This name exists in several other forms, e.g. Mihell, Myhill, Mighill, and most frequently of all as Miles (Chapter VIII). The reader will remember the famous salient of Saint-Mihiel, on the Meuse, held by the Germans for so long a period of the war. From Gabriel we have Gabb, Gabbett, etc. The common rustic pronunciation Gable has given Cable (Chapter III).

Among female saints we find Agnes, pronounced Annis, the derivatives of which have become confused with those of Anne, or Nan, Catherine, whence Call, Catlin, etc., Cecilia, Cicely, whence Sisley, and of course Mary and Margaret. For these see Chapter X. St. Bride, or Bridget, survives in Kirkbride.

FEAST-DAYS

A very interesting group of surnames are derived from font-names taken from the great feasts of the Church, date of birth or baptism, etc. [Footnote: Names of this class were no doubt also sometimes given to foundlings.] These are more often French or Greco-Latin than English, a fact to be explained by priestly influence. Thus Christmas is much less common than Noel or Nowell, but we also find Midwinter (Chapter II) and Yule. Easter has a local origin (from a place in Essex) and also represents Mid. Eng. estre, a word of very vague meaning for part of a building, originally the exterior, from Lat. extra. It survives in Fr. les êtres d'une maison. Hester, to which Bardsley gives the same origin, I should rather connect with Old Fr. hestre (hêtre), a beech. However that may be, the Easter festival is represented in our surnames by Pascall, Cornish Pascoe, and Pask, Pash, Pace, Pack.

Patch, formerly a nickname for a jester (Chapter XX), from his motley clothes, is also sometimes a variant of Pash. And the dim. Patchett has become confused with Padgett, from Padge, a rimed form of Madge. Pentecost is recorded as a personal name in Anglo-Saxon times. Michaelmas is now Middleman (Chapter III), and Tiffany is an old name for Epiphany. It comes from Greco-Latin theophania (while Epiphany represents epiphania), which gave the French female name Tiphaine, whence our Tiffin. Lammas (loaf mass) is also found as a personal name, but there is a place called Lammas in Norfolk. We have compounds of day in Halliday or Holiday, Hay-day, for high day, Loveday, a day appointed for reconciliations, and Hockaday, for a child born during Hocktide, which begins on the 15th day after Easter. It was also called Hobday, though it is hard to say why; hence the name Hobday, unless this is to be taken as the day, or servant (Chapter XIX), in the service of Hob; cf. Hobman.

The days of the week are puzzling, the only one at all common being Munday, though most of the others are found in earlier nomenclature. We should rather expect special attention to be given to Sunday and Friday, and, in fact, Sonntag and Freytag are by far the most usual in German, while Dimanche and its perversions are common in France, and Vendredi also occurs. This makes me suspect some other origin, probably local, for Munday, the more so as Fr. Dimanche, Demange, etc., is often for the personal name Dominicus, the etymology remaining the same as that of the day-name, the Lord's day. Parts of the day seem to survive in Noon, Eve, and Morrow, but Noon is local, Fr. Noyon (cf. Moon, earlier Mohun, from Moyon), Eve is the mother of mankind, and Morrow is for moor-wro, the second element being Mid. Eng. wra, comer, whence Wray.

MONTH NAMES

We find the same difficulty with the names of the months. Several of these are represented in French, but our March has four other origins, from March in Cambridgeshire, from march, a boundary, from marsh, or from Mark; while May means in Mid. English a maiden (Chapter XXI), and is also a dim. of Matthew (Chapter IX). The names of the seasons also present difficulty. Spring usually corresponds to Fr. La Fontaine (Chapter II), but we find also Lent, the old name for the season, and French has Printemps. [Footnote: The cognate Ger. Lenz is fairly common, hence the frequency of Lent in America.] Summer and Winter [Footnote: Winter was one of Hereward's most faithful comrades.] are found very early as nicknames, as are also Frost and Snow; but why always Summers or Somers with s and Winter without? [Footnote: Two other common nicknames were Flint and Steel.] The latter has no doubt in many cases absorbed Vinter, vintner (Chapter III) but this will not account for the complete absence of genitive forms. And what has become of the other season? We should not expect to find the learned word "autumn," but neither Fall nor Harvest, the true English equivalents, are at all common as surnames.

