U?u^i^e^^U^ yi^ ^a^^rm^ c^rc^. c,x^^>^^t;t^ ^/?-i /// f ( • /j/yi/ ::■;•/: JB.5^rriSH PLACE-NAMES On a superficial survey, all the names of places recorded in our Ordnance Maps of England would seem to be of English origin. A minute examination of these will show, however, that numerous elements involved in them are not to be traced to an English source. A further investigation of these " foreign " elements will enable us to class them roughly into Celtic, Latin, and Scandinavian. There is, if we take in the rest of Britain, probably also a number of them which belong to a pre-Celtic language ; that is to say, designations handed down from a race which was in possession of these islands when the first Celtic invaders entered them. Some of our linguistic experts think that these pre-Celtic peoples were represented by the Picts, who appear to have occupied certain parts of Scotland and of Ireland when our first great English historian — the Venerable Bede — was writing his History, nearly twelve hundred years ago. In his time, Bede says, and the Saxon Chronicle adopts the statement, the following races, as represented by languages, were known in Britain : English, British, Scot-ish (i.e. Irish), Pictish, and Book-Latin (i.e. the language of the Romanized Britons). The Celtic languages to which Bede here refers under the designations British, Scot-ish, and Pictish, have been scientifically classified of late according to their local survivals into Goidelic, (which embraces Irish, Manx, and Scottish Gaelic), and Brythonic which includes Ancient British (and Gaulish), Cornish, Breton, Cymric (i.e. Welsh), and probably Pictish. The ancient Continental home of the Celtic-speaking people was, according to latest authorities, the territory lying between the Rhine, the Main, and the Danube, whence according to D'Arbois de Jubainville they spread into North- West Germany, the GoideHc branch about 800 b.c passing into the British Islands. These were followed, the same writer alleges, about 630 B.C. by a Brythonic (?) people called Belgae, who, driven by Germanic tribes out of the region between the Elbe and the Rhine, occupied North-Eastern Gaul, from whence some passed into Britain. According to Julius Caesar {De Bell. Gal. i) the INTRODUCTION 9 Belgae differed from the Gauls (called Galli, he says, in Latin, Celtae, in their own speech) in language and customs, and were divided from them by the Marne and Seine. Caesar tells us, moreover (ibid. v. 12), that the maritime parts of Britain were occupied by those Belgae who, passing over for booty, had settled there, and that they were there known by the names of the states whence they had sprung. This is confirmed by the names Venta Belgarum (Winchester) and Calleva (or Galleva) Atrehatum (Silchester), the Atrebates* occupying at this period the region round Arras in which their name is still preserved. According to D'Arbois de Jubainville, Celtic invasions of Spain (500 B.C.) and of Bohemia and Italy (400 b.c.) followed. Greece and Galatia in Asia Minor were invaded by the Celts, in the third century b.c As to the Belgae being the first Brythonic invaders of Britain, it is clearly against the statement of Caesar, and other- wise not credible. A tribe called Scots belonging to the Goidels are known to have passed from Ireland into Alban (Scotland) in the fourth century of our era, the latter country being then occupied by Bry- thons and Picts.'^ In the fifth and sixth centuries a.d. a southern Brythonic people, under pressure from invading Teutonic tribes, passed over to Armorica in Gaul, in which Gallic speech had then died out, and gave their name to the country — Brittany. The influence of each of these Celtic peoples is still to be traced in our place-names. A few native territorial and Celtic tribe-names were familiar to the Romans before our era. Kent, with its white cliffs, was known to Julius Caesar in 54 b.c by the name Cantium {Cant'to-n\ a Latinized form of the native appellation. The * Rolemy (ii. 3. 13) places among the Belgae the towns Iskalis, Hydata Therma (i.e. "Hot Waters" = Bath) and Venta ( = Winchester). See later on the Summary of Ptolemy's Geography of Albion. ^ But see later on Sir John Rhys's view that Goidels were in Alban much earlier. 10 BRITISH PLACE-NAMES Tameses {Tamesa in Ptolemy) was also known to him, and the tribes Belgae and Trinovantes dwelling not far from it, as was also by report Mona (the island of Anglesea) which was captured a little over one hundred years later (a.d. 6i) by the Roman General Suetonius Paulinus, and of which the name still survives in the last syllable of Carnar^'o«. Julius Caesar received the submission of certain tribes occupying the southern part of Britain; the Cenimagni^^ the Segontiaci'^ , the Ancalites^^ the Cassi*, and the Bibroci, and the last seems still to survive in our Berk-^\xt. The early form of the latter is Berruc-scire, Baerruc-scire, Bearruc-scire (in a charter ascribed to A.D. 931), and Barroc-scire (in a British Museum MS. of A.D. 1 050-1 065). We shall see, from instances given later on, that an internal "b" in a Latinized Celtic name was pro- nounced as "v" and that it tended to disappear altogether. Thus Bivroc would naturally become Birruc with a compen- ^ Gluck (Celtic names in Caesar's de B. G.) sees in this name Ceno — distant, and in magni, the antecedent of Welsh man (Irish wa^) conclude the list. Pedersen (ibid. p. 31) describes thus shortly the modifications to which this system was subjected in Celtic. The distinction in the first place between long and short vowels is maintained. The short and long diphthongs are no longer distinguished. The diphthongs are in part resolved into simple vowels out of which later new diphthongs are formed. From the syllable-making liquids arise groups of consonantal i6 BRITISH PLACE-NAMES liquids and vowels. " p " has become a breathing {h) where it has not vanished entirely. The velars and palatals have ceased to be distinguished. The labio-velars have in part become labials (^" in all Celtic languages, ^" and /^% in Gaulish and British, while, on the other hand, g^h had lost its labialization in all Celtic languages, and also k^ and k^h in Irish after the period of the Ogam-inscriptions). The tenues had fallen in with the aspirated tenues completely at an early date. The mediae likewise had coincided with the aspirated mediae, but traces of a special articulation of the aspi- rated mediae were still to be detected, w (v), j, the explosives and the consonantal liquids were represented in Celtic in a double manner, that is, besides the normal pronunciation in the beginning of a word, there was a relaxed languid utterance (with wider opening of the mouth), which occurred, among other instances, between vowels. This modification of utterance is called Lenition or aspiration. Pedersen traces, in vol. i. of the work referred to, the modifications of the foregoing sounds in the various branches of Celtic speech. Short a is preserved, for instance, in Irish Caihir { = a town), and in its Welsh form Cader (=a hill-fortress), both of which are compared with Latin Cater-va (=a troop). Short e is also preserved, but can become by umlaut y. Before a nasal and an explosive it becomes i, as instanced in Irish Linn {^-^Lenda, liquid, and Lind^ a pool), Welsh Llyn, the AiVSov of Ptolemy =. Lindocob'na (Bede, ii. i6) = Lincoln, and in Lindi-macus {'=.*Lmdi magos^ now the Limmat, a stream running out of Zurich lake). Middle Irish Glend, gen. Glinde (= valley) and Welsh Glynn [together with Welsh Glann {=bank) cognate with Danish Klint (steep-bank) and old Norse Kleti-r (rock)] are further instances which enter into place- names in Great Britain. Short i is preserved in Gaulish but becomes sometimes^ or e in later Celtic. Compare Vindo-gladia {Ant. liin,) now (.?) ''Bad- bury-rings", Vindo-hona, now Vienna, with the Gaulish personal name UuenA^ni and the Dorset stream Wenfrith. o is pre- served, but becomes before certain consonants ^, written y in all syllables before the last. Cf. Welsh mynydd (mountain) and old Breton, -monid, Cornish meneth, the latter with umlaut. Pedersen proceeds to show the fortunes of all the vowels, together with those of the consonants, in the various branches of Celtic speech. As references will be made to this and other INTRODUCTION 17 authorities as each case arises it is not necessary to proceed here further with Celtic sound-changes. A few illustrations, however, of the changes to which the Indo- European vowels and other sounds were subjected in the Teutonic languages may well find a place here (see Brugmann's Grundrtss, § 35 et seq., and Kluge, in Paul's Grundriss^ pp. 300 et seq.). Indo-European 2 became Germanic /; before r and h =€ # „ „ « ; after a in the following u „ „ tt [syllable =6 ^ i) j> ^ ei „ „ t eu „ „ eu ^ >» >» ^ „ ,, a; this and a too are peculiar bi „ „ fl/ [to Germanic vocalization bu „ „ au ai „ „ at au „ „ au The Indo-European explosives k, /, p became respectively ;^ ( = h) /k,/. The first stage in the/>, k (kv) " Shift " according to Brugmann was when they came before / and s. ts became ss^ and ik became sk at about the same time. Then followed the general shift of the tenues, except when preceded immediately by J, or when they came under *' Verner's Law '*. The original mediae g, d, b, became k, /, /. As a consequence of the shift of original k, /, p, it may be assumed, for instance, that historical Teutonic words beginning with any of these letters had originally other initials. Torp's Worlschalz der germanischen Spracheinheit, Gottingen, 1 909, gives, for instance, some forty-five early Teutonic stems beginning with />, with their equivalents in the various dialects. Most of these are traced to Indo-European originals beginning with bh (3), others are onomatopoetic, others have been borrowed from non- Teutonic languages, and the residuum consists of a few unex- plained words, such, to take English examples, as play^ plough^ i8 BRITISH PLACE-NAMES and path. The last word seems to represent an Indo-European original, instanced in Greek TraVos and Irish ^M = ford (=orig. "^Pdia) and was possibly borrowed, as ^chT2iditT{Handehgeschichte, 13) suggests, from the Celtic, before initial/* had disappeared from the latter language. Sufficient instances of Teutonic sound- changes have been given here to show the principles. Further examples will be cited as each case discussed requires. I. THE ROMAN OCCUPATION After the departure of Julius Caesar from Britain 54 b.c, the history of this country remains a blank till a.d. 43. It is true that the Roman Emperor Caligula — who had harboured an exiled British chief — wrote to the Senate from Gaul in a.d. 40, boasting that he had conquered the entire country, but he had never really set foot in Britain. For the progress of the Roman arms in Britain from a. d. 43 to 93 we are indebted in the main to the almost contemporary historian Tacitus. He had exceptional opportunities for acquiring information, as his father-in-law was the famous Roman general Agricola, who reduced almost the whole of Britain to the Roman obedience. The army sent into Britain by the Emperor Claudius in A.D. 43 was commanded by Aulus Plautius, who brought into subjection the tribes in the neighbourhood of the Thames, and finally captured Camulodunum, the royal residence of the late Chief Cunobellinus. The Emperor Claudius, who spent alto- gether only some sixteen days in Britain, was present at the capture. In the campaign of this period, the future Emperors Vespasian and Titus (father and son) took part, the latter some twenty-five years later (a.d. 70) becoming associated with a more important historical event in his capture of Jerusalem. In the year a.d. 49 Ostorius Scapula took the command in Britain, and on finding that the neighbouring tribes had broken into the lands of those who were subject to Rome, prepared to disarm the suspected, and to occupy with camps the whole country to [? between] the Avon and Severn. He was opposed B 2 20 BRITISH PLACE-NAMES by the Iceni, already referred to, but having inflicted defeat upon them, set out against the Ceangi, and reached at length the neighbourhood of the sea facing the island of Hibernia (Tacitus, Annals, ii. 24). Can we find any traces of these events in surviving place- names? Camulodun-um (Tacitus, Annals, and Ani. Itin.) was the earliest name of Colchester, i. e. Coln-chester, but was supplanted by the term Colonia, a dignity conferred upon it later on. It means the dun or stronghold under the pro- tection of the god Camul-os (see Holder's Sprachschatz, s. v., and Rev. Celt.). The gods of the Britons were sometimes associated with rivers, as we gather from the several streams in Britain called Dee (Deva^), and from Belisama, the ancient name of the Mersey; and the stream running past this stronghold doubtless involved the name Camul.'* The same designation still survives as the name of rivers in Somerset and Cornwall, &c. As the m in Camul did not become aspirated between vowels in British speech, the first syllable must have been originally Camb (see Pedersen, V. Gravim. § 73) and have involved the idea of " Crooked ", an idea preserved also doubt- less in the designation of the god ; cf. the Irish Idol destroyed by St. Patrick, Cromm-CrHaich, in which the same notion of * Sir John Rhys {Celtic Folklore, p. 448) connects problematically Lydney in Gloucestershire with the Celtic god Nodons (found on an inscription in Lydney Park many years ago), equating it with a Cymric Ludo. Lydney, however, appears in a Charter of 972 as Lidan-ege, that is, the island of the river Lidan or Litan, a name not uncommon as a Celtic designation of streams, &c., and meaning in this connexion "broad". Cf. the Gaulish Litanobriga (" Broad-burg" according to Pedersen, Gramm. § 30) of the Antonine Itinerary. It would seem, therefore, to have nothing to do with the Nudd of Tennyson's " Idylls" or with " Ludd of the silver hand ". Torp {Wortschatz, p. 299) traces Nudd or Nuada (stem *Noudent), the name of a Celtic sea-deity, to the root involved in Anglo-Saxon Neotan = to use, to enjoy. 2 Mr. J. S. S. Glennie, in his Arthurian Localities (E. E. Text Soc, 1869), says, "Cadbury (Somerset) is mentioned in old records under the name Camelot, a name still perpetuated in the adjoining villages of Queen's Camel ROMAN OCCUPATION 21 crooked is involved. Was it this Camulodun, or some other place of the same shortened name, to which reference is made in Tennyson's lines ? — ** On either side the river lie Long fields of barley and of rye, That clothe the wold and meet the sky; And through the field the road runs by To many-tower'd Camelot." Tennyson (in Elaine) makes London the chief court of Arthur : — " . . . Arthur holding there his Court Hard on the river, nigh the place which now Is this world's hugest, let proclaim a joust At Camelot." Malory places the latter at Winchester, but his geography is not to be trusted.* The name may linger still in that "Camlet way" of which Newcome, in his Hist, of St. Albans^ writes, ''There is still (1795) visible, beside the Watling Street way, another original Roman road through the forests of Enfield Chace, called at this day Camlet way, which seems to have been the road from Verulam to Camalodunum." There is now no trace of the course of this " Camlet " way, which and West Camel." He does not give these " old records", so that I am unable to test his statement. There is another river in Somerset involving •* Camel". It was called the Camelar, on which are situated Camely and Camerton (anciently Camelarton, CoUinson's Hist, of Somerset ^ iii, p. 329). This place is mentioned in a charter ascribed to the year 961 {Cart. Sax. iii, p. 300). The river runs into the Avon above Bath. Campton, in Bedford- shire, on a tributary of the Ouse, appears, in a British Museum MS. of 11 50, as Camelton, suggesting that the stream flowing by it may then have been known as the Camel. The Hamble river in Hants, Bede's Homelea, may be an early borrowing of Camuly or a cognate word. The element " Camul " is also involved in the ancient form of Chamblay (Camul-oscus, Camblosco), of which name there are several instances, in France. The " Camulos-sesa " of the Ravenna Geographer, which was in North Britain, has not been identified. * The Arthurian geography will be dealt with in its place. 22 BRITISH PLACE-NAMES crossed the Watling Street near St. Michael's Church in Verulam, yet it is difficult to believe that it and its direction were an invention of Newcome's, who evidently saw no con- nexion between its name and the " Camulod " of Camulo- dunum, of which it would be a possible survival. The account in the Annals of Tacitus of the design of Ostorius to occupy with camps the whole country to [or between] the Avon and Severn has created difficulties. The manuscript authorities for the Annals give two readings here for the former river. In one it is called the Antona, and in the other the Avon. The seeming impossibility of identifying either of these names with any existing river in this region has led to several conjectures. Dr. Henry Bradley proposed {Academy, April 28, 1883) a most ingenious reading of the passage — previously (?) advanced by Miiller, Ptol. Geogr,^ Paris, 1883, i, p. 87 n, — making castris (by forts) into cis Tris, and pre- fixing the latter syllable to Antona-m, thus rendering the river- name Trisantona. Sir John Rhys is prepared to accept this conjecture, and thinks that this may be the early form of the Tre(h)anta, Trenta, now the Trent; but see p. 118. The geographer Ptolemy, who about a. d. 120 furnished in his Geography a list of localities in Britain, places the river Trisantona in the neighbourhood of what is now Southampton, and reasons for the correctness of this position will be given later on. The form Trahannon occurs, it is true, in Nennius (ninth century), but it seems to refer to the Trisantona of Ptolemy. Ostorius had in his advance the Iceni on his front and right. He wanted a river of some importance to protect his right flank in his advance. The Trent seems too far north, but the Nene might have served his purpose. Previous inves- tigators had accepted, without etymological considerations, this river as the Avon of Tacitus. This river, for a certain part of its course at least, has been known, from the middle of the tenth century at latest, as the Nyn or Nen. But there is good ROMAN OCCUPATION 23 reason to think that for some part of its course, possibly as far as Peterborough, it was known from early times as the Avon ^ Leland, in his Itinerary^ written about 1542, calls it, over and over again, the Avon. He says in one passage : " The ryver of Avon so windeth about Oundale Town that it almost insula- tethe it, savyng a litle by west-north-west." Oundale (Oundle) is suggestive. In the Eccl. Hist, of Bede (died a. d. 735) we have " in the province Undalum " '', but in a charter (now in possession of the Society of Antiquaries) ascribed to the tenth century, it is called in the accusative case Undalan. Dael {neut^ is a valley, and appears often in the charters as dell, which Sweet considers to be a weak form of dsel, and thus to form its accusative in -an. Dail was, like its Scandinavian equivalent daUr in the north of England, generally coupled with river- names, e. g. Dover-dael, Worcester (in a ninth-ceniury charter), and the familiar Teesdale, Wharfdale, &c. Oundle would there- fore be a worn form of Avon-dael, and thus at once confirm Leland's name of the river, and the accuracy of the text of Tacitus which reads Avonam, In Speed's Theatrwn Britanniae (16 10), moreover, Avon-well is given as in Roth well Hundred, and is doubtless the well or source of the stream that, running by Kettering, falls into the Nene. Drayton in his Polyolbion (1613) says of the Nen: "... Avon, which of long, the Britons called her," " whom by Aufona's name the Roman did renown." ^ Dr. Henry Bradley thinks that "Avon" was a generic term applied to British rivers, and that there was in all cases a specific name added to it. The Avon-Nyn would, therefore, be a transitional form, leaving either " Avon" or " Nyn", as one or other form fell out in the course of time. " S. Cett resteth on Undola near the river Nen," says The Saints of Eng- land, a work drawn up shortly before A. D. 1000 i^Die Heilig. Eng., Lieber- mann, Hanover, 1889), showing the truth of Dr. Bradley's view. '* It is right to add that the latter element in this word may be -disl (masc), meaning "portion or division ". Numerous instances of this -dcel occur in place-name compounds in the North and in such forms as Femdel or Ferthingdel = a fourth part. The A. -Sax. dal ( = valley) is traced (Torp_) to a stem meaning to bend. 24 BRITISH PLACE-NAMES As to the Severn, the Romans probably learned the name Sahrina from the Goidels (or Gaels, i. e. the Irish branch of the Celts, who are known to have occupied in the Roman period the greater part of what came afterwards to be designated Wales) and not from the Britons, who knew the river in historic times by the name Hafren. It would be difficult to say when the Britons began to use H for an initial S. The parallel between the Greek initial aspirate and the Ladn S tends to show that the change must have been very ancient. Sir John Rhys {CeUi'c Britain) would read for Ceangi, "Deceangli," who, he thinks, probably inhabited Flintshire and Cheshire, as the abbreviations Decea, Dtceang, Dece^ngl, appear on pigs of lead ^ found in the latter county, and in Staffordshire. It has been conjectured, with some degree of probability, that this name is perpetuated in the Deanery of Tegingel (now Holywell, and in the Taxatio, " Englefild "). The Deaneries often involve in their designations very ancient district names. Bowel's Hist, of Cambria, published in 1584, says (p. 10), "the feft Cantref (of North Wales) is Tegengl, and is now a part of Flintshire, having these Comots, Counsylht (now corrupted to Coleshill, Arch, Camb. Suppl., April, 1909, p. 71), Prestatyn and Ruthlan." Ptolemy locates in this region the Promontory of the Cseanganoi '\ a form which seems to confirm the ^' Ceangi '' of Tacitus. Ostorius had, therefore, probably proceeded to, and beyond, the Dee, to what is now Chester, as a funerary inscription found in the walls there shows, according to Dr. Haver- 1 The excavations at Charterhouse-on-Mendip {Times, Aug. 24, 1909) seem to show that the smelting of lead was an industry even before the amval of the Romans in this country. The pigs of lead found there with emperors' names on them are associated with silver coins, of which two are British and two are Republican. The silver of the British coins was probably procured from the lead ore. It is not without significance that the name Mendip seems to preserve in its first element the Breton Men = ore, crude-metal. The Irish form is MHn and the Welsh Mwyn, representing an earlier ^Men. 2 For variants of this form and for further discussion of the place-names occurring in Ptolemy's Geography see p. 91. ROMAN OCCUPATION 25 field {Catalogue Chester Museum, No. 54), that Roman troops were possibly in this locality about a. d. 50. The Roman General Ostorius, Tacitus tells us, was recalled from his Western campaign by the outbreak of disorders among the Brigantes. He had not mentioned these people before, but his language implies a previous contact between them and the Roman troops. Indeed, their name was seemingly well known in Rome at an early date, for there is a passing reference to them in lines of Seneca {Apocol. 12. 13-17), who died a.d. 65, not to speak of the allusion in the later (a. d. 90) writer Juvenal (Sat. xiv), who probably was prefect of a cohort in Britain. Tacitus seems to indicate that the outbreak of disorders among the Brigantes was a concern to Ostorius, because he had previously established some kind of order among them. He was not long, moreover, in bringing them into subjection again. Could these, we might naturally ask, be the Brigantes of the north, or were they some tribe of a similar designation nearer the centre of Roman dominion at this lime ? The name, in an original charter of a. d. 705 now in British Museum (*' Brigunt ford", Cart, Sax. i, p. 169), of the Brent river, which flows into the Thames below Brentford, might favour the latter hypothesis, as furnishing a possible source of the tribe-name, but the con- tinued employment by the Roman historians of ''Brigantes" to designate the tribe or tribes occupying the region north of the Humber, and stretching to the Forth, seems decisive as to the locality.^ ^ The application of a tribe-designation, as well as a river-name, to a ford is not unusual; cf. Wallingford, Britford, and on the Continent, Frankfort. Ptplemy (a. d. 150) places the Brigantes between the Selgovoi (that is the tribes who gave their names to the Solway Firth) and the Otadinoi, and makes them extend from sea to sea. He includes the following places in their territory : Vinovium = Binchesler, Catarac- tonium = Catterick, Isurium (elsewhere Isu-brigantum) = Aid-borough, Eboracum = York, Olicana = Ilkley, and a certain Camulodunura (else- where Cambodunum), which has not been identified, but which was probably near Slack ; but see note on Ptolemy's Geography, pp. 91 at seq. 26 BRITISH PLACE-NAMES Ostorius probably proceeded from Deva (Chester) via Man- cunium (Manchester), into the region of the Brigantes, by a route which was, or afterwards became, the Roman road connecting these two places. From the fact that the early form " Brigunt " has passed into the " Brent " of Brentford, one might expect that a modern survival of the ancient tribe-name ''Brigantes" would exhibit a similar modification. We find in an early charter {Cart, Sax. i, p. 177) a " Brente " hill, but even if this could be proved to represent an earlier " Brigantia ", it is in Somerset, and therefore, not at all in the locality occupied by the Brigantes. Sir John Rhys thinks that the kingdom of the Bernicii mentioned by Bede represents an Anglicized form of Brigantes. Starting from a stem which has become in modern Welsh " bri " ( = renown), he follows the possible changes of the name until he reaches a form (Brenneich) which might well be represented by Anglo-Saxon lips as Boermcas, from which would naturally arise Bede's Bernicii. Dr. Whitley Stokes would connect the "Brig" in Brigantes with Irish Brl, accusative Brigh (Glossed tulach, i. e. hill), the modern Welsh bre (a hill), which appears in such place-names as '^Pen- bre", "Moel-fre", ''hill-top", and "bald hill" respectively. "Bre" and ''Bri" may possibly come from a common root meaning " high ", but it seems more natural to regard the Brigantes as thus denominated because of their dwelling in mountain fastnesses, than because they were renowned or privileged, and this for the following considerations. In attempting an explanation of the origin of tribe-names it is well to keep in mind two things : (i) that the forms of such names come to us generally through a neighbouring people, and not from the tribes themselves. Thus, the name " Welsh " ( Wealas), which means simply " foreigners ", comes to us not from the people so designated, but from the English, who were their neighbours. The Welsh call themselves " Cymry ". The Highlanders were so called bv neighbouring English- speaking people, and never use in their own speech this de- ROMAN OCCUPATION 27 signation of themselves. Gaul and Gauls were applied to the country and people of the land now called France, by Celtic tribes on the borders with whom *' Galli " would mean simply (like the English word "Welsh") foreigners; and the name German is also, according to Grimm, a Celtic designation, and is not used in their vernacular speech by any Teutonic people on the Continent in speaking of themselves.^ (ii) That tribe-names yield generally on analysis certain dis- tinctions which would naturally be used by one people speaking of another. Thus the Saxons were so called from the " seax *' or sword they used in warfare. The Franks were thus de- signated from the javelin they carried, which was called " franca " by the Anglo-Saxons, and not from a boast of their being a frank or "free" people. The Longobards (after whom Lombardy is called) took in the same manner their name from their long " bards " or spears, and not from their beards. In a similar manner the nature of the locality, or mode of dwelling, of a tribe furnished to their neighbours an obvious name for them. Thus the Burgundiones, who invaded in the early fifth century that part of Gaul which is now called Burgundy, were so called because they dwelt in a hilly country. Professor Fick associates, in this connexion, the " Burg " in their name with the ancient Irish Brtg, which is represented in modern Welsh by the Bre, already mentioned, and thus shows the equivalence of Burgundiones and Brigantes. The Brigantes mentioned by Tacitus derive their name naturally, therefore, from the hills, or rather hill-forts, wherein they dwelt. This is confirmed somewhat by Juvenal's reference to them (Sal. xiv), " Break down ... the Castles of the Brigantes." The ultimate connexion of Brig and Burg is further * The " Insi-Gall " applied to the Northmen of the islands of the West of Scotland means simply " Island foreigners ", and the same meaning is also perpetuated in Dugall (generally written Dugald), "the black stranger," or Dane, and in Fingal, ''the fair stranger," or Norseman. Gall chnu in Gaelic, like its cognate form in English (wal-nut), means " foreign nut". ?8 BRITISH PLACE-NAMES strengthened by the fact that Briga, an extended form of Brig, ""Brigs, was in use by the Celts of Spain from a time long before our era to designate what the Teutonic nations call Burg; for instance, Julio-briga, that is, the '*Burg of Julius", which {Not. Dig. 42) was previously called Brigantia. To return to " Bernicia " — the alleged later form of the tribe of the Brigantes — Bede speaks of the people as Bernicii, and assigns them a place in the present region of Northumberland. The boundaries of the Anglian kingdom thus designated are given in the twelflh-century Life of Oswald, and embrace the country between the Tyne and the Firth of Forth — Deira ^ another Anglican kingdom, occupying the region between the Tyne and the Humber. Richard, Prior of Hexham, writing about 1 154, makes the Tees the southern boundary of Ber- nicia, and the Tweed its northern Hmit. There is apparently no surviving place-name in all this region preserving the designation Bernicia. If such a form does exist, it is too much obscured to be easily recognized. I have gone carefully through the charters of the Monastic houses in these districts, but can find nothing at all representative of Bernicia. Such forms as Branx-holm or Brance-path, which might suggest a survival, are to be otherwise explained. The numerous Birrens Camps or Burrins Camps, on the Scotch border, are to be interpreted by a passage, p. 171, of Burton's Monasticon Eboracense in which he translates Barganes lapidum by " Sheep-fold[s] of Stone[s] '\ This word may be of Celtic origin, for Adamnan's Life of Columba, p. 113, gives Bairind as a '' great rock", and Burren is the common Irish designation of a stony district. (But see note, p. 272, and Dialed Diet, sub Burren) Having established order among the Brigantes, Ostorius, as Tacitus tells us, began a campaign against the Silures, who ^ It was from Deira, as we remember, the young Anglian slaves were taken who called forth Pope Gregory's pity in the Roman Forum. The name Deira is seemingly preserved in the Deorstrete, Dere-street, of Sim. Dunelm, 70, and of the Melrose Charters. It was Watling Street north of the Wall. ROMAN OCCUPATION 29 had offered a strenuous resistance to the Roman arms, and who were neither amenable to kindness nor severity. Tacitus describes the Silures as having swarthy complexions (using the same colour adjective as Virgil applies to the Indians) and curly hair. He is inclined to think, for this reason and because they lie over against Spain, that they came from the latter country and were an Iberic people. This tribe occupied the region to the west of the lower Severn and to the north of the Bristol Channel. Ptolemy (a. D. 120) places them to the east of the Demetae, whose territory extended from St. David's Head as far, at least, as Carmarthen (Maridunum), which he makes one of the strong- holds of the latter. Ptolemy gives Bullaeum as one of the towns of the Silures, and this place has been identified by some with the Burrium (cf. Din-birrion of the Book of Llan Dav) of the Antonine Itinerary, which, according to the distances there given, would be near Usk, but no distinctively Roman remains, nor a Roman road, have been found here, as far as I know. The Book of Llan Dav (printed from a MS. of the twelfth century) gives among the boundaries of the See of Llandaff, " Buell " on the Wye, now printed Builth, but in the sixteenth century Byellt. This is a place of some strategical importance, and may represent Ptolemy's BuUcBum, It is celebrated as the spot where the Prince Llewellyn, the last native ruler of Wales, met his death at the hands