cadintone (1210), R.B. = Saddington (Li.); ColeU, R.B. = Sulby (Np.), Side-, Solely, Index; Callay,Yo. N.V. = Sawley (Yo.), Sallai, D.B.; Sampson, the name of a Wo. bishop, appears in the Fr. Ch. as Canson, Sanson.

b. Before a consonant: Clipton (1428), F.A. = Slipton (Np.), besides always Slipton in the F.A.; Kectevene, Kesctevene, Kestevene, H.R. = Kesteven (Li.). For similar instances see RUDOLF, Die engl. Orthographic, p. 37.

§ 2. Anglo-Norman (t) is substituted for English initial ([>).

1. The Anglo-Norman forms have prevailed.

1 . Tingrith (Bd.) < O.E. ping, Mdf., p. 139. Of. pingleah, Kemble, pinghow, Thorpe. 2. Turnworth (Do.) < O.E. porn, Mdf., p. 139, very usual both as the first and the second element of a place-name. Cf. Thornby Np.). Turneby, Ch.R. I; Thornbury (Gl.), Turneberie, D.B., Thornebir', Torneber, H.K., Torneb'i, T.N., Thornebiria, Ch.R. II, Torbyri (1284), F.A. 3. Tormerton (GL). The first element is, no doubt, due to O.E. Thurmcer, Thurmer, Searle. The D.B. form, on the other hand, points to O.E. Thurmund, though n may be an error for r. That we really have to do with an interchange or perhaps rather confusion between -mcer and -mund is, however, evident from the

40 R. E. Zachrisson

form Tormanton on Camden's map of Gloucesterhire, and still more so if we scan the early forms of Farmington in another hundred, which appears in the F.A. as Thormanton (1284), Thormerton (1303, 16, 46), on Camden's map as Farmynton, and in Magna Britannia as Thormerton alias Farmington. 4. Torleton alias Tarlton (GL). The etymology may be an O.E. *Thurhild-, * Thurwcaldtun. 5. Tealby (Li.). All the early spellings point to O.E. Theabul (=peafol\ Sweet, O.E.T., p. 494. There is in Lincolnshire another place called Thealby, which shows exactly the same early forms as Tealby (Teuelby, 1316, F.A. etc.), but in which (f>) has been kept. 6. Torkesy (Li.)? < O.E. Thurkill, Searle. 7. Thorpel or Torpel in Ufford (Np.). The C.R. and the F.A. spell the modern form with £, the Ch.R. with th. The first element seems to be O.E. porp, the second may originally have contained O.E. leah or hyll. Used as a personal name the place appears already in the H.R. and the T.N. as Torpel, in the F.A. (1346) as Torpell, and in the C.R. (1227—31) as Thorpel. 8. Torworth (Nt.). To judge from the early spel- lings, O.E. Thord, Thored, Thured, Searle, may be contained in the first element of the place. 9. Tusmore (Ox.) < O.E. Thur, Searle. 1 0. Tor- weston (Sm.)? < O.E. Thurwulf, Thurweald, Searle.

D.B. Tingrei1, Torneworde2, Tormentone*, ' Tavelesbi'0, Turden- worde*.

12th c. — Fr.Ch. Tablesbeia*. R.B. Tunewrthe*.

\ 3th c. — Ch.R. Torpel7 f Thordworth* I, Torreivorth* II. P.A. Thouesmer9 (1284), Thorveston10 (1284). H.B. Tyngri\ TorkeseyQ (7 X), Th-or- keseye* (6 X), Tursmere*. P.B. Tormer-, Thormerton9 (1292—1301). T.N. Ting'e1, Tormarton*, Tures-, Thuresmere*.

14th c. ~ C.R. Tormerton* (37—39, 39—41, 43—46), Torleton4 (37—39). Cat.A.D. Tevelby*. F.A. Tyngrie1 (1302), Thorneivorthe- (1316), Thormerton*, Theuel- Teuelby*, Teelby'a, Torptll'1 (1346), Tordworth* (1316), Toresmere9 (1316), Thouresmere9 (1346). Gl.Recs. Thor- mertone* (1317). P.R. Tormer-, Thormerton3 (01—07), Tormerton9 (34—38), Thormarton* (21—24), Therleton* (24—27), Torle-, Thor- leton4' (43—45), Theuelby* (27—30), Torworth* (43-45).

1 5th «c. ~ F.A. Tingrie1 (1428), Turneworth* (1428), Thormerton* (1428), Torkeseif (1428), Torpell1 (1428), Thoresmer9 (1428), Torveston (1428). Index. Thormerton5 (1432), Thorlston* (1433).

™

J) The above forms are meant to instance A.N. (t) for O.E. (p) in English place-names. As no O.E. forms are recorded, and there is a contin- ual interchange between initial t and th in n great many M.E. documents, the greatest cautiousness is required in judging of the forms. We take Tealby and Tormerton to be absolutely safe instances; as for the others it is at least highly probable that the O.E. ground-form contained (J>).

Anglo-Norman influence on English place-names 41

2. The A.N. forms have not prevailed 1.

1. Fingerst (Bu.) < O.E. ping. 2. Catthorpe (Le.) < O.E. porp. 3. Winthorp (Li.) < O.E. porp. 4. Cawthorpe (Li.) < O.E. porp. 5. Finedon (Np.) < O.E. ping.

D.B. Catorp*, Tingdene*.

1 2th c. — Index. Winetorp* (1150-60), Wintorp* (1175). 13th c. — F.A. Tinghurst1 (1284), Thorpcat* (1284). H.R. Tingehurst1, Torp-

ha?, Calthorp\ Thingden*. Index. Winetorp* (Hen. III.), Win- • thorp* (Hen. III.), Kaltorph*, lingd* (Hen. III.), Thinderi* (1296).

T.N. Cat-, Cathorp*, Wynnetlwrp*, Catthorp, -torp*, Thingden\ 14th c. — F.A. Tyngeliurst1 (1346), Calthorp4 (1303). Index. Wynthorpe*

(1331), Wynthorp* (1347, 64), Calthorpe* (1335), Thyngden* (1393)

The voiceless lip-teeth consonant ([)) of so frequent occur- rence in the Germanic languages was, at least in initial position, absolutely unknown to the Normans, who in the course of one century and a half had so completely forgotten their own language, that if we except a few loan-words it has only left some traces in the local nomenclature of the districts they occupied. Cf. NYKOP, Gramm. § 13.

The phonology of Germanic loan-words in O.Fr. shows that (t) was always substituted for initial (JD). Cf. MEYER-LUBKE, Gramm. I § 18, p. 41, MACKEL, p. 164 f., and in the case of Normandy such place-names as le Torp, le Tourp < Scand. porp, Torgistorp < Scand. forgilsforp etc.

Not otherwise did the Normans treat initial (\>) in English place-names; (t) was always substituted for ({)). In the earliest documents this (t) is mostly rendered by t. In D.B. t is by far the more usual spelling, though a few instances of th have also been noted. Cf. STOLZE, p. 41. Most 12th cent, records (Inq. El., I.C.C., Fr.Ch. etc.) do not differ much from D.B. in this respect. In the Pi.R. of 1189—90 we do not find one single instance of th for (t) in the numerous place-names < O.E. porp, porn there recorded. Also in the Li.S. t predominates, though we have found th in a few cases, e.g. Hacatorn, Hachethorna = Hack- thorn.

There are, howeyer, two feudal records of the 12th cent, in which we perceive a strong and, no doubt, intentional ten- dency to denote ({>) with th, viz. the Le. and Np. Surveys.

1 Cf. also the account of the use of t and th given below.

42 R. E. Zachrisson

Nearly all place-names in which O.E. (J)) is rendered by t in D.B., appear in the Np. S. with th. The exceptions are very few. In the short Le. S. place-names containing O.E. (J)) occur only four times; all of them are spelled with th.

This tendency becomes general in the course of the 13th cent. Obviously the Norman scribes had now acquired a better knowledge of the difference between English ([)) ancj (t), and seriously endeavoured to keep them apart by denoting one with th the other with t. An examination of the T.N. for Leic., where owing to strong Scand. influence Thorpe is a very usual designation of a place, shows that Thorpe is by far the more usual spelling, and some ten years later in the H.R. for the same country we find only a few instances with t. Interchange between th and t is, on the other hand, not uncommon in the T.N. and the H.R.: Turgtun, Thurgereton, H.R. == Thurgarton (Nf.); Althorp, -torp, -trap, H.R. = Althorp (Li.); Eylestlwrp, -torp, H.R. = Elsthorpe (Li.); Ha- Jcethorn, -torn, T.N. = Hackthorn (Li.) etc. - - In the first vol. of the Ch.R. we have, by the side of regular forms in th, found a few isolated instances of t, viz. Tacheam = Thatcham (Be.); Taxied = Thaxted (Es.); Tefford = Thetford (Li.); Taydon = Theydon (Es.); Tymelesby = Thimbleby (Li.); Turneby = Thornby (Np.).

In the 14th c. t is still more exceptional. We have gone through the second vol. of the F.A., and, in addition to those given in our list, only found the following cases: Timing = Thurning (Ha.); Tyteburst = Theobald street (Ht.); Terefeld = Therfield (Ht.). Trehus and Trolcking, ib. = Three Houses (Ht.), Throckiug (Ht.) are no certain examples, dialectal sound-change being not improbable.

It goes without saying that we cannot expect to find many instances of t for th, fi in purely English works. The following examples from the A.S. Chr. are worthy of note: Torn MS. E = The Isle of Thorney; Turstan, Teobald MS.H., Turstein MS.E. R.Br., p. 27, has Teteford = Thetford (Nf.).

The forms in t due to A.N. sound-substitution have very seldom been generally accepted, probably because the native forms vere often supported by words still living in the spoken language (thorp, thorn etc.) When the etymological identity of the word was not so obvious the French forms might more easily supplant the native ones. In fact, it is particularly with

Anglo-Norman influence on English place-names 43

place-names containing a personal name that this has been the case. This is the less astonishing in consideration of the cir- cumstance that a great many O.E. names in f were ousted by their French equivalents in t.

It still remains to be seen if (t) for (h) in Tormerton, Tealby etc. may have originated in some other way than by A.N. sound-substitution .

Dialectal influence must not be left out of the question altogether, for in the vernaculars the interdental spirant has often been replaced by the dental stop.

In the South of England initial (h) was voiced, and, accord- ing to SWEET, H.E.S. § 728, the voicing had been developed before the llth c. KLUGE, Paul's Gr., p. 1007 Anm., thinks that at least in Ayenbite initial (hj was generally voiced. At any rate, we must start from voiced initial (JD) to account for the sound-change (ct) > (d), which is well evidenced for the South of England in late M.E. times, See DIBELIUS, p. 451; NEUMANN, p. 75. In Octavianus, ed. SARRAZIN (Alteng. Bibl., herausg. von Kolbing, III) there are a great many spellings with d for p in initial — as well as medial — position: de, dan, do, dare, dyder, denJc ; dour,hte, drew etc., as also a few with f for d: fere,

1 Some spellings in D.B. tend to corroborate Sweet's statements as to the early date of the voicing. Initial d •- sound-substitution for (d)? - appears in Dereuelde = Therfield (Ht.), Edrope = Heythrop (Ox.), and Der- nescop = Thurnscoe (Yo.). All these places are situated on a territory where initial (J)) remained unvoiced. STOLZE, p. 49, has, however, pointed out that in the case of N. places S. forms are not uncommon in D.B., so Dernescop etc. may really be instances of (d) for initial (d). On the other hand, in view of the inconsistent use of t and d in D.B., we are scarcely justified in drawing any far-reaching conclusions from only a few spellings.

How are we to account for the following cases of t for initial d, and vice-versa: Todingetona, Dodingedona, Fr.Ch. == Toddington (Bd.); Torsete, Wace, v. 1463 = Dorset; Tantes-, Dantesborne, D.B. = Duntisbourne (GL); Totm'ton, Dodinton, D.B. = Totton (Ha.), Todintona, Ch.R. II; larenteford, D.B., Ch.R., R.B. =Dartford (Ke.) < O.E. Diorente, Earle; Danefelt, Tanefelt, D.B. = Tanfield (Yo.); Tiverington, Deverington (Hen. III.), I.P.M. = Terrington (Yo.). Has the usual orthographic interchange between final and medial d and t been thoughtlessly transferred to d and t in initial position as well, or could the Normans not properly distinguish between Engl. (d) and (t)? SKEAT, Ht. Pl.-Ns., p. 55, thinks that initial t for d sometimes may be due to the influence of the preposition cet, at.

44 R. E. Zachrisson

pefende, forst etc. All of them are looked upon by the editor as errors. See Introduction, p. XII. The same peculiar spellings appear also in another Kentish MS. from about the same time (the second half of the 14th c.), Cott. Jul. D. IX. Cf. also theer = deer, Libeaus Disconus, v. 1133. We are of opinion that d in derik etc. denotes a then common dialectal pronunciation (d)

< (d), while the less numerous cases of J5 for d.may be explained as inverted spellings. In the 16th cent, initial (d) for (d) is evidenced by the English orthoepist Bullokar for Kent and Sus- sex. See KLTJGE, Paul's Gr., p. 1007 Anm. In the present Eng- lish dialects (d) for initial (d) occurs sporadically in the whole S. and S.W. England: (1) in disle and datch = thistle, thatch (Sm., Dv., He.); (2) in pronouns and adverbs (Ke., Ss.); and is (3) par- ticularly well instanced in the combination thr. Cf. E.D.Gr. §§ 310, 311, 313; KRUISINGA § 358. An English grammar of the year 1701 warns against dialectal dunder for thunder. See D. and M. Forster, Anglia XXIV, p. 116. In place-names drop for throp etc. occurs not unfrequently in S. England; Hilldrop (WL), Hullethrop (1310), Index, is one instance.

No doubt, the modern instances of (d) for initial (d) are only remains of a once very common dialectal sound-change. On the gradual disappearance of (v) for initial (f) from many S. Engl. dialects see EKWALL, Jones's Phonography § 588. In like manner (z) for initial (s) is hardly to be found any longer in the Kentish dialect, of which it once was a characteristic feature. See KLUGE, Paul's Gr., p. 1010; E.D.Gr. § 320.

In some Midi. , W. and N. dialects, in which initial th remains ({>), we find (tr) < (t>r), exactly answering to S.English (dr) < (dr). On this point see E.D.Gr. § 313, where we are told that the sound-change occurs in Yo., La., Nf., Sf. FRANZ- MEYER, p. 54, notes one instance also from Bucks.

We have found several cases of (tr) for ([>r) in the names of places which are situated within or in the immediate neigh- bourhood of those districts, so (t) for (|D) is here, no doubt, a provincialism: Threo or Treo (Nf.), Trehoo, D.B., Trehow, H.R.

< O.E. freo + hoh, Mdf., p. 74; Thrandeston (Sf.), » vulgarly pronounced Transon» (see HALL, Mod. English), »alias Franston» (M.Br.) < O.E. Thrandeston, Thorpe; Heythorp (Ox.), ^commonly called Aytrop» (M.Br.), Ethrop, Hetrop, T.N. < O.E. forp\ Trus-

Anglo-Norman influence on English place-names 45

thorpe (Li.), Thursthorp, Thrusthorp, Trussethorp, F.A. < O.E. Thur, Searle. Here belongs perhaps also Tring (Ht.) < prifing. Cf. .§ 6.

In the same manner as, though more seldom, we find ini- tial (d) for (d), by the side of the more common (dr) for dr), in the South of England, we might expect to hit upon some instan- ces of initial (t) for ({>) in those parts of England where (tr) occurs for (pr). However,, cases of initial (t) for (f>), as a parallel to (tr) < (t>r), have hardly been noted here either by Wright or Franzmeyer. Nevertheless, we consider it highly probable that such a sound-change formerly existed, and as a matter of fact it has been evidenced in a few instances from M.E. and e. N.E. times: tanke, Tyndale; ire, tynge, tynlcen, Pride of Life (see DIBE- LIUS, p. 451); tanked tynggs, Paston Letters (see NEUMANN, p. 91); taught, Colet, see SAM. BLACK, Die Schriftsprache in der Lon- doner Paulsschule (Halberstadt 1905), p. 41. In Bartholomew Fair Captain Whitley uses (t) for (|D) in tank and ting. Cf. PAN- NING, Dial. Engl. in Elizabethanischen Dramen (Diss. Halle 1884), p. 34.

From the evidence of the material above — which cer- tainly could be much added to — we seem justified in assuming that initial th in late M.E. has become t and d, respectively, in several English dialects. In the present vernaculars there are still some lingering traces of the sound-change, particularly in the combination thr. The more general occurrence of initial (t) for (J>) on the Isle of Man (see E.D.Gr. § 313; FRANZMEYER, p. 54), where owing to insular isolation dialectal changes have greater possibility of being preserved, may perhaps be pointed to as the last remnants of the transition of initial ([>) to (t). The vernacular of the Isle of Man is placed by Ellis in the same division as the Lancash. dialect, where (tr) frequently occurs for for).

This whole question — as well as that of medial (d) > (d) presents many obscure and difficult points, and would cer- tainly be well deserving of further investigations. It would be of particular interest to fix the exact time of the transition. Some more material is adduced in § 6. The M.E. forms of numerous places in -thorp, -throp would prove a very useful material.

46 R. E. Zachrisson

In the instances we have quoted to illustrate French sound - substitution, no dialectal distribution of t- and d- forms is to be traced, t occurs for th not only in the Midi, and W., but also in the S. parts of England. Consequently, dialectal sound-change is not very probable. The early appearance of the spellings with t also tells against such an assumption.

One more objection may be made to our explanation of the /-forms; perhaps they are due to assimilation in connected speech if, for instance, the preposition at was prefixed to a place- name. Cf. atte < at pe, Ormulum etc. This explanation is, however, not very probable for, as far as we know, no cases have been recorded of initial (p) having been assimilated to (t) in stressed syllables. Cf. SCOTT, Attractions in English in Trans, of the Amer. Phil. Ass. XXV, 1894, p. 88. Such forms as Eandeston, T.N. = Thrandeston (Sf.), Redland (Gl.) = Thirdelond (1346), Yriddelond (1284), F.A., are perhaps best explained from at Trandeston, at Triddelond with French or dialectal tr. Even if ({)) was first assimilated to (t) the result has been complete loss, not alteration of ({)) to (t). Cf. qwat at < hwcet fat, Paston Letters, DIBELIUS, p. 451.

NOTE. In this place we will point out two instances in which the interchange of t, th in the early spellings is merely due to graphic confu- sion between two personal names from different roots: (1) Tibaldstone Hundred (Gl.), Tedboldestan , D.B., Thebaldestan, Tedbaldestan, H.R., Tibald- stone,: Tedbaldston, Theobaldeston, F.A.; (2) Tilshead (Wl.), pron. [:ta'lsad]? ELLIS, V p. 58*, Theodulveside (1 XX Tidulfhide (4 X), D.B., Thidulfhide (1 X)! Tidulfeshid (8 X), T.N., Tidolesside, Tydellesside, H.R., Tidulveshida, R.B., Ty- dulveshide, Thidulveshide (1216—25), Tydolveshyde (1381—85), P.R.

To judge from such forms as Theodulveside and Theobaldeston, the first element of the above-mentioned place-names would contain O.E. peodbeald, and modern t might be due to A.N. sound-substitution. Cf. Germanic piud- bald, peodbald = O.Fr. Tibal, Tibalt, Tedbalt = Mod. Fr. Thibaut, MACKEL, p. 128, p. 165. But if we scan the O.E. records it appears that peod- is rather an unusual compound in personal names, whereas compounds with tid- are very usual from the earliest times. In the Liber Vitse of Durham the only compound with peod- is theodric (instanced 4 times), whereas tid- occurs more than 70 times with a great variety of suffixes. Cf. MULLEK (Palaestra IX), p. 86, p. 90. The O.E. ground-form of Tilshead is, no doubt, *Tidwulfeshigid or -hid. As to Tibaldstone the spellings pointing to peod- beald are undoubtedly in a majority, but for all that we do not hesitate to derive the name of the place from an O.E. *Tidbealdestune. Some early forms with e for i may be due to A.N. spelling, cf. STOLZE, p. 17. Nor is loss of d before fr,

Anglo-Norman influence on English place-names 47

as in Tibaldstone, F.A., necessarily, due to French influence, but may be a case of assimilation in English itself. Similar cases are Tetbriston, D.B., Tibbrihtonia (1181), Index = Tibberton (Sh.); Tidbrihtingctun (a 978), Kemble 603, Tidbertun, D.B., Tibrihton, H.E. = Tibberton (Wo.).

Appendix. Orthographic interchange between t and th for etymological (t). -

Particularly in some 13th and 14th c. records we note the curious circumstance that th is much more often found for ety- mological t than t. for th. Thus in the F.A. we find Toft as well as Thoft for O.E. toft, Theversham and Teversham = Tevers- ham (Cb.) etc. Similar instances occur also in the Index, e.g. Thateshale (1304), Tatheshule (1332) = Tattershall (Li.), Stoivlange- thot (1393) = Stowlangtoft (Sf.). Though more seldom, th for t appears as early as in D.B.: Toffh = Toft (Cb.); Bagathora, Exon D.B., Bagatora, Exch. D.B. = Bagtor (Dv.) < O.E. tor(r), Mdf.T p. 135, etc. - - th for O.E. t has, moreover, been noted in Eng- lish place-names from many O.Fr. texts. For instances see WEST- PHAL, p. 17. Rhymes like north: Oxenford, Gairnar, v. 5076r etc. indicate that this th was pronounced (t).

In our opinion this confusion between th and t is partly due to the inability of the Normans to pronounce English (|j). Initial th as an orthographical symbol was known in the French language, where it had the value of (t). Cf. SwEEt, H.E.S. § 600, N.E.Gr. § 380. This accounts for the use of th = (t) in early texts as D.B. In the 13th and 14th cent, the interdental spirant was lost in A.N. in all positions. The grea,t inconsistency in the use of t and th in records from that time may be due to the vain attempts of the Norman scribes to keep the two sounds orthographically apart. It is, however, very curious that at least in the 14th cent, records th is much more often used for t than t for th. Under such circumstances the most probable explanation of the interchange would perhaps be that th was used: (1) to denote ({>); (2) alternating with t to denote (t).

WESTPHAL, p. 18, assigns a special phonetic value to the symbol th for t in the Engl. place-names of the O.Fr. chronicles. He is of opinion that h is put after t as a diacritic mark to distinguish the Engl. aspirated (t) from the French one. This cannot very well be correct, for t and th are interchangeable

48 R. E. Zachrisson

also in the spellings of French place-names. Any number of instances might be adduced, but the following will suffice: Avrehuth (1170), Averhout (1139) = Avroult (Pas de Calais), see

A. DE LOISNE, Les formes originales des noms de lieu du P. de C., p 6; Torigne, Thorignie, Benoit, v. 33641; TiergeviUe (1308), Thiergeville (1338) = Tierceville (Eure); Tylli (1293), Thilly(Uc.) = Tilly (Eure); Totteneium (1061), Thonaium (1186), Toeni (1205), Thooniacum (1227), Thoeneium (1239), Thony (1392), Tony (1419) = Tosny (Eure). See Die. Topogr. de 1'Eure. Of. also the various old spellings of Thevray, Thibouville, and Thuit in the same Dictionary.

In many M.E. texts we also find the spelling th for /, particularly in foreign names and in words more or less directly borrowed from Greek and Latin. For examples see BEHRENS, p. 171, DIBELIUS, p. 448, KOEPPEL, Spell.-pron., p. 19 f.

But th for etymological t is well instanced in native words as well. Cf. MOESBACH, Urkunden, p. 105; DIBELIUS. p. 448; NEUMANN, p. 92 f. All these authors are of opinion that th is is a phonetic spelling (»t + Hauch», Morsb.; »eine stark aspi- rierte Aussprache», Dib.). We do not think it is necessary to assign any special phonetic value to Several sporadic instances of th for t. The symbol th being used in some (= learned) words to denote (t), it seems quite natural that the same spel- ling should sometimes be transferred also to the native words: winther for winter analogous to throne for trone, Neumann seems inclined to adopt a similar explanation for some of his examples. Cf. p. 93, »oft mag aber bei der mannigfachen verwendung des th, th fur t blosser schreibfehler sein.»

We must keep in mind that there was an ambiguity as to the use of th. To the French it was only a variant of t\ by the English it was generally associated with the pronunciation (l>, ct). We know that in course of time this ambiguity led to a pronunciation ([)) for a great many words in which th origi- nally denoted (t). To the instances adduced by KOEPPEL, Spell.- pron., p. 19 f., the following may be added: (1) Thanet (Ke.), pron. with ({>), Tanger. All the early forms - - Tenet, Tanet, A.S. Chr., Tenid (a. 679), Tcenett (a. 949), Index, Tcenate A, Tanate

B, La3amon, v. 14737 point to original t. As late as in Camden's time t seems still to have been kept in the pron. Cf.

Anglo-Norman influence on English place-names 49

OAMDEN I, p. 242, »we call it Tenet». Our first instance of the spelling th is from the Ch.R., Thanet (1227). (2) Thame on the river Tharne (Ox.), pron. with ({>), seldom with (t), Tanger, Tamu, A.S. Chr., Tame (1473), Index. There is a river Tame in Staff. — no doubt derived from the same, probably Celtic, root as Thame and Thames — in which etymological t has been kept in spite of numerous early spellings with th.

§ 3. Some difficult initial consonant-combinations are altered in various ways to suit Anglo-Norman pronunciation.

In the Germanic languages there existed several initial •consonant-combinations which were unknown to Romance-speaking people. In the case of loan-words such combinations were, na- turally,, got rid of in one way or another. Thus a vowel - generally e was sometimes either prefixed to (e.g. the pro-

thetic e before s impurum) or inserted between (= svarabhakti) the two consonants. In this paragraph a number of initial consonant-groups in Engl. place-names the pronunciation of which presented difficulties to the Normans of the llth cent, will be dealt with in due order.

A. Anglo-Norman Jcen, Jean, kun, n is substituted for

English kn.

1. The Anglo-Norman forms have prevailed.

Cannoc (St.) < Celtic cnoc. Cf. DUIGNAN, St. Pl.Ns., p. 30 f. D.B. Chenet. Records (12th c.), see DUIGNAN I.e. Chnoc, Cnot, Chnot, Chenot, Canot, Canoe. Index. Gannok (1245).

2. The Anglo-Norman forms have not prevailed.

1. Knapwell (Cb.) < O.E. Cnapa; cf. SKEAT, Pl.-Ns of Cb., p. 36. 2. Knutsford (Ck.) < O.E. Cnut, Searle. 3. Knowlton (Do.) < O.E. cnoll, Mdf., p. 28. 4. Church Knowle (Do.) < O.E. cnoll. 5. Knowston (Do.) < O.E.

1 Cf. MEYEK-LTJBKE, Gramm. I§ 387: — »Le de"veloppement d'une voyelle «ntre deux consonnes se produit dans diffe>entes circonstances. II a lieu tout d'abord dans les mots 6trangers qui renferment des combinaisons de sons insolites.»

2 Cannoc Chase formerly contained a castle and hunting-seat; BAE-

THOLOMEW.

4

50 R. E. Zachrisson

Cnut. 6. Knighton (Dv.) < O.E. cniht. 7. Knowle (Ha.) < O.E cnoll 8 Knapp (Ha.)< O.E. cnap, Mdf., p. 28. 9. Knebworth (Ht.) < O.E. Cne&fca; cf, SKEAT, Pl.-Ns. of Ht., p. 56. 1O. Knaith (Li.). 11. Knapton (Nf.) < O.E. cnap. 12. Kneesal (Nt.), perhaps < O.E. cneo, cf. SKEAT, Pl.-Ns. of Cb., p. 26. 13. Knightley (St.) < O.E. cniht. 14. Knook (Wl.)? < Celtic cnoc. 15. Knaresborough (Yo.). 16. Knarford (Yo.) now lost; cf. SKAIFE, The Survey of the county of York etc. 17. Kneeton (Yo.)? < O.E. cneo. 18, Knottingley (Yo.).

D.B. Chenepewelle1, Cunetesford*, Chenoltune*, Chenolle*, Cnolle*, Chenudestane*, Chenistetone6, Chenol\ Chenep*, Chenepeworde9 , Che- neide10, Ranaptone11, (B error for K\ Chenistelei1*, Cunucheuf Chenaresburg™, Kenares-, Neresforde*6, Naton1"1.

12th c. — Index. Eneshala1*. Pontefract Oh. Nottingham18. Yo.P.F. Gnareburc™.

13th c. — Ch.R. Kneye10, Cnaresburg15, Nottingleya1*. Index. Knapeton^, Keneshale1*, Knyhttesle1* . B.B. Canardesburgus13. T.N. Knolton3.

14th c. — P.A. Knolton*, Knouston6, Kneth10. La. Bees. Knottesford~> forth*. Yo. N.V. Kneton11 (1316).

The group -A*w, which was unfamiliar to French ears, is regularly turned into ken, Jean etc. also in Germanic loan-words. Examples are adduced by MEYER-LUBKE, Gramm. I § 18;NYROP, Gramm. § 494. The name of a place in Normandy, Canapville, Canapevilla (1196), is traced back by JORET, p. 64, to Scand. knappr. - As far as we know, loss of k has not been noted in the Germanic loan-words.

The French forms have prevailed only in the case of Cannoc < cnoc, which with vowel- insertion and change of suffixes appears in D.B. as Chenet. That the Normans substituted the well-known diminutive suffix -ot, -et for the unfamiliar -oc, is by no means astonishing. Cf. Introduction, p. 12. In the 13th cent, the native ending was restored, while the parasitic vowel was kept.

Later loss of h in the combination kn must, of course, be kept strictly apart from the early instances we have ascribed to French influence. In the latter case loss of k is sporadical, and ~k reappears in later spellings.

NOTE. Though O.E. Cnut appears as Chenut etc. in D.B., this French form never prevailed in the place-names. On the other hand Cnut the Great, the famous Danish king who in the beginning of the llth cent, extended his rule also to England, is spoken of in modern works of history both as

Anglo-Norman influence on English place-names 51

Cnut and Canute, pron. (kanjuwt), Tanger. With reference to these double forms FREEMAN I, p. 399, makes the following reflections: — >Cnut or Cnud is this king's true name, and the best Latin form is Cnuto. The form Canutus seems to have arisen from Pope Paschal the Second's inability to say Cnut* Pope Paschal was certainly not alone in mispronouncing this difficult word. Canute is obviously a French rendering of the true English form, as it appears in the A.S. Chr., Knut, or in Rob. of Brunne, Knout, p. 45. We have noted the French form in Langtoft, Kanute, Kanutus, I, p. 354, p. 368, and in Mousket, Konus, v. 15309.

B. Anglo-Norman wer, war, r is substituted for English wr.

1. Wrenbury (Ch.) < O.E. wrenna, Mdf., p. 153. 2. Wroxall (Ha.); cf. Wrocces heal, Kemble 768. 3. Wragby (Li.), probably < O.E. wearg, Mdf., p. 147. Wragholm (Li.) appears as Wargholme (1295), Index. 4. Wrawby (Li.)? < O.E. wearg. 5. Wrangle (Li.); cf. on wrangan hylle, Birch 801. 6. Wretham (Nf.). 7. Wramplingham (Nf.). 8. Wroxeter (Sh.). 9. Wrington (Sm.) < O.E. Wrington, Kemble 338. 1O. Wraxall (Sm.). 11. Easwrith Hundred (Ss.) < O.E. wrtj), wrid, Mdf., p. 153. 12. Wraxhall (Wl.).

D.B. Wareneberie1, Warochesselle* , Waragebi5, Waragebi*, We- ranghe5, Wrel-, Weretham*, Wran-, Waranplinchani1 , Rochecestre* , Weriton9, Werocosale10, Esweride11, Werocheshalle™.

12th c. — Index. Warkebi*, (Hen. II.), Wraggebi*.

13th c. — Ch.B. Wragheby*, Wrangle*. H.B. Wraggeby*, Wroccest'*, Es- tvrythel\ Esewrith11. Sm.K.Q. Wroxhall10 (1284). T.N. Wrag- geby*.

1 4th c. — F.A. Wraggeby5, Wragheby*, Wrauby*, Wrotham*, Wrampling- ham\ Wroccestre*, Wryngton9. Index. Wrangel*, (1302), Wrox- hale10 (1331).

Instances of vowel-insertion between w and r in Germanic loan-words are noted by NYROP, Gramm. § 494, and MACKEL, p. 184. Loss of w does not seem to have taken place in loan-words.

In this particular case the French forms enjoyed a very brief existence. We have not found them anywhere but in D.B.

C. Anglo-Norman Z, w, n, sel, sen is substituted for English, si, sm, sn.

1. The Anglo-Norman forms have prevailed. Nottinghamshire, Nottingham (Nt.) < O.E.- Snotingaham, -stir, A.S. Chr.

52 R. E. Zachrisson

D.B. Snoting'scire, Snoting-, Snotingeham. Pi.B. Notingeham (1131). Index. Notingeham (1153). Florence of Wo. (Mon. Brit.). Snotingaham (MSS. 12c.). Symeon of Du. (Mon. Brit,). Sno- tingnham (MS. 12c.). Henry of Huntingdon (Mon. Brit.). Snotinghamscire, Snotinga-, Notingeham (MSS. 13 and 14c.). MS. Jesus Coll. Oxford (O.E. Miscellany, E.E.T.S. 49), Snoting- ham (13c.). Le. Bees. Snolyngham (1353), Notingham (1336). Index. Notingham (1304). Oapgrave. Notingham, p. 101.

2. The Anglo-Norman forms have not prevailed.

1. Snelson (Ch.); cf. Snelles cumb, Kemble 1052. 2. Sleaford (Li.) < O.E. Sliowaford, A.S. Chr., Slioforda, Thorpe 1. 3. Snarford (Li.), perhaps connected with M.E. snart, adj. severe, used a personal name. 4. Snettis- ham (Nf.)? <; O.E. snytre, adj., wise, used as a personal name. 5. Snoring (Nf.). 6. Smithdon (Nf.) < O.E. smepe; cf. on smepandune, Birch 748. 7. Sneinton (Nt.) < O.E. *Snotingatun 2. 8. Slingsby (Yo). 8. 9. Sledmere (Yo.), perhaps < O.E. sled, Mdf., p. 117 4.

1 There are no less than three places with that name in Line, viz. Sleaford, New Sleaford, and Old Sleaford. For the two latter the French forms were still in use as late as the 18th cent. In Magna Britannia the places in question are called Sleford (Flaxwell Hundred), Lafford, and New Lafford (Ash warden Hundred). It is interesting to note that in the F.A. Old Sleaford appears in 1316 as Vetus Lafford, and in 1401 as Sleaford. The Hundreds of Ashwarden and Flaxwell are contiguous, and the three places are situated on the boundary. -- At Sleaford there are ruins of an old castle famous in Norman times.

2 We cannot refrain from quoting here the following interesting note on Sneinton from Viet. Hist, of Nt., p. 245: — »There are a few difficult questions of identification in the Nt. D.B., one of which has the rare distinc- tion of being raised in a mediaeval law-suit. In 1285 the lords of Sneinton were impleaded by the men of the vill on the ground that, while Sneinton was ancient demesne of the crown, they had increased the rent due from tenements there (etc.). In reply it was urged that Sneinton was no ancient demesne, upon which D.B. was ordered to be searched, when Notintone was found to be duly entered on the king's land. In 1287 a jury found that Sneinton was always called by that name, and that Notintone was part of Nottingham. However, as in documents of the time the name appears as Snotintone or Snadinton, we need not attach much weight to the jury of 1287. »

1 We have, moreover, found two entries of the French form used as a pers. name, viz. Lengebi (13c.), Yo. I.P.M.. and Lengebi (1301), Yo. S.R. — In the vicinity is Slingsby castle, a ruined fortress built in the llth cent; BABTH.

* We do not hesitate to consider Ledemare and Slidemarc, D.B., as Identical. Both are entered* tinder East Riding, the former in Dickering, the

Anglo-Norman influence on English place-names 53

D.B. Senelestune1, Eslaforde*, Snardes-, Snerteforde*, Netesham*. Snes-, Sneteskam4, Esnaringa5, Snaringa5, Metheduna*, Smetheduna* Smeze-, Smeteduna?, Notintone1, Eslingesbi*, Selungesbi*, Ledemare9, Slidemare9.

1 2 th c. — Index. Laford* (Hen. II.), Narteford* (1163). Li.S. Snarteford*. Sempring-ham Oh. Lafford*, -Slafford*.

13th c. — C.B. Snointon1 (1217). Ch. B. Snetesham4* II, Leddemere9. H.R. Laford*, Laffordia?, Sfaarford*, Sneteres-, Sneders-, Snetisham*, Naring'*, Smethe-, Smethes-, Smesdon*, Sneynton1. Index. Snarte- ford9 (Hen. III.), Nartesford3 (1203—05), Slengesby*, Sclengesbys, Eslenggesbi*, Lingeby*. B.B. Lajforde*. T.N. Lafford*, Snetes- ham*, Naringes5, Smethedun*, Snenton1. Yo. I.P.M. Slengeby*, Sledemer9. Yo. K.Q. Slingesby*. Yo. S.B. Sledemer9 (1297).

14th c. — P.A. Vetus Lafford* (1316), Nova Lqff'ord* (1316), Snartford* (1316), Snetesham4-, Naring6 (1302, 16), Snaryngg5 (1346), Naringg^ (1346), Smethdon*. Index. Snoryng'0 (1388), Lengeby* (1301), Slengesby* (1348), Slengisby* (1364). P.B. Lafford* (01—07), Sle- ford*, New or Novel Lafford* (13—17), Sleaford* (30—34), Lajford*, Sleford* (85-89), Snartford* (13—17, 34—38, 61—67), Slyngesby* (01—07), Slengesby*(±8—50}. B.Br. Sleford3. Yo. N.V. Slingesby* (1316). Yo. S.B. Slengesby* (1301), Sleddemer9 (1301;.

15th c. — F.A. Sleford* (1401), Lafford* (1428), Snartford* (U31\Snorryngg* (1428), Naringg* (1428). P.B. Snetesham* (67—77).

At the time of the Nor man Conquest ,9 before /, m, ^ had already been dropped, or changed into another sound. Cf. MEYEB-LUBKE, Gramm. I § 529; NYROP, Gram in. § 462; BEHRENS, p. 182 f. Consequently, the combinations si, sm, sn in English place-names were unknown to the Normans, and had to be altered. Sometimes a svarabhaktic vowel is inserted: Senelestune, D.B., Selungesbi, D.B.; more frequently s is dropped: Ledemare, D.B., Lafford, Index, Metheduna, D.B., Netesham, D.B. Notintone, D.B. etc.

Though ,9 was mostly kept in the orthography - - and even a prothetic e not unfrequently added this does not

prove it to have been pronounced. The occurrence of several inverted spellings decidedly tell against such an assumption 1.

latter in Buck cross Hundred, which, however, are contiguous. Moreover, we find in the Yo. I.P.M. (1277) Petrus de Ledemer as one of the jurors for Buckcross H. In the Yo. I.P.M. Sledemer is incorrectly identified with a non-existent modern Ledmere.

1 Also in French place-names s is often kept in the spelling, though it had ceased to be pronounced long before: Esneutrevilla (1233), Esnitrevilla, Netrevilla (1254) = Ne"treville (Eure); Esmalville (1170) = Emalleville (Eure). See Dictionnaire Topographique de 1'Eure.

54 R. E. Zachrisson

We have noted the following cases of additon of an unetymolo- gical st es in D.B.: Esledes = Leeds (Ke.), Ledes, Ch.R. I, II, Ledes (1316, 46), F.A.; Snateshird, D.B. = Neatishead (Nf.) < O.E. Neteshirde, Netheshird, Kemble 740, 785; Slapeford, Exchequer D.B., Eslapaforda, Exon D.B = Lapford (Dv.), Lapeford, Ch.R. II, Lappeford, F.A. Cf. O.E. Hlappawurth (DUIGNAN, St. Pl.Ns., p. 90, under Lapley) = Lapworth (Wa.).

NOTE. It seems very probable that from the earliest times s was ignored also medially before such consonants where it was no longer pronounced in A.N. itself. In support of this conjecture the following instances of spellings without s may be quoted: Chileborneford (Hen. L), Index = Chesilborne (Do.), Chiselburneford (12 c.), Index, < O.E. cysleburna, Birch 596; Leinnes (Stephen), Index (reminds of such spellings as O.Fr. arcinnad, where sn > n, see MKNGEE, p. 104) = Lessness Abbey (Ke.), Lesnes Ch.R.1., T.N., perhaps < O.E. Ices, Mdf., p. 85; Gowberkirche (I3c.), Index = Gosberton (Li.), Gozeberdecherca, Gosebertechirche, D.B., Goseberkirke, H.R. <; O.E. Gosbeorht, Searle; Ouberneby (2 X), T.N. = Osbournby (Li.), Osberneby, H.R., F.A. < O.E. Osbern, Searle; Obernestun (Stephen), Index = Osbaston (Sh.), Osbernestun (12c.), Index *. — Addition of an unetymological s has been noted by us in Haslingueham. Fr.Ch. = Ellingham (Ha.), Meslingues (1201), Index = Mailing (Ke.), Cameslingeham, D.B. = Camrneringham (Li.).

A parallel to the changes the sound-combinations si, sm, sn underwent in English place-names is offered by the -treatment of several Germanic loan-words, adopted into the French language at a period when s before a consonant was no longer pronounced. Often a svarabhakti is inserted, sometimes s is entirely lost. For instances and references see DIEZ, Gramm. I4, p. 315; NTROP, Gramm. § 494; MACKEL, p. 34, p. 173 (svarabhakti)2 — NYROP, Gramm. § 461 (loss of s). Here may belong also Mod. Fr brindestoc < Dutch springstoc.

Only in the case of the important town of Nottingham have the French forms gained the victory. Isolated forms with s are to be found as late as the 14th cent. We have already pointed

Two instances where 8 has seemingly been dropped before /, may be accounted for as mere errors of the scribes, s having been taken for /: Wafforda, Wesforda, D.B. = Washford (Dv.), Wasford (1303), F.A. The same place in Sf. is called Cesforda and Ceffordd in the Inq. El.

1 Cf. also Fr. sequelin by the side of skellin (Godefroy) < Germ. skilling.

Anglo-Norman influence on English place-names 55

out that in the case of Sleaford the English forms have had a narrow escape of being superseded by the French.

To connect the change Snotingaham > Notingaham with the dialectal addition of s which is to be seen in such words as snotch < notch, snowl < nowl etc. (see RITTER, Archiv 116, p. 47), cannot be justified. The very time of the appearance of the s- less forms tells in favour of French influence. Indeed, Noting- ham etc. (always with loss of s) is the only form recorded by WESTPHAL, p. 35, in numerous examples from O.Fr. texts.

D. Anglo-Norman esc, esp, est is substituted for English sc, sp, st.

1 . Stow (Cb.) < O.E. stow, Mdf., p. 127. 2. Sturminster (Do.); cf. O.E. Sturemupa, A.S. Chr. 3. Okeford Shilling (Do.) < O.E. stilling, Sweet, A.S. Die. *. 4. Speechwick (Dv.). 5. Sporle (Nf.). 6. Stretton (St.) < O.E. street, Mdf., p. 127. 7. Stubbs (Yo.). 8. Scrayingham (Yo.). 9. Scar- borough (Yo.).

D.B. Espicewita*, Estretone*, Eistop\ Istop1, Escraingham*, Scring-

ham*, Escringham*.