I regard this group, viz. days, months, seasons, as one of the least clearly accounted for in our nomenclature, and cannot help thinking that the more copious examples which we find in French and German are largely distorted forms due to the imitative instinct, or are susceptible of other explanations. This is certainly true in some cases, e.g. Fr. Mars is the regular French development of Medardus, a saint to whom a well-known Parisian church is dedicated; and the relationship of Janvier to Janus may be via the Late Lat. januarius, for janitor, a doorkeeper.

[Footnote: Medardus was the saint who, according to Ingoldsby, lived largely on oysters obtained by the Red Sea shore. At his church in Paris were performed the 'miracles' of the Quietists in the seventeenth century. When the scenes that took place became a scandal, the government intervened, with the result that a wag adorned the church door with the following:

CHAPTER X METRONYMICS

(Bleak House, ch. iv.)

Bardsley first drew attention to the very large number of surnames derived from an ancestress. His views have been subjected to much ignorant criticism by writers who, taking upon themselves the task of defending medieval virtue, have been unwilling to accept this terrible picture of the moral condition of England, etc. This anxiety is misplaced. There are many reasons, besides illegitimacy, for the adoption of the mother's name. In medieval times the children of a widow, especially posthumous children, would often assume the mother's name. Widdowson itself is sufficiently common. In the case of second marriages the two families might sometimes be distinguished by their mothers' names. Orphans would be adopted by female relatives, and a medieval Mrs. Joe Gargery would probably have impressed her own name rather than that of her husband on a medieval Pip. In a village which counted two Johns or Williams, and few villages did not, the children of one might assume, or rather would be given by the public voice, the mother's name. Finally, metronymics can be collected in hundreds by anyone who cares to work through a few early registers.

FEMALE FONT-NAMES

Thus, in the Lancashire Inquests 205-1307 occur plenty of people described as the sons of Alice, Beatrice, Christiana, Eda, Eva, Mariot, Matilda, Quenilda, [Footnote: An Anglo-Saxon name, Cynehild, whence Quennell.] Sibilla, Ysolt. Even if illegitimacy were the only reason, that would not concern the philologist.

Female names undergo the same course of treatment as male names. Mary gave the diminutives Marion and Mariot, whence Marriott. It was popularly shortened into Mal (cf. Hal for Harry), which had the diminutive Mally. From these we have Mawson and Malleson, the former also belonging to Maud. Mal and Mally became Mol and Molly, hence Mollison. The rimed forms Pol, Polly are later, and names in Pol- usually belong to Paul (Chapter IX). The name Morris has three other origins (the font-name Maurice, the nickname Moorish, and the local marsh), but both Morris and Morrison are sometimes to be referred to Mary. Similarly Margaret, popularly Mar-get, became Mag, Meg, Mog, whence Meggitt, Moxon, etc. The rarity of Maggot is easily understood, but Poll Maggot was one of Jack Sheppard's accomplices and Shakespeare uses maggot-pie for magpie (Macbeth, iii, 4). Meg was rimed into Peg, whence Peggs, Mog into Pog, whence Pogson, and Madge into Padge, whence Padgett, when this is not for Patchett (Chapter IX), or for the Fr. Paget, usually explained as Smallpage. The royal name Matilda appears in the contracted Maud, Mould, Moule, Mott, Mahood (Old Fr. Maheut). Its middle syllable Till gave Tilly, Tillson and the dim. Tillet, Tillot, whence Tillotson. From Beatrice we have Bee, Beaton and Betts, and the northern Beattie, which are not connected with the great name Elizabeth. This is in medieval rolls represented by its cognate Isabel, of which the shortened form was Bell (Chapter I), or Ib, the latter giving Ibbot, Ibbotson, and the rimed forms Tib-, Nib-, Bib-, Lib-. Here also belong Ebbs and Epps rather than to the Anglo-Sax. Ebba (Chapter VII).