of the English in a.d. 1282. It would appear, therefore, that the region occupied by the Silures was roughly co-extensive with the ancient Diocese of Llandaff, and included Monmouth and Glamorgan, besides parts of Herefordshire and Brecknock. The Silures placed themselves under the command of Cara- tacus, who had won, among the Britons, a high reputation as a general. He succeeded, by drawing the Roman legions after him, in involving the latter in war with the Ordovices, who were a British tribe occupying the country to the north and west of the Silures. Notwithstanding this strategical move of Cara- 30 BRITISH PLACE-NAMES tacus, the combined forces under him were defeated, and his wife and daughter captured by the Romans. Caratacus having taken refuge with Cartismandua, the Queen of the Brigantes, was handed over by her, bound in fetters, to the victors, and was sent to Rome to grace the triumph of the conqueror. Ostorius had not, however, done with the Silures, who fell upon certain cohorts engaged in constructing camps in Silurian territory, and succeeded in killing many, including the prefect of the camp, and putting the rest to flight. Their attacks were such, indeed, that Ostorius, worn out by the arduous and anxious character of the campaign, at length succumbed, and died about the year a. d. 53. Ostorius was succeeded by Aulus Didius, of whom the chief exploit recorded is his saving Cartismandua, the Queen of the Brigantes, from the attacks of her husband, Venutius, a native, Tacitus tells us, of the State of the Jugantes^. Veranius, who succeeded Didius a. d. 57, died in the following year, after having ravaged the territory of the Silures. Suetonius Paulinus, who was sent by Nero to succeed Veranius, led the Roman legions to the Menai Straits, and, cross- ing to Mona (Anglesea) in flat-bottomed boats, reduced the island to subjection. Mona was the chief seat of Druidism, and Tacitus describes the Druids as pouring forth, from among the hosts drawn up on the shore to oppose the Romans, dire imprecations upon the latter. The Romans established a garrison on the 1 The MSS. read Jugantes or Evigantes, the latter reading appearing in the Editio Princeps. Camden, on the ground that no such people as the Jugantes were known and that the ''State of the Brigantes" appeared elsewhere in Tacitus {Agricola)^ substituted Brigantes here, and he has been followed by all subsequent editors. This is a dangerous method and tends to stop further research. We have several instances (Anscombe's Welsh Genealogies in Archivfur Cettische Philologie) oiju or Giu being read by transcribers as Evi^ but no transition from Bri to Evi, and we may therefore assume that Jugantes or Evigantes is the right reading. If Venutius had been, like Cartismandua, one of the Brigantes, there would have been no need to specify the fact. ROMAN OCCUPATION 31 island, and strove, by destroying the Sacred Groves, to put an end to the sanguinary rites performed within them. This brings us down to the year 60 of our era. Have we any memorials of these events, we may ask, in the existing topographical nomenclature of the country ? Solinus, writing about A. D. 80, tells us of an island Silura lying off the coast held by the Dumnonii (that is, the Devonians, meaning the coast of Devon and Cornwall). Sir John Rhys thinks that the name of this island is of the same origin as that of the Silures, whether Solinus meant by it the Sciily Isles, or the region on the north of the Bristol Channel. The quantity of the " u " in Silures is doubtful. In Holder's Sprachschaiz it is given as short (Siliires), but in Scheller's Lexicon it is Silures. If the vowel were short it would facilitate the contraction of the word, and Sir John Rhys's conjecture that the name of the chief man connected with the temple of Nodons discovered at Lydney Park (in the region of the Silures) — Silulanus — involves a term which may be equated with Silur^ seems very probable. The name appears also in \}i\t Antonine Itinerary in conjunction with "Venta", i.e. Venta Silurum^ now Caer Went in Mon- mouthshire, and in an inscription found here lately. The word ** Venta", like the term " Silures'*, is somewhat obscure. The note on the next page is an attempt to explain it. The island Silura, mentioned by Solinus, is probably the same as Susura, an island in the British seas, found in the list of the Ravenna Geographer. No explanation of the meaning of the term '* Silures" has yet been offered. It does not seem even to be a Celtic word, and the probability of its survival in any modern form is remote. Jornandes, who wrote about A. D. 552, read in the passage he quotes from Tacitus, " Sylorum " in place of Silurum, but this was a mistake. Perhaps he confused the name with that of the Silura referred to by Solinus, possibly now a little island off the coast of Glamorgan, called in Speed's map (a. d. 16 10) Sylye, now Sully. The "stormy strait" which, according to Solinus, 32 BRITISH PLACE-NAMES separated Silura from the coast of the Devonians, might mean the Bristol Channel and not the sea between Scilly {Sorlinga in the earliest maps given by Nordenskjold and still les Sorlingues of French navigators) and Land's End. Sully Castle and the parish of Sully on the adjoining mainland would seem to indi- cate that the name covered a region sufficiently large to form a designation for a tribe. The origin and meaning of this name are, however, not clear, and I know of no earlier form than that preserved in the Taxatio of Pope Nicholas (late thirteenth century) " Sulleye ". As to the Ordoviccs mentioned by Tacitus, Sir John Rhys considers them to have been a Brythonic and not a GoideHc tribe. He places them in what was later called Powys, a region which interposed between North and South Wales, both at this time inhabited by a Goidelic people. The name sur- vives, he thinks, in Rhyd Orddvvy (or the "ford of the Ordovi"), near Rhyl, and in Din-orwig (formerly Dinorddwig), in the neighbourhood of Carnarvon, meaning the fortress of the Ordovices. The district between the Dovey and Gwynedd (= North Wales) is called in the lolo MSS. Cantref Orddwyf, or " Hundred of the Ordovi ", and this, Sir John Rhys thinks, is a further survival of the term. NOTE B. The meaning of " Venta *' in the British place-names, Venta Silurum, Venta Belgarum, Venta Icenorum : — 1. It is evident from the forms of these names that they were imposed, and intended to be understood, by a Latin-speaking and not by a Celtic-speaking people. 2. Putting aside place-names in regions occupied by the Romans, the word " Venta " does not occur in Celtic literature, either in the form " Venta ", or in its equivalents *' Guent " or " Went ". There is, moreover, no Goidelic equivalent with any topographical meaning. 3. The word "Ventum" or "Venta", however, occurs in place-designations in Romanized regions on the Continent. We ROMAN OCCUPATION 33 have, for instance, ''Beneventum "^ in South Italy (now " Bene- vento "). In Northern Italy, '' Mutatio Beneventum" {Anton. Itin., 558, 14), between Brescia and Verona. In Spain we have " Et oppidum nobilissimum Beneventum " (L. Marin. Siculus, De Rebus Hi'spamae, Book III), now " Benaventa ", south of Astorga. There are also in Spain and Portugal several places of this name, e.g. '' Benavent," north of Lerida in Cata- lonia ; " Benaventa,'' on a tributary of the Tagus, not far from its mouth ; and "Benaventa", west of Oviedo in Galicia. The Abbey of Benevent (Creuse) is a French survival of the name and the word also occurs in Germany, e.g., *' Beneventenreut" (Oesterley, Geog. Worker buck), where rent (OHG. riuti) is the equivalent of our North Country royd = clearing. There are, besides, dozens of places called "Venta" in Spain, such as "Venta la Reina", '* Venta del Marquis", ''Venta Moral", *' Venta la Vadera ", '' Las Ventas ", &c. •' Venta " in Spanish means now an " inn '', or a place where food and drink are sold. The original meaning was probably " market '', the word being possibly derived from " Venum-eo '\ contracted to "veneo" = "I go for sale": its past participle in the neuter singular would be " venitum ", and the neuter plural " venita ". 4. Ducange confirms this import of the word in a passage where, sub voce "Venta", it is said that it means ** a place where goods are exposed for sale, or where tribute is received from things sold '\ and a quotation is given from the obituary notices of the Church of Langres (France), "John de St. Sequano gave to the Church of Langres sixty shillings (soldi) of Touraine in the ' Venta ' or ' Hall ' (' aula ' = here * market ', cf. * les Halles ', the Paris * Covent Garden ') of Montissalio." 5. The word "Beneventum" meant probably "a good market place ", for in the Ravenna Geographer (p. 280, 5) " forum novum ", or " new market ", is said to adjoin Bene- ^ In 268 B. c, when the Roman colony was there founded, 'Makofevr was changed to Beneventum. The name occurs on a bronze coin as an abl. sing. Benuentod (in Samnite, which was subject to Oscan influence, accord- ing to Conway, /talt'c Dialects ^ i. p. 171). Keller, Lat. Volkselym. p. 14, records Benuuent: as in Oscan the Ven in Venio becomes Ben, Ventum cannot well be from Venio. It was Malovent, moreover, before it was colonized by the Romans. Pliny, iii. 12. 107 gives Terventum, Tereventum, as a Samnite place-name now Trivento, and Livy cites a Carventana Arx (xxii. 15. 16) as a place -designation of the Latini. C 34 BRITISH PLACE-NAMES ventum, that is presumably the " old market *' ; and Pliny i^Hist. Nat.\\\. ii), speaking about the southern Bene ventum, says thus more auspiciously named now, but formerly called " Male ventum". In the Life of St. Cadoc {Camhro-British Saints, Rees, pp. 70 et seq.) already referred to, it is related that the saint was carried on a cloud from Llancarvan (i. e. the Church of St. Carfan, the later Welsh form of Corbagni — genitive) to " Beneventana Civi- tas", where he was made bishop; but as no bishop of Bene- ventum of that name is to be found in Ughello's Italia Sacra, the place must be sought elsewhere. It seems clear that Bene- ventum was not far from Llancarvan, for, in the same Life, it is said that St. Elli, Cadoc's successor at the latter place, *' was accustomed to go very often with his disciples to the City of Beneventum," a practicable thing if the locality were in Britain, but not so if it were in Italy. One is tempted to think of the "Bannaventa" (given in MSS. A and F as Bennavento) of the Antonine Itinerary which has been identified by some with "Daventry", but there were possibly more places than one of this name in Britain : cf. also the Glannibenta of the Not. Dig., given as Clanoventa in the Ant, Itin. The " Venta" of the Silures, however, may have been meant, or rather, that of the Belgae. From the foregoing it seems clear that the '' Venta " of the tribes named was the locality where they sold and bought what they needed or where they paid tribute. As far as Venta Silurum is concerned, it is not without significance that its near neighbour was called Chepstow (that is, " C^apstow " = market), in earlier times "Emricor va" {Lib. Land.), where there may possibly exist an echo of " Emporium " coupled with ma {va = place), the later form of mag-os. THE ROMAN OCCUPATION II Tacitus goes on to tell us of the great disasters which the Romans experienced in Britain in the following five years. The Iceni under their King Prasutagus and their Queen Boudicca had become friends of the Romans, thinking to purchase by ROMAN OCCUPATION 35 submission freedom from further attack. They found, however, that they were treated by the Roman centurions as a conquered people. The queen and her daughters were subjected to all manner of indignities, and the people were exposed to robbery and oppression on every hand. Owing to this state of things and the fear of worse the Iceni took up arms. They were joined by the Trinovantes and other tribes who had not yet had their spirit broken by servitude, and who were actuated by bitter hatred of the veterans recently planted as colonists at Camulodunum (Colchester). The latter had been driving the natives from their lands and treating them as captives or slaves. Paulinus Suetonius, who had succeeded to the command of the Roman troops in Britain, was away at this time on an expedition to some other part of the country. Camulodunum, which had not been fortified, was taken by the allied British tribes, and the temple therein (erected under the Emperor Claudius), to which the Romans retired as to their citadel, was carried by storm. Suetonius hastened to the rescue of the colony and was able to make his way to Londinium (London), which, although it had not been raised to the dignity of a colony, was then celebrated as the greatest mercantile centre in the island. Suetonius, finding that with his comparatively small body of troops he could not defend the place, left it to its fate — notwithstanding the prayers and tears of the inhabitants — judging it preferable to lose one city rather than endanger the whole region occupied by the Romans. Accompanied by the able-bodied men of the place he hastened to take up a position in which he could defend himself. The Municipium of Veru- lamium (close to St. Albans, where there are extensive Roman ruins, and where, according to tradition, St. Alban was martyred in A.D. 286), which had then a large number of inhabitants, was also left at the mercy of the enemy. Both London and Verula- mium were captured by the Britons, and, as Tacitus tells us, some 70,000 Romans and their allies were slain. The hopes of the Britons were naturally excited by their success, and they c 2 36, BRITISH PLACE-NAMES hastened to crush by superior numbers the Roman troops drawn up in an advantageous position under Suetonius, but numbering not more than 10,000 men. The generalship of the Roman leader and the strategical position he had chosen enabled the Romans to win a decisive victory, in which fell some 80,000 Britons. Boudicca, who with her daughters, had passed in her chariot from tribe to tribe to fire them with her wrongs, could not endure defeat, and put an end to her existence by poison. The generalship of Suetonius was not, however, fully appreciated at Rome, and he was ordered by the Emperor Nero to hand over his command to Petronius Turpilianus (a.d. 62). Of the tribe named Trinovantes we seem to have no survival in our topographical nomenclature. Their capital was, as we have seen, Camulodunum (Colchester), and the area occupied by them embraced all Essex and Hertfordshire. The name itself seems to involve the numeral Tri- (three), whatever " novantes '* (cf. Ptolemy's Novantoi) may mean. A similar use of this numeral appears in the names of ancient tribes of Gaul, e.g. Tri-cori (Pliny: to be compared with Petor-corii^ that is, " four hosts," now Perigord), Tricassi, now Troyes, Tri- boci, Tri-castini, &c. We have also the same numeral in the Gaulish Trinanto (which is glossed Tres-Valles= Three Valleys), and in the Nant tri neint (= Valley of the three Valleys), which appears in the Book of Llandaff of the twelfth century. " Nant tri neint" survives in " Turnant brook ". ** Londinium " still appears but slightly altered as " London " and continues, as in the days of Nero, to be the greatest mercantile centre in the island. There has been evidently a continuity of civil life here since Roman times, and this fact furnishes a comment on the theory that in Britain the Roman centres, which preserve relics of their ancient names to our own day, were not continuously occupied. Is there a modern representative of the name Verulamium ? 1 Korio — A.-Sax. Here, see Stokes {Wortschatz d. kelt. Sprach.), and later, p. 155. ROMAN OCCUPATION 37 The answer to this question depends upon several considerations. We cannot, for instance, predict from purely philological grounds the form a place-name will assume in passing into use among people speaking an alien language. The law of attrac- tion, by which the unfamiliar place-name is drawn into the orbit of the familiar, has produced some strange metamorphoses in our nomenclature beyond explanation by linguistic laws. It began to exercise its influence in Britain at the very beginning of the Teutonic invasion. Eburacutn \ or rather Eburac, for the um was Latin and evanescent, was not long in passing into Eoforwic^ which would mean in English "Boar- shelter or house ", and became finally ** York ". Regulbium could not long be preserved by Teutonic lips in this uncouth form, and soon found an intelligible equivalent in civitas Recuulf, also Racuulfe (Bede), and Raculf-cestre [also Raculfs Cestre], which finally became the modern form Reculver or " the Reculvers ", as we find it in the seventeenth century, from the two towers of its Church serving as landmarks to fishing vessels. RitupicE also was made to have some kind of meaning by taking on Bede's form Repta caester, which finally degenerated, through the Rales-burgh of Leland's Itinerary (vii. 137), into the present Rich borough. But these changes are nothing to the modifications which the ancient name of Rochester suffered. The form by which this place was recorded in the Roman Itinerary was Duro-brivis, which means " the Stronghold at the Bridges "? In the Peutinger Tafel it appears as Roribis, showing that by this time the ^ Dr. Henry Bradley would regard Eboraciim as representing a Celtic accusative Eburacon. * Duro as a prefix is most probably the same as durum — a suffix in such names as Augusto-durum, &c., and meant stronghold. Sir John Rh^s sees in it a word cognate with our " door" and with the " for " in the Latin Fores and Forum. In Dnrobrivis as in Durocasses the accent was probably on the When Agricola handed over the administration of Britain (a.d. 86) to his successor the country was, as Tacitus describes it, quiet and secure. It is a matter of conjecture who this successor was, but Suetonius, who compiled the lives of the twelve Caesars about a.d. 120, mentions, as Governor of Britain at this time, a certain Sallustius Lucullus, who was subsequently * " The Texel," an island off North Holland, of which we have early similar forms, may have been furnished to Ptolemy by navigators in the North Sea, and been placed by him in Eastern Scotland as ' * Taizaloi ", but see the discussion of Ptolemy's Geography further on, pp. 91 et seq. 48 BRITISH PLACE-NAMES put to death by Domitian. The Emperor Domitian died in A.D. 96 at the age of forty-five, and was succeeded by Nerva, who ruled only two years. On his death he was followed by his adopted son Trajan. Of aflfairs in Britain we know little or nothing until the accession of Hadrian, the successor of Trajan, in A.D. 117. Spartianus, who wrote, about the beginning of the fourth century, the Lives of the later emperors, tells us in his biography of Hadrian that the Britons at the outset of his reign could not be maintained under Roman authority. Julius Severus was at this time (a.d. 120) in command in Britain, but the Emperor Hadrian thought it wise to appear there (a.d. 123) in person. He set things in order there, and was the first, as Spartianus says, to construct a wall, of some eighty miles in length, which was to serve as a boundary between the bar- barians and the Romans, that is, those who accepted Roman rule. The length of this wall, corresponding as it does with that which runs from Wallsend on the Tyne to Bovvness in Cumberland, compels one to identify it with the latter. The inscriptions found in the neighbourhood of this wall — which consisted of a stone wall with large and small forts, together with a ditch — date, as Dr. Haverfield says, " mainly if not wholly from Hadrian's reign." South of this wall, and separated from it by an interval varying from 30 to 1,300 yards, is a vallum consisting of three ramparts and a ditch. Mommsen thinks this was thrown up in connexion with Hadrian's wall, but other authorities look upon it as an earlier structure. Hadrian died in a.d. 139, and was buried by his successor in that magnificent mausoleum which is now known as the Castle of St. Angelo at Rome. Antoninus Pius was his succes- sor. Julius Capitolinus, who wrote about the beginning of the fourth century, tells us that during the reign of Antoninus the Britons were reduced to order by Lollius Urbicus, who had been sent to Britain as Governor in a.d. 139. He constructed a second wall of sods between the Clyde and Forth, after he had driven away the barbarians. This wall was not apparently ROMAN OCCUPATION 49 intended to supersede Hadrian's wall, but to relieve pressure on the latter. Dr. Haverfield infers that it was soon abandoned. The inscriptions found along its course show that the legions employed in constructing it (the twentieth, the second, and the sixth) finished the work in the time of Antoninus Pius. Pausanias, a Greek writer of about the middle of the second century, inci- dentally mentions that the Emperor Antoninus deprived the Brigantes of much of their lands because they had begun to overrun the Genunian territory or patrimony (JMoird)^ of which the inhabitants were subject to the Romans. The construction of ihe second wall by Lollius Urbicus seems to have been connected with his efforts to punish the Brigantes, whom we find in previous records to have been included in the region now called Northumberland, and to have extended to the Forth. The territory taken from the Brigantes was, there- fore, in that region ; and the Genunian land, which was open to their attacks, may have been either north or south of the Tyne- Bowness wall. Sir John Rhys would make the Brigantes here mentioned the northern section extending to the Forth, and he would look for the Genunian possessions somewhere in that quarter. In the absence of any inscriptions near the Clyde-Forth wall later than Antonine, Dr. Haverfield's conjecture that it was quickly abandoned seems very probable, and, as the Romans appear to have been hard pressed in this region all the time, it is not likely that the Genunian territory, occupied by a people subject to Rome, should be found in this part of the country. Where then was this Genunian (or as some editors would read, Venunian or Venuvian) territory ? The late Dr. Reeves and Sir John Rhys were disposed to see a connexion between it and a certain Geona Cohort mentioned by Adamnan in his Life of Columba, Columba, the Irish apostle of the Picts of North Britain, died within not many days after the arrival of St. Augustine in the island of Thanet (a.d. 597). His biography was written by Adamnan about one hundred years D 50 BRITISH PLACE-NAMES later, and the earliest MS. containing his biography is of the first quarter of the eighth century, that is, it was extant before the death of the Venerable Bede (a.d. 735). Adamnan tells us that when Columba was in the Isle of Skye, an old man, a com- mander (Primanus) of the " Geona Cohort", landed on the island and was converted through an interpreter by Columba. His name is given as Artbrannan, a thoroughly Celtic appellation, and yet Columba required an interpreter to talk with him. It is therefore argued that he was a Pict and spoke Pictish, which presumably was a non-Celtic speech — but this would leave his Celtic name unexplained. Of the Geona cohort, Dr. Reeves wrote (Adamnan's Lt/e of Columba, p. 62) that it was " probably a Pictish corps deriving this name from the district to which it belonged '\ Sir John Rh;^s is inclined to think that Geona " is a defective spelling of Genona", and that it referred to the Genunian territory of Pausanias. This he would locate in the Pictish mainland opposite Skye, but it is hard to believe that Roman institutions could survive so late, as the words " Primarius " and " Cohort " imply, in this wild and rather inaccessible country, and it is still more difficult to believe that this could be the Genunian territory which was occupied by a tribe subject to Rome in a.d. 140, and exposed to attacks from the Brigantes at this time. The editorial conjecture of " Venunia " for *' Genunia " suggests the place called in the best MSS. of the Antonine Itinerary Vinonia, and now known as Binchester. This is sufficiently near the Brigantes, whether these were within or without the Tyne-Solway wall, to account for their attack on Vinunian or Genunian territory. The Greek word used by Pausanias, Moira, was employed, it is true, by Xenophon to denote a division of an army, and suggests thus the Latin Cohort, but it is far more frequently used for patrimony, share, or portion. We learn from Julius Capitolinus, before mentioned, that the Britons gave trouble again about a.d. 161, and that Calphur- ROMAN OCCUPATION 51 nius Agricola was sent against them by the Emperor Marcus Antoninus, better known as the Philosopher Marcus Aurelius (a.d. 1 61-9). In the year a.d. 181 also, as the same author tells us, the Northern tribes broke through the wall, but were worsted by Marcellus Ulpius, who had been sent against them by the Emperor Commodus (a.d. 176-92). But it was to the Emperor Septimius Severus (193-21 1) was due the actual pacification of the northern part of the country. He made a wall across the island, 32 (or 132) miles in length, says Eutropius* (writing in A.D. 370), and the same statement is made by Spartianus, who adds that it was fortified at each end where it met the sea. Aurelius Victor (a writer of a.d. 360 (?)) repeats the statement, which appears also in the Chronicon of St. Jerome (who died A.D. 420), in Orosius (born in Spain about a.d. 390), in Cassio- dorus (died a.d. 573), in the so-called Nennius (eighth century), and in Bede (died a.d. 735). This is a statement which it is difficult to reconcile with evidence from other sources. In the epitome of Dio Cassius, which we owe to Xiphilinus, who com- piled it about A.D. 1071, we find no reference to a wall con- structed by Severus, though mention is made of some earlier fortification. No inscriptions, moreover, connected with Severus have been found on either of the walls. Until some further discovery is made one may assume, with Mommsen, that Severus merely reconstructed the Clyde-Forth wall during his northern campaign. He doubtless found it a difficult matter to subjugate the northern tribes, and provided some sort of bulwark against them, but it is difficult to connect him with either wall. The tribes which gave him most trouble were, as we find in the * In the extension of Eutropius ascribed to Paulus Diaconus (died 782) and in other writers, Severus is said to have fully secured the regions obtained in this war by a great ditch and a strong Vallum with towers at long intervals, stretching 132 miles from sea to sea. The mention of the ditch, Vallum and Towers, appears first in a MS. of the tenth century. D2 52 BRITISH PLACE-NAMES epitome of Dio Cassius, the Caledonians and Maeatae, and he proceeded against them in person, almost reaching the northern limit of the island in his march. He was old and sickly, and had to be carried in a litter all the way. At length, on re- turning from that campaign, he succumbed to his malady, and died at York in the year a.d. 211. His body was cremated there, and the remains carried to Rome. He was succeeded by his son Antoninus, known better as Caracalla (a.d. 198-217). Archbishop Ussher was the first to see a connexion between the Maeatae of Dio and the Miathi (De Bello Miaihorum) of Adamnan's Life of Columha, The battle here referred to, according to Dr. Reeves (Adamnan, p. 35), is that called in the Annals of Tighernach Chirchind, which Dr. Reeves would identify with Kirkintulloch on the Clyde-Forth wall.^ Sir John Rh^s has no hesitation in placing the Maeatae immediately north of the wall (where Dio locates them), and finds in Dunmyat (alias Dalmyot and Demyat), a few miles north-east of Stirling, a stronghold of these people, preserving their name, as Dunkeld preserves that of the Caledonians, and Dumbarton (Dun-bretane) that of the Britons. He also connects the two rivers and the island called May with the Maeatae also, to which might be added Dunmay in Fife. From the death of the Emperor Septimius Severus in 211 to the year a.d. 287, that is for seventy-six years, there is prac- tically nothing known of affairs in Britain. Lampridius, a writer of about the end of the third century, tells us, it is true, that the Emperor Alexander Severus was killed (in a.d. 235) by his own troops in a town called Sicila (or Sicilia), in Britain, but adds that others assign the place to Gaul. Aurelius Victor, a later author, definitely asserts that Sicila was in Britain, but all other writers agree in describing the death of Alexander Severus as having taken place in the neighbourhood of Mainz on the Rhine (Bretzenheim, as some conjecture) whither he had been called ^ But compare Circinn and Magh-girginn referred to later on, p. 77. ROMAN OCCUPATION 53 from the East to repel the Germanic tribes which were then threatening the Empire. Severus was succeeded by Maximinus, who was declared an outlaw by the Senate in 238, and was followed in rapid suc- cession within that year by four Emperors, whose names hardly require to be mentioned. Gordianus III ruled from 238 to 244 and had as his successor Philippus (244-9), Decius (249-51), Trebonianus Gallus (251-3), Valerianus (253-60), Gallienus (253-68). Postumus, who ruled, wdth Victorinus, in Gaul only, was Emperor from 258 to 267. Claudius II and Tetricus ruled from 268 to 270, and were followed by Aurelian (270-5), Tacitus (275-6), and Probus (276-82). Diocletian and Maxi- mianus were Emperors from 284 to 305, and it was during their rule (a. d. 287) that Britain again came within the horizon of history. This period of seventy-six years, although almost a blank ' as far as written history is concerned, is not without other indica- tions of the persistence of Roman rule and activity in the island. Contemporary inscriptions on stone monuments discovered at various places show that the jurisdiction of Rome was main- tained in the region south of the Tyne-Bowness Wall. There have been found, for instance, inscriptions of the time of Elagabalus (a.d. 218) at Riechester, Northumberland, and of the time of the Emperor Alexander Severus at Old Penrith, as well as on the Roman wall. Of the time of the Emperor Gordian III inscriptions have been discovered at Lanchester (Durham), at Bittern near Southampton and at Old Carlisle. * The historian Zosimus, writing in Greek about the beginning of the fifth century, tells us that the Emperor Probus sent into Britain (a.d. 280?) a number of Burgundians and Vandals to sustain Roman authority there. This is significant as showing that Teutonic tribes were in Britain before the time of Zosimus, and throw-slight, possibly, on Geoffrey of Monmouth's "Africans". 54 BRITISH PLACE-NAMES Inscriptions of the reign of Philip have been found at Old Penrith, and others of the time of the Emperor Gallus and Volu- sianus at Bowness (Cumberland) and at Bittern, Southampton. Of the reigns of Valerianus and Gallienus there has been found an inscription at Caerleon (Monmouthshire), and other inscrip- tions have been discovered at Pyle, near Neath, Glamorgan, at Bittern, at Castor (Northampton), and at Kenchester (Hereford), representing respectively the reigns of Victorinus, Tetricus, Florianus (276), and Numerianus (284). There is possibly further indirect testimony to the Roman occupation of Britain at this time in the Ancient Welsh Genea- logies, if we could disentangle from them the historical elements which they seem to contain. The Bards, who are referred to by classical writers from 200 b.c. onwards, are uniformly represented as the annalists and genealogists of the Celtic chiefs. Giraldus Cambrensis, writing in the twelfth century, says (Descr. Ca?nbriae^ cap. iii) " I ought to record that the Welsh Bards or reciters have in certain ancient and authentic books of theirs the genealogies of the aforesaid Princes, but written in Cymric, and the same they know by heart*'. The genealogies of Cymric (Welsh) notables which have come down to us, contain besides native names, many of Latin origin. See Note C. NOTE C. The early British Pedigrees which have come down to us have been probably preserved by the Bards. These pedigrees are not consistent with each other, and there are indications that they were edited from time to time in the interest of certain families. They contain names, however, which go back apparently to Roman times and could not well have been introduced at a later date. Some of these names seem to have come into use at the period with which I am dealing and to have persisted as appellations for a long time after. Hence a difficulty arises of identifying the personages who bore them. These pedigrees are found, with one exception, in MSS. from the eleventh century onwards (see Mr. E. Phillimore's transcript of the earliest — Harleian MS. 3859 — in the Cymmrodorion Mag. vol. ix, ROMAN OCCUPATION 55 for 1888, pp. 141 et seq.). The exception, which is inscribed on stone, dates from not later than the middle of the ninth century, and may, therefore, be dealt with here first, as it seems to refer to matters of the end of the fourth century. This inscription in memory of an ancestor called Eliseg (Elized, according to Mr. Phillimore) was carved at the instance of a certain chief called Conchenn ^ on a stone cross which gave the name to, and still stands at, Valle Crucis near Llangollen. The monument, according to experts, belongs to a period some four hundred years after the Roman troops had left Britain, but there are indications on it which would lead one to assign some of the sources of it to the end of the fourth century. Only a portion of the inscription was made out in the seventeenth century by Mr. E. Lhuyd, whose MS. was copied by Westwood and Hubner. Professor Sayce i^Arch. Cambrensis, Jan. 1909) has advanced a most ingenious restoration. At the end there is a pedigree containing three Roman names, Pascen, Maximus, and Sev[i]ra (?), and of these Maximus has been identified with Maximinus who was a companion in arms with Theodosius in Britain, and in a.d. 367, according to Bede [epiL\ was made Emperor. Sev[i]ra is given as his daughter. Bede also tells us that Maximus went into Gaul and slew there the Emperor Gratian, who was then (389) at Lyons.' The end of the * This is the same name as Conchend (Doghead) occurring in the Irish annals. See the F^lire of Oengus (published in a revised edition by the Henry Bradshaw Society) sub voce " Conchenn ". The name is undoubtedly Irish in this form, but it took on afterwards a Cymric shape, Cyngen, which disguises its origin but is otherwise instructive, as it suggests that the Celtic prefix cuno-, Welsh Cyn, involves "hound". Conchenn is repre- sented on the stone [Haddan and Stubbs, Councils, i. 626] as King of Powys, and the Attnals of Cambria at a.d. 814 calls his father Catell, who died in 808. Sir John Rh^s has shown {Epigrapkic Notes in Archxo- logia Camb., Jan., 1907) that in the ancient Celtic personal names involving cuno, such as Cuno-maglus, Magli-cunas, Cunotamos, Cunovalos, &c., the element Cun-o means " hound ", and has nothing to do with a supposed Czvn = high. ^ Maximianus (or Maximinus) was of British origin according to Socrates the ecclesiastical historian, and in Sigebert's Chronicon he is made to be of imperial descent, and a kinsman of Constantine the Great. All this falls in with the traditions (taken from Geoffrey of Monmouth) preserved by Matthew of Westminster {Chron. Afajora, under A. D. 376). 