12th c. — Fr.Ch. Esparlaium5. Ramsey Ch. Estow1. 1 3th c. — Ch.R. Escardeburg9, Eschar deburg*. Index. Estrumenistre* (1204?). 1 4th c. — P.A. Sporle* (1316). Index. Esturminstre* (1327), Acford Eskyl-

ling* (1303), Okford Eschylling* (1319), Okford Skylling* (1384).

At the time of the Norman invasion s was kept before c, p, t. Cf. MEYER-LUBKE, Gramm. I § 468; NYEOP, Grarnm. 462; BEHRENS, p. 182. But the difficulty the Normans had in pronouncing an s impurum accounts for the appearance of a prothetic e before sc, sp, st. The distribution of this e in early A.N. texts renders it probable that it was added only when in connected speech a consonant immediately preceded the s- com- bination. See NTROP, Gramm. § 493. This state of things must have given rise to a general uncertainty as to the orthographical use of this e. It was, no doubt, often put where it was not pronounced, and vice-versa. In records of later date than D.B. it is rarely found in English place-names.

1 sk instead of sh, sch in the forms quoted from the Index seems to be due to confusion with O.Fr. escalin, eskallin (Godefroy), Schillingus (du Gauge) < Germanic skilllng.

56 R. E. Zachrisson

On the other hand, a svarabhaktic vowel is never inserted between s and c, p, t. This is one more proof - - if proofs are needed - - that s before c, p, t was still pronounced towards the end of the llth cent. In like manner the appearance of such a vowel between s and I, n may be pointed to as a test of loss having taken place in the same position.

•

NOTE. BEHRENS' statement, p. 182, that »bereits altere normannische und anglonorraannische Hss. die e- Prosthese ziemlich allgemein durchgeftihrt zeigen» is not correct. An examination of the purely French word-material in D.B. shows that e before s impurum is omitted much more often than it is put. According to BUSCH, Laut- und Formenlehre der A.N. Sprache des XIV Jahrhunderts (Diss. Greifswald 1887), p. 9, it is often neglected in many other early A.N. texts, and has not been generally accepted even in the 14th cent. Cf. also P. SCHLOSSER, Quatre Livres des Hois, p. 71. From the modern and old forms of several place-names in Normandy we may conclude that also here the forms with e were not used to the absolute exclusion of the others: Sturvilla, (1179) = ^treville (Eure), Torsel (1310) = Estorsel (Eure), Sprtvilla = l£pr4ville (Eure), Netreville (Eure) < Esneutrevilla etc. See Die. Topogr. de 1'Eure. It is a well-known fact that the prothetic e hardly ever appears in the French loan:words in English. This is explained by Behrens as due to English apharesis of the Norman e. Without denying that such may have been the case, we will take this opportunity of suggest- ing another explanation. If in connected speech there was a continual interchange of forms with e and without, it would seem highly probable that, in the very act of borrowing, the forms without e were preferred by the English, as being more natural and more in accordance with their own vocabulary.

E. Anglo-Norman ser, sir, sar is substituted for English (/r).

1 . Shropham (Nf.) < O.E. scrap, Mdf., p. 114. 2. Shrewsbury (Sh.) < O.E. Scrobbesburh, A.S. Chr. 3. Shropshire or Salop < O.E. Scrobbes- byrigscir, A.S.^_Chr. 4- Shrawardine (Sh.), pron. [: shrecd'n], ELLIS V, p. 183.

D.B. Screp-, Scerep-, Serepham1, Sciropesberie3 , Sciropescire*,

Saleurdine*. 1 2th c. — Florence of Wo. (Mon. Brit.). Scrobesburia*. Inq. El. Scerep-

ham*. Henry of Huntingdon (Mon. Brit.). Scropsire*, Salop-

scire*, Salopsyre*. Pi.B. Salopescira* (1155). 1 3th c. — O.B. Salopia^ Salopsire9. Oh.B. Screpham1 I, Salop* (1225),

Screwardin4. H.B. Salopisbur'*, Salop'9, Schrewardin*. Index.

Scropham* (1286). MS. Jesus Coll. Oxford. Slobschire*. P.B.

Salop* (72-81), Salopesbiry* (16—25), Salopsire9 (32—45). T.N.

Salop'*, Salop'*, Sre.wardin*.

Anglo-Norman influence on English place-names 57

1 4th c. — F.A. Salopia*. P.B. Shreusbury*, Shrouesbury*, Salop9 (91—96),

Shrouesbury* (1399—1401). 15th c. — Capgrave. Schreuisburyz, p. 282, Schrovysbury* , p. 157. I.P.M.

Salop9. P.B. Salop*.

The combination (/r) presented especially great difficulties. There existed no initial (/) in the A.N. dialect of the llth cent. Hence (s) seems to have been substituted for (/). To facilitate the pron. a svarabhaktic vowel was', moreover, inserted betveen s and r: Serepham, D.B., Sciropesberie, D.B.

The form Salop, which is still often used by the side of Shropshire, exhibits a third A.N. peculiarity, dissimilation of r to I

The local pron. of Shrewsbury is [: sroozbri] or t[: soozbri], with dissimilatory loss of r. See ELLIS V, p. 183. This (s) is hardly due to French influence. It merely reflects a change of (/r) > (sr) not uncommon in the Midland dialects. See E.D. Gr. § 338.

The peculiar appearance of ew by the side of ow in forms of the 14th c. etc. has been explained by Professor EKWALL, Jones's Phonography § 264 foot-note, as due to the analogy of such pairs as shrgu: shrpu etc. According to Rippmann, Shrews- bury is still pronounced (/rouzbari) and (/ruwzberi).

§ 4. Anglo-Norman ((h) is substituted for English initial (j).

1. The Anglo-Norman forms have prevailed 1.

1. Jarrow (Du.), pron. (o^aerou), Tanger <C O.E. gyruum, Bede, Sweet O.E.T. p. 564 2. 2. Jesmond (Nb.) on the Jess, from which it probably has

1 The symbol 3 hardly occurs in the records we have consulted. Whether 3 is unknown to the MSS. themselves, or the respective editors have not made any distinction between g and 3, we are unable to decide. j is much more usual than »', and, here again, it is doubtful whether this represents the original state of affairs. However, the mere circumstance that i does occur seems to prove that the editors have distinguished between the two symbols.

2 Jarrow, the birth-place of the Venerable Bede, once possessed a famous monastery & (founded in 680), of which the ruins still remain; BAKTH.

58 R. E. Zachrisson

derived its name1. 3. Jagdon (Sh.). 4. Jevington (Ss.); cf. O.E. com- pounds with Gef- like Gejwine, Gefwulf, Searle. 5. Jervaulx (Yo.), pron. (d3aavou), Rippmann < O.E. eofor and O.Fr. val, vaus 2.

D.B. Lagedone*, Lovingetone*.

12th c. — Florence of Wo. Girvum1. Symeon of Du. Girvum1.

1 3th o. — C.B. Gese-, Jessemuth* (88-96), Yeverton * (79—88), Gevyngton4, Jevyngton4 (88—96). Oh.B. Gevyngton* (1295), Jorevallis5, Jero- vaU* I, Jorevall* II. H.B. Gesemuthe*, Jagedone*, Gewington*. I.P.M. Jagedjn* (Hen. III.). Index. Jezemue*. Nb. Bolls. Jesemm* (1256), Gesemue* (1279). P.B. Gervaus5, de Gerrevalli- bus6, Gyrivall* (32—47). T.N. Josemouth*. Yo. I.P.M. Jerveaux5.

14th c. — C.B. *Jesmuthz (41—43), Gervaux* (46—49). F.A. Zesemuth* (1346). Index. Jesemuth* (1334). Inventory of Jarrow (Surtees Soc, 29). Jarowe1, Jarue\ Jarvwe* (1303). P.B. Jarrow^ (27— 30), Gesmonth* (1399—01). Yo. S.B. Jerovallis, -walk, -valle* (1301).

15th o. — I.P.M. Jevyngton*. Index. Jesmowethe* (3421). Inventory of Jarrow. Zarowe1 (1427), P.B. Jesmounde* (01—05), Yakedon* (01—05), Jervaux* (01—05), Jorvall* (67—77). Ss. S.B. Jevyn. ton* (1411).

1 6th c. — Inventory of Jarrow. Jarow1 (1508), Yarow1 (1526). V.B. Jorevalle*.

2. The Anglo-Norman forms have not prevailed.

1. Yaxley (Hu.), probably < O.E. geac, Mdf., p. 56; cf. also O.E, Gcecceledh, Geakeslea, laceslea, Kemble 579,591 etc. 2. Yawthorp (Li.) < O.E. lol, Ma, Searle. 3. Yarborough (Li.) < O.E. eorjt; cf. eordburge, Eorthebytiy}g, Birch 758, 598. 4. Yaxham (Nf.) < O.E. geac. 5. Yarmouth (Nf.) has derived its name from the River Yare, the Garriennus of the Ptolemy (FOBBIGER, III, p. 282), afterwards called the Gerne, Yerne; TAYLOB, Names, p. 298. 6. Yockleton (Sh.). 7. Yoxford (Sf.) < O.E. geoc, Mdf., p. 60. 8. Yarmouth (Sf.), suburb of Yarmouth (Nf.), with which it is connected by a bridge over the Yare. 9. Yaxley (Sf.) < O.E. laceslea, Thorpe. 1O. Yokefleet (Yo.) < O.E. gcoc. 11. Youlton (Yo.) < lol, Tola. 12« Yarm (Yo.). 1 3. Yapham (Yo.). 1 4. Yafforth (Yo.), perhaps < O.E. ea, Mdf., p. 45.

The immediate neighbourhood of the important town of New- castle, which derived its name from a fortress built in 1080 by a son of William the Conqueror, may have contributed to the victory of the Nor- man forms.

2 Jervaulx was the seat af a celebrated Cistercian abbey founded in 1156; EARTH.

Anglo-Norman influence on English place-names 59

D.B. Lolestorp*, Loletorp*, Gereburg*, liachesham4, Jachesham*, Jakesham*, Gernemwab, Loclehuile* ', Gokesford1, Jernesmua*, Gerne- mutha*, Jacheslea9, Lugufled™, Lucufled10, Jugufled10, Loletune", Larun1*, Jaforde1*.

1 2th c. — Index. Yerbitrg* (c. 1160), Jerdeburch9, Joscefort1 (1156). Li. S. Jerburc*. Sf. P.F. Jaches-, Jakele9. Yo. P.F. Jaforde™ (1191).

1 3th c. — G.R. Yolthorp* (72—79), Gernem"0 (27—31). Ch.B. Jackeslada1, Jakle1 I, Gernemuth, -mue5 II, Yokethil* I, Jokesford1 II, Yuc.ke- flet10 I, Japum13 II. H.B. Jake-, Jackele\ Yolthorp\ Yerdeburch-, burges, Jordeburge*, Jertheburge*, Jerne-, Gernemuta5 , Gernemue*, Yokelcul6, lakes-, loxford1, Gernemuta*, Jakele9, Yarom1*, Jarwm12, Yapam13. Index. Jakesle1 (1227), Jaxham4* (Edw. I.), Jokes-, Joke- ford1, Jakele9 (1296). La3amon. Gernemude5 A, lernem'* E, v. 30543. P.B. Jakele^ (25—32), Gernemue5, Jernemue* (16—25). B.G1. yarnemouthe5, ^ermouthe'0 , v. 3429. Sf. P.P. lokeford1, Yoxford1, Jeme-, Gernemuth*, lakeslee9, Jakele9. T.N. Jerne- muta5, Gernemue*, Jernemuth*, Jaru'1*, Yappu'ls, Japumls Yo. I.P.M. Yokflet*0, Jolton™, Yolton", Jarum1*, Yapum1*, J afford" Eiford^. Yo. K.Q. Jafford™. Yo. S.B. Yucflet10 (1297). '

1 4th o. — O.B. Yakes-, Jakesle1 (46—49), Yerdeburg* (07—13), Jokethull* (18—23). F.A. lakeleghe1 (1316), Yolthorpe*, Yerdburgh3, Yord- burgh3, Jordeborw3, Yakesham*, laxham*, Gervemerthe5 (1346). Index. Jernemuth'0 (1353), Joxford7 (1356), Jakesle9 (1343), Yaxle9 (1335), Jarum1* (1327). P.B. Jakele1 (27—30), Jakesle1 (43—45), Yakeslaye1 (45—48), Jakes-, Yakesle* (77—81), Joldethorpe* (45—48), Jerdeburgh* (27—30), Jordeburgh* (45—48), Jernemutha* (43—45), Jernemuth5 (81—85, 1399—01), Yokelton6 (77—81), Yocul-, Yokel- ton« (81—85), Yoxford1 (27-30), Joxford1 (43—45), Jakeslee9 (43—45). BBr. Germue*, Gerncmeue5, mue\ p. 252. Sf. P.F. Yoxforde\ Jernemuta*, Jakesle9, Yakesle9. Sf. S.B. Jernemutha* (1327). Yo. N.V. Yafforde". Yo. S.B. Yiolton11, Yolton11 (1301), Jarum1* (1301), Jafforde^ (1301).

15th o. — Capgrave. Yermoth*, p. 208. F.A. lakesle1 (1428). I.P.M. Yaxle9 (Hen. VII.). Index. Yermouth* (1464), Jernomewth'0 (1420), Yakesle9 (1412). P.B. Yakyslee1 (01—05), Yarbourgh8 (61—67), Yerburgh* (67—77), Jernemuth* (01—05), Jernemouth^ Yermothe* (61—67), Yoxford1, Jernemuth* (1467—77). Sf. P.F. Yarmouth* (1404), Jernemuth*, Yarmouth* (1434), Yakesle9.

In the earliest O.E. Germanic initial g had heen fronted before primitive front vowels, and was soon levelled under O.E. (j) < Germanic j. Of. O.E. spellings like gung by the side of iung, and ieces, iarwan alternating with peaces, yearwian. Of. SWEET, H.E.S. § 543 f. The value of both was, no doubt, that of a palatal spirant (j). Cf. SIEVERS, Ags. Grarnm. § 211; CHAD- WICK, Studies in O.E., p. 170 ff. Another source of O.E. initial

60 R. E. Zachrisson

(j) was ja, jo < ea, eo. See SIEVERS, Ags. Gramm. § 34 Amu., § 212 Anm. 2.

Initial (j) was unknown to the Normans, for though Latin g had been fronted before a, e, i, the result of the development was O.Fr. (d3), which sound was kept till towards the beginning of the 13th cent. Cf. NYROP, Gramm. §§ 423, 424. In like manner Latin j before «, e, i had been developed to (d3). Cf. NYROP, Gramm. § 469.

Consequently, when the Normans had to pronounce initial (j) in English place-names, this sound had to modified in some way or other. Here two possibilities might easily offer them- selves. Such a combination as (je) might be replaced by either (ie), or (d3e) l. In D.B. spellings like liachesham, on the one hand, and Jernesmua, Gernemwa, GroJcesford, on the other, /and J, G may very well be taken to denote the sounds (i) and (d3), respect- ively. We believe STOLZE, p. 46, to be. wrong in assigning the value of [#], [j] to the above-mentioned spellings. Where in purely French texts like - - to take one instance - - Langtoft's Chronicle such forms occur as G-ernemue, II, p. 195 = Yarmouth (Nf.), g was here certainly not pronounced differently from g in geter (= Mod.Fr. jeter). And from a linguistic point of view we have to look upon D.B. as an essentially Norman and French work 2.

In this connection it may be pointed out that the change of (j) > (d3) appears also in loan-words in French itself. Cf. NY- ROP, Gramm. § 469. In support of our theory we will, moreover, draw attention to a number of half popular French words in which for a long time the pronunciation (i) + a vowel inter- changed with that of (63) + a vowel: »Pour les mots qui conmenQaient par HI A ou HIE, on a hesite longtemps entre Ma-, hie et ja-, je: HYCINTHUS > hyacinthe, jacinthe (etc.) L'usage s'est determine' pour la forme savante dans les mots les plus

1 The substitution of (d3) for (j), and vice-versa, is a very natural one. Many Swedes, when speaking English, are apt to replace the unfamiliar Eng- lish combination (dj) by Swedish (j).

1 When, for instance, it is pointed out by VAN DEE GAAF, Engl. Studien 39, p. 282 f., that O.E. final d and ng in D.B. often appear as t,nc, this is only one more A.N. feature, and does not justify us in concluding that final d and ng had been unvoiced in the corresponding O.E. dialects.

Anglo-Norman influence on English place-names 61

rares, tandis que la forme populaire a ete adoptee dans les mots les plus usites»; NYROP ib. Here we are expressely told that the forms in (d}) were the more popular ones, and it is, indeed, not impossible that also i in D.B. denotes this sound l.

The modern pronunciation of Jarrow etc. is our best test of French influence, (d}) < (j) cannot well be due to internal English sound development. As far as we know, (j) was never- developed to (63) either in literary English or in the dialects. The occurrence of (d3) in districts so widely separated as Sussex and Yorkshire also tells against the assumption of dialectal influence.

As to the M.E. forms, we consider those in y to represent the native pronunciation (j), those in g (supposing the editors have properly distinguished between g and 3), j, and possibly also i to indicate the French pronunciation (63) 2. The distinc- tion between 3 = (j) and g = (g,d3) is well kept up in most M.E. MSS. after the year 1200. Of. WTLD, Guttural sounds, p. 150. Though in some early texts i was sometimes employed to denote (j), it never became popular in this use. Cf. SWEET H.E.S. § 608. j = (j) we have noted only in a few isolated cases: burje, La3amon B, v. 4264;t;'afe, Hoccl.; jondur, Egl. (Stratinann-Bradley.). In French loan-words (63) is denoted by g, j, i (more seldom y) from the earliest times. Cf. BEHBENB, p. 177; HGGVELMANN, p. 75 ff.; NEUMANN, pp. 78, 80, 81, 97.

The French forms in j, g are of comparatively rare occur- rence in late M.E. records. An examination of the first three volumes of the F.A. shows that, in addition to the cases given in the lists, g and j are to be found only in Gyvelden = Yelden (Bd.), Jerdel = Yardley (Es.), Gelham = Yealdham (Es.) and Grilling =

1 Thus much is certain, that the symbols I and J in D.B. were often used to denote the same sound. The two are very much alike, and one is frequently put for the other. Cf. the facsimile ed.

2 From the evidence of such spellings as Gilvertoft, Chelveristoft, Jelveristoft Cat. A.D. = Yelvertoft (Np.), Yelvertoft (always), F.A., we may also conclude that j really was pronounced (d^>, ch being a rare variant of g, j in French loan-words. Examples are adduced by BEHEENS, p. 177; HCEVELMANN, Zum Konsonantismus der altfranz. Lehnworter etc. (Diss. Kiel 1903), p. 77; DIBELIUS, p. 456. For A.N. instances see STIMMING, p. 237; MENGEE, p. 99.

62 R. E- Zachrisson

Yelling (Hu.). This latter place is called Gellyng alias Tellinge in the Index as late as 1545. The 15th cent, forms of list 2 exhibit.;', only for Yarmouth (Nf.), where owing to the consider- able importance of that place French forms were more likely to persist.

To the five places mentioned in list 1, where in our opinion j __ (^3) is due to A.N. sound-substitution, may possibly be added Jedburgh (Roxb.) called Jedworth and G-edworth in old records (see TAYLOR, Names, p. 156), and Gerse Hill (Gl.), the early forms of which (Yerdeshull, 1303, Zerdeshutte, 1346, Jerdushulle, 1402, F.A.), point to O.E. ^eard. We do not know if here g is pronouftced (d}).

The modern form of Jesmond exhibits one more, as it would seem, A.N. peculiarity. The original suffix- mouth has been replaced by -mond. Whether this -mond is equal to A.N. mont, mond < Lat. montem - - possibly suggested by such a spelling as Gesmonth, where n is a mistake for u - - or represents an intro- duction of the word mond = mouth of a river, we cannot decide J .

Of particular interest is Jervaux (Yo.), for here we can follow the various phases of A.N. influence, the etymology of the place being quite certain. From its situation on the Ure we may safely infer that Jervaux is equal to English Yordale, Barth. = the valley of the Ure 2. Ure for Yore is quite a mo- dern spelling, not used until ur had been levelled under pr. The following old forms of the river are on record: Jor (Hen. II.), Index, Eowere (Hen. III.), Index, Yore Aqua, H.R., Aquam de Yore (1327), SKAIFE, The Survey of York etc., Yore (1399-1401, 1467—77), P.R. The etymology is obviously O.E. eqfor,-e, and there is strong reason for believing that the name of the river is identical with the first element of Q.E. Eoforwic, A.S. Chr., = York. The city of York is situated on the Ouse, which is

1 Of. COCHERIS, Origine et form, des noms de lieu, p. 23: — >Le mot embouchure se dit mund en allemand, mond en flamand et mouth en Anglais: Deulemont (Nord) a Tembouchure de la Deule; Tenremonde (Belgique) a 1'embouchure de la Dendre.

2 In like manner the adjacent Rievaulx (= the valley of the Rye with R. Abbey has derived its name from the River Rye, on which it is situated.

Anglo-Norman influence on English place-names 63

formed by the confluence of the Ure and the Swale. In some way or other Eofore became Yore, which was turned into Jore by the Normans. Hence such forms as Jorevallis, Jorevall. After this, pretonic o was changed to e: G-errevallibus, and, with later syncope of the second e, G-ervaus. The appearance of e for o may be due to the influence of the following e. For A.N. in- stances of o for e owing to assimilation, or e for o owing to dissimilation, see STIMMING, p. 177, p. 191. Analogical transfor- mation after numerous French place-names in Ger- may also have been the cause of this change.

NOTE. In this place we will devote a few lines to a discussion of the etymology and development of York < O.E. Eoforwic A.S. Chr. By most authorities (CAMDEN II, p, 876; TAYLOE, Names, p. 299 f.; MADDEN, La3amon III, p. 315) this city is supposed to have derived its name from the river Ure. We do not know if any historical evidence can be adduced in support of this theory; if not, it seems to us more probable that the river has been named after the city. Taylor's argumentation is utterly wrong. He thinks that York is an analogous case to the French place Ivry (Eure) named after the Eure <. Ebura, on which it stands. But the earliest forms' of the Eure are Audura, Authura, :ind in Voltaire's Henriade Eure rhymes on nature (cf. Die. Topogr. de 1'Eure), so it is, or was at least for- merly, pronounced with y < u. The form Ebura recorded in the year 1236 is merely an instance of false etymological spelling.

By the Romans York was called Ebur-, Eboracum (cf. FORBIGEE III, p. 279) < Celtic Eburus, a personal name, -|- the Celtic suffix -acum = >the abode of Eburus* (cf. D'ARBOIS DE JUBAINVILLE, Revue Celtique VIII, p. 112, p. 134), which, owing to the fact that intervocalie b in late Latin was pronounced v, easily might have been turned into Eoforwic by the Saxons.

In the above-mentioned work by Taylor we are told that Eoforwic was changed into Jorvik by the Danes, and this seems, indeed, to have been the case. Professor Finnur Jonsson of Copenhagen has kindly informed me that the form Jorvik appears in Emits drapa (c 1040), v. 11 and, about a hundred years earlier, in another Icelandic poem called Arinbjarnardrapa by Egill Skallagrimsson. According to NOEEEN, Altisl. und Altnorweg. Gramm. 3 § 227, 2, j6r is a contraction of eofor. The city of York was one of the principal Danish settlements in England, and the seat of the Da- nish jarl- On the stage *Torik(e) < Jorvik contraction to York has taken place.

The first time we have found the present form is in the following interesting passage from La3amon A, vv. 2662—73: —

64 K. E. Zachrisson

jErest he makede ane burh ;

& Kaer Ebrauc heo clepede.

fa ofier on are dune;

Adud he heo hahte.

pa heo wes icleped Kaer Ebrauc;

setitien wes icleped Eborac.

Seofiden comen uncude men;

& Eoverwic heo hahten.

& Pa norferne men;

nis hit nawiht yeare.

purh ane unpewe;

yeorc (yore) heo ihcehten.

MADDEN, I.e., draws attention to the following lines from K.G1. (ed. HEAKNE, p. 27): — And suth me clefeth Enermcke this ilke toun y-wis, And York also, thorgh light speche, y-hote also hit is.

Robert of Brunne, p. 301, has ^ork, Capgrave, p. 264, ^ork, York; Havelok, v. 1178, yerk: clerk. The expression pa norperne men in Lasamon certainly refers to the people of Yorkshire, who are said to have adopted and given credit to the Scandinavian form.

According to Hope, the local pronunciation of the place is »Yark», »Yorrick>.

As far as we know, the theory of Norman (d$ having been substituted for initial (j) in English place-names has not been advanced until now. - - On French g = (63) for English 3 = (j) in several names from the Orrmulum cf. REICHMANN, Die Eigen- narnen in Orrmulum, pp. 32 f., 50 f., 82 f. x. - - In his treatise on English personal names in Gaimar RATHMANN, p. 46, in our opinion quite correctly, assigns the value of (d3) to g < O.E. 3 -f a palatal vowel.

Several times and in various places Prof. Skeat has dwelt upon the difficulty the Normans had in pronouncing English initial (j). Cf. in particular Transactions of the Phil. Soc, 1899 — 1902, p. 457, p. 452; Proverbs of Alfred (Oxford 1907), p. XIV. He seems to be of opinion that the Normans either substituted hard (g) for (j), or altogether dropped it. We do not think he

We have noted what seems to be a very curious instance of A.N. (d3) for English (j) in a Mod. Engl. surname, viz. Jung, Junge. Entries from the H.R. like Johannes le Junge, Robert le Jevene, Ralph le Younge (BABDSLEY, Die., p. 438, p. 831), render it highly probable that Younge has been altered to Jounge either by French mispronunciation, or owing to the influence of the corresponding French word Jevene.

Anglo-Norman influence on English place-names 65

lias succeeded in giving any conclusive proofs of such having been the case. Mod. Engl. Guildhall for M.E. Yeldehalle is, no doubt, best explained as due to Scand. gildi. Cf. BJORKMAN, Scand. Loan-words, pp. 154, 307. — As to Etiminstre, D.B. = Yetminster (Do.), which is adduced to illustrate loss of (j) caused by A.N. influence (cf. Student's Pastime, p. 349), it is by no means cer- tain that the first Compound contains O.E. ^eat\ it may as well be derived from O.E. Eata, Eating, Searle. Cf. Yatesbury (Wl.), JEtesberie, D.B., Yetesbir, Ch.R.I; Yatton (WL), Etone, D.B. Also hsre (j) may depend on the change of a falling into a rising diphthong. It is, moreover, to be kept in mind that such a spelling as eat for ^eat might occur in O.E. itself owing to the dialectal levelling of ea, eo under 300, 300. - - Real loss of (j) we have noted in Epesuuic (12. c.), Inq. El. < O.E. Gipeswic, A.S. Chr. = Ipswich (Sf.), and Ivelcestre, Giuelcestre (Hen. II.), Pi.R. = Ilchester (Sm.). In our opinion this loss may be due to the dialectal development 32 > i. Though the first instances of itch and icicle (< O.E. jiccan, yiecel) without 3 noted in N.E.D. are not older than the 14th cent., i < 31 in verbs occurs as early as in Herb. Apuleii, 12th c. (see BERBEEICH, p. 30), and in the Jesus (Oxford) MS. of the 13th cent, we find yvernesse, yssing < O.E. yifernes, ytsung (An O.E. Miscellany, p. 38), and ymstone < O.E. gimmstdn (ibid., p. 98). Cf. also SIEVERS, Ags. Gramm. § 212 Anm. 2.

NOTE 1. In this paragraph we have had occasion more than once to dwell upon the dialectal development ea, eo > id, io, which, to judge from the evidence of the early spellings of place-names, seems to have been quite usual in early M.E. For O.E. instances we have already referred to SIEVERS, Ags. Gramm. More examples and references are to be found in SCHROER, Die Winteney-Vers. der Regula S. Bened. (MS. c. 1225), p. XXVI. -Of. also TACHAUER, Laute und Flex, der Winten.-Vers. der Reg. S. Ben., p. 15. As up to the present this question has been paid very little attention to by English philologists, some more material calculated to illustrate the sound- change would perhaps not be unwelcome: Hartleford (GL), lorotlaford (a. 779), Index (probably a late copy). - Yardley or Ardeley (Ht.), Erdelei, D.B.? Erdeleye, H.R. — Yalding (Ke.), Uuest aldingis (1072), Index., Ealdyngge (1411), Index. — Heapham (Li.), Jope-, Lopeham, D.B., Jopheim, Li.S., Hep- ham, T.N., F.A. — Earlham (Nf.), Erlham, Jierlham, D.B., Erlham, F.A., Index. — Easington (Nb.), Yesington (1346), F.A. — Edingworth (Sm.), Yaden- worlh (1316), Edeneworth (1428), F.A. - - For Sm. the Crawford Ch., ed. NAPIER-STEVENSON, which exist in parallel versions from different times,

5

66 R. E. Zachrisson

afford some excellent instances. To earnes hrycg (a. 739) corresponds Yearnis Rygge (loth cent, copy), see p. 55; to eorpgeberst (a. 739), yeorfiberst (15th cent.), ibid. The River Yeo is called Eowan in 739, ewyn in the 15th cent.

- Urchfont (Wl.), lerchesfonte, D.B., Erchesfont (14 c.), Index. — Yearsley (Yo.), Eureslage, D.B., Eversley, Yeverslay (1301), Yo. S.R. — Yeddingham (Yo.), Edingham (1301), Yedingham (1297), Yo. S.R. - - Everthorpe (Yo.),. Jurethorp, Ch.R. I. - - Arncliff (Yo.), Erneclive, Gerneclif, D.B. — In the dialects the phenomenon is widely diffused; cf. E.D. &r. §^248. The examples recorded under that paragraph, naturally, belong to different periods of the development of the language, and require different explanations. In some instan- ces the diphthong has, no doubt, originated quite recently, in others it may date very far back. — We do not think BOEENER, Die Sprache Roberd Mns. of Brunne (Studien zur engl. Phil, herausg. von Morsbach, XII), p. 301, is right in assuming Scand. influence not only for such forms as yearth, yarl, but also for the whole phenomenon eo, ea > id, to; it has much too wide a. range to admit such a general explanation. We feel convinced that this whole question, as well as that of 31 > i, would easily be solved by means of a thorough special investigation based on the evidence of the early spel- lings of English place-names.

NOTE 2. The numerous forms with L for J, J in D.B. (Loletorp,. Lucufled etc., see list 2) are merely due to orthographic confusion between initial J, J and L, which are extremely alike both in the MSS. of D.B. and that of the Li.S. The only difference is a small and almost invisible stroke, by means of which J or I is made into an L. Loversall (Yo.) appears in D.B. both as Geureshale and Loureshale. The copyist has here used the variant G to render what he imagined to be a J" in the original. That the instances of L here in question are merely due to careless orthography is proved be- yond all doubt by the fact that L for J occurs also in purely French words. Thus Ivry, a place in Normandy, is called in the Np. part of D.B. Juri, Luri.

• In other records the same mistake occurs, though more rarely. Thus Yaxham (Nf.) appears in the Inq. El. as Lachesham, lakesham, Jakesham. Ewhurst (Ss.) is called Luvehest in the Fr. Ch., and for one and the same person we find the spellings Lagon in La^amon, v. 2694, corresponding to- Jagon in Wace, v. 1583.

II. Changes in English place-names due to Anglo-Norman sound-development.

§ 5. s is lost before c, p< t.

A. In initial position. 1. The Anglo-Norman forms have prevailed.

1 . Tutbury (St.), pron. [ti'dburt], ELLIS V, p. 482, perhaps from an unrecorded O.E. *Stutta *, pet-name form of the numerous 'compounds with Stut-. 2. Crane End or Scrane (Li.).

D.B. Toteberie1. 12th c. — Fr.Ch. Stutesberia1'', Totesbery1, Tuttesberie1 . Index. Stutesbiria1

(1180), Screinga* (1158). 1 3th c. — Ch.B. Tutiebury1, Tuttebyri1 II, Screnge* II. Index. Tuttebury^

(1290). B.B. Scrahinge*. T.N. Estrahinge*.

2. The Anglo-Norman forms have not prevailed.

1. Spettisbury (Do.). 2. Stragglethorpe (Li.)? < Stergar, Searle. 3. Sturston (Nf.). 4. Stafford (St.) < O.E. Stce/ord, A.S. Chr. 5. Stitchbrook (St.). Numerous early forms, all of them with s kept, are adduced by DUIGNAN, St.] Pl.Ns., p. 143. 6. Stanley (Yo.) < O.E. stan. 7. Stubbs Walden (Yo.) < O.E. stybb, s^/Mdf., p. 123. 8. Stapleton (Yo.) < O.E. stapol, Mdf., p. 123. 9. Scutterskelf (Yo.). 1 O. Scampston (Yo.). 1 1. Scraying- ham (Yo.).

D.B. Sterestuna3, Tichebroc5, Stanlei*, Stubusun1, Staplendun*, Codreschef9, Codeschelf9, Godreschelf*, Scameston10 , Escraingham^.

\ 2th c. — Fr.Ch. Postebere1, Posteberies1.

13th c. — C.B. Stichebrok* (72-79). Gl.Becs. (Geoffrey de) Stafford* (1200), TefforV (1210), Steffort*. H.B. Stragerthorp*, Traggart*, Terston*.

Of. Stutsbury (Np.), Stutesbyr (early 13 c.), Stotesbur (c. 1230), Index.

68 R. E. Zachrisson

Yo. I.P.M. Stobis\ Stubbes1, Scudirschelf*, Scotherskalfe9. Yo. K.Q. Stapelton*, Tapelton* (pers. name), Scuterskelfe9, Scothri- kelf9, Scameston10.

1 4th c. — C.B. Stragertiwrp* (46—49). F.A. Stragerthorp* (1303), Thra- gerthorp* (1316), Thragelthorpf (1346). P.B. Thragelthorpez, Trey- garthorpe* (43—45), Skamston10 (81—85). Yo. N.V. Taneley*, Stubbes\ Stapilton*, Tapelton* (pers. name), Scotherskelf9, Tames- ton10, Traingham11 (in the latter two forms T is an error for C). Yo. S.R. Tubbeswaldyng'1 :, Scotherskelfe*, Scameston10.

15th c. — F.A. Shragulthorpe* (1401), Spectebury1 (1428).

In the above instances we have no doubt to do with real loss of s, for in the very same records, or in records of about the same date, we have noted several cases of addition of an unetymological s. As it is not always possible to establish the O.E. ground-form of the place-names here concerned, some of the following instances ought perhaps. to have found a place among the forms illustrating loss of s, but we trust that the distinction we have made will in most cases prove trustworthy.

An unetymological s is added before c, p, t, th 1.

1. Tabley (Ch.) < O.E. Tcebba, Searte. 2. Knolle (Co.) < O.E. cnoW, Mdf., p. 26. 3. Totham (Es.) < O.E. Totta, Searle. 4. Testwood (Ha.), cf. on Tcerstan stream (1045), Index. 5. Cliff (Ha.) < O.E. clif, Mdf., p. 27. 6. Toxteth (La.) < O.E. Toe, Toce, Searle. 7. Thimbleby (Li.). 8. West Tofts (Nf.) < O.E. toft, Mdf., p. 135. 9. Bircham Tofts (Nf.) < O.E. toft. 1O. Itteringham (Nf.). 11. Otringhithe (Nf.). 12. Chedgrave (Nf.)<O.E. Ceatta, Catta, Searle. 13. Tharston in Depewada Hundred (Nf.). 14. Epper- stone (Nt.). 15. Cliff (Wl.) < O.E. clif. 16. Crofton (Yo.) < O.E. croft, Mdf., p. 31. 17. Cusworth (Yo.)< O.E. Cusa, Cussa, Searle. 18. Criggles- tone (Yo.). 19. Tocketts (Yo.).

D.B. Stabslei1, Stabled, Sclive*, Stochcstede*, Stinblebi7, StimbleW, Toftes*, Stoftam*, Stofsta9, Vtrincham", Strincham10, Otringeheid1 \ Strinkeehia11, Scatagraua^ , Sterestuna™, Sclive™, Scrotone1*, Scusce- uurde" (1 X), Cuzeuuorde11 (2 X), Crigestone1*.

12th c. — Index. Catkegrave1*, Yo. P.F. Crigleston™.

13th c. — Ch.B. Tyrndc-, Thimelesby1, Epereston1* II, Crofton™ II, Cusse- worthaml\ H.B. Themel-, Tymel-, Thymelby\ ItringJiam™, Thers- ton18, Sterstnn1*. Index. Totham3, Totteham0, Otrigeide11 (1203). La. P.P. Toxtath* (1298). B.B. Timbleby'1 (1210). T.N. Tocke- stach* (ch for th\ Itringham10, Eperiston". Yo. K.Q. Thocotes™

1 From a Norman point of view initial p, th was equal to t. Cf. § 2.

Anglo-Norman influence on English place-names 69

(pers. name), Tocottes™, Toucotes™, Stoucotes1*. Yo. S.R. 1297.

Crofton16, Cusceivrth11, Krigelston18 . 1 4th c. -- F.A. Scnoll* (1303), Knolle* (1346), Tersteivode* (1316, 46), (Ri-

cardus) Sterstewode* (1316), Toftes*, Itringham10, Oteringhithe*1,

Chategrave™, Therston1*, Sterston1* (1346), (Ricardus de} Therston™,

Sterston™ (1302, 16, 46), Eperston" (1302, 16), Sperston1* (1346).

P.B. Westoftys* (1399—01), Crofton™ (1399—01). Yo. N.V.

Croffeton™. \ 5th c. — F.A. Knolle* (1428), Totham3 (1428), (Thomas') Stotham* (1428).

P.B. Crigilston™ (67—77). 1 6th c. — Index. Scregylston™ (1555).

That s was pronounced before c, p, t, when the Normans first settled in England is evident from the phonology of the French loan-words in English, in which almost without excep- tion s is kept in this position. Where s is lost, the loan is of a more recent date. In O.Fr. loss of s in this position dates far back. Already towards the end of the 12th cent, we find instances of s having been dropped or unetyrnologically added before c, p, t, and in the course of the 13th cent, the loss seems to have been completed. Cf on this, as far as details go, very difficult question, MENGER, p. 105 ft; STIMMING, p. 225 f.; NY- EOP, Gramm. § 462; NEUMANN, Zur Laut- und Flexionslehre des Altfranz. (Heilbronn 1878), p. 108 ff; W. KORITZ, Uber das s vor Kons. im Franz) (Diss. Strassburg 1885).

This A.N. sound-law did not, as a rule, extend its sphere of action to the French loan-words in English1. The English place-names on the other hand have, naturally, participated in this general sound-change, and the instances of loss or addition of an inorganic s noted above are, in all probability, due to the circumstance that the Normans of those times could no longer properly pronounce English s before c, p, t.

1 Only in a few cases sporadic loss of s, owing to later French in- fluence, has been noted: hopctall, Machyn 25526 (a. 1550—63), see DIEHL, Anglia, XXIX, p. 188, and ametite: sersolitte, Ipomedon, see KOLBING. Inorganic s appears in irritaste, Letters of Queen Elizabeth (DiEHL, ib.). One more example adduced by Diehl, tenstes (= tentes\ Machyn 1581 anm., is to be compared with mestes (= metes), Alfred's Proverbs, v. 615 (ed. SKEAT, 1907), wcestres (= wcetres) Inaction, v. 11770, and here s is best explained as due to anticipation of the following s. Thus LUHMANN, La3amon, p. 51.

70 R. E. Zachrisson

We admit that the s of a preceding word may sometimes have caused addition or subtraction of s. Thus Stochestede, D.B. (for Tochestede) < Wes-, Estochestede = West and East Toxteth (La.) ; Stoftam D.B. (for Toftam) < Westoftam = West Tofts (Nf .) etc. When the name of the owner immediately preceded the place-name the s of the genitive might easily be prefixed to the latter : Sclive (for Clive = Cliff) < *Kingesdive etc. In like man- ner s may have been transferred from the name of the person to that of the place in such entries as Thomas Stofham (1428), F.A. = Totham (Es.), Eicardus Sterstewode (1316), F.A. = Test- wood (Ha.) x.

But by far the greater part of our instances do not admit of such explanations. Sperston (1316), F.A. = Epperston (Nt.)> Strincham, D.B. = Itteringham (Nf.), StrinJceehia D.B. = Otring- hithe (Nf.) can only be accounted for as A.N. spellings indicating a pronunciation Eper-, Etr- 2. And the occurrence of Posteberies, Fr.Ch. = Spettisbury (Do.) in some purely French charters goes far to prove that also loss of s is a sign of French influence. Cf. also spellings of French place-names like Sparnascum (1223), Pernacum (1257), Pernascum (1252) = Epernay (Marne); Stricheium (1189), Tricky (1300) = fitrechy (Marne); Torsel (1310) = Estorsel (Eure); Sprevilla = Epreville (Eure). Cf. the series Dictionnaires Topographiques.

With some hesitation we have pointed to Crane (Li.) and Tutbury (St.) as two instances of the French forms having pre- vailed. In the case of Crane the circumstance that all early references we have found exhibit s renders French influence rather doubtful. Tutbury (St.) < O.E. * Stuttesbeorh, -byrig seems to be one certain example. Numerous 12th cent, forms like Slutesberia etc. (cf. list A 1; a few more early forms with s are given by Duignan) undoubtedly point to original st 3.

1 In the English Sir Beves (E.E.T.S., E.S. 46, 48, 65) Beves' horse is ge- nerally called Trenchefise, Trinchefia etc. Once only occurs Strenchefis, and the addition of 8 is here, certainly, due to the cirumstance that the preceding word ends in «. The whole passage runs: Beves Strenchefis bestrit, p. 90.

1 Cf. D.B. spellings like Sbernestune, Esberneston = Osbaston (Sh., Li.); Stope = Easthope (Sh.).

8 In the case of Tutbury historical evidence strongly supports our theory of French sound-change. Tutbury was in Norman times a very im-

Anglo-Norman influence on English place-names 71

In his work on Staffordshire place-names Duignan, refer- ring to SKEAT'S Principles of English Etyrn. II, p. 234, advances the theory that s in the 12th cent, forms of Tutbury, as well as present s in Stitchbrook, are excrescent, and have originated in the same way as s in splash < plash etc. In our opinion such an explanation is highly improbable. - - In Bardsley's Dictionary we find several pairs of names with s and without, which are stated to be etymologically identical: /Scripps < Crips (p. 673), /S'pur- <3ance < Purdance (p. 710), /Sturgis < Turgis (p. 725), Pichfatt< Spichfat (p. 707), Pilsbury < /S'pilsbury (Ox.) (p. 606), Pickernell < tfpicknell (p. 603), tfnazel < Kneesall (Li.), (p. 700). In an article on movable s (»das bewegliche s»), Archiv 116, p. 41 ff., EITTEE quotes all these instances, and also adds /Swithold < Withold. Of Bardsley's examples only /Snazel < Kneesall is by Hitter marked as dubious. In fact, Bardsley has proved the supposed identity only in the case of Cripps and Scrips, $purdance and Purdance, which really occur in entries relating to the same person. We cannot but feel somewhat sceptical towards a theory according to which s has been added to personal names and place-names, in the same way as it is added dialectally to ordinary words 1. For all we know, /Shazel has nothing whatever to do with Knee- sall. Nor has Pilsbury with /S'pilsbury. There is one place in Oxfordshire called /S'pilsbury, another in Derbyshire called Pils- bury, from which the two personal names, no doubt, have originated. The old forms we have ascertained do not show any interchange of p and sp. - Why not derive Sturgis from an O.E. *Siurgisl? It is true this form is not recorded, but Searle cites Stur and Sturstan as O.E. names. - - Why could not Swith- old be due to an O.E. name compounded with Switli-, e.g. Swith- weald, Searle, recorded as early as in the Liber Vitse? Loss of • s in tfpichfat and /Spicknell (? < spile, »lardum», Spigurnellus,

portant place, often mentioned in the old records. Cf. the following pas- sage from CAMDEN I, p. 643: • »Tutbury-castle (also called Stutesbury) formerly very large, and commanding the lower Country by its high situa- tion on an Alabaster-hill. It was built by Henry cle Ferrars, a noble Nor- man, to whom William the First gave large possessions in this County. »

1 As to the question of movable s cf. PAUL und BBAUNE, Beitrage 29, p. 479 ff . ; E.D.Gr. § 323, KBUISINGA § 363, and the above-mentioned article -by Ritter.