Many names which would now sound somewhat ambitious were common among the medieval peasantry and are still found in the outlying parts of England, especially Devon and Cornwall. Among the characters in Mr. Eden Phillpotts's Widecombe Fair are two sisters named Sibley and Petronell. From Sibilla, now Sibyl, come most names in Sib-, though this was used also as a dim. of Sebastian (see also Chapter VII), while Petronilla, has given Parnell, Purnell. As a female name it suffered the eclipse to which certain names are accidentally subject, and became equivalent to wench. References to a "prattling Parnel" are common in old writers, and the same fate overtook it in French—

"Taisez-vous, péronnelle" (Tartufe, i. 1).

Mention has already been made of the survival of Guinevere (Chapter VIII). From Cassandra we have Cash, Cass, Case, and Casson, from Idonia, Ide, Iddins, Iddison; these were no doubt confused with the derivatives of Ida. William filius Idae is in the Fine Rolls of John's reign, and John Idonyesone occurs there, temp. Edward I. Pim, as a female font-name, may be from Euphemia, and Siddons appears to belong to Sidonia, while the pretty name Avice appears as Avis and Haweis. From Lettice, Lat. laetitia, joy, we have Letts, Lettson, while the corresponding Joyce, Lat. jocosa, merry, has become confused with Fr. Josse (Chapter I). Anstey, Antis, is from Anastasia, Precious from Preciosa, and Royce from Rohesia.

DOUBTFUL CASES

It is often difficult to separate patronymics from metronymics. We have already seen (Chapter VI) that names in Ed- may be from Eda or from Edward, while names in Gil- must be shared between Julian, Juliana, Guillaume, Gilbert, and Giles. There are many other cases like Julian and Juliana, e.g. Custance is for Constance, but Cust may also represent the masculine Constant, while among the derivatives of Philip we must not forget the warlike Philippa. Or, to take pairs which are unrelated, Kitson may be from Christopher or from Catherine, and Mattison from Matthew or from Martha, which became Matty and Patty, the derivatives of the latter coalescing with those of Patrick (Chapter VI). It is obvious that the derivatives of Alice would be confused with those of Allen, while names in El- may represent Elias or Eleanor. Also names in Al- and El- are sometimes themselves confused, e.g. the Anglo-Saxon AElfgod appears both as Allgood and Elgood. More Nelsons are derived from Neil, i.e. Nigel, than from Nell, the rimed dim. of Ellen. Emmett is a dim. of Emma, but Empson may be a shortened Emerson from Emery (Chapter VIII). The rather commonplace Tibbles stands for both Theobald and Isabella, and the same is true of all names in Tib- and some in Teb-. Lastly, the coalescence of John, the commonest English font-name, with Joan, the earlier form of Jane, was inevitable, while the French forms Jean and Jeanne would be undistinguishable in their derivatives. These names between them have given an immense number of surnames, the masculine or feminine interpretation of which must be left to the reader's imagination.

CHAPTER XI LOCAL SURNAMES

There is an idea cherished by some people that the possession of a surname which is that of a village or other locality points to ancestral ownership of that region. This is a delusion. In the case of quite small features of the landscape, e.g. Bridge, Hill, the name was given from place of residence. But in the case of counties, towns and villages, the name was usually acquired when the locality was left. Thus John Tiler leaving Acton, perhaps for Acton's good, would be known in his new surroundings as John Acton. A moment's reflection will show that this must be so. Scott is an English name, the aristocratic Scotts beyond the border representing a Norman family Escot, originally of Scottish origin. English, early spelt Inglis, is a Scottish name. The names Cornish and Cornwallis first became common in Devonshire, as Devenish did outside that county. French and Francis, Old Fr. le franceis, are English names, just as Langlois (l'Anglais) is common in France. For the same reason Cutler is a rare name in Sheffield, where all are cutlers. By exception the name Curnow, which is Cornish for a Cornishman, is fairly common in its native county, but it was perhaps applied especially to those inhabitants who could only speak the old Cornish language.