56 BRITISH PLACE-NAMES inscription refers evidently to this fact, although Gratian's name does not appear. '' Pascen(t) " and '' Eliseg " are names that occur several times in the Welsh pedigrees, and Nennius (Hisi. Brit., 70) applies the Latin name Pascent to the son of Vortigern, the British king at the time of the Teutonic invasion, and the Irish version of Nennius makes the latter the son of Guital (? Vitalii). The transcript of the Genealogies edited by Mr. E. Phillimore was originally compiled, according to the late Henry Bradshaw, by an Anglo-Norman scribe, in the eleventh century, from older documents of varying age. It furnishes a considerable number of Roman, and possibly historic, names associated with those of the families of British notables, among the latter being Eliseg (Elized). The Roman names are somewhat disguised, but their classical forms can be readily restored. In "Aircol son of Trephun", for instance, the former represents, according to Sir John Rh^^s, Agricola, and the latter is apparently the British equivalent of Tribonianus, a name appearing in 235 as a designation of the Emperor Gallus. Tribon- becomes quite normally Trifun- in Welsh : cf Tafarn and Taherna. In the sequence of " ^tern, the son of Patern pesrut (that is, of the red tunic — Pets from Low-Latin Pexa — i.e. Chlamys purpurea) son of Tacit" we have possibly a reminiscence of the Emperor Tacitus (275-6) and of Ovinius Paternus, a consul of his time. Anthun, which occurs more than once in connexion with Maxim Guleiec (that is, Emperor), represents perhaps M. Antonius Gordianus, Emperor a. d. 238. The Maxim referred to in the pedigrees, as well as on the pillar, is given as he '^who slew Gratian King of the Romans ", i. e. the Maximus who, elected Emperor in Britain, was killed afterwards by Theodosius (a. d. 388), but possibly Caius Julius Verus Maximinus is meant, who was Emperor from a. d. 235 to 238. In another sequence, " Dimet son of Maxim Guletic son of Protec," we have probably the latter Maximus associated with the Emperor Domitius Aurelianus (270). The short of Domitius would give normally a Welsh o, but we have instances where, like long 0, it becomes ii = 1 (Pedersen, Gramm. i, p. 195). louanaul (cf. Juvenal, Haddan and Stubbs, Councils, \, p. 166), Grat\ and Urban '^ also occur. ^ Gratus was Emperor in a.d. 222. Cf. also Annins Grains, Emperor in A.D. 251. ' Urbanus was the colleague of Maximus in 236 A. D. ROMAN OCCUPATION 57 Another name, Serguil, occurring in the pedigrees, is the British form of Serv'ilius, which is the name of a consul under the Emperor Trajan. " Pappo Post Priten " is the mysterious designation of an important personage, " son of Ceneu," occurring in the genealogies, and is perhaps a reminiscence of Pappus, a consul under Gordianus, who may have been a Praepositus (Post?) in Britain at the time. Praepositus, which was applied to an Admiral of the Fleet, as well as to other officers, gives the modern Welsh Prawbst. The name of the Emperor Postumus might be compared here, but he ruled only in Gaul. Several other Latin names of civil and military functionaries, many of them not such as would occur to a later compiler, appear in these genealogies, and thus reflect the days of the Roman occupation. For instance, we have more than once '^Protector " (and the shortened form Protec) meaning one of the imperial life guards, and applied in this sense by Spartianus in his Life 0/ Car ac alia to a certain Marcellus. "Stater^" is another official designation, meaning one who attends on the Provincial magistrate. There occurs also a mysterious designa- tion " Pincr misser ", a term which seems to have baffled all commentators ; but may it not be for Pincerna mensarius^ a buder? Lampridius uses, in his Life of Alexander Severus, " Pincerna " in this sense, and the term Pincerna was applied also to a subordinate officer of the Prankish Court : Brunner, Rechtsgeschichte^ ii. 102 ; cf. Stevenson's Asser, p. 164 n. ; and Duddan Pincerni {sic, for genitive) in a Worcester Charter ascribed to 779, Carl. Sax. i, p. 325. *' Peretur " (later, according to Sir John Rh^s, Pryderi) is another name occurring in the genealogies as one of the sons of Eleuther [Eleutherius ?] of ** the great retinue ", [i.e. cascord = gosgordd], and, as far as I know, has not been explained. May it not be an attempted transliteration of Praetor, all the more readily retained because of its seeming resemblance to the British pryderus = "solicitous", "caring about"? These are only a few of the probable Roman names in these genealogies, which are supposed to preserve some survivals of the oldest materials of British history. List xvi is instructive in this respect. It gives the chronological sequence of the Roman Emperors down to 337, with some omissions, and makes each ' The form Stator Practorius or Praetorian Stator occurs. S8 BRITISH PLACE-NAMES emperor the son of the preceding one. The forms given to the classical names are instructive. I place the list in one column and the actual succession in another. List " Harteian MS." 3859. Map ( = Son of) Constantis Map Constantine the Great Map Constantine Map Galerii Map Diocletiani ** In his time suffered the blessed martyrs in Brit- tannia, Alban, Julian, Aron.with many others " Map Caroci (glossed in later hand " i imp^ro^ms ") Map Probi Map Titti Map Auriliani Map Antun dv. and Cleopatre. Map Valeriani Map Galli Map Vitzxtis. mus Map Philippus {sic) Map Gordianwi- Map Alaximwj ^ Map Alaxander Map AurelianMJ- Map Mapmau cannz^j ( — Mau- cannus) Map Antoniw^ Map Severus Map Moebus (? for Marcus Anto- ninus) Map Cowmodius Map Antonius Map Adiuuandwj (=Adrianus?) Map Troianw^ Map Nero : under whom suffered the blessed Apostles of our Lord Jesus Christ, Peter and Paul. Map Domitianus Actual Succession. Constantine, 306-337 Constantius and Gallerius, 294-306. Diocletian, 286. Carus, 382, and his sons Carinus and Numerianus, 283. Probus, 276. Tacitus, 275; Florianus, 276. Aurelianus, 270. Aurelius, Quintillus, and Aurelianus, 268, 270. Galienus, 253; Postumus, 258. ^milianus and Valerianus, 253. Trebonianus, Gallus, and Volusianus, 251. Decius, 249. Philippus, 244. Gordianus, 238. Maximinus, 235. Alexander Severus, 232. Elagabalus, 218. Macrinus, 217. Antoninus (Caracalla), 198. Pertinax, Julianus, and Severus, 193. Commodus, 176. Marcus Antoninus and Lucius Verus, 161. Antoninus Pius, 138. Hadrianus, 117. Traianus, 99. Nerva, 98. Domitianus, 81. 1 Mr. Phillimore rightly read A/ as a. mistake for M. ROMAN OCCUPATION 59 List " Harleian MS." 3859. Actual Succession. Map Titus Titus, 79. Map Vespassianwj Vespasianus, 69. Galba, Otho, and Vitellius, 68. Nero, 54. Map Claudius Claudius, 41. Caligula, 37. i^/ Tiberias: under whom suffered Tiberius, 14. our Lord Jesus Christ. Octavianwj- Augusti Cessarus. Octavianus Augustus, 23 B. C. In his time our Lord Jesus Christ was born. The foregoing list is instructive in showing how these genealogies were made up, as well as in its transcription of names, even when these are manifest blunders. The misplacing of Antony and Cleopatra and of Nero is curious. The scribe had apparently Nerva before him and, making it Nero, added the Latin account of the martyrdom of SS. Peter and Paul. The omission of all notice of Carausius and Allectus tends to suggest that all the information about the emperors was acquired from continental information derived through foreign ecclesiastics or books. It has not certainly the air of home tradition. The list will serve to explain also why imperial names were foisted into the previous pedigrees. It is not, therefore, too rash to connect, as I have done, the names of known Latin rulers with the indigenous appellations appearing in the genealogies, seeing that commentators assign several of these Latin forms to British saints who lived at a much later date, and of whom inscriptions still survive. For instance (see Haddan and Stubbs, Councils^ vol. i, pp. 162 et seq.), an inscribed stone bearing the name ** Pascent " exists at Tywyn in Merionetlishire and is not earlier than the sixth century. The conjecture about Pappo Post Priten may also seem rather ridiculous in face of the fact that at Llanbabo (i. e. the church of Pabo) in Anglesea, there exists a tombstone with this inscription, Hie iacet Pabo Post Prud, &c., that is, ' Here lies Pabo Post Prud." Unfortunately, however, this tombstone, while it evidently preserves, in the figure upon it of a sceptred king, older traditions, belongs to the thirteenth or fourteenth century only, and is probably an instance of how pious ecclesiastics made use of early names to found upon them claims for the antiquity of sacred sites. The Lives of the British saints contain many such associations, but they were all written centuries after the saints lived. It ought to be 6o BRITISH PLACE-NAMES here noted that certain ancient place-names are found in the Genealogies, e. g. in List v dimor meton (i. e. " from mor melon " = ? the Middle Sea, Mediterranean. In xxv Glastenic (Glaston- bury) and Loyt Coyt (Lichfield). In xxxii manau Guodoton^ i. e. Manau of the Votadinoi (see p. 67). THE ROMAN OCCUPATION III With the year 287 Britain not only comes again into the light of direct history but also appears then, and for some eight or nine years afterwards, as independent of Roman authority. This is the alleged year of the martyrdom of St. Alban at Verulamium, and the baptism of the land by Christian blood was a prelude to British independence. It was, however, but an independence in name. The new ruler, Carausius, assumed the Roman purple, and took up the position of an emperor. The story of Carausius is invested with all the elements of romance. Eutropius the historian, who died about a.d. 370, tells us (ix. 21), that Carausius sprang from the vilest of the people, and we learn from Victor (a.d. 360) that he was a citizen of Menapia, that is, roughly, of the region extending between the river Meuse and the sea. The Menapii, who gave their name to this district, extended in the time of Julius Caesar even to the Rhine, and gave him no little trouble in his GaUic wars, but by a.d. 287 they had become Romanized. The position of Menapia, with its numerous islands in the estuaries of the Rhine and Meuse, was favourable to the development of a seafaring population, and the Romans would naturally avail themselves in manning their fleet of this aptitude of the people. Carausius was a seaman, having exer- cised, as Victor tells us, the calling of a pilot from his youth upwards. For this reason, and because he had already obtained celebrity for his prowess in war, he was placed in command of the Roman fleet, which had been got together at Boulogne ROMAN OCCUPATION 6i (Bononia), to drive from the seas the Germans who were then infesting them. Eutropius is more specific and gives us the nationalities of those German raiders, and defines the coast also upon which they were making iheir descents. They were Franks and Saxons, and it was to act against them that Carausius had his commission given him (Eutropius iv. 21), to maintain Roman authority on the seas which washed the coasts of Belgium and Brittany (Armorica).^ These Teutonic pirates infested not only the Belgic and Armorican coasts but also those of south-eastern Britain, and a special official, with the designation of " Count of the Saxon Shore ",* ^lad to be appointed to keep them in check. The Saxons were, therefore, pirates in the Channel long before they made a permanent settlement in Britain, and this fact, coupled with the notice by Ammianus Marcellinus of the com- bined attack of Saxons and Scots upon the Britons, tends to explain why the Celtic peoples of Ireland, Scotland, and Wales have no other name for the Teutonic invaders of Britain than Saxons (Sassenach, Saisnec). Carausius, making good use of his old skill as a pilot, succeeded in following up the Teutonic pirates, and in wresting from them no inconsiderable booty. He was accused of having appro- priated this booty to himself instead of sending it into the provincial or imperial treasury, and as a suspicion then arose that he was enriching himself by compounding with the enemy, * The loss in the Celtic language at an early date of the original initial P disguises the cognate forms of certain native words. Armorica or Aremorica is an adjectival form from " Are" = (p)arei " upon" and " Mori " = the sea. " Are" is thus cognate with the Greek Para = '-^ beside" and (P)areimorica is thus the equivalent of the Greek Parathalassios or Para-alos. Po-mer-ania, on the Baltic, is a similar form in Slavonic (but with the preposition Po, of a similar meaning to Para) = sea coast. ^ This term appears first in the Notitia Digftitatum, drawn up about A. D. 420, but the Comes Maritimae Tractus occurs in Ammianus Marcellinus (writing about a. d. 390) who tells of Saxons, together with Scots and Attacotts, attacking the Britons in A. D. 368. 62 BRITISH PLACE-NAMES an order was given by Maximianus for his destruction. On this, Carausius boldly assumed the purple and took possession of Britain. Here he maintained his rule for some seven years, and forced the Roman emperors finally to come to terms with him and to acknowledge his authority. Numbers of coins of his reign have been found on which he designates himself as emperor, but none in which he limits his authority to Britain. One of the later coins has the inscription " Carausius and his brothers", that is, as has been surmised, the emperors Dio- cletian and Maximianus. A few — some of them found at Rouen — have a Roman Galley on the reverse. Eutropius goes on to tell us that Carausius was at length (293) slain by his companion Allectus, who, after a reign of three years, was overthrown by the praetorian prefect, Asclepiodotus, and Britain was restored shortly afterwards, by the capture of London, to the Roman dominion, after an independence of ten years. Some further details of the career and fate of Carausius, and of his successor Allectus, are to be found in a panegyric, pro- nounced upon the Emperor Constandus Chlorus by Eumenius, who was born, of Greek origin, at Autun (Augustodunum) in Gaul, and who became a kind of private secretary to Con- stantius. Constantius had been sent by Aurelian into Britain in A.D. 271, where he is said to have married his first wife Helena, whom legend makes to have been the daughter of a British chief, Coel, and the mother of Constantine the Great.^ When Constantius was associated with Diocletian and Maximianus in the Empire in a. d. 291, Maximianus bound him still closer to his interests by giving him his step-daughter, Theodora, in ^ Sozomen (writing about 400) and Theodoret (about 450) describe Helena's finding of the Cross at Jerusalem, but they say nothing about her parentage. It is not until the ninth century {Life of Si. Helena by Altmann, who died 882) that we have any information on this point: some early records of her birth make her a native of Treves, and others of a place in the present Bosnia, &c., but Geoffrey of Monmouth and those who followed him make her a Briton and the daughter of ''King Coel". See Diet, of Christ. Biography^ sub Helena. ROMAN OCCUPATION 63 marriage, Helena ' having been repudiated. Constanlius, having been sent by Maximianus into Britain to overthrow Carausius, moved \vith such rapidity that his arrival in Gaul anticipated any announcement of his coming. Carausius had strongly fortified Boulogne (called Gesoriacum '^ as well as Bononia), to which Constantius laid siege, and, to prevent the fleet of Carausius coming to its succour, blocked up the entrance to the harbour with large rocks and beams. Eumenius tells us that his prosecution of the campaign (a. d. 292) was delayed by the necessity of building ships for his transit into Britain. Although his panegyrist makes the best of it, ii is clear that Constantius was unable for a time to make any headway. The Roman fleet appropriated to protect the coasts of Gaul had followed the fortunes of Carausius. Increased by new vessels built after ' She is called concubine by Orosius and Sozonien. " The oldest MS. (eighth century) of the Itinerariiim Anton, gives in two places a various reading of this word, which seems to be the right one, viz. Gesorigensi and Gessorigiaco. Celtic place-names ending in -aiumy as D'Arbois de Jubainville has shown, contain in the majority of instances the personal name of the owner of the locality, but it is often attached to common nouns, giving them the significance of *' belonging to" as in Tavemiaaim^ now Taverny. Gaisorix is such a name and actually occurs, according to the reading of Zeuss, in Orosius, v. 16, 20. Gaiso-n meant spear in Celtic and was borrowed by the Latins in the form Gaesum, which is used by Vergil and also in the Vulgate (Joshua viii. 18). The Greeks also had the word, and cognate forms are found in Teutonic personal names^ e. g. Hariogaiso-Sy Laniogaiso (a Frank and a Roman tribune, A. D. 354). The Teutons obtained their first knowledge of iron from the Celtic tribes, and borrowed the name for spear from them also, that is, at a time when intervocalic s was still retained in Celtic. How early this was is shown by the employment of the stem gaiso in Teutonic proper names, Schrader, Aryans, p. 235. The Early English equivalent is Gar, which also appears in personal names, e.g. Wulfgar and Garwulf. The form Coisis of the Todi bilingual stone is read by Sir John Rhys {Proc. of British Academy, vol. ii.) as Goisi(o), which he is inclined to connect with the Gaulish Gaiso-n, Gaiso-s ; Irish, 6^ = plain, place. Duroco-brivis seems to contain an adjective formed from Duro (stronghold) and bHvis^ an oblique case of Briv = bridge. The site may have been on the low ground below Dtmstable, as no bridge could well he on the height there. Sir John Rhys draws attention to Dvorico, in a Gaulish Inscription, meaning some kind of portico {Celtic Bj-itain^, p. 301). For Verolamio see p. 39, Sulloniacis seems to be the oblique case of a place called after its pro- prietor, Sullon-i. It is somewhat striking that Brockley Hill is called in Cart. Sax. iii. 605, "The old Tnnsteall,^' i.e. "Town site". Names ending in -acus^ although made thus simply adjectival, seem to be generally used for patronymics corresponding to English names in 'ing, -itigas. The meaning of " Londin-ium " cannot be satisfactorily explained with our present knowledge. Novio-mago means '* new-place ". Vagniacis (variant Vagnacis) is probably an oblique case of a patronymic from a personal name, *Vagn-o, cf. Vaco in Vacomagi. Durobrivis, see p. 37. Duro-levo is obscure. For Duroverno see p. 38. For Ritupis see p. 37. ' Dubris is in the locative plural, meaning " at the Waters", i.e. the two io8 BRITISH PLACE-NAMES Route IV. From Londinio ad P or turn Lemanis (Lympne), 68 miles. This is the same as the preceding Route as far as Dubris ; whence to Portum Lemanis is 1 6 miles.^ Route V. From Londinio to Luguvalio — to the Wall, 443 miles. The first station is Caesaromago (? near Widford, south-west of Chelmsford), 28 miles; after which follow Colonia (called Camoloduno in Route IX = Colchester), 24 miles; Villa Faustini (? Woolpit), 35 miles; Icinos (Pickworth near Bury St.Edmunds), 18 miles : Camborico (unknown), 35 miles ; Duroliponte (? God- manchester on the Ouse), 25 miles; Durobrivas (? Castor on the Nene), 35 miles; Causennis (PAncaster), 30 miles; Lindo (Lincoln), 26 miles; Segeloci (Ageloco in Route VIII = Little- borough), 14 miles; Dano (Doncaster), 21 miles; Legeolio (Lagecioin Route VIII=:Castleford), 16 miles ; Eburaco (York), 21 miles ; Isubrigantum (=:Isurium of Route II, = Aldborough), 17 miles; Cataractone (Catterick), 24 miles; Levatris (La- vatris Route II, Lavatres Not. Dig. near Bowes), 18 miles (16 in Route II); Verteris (Brough), 14 miles; Brocavo (near Brougham Castle), 20 miles; Luguvalio (Carlisle), 22 miles.^ small streams which found their exit here. Dubron is the earliest Celtic form for Water. The French Douvres still preserves the plural. ^ Lemanis has been dealt with else#here. "^ The identification of the places on this route after Colchester is a be- wildering work. Assuming the Villa of Faustmus to have been at a place where the Peddar's Way forked, this might be Woolpit, in Domesday ^F^^/a = Wolfpit. Although no trace of a Roman road is found between Woolpit and Bury St. Edmunds or Lackford, it is probable, from the dis- covery of Roman remains at Ickworth (? involving Icinos, cf. the Icenan river of the Cod. Dip. iii. 316, which gave the name to Itchington), that a route lay in this direction. The road is plain onwards toward New- market and Great Chesterford. The latter place may be Camborico which involves Camb = a bend. Prof. Skeat has shown that Cambridge arises from Grantabrycge, and hence Camborico cannot be identified with it. The next station, Duroliponte, is 'given as twenty-five miles from Cam- borico. Sir John Rhys {Celtic Britain^, p. 300) thinks that -lipons may be a Celtic termination and have nothing to do with Latin Pons^ but if Godmanchester on the Ouse represents the site, the idea of a bridge is natural. Duroli might thus represent the genitive of the early name of the Ouse. Durobrivas seems to be Castor, where Mr. Codrington {Roman ROMAN OCCUPATION 109 Route VI. From Londinio to I.indo, 156 miles. The first stage given is Verolami (St. Albans), 21 miles; then follow Durocobrivis (Dunstable), 1 2 miles ; Magiovinio (Magiovinto in Route 11 = Fenny Stratford), 12 miles ; Lactodoro (Towcester), 16 miles; Isannavantia (the oldest MS. reads Isannantia. This is the Bannaventa of Route II, and the Bannavento of Route VIII = ? near Daventry), 1 2 miles ; Tri- pontio (? where Watling Street crosses the Avon), 1 2 miles ; Venonio (High Cross), 8 miles; Ratas (Leicester), 12 miles; Verometo (Vernemeto, evidently the correct form, in Route VIII = Willoughby), 13 miles; Margiduno (Castle Hill), 12 miles ; Ad Pontem (? E. Stoke-on-Trent), 7 miles ; Crococalana (near where the Devon joins the Trent, PBrough), 7 miles; Lindo, 12 miles.^ Roads ^ J p. 140) says the Nene here "was crossed by a timber bridge on stone piers, the remains of which were removed when the river was made navigable". The briv-as is thus explained and the extensive Roman remains (including a mile-stone with M. P. L (or LI), i. e. 50 or 51 miles, possibly the distance from Lincoln) show the importance of the place. Causennis is at the correct distance for An-Caster, where the an- may still preserve a relic of the former name. Lindum has been explained. Segeloci or Ageloci probably involves the antecedent of the Welsh Uwch (= lake), which Dr. W. Stokes would trace to a stem (P)loug ( « to fly) : the Penne-locos ( = " Lakehead ") on the Lake of Geneva involves the same word (Thumeysen, Handbuch 47). Dano involves probably the old name of the river Don. Legeolio is obscure. The remaining names on this route have been already explained, except Lavatres (which may be allied with the Ancient Irish Ldthur^ glossed Canalis (Pedersen, V. Gramm. § 42), and Breton Laouer, trough. Cf. the river Lowther) and Brocavo, where we have probably the same termination as in Ande-gav-ia (Angers), the Ande- here representing an intensive prefix or the early form of the definite article (Thumeysen, Handbuch des alt. Irischen, p. 471). ' Magio-vinto seems to involve Mag-o — plain, place. Lacto-doro: the variants give -duro, hence this is the stronghold possibly on a river *Lact-o, now the Tove, which gave its name to Tovecester, Towcester. Isannantia of the oldest MS. can hardly be a misreading of Banna- venta (= ? "chief market"), but seems to be a synonym composed oi Isa (= Low, at foot of) and nant-ia, possibly "valley". Tripont-io = "Three Bridges", cf. Tri muntium of Ravennas and the numerous other compounds with Tre = three. Venonio is obscure. Ratas seems to be related to the Modern Irish Rath^ an earthwork, found also in Pembrokeshire. Ver-nemeton is the correct reading doubtless, and means "Chief Temple" as it stands: see no BRITISH PLACE-NAMES Route VII. From Regno (Chichester) to Londinio, 96 miles. The first station is Clausentum (? Bittern), 20 miles; then follow Venta Belgarum (Winchester), 10 miles; Calleva Atre- batum (Silchester), 22 miles; Pontibus (Staines), 22 miles; Londinio, 22 miles.' Route VIII. From Eburaco to Londinium, 227 miles. The first station is Legecio (Legeolio, Route V), 2 1 miles ; Dano, 16 miles; Ageloco (Segeloci, Route V), 21 miles; Lindo, 14 miles; Crococalana, 14 miles; Margiduno, 14 miles; Vernemeto, 12 miles (13 in Route VI); Ratis, 12 miles (13 in Route VI); Venonio, 12 miles; Bannavento, 18 miles (20 in Route VI, via Tripontium); Magiovinio, 28 miles; Durocobrivis, 12 miles; Verolamo, 12 miles; Londinio, 21 miles." Route IX. From Venta Icinorum (Caistor near Norwich) to Londinio, 128 miles. First station, Sitomago (? where Roman road from Caistor reaches the river now called Gipping), 32 miles; followed by Combretonio, 22 miles; Ad Ansam (.? Stratford), 15 miles; Camoloduno (Colonia in Route V), 6 miles ; Canonio (.? Kelve- don), 9 miles ; Caesaromago, 12 miles ; Durolito {? at Writola- burna, original Charter of 692-3 in B. M., now Writtle), 16 miles ; Londinio, 15 miles.^ p. 1 30 on Netnet. The Margi in Margi-duno is obscure. Croc-o-calana seems to involve Cruc = tumulus. ^ Regno : cf. Regina, now Regensburg ; it possibly involves Rig = King. The famous Chichester inscription gives the name of Cogidubn-i {Phil. Trans., 31 Oct., 1723), and this was doubtless his residence as king of the "royal" people of the district. Clausentum is obscure. Venta Belgarum and Calleva have been explained elsewhere. Pontibus means at the Bridges, the ruins of which probably gave origin to the Saxon Staines = Stones. 2 The names in this route have been already dealt with. 3 Venta-Icenorum. Sir John Rhys thinks Eceni was probably the correct name of the Iceni who occupied the Norfolk and Suffolk district. He equates the name with the Welsh egin = bladed, which possibly comes from the root ak ~ sharp. The Welsh word is cognate thus with A. -Saxon Rginu = English Awns, as in the awns of barley. Sir John thinks these people, like the Saxons, were called from the weapons they bore (see p. 27), probably swords. ROMAN OCCUPATION iii Route X. From Clanoventa (Glannibanta Not. Dig) to Medio- lano (? Whitchurch), 150 miles. The first station is Galava (unknown), 18 miles; then follow Alona (unknown), 1 2 miles ; Calacum (? Overtown on Lune), 19 miles; Bremetonaci (Ribchester), 27 miles; Coccio (Wigan), 20 miles ; Mancunio (Manchester), 1 7 miles ; Condate (North- wich), 18 miles; Mediolano, 19 miles.^ Route XI. From Segontio (Caer Seiont, near Carnarvon) to Devam (Chester), 74 miles. Conovio (Conway), first station, 24 miles ; then Vans (? Bod- fari, Flint), 18 miles; Deva, 32 miles.* Sitomago seems to mean Corn-plain. The initial "s" has vanished in the modern Irish Ith = corn, as in Ithtige - comhouses (Thumeysen, Handbuch^ p. 83), cf. Mag-ith. Combretonio appears to involve Com-ber = a bearing together, a con- fluence, as in numerous place-names in Ireland and Wales, Commer, Cumber, plur. in Cumrew, Cumberland. This might lead to the identifi- cation. Ad Ansam is Latin for "at the Bay'*. The French Anse preserves the signification. Stratford would satisfy the conditions, as the tidal waters came up thus far. Canonio is an obscure word. Caesaro Mago =« •* Imperial place " and must have been of importance as the seat, probably, of authority. Zaragoza for " Caesar Augusta " indicates the disguised form in which this name, if it survived, would now appear. Durolito may be represented in a degenerate form in the VVritola burna. An intrenchment and other remains have been found at Writtle, which is about a mile from the Roman road to Colchester and seemingly on that which converges on it from Braintree. The Itinerary may deal with both these routes, and hence the difficulty about the mileage. The fate oiDuro in Durobrivis prepares one for such a transformation. * This route is full of difficulties. We do not know where Clano-venta (or Glannibanta) was. If Clan- represent Glan, as the Not. Dig. indicates, it would mean " market on the bank " of some river, ? the Lune, of which the early name seems preserved in Alone (cf. the Aenona of RavennaSy now Nona). Calacum has been placed at (?) Overtown on the Lune, and Bremetonaci at Ribchester. Coccio is supposed to be Wigan, but these are pure conjectures. From Mancunio the places are clear. 2 Caer Seiont and the river Seiont ("Seintes" in Cah Chart. Rolls, 11. 280), near Carnarvon, preserve a relic of Segontium (= ?the place of the God of Victory). Conovio is obscure, as is also Varis. of which a reminis- cence may be preserved in Bod-fari, Bod = dwelling. 112 BRITISH PLACE-NAMES Route XII. From Muriduno (? an error for Ptolemy's Man- dunon = C2,x- Mar then) to Veroconium (Wroxeter), i86 miles. (The aggregate makes up only i66 miles.) The first eight stations are a repetition of those in Route XV, and do not belong here. The first station of this route is Leucaro (PLoghor), 15 miles from Maridunum; then follow Nido (? Neath), 15 miles; Bomio (? Cowbridge), 15 miles; Isca Leg. II. Augusta (Caerleon), 27 miles; Burrio (?Usk, but of. Din birrion of Lib. Land.), 9 miles ; Gobannio (Abergavenny), 1 2 miles; Magnis(?Kenchester), 22 miles; Bra vonio (Brandon), 24 miles; Virocono, 27 miles.* Route XIII. From Isca (Caerleon) to Calleva (Silchester), 109 miles. Burrio, the first station, 9 miles ; then follow Blestio (Mon- mouth), 1 1 miles ; Areconio (cf. Deanery of Erchen^itldi), 1 1 miles; Clevo (Gloucester), 15 miles; Durocornovio (Cirencester), 14 miles; Spinis (Speen), 15 miles; Calleva, 5 miles.* Route XIV. Another route from Isca to Calleva. The first station is Venta Silurum (Caerwent), 9 miles ; Abone (Mr. Codrington makes this Sea Mills), 14 miles; Trajectus (Mr. ^ Maridunum is another form of Muridunum, Mart = the Sea : early Irish Muir. Leucaro is obscure. Nido, = Neath, has its parallel in the German JVied, which is the survival of the Nida of the Itinerary. Bomio would become later Bovio and may have suggested the Cow (as a translation) in Cowbridge. Isca, of the second legion called Augusta, i.e. Isca Legionis or Caer (= Castra) Legionis. Burrio is obscure but cf. the Din birrion of the Lib. Land. Gobannio ( = ? Smithy) is doubtless from Gobann, genitive of Goibniu = Smith, perhaps, an Irish Vulcan ; cf. Gaulish Gobannitio (Caesar, De Bell. Gall.). Goibniu is an Irish proper name = Smith : Mac Gowan (i.e. Goibnenn, genitive) = Smith's son. Magnis would seem to be Latin, but it is possibly an ablative plural of Magen — Old Irish for " place ", which is the equivalent of Welsh Maen = stone, the common significance being " earth surface " ; cf. Magos (Peder- sen, V. Gramm. % 59). Bravonio is obscure. 