72 R. E. Zachrisson

Du Cange) may depend on A.N. sound development. In our opinion addition of so called movable s to personal names and place-names can only take place if the name itself or one of it& elements is identical with a word which in the spoken language exhibits both forms. Thus tfcripps may really be one instance of movable s having been added to Crips. Among the instances we have adduced of addition of s to place-names, movable s- would be admissible, at the utmost, for Scrigglestone < Criggles- tone, supposing there existed in the dialect two words criggle and scriggle with about the same sense.

NOTE. In this place we want to draw special attention to a very curious and hitherto unnoticed circumstance, viz. that loss ef s before c, p, t appears as early as in D.B. And this loss is not confined to the English words, it also appears in one French name: Ricard Estormid, Estormit, Sturmid; Radulf Sturm, Turmit (HILDEBEAND, p. 335) = Mod. Fr. Le^ourmi < Scand. storm, see JORET, p. 93. This remarkable form is not even men- tioned by Hildebrand in his account of the language of D.B.

We are quite aware that some of the D.B. instances we have adduced in the lists heading this paragraph are open to several objections. It may be reasoned thus, c for sc and sc for c are mere spellings. Cf. the analysis of the use of s, c, sc in D.B., § 1, Appendix B. Other instances may be due to occasional addition or subtraction caused by the s of a pre- ceding word. Lastly, loss of initial s is perhaps better explained as due to sound-substitution \ Nevertheless, two instances remain for which none of these explanations would hold good: Strincham and Strinkeehia = Ittering- ham (Nf.) and Otringhithe (Nf.). Moreover, loss and addition of s is not confined to initial position. The following examples may serve to illustrate the appearance of the same phenomena in medial position as well: Witenesto, D.B. = Wistow (Le.), Wisletiestowe, T.N. < O.E. Wigstan, Wistan, Searle; lest, D.B. = Lathe, division of land in Kent etc.; Otrinctun, Ostrinctun, D.B. = Otterington (Yo.), Oterington, Yo. K.Q. and N.V., probably < O.E. oter, Mdf., p. 100; Eschintone, Echentone, D.B. = Eckington (Ss.) < O.E. Ecca, Searle; -hert for -herst is very usual in D.B. Cf. STOLZE. p. 42. One more thing is well worth mentioning here. O.E. (yt) in cniht etc. is very often rendered by st in D.B. Examples might be quoted from nearly every county. This also points to a weak pronunciation of s 'before c, p, t 2.

There are a few rare instances of very early loss of initial 8 even in French words derived from Latin. Cf. DIEZ, Gramm. I4, p. 243, p. 454. 1 Though sporadic loss of s before c, p, t is evidenced already in late Latin, this can hardly be placed in any connection with its later disap- pearance in O.Fr. Cf. on this point SEELMANN, Die Aussprache des Lat., p. 316; SCHUCHAEDT, Der Vokalismus des Vulgarlateins II, p. 354 f.

Anglo-Norman influence on English place-names 73

From what has been said above, we seem justified in concluding that at the time when the existing copies of D.B. were written the sound-change owing to which s was lost before c, p, t had already commenced. As early as 1086 such a loss is not admissible. Consequently, we have here another proof that the D.B. MSS. are not as old as is generally supposed.

B. In medial position.

1. The Anglo-Norman forms have prevailed.

\ . Exeter (Dv.). 2. Gloucester (Gl.). 3. Cirencester (Gl.). , 4. Fro- cester (Gl.). 5. Bicester (Ox.). 6. Leicester (Le.). 7. Wroxeter (Sh.). 8. Kocester (St.). 9. Mancester (Wa.). 1O. Worcester (Wo.). Here belong also Alcester parish (Do.), Craster (Nb.), Towcester (Np.), Alcester (Wa.), of which we have found no spellings without s.

1 3th c. — Ch.B. Exeter1 (1292). Lasamon. Axcetre1 A, v. 30865, Glocet'e2 A, v. 16708, Leycetre6 B, v. 2915. R.G1. Excetre\ Gloucetre\ Claucetre*, Leicetre*, Roucetre*, Wircetre10 (cf. the Index to Wright's edition). Tax. Eccl. Mancetre9,

14th c. — O.B. Exeter1 (07— 13). Index. Glousetre* (1374), Rameter* (1392). William of Palerne. glouseter*, v. 166.

15th c. — Capgrave. Excetre1, p. 122, Excetir, p. 196, Glouceter*,' p. 136, Cicetir*, p. 275, Leycetir*, p. 118, p. 157, Wyscetir10, p. 281, Wy- cetir10, p. 292. I.P.M. Exeter* (Hen. VII.), Bysseter6 (1495). Nt.Becs. Glowecetter* (1484). Paston Letters. Exsater1 (see NEUMANN, p. 108) Worceter10 (ibid., p. 52), Worseter10 (ibid., p. 87).

16th c. — V.B. Glouccfer*, Glociter*, Glocitershire2, Cirincetre*, Frowceter*, Froucettour4, Biscyf'0, Leiceter®, Wroxcetor1, Wroxcetour1, Wroxtor1, Rocettre8, Mancettur9, Worceter10.

2. The Anglo-Norman forms have not prevailed.

1. Colchester (Es.). 2. Winchester (Ha.). 3. Godmanchester (Hu.). 4. Horncastle (La.) l. 5. Chichester (Ss.).

12th c. — Index. Gutmuncetre3 (1146—54).

1 3th c. — H.B. Colcetre\ Hornecatre*. B.G1. Wincetre*, Wincheter*, Cicetre*.

Though owing to A.N. influence medial s before c, p, t was, no doubt, lost in many other English place-names 2, the forms

1 The second element originally contained O.E. ccester, which in comparatively recent times was replaced by castle.

2 To the instances we have already adduced from D.B. may be added Glatirioergere, Mousket, v. 18506, Witenehet, Fr.Ch. = Wheatenhurst (Gl.), as

74 R. E. Zachrisson

without s do not seem to have prevailed except in the case of ceter < cester. We have therefore devoted this section solely to the study of the cefer- forms. Under this heading we will, more- over, consider the further development of these forms, thus trying to define in how far their present spelling and pronun- ciation in general depends on English or French sound-develop- ment.

That cester was pronounced ceter is proved not only by the spellings without 5, but also by rhymes like Wyrceterre: letter, R.G1. (see F. PABST, Die Sprache des R. von Gl., Diss. Berlin 1889, p. 136), as well as the modern pronunciation of Exeter and Cirencester.

The first instances of loss of s, to which we can assign a comparatively safe date, are from La3amon A (c. 1220): Axcetre, Glocet'e, and, in all probability, also Glocecre (= Glocetre), v. 16482, which form LUHMANN, p. 51, marks as dubious. From this time downward s was certainly no longer pronounced, although it was mostly kept in the spelling. In Capgrave's chronicle the s- less forms are, curiously enough, the most usual ones.

It is worthy of note that some places in modern Chester, as also one in -castle < earlier -caster, exhibit M.E. forms in -ceter, -cater: Colcetre, Ciceire, Wincetre, Horncatre, Gutmuncetre. The latter form would be a very early instance of loss (c. 1150), supposing it is taken from an original charter and not — which is more probable - - from a late copy.

In his above-mentioned treatise on the language of R.G1. PABST, p. 136, accounts for the loss of s in cester in a way quite different from ours: »Wir konnen also den Schluss

ziehen, dass der Ausfall des s vor dem tr einzig und allein durch das Vorhaudensein des vorhergehenden c herbeigefiihrt worden ist, und dieser Schluss wird auch fiir das obige auncetres x gelten.

also some inverted spellings like cet Cynestan, A.S. Chr., MS.D. = Kintbnry (Be.), Cheneteberie, D.B.; Lastus de Sutton (1316), F.A. = The Lathe (division -of land), of Sutton (Ke.).

1 As to ancetcr <. ancester we are of opinion that the transition is due to the analogy of the numerous -ceter forms. Spellings like ancetcr appear at the very time when cester must have been generally pronounced ceter: aunceter, Ch., C.T., v. 6737 awiceterez, All. P. 2., 258 aunceteres, Will , v. 5132;

Anglo-Norman influence on English place-names 75

Wir haben es mit einer Dissimilation zu thun, die wahrscheinlich erst auf englischem Boden eingetreten ist, und die jedenfalls mit dem Schwunde des gedeckten s in Frankreich in gar keinem Zusammenhang stehtU In our opinion this theory is absolutely to be rejected. Instead of explaining the transition in question as due to A.N. sound-change, Pabst invents an English sound- law which has never existed. As far as we know, dissimilatory loss of s has not been evidenced either in Standard English or in the dialects. Cf. also POGATSCHEK, Englische Studien, 27, p. 275. On the other hand there is nothing astonishing in the circumstance that places like Gloucester and Leicester, which were daily spoken of by many thousands of people, and which, more- over, not uufrequently were used as personal names (cf. Joh. Leycetre (1409), Index, Tli. Gloucetre (1404), ib.), should be more subject to renewed French influence than ordinary words the in- troduction of which into the English language was long ago a »fait accompli » 1.

Nearly all places in -cester are nowadays pronounced in two syllables. MAYHEW, Academy, 1240, p. 17, has advanced several theories to account for this state of things, without arriv- ing at any definite results. It is, however, a well-known fact that in English words of three syllables the unstressed vowel of the penultimate was syncopated at various times 2. Cf. LUICK,

auncetres, E.GL, p. 193 etc., see Stratmann-Bradley and Matzner. More forms without s are noted in N.E.D., from which the following passage may be quoted : — »In English ancessour soon became obsolete: ancestre became phonetically ancetre, of which the regular form, now dialectal, is anceter or anster*.

1 We may point out here that owing to later French influence s has been lost before t also in a modern English surname, Grossetete = >Great- head>, which in the H.R. and the K.Q. appears as Groceteste, Grossetest. See BARDSLEY, Die., p. 340.

2 Mayhew raises the following objection: - - »It may be that the pronunciation Lester stands for an earlier *Leiceter (we have already seen that such forms are by no means theoretical) with syncope of the vowel e in the unstressed penultimate. But this hypothesis entails the difficulty of the retention of the historic spelling -cester in spite of the intervening -ceter stage*. It is indeed, very curious that Mayhew is so astonished at the traditional spelling being kept in a language like his own, which still to a very great extent preserves the mediaeval spellings. Besides, even in French s was kept orthographically before t as late as the 18th cent., when it had ceased to be pronounced more than 400 years ago.

76 R. E. Zachrisson

Anglia .XX, p. 353: - - »Jedes einfache dreisilbige wort erleidet im lauf der englischen sprachentwicklung synkope des mittels- vokals, wofern nicht kunstliche einfliisse sie hemmen, und wo- fern die lautliche gestaltung des wortes sie zulasst.» Consequently, Glouceter, Leiceter (etc.) were developed to »Gloster», »Lester» in perfect conformity with the phonetic laws *)f the English language.

The new pronunciation is indicated by the spelling in one single case, viz. Craster < Craucestre1, F.A., the little Northum- brian fishing-place with the seat of Craster Tower. Here the etymological connection with -cester has been forgotten, and the name of the place appears in phonetic spelling.

In all other instances the M.E. spelling has been kept in spite of the modern pronunciation: Worcester, pron. (wusta); Gloucester, pron. (glosta); Leicester, pron. (lesta); Towcester, pron. (toista), Rippmaun.

For the following places, most of them small towns or vil- lages, there exists, by the side of the phonetic pronunciation, another due to the influence of, the spelling: Alcester, pron. (oolsta), Rippmaun, (oo(l)sta), Tanger, phon. pron., (selsesta), Tanger, spell.-pron.; Rocester, pron. » Roaster », Hope, phon. pron. (rosesta), Tanger, spell.-prou.; Bicester, pron. (biste), phon. pron., (baisesta), spell.-pron., Tanger; Cirencester, pron. (sista), phon. pron., (sairansesta), spell.-pron., Rippmann. Tanger and Rippmann besides give the pronunciations (sizita), (sisesta), which are to be considered as intermediate stages in the development Cirencester > »Cister».

We have not been able to obtain any information about the pronunciation of Mancetter, Frocester, and Wroxeter. To judge from the previous cases, it. is highly probable that there exists a local pronunciation (frosta), (msensta), (roksta) by the side of spelling-pron. like (frosesta), (msenseta), (rokseta).

As to Exeter the only pron. given by our authorities is (ekseta). It is, however, by no means probable that (ekseta) represents the original local pron. Tanger gives (eksanz) < Anglo- Latin Exonia, as the pron. used by the inhabitants of the place, and in maps and geographical works of the 17th aud 18th cent. Exonia is by far the more usual designation of the place. If

1 Magna Britannia still has the spelling Craucestre.

Anglo-Norman influence on English place-names 77

the pron. (eksete) was mostly used of the place not in the place it is easier to understand why the unsyncopated form has prevailed l.

The names in -cester occur as personal names in many interesting, partly phonetic spellings, some of which may well be worth mentioning. The following instances are all of them taken from Bardsley's Dictionary of English surnames. We give first the modern forms, then a few early spellings: Wurster, Worcester, Worcester or Worcettor (1569), Worster (1619), p. 828; Lester, Leicester: Leicester or Lester (1604), p. 478; Rossiter, Eociter (1581), Eossiter (1641), p. 655.

We will now proceed to examine whether the modern spel- ling and pron. of the place-names here treated exhibit any other A.N. peculiarities than s for ch and loss of s before t.

ou in the first syllable of Gloucester is generally supposed to be an A.N. alteration of au. Cf. STOLZE, p. 28, and WEST- PHAL, p. 16, who even compares Glaw- > Glou- with audire > ouir. But the Normans were familiar with the sound-combina- tion au (< al -f- cons.), so there was no reason for a substitution. Would it not be possible to consider Glouuecestre, D.B. etc., as a continuation of Britain Cair Gloui, Nennius (Mon. Brit.)? Our Latin authorities have Clevum, Itiner. Anton., Glebon Colonia, Oeogr. Rav.; see FORBIGER. Ill, p. 289. Whether there is any relation between the forms Gloui and Clevum (for Glevum?), we -cannot decide. This question may perhaps be settled by Celtic philologists. In many 12th cent, records au, aw, are usual spel- lings by the side of ou. The A.S. Chr. has Gleaweceastre and Glotveceastre, Henry of Hunt, (Mon. Brit.) has Gloucestre, Gloue- ceastria, Florence of Wo. (ib.), Glavorna, Glauweceastre, Symeoii of Du. (ib.), Glocestria, Ethelwerd (ib.), Gleaucestre etc. As late

1 I think I have found at least some traces of the phon. pron. Thus, an English orthoepist of the 18th cent, warns against the dialectal pron. Cassel of Exter for » Castle of Exeter*. See FOKSTEB, Anglia XXIV, p. 116. - Hexter, a not uncommon Devonian surname, is derived by BAEDSLEY, Die., p. 379 from Exter < Exeter with addition of an inorganic h. The tendency to drop h's is, indeed, very strong in Devon. During a stay there I have noticed it in people of very good education, whose English in every •other respect was quite correct.

78 R. E. Zachrisson

as in Rob. of Gloucester's chronicle we find forms like Claucestre, Claucetre1. Note that the true English form in La3amon, Gloi- ehestre, see § 1 p. 19, appears with o as well as ch.

To account for the development of Cirencester to Cice(s)ter, Birencester to Bice(s)ter, Worcester to Woce(s)ter, it is hardly necessary to assume French influence. The n of Cirencester and Birencester has been dropped owing to *a general M.E. tendency. Cf. HORN, Gramm. § 224. Instances of loss of r be- fore s are noted by HORN, Gramm. § 235 Anm. Cf. also EK- WALL, Jones's Phonography §§ 583, 586 2.

Appendix. Some cases of analogical transformation on the pattern of forms in -cester, -Chester.

1. Uttoxeter (St.). Of the name of this place we have noted the following early forms: Wotocheshede, D.B.; Utukeshes- tra (12 c.), Fr.Ch.; Uttoclceshedere, Utukeshere (12 c.), see DUIGNAN, St. Pl.Ns., p. 157; WittoJceshather, T.N.; Huttolces-, UittoJcesather, Ch.R. I; Ottokeshather (13 c.), see Duignan; Utoxatre, Uttoxhather, UttocJccester, Utcheter (14 c.), Duignan; Vttolceshather (1348), Index; Uttoxather (1401), ib.; Vttexatur (1536), ib.; Vttoxatur alias Vtset- tur (1567), ib.; Utccster, Utseter, Uttecester (16 c.), Duignau; Vttoxather (1608), Index; Utcester, Camden.

Uttoxeter has been derived from an O.E. Uttocceaster by several authors (Camden, Taylor, Barth. etc.). Whatever the se- cond element of this word may be, it has certainly never con- tained O.E. ceaster. The supposed O.E. form Uttocceaster must have been invented by Camden. As is correctly pointed out by Duignan, all the mediaeval terminals point to M.E. hather = N.E.

The forms in Clau- may, however, owe their existence to the old myth that Gloucester was once the seat of the emperor Claudius, who is said to have given it to his son Gloi or Gloui. See MADDEN, La3amon III, p. 339. In the Latin history of Lanfranc (see A.S. Chr., ed. Earle-Plummer, I, p. 287) Gloucester is called Claudia civitas.

2 On the other hand loss of r before 8 in the following early instances from D.B. etc. is, in all probability, due to French influence: Holehest, Linhest, Eisseburne, Esselorne, Messe, D.B. = Holdenhurst (,Ha,), Lindhurst (Ha.), Hurst- bourne (Ha.), Marsh (Sh.); Luvehest, Fr.Ch. = Ewhurst (Es.); Brockehest: Nove Forest, Gaimar, v. 5799 = Brockenhurst (Ha.); Dosete, Benoit, v. 14, v. 22 = Dorset. The question of loss of r before s has recently been made subject of further investigations by POGATSCHER, Engl. Stud. 31, p. 261 ff.

Anglo-Norman influence on English place-names 79

heather, which in place-names may have had the meaning of heath.. As to the mutual relation of the words heath and heather see EKWALL, Archiv 121, p. 135 if. Forms like Uttockeshederc, Ut- toxather had, undoubtedly — in spelling and pronunciation — some resemblance to certain place-names in -caster, e.g. Exeter and Wroxe- ter. Hence -cester took the place of -hather in some spellings of the 14th cent. The modern pronunciation, usually (raksete), Tanger, has obviously arisen from such a form as Uttockcester. Hope gives as local pronunciation the extremely interesting notations »Uttcheter», »Ucheter». Cf. the 14th cent, spelling Utcheter. We can suggest no other explanation of this extraordinary form than that it is due to a blending of A.N. -ceter and English -Chester. If this conjecture is right the form Winchetre in Robert of Gl. is probably no scribal error, but has originated in the same way as Utcheter.

2. Grantchester (Cb.). The early spellings are as follows: Granteseta, Grantesefe, D.B.; Grenteseta, I.C.C.; Grenteshet (1210), R.B.; Grancete, Granteset, T.N.; Grantecef, H.R.; Grantecete (1284), F.A.; Gransete (1302), ib.; Grantesete, Grancete (14 c.), Cat. A.D.; Grauntcestr' (1327—30), P.R.; Grantecestre (14 c.), Ramsey Ch.;. Graunceter, Grancesfer (15 c.), Cat. A.D.; Graunsete (1428), F.A.; Gratvnceter, Graundeceter, Graundecestre (16 c.), Cat. A.D.; Graun- cestor (1549), Grauncester (1592), Index.

Grantchester is identified with an O.E. Grantaccestir by MILLER, Quellen und Forschungen 78, pp. 60, 61, and TAYLOR, Names etc., p. 137. Even in such an important and quite recent work as R. L. POOLE, Historical atlas of modern Europe (1896—1900) we find, in the part called England and Wales be- fore 1066 (by W. H. Stevenson), the same, certainly erroneous,, etymology.

The first time we meet the form Grantaccestir in the O.E. literature is in the Latin Bede, Lib. VI, Chapter XIX, where it is related that Queen JEthelthryth's body was removed to Granta- ccestir to be buried: »ascensa navi venerunt ad civitatem quandam desolatam non procul inde sitam, que lingua Anglorum Grantaccestir vocatur*. In our opinion the place Bede alludes to is Cambridge not Grantchester. On Bede's authority the name occurs again in some Anglo-Saxon charters: into Grantanlricge-

80 R. E. Zachrisson

(original of 970), Birch 1267; in provincia Grantaceaster (Latin version of the same), Birch 1266; in comitatu Grantecestrice, Kemble 907; comitatus Grantebrigice, Birch 872. It is, however, obvious that the locality here referred to is Cambridgeshire, not the town of Cambridge. Gaimar, when defining the extent of the Kingdom of the Southumbrians, mentions Grantaceaster in the following way:

Neis la meite de Grantcestre

i jut jadis et devereit estre; v. 1605 f.

We need not hesitate to identify Grantcestre with Cam- bridgeshire also here. The very wording of the passage from Bede quoted above indicates that the author was not very well acquainted with the locality in question. He may have known the river Granta, and then added the usual -ceaster, applied by the Saxons to almost any place where the Romans had settled. Of. Lundenceaster (see POOLE, Historical Atlas) for modern Lon- don etc.

The usual form for Cambridge, and the only one in the A.S. Chr., is Grantaribrycg l. After Bede's time Grantaceaster is used by the side of Grantanbricge, but only as a designation of Oambridgeshire.

In his Place-Names of Cb., p. 39, Prof. SKEAT also iden- tifies Bede's Grantaccestir with Grantchester, but at the same time draws attention to a passage in LYSON'S History of Cam- bridgeshire, where is noted that Walter de Merton gave to Mer- ton College, Oxford, a certain manerium de Grauntesethe. »It has often», he continues, »been supposed that this form is only another spelling of Grantchester. Such seems to be the fact, though there may have been some confusion with A.S. scete.» We be- lieve the learned author would have come closer to the truth had he inverted his statement, and said that Grantchester is only another spelling for Grauntesethe. All the early forms unambiguously point to an O.E. * Grantanscete. From a histori- cal point of view it is also highly improbable that the manerium de Grauntesethe, the little village of Grantchester with a popula- tion of a few hundred inhabitants, should be identical with a

1 However, in some entries from the I.C.C. the terminal -burg appears by the side of -brigge: Grenteburga, Granteburga, Grantebrigge.

Anglo-Norman influence on English place-names 81

place of considerable importance already at Bede's time when, moreover, there is in the immediate neighbourhood the old Roman Camboricum (= Cambridge, FORBIGER III, p. 292) 1.

The modern form of Grantchester has certainly originated in the same way as Uttoxeter, i.e. by false association with •cester, -Chester. Gransete > Grance(s)ter, whereupon an English form Grantchester came to be used by the side of the French one. The difference between Uttoxeter and Grantchester is that in the former case the French form has prevailed - - at least in the spelling - - in the latter, the English.

3. Penshurst (Ke.), Pensherst (1072), Index; Penecestre (c. 1270), Ke. P.F.; Penecestre (1271), I.P.M.; Penesherst, H.R. Penecestre (1316, 46), F.A.; Penshurst (1343—46, 77—81), P.R.; Penseherst (1467—77), P.R.; Penshurst (1512), Cat.A.D.

The second element of this place-name is obviously O.E. hyrst with Kentish e. The confusion between (Pen)sherst and -cester may be accounted for in different ways: Penserst may have been changed into *Pensert (cf. D.B. forms like Salhert (Ss.), Witenhert (Gl.), as well as Chilbehert, Fr.Ch., = Chippenhurst (Ox.)), which subsequently became *Pensetr by the metathesis of r. There is another possibility, viz. that s was originajly kept, and r shifted : Penserst > *Pensestr (cf . Bolehestre, D.B. = Bolnhurst (Bd.)). In this case the analogous forms were soon ousted by the etymological ones.

4. Bilchester or Bilchurst in Hawking (Ke.) is one more instance of confusion between s -f herst and cester, -Chester. The only old spelling we have been able to find is Bylcherst (1346), F.A.

5. Cruwys-Morchard (Dv.), Morceta, Morceth, Morchet, D.B.; Morcestre, H.R.; Morcestre, Morceth, T.N.; Morcestre Crues (1284, 1316, 46), F.A.; Morchard (1303), ib.; Cruwesnorchard (1381—85), P.R.; Morchestre (1428), F.A.

As to the etymology of this place-name, we can only sug- gest that it is due to the O.E. formula cet f&m ortgearde or

1 After writing this, we have found the following interesting passage in Henry of Huntingdon (Mon. Brit., p. 692 a), which eminently corroborates our conclusions as to the identity of Cambridge and Grantanccestir : — »Kair- Grrant, id est, Granteceastria quce modo dicitur Cantebrigia.*

6

&2 R. E. Zachrisson

orce(a}rde. If our conjecture is right the m of the article was pre- fixed already in O.E. time. Macclesfield (Ch.) would he a parallel instance, supposing RITTER, Archiv 116, is right in deriving it from <Bt pcem JEcdesfelde. *Morcerd appears in D.B., with sup- pression of r and change of final d into t, as Morcet, Morchet. Morcet or *Morcert is turned into *Morcetr \ Morcestre, which gives origin to an analogous English form Morchestre.

From the above instances we have seen that the termina- tions of several English place-names are liable to be trans- formed on the analogy of the numerous -cester, -ctiester forms. Such a transformation may take place if the termination, in itself or owing to native or French development, shows some resem- blance to -ce(s)ter, more seldom -Chester. In its turn the spelling -cester gives rise to its English correspondence -Chester. In two cases such forms, due to the operation of analogy, have ultima- tely prevailed over the etymological ones, viz. Uttoxeter for Uttoxhather and Grantchester for Grantsete.

§ 6. Loss of the interdental spirant -. A. In place-names containing O.E. mu]>(a) 3.

1 . Cockermouth (Cu.). 2. Dartmouth (Dv.) < O.E. Dertamutia, A.S. Chr. 3. Axmouth (Dv.) < O.E. Axamupa, A.S. Chr. 4. Plymmouth (Dv.). 5. Portlemouth (Dv.). 6. Sidmouth (Dv.). 7. Teignmouth (Dv.). 8. Wey- mouth (Do.). 9. Yarmouth (Ha.). 1 0. Portsmouth (Ha.). 1 1 . Stourmouth (Ke.) < O.E. Sturemupa, A.S. Chr. 1 2. Monmouth (Mo.). 1 3. Jesmond (Nb.). 14. Learmouth (Nb.). 15. Tynmouth (Nb.) < O.E. Tinanmuj, A.S. Chr. 16. Alnmouth (Nb.). 17. Yarmouth (Nf.). 18. Yarmouth (Sf.).

D.B. Alsemuda, -e3 (Is is possibly meant for MS = x\ Porle- muta, -e5, Ermud9, Gernemiva11, Jernesmua1*, Gernemutha1*. Geld. Inq. Alsemuda?, Axrnuda*.

1 Such a shifting of r from one syllable to the other is very usual in the early Anglo-Latin records : Ferdeston, D.B. = Featherstone (Sh.); Congrethorp, Cronkethor, D.B. = Crownthorpe (Nf.) etc.

2 From reasons stated below forms exhibiting A.N. (?) (t), (d) for M.E. final QO and intervocalic (d) have also found a place in the lists belonging to this paragraph.

8 In consideration of the great bulk of the material we have brought similar instances together under the same heading, which arrangement we hope will lead to a more comprehensive view of the whole.

Anglo-Norman influence on English place-names

83

12th c. 13th c.

Fr.Ch. Portesmue, -mues™, Monemuta, -mud, -mue18, Munemud1*. Index. Gernemuta11 (1163).

C.R. Cocremuth1, Kokermul (04— 27), Cokermue1 (88— 96), Waymue* (ib.), Waymo* (79—88), Portesmua, -muha, -mues10 (04—27), Portes- mue10 (27—31), Porthesmuth10 (72—79), Portismuth10, Portesmue10 (79—88), Portismuth10 (88—96). Ch.R. Cokermouth1 (1227), Porte- muth* (1280), Teingnemuth1 (1253), Weymuth* (1258), Weymue* (1284), Portesmue, -muda, -rnuta10!, Sturnmthe11 II, Levermue1*!, Tinemuda, -mue, -muthr°, Gernemuth, -muesli. F.A. 1284. JL^emwe3, Porle- mue, -muwe6, Sydemue6. H.R. Dtrthamuth*, Dertemuth, -mue*, Axe- mues, Portlemuive5, Tengemue'1, Waymue*, Weymuth*, Gesemuthe13, TynemutW0 , Gernemue*1, Jernemuta11, Gernemuta^. Index. Eretmia"3 (1206), Partenme™ (1231), Portesmues10 (1200), Jezemue™ (Hen. III.), Gernemothe1*. La3amon. Dertemufie2, v. 1786, Gernemude1', v. 30542. Nb. Rolls. Gesemue1* (1279), Alnemue™, Allemue" (1256). P.R. Horemue9, Eremue* (25—32), Gernemue, -muth11 (16—25), Jernemwe11 (25—32). R.B. Cocremuce1, Portesmue™, Sturre-, Stormue11. R.G-1. ^arnemouthe11 , v. 3429, ^ernemoupe^1 , v. 4670. T.N. Porlemue5, Sidemue9, Teigemue1, Portesmue10, Jose- mouth18, Levermuwe1*, Jernenmth, -mula11, Gernemive11, Jernemuth,

14th c. — C.R. Eremuth9 (07—13), Yaremuth* (41-43), Jernemuth1'1 (43-46). P.A. Axmue* (1303), Axemuth* (1316), Axemouth* (1346), Portle- tnue* (1303), Portelemouth* (1346), Sidemue* (1303), Tengemwe1 (1303), Teynghmue'1 (1346), Ermuth9 (1316), Sturmouthe11 (1346), Monmue™ (1316), Levermuth™ (1346). Index. Jesemuth1* (1334), Jernemuth, -muta1*. P.R. Teignemouth1 (1399—01), Yaremuth9 (43-45), Iremuth9 (45—48).

15th c. — Capgrave. Dertemouthe*, p. 239, Plummouthe*, p. 284, Fortes- mouthe10, p. 239, Tynmouth15, p. 240, YermoW, p. 208. F.A. Cokermouth1 (1428), Waymouth* (1431), Levermuth14' (1428). Index. Sydemouth" (1420), Eruemouthe* (1488—89), Jesmowetlie™ (1421), Jernomewth, -moth11 (1420). P.R. Cokermouth1 (67— 77), Waymouth* (61-67), Portesmuth10 (ib.), Portesmouthe10 (67—77), Tyngmouth1* (ib.), Jernemuth17 (01—05), Jernemouth11 , Yermothe11 (61—67).

16th c. — Index. Yearmouth* (1593), Portesmouthe10 (1578).

B. In place-names containing O.E.

1.

present forms exhibit th.

1. Southwick (Ha.). 2. Southampton (Ha.) < O.E. Sudhamhin, A.S. Chr. 3. Southington in Selborne (Ha.). 4. Southwell (Nt,). 5. Southwark (8r.). 6. Southam (Wa.) < O.E. Sutiham (1101), Index.

fc4 R. E. Zaehrissori

D.B. Sudwelle*, Sudwerca, -iverche*. 12th c. — Fr.Ch. Suwice1. Index. Sudwelle* (1130), Sudwerk*, Sutwerc*,

Suicerch*, Suiverk6. 13th c. — Ch.B. Suwick1, Suhanton*, Suivell*, Suham*, Sutham*. Index.

Sudiverk* (c. 1200), Swotrk* (Hen. III.). T.N. Suwerc*. Sel-

borne Oh. Soynthone* (1230), Suthintune* (c. 1250), Sudington*

(c. 1260), Sudintone* (c. 1260). 14th c. — Index. Sutheivell4.

2. The present forms exhibit d.

1 . Siddington (Gl.); * in the first syllable is very curious and diffi- cult to account for. Cf., however, KHUISINQA § 235. 2. Sodington (Wo.) < O.E. Suthlune, Suthinton (a. 825, 957), Birch 386, 1007.

D.B. Sudin-, Suintone1.

1 2th o. — Fr.Ch. Suthintuna1, Suinthon1, Swinton\ Suenton*. 1 3 th c. — T.N. Sudin-, Suthinton1, Sutin-, Suthinton9. "Wo. S.B. (see

DUIGNAN, Wo. PJ.Ns., p. 151). Suthintone*, Sodinton2 (1275). 1 4th c. — F.A. Sotingtori*, Sodyntori*.

C. In place-names containing O.E. wipfe, wifig, wiping.

1. The present forms exhibit th.

1 . Withycombe (Dv.). 2. Withington (Gl.) < O.E. Widian dun (a. 700), Earle. 3. Withern (Li.). 4. Withington (Sh.). 5. Withyford (Sh.). 6. Withyconib (8m.). 7. Withypool (Sm.). 8. Withyham (Ss.). 9. Withy- brook (Wa.).

D.B. Widecoma1, Widindune\ Widerne*, Witntone*, Wideford*,

Widecombe*, Inwidepolla1 .

1 2th c. — Index. Wiernia*, Wihemia*, Widebroc9.

13th c. -- Oh.B. Wichenton*, Witene*!. F.A. Widecombe1, Wydinton* . (1284), Videcombe*. H.R. Wydecumb, -counibe1, Wythindon9, Wy-

thyford*. Index. Wydecome1 (1273), Wierna*, Wyern*, Within-

ton4 (1267), Widebroc9 (1208). 1 4th o. — F.A. Widecombe1, Wydinton* (1303), Witherne* (1316), Wythin-

ton* (1316), Wydecomb* (1303, 46). Index. Wythcrne* (1343),

Withihamme* (1314), Wydyham* (1326). 1 5th c. - F.A. Widecombe1 (1428), Wyerne* (1428), Wymccombe*, Widepole1

(1428). P.B. Wythyrn* (61-67), Wythern* (67—77). 1 7th c. — Index. Wythyham* (1578).

Anglo-Norman influence on English place-names

85

2. The present form exhibits loss of th.

Wyville (Le.). T.N. Wiwell. H.B. Wywell (pers. name). P.A. Wywett (1316), Wydewell (1428), (Rob.) WyviWas (1428),

' 3. The present, forms exhibit d.

1 . Widdicombe (Dv.). 2. Widecombe (Dv.), called Wethicombe on the map ' of Dv. in Camden, and in a book called Index Villarum (17th c.). 3. Widford (Ht.). 4. Widdial (Ht.). 5. Widford (Ox.) < O.E. Widigford, Kemble 426, 177. 6. Widecombe (Sm.).

D.B. Widecombe1, Widecombe*, Wideford', Widihale*, Widiforde\ 13th c. — Ch.B. Widecumba*. H.B. Vydyford*, Wydeford6. H.B. Wide- ford3. Tax. Eccl. Widiford*. 14th c. — F.A. Widecumbe1, Widecumb*. 15th c. — F.A. Wydecomb*, Wydiford*, Widecombe6 (1428).

D. In place-names containing O.E. hyf, hcep, rip,

1. The present forms exhibit th.

1. Tingrith (Bd.) < O.E. rip. 2. Shepreth (Cb.). SKEAT, Pl.-Ns. of Cb., p. 43, thinks the second element contains O.E. tvrceth = ^enclosure* etc.; the many early forms with i point, however, to rip. 3. Horsheath (Ob.) < O.E. Keep. 4. Meldreth (Cb.). SKEAT, I.e., derives also the second element of this place from O.E. wrcep, here, no doubt, with greater probability, as all the early forms have e. 5. Aldreth (Cb.) < O.E. hyp. Q. Rawreth (Es.), probably < O.E. rip. 7. Chawreth (Es.), also here the terminal may contain O.E. weep. 8. Earith (Hu.) < O.E. Earhyp, Earhyth, Kemble, Earhip (a. 988), Facsimile of Charters. 9. Oteringhithe (Nf.) < O.E. hyp. 1O. Rack- heath (Nf.)? < O.E. hap. 11. Lakenheath (SI) < Lacyngahyp, Thorpe. 12. Lambeth (Sr.) < O.E. Lambhyp, A.S. Chr., Lambehip, Thorpe. 13. Reeth (Yo.) < O.E. rip.

D.B. Tingrei1, Escepride*, Esceprid*, Horsei*, Melrede*, Otringe- heia?, Rachdtha10, Racheia™, Lakingaheda, -hethen, Lanchei" (nc = m\ Riels.

12th c. — I.C.C. Sepeia*, Horseda*, Melreda*. Inq.El. Sceperia*, Sceperee, -eie*, Melrede*, Melrede*, Meldrcthe*, Lakincgeheda11, Lacking ehetha11, lagingehyda11, Lachingehydali, Lachinchede11, Lagingchee", lachincg- heda11, lachingehea11. Index. Melreda* (1103), Lachingeia11, Lach- ingdhutha11. Pi.B. Alreheda, -hedra, -hudra*. Sf. P.F. Lakinge- hee11 (1199).

86 R. E. Zachrisson

13th c. — Annales Mon. Tingrize, -rithe, -ri, -rie1. C.R. Chaurethe (88-96), Erktz* (79—88). Ch.R. Meldred', Melre* I, Raureth* II, Chaure1 (1246), Racheth™ I, Lakingheth11, Lamhee™ I, Lambithd1* II. H.R. Tyngri1, Schepere*, Scheppere2, Horsefh*, Raere6, Chaure1, Schaure1, Erhcth*, Erhee*, Rakhey10, Rakkeye10, LakingheW1. Index. Chaure1, Otrigeide* (1203), Lamhethe™. R.B. Lambhyeth**, Lamhe1*. R.G1. Lamhupe1*, Lambhepe™. Ramsey Ch. Alderhithe,-hethe6, Alreheda\ Herhelhe*, Herhythe, -ithe*, Eritlte*, Erethe*, Erhe*. Sf. P.F. Lakingheye11 (1219). T.N. Tingre1, Sepree*, Rareg'6. Yo. K.Q. Rithe1* (1284).

14th c. — F.A. Tingrie1 (1302), Scheperethe* (1302), Melreth*, Raureth, -ryth« (1303), Chaureth1 (1303, 46), Oteringhithe*, Racheyth10 (\3lfy. Index. Horseth* (1305), Raurethe" (1370), Lakynghethe11 (1390), Southlam- hethe". P.R. Lakynghithe11 (45—48). Sf. P.F. Lakynghethe11 (1313), Lakynghythe11 (1330). Yo. N.V. J&WS (1316).

15th c. — Capgrave. Lambhithe1*, p. 313. F.A. 2>?#re (1428), Chaureth1 (1428). Index. Rakhith10 (1451), Lambehith1*. P.R. Shepereth* (61—67), Malrede*, Melreth4 (61—67), Chawreif (67—77), 7ic^11 (67—77).

16th c. — Index. Horseth*, Rawriih* (1541).

2. T/ie present forms exhibit loss oj th.

1. Childrey (Be.) < O.E. Cilia-, CyllanriJ), Kemble 746, 1133 etc. 2. Cricksea or Creeksea (Es.) < O.E. liijp. 3. Sawtry (Hu.) < O.E. wrcep; cf. SKEAT, Pl.-Ns. of Ht., p. 62. 4. Cbelsey (Mi.). In Speculum Britannia )c. 1600) we find the following note on the place: — »Chelsey also called Cheselsey for the sake of the pebbles >. The etymology is, however, O.E. Ctelic hyp, Earle. As to present s for older ch=(tf), cf. Appendix C. 5. Stepney (Mi.) < O.E. hyp. 6. Setchey (Nf.), probably < O.E. hyj). 7. Boul- ney (Ox.), probably < O.E. hyp. 8. Shottery (Wa.) < O.E. Scottarip, Earle

D.B. Criccheseia\ Saltrede*, Chelched*, Stibenhede*. 12th c. — Fr.Ch. Stebeheie*. Index. Saltereia8, Saltreia*, Sauteria*, Ste-

benheth*. 13th c. — Ch.R. Sechie* II. F.A. Chelchehuth* (1284). H.R. Sechize*,

Bulhud1. Index. Sechithe* (Hen. III.). Ramsey Ch. Saltreche

(ch = th\ reye, -reio, -rethe, -rea8, Sautre3, Sautreia*, Sautereie*.

T.N. Sethiche* (error for Sechithe], Bulchude1. 14th c. - Cat. A.D. Scotrith*, Schotrythe*, Shotriche* (ch = th\ Schot-

rith* (all of them from the year 1318). F.A. Chelrey1 (1316),

Sautre* (1316), Stebcnhuth* (1316), Sechhithe* (1316), Buletchurche1

(1316). P.R. Chelcheheth* (1348).1 15th c. — F.A. Chelrey1 (1428), Crikesheth* (1428), Chelcheth* (1428), Steben-

hith6 (1428), Bulnehith1 (1428). Index. Chclchithe* (1465). P.R.

Kyrk Hithe*, Crykeshithe* (67—77), Setcheif, Segthe* (61—67).

Anglo-Norman influence on English place-names 87

1 6th c. — Index. Chellsayt* (c. 1550), Seche* (1564). Mi. PF. Stebenhithe* (47—53), Stebunheth* (1568), Strpneif (1575), Stepney* (1587).

3. The present form exhibits t.

Orsett (Es.) < O.E. *Horshce]>. D.B. Orseda. H.B. Horsete. Cat. A.D. Orsete (c. 1242). C.B. Horseth (1343—46), Orseta (1318—23, 46—49). F.A. Horseth, Horsete (1303, 1428).

4. TAe present forms exhibit d.

1 . Maidenhead (Be.) < O.E. liyp. 2. Hendred (Be.) < O.E. Henne- Hennanrip, Birch 165, 391 etc. 3. Cottered (Ht.), perhaps < O.E. iorcv]); cf. SKEAT, Ht. Pl.-Ns., p. 61, p. 64. 4. Oxnead (Nf.)? < O.E. Keep. 5. Leather- head (Sr.) < O.E. Lcodrithan, see TAYLOR, Names,, p. 170.

D.B. HenreP, Chodrei5, Oxenedes*, Leret*.

1 2th c. — Fr.Ch. HeinreiP, Henreth*. Index. Henreda* (1157). 14th c. — Ch.B. Heanred* (1227), Hanred* (1274). H.B. Henreth*. Index.

Hendred2 (1225), Ledered*, Leddred6, Ledrede5. B.B. Henred*,

Coddre*, Codreye*, Ledrede*, Ladrede*. T.N. Hen-, Hanreth*,

Hanred*, Leddred'a. 14th c. — C.B. Oxeneddes* (07—13). F.A. Hanreth* (1316), Codrede3 (1303),

Oxneth* (1316), Oxenedis* (1302). Index. Oxnegges* (1312). P.B.

Oxtned4 (77—81), Oxe?iedfe4 (81-85). 1 5th c. — Cat. A.D. Codreth* (1428). F.A. MaydenJtythe1 (1428), Hanreth*

(1401), Codrethe* (1402), Codreth9 (1428), Oxenede* (1401, 28). I.P.M.

Henrede* (Hen. VII.). P.B. Madenhithe1 (01—05), Maydenhede1

(67—77), Esthenrith* (61—67), Henrede* (67—77), Occn^4 (61-67),

Oxeney* (67—77).