CLASSES OF LOCAL NAMES

The local name may range in origin from a country to a plant (France, Darbishire, Lankester, Ashby, Street, House, Pound, Plumptre, Daisy), and, mathematically stated, the size of the locality will vary in direct proportion to the distance from which the immigrant has come. Terentius Afer was named from a continent. I cannot find a parallel in England, but names such as the nouns France, Ireland, Pettingell (Portugal), or the adjectives Dench, Mid. Eng. dense, Danish, Norman, Welsh, (Walsh, Wallis, etc.), Allman (Allemand), often perverted to Almond, were considered a sufficient mark of identification for men who came from foreign parts. But the untravelled inhabitant, if distinguished by a local name, would often receive it from some very minute feature of the landscape, e.g. Solomon Daisy may have been descended from a Robert Dayeseye, who lived in Hunts in 1273. It is not very easy to see how such very trifling surnames as this last came into existence, but its exiguity is surpassed in the case of a prominent French airman who bears the appropriately buoyant name of Brindejonc, perhaps from some ancestor who habitually chewed a straw.

An immense number of our countrymen are simply named from the points of the compass, slightly disguised in Norris, Anglo-Fr. le noreis, [Footnote: The corresponding le surreis is now represented by Surridge.] Sotheran, the southron, and Sterling, for Easterling, a name given to the Hanse merchants. Westray was formerly le westreis. A German was to our ancestors, as he still is to sailors, a Dutchman, whence our name Douch, Ger. deutsch, Old High Ger. tiutisc, which, through Old French tieis, has given Tyas. [Footnote: Tyars, or Tyers, which Bardsley puts with this, seems to be rather Fr. Thiers, Lat. tertius.]

But not every local name is to be taken at its face value. Holland is usually from one of the Lancashire Hollands, and England may be for Mid. Eng. ing-land, the land of Ing (cf. Ingulf, Ingold, etc.), a personal name which is the first element in many place-names, or from ing, a meadow by a stream. Holyland is not Palestine, but the holly-land. Hampshire is often for Hallamshire, a district in Yorkshire. Dane is a variant of Mid. Eng. dene, a valley, the inhabitant of Denmark having given us Dench (Chapter XI) and Dennis (le daneis). Visitors to Margate will remember the valley called the Dane, which stretches from the harbour to St. Peters. Saxon is not racial, but a perversion of sexton (Chapter XVII). Mr. Birdofredum Sawin, commenting on the methods employed in carrying out the great mission of the Anglo-Saxon race, remarks that—

The name Cockayne was perhaps first given derisively to a sybarite—

but it may be an imitative form of Coken in Durham.

Names such as Morris, i.e. Moorish, or Sarson, i.e. Saracen (but also for Sara-son), are rather nicknames, due to complexion or to an ancestor who was mine host of the Saracen's Head. Moor is sometimes of similar origin. Russ, like Rush, is one of the many forms of Fr. roux, red-complexioned (Chapter II). Pole is for Pool, the native of Poland being called Polack—

"He smote the sledded Polack on the ice" (Hamlet, I. i).

But the name Pollock is local (Renfrewshire).

COUNTIES AND TOWNS

As a rule it will be found that, while most of our counties have given family names, sometimes corrupted, e.g. Lankshear, Willsher, Cant, Chant, for Kent, with which we may compare Anguish for Angus, the larger towns are rather poorly represented, the movement having always been from country to town, and the smaller spot serving for more exact description. An exception is Bristow (Bristol), Mid. Eng. brig-stow, the place on the bridge, the great commercial city of the west from which so many medieval seamen hailed; but the name is sometimes from Burstow (Surrey), and there were possibly smaller places called by so natural a name, just as the name Bradford, i.e. broad ford, may come from a great many other places than the Yorkshire wool town. Rossiter is generally for Rochester, but also for Wroxeter (Salop); Coggeshall is well disguised as Coxall, Barnstaple as Bastable, Maidstone as Mayston, Stockport as Stopford. On the other hand, there is not a village of any antiquity but has, or once had, a representative among surnames.