2 Blestio is obscure. Areconium preserved in Erchen seems to involve a personal name. See previous note, p. 107. Clevum alias Glevum is preserved in the Glou of Glou-cester. Durocornovium is seemingly the stronghold of the Cornovii. Spinis is the locative plural of Spinae = thorns. ROMAN OCCUPATION - 113 Codrington makes this Bitton), 9 miles ; Aquis Solis (variant Aquis Sulis = Bath), 6 miles; Verlucione (Wans), 15 miles; Cunetione (near Marlborough on Kennet), 20 miles; Spinis, 15 miles; Calleva, 15 miles.^ Route XV. From Calleva to Isca Dumnuniorum (that is Isca of the Devonians = Exeter), 136 miles. The first station is Vindomi (? unknown), 15 miles ; then follow Venta Belgarum (Winchester), 2 1 miles ; Brige (variant Brigae ; position unknown, but between Winchester and Salisbury), 1 1 miles ; Sorbiodoni (in the misplaced stations placed before Route XII it is Sorvioduni = Old Saresbury), 8 miles ; Vindo- gladia (Badbury Rings), 1 2 miles ; Durnonovaria (Dorchester), 8 miles ; Muriduno (somewhere near the Otter river), 36 miles ; Isca Dumnuniorum, 15 miles.' Place-names in the NoHHa Dignitaium. Chapter XXV gives the following places on the Saxon shore {^Litus Saxoniais), that is on the coast exposed to the raids of Saxon and other pirates : — Othona (the Ythancaster of Bede, discussed p. 179), Dubrt's (Dover : p. 107), Lematmis (Lympne, p. 215), Branoduno (Bran- caster : so called probably from a personal name Bran. Cf. Bran, son of Lir (Llyr), in the Mabinogion\ Garriano (Burgh: see p. 176), Regulbi (Reculver, see p. 37), Rutupis (Richborough, p. 37), Anderitos (Anderida and Andredes Wald: p. 143. Cf. Andereton of Ravennas, p. 238, and Anderitum of the Pent ^ Venta Silurnm is the " Market of the Silures ". Abone is the locative case of Abon = river : it would seem to mean the Severn, but 14 miles can hardly be accounted for, except the route was along the shore as far as the Trajectus (i. e. the Ferry), which would be at Aust Passage (Aust possibly from the Legio Augusta), see p. 203 ; but then this is too far from Bath (Aquae Sulis, i.e. "The Waters of the God 5'm/-"). Verlucio contains the intensive Ver and possibly the locos mentioned above = lake, probably the spreading out of a river. Is there such a thing at Wans? Cunetio involves the river-name, now the Kennet. 2 Vindomi is obscure, but seems to involve Vindo = White. Brige is evidently Briga = Burg or Stronghold. Sorbio-dunum, see p. 146, ? from Irish Soirb = easy, perhaps earlier " level". Vindo-gladia = White ditches, i. e. fosses cut in the chalk. Cf. Cladid = to dig, in Old Irish, and Welsh Clawd = trench, grave. Dumovaria is dealt with elsewhere, see p. 138. H 114 BRITISH PLACE-NAMES Taf^^ For turn Adurni (position unknown: probably west of Brighton. The river " Adour " has seemingly been coined from Camden's identification). Chapter XXXVIII furnishes the names of the stations of troops under the command of the Dux Britanniarum — a dignity which Booking thinks {Not. Dig. II, pt. 2, Occid. p. 851*) was first instituted by Constantine. Sexlae, the first on the list, is the station of the Sixth Legion and probably means York, the station of this legion according to Ptolemy. Bocking (ibid. 853*) shows that York was called a Colonia on coins and inscriptions, a Municipium by Aurelius Victor, a Civitas by Spartianus, and an Oppiduin by Bede. Praesidtum, where was stationed the Prefect of the Dalmatian cavalry, is unknown, but cf. Sir John Rhys's Celtic Britain^ ^ pp. 229, 302, where he shows that the Epiacon of Ptolemy and the Vereda of the Ant. Itin. may mean places for horses or cavalry. Dano (near Doncaster, p. 10 8), Morbio (site unknown, but cf. Moerheb, p. 216, and Murief^ i. e. Moray, which in the " Story of Arthur " [Red Book of Hergest) is said to be another name for Reged). Arbeia, the station of the Tigris boatmen, was prob- ably South Shields, where an inscribed stone was found with the designation of these boatmen, who were doubtless required to carry supplies by the North Tyne to the garrisons on the Wall. The attempted identifications of these last two places with Moresby and Ireby on account of similarity of termina- tions is baseless, because the -by is Scandinavian. Dictim, the station of a Numerus of the Dictensian Nervii, is unidentified. Concangios (see p. 95). Lavatres (see s. v. in Ant. Itt'n.). Verterzs (see p. 106). Braboniaco (this is the Brovonacis of the Ant. Itin. It is supposed to be Kirkby Thore, Cum- berland; cf. the Ravonia of Ravennas). Maglone (not men- tioned elsewhere, site unknown. If the name involved Mag^ plain, place, there would be a thematic vowel following it). Magis (locative plural = at the plains, but where these were is not known, but cf. Maes^ p. loi). Longo-vicio (variant Longo- vico =(?)"Long row of houses." Camden identified this, and afterwards relinquished the identification, with Lanchester on the Lune. The site is unknown. It may have been the Lanchester north of Binchester). Derventione, where were stationed troops called Petueriensis, ?from the Petuaria of ROMAN OCCUPATION 115 Ptolemy. There were several rivers in Britain called Dervenito. This may have been that mentioned by Bede, H. E. ii. 9, " Close to the river Derventio where was then a royal vill." Stations along the Wall according to the Noiitia Digniiatum. Segeduno (the Serduno, probably, of Ravennas : it has been identified with Wall's End). Ponte Aelii (Newcastle), called after Aelius Hadrianus. Conderco {Condecor, Ravennas = ? Benwell Hill). Vindohala ( Vindovala, Ravennas = ? White-wall, identified with Rutchester, but the site is unknown). Hunno {Oniio in Ravennas — near Halton). Cihirno (Celun7io^ Ravennas. This was at Chesters near Cholierioidij which, with Chollerion, both on the North Tyne, seems to preserve the early name of that river, " Cilurn " — so-called, probably, as suggested by Sir John Rh^s, from the Cavities in the rocky river-bed here. Welsh Cilwrn = Chaldron). Procolitia {Brocoliti in Ravennas = Carrawburgh. Broco-litt\ from a comparison with such forms as Broco-mago (now Brumat) in the Ani. liin., seems the correct form. Broco- may be associated with Welsh Brog = country, and the ht-i may involve the stem /// = broad). Borcovicio (= House-steads). It is possible that we have in Borco a case of metathesis of Broco. Vic-io is from the Latin Vicus, Vindolana ( = Chesterholm : here we seem to have lan-a^ plain, or land, and Vindo = white). ^sica (Great Chesters). Magnis (Magnis in Ravennas. This is placed at Carvoran). Ambo-Glanna ( = Burdoswald, ?the Gaba Glanda of Ra- vennas. The name contains ^w^ = about. Anglo-Saxon Vmb-, preserved in " Evib-tx days ", and Glann = shore, bank). Peiriams (unknown). Aballaba (Avaiana of Ravennas, unknown site) seems to involve the Celtic Aba/l, 6^<$a// = apple : Appleby is a Scandinavian name. Congavata (Chester-le-Street was called by Richard of Hex- ham " Kunka-Cester " and may preserve this Congavata. It lies about 8 miles south of the Wall). Axeloduno (= Uxelludamo of Ravennas. It is placed by some at Hexham and by others as far away from the Wall as Ellenborough. The Hex in Hexham is a relict of Hagtistald — if we accept the Earliest English form — but it is possible that H2 ii6 BRITISH PLACE-NAMES the latter intelligible word may represent the unintelligible Axelo-dun, or rather Uxello<^\in = '' high fort "). Gabrosenie (Gabrocentes of Ravennas, supposed to be Gates- head. Gabr- = Goat, and seni, Irish set, with the usual loss of « in «/ = path, = Gothic Stnths). Tunnocelo (supposed to be the Fano-Cocidi of Ravennas. Cf. the inscription (Camden-Gough) to the God Cocidius found in Cumberland. Mommsen makes Cocidius a by-name of Mars (Prov. Rom. Empire, i. 193). Watling Street crosses the Coquet, which may be a deified river). Glannihanta (this is seemingly the Clanoventa of the Ant. Itin., which see. It is placed by some at Cockermouth, but the site is unknown). Alione (this is the Alone of the Ant. Itin., which see). Bremetennaco (probably the Bresnetenace Veteranorum of Ravennas. It has been identified with Ribchester). Olenaco (this can hardly be the Olikana of Ptolemy, one of the cities of the Brigantes, which has been identified with Ilkley, but this is far from the Wall). Virosido (site unknown. For the significance of Viro see Vero-lam-ium : and cf. Viro-dun-um {Ant. Itin) now Verdun and Viro-viacum (ibid.) now Werwick. Sido probably repre- sents Irish Suide, Welsh Sedd = seat). The Ravenna Geographer furnishes longer lists of British place-names in Roman times, but it is difficult to identify many of them owing to a seeming lack of method in their arrangement. Instances have been given above which tend to show that he had one or more Itineraries before him, but he is not consistent in following these. In his list of names he seems to associate together sometimes those which are more or less near to each other, but he often departs from this practice. An article on the Roman place-names of Derbyshire by W. B. Anderson, M.A., in Melandra Castle (Manchester University Press, 1906)^ raises this question: the sequence of rivers Dorvantium (Derwent), Anava, and Bdora seem to belong to Derbyshire, so do the places Nanione (or Navione), and Aquis = Buxton. There are several Derwents, it is true, but a mile-stone, discovered in 1862 near the Silverlands of Higher Buxton (now in Buxton Museum), furnishes, as we see from the article, the name Anavione. A fragment, moreover, of a sepulchral inscription found at Foligno in Italy records ROMAN OCCUPATION 117 (1102 in Eph. Ep. vii) a Cemiior Brittonum Anavion\ensium\ Mr. Anderson, from the facts (i) that the river Noe joins the Derwent about one and a half miles below Brough, and (2) that there was a Roman fort at Brough, to which the Roman road (now Batham Gate) leads, (3) that Brough is about the right distance from Buxton, viz. ten or eleven miles according to the miliary, infers that Brough represents Anavio, so called from the Anava now the Noe. The third name Bdora, although it has been placed far afield, may have its reflex in the Derbyshire Dore, otherwise (see p. 239) hard to explain. There is a stream at Dore, but it is now called '' the Sheaf". NOTE E. The Legions in Britain, see p. 81, ante. Mommsen (Prov. Rom, Empire^ i. 174, London, 1909) says that for the subjugation of Britain three of the Rhine Legions and one from the Danube were destined thither, and adds : — "The three legions of the Rhine were the 2nd Augusta, the 14th, and the 20th ; from Pannonia came the 9th Spanish. The same four legions were still stationed there at the begin- ning of the government of Vespasian; the latter called away the 14th for the war against Civilis, and it did not return to Britain, but, in its stead, probably the 2nd Adiutrix. This was presumably transferred under Domitian to Pannonia; under Hadrian the 9th was broken up and replaced by the 6th Victrix. The two other legions, the 2nd Augusta and the 20th, were stationed in England from the beginning to the end of the Roman rule." B. W. Henderson {Life of Nero, pp. 202 et seq., and Eng, Hist, Review, Jan., 1903, states that in a.d. 49 the 9th legion was at Lincoln, the 14th and 20th at Wroxeter, the 2nd at Gloucester. The 2nd was shortly afterwards pushed forward to Caerleon and the 14th and 20th to Chester. After Nero's death, in a.d. 68, the 14th was withdrawn from Britain and the 9th removed to York. NOTE F. Tacitus, Ann. xii. 31, see p. 22, ante. Mommsen, Prov. Rom. Empire, i, p. 178, reads, " Ostorius makes ready to secure all by camps (extending) to the . . . ntona ii8 BRITISH PLACE-NAMES and Sabrina rivers," and places the camp for the 14th legion at the confluence of the Tern with the Severn near Viroconium (Wroxeter). NOTE F. 2. Some Latin words which had entered into place-names before the Norman Conquest : — Calx, as *' Cealc" in Cealchythe. Castra, as Castor, Chester. Castellum^ as '' on tha Csestello", Cart. Sax. iii. 338. Castellehim, as Cistelet. Chors {Cohors), as in '' Dovor-Cortae ", Cart. Sax. iii. 603. Cerasus, in numerous Cheritons. Fons, as in " Ceadeles funtan ", now Chalfont, Cart. Sax. iii. 40; "Teofunten", now Teffont, Wilts., ibid. 385; "Funtines burnan," ibid. 588, cf. also " Fobbefunte ", now Fovant, Wilts., ibid. ii. 232; Funtg^all", now Fonthill, Wilts., ibid. ii. 234; " Hamanfunta", now Havant, &c. Fossa, as '* fossa", cf. Foss-way. Palantiwn, as in " the Palant " at Chichester. See Index. Palus, ? as in '' Plash ". Pirus, as " Pere " and " Pirige " in Pirigtun, now Purton. Populus, as in ''Populfinig", Cod. Dip. 652. Portus, as in Portsmouth. Porta, as in *' Port straet ", &c. Prunus, as in " Plumstede'', now Plumstead, Cod. Dip. 562. Puteus, as in '' hwytan pyt", Cart. Sax. ii. 81. Silva, as in " Monks' silver ". See Index. Sptnae, as in ** Spene ", now Speen, Cart. Sax. i. 506. Spinetum, as in Spinny. Strata {via), as in " Straet ", passim. Sulcus, as (?) in " Sulhford ", Crawford Charters, i. 1 1 . Vicus = row of houses, dwelling, or market in the " Wich " of place-names. Latin topographical survivals in Celtic districts are naturally more numerous ; some of them, like Cornish pras, prat (for pratum), are incidentally dealt with as they occur in this work. It is but right to add that the Teutonic invaders may have imported a few of the Latin terms here given. See Pogatscher {Lautkhre). Ecclesiastical elements in early English topography are ex- cluded from this list, as they could not well be survivals from Roman times. II. THE TEUTONIC INVASION It is commonly believed that the Teutonic tribes — Saxons, Angles, &c. — invaded Britain about the year a. d. 449. It is clear, however, that some of them had entered this country long before. Apart from the Saxon and Prankish pirates who were making continuous expeditions against the " Saxon shore " from as early at least as a.d. 297, and who also took service in Britain under Carausius, there were, according to Ammianus Marcellinus (writing about a.d. 378), Saxons,^ in alliance with Picts and Scots, fighting against the Romano-Britons in 365. Alliances, begun so early, seem to have established a precedent, for Bede tells us of two subsequent occasions when the Teutonic invaders had Picts as their allies, one which, although it is ascribed by Bede to 449 (Bede, H, E. i. 15), is relegated by Plummer (ii. 33) to possibly the fourth century, the other in 430, the occasion of St. German's^ *' Hallelujah victory ", which shall be dealt with later. The entente cordiate between the Northumbrian English and the Southern Picts was later on cemented by marriage. Eanfrid, who was a great- grandson of Ida (King of Bernicia, a. d. 547), was, according to Skene, the father of Talorg Mac An frith. King of the Picts, who, following the Pictish custom, inherited the kingdom from his mother's side. The personal names (involving Peoht) of distinguished Northumbrians about this time, to which refer- * " Picts, Saxons, Scots, and Atacotti harassed the Britons with con- tinuous calamities." (Bk. xxxvi. 4.) 2 Bede's account of the "Hallelujah Victory*' is taken almost word for word from the Life of St. Gervian^ ascribed to Constantius, a presbyter of Lyons : see later note, p. 125. I20 BRITISH PLACE-NAMES ence was previously made, testify to friendly relations. Peoht- helm, for instance, who was Bishop of Whitern in the eighth century, means the helmet, or protector of the Pict. The easy conquest of Galloway (the region of the ''Niduarian Picts " or Picts west of the river Nid, or Nith ^) by Oswy, and the subsequent strong Anglicizing influence exercised over this region, are thus explained. Here, at least, the Teutonic invaders were not "exterminators". The first Teutonic settle- ments on the East coast, south of the Forth and north of the Humber, were made probably as early as the fourth century. They consisted to some extent presumably of Frisians, whom Procopius, writing about a. d. 560 {De Bello Goth, iv. 20), associates with Angles as invaders of Britain; and Dumfries, i.e. the Dun or stronghold of the Frisians on the Nith, is considered still to preserve their name. The German Ocean is called in the Historia Britonum of Nennius (c. 38) the Frisian (Fresicum) Sea, and (c. 37) the Tythica Valley. In Jocelyn's Life of St. Kentigern Liius Fresicum ("Frisian Shore") is applied to the district of Culross (see Sir H. Maxwell's Scottish Land-names^ p. 73, who also makes the Caer P hen's of Nennius to be Dumfries). The Lowland Scotch has some points of resemblance to the Frisian dialect, and certain old Frisian place-names seem to find an echo in the Lowlands. Leith on the Forth finds, for instance, its satisfactory signification in the Frisian Leith ^ = passage or ferry, which corresponds to the southern **Lode" (compare the Lodes on the Severn) repre- senting the Anglo-Saxon "Z«^, Geldd ", and still surviving in the present Yenlet (for Bede's Genlade) in the Thames-mouth, and * Ptolemy places the river Novios about fifteen miles to the west of the estuary of the Ituna, which has been identified with the Eden, the river flowing past Carlisle into the Solway Firth. The Novitia of the Ravenna Geographer is supposed to be the same river and to represent the early form of the Nith. But Novios could not become Nith, and it is pure conjecture that it is the Novitia of Ravennas. The origin of " Neath " given in a note further on suggests another source. ' See Outzen, Friesisches Glossarium^ p. 186. THE TEUTONIC INVASION 121 also in Inlet. In the last syllable of Prender^^j/ (Berwickshire)* there is possibly a representative of the Frisian Gaast = morass, but as the remainder of the word is presumably Celtic, it is also possible that Gast or Gest may be a corruption of some Celtic form. Although Gasi is of early occurrence (Prene-gest, Cold- ingham Charter, a. d. hoc), it seems, like -Wyndes, mentioned in note below, of Teutonic origin, but it may be a corrupt form of the termination -gask (= Casach, Gaelic for "outlet of a lake", or ford), which appears, for instance, in composition with Teutonic words, as in (Buttergask) " Butter-gask ", where " Butter " prob- ably means Bittern, cf. the Lincolnshire " Butterbump " as a name for that bird, and the English Buttermere. The difficulty of finding topographical evidence of the existence of Frieslanders in the Lowlands of Scotland is increased by the fact that this region, originally the home of Picts and Britons, was successively overrun by Goidels and Norsemen, the existing place-names furnishing abundant testimony to the influence of both these peoples. The earliest direct notice of the invasion of Britain by Teutonic tribes is in T/ie Destruction 0/ Britain, a Latin work written about a. d. 537 by a Briton called Gildas. The Britons, fearing a return of their troublesome enemies the Picts and Scots, invited the help, as Gildas bitterly says, of a " most ferocious people of the execrable name of Saxons, hateful alike to God and man ". " Then," he continues, " bursting forth from the lair of the barbarous lioness, a flock of whelps, with favouring gales and omens ... set forth, at the instance of the ^ Plenderleith in Roxburghshire was earlier (1587) Premderleith. Plen- derguest and Prandergaist occur also as equivalents; cf Pranderwyndes. The word Preinder is suggestive. The early Irish word for priest was Cruimther, which appears even in an Ogam inscription, Rh^s, Lectures '', 349 et seq., and is from Preshiter^ through Frebiter. Cormac, writing in the ninth century, gives in his Glossary an Old Welsh form Premier (see Pedersen, V. Gram7n. § 124. 6). Premderleith would be thus "Priests'- ferry", and Pranderwyndes, " Priests' -lanes." "Priest" enters into many English place-names. For English words having exclusive parallels in Frisian and Dutch see Kluge in Germanische Philologie, vol. i, p. 793 ; Siebs, ibid,, pp. 724-9, treats of Frisian and Northumbrian. 122 BRITISH PLACE-NAMES unfortunate Tyrant (? Vor-tigem = Over-Lord) in three Cyulas, as expressed in their language, in our ^ language, Longae Naves (i. e. long ships), and fixed their terrible claws at first upon the -eastern part of the island, ostensibly as if to fight for the country, but in reality to assail it." Supported by further arrivals from the Continent the Saxons gradually encroached upon the Britons, and in a few years overran the country, probably only at first in marauding bands, " the eastern fire, heaped up moreover by sacrilegious hands, laying waste," as Gildas says, " the adjoining towns and country — an un- ceasing kindling — until consuming at length almost the whole face of the country from sea to sea, the flaming and fierce tongue was licking up the western ocean." Some of the natives, he says, were slain, or perished from hunger — while others gave themselves up to slavery or passed beyond the sea.^ At length the miserable remainder taking heart, and with Am- brosius Aurelianus, a man of Roman race, at their head, gave the invaders battle and obtained a victory over them. This was, as we gather, in 466. " From that time," Gildas says, " some- times the natives (Owes) and sometimes the enemy were the victors, until the year of the siege of the Badonic Mount [which was near the Severn mouth], the year of the latest and not the least slaughter of the Gallows-rogues; and the year which is now beginning, its first month having already elapsed, is, as * This is one of the proofs that the Britons had become in the main Latin-speaking under the Roman occupation of 400 years; "Nennius" also uses the Teutonic word Chiulas not only for the ships of the Saxons, but also for those of the Roman fleet. The Old English form was ceol, still used on the Tyne for a flat-bottomed boat (keel). Bede.(i. 25) employs the Latin term Longae Naves. The Welsh Llong, Irish Long = ship, was derived at an early date from the adjective. It gave rise in Irish to Longeas, a voyage, and L.oingseac h^voysigtr, sailor, now "Lynch", a personal name. Loingseach Lie ocus Arann in Chron. Picts and ScotSy p. 99, is rightly translated by Skene, " mariner of Hay and Aran." 2 This doubtless refers to the British emigration in the early sixth century to the Continental district called then Armorica ; see later on where this subject is dealt with. THE TEUTONIC INVASION 123 I have discovered, the forty- fourth from that year, the year also of my birth." ^ Mons Badonicus has been placed by con- jectures in various localities. All we know is that it was pre- sumably east of the Severn mouth and high land. If the name survives I would suggest that it is preserved in Bown Hill. Badon would naturally become Bown, and the neighbouring Frocester and Woodchester indicate Roman ways accessible to the invaders. Prosper of Aquitaine (a. d. 403-63) tells us, under the year 413 (Mon. Germ, Hist,, Chron. Min,^ p. 467), that " Pelagius Britto (that is, a Briton) ^ put forth at this time the doctrine called after him (Pelagianism) against the grace of Christ, and that, aided by Caelestius and Julian, led many into his error. It was this Pelagius and his teaching which were the occasion of Britain's coming again into the light of history.' For, as the same * This is my translation of Plummer's reading of the original (see his Bede). Some critics (Mommsen, for instance) regard as a later gloss the words " which was near the Severn mouth ", but the passage represents in any case ancient tradition. Bownhill is one of a range bristling with formidable earthworks which seem to have owed their existence to the need of commanding the passages across the Severn mouth. Woodchester, which is close to Bownhill, is possibly the Kaer zntdei of the Ystorya Brenhined, pp. 184, 204, 234 of Red Book of Hergest, vol. ii. Meugant was made Bishop of Caer Viidei, see excerpt from The Story of Arthur in Introduction to Early Welsh, p. 165, but although Dr. Strachan identified the other Caers mentioned therein this is not explained. Mommsen {Chronica III, Mon. Geiin. Hist., p. 8) gives the date of the Epistola of Gildas as a few years before 547. The date of his birth and of the battle of Mons Badonicus is, according to both Mommsen and Zimmer, a few years before 504. Bede's date works out about 499. The Annales Cambriae place the battle at 516, The A.-Sax. Chron. records a battle between Wulfhere, King of Mercia, son of Penda, and yEscwine, King of the West Saxons, at Biedan Heafde in 675. This may refer to a second battle at Mons Badonicus given in Annales Cambriae under 665. Mr. Plummer makes Biedan Heafde to be Bedwin, Wilts., Bedvinde in Domesday, and seemingly Celtic. 2 Orosius (writing about A. D. 415) and St. Augustine (about A.D. 417) call him also a Briton. ^ Earlier references to British affairs are to be found in the notices of St. Ninian (died A. D. 432), of Palladius (to whom Prosper of Aquitaine 124 BRITISH PLACE-NAMES writer tells us, under a. d. 429, " Agricola, a Pelagian, the son moreover of a Pelagian Bishop called Severianus, corrupted the refers under 431), and of St. Patrick, whose mission to Ireland is assigned to A. D. 432. Bede {11. E. iii. 4), writing in a. d. 731, states that " a long time ago, as it is maintained, the Southern Picts, having given up their idolatry, had received the true faith by the preaching of the Word through the most reverend Bishop Nynias, a most holy man of the British race, who had been fully instructed in the faith and in the mysteries of the truth at Rome ; whose episcopal See and Church were distinguished by the name of St. Martin the Bishop, and in the place, which now the English hold, his body and those of many saints rest, — which place belonging to the Bemician province is now commonly called Candida Casa ( Whitehorn — in Galloway), because he had there built in a manner foreign to the British, a church of stone." Alcuin (born at York A. D. 735, died Abbot of Tours a. D. 804), tutor of Charles the Great, also mentions Ninian in his epistles, and asks the intercession of St. Nyniga, and that of the monks, at Candida Casa, to whom he sends a silk veil for covering the relics of the saint (Usher, De Frimord. Britt, Eccl.y p. 669, and Haddan and Stubbs, Concilia, ii. 8). Alcuin mentions also, in his letter to the nobles of Kent (Haddan and Stubbs, Concilia, iii. p. 510), Gyldum Brettonuni Sapientissimiim. The life of St. Ninian, written by Ailred of Rievaulx (died 1166) is too late to be of historical value, but the traditional association of Ninian with St. Martin, who died about 400, helps to fix his date. The Anglo-Saxon Whitem (Whithorn) is a translation of Candida Ca^a = white house. In the Martyrology of Oengus it is " Futerna in the ' Rinns of Galloway ' " (p. 212). *' Futerna" is an instance of the difficulty to Gaelic lips of pronouncing the wh (or Hw) in Whitern. The substitution of initial " F " for " W " was not uncommon also, after the eleventh century, in the dialect of Kent {Grund. Germ. Fhil. i. 857). The "Rinns" are points of land jutting out into the sea. The ancient Irish form is Rind (masc.) = point, Rind (fem.) =Star (Pedersen, V. Granwi. p. 37). There is a further refer- ence in the Martyrology of Oengus (compiled about A. D. 800) to a place in this neighbourhood called Dun-rechet, now probably Dunragit in Galloway, which may contain the name of the region of " Reged " referred to in Welsh tradition as imposed upon the district of Gower after having been presum- ably used elsewhere (in the North). Reged has been located ( Celtic Scotland, i. 153) in the country near the mouth of the Clyde. The Welsh version of Geoffrey says Reget is another name for Murief { = ^ox2iy). Whithorn was also known according to Skene {Celtic Scotland, ii. 47), as Rosnat or Alba (white) and Magnum Monasteritim. The evidence of this is strong, but not convincing. Roseneath in the estuary of the Clyde has an outward resemblance to Rosnat, and tradition assigns to it an ancient ecclesiastical importance {New Statistical Account^ vol. viii), but it was, however, in THE TEUTONIC INVASION 125 British churches by the introduction among them of his teach- ing, and that, at the suggestion of Palladius, a deacon, Pope Caelestine sent German, Bishop of Autissiodorum (Auxerre in France), in his stead and guided the Britons, perplexed by heresies, into the CathoUc fsiiih" {Af on, Ger??t. Hist, ix, p. 472). In the Life of St. German {Suritis, July 30), which was written by Constantius, a priest of Lyons, about a. d. 480, this event is alluded to in different words. The Britons, according to Con- stantius, sent delegates to St. German, asking him to come to their aid, and after a Synod had been called, it was agreed that St. German, accompanied by Lupus, Bishop of the Tricassini (that is, Troyes in France) should go to Britain. They travelled through the Parisii (Paris), where St. German persuaded St. Gene- vieve,^ afterwards the Patron Saint of Paris, to dedicate herseh to God, and, setting sail from some port on " the Gallic Gulf", made for Britain. A violent tempest arising on the way, he calmed the waves, after invoking the Holy Trinity, " by pouring oil upon them," and reached safely the British coast. The heretical teachers, who are depicted as conspicuous by their wealth and splendidly robed, appeared, with great multitudes of the people, including women and children, before St. German, who proposed an open discussion, which was accepted. The heretics were beaten, and St. German, proceeding on his way back, stopped at Verulamium (as we have already said) to return thanks to God at the tomb of the Martyr St. Alban. The place of meeting was, it would thus appear, not very far from St. Albans. a district called in early charters Nevet, and was doubtless originally Ros-nevet or the promontory of Nevet (representing possibly an older A^', as the Welsh loses the THE TEUTONIC INVASION 171 followed — the baptism of Blaecca, the prefect of the City of Lincoln {Lindocolina), and of his whole household being the first-fruits of the preaching of Paulinus, who afterwards (a.d. 633) was made Bishop of Rochester — he having left in York Jacob, a deacon, who gave his name, Bede adds, to the vi'cus where he dwelt (conjectured to be Akeburg). Bede acknowledges his indebtedness for his history of the beginnings of Christianity in this region to a certain Deda, Abbot of Peartaneu,* who had equivalent gu, and becoming ^;rrt«, as in the several islands of that name and in the designations of 'Mand" in the West of Ireland; cf. Joyce, Irish Place-Names^ ii. 360). Lindisfam (also called Medcaut, and in Tigernagh's Annals under a. d. 632, Inis Metgo'it. In the Felire of Oengus it is Inis Medcof) is therefore of the same significance as the Anglo-Saxon Ea-landy that is water-land, island. Fame is the name of an island close to Lindisfarne. The Guern in Guern-s-ey, and "the Wame", although no longer above water, may possibly belong to the same stem. Simon of Durham, i. 51, derives the Lindis of Lindisfam from the name of a river flowing into the sea opposite it ; Leland calls the Witham (early form Widmd) river flowing past Lincoln the "Lindis", and also the Hhee, a generic word ( = also Kee, Key, Rei) for stream in late English, repre- senting Anglo-Saxon Kith cognate with Irish Kiathor ( = torrent). Old Welsh Reatir now Rhaiadr ( =• waterfall), Latin 7?/-vus, Pedersen, v. 9, § 45. The transition from rith to rey is shown in the Berkshire Child-rey from the form Cilla-rith of the early Charters {^Cod. Dip, 746, 1133, &c.). In Holinshed's Chrotiicle, Rhee is said to be a general word for stream. The New English Dictionary pronounces the word of obscure origin, and the Eng. Dialect Diet, has no light to throw upon it. (i) The frequent occurrence of Rith in Anglo-Saxon place-boundaries implies that it must have a modern representative. (2) If Ree be not that representative there seems no other to hand. (3) The instance given above shows that it has become rey^ and the conclusion that lith = ree seems fully justified. Rye, which occurs in place-names, is quite a different word. On the Welsh border it represents /'^zw = slope, as in Rhiw-las (= green slope) in Hereford, which is represented a few miles away (near Dymock) by Rye-lass. Lind-Colina {Ang\o'Sa.xon Zind-Cylene) contains "Colonia", although the existence of an Anglo-Saxon Lind-Cylne in Kent, where Lind = Limetree, makes the connexion obscure. * The fact that Peartaneu (now Partney, not far from Spilsby) begins with a /*, and has an ending {eii) of which the significance is not clear, suggests that it is not an English word. Parthenay {tzxXxtxPartenactmi) too, occurs in France, and is derived by Holder {Sprachschatz) from a personal name Parthenus, 172 BRITISH PLACE-NAMES had from the lips of an old man an account ot the latter^s baptism and that of a great multitude by Paulinus in the River Trent {Treanta) near the city called in the language of the Angles Tiouulfingacaestir/ King Eadwin himself being present on the occasion. Eadwin seems to have owed his advancement to Christian influence. For it was while he was a fugitive from ^thelfrith's persecution at the Court of Redwald, King of the East Angles, that he secured, as Bede tells us {H. E. ii. 1 2), by divine inter- position, the help of the latter, who, in a battle on the east bank of the river Idla** (a.d. 627), slew ^thelfrith and helped to place Eadwin on the throne of Northumbria. Eadwin's influence with Redwald led afterwards to that King's conversion and to a tem- porary spread of Christianity in East Anglia. But it was to Sige- berht, who had been baptized when an exile in Burgundy, that the conversion of the people was mainly due. He succeeded in a Latin form of the Greek word for Virgin. Most 01 our words connected with the cloister are derived from the Greek, viz., "monastery," " coeno- bium," *'monk," "nun," "anchorite," &c., and it is not quite impossible that the Greek Parthendn, which is used for a nunnery (Life of Anthony in the Lausiac of Palladius), should also have passed to the west. The form in the Anglo-Saxon of Bede's History is, in this respect, suggestive. Peortan-ea tham ham = of the ham or home of Peortanea. The eu termination is here made into ea, which would indicate " river ". The oldest Anglo-Saxon form, however, of the latter word, as far as I can find, is ea, as in Bede's "Homel-ea" (now the Hamble river), and eu does not become ea, but eo, as in Bede's Hreutford for Hreodford (Redford or Redbridge, Hants). The eu represents, therefore, a different and probably a foreign element. All this is pure speculation, and the name in a Sussex charter Peartinga-wyrth (identified by Mr. Birch with Petworth) leads to further surmises. ^ Tiouulfingacsestir has not left any relic ot its name, but it has been identified with "Little borough" (the Roman Segeloaim or Agelocuni) where the Roman road crosses the Trent. It has nothing to do with Torksey, which in 873 was Turcesig. 