E. In place-names containing O.E. ^Z, ^(e)r.

1. The present forms exhibit th, d.

1. Wetherley (Cb.) < O.E. w$<r; cf. SKEAT, Pl.-Ns of Cb., p. 66. 2. Hatherton (Ch.)?< O.E. Hathor, Searle. 3. Netherexe (Dv.) < O.E. niperra, neopera, Mdf. p. 98. 4. Headley (Ha.) < O.E. A^jB. 5. Southry (Li.) < O.E. SM£. 6. Blatherwick (Np.). 7. Netherfield (Ss.) < O.E. niferra etc. 8. Featherstone (Yo.). Cf. Featherstone (St.) <C Feotherestan, DUIGNAN, St. Pl.Ns., p. 60. 9. Rotherfield (Yo.).

D.B. Wederlai1, Haretone2, Niressa, -es, Hallege*, Sutreie6, Sudtrie5, Blareiviche*, Nedrefelle1, Nirefeld1, Ferestane*.

12th c. — I.C.C. Werleia1, Werleie1. LEI. Wederlai1. Index. Blarewihc* (1199). Np.S. Blatherwyk*.

88 R. E. Zachrisson

13th o. — O.B. Suthreya* (88—96). Ch.B. Fetherstane8 II, Sudereie'0 I. H.B. Suytray*, Neddrefeld'\ Index. Sutherayeb, Surrey a*, Suth- reya* (Hen. III.), Blathinvyk* (1248). Yo. S.B. Rorefeld9 (1297).

14th c. — F.A. Nyther-, Nitherexe*. Index. Suthereya^ Blatherwyke* (1307).

15th c. — F.A. Sotheray* (1401).

2. Tfte present forms exhibit loss #f th.

1. Harrington (Np.). 2. Werrington (Np.)< O.E. Witiringtun, A.S. Chr. 3. Surrey < O.E. Suprige, A.S. Chr.

D.B. ^.riwfone1, fF&iertfltofM1, Sudries. 12th c. — Np.S. Hetherington1. 13th c. — O.B. Hetherinton1 (27—31). Ch.B. Hetherinton1 I. F.A. #e-

therington\ Witherington* (1284). Index to Sloane MSS. (ed.

J. L. SCOTT, 1904). Hetheringtone1 (1276). B.B. Surreia*. T.N.

Hetherintone1, Surr'3. 14th c. — F.A. Hetherington1 (1316). Index. Hetryngton1 (1342). P.B.

Haryngton1 (77—81).

15th c. — Capgrave. Surry*, p. 160, Sotheray*, p. 271. 16th c. — Index. Hetherington1 (1541).

F. In other place-names containing O.E. th, p.

1. The present forms exhibit th.

1. Witham (Es.). The name of this place does not contain original p. The forms Witanham, Witham, A.S. Chr., point to O.E. wita -\-ham-, cf. Mdf., p. 152. No doubt th in such a form as Witham has been taken by copyists or scribes not so well acquainted with the locality to represent the sound (d). Hence the spellings with d, or loss of th. Such confusions be- tween t-\-h and p are to be found also in A.S. charters. Cf. Taticre = Tathere, see Mdf., p. 133, and Waltiam, Kemble 988 = Waltham, ib. 844. 2. Rothley (Le.). 3. Knaith (Li.). 4. Rothwell (Li.). 5. Witham, N. and 8., also the name of a river (Li.); as to the etymology, cf. Witham (Es.) above. According to Hope, the local pron. is »Wittn», which presupposes original t. 6. Seething (Nf.). 7. Rothwell (Np.). 8. Rothley (Np.). 9. Blyth (Nt.). 1O. Bath (Sm.) < O.E. Badum, Batie, A.S. Chr. 1 1. Routh (Yo.).

D.B. Witham\ Rodolei*, Cheneide*, Rodouuelle*, Rodewelle4, 'Wirne*, Widme*, Witham*, Sifhinga*, Sithinges*, Rodewelle1, Rute11. Geld. Inq. RotheweW.

12th c. — Fr.Oh. Blia», Blyag (c. 1175). Index. Roelay*, Rothewelle* (1156), Rodewelle1 (1150-55), Widme* (c. 1170), Wieme* (Hen. II.), Wiema* (1183—84), Normme*, Nortwidhem*> Nordwidma*, Widheme*. Le.S. Rodeleia*. Li.S. Rodewella*. Np.S. Rowell1.

Anglo-Norman influence on English place-names 89

13th c. - C.B. Wyham1, Wytham1 (72-79), Roelay*, Roleg* (04—27), Wit- ham5 (88-96), Senges* (ib.), Sithinge* (79—88), RoirelF (72—79, 88—96). Ch.B. Witham1, Wyham1 I, Eole\ Roeleg*, RoeP I, Rothelee* (1284), Cneye*, Kneye* I, Senges6 (1267), Sitting9 (1285), Rolce*, Rotheley* II. Gl. Bees. (Geoffrey de) Baw (1230). H.B. Wyham1, Wytham1, Rothele\ Rotheleye2, Kney3, RoutheweW, Wit- ham5, Wime*, Wytme*, Seynges", Rowell\ Index. Rothleya2, Roleya* Rothele*, Wydma\ Witme*, Wyme\ Senges6. Le. Bees. Roleya* (1251). B.B. Rolega*, RouweW. T.N. Rottel\ Role\ Rolle\ Rou-> RoweW, Rowell\ Rotheley*. Yo. K.Q. Rue11 (1284).

14th c. — P.A: Wyham1 (1303), Kneth* (1316), Rothewell*, Nortwyme*, Wy- ham*, Wythme\ Senges" (1302), Seng6 (1316), Sithyngg6 (1346), RothewelF (1316, 46), Rotheley* (1346), Slid9, Blida9. Index. Wyma5 (1343), Sithyng6, Sythinge, Rothewelle1 (1335). P.B. Rothe- wcW (1399—01).

15th c. — P.A. Wytham1 (1428), Kneth* (1428), RotheweW (1428), Sythinge6 (1428), Rothewell7 (1428). Index. Sethyng6 (1450). P.B. T^am1 (61-67, 67— 77), Rotheleif (61— 67), Rothewell* (ib.), (ib.), Sethyng* (67—77), Rothewell\ (61-67, 67—77).

2. T/ie present forms exhibit loss of th.

1. Wingrave (Bu.). 2. Wing (Bu.). The early forms in D.B. and the Index point to original th. The etymology may be an O.E. wljmng for cf. Mdf., p. 152.

D.B. Withungrave1, Witehunge2. 12th c. — Index. Wiungua2 (Hen. II.). 13th c. — C.B. Wing2 (88— 96). H.B. Wengraue\ Weng2. Index. Wenghe*

(1208). T.N. Wengraue\ Weng\ 14th c. — F.A. W engrave1, Weyngrave1, Wenge*, Weynge*.

3. The present forms exhibit d.

1. Roothing or Roding (Es.), possibly < O.E. *Hrofingas < Hroth, Searle. 2. Edington (Wl.). According to Stevenson, the place is identical with EJmndun, A.S. Chr. and Crawford Ch. Epandun has also been identi- fied with Edington (Sm.). This identification must be wrong, for the place in question appears in D.B. as Edivinetona. In our opinion it is not altogether excluded that also Edington (Wl.) is due to an O.E. Eadivinetun. If so, it does not belong here.

D.B. Rodinges1, Roinges1, Edendone*. 1 2th c. — Index. Roing1. 1 3th c. — H.B. nothing1, Roing1, Edindone*. B.B. Roinges*, Roynges1,

Roynge1. T.N. Roing'1, Edinton*.

90 R. E. Zachrisson

1 4th c. — F.A. Rotkyng1 (1303 etc.). Index. Rotting1 (1389). P.R. Edyn-

rZon2 (1399-01). 1 5th c. — F.A. Rothyng1 (1428). Index. Edyngdon* (1432). P.B. Edyn-

don* (61-67).

In early O.Fr. d < Lat. t, d is lost, finally after a vowel, medially between two vowels and before liquids. It has been proved beyond all doubt that, before d-was lost, itVas changed into (3), (J)), which sounds were probably distributed so, that (ct) occur- red between two vowels and before the liquids, whereas there was in the spoken language an interchange between final (d) and ({>). Cf. MEYER-LUBKE, Gramm. § 436; NYROP, Gramrn. §§ 386,387; MENGER, p. 94 ff.; GASTON PARIS, Extraits de la Chans, de Ro- land (Paris 1899), p. 15 f. The existence of such an interme- diate sound is best proved by the fact that in O.E. and M.E. loan-words Lat. and O.Fr. d, respectively, appear as J5, ct. For instances see POGATSCHER, Lehnworte, p. 174, BEHRENS, p. 175 f., and especially KLUGE, Zeitschrift fur Rom. Phil. XX, p. 322 ff., where the whole matter is discussed at length, and also numerous references given 1. Cf. also REICHMANN, Eigennamen im Orrmulum, p. 28 f., who considers p in several Biblical names from this text as a criterion of French influence. The spirant is still preserved in Mod. E. faith < M.E. feif, feicf, Jei£h(e) < O.Fr. feid, feit < Lat. jidem (see N.E.D.), asseth < M.E. aseeth, assethe < O.Fr. as(s)et < Lat. ad satis (see N.E.D.), as probably also in Scotch dainteth -ith < M.E. deinteth(e\ -ith(e) < O.Fr. dain-, deintiet < Lat. dignitatem, bounteth -ith < O.Fr. bontet < Lat. bonitatatem, poortith < 0 Fr. povretet < Lat. paupertatem.

In O.E. the interdental spirant was voiced between two vowels, finally after a vowel it was voiceless. Cf. KLUGE, Paul's Gr., p. 1006f.

As for the various ways in which the sounds of ([)) and (ct) are rendered in early E. and A.N. MSS., cf. Appendix A.

In the beginning of the 12th cent, the Normans had, to judge from the orthography of the earliest A.N. MSS., two ways of pronouncing such words as vide < Lat. vita. Sometimes

2 To the instances of A.N. (|>), (d) having been kept in La3amon which are adduced by LUHMANN, pp. 191, 193, we may add one more, viz. Ucert Escud, A, v. 2769, corresponding to Vertescu, B — Mod. Fr. £cu < Lat. scutum.

Anglo Norman influence on English place-names 91

d was kept in the pronunciation, sometimes not. Of. KLUGE, Zeitschrift, p. 322, and NYROP, Gramrn. § 386, who draws atten- tion to the fact that in Mod. Spanish there is a similar vacilla- tion with reference to the pron. of the interdental spirant in an intervocalic position. The spirant seems to have been lost earlier when medial than when final. Already in Computus there are rhymes like signifie: vie (cf. MEYER-LUBKE, I.e.), and soon after the middle of the 12th c. d, t disappear in all positions.

In the French word-material of D.B. intervocalic d = (ct) is often lost medially, while in a final position it is always kept. Instances like Todeny, Toeny = Tosny (Eure), Buenvaslet, -vasleth etc. are adduced by HILDEBRAND, p. 360 f.

Also in the English place-names loss of intervocalic (d) is usual, not only in D,B. but also in records of a more recent date: -mude, -mue, (see list A); Suthinton, Suintone (see list B 2); Widerne, Wiernia (C 1); -hyda, -heda, -hetha, -hea, -hee, -Tithe, -rie, -rede, -reihe, -rede, -ree (D 1); Blida, Blia, Rode-, Roihe-, Boe-, Elite, Bue, Sethyng, Seynges, Wytham, Wyham (F 1); Rodinges Boinges, (F 3).

Loss of the interdental spirant before liquids appears in such instances as Hallege, Haretom, Arintone etc. (list E 1, 2). Here it is worthy of note that the combination per is often tur- ned into pre, whereupon pre > re.

Finally we find in D.B. an interchange of t, d just as in the purely French words: -rid, (D 1), -hed (D 2), -ret (D 4). Loss of the spirant we have not noted as early as in D.B. save in one single case, viz. TocJcvi = Tockwith (Yo.). This is probably a mere error in view of the many other cases in which the den- tal was kept. In later records loss is usual also in final posi- tion: Waymo, C.R., Cokermu, C.R., (A); Lamhe, R.B. (D 1); Chaure, Mdre, Ch.R. (Dl); Ba, Gloucester Records (F 1) etc, Cf. also rhymes, like Jesu: ju: Adeldru, Gaimar, v. 1341, v. 1407, Edelfrid: saisi, ib., v. 1147.

When owing to A.N. sound-development intervocalic (d) is dropped, two vowels meet, a phenomenon which is generally known under the name of hiatus. During all periods the French language has never been very partial to such a com- bination, and several devices have been used to get rid of it:

92 &• E. Zachrisson

(1). Sometimes the second vowel is entirely lost. This seems to have been the case in Rolega, R.B. < Roelay < Rothe- ley (Fl); Wime, D.B. < Wieme < Witham (Wime may also be due to Widme) (F 1); Wing, C.R. < Wiungua < Withung (F2); and perhaps also in Seng, F.A. < Seijnges < Sethyng (F 1), cf. however below.

(2). The two vowels are amalgamated into one sound. A diphthong may really have been pronounced in Soynthone, Sel- borne Ch. < Suthintune (B 1) and Boing, Index < Rothing (F 3).

(3). Both in O.Fr. and Mod. Fr. a glide is often developed between the two vowels. This glide before a following palatal vowel is (j), and is either not noted in the spelling or rendered by i, y. Cf. NYROP, Grarnm. § 279; PASSY, Changements phone'- tiques, p. 214. STIMMING, p. 237, adduces several A.'N. instances of i, y, as a glide between unstressed a, o and a following pala- tal vowel. In the case of the English place-names j as a glide appears with certainty only after what from a Norman point of view was no doubt stressed e, whether this e (< y, &, e) already existed in English or was due to A.N. adaptation. On this point cf. STOLZE, p. 18 Anm. 2. The following examples may serve to illustrate the development in question: Scepereie, LEI. < Sce- peree, Lachingeia, Index < lachingehea < Lackingehetha (D 1); Saltreye, Ramsey Ch. < Saltrede, Stebeheie, Fr.Ch.< Stibenhede (D 2); Codreye, R.B. < Codrede (D3). The above-mentioned forms are perfect analogies to A.N. espee, espeie < Lat. spatha (see STIM- MING, p. 175, p. 238) or M.E. contreie (< contree): waye\ see BEHRENS, p. 82.

The glide is still kept in Mod. E. fay < M.E. fay < O.Fr. foe, faie, fayee, fed, fee (Godefroy) < Lat. fata (N.E.D.). An excellent parallel to the twofold development of -e]), -ecte in the second element of English place-names is also afforded by such O.Fr. doublets as deintie, dente - - daintiet, bunte, bountee — bon- tet corresponding to M.E. deinteth(e) - - daynte, Mod. E. dainteth dainty, bounteth bounty. Cf. also O.Fr. feid, feit - - fei = M.E. jeiti — fei, Jey = Mod. E. faith, fay.

A few peculiar spellings of the diphthong ei (ai) are discussed in Appendix B.

The instances of loss of ({)), (it) we have just been discus- sing can hardly be explained in any other way than as due to the

Anglo-Norman influence on English place-names 93

operation of A.N. sound-laws. In O.E. and M.E. the interdental spirant is hardly ever lost in position between two vowels or finally after a vowel. Also in the Mod. E. dialects loss of ({)), (ct) is rarely found. Of. E.D. Gr. § 314; GRONING, p. 47. What also tells in favour of French influence is the circumstance that the forms without spirant are best and most frequently repre- sented in documents which in every other respect exhibit strong A.N. peculiarities (D.BV Fr.Ch., R.B., I.EL, T.N. etc.). Exam- ples from O.Fr. and A.N. texts are given by WESTPHAL, p. 20 and p. 37.

There are, however, some instances in our lists for which A N. influence can hardly be claimed. Before I and r (see list E) loss of th as early as in D.B. is most likely due to French influence. Modern forms like Surrey < Sutirige etc. (see E 2) have been explained as due to assimilation in English itself (cf. KLUGE, Paul's Gr., p. 1007), though in our opinion French sound-development is not altogether excluded 3. We must keep in mind that assimilatory loss of th sets in mainly before /, s, d, t, more seldom before other consonants.

The French forms without th existed for a long time by the side of the native ones in which th was kept. In the 13th cent, the former were still very well represented. Even in the 14th cent, they are not unusual, but a hundred years later they seenT to have altogether disappeared except in the few cases where they ultimately prevailed.

The French forms are, of course, much more numerous in the Anglo-Latin feudal records than in the purely English chro- nicles. A comparison between the chronicle of Peter Langtoft and that of Rob. of Brunne, of which the latter is to a great extent based on the former, shows that Langtoft. always prefers the forms in -mue < O.E. mupa, whereas R. of Brunne, with very few exceptions, makes use of the native forms : Portesm ue, Lang- toft I, p. 450; Portesmouth, R.Br., p. 96 etc. The form -mue is very rare in R.Br. One instance is Germue, Gernemeue, -mue, p. 252, ansvering to Gernemue, -meue, Langtoft II, p. 196.

1 In all the O.Fr. texts we have had an opportunity of examining Surrey always appears without the dental. Thus Surree, Ron, v. 7736; Surr(i)e, G. le Mar., v. 1552; Surye, Langtoft I, p. 170.

94 H. E. Zachrisson

We will now proceed to give a short survey of the cases iu which the French forms have gained the victory.

That, as far as compounds with Sup-, •muf(a) are concerned (see lists A, B), the French forms have never prevailed, is, of course, due to the fact that the etymology of the words was quite clear. South and mouth were parts of the general vocabu- lary, and supported the native forms.

In Wy ville (see list C 2) it is not only the spirant that has been lost owing to A.N. sound-development, but in the suffix French ville has supplanted older well. When ville appears as a suffix in English place-names it is mostly due to substitution of the French etymon ville for English well or Southern vel(d) < feld. Acoustic resemblance is probably often the cause of the substitution.

It is above all in place-names containing O.E. rip, hyp, h&p etc. that the French forms have been victorious (see list D 2). This is perhaps best explained thus: the forms in -ee, -eie did not appear strange or unfamiliar to the English. They were easily associated with a similar and very usual suffix, -e, -eie < O.E. ea, eg, Mclf. p. 44, 45. To the instances quoted in list D 2 we may add one more, in which the forms without tin, also have prevailed. In the English Sir Beves of Hamtoun (ed. KOLBING, E.E.T.S., E.S. 46, 48, 65) we note the following interesting forms of Putney (Sr.), nowadays the name of a London district south of the Thames: Pofenhipe: swipe, p. 202, MS. A (c. 1327); Poun- tany, ibid., MS. S (end of 14 c.); Ponteney: they, ibid., MS. C; Poivnteneth, ibid., MS. M (loth cent.); Putneth, ibid., 0 (printed copy, closely related to M); Pountenay: lay: way: tway (= two), p. 208, 210, 214, MS. E. (sec. half of 14th cent.) l.

1 As late as the 17th cent, double forms still occur for some of the place-names here in question. The same person is called R. Stepneth or Stepney in an entry of 1600; see BARDSLEY'S Die., p. 716. This being so, it would really be curious if we did not sometimes find -eth for etymological •ei on the analogy of those cases where the interchange was original. I think I have found a trace of such analogous forms in Magna Britannia, where Brent Eleigh and Monk Eleigh (Sf.) are called Brent Ely or Brentil- leth alias combusta, Monks Ely or Munselleth. The etymology is O.E. Illan leah (leage] (a. 999), Earle. M.E. forms are Illeya, Ille.ye, H.R., Illeg, T.N. Cf. also Appendix B.

Anglo-Norman influence on English place-names 95

According to Tanger, Rothwell as a place-name is pro- nounced with (|D), as a personal name, without. Hope gives the notation »Rowell» for Rothwell (Np.). Both the latter pronun- ciations may depend on A.N. forms without th (see list F 1).

A.N. (t), (d) is substituted for M.E. final ([>), and intervocalic (d).

Though in the course of the 12th cent, the interdental spirant was lost in A.N. in all positions, we, nevertheless, find spellings t for (p) finally, and d for (d) medially. To assign the value of (ID), (d) to these spellings is not possible, for some of them, particularly those with d, have been handed down to our time, and d is here still pronounced (d). In our opinion such instances of t and d are now due to A.N. sound-substitution, and should, consequently, have been treated in Part I. To avoid unnecessary repetitions of forms and to give a better survey of this whole question, we have considered it the best plan to deal with the eventual sound- substitution of (t) for final (|D) and (d) for intervocalic (d) in this context. In the sequel we will, in the first place, treat of some cases of (t) for ([>), then proceed to (d) for (d).

1. A.N. (t) is substituted for M.E. final (|D).

That (t) was used by the Normans as a substitute for ({>), not only initially but also finally, is indisputable. Rhymes of t: th < p are usual in Gaimar etc.; cf. VISING, Le dial. Anglo- Normand, p. 90, and RATHMANN, p. 39.

The only case of t being preserved which may belong here is Orsett (Es.). Forms like Orseda, D.B., Horseth, C.R. (see list D 3) undoubtedly point to an O.E. *HorshcBp !.

We note, however, in many purely English works of the llth cent, and later a peculiar orthographic interchange of final ^, th and t, d:

Rule of S. Benet (MS. 1020—30), t for f, as in det (= doth); d for #, as in wurdmend, hwced (= what), see LOGEMAN, p. LIT.

1 As regards Ayot < O.E. egofi, ef. the explanation given by SKEAT, Pl.-Ns. of Cb., p. 58.

96 &. E. Zachrisson

La3amon, / for p, in A 6 X, in B often; ft for t, a few times both in A and B, see LUCHT, p. 65, 72. On page 65 Lucht also draws attention to similar spellings from several llth and 12th cent. MSS., e.g. ^Elfric, Textus Roffensis, M.E. version of the Rule of S. Beuet etc.

Norfolk Guilds (14th c.), t for th is usual, see SCHULTZ (Diss. Jena 1891), p. 28 f.

London deeds (1384—1430), t for J5, th, as in hat, levet, strecchet; d for th, as in had, see MORSBACH, Urkunden, p. 134, p. 137.

Many more examples are noted by DIBELITJS, Anglia XXIII, p. 452 f., and NEUMANN, Paston Letters, p. 91.

This interchange is generally confined to unstressed sylla- bles, particularly the verbal ending -eth, and in consideration of the very usual inverted spellings (at, f for t) it may represent a real transition of ([)) > (t), which perhaps has taken place in con- nected speech before consonants like ,9, f, t, d etc. Having thus arisen in some positions, the dental stop might easily have been transferred to others, and have even sometimes transplant- ed original th. A similar explanation would seem possible for some sporadic instances of t for final th also in the present English dialects. Cf. E.D. Gr. § 316; FRANZMEYER, p. 75; KRUIS- INGA § 362 and Additions p. 178.

As to d for final th, p I do not think we need assign the value of (d) to this spelling in the cases here in question. Thus MORSBACH, Urkunden, p. 134. It seems more likely that d is nothing but an inverted spelling for t. In fact, in nearly all those texts where we find d for final th we also find some in- stances of t for d in the same position. Cf. LOGEMAN, Rule of S. Benet, p. LI; LUCHT, La3amon, p. 68 1; MORSBACH, Urkun- deu, p. 144, fulfyllyt by the side of fulfyllyd. Now, if both the ending of the past tense -ed and that of the present tense -eth were changed to -et, one might sometimes be put for the other. For such instances see BLACH, Paulsschule, p. 41. DIBE- LIUS, p. 452, notes rhymes like longith (present tense): hongit (past part.).

From the above examination we may conclude that, though A.N. influence in itself is not impossible with reference to

Anglo-Norman influence on English place-names 97

Orsett < Horsheth, t in this and similar cases can also be explained as due to dialectal sound-change.

2. A.N. (d) is substituted for M.E. intervocalic (ct).

Intervocalic d for th is much more usual than final t for ih, and such forms have also been adopted much more generally. In his treatise on the Germanic elements in French and Pro- venc,al, p. 168, MACKEL, with reference to the different treatment of (ct), distinguishes between early loan-words in which the spi- rant was lost, and comparatively late ones in which it, probably owing to sound- substitution, appears as d. After observing that Gaimar generally uses d for O.E. (d) in personal names, RATH- MANN, p. 42, no doubt correctly, assigns the value of (d) to this spelling. Numerous instances of d for etymological ct, and vice- versa, in both the MSS. of La3amon are looked upon by LUHMANN, p. 38, as due to French misrepresentation and, no doubt, also mispronunciation. We have then to assume that from about the middle of the 12th cent, the Normans commenced to substitute (d) for (ct) in English place-names.

The following typical 13th cent, instances have been chosen from the lists to illustrate the sound-substitution: Tinemuda, Ch.R. (A); Sudwerk, Index, Sudington, Selborne Ch. (Bl); Wydecunib, H.R., Wydinton, F.A., Widebroc, Index (C 1); Widiford, Tax. Eccl. (03); Meldred, Ch.R., Alreheda, Ramsey Ch. (D 1); Bulchude, T.N., Bulhud, H.R. (D2); Heanred, Ch.R., Ledrede, Index, (D 4) etc. O.E. Wiftigslad (a. 739), Crawford Charters, appears as Wydeslade in a late copy of the 15th cent. O.E. Bade, A.S. Chr., is called Ba, Bae in Langtoft I, p. 34, and Roman de Rou, v. 7750, with loss of (cT), Bade in Wace, with d for (ct).

Though we may be fairly sure that comparatively early in- stances of d for th in English place-names are due to A.N. sound- substitution, we are not therefore always justified in considering d in the present forms of the place-names in question as an A.N. peculiarity. The transition of (ct) to (d), particularly before Z, and when -er follows, is well evidenced in M.E., and has also taken place in several words belonging to the literary language. Of. KLUGE, Paul's Gr., p. 1008. d m Headley (Wo.) < O.E. HaWege, Birch 455, and Widley (Ha.) < O.E. Withiglea, Birch

7

98 R. E. Zachrisson

142, is, no doubt, due to this sound-change. In the M.E. texts mentioned in § 2, p. 43 f., where initial p often appears as dr intervocalic d for f is also quite usual. Promptorium Parvulorum has schede (= sheath), see Stratmann-Bradley. In our opinion the above-mentioned spellings with d for f, th may very well indicate a dialectal pronunciation, of which now only some traces are left. As a matter of fact, in the present English dialects intervocalic (d) for (d) occurs, independently of position before one consonant or the other, both in N. and S. England. Cf. E.D. Gr. § 315; FRANZMEYEK, p. 64; HIRST, Dialect of Kendal (Anglist. Forschungen 16), p. 89. In his edition of Pedes Fi- uium for Kent (Publ. of Kent Archseol. Soc. I, p. 230) the Rev. L. B. LARKING tells us that South Ash is in the vernacular pronounced »Soudaisse».

But the great question is: when and where did this change of (d) > (d) take place? Here again, we find ourselves in face of a problem which is not to be solved easily *. Only by a thorough special investigation could it be decided if d in such modern forms as given in the lists B 2, 03, D 4, F 3 is due to A.N. sound-substitution or dialectal sound-development.

In one or two cases some evidence may be adduced which tells in favour of French influence. Rodiug is, according to Hope, dialectally pronounced »Roothing», and from some th spellings of the 17th and 18th cent. (Index Villarum, Camden) we may perhaps conclude that this is or was the case also with Wide- combe (Dv.).

Some very interesting examples of the double A.N. treat- ment of intervocalic (5) are offered by the various modern and M.E. forms of O.E. friping as tabulated below.

1. Tring (Ht.), SKEAT, Pl.-Ns. of Ht., p. 38, is unable to arrive at any certain etymology. Curiously enough, he has overlooked the important D.B. form Tredunga, which makes the derivation from O.E. griping obvious.

1 JELLINGHAUS, Anglia XX, p. 258 Anm., thinks that d already in D.B. stands for (d), a dialectal development of (d). LUCHT, Lajamon, p. 97, and LANGER, Abingdon Chart., p. 66, are of opinion that the occurrence of r) as an inverted spelling for d in several 12th and 13th cent. MSS. necessarily must indicate a transition of (d)>(d), We cannot subscribe to this view. Both in O.E. and e. M.E. d is often used for J5, d, merely owing to confusion of letters. Cf. Appendix A.

Anglo-Norman influence on English place-names 99

2. Division of land in Line. We have not been able to find any more entries after the year 1401. 3. Riding, division of land in Yorkshire.

D.B. Tredunga1, Treunga1, Trevinga1, Norttreding* , Sudtreding* , Westredinge* , Nortreding5 vel Norttreding*, Estreding*, Westreding*.

1 2th c. — Li.S. West Triding*, Nortriding*. Pontefract Ch. de Tridings3 et de Wapentaciis (c. 1150).

13th c. — H.R. Norththrything*, Northtryyng*, Norttrihing*, Norttrifhyng'\ Sud Thrything*, Sudthrything* , Suththreheng* , Suthtreing'*, Suth- tring'z} Suttring'\ Sutt'ying*, Westthreylg*, Westreingge*, Westre- thing\ Westrithlg*, Westthrithing*, Westtrying*, Northtithing\ Northriding*, Estriding*, Estridding*. R.B. Traynge*. T.N. Treing1, Treng1.

14th c. — F.A. Trehynge1 (1303), Northrithing* (1316), Suthrithing* (1316), Westrithing^ (1316). Yo. Recs. (see SKAIFE, Survey of Yo. etc.r Appendix). North striding5, Nortridinge*, Westriding* (c. 1300).

15th c. — F.A. Northrithing* (1401), Southrithing (1401), Westrithing* (1401). Index. Trynge1 (1441).

On the evidence of the D.B. forms STEENSTRTJP has derived Mod. E. Riding (Yo.) from O.E. frifing < Scand. priffjungr, in spite of the phonetic difficulties involved by such a derivation. There was, however, the same division of land in Line., and in entries relat- ing to this county from the H.R. and the F.A. we find the true English form thrithing, which hitherto has escaped the attention of philologists.

Forms like Estriding, Northriding, H.R. etc., corresponding to modern Riding, in all probability, exhibit French d for Eng- lish (3) ^

Tring (Ht.) is an instance of the early French forms with- out spirant having prevailed. As to the intermediate stages in the development Thrithing > Tring there are several possibilities to be taken into consideration, Triing2 may have been contract-

1 Cf. BJOEKMAN, Archiv 116, p. 105, who also points to the lengthening of i as a criterion of French influence. This may be true, though, on the other hand, (ai) for (i) in the present pron. may be due to spell.-pron. In like manner (ai) in Mod. E. Clive < clif and Ticehurst < ticcen (see § 1, p. 31) may be due either to spell.-pron. or French vowel-lengthening.

2 Tr instead of Thr may be due to A.N. sound-substitution or dialectal development. Cf. § 2, p. 44 f. In the personal name Thring Hi has been kept to the present day. Cf. BAEDSLEY, Die., p. 747, where numerous early forms are recorded.

100 R. E. Zachrisson

ed to Tring, whereupon, owing to a general M.E. tendency, i was shortened before ng. On this point cf. MORSBACH, Mittel- engl. Grainm., p. 69, p. 73. If we start from early forms like Trayinge, Treinge < Tredinge (with A.N. e for E. i, cf. STOLZE, p. 17), ei may here have been made into a diphthong ei, ai, whence by monophthong! zation e, which became i before ng. On the other hand ei may also be a Frenoh spelling for e, which in late M.E. often interchanged with i. Cf. MORSBACH, Mittelengl. Gramrn., p. 150.

The Lincolnshire forms in the H.R. are of particular interest, for here we find all three forms as well as many curious blendings *.

G. In place-names containing O.E. cepel.

Nearly all place-names here in question are compounded with personal names in cepel. The greater part of of this sec- tion must, therefore, be devoted to the study of such personal names. We admit that in the following discussion a few mat- ters have been entered upon which are not absolutely necessary in order to obtain a right understanding of the various forms of

1 Instances of h for th, d appear in entries like Norttrihing, H.R., Trehynge F.A. Such forms are to be compared with Portesmuha C.R. (see list A.)

In some early M.E. texts several cases of initial and final h for J) have been noted, which, no doubt, are best explained as mere confusions of letters. Thus SCHROEE, Regula S. Bened, p. 18; BERBEKICH, Herb. Apu- leii, p. 8. Of different opinion are LUICK, Deutsche Lit. Zeitung 1890, p. 746, and LUHMANN, Lasamon, p. 46, who think that h in such spellings represents a development of p. Their reasons do not seem to us very convincing.

In Trehynge, Portesmuha etc. h has certainly no phonetic value; it merely stands to indicate that e and i, u and a belong to two separate syl- lables. Both in O.Fr. and Mod. Fr. we often find h used as a mere hiatus- filler. Cf. NYBOP, Gramm. § 279 Remarque; MACKEL, p. 133.

Initial h is often omitted in D.B., and here, as well as in later records, we also find instances of an unetymological h both initially and medially. Cf. spellings like Cornehude, Bradehode, D.B. = Cornwood, Bradwood (Dv.) — in D.B. ode, ude are usual spellings for O.E. wudu — Suavesia, Sivavesehe, Fr.Ch. = Swavesey (Cb.); Clixeby, Clixheby, H.R. = Clixeby (Li.); Cameshing, R.B. = Kemsing (Ke.) < O.E. Kymesinc, Index; Messingham, Meshingham, H.R. = Messingham (Li.) etc.

Anglo-Norman influence on English place-names 101

the place-names, but, as the entire question exhibits many points of special interest, we have * thought it admissible to treat it somewhat more exhaustively than would have been required solely from the point of view of the place-names.

1. cefel in personal names.

Two questions, in particular will have to be closely inquired into: (1). The transition of cefel to cegel. (2). The confusion of cefel with celf, eald, ealh.

a. Transition of cefel > cegel.

This transition has already been dealt with by Prof. MOES- BACH in Festschrift fur Wendelin Fcerster (= Die angebliche Origiualitat des Fruhmittelenglischen King Horn, Sonderabzug, Halle 1902). The chief aim of the author is to prove that, though English as to its contents, the original version af the M.E. story of King Horn was the work of one, or more than one, French, probably Anglo-French, writer (p. 299). In his opinion the French origin is clearly indicated by the fact that the metre requires such forms as Ailmar etc. pronounced in two syllables < original ^Ethelmcer etc. in three syllables (p. 322). To account for the transition of cefel > cegel MORSBACH assumes French influence (p. 305) *. For examples he refers to Searle's Onomasticon

1 Indeed, this seems to be the only possible explanation. The theory according to which O.E. cegel, eg el (< cege, cf. MULLEK, Palaestra, IX, p. 105) has taken the place of cepel is highly improbable, the latter being a much more usual compound in O.E. personal names than the former. Cf. the examples in Searle. - - In Viet. Hist, of Np.,' p. 297, Mr. ROUND, a well-known English scholar, accounts for cegel < cepel as due to confusion of letters. Had we to do only with a few sporadic spellings, this explanation would in itself be possible. The O.E. symbols p, #, 3 were unknown to the Normans, and for that reason are often put one for the other. Instances are adduced by SKEAT, The Proverbs of Alfred (Oxford 1907), p. XIV; SCHROEK, Winteney-Vers. der Regula S. Bened. p. XXII; LUHMANN, La3amon, pp. 44, 45. In all their instances 3 occurs as an occasional spelling for J), which has hardly given rise to any new pronunciation, whereas cegel for cefel is extremely usual, and has been kept to the present day. Consequently, the explanation of cegel as due to mere confusion of letters does not seem to be admissible.

102 R. E. Zachrisson

(pp. 5, 33 ff., 534 ff.), where we are told »that cegel, is a late form of (ethcl, which first appears on coins of ^Ethelred II. (978— 1017). »

It seems, indeed, surprising that French should have exercised any influence on English names as early as 1000, and, as far as we can see, Searle has not adduced any certain examples of agel for (Bpel of so early a date. Perhaps he has overlooked that early cegel-toTms on coins etc. may be due to the above-mentioned O.E. cegel, egel. If we except the charters, which often exist only in late copies, Searle's instances of cegel for cefel are mainly derived from Anglo-Latin chronicles, which exist in MSS. of the 12th cent, and later, e.g. Henry of Huntingdon, Will, of Malmesbury, Flo- rence of Worcester etc.

With a view to establishing the chronology of the transition, we have scanned the various forms of personal names com- pounded with cepel in the MSS. of the A.S. Chr. The sec. part of Eafle and Plummer's edition contains an excellent Index of Names. In the Introd. to the same part the date of the MSS. is noted. It appears that in the MS. A (entries nearly contem- porary with events, latest entries in 1075) there are no instances of cepel > cegel. The same is the case with B (c. 1000), C (c. 1050), as also with E (1121—1154). D (second half of lithe.) has one single example : j&gelmer = ^Efelmcer E. First in F (pro- bably contemporary with D.B., there being, a great resemblance between the handwriting of the two), the transition is very well instanced : JEctelberht, ^Egclberht, Egelbert = 2Eftelbrylit A ; j?Egel- Eftelno$, -nod C, E; JEgelred = JE&elred A, C, D, E; = JEpelric C, E; ^Egelword = JEtielward E. - - We have not been able to find any instances of interchange in Sweet's O.E.T. Thus, as cegel for cefel does not seem to appear until after the Conquest, this seems to support the theory that the change in question is due to A.N. influence.

In the A.S. Chr. we also find cefe, cep as alternative forms of ccfel: rfftebald (a. 778) E = JStfellald D; JEfrelm (a. 837) B = ^Epdhelm A, E; JEpered (a. 675) A = JEfelred A; JEfered C = JEfelred C; JEfrered a (contemporary with F) = ^]felred F; JEdestan C = ^Efelstan A, E ; ^Efestan B = ^elstan A , E ; Apewold E = ^Ef el- wold A; AfulfA, B, C, D, ActewulfF = JEfclwulj A, ^delwulfE.

The loss of I in cepel is explained phonetically by some authors. BCLBRLNG § 552 a, accounts for Aecterred < ^Eftilred

Anglo-Norman influence on English place-names 103

as due to assimilation of Ir to rr. In the Crawford Ch., ed. NAPIER-STEVENSON, p. 99, we find the remark that I seems to be frequently omitted in the name of ^Epelstan after the beginning of the llth cent. The forms above indicate, however, that cepe for cefiel is not confined to any special position or any special time. We feel inclined to look upon ape, &£ as shortened (pet- name) forms of cefiel. Or*. &]) < cepel in the Liber VitaB, MULLER (Palaestra IX), p. 99; ad < adal, mad < madal, STARK, Die Kose- namen der Germanen (Wien 1868) pp. 40, 421.

Be that as it may, this much is certain, that in records of the 12th cent. 'both cepel and cefie often appear as cegel, cege, ail, ai. In Morsbach's opinion this change is due to French sound- substitution: »Das fur den Franzosen unaussprechbare inter- vokalische (stimmhafte) f fallt entweder aus, oder wird durch die stimmhafte Spirans j ersetzt» (p. 305). For various reasons this explanation can hardly be the correct one. We have already seen that Norman (ct) < Lat. t, d was still kept at the time of the Conquest, and, moreover, j does not seem to be any likely substitute for (d). We are of opinion that, after the loss of ^, ceel < (epel was turned into cpgel, ceil,, owing to a desire to avoid hiatus.

We will try and support this theory by adducing some parallel instances. In the first place, we may draw attention to Mod. Fr. »aimant», Engl. »adamant». The etymology of this word is Lat. adamantem, which, with loss of d and transition of •ae to ai(e), appears in O.Fr. as aymant, aiemans, aemans (Gode- froy). The O.E. form adamans, Cura Pastoralis (see N.E.D.), of which athamant, Chaucer, may be a continuation was borrowed at a time when the spirant was still preserved. Mod. English »adamant» corresponding to adamant, Chaucer, seems to owe its d to learned influence from the Latin word. M.E. aimont, Ayen- bite, is, as regards its form, identical with O.Fr. aymant. < — Another parallel is offered by O.Fr. Aimer < Ademarus, Adamar, containing a Germanic root *ada or *hactu. Cf. A. LONGNON, Polyptyque de 1'abbaye de Saint-Germain (Paris 1895) vol. I, p. 276; MACKEL, p. 15; WALTEMATH, pp. 12, 38. — Lastly we may

1 Corresponding to cepe < cepel we also find cege < cegel in the A.S. €hr.: uEgebertus F = jEgelbryht A. In the Liber Vitse we have noted oid < oidil (= epel): oiduald == oidiluuald, Sweet, O.E.T., p. 145.

104 R. E. Zachrisson

quote from D.B. itself Oilard, Oidelard, probably correspond- ing to O.H.G. Othalhart, see HILDEBRAND, p. 348; LONGNON, p. 355 *.

From what has been said above it appears that at the time of the Conquest there must have existed various pronunciations of ce^el, cepe: One with the dental retained, which was used both by Englishmen and Normans, another with ioss of the dental owing to A.N. sound-change, and a third with cei(e) < aee like- wise in accordance with Norman habits of speech. The existence of all these pronunciations may easily be proved by spellings from D.B. and other early records 2.

1. The interdental spirant is kept (spelled th, d, cf. Appen- dix A): JEdelftete ; Adelid; Adelard; Afhelwold, Adeluuald, Edel- uuold, Adeold; Adelulf, Adulf, Edeulf; Adelric, Edelric,

1 Though it seems to be a matter beyond all doubt that cepel Z> cegel is due to French influence, some uncertainty remains as to the way in which the transition has taken place. A glide (j) may have been developed be- tween a and e. Cf. above p. 92 hee > heie etc. This theory seems to gain in probability when we consider the occurrence of such a form as aiemans < aemans. On the other hand, the opinion generally held by French scholars is that the unstressed e was changed to t, whereupon i coalesced with the pre- ceding a to a diphthong ai (adamantem > aemant > aimant ^> aimant}. Cf. NYROP, Gramm. § 275. In this case the spelling cegel in early Engl. records would signify a pronunciation cnl in one syllable. Obviously, the most handy way of expressing the sound-combination ceil was to alter cepel into cegel. - To the Normans the spelling agel for ail might have been easily suggested by the coexistence of such name-forms as Agelberga, Ailberga, Agilmundus, Ailmundus etc., agel being a more archaic spelling when contrasted with ail, the purely phonetic one. More instances are afforded by LONGNON, Po- lyptyque, p. 278; WALTEMATH, p. 12. As to the sense and origin of O.Fr. agel cf. LONGNON I.e., and ARBOIS DE JUBAINVILLE, Etudes sur la langue des Francs, Paris 1900, p. 9 f. The definite solution of the whole question is, no doubt, to be found in the name-forms of O.Fr. records of the llth an4 early 12th cent. Such records have unfortunately not been acces- sible to us.

2 Mr. H. ELLIS, the learned editor of the two last volumes of D.B., has brought together the various names occurring there in three lists: A tenants in capite in 1086, B owners of land under Edward the Confessor, C subtenants in 1086. These lists are to be found in vol. 3 of D.B. They are also printed in H. BARBER'S British Family Names (London 1903). Most forms are incorporated with the material given by Searle. We have chosen our examples from the lists B and C, it being fairly certain that persons whose names are recorded there were mostly English by birth.

Anglo-Norman influence on English place-names 105

cus etc. (< ^Ethel-, ^Ethe- flced, gyth, -heard, -weald, -wulf, -ric, Searle). To render O.E. #, . a sound between e and a, the Nor- mans used their own symbols e, a, as well as O.E. CB. In D.B. the value of all three spellings, is, no doubt, (se). Cf. STOLZE, p. 13.