NAMES PRECEDED BY DE

The provinces and towns of France and Flanders have given us many common surnames. From names of provinces we have Burgoyne and Burgin, Champain and Champneys (Chapter II), Gascoyne and Gaskin, Mayne, Mansell, Old Fr. Mancel (manceau), an inhabitant of Maine or of its capital Le Mans, Brett and Britton, Fr. le Bret and le Breton, Pickard, Power, sometimes from Old Fr. Pohier, a Picard, Peto, formerly Peitow, from Poitou, Poidevin and Puddifin, for

Poitevin, Loring, Old Fr. le Lohereng, the man from Lorraine, assimilated to Fleming, Hammy, an old name for Hainault, Brabazon, le Brabançon, and Brebner, formerly le Brabaner, Angwin, for Angevin, Flinders, a perversion of Flanders, Barry, which is sometimes for Berri, and others which can be identified by everybody.

Among towns we have Allenson, Alençon, Amyas, Amiens, Ainger, Angers, Aris, Arras, Bevis, Beauvais, Bullen, Boulogne, Bloss, Blois, Bursell, Brussels, Callis and Challis, Calais, Challen, from one of the French towns called Chalon or Chalons, Chaworth, Cahors, Druce, Dreux, Gaunt, Gand (Ghent), Luck, Luick (Liege), Loving, Louvain, Malins, Malines (Mechlin), Raynes, Rennes and Rheims, Roan, Rouen, Sessions, Soissons, Stamp, Old Fr. Estampes (ttampes), Turney, Tournay, etc. The name de Verdun is common enough in old records for us to connect with it both the fascinating Dolly and the illustrious Harry. [Footnote added by scanner: Some modern readers might not realise that Weekley was referring to Harry Vardon, a famous golfer of the late nineteenth and early twentieth century. Dolly Varden was a character in Dickens’ "Barnaby Rudge", a pretty girl in flowered hats and skirts. Her name was borrowed for various clothing styles, breeds of flowers, shows, theatres and even angling fishes among other things. There seem to have been several references to "the fascinating Dolly Varden", though the expression does not occur in the book.] To the above may be added, among German towns, Cullen, Cologne, and Lubbock, Luebeck, and, from Italy, Janes, Gênes (Genoa), Janaway or Janways, i.e. Genoese, and Lombard or Lombard. Familiar names of foreign towns were often anglicized. Thus we find Hamburg called Hamborough, Bruges Bridges, and Tours Towers.

To the town of Angers we sometimes owe, besides Ainger, the forbidding names Anger and Danger. In many local names of foreign origin the preposition de has been incorporated, e.g. Dalmain, d'Allemagne, sometimes corrupted into Dallman and Dollman, though these are also for Doleman, from the East Anglian dole, a boundary, Dallison, d'Alençon, Danvers, d'Anvers, Antwerp, Devereux, d'Évreux, Daubeney, Dabney, d'Aubigny, Disney, d'Isigny, etc. Doyle is a later form of Doyley, or Dolley, for d'Ouilli, and Darcy and Durfey were once d'Arcy and d'Urfé. Dew is sometimes for de Eu. Sir John de Grey, justice of Chester, had in 1246 two Alice in Wonderland clerks named Henry de Eu and William de Ho. A familiar example, which has been much disputed, is the Cambridgeshire name Death, which some of its possessors prefer to write D'Aeth or De Ath. Bardsley rejects this, without, I think, sufficient reason. It is true that it occurs as de Dethe in the Hundred Rolls, but this is not a serious argument, for we find also de Daubeney (Chapter XI), the original de having already been absorbed at the time the Rolls were compiled. This retention of the de is also common in names derived from spots which have not become recognized place-names; see Chapter XIV.