2 Idla is the Idle river in Notts.; as the form in 1200 was Yddil (B. Mus. Index of Charters) it looks as if the a termination is for ea, river (Old Frisian a). In any case the origin and meaning of the word are obscure. THE TEUTONIC INVASION 173 A.D. 631, and made Felix, who was born and ordained in the same region, to be bishop of his kingdom, assigning to him Dunwich (JDommoc or Domnoc) ^ as his See. The battle of Haethfelth- (Oct. 12, 633) in which Caedualla, the King of the Britons — Penda of the royal house of Mercia assisting him — overcame and slew Eadwin, put a temporary check to Christianity in Northumbria. Paulinus, who had been a missionary in this region for some years, then took ship and sailed, with Eadwin's widow, to Kent. A kinsman of Eadwin, Osric, who had been converted by Paulinus, succeeded to the southern part of Northumberland (Deira) and Eanfrid, son of ^thelfrith, who, having been an exile with the Scots and Picts, had been brought up as a Christian among them, succeeded to the northern portion called Bernicia. Both these kings, how- ever, returned to idolatry, and Caedualla, who slew them, was regarded by Bede as an instrument of divine vengeance {H. E, iii. i). Oswald, the brother of Eanfrid, who had been in exile with him among the Scots, succeeded to the Kingdom of Northumbria and slew Caedualla at a place called, in the language of the Angles, Denisesburna,' that is, the '* Stream of ' Dumnoc is Celtic and involves a term which means " deep " {Dumnos or Dumno-s = the modern Welsh Dwfn). It probably signified a port with a deep-water approach as distinguished from shallows with shifting sands. The oc is an adjectival termination. The use of the cognate Dutch (Frankish) word Diep in Nieuwe Diep^ and Dieppe^ is a parallel instance of the employment of such a word to designate a port, ^ Haethfelth is supposed to be Hatfield Chase near Doncaster. Nennius and the Annales Cambriae call this the battle of Meicen, The Welsh and Irish Chronicles are two or three years behind, and place the battle in A. D. 630. Caed-ualla is represented in Welsh history as having been King of North Wales {Gwenedd), but if this were the case this territory probably reached further north at that time. Chester lay waste (hence West Chester) after ^thelfrith's victory, and Eadwin's writ hardly ran there. The point of contact between Caedualla's and Eadwin's territories was probably to the north-west of Chester. ' The site of this battle is now well known. In a thirteenth-century charter we find the Archbishop granting ** xx acras terrae . . . in Rtileystal 174 BRITISH PLACE-NAMES Denisus ". This battle began at a place a little north of the Roman Wall called by Bede Hefenfelth, which he places not far from Hexham (Hagustaldensis Eccksia), but ended in the victory over Caedualla at Denises-burn (a.d. 634). Christianity was restored in Northumbria by Aedan, sent, at Oswald's request, by the Scots, to whom Oswald assigned the island of Lindisfarne as his See. After reigning nine years Oswald was slain at Maserfelth ^ by Penda (642), and was succeeded by Oswiu. His relics, Bede tells, were taken to the monastery of Bardney {Bardaneti),^ . . . inter Denisesburne ei Devilis'\ that is, twenty acres of land — in Ruleystal between Denises-burne and Devilis. Denises-burn is now Rowley Water, and Devilis, which is one of the many forms into which Dubglas ( = black or peaty stream) has come, is now, by a natural attraction. Devil's Water (Speed's Map, 1610, was Dovols fi.), and gives its name also to Dilston (see previous reference to Dubglas). The battle began at Hefenfelth seven or eight miles north of Hexham beyond the wall, where a chapel existed for centuries after, built, as alleged, on the site of the cross which Oswald raised there. Leland identified the spot with Halidene, now Hallington. Hexham is the modern form of Hagustaldes- ham. Hagosteald meant a young soldier, or a bachelor. The modern German Hagestolz = old bachelor, preserves the latter meaning of the word. It gave name also to a place, Hcegstaldes-cumb in Somerset (Cari. Sax, i. 97). It is possible, however, that we have here a case of attraction, and that the original word was Romano-Celtic, perhaps some compound of Uxello = high, as in the Axeloduno of the Not. Dig. ( Uxelludavio ol Ravennas) with which Hexham had been identified, although there are difficulties with the surmise. Nennius calls this battle Cat-scaul^ " Cat " = battle, and *' scaul" may be for is {g)uaul =■ beneath the zuall. ^ Maserfelth is commonly assumed to be Oswestry, which in Welsh is Croes-oswald, or Oswald's Cross. Nennius and Ami. Cambriae call it Bellum Cochboy. And Simon of Durham places it at " Waneloc" (? Wenloc) which he makes = Croes-Oswalt. 2 Bardaneu is distinct from Fartaneu, but owes, seemingly, its form to the same original influence. What that influence was is not easy to say. In Die Heiligen Englands^ p- ii> the name is " Bardanege^ where rest near the river Withma, St. .^thered, the king, and St. Ostryth, sister of King Oswald ". The termination eu is here made ege — island. The Withma is now the Witham river, the earlier name suggesting an obscure Celtic original. THE TEUTONIC INVASION 175 Oswiu fought and slew Penda at Winwidfeld^ in 654 (at the river Uinuaed, according to Bede). Mercia had been Christianized, according to the Saxon Chronicle, \n 652, and Penda was there- fore nominally a Christian at the time of his death. Wessex during these years had been extending to the westward, the Chronicle recording a victory over the Britons (\Valas) at Beandune [somewhere probably on the high lands north of Bath, and not at Bampton (Oxf.) or Bindon (Dorset)] in a.d. 614, but Mercia was coming into conflict with Wessex, for in 628 the Kings of Wessex, Cynegils and Cuichelm, fought with Penda at Cirencester {Ct'renceasire), the battle ending, however, in a truce. Pope Honorius sent Bishop Birinus in 634 to evangelize the West Saxons, and Cynegils was baptized in the following year at Dorchester {Dorcic or Dorcic-ceastre^ in Oxfordshire, and Cuichelm in 636 at the same place. Cenwalh, according to the Chronicle, succeeded to Wessex in 643, but was driven out by Penda two years after. He was not baptized until 646. Birinus was succeeded in 650 by ^gelbeorht (JElgelbiryth), a Prankish Bishop, who had been educated in Ireland. In 652 Cenwalh is represented in the Chronicle as fighting at Bradford-on-Avon {Bradan forda be Afne), and six years later he overcame the Britons (Walas) at the Penns (Ptf^«;//^7;/ = Penselwood), and drove them as far as the Parret (to the Pedridan)? The extension of Christianity in East * Uinuaed is supposed to be the river Broad Aire which flows by Leeds. The form suggests some Celtic river-name, beginning with Vind-o ( = clear, white, as opposed to Dub-o, "dark"), of which there were numerous instances, some of them coming down to our own day, e.g., \Vinfrith Win-ster. ^ Dorcic-ceastre appears in the Chronicle also as Dorces-cosiSitQ, illus- trating the tendency of the Teutonic settlers to make the Celtic prefix in a name they did not understand into the genitive case ot an imaginary holder, cf. Rofes-ceastre and Raculfs-ceastre. Dorcic here may represent the Durcinate of the Ravenna geographer, but the latter seems from the sequence in which it appears to be in East Anglia. Penn is here, as often elsewhere, made Peon. It represents an early Pend- (which in late Cornish becomes Pedn) meaning *'head". It is 176 BRITISH PLACE-NAMES Anglia at this time is shown by the entry in the Chronicle under 654, " Botulf began to build the minster at Icanho," which was afterwards, it is assumed, called Botulfston, that is, Boston, in Lincolnshire.^ In the records of the spread of Christianity in Britain, during the seventh century, numerous localities of an interesting character are mentioned. Bishop Felix, who came from Bur- gundy, then but lately brought by Irish monks to a knowledge of Christianity, laboured in East Anglia, where he was assisted later on (a.d. 633) by a small band of Religious from Ireland. Fursey, the leader of these, erected a monastery within a Roman fortress, probably that of Garianno S^Not, Dig\ then known, according to Bede, as Cnobheresburg,^ and now identified with here in the English dative or locative plural ending in -um. The Pedrid-an river is Celtic, but comes to us through Teutonic sources. It may mean the " chief ford " -river, that is, Pen-rit, which would pass naturally into *Perrit, the Parret— but this is purely speculative. ^ Ican-ho is made up seemingly of the genitive of some shortened personal name, lea, and ho {kdh), a promontory, but the lean may preserve a remi- niscence of the people called Icetti, whose chief centre was at Venta Icinorwn, near Norwich. St. Botulf is said in Die Heili^en Englands (p. 11) to have been buried at Medeskamstede near the river Nen: and to rest (p. 15) at Thornige (Thomey Abbey) with several other saints. The battle at Bradford-on-Avon seems to be that described by William of Malmesbury (i. 23) as *'in a place called Wirtgemesburg ". William lived sufficiently near to be acquainted with any local tradition, but nothing is known of a citadel of Vortigern in this locality. Nennius makes the Citadel {Arx) of Vortigern to be " beside the river Teibi in Demetia ", that is apparently Cardigan, which is by the Tivy in South Wales. '' Caer Guor- thigim," in his list of Civitates, is elsewhere possibly Tintern for Dinthigirn or Dindeym, or fort of the chief. Vortigern = Over-lord. 2 The personal name Cnobhere involved in this name is seemingly made up of Cnapa = boy, servant (now English '' knave ") and Here = Army, host. Cf. Marah-Scalc (= Marshall = horse-boy) Gottes-scalc, Sene-scalc (= Seneschal), &c. Cenapa occurs as a moneyer in the time of ^thelstan (Grueber, A.-Sax, Coins), and Cenepes-moor appears in a charter {Cod. DipL 120). The name Cnofwealh (Cnobwalch) is in the Durham Liber Vitae. There are, as far as I can discover, no continental forms similar to Cnobhere, and it is quite possible that we have here a pre-Anglian name attracted into a form intelligible to Teutonic ears. The monastery THE TEUTONIC INVASION 177 Burgh Castle close to the junction of the Yare and Waveney rivers. In the life of Fursey (Capgrave, Nova Legenda Angliae) — of which some form was used by Bede — he is said to have worked twelve years in Britain, after which, in his desire for a more secluded life, he left his monastery to his brother Foillan and to two Presbyters, Cobban and Dicul, and proceeded to Gaul. Here he founded a monastery called Latiniacum (now Lagny) not far from Paris, and died about 650, his reHcs being afterwards enshrined in a newly built church at Perrona (Pdronne) in Picardy. According to a MS. [E] of the Saxon Chronicle^ written in the first half of the thirteenth century, the King of Northumber- land and the King of Mercia erected in East Anglia a minster in 661, to which ** the name of Medeshamstede was given, be- cause [as the MS. says] there is a well there called Medes-wael ".^ This was known afterwards as Peterboro. The same MS. alleges that a grant of lands and privileges was then made to thg minster, but it is clear from the names of the places cited, &c., that no such grant could have been made at this time. Bishop Tuda, who was present at the consecration of this minster, died the same year, and according to this MS. was buried at Wagele," but according to Bede at Psegnalaech. Peada, King of Mercia, was built within the "Chester" {on Sunire Ceastre, A.-Sax. version), which possibly had then a name — perhaps the *^ Garian" of Gariannum, now preserved in the river " Yare ". ' Medes-WKl suggests tlie personal name McBthj a contracted form for some such appellation as Mocthhelm, which occurs in the Lib. Vitae. ^ It is clear that the early form in Bede, " Pcegnalaech," is the correct one, and that the twelfth-century chronicler has made, owing to a common mistake, the /'into a W, which it resembles in the A.-Sax. character. Pxgna is in form a genitive plural of Paeg, and probably represents a British tribal name. Laech = Leah in A.-Sax., which came to mean field, meadow, but was originally " wood "; see Stevenson's Assers Life of Alfred, p. 272, and Kurths {Fronii^re Linguisiique, i, p. 365), who cites the examples of Belgian place-names ending in -loo (e.g., Waterloo), earlier -log. The place has not been identified, but Paull in Yorkshire, in Domesday Book Paghal, and in later documents Pagula (see Poulson's Holderness\ may represent the name. It has nothing to do with Whalley. M 178 BRITISH PLACE-NAMES and his following were baptized, according to Bede, by Bishop Finan (one of the Irish Clergy from Lindisfarne) at the royal vill called Ad Murtim (Walton) where, Bede tells us, Sigebert also was baptized by Finan. The conversion of all Mercia by the Lindisfarne missionaries followed. Diuma, one of these missionaries, was made by Finan Bishop of the Southern Angles and Mercians, but died shortly after (658) in a region called In Feppingum^ being succeeded by another Irish Bishop called Ceollach (now Kelly), who, shortly afterwards returning to lona (Hy), was followed in the See by Trumher, an Englishman, ordained by the Irish of Lindisfarne. He had been abbot, as Bede tells us (i7. E. iii. 24), of a monastery called Ingetlingum [the place where Osuiu was slain, now Gilling, Yorkshire]. After the death of Felix, Bishop of Dunwich, in 647, Honorius, Archbishop of Canterbury, appointed Thomas, of the province of the Gyrvii,'^ to succeed him among the East Angles. Mission work in this region, however, passed soon to the Lindisfarne ^ This has not been identified. " Fobbinges" occurs (but in Essex), and the word here is a similar form in the locative plural. The resting-place of Diuma is given in Die Heil. Eng., p. 11, as Ceorlingcburh near the river Wenrisc, that is, Charlbury near the Windrush, Oxon. This river-name is doubtless Celtic and involves probably -isc «= water. 2 Gyrvii means "fen-dwellers," "Gyrwe-fenn" is glossed "marsh", and Yarrow is given by Bede as In gyrnum^ at the mouth of the Wear river, that is the marshy land at Wearmouth. For the land boundaries given in Cart, Sax, iii, p. 203, we have *' on Gyruwan fen that eft (thence directly) on Bugla fen". Cf. also the " Gyrwa river" in the Chrcjiicle of Melrose. Kluge {Gesch. der Eng. Sprache, p. 781, in Paul's Germanische Philologies would make the Gyrwe, as well as the Hwicce, Teutonic tribe-names, possibly of continental origin, although they have not been identified in the ancestral home of the Teutonic invaders. Of the conti- nental Rugii (cf. Holm-ryge in Widsith) Kluge finds a representative in SUth-ryge (now Surrey), and in East-ryge (now Eastry in Kent). For the former Bede has " in regione sudergeona ", in the Anglo-Saxon version Suthrigna-lande, both in genitive plural. In an original charter of 780 {Cart. Sax. i. 353) occurs "in regione Eastrgena" (= Eastry), Croyland {Cr Aland) is placed by Die Heil. Eng., p. w^on middan Girwanfanne, i. e. in the middle of the Girwanfenn. The same authority records from a MS. written before 1000, that Thunor, who had at the command of King THE TEUTONIC INVASION 179 clergy, one of whom, Cedd, who had founded a monastery at Lastingaeu ^ (Lastingham near Whitby), was ordained Bishop by Finan, and became most successful in the conversion of the people, especially in the City {civiias) of Ythanceastir,^ which, Egbert slain the brothers of Eormenburg {Domncva), had buried them at Eastrige under the king's " high setl" in his Healle (Hall). ^ The termination eii in Lastinga-eu, Hemt-eu, &c., is of obscure origin and meaning. In Bede's time the ending in the latter word had come to be identified in meaning with ig = island, but when the A.-Sax. version of Bede's History was made, Herut-eu became Herot-ea, that is, Hart- water, or, as it is now, Hartle/^^/. Ea (= water) and ig {= island) are both traceable to a common Gothic stem ahwa (cognate with Latin Aqim), which became Old German aha, and later a, as in Fuld-a, but an A.-Sax. form eu for the normal ea is not found. It is possible that we have in this termination a remnant of a Celtic ending, and that the first part of the combination is of similar origin, but attracted into intelligible Teutonic forms. Compare Lastinga-cu and Lastingham with Incuneningura and Concha-ceastre ; Coldingham, and Bede's Coiudi urbs — in the Chran., Coludes-biirh. It was in this " Town of Coiudi" that i^ilthryda, daughter of Anna, king of the East Angles, became a nun. The abbess at the time was ^bba, whose name survives in St. Abb's Head, near Coldingham iEdilthryda was subsequently made abbess in a region called Elge {Eliga- birigj Die Heil. Eng. 7) — now Ely. This name was given to the place, Bede says (J7. E. iv. 19), because of the abundance of eels in the marshes surrounding it. Kluge {Etymol. Worterbuch, sub Gau) thinks that the ge in the name represents the German Gati = district, and that the complete word means Eel-district. The A.-Sax. Chron. [E] variant Ileiige is late, or one might think of the British Helig = willows. ^ Ythan-ceastir has been identified, and rightly, with the Othona of the Notitia Dignitatum. It is now marked by St. Peter's Chapel in Bradwell, Essex, which is called St. Peter's ad muriim. This is referred to in Die Heil. ^«^., whereon p. 13 it is said, there resteth S**Osgith (Osyth) inCicc (Chich) near the sea at St. Peter's Minster. There is a wall there, but whether of Roman origin is not certain, but the ancient Chapel, now a barn, has Roman bricks built into it. The Blackwater, on which the place is, forms the lower reaches of the Pant. I suggested some years ago (see Plummer's Two Sax. Chron. ii. 149) that this is the ludan-byrig of the Chron. under 952 [D], where King Eadred imprisoned Wulstan, Archbishop of York. The mention immediately afterwards of Thetford suggests that the place was in East Anglia, not at Ged-burgh (Jedburgh), where the King's writ did not run. A similar form, Geotha-j occurs among the A.-Sax. mints (Grueber, A.-Sax. Coins). The form in Fl. of Wore, is futhan birig, M2 i8o BRITISH PLACE-NAMES Bede adds, is on the shore of the river Pent [Fen/a], and at Tilaburg (Tilbury). Siudhelm, who had just succeeded to the throne of East Anglia (660), was baptized also by Cedd at a place called, according to Bede (ff. E. iii. 29), " Rendlaesham, that is the dwelling of Rendili.'* Cedd died in 664, and was succeeded by his brother Ceadda, then abbot of Lastingham. Finan, Bishop of Lindisfarne, died in 661, and was succeeded in the See by Colman, another Irishman, who became the champion of the Celtic mode of keeping Easter as against the Roman. A synod was summoned to settle this question, Bede tells us (iii. 25), " in the monastery which is called Strenaeshalch, which means the bay of the watch tower \Sinus Fari\ ".^ Bede makes the Pharos of Alexandria the second of the seven wonders of the world, the Capitol being the first. So he was familiar with the word. Hild, who had formerly ruled the monastery at Heruteu (" the Isle of the Hart ", according to Bede), was the abbess of this place, and took the part of the Irish in the controversy. Bishop Cedd, who acted as interpreter, maintaining, as Bede says, a most watchful attitude on both sides. Wilfrid was the most powerful advocate of the Roman position, and through him an agreement was arrived at, but Colman would have no part in it, and went with his followers (some thirty monks) to an island in Ireland called, says Bede, Inisboufind, that is, " the island of the white cow." He after- wards found a more suitable place called Mageo, where he built a monastery which was known for centuries after as " Mayo of the Saxons '*, now Mayo. When Theodorus (like St. Paul, a native of Tarsus) became Archbishop of Canterbury in 668 he set about reforming * Hale (preserved in halec, haiigh in Scotland) = Healh, which Sweet renders hiding-place, bay, but its modern form means " low lands by a river ". Streon = treasure, gain, wealth, and it is not easy to see how it could mean "watch tower" (P/iarus). Mr. Mayhew makes the meaning ** the rock of gain ". It is possible that Bede's/ari may represent a Teutonic word, such as far, which is given by Outzen as Frisian for boundary, or it may be the same/ari we find in Lindisfarona, see p. 1 70, note. THE TEUTONIC INVASION i8i irregularities, and finding that Ceadda had not been rightly consecrated Bishop [he had been consecrated by a Bishop of Wessex, assisted by two British Bishops (Bede, H. E. iii. 28)], he consummated the act after the Roman manner. Ceadda erected a monastery at a place called, by Bede, Ad Baruae, that is, '* at the wood/' but his See was at Lyccidfelth * (Lichfield), where he died and was buried (a.d. 672). ^ Among the Civiiates Britanniae of Nennius we have Cah Luit Coyt^ that is, "the stronghold of the Grey wood." Luit coyt represents, as Dr. Bradley pointed out, the earlier Leto-cetum {lecto-cettim of Ravennas ; cf. [L] Etocetum of the Attt. Itin.) of which Hede gives the shortened form Lyccid in his Lyccidfelth. In an original charter of a.d. 803, now in the British Museum, the form is Liccidfeld, for possibly *Litcidfeld. This has become Lichfield, which has nothing to do with Lie (body or corpse), with which it is usually associated. The Let may also mean "broad" and the name would thus = English "Broadwood", but Llwyd coed occwxs in Wales to-day. The origin of the Litchfield in Hants seems to be from an early Ludeshulfe ("Church of Lydeschulv" in temp. 17 Edw. II. Hist. MSS. Report XV. App. x. p. 173). The oldest form of Coed is preserved in the Continental place-names, KaiTo^pi^ (in Portugal), Ptolemy ii. 5. 2, given by the Ravenna Geographer as Ceto-bricca. In the Brythonic districts of Scotland we find it modified, probably by Goidelic lips, into Chet or Keith. The early form, as already stated, of modern Bathgate is "Bat chet" (Holyrood Charters), and the early antecedents oi -Keith take the same form, see note, p. 86. Several modern English place-names, involving chety chute, such as Chet-wood, Chute Forrest, appear early with Cet in the compound. For example, we have in the Calendar of Charter Rolls^ "Forest of Chett" (1248), "In Bosco de Cett," 1270, where Chute is the modem form; " Chetwod," Bucks. (1290), and "The wood of Cett-red", Dorset (1243) ; cf. Lasket grove (Monmouth), Lasket in Cumberland (where probably the originally prefixed definite article has caused the dropping of the initial ^), and Glascoed ; Kesteven in Lincoln appears {^B. Mus. Charters) in twelfth century as Ket-stefena {Chetstevett : Domesday), in which the latter element is seemingly an Anglian genitive plural of Stemn or Stefn = stem, and the whole may represent " the wood of trees large enough to be used in ship-building". Another origin is, however, possible. The A.-Sax. SteOt means also " period ", and dialcctically in Steven " appointed time or place of meeting". C^// appears also in an early Charter as /mww/wj-. We have thus " the district of the people who assembled at the Cett; cf. names of Hundreds. The form Ceoftefne'm Ethelwerd's Chronicon isevidently a mistake of the Copyist. In a copy of a Westminster charter of 1067 {B. M. Charters) i82 BRITISH PLACE-NAMES Sussex and the Isle of Wight were not converted till the close of the seventh century. Wilfrid, who, when he was a Presbyter, was sent by Alchfrid, son of Oswiu, to Gaul, to be consecrated Bishop, received Episcopal Orders at Compiegne {Compendio) at the hands of Agelberct (Agelbeorht), then Bishop of Paris, but formerly of Wessex. Wilfrid was Bishop of York from 669 to 678, when, as Bede says {H.E. iv. 13), "driven from his Bishopric, and having wandered about for a long time through many places, he went to Rome." On his return he set about converting the South Saxons. The king of these people (^Edil- walch) had been recently baptized in Mercia — Wulfhere, the king, being his sponsor, from whom ^dilwalch received as a gift the Isle of Wight and the Province of the Meanuari * {Afeanwara, Anglo-Saxon version), that is, of the region of which East and West Meon and Meonstoke preserve the name. At this time, A.D. 681, Sussex, as Bede says, was ignorant of the Divine Name and of the faith, but a certain Irish monk named Dicul (probably the companion of Fursey) had a very small monastery in a place called Bosan-hamm,^ where, surrounded by woods and the sea, he, with five or six others, lived a life of meekness and poverty. Wilfrid's success was great, and King ^dilwalch we have "Penceat Wood in Battersea Manor", which seems to be the same as " se wude the hatte Poenge ", i. e. the wood called Penge be- longing to Battersea according to an almost contemporary charter of 957. {Cart. Sax. iii. 189). In a charter of 1308 we have "Penge in parochia de Badricheseye ". It is probable, therefore, that Penge is the worn equivalent of Penceat = chief wood, the ceat as in the Letcet of Lichfield being softened to che = ge : cf. Pen-cet, Lanes., p.- 86. The Westminster Charter of 693 does not include Penceat. This derivation which explains an early obscure form is both startling and suggestive. In Die Heil. Eng., p. 1 1, we have the entry "There rest St. Ceada and St. Cedde, and St. Ceatta in the minster named Luetkld near the river called Tamer (Tame) ". ^ Mean represents the district, and may have taken its name from a river ; wara — " of the people ". The Moenus, now the Main (Germany), and numerous rivers called Afam in Ireland, show that a Celtic etymon is possible. ^ Now Bosham. The diocese of Selsey, afterwards Chichester, took its beginning from this small monastery. THE TEUTONIC INVASION 183 gave to him Selaes-eu/ in the A. -Sax. Chron., Seoks igge^ which, Bede says, means Sears-island. A new authority now appears on the scene. Caedualla, " a most vigorous youth of the royal race of the Gewissi," after having raided Kent, according to the A.-Sax. Chron. in 685, invaded Sussex, slew .^dilvvalch, and ravaged the whole country. He also took possession of the Isle of Wight, and put two brothers of the King Arualdi to death, the youths having taken flight to the mainland, where they had hidden themselves at a place called Ad Lapidem, but were discovered. The abbot of a monastery called Hreutford^ (that is, "the ford of the Reed," says Bede), obtained permission to baptize them before their execution, and they were thus the last martyrs of early Britain. Caedualla went to Rome three years afterwards (689) and was baptized by Pope Sergius, and there he died, furnishing material for Welsh legend. The names ''Caedualla" and "Gewissi'' (see previous note) suggest alliances, if nothing closer, between the Welsh and the West Saxons. It is noteworthy that his predecessor, Kentwine, is recorded as having driven the Welsh {Brit-wealas) to the sea (A.-Sax. Chron. 682). The Wessex kings for some time had been extending their sway over the whole West country. In 661 {A.-Sax. Chron) Cenwalh won a battle over Wulfhere, King of Mercia, at Posentes byrg (now * Selsey is the modem name. Here Bede again understands ^m to be the same as ig =» island. He makes a similar assumption (//. E. iv. 6) in regard to Chertsey, which he says ** is called Cerotaes-ei, that is, Cerot's- island". The A.-Sax. Chron. under 964 makes it ^^ of Ceortes tge". Cerot may be a doublet of Herat = Hart, as in Hertit-eu ; or Cerot may be a personal name. The root Ke> (=horn) involved in Hart (= homed deer) is the same as that found in /brother ( = homed cattle) — the initial h in both cases representing an Indo-European k which appears in the Latin cervus and the Welsh carw ( = stag) cognate forms. Rother appears in several English place-names, e.g. Rotherhithe (the earliest form contains Rether), Rotherfield, Rutherford, &c. ' Ad Lapidem has been identified with Stoneham on the Itchcn: Hreutford is Redbridge near Southampton. i84 BRITISH PLACE-NAMES considered to be Pontesbury, south-west of Shrewsbury; but compare Poscetem tune in Hereford Domesday, in which seems to be involved a genitive plural of Cef), and drove him as far as -^Escesdun, that is, as Mr. Plummer has shown {Two Sax. Chronicles, ii, p. 23), the Berkshire downs. Ina, Caedualla*s successor, according to a contemporary note to the Chronicle, "built (? repaired) the minster at Glastonbury," which had probably ceased to be British about this time : see Note G, p. 197 . Wilfrid returned to York in 686, where he remained till 692, when he went to Leicester. From Leicester he went to Hexham in 705, where he died in 709. All England was now Christian, and up to 686 even the Picts of Galloway received Bishops from the English. The last years of the seventh century are marked by some interesting events, among them being the consecration of St. Cuthbert as Bishop of Lindisfarne. As a youth he had entered the Monastery of Mailros, of which the abbot, as Bede tells us, was then Eata, afterwards Bishop of Hexham, but Cuthbert spent many years afterwards as a recluse on the Fame Island, about nine miles from Lindisfarne, now Holy Island/ Cuthbert's teacher was the famous St. Boisil, " a priest of great virtue and of a prophetic spirit,'* whose name is still preserved in St. Boswell's-on-the-Tweed. Boisil, and another Bosil, Bishop of the Hwiccas, seem to be Celtic and not Teutonic names. In- deed, according to the Libellus (Hardy, Materials, i. 310-13), Cuthbert himself was also Irish, although bearing an A.-Saxon name.2 King Ecgfrid, who had caused Cuthbert to be conse- crated bishop, was, however, no friend to the Irish, for in the year A.D. 684, a year before Cuthbert's consecration, he had sent an ^ Mailros and Hexham have been already explained. Bede, by coupling Hexham with Lindisfarne in designating the Bishopric of Eata (" after- wards Bishop of Hagusialdensis or Lindisfarnettsis Church "), seems to indicate that Hexham was then the bishop's see. 2 Ciith is A.-Saxon for known, cleat (still preserved in xnacouth), and Beret (later Beorhf) = distinguished, now bright. Ireland was called Scotia and the people Scots in Bede's lime and much later. THE TEUTONIC INVASION 185 army into Ireland {Hiberma in the text ; " on scotta ealonde '* in the A.-Saxon version) under a leader called here Beret, but afterwards " Berctred", who, according to the A -Sax. Chron.^ was slain by the Picts in 699 (698 Bede). The Northumbrian host under Beret devastated the country of the Irish, whom Bede designates " as a harmless race and always most friendly to the nation of the Angles ". Neither churches nor monasteries were spared in this invasion. The Irish defended themselves with their weapons, but also trusted to their continuous prayers for heavenly help. The divine vengeance was not long delayed, for in the following year, as Bede tells us, Ecgfrid, having set out on an expedition to devastate the Pictish province — against the protests of his friends, Cuthbert included — and having been led by a simulated flight of the enemy into the fastnesses of the mountains, was there, together with the greater part of his army, slain (May 20, a.d. 685). Bede adds, *' Because in the previous year he had turned a deaf ear to the most reverend Father Ecgberct, enjoining him not to attack Ireland {Scotia), which was not injuring him, he had now to suffer the penalty of that sin, by not hearkening to those who were anxious to recall him from his own destruction." ^ The Picts, and some portion of the Britons in the North, * The Annals of Ulster calls the place of the battle Duin Necthan, and states that Ecgfrid was buried at lona, a somewhat surprising statement from his known enmity to the Irish. The place of the fight was Dunnichen in Forfarshire (called in a charter of William the Lion, Dunnechtyn), also marked now by Dunnichen Moss, the equivalent of Nechtanesmere, the name of the battle in Simeon of Durham, who lived in the early part of the twelfth century. The MS. is of the same century. The A.