2. The interdental spirant is lost. There are no certain examples in D.B. Unequivocal instances are offered, however, by spellings like JEelman, 2Eelric (< JEthelman -ric\ on coins, and ^Eelmund, Meltwine (< ^Ethelmund, -wine) from charters, Searle. Cf. also Aelwyn, Ramsey Ch. < ^Ethelwine, and Aelesme, Adeelme < ^Ethelhelm, Gaimar, v. 2411.

3. After the loss of the interd. spirant aee > cei(e) : Ailbold; Ailbriht; Egilfride, Aifride; Ailiet; jEileva, Aileua, Eileua; Ailith; Ailardus; Agelmarus, jEilmar, Eilmer, Aimar; Agelricus, JEH-, Ail-, Eilric; Ailsi; ^Egeluuardus, Ageluuard, Aielvert, Ailvert, Ailuuard; Ailof, Aiolf etc. (< ^Ethel-, ^Ethe- beald, -beorht, -frith, -geat, gifu, -gyth, -heard, -mcer, -ric, -sige, -weard -wulf). Though some of these instances may be due to O.E. cegel, <$ge, most of them, no doubt, represent a development of cepel, cepe.

In records of a more recent date than D.B. we note that the aefel- names so frequent in O.E. l rapidly grow less in num- ber. At the same time ail becomes more usual than afel. The old forms with d (adel) subsist, but d has, no doubt, in later records mostly the value of (d). Early instances (end of 12th and 13th cent.) of this (d) may be accounted for as due to French sound- substitution 2. In later M.E. the possibility of English sound- development is not excluded 3. Cf. above p. 97 ff.

Some instances adduced from various M.E. sources may serve to illustrate the mutual frequency of athel, adel, ail In the Index to Ramsey Ch. (12 and 13 c.) the following ortho- graphic variants for the name of one and the same person are noted: -ZEthelwine, Ealdorman of East Angles, Ethelivin, Egeluuinus,

1 In Searle they fill nearly 30 pages ; then comes celf-with 23, whereas eaM-only takes up 8, ealh- 5 pages.

2 Cf. O.Fr. Adelaide < Adalheid, Lodier < Lofihari, see MACKEL, p. 169. In Gaimar O.E. cepel nearly always appears as adal, edel, where d, no doubt, has the value of (d). Thus RATHMANN, p. 42.

3 In an English grammar of 1701 there is a warning against the dia- lectal (Somerset) pron. »Addleston> for »Athelston». Cf. FOKSTEE, Anglia XXIV, p. 116. Here > Addle- », no doubt, represents the dialectal development fl > dl.

106 R. E. Zachriseon

in, Ailwin; ^Ethelwine the Black or Swerte, Ailwine, Eylwyn, Aelivyn; ^theliiiaer bishop of Elm ham < Athelmar, jEgelmcer, Ailmar; ^Ethelsige 8th abbot of Ramsey, sEgelsin, JEgelsi, Ailsi, Ailsy, Eilsin, Aelsi.

In vols. 3 and 4 of the F.A. (1284—1431) we only find the following instances of names containing O.E. cepel: Athelard, Athelwald; Adelarde, Adelwald; Aylward, Aylmer * Aymer , Eyldrich.

The forms in adel, ail find their way also into the M.E. chronicles, where they occur side by side with the original ones. In the various MSS. of R.GL, v. 5705, the forms Afelwold, Adelwold, Adfelwold (blending) appear for one and the same person. ^Ethelbeorht of Kent is called Aylbri^t, v. 4741, Ethel- red II, Aildred, v. 6006. Note also Adelbritf, v. 5240, adel- ston, v. 5283. In R.Br., p. 46, we find Eilred. for Ethelred, cor- responding to Eylrede, Eldrede. Langtoft I, p. 344.

In course of time athel was superseded by adel and ail. Whereas in Bardsley's Dictionary of Mod. Engl. surnames seve- ral instances of adel and ail are noted, not a single example of athel is to be found. We will now give a short survey of the present forms, as recorded by Bardsley, followed by a few early references mainly from the Hundred Rolls, one of our principal sources of M.E. personal names.

1. Adel.

Adlin, Adling, BARDSLEY'S Die., p. 41, Edlin, Edling \ p. 266, Adelyne, Edeline, Atheline, Adeline, H.R. < O.E. JEthelwine (or possibly cefeling).

Adlard, p. 40, Adelard, Athelard 2, H.R. < O.E, ^Ethelheard (or possibly JEtkelweard).

2. Ail.

Aylen, Ayling, Aylin, Aylwin, p. 70, Ayline, Ailwine, H.R. < O.E. jEthehvine (or possibly Deling).

1 Edel for Adel may be due to (1) dialectal differences, cf. BULBKING § 91, MORSBACH, Mittelengl. Gramm., p. 128 if., E.D. Gr. § 23; (2) another O.E. ground-form, viz. epel = oipil, Liber Vitse.

8 Athelard, Atheline, Athelston, and Athelem < ^Ethelhelm are the only forms with th in the H.R.

Anglo-Norman influence on English place-names 107

Aylrner, Aymer, Aymar, p. 70, Ailmar, Aylmer, Eylmer, H.R. < O.E. jEtM-, uflthemcer.

Ayliffe, Aylieff, p. 70, Ayljeva (c. 1300), Gl. Recs. < O.E. Mthelgifu or jEthdwulf. This latter is certainly the source of Eylof (c. 1300), Gl. Recs., Ayloffe (1500, 1698, 1702), see BARDSLEY.

Aylett, Ayllett, Ayllot, p. 70, Ayllyth, Ayllyht, Aylet, Aylede, Ailot, H.R., no doubt < O.E. jEthelgyih and jEthdnoih 1.

Aylward, Aylard, p. 70, Aylward, H.R. < O.E. JEthelweard.

b. Confusion of cefel with cdf, eald, ealh.

We will now pass on to the second question: How is it that afel was soon confused with cdf, eald, ealh? No doubt the reason is that cefel was changed into eel. Hence names like jEthelflced and JElflced, were at once levelled under the same form, and this must have been the case with many others, by the loss of one consonant in position between two others 2. This levelling, which, consequently, was phonetic in many names the second element of which commenced with a consonant (JEthel-, JElf-, Ealhsige etc.), might then easily be analogically extended to those in which the second element commenced with a vowel. We must also bear in mind that owing to A.N. influence eel/, eald, and probably also ealh, were liable to lose their final con- sonant independent of position before a consonant or a vowel. Cf. felle < Engl. /eld in D.B., STOLZE, p. 41, and ol < o// < wulf instanced by such D.B. forms as Ardul and Ardulj, Torul and Torolf etc., HILDEBRAND, p. 350 3.

Our first undoubted instances of eel, al < cepel are from D.B. Though such name-forms as Alstan, Alsige, Alwine etc.,

1 A person, who in Florence of Wo. (Mon. Brit.) is called Dunstan, son of ^Ethelneth, appears in Langtoft I, p. 396, as D. fiz. Ageleth, correspon- ding to Agilet, K.Br. In Aylet, Allot A.N. t is substituted for English th. The loss of g, n after / may also be due to French influence.

2 Such a loss is recorded very early: uulsig < uulfsig; seoluini < seolhuini, Liber Vitse (MtJLLER, p. 28, p. 34); Ml&tanus F, A.S. Chr. == jElfstan D; Alwold ¥ = mfwold E. Cf. also M.E. halter < O.E. hcel/ter.

3 In some cases cepel is levelled under eald by metathesis : Audebert < ^Ethelberht, Gaimar, see EATHMANN, p. 54. Cf. also O.F. Maudebert < Mal- debcrt < Mctialbert, MACKEL, p. 169.

108 R. E. Zachrisson

D.B., may be due to the corresponding O.E. compounds in eald, ealh — which, besides, frequently occur as celf, alf, aid, aid, alch, alche (cf. Searle) with their final consonant kept - it is, on the other hand, a matter beyond all doubt that in many, not to say most instances, they go back to an O.E. ground-form1 aefel. In the Viet. Hist, of Surrey (see the Index to D.B. names) we are expressly told that Alfred, D.B., is iden- tical with ^thelnoth of Canterbury. Searle, p. 66, identifies Almar, D.B., with .ZEthelmser, bishop of Elmham (in Ramsey Ch. called Ailmar etc., cf. above). — In some cases we have established the identity by comparing the nearly contemporary Exon and Exchequer versions of Dv. D.B. For the name of one and the same person we have found the following alternative spel- lings: Ailnod, Exon = Alnod, Excheq.; Aimar (< tefe), Exon = Almar (< afel), Excheq.; Almar ( <_ce])el\ Exon — JEdmter (< ce]>e), Excheq.; Ailsi, Exon = Alsi, Excheq.; Adestan, Exon = Alestan, Excheq. (often).

It still remains to decide how aefel has been changed into eel. This transition can hardly be due to English sound-change. According to several authors, there should have existed in O.E. besides afe one more shortened (pet-name) form of cefel, viz. (el, which, though of rare occurrence in the written language, might have been quite usual in everyday speech *. Nevertheless, as in the records eel for aefel did not become usual until the time of the Conquest, we must strongly suspect French influence. Then cd may have originated either directly from O.E. &fl, with already syncopated e, or from ceel, with later syncopation of e, another means of avoiding the hiatus. This is indeed the regular

development of O.F. names containing Germanic adal etc. Cf . LONG- NON, Polyptyque, p. 277, MACKEL, p. 16, p. 135, and such D.B. forms as Matdger, Maelger, Malger < Madalger, HILDEBRAND, p. 353. In this place we may add that to account for Aaluf in the French version of King Horn we need not assume with Mors-

1 Cf. MULLER, p. 4, who points to celberct, Liber Vitse, as one instance ; Searle has a few more references. Some instances of mal < madal, al < adal are also explained as pet-name forms by STAEK, Die Kosenamen, p. 46, p. 49. If those really are taken from original MSS. of the early date stated (7th and 8th cent.) the existence of al as a pet-name form of adal must be considered as proved.

Anglo-Norman influence on English place-names 109

bach (p. 311) an O.E. ground-form Apalwulf. AaluJ may be due to Aeluf < sEthelwulf, with subsequent assimilation of ae to aa. For similar O.Fr. instances cf. MACKEL, p. 16, p. 135.

The levelling of cefel and celf, eald, ealh under eel, al gives rise to great orthographic confusion. In MSS. of the 12th and 13th cent, old and new forms occur promiscuously for the name of one and the same person. This state of things may be illustrated by some instances derived from various sources. jElfno]), A.S. Chr., sheriff of He., occurs in Florence of Wo. (Mon. Brit.) as Agelnofh. Cf. Searle, p. 45. The same person is called Elfgiva in Wil- liam of Malmesbury, and ^Ethelgeovu in Florence of Wo. Cf. Searle, p. 38. In the Inq. El. we have noted the following

variants: JElricus = JElfricus = Alricus; Aljpinus = Alwinus = ^Iwinus; Aluuold = cedelwold — alfwold. - - ^Ethelwulf, bishop of Carlisle, is recorded in the Fr.Ch. as Adelulf, Adelolf, Aelolf, Aldulf. — Ramsey Chartulary, so rich in variants, offers some instances of alternative forms which may all of them be con- sidered as developments of cefiel: Ethelwin = Aihvin = Alwin, Ethelric = Etheric = Helric, though, on the other hand, the con- fusion is also well instanced: Aylsi = Elsin = Alxi (< ealh), Aif win = Ailwine, Athelwold = Alfwold, Ailwin = Alfivine = JElwyn, Athelred = Ealdred = Ealriedus. In R.G1. we have

noted the variants adelstone, aleston, alcston, v. 5233; Elfled, ib., v. 5428 and Alfled, v. 5474, are, according to Wright (see Index to names), identical with O.E. JEfhelfleed. - - Lastly we will ad- duce a few instances from O.F. chronicles: Alestain, Benoit, v. 1656 = O.E. JEthelstan\ Adellrict, Gaimar, v. 87 = Albrict, ib., v. 90, Adlestan, Elstan, ib., v. 2259 = Ealhstan, A.S. Chr., Adelher, AlJcer, ib., v. 2498 = Ealhere, A.S. Chr.; Eldres d' Engletierre, Mousket, v. 15254, Eldre: De, ib., v. 16508 = King Ethelred; Aired, Rou, v. 260, Milred, ib., v. 1255 = King Ethelred II.; Eifrid, Wace, corresponding to ^Edilfrid, Bede, according to MADDEN, La3amon III, p. 422.

Present English surnames still preserve some traces of the old forms in al < cefel etc. The following examples are taken from Bardsley's Dictionary. Also here we have added some references from the H.R.

Aldrich, p. 45, AMrich, H.R. < O.E. Mlhelric etc.

All-, Albright, p. 47, Albert, H.R. < O.E. JEthelbeorU.

110 R. E. Zachrisson

Allnutt, p. 49, Alnath, H.E. < O.E. • JEihelnofh; t is due to French sound-substitution.

Elmer, Elmar \ p. 271, Elmer (1284), K.Q. < O.E. ^Eth- elmcer.

Allard, Allardson, Ellard, p. 47, p. 269, Alward, Elward, H.R., Alardus, T.N., Alard, F.A. < O.E. jEtlielweard.

Alwin, Alwine, Alwyne, Elwin, Elwyn, p.. 70, p. 273, Alwine, Elwin, H.R. < O.E. JttMwine.

If eZ, al in the above-mentioned instances is due to original cefel or celf, eald< ealh cannot be definitely settled.

Our conclusions as regards the development of cefel may be summed up as follows. At the end of the llth cent. O.E. cefel assumes four different forms:

1. Original cefel with the interdental spirant kept. This purely English form exists throughout M.E. times. From the middle of the 12th cent, or somewhat earlier a form appears in d (= the dental stop), probably due to the combined effects of A.N. sound -substitution and English sound-development. Con- trary to the fate of cepel, this form is still kept in the present English surnames.

2. ceel with loss of f owing to A.N. sound-change, ceel may be considered as a transitional form of short duration. The Normans were not very partial to the combination ceel. Various means were used to get rid of the hiatus which had resulted from the dropping of f.

3. ceil, no doubt, a further A.N. development of ceel. Hiatus was avoided by turning cee into cei. This form seems to have been the most popular one in M.E., and as ail or el it has been preserved to this day in various English surnames.

4. eel, either due to an O.E. shortened form of cefel, which, though very rarely recorded, may have been usual in everyday- speech, or what is more likely, another A.N. development of ceel. Hiatus was here avoided by dropping e. Present English sur- names preserve some traces also of this form.

1 As to el by the side of al, cf. our previous remark on adel, edeL d may also have arisen from ail either by A.N. monophthongizing (cf. M.E. frele = N.E. >frail») or by shortening in English itself after the transition of ai > .

Anglo-Norman influence on English place-names 111

The development may be illustrated by the following table.

^»7i 7 -. (a M T? ^ J ^hel, ethel (now lost). <epa (e. M.JM | ^^ c^ > } ^ ^

O.E <£/>eZ > (end of lithe.)

6e? > (e. M.E.) | ^?'

lltnc-) l^Z>(e.M.E.&N.E.

\

ceel > (end of

2. cefel in place-names.

After this somewhat detailed discussion of the development of O,E. cefel in personal names we may pass on to place-names containing cefiel, the various forms of which will be easily under- stood after the foregoing investigation.

We have divided our material into two groups, according to the greater or less degree of certainty with which the etymo- logy may be established.

a. Place-names which to judge by old forms on record contained O.E. tffel.

a. The present forms exhibit th.

1. Athelhampton (Do.), pron. »Athelton», Hope; called Athelhamstone in Index Villarum (1668), Athel Hampton on the map in Camden. 2. Athel- ney (8m.), pron. (aedelni), Tanger < O.E. JEpelinga eigg, -ig, A.S. Chron., JEthelinga ege, Kemble. 3. Atherstone (Sm.) pron. (sedastan), Tanger < O.E. jEthelheard, Searle. 4. Athelington (8f.), pron. »Allington>, Hope, < O.E, cepeling.

D.B. Adelingiz, Adelingensis2 . Geld. Inq. Adelingeres* (1086). 13th c. — Athelney Beg. Alardeston* (1225), Athelardeston* (1251). O.B.

IPideV athelamston1 (79—88). Ch.B. Athdingeneye* (1267). Muchel- ney and Athelney Ch. Athelingeny* (1254), Athelney*, Alynge- nie"*. Sm. K.Q. Athelnigg*, Aveingleye2, Alegine*, Athelingleye*, Aling'2. Sm. S.B. Athelarstone* (1272). Sf. P.F. Atliton* (1202\ Allingeton* (1218), Athelinton* (1234), Athelington* (1254). 4th c. — F.A. Alhelhameston1, Athelmanstone* , Athelardeston* (1303), Ade- lardeston* (1346). P.R. Athelney^ (99—1401). Sm. N.V. Athel- ynge* (1316). Sm. P.F. Athelardeston*. Sf. P.F. Allington4. Sf.S.B. Ayllington* (1327). 15th c. - Index. Alyngton4 (1477). P.B. Athelney* (67—77).

112 R. E. Zachrisson

P. The present forms exhibit d.

1 . Admington (Gl.) < O.E. Mihdm or ^Ethelhelm, Searle. 2. Adling- ton (La.) < O.E. cepeling or JEihelwine, Searle. 3. Edlington (Li.) < O.E. cppeling. 4. Adlingfleet (Yo.) < O.E. cepeling. 5. Edlington (Yo.) < O.E. apeling.

D.B. Edelmintone1, Ellingetone*, Tedlintune,3, Adelengesfluet', El-

lintone5.

12th c. — Index. Hedlingt* (1170), Adelingfleoth* (1154). 13th c. — Ch.B. Adilmington1 1, Edlinton* I, Edelington* II, AdlingflePTL.

H.B. Edeling-, Edlington5. Index. Edelineton* (1297). La.

P.F. Adelvinton* (1202). T.N. Edlington*. Yo. I.P.M. Athe-

lingflet* (1274). 1 4th c. — P.A. Adelminton1, Edelington*. Index. Athelyngflet* (1366). Yo.

N.V. Adthelingflet*, Edelington5. 1 5th c. — P.B. Adlyngflete4' (67—77).

7. The present forms exhibit al, el.

1. Allington (Do.)< O.E. cupeling. 2. Ellingham (Ha.) < O.E. cepe- ling. 3. Elton (Hu.) < O.E. cepeling. 4. Aloes Bridge Hundred (Ke.) < JEthelwulf, Searle. 5. Allexton (Le.)? < O.E. *JEiheloc or Heafiulac, Searle. 6. Allington (Li.) < O.E. wpeling. 7. Ellerton (Sh.) < O.E. JEfelheard, Searle. 8. Albrighton (Sh.) < Jtthelbeorht, Searle. 9. Albrightlee (Sh.) < O.E. jEthelbeorht, Searle. 1 0. Ellaston (St.), cf . Allexton (Le.). 1 1 . Allington (Wl.) < O.E. adding.

D.B. Adelingtone1, Adding eliam?, Adelintune*, Adelovesbrige*, Adi- lovtesbrige*, Adelachestone5, Adelingetone*, Adelinctune6, Etbritone*, Albricstone*, Etbretclie9, Edelache-, Elachestone10, Adeling-, Allen-, Alentone11.

1 2th c. — Fr.Ch. Alinge-, Elingueham*. Index. Halak.^ Alakestona5. Le.S. Adelactone6.

1 3th c. — Ch.B. Albrichton* I, Edbrittleg9, Aling-, AIU-, Aleton11. P.A. Athelinglone1 (1284), Aluffesbrogg*, (1284), Ethelarton"1 (1284), Al- bricton*. H.B. Aylingtone*, Alingtone8 Adlig-, Alinton*, Albriton*, Alington11. I.P.M. Harthelakeston* (Hen. III.). P.B. Atheloxton* (72—81). Bamsey Ch. sEthelyngtone*, Hathelintone\ Athelinton*, Adelyngtona6, Adelingtune* , Ayllingtone* (1251), Ailingtona* Aylin- tone* (1244), Alitona*. T.N. Ali'geton1, Alin-, Alington*, Albrit- fon8, Attrition*, Adlac-, Adlas-, Adthelaxton10, Alingeton11.

I4thc. — C.B. Athelocston* (13— 18). Cat. A.D. Aylington* (1316). P.A. Elyngham* (1316, 46), Aillington* (1316), Aloluesbrigge* (1316), Aloucsbregg* (1346), Adlington* (1316), Alyngton* (1303, 46), Al- bryghton* (1316). Index. Alyngton* (1392).

Anglo-Norman influence on English place-names

113

15th c. — F.A. Athelington1 (1428), Elyngham* (1428), Aylington3, Adelokston5

(1428). I.P.M. Adloxton* (Hen. VII.). 16th c. — Index. Aylyngton3 al. Aileton3 (1517).

b. Place-names which, though no old forms with th, d are on record, in all probability contained O.E. cefel.

a. The present forms exhibit ai(l).

1. Aylscot (Dv.) < O.E. Mthelwulf, Searle. 2. Aylsford (Dv.) < O.E. JSthehvulf, Searle. 3. Aylburton (Ql.) < O.E. ^Ethelbeorht, Searle. 4. Aymes- trey (He.) < O.E. ^Ethelmund, Searle. 5. Aylmerton (Nf.) < O.E. ^Ethelmcer, .Searle.

D.B. Ailevescota1, Exon, Eilevescote1, Excheq., Ailavesfort* ', Exon,

Ellaves/ord? , Excheq., Almartune5. 1 3th c. — C.B. Albrithton3 (79—88). Ch.B. Ailbricton* (1224), cf. corrigenda

in vol. II., Aylmerton* (1281). F.A. Ayberton3 (1284). H.B.

Albricton3, Ailm'eton5. T.N. Albricton3, Elbrichtonas. 14th c. — C.B. Aylmerston5 (18—23). F.A. Ales-, Alscote1, Aylesford\

Ailberton3 (1303), Aylfo-izton*, Aiberrton3 (1316), Ayiverton* (1346),

Aylmondestre* (1303, 46), Aylmerton5 (1316). 15th c. — F.A. Ailberton3 (1402). I.P.M. Aylemertori* (Hen. VII.). P.B.

Aywarton3 (61—67).

(3. The present forms exhibit a(l), e(l).

1 . Arlsey (Bd.) < O.E. JSbhelric, Searle. 2. Elvaston (Db.) < O.E. JEthelweald, Searle. 3. Alveley (Es.) < O.E. ^EUelgifu, Searle. 4. Elberton (Gl.) < jEthelbeorht, Searle. 5. Elmston (Gl.) < O.E. ^Epelmund, Searle. 6. Elms- tree (Gl.) < O.E. JEthelmtind, Searle. 7. Abbotsley (Hu.) < O.E. ^Ethelbeald, :Searle. 8. Allington Hundred (Ke.), called Elnothington in LAEKING'S D.B. of Kent < O.E. JEthelnoth, Searle. 9. Elmsthorpe (Le.) < O.E. ^thelmcer, Searle. 1 O. Aldrebury Hundred (Wl.) < O.E. ^Ethelweard, Searle.

D.B. Alricesc-ia1 , Alevvoldestune\ j<Elvvoldestune* , Eldbertone*, Al- mundestan5, Alnoitune*, Alwarberie10.

.2th c. — Fr.Ch. Aelivaleio3, Aliuelia3. Index. Aluualdestona*.

.3th c. — Ch.B. Ailricheseye1 I, Alverycheseye1 II. F.A. Aylmerthorpe9 (1284). Gl. Bees. Ayl-, Almundeston* (1240). H.B. Aylwarde-, Alwardebyr10. Index. Ailricheseya1 (1206), Alrichesheye1 (1251), Haurich1 (1255), Arlicheseie1 (Hen. III.), Alwaston2 (1282), Elmun- destruo* (1201), Ailmundestre6, (1212). Bamsey Ch. Alboldesle1, Albodesleye1, Aylboldelle1 , Addeboldesleye1 . T.N. Alwaldes-, Alval- deston2, Aywaldeston2, Eylnothynton*, Altcarebir™, Alwardebur™.

8

114 R. E. Zachrisson

14th c. — F.A. Alwaston*, Aylwaxton*, Aylberton* (1303), Aylbryghton* (1346),.

Eylnothinton* (1346). Index. Arlechy* (1386), Ayllewaston* (1322).

P.R. Ail-, Aylwaston* (77—81). 1 5th c. — P.A. Aylmerthorpe9 (1428). G-l. Bees. Ayleberton4- (1437). Index.

Arlichesey1 (1402), Erleshey1 (1438).

No doubt, the majority of instances brought together in list b represent developments of original <epel. A few criteria may be pointed out. If the second element of the name con- tains -mund, celf cannot well be contained in the first, no O.E. *JElfmund being recorded. If early spellings exist in al, as well as ail, which, indeed, is the case with the majority of our exam- ples, this seems to point to original cefiel. Cf. the early forms of a Hundred in Np. now lost: Egelweardesle (1076), Geld. Inq.; Alwardeslea, D.B.; Aylwoldesle, Np.S., no doubt < O.E. ^Ethelweard, Searle.

As we have been unable to ascertain the modern pronun- ciation, except in a very small number of cases, we have classi- fied our material according to the present spelling, though we are quite aware of some incongruities having resulted from this arrangement. Thus Athelington (Sf.), pron. »Allington», Hope, would perhaps better have been placed with the examples in list a Y, Elton (Hu.) and others, in which e/, in all probability, is due to earlier ail, with those in b a etc.

As to the explanation of old and modern forms we need not add much to what has been said already. All the develop- ments we have found in personal names compounded with aefet are represented also here. Original (ct) is still pronounced only in Athelney, Athelhampton, and Atherstone (a a). More usual is adel, but whether the d is due to French sound-substitution or English sound-development cannot well be decided in every special case. Undoubted instances of ail are few, but their number will swell considerably if we add the numerous forms in el, which often go back to earlier ail. Here may belong Elton (Hu.), (a 7), Elberton (GL), Elmston (GL), Elmstree (GL), Elms- thorpe (Le.) (b p). at and el are astonishingly well represented. It may be pointed out that to O.E. cepeling corresponded O.Fr. edelin, edeling, elin (Godefroy). Note that with reference to the same place, Elton (Hu.) (a 7), all four forms afel, add, ayl, al occur in 13th cent, spellings iu Ramsey Ch.

Anglo-Norman influence on English place-names 115

Appendix A. Various ways of rendering ({>), (a) in early English and Anglo-Norman MSS. -

As to the various ways of rendering (h), (d) in A.N. texts, we find above all d, which, of course, is the historic spelling. Finally there is an interchange of t and d = (h), (d). l Of more scarce occurrence are the symbols p, cT, which, naturally, must have been borrowed from English. Cf. MENGER, p. 39; VARN- HAGEN, Zeitschrift fur Rom. Phil. X, p. 296 ff. Lastly, we also find th and dh.

th as an orthographic symbol was well known in O.F., where it was employed particularly in learned Latin words. Alternating with t it was also used to render Germanic (J>) in proper names. Cf. SWEET, H.E.S. § 600; WALTEMATH, p. 56. Thus used, it had certainly in both cases the value of (t). Now we may very well ask: how came th to be used in order to denote (J)), (3)? GASTON PARIS in his Introduction to Alexis, p. 95, assumes that the Normans borrowed it from English. This sug- gestion leads us to examine the various ways in which the interdental spirant was rendered in O.E. and early M.E. MSS.

On this point SWEET, H.E.S. § 516, gives the following in- formation: »f is generally denoted by the Latin th in the oldest mss, especially initially and finally, which th is sometimes

1 A few times intervocalic (d) is, curiously enough, rendered by t: Porlemute, Exch. D.B., Porlemuta, Exon D.B. (see list A); Bute, D.B. (see F 1); Smethedune, Smetedune, D.B. = Sinithdon (Nf.) < O.E. smefe} etc. It is not impossible that intervocalic t for d is often due to mere carelessness. Finally there was a regular interchange of t and d — not only of £, d = (J>), (d), but also of t, d = (t) — and by thoughtlessness this interchange may have been graphically transferred to other positions. But there is another possi- bility which must also be taken into account. We know that in O.F. Latin t and d under certain circumstances were changed to spirants, and subse- quently lost. Nevertheless, the French scribes — to whom the etymological ground-forms not only of many Latin words, but also of a great many Ger- manic personal names and place-names were well known - - when they wrote in Latin endeavoured to keep t and d etymologically apart, which, naturally, often caused one to be put for the other. That this really was the case is clearly seen from the early spellings of several places in Normandy. Thus Tosny (Eure) occurs in the 12th cent, both as Totteneium and Todi- niacum. All other forms exhibit loss of the dental. See Die. Topogr. de 1'Eure.

116 R. E. Zachrisson

abbreviated to t (Ep.GL). Medial p is often denoted by d, which also occurs finally, and even initially (Ep.GL). The Runic p is rare in Ep., as also ft. Ep. has once dh\ also ph, fth occur (blendings). In the later Anglian texts ft is universal, as also in most of the oldest W.S. and Kt. mss. In some early W.S. mss p, however, predominates. » So far Sweet; we have, besides, ascertained that, though more seldom, d occurs by the side of p, ft in the following texts : Liber Vitse of Du. (8th c.), cf . MILLER (Palaestra IX), p. 29; Vesp. Ps. (9th c.), also a for d, cf. SWEET, O.E.T., p. 187; Kit. of Du. (10th c.), cf. LINDELOF, p. 67; Gospel of St. Mark (Rushworth MS. 10th c.), also It for d, cf. Miss LE^ (Anglia XVI), p. 125, p. 127; Rule of S. Benet (c. 1020), also & for d, LOGEMAN, p. LII f. Also in several e. M.E. texts d occurs for <f, and vice-versa. Cf. BEKBERICH, Herb. Apul., p. 9; LUCHT, La3amon, p. 65.

The result of our examination is that, though p, ft is the most usual way of denoting the interdental spirant in O.E. and early M.E. MSS., also d occurs sporadically all through. Besides, th has been noted not unfrequently, particularly in names, where it certainly had the value of (]}), (d). When used in A.N. MSS. to denote the interdental spirant, th may, consequently, very well have been borrowed from English.

Appendix B. Orthographic confusion between the symbols d and i, g in the diphthong ei, ai etc. -

When for the same place in MSS. of about the same time or even in the very same MSS. we find such alternative forms as Lachinchede, Lagingehee, Lachingeia, Saltrede, Saltreye (see list D 1, 2) it goes without saying that spellings like -hede, -rede could indicate three different pronunciations : he&e, hee, heie; refte, ree, reie. Hence it follows that on the one hand d could some- times be inserted between two wowels in words where it was not etymologically justified, while on the other the notion might easily be entertained that in spellings like -hede, -rede d really denoted the glide (j). Consequently, as -ede was equal to >ee, -eie C O.E. -epe, we also, conversely, find -ede for O.E. -eie, -ege- Thus Eastry (Ke.) appears in D.B. as Estrede, Estrea, Estrd. O.E. forms of this place are Eastrege, Eastorege, Easterge, Kemble

Anglo-Norman influence on English place-names

11?

153, 191, 195, Eosterege (a. 811), Index, so the suffix obviously contains O.E. eg, Mdf. p. 49. — Swavesey (Cb.), Svavesye, Svavesy, D.B., Suauiseye (1260), P.F., Swavesey e (1346), F.A., is derived by SKEAT, Pl.-Ns. of Cb., p. 54, from an O.E. *Swcefeseg, -ig. The forms Suauesheda, I.C.C., which has escaped his attention, is to be explained in the same way as Estrede, D.B.

Other instances of d for i, g, in the diphthong ei, ai, are Cuselade, Crucheslaga, D.B. = Whixley (Yo.), Quixelay, T.N., Quixley, H.R., Whixeley (1307—13), C.R.; Trulleda, Trullea, Thru- lege, Fr.Ch. = Throwley (Ke.), Thruleghe, Thurlegh, F.A.; JacJceslada, JaUe, Ch.R. I = Yaxley (Hu.), Jdkele, H.R. The second element of all these three places is O.E. leah, leage. Also in such forms as Ledecestre, D.B., Lefeccestre (1124), A.S. Chr. E = Leicester (Le.), Legra-, Ligraceaster, A.S. Chr. B, D, E., d, f, in all proba- bility, stand for i, g. Loss of r before the following consonant — which, moreover, is very usual in A.N. - - may here be due to dissimilation; another possibility is confusion with Legeceaster, A.S. Chr. = Chester (Ch.), of which also a form with unetymo- logical r appears in the A.S. Chr., viz. Legercyestre, MS. a. A second example of d for g in the A.S. Chr. is Widrecestrescir E = Wigrecestrescir, ib.

The opposite phenomenon, g for d, we have also noted in a few cases. Nether Avon (Wl.) occurs in D.B. as Nigravre, which spelling has been interpreted »nigrum arvum» by JONES in his D.B. of Wl. In our opinion Nigravre stands for Nidravre < O.E. niperavene (in the Fr.Ch. the place is called Nideravena), with g for d and transition of n to r. Cf. O.Fr. havre < Germ. htefene (Korting). - Stevenage (Ht.) is called Stigenace in D.B. Other early spellings are Stichencece, Kemble 178, corrected to Stithencece by Thorpe, Stidenach (John), Rotuli Curise Regis. With some hesitation SKEAT, Pl.-Ns! of Ht., p. 66, derives the name of the place from the O.E. formula cet fare stifan hcecce, but Ids that g in D.B. seems to point to Stichen-. The supposi- ion that g stands for d at once solves every difficulty. Later >rrns Steben-, Steven-, Stiven- exemplify the sporadic transition )f eT> v\ then popular etymology may have connected the first element of the word with the usual Norman name Stephen. - A third example of g for d is Sichemug, Fr.Ch. = Sidmouth (Dv.).

118 R. E. Zachrisson

Appendix 0. Loose s between two place-name compounds. -

That in Chelsea (Mi.) < O.E. Ccelic hyp, Earle (see list D 2) etymological (t/) has been replaced by s is perhaps best explained thus: an inorganic s has been inserted between the two com- pounds Chelch- and -ey (< -heth), whereupon (£/) was lost before ,9: *Chelchey (< Chelcheth) > *Chelchsey > Chellsaye. A similar development has taken place in the following place-names: Arlsey (Bd.), Ailricheseya (1206), Arlechy (1386), Index < O.E. ^Ethelric, Searle; Hildersham (Cb.), Hildricesham, D.B., Hildrichesham, I.C.C. < 'O.E. Hilderic, Searle; Aldersgate (London), Aldrichgate (1428), F.A.? < O.E. Ealdric, Searle; Gothersley (St.), Godrichesky, Gode- richeley (14 c.), see DUIGNAN, St. Pl.Ns., p. 69 < O.E. Godric, Searle; Atherstone (Wa.), Adericheston, T.N., Edriceston, R.B., Atherestone, H.R. Cf. also e. N.E. voutsafe < vouchsafe (see HORN, Gramm. § 220 Anm. 2), where, however, the assimilation has not been fully carried out.

The uncertainty as to the use of s in the old forms of the above-mentioned places is noteworthy. In the following two place- names the forms with s have been kept in the local pron., whereas the present spelling presupposes loss of s and change of (t/) to (d3): Idridgehay (Db.), pron. »Ithersea», Hope, Iddersey (1574), Index, Edrichey (1638) ib. < O.E. Eadric, Searle; Sawbridgeworth (Ht.), pron. (ssepsa) < O.E. * Scelerhteswyr]), cf. Introduction, p. 7.

Indeed, in numerous instances from our earliest sources there appears on the one hand an inorganic s between the two compounds of a place-name, while on the other s is often omit- ted in the same position, not only when it is a mark of the geni- tive, but also when it forms part of the word-stem itself. We will content ourselves with adducing the following instances of this very usual phenomenon: Cheretowe (1402), Index = Church- stow (Dv.) < O.E. stow\ Hechesburne, Echeburne, D.B. = Exbourne (Dv.), Ekes-, Ekysburn, F.A.; Boystok, Boytok, T.N. = Bystock (Dv.), Boystok (1303), F.A.; Borchetelle, D.B. = Borstall (Ke.) < O.E. borgsteal, Mdf., p. 16; Kate-, Katesby, H.R,, Gate-, Catesby, F.A. = Cadeby (Li.); Cnichtebrugge (Hen. III.), Index = Knights- bridge (London) < O.E. cniht- Harningsey (Edw. VI.), Mi.F.F., Harengey (1316), F.A. = Hornsey (Mi.); Smethe-, Smethes-, Smes-

Anglo-Norman influence on English place-names 119

don, H.R. = Smithdon (Nf.) < O.E. smepe- Eincsteda, Eincteda, D.B. = Ringstead (Nf.) < O.E. stede; Withesfeld (1150), In- dex = Whitfield (Np.) < O.E. }iwlt\ Winemereslea, Wimeres-, Wimerleu, D.B. = Wimersley Hundred (Np.) < O.E. Winemcer, Searle; Clailei, Claislea, D.B. = Cleley Hundred (Np.), Klegele, Geld. Inq. < O.E. dag; Torlaston, Torlaxton, F.A., Torlaton (1539), Nt. Recs. = Tollerton (Nt.); Timbrecumbe, D.B., Tymbercome {1384), Index = Tirnberscombe (Sm.); Frodeley (1305), Index, Fodresleye (1262) = Fradley (St.). Cf. O.E. Frod, Froda, Searle. According to Hope, Alnmouth (Nb.), at the moutb of the Aln, is dialectally pronounced »Aylesmouth». WESTPHAL, p. 19, p. 33, has noted several instances from O.Fr. texts: Banesburc < Bcebbanburh, Everswic < Eoferwic, Lichesfeld < Licetfeld etc.

The appearance or non-appearance of s in the examples tibove has probably no connection whatever with the instances of loss or addition of an unetymological s caused by A.N. influence which are discussed in § 5. In the present case we have, no doubt, to do with a kind of analogical transformation. The mere circumstance that of a great many O.E. personal names frequently used as the first element of place-names there existed two forms of the genitive, one with s, another without (Frodes — Frodan, Haddes — Haddan, Lulles — Lullan etc.) may easily have led to confusion and uncertainty as to the use of s between two place-name compounds in general. Thus Wimerleu for *Wimers- leu (< O.E. *Winemareskak), Smethesdon for Smethedon (< O.E. *Smef)edun) on the analogy of Frodeley, Frodesleye (< O.E. * Fro- dan-, *Frodesleage}.

When establishing the etymology of place-names on the evidence of early M. E. spellings it is of great importance fully to realize that 6? is » loose » in position between two compounds. Consequently, if s occurs in e. M.E. forms it does not always follow that a personal name is contained in the first compound, nor does the absence of s absolutely tell against such a supposi- tion. It seems to us as if Prof. SKEAT in establishing the etymo- logy of place-names were apt to attach too much importance to the appearance or non-appearance of s in the early spellings. Cf. Pl.-Ns. of Cb., p. 12 under Rampton, and p. 37 under Snail well.

120 R. E. Zachrisson

§ 7. Change or interchange of the liquids (/, n, r).

The collections of material which precede the various sec- tions of this paragraph have been arranged in the following way. The altered forms of the various place-names are given first* then the modern ones, and lastly some references for the ety- mological forms. We have denoted doubtful instances by placing them inside square brackets.

A. Interchange of I, n, r. 1. Dissimilation of r — r.

a. r — r > I — r.

Salopescira (1155), Pi.R.; Salopescira, Fr.Ch.; Salop', H.R., T.N.; Salopia- (1231—34), C.R.; Salopsire (1232—47), P.R.; Slobschire (13th c.), Jesus Coll. MS. (Oxford); Salop (Hen. VII.), I.P.M. — Shropshire or Salop < O.E. Scrobbes- byrigscir, A.S. Chr.; Sciropescire, D.B.; Schorpschire, Scropssire, R.G1.

Salopesbiriam (1139), Index; Salopesberia (1155), Pi.R.; Salopesberia, Fr.Ch.;. Salopisbur, H.R.; Salop, T.N.; Salop' (1225), Ch.R.; Salopesbyry (1232—47), Silop (1272—81), Salop (1391—96), P.R. = Shrewsbury (Sh.) < O.E. Scrobbes- burh, A.S. Chr.; Sciropesberie, D.B.; Scroberiensis, Fr.Ch.; Ssrofsburi, R.G1.;. Shrovesbury, Shrewsbury (1318—23, 43—46), C.R.; Shrouesbury (1391— 96), P.R.;. Schrobesbiri, R.Br., p. Ill; Shrewsbury (Hen. VII.), I.P.M.

Saleurdine, D.B. = Shrawardine (Sh.), Schrewardin, H.R.; Srewardin, T.N.; Schrewardyn, F.A.

Scelesberi, Seleberi, A.S. Chr. MS. F; Salesberia, Fr.Ch.; Salesburi A, Salusburi B, La3amon, v. 15290; Salesbir, T.N.; Salesbury, R.GL; Salesbury (Hen. VII.), I.P.M. = Salisbury or New Sarum (Wl.) < O.E. Searoburg, B, C, Sereberi E, A.S. Chr.; Saresburia (1155, 89), Pi.R.; Saresberie, Fr.Ch., Ch.R.; Saresbur', Sarru', H.R.; Saresbury (1232—47), P.R.; Saresbury^ Jesus MS.;. Nova Sarum (1391—96), P.R.; Sarum, V.E.

Walegrava, Fr.Ch. == Wargrave (Be.) < O.E. wer, Mdf., p. 146; Were- grauce (1061—65), Index; Wergrave, F.A.

Bilichangram, D.B.; Bilchangr (1200), Index = Birchanger (Es.) < O.E. byric, Mdf., p. 23.

[Belford, F.A. = Barford (Nf.); Berford, F.A.; Berford, T.N.]

b. r — r > r — I.

[Brithlinges, T.N.; Britlingeseye, H.R.; Bridlinges (1231—34); C.R.; Bryt- lingseye (1303), F.A.; Brickelsye, Brykkylsay , Brikelsaye (1476-85) P.R. = Brightlingsea (Es.), pron. (brikalsi) < O.E. Beorhtric, Settle; Brictrtceseia, D.B.]

Anglo-Norman influence on English place-names 121

c. r — r > n — r.

[Henert, D.B. = Herriard (Ha.); Herierd (1190), Selborne Ch.; Here- hyrdc, Herierd (1316, 1428), F.A.]

[Kencbroc, Fr.Ch. = Carbrooke (Nf.); Kerbrok (13 and 14 c.), F.A.; Ker- broc, -brock, H.R.]

2. Assimilation to r — r.

a. I — r > r — r.

[Bergemere, Borgemere, Burgemere, D.B.; Burgamera, Anc.Ch. = Bolmer or Baliner (8s.), cf. O.E. bulemere, Birch 687; Bulmer, D.B.]

b. n — r > r — r.

[Estarforda, D.B. = Stanford (Nf.) < O.E. stan; Estanforda, D.B.; Stan- ford, F.A.] "

3. Dissimilation of 1 — 1. a. I — I > r — I.

Erwadeston, Fr.Ch. = Elvaston (Db.) < O.E. Mthelweald, Searle; Elwal- destone, Fr.Ch.; Aluualdestona (Hen. II.), Index.

Ciresfel, D.B. = Chelsfield (Ke.) < O.E. Ceol, Searle ; Cheleffeld, T.N.; Chellesfelde (1316, 46), F.A.

Herelou (1254), Index = Belleau (Li.); Elgelo, D.B.; Helghelowe, T.N.; Hellowe (1355), Index. The first element may have contained O.E. Helga, Searle. As for the modern form, see Introduction, p. 13.

Brumhelle,- D.B. = Blymhill (St.); Blumenhulle (c. 1250), Index; Blumon- hull (14 c.), see DUIGNAN, St. Pl.Ns., p. 19.