But to derive a name of obviously native origin from a place in France is a snobbish, if harmless, delusion. There are quite enough moor leys in England without explaining Morley by Morlaix. To connect the Mid. English nickname Longfellow with Longueville, or the patronymic Hansom (Chapter III) with Anceaumville, betrays the same belief in phonetic epilepsy that inspires the derivation of Barber from the chapelry of Sainte-Barbe. The fact that there are at least three places, in England called Carrington has not prevented one writer from seeking the origin of that name in the appropriate locality of Charenton.

CHAPTER XII SPOT NAMES

Verstegan's couplet, even if it be not strictly true, makes a very good text for a discourse on our local names. The ham, or home, and the ton, or town, originally an enclosure (cf. Ger. Zaun, hedge), were, at any rate in a great part of England, the regular nucleus of the village, which in some cases has become the great town and in others has decayed away and disappeared from the map. In an age when wool was our great export, flock keeping was naturally a most important calling, and the ley, or meadow land, would be quickly taken up and associated with human activity. When bridges were scarce, fords were important, and it is easy to see how the inn, the smithy, the cartwright's booth, etc., would naturally plant themselves at such a spot and form the commencement of a hamlet.

ELEMENTS OF PLACE-NAMES

Each of these four words exists by itself as a specific place-name and also as a surname. In fact Lee and Ford are among our commonest local surnames. In the same way the local origin of such names as Clay and Chalk may be specific as well as general. But I do not propose to deal here with the vast subject of our English village names, but only with the essential elements of which they are composed, elements which were often used for surnominal purposes long before the spot itself had developed into a village. [Footnote: A good general account of our village names will be found in the Appendix to Isaac Taylor's Names and their Histories. It is reprinted as chapter xi of the same author's Words and Places (Everyman Library). See also Johnston's Place-names of England and Wales, a glossary of selected names with a comprehensive introduction. There are many modern books on the village names of various counties, e.g. Bedfordshire, Berkshire, Cambridgeshire, Hertfordshire, Huntingdonshire, Suffolk (Skeat), Oxfordshire (Alexander), Lancashire (Wyld and Hirst), West Riding of Yorkshire (Moorman), Staffordshire, Warwickshire, Worcestershire (Duignan), Nottinghamshire (Mutschmann), Gloucestershire (Baddeley), Herefordshire (Bannister), Wiltshire (Elsblom), S.W. Yorkshire (Goodall), Sussex (Roberts), Lancashire (Sephton), Derbyshire (Walker), Northumberland and Durham (Mawer).] Thus the name Oakley must generally have been borne by a man who lived on meadow land which was surrounded or dotted with oak-trees. But I should be shy of explaining a given village called Oakley in the same way, because the student of place-names might be able to show from early records that the place was originally an ey, or island, and that the first syllable is the disguised name of a medieval churl. These four simple etymons themselves may also become perverted. Thus -ham is sometimes confused with -holm (Chapter XII), -ley, as I have just suggested, may in some cases contain -ey, -ton occasionally interchanges with -don and -stone, and -lord with the French -fort (Chapter XIV).

In. this chapter will be found a summary of the various words applied by our ancestors to the natural features of the land they lived on. To avoid too lengthy a catalogue, I have classified them under the three headings—

reserving for the next chapter the names due to man's interference with the scenery, e.g. roads, buildings, enclosures, etc.

They are mostly Anglo-Saxon or Scandinavian, the Celtic name remaining as the appellation of the individual hill, stream, etc. (Helvellyn, Avon, etc.). The simple word has in almost all cases given a fairly common surname, but compounds are of course numerous, the first element being descriptive of the second, e.g. Bradley, broad lea, Radley and Ridley, red lea, Brockley, brook lea or badger lea (Chapter XXIII), Beverley, beaver lea, Cleverley, clover lea, Hawley, hedge lea, Rawnsley, raven's lea, and so ad infinitum. In the oldest records spot names are generally preceded by the preposition at, whence such names as Attewell, Atwood, but other prepositions occur, as in Bythesea, Underwood and the hybrid Suttees, on Tees. Cf. such French names as Doutrepont, from beyond the bridge.