-Sax. Chron, [E] says that Ecgfrid was slain " be northan See", that is "to the north of the Sea of Forth" (Plummer, A.-S. Chron., ii, p. 32). Nennius calls the battle " the action of Linn Garan ". Nechtan was the name of several Pictish leaders, and the appellation still survives in the families of Mac- Nachten or McNaughten. Holder derives Nechtan from *Nect-a(g)nos = Neit-anus, which he connects with old Irish Necht glossed glan, that is " pure", "clean ", but Nit (also Neid) was, according to Cormac^s Glossary (ninth century), the Celtic God of War. Cf. Mog'^i\. = Slave of N^t. See Celtic Britain ', 265. i86 BRITISH PLACE-NAMES recovered after this victory their territory (which had been previously subject to Angles and Scots), and were still in posses- sion of it in Bede's time. Trumwine, who was bishop in this locality, had to retire from it with his following, ''who were in the monastery of Abbercurnig,^ situate, indeed, in the region of the Angles, but close to the strait which divides the lands of the Angles and Picts (that is, the Forth)." He took refuge at Whitby (Strenaeshalch) where he afterwards died (Bede, H. E, iv. 26). Cuthbert died in the Island of Fame in the year 687, but his body was removed afterwards to Lindisfarne, and thence, after many wanderings during the Danish invasion, to Durham where it still remains. Bede tells us that a certain Presbyter named Herberct, who led the life of a solitary "in an island of that great lake out of which the Derwent {Derveniw) river flows " (now known as St. Herbert's Isle in Derwentwater), was accustomed to visit Cuthbert, and on one occasion, coming to Carlisle (Lugubalia) to see him, obtained from him the promise that he and St. Cuth- bert should die on the same day, which came to pass on March 20 in the year mentioned.^ Bede ascribes several cures ^ This is now Abercorn. The first part ot the compound {Ader) is made up of the Celtic preposition ad (Latin = ad) and bero-s, cognate with Latin fero and its English equivalent, " I bear.'' It means a junction of streams, and contains the same element we find in Irish Commar {Ko77i-bero-s)^^€^^ Cy miner, Breton A'^fw/^r (giving its name to Quimper), meaning " a coming together of valleys, streams, or ways ". The word Inver, of similar meaning, is made up of the same element with the prefix Eni = in ; "^Eni-bero-s becoming the Irish Inber and the Welsh Ynfer. The Welsh Gofer ( = stream), Cornish Guvei (plural Goveroii)^ Breton Gouer are made up of the preposition Vo (now Welsh Gud) = (under) and the same bero-s. (See Dr. W. Stokes, Wortschatz d. Keltischen SpracheinheW). The Corn in "Abercorn " represents the name of a river, possibly the Carron, involved apparently in the Linn-garan or *' Carron- Water '* referred to in preceding note. 2 An anonymous life of Herbert speaks ot him as coming to visit Cuthbert ''from the Islands of the Western Sea" {Maris), which Mr. Plummer makes to mean " West Mere ", that is, Westmoreland. See previous note on this region on p. 137, and Geoffrey of Monmouth's THE TEUTONIC INVASION 187 of various diseases to the relics of St. Cuthbert, the accounts of some of which he himself had heard from witnesses. Among these cures he mentions that of a young man suffering from a tumour, who dwelt *' in the monastery of Dacore, so called after the river on which it was built ".^ fanciful origin of the name. Derwent was the name of several rivers, one of them giving designation to the Roman station Derventio, supposed to be Stamford Bridge, near York. Dervent-io contains, seemingly, the stem Derv-a or Darv-ac, represented in old Irish Daur (Gen. Daro) = oak, Welsh Derw, and in the Anglian Treow = tree. The form Dor, in Mapulr/(?; (also Mapuldre and Mappowdre) = Maple-tree, and in Appel- dor (Appuldre) = Appletree, making English place-names like Appledore, Appuldrecombe, Mapuldre-ham (pronounced now as Maple-durham), is an old termination, but has no linguistic connexion seemingly with "tree". Derv occurs also in Derven-tum, now Drev-ant in France, and also " In foreste Dervo", now la forH dt Der in Haute-Marne, France. See Pertz, Dipt., n. 31 (a.d. 673), p. 30. Dr. W. Stokes thinks Daruernon to be, as Ptolemy gives it, the early form of what is called in the Ant, Itin. Durovernuni (Canterbury), and ascribes the first part of it to Daru -» oak, the Old Irish Daur. The form Durovemum, besides being better established, is more easily explicable : Dur-o (in old Gaulish Duron) = citadel, as in the place-names Augusto-durum, Brivodurum, &c., and F^rw-ww = alder tree, or originally rather "Swamp", since the Welsh adjective formed from its modem equivalent, Gwern-og = swampy, boggy : tlie Breton Gnernog has the same signification. Compare Verno-dubram, now Verdouble in France. Sir John Rh^s sees in Duron a cognate form of Irish DoniSy Old Cornish Dot = door, gate ; cf. Darat already referred to. River-names may take their form from any accidental circumstances, such as the trees upon their banks, of which we have familiar instances in Ash-brook, Asche-bum (twelfth century), now Ashbourne, but the Celtic Esk (= water) may in some cases be the antecedent as already stated, compare the Welsh Onn — ashes, and the Onn-ey (= Onn-ea river), Shropshire. " Lugu-vallium " referred to in the text is the form in the Ant. Itin., and means the "wall of the God Lugus" (Dr. Stokes). The British forms were Caer Ligualid, Liwellid. Simeon of Durham calls it Luel, and applies to a certain Eadred (a. d. 867) the designation Lulisc because he had been educated in a monastery founded by Cuthbert at Luel. The God Lugus and derivations from it are referred to p. 12. ^ This small stream is called Dacre, near Penrith. The early form was possibly Dacara, which contains the suffix drd, a frequent ending of Celtic river-names, such as Isara, Iscara, Oscara, Samara, Tamara. i88 BRITISH PLACE-NAMES Miracles ascribed to the virtues or the relics ot early English saints abound in Bede and later writers. St. John of Beverley (died 721) is a noteworthy miracle- worker of whom Bede heard many wonderful things (Bk. v. 2) from one of his deacons called Berct-hun, afterwards abbot of the monastery In-dera-uuda (= Beverley), that is " in the wood of the Deri " (Deira province). St. John's favourite resort in Lent, when he was Bishop of Hexham, was a retired dwelling not far off, to which the cemetery of St. Michael the Archangel belonged, and where he performed some miracles. This was identified by Richard of Hexham {Decern Scriptores^ col. 291) with Herneshaw, that is probably Earnes-Scaga or Eagles-Copse, and now thought to be St. John's Lee near Hexham. Berct-hun informs Bede ot another miracle of St. John, performed (a.d. 686) at a nunnery in a place called Vetadun of which Heriburg was the abbess. Folcard, who wrote the life of St. John of Beverley before 1084, makes the name Beiendune, with a variant Yaiadini. Smith, in his Bede's H. JE., identifies the place, correctly, it would seem, with Watton, between Driffield and Beverley. There seems to be in this name a reminiscence of the people called Otadinoi (or Votadinoi) of Ptolemy, who places the Brigantes between them and the Selgovoi, a. name still preserved in the Solway Firth.^ Bede tells us {H. E. v. 1 2) of another wonderful occur- rence which took place in the region of the Northumbrians called Incunem'gum, where a certain Drycthelm having been raised from the dead, gave afterwards his experience of hell, purgatory, and heaven, in language which has a certain resem- blance to that of the great trilogy of Dante.'^ 1 For the Votadinoi, see back, p. 100. The name is represented in early Welsh literature as Gododin and Guotodin. The locality is defined by the fact that Bremenion (now Riechester) was in it (see Ptolemy). ^ Mr. Moberly identifies " In Cuneng-um '', which is a locative plural, with Cununga-Chester, called also Cuncha-Chester (in Simeon of Durham and in R. of Hexham), and known now as Chester-le-Street. Congavata of the Notitia Dignitatum maybe the locality, although it has been identified with Moresby. Congangios of the Not. Dig. has been identified by Holder, follow- ing Sir J. Rhys, with Ceangi, a people of North Wales mentioned by Tacitus. THE TEUTONIC INVASION 189 The great Archbishop of Canterbury, Theodore, died in 692, and was succeeded, as Bede says {H. E, v. 8), by Berctuald, abbot of a monastery at Reculver {in Racuulfe) '* which lies on the northern mouth of the river Genlada'*\ In what Mr. Plummer calls ''the oldest extant native charter of which we possess the original (dated a.d. 679)", a grant is made by Hlothari, King of Kent, to this Berctuald, and '* in civitate Recuulf '* is given as the place at which it was executed, that is the Regulbhim of the Notiiia Digniiatum} The Genlade has already been explained — Gen=.Gegn, opposite, and lada {Gelad) = passage, way; the modern name is Yenlet, which is given to two or three places in the Thames mouth ; it is the antecedent of Inlet (see Oxford Did., sub " Inlet "). The numerous lodes on the Severn, " Saxon-lode," &c., preserve the word in its sense of *' passage, and it also appears with the sense of " way " in place-names ending in -let. Cf. gelat = a junction of roads. The Teutonic tribes of Britain had not only become all Christian towards the end of the seventh century, but efforts were made by some of them to convert their kinsfolk on the Continent. The priest Ecgberct, previously mentioned, who had been leading the life of a pilgrim in Ireland, proposed that the word of God should be made known to the natives from whom the Angles and Saxons had sprung — that is, to those "who were corruptedly called by the neighbouring people of the Britons, Garmani, namely, the Fresones, Rugini, Danai, Hunni, Old Saxons, Boructuari " (Bede, H. E. v. 9).^ Ecgberct was twice ^ This word seems to contain the Celtic Gulb (Old Irish Gulba) = beak, point, and a prefixed preposition cognate with the Latin Prae, represented here by Re : compare Rak ( = ? Prae-ek) in Rac-ynys = " insula prae-jacens " (Dr. W. Stokes), and the Irish Island of Rachrie. Ru = point [it occurs in a list of places in Myvyrian Archaeology], so frequent in place-names on west coast of Scotland, is possibly of kin. ^ The Frisians, as we have seen, formed a section of the invaders 01 Britain, and their name is probably preserved in Dun-fries (Dumfries)' For the Rugini, who gave their name to the Island of Rugen, in the Baltic, see previous note on Suth ryge and East-ryge = Surrey and Eastry respectively. iQO BRITISH PLACE-NAMES warned, it is alleged, by an appearance in a vision of Boisil, former Abbot of Melrose, to desist from his enterprise until he had received instruction at the Monastery of Columba in Hii (lona), the founder of which was now called, as Bede says, " Columcelli, that is Columba of the Cella (=Cell) ", represented by the Celtic prefix " Kil " in names of churches, as in Kil- patrick, &c., and in the parts of the Continent where Irish missionaries worked by -Zell^ as in Appenzell, &c. Notwith- standing the repetition of the vision, Ecgberct persisted, and was about to sail forth when a tempest destroyed his ships and cargo, Ecgberct and his companions being, however, saved. One of these companions, named Victberct, who had spent many years in Ireland (Hibernia) as an anchorite, managed to set sail for Fresia (Friesland), whence, after preaching fruitlessly for two years to King Rathbod and his people, he returned to the contemplative life. Vilbrord and twelve companions then took up the work, and having been received favourably by Pippin, King of the Franks, who had conquered Fresia and expelled Rathbod, he was rewarded with success.^ Following his example two English Presbyters named Heuuald (one called Black Heuuald and the other White Heuuald), who had spent much time in Ireland, set out on a similar mission, but were slain by the Pagans and thrown into the Rhine (a.d. 675). Their bodies were afterwards recovered and enshrined in the church of Cologne. Vilbrord was consecrated by the Pope and made Bishop of Fresia, and Pippin gave to him a place for his cathe- dral in his celebrated castle, known in the language of the inhabitants as Viltaburg, i. e. " the City of the Wilts ", but in Gallic speech as Trajectum^ ^ Bede says nothing about the mission of Boniface (Winifrith) to the Frisians. He was martyred at Doccum in Friesland. Willibald, in his Life, says Winifrith was educated at Adescancastre, that is at Exanceastre, Exeter, which seems to have been under English dominion at the time, that is before the end of the seventh century. ^ The A.-Sax. version of Bede says Wiltenburg and iEt-Treocum, that is Utrecht. THE TEUTONIC INVASION 191 Bede*s ecclesiastical history extends to the year 731. He died in 735 in his monastery at Wearmouth {Viurae-muda) in Yarrow {Ingyruum), where he was buried ; his relics were after- wards removed to Durham, but they and the rich reliquary which contained them were carried off from the cathedral in 1 54 1. He chronicles comparatively few events during the last thirty years of his life. We have among these the record of a visit to Aldfrid, King of the Angles, paid by Adamnan,^ abbot of the monastery of Hii (lona), who, after a sojourn of some time in Northumbiia, became persuaded that the Roman method of keeping Easter was the right one. Endeavouring in vain, however, to convert to his opinion the monks of Hii and those who looked to them for guidance, he took ship to Ireland, where he was more successful, and died within a year after his return to lona. Bede mentions also that Adamnan presented to King Aldfrid an important work on the Holy Places, the matter ot which he had obtained from a Gaulish bishop named Arcuulf, who, after visiting the Holy Land, Damascus, Constantinople, Alexandria, and many islands of the Sea (Mediterranean), at length was wrecked on the western shores of Britain, whence he made his way to Adamnan's monastery at Hii. This work, from which Bede gives some extracts, has come down to us, and is important for its description of the holy sites in Jerusalem in the seventh century. Bede chronicles also the death of Wilfrid in 709 at Oundle (In Undalwni)^ and his burial in his own first monastery at Ripon {Inhrypiim) after a stormy Hfe, of which Bede gives a summary (H. E. v. 19). * Upon Adamnan see p. 49. " For "Oundle" see p. 23. It may be well to point out here that In Un-dalum may contain in its final element the locative plural of dal^ meaning shares or divisions. " In Hrypum " is also in the locative plural. Its origin is obscure. There is no Saxon word that seems near to it, but it is possible that the Norse place-name Ripar (= Crags), dative plur. Ripum, may have had an English equivalent with initial H. The Latin Rtpa = " river bank " would be far-fetched. This place is now Ripon 192 BRITISH PLACE-NAMES We have also his account {H. E. v. 21) of the conversion of Naiton (Nechtan), King of the Picts, to the Roman observance of Easter and to the Roman mode of tonsuring (a. d. 7 i o). Naiton, he tells us, wrote to the abbot of Bede's own monastery at Jarrow to obtain architects to build future churches, after the Roman custom, for his people, a request which was granted. These peaceful relations with the Picts did not continue long, for in the following year (711) Bede records {Epit.) that ** Berctfrid fought with the Picts '\ a statement repeated by the A.-S, Chron, (under 710) with the addition that the battle took place between Haefe and Caere, and that Ine, and Nun his kinsman, fought against Cerent, Wala-king (king of the Welsh). The Irish annals (Tighernagh and Ulster) record the defeat of the Picts by the Saxons in 711 at the field of Manann.^ Although Adamnan had not been successful in Romanizing his monastery, Ecgberct, who has been just referred to, had, according to Bede, convinced the monks of Hii of their insular errors and persuaded them to follow Roman customs (a.d. 716), so that, eighty years after they had sent Bishop Aedan to Northumberland, they accepted, under their abbot Dunchad,^ * Mr. Plummer {Two Sax. Chron. ii, p. 36) surmises, and I think rightly, that " between Hsefe and Caere " means " between the rivers Avon and Carron", which include the plain of Manann, corresponding with the site of the battle in Catnpo Manand given by Tighemach. Gereiit, if it is not a British name, is for Gerontius, a name derived from the Greek, but in use among the Romans. He was seemingly king of the Welsh (that is Latinized Britons) in Cornwall. The Liber Landav (pp. 108-114, ed. Evans-Rhys) says St. Teilo, when he crossed rom Armorica to visit King Gerennius on his death-bed, landed at Din-Gerein, i. e. the Castle of Geraint, now Garrans, Cornwall. Haddan and Stubbs conjecture that this is the Dinnurrin mentioned in the Canterbury Scribe's record of the Pro- fession of Kenstey the Bishop of Dinnurrin, H. and S. Councils, i, p. 674 n. H. and S. add cf. " Din-sol ", the Cornish name for St. Michaels Mount (Camb., Brit. SS. 65). 2 Dun chad : this, in its early form (as on the Ogam Stone at Glan Usk near Crickhowel) was Duno-Catus, meaning "the fortress warrior". It became Anglicized to Duncan, but its derivative Mac Dhuncadh, now pronounced MacConkey or Maconochie, preserves the original sound of the ending. THE TEUTONIC INVASION 19J the Catholic traditions. Ecgberct, after spending thirteen years among them, died in Hii on April 24, 729. Bede ends with a description of the English Church in 731. Tatwini, who had been a monk in the monastery of Bredon {BriudunY was now Archbishop of Canterbury, Inguald was Bishop of Essex, Aldberct and Hadulac were Bishops of the East Angles, Daniel and Fortheri Bishops of Wessex, Aldwini Bishop of Mercia, and Walchstod Bishop of the people west of the Severn (that is, of Hereford), Wilfrid Bishop of the Huiccii, and Cyniberct Bishop of the Lindisfari * (that is, of the people of Lindissi, or Lindsay in Lincoln). " Four bishops ruled in Northumbria : Vilfred in the Church of York, Edilwald in Lindisfarne, Acca in Hexham, and Pecthelm in Candida Casa (Whitern, now Whithorn), which latter, owing to the increase of the faithful, lately made into a bishoprick, has now as its first bishop the same " (i. e. Pecthelm). The Picts, Bede adds, were now allies of the Angles not only by treaty, but by the tie of Catholic peace and truth, and as rejoicing in belonging to the universal Church. The Irish {Sco//i) inhabiting Britain were also content with their boundaries and meditated no evil designs against the Anglian people. The Britons alone, by their opposition to the Angles and to the Catholic Church, remained unsatisfied. " This," says Bede, '* is the state of Britain in this year, the 285th after the coming of the Angles into the country." We have now to go to the Saxon Chronicle and later sources for the course of events in subsequent times. A few events, however, in the early part of the eighth century which are not ^ This is Bredon (Wore). A charter of A.D. 780 (Brit. Museum, Facs» i. 11) gives " Breodun". There are several hills of this name in England. The first part seems to represent the Welsh Bre (plural ^r^^«)= hill, appearing in many compounds, such as Penbre (giving Pembrey, Pember- ton, and possibly the Pepper in early place-names), Moelfre = bald hill. The early form was Brig. Holder {Alt Celt, Sprachschatz) says that "Brigden" in Belgium represents an early Brigo-duno-n, and is thus of the same origin as Bredon. ' For Lindisfari as distinct from Lindisfarne see note, p. 1 70. N 194 BRITISH PLACE-NAMES found in Bede are to be found in the Chronicle. We have, for instance, the death of St. Guthlac recorded under 714, and under the year 715 we have the entry that Ine (King of Wessex) and Ceolbred (King of Mercia) fought at Woddes- beorg 1 (variant Wodnesbeorg). The death of Ceolred, who was buried at Licetfelda (Lichfield), and of ^thelrsed, who was buried at Beardanig ^on Beardan iggc)^ are recorded under the following year. Under 718 a religious community at Win- burna (Wimborne)^ is referred to. The overturning of the defences of Tantun (Taunton) is mentioned under 722. The Chronicle records also under 733 the occupation of Sumertiin by ^Ethelbold, King of Mercia. The Chronicle [F] under 742 says, '* There was a great synod gathered together at Cloves- hou,* at which were present ^thelbold, King of Mercia, ^ Woddesbeorg. Mr. Plummer (op. cit. ii. 38) makes this to be Wan-' borough, presumably on Kemble's identification {jCod, Dipl.j Index) of Wodnesbeorg with "Wanborough, Hants" (? Surrey), but Wanborough, Surrey, was, in 1147, Wenebergia, and Wanborough, Wilts., was, in 1245, Wamberge, both forms excluding an original Woddes or Wodnes-beorg. Wodnesborough (Wodnesberge in thirteenth century) in Kent agrees so far as form is concerned, but it does not suit the locality, which was somewhere on the confines of Wessex and Mercia. William of Malmesbury gives Wodnesdic, now "the W'ansdyke", but see previous note, p. 148. 2 It will be observed that the Beardan-eu of Bede has here become Beardan-ige, where Beai'dan is treated as the genitive of a personal name. ^ The Win in Winbuma (that is Wimbome) is the same word as the Celtic Win in Winfrith (Dorset) that is Win (or Wen) fnit (= white stream), and means white or clear. The burna is of course English. Tantun takes its origin from the Celtic name of the river flowing through it, the Tan^ possibly for an earlier Tatn^ of which the meaning is obscure, but from the number of river-names in which it occurs it seems to have been almost a generic word for a stream, e.g., Tam-eses, Thame, Temede (now the Teme), &c. A river Thame in Durham has a Tanton upon its bank. * This is given in Bede as Clofes-hoh or hoch. The first element is seemingly a personal name in the genitive, and Hoh means promontory. Cliffe at Hoo is linguistically impossible. Cloves or Clofes could not become cliff. The early form of Closeworth (Somerset) was in 1252 Cloves- wurthe {Cal. Chart, Rolls, i. 408), and in 1270 Cloves*uude (ibid. vol. ii, p. 136). See p. 231. THE TEUTONIC INVASION 195 Cuthbert, Archbishop, and many other wise men." iEthelbald, together with Cuthred, King of Wessex, fought against the British (Walas) in 743, but Cuthred nine years later turned his arms against ^thelbald, defeating him at Beorg-ford ^ (752), and the year after we find him again engaged in a conflict with the Walas. Cuthred died the following year and was succeeded by Sigebryht, who was deprived of his throne by Cynewulf — his Witan, owing to his wicked deeds, consenting — and driven by him into Andred (that is, Andred's Wood), where he re- mained until he was slain at Pryfetes ^ floda. Cynewulf, who, we are told, had fought many battles with the British {Bre/- walas), received his death, after a long reign (a. d. 784), at the hands of Sigebryht's brother, Cyneheard, at a place called Merantun. The account in the Chronicle of Cyneheard's attack on Cynewulf is interesting (as Mr. Plummer has shown— Ilvo Sax. Chron, ii. 45), for its illustration of the arrangements of a Saxon residence. The chief building was the hall {Heali) with the other apartments around it, the whole surrounded by a rampart of earth i;weall) constituting a burg {burh). The gate {geat) was an opening in the weally but the entrance to any of the buildings within it was called a door {duru). Cyne- heard captured first the " bur " or ladies' chamber, at the door of which Cynewulf was slain, and in the fight which followed the king's supporters were all slain, " with the exception of a British hostage {Gisl, which enters into many Teutonic per- sonal names, e. g. Gislbert, Gilbert, but, according to Pedersen, was originally borrowed from the Celtic, see p. 200), who was ^ Kemble gives Berg-ford as the early charter form of Burford (Oxford). " Pryfetes floda is now Privet, Hants, where are two places of this name. The first element seems a personal name in the genitive case, and the second is the plural of Flod, meaning, according to H. Sweet, "water"; thus " Pryfetas Water" or Creek. Pryfet as a personal name is non-English, and a British original is equally hopeless, except we may see in it Pryvet, the Old Welsh plural of Fryf = worm (Strachan's Early Welsh, p. 24), Cf. Kerrig y Piyved in Suppl. Camb. Arch. Journaly April, 1909, p. 31, but see p. 210 where Prihta = a young steer, and cf. Privates Moreshed {C. Sax, ii. 422). N2 196 BRITISH PLACE-NAMES severely wounded." Cyneheard was shortly afterwards slain by Cynewulfs thanes and was buried at Ascan-mynster (Ax- minster), the king's body being entombed at Wintanceastre (Winchester). Beorhtric, who succeeded Cynewulf as King of Wessex in 784, died, after reigning sixteen years, and was buried at Werham (Wareham). In the same year (784), says the Chronicle^ ^thelbald, King of Mercia, was slain at Seccan-dun and buried at Hreopa-dun.^ Under 761 the Chronicle [E] records the slaying of Oswiu by Moll, King of Northumbria, at .^dwines-clif,'^ and under 762 the death of Frithewald, Bishop of Witern (who had been consecrated at Ceastrum, locative plural = York) and the con- secration of Pyhtwin, his successor, at ^Ifetee.^ In 733 the Mercians and the men of Kent fought at Ottan- ford (Otford, Kent), and in 777 Cynewulf and Offa fought around Benesingtun (that is Benson, Oxford), of which OfFa succeeded in getting possession. Seletun (now Silton, York) is mentioned under 779 as where Beorn Alderman was burnt, and Soccabyrig under 780 as the place of consecration of Higbald to Lindisfarna ee. A synod was held at Aclea in 782 and another at Cealc-hythe in 785.* * Seccandun is now Seckington, Warwickshire, and Hreopa-dun is now Repton. "^ ^dwines-clif is in Simeon of Durham ''Eldunum near Melrose", now Eildon (Eildon hills). Ail is Celtic for " rock ", from an early [P]alek. It occurs in Ail-Clud = Clyde Rock, now Dumbarton, and in this Eildon = "Ail-dun", rock fortress. Cognate forms in other languages zxo. fels (from [P]allos) in German = rock ; French /alazse = cliff. J^'a/l in Norse (" Fell" in North of England) = hill. ^ v^lfetee. This seems to contain the British El-vet, and ee another form of ea — river or for ige - island, but the first part may be A.-Sax. Elfetu (Norse ^^) = Swan, and the whole " Swan-river". An Elvet forms part of Durham, but the name occurs in this, and in an earlier form, Elmet, elsewhere. See p. 1 70. * Soccabyrig is now possibly Sockbridge on the Eamont {ea = river and ntont for mufid w^ mouth, see note, p. 269. Mr. Plummer suggests Sockburn on the Tees. In Lindisfarna ee the last syllable seems to represent island. THE TEUTONIC INVASION 197 NOTE G. Glastonbury. There is no place in these islands which can boast of a longer continuance of religious life than Glastonbury. Although it does not figure among the Civitaies Briianni'ae appended to the Historia Briitonum of Nennius, its antiquity is unquestionable. In William of Malmesbury's (twelfth century) De Antiquitate Glastoniensis Ecclesiae^ we have an attempt to trace the origin of Christianity there to the time of Pope Eleutherus (Pope from 177-92). This writer knows of the legend in Bede {H, E. i. 4; V. 24), who, following the entry in the Liber Pontifi- calls (drawn up according to Duchesne in 530), relates how the British King Lucius asked Pope Eleutherus for mis- sionaries to convert his people, and that the Pope sent them. William of Malmesbury gives the names of these missionaries as Phaganus and Deruvianus, using as his authority the forged Charter of St. Patrick hereafter referred to. The question of King Lucius and the mission of Eleutherus has been dealt with fully by Zimmer {Nennius Vitidicatus, pp. 140 et seq.), and by Haddan and Stubbs, Councils^ vol. i, pp. 25, 26), and may be relegated to the region of pure fable. The Apostles St. Philip and Joseph of Arimathea are also brought in by WilHam of Malmesbury, the latter being made the leader of an earlier mission. The forged Charter, ascribed to Ireland's Apostle, gives besides the names of the assumed founders, Fagan and Deruvian, a list of their early successors. This list is an instance of a pious fraud unsupported by knowledge. Some of the assumed personal names therein are designations of The Norse ey\% to be excluded, as the form here was in use before the Northmen came. There are numerous Acleys, but this was presumably in the south. It may be Ockley in Surrey. Cealchythe can be traced through Chalchithe (1465), Chellsaye (sixteenth century) to Chelsea. (See B. M. Charters and Rolls and cf. the form in the Taxaiio of Pope Nicholas, wherein its position establishes its identity beyond doubt.) Challock in Kent was, in the ninth century, Cealf-loca or calf-enclosure, and Mr. Plummer's mention of Mr. Karslake's view is, therefore, superfluous. The term loke still survives in East Anglia. A further instance of this use of loca is given in Hist. AfSS. Report XV, App. x. 148, " Recourse with cattle to the Pynloke" (i.e. Pound-enclosure). 198 BRITISH PLACE-NAMES localities (e.g. WelU'as, Breden, and Swelwes, where we have most probably a reminiscence of Wells, Bredon, and the Swelle common in English place-names, from Swelg = swallow, indi- cating the disappearance of a stream underground). Some of the other names, as Adelwolred, are Anglo-Saxon compounds and impossible in this region at the date. Htn-Locrifius (thus in William of Malmesbury : Gale's MS., thirteenth century, now in Trin. Coll. Camb., reads htn loernius) seems borrowed from Geoffrey of Monmouth's Locrinus — formed from the Welsh word for England Lloegr — with hin^ i.e. hen (= old) prefixed. Glastonbury is called in this Charter Ynsgitrin (in Gale's MS. Insgytrin) that is, hiis (= island) and gytrin for wytrin (i.e. Vitrinus — " of glass "), which quasi translation the composer of the name uses, assuming that the Glas in Glastonbury repre- sented the English word Glass. The Welsh Triads calls also the place '* Bangor Wydrin in the island of Afallen (Avallon) ". St. Patrick's Charter is ascribed to 430 {Cart, Sax. i, p. i), and it did not strike the scribe as anything extraordinary that an English place-name should be found there at so early a date. Professor Freeman, if I mistake not, also made Glastonbury a pure English name. The names Fagan and Deruvian (variants Dy/an, Devinianus) are late forms — the latter probably corrupt — of possible Irish names. Arniilf and Ogmar, which also appear in the document as Irish brethren, seem to be Scandi- navian forms. There is a mention in the Annales Cambriae of Glastonbury, and it is alluded to in the Scholiast to Fiech's Hymn (see p. 130, Irish Liber Hymnoruvi, H. Bradshaw Soc. and vol. i, p. 103), where a Sen Pairaic (i.e. Old Patrick) is associated with the place. This is not the Irish Apostle, but, according to the editor of the Annals of Ulster (Rolls Series), p. 1 7,the Founder of Armagh, who was followed there by S. Benignus (Beonna) ; also associated with Glastonbury by William of Malmesbury who makes him Abbot there immediately after St. Patrick. There are numbers of Charters of Grants to this place, many of doubtful authenticity, in the Cart. Sax. and in William of Malmesbury's work already cited. Some of these throw some light on place-names in the locality. Among the charters of Grants by English kings, bishops, &c., we have one ascribed, but with doubtful foundation, to a.d. 680 (Cart. Sax. i, p. 74), in which a piece of land called Lantocal and a marsh called Ferramere (now Meare) are alleged to have been given to the Abbot Hemgisl by the Bishop of Winchester. THE TEUTONIC INVASION 199 Ferramere is clearly an English form and suggests English occupation of the district, which could not have been earlier than about a. d. 658. Another charter ascribed to 681 {Cari. Sax, i, p. 96: Ord. Survey facsimile) records a grant by Baldred King of Mercia of land on the top of the hill called Pengerd (Pennard, Somerset) to the same Hemgisl (Hemgils) for the completion {supplemen- tum) of the Church [at Glastonbury] of the Blessed Mary and St. Patrick. The bounds of the land — which are given in Anglo-Saxon — are all English, except the name of the river Brue {Briuu), implying that the West Saxons had then settled in the region about the Church and had given names to localities there. To CuUanbyrg^ to Siangedel/e (stone quarry), and io Stcenenanbrycge (stone-bridge) occur among them. Another charter ascribed to 682 is a grant by Centwm King of the Saxons {Cart. Sax. i, p. 97) to the same Hemgisl (Hamegils) of twenty-three measures of land {Manstones) in a " place close to the celebrated wood called Cantuc-uudu (the Quantocks), bounded on the south by the Tan (the Tone river) . . . with other boundaries as far as the ford called Weala- /ord" {Wea/a, gen. pi. of PF^jM = Welshman). A HcBgstaldes- Cumh (see p. 116) occurs among the boundaries, and also in Cart. Sax. ii, p. 77. We now come to a charter (Cart. Sax. i, p. 165) ascribed to Ini, King of Wessex, who is stated in the Chronicle to have rebuilt the monastery there (before 680). The charter which is ascribed to 702 (for 705) grants a piece of land on both sides of the Doulting stream to the Abbot of Glastonbury. Among the boundaries is a Crich hulk (now Creech-hill, see note, p. 23). This is followed {Cart. Sax. i, p. 166, Ord. Survey facsimile) by a grant of land near the Tan (Tone) at a place called Pouelt (now Pawlet) and of land on the Doulting stream — here called Dulut-ing. Pou seems to be the British form of the Latin Pagus, universally so used in Brittany, and the -elt seems to be the same as the -ulut in Duluting, representing possibly the remains of some Celtic land designation. A further grant of Ini appears in a charter {Cart. Sax, i, p. 177) ascribed to 663 (Dr. Birch suggests for 693 or 708) in which land in and about the mount called Brente (now Brent Knoll) is given to the same Abbot Hemgisl (here Hengisl), The bounds are the Sabrina (Severn estuary), the Axa (Axe), 206 BRITISH PLACE-NAMES the Termi'c, and the Stger. The name Ternuc occurs in another Charter, and it is doubtless this Termic, The " Tarnock Common *' (Tornoch in 1533) near Biddenham on Ordn. Survey Map pre- serves the name. Another river called Torric (which seems to be the same) appears in the boundaries. The Axa {Exa in a charter of A.D. 944, Cart. Sax. ii, p. 553) is probably for Isca, which appears in Charters also as Esce, as in the early forms of the Berks and Derby -^j-^fords {/Esc/or d). Hengisl here may be the correct form and the equivalent of the Cornish Hen-gwystl = Old hostage. The names of the notables Tangisil and the dux Bamgisil of the Kent Charter ascribed to a.d. 605 (Cart. Sax. i, p. 10) have the same termination and there are numerous similar names in the Liber Vitae. Gul ( = hostage) is also used as a prefix in names, e.g. Gislhere Gislbeorht (= Gilbert), &c. The similarity of Hengisl and Hengwystl suggests that the abbot may have been British, and it is significant in this con- nexion that a grant by the King of Devon is referred to in William of Malmesbury (De Antiq. EccL Glast.) of the land called Ineswyirtn to Worgret Abbot of " the old Church," and that Abbot Bregored followed him after one interval. Worgret is a manifest Celtic name with many parallels in Breton nomen- clature, e. g. Uuoruuoret, see Cartulaire de Redon, passim. The possible identification is not strengthened, however, by Pedersen's opinion (F. Gramm. § 87) that the Old High German Gisal and A.