[Guldesmere, D.B.; Ingolvesmera, Anc.Ch. = Ingoldsmells (Li.); Ingolde- meles, Ch.R. II; Ingoldemoles, Ingoldesmolis, T.N.; Ingoldmels, -mesles, F.A. The first part of the name seems to -contain O.E. Ingulf or Ingweald, Searle *. The second element is, no doubt, due to Scand melr, dial. Engl. meol = »a sand-bank» or >sand-hill», frequent in proper names. See E.D.D.]

C. I — I > n — I.

Soanesfelt, D.B. = Swallowfield (Be.) < O.E. swealwe, Mdf., p. 131; Solafel, D.B.

1 In having been taken for the Latin preposition in accounts for its non-appearance in D.B. Cf. Gemunstorp, D.B. = Ingmanthorp (Yo.).

122 R. E. Zachrisson

[Nyland or Hand (Do.), Hand, Inlande, D.B.; Liland, T.N.; La Eslond <1284), Layelond (1303), Le Ylond, La Ilond (1431), Laylond (1428), F.A. The etymology is O.E. legland, Sweet, A.S. Die.]

Nicol (1196), Le. Recs.; Nicholia,' Ramsey Ch. = Lincoln (Li.) < O.E.; Lincolne, Lincolle, A.S. Chr.

4. Assimilation to I — I. a. r — I > I — I

Belissolt, Belicolt, D.B. = Bircholt (Ke.) < O.E. llerce, byric-, Beriscolt, D.B.; Biricholt (1204), Ke. P.F.; Byrcheholt (1346), F.A.

Clawelle D.B. = Crowell (Ox.) < O.E. crawe, Mdf., p. 30; Crawell, H.R.; Vrowell (1316, 46, 1428), F.A.

Fallemella, Anc.Ch. = Falmere (Ss.) < O.E. mere, Mdf., p. 92 ; Felesmere, Falemere, D.B.; Fallemere (1377—81), P.R.

Levechanole, D.B. = Lewknor (Ox.) < O.E. ora, Mdf., p. 99; LevJcenor, H.R., Leukenore (1316), F.A.

Ellcsmeles, D.B. = Ellesmere (Sh.) < O.E. mere; Ellesmere, H.R.; Ellesm'e, T.N.; Ellesmere (1402), F.A.

C. W — Z > Z — Z.

Glole, D.B. = Church Knowle (Do.) < O.E. cnott, Mdf., p. 28 ; Cnolle, D.B.

Hanenchelole, Exon D.B.; Hanechelole, Excheq. D.B. = Honicknowle <Dv.) < O.E. cnoll.

Benclolle (1350), Index = Bincknoll (Wl.) < O.E. cnoll; Bienknolle (1362), Index.

Latelie, D.B.; Lettelegh, Latelie, Ch.R. I; Letele (1316), F.A.; igffefe, Le«% (1461—67, 67—77), P.R. = Nately, Netley, and Netley Abbey (Ha.) < O.E. Natanleaga A.S. Chr.; Neddan-, Nceddanleage, Kemble 624; Nataleie, D.B.; Natale, F.A.; Nateley, Netdeye, H.R.; Natelegh (1279—88), C.R.

5. Dissimilation of n — n. A. n — n > I — n.

Lantalan (1412), Index = Nantallan (Co.) < Celtic *nanton, Cornish nans = »valley> (cf. WILLIAMS, Die Franz. Ortsn. Kelt. Abkunft, p. 67; cf. also Mdf., p. 97); Namtalen (1392), Index; Nanstallan (1594), Index.

Lclantt (1006), Index = Lelant (Co.) < Celtic *nanton; La Nante(123Q\ Index; Lanantha (1331), ib.; Lananta (1349), ib.

Lantien, D.B.; Lantyen, T.N.; Lantien, F.A. = Lantyan (Co.) < Celtic *nowfon; Nauntyanc, Nantyan, F.A.

Anglo-Norman influence on English place-names 123

Tiltetona, Fr.Ch. = Tenterden (Ke.); Tenterdenn (1316, 46), F.A.; Ten-

»,, H.R.

Clotone, D.B. = Knutton (St.); Cnotton, Cnoton (12 c.), see DUIGNAN, St.

.Ns., p. 89.

b. n — n > r — n.

[Carenduna, D.B. = Canewdon (Es.); Caneuedon, H.R.; Kenewedon (1227), i.R.; Canwedon, Caneudon, F.A.]

[Morchelona, D.B. = Monkton (Sm.) < O.E. munuc, Mdl, p. 96; Moneche- tona, D.B.; Moneketon, Monketon, Moncketon, F.A.]

6. Interchange of 1, n, r «'w o^er cases.

a. r — w>Z — n.

Melestun, D.B. = Murston (Ke.); Merston, T.N.; M'sfow, H.R.

b. n — r > I — r.

Elesford, D.B. = Eynsford (Ke.) < O.E. ^Enes ford, Index; Aynesford, €h.R. II.

Boleberie, D.B. = Bunbury (Ch.); BoneUr, Ch.R. II.

c. n — I > r — I.

[Tarefeld, Ch.R. I = Tanfield (Yo.); Tanefeld, D.B.; Tanfeld (1284), Yo. K.Q.]

I, n and r in two syllables of the same word exercise in many Lido-Germanic languages a mutual dissimilatory or assimi- latory influence; I is often also exchanged for r, and vice- versa, a phenomenon already known in the »Ursprache», as is shown by the coexistence of the suffixes -tro-, -tlo-, -dhro-, -dhlo-. Of. BEUGMANN, Kurze vergl. Gramm. der indog. Sprachen, p. 117. This interchange of the liquids is by no means a characteristic of the O.E. dialect. On the contrary, O.E. sometimes preserves the original sequence of consonants in words which in other Germanic dialects have undergone dissimilation. Cf. O.E. frqfor against O.H.G. fluobra.

KLUGE, Paul's Gr., p. 1015, draws attention to some late O.E.' cases of dissimilation and assimilation between r and I. The instances he adduces are in our opinion not wholly conclusive. Slapor for slapol and slaporness by the side of sldpolness may

124 B. E. Zachrisson

represent the same interchange of suffixes which appears in waccer, wcecer together with wacol. The same is perhaps the case with diper and clypol answering to Dutch Mepel, M.H.G. Uepfel, Uepfer. Gyrder < gyrdel seems to be a ghost-word, as it is ignored both by Sweet and Bosworth-Toller. Stratmann's M.E. Die. and Matzner's Worterbuch have only forms with /.

There is hardly even in modern literary. English any word of native origin which may be pointed to as a safe instance of the interchange here in question. » Smoulder » is dubious; cf. Skeat's Concise Etym. Die. » Riddle » < O.E. hrldder is explained by Skeat ib., as due to the change of suffixes. In some Mod. E. dialects the original r is still preserved: ridder (Sm.), see E.D.D.

Where interchange or change appears in French loan- words, this is very often a mere reflection of the uncertainty as to the use of the liquids which prevailed in O.Fr. itself l. The frequent appearance of / for r, n, and conversely, may, indeed, be considered as a characteristic feature of the Romance langu- ages. In Latin the dissimilation of two consecutive r's was regular (cf. J. PASSY, Sur les changements phonetiques, p. 201), and in the French, Italian, and Spanish languages we find at different periods the same tendency: /, n, r in two syllables of the same word are often interchanged.

For our purposes it is of particular importance to know (1) if at the time of the Conquest the interchange of the liquids was a usual phenomenon in the Norman dialect (2), and, if so, under what circumstances this interchange took place, i.e. by what phonetic laws it was regulated.

1 Of Mod. English > purple*, which in e. M.E. appears as purper, doublets existed in O.Fr. itself (see Godefroy). In this case, however, another expla- nation would seem possible. The late occurrence of the form in el is per- haps best accounted for by assuming it to have arisen in the compound purpelpelle, by a process which might be called regressive syllabic assimila- tion. According to H<EVELMANN, Zum Konsonantismus etc., p. 28, the word in question appears in l.M.E. always as purper, while in the above-mentioned compound purpel is the regular form. Cf. also O.Fr. altel < principel alter

> principel altel, NYBOP, Gramm. § 509. The change of r to I in Mod. Engl.

> laurel* and * marble » < O.Fr. marbre, lorer etc. may have taken place in. the very act of borrowing, possibly owing to mishearing. The circumstance that both these suffixes existed in native words has, no doubt, contributed to facilitate the substitution.

Anglo-Norman influence on English place-names 125

In D.B. 2, n, r are interchanged in the following French words: Morel - Morinus, Columbels - • Columbers, Helion — Herion. See HILDEBRAND, pp. 339, 361. Other examples quoted by MENGER, p. 87, and STIMMING, p. 212 with numerous

references - - show that there must have been a great confusion between I, n, r in A.N.

Under such circumstances we seem justified in ascribing to A.N. influence the many instances of a similar confusion in the forms of English place-names recorded in Anglo-Latin MSS. from the time after the Conquest.

The second question still remains. Under what circum- stances did the interchange take place? In a treatise called »La dissimilation consouantique dans les Langues Indo-Europeennes» a French scholar, M. GRAMMONT, has formulated no less than twenty laws of dissimilation, illustrated by numerous examples, especially from the Romance languages. The correctness of these laws is, however, very problematic. Cf. MEYER-LUBKE, Ltbl. 1896, p. 409, and MERINGER, Indogerm. Forsch. 12, Anzeiger, p. 12. The true nature of dissimilation and similar changes is better denned by METER-LUBKE, Gramm. I § 570, ^as »une serie de changements qui ne se produisent pas selon des regies fixes, mais qui sont, dans une certaine inesure soumis a 1'arbitraire et au hasard.» The great variety of forms seen in the above lists strongly supports Meyer-Liibke's opinion. Meringer particu-

larly points out what an important part faulty speech or hearing play, especially where I and r are concerned. Cf. also MERINGER und MAYER Versprechen und Verlesen (Stuttgart 1895), pp. 89, 95 fL It is consequently very often just in loan-words that I, n, r are interchanged, and this circumstance certainly deserves to be taken into consideration in the present case. The English place-names were unfamiliar words to the Normans; they did not know anything of their etymology or sense, but wrote and pronounced them as they heard them.

The anomalous forms are also not so numerous in any

other record as in the earliest of them all, Domesday-Book. In

the 12th cent, they are still very well represented, particularly in

the French charters - - by no means an unimportant circumstance

- but after the middle of the 13th c. they may be said to have

126 R. E. Zachrisson

disappeared, if, of course, we except the few cases where they ultimately prevailed.

Though we have expressed it as our opinion that no fixed rules can be laid down according to which the changes we have to deal with in this section may be said to have taken place, our material has been classified under various headings so as to present a more perspicuous view of the whole. At the utmost^ a few general tendencies may be noted. Dissimilation seems ta be more usual than assimilation, and the change of r — r > I — r is of particularly frequent occurrence 1. For interesting O.Fr, instances parallel to the ones noted by us - - both in place- names and ordinary words, see NYROP, Grarnm. §§ 327, 328, 340, 359, 511, 512; EUREN, fitude sur YE francos (Diss. Upsala 1896), pp. 18, 37 ff.; WILLIAMS, Franz. Ortsnamen, pp. 22, 23; LINDSTROM, De obet. vokal. bortfall i nagra nordfranska ortnamn (Diss. Upsala 1892), p. 13, Anm. 2; NEUMANN, Zur Laut- und Flexionslehre des altfranz., p. 69 2.

Before going further, it will be convenient to try and establish whether some of our instances may be accounted for as due to mere confusion of letters. Cf. Introd., p. 6 f. A glance at the facsimiles of D.B. and some early 1 2th cent, records, like the Li.S., will suffice to show that I is quite distinct in form from n and r, so that all interchanges between I, on the one hand, and n, r, on the other, seem to require a phonetic explana- tion. Moreover, no graphic confusion between I and n, r has

1 STOLZE, pp. 37,38, has also noted a few cases of interchange between I and r in D.B., which he rightly ascribes to A.N. influence. One of his examples is not to the point. He thinks that Warborgetone and Walburgeton in SB. stand for the same place, but Walburgeton is identical with Walberton in the Rape of Benested, later on called Avisford, while Warborgetone stands for Warbleton in the Rape of Hawksford. Warborgeton D.B. corresponds to later Warbillhun (13 c.), Index, Warbleton (14 c.), P.R., C.R., but in the early Fr.Ch. we find the form Warborgultona, containing both I and r; so it is not safe to assume the change of I > r in the D.B. form.

2 We may add a few more: Linglonia (1162), Lingronia (1190), Ningronia (1278) = Langrune (Calv.), see Die. Top.; Berebrona (1116), Belebrone (1206), Bereborna (13 c.) = Bellebrune, see A. DE LOISNE, Les noms de lieu du P. de Calais, p. 7. n has been changed to r in Marville < Manulfi villa, see QUICHERAT, p. 62. In the R.B. the same person is spoken of as Hug. de Anvers and Alverso (n — r>l — r), but this instance also admits of another interpretation: Alvers(o) — Auvers = Anvers (u in Auvers being an error for n)~

Anglo-Norman influence on English place-names 12T

been noted by tbe authors referred to in the Introd. It is dif- ferent where n and r are concerned. Particularly in a medial position these letters are very much alike in some MSS. of that time. For instances of confusion between the two see SCHROER, Anglia, Beiblatt XVIII, p. 97 f., Engl. Stud., 14, p. 249, Ltbl. 1902, p. 288; BERBERICH, Herb. Apul., p. 8; LTTHMANN, La3amou, p. 50. This being so, the appearance of n for r, and vice-versa, might sometimes be explained as a scribal error 1. We must, however, not forget that the change of n to r by dissimilation is well evidenced in O.Fr. For instances see EUREN, p. 24; LIND- STROM, p. 13, Anm. 2; WILLIAMS, p. 23; Jahresbericht der rom. Phil. 1891—94 I, p. 157.

We will now pass on to discuss a few instances that be- long here, which for various reasons have not found a place in the lists. Combined dissimilation and assimilation appears in Erwardeston (1246), Ch.R. = Elvaston (Db.), ^Elvvbldestune, D.B. < O.E. JEthelweald, Searle (I — I > r — I > r — r) 2. Of. dialectal French verocipele (see NYROP, Gramm., p. 406) < velocipele < velocipede (I — d > I — I > r — I).

Apparent loss of n after o in Mold (Flintshire) and Mowbray, pron. (moubre), Tanger (a personal name, see BARDSLEY'S Die.,, p. 544) is, in all probability, only an instance of n having been turned into I.

Of Mowbray we have noted the following early spellings: 6r. de Molbray, Mubray, Mumbray, Munbray, Ch.R. I; N. d& Mumbray, Moubray, H.R; John de Mun-, Mount-, Mou-, Mum- bray, F.A., vol. IV. We take Munbray (= Mod. E. Mumbray, see Bardsley, I.e.) to be the original form, from which arose

1 For that reason we have put between brackets all instances of n for r and r for n in the lists heading this section.

* The same place offers an instance of what BEHEENS (cf. his Uber reciproke Metathese im Roman., Greifswald 1888) calls reciprocal metathesis, viz. Aswalton (14 c.), F.A., for Alwaston. According to Behrens, this pheno- menon is of very frequent occurrence in Italian dialects and, besides, a usual mistake in ordinary speech (French feciliter < feliciter etc.). Cf . also MEEINGEE und MAYEE, Versprechen etc., p. 20 f. English instances are- caliban < cannibal and possibly O.E. eolone < *iluna < inula, see POGATSCHEE. Lehnworte, p, 61.

128 R. E. Zachrisson

Molbray (n — r>l — r), whence, by the vocalization of I to u, Mowbray, Mubray 1.

The Index gives the following old spellings of Mold {Flintshire): Mom altus (1338), Mohaut (1337), Mohaute (1400). The development seems to have been Montalt > *Moltalt (n — I > I — i) > Mohaut. The reappearance of I before t and the transition of It > Id does not offer any difficulties.

In a few cases we are at a loss to decide whether a form is <iue to sound-change or to the substitution of one word for another. Aylwoldesle, Np.S. = Egelweardesle, Geld. Inq., Alwardeslea, D.B., .and Chenwardestone, D.B. = Kinvaston (St.) < O.E. Cynwaldestun (a. 944) (see DUIGNAN, St. Pl.Ns., p. 87) look like instances of I — r > I — /, n — I > n — r, • but, on the other hand, we have noted the same interchange where no other liquid is found in the same word: Aswoldeby, Aswordeby, T.N., Asewardby, Asewaldby, H.R. = Aswardby (Li.). So some confusion between the very usual components -weald, -weard may have been the cause of all the above changes.

Lastly a few instances are well worthy of note in which, &part from the proximity of other liquids, I, n, r are confused in the A.N. spellings of English place-names: n for I appears in Devenis, D.B. = Dewlish (Do.) < O.E. develisc, Birch 525, Deve- lyssh, F.A.; r for I in Sperehou, Spelehou, D.B. = Spelhoe Hun- dred (Np.), Spelhoh (1076), Geld. Inq., Speleho, Np.S., Spdho <1316), F.A.; / for r in Kiemela, Chimela, Anc.Ch. = Keyrner (Ss.) < O.E. mere, Kymere (1316), Index, and Motbilie, -berie, D.B. = Modbury (Dv.X Modbijr, H.R., Modbury, F.A. The last instance strongly recalls a similar interchange in some French words of the same type: concile, concire etc. Cf. EUREN, p. 22.

The ultimate source of the surname is, no doubt, a place in Calva doe called Mombray of which old spellings, like Molbrai (1198), Monbreium (1231), Monibray (1461), are noted in the Die. Topogr. du Calv.

8 From an etymological point of view Mold is identical with a Mod. E. surname Mowatt, pron. (mauat), Tanger, which in the early records ap- pears as Muhaut, Muhalt, Mohaud, MouJiaud, Rievaux Ch.; Montr. Alto, Mou-, Mow-, Muhaut, Ch.R., F.A. n is still preserved in St. Mary Mounthaw, the name of a London parish, the first element of which obviously contains the word >mounlain>, and of which entries occur with n and without: St. Marie Mountenhaute (1428), F.A., 8. Mary Mowthaw, Speculum Britannise, S. M. Mountenhaut (1463), Index.

Anglo-Norman influence on English place-names 129

In fact, if we consider the very nature of the sounds in question, we find another reason why they were so often misun- derstood and misrendered by the Normans. On the one hand, there may have existed a considerable difference between the English liquids and the corresponding French ones, both from the point of view of articulation and in other respects, whilst, on the other, we have strong evidence for believing that the acoustic effect of I, n, r in O.Fr. itself was very much the same. Rhymes of vowel + 1: vowel + n: vowel + f are extremely usual both in A.N. and Continental French. For instances, see MENGER, p. 87, WAHLBERG, Le Bestiaire, p. LV *. Even in literary French there was a hesitation between I and r as late as the 16th cent, and a great uncertainty as to the use of I, n, r is still a characteri- stic feature of Parisian vulgar speech. Cf . NYROP, Gramm. §§ 339, 340, 342. Investigations into the modern French dialects show that the same uncertainty prevails also there. On this point see EUREN, p. 16, who gives numerous references.

We will now proceed to a discussion of those instances in our lists which offer some particular points of interest. A con- siderable number of the anomalous forms seem to have been of a very ephemeral nature, as it would seem, solely due to mis- hearing. In that case they, no doubt, disappeared, very soon after they had been brought into existence.

Association with words usual in the place-nomenclature of Normandy may sometimes have facilitated the change, or even influenced it in a certain direction. Thus Belford (see list 1 a) (==. Barford, Nf.) recalls O.Fr. bel, Ingolvesmera (3 b) (= Ingolds- mells, Li.), O.Fr. mer. Bur gamer a, Bergemere etc. (2 a) (= Bol- mer, Ss.) has, no doubt, been influenced by O.Fr. bourc, bourg (cf. KORNMESSER, p. 43) and berg, cf. JORET, p. 86.

A tendency to avoid difficult consonantal combinations may be traced in the numerous forms glolle, lolle (4 c) for cnolle, as well as in Clotone (5 a) for Knotone (= Knutton St.).

One French form, formerly widely used, has now fallen

1 We have noted some rhymes of this kind in the Chron. of Mousket: Sulie: pcenie, v. 31730; Sulie: chevalrie, v. 36429; more are given by TH. LINK, Uber die Sprache der Chron. von Mousket (Dies. Erlangen 1882), p. 5.

9

130 R. E. Zachrisson

into complete oblivion, viz. Nicol (3 c) for Lincoln 1. The former is peculiar to purely French works, whereas English and gene- rally also Latin works make use of the latter. Thus Robert of Brunne has Lincolne, p. 6, Lyncolnschire, p. 119, against Nicole, Nichole, I, pp. 148, 232, and Nicoleschire, I, p. 484 in Langtoft Numerous instances from O.Fr. chronicles quoted by WESTPHAL, p. 34, only exhibit n. According to Camden, 1, p. 563, the form Nichol occurred many times »in ancient Charters, and even as low as in Edward the fourth's time. William Caxton, in his Chronicle, calls it Nichol» 2.

The French forms have gained the victory only in a few rare instances: Salisbury (la), Salop (la), Lantyan (5 a), Lelant (5 a), and possibly Brightlingsea (1 b).

The first time we have hit upon the dissimilated form of Salisbury, is in the MS. F of the A.S. Chron., so remarkable for its many Norman peculiarities. The two forms existed side by side for a long time. As late as 1403 a certain J. Sarislury is mentioned in the Index, and r still lingers on in the shortened Latin form Sarum. Jones (1701) prefers to spell the name of the town » Sarisbury* , which, however, »is sounded Salsbury*. Cf. EKWALL, Jones's Phonography, p. 72. All O.Fr. works we have consulted favour the dissimilated form: Salesbire, Gaimar, v. 4390; Salesbyry, Langtoft I, p. 66; Salesbires, Guillaume le Mar., v. 1558 etc.

In Salop s for (/) and the svarabhakti between s and I are also due to A.N. influence. Cf. § 3 p. 56 f. This form is used for the town of Shrewsbury till towards the beginning of the 16th cent., and is still a common appellation for the county.

A parallel to Lantyan < Nantyan is offered by several French place-names, where original Celtic nant has been turned into lant. See WILLIAMS, p. 22. Is the dissimilation here due to Celtic sound-development?

1 The intermediate form Nincole appears in Fantosme, v. 1551. Nincole was soon transformed to Nicole, probably on the analogy of the usual per- sonal name Nichol.

3 Prof. Ekwall has kindly pointed out to me the interesting circum- stance that Kingsley, the well-known English novelist, was not ignorant of the existing doublets of Lincoln, as appears from a passage in his Here- ward the Wake, chapter XXI, p. 280: »Lic- nic- cole? replied Ascelin, who, like the rest of the Frenchmen, never could manage to say Lincoln.*

Anglo-Norman influence on English place-names 131

As for Brightlingsea it is doubtful whether such forms as Brytlingseye and Brickelsye are really due to Brictriceseia. It would perhaps be safer to start from an unrecorded O.E. ground- form *BrMlingaeg, the former element of which may have de- noted the whole family to which Brihtric belonged.

Though Nyland (3 c) looks like a dissimilated form of Liland, we prefer to keep the two apart. Obviously both of them go back to an O.E. ground-form Igland, but, while the for- mer has originated by prefixing the French article (cf. Introd., p. 12), we construe the latter as due to the M.E. formula *attenilande *.

It remains to be seen whether some of the cases we have noted may be explained a§ due to dialectal sound-change. Though 7, n, and r are sometimes confused in the English dialects, this confusion is mainly confined to two distinct classes of words: (1) Words which only occur in the dialects, and the etymology of which is often very uncertain; (2) words borrowed from foreign languages, in which interchange may already have existed - this is often the case with the French loan-words - - or else change have set in at the very time of the borrowing, as well as afterwards, owing to ignorance of the word and want of practice in pronouncing it correctly. Here, again, we have to pay due regard to the physiological fact that I, n and r were articulated in a similar manner, so that the acoustic difference between them was very slight. The comparatively small number of words belonging here which are quoted by FRANZMEYER, p. 1—10, and KRUISINGA §§ 328, 330 may, nearly all of them, be referred to one of the above-mentioned groups 1.

NOTE. Some of the instances adduced by Franzmeyer and Kruisinga are to be eliminated, as being not to the point.

1 Cf. the following early forms of Ewhurst (Sr.): La Slefherst (here the French article seems to have been prefixed twice) (13 c.), Index; New- hurst (< *attenewhursf) (13 c.), Index. For more similar instances see DUIG- NAN, Wo. Pl.Ns., p. 113 ff.; St. Pl.Ns., p. 108.

2 To group 1 may belong: bourtree, bultrey, bountree (see E.D.D.); to group 2: calavan < caravan, gualantee < guarantee, paltridge < partridge, oliprance < orprance, cravel < clavel,/raiJ < flail, sy noble < syllable; chander < chalder, all from French; frannel < flannel, from Welsh; keerhauling < keelhauling, from Dutch.

132

R. E. Zachrisson

Nittle for little (see FBAZMEYEB, p. 3) by no means exemplifies a dis- similation of I — / > n — L The form in question, which is often used by grown up people, when speaking to children, has, naturally, originated through an intentional desire to imitate the imperfect language of children Of. Swe- dish »stackars niten>, for >stackars liten», similarly used. Under the same circumstances I is often substituted for the difficult sound of r (cf. Swedish *Helle dud sa lal> for »Herre gud sa rar>\ and this is no doubt the way in which the English pet-name forms Hal, Sal, Mai for Harry, Sara and Mary, as also doll < Dorothy, have arisen, i.e. from being originally used only by the children themselves, such hypocoristic forms have been adopted also by grown-up people.

Emeratid < emerald (FBANZMEYEE, p. 6), of which the former only occurs in the Scotch dialects (cf. the instances in N.E.D. and E.D.D.), is best explained as due to the change of suffixes, the usual Scotch ending and having been substituted for original -aid.

KEUISINGA § 328 points to the dialectal forms armenik and arfabet for almanac, alphabet as instances of I > r. Here r has probably been in- serted after the loss of I, and this is the more likely, as in the Somerset dia- lect we find an inorganic r also in larf <C laugh, pronounced with the same vowel -sound as that of arfabet.

In like manner we would explain the form armere < almonry, Machyn 2544 (a. 1561), cf. DIEHL, Anglia XXIX, p. 194. For more instances of a similar insertion of r, see HOEN, Untersuchungen, p. 61.

At least in one case the change of I > n seems to depend on English development, viz. in the combinations nm, mn. This transition may be explained either as a dissimilation of n to I, caused by the neighbouring m, or as a tendency to replace a consonantal combination of rare occurrence in the language by a more familiar one *. M.E. instances are elmyes, garlementes, remwlant, Paston Letters (cf. NEUMANN, p. 82), remelaunt, rem- lawnt, see BEHRENS, p. 199. In the dialects and in vulgar Eng- lish chimley, omlibus, umlibush (cf. FBANZMETEB, p. 9; STORM, Engl. Phil., p. 823) may be pointed to as illustrating the trans- ition. Here may also belong Standard English martlemass < martmrnass.

In the place-names we have noted the following cases: Kil- mersdou (Sm.) < O.E. Cynemar, Searle, Cenemeresduna, D.B., Kinemeresdon, Ch.R. II, Kynemeresdon (14 c.), P.R.; Wilmots-

In O.Fr. the unusual sound-combination nm in also done away with in various manners; cf. Lat. anima > anme, which latter is changed to ame, aime, and arme. See EUEEN, p. 25; GEAMMONT, p. 56.

Anglo-Norman influence on English place-names 133

ham l (Sm.) < O.E. Winmcer, Searle, Winemeresham, D.B., Wy- nemerham (1284), Wynemersham (1302, 46), F.A.; Kylmerston (1346), F.A. = Kimmerston (Nb.) < O.E. Cynemar, Searle, Kin- merston, Kymerston, F.A.

Present n for older I in Muncaster (Cu.), Mulcastre (1288 — 96), C.R., Mulcastre (1385), Cat. A.D., »Moncaster but its true name is Mulcastre*, M.Br. I, p. 367, hardly arose owing to A.N. in- fluence, our first reference for the n- form being as late as the 14th cent. = Muncastre (1318—23), C.R. 2. The alteration may be due to association with numerous places in Monk-, like Monk- ton etc. < O.E. munuc 3.

According to Hope, Alderwasley (Db.) and Needham (Sf.) are dialectally pronounced »Annerslee», »Allerslee» and »Leedum» Whether this is correct we cannot decide, nor do we venture to suggest any explanation, as we have no instances of early spel- lings of the two places in question.

B. Dissimilation of n to I, r by a following m.

The French forms have prevailed only in the case of Dur- ham (Du.) for which we have found the following early re- ferences.

1. Forms with n: Dunelm, Boldon Book (late 12th cent.); Dunolm, Donelme (1191), Pedes Finium for Du.; Duriholm (1227 -31—34) C.R., Dunolm (1343—46), ib.; Dunelm Civitas, V.E.

2. Forms with r: Durelme, Gairnar, v. 1296; Durealme, Fantosme, v. 534, Dureaume: Uasme, ibid., v. 1603; Dur erne, Dur- ham, Duram, Langtoft 1, p. 414; Durham (1231), Ch.R.; Durham (1279—88), C.R.; Duram, R.G1.; Durham, R.Br., p. 74; Durham, Durem, Duresme (1313 — 18), C.R.; Dorham, Capgrave, p. 145; Duriaume, Durialme, Mousket, vv. 14031, 13311; Dureaume, Du- reiaume, Benoit, vv. 38663, 38387; Durham, (Hen. VII.), I.P.M.;

1 Wilmotsham is a mere fancy-spelling; cf. Headless Cross for Head- ley's Cross in Wo., DUIGNAN, Wo. Pl.Ns. p. 80.

2 The old form Mulcastre still survives in the English surname Mulcaster. See BABDSLEY'S Die., p. 545.

3 It may be that this alteration was arbitrary, caused by the somewhat vulgar sense of the first element wm£=»dust», >rubbish». See E.D.D. For similar reasons Ugley (Es.) has in quite recent times been altered to Oakley. (Bartholomew's Gazetteer gives both forms). Old forms, like Oggele (1303), F.A., Uggele (1428), ib., point to an O.E. ground-form *Ugganleah.

134

R. E. Zachrisson

Duresme (17th c.), see DIEHL, Anglia XXIX, p. 188. - In a record called Registrum palatinura Dunelmense, 1311 — 16(=Chron. and Mem., 62) Durham in the French parts regularly appears as Duresme, in the Latin ones as Dunelm.

The original O.E. form is DunJwlm, A.S. Chr., which in Bede appears as Dunelma < the adjective Dunelmensis, just as Palermo < Panermitanus, adj. derived from Patiormus; cf. MEYER- LUBKE, Einfuhrung in das Stud, der rom. Sprachen, p. 187. Dunelm, BoJdon Book, is possibly a continuation of this Latin form, but may also very well be due to O.E. Dunholm with weakening of unstressed o to e. The form Durelme, Gaimar, has originated by dissimilation of n to r, caused by the follow- ing m.

This kind of dissimilation, which is not quite as usual as the others treated of before, has been noted by GRAMMONT in several cases, of which Dutch Zaandam (name of a town) > French Saardam (see p. 19, p. 25) is the most interesting case, as being quite analogous to Dunelma > Durelme.

Fantosme's Dureaume rhyming with blasme is a regular development of Durelme. Duresme may have originated from Dureaume by the dropping of u and insertion of an inorganic s l.

Though the modern form Durham is, no doubt, due to French Dureaume, Dureme 2, the loss of / before m might easily be accounted for by English sound-development. Indeed, the same phenomenon appears in several cases where A.N. influence is out of the question. Cf. Cuckamslow (Be.) < O.E. Cwicelmes- hlcewe, Index; Brickampton (Gl.) < O.E. Beorhthelm, Searle, Brithel- meton, Brihtamtunn (1200 — 1300), Gl. Recs.; Brighthampton (Ox.) < O.E. Beorhthelm, Bristelmestone, D.B., Buthampton (1350), Frides- wide Ch.; Wulfhem, Byrhthem < Wulfhelm, Beorhthelm, LANGER, Abingdon Chart., p. 59; MyJcames (= Michaelmas) (14 c.), Norfolk Gilds, SCHULTZ, p. 29. According to Jones, I is silent in Dunelm Kenelm, see EKWALL § 577. From the above instances we may

The spelling Duresme for Durem may possibly be due to the ana- logy of the numerous French place-names in -esme < Celtic -isma. Cf. WIL- LIAMS, p. 8.

1 Cf. such usual O.Fr. forms as Willeaume, Willame, Willeme < Wilihelmus.

Anglo-Norman influence on English place-names 135

conclude that English I has a tendency to disappear before m in weakly stressed syllables.

In the 13th and 14th cent, the A.N. form of Durham had gained a complete victory, as it is the only one that appears in the English chronicles of that time (R.GL, R.Br.). All forms with n are now to be considered as mere archaisms used by scribes who had some knowledge of the O.E. form. The archaic spelling Dunholm still occurs on the map in Camden's Britannia.

NOTE. In this place we will consider one more possibility of explain- ing the change of n to r in Durham. Dur < Celtic *duros has perhaps "been substituted for] dun. In French place-names dun and dur are very usual, and the sense of both was »a strongly fortified place*. Cf. WILLIAMS, p. 54, p. 57. Interchange between the two is, however, very rare in the early spellings of place-names. The MS. A of La3amon has Waladunes dune, v. 2823, against Waledures doune in B. The very same place is called Paladur in Wace, v. 1563, palidour in R.Gl., Appendix, p. 781. The French place Roquedur is recorded as Rocdun in 875, as Rocadunum in 1156; see MEYNIER, 1897, p. 361. In all the other cases we have examined, dun and dur are kept strictly apart. As regards Durham, it is, therefore, much more probable that n > r is due to the dissimilatory influence of the follow- ing m than to a mechanical substitution of one etymon (dur) for an- other (dun).

We will add to the above some instances of dissimilation, similar to the one in Durham, which, however, have had no in- fluence upon the modern form of the places in question.

1. n — m>r — m.

Therham, D.B. = Teynham (Ke.) < O.E. Tenham, Kemble 1019; Tenham (12 c.), Index.; Tenham (1316), F.A.

Lerham, D.B. = Lenham (Ke.) < O.E. Leanaham (a. 858), Earle; Lene-, Lenham, Ch.R. I, II, F.A.

2. n — m > r — m.

Winessamstede, D.B. = Willshampsted or Wilsted (BdL); Wyles- hamstede, Tax. Eccl.; Wilshampsted, F.A.

Limet, D.B. = Nimet or Intake (Dv.); Nymet, D.B.; Nimet, R.B., T.N., H.R., F.A.; Nimeth (1216—25), P.R.

136 R. E. Zachrisson

Eglesham, D.B. = Eynsham (Ox.) < O.E. Egonesham, 'A.S. Chr.; Eymsham, T.N., Ch.R. I.

C. Dissimilatory loss of r.

The most interesting instance belonging here is Cambridge (Cb.) < O.E. Grantebrycg, A.S. Chr., which development has already been explained by Prof. SKEAT, Pl.-Ns. of Cb., p. 30, as due to A.N. influence. »The use of C for G-r arose from a Norman mispronunciation; the dropping of r, in particular, is clearly due to a wish to avoid the use of gr and br in the same word.»

Prof. Skeat has recorded so many old forms that we need only supplement his list by adding a few early instances from the I.C.C. and the I.EL: Grenteburga, Grantebrigge, Granteburga, I.C.C.; Grantebrige, Grantebrugge, Grantebrice, Cantebrigia, I.EL

The English forms existed for a long time by the side of the Anglo-Norman ones, and do not seem to have been definitely lost until the beginning of the 15th cent.

Dissimilatory loss of r is found as early as in D.B.: Kachene- foda =.Rackenford (Dv.); Hardes = Hardres (Ke.), Netherardres, Ch.R. I, Hardres, T. N., (1316, 1346, 1431), F.A.; Stagrigesholle = Stramshall (St.), Strangricheshall, -hull (13 c.), see DUIGNAN, St. Pl.-Ns., p. 145; as also in various other records: Bricthicesfel (13 c.), Index = Brushfield (Db.), Britreichfeld (13 c.), Index; Tor- teoda, Fr.Ch. = Tortworth (GL); Chelysforde (1304), Index = Chels- wortb] (Sf.) < O.E. Ceorles wyrfre (a. 962), Index.; Sapy Pychar (1519), Index = Sapey Pritchard (Wo.).

It is, however, not absolutely certain that loss of r in all the above instances depends on A.N. influence. Though the phenomenon in question was a usual one in O.Fr. (cf. NYROP, Gramm. § 513) it is well represented also in English itself. Cf. HORN, Archiv 115, p. 324, Gramm. § 237. That, at least with reference to Grantebrycg, the disappearance of r is due to A.N. influence is rendered very probable by the circumstance that the present form and pronunciation 6f the place exhibit other French peculiarities, viz. (ei) for (se) (cf. § 9) and possibly also c for g.

Anglo-Norman influence on English place-names 137

NOTE. We do not think that Prof. LUICK has succeeded in giving a satisfactory explanation of the appearance of c forg. Cf. Anglia XVI, p. 489 >der ersatz des g durch k kann daher kommen dass im Agn. anlautend die folge ga- (die streng von gua- zu scbeiden ist) lautgesetzlich nicht moglich war und auch thatsachlich kaum vorkam, wahrend ca zwar auch lautgesetzlich unmog- lich, aber nach den englischen reflexen zu urteilen doch nicht selten war». To this we may object that ca was the regular representative of Lat. ca in words like camp etc. in the North of Normandy (cf. § 1 p. 23); ga was com- mon to the various O.Fr. dialects in some Germanic loan-words of late intro- duction, gdber etc. To account for the change we have, no doubt, to look elsewhere. In the first place, it is worthy of note that k and g are inter: changeable in O.Fr. itself. For instances, see NYEOP, Gramm. §§ 399, 405, 409, 423, 434; MACKEL, pp. 139, 140, 145; BEHKENS, pp. 202, 203, 207. One more circumstance is to be taken into consideration. We know there is nowadays a marked difference between English k, g and the corresponding French sounds, and such a difference may have existed even at the time of the Conquest. Consequently, k for g, and vice-versa, may sometimes de- pend on mishearing. We have collected some material, calculated to illustrate the interchange between k and g which often appears in the early spellings of English place-names. The instances are classified according to the sound — consonant or vowel — that follows after c, g. All of them can hardly be interpreted as orthographic misrepresentations. (C and G are much alike in the MSS).

Interchange of 1. cr, gr:

Crauenhest, D.B., Cravenhurst (1377), Index, Graven- Gravehurst, F.A., T.N. = Gravenhurst (Bd-); Gramborne, D.B., Cramburne (1346, 1428), F.A. = Crambourne (Ha.); Granewrthe, H.K., Cranewrth (13 c.), Craneworth (14 c.), Index = Cranworth (Nf.); Gresinga-, Cressingham, D.B., Cressingham, H.R., F.A. = Cressingham (Nf.); Grenedona, D.B., Crandone, Sm. K.Q. = Crandon (8m.); Granforde, Craneford, R.B. = Cransford (Sf.); Gretinge, Cretinge, R.B., G-ratinges, Fr.Ch. = Greeting (Sf.); Gretingeham, D.B., LEI., Cretingham (1219), Sf. P.F., Gretingham, Ch.R. I, II, T.N., (1307, 33) Index, Cretyngham (1461-67), P.R. = Cretingham (SI); Crigeston, D.B., Grigelstone (1327), Yo. S.R. = Crig- glestone (Yo.).

2. cl, gl:

Glauton (1284), F.A., Clauton (1303), ib. = Clawton (Dv.); Cloecistra, Cleucestra, A.$. Chr. MS. i, Claucestre, Claucetre, R.G1., v. 1517 = Gloucester il.), cf., however § 5, p. 78, foot-note 1; Gliderhov (1232), Index = Clitherhoe i.)1; Clenyndon (1237), Index, Clendon, F.A. = Glendon (Np.); Clachestorp, ).B. = Glassthorpe (Np.); Glanefeld, T.N., Clane-, Clanfeld, F.A. = Clanfield

Glitherhoe still exists as a personal name; see BARDSLEY'see Die., p. 321.

138 R. E. Zachrisson

(Ox.); Glademutha, Henry of Hun. (Mon. Brit.) = the mouth of the river Cleddau (Pembroke), Cledemupa, A.S. Chr.

3. en, gn\

. Gnatishale (1230—40), Gnateshale (Edw. I.), Index = Knettishall (Sf.); Gnolla (1200), Index = Knolle (Wa.); Gnareburc (1191), Yo. P.F., Gnareburg (1176—77), Index, Gnaresburg (1272—79), C.R. = Knaresborough (Yo.).

4. ca, ga:

Gannok vel Cannok (1201—16), P.R. = Cannok (St.); Camelegeia (1152—53), Index = Gamlingay (Ob.).

5. Ice, ge: Chevelestune, D.B., Geuelston, F.A. = Gelston (Li.).

6. gi, lei:

Gildecrosse, Childecrosse, Inq.El. = Guiltcross Hundred (Nf.); Kildincham, D.B., Gilyngham, F.A. = Gillingham (Nf.).

7. co, go:

Godrintona, D.B., Codcrington (1284, 1428), F.A., Godryngton (1303, 46)r F.A. = Goodrington (Dv.); Goldhangra, D.B., Caldhangre (13 c.), Index, Gold- hengre, H.R., T.N. = Goldhanger (Es.); Golchestre, R.G1V v. 38 = Colchester (Es.); Godbritleye, Codbricleye (1284), F.A. = Coberley (Gl.); Godrinton, D.B., spelled with G in the F.A. (1303, 16, 46), with C in the Index (1189), the H.R, the F.A. (1303), the Gl.Recs. (1465) = Codrington (Gl.); Coxhill (1467— 77), P.R. = Goxhill (Li.); Gossale (1284), F.A. = Cossal (Nt.). etc.

Many more instances of this interchange have been recorded by BARDSLEY. See his Dictionary, p. 19, and under Chrimes, p. 180, Cladish, p. 183, Carbutt, p. 159.

D. Interchange of -ling, -ning, -ring.

1. In the middle of the word.

a. ring > ling, ning.

Telingham, Tedlingham (d is excrescent), D.B.; Tiningon, Fr.Ch. = Ty- ringham (Bu.); Tyringeham, T.N., H.R.; Tyringham, F.A.

Gothelyngton (1316), F.A. = Goodrington (Dv.); Godrintona, D.B.; God- ryngton (1303), F.A.; Gotheryngton (1414), Index.

Anglo-Norman influence on English place-names 139

Cudelintona (1189), Index = Codrington (Gl.) ; Godrin ton, D.B.; Goderynton Coderinton, F.A.

Walintune, D.B. = Warrington (La.); Weryn-, Weryngton (1200—1350), La. Pi.E.; Werirtgton (1303, 41), F.A. Wcer- as the first element of a personal name is very usual in Searle's Onomasticon.

Cameling ehatn, D.B. = Cammeringham (Li.); Cameryngham, Li.S.; Camer- ingham, H.R.; Cameryngham (1316), F.A.

Coclyngton, F.A. = Cockerington (Li.); Cokerinc-, -ring-, -inton, H.R.

Simplingham, Rievaux Ch., Semplingham, H.R. = Sempringham (Li.); Sempringham (c. 1100), Sempringham Ch.; r after p looks like a late insertion, for many of the earliest forms occur without r, viz. Sempigaham, A.S. Chron., Sempingham (cv1100), Sempr. Oh.

Silingeham, D.B. = Sherringham (Nf.); Shiryngham, F.A.; Shyringham (1379), Index.

Felinghefort, Fr.Ch. = Fringford (Ox.); Ferryngforde (14 c., 15 c.), F.A.

Tedelintone, D.B. = Tytherton (Wl.); Tuderington (c. 1300), I. P.M.

b. ling > ning, ring.