One curious phenomenon, of which I can offer no explanation, is that while many spot names occur indifferently with or without -s, e.g. Bridge, Bridges; Brook, Brooks; Platt, Plaits, in others we find a regular preference either for the singular or plural form. [Footnote: In some cases no doubt a plural, in others a kind of genitive due to the influence of personal names, such as Wills, Perkins, etc.] Compare the following couples:

to which many more might be added. So we find regularly Nokes but Nash (Chapter III), Beech but Willows. The general tendency is certainly towards the -s forms in the case of monosyllables, e.g. Banks, Foulds, Hayes, Stubbs, Thwaites, etc., but we naturally find the singular in compounds, e.g. Windebank (winding), Nettlefold, Roundhay, etc.

There is also a further problem offered by names in -er. We know that a Waller was a mason or wall-builder, but was a Bridger really a Pontifex, [Footnote: An example of a Latinized name. Cf. Sutor, Faber, and the barbarous Sartorius, for sartor, a tailor. Pontifex may also be the latinized form of Pope or Bishop. It is not known why this title, bridge-builder, was given to high-priests.] did he merely live near the bridge, or was he the same as a Bridgman, and what was the latter? Did Sam Weller's ancestor sink wells, possess a well, or live near someone else's well? Probably all explanations may be correct, for the suffix may have differed in meaning according to locality, but I fancy that in most cases proximity alone is implied. The same applies to many cases of names in -man, such as Hillman, Dickman (dyke), Parkman.

Many of the words in the following paragraphs are obsolete or survive only in local usage. Some of them also vary considerably in meaning, according to the region in which they are found. I have included many which, in their simple form, seem too obvious to need explanation, because the compounds are not always equally clear.

HILL AND DALE

We have a fair number of Celtic words connected with natural scenery, but they do not as a rule form compounds, and as surnames are usually found in their simple form. Such are Cairn, a stony hill, Crag, Craig, and the related Carrick and Creagh, Glen or Glynn, and Lynn, a cascade. Two words, however, of Celtic origin, don, or down, a hill, and combe, a hollow in the hills, were adopted by the Anglo-Saxons and enter into many compounds. Thus we find Kingdon, whence the imitative Kingdom, Brandon, from the name Brand (Chapter VII), Ashdown, etc. The simple Donne or Dunne is sometimes the Anglo-Saxon name Dunna, whence Dunning, or a colour nickname, while Down and Downing may represent the Anglo-Sax. Duna and Duning (Chapter VII). From Combe, used especially in the west of England, we have Compton, and such compounds as Acomb, at combe, Addiscombe, Battiscombe, etc. But Newcomb is for Newcome (Chapter II). See also Slocomb (Chapter XXI).

HILLS

The simple Hill and Dale are among our common surnames. Hill also appears as Hull and is easily disguised in compounds, e.g. Brummel for broom-hill, Tootell and Tuttle for Toothill, a name found in many localities and meaning a hill on which a watch was kept. It is connected with the verb to tout, originally to look out

"David dwellide in the tote hil" (Wyc, 2 Sam. v. 9).

We have Dale and its cognate Dell in Swindell (swine), Tindall (Tyne), Twaddell, Tweddell (Tweed), etc. —

Other names for a hill are Fell (Scand.), found in the lake country, whence Grenfell; and Hough or How (Scand.), as in the north country names Greenhow, Birchenough.

This is often reduced to -o, as in Clitheroe, Shafto, and is easily confused with scough, a wood (Scand.), as in Briscoe (birch), Ayscough (ash).

In the north hills were also called Law and Low, with such compounds as Bradlaugh, Whitelaw, and Harlow. To these must be added Barrow, often confused with the related borough (Chapter XIII). Both belong to the Anglo-Sax. beorgan, to protect, cover. The name Leatherbarrow means the hill, perhaps the burial mound, of Leather, Anglo-Sax. Hlothere, cognate with Lothair and Luther.