- Sax. Gisl are borrowed from the Celtic, for this borrowing must have taken place in pre-historic times. A grant by Fortere, Bishop [of Sherborne] to the Abbot [of Glastonbury] is given in Dr. Birch's Cart. Sax. i, p. 189, in which the river Aesce (the Axe referred to above) and a "port" called Bledenithe (i.e. *Bleden-hithe ; hithe ■=. ^ox^ and also the Church of the Blessed Martin Confessor are mentioned. This charter is ascribed to 712. The A. -Sax. bJed^ meaning '* branch ", may be used in Bledtm\\\Q for an inlet of the Axe. Dr. Birch makes the Bledone granted to Winchester (?) 975 {flart. Sax. iii, p. 647) to be Bleadon on the Axe ; Bleden-ithe is more likely to be the modern Bladney standing on rocky ground and communicating with the Axe seemingly by one of the " Rhines ". The association of St. Martin as well as St. Patrick with Glastonbury district is a note of antiquity in itself. Another charter of Ini (here Ina) granting lands to Glaston- bury, ascribed to 725, appears in Cart. Sax. i. 207. It is very corrupt and the place-names are much distorted, possibly by THE TEUTONIC INVASION 2or successive copyists. Budecalech (now Butley, which may be for Budecan-legh, which, though English in form, may be Celtic) appears among them in probably its correct form. Soweie is for Sow-ige (cf the variant of the next charter), i. e. ? Su-ige or Sug-ige = "Sow island", now "Zoy". William of Malmesbury gives a legend to account for the name stating, on the authority of ancient British history, that a certain Glasteing (the Welsh eponymus of Glastonbury) following his sow with eight legs through Escebtiorne (? Iveythorn in Street) and Wells by the Sugewege (Sow's way) found her at an apple-tree near the old church, and gave to the island hence the name of Avallonia. Lantocai also occurs for the Lantocal previously mentioned involving the Celtic Z^;/ = " enclosure " = finally "church". The next charter (ibid. 210) is ascribed to the same date and furnishes an early form of the river Parret (i.e. Pedredt-sirem), and gives the river Kan (Cary) as a boundary. Pouell has become here Poholte^ but it is Pouholt in a subsequent charter {Cart. Sax. i, p. 213) which gives us also Cengisl (variant Hemgisl, but it is Coengils in a later charter) as the Abbot in 729. The latter charter is alleged to have been signed at Pen-crik, hardly the Pencric (now Penkridge) in Staffordshire, for the writ of the King of Wessex would not avail at this time so far north. It was pro- bably one of the Somerset " Creeches ". A certain Hilla makes a grant in 744 with the permission of ^thelbald (called by a shortened form Elbald in the same charter, Cart. Sax. i, p. 242) of Balteresberghe (Balsboro in Speed's Map of 1610, now Baltonesboro) and Scobbanwirth to Glastonbury. Among the boundaries, which are very corrupt, occur Ni'met and the river Olan or Olitane, which seems to have given its name to the Leden ford occurring in the same charter, Litana (= broad). The Nimet here is suggestive. It still occurs in the names of several localities in Devon (Cart. Sax. iii, p. 624, gives the early form), and we find it in the Nemeto-bn'ga in Spain, and the Nemetacum Atrebatum in Belgium (recorded in the Itinerary of Antonine) and in the early name of Nanterre — Nemeto-dHron. Its meaning is given by W. %\.o\Q%{Wortschatz d. kelt.Spracheinheit, p. 19 2) as "noble", " holy ", who cites the Irish ; in bratha nemed (" the judgements of the nobles ") : it occurs in the Old Welsh personal names Gornivety Eidnivet, &c., and in the early Breton Catnimet^ ludnimet, &c. Compare the old Cornish Carn-nivei = "heaven-bow", that is, rainbow. There was probably a religious shrine at the spot 202 BRITISH PLACE-NAMES before the foundation of Christian Glastonbury. The place may be the St. Michael's Tor, close to Glastonbury (and near enough to Baltonesboro to be a boundary) which elsewhere I have connected with the Nen-ihur (i. e. (?) NemeHor) of St. Patrick's birthplace. There is a Nempnet north of the Mendips, but this is out of the question. The charter mentions further that the Glastonbury monks (who are called there Hen- gisling, i.e. followers of Hengisl) had chosen a wooden Church at Glastonbury as a resting-place for the sarcophagus of Hengisl. William of Malmesbury describes all the churches here {De Antiq. Glasi.). A letter of Pope Leo III (a.d. 798) confirming the monastery of Glastonbury to Kenelm, King of Wessex, is given in the Cart. Sax. i, p. 392. There is a charter of King Edbert ascribed to 801 (Cart. Sax. i, p. 418) granting a portion of land at Bodecaleie (the Butley previously mentioned) to a certain Eadgils, and among the boundaries the following names occur : — Bregedes-were (i.e. (?) the weir of a man named Breged : could this be the Abbot Bregored? A Brudenewere appears in Sock Dennis in 1280, B. M. Chart, and Rolls), Lang-Echer (= Acre) and Ucktng Echer (cf. Ucingford, B. M. Chart^, Swelle, Lane her pille (? = modern Lancherly : for Lancher, see p. 298) and Stan-pille {pille is here, from its association with "stone", most probably a boundary pillar and not /z7/ = stream : cf. *'on thone Stenenan Stapol " = stone pillar, Cart. Sax. i. 47), Olde-lake^ Hoctan- (h)yth, the Old yo (not, I think, eoh = poetical for yew, but the Fee river, a possible dialectical form of ea, eo, cf. old- Lake), sel/(h)ith cf. Shuthelve Hill, Axebridge), Weles(h)ith and "from the {h)ythe*\ &c. This charter was confided to the church of Glastonbury. A charter ascribed to 842 {Cart, Sax. ii, p. 13) mentions some localities near the Brue river, among others the river Alum (upon which is Dicheford : there is a Ditcheate, i. e. Ditch-gate, two miles north of the Brue) which has its confluence {Gemido — Gemythe) with the former river. The name occurs in another grant of lands near Glastonbury {Cart. Sax. ii. 471) along with a curious place-name (near Batcombe) "a/ Austien on than put " i. e. puteus = well. Here is another possible claimant for ''Augustine's Oak" or rather ''Austin's Well", but the Wessex boundary was not so far west in a. d. 604. A Worcester Charter {Cart. Sax. i. 109) has cet Austin^ of which we may still have an echo in " Austins rick " of the Ord. Map. The Penryn Awstin THE TEUTONIC INVASION 203 in Kernwy of the Welsh Triads (ii. 183) seems to contain the same word, and if it was meant to designate "Aust Cliff'* {Penrhyn = Promontory) on the Severn mouth it shows that Kernwy (Cornwall) extended further north than is assumed. Alum may be a form of the common Celtic river-name, Alauna, now generally ^/«^ (Aluuinn — as in Cart. Sax. i. 227 — Warwickshire) as in Alnwick, Alne-Chester (now Alcester, Warw.). Nimei appears in the same charter, attracted into the intelligible form Nymede, as if "new meadow". The charter is alleged to have been written at Andredes-eme^ which suggests the Andred in Andreds-wood — a more correct ending is given below — but William of Malmesbury explains it as St. Andrew's (Chapel). The charter dated 904 {(2ari. Sax. ii. 263) is alleged to have replaced a lost charter showing the lands which Athelstan DuXy son of Etheredi (for ^thelredi) had given to Glastonbury when he became a monk there. It contains the boundaries of Wrington, Somerset, but the names are very corrupt. The Wring is a river which gives its name to Wrington and Wringford. Preosteseluoey with which the boundaries begin, is not intelligible : it may be for Priests leak, now Priestleigh, Somerset. The remaining names are not instructive. King Edmund granted to St. Dunstan, Abbot of Glastonbury, certain lands at Christian-Mal-ford, Wilts. ( = Crisiemal = " Christs Cross "). *' On tha olde oden missenne " occurs in the boundaries, which is an obscure sentence ; Oden may be for the Welsh Odyn — kiln, although it is given as Anglo-Saxon for "threshing-floor" (see Oden Cole, p. 157). The last term suggests a borrowing from the Latin messis (= a reaping) either by the British or English. The Welsh and Cornish Mesen (= acorns) can hardly apply. In Cart. Sax. ii, p. 544, there is a grant of privilege to Glastonbury by King Edmund in a.d. 944, and also a grant of land at Wootton, Somerset, by King Edmund in 946, to the thane Athelnod {Cart. Sax. ii. 578) which land is liable to charges on behalf of Glastonbury. The boundaries are in corrupt and late English. Ashbury (in which is Ashdown, Berks.) is granted in another charter to Glastonbury by Edric — see Cart. Sax. ii, p. 594. The Dean and Chapter of Wells were formerly in possession of a charter, now lost, of King Eadred, granting in a.d. 950 land at Pucklechurch {Pucelancyrca) to Glastonbury. A copy exists 204 BRITISH PLACE-NAMES in the Record Office from which {CarL Sax. iii, p. 44) I extract the more important boundaries : — Stanford^ Hennamere {henna = fowl, i. e. perhaps here " wild-fowl mere ") — Gosamere (Gos = '' goose " f gives an irregular plural Ges, but here we find a regular genitive plural Gosa), Stanora (^rfl: = bank); Hlypgei {hliep = leap, geal = gate, " ? a gate that can be leaped one way only." The place-name Lipyate is not uncommon). Further boundaries are : — Havima (gen. pi. of hamm = meadow or enclosure) on Kynges Rode {= to King's Cross or possibly "clearing "). Byd in the boundaries seems a river-name, but it may be for hyht (masc. = a bend, bight : neut. = a dwell- ing). It appears here in Bydincel {wincel = corner), in Bydewyllon (Byd- well), and in "///^ hamm on Byd". "Of Loddera-wyllon " {Loddere = beggar : Wiella = well). On Loddra-welhm occurs in a grant to Pershore, Cart. Sax. iii, p. 589 ; cf. Beggar es-ihome in charter granting land at Bledon, Somerset, Cart. Sax. iii, p. 648. Hreod-mor ivyllon (" Reed-moor springs "), " Where the wyriwala runs to Fearfi beorhg''' (i.e. Fern hill). Wyrtwala appears in the dictionaries as "root", i.e. as the equivalent of German Wurzel, but this cannot be its significance here or in the many other cases where it appears as a boundary. Weall-wala ( = (?) " Wall foundation ") occurs only in poetry, but the Wala here may mean " foundation *' and Wyrt-wala thus = " root- foundation", or the ground base perhaps of old structures. The compound, however, may mean simply vegetable {wyrt) garden. See Crawford Charters (Napier and Stevenson), p. 68, on a parallel use of Wyrtrum. On tha eggce = " on to the edge," a common element in place-names. On thone graf = " on to that one grove ". On haran stan = " on to the hoar (gray) rock ". On hric weg = " on to the ridge way ". On oxna healas = " on to the oxen shelters " [healh, pi. healas = " corner, shelter "). On Heges- taldes setl = to the residence of Haegestald (see Hexham, p. 174), and lang Fromes along the Frome river. On Cweorn Cleo/u=.^' to the quern-cliff", i. e. the cliff from whence querns or stone hand- mills were obtained. On Befer pyttas = " to the Beaveri.holes '*. On ^schurhg = " to Ashbury ". On Sinder ford = (?) a ford where " slag " is used for stepping-stones. The obscurity of the name Puckle Church has been referred to elsewhere (see p. 275). To the explanation there advanced may be added the hypothesis that the latter element (flyrca^ may not be for Church at all, but a modified form of the British Cruc (= tumulus), in which case the term Pucel would be THE TEUTONIC INVASION 205 a natural determinative, the whole meaning possibly " fairy mound"; cf. Pukelonde, B. M. Ch. and Rolls. The names in this charter have been analysed at some length because they afford fair instances of Saxon name-giving. In Cart, Sax. iii, p. 89, we 4iave a confirmation by King Edwith (var. Eadwig) of the grant in 956 of a vineyard at PaihenehergHe to Glastonbury. Dr. Birch places this, I know not upon what authority, at Mere in Wilts., and says the vine- yard is noticed in Domesday. I should be inclined to look for it closer to Glastonbury, and if the name Mere is implied any- where it might more fitly be Meare close to the monastery and on the Ea {= river) mentioned in the boundaries, that is the Brue. This is confirmed by the fact (see below) that Pathene- berga is placed " in insults " in King Edgar's grant of privilege to the abbey. Pathene seems to be for Pathena, a genitive plural of a word which is not seemingly English ; cf. Peatanige now Patney, Wilts., Cart. Sax. iii, p. 354. Pamborough, the modern surviving form, shows that the place was some four miles north-west of Glastonbury. Oslakes-leag, in the boundary, may still survive, but I have not found it on the map. In Cart. Sax. iii, p. 452, is another Glastonbury charter in which the Cari stream is mentioned, and in another charter of Edgar dated 968 {Cart. Sax. iii, p. 493) a grant is made to Glastonbury of land on the Dorset Stour, in which Dewlish {JDeveliscK) is mentioned (see p. 150). King Edgar granted again certain privileges to Glastonbury in 971 according to Cart. Sax. iii, p. 574, in which certain places in the islands (in insulis, i. e. in the low insulated lands near the Abbey) are mentioned, among them ^^ Bekeria^ which is QdiiXtdi parva Fbernia {\\\.\.\e; Ireland)". It is somewhat striking that we have in this word, not only a link connecting Glaston- bury with Ireland, but in the term itself a form but slightly modified of the Irish Bec-eriu ( = Little-Erin : erinn being the genitive of Eriu). The term occurs also in the Felire ofOengus (Henry Bradshaw Soc, p. 118), with the same significance, as the designation of a little island off the Wexford coast which is now known as Beggary-island. The Glastonbury form is pre- served still in the name of a village not far from the Abbey, called Bickery in the new Ordnance Survey Map, replacing a misnomer in the older maps, where, doubtless by the influence of some local antiquary, the surveyors set it down wrongly as Bickworthy. William of Malmesbury says that St. Bridget 2o6 BRITISH PLACE-NAMES spent some time on this island, and that certain relics of her and of St. Indracht were preserved in the church. Ind-recht = *' the Just " ; the modern form in Ireland is Hanratty. Patheneberga also occurs here as in i?tsuHs, and also Adredes-eya for the form given above as Amdredesevie. The eya termination is the usual Latinized form of the Anglo-Saxon ige = island. King Edgar grants privileges here to Glastonbury at the instance of St. Dunstan Archbishop of Canterbury, Oswald Archbishop of York, Brictellinus (Beorhtholm) Bishop of Wells, and others, for the soul of his father whose body rests there, and for the souls of his ancestors. Among the signatures — at London it is said in one of the MSS. — appear the names of Keneth [Kinadius rex Alhanie) King of Scotland, and Mascusius, " High Admiral " or "Wicing Chief" {Archtpirata), the latter name probably a distorted form of the name Magnus common among the Scandinavians. ^thelflaed(Car/. Sax.'iu. 600), second wife of King Edmund I, she gives for the souls of Edmund and of King ^Edgar and for her own soul, land at Domerham to Glastonbury. A charter of King Edgar {Cart. Sax. iii. 608), ascribed to 973, exchanges land at Brancmtnstre, elsewhere Braniic minstre involving a Celtic personal name (? St. Branoc, diminutive of St. Brandan, associated with Bristol and with Braunton, Devon), for land at Hamme. Branuc Minstre recalls '' Branok hyalf hiwisce" (Thorp, Dipl. p. 107), belonging to Glastonbury. In this charter we have as bounds Wernan ford and Wernan- strem and the river Wern ; also a Stanwei (stone-way, perhaps a Roman road) and the river Peret (Parret). The names are corrupt, showing the results possibly of frequent copying. There are several other charters given in William of Malmesbury's De Aniiq. Glasion. Some of these contain seemingly early forms of place-names in the locality, e. g. Ceddren (Cheddar, where a grant of Elleanhoro is situate); Godenice. (now God- ney) which is said to be so called from the Holy Trinity ; Mar/enesi'e (from a chapel of St. Martin on the island) ; Bade- cumhejuxta Montem Munidop (Mendip, Mened-ipp in 1290 {Cal. Ch. RollSy ii, p. 363), see note, p. 24); Bedul, Branuc Dunhead (= Downhead, where probably Branuc Minstre may have been, cf. Branch Huish (Huish = Hiwisc z=zfamilia and also a land measure: see above), N.E. of Downhead) ; Lemucerine (probably a corruption of some Celtic name) ; Eatumberg = ? Emborough, Eteneberga in 1270 CaL Ch. Rolls, ii. 137 (cf. ^c?/^^ = Jutes THE TEUTONIC INVASION 207 and Eo/en = g\3.nt. See Asser, Li/e of Alfred^ Stevenson); Occemund (? Oakhampton) ; Occenefeld (? Okenhede of B. M. Ch. and Rolls) ; Peasuanere (Peasuc = ? a personal name) ; Lidencge (cf. Lidenford above) ; leholt, AttahoU, Lupwite (? Lopen); Z^^ p. 234) makes the original signification to have been " wooded hill ". The Old High German Fergunea is the Henynia silva. The stem is /«?r^« = oak, A.-Sax. Furk, replaced by Danish Fir in Fir tree. ' Connected by Torp with Anglo -Sax. hage ( = hedge, wood, enclosure), a suffix == hate. Cf. Hagu -staid. THE TEUTONIC INVASION 213 Syla = wallowing-places, cf. Sole in place-names. Wyla = ? gen. pi. " of Caves ", cf. Wygel, gloss, to Cavea in Diefenbach's Glossarium, cf. Wyle-cop, Shrewsbury. Grundwealle, dat. = foundation. Pintreow = pine-tree, from Lat. Pinus, Wilegum, dat. pi. = baskets. Landgemacena, gen. pi. = of neighbouring lands. Dimhus and Dimhof = hiding or dark place, cf. Dimchurch. Warena, gen. pi. = citizens : Old Norse -ven\ as in R{im- verjar = Romans. This is the Varit in Chatuarii, &c., and is the final element in Caniwarena. Snaeda = parts. Hsethen = heathen. Healle, gen. = stone : also = palace. Hweras and Crocc = pots. Moldum, dat. pi. = sands. Cocca, gen. pi. = "chickens, cf. cei Scyte Cocceajnd cock shot", (variant) " Cock shade" in place-names. The Dialect Diet, makes the latter to mean open spaces in woods where wood- cocks may be shot. Edwinde = whirlpools. Watelum, dat. pi. = tiles. II. Genuine Charters. Dr. Sweet printed in his Old English Texts (pp. 426 et seq.) the vernacular words occurring in what he considered to be genuine Old English Charters. As some additional genuine charters have come to light since 1885, when Dr. Sweet's book was published, the following selection of place-names is taken from authentic charters up to a. d. 800, thus excluding any of Scandinavian influence. In an original charter of 679 now in British Museum we have the following names in Kent : — Tenid (the oldest English form of Thanet) : Uuestan-ae (i. e. Westan-ea = river running from the west) : Recuulf (= Reculver Regulbium, see p. 37) in Sturia (latinized form of Stur, or Stur-ea, now Sturry). In Essex charter of 692-3 now in British Museum we have: monastery called Beddan-haam (Beddan is here the weak genitive of a man's name Bedda: — the haam is properly the inner bend of the knee, and possibly came like the Celtic Camb- us to be applied to land on river-bends and then to pieces of land generally. The East Frisian Hamm meant a piece of 2 14 BRITISH PLACE-NAMES land drained by dykes. It is not the same word as Ham = home, house); Ricinga-haam (the Ric-inga here is a genitive plural of a family name). In Deccan-haam (now Dagenham) we have the genitive of a personal name Decca. In A«gen-lab-es haam we have a personal name made up of Anga = only, unique, and lah, later laf = left, remaining : thus " only surviving son ". Cf. Angen-geot-ing of the Northumbrian genealogies in Sweet, which may mean " the descendant of the only begotten ", cf. be-gi^tan, &c., the vulgar use of "get'* substantive; but Dr. Sweet makes Geoting = son of Geot. The " field in the wood called Uuidmundes felt "means the field of Widmund. T or p{Wor/sc/ia/z) traces field " to a stem meaning open, cf. Ladn Fal-am = openly". Among the boundaries are Writola-burna (in which Writola seems a genidve plural, but it may be a mangled Celtic word). Centinces treow (= Centing's tree), Hanc-hemstede (the hanc here seems to represent the Teutonic stem kan/i = to hang ; A.-Sax. later forms I/on, heng ; cf. Hanger, Ongar, &c. Hem- sted is home-stead). The Tamisa is also given as a boundary. The grant is endorsed ^' to Bercingon ", i. e. the people of Berc = Barking. An original charter of 704 now in B. M. gives the name of Tuican hom (Twickenham? = Twih-ean-ham, i. e. the dwelling between the rivers Tamisa and Frocesburna, which are men- tioned. Twicen = a place where two roads meet, is not involved). In a letter of 705 (original in British Museum) Breguntford, now Brentford, Middlesex, is mentioned (see p. 25). A Kent charter of 700 or 7 15 mentions Liming-ae (see below) and Pleghelmestun (i. e. the enclosure of Pleghelm). The boundaries are Bere-weg (i. e. way along which the produce of the land is brought home, Hei-weg = Hay-way), Meguuines paeth (i. e. Megwines path), and Siret leg (i. e. the wood, after- wards meadow, on the Street = the Roman-road. A portion of land called Rumining Seta (see Seto in Epin. Gloss. Rumin-ing, now Romney Marsh, may involve " Romans "). The grant is made to the Church of B. V. M. at Lyminge. The Dean and Chapter of Chichester have the original charter of the year 725, in which the bounds ofHugabeorg — some unreadable — are given, e. g. on theodweg (theod = people, public- way) (to) the East-end of Lavingtunes dices, i. e. Lavington dyke to Frecce-hlince (Frecce, = bold, is obscure, but Mine = slope, ridge) to Halignesse beorge (i. e. to the hill of Sanctity) of tham garan (i.e. from the Gore, '*a piece of land," cf. Kensington THE TEUTONIC INVASION 215 Gore) to tham Byrgelsun (= to the burying-places) to Billinga hyrig (to the burg {Dat. byrig) of the Billings) — where Scealces burna and Bollan-ea come together (scealcs = servants as in Marah-scalc = marshall) ; — and Isenan aewelm (iron-spring, ea- welm = river source) and Saengelwicos and other names which are not intelligible. An original charter in Brit. Museum, dated 732, grants land near Limin-ae (i. e. the river Limen which appears in Liminge and Lympne, near Hythe, and is ultimately traceable to ihtporius Lemanis, of the Itm. of Antonine), The names Sand tun (Sand- town) and Hudan fleot (Huda's fleet or estuary) occur in the copy published in Cart. Sax. i. 215, but as Dr. Sweet rejects these, they are not probably of the date of the charter. Another original charter, dated 740, grants fishing at the mouth of the Limin-aea river, and **a portion of land in which is situated the oratory of St. Martin, with the dwellings of the fisher- men, and another portion near the Marsh called Biscopes uuic (i. e. Bishop's dwelling) as far as the wood called Ripp (unknown and probably not an English name), and to the bounds of Sussex, as formerly possessed by Romanus, priest of the Church of the most blessed Virgin Mary, which [qtwd (sic)) is in Liminia -eae (Lyminge) ''. The oratory of St. Martin at Lyminge-ae, and the name of the presbyter, Romanus, suggest early continental connexions, while two of the signatories Dun-uualhi Pincerni (gen.) and Dunuuallan (gen.) seem to point to the Celtic names Duno-vali and Duno- vellaunus, cf. Cassi-vellaunus, Catuuallon, &c. A Kent charter of 770 (in B. M.) mentions the monastery of Recuulfi (attracted into this form on account of the common name Rac-uulf see above), Perhamstede (where Per = modern Pear) in the region er (?) of the Caestri-uuara (gen. pi. ="of the Castlemen"), where is the place called Heahhaam (heah = high, Higham). The Hehham (= Highham) ot the original charter of 774 (in B. M.) is possibly the place here mentioned. The boundaries there given are Msedham (i.e. Meadham), Ac-leag(= Oakley), Wseter lea (Waterley), Colling (= Cooling), Eohinga burh (Eoh = steed, Eohinga gen. pi. of patronymic : Burh = burg), Mersc- tun (Marsh-ton), Mercfleot = " boundary river ". All these places are near Rochester. An original charter of 774 (in B. M.) mentions the Mersc- uuare (i. e. the marsh-people) of Romney,=here Romenal, among 2i6 BRITISH PLACE-NAMES whom Hlid (Lydd, Kent ? = litus^ shore) is placed. Denge marsh is also mentioned. The earliest Mercian charter extant seems to be of the year 736. It is now in the B. M. and deals with lands in the ** province of Husmerae ", that is in the region of the Stour (Stur) above Stourbridge. It gives the ancient Celtic names of two forests, Cynibre and Moerheb, which still survive in Morfe and Kenver forests ; see p. 67. Brochyl (i. e. Badger-hill) is mentioned as being " in silva moreb ". The synod of Clofeshos (Cloveshou in later charters) is men- tioned in a Mercian charter (now in possession of Dean and Chapter of Canterbury) of the date 742. An original charter in B. M. of a. d. 755-7 contains a grant by the King of Mercia to Eanberht, of land near the wood called Toccan-sceaga (i.e. Toccas Shaw or wood) not far from the hillock {tumulus) called Reada-beorg (i. e. Redhill). A Tyccaea abbot signs the charter, probably the Tocca here mentioned. The locality has not been identified. A Mercian charter of 759 (now in B. M.) grants land in Worcester at Onnan-forda, bounded on the south by Uuisleag (Wisley), on the west by Rind-burna (Rind, which is the OHO. form, is probably a dialectic word here for Hrith ( = Hri- ther = Rother = cattle), on the north by Meosgelegeo(=: mossy- tracts : Dr. Sweet), and on the east by Onnan duun. The river Oney giving its name to Onibury, Shropshire, is perhaps too far west. Onn is the Welsh for Ash-trees : cf, English " ash" in stream-names. In a Mercian Charter of 767 we have a grant of land in Middtesex lying between Gumeninga hergae (that is the harrow or pagan temple of the Gumenings = ? Harrow-on-the-hill) and Lidding brook, in exchange for land in Ciltinne (? Chilterne hills or Chiltern All Saints) in a place called Wichama. A Worcester Charter of 770 (in hands of Dean and Chapter of Worcester) mentions the Huiccii (see p. 164), the river Saluuerp (Salwarpe), Cymedes halh, Huitan stan (= white [pillar] stone), and Readan-Solo. Saluuerp seems to mean, like Sal-gewearp, " Salt-heap," i. e. the river that ran by the Salt, packed ready for transport at the Salt wic or Salt market — now Droitwich — was thus called from the fact that river-names frequently take their designations from such incidental factors.^ Redan solo = ^ The halh in Cymedeshalh is given by Dr. Sweet as = ^^rt/// = corner. It is evidently the same word which we find in the North in the forms THE TEUTONIC INVASION 217 Red cattle-ponds. Sole occurs with this meaning in Kentish place-names, see Syla, pi., in Napier Glossaries. A Kent Charter of 778 (in B. M.) mentions the following names: — Brom-geheg (Broom-meadow, Dr. Sweet), and a marsh called Scaga (= Shaw), a meadow called Hreod-ham {Hreod = reed), a place east of the bounds of the Clif wara (gen. pi. = "the people of the Cliff,*' which place, if it existed now, would be called, like Clif wara near Cheddar, Clewer) **and to the south of Tucincg-naes".^ To the west of the bounds of the Culinga (the name of this family is still preserved in Cooling Castle) is Mearcfleot.'' Another, a Wessex, Charter of 778 gives the names of Bedewind (Bedwin, Wilts. : which looks like a British name involving Celtic Vind = white), Cymenes denu ( = Cymen's Valley), ad Peadan-Stigele (= Peada's style), Tatan Edisc (= Tata's deer park). Several words then follow involving gcBt, geatf get = English gate, e. g. Rames dene geat (= Ram's personal name, ? -den-gate), Holh-ryge-get (= Hollow-ridge- gate), (fee' In Puttan Ealh (= in Putta's (heathen) temple, of. Alh-mund = Temple Guardian), to Bulcan pytte (Bulca's pit : pit from Latin Puteus). Baldwines Healh (see Halh above), " into the Antient Monuments in a place called cet tham Holm- stypbum (e= at the Holly Stumps) to Wadbeorge (Woad-hill), Healc, halec, Halech, Hauch, and Hough. It is always there associated with rivers, and Jamieson in his Scot. Diet, defines it " low-lying flat ground, properly on the border of a river". In the Records of Kinloss, p. 112, we have " the third part of the Halech of Dundurcus and the third part of the fishery of the same halech ". The old name of Whitby given by Bede (see p. 180) bears out this signification. ^ Tucincg, a patronymic, and naes = nose or point. This name shows that nces, like wicing and Tkrop, existed in English place-names before the coming of the Scandinavians. ^ This is ihe mercfleot of the A. D. 774 charter cited above, which gives also Colling for Culinga. An original charter of 808, in B. M., also mentions Culingas, with its bounds Cyninges-tun-tih ( = ? Kingstown paddock = Tih = Teah = Teag, now Tye and Tey in place-names, e.g. Marks Tey, Essex, which involves, as Mr. J. H. Round shows, Fettdal England^ 464, the name of the village of Marck in the Pas-de-Calais, borne by Adelolf de Merk, the owner of Marks Tey district in Domesday 3. Cf. Tigan, now Tey river, Essex, Cart. Sax. iii. 602. The 808 Charter mentions Genlad, Biorn heardes lond (i. e. Bernard's land), and Thomdun. ' These gates were probably entrances to the enclosed Edisc or park. 2i8 BRITISH PLACE-NAMES thence to the spring which is called Forsca burna (= Frogs' brook)/' 1 A Gloucestershire Charter of 779 (in B. M.) gives the follow- ing names: — Sulmonnes burg (= Suhl monnes-burg = Plough- man's burg), the stream called Theodningc (=? a patronymic from (?) Theodnoth), the river Uuenrisc (now Windrush: the last element seems Celtic -isc = water, and the first element Gwen, wen, fem. adj. = white ; cf. the rivers Winster where Ster is probably the early form of S/ur), Uuithig-ford (= Withy- ford) a valley called Turca-denu (that is the den on the river Turca ; cf. the Welsh Twrch = boar, which is also the name of a river : see Mabinogion story of Twrch TrwytK), Slohtran- ford (= ford of the Sloe-trees, now Slaughter-ford). This Charter was signed at loratla-forda (Hartleford, Glouc, seems to involve Heorot ( = hart, stag) : possibly a genitive plural of a form Heorotlu, a species of stag). An early grant (785) to Westminster Abbey, published by the Ordnance Commission in Photozinc facsimile, gives the name of the island on which the minster was built, Torneia (i. e. Thorn-ige = thorn-island), which is called in the Charter " a terrible place " (perhaps in the sense of holy-dread). The grant is of land at ^Idenham (Aldenham, Herts.) of which the bounds are given : among them : — Colen-ea (that is the Colne river), the middle of the street (i. e. Watling Street, which crossed the Thames at Thorney island), Tidulfes treow (= Tidulf 's tree; cf. Elstree), Haesel-hyrst gate (Hasel-wood), Lusebyrge. The names are manifestly late and most of them unintelligible. This is probably why Dr. Sweet does not include them in his early charters. In Cart. Sax. i, pp. 342, and 344, two charters which Dr. Birch says are almost contemporary (a. d. 785) are pub- lished, in which loccham (Ickham) and Per-hamstede (men- tioned above) are placed in the wood called Andred (seep. 143). Dens for feeding swine in this wood are mentioned ; e. g. Dunwal-ing daenn (where we may have a reminiscence of the Dun-walh(i) before mentioned) Suithhelming daenn : a stream called Heorat burna ( = Hartbourne), a Snad ( = Snaed, a piece of a wood), and fishing in the Pusting uuerae (Wer = Weir). Hroching (now Rocking) also occurs, and Limen wero weald ^ The Peada and Putta in this Charter are not normal English personal names and imply a foreign connexion. THE TEUTONIC INVASION 219 (= ? Limen (Liminge), men's wood) and burh waro uuald (i. e. the Burgmen's wood: Sealterna Steallas (the salt houses' steads). The names are late forms and Dr. Sweet rightly does not include these in his early charters. The charters pp. 350 and 360 Cart. Sax. i (taken from the ancient Chartulary of St. Denis near Paris) are manifestly late forms, notwithstanding the seal of Offa affixed to the latter. The forms Chichestra, Hastingas, and Successa (for Sussex) amply prove this. Pevenesel may preserve the ancient name of Pevensey, and Lunden uuic may be Sandwich. An original Charter of 788 (in B. M.) gives the name of a region called Eastrgena (now Eastry ; the rgena is supposed to be the genitive plural of the continental tribe the Rugii, who also appear in the early name of Surrey ; see p. 178). Duningc- land and Celchyth (Chelsea, see p. 157) are also mentioned. Another original Charter of 793 (in Canterbury Cathedral Library) grants several places in Middlesex : — on Linga Haese (= Hayes, Middlesex), Geddingas (Yedding, Middlesex), Froces burna and Tuican ham (see above, for the repairs of the Church of the Holy Saviour, Canterbury). An original Charter (in B. M.) of 793-6 grants lands in the province of the Huuiccii at Uuestburg near the river Aben (Westbury, Glouc, near the Avon). The Charter is recorded as having been written in the celebrated place called Clobeshoas. Another original Charter in B. M. dated 799 grants land at Ciornincge (Charing, Kent : another village called Charing gave its name to Charing Cross. " Ch^re reine " is a survival of days of ignorance), Seleberhtes Cert (perhaps Great Chart). Cert is a mysterious word. It seems to mean " wood ", but its origin is probably not Teutonic : Kent was latinized early and the word (cf. numerous wics in Kent) may be from that language : cf. Carecium = reed-land or sedge-land. Bryning lond (called Biringland also), Humbing lond on Bioraham (? = Barham, Kent) aet Burnan (at bourne). An original Charter of 798 (in B. M.) mentions several places in Kent : Hremping-uuic, called also Hafing seota (see Seota above), south of the Limin-ea (i. e. the river Limin at Liminge) : also Bobing-seata. In the contemporary endorsement seota is made of the weak declension genitive singular seotan. 