Gameningeye, R.B. = Gamlingay (Cb.) <c O.E. Gamel, Gamela, Searle ; Gamelingei, D.B.; Gamelingeye, R.B.

Greningewell, Ch.R. I = Graylingwell (Ss.); Greylingeivell, Ch.R. I.

Boniton, Bolintun, Bollintune, D.B.; Boniton, Volintona, Fr.Ch. = Bul- lingtons in Bexhill (Ss.).

c. ning > ling, ring.

Herlingtone (1087), Fr.Ch.; Erlingetona, I.C.C. = Arrington (Cb.) < O.E. earn, Sweet, Die.; Erningtune, D.B.; Erningetone, cerningetune, LEI. 1.

Belingehou, D.B. = Bengeo (Ht.) < O.E. Benning, Mdf., p. 12; Beningeho, R.B.; Beningho, Tax. Eccl.

Rauelinc-, Raverincham, D.B. = Raveningham (Nf.) < Reef en, Rauan, Searle; Ravenicham, D.B.; Ravenyng-, Ravyngham, F.A.

Siverinctun, D B.; Sevelyngton (1284), Yo. K.Q.; Sivillington, Ch.R. I = Sinnington (Yo.); Sevenictun, D.B.; Siniton, H.R.; Synnyngton (15 c.), P.R.

1 Loss of n in the present form of the word may have been caused either by dissimilation or by the substitution of -ing for ling. We have noted a similar interchange between -ing and -ling in Barming (Ke.), Berme- linge, D.B., Barmling, T.N., Barmlinge (1316, 46), F.A., Barmyng (1556), Index. Cf. also the 16th century form Fremyngham, Machyn 705 anm., see DIEHL, Anglia XXIX, p. 1 93 = Framlingham (Sf.), pron. »Frammingham», Hope.

140 R- E. Zachrisson

2. At the end of the word. a. ring > ling.

Wachelingam, D.B. = Wakering (Es.); Waclieringa, D.B.; Wakeryng, H.R., T.N.; Wakering, -es (1251, 52), Ch.R.; Watering (1279-88), C.R.

Clauelinga, D.B. = Clavering (Es.) < O.E. Clcefring, Kemble; Clavering, -e, H.R., T.N., R.B., F.A.

HoMinka, D.B. = Hockering (Nf.); Hokering, (14*c., 15 c.), F.A.

b. ling > ning, ring.

Haringeia, Fr.Ch. = Hayling (Ha.); Heglingaig, Heilincigce, Codex. Vint. (MS. 12 c.), see Anglia, XXV, p. 435; Haylyng (1316), F.A.

Diceninges, D.B.; Dychening ', H.R.; DicJienyng (1411), Ss. S.R. = Ditch- ling (Ss.) < O.E. Diccdingas, Kemble 314, Dicelinges, U.B.

Graveninge, R.B. = Groveley (Wl.); Gravelinges, D.B., R.B.; Graveling, -ling', T.N., H.R., Ch.R. II.

c. ning > ling.

Avelingnes (1246), Ch.R.; Ave-, Havelinges, Fr.Ch. = Avening (Gl.) < O.E. jfifeningas, cf. TAYLOK, Names, p. 354.

heselinge, LEI.; Ixelinge (1279—88), C.R. = Exning (Sf.); hexeninge, I.E1.; Exning', H.R., Exninge (1272—79), C.R.

It is true that many of the above instances might be explained as due to the change of endings, particularly as in most cases •ning and -ring have been replaced by the more familiar -ling. There is, however, one circumstance which decidedly tells in favour of A.N. influence, viz. that the irregular forms are mainly confined to D.B. and other early records, like Fr.Ch., I.EL, I.C.C., R.B., while the normal and present forms regularly reappear in MSS. of a later date, which in a much lesser degree are affected by this influence.

Besides, a similar interchange occurs also in French place- names. Cf. Frasnines (1101) = Fralignes (Aube), see MEYNIER, 1898, p. 141; Pons Berengarii (1203), Pons Bellengerii (1278) = Pont-Bellenger (Calvados); Berengreville (1418), Bellangreville (1454) = Bellengreville (Calv.), see Die. Topogr. du Calvados. Cf. also the various O.Fr. forms of galingal mentioned below.

Our lists exhibit no instance of the unetymological forms having prevailed.

On the other hand, we have noted one case where at a late date -ning was substituted for -ling, viz. Coney swick or Connings-

Anglo-Norman influence on English place-names 141

wick (Wo.), Colingwic, D.B., Collingwike (1275), Colyngwyk (1327). Cf. DUIGNAN, Wo. Pl.Ns., p. 41. The first reference for n given by Duignan is Conisweeke (1603). Here the change may be due to some popular etymology.

An interchange of -ling, -ning, ring is, however, not un- known in the English dialects, not only where French loan-words are concerned \ but also in native words : reckoling < reckoning, yeaveling, eveling < evening (see KRUISINGA § 366; FRANZMEYEK, p. 8), pollinger < porringer r at least is here due to native

development - - in which -ning, -ring has been replaced by -ling. It is more astonishing that the less common -ring should some- times appear for -ning, -ling, as in everin < evening (see FRANZ- MEYER, p, 8), and, if Hope is to be trusted, in Hellingly (Ss.), dialectally pronounced »Herringlye», »Hellinglye».

E. Interchange of unstressed le (el), ne (en), re (er).

1. In the middle of the word.

a. n > r, I.

Witrehame, Fr.Ch. = Wytham (Be.) < O.E. Witanham, cf. Mdf., p. 152.

Hedrehan (1088), Index; Hadrehan, Ch.R. = Haddenham (Bu.) < O.E. Headda, Searle; Hadenam, D.B.; Hedenham, H.R.; Haddenham, F.A.

Beccheretona, Fr.Ch. = Beachanipton (Bu.) < O.E. -hantun = ham -\- tun; Bechantona, Fr.Ch.

Wandrie, D.B. = Wendy (Cb.) < O.E. * Wendanige (cf . SKEAT, Pl.-Ns. of Cb. p. 55); Wendye (1316), Wendeye (1346), F.A.

Hadreham, D.B.; Haddreham, Hcederham, I. El. = Haddenham (Cb.) < O.E. Headda; Hadenham, I.E1.; Hadenham (c. 1300), Cat. A.D.

Sidreham, Sidelham, D.B. — Sydenham (Dv.); cf. O.E. Sidanham, Birch 759; Sydeham, H.R., T.N.; Sydynham, F.A.

Wigrehala, D.B. = Wiggenhall (Nf.) < O.E. Wicga, Wigga, Searle; Wygenhal, F.A.

Mundreham, D.B.; Mundre-, Mondreham, Fr.Ch. = Mundham (Ss.) •< O.E. Mundan ham, Index.

Ackiltone, D.B. = Eckington (Ss.) < O.E. Ecca, Searle ; Echentone, D.B.; Echinton, T.N.

1 Cf. Mod. Engl. galingale, M.E. ganyngale, galyngale (Matzner) corre- sponding to O.Fr. garingal, galingal, ganigal (Godefroy) < Arab, khaulinjdn N.E.D.).

142 R- E. Zacbrisson

b. I > r.

Citremetona, D.B. = Chittlehampton (Dv.) < O.E. Cytel, Searle (cf. also cytel, Mdf., p. 37); Chitelhamton, F.A.

Tistertunc, D.B. < O.E. fistel, Mdf., p. 139; Tisteltone, D.B.

Wifretorp, D.B. = Wilsthorpe (Yo.) < O.E. wt/rf, Mdf., p. 150; Wifles- torp, D.B.

Ampreford, D.B. = Ampleforth (Yo.); Ambleford, D.B.; Amplesford, T.N.; Ampelford, Ch.R. I, Yo. K.Q.

> Z.

Liuelmera, D.B. = Levermere (Sf.); Liuermera, D.B.; Leuremere, Inq.El. Amblesberie, D.B. = Amesbury (Wl.) < O.E. Ambresbyri, A.S. Chr.; ,4m bresberie, D.B.; Ambresbur, H.R., T.N.

2. JL# £fte ewd o/ #/i a. w > r.

, D.B.; Auren, La3amon, v. 2498 = the river Avon < O.E. Afen, Afene-, Afenan-, A.S. Chr.

Bucfastra, D.B.; Bufestre, T.N.; Bufestre, Ch.E. II = Buekfast (Dv.); Bocfasta(l3n— 81), P.R.; Bukfast (1318— 23) O.K.; Bucfaste, Bukfast (Hen. VII.), I.P.M. Bucfastra, D.B. etc., is in our opinion best explained from a M.E. form Bucfasten, -ne, where -fasten represents O.E. fcesten. In fact, this very form appears in Bugivastene (1292), Index.

Nigravre, D.B. = Nether Avon (Wl.); Netheraven, H.R.

Oppavrene, D.B. = Upavon (WL); Uphavene, H.R.

Upelider (er <; en <C an, weakened form of aw often = ham in the A.N. recs.), D.B. = Upleatham (Yo.); Lethum, Lithum, Yo. K.Q., Yo. N.V.

b. l>r.

Tanbre A, T... rnb-e B, La3amon, v. 28532 = Camelford (Co.), Camelford> La3amon, v. 28534 < Britain Camlan, cf. MADDEN, p. 408, where more old forms are noted.

Pidre, Pidere, Pidrie, Pitretone, Piretone, D.B. = Puddle, Puddletown (Do.)< O.E. Pydele, Uppidelen, Kemble 656, 522; Pidele, D.B.; Pidelc, Pudele, Pedyll, Pudull, F.A.

Filunger, D.B. = Fillongley (Wa.); Fillmgle, Ch.R. I;Fylingele, H.R. < O.E. leah.

Anglo-Norman influence on English place-names 143

C. r > I.

Aissele, Aissela, D.B. = Esher (Sr.); Esshere, Essere, Kemble 812, 988; Assere, D.B.

As for the above forms A.N. influence is hardly to be questioned when we consider the time and place of their appea- rance. Moreover, most instances cannot well be explained by the aid of English sound-laws, whereas numerous parallel cases may be pointed to in O.Fr. In the majority of cases unstressed ne, le has been changed to re, more seldom the other way.

In O.Fr. the transition of unstressed ne to re is very usual particularly after the consonants m, p, f, v, d, t,

and the palatals — not only in words of Latin origin (joevres- < junenis etc.) but also in the Germanic loan-words (havre < hcefene). Cf. EUREN, p. 26 1; MENGER, p. 85; MACKEL, p. 63, This agrees well with such instances as Witrehame, Hedrehan, and others brought together in 1 a and 2 a.

We believe GRAMMONT, p. 139, is right in considering this whole transition as the work of analogy. The combination cons. + ne was unusual, and was, therefore, replaced by the more familiar one of cons. + re, more seldom le, as in havle < hcefene. With this latter we may compare Sidelham (1 a) by the side of Sidreham (ib.).

The" French form Londres < Londinum (cf. Forbiger III, p. 283) = London, which also belongs here, we have not found except in purely French works. On the other hand, Lundreis as the denomination of the inhabitants of London is used by R.Br., p. 120. The English equivalent is Lundenisc, A.S.Chr.

Also unstressed le passes into re in O.Fr. after certain con- sonants (dentals, labials, gutturals; after t the transition seems to be regular). Cf. EUREN, p. 21 if.; NYROP, Gramm. § 341. Double forms existed of a not inconsiderable number of words: title, litre; escandle, escandre; idle, idre (< idolum) etc. It is obviously this O.Fr. sound-change that has operated in Wifre- torp, Pidre, and other instances noted in 1 b and 2 b. The occa- sional appearance of / for r (1 c, 2 c) may be due to the ana- logy of the above-mentioned French doublets: title, litre etc.

As for the explanation of the transition, we are once more ready to subscribe to the view of GRAMMONT, p. 130, who has

144 R. E. Zachrisson

characterized it as a process of analogical transformation in the manner of the many words in cons. + re already existing in the language. Acoustic resemblance between I and r, and the lack of distinctness in articulation which must, necessarily, have resulted from the absence of stress have, no doubt, facilitated the change.

A similar interchange of unstressed I, n, r is of rare occur- rence in English. At the utmost, a few cases due to the change of suffixes may be pointed to: M.E. dronkele for droncnen, nemele for nemnen, scorhle for scorlmen, brutle for brutne. See KLUGE, Paul's Gr., p. 1022. A few more instances from the English dialects are adduced by FEANZMEYER, p. 5.

A number of cases, where in the spelling of place-names older le has been replaced by modern re, ne, and conversely, or where interchange appears in later forms unaffected by A.N. influence call for various explanations.

In the following instances the change may be due to po- pular etymology: Childerditch (Es.), Ciltendis, D.B., Chiltendich, T.N., Chiltendich, -dik, H.R. Forms with r do not appear until the 14th cent. (Childerdyche, F.A.), and have probably arisen by association with childer, a usual dialectal plural form of »child»; see E.D.Gr. § 380. - In all early spellings Londonthorpe (Li.) appears without n, as in Lundetorp, D.B., Lundthorp (1227), Ch.R., Lunderthorp, H.R., T.N., F.A. etc. London Thorpe, M.Br., is the first reference we have found for the modern form, which seems to be due to the analogy of London.

We will now draw attention to another procedure by which many cases of the late interchange in question may have origi- nated. It often happens that under circumstances not yet fully cleared up an unstressed I, n or r is dropped in M.E. Ortho- graphy is, however, particularly where place-names are concerned, very conservative, and the scribes, in their vain attempts to pro- duce what in their opinion was the etymologically correct form, may often have substituted another consonant for the one which was already lost in pronunciation. For want of a -better name we will call this method of procedure false reconstructive spel- ling. It goes without saying that such a spelling sometimes has called forth a new pronunciation. An instance or two may be adduced to illustrate our theory.

Anglo-Norman influence on English place-names 145

Loss of I in Prittlewell (Es.) (cf. Pritewell, 1303, F.A.) gives rise to graphic uncertainty, both r and n being introduced as substitutes for I: Priterwell, F.A., Preterwell (1467—77), P.R, Putenewell (1267), Ch.R,. - I is lost also in Chivelston (Dv.), whereupon r appears by the side of I in the spelling, Cheverston (1409), Index, Chyvelston, F.A.

In the above-mentioned two cases the erroneous forms have had no effect upon the modern spelling and pronunciation. This has been the case, however, with the following places. Farmington (Gl.) certainly goes back to an O.E. *furmundtun (cf. old forms like Tormentone, D.B., and Thormanton, K.Q., 1284). Such a form as Thormanton has undergone a twofold development: (1) n is kept; en > in > ing = Farmington; (2) n' is lost, and replaced by unetymological r = Thormerton (1303, 16), F.A. The place is still called by both these names, see BARTH. - - The oldest spellings we have found of Thurlton (Nf.) are Thuruerton, D.B., Thurvertun (Hen. III.), Index, r is dropped (cf. Thurweton, 1384, Index), and I is now introduced instead of r. The dialectal pronunciation seems to ignore this /, to judge by the form Thurton on the map of Nf. in Camden. - - Hilderston (St.) < O.E. Hildewulf (cf. Heldulvestune, D.B.), and Kidderminster (Wo.) < older Kydelminster also belong here. For numerous references to old forms, see DUIGNAN, St. Pl.Ns.. p. 79; Wo. Pl.Ns., p. 96. - Ashmansworth (Ha.) and Rickmausworth (Ht.) are two more instances of false reconstructive spelling. That n is unoriginal is evident from such early spellings as jEscmeres weorf, Earle, Asshmeresworth (1316), F.A.; Prichemare-, Ricemareworde, D.B., RiJcemers-, Rykemerysworth, F.A.

Both popular etymology (association with » straggle ») and false reconstruction (we have noted a form without r as early as in the T.N.) may have caused the change of r to I in Straggle- thorp (Li.), all the oldest forms of which (Strager-, Stragar-, H.R., T.N., F.A. etc.) unambiguously point to original r. Our first forms with I are Stragelthorp, Shragulthorpe (1346, 1401), F.A. Worston (St.) also admits of this double explanation.

The etymology is *Wig/rithestun (cf. Wyfridestone, 13 c., DUIG- NAN, Pl.Ns. of St., p. 176). Later forms with I ( Wiveleston, 13 c. etc.) may be due to association with O.E, wifel, or false insertion of 7 as a substitute for r.

10

146 R. E. Zaehrisson

§ 8. / before a consonant is vocalized or lost.

A. I is vocalized *. 1. After a Jrom various O.E. sources.

1. Stotfold (Bd.) < O.E. fald, falod, Mdf., p. 49. 2. Wyboston (Bd.)

< O.E. Wigbeald, Searle. 3. Aldwark (Db.) < O.E. <?aM. 4. Faulkbourne (Es.)

< O.E. fealcen, Mdf., p. 50. 5. Waltham (Es.) < O.E. Wealtham, Kemble 608. 6. Oldbury upon Severn (Gl.) < O.E. eald. 7. Waltham (Ha.). 8. Calbourne (Ha.). 9. Harbledown (Ke.) < O.E. Herebeald, Searle. 1O. Aldington (Ke.) 11. Mablethorp (Li.), perhaps from an unrecorded O.E. *Mcelbeorht. 12. Cux- wold (Li.) < O.E. weald, Mdf., p. 146. 13. Alford (Li.). 14. Walsingham (Nf.) < O.E. Wcplsingaham (TAYLOR, Names, p. 353). 15. Scaldwell (Np.), 16. Walkeringham (Nt.). 17. Shalford (Sr.) < O.E. sccalde, Mdf., p. 111. 18. Cowfold (Ss.) < O.E. fald.

D.B. Cauborne*, Audintone10, Aldintone10.

12th c. — Index. Maubertorp" (1191).

13th c. — C.B. Wautham5 (72—79). Ch.B. Wautlwm*. F.A. Aldinton10 (1284). H.B. Walt-, Wautham*. Index. Stotfaud1, Wibaudeston* (1208), Audewerk* (1289), Mauberthorp11 (1263), Malberfhorp" (1256), Auford1* (1220), Wausingham1* (1262), Scaudewell16, Waucringham1*, Scaudeford11, Cufaude1*. T.N. Gukewaud, -waldl\

14th c. — F.A. Faukeburne4, Falkeburn4 (1303), Kokewald™. Index. Aide- teerk* (1309), Audebyre* (1301), Wautham'1 (1304), Schaldewell15 (1384).

15th c. — Index. Wybaldeston* (1444). P.B. Harbaldowne9, Herebaudon9.

2. ^L/^r e from various O.E. sources 2. a.

1. Bradfield (Be.) < O.E. cet Bradanfelda (c. 990), Index. 2. Shef- field (Be.). 3. Hatfield Peverel (Es.). 4. Kelfield (Li.). 5. Smithfield (Li.). 6. Catfield (Nf.). 7. Whitfield (Np.). 8. Fressingfield (Sf.). 9. Stanning- field (Sf.). 10. Waldingfield (Sf.). 11. Westerfield (Sf.). 12. Whatfield (Sf.). 13. Ifield (Ss.). 14. Uckfield (Ss.). 15. Bradfield (Wl.). 16. Wake- field (Yo.). 17. Sheffield (Yo.).

1 2th c. — Index. Smetfeud*, Smethefeld5, Withes/eld1 (c. 1150). 1 3th c. — Index. Bradefeud1 (1277), Bradefeld1, Scheufcud* (1284), Scefeld* (1202), Hathfeud\ Kelkefeud* (1281), Catefeld* (1297), Whitefeud1,

1 We have arranged our material according to the quality of the vowel that precedes I.

* We will first adduce some instances of place-names containing O.E. feld; then proceed to deal with some other examples.

Anglo-Norman influence on English place-names 147

Fresingfeud*, Fresingfeld*, Waudingfeud™, Wallynfeud™, Wal- dingfeld™, Westerfeud11 (1287J, Watefeud™, Ifeud1*. T.N Brade- feud, -fel\ .feldl'». Yo. S.B, Wakefeld, -feud", 8ehefeW\ Scefeud" (1297).

1 4th c. - Index. Catefeud* (1308), Stanefeld, -feud9, Waldingefeld -feud10 WatfeW (1398), Ifeld" (1316), Ifeude™ (1352), OckefeU" (c. 1306)' Ukkefeud" (1322), Wakcfeud1' (1303).

b. In other cases.

1 . Elstow (Bd.), perhaps < O.E. ellen, Mdf., p. 45 \ 2. Elmton (Db.)

< O.E. elm, Mdf, p. 45. 3. Chelborough (Do.) < O.E. Cheolesburh, Kemble 366, 412, 454. 4. Elmstead (Es.)< O.E. elm 2. 5. Chawton (Ha.). Either

C O.E. Ceol, Searle (cf. also ceole, Mdf., p. 26) or Ceawa, ibid., in which lat- ter case el (D.B.) would be an inverted spelling for au. The latter alternative is perhaps the more probable one when we consider that in no other case have the A.N. forms prevailed. 6. Hemswell (Li.) < O.E. Helm, Searle. 7. Belchford (Li.) 8. 8. Feltwell (Nf.). 9. Elmsett (Sf.) < O.E. Ylmesceton (10 c.), Index. 1 0. Bexhill (Ss.), cf . O.E. Becca, Searle. 1 1 . Elmley Castle (Wo.)

< O.E. Elmlege, Birch 235. 1 2. Helmsley (Yo.) < O.E. Helm, Searle.

D.B. Elnestou\ Celberge*, Elm-, Almestedam*, Celtone*, Helmes- welle6, Beltesford\ Feltwella*, Bexelei10, Elmeslac™.

1 2th c. — Index. Fletuuella* (1147), Fautewelle* (1162), Feute-, Fauteivell* (1179). Li.S. Helmes-, Halmesuuella*.

1 3th c. — Annales Mon. Elm-, Alne-, Aunestoive*. C.B. Chauton* (79 -88), Chaul-, Chauton* (72—79), Chauton* (79—88). Ch.R. Aume- ton* I, Elmesette9, Aumesete9 I, Aumesete, -scete9 II, Aumeleghiz II. H.R. Elmestede, -stade*, Belcheford1, Byxle™. Index. Elnestou1 (1203—06), Elmestede4' (1243), Beudeford1 (1247), Beltesford1 (1297), Feuteuuelle*, Felteivell* (1272). P.B. Hamelac1* (25—32). Bievaux Ch. Helmesleia™, Helmeslac1*, Halmelac12, Hamelak^, Haumeslac™,

1 SKEAT, Pl.-Ns. of Bd., derives the first compound of this place from O.E. Mihdnoih ; we consider a derivation from O.E. ellen as equally probable,, and account for the forms with «, au as due to A.N. sound-change.

2 STOLZE, p. 15 Anm. 2, thinks the D.B. form aim- is due to Scand. almr, aim. In fact, this word occurs in M.E. in the compound alme-barke. See HalliwelFs Die6. (1904), p. 47. Forms like Elmsted being in such a great majority, Aim- is probably best explained as an A.N. form.

8 The etymology of this word is obscure. 1 think we are on the safe side in considering the combination -lt(e)s- to be the original one. Modern ch is perhaps to be compared with >linchpin», »henchman> < linspin, hensman, which are looked upon by HORN, Gutturallaute, p. 64, as >hyperliterary> forms. If this conjecture is right, we have to start from such a form as Belseford, whence Belcheford.

148 R. E. Zachrisson

Hamelenc™, Hamylac™ S.B. Elmeleye11 (1275) (see DUIGXAN, Wo.

Pl.Ns., p. 58). Ss. Becords. Beause10 (see Viet. Hist, of 8s.,

p. 537). T.N. Chaub'ge', Elmstede*, Cheleton5, Chauton* , Beautes-

ford\ Yo. K.Q. Hammelec™, Haunelec1*. 1 4th c. — O.B. Beltesford1 (81—85), Hamelak1* (13—18, 43—46—49), Hel-

mesleye™ (46—49). Cat. A.D. Beltesford1 (1361, 62), Hamelak1*.

F.A. Chawberge*, Schaiveberge*, Chelberg*, Chelbury*, Cheldeborowes,

Elmestede\ Helmeswell", Beltesford1. P.B. Estchelberewe*(43—4S),

Beltesford'' (81—85), Bexle10 (30—34), Hamelak1* (43—45—48).

Yo. N.V. Helmeslegh™. 1 5th c. — F.A. Belseford1 (1401). I.P.M. Helmesley1* (Hen. VII.). P.B.

Belcheford1 (67-77).

3. After o < O.E. o.

Bircholt (Ke.) < O.E. holt, Mdf., p. 75. D.B. Beriscolt, Berisout. Ke. P.P. Biricholt (1204).

B. / 1st lost.

1. After a from various O.E. sources.

\ . Smallridge (Dv.) < O.E. smcel, Mdf., p. 118. 2. Alconbury (Hu.) < O.E. Alkmund, Searle. 3. Palgrave (Nf.) < O.E. Palegrave (a. 962), Birch 1084. 4. Scaldwell (Np.). 5. Halnaker (Ss.). 6. Calne (Wl.). 7. Helmsley (Yo). < O.E. Helm, Searle.

D.B. Esmartge1, Esmaurige1, Acumesberie* (um = mu), Scalde-,

Scadewell*, Canna6.

1 2th c. — Fr.Ch. Pagava?, Halnac*, Hannah. 13th c. — C.B. Hamelak\ H.B. AlkmundeUr\ Calna*, Calne9. P.B. Hame-

lak\ Bievaux Ch. Hamelak1, Hamylac1.

2. After e from various O.E. sources.

I . Silverley (Cb.) < O.E. seolfer, silfer, Sweet, A.S. Die. 2. Shilving- ton (Do.) < O.E. scylf, Mdf., p. 114. 3. Canfield (Es.) < O.E. feld.

D.B. Severlai1, Sevemeton*, Scilfemetune* .

1 2th c. — I.O.C. Seuerlai1. Inq. El. 8euerlaio\ Seilverleia1. Index. Kanefed*. 1 3th c. — Index. Childeskanefeld* (1285). 1 4th c. — F.A. Silverle1 (1302), Schilfamtun*, Shulfhampton\ Shilvington*,

Canefeld*.

3. After o < O.E. o.

Rolvenden Hundred (Ke.) < Rolf, Searle. D,B. Rovinden. F.A. Rol- vynden. T.N. Rolvendenn.

Anglo-Norman influence on English place-names 149

4. After i < O.R i.

1 . Wilbraham (Cb.) < O.E. Wilbur geham, Birch 630. 2. Wymondley (Ht.) < O.E. Wilmund, Searle. 3. Wilting (Ss.).

D.B. Wiborgham1, Wimundeslai*, Witinges3.

1 2th c. — I.C.C. Wilburgeham1. Inq. El. Wilbergeham1, Wiborgeham1. In- dex. Wilburcham1 (1150).

13th c. — H.B. Wilmundele*. R.B. Willemundele*. T.N. Wilemundeslea*.

1 4th c. — Index. Wylburham1, Wilburgham1 .

5. 4/ter U < O.E. U.

Bulkington (Wa.), cf. O.E. Bulcan pyt, Kemble 133. D.B. Bochintone. Index. Bulkinton (1232).

It is a well-known fact that I before a consonant was gene- rally kept in M.E. It was hardly until late M.E. or early N.E. times that a glide u developed before I in the groups al, 61, gl. At about the same time loss of I is evidenced. This loss, which in some dialects has a much larger extension than in Standard English, appears almost exclusively after a guttural vowel. Cf. on this question HORN, Gramm. §§ 42, 231 ff.; LUICK, Anglia XVI, p. 462 ft.; GRUNING, p. 21.

In O.Fr. vocalization of I before a consonant set in very early, and the process had certainly commenced at the time of the Conquest. Cf. MEYER-LUBKE, Gramm. § 482; NYROP, Gramm. § 343.

Most instances of I > u which we have brought together in list A can only be accounted for as due to this sound-change. The etymological spelling with I is kept very long in A.N. u does not become general until the 13th and 14th cent. Cf. SUCHIER, Les voyelles toniques du vieux fran^., p. 144; STIMMING, p. 210. This fits in with the circumstance that I for u in English place-names is of sporadic occurrence until the 13th cent., when it becomes much more usual. In the 2nd vol. of the F.A. 9 out of 17 places in feld occur only with /-forms, one only with u- forms; for the remaining names we have found both I- and u~ forms. It goes without saying that, though I was orthographically retained in A.N., it nevertheless was pronounced u. Cf. the variant spellings of the River Eaulne in Benoit, v. 21372: JEvne, Elne, Eiaulne.

150 R. E. Zachrisson

From the earliest time loss of u < I before a consonant is attested in A.N. In one and the same text we find instances of loss as well as of I > u. Cf. MENGER. p. 85. The loss seems to have been general after i, u (< Lat. u), e (<. Lat. a), but is well instanced also after other vowels. Cf. STIMMING, p. .211; MENGER, p. 85 f.

This phenomenon appears, though more rarely, in English place-names, and is particularly well attested in examples from D.B. We are well aware that some of the instances of loss given in list B may be explained as due to dissimilation, but as so many remain which do not admit such an explanation, we have not considered it necessary to distinguish between the two.

The following cases of addition of an unetymological I may be adduced as a further proof of loss having taken place : Olpessa \ Exou D.B., Ulpesse, Exchequer, D.B. = Up Exe (Dv.), Opexe, Upexe, F.A. < O.E. up, Mdf., p. 142; Fulsescote, D.B. = Foxcott (Ha.), Foxcote (1316, 46), F.A. < O.E. fox, Mdf., p. 53 (s for x may be another A.N. peculiarity); Holstune, D.B. (s = (•/)) = Hough ton (Ha), Hocthon, Ch.R. II, Hough-, ffoghton, F.A. < O.E. hoh, Mdf., p. 75; Horsted, -stdd, Fr.Ch. = Horstead (Nf.), Horstede, H.R. < O.E. stede, Mdf., p. 125; Falvesham, R.B. = Faveraham (Ke.) < O.E. Faures(feld), A.S. Chr.; 07/oro, Fr.Ch. = Offham (8s.), Of- ham (1200), Index < O.E. 0/a Searle.

In this place a few instances of loss from O.Fr. works may also be quoted: Suefoc, Sufoche, Rou; v. 7737 = Suffolk; Norfoc, •foche, ib., v. 7738 = Norfolk; Watehan, ib., v. 8994 = Waltham (Es.); Gedefort, Geldefort, ib., v. 4722 = Guildford (Sr.) (Guill. le Mar., v. 13801, has Geudefort); Mammesbere, Guill. le Mar., v. 15257 = Malmesbury (Wl.).

Between O.Fr. ^ and / -f- cons, a was developed as a glide. The new combination is orthographically represented by eal, eau, at, au. Often el is kept, more seldom — and, as it seems, hardly ever except in the A.N. dialect - - we find eu. All the above- mentioned spellings are to be found in our earliest A.N. texts. Cf. SUCHIER, Les voyelles etc., p. 151 ff.; MENGER, p. 59. Under

That I in Olpessa etc. is inorganic, and does not stand for ou = 0, is the more probable as ou for o was rarely used in A.N.; it hardly even appears in the earliest texts. Cf. MENGER, p. 67. STOLZE, pp. 19, 20 f., records no single instance of ou for O.E. ft, o in D.B.

Anglo-Norman influence on English place-names

151

such circumstances we need not hesitate to ascribe to French in- luence such forms as Halmelac, Aunestowe, Feutewell, Beautes- Jord, and others recorded in list A 2 b.

In no instance have the French forms prevailed. As to Celtone D.B. = Chawton (Ha.) (A 2 b) we have already hinted at the possibility of el being an inverted spelling for au. In this par- ticular case we have, moreover, not been able to ascertain if the identification is absolutely certain. On the change of Dunelme > Durelme >Dureaume > Duram, cf. § 7 B., p. 134 f.

NOTE. Loss of I in the present pronunciation of Wymondley (Ht.) (B 4), Hemswell (Li.) (A 2 b), Calne (Wl.) (B 1) (pron. without I, according to GRUNING, p. 22) is, no doubt, of a comparatively recent date, and by no means due to A.N. influence. In the first two cases dissimilation may have caused the loss; besides, there are instances of the groups Im, In having been simplified in English independent of dissimilatory influences. Cf. on this point EKWALL, Jones's Phonography §§ 571, 577.

In the case of Helmsley (Yo.) the A.N. forms are exceedingly well represented *. Hamelac etc. is often found as late as the 14th cent. In Camden II, p. 912, the place in question is called Elmesley and Hamlak. In the first element the following stages of development are represented: Helm > Heaum > Haum > Ham with loss of u. The second element is obviously O.E. ledh, ledge, leh, lege, which was turned into lac, lee by the Normans 2. To account for lac < leh, lege two explanations seem to be equally >ossible: (1) lac, lee may be taken to represent the nominative leh, in which case Norman Jc has been substituted for English (x)3. (2) If we start from the oblique cases, lac may as well be a Latini- zed form of lege, leie. Cf. Gurnai = Gornaco (pers.name), R.B. 4.

1 In the immediate neighbourhood are still to be found the ruins of [elmsley castle, constructed in the 12th cent.; BARTH.

2 In Gaimar, v. 6448 L, Helmeslac rhymes with Espac (a pers. name).

3 A similar substitution is indicated by such D.B. spellings as Pochelac Pockley (Yo.); Asmundrelac = Osmotherley (Yo.); Wisoc = Wysall (Nt.)

(I is a later addition), Wishoiv, -owe, F.A.; as probably also by Spelehot = Spelho Hundred (Np.), Spelhoh (107B), Geld. Inq., and Caissot = Keysoe (Bd.). so, t in the two latter cases is a mistake for c. On the other hand, -ot my also be the French diminutive suffix.

4 The Celtic base acum is of frequent occurrence in French place-names, id results in ay, ey in the North of France. Cf. QUICHERAT, p. 41.

152 R. E. Zachrisson

That official Latin forms, like Hamelac for *Hamelei, may be adopted into the spoken language, and even eventually supersede the phonetic forms, is proved fully by the occurrence of parallel cases. Thus Mod. Fr. Orleans for O.Fr. Orliens, -Quill, le Mar., v. 4490, etc. < Lat. Aurelianum, is due to the influence of the Latin ending -ianum (cf. MEYER-LUBKE, Einfuhrung in das Stud, der rom. Sprachwissenschaft, p. 194), and in the English place- nomenclature we may point out two cases where the whole name appears in a true Latin form, viz. Pontefract or Pomfret (Yo.), Melsa or Meaux (Yo.)f

NOTE 1. A perfect analogy to the treatment of the second element of Helmsley is offered by the following early references for Beverley (Yo.)r Beureli, D.B.; Beverlacum (c. 1199), Fr.Ch.; Beverley (John), Yo. P.F.; Bever- lacensis (1286), Beverlacum (1305), Beverley Chapter Act Book (Surtees Soc., 1898); Beverli, Kievaux Ch.; Beverlacum, Yo. N.V. Forms like Beverlay: layy Beverle: pre, R.Br. p. 29, p. 235, point to an O.E. ground-form left, le, lege. Beoferlic, A.S. Chr., may he due to the Latinizing of such a form as Beureli, D.B., where unetymological -li has been substituted for -ley 1. TAYLOR, Names, p. 68, is considerably disturbed by the forms Beoferlic and Bever- lac. He points to a hypothetic Celtic ground-form * Pedwarllech. Our expla- nation solves every difficulty, as far as the forms -lac, -ley are concerned, though we admit that Beoferlic, A.S. Chr., complicates matters. To assume French influence on the forms of personal-names and place-names in the later MSS. of the A.S. Chr. is, however, quite permissible. We have had occasion more than once to dwell upon such forms, and a thorough special investigation might perhaps bring forth more instances of the same kind.

NOTE 2. STOLZE, p. 38, adduces some instances of I having been dropped in weak-stressed position, which in his opinion are due to mere carelessness of the scribes. In consideration of the great number of similar instances we would rather suggest that the phenomenon in question reflects an A.N. phonetic tendency, by means of which the combination /-[-cons, is avoided. We must keep in mind that the change of I > u took place also in weak-stressed position. On the other hand, unstressed I before a cons, was sometimes lost in English itself, though we do not yet know under what circumstances this loss took place. Without presuming to decide in

1 Cf. Catereio, Cetereio by the side of Ceterio, LEI. < O.E. Ceateric, Thorpe = Chatteris (Ob.). Is this curious interchange due to the analogy of such French forms as Abbaneio, Aubeny, R.B., Albineium (1108), Aubeni (1160) = Aubigny (Calvados); Aniseium (1198), Anisie (1155) = Anisy (Calv.), Aguerneium (1180), Agerni (1190) = Anguerny (Calv.) (see Dictionnaire Topogr. du Calv.) < Celtic -iacum, which on Norman territory seems to have given both -ei and -». Cf. QUICHERAT, p. 37.

Anglo-Norman influence on English place-names 153

each particular case whether the loss is due to French influence or to Eng- lish sound-development, we will content ourselves with adducing some instances calculated to illustrate the loss. The chronology may sometimes serve as a criterion. If the occurrence of forms without I is confined to D.B. and other early Anglo-Latin records French influence is highly probable, whereas it must be altogether out of the question when the loss is not evidenced until the 14th and 15th cent. We have noted the following cases: Liticote, Fr.Ch. = Littlecote (Bu.) < O.E. lytel, Mdf., p. 90; Witesie, D.B. = Whittlesea (Cb.), Witleseye (1389) (Conybeare's Cbs., p. 147) < O.E. *Hwitel (cf. SKEAT, Pl.-Ns. of Cb., p. 55); Cingefort, D.B., Chinge-, Chingelford, H.R., Chinkeford (1279—88), C.R. = Chingford (Es.), Chyngelford (1288—96), C.R., Shingelford (1346), F.A.; Pritteuuella D.B., Prit-, Pritewell (13 c., 15 c., 17 c.), Index = Prittlewell (Es.), Pritelwelle, H.R., Pritelewell, F.A.; Brempschott (1428), F.A. == Bramshot (Ha.), Brembelshute (1316), F.A. < O.E. brembel, Mdf., p. 17; Puteorde, Puteleorde, D.B. = Prittleworth (Ha.); Fengesham (1206), Ke. P.F. = Finglesham (Ke.)< O.E. fenglesham (c. 831), Index; Stapeford, Fr.Ch. = Stapleford (Le.), Estaple- ford, Fr.Ch. < O.E. stapol, Mdf., p. 123; Brokesby, F.A. = Brocklesby (Li.), Brochelesbi, D.B., Broclesby, T.N., Brokelesby, F.A.; Hundebi, D.B. = Hundleby (Li.), Hundelbia (1145), Index, Hundelby (14 c.), Index, F.A.; Cringaforda, Kri'gelforda, D.B. = Cringleford (Nf.), Cryngylford (14 c., 15 c.); Esbece, D.B. = Haselbeech (Np.), Haselbech, Np.S. < O.E., hcesel, Mdf., p. 66; Hangetone, D.B. = Hangleton (8s.), Hangel-, Hangleton, T.N.; Chinewrde, D.B. = Kenilworth (Wa.), Kenileivorth, Ch.R. I < O.E. Cinildewyrth, according to TAYLOB, Names, p. 160; Hasbury (Wo.), Haselburi (13 c., 14 c.) (cf. DUIGNAN, Wo. Pl.Ns., p. 78); Huerleberie, D.B. = Hartlebury (Wo.)< O.E. Heortldbyrig , Birch 360, 361; Eybecestre (1232—74), P.R. = Ribchester (Yo.), Ribelcastre, D.B.; Crigestone, D.B. = Crigglestone (Yo.), Crigleston (1191), Yo. P.F.

§ 9. a before nasal groups becomes au.

After the year 1250 the spelling aun for an appears in nearly all English place-names exhibiting this sound-combination. We will therefore content ourselves with adducing some instances where the forms in au have been retained either in spelling or in pronunciation.

1. Saunderton (Bu.). 2, Cambridge (Cb.)< O.E. Grantebrycg, A.S. Chr. 3. Launceston (Co.), > Lanstupha don, i.e. the Church of Stephen, com- monly call'd Laumton* (CAMDEN I, p. 25); »a corruption of Lan- Stephen- dun, 'the hill by St. Stephen's Church', an adjacent monastery being dedicated to St. Stephen* (TAYLOB, Names, p. 169) *. 4. Taunton (Co.) < O.E. Tantun,

1 Launceston (formerly Dunneheved) was the ancient capital of Co. At the time of the Conquest it contained a famous castle, and was also the seat of an Augustinian priory; BAETH.

154 R. E. Zachrisson

A.S. Chr., Thorpe1. 5. Penzance (Co.), »means 'Saint's head' or 'holy head' from a chapel dedicated to St. Anthony > (TAYLOR, Names, p. 220). 6. Saunton (Dv.) < O.E. sand, Mdf., p. 109. 7. Braunton (Dv.) < O.E. brant, Mdf., p. 17. 8. Caundle (Do.)? < O.E. candel, Mdf., p. 24. 9. Staunton (Gl.)

< O.E. stem, Mdf., p. 122. 1 O. Staunton (He.) < O.E. stun. 1 1 . Braunce- well (Li.). 12. Braunston (Np.), probably < O.E. Brand, Searle. 13. Saundby (Nt.) < O.E. sand. 1 4. Staunton (Nt.) < O.E. stdn. 1 5. Staunton (Wo.)

< O.E. Stan tune, Birch 1281.

D.B. Santesdune1, Lancavetone* , Santone9, Brantone1, Candel*, Stantune9, Standone10, Branzeuuelle11, SandeU™.

12th c. — Index. Santersdune1 (Hen. II.). Pi.R. Cantdbrigia*, Lanzauenton* (1189—90), Branton1 (1184).

13th c. — C.B. Cantebrug* (1218), Lanzaveton* (31—34), Tanton* (31—34). Oh.R. Laustaneton3 (1229), Bramton1 (1229), Braundeston1* II. F.A. Staunton9 (1284). H.K. Sant'sdon1, Lancaneton3, Branton\ Stanton?,Brancewelln. Index. Kandel*, Cmmde/8 (1280). P.B. Lan- caveton* (32 — 47), Tanton* (ib.). K.B. Lansanestone*, Lanceneton9, Landstanelle*. B.G-1. Grauntebrtigge*, v. 132. T.N. Sant'don1, Sauntesdon*, Castrum de Landstavesfs, Landstanest's, Branton1, Brampton\ Stanton*, Sandeby13, Stan-, Staunton1*. Wo. S.B. Stantone16 (1275) (see DUIGNAN, Wo. Pl.Ns.; p. 153).

14th c. — F.A. Sauntdresdone1 (1302), Sauntresdene* (1346), Staunton9 (1316), Staunton10 (1316, 46), Braunceswett11 (1303, 46), Braundeston1* (1346), Saundeby13 (1302, 16). Index. Saundreton* (1317), Saun- dreston1 (1389), Launceton* (1369), Brampton1 (1322).

15th c. — F.A. Lanceston3 (1428), Stanton9 (1402), Staunton9 (1428, 31), Staunton10 (1431), Braundeston1*, Saunby}3 (1428). I.P.M. Saunton* (Hen. VII.), Brampton'1 (ib.), Caundell* (ib.). Index. Caundell* (1419) Paston Letters. Kawm-, Caumbrege* (cf. NEUMANN, pp. 21, 22).

16th c. — Index. Lanceston3 (1573), Pensaunce* (1552).

In M.E. as well as A.N. texts from the middle of the 13th cent, downward we often find the spelling au + nasal + cons, in words containing an original a. This spelling is in English mainly confined to French loan-words, though it also occurs in some words of native origin. In the Orthographica Gallica, an A.N. spelling-book of the 13th or 14th cent., we are expressly told that the value of this spelling was aun. Cf. STOKZINGER, p. 19. Statements to the same effect are made by

1 Taunton is a very ancient place, and was a favourite residence of the West Saxon kings. Its castle, now used as a museum, was founded about 700 A.D.; EARTH.