A hill-top was Cope or Copp. Chaucer uses it of the tip of the Miller's nose

Another name for a hill-top appears in Peak, Pike, Peck, or Pick, but the many compounds in Pick-, e.g. Pickbourne, Pickford, Pickwick, etc., suggest a personal name Pick of which we have the dim. in Pickett (cf. Fr. Picot) and the softened Piggot. Peak may be in some cases from the Derbyshire Peak, which has, however, no connection with the common noun peak. A mere hillock or knoll has given the names Knapp, Knollys or Knowles, Knock, and Knott. But Knapp may also be for Mid. Eng. nape, cognate with knave and with Low Ger. Knappe, squire—

Redknap, the name of a Richmond boat-builder, is probably a nickname, like Redhead. A Knapper may have lived on a "knap," or may have been one of the Suffolk flint-knappers, who still prepare gun-flints for weapons to be retailed to the heathen.

Knock and Knocker are both Kentish names, and there is a reef off Margate known as the Kentish Knock. We have the plural Knox (cf. Bax, Settlements and Enclosures, Chapter XIII). Knott is sometimes for Cnut, or Canute, which generally becomes Nutt. Both have got mixed with the nickname Nott.

A green knoll was also called Toft (Scand.), whence Langtoft, and the name was used later for a homestead. From Cliff we have Clift, [Footnote: This may also be from Mid. Eng, clift, a cleft.] with excrescent -t, and the cognates Cleeve and Clive. Compounds of Cliff are Radcliffe (red), Sutcliffe (south), Wyclif (white). The c- sometimes disappears in compounds, e.g. Cunliffe, earlier Cunde-clive, and Topliff; but Ayliffe is for AElfgifu or AEthelgifu and Goodliffe from Godleof (cf. Ger. Gottlieb). The older form of Stone appears in Staines, Stanhope, Stanton, etc. Wheatstone is either for "white stone" or for the local Whetstone (Middlesex). In Balderstone, Johnston, Edmondstone, Livingstone, the suffix is -ton, though the frequence of Johnston points to corruption from Johnson, just as in Nottingham we have the converse case of Beeson from the local Beeston. In Hailstone the first element may be Mid. Eng, half, holy. Another Mid. English name for a stone appears in Hone, now used only of a whetstone.

A hollow or valley in the hillside was called in the north Clough, also spelt Clow, Cleugh (Clim o' the Cleugh), and Clew. The compound Fairclough is found corrupted into Faircloth. Another obscure northern name for a glen was Hope, whence Allsop, Blenkinsop, the first element in each being perhaps the name of the first settler, and Burnup, Hartopp, (hart), Harrap (hare), Heslop (hazel).

Gill (Scand.), a ravine, has given Fothergill, Pickersgill, and Gaskell, from Gaisgill (Westmorland). These, like most of our names connected with mountain scenery, are naturally found almost exclusively in the north. Other surnames which belong more or less to the hill country are Hole, found also as Holl, Hoole, and Hoyle, but perhaps meaning merely a depression in the land, Ridge, and its northern form Rigg, with their compounds Doddridge, Langridge, Brownrigg, Hazelrigg, etc. Ridge, Rigg, also appear as Rudge, Rugg. From Mid. Eng. raike, a path, a sheep-track (Scand.), we get Raikes and perhaps Greatorex, found earlier as Greatrakes, the name of a famous faith-healer of the seventeenth century.

WOODLAND AND PLAIN

The compounds of Wood itself are very numerous, e.g. Braidwood, Harwood, Norwood, Sherrard and Sherratt (Sherwood). But, in considering the frequency of the simple Wood, it must be remembered that we find people described as le wode, i.e. mad (cf. Ger. Wut, frenzy), and that mad and madman are found as medieval names

As a suffix -wood is sometimes a corruption of -ward, e.g. Haywood is occasionally for Hayward, and Allwood, Elwood are for Aylward, Anglo-Sax. AEthelweard. Another name for a wood was Holt, cognate with Ger. Holz—

Hurst or Hirst means a wooded hill (cf. Ger. Horst), and Shaw was once almost as common a word as wood itself—