Anglo-Norman influence on English place-names 155

the English orthoepists of the 16th cent. Of. LUICK, Anglia XVI, p. 480.

This au has in Modern English given different results be- fore different consonantal combinations:

1. (ei) before mb, as in »chamber».

2. (aa), (oo) before nt, nd, as in »aunt», » grant »; »haunt», »taunt» - - »demand», »slander»; »Maundy Thursday*.

3. (aa) before ns, as in »chance», »dance».

It is, however, not to be forgotten that in a few English words (»answer», »ant» etc.) we find the same sound-development as in the French loan-words, according to Sweet, H.E.S. § 860, due to the analogy of the numerous French words with ant, ans.

To judge by the present orthography, most English place- names of the kind here in question have been developed accor- ding to English sound-laws. Bartholomew's Gazetteer has about 40 Stantons and only 9 Stauntons, several scores of places be- ginning with Sand- against 7 in Saund-, 14 Brans- against 4 Brauns- etc. The comparatively small number of forms with au is, however, sufficient to prove that A.N. forms have been adopted in spelling and pronunciation, and often superseded the native ones.

Where such has been the case the result of the further development of au has been the same as in the French loan- words:

1. (ei) in Cambridge, pron. (keimbrid}), Tanger.

2-. (aa) in Staunton, pron. (staanten), Tanger 1, Launce- ston, pron. (laansten), Tanger, Peuzance, sometimes pron. (pen- zaans) 2.

3. (oo) in Taunton, pron. (toontan), Tanger.

Though we have not been able to ascertain the pron. of Braunton, Saunton, Caundle etc., we are hardly much mistaken

1 Close to the village of Staunton (Wo.) there is a manor-house, Staunton Hall, which has belonged to the family of Staunton ever since the Norman Conquest. It is, indeed, very probable that the name of the place owes its present spelling and pronunciation to the family-name. Cf. Introduc- tion, p. 15. In like manner Saunderton (Bu.) may have been influenced by the family-name Saunders, pron. (saandaz), Tanger <: Alisaundre. Cf. BAKDSLEY'S Die., p. 668.

2 Prof. Ekwall kindly informs me that the local pron. is (penzams); this being so (penzaans) is, no doubt, a mere spelling-pron.

156 K. E. Zachrisson

in assuming that (se) is the local proii., (oo) or (aa) the one used by people unacquainted with the respective localities (spelling- pron.). As a matter of fact, several Stauntons are locally pro- nounced (staentan), see Tanger, Hope. In Magna Britannia and Catnden we often find the spelling a where a modern Gazetteer has au, a circumstance which also seems to indicate a local pronunciation (se): Saunderton = Sanderton,. M.Br., Senderton, Camden; Saunton = Santon, M.Br.; Braunton = Branton, M.Br., Camden; Brauncewell = Branswell, M.Br., Camden; Saundby = Sandeby, M.Br.; Staunton (He.) = Stanton, Camden. In many of these cases present au may be considered as an archaic spelling, a remainder from that time (13th — 16th c.) when in the La- tin and English records the A.N. au- forms were much more widely diffused than they are now.

In several Mod. English dialects, especially in the Midlands, O.E. a, o appears as o before a nasal (Cf. E.D.Gr. § 30 f.), and au in the present spelling may sometimes very well be a symbol of this sound 1. That such is not the case in the instances treated of above is clearly shown by the circumstance that au makes its first appearance at the same time as in the French loan-words, and is, moreover, not confined to any dialect, but is to be found all over the country.

§ 10. Change of (t/) > (/).

1 . Chalfont (Bu.). 2. Chelmorton (Db.) < O.E. Ceolmcer, Searle. 3. Chesterfield (Db.) < O.E. ceaster, cester. 4. Chelborough (Do.) < O.E. Ceol, Searle. 5. Chingford (Es.). 6. Chedworth (Gl.) < O.E. Cedda, Searle. 7. Churchdown (Gl.) < O.E. cyrice, Mdf., p. 37. 8. Charlton (Ha.). 9. Chart- ham (Ke.) < O.E. ceart, cert, Mdf., p. 26. 1O. Charwelton (Np.). 1 1. Checken- don (Ox.) < O.E. Ccecca, Searle. 1 2. Chipstead (Sr.).

13th c. — F.A. Shedeworth* (1284), Ghertheham9 (1284), Chakendon11 (1284). H.R. Chingc-, Chingelford*. Index. Chalfhunte1 (1208), Cherte- ham9 (1263), Scharwelton10 (Hen. III.), Chanvilton™ (1203), Chep- sted™ (1250).

1 Cf. Launton (Ox.), Langtun (1066), Index; Langetune (Hen. III.), ibid.; Langtun, V.E. The present spelling seems to date only from the beginning of the 38th cent The transition of (g) > (n) is, no doubt, due to the assi- milatory influence of the following t.

Anglo-Norman influence on English place-names 157

14th c. - F.A. Chalfunte* (1303), Chalphunte1 (1316), Chalfhvnte1 (1346), Chaster/eld* (1346), Schawebe.rgc* (1303), Chelbergh* (1346), Shingel- ford* (1346), Chinggeford* (1303), Cheddeivorth* (1303, 46), Ched- worthe6 (1316), Schurchesdon1 (1303), Churchesdori1 (1316), Schalston* (1316), Chalghton* (1346), Chertham* (1316), Charewelton™ (1316), Chakindon11 (1316), Chakendon11 (1346). Index. Shalfount1 (1375), Scheilmardori* (1317), Chepstede1* (1365). Nt. Bees. Shasterfeld* (c. 1300).

1 5th c. — F.A. Chelmardon* (1431), Chelburgh* (1431), Cheldeborowe* (1428), Chynegford\ (1428), Churchedon1 (1402), Chaulton* (1428), Cftare- t^faw10 (1428), Shakynden11 (1428). Index. Scherthom* (1423), Chipstede" (1429), Shipstede1* (1424).

In the course of the 13th cent. O.Fr. (t/) was changed to (/). Cf. NYROP, Gramm. § 402. The above instances of s&, s<?/& for c/i are, no doubt, due to this sound-development. In the French loan-words in English BEHRENS, p. 178, notes several instances of sc/i for c&, which really may indicate an incidental pron. (/) for (t/) in imitation of later A.N. habits of speech 1.

The opposite phenomenon, ch used for sh, sch, we have also noted in several cases, e.g. Chi/orde, F.A. = ShefTord (Be.); Chelford, ib. = Sheli'ord (Cb.); Chettisbeare, ib. = Shittisbeer.(Dv.); Cherynge (14 c.), ib. = Sheering (Es.); Chepehale (14 c.), ib. = Shep- hall (Ht.); Chymplynge (1371), Index = Shimpliug (Sf.); CUpley (1562), Index = Shipley (Ss.); Chustoke (1323), Index = Shu- stoke (Wa.).

The spelling ch for sh, sch is, besides, not uncommon either in French loan-words (see BEHRENS, p. 190) or in native words. See SWEET, H.E.S. § 607; DIBELIUS, p. 462; NEUMANN, p. 73. Cases of this kind are probably to be considered as inverted spellings, originally used by scribes who were Frenchmen by birth or at least possessed some knowledge of the French language.

We are more surprised at finding sh, sch for ch in several 14th and 15th cent, spellings of native words. For instances see, DIBELIUS, p. 461 f.; NEUMANN, p. 88; HORN, Gutturallaute, p. 18.

1 French loan-words in English of early introduction, naturally, keep the old sound (t/). In dialectal schesel (E. Angl., see E.D.D.), corresponding to literary English » chisel » < Picard chisel, later A.N. (/) seems, however, to have ousted earlier (t/). As a personal name O.Fr. chanoine, chanonis (plural) Oodefroy) has survived to the present day in both forms,, Shannon and Channon. See BARDSLEY'S Die., p. 681.

168 R. E. Zachrisson

Many more are noted in N.E.D. It is very hard to tell if we here have to do only with faulty spelling, caused by the inter- change of ch, sch in words of French origin, or if original (t/) really had been turned into (/).

As a matter of fact (/) does occasionally occur for (t/) even in the South E. dialects where it is not to be accounted for as a dialectal adaptation of (t/) in Standard English Cf. HORN, I.e. To explain South E. dial. (/) < (t/) HORN, ibid., assumes that the French loan-words, have influenced the pron. of native words containing (t/). However, owing to their very sporadic occurrence, it would probably be safer to explain the few cases which have been recorded as mere individual mispronunciations; in connected speech (t/) in certain positions (cf. below, (t/) before a consonant) may easily have been assimilated to (/). The great acoustic resemblance between (t/) and (/) is also to be taken into con- sideration.

In a few instances a place-name with original ch = (t/) occurs in Mod. E. with sh =(/)1:

Fiendish Hundred (Cb.); Flamingdice, D.B., Flammigedic, flamingedich, Inq. EL, Flemedich (1158), Pi.R., Flemdiche (1302), F.A. < O.E. die. The true English form is kept in Fleam Dike, cf. SKEAT, Cb. Pl.-Ns., p. 40 f.

Lashbrook (Dv.), Lachebroc, D.B.? Lecchebrok, Lachebroch, F.A., possibly < O.E. lace, Mdf., p. 85.

Washlingston Hundred (Ke.), Wacheleston, D.B., Wechelstone (1316), Wachelstane (1346), F.A., WatMingston (1587), Index; cf. O.E. wcecel, wacel, Mdf. p. 140.

1 In Shelmerdine, a personal name derived from a locality, present sh stands for earlier ch. BARDSLEY, Die. p. 683, who gives some late refe- rences for the word, is unable to find the spot, but remarks that the habitat must be sought for in S. Lane, or E. Chesh. We do not hesitate to identify Shelmerdine with Chelmorton, a town in the adjacent part of Derbyshire, of which, besides numerous forms in ch (Chelmerden, Cheilmardon, Chelmordon, Index), we have also found one in sh, viz. Scheilmardon (1317), Index. The etymology is obviously O.E. Ceolmcer (cf. above), and sh in the present spelling and pron. of the personal name is, in all probability, due to French influence. On the other hand, several instances where present sh is explained by BARDSLEY as due to earlier ch are wrong. Sherrington and Shillingford (see Bardsley, p. 685) must by no means be derived from *Cherrington, *Chillingford, as the early forms of the corresponding localities exhibit nothing but s, sch, sh.

Anglo-Norman influence on English place-names 159

Washford (Sin.), Wachetford (1367), Index; cf. Watchet (Srn.)> Weced, Wceced, A.S. Chr., Wechet, Wachet (13 c.), H.R., Index.

Wishford (WL), Wicheford, D.B., Wyche-, Wichford, H.R.r T.N., Wiche-, Wichford, Ch.R. I, II, Wichford (1351), Index, Wycheford (1445), ib., Wyssheford (1513), ib., Wisheford (Hen. VII.), I-P.M. All the old forms point to original ch, probably O.E. tvic.

Etchilhainpton (Wl.); locally pron. »Ashelton», »Ashlington» (see JONES, Domesday Book for WL, p. 213), Eckel-, Hechelhampton, H.R., Hechelhampton (1318—23), C.R., Echel-, HecMampton (Hen, VII.), I.P.M.

Sharncott1 (Wl.), Cernecote (1327— 30), C.R., Cernecote (1536), Index, Sharncote, Camden, on the map of Wl. This little place is situated on the Churne, from which it seems to have derived its name 2. The forms Sharncote, Cernecote presuppose the fol- lowing development:

O.E. *Cirencote > M.E. *™™> Sharncote.

To account for the above instances of (/) < (t/) French influence can hardly be assumed but in the case of Sharncott, supposing our guess at the etymology to be correct. It is remark- able that in nearly all the remaining cases 3 ch was followed by a consonant, and a transition of ch > sh before a consonant seems very plausible.

This leaves us with Fiendish unexplained. It would perhaps not be impossible to assume confusion with O.E. disc, which, according to SKEAT, Bd. Pl.-Ns. (see under Farndish), occurs in English place-names in the sense of »cup»

1 To be kept apart from Sharcott (Wl.) near Pewsey.

2 Charmouth (Do.) appears in old spellings as Chernemeive, Charnemouth, F.A., Cernemue, Ch.R. II. Cf. also such French forms as Cernay, Charny (see Die. Topogr.) = Cernay (Marne).

8 The two places Chatton and Chillingham, both of them situated in Northumberland, are dialectally pronounced with (/) (see ELLIS, English Dial. Soc. 50, p. 128), which here, no doubt, is best explained as a dialectal adaptation of the Standard pronunciation.

4 Cf. Farndish (Bd.), Fernadis, D.B., Farnesdisch, T.N., Farnedis, -Aiche, -dich, F.A.; Bendish (Ht.), Benedis, D.B.; Standish (Gl.), Stanedis, D.B., Stanedix (c. 1200), Gl. Recs., Stanedis (1260), ib., Stanedyssh, -dich (1290), ib., Standyshe (1551), ib.

160 R. E. Zachrisson

this particular case there is another way of accounting for the change. The mere circumstance that in the M.E. spellings of place-names -is could stand for -iz < -ich (by A.N. sound- suhstitution, see § 1 list C, p. 22), as well as -ish, might easily have induced a scribe to put -ish for -ich, and vice- versa. And such a form, which owes its existence to a scribal error, may have' succeeded in superseding the etymological one.

A few instances may be adduced to show what important part spelling plays in a case like this. Of Barton Bendish, a small place in Nf., we have ascertained the following early forms: Bynedich (1316), F.A., Bendyssh (1467—77), P.R. On so little evidence it is quite impossible to decide which of the forms is the original one. (Byne)dich, F.A., points to O.E. -die, though, for all we know, -disc may really be the true etymology. Only a sufficient number of early references could decide the question. Nothing but modern -idge, which, at least from a phonetic point of view, cannot be due to -ish, makes us assume original ch for the following places: Pentridge (Do.), Pentrich (1303), Pentryssh (1346), F.A.; Whitnage (in Uplowman, Dv.), Witenes, D.B., Wkytmeih (th=ich), T.N., Hwytenych (1316), F.A., Wytenisshe (1346), ib. - - In Cranwich (Nf.) etymological sc, sch has, at least in the spelling, been ousted by ch. All the early spellings of the name of this place exhibit -isse, -ys -ysh etc.1, and, according to Hope, it is still dialectally pronounced with s. Two forms in •iz from the F.A. give us the clue to the transformation. This -iz may originally have been a spelling for -is; however, to French -iz corresponded very often English -ich, hence the modern form.

NOTE. Perhaps this whole question is in some way connected with the cur- ious interchange of the endings -is, -ich, -ish, -idge which has been noted in several English words. Thus > notice > and » rubbish » appear in several dialects with -idge. See E.D.D. On the other hand -ish sometimes appears for -idge, an in podish for >porridge>, demish for »damage». See N.E.D. and E.D.D.

In the place-name Sundridge (Ke.) earlier -esh -ish has, in a similar manner, been replaced by -idge. Old forms are Sunderhirse (1072), Index, Sonderesse, D.B., Sunderesshe (1316), F.A., Sonderershe (1346), ib., Sundrisse, Sundrissh (1461—67), P.R. Our first reference for -idge is Sundrygge (Hen. VII.), I.P.M., by the side of Sondrissh, ib. BARDSLEY, Die., p. 712, derives Standage, a personal name, from Standish.

1 Such early spellings are Craneiciase, D.B., Craneivisse, Creneuuis, Inq. El., Cranewes, Cranewys, Craneivyz, Carnuiz, F.A., Cranwyshe (Hen. VII.), I.P.M.

Index to names.

Only the modern forms of the names are given. The figures refer to pages.

Abbreviations : f.-n. = foot-note; n. = note.

English place-names and river-names.

Abbotsley (Hu.), 113.

Adling-, 112.

Admington (Gl.), 112.

Albrightlee (Sh.), 112.

Albrighton (Sh.), 112.

Alconbury (Hu.). 148.

Aldersgate (London), 118.

Alderwasley (Db.), 133.

Aldington (Ke.), 146.

Aldrebury Hundred (Wl.), 113.

Aldwark (Db.), 146.

Alford (Li.), 146.

Allexton (Le.), 112.

Ailing-, 112.

Allington Hundred (Ke.), 113.

Alnmouth (Nb.), 119.

Aloes Bridge Hundred (Ke.), 112.

Alveley (Es.), 113.

Amesbury (WL), 142.

Ampleforth (Yo.), 142.

Amport (Ha.), 12.

Ardeley or Yardley (Ht.), 65 n.

Arlsey (Bd.), 113, 118.

Arncliff (Yo.), 66 n.

Arrington (Cb.), 139.

Ash, Ash-, 10.

Ashmansworth (Ha.), 145.

Aswardby (Li.), 128.

Atchlench (Wo.), 26.

Athel-, 111. Atherstone (Sm.), 111. Avening (GL), 140. Avon, 142. Ayl-, 113.

Aymestre}?' (He.), 113. Ayot, 95 f.-n.

Bagtor (Dv.), 47.

Barford (Nf.), 120, 129.

Barming (Ke.), 139 f.-n.

Barton Bendish (Nf.), 160.

Barton-le-Street (Yo.), 12.

Bath (Sm.), 88, 97.

Beachamptou (Bu.), 141.

Beauxfield or Whittield (Ke.), 13.

Belchford (Li.), 147.

Belleau (Li.), 13, 121.

Belle Hatch (Ox.), see Boulney.

Bendish (Ht.), 159 f.-n.

Bengeo (Ht.), 139.

Beverley (Yo.), 27 f.-n., 152 n.

Bexhill (8s.), 147.

Bicester (Ox.), 78.

Bilchester or Bilchurst (Ke.), 81.

Bircham Tofts (Nf.), 68.

Birchanger (Es.), 120.

Bircholt (Ke.), 38, 148.

Birmingham (WTa.), 2.

11

162

R. E. Zachrisson

Blatherwick (Np.), 87.

Bletchingley (Sr.), 21.

Blymhill (St.), 121.

Blyth (Nt.), 88.

Bolmer (Boulmer) or Balmer (8s.),

121, 129.

Borstall (Ke.), 118. Boston (Li.), 3. Boulney (Ox.), 86. Bramshot (Ha.), 153. Brauncewell (Li.), 154, 156. Braunston (Np.), 154. Braunton (Dv.), 154, 156. Brickampton (GL), 134. Bridgewater (8m.), 12. Brighthampton (Ox.), 135. Brighthlingsea (Es.), 120, 131. Bristol (Sm.)f 1 f., 3. Brocklesby (Li.), 153. Buckfast (Dv.), 142. Budleigh (Dv.), 17. Bulkington (Wa.), 149. Bullingtons in Bexhill (Ss.), 139. Bunbury (Ch.), 123. By stock (Dv.), 118.

Cadeby (Li.), 118. Caithness (Scot!.), 24. Calbourne (Ha.), 146. Calne (WL), 148, 151. Cambridge (Cb.), 79 f, 136 ff.,

153 f.

Camelford (Co.), 142. Cammeringham (Li.), 54 n., 139. Canewdon (Es.), 123. Canfield (Es.), 148. Cannoc (St.), 49, 138. Canterbury (Ke.), 24. Carbrooke (Nf.), 121. Castlecombe (Wl.), 11. Catehill (Sr.), 24. Catthorpe (Le.), 41. Caundle (Dv.), 154, 156. Cawthorne (Yo.), 39. Cawthorpe (Li.), 41. Cerne (Do.), 19, 30. Cerney (Gl.), 19, 30.

-cester, 18/20, 29 f., 73, 76.

Chaddesley (Wo.), 24.

Chadwick (Wo.), 24.

Chagford (Dv.), 24.

Cbalfont (Bu.), 156.

Chard (Sm.), 28.

Charlton (Ha.), 157.

Charmouth (Do.), 159 f.-n.

Chartham (Ke.J, 156.

Charwelton (Np.), 156.

Chatteris (Cb.), 22, 24, 31, 152

f.-n.

Chatton (Nb.), 159 f.-n. Chawton (Ha.), 147, 151. Checkendon (Ox.), 38, 156. Cheddar (Sm.), 28. Cheddon (Sm.), 28. Chedgrave (Nt), 68. Chedworth (GL), 156. Chelborough (Do.), 147, 156. Cheldon (Dv.), 24. Chelmorton (Db.), 156, 158 f.-n. Chelsea (Mi.), 86, 118. Chelsfield (Ke.), 121. Chelsham (Sr.), 25. Chelsworth (Sf.), 136. Cheriton (Ke.), 28. Chertsey (Sr.), 25, 28. Chesham (Bu.), 20. Cheshunt (Ht.), 20. Chesilborne (Do.), 54 n. Chester-, -Chester, 3, 18 f.-n., 20,

25 29 f 73 Chesterfield '(Db.), 156. Childerditch (Es.), 38, 144. Childrey (Be.), 86. Chillingham (Nb.), 159 f.-n. Chingford (Es.), 153, 156. Chinnock (Sm.), 28. Chippenham (Cb.), 25. Chipstead (Sr.), 156. Chiselborough (Sm.), 28. Chistlett (Ke.), 28. Chitterne (WL), 28. Chittlehampton (Dv.), 142. Chivelston (Dv.), 145. Church Lench (Wo.). 26.

Anglo-Norrnan influence on English place-names

163

Churchdown (Gl.), 156.

Churchill (Ox.), 28.

Churchstow (Dv.), 118.

Churne River, 19 f.-n.

Churnet River, 19 f.-n.

Cippenham (Bu.), 19.

Cirencester or Cicester (GL), 18 f.,

78 f.

Clanfield (Ox.), 138. Clavering (Es.), 140. Clawton (Dv.), 137. Cleddau River (Pembroke), 138. Cleley Hundred (Np.), 119. Cliff (Ha.), 68, 70. Cliff (WL), 68. Clitherhoe (La.), 137. Clixby (Li.), 100 f.-n. Coberley (GL), 138. Cockerington (Li.), 139. Codrington (GL), 138, 139. Colchester (Es.). 138. Combermere (Ch.), 16 n. Comberton (Wo.), 16 n. Compton (GL), 16 n. Coneyswick or Conningswick (Wo.),

141.

Cornwood (Dv.), 100 f.-n. Cossal (Nt.), 138. Coventry (Wra.), 2. Cowfold (8s.), 146. Crambourne (Ha.), 137. Crandon (Sm.), 137. Crane End or Scrane (Li.), 67, 70. Cransford (Sf.), 137. Cranwich (Nf.), 160. Cranworth (Nf.), 137. Craster (Nb.). 18, 73, 76. Greeting (Sf.), 137. Cressingham (Nf.), 137. Cretingham (Sf.), 137. Cricksea or Creeksea (Es.), 86. Crigglestone (Yo.), 68, 72, 137, 153. Cringleford (Nf.), 153. Crofton (Yo.), 68, Cromwell (Nt.), 16 n. Crowell (Ox.), 122. Crownthorpe (Nf.), 82, f.-n.

Cruwys-Morchard (Dv.), 81. Cuckamslow (Be.), 134. Cusworth (Yo.), 68. Cuxwold (Li.), 146.

Dartford (Ke.), 43 n. Dewlish (Do.), 128. Diss (Nf.), 22, 31. Ditchling (Ss.), 140. Dorset, 43 n., 78 f.-n. Droitwich (Wo.), 11. Duntisbourne (GL), 43 n. Durham (Du.), 133 ff., 151. Dyse worth (Le.), 7.

Earlham (Nf.), 65 n. Easington (Nb.), 65 n. Easthope (Sh.), 70 f.-n. Eastry (Ke.), 116. Easwrith Hundred (Ss.), 51. Eckington (Ss.), 72 n., 141. Edington (WL), 89. Edingworth (Sm.), 65 n. Edling-, 112. Elberton (Gl.), 113. Eleigh (Sf.), 94 f.-n. Ellastone (St.), 112. Ellerton (Sh.), 112. Ellesmere (Sh.), 122. Elling-, 112. Ellingham (Ha.), 54 n. Elmley Castle (Wro.), 147. Elmsett (Sf.), 147. Elmstead (Es.), 147. Elmsthorpe (Le.), 113. Elmston (GL), 113. Elmstree (GL), 113. Elmton (Db.), 147. Elstow (Bd.), 147. Elton (Hu.), 112, Elvaston (Db.), 113, 121, 127. Epperstone (Nt.), 68, 70. Esher (Sr.), 143. Essex, 10.

Etchilhampton (WL), 159. Evercreech (Sm.), 22. Everthorpe (Yo.), 66 n.

164

R. E. Zachrisson

Ewhurst (Sr.), 66 n., 131 f.-n. Exbourne (Dv.), 118. Exeter (Dv.), 18, 73, 76. Exning (Sf.). 140. Eynsford (Ke.), 123. Eynsham (Ox.), 136.

Falmere (Ss.), 122. Farmington (Gl.), 40, 145. Farndish (Bd.), 159 f.-n. Faulkbourne (Es.), 146. Favershara (Ke.), 150. Featherstone (St.), 82, f.-n., 87. Featherstone (Yo.), 87. Feltwell (Nf.), 147. -field, 146.

Fillongley (Wa.), 142. Finedon (Np.), 41. Finge(r)st (Bu.), 41. Finglesham (Ke.), 153. Fleam Dike (Cb.), 158. Fiendish Hundred (Cb.), 158, 160. Fletching (Ss.), 21, 38. Foxcott (Ha.), 150. Fradley (St.), 119. Framlingham (Sf.), 139 f.-n. Fringford (Ox.), 139.

Gamlingay (Cb.), 138, 139. Gelston (Li.), 138. Gentleshaw (St.), 12. Gerse Hill (Gl.), 62. Gillingham (Nf.), 138. Glassthorpe (Np.), 137. Glastonbury (Sm.), 73 f.-n. Glendon (Np.), 137. Gloucester (Gl.), 77 1, 137. Goldbanger (Es.), 138. Goodrington (Dv.), 138. Gosberton (Li.), 54 n. Gothersley (St.), 118. Goxhill (Li.), 138. Grantcbester (Cb.), 79 f. Gravenhurst (Bd.), 137. Graylingwell (Ss.), 139. Greetham (Li.), 14. Groveley (Wl.), 140.

Guildford (Sr.), 150. Guiltcross Hundred (Nf.), 138.

Haddenham (Bu.), 141. Haddenham (Cb.), 141, Halnaker (Ss.), 148. Hanchurch (St.), 38. Hangleton (Ss.)> 153. Harbledown (Ke.), 146. Hardres (Ke.), 136. Harrington (Np.), 88. Hartlebury (Wo.), 153. Hartleford (GL), 65 n. Hasbury (Wo.), 153. Haselbeech (Np.), 153. Hatherton (Ch.), 87. Hayling (Ha.), 140. Headley (Ha.), 87. Heapham (Li.), 65 n. -heath, 85. Hellingly (Ss.), 141. Helmsley (Yo.), 147, 148, 151 f. Hemswell (Li.), 147, 151. Hilderston (St.), 145. Herriard (Ha.), 121. Hildersham (Cb.), 118. Hitcbin (Ht.), 22. -hithe, 85. Hiz River, 22, 31. Hockering (Es.), 140. Hornsey (Mi.), 118. Horstead (Nf.), 150. Houghton (Ha.), 150, Hundleby (Li.), 153. -hurst, 72 n., 78 f.-n., 81. Hurstmonceaux (Ss.), 12. Hurstpierpoint (Ss.), 12. Huttoft (Li.), 13. Hyde Abbey (Ha.), 12.

Idridgehay (Db.), 118. Ilchester (Sm.), 65. Ingman thorp (Yo.), 121. Ingoldsmells (Li.), 121, 129. Instow (Dv.), 6. Ipswich (Sf.), 65. Isleworth (Mi.), 6.

Anglo-Norman influence on English place-names

165

Itteringham (Nl), 68, 70, 72 n.

Jagdon (Sh.), 58. Jarrow (Du.), 57. Jedburgh (Roxb.), 62. Jervaulx (Yo.), 58, 62 f. Jesmond (Nb.), 6, 57, 62, 82. Jevington (Ss.), 58.

Kelmarsh (Np.), 38.

Kemsing (Ke.), 25, 100 f.-n.

Kenilworth (Wa.). 153.

Kerne (Ha.), 12.

Kesteven (Li.), 39.

Keymer (Ss.), 128.

Keysoe (Bd.), 151 f.-n.

Kidderminster (Wo.), 145.

Kilmersdon (Sm.), 133.

Kimmerston (Nb.), 133.

Kintbury (Be.), 74 f.-n.

Kinvaston (St.), 128.

Knaith (Li.), 35, 50, 88.

Knapp (Ha.), 50.

Knapton (Nl), 50.

Knapwell (Cb.), 49.

Knaresborough (Yo.), 50, 138.

Knebworth (Ht.), 50.

Kneesal (Nt.), 50,

Kneesall (Li.), 71.

Kneeton (Yo.), 50.

Knepp Castle (Ss.), 12.

Knettishall (St.), 138.

Knightley (St.), 50.

Knightsbridge (London), 118.

Knolle (Co.), 68.

Knolle (Wa.), 138.

Knock (WTL), 50.

Knottlngley (Yo.), 50.

Knowl-, -knowl, 49, 50, 122, 129.

Kno\\ston (Do.), 49,

Knutsford (Ch.), 49.

Knutton (St.), 123, 129.

Lantyan (Co.), 122, 130. Lapford (Dv.), 54. Lapworth (Wa.), 54. Lashbrook (Dv.), 158.

Latchingdon (Es.), 21. Launceston (Co.), 153, 155. Launditch Hundred (Nf.), 22. Launton (Ox.), 156 f.-n. Leatherhead (Sr.), 87. Leeds (Ke.), 54. Leicester (Le.), 117, Leigh (Gl.), 12. Lelant (Co.), 122, 130. Lench (Wo.), 26. Lenham (Ke.), 135. Lessness Abbey (Ke.), 54 n. Levermere (Sf.), 142. Lewknor (Ox.), 122. Lifford (Wo.), 12. Lincoln (Li.), 14, 122, 130. Littlecote (Bu.), 153. Lolworth (Cb.), 16 n. London, 14, 143. Londonthorpe (Li.), 144. Loversall (Yo.), 66 n. Lutwyche (Sh.), 38.

Mablethorp (Li.), 146.

Maidenhead (Be.), 87.

Mailing (Ke.), 54. n.

Malmesbury (WL), 150.

Marsh (Sh.), 78 f.-n.

Matching (Es.), 21.

-mere, 13.

Messing (Es.), 21, 26.

Messingham (Li.), 100 f.-n.

Methwold Hithe (Nf.), see Otring-

hithe.

Modbury (Dv.), 128. Mold (Flintshire), 127 f. Monkton (Sm.), 123. Montacute (Sm.), 11. -mouth, 82, 93 f. Muncaster (Cu.), 133. Mundham (Ss.), 141. Murston (Ke.), 123.

Nantallan (Co.), 122. Nateley (Ha.), 122. Neatishead (Nf.), 54. Needham (Sf.), 133.

166

R. E. Zachrisson

Nesfield (Yd.), 38,

Nether Avon (WL), 117.

Netherexe (Dv.), 87.

Netherfield (Ss.), 87.

Netley (Ha.), 122.

Newcastle (Nb.), 11.

Newland (Wo.), 3.

Nimet or Intake (Dv.), 136.

Nobury (Wo.), 30.

Norfolk, 150.

Nottingham, -shire, 51, 55.

Nyland or Hand (Do.), 122, 131,

Oakley (Es.), 133 f.-n. Offham (Ss.), 150. Okeford Shilling (Do.), 55. Oldbury upon Severn (GL), 146. Oldcastle (Mo.), 11. Orsett (Es.), 87, 95, 97. Osbaston (Sh.), 54 n. Osbournby (Li.), 54 n. Osmotherley (Yo.), 151. Otringhithe (Nf.), 68, 70, 72 n. Otterington (Yo.), 72 n. Oxford (Ox.), JO, 47. Oxnead (Nf.), 87.

Palgrave (Nf.), 148. Penshurst (Ke.), 81. Pentridge (Do.), 160. Penzance (Co.), 154, 155. Pilsbury (Db.), 71. Pleshy (Es.), 13. Pockley (Yo.), 151 f.-n. Pontefract (Yo.), 11. Prittlewell (Es.), 145, 153. Prittleworth (Ha.), 153. Puddle, Puddletown (Do.), 142. Purbeck (Do.), 27 f.-n. Putney (London), 94.

Rackenford (Dv.), 136. Raveningham (Nf.), 139. -red, 87, 97. Redland (GL), 46. -reth, 85. Ribchester (Yo.), 153.

Richmond (Sr.), 11. Rickmansworth (Ht.), 145. Riding (Yo.), 99. Rievaulx (Yo.), 62 f.-n. Ringstead (Nf.), 119. -rith, 85.

Rolvenden Hundred (Ke.), 149. Roothing or Rpding (Es.), 89, 98. Rotherfield (Yo.), 87. Rotherhithe (Sr.), 16 n. Rothley (Le., Np.), 88. Rothwell (Li., Np.), 88, 95. Routh (Yo.), 88.

Saddington (Le.), 39.

Salisbury or New Sarum (Wl.), 120,

130.

Sapey Pritchard (Wo.), 136. Saundby (Nt.), 154, 156. Saunderton (Bu.), 153, 155. Saunton (Dv.), 154, 156. Sawbridge worth (Ht), 7, 118. Sawley (Yo.), 39. Sawtry (Hu.), 86. Scaldwell (Np.), 146, 148. Scampston (Yo.), 67. Scarborough (Yo.), 25, 55. Scrayingham (Yo.), 55, 67. Scutterskelf (Yo), 67. Seething (Nf.), 88. Sempringham (Li.), 139. Setchey (Nf.), 86. Shalford (Es.), 38. Shalford (Sr.), 146. Shalstone (Bu.), 38. Sharcott (WL), 159 f.-n. Sharncott (WL), 159. Sheering (Es.), 157. Shefford (Be.), 157. Shelford (Cb.), 157. Shelsley (Wo.), 38. Shelvestrode (Ss.), 38. Shenfield (Es.), 38. Shephall (Ht.), 157. Sherringham (Nf.), 139. Shilvington (Do.), 148. Shimpling (Sf.), 157.

Anglo-Norman influence on English place-names

167

Shipley (Ss.), 157.

Shittisbeer (Dv.), 157.

Shottery (Wa.), 86.

Shrawardine (Sh.), 56, 120.

Shrewsbury (Sh.), 56 f., 120.

Shropham (Nf.), 56.

Shropshire or Salop, 56 f., 120,

130.

Shustoke (Wa.), 157. Siddington (GL), 84. Sidmouth (Dv.), 117. Silverley (Cb.), 148. Sinnington (Yo.;, 139. Sleaford (Li.), 52. Sledmere (Yo.), 13, 52. Slingeby (Yo.), 52. Slipton (Np.), 39. Smallridge (Dv.), 148. Smithdou (Nf.), 52, 115 n., 118. Snarford (Li.), 52. Sneiuton (Nt.), 52. Snelson (Ch.), 52. Snettisham (Nf.), 52. Snoring (Nf.), 52. Sodington (Wo.), 84. South-, 83, 94. South Ash (Ke.), 38, 98. Speechwick (Dv.), 55. Spelhoe Hundred (Np.), 128, 151

f.-n.

Spettisbury (Do.\ 67, 70. Spelsbury (Ox.), 71. Sporle (Nf.), 55. St. Mary Mounthaw (London), 128

f.-n.

Stafford (St.), 67. Stallenge (Dv.), 26. Standish (GL), 159 f.-n. Stanford (Nf.), 121. Stanley (Yo.), 67. Stapleford (Le.), 153. Stapleton (Yo.), 67. Staunton (Gl. etc.), 154 ff. Stepney (Mi.), 86. Stevenage (Ht.), 117. Stitchbrook (St.), 67, 71. Stoke (8m.), 12.

Stoke Mandeville (Bu.), 12. Stotfold (Bd.), 146. Stow (Cb.), 55. Stowlangtoft (Sf.), 13. Stragglethorpe (Li.), 67, 145. Stramshall (St.), 136. Stretton (St.). 55. Stubbs (Yo.), 55. Stubbs Walden (Yo.), 67. Sturminster (Do.), 55. Sturston (Nf.), 67. Stutsbury (Np.), 67 f.-n. Suffolk, 150. Sulby (Np.), 39. Sundridge (Ke.), 160. Surrey, 88, 93. Swallowfield (Be.), 121. Swavesey (Cb.), 100 f.-n., 117 Sydenham (Dv.), 141.

Tabley (Ch.), 68. Tachbrook (Wa.), 38. Tame River, 49. Tanfield (Yo.), 43 n., 123. TattershaU (Li.), 47. Taunton (Co.), 154, 155. Teaiby (Li.), 40. Tenterden (Ke.), 123. Terrington (Yo.), 43. Testwood (Ha.), 68, 70. Teversham (Cb.), 47. Teynham (Ke.), 135. Tbarne River, 49. Thames River, 49. Thanet (Ke.), 48. Tharston (Nf.), 68. Thatcham (Be.), 42. Thaxted (Es.), 42 The Down (Ha.), 12. Theobald Street (Ht.), 42. Therfield (Ht.), 42, 43 n. Thetford (Li.), 42. Theydon (Es.), 42. Thimbleby (Li.), 42, 68. Thorn-, -thorn, 39, 41, 42. Thorp-, -thorp, 42, 43 n., 44. Thrandeston (Sf.), 44, 46.

168

R. E. Zachrissori

Three Houses (Ht.), 42.

Threo or Treo (Nf.), 44.

Throcking (Ht.), 42.

Throwley (Ke.), 117.

Thurgarton (XL), 42.

Thurlton (Nf.), 145.

Thurning (Hu.), 42.

Thurnscoe (Yo.) 43 n.

Tibaldstone Hundred (Gl.), 46 n.

Tibberton (Sh., Wo.), 47 n.

Ticehurst (Ss.), 31, 99.

Tilshead (WL), 46 n.

Timberscumbe (Sm.), 119.

Tingrith (Bd.), 39.

Tisted (Ha.), 31.

Tocketts (Yo.), 68.

Tockwith (Yo.), 91.

Toddington (Bd.), 43 n.

Toft (Cb. etc.), 13, 47.

Tollerton (Nt), 119.

Torksey (Li.), 40.

Torleton or Tarleton (GL), 40.

Tormerton (GL), 39.

Torpel or Thorpel in Ufford (Np.), 40.

Torridge (Dv.), 27 f.-n.

Torworth (Nt.), 40.

Tortworth (GL), 136.

Torweston (Sm.), 40.

Totham (Es.), 68, 70.

Totton (Ha.), 43 n.

Toxteth (La.), 68, 70.

Tring (Ht.), 45, 98 f.

Trusthorpe (Li.), 45.

Turnworth (Do.), 39.

Tusmore (Ox.), 40.

Tutbury (St.), 67, 70, 71.

Tyringham (Bu.), 138.

Tytherton (WL), 139.

Ugley or Oakley (Es.), 133 f.-n. Up Exe (Dv.), i50. Upleatham (Yo.), 142. Urchfoht (WL). 66 n. Ure River, 62 f. Uttoxeter (St.), 78 f.

-ville, 94.

Wakering (Es.), 140. Walberton (Ss.), 126 f.-n. Walkeringham (Nt.), 146. Walsingham (Nf.), 146. Waltham (Es.), 146, 150. Waltham (Ha.), 146. Warbleton (Ss.), 126 f.-n. Wargrave (Be.),. 120. Warrington (La.), 139. Washford (Dv.), 54 f.-n. Washford (Sm.), 159. Washlingston Hundred (Ke.), 158. Wendy (Cb.), 141. Werrington (Np.), 88. West Tofts (Nf.), 68, 70. Westminster, 13. Wetherley (Cb.), 87. Wheatenhurbt (GL), 73. Whissonsett (Nf.), 21, 26. Whitchurch (Sh.), 3. Whitfield (Np.), 119. Whitnage (Dv.), 160. Whittlesea (Cb.), 153. Whixley (Yo.), 117. -wich, 22, 27. Wickmere (Nf.), 12, 13. Wide-, Widde-, Wid-, 85, 97. Widecombe (Dv.), 85, 98. Wiggenhall (Nf.), 141. Wilbraham (Cb.), 149. AVillitoft (Yo.), 13. Willshampsted or Wilstead (Bd.).

135.

Wilmotsham (Sm.), 133. Wilsthorpe (Yo.), 142. Wilting (Ss.), 149. Wimersley Hundred (Np.), 119. Winch (Nf.), 26. Wing (Bu.), 89. Wingrave (Bu.), 89. Winthorp (Li.), 41. Wistow (Le.), 72 n. Wishford (WL), 159. Witchingham (Nf.), 21. Witham (Es., Li.), 88. Withy-, Withing-, 84. Wiveliscombe (Sm.), 12.

Anglo-Norman influence on English place-names

169

Worcester, 78. Worston (St.), 145. Wragby (Li.), 51. Wragholm (Li.), 51. Wramplingham (Nf.), 51. Wrangle (Li.), 51. Wrawby (Li.), 51. Wraxall (Sm.j, 51. Wraxhall (WL), 51. Wrenbury (Cb.j, 51. Wretham (Nf.), 51. Wrington (Sm.), 51. Wroxall (Ha.), 51. Wroxeter (Sh.), 18, 51, 73. Wyboston (Bd.), 146. Wyke Regis (Do.), 3.- Wymondley (Ht), 149, 151, Wysall (Nt.), 151. Wytham (Be.), 141. Wyville (Li.), 85, 94.

Yafforth (Yo.), 58. Yalding (Ke.), 65 n. Yapham (Yo.), 58. Yarborough (Li.), 58.

Yard (Dv.), 12.

Yardley (Es.), 61.

Yare River, 58.

Yarm (Yo.), 58.

Yarmouth (Nf., Sf.), 58, 60, 62.

Yatesbury (WL), 65.

Yatton (WL), 65.

Yawthorp (Li.), 58.

Yaxham (Nf.), 58, 66 n.

Yaxley (Hu.), 58, 117.

Yaxley (Sf.), 58.

Yearsley (Yo.), 66 n.

Yeddingham (Yo.), 66 n.

Yelden (Bd.). 61.

Yeldham (Es.), 61.

Yelling (Hu.), 62.

Yelvertoft (Np.), 61 f.-n.

Yeo River, 66 n.

Yetminster (Do.), 65.

Yockleton (Sh.), 58.

Yokefleet (Yo.), 58.

Yordale (Yo.), 62.

York, 12, 27, 63 f.

Youlton (Yo.), 58.

Yoxford (Sf.), 58.

French place-names and river-names.

Anguerny, 152 f.-n. Anisy, 152 f.-n. Aubigny, 152 f.-n. Avroult, 48. Bellebrune, 126 f.-n. Bellengreville, 140. Bicetre, 25 f.-n. Canapville, 50. Cernay, 31 f.-n., 159 f.-n. Damblainville, 12. Eaulne River, 149. Emalleville. 53 f.-n. ^pernay, 70. fipreville, 56, 70. Estorsel, 56, 70.

Etrechy, 70. Etreville, 56. Eure River, 63 n. Fralignea, 140. Fresne -la- Mere, 13 f.-n. Roquedur, 135 n. Langrune, 126 f.-n. Mombray, 128 f.-n. Montamy, 12. Netreville, 53 f.-n., 56. Orleans, 152. Pont-Bellenger, 140. Tierceville, 48. Tilly, 48. Tosny, 48, 115.

243 10

DA 645 .Z3 IMS

Zachrisson, R. E. (Robert Eugen), 1880-

A contribution to the study of Anglo-Norman influence on English place-names. —

PONTIFICAL INSTITUTE OE MEDIAEVAL STUDIES

59 QUEEN'S PARK TORONTO 5, CANADA

t I |

luiHl II fl

. i ;

it!

I j