220 BRITISH PLACE-NAMES III. Ancient List of English Territorial Names. There is a list of English territorial names (given in Cart. Sax, vol. i, pp. 414-16) copied, presumably from an ancient form, in the tenth or eleventh century. That it is a copy of an earlier document seems clear from the distorted forms of some of the names, which are difficult, if not impossible, to identify. I append the list with such notes as occur to me. Myrcna landes (i. e. of the land of the Mercians Wocen Saetna^ ...... Westerna '^ (Porcensetene in later Latin form = ? Worcensetene) Pec Saetna^ . Elmed Saetna* Lindes farona^ Suth Gyrwa^ North Gyrwa [East Wixna"^ West Wixna] Names in brackets from another list, C. S. i. 416. Spalda® ....... HIDES 30,000 7,000 7,000 1,200 600 7,000 600 600 300 600 600 1 Satna is a genitive plural and is equivalent to ivara, gen. pi. of wer= man. Cf. Ceaster setna {Cart. Sax. ii. 217) and Caestruuara {Cart. Sax. i. 282). It means thus in the connexion here " of the people or settlers of". Wocen is possibly an error of the copyist for Wrocen = the Wrekin district, for the people of which Wreocen Setna actually occurs {Cart. Sax, iii. 355). ^ This seems an error of the copyist. The alternative name Porcensetene is clearly a mistake for Worcen setna, a duplication possibly of the previous name. ^ These are the settlers of the peak district. * Elmed Ssetna would seem to refer to the Elmet district in Yorkshire mentioned by Bede. We have an Elme-setena (Elmley Lovell) in Cart. Sax. i. 502, but this, as the list seems to proceed to the north-east, is too far to the West. An Ylmesceton in Essex (?) appears in Cart. Sax. iii. 601. ' Lindes farona (i. e. Lindsay-farona) seems the gen, pi. of Faru (f.), one of the meanings of which is '' troop ". * Gyrwa, Gyrwe fenn is glossed Marsh Gyrwa = of the fens. ' Wixna is gen. pi., a contraction of Wicsetna, that is, people of the market. There was a Wixenabroc in Worcester {Cod. Dip. 57o)> now Whitsunbrook. * Spalda is evidently for Spalding district. THE TEUTONIC INVASION 221 HIDES Wigesta^ 900 Herefinna'* 1,200 Sweord ora ' . , 300 Gifla[Eyflain C. 6*. i. 416]* . 300 Hicca [Wicca in C. 5. 1.416] = 300 Wiht gara [-gora in C. 5. i. 416] « 600 Nox gaga'' 5,000 Oht gaga 2,000 Total = 66,ioo That is, says the oldest MS. which does not give East and West Wixna with their 900 hides, sixty-six thousand one hundred hides. It is clear from this that the copyist of the oldest MS. had omitted East and West Wixna although his total number of hides is only correct with their addition. The list then proceeds : — HIDES Hwinca* ........ 7,000 Ciltern Saetna* 4,000 Hendrica [the second list gives hides 3,000] . . 3,500 ^ Wigesta might be an error for Wigerna (Worcester), but there was a Wicgestan near Burton-on-Trent {jCod. Dip. 1298). " A probable mistake for Hereford. ' There was a Sueordleag in Dorset (orig. Charter Cart, Sax, ii. 34), and a Sweordestaii in Gloucester {Cart. Sax. ii. 174). * Gifla may represent (i) the district around Biggleswade where the name of the river Ivel or Givel still survives in North-ill and South-ill parishes; or (2) it may mean the district around Ilchester to which the river Givel = Ivel gave its name. ' Hicca, with its variant reading Wicca, could not mean, owing to the small number of hides, the province of the Wiccii. ® This name may have nothing to do with Wight. ' Nox and Oht gaga are mexplicable in these forms. * There is no light on Hwinca. * Ciltern, or the Chiltern hill district. The remaining names down to Widerigga are unintelligible. Widerigga may be the Withering-Sett, Suffolk. East and West Willa seem to refer to the district near the Wiley river (Wilts.). The name in the Charters is Wilig, Uuielea. The remaining names explain themselves. 222 BRITISH PLACE-NAMES Unecunga Aro Saetna Faerpinga [Fearfinga in list C. S. i. 416], which says the early list, is among the Middle Angles Bilmiga Widerigga ...... East Willa WestWilla East Engle (of the East Angles) . East Sexena (i. e. of the East Saxons) . Cantwarena (of the inhabitants of Kent) ^ Suth Sexena (of the South Saxons) West Sexena (of West Saxons) . HIDES 1,200 600 300 600 600 600 600 30,000 7,000 15,000 7,000 100,000 Total = 244,100 Amounting in all, says the list, to two hundred and forty-two thousand seven hundred hides, falling short of the correct total by 1,400 hides, or, if we estimate Hendrica at 3,000, by 900 hides. 1 This term involves the Warena in Prof. Napier's Glosses, see p. 213. It is different from the genitive plural {ward) of iver = man. III. THE COMING OF THE NORTHMEN Norse Influence on Topographical Nomenclature AND Language. The entry in the Chronicle under 787 is remarkable for the announcement that in Beorhtric's days, who, as we have seen, succeeded Cynewulf as King of Wessex in 784, the first ships of Danish men reached the land of the Angles. The actual year is between 784 and that of his death — 800. The '^ three " ships, which came from Heretha-land ^ put into some place on the Dorset coast, according to Eihelwer^s Chronicle^ whither, the same authority tells us, the King's Reeve, Beaduheard by name, rode from Dorchester, believing that they were traders and not foes, and was there slain. From this date forward until the eleventh century the British Isles were subject continuously to the raids of these pirates, and large portions of the land were from time to time occupied by them. Their first appearance in the Irish Seas was, according to the Irish Annals, in the year 795, but they had most probably been making southerly voyages from a much earlier date. The ^ Heretha-land, "strictly Hor^aland on the Hardangr-fjord in Norway," says Mr. Plummer {Two Sax. Chronicles, ii. 58), "the country of the HorSar or Hawrds (Chamdes, Harudes)." In the Irish Annals it appears as Irruaith, and is there a general term for Norway. It is to be noted, however, that the **Hreth" Goths are placed in Jutland or in S. Sweden {Iceland, Diet., sub Goti and ReiS Gotar), and hence the reason possibly why these invaders are referred to as Danes by the Chronicle. The Annals of St. Neot (compiled towards the end of the twelfth century) say that the ships put into the '* island called Portland". 224 BRITISH PLACE-NAMES causes of these expeditions are not easy to discover. Dr. Vig- fusson was probably right in ascribing the main cause, as far as Norway was concerned, to over-population.^ The Danes, although they are thus named in the entry of the A .-S. Chron- icle referred to above, and are confused throughout with Norse- men proper, were not among the first invaders. In the Irish Annals, the Norsemen and Danes are distinguished as "Fin- Gall " and " Dubh-Gall ", or fair and dark foreigner respec- tively. This distinction of fair and dark hair was probably of a racial character, for ethnologists tell us that the skull of the average modern Dane is not by any means of Aryan type. Possibly in pursuit of the whale, the early Norse sailors may have been led to the soutn and west, and made as their first land-fall the Faroe or Shetland Islands. In an age when the compass and the log were unknown, these early navigators -would normally hug the shore, but once they had reached the Shetlands, these, and the Orkneys, would become a base for further explorations. It was from such a base, as we know, that the Norsemen harried the coasts and islands of Scotland and Ireland, and it was here that their sovereignty over Sodor and Man began, which survived until 1470.^ Nowhere else in these islands do we find such undisputed evidence of the presence of Norsemen as in the modern place-names of Shetland and Orkney. The previous occupants, as we infer from the existing Pictish '' Brughs " and Barrows, were Picts, and the early names 1 " The Northmen," said Dr. Vigfnsson, *' were cribbed up in a narrow, overpeopled strip of land : a powerful race of men, with vast pent-up strength lacking all outlet, a great human steam-boiler without the requisite safety-valve : till in the Wiking Age the great exodus to west and south opened a new field of action to them" {Corp. Poet. Boreale, i. 426). ^ What is known in Norwegian history as " the flight of the Jarls", who emigrated to these islands in 872, was the beginning of the Kingdom of the Sodor (Sudr-eyjar = south islands, or " Sodor islands ", applied to the Orkneys [_Orig. Iceland, i, p. 253]) and Man. Man and the South isles were made subject by treaty to Scotland in 1266, but Shetland and the Orkneys were not really handed over till 1470. THE COMING OF THE NORTHMEN 225 of these groups of islands^ probably preserve a relic of their original language. Irish missionaries, probably from Hy (lona), had laboured in these islands from about the end of the sixth century until the arrival of the Norsemen some 200 years later, and indications of their presence there are still preserved by a few Ogam inscriptions, place-names, and dedications of churches. Irish hermits were found also in Iceland when the Norsemen discovered it about a. d. 850.* An account of the voyages of the Norsemen prior to the time of Alfred the Great (a. d. 849-900) was furnished to that king by the two Scandinavian travellers Wulfstan and Ohthere, and their account was added to Alfred's Description of Europe, No mention of Britain occurs in this account, but it is clear from the whole narrative that Norse seamen went far afield, and that they were acquainted with the White Sea as well as the Baltic. Though they were pagans, and remained so until the eleventh century, they must have been brought into contact, in their eastern wanderings, with Gothic Christianity — that is, Greek and not Latin Christianity. This is evidenced from the word ^ The name of Shetland in the sagas is Hjallt-land. Hjallt and Shialt (the latter preserved in the existing designation of "Shelty" for a Shet- lander or a pony from the islands) are evidently the same word, and are paralleled by the Cymric and Goidelic names Hafrett and Sabem for the river Severn. The Picts, who in historic times spoke a language more akin to Welsh than Gaelic, would follow the former in using in place of Shialt the form Ilialt, The earliest Scandinavian maps {Nordenskjold Atlas) give Hjialt-land. An analogous instance is the parallel between the Hjal- pands-ey of the sagas and tlie Shapinsay of our maps. As Orkn in Norse means a Seal, Orkn-eyjar would seem a natural designation for these islands, but the term Ore in Orcades goes back to classical times, long before a Northman had put his foot upon them, and its meaning must be sought in the language of the earliest inhabitants. If these were Celts ore, oirenin (glossed Porcellus) has the meaning of Pig in their language, and is cognate with the Latin Pore-us, and the A.-Sax. Fearh = farrow. Compare Latin orca = Grampus, but the word is probably non- Aryan. 2 The Landndma B6c (that is "Land Settlements Book") shows that there was a considerable Irish element among the early Norse settlers in Iceland. Place-names like Ir-oe = Irishman's river, occur also in this work P 226 BRITISH PLACE-NAMES ** Papa ", which they applied to the Irish priests they encountered in the islands, and also from the term ''Kirkja" with which they were conversant as the designation of church.^ The place- 1 The Goths on the south of the Danube had become Christianized in the fourth century. They had, as their historians tell us, come originally from the Baltic coasts, and probably some connexion with their Norse kinsmen was maintained in later times. There is at any rate no likelier source for the knowledge by the Norsemen of such terms as Papa and Kirkja. Professor Kluge thinks the former is from the vocative Papa of the Greek Pdpas'. the latter is from a Greek adjective formed from K^rios = Lord. The names of some islands in the Shetland and Orkney group perpetuate the former appellation, e.g. Papa Isle (Shetland), Papa Stour (Shetland), Papa Stronsay Isle (Orkney), Papa Westray (Orkney). The Flateyjarbok, ii. 417-18, gives us some of the saga-forms Pap-ey hin litla = " Priest-island the little ", and Pap-ey hin Meiri = " Greater Priest island ", which are now represented by the two Pabeys of the Hebrides. The name occurs also in Iceland, e.g. Papafjordr. It is possible that we have in the Domesday "Popeselle" or "Popesalle" an echo of the same word. It was in Kent, where one would expect from its Jutish occupiers, a closer connexion with early Scandinavian traditions and language. I ventured to point out some years ago in the Athenceum that there was distinct evidence in the place-names of Kent of a Jutish element therein. I have seen no reason since to change my opinion. The most important evidence is in the early name of Bapchild. Haddan and Stubbs, Councils, iii, p. 244, cite from the Anglo-Saxon Chronicle the account of the " Great Council, at the place which is called Baccan-Celde " held about 694, long before the Norse invasion. The element Celde in this name seems to be the Scandinavian Kelda = a well, of which Outzen {Friesisches Glossarium^ p. 159) gives the Jutish form Kild with the same meaning. The only parallel for the term is the Keld of the North ( = well), of Scandinavian origin, for undoubtedly Norse is the term St. Kilda, which applies not to a holy person of this name, but to the known Holy Wells on the island ; cf. Halikeld, Yorks., Cal. Chart Polls, ii. 240. It is a strange thing to find in Kent alone — apart from the North — a word closely allied to a Scandinavian form, but the strangeness is explained by the assumption of a Jutish origin. The Baccan in Baccan-Celde is seemingly the genitive of a shortened personal name Bacca. Celd occurs also else- where in Kent, as in an original charter of 858 (British Museum Facsimiles') in the form Hwyte Celdan ( = white well), it being coupled there with another well Wassingwell, as the site of a water-mill. There is also a " Honey Child " in Romney Marsh (^Honi Child 1227), where we have also a " Snargate " which receives its explanation (see p- 232) as a Scandinavian THE COMING OF THE NORTHMEN 227 names in Iceland referred to in the Landndma Boc, of which we have a thirteenth-century MS. (the materials therein going back to the ninth or tenth century), and those in the various sagas, enable us to show, by comparison with the existing place- names in Shetland and Orkney, how completely the Scandi- navian tongue had displaced the original language of these islands.^ The historical statement that Norwegian was the term. "The Note " also seems to bear indication of a similar origin. It is noteworthy also that the Scandinavian Gil (see p. 331) is used in Kent. ^ The topographical nomenclature of Shetland and Orkney is almost exclusively Scandinavian. As instances we may take the names containing such Norse forms as Voe (elsewhere -way = Vdgr = creek), Skaw {Skagi = a low headland as opposed to Hdj'de = a high headland), ness ; wick (- Vik, a bay),yfr//5 {=fjor6r), Holm (applied to uninhabited islands), Sound (Sund), ey {= ey, island, plural eyjar)^ exemplified in Lang-ey, Flat-ey, Sand-ey, Ha-ey (that is " high-island ", now Hoy), Sker-ey (that is "rock-island''), Hun-ey (for Hafn-ey, i.e. Haven island), Urfas-ey (probably for Orfirisey, from Orfiri •= ebbing). Stad-r (stead) which takes the form of ster (plural) and sta, Bustadar (Home steads = Bowster and Bister, as in Isbister, &c.), ty^r (Danish Ore^ Swedish Or ; - air, ayre, a gravelly bank or point running out into the sea, as in Helsingor, Shake- speare's Elsinore), thing (also ting = a meeting of the people, as in Sands-ting, Ting-wall, where wall = Vollr = field, compare Dingwall in Sutherland, Cheshire, &c.) : BH (also By = habitation) : CartJr (= garth *= enclosure) : Scet-f^ mountain pastures: Dal-r, a dale : Skdli, a shieling or temporary dwelling in the pastures : Skaill = Skal, a bowl, a hollow: Skdgr (also Skew and Shaw ■= wood — brushwood, Danish Skov) : Klettr^ Danish klint (Clett, Cleat, an isolated rock) : Gnupr (Noup, or Nupe or Nip = a peak, as in Gaitnip = Goat-peak) : H6p (^a sheltered haven, or spot) : Hdll^ a hill : Borg (Burg), Hus ( = house, e. g. Bon-us ? = *Boon- hus = Prayer-house, the equivalent of English Bede-hus, Bed-em. Bettws in Wales is the English Bedhus, which occurs also in Scotland), Lax ( = Salmon) in Laxforth : Hamarr^ a rock, as in Hamarr-Gnipe in Iceland = Cragpeak. We should not expect to find on these wind-swept islands words meaning wood, such as Lund, Holt, &c., and they are absent. The parallel names in Iceland show how thoroughly the Norse has prevailed in these islands. These place-names in their turn have become surnames, as in Foubister, Isbister, Inkster, Cursitter. We find also in these islands other personal names which have a distinctively Scandinavian character, e. g. Swan-son, Man-son, Simon-son, Hender-son, OUa-son, which preserve respectively the well-known Norse personal names, Svein, Magnus, Sigmundr, Heinrekr, P 2 228 BRITISH PLACE-NAMES vernacular there down to the seventeenth century is further confirmed by Brand, who found in Hara in 1701 people who knew only Norse. Caithness and Sutherland — the parts of Scotland lying nearest to Orkney — furnish also indubitable evidence of Norse occupation,' but it is on the chart of the west coast of Scotland and the adjacent islands that the Scandi- navian sea-rovers have left their most permanent mark. There is hardly a headland, or a half-sunken rock, or the scrap of an island, or a bay or inlet, in these western waters which had not been charted— although their mode of doing so is unknown to us — by these daring navigators, and handed down to later map- makers.^ It is to the same navigators, too, we owe the names and Olaf. Many such names have become shortened, as in Gunn (from Gunn-bjom, or Gunn-laugr). Oman is from Amundr, which also appears in the western islands, in a Celtic dress, as MacAmond and McCammond, where quite a number of similar Norse appellations are perpetuated, as McCorquodale for MacThor-Ketel, McCorkell for MacThorkell — the Th in Gaelic being silent, MacAlden for Mac-Halfdene ; Half-dan (now Haldane = Half-Dane). There are pitfalls for the philologist in the place-names of this region. What could one make, for instance, of Pomona, a designation of the main- land of Orkney ? Although it occurs early it seems the application to this island, by an antiquary, of the Pomona associated by Solinus with Thule, and it remains there to disturb future investigators. ^ In Caithness and Sutherland the following small selection of place- names supply ample evidence of Norse influence there. Cannis-by, Seater- hill, Wick, Ny-bster (for New Bustadar, plural), Ul-bster, Lybster, Rester, Bilster, Bilbster, Thuster, Achilbster, Scrabster, Halla-dale, Armadale, Torrisdale, Erri-boll {bol = English Botl^ Bold = house), Fres-gill, Lax- ford, &c., &c. There appear many other Norse forms, such as Clett = rock ; Geo ( = Gjd, a cleft) which becomes Goe, and in the Hebrides is Celticized to Geodha, pronounced Geyo [Giau in Manx], and Gil = stream. These names tend to confirm the assertion of the Saga that *'01af Feilan (= Irish Faol-an = the Wolf), Thorstan, and Sigurd con- quered Katness, Sudrland, Ross, Moray, and more than half Scotland", Landndma B6c, ii. 14. ' The Skeirs and Skerries, the Ha-Skeir (High-reef) and Deas-Skeir {Des - shaped like a hay-rick, or Irish Deas = south), all significant as things to be avoided by the navigator, stud the western seas. There, too, we find Shellay with its shell beach, and Sanda, with its sand, Pabbay, THE COMING OF THE NORTHMEN 229 of Ireland and of three of its four provinces, together with the appellations of many of the isles and bays on its eastern coast and on that of the Isle of Man.^ These expeditions were not made, however, for the purpose of acquiring a knowledge of which marked the lone monastery of world -renouncing Irish monks, Vallay, which promised Vellir, or fields of rich grass behind its rocky shore — every name significant if we could interpret them. In the Western Isles, Skye, Rona, Raasay, Uist, Lewis, Harris, &c, we have fuller illustrations : e.g. Ose Point {/iss = ridge; Oss, mouth, also occurs as Ois, as in Ois Gillbay), Scara-vay {Skarar = edge : Fa^= bay), Causa-Mul {A/u/t probably borrowed from the earlier Celtic inhabitants = jutting crag), Sletta-val (= level field), Ha-skeir, Ha-clett (« High reef and rock), Geo (cleft). Valla-rip (rt^r — crag), Greanascore {Grain = forked valley = Grain in Northumberland, York, and Lincolnshire place-names, and skor = cleft), Husa-bost, Steis-buist, Kirka-bost {dosi is the shortened form of Btisfad-r), Staflfa island {stafar = posts of a building), Stack {stakkr = stock), Bola- val and Bola bratt (= Bull field and Bull hill: Eng. Brant = Brattr), Vater-galt {vatr = vatre = water, and Gait = Hog = Hogsback, ridge, but here the Celtic Gait = cliff may be involved ; and countless Sgeirs ( = sker = reef). ^ The Ster in Ul-(for Uladh-)ster, Lein-(for Leigin)ster, Mun-(for Mumh)ster, is Norse — so is the termination of the islands, Ireland's eye — as opposed to Angles-ey — and Lamb-ay. Howth Hill, Wicklow (earlier Wiking-low), Wexford, Water-(that is weather)ford, Leixlip, Carling-ford, Strang-ford, &c., all show Norse impress. The Isle of Man is full of Norse appellations, including such Icelandic forms as a = river, Brekkr ■= brink or slope, Dalr, Ey, Eyrr, Fjoi©-r, GartJr, Gil, Hammar, Haugr (Tumulus, How), H611, Setr, VaC (a ford), Vik, Vollr, and the surnames there show also a large Norse influence; e.g. Casement (from McAsmundr), Caskell and Gaskell for McA skill, short for McAsketill; Cottar for McOttar, Corlett for McThor-ljotr, Crenill for McRagnild, &c. In Manx personal names beginning with c {k) we have the last letter of Mac (= Son), the initial i^ becoming aspirated in oblique cases and finally vanishing. The Welsh Map, Ap, and P initial represents the same change, e.g. MapRichard, ApRichard, and Prichard. The spread of Norse influence in Southern Britain is also marked by Scandinavian place-names, which will be dealt with in due course. Great caution is needed here, however, as many English names are similar to the Norse. The extent to which our Modem English vocabulary is indebted to imported Norse words exhibits further the pervasive character of Scandi- navian influence. A short list of familiar words traced to a Norse ancestry is appended (see Note I). 230 BRITISH PLACE-NAMES geography or for peaceful commerce. They were uniformly piratical, and the main objects of their pursuit were the costly shrines, the richly adorned books, and other precious objects of religious art belonging to the churches and monasteries. The Wars of the Gaedhill and the Gaill (i.e. the Irish and the foreigners), and the Annals of Ulster and other Irish Annals published under the direction of the Master of the Rolls, are filled with records of the pillaging of religious houses in Ireland, and the A.-S. Chron. contains records of similar atrocities in England.^ The early expeditions were of the nature of summer raids, and the person engaged in them was distinguished by the name of Sumar-lithi (Somer led), that is, Summer-sailor. The Norsemen soon came, however, to make permanent settlements in all parts of Britain, succeeding eventually in asserting their sovereignty in the early years of the eleventh century (1016) over all England. NOTE I A SHORT List of Words (abstracted chiefly from Torp's Wortschatz, Gottingen, 1909) borrowed from the Scan- dinavian Invaders of Great Britain. Norse.4»^-r is involved in place-names of the western islands, and a cognate form, Norse eng (grass-land), is the origin of our northern eng and tng = meadow. Bahki (= bank) is found in northern place-names. Beck (of which the English form Bece gives " beach ", and " batch ", ** bach " = Brook : the early form of the Chester Sandbach was Sandbec); Billow; Bleak; Bloom; Bole (of a tree); Bunker (heaps, cf Bunker-hill or Golf bunkers) ; Booth ; Boon (A.- Sax. form is Ben) are from the Norse, as are also — * Mr. Plummer shows {Two Sax. Chronicles^ ii. 156) how it had become necessary, owing to the ravages of the Scandinavian pirates, to fortify the monasteries. Peterborough was, as he says, one of the earliest instances, which from being a Home-stead (Medes-hamslede) and exposed to attack, was made a Burh, by being encircled with a wall (an earthwork) with defences. It was henceforth Vtitx-borough. . THE COMING OF THE NORTHMEN 231 Cake; Car, in northern place-names, Norse Kjarr (= low bushes); Carl (churl); Cast (to throw) ; Kid; Crook. Down (feaihers). Fir (tree) ; Fors = foss ( = waterfall). Gate (in the sense of way); Gable \- Game! (= old, in personal and place-names) ; Girth ; Gills (of a fish) ; Guest ; Gil (= narrow valley in northern place-names and in Kent). Hope (=a bay and river valley); Holm (a sea- or river- island in the north. It is also English); Haug-r* in northern place-names. 111. Kindle. Law; Lceck-r (= brook: in northern place-names, as Leek: the cognate English Leach, Lech, Letch, and Latch appear in the south and in the Charters as lacu = also '* lake ") ; Loft ; Loan. Meols (Norse MeUr, a sand-bank, appears in the Wirral dis- trict); Muggy; Mire. The Meoles at Shrewsbury, e.g. Cruck MeoUy seem to involve the Welsh 3foel = bald, i. e. treeless. Nore (= Norse NSr, a bay with narrow entrance).' 6ss (river-mouth: in northern place-names); Orm (**Orm*s Head " = Serpent's head). Raise (to elevate) ; Royd and Roothings (Norse Rjo5 = a clearing : in northern place-names : cf. German Riute, Rente) ; Root; Ra, R6 (= Corner: in northern place-names). Set and Saet-r (= summer grazing places : in northern place- names) ; Skagi (tongue of land : in northern place-names) ; Skog-r (Wood : in northern place-names as Sko and Skov, the latter a Danish form) Skin; Scab; Skirt; Scrape; Skull; Sky; Score (a notch) ; Scum ; Screech ; Scrannel ; Scraggy ; Squatter; Squander; Squall (to cry); Stith (= Anvil) and * A. -Sax. Gafol = fork, is derived according to Torp {IVortschatZj 125), from Latino-Celtic Gahalus = Gallows. "^ The A.-Sax. Hoh^ pi. Hdas ( = promontory), as in Clobes-hoas, Cloves- hoh, may be cognate, but Sweet makes this to be the same word as Hdh = heel, which Torp ( Wortschatz) traces to a different Teutonic stem. The Hill of Howth involves the Norse Hojbe = a high headland ; an OHG. form appears in Donners haugk, and suggests that the Clob in Clobes-hoh may be mythological. ' As ** the Nore " in Kent seems the only instance of this term in Britain it is possibly a Jutish word (see p. 226). 232 BRITISH PLACE-NAMES Stithy; Sly; Sleuth (hound) ; Snap; Snar (= quick: in Snar- gate [= the Shakespearian "next way"], which being in Kent may be Jutish); Spae -woman (prophetess); Spike; Spill; Sling; Swain; Sway. Tarn ; Teig-r (piece of meadow : with ablaut, A.-Sax. Tig^ Tih\ A.-Sax. Teag (= paddock, = Tey in place-names)) is traced by Torp to a different stem (see p. 217) ; Take; Wapen- take; Thrive; Thrift; Trash. Wand (a rod) ; Wath (a ford, in northern place-names : the English form is Wade (A.-Sax. Wded) in Biggleswade, &c.); Whin, Whirl. IV THE WARS OF ENGLISH AND NORSEMEN The events ascribed by the Anglo-Saxon Chronicle and other authorities to the closing years of the eighth century are, for the purpose of this work, comparatively few. Under a. d. 788 we have the record in the Anglo-Saxon Chronicle of the meedng of a Synod at Pincanheal (in Simeon of Durham, according to the Surtees Society Editor, Winchanheal and Wtncahala) and of the assembling of another at Aclea in the following year — a year under which the Chronicle notes the slaying of ^Ifwald, King of Northumbria, Simeon adding that this was " in a place called Scythles Cester, next the Wall ",^ and that he was buried in the ^ Pincanheal appears in one MS. of the Anglo-Saxon Chronicle as Wincanheal. The confusion between the Saxon P and W has often led to misreading of proper names, and we have no guide here as to the original form, except the assumption of Matthew of Paris (following R. de Wend- over), written before 1269 (Rolls, Ed. i. 354), that the place was Finchale, near Durham, in his days a Benedictine Priory, founded about 1196 on the site of the hermitage of St. Goodric, who died about 1 1 70. From the lives of this saint (Capgrave, N'ova Legenda Angltcte, and Ada Sanctorum, May 21) the place was fixed upon by St. Goodric because it was a primitive solitude, a resort of wolves. It was not, therefore, likely that it would be selected for a Church Council some 370 years before this hermit's time. There is, moreover, as far as I know, no instance of an initial P becoming F among the Teutonic peoples of Britain in historic times. A change ot initial ^ to ^ is almost without precedent. An initial Th in Early English, it is true, became often P-, Thelbrig (i.e. plank or deal bridge) is now Felbrig; Thindon (1296) is now Finedon (Northants) ; Thenglesham (831) is now Fenglesham (Kent). The earliest form of Finchale seems to have been Fink-halh, which became later Fink-halgh, Fink-halugh {^Priory of Finchale, Surtees Soc, p. xiii), where yf«>^ seems to be theAnglo- Saxon_/?«ir = finch, the Anglo-Saxon halh representing, according to H. Sweet, 234 BRITISH PLACE-NAMES Church of St. Andrew, Hexham (Hehstealdes-ige). The death of Jaenbryht, Archbishop of Canterbury, is noted in the Chronicle *' a corner ", «' a hiding-place ", but in its modern equivalent, "haugh''^ low lands by a river. The final h in halh is a guttural = ch. The latinized form is Halha^ which appears in a charter of the early thirteenth century, as de- scriptive of thirteen acres " infra Halha " ( = under the Haugh) between the rivers Wear and Gaunlesse {Feod. Prior. Dunelm. p. 7 note). This Halha seems to have been part of Bishop Auckland, of which the earliest form in the Durham Charters and in the *' Boldon Book " is Alclet, probably for an earlier Hak-let or Halc-clet, that is Haugh-rock, so that neither "oak" nor " land" seems to enter into the name. For a parallel transformation of Nolt-clet to A^oltland see Viking Mag., Jan., 1909. Halugh-ton, near Dar- lington, a similar compound of Hale, is now Haughton. The form of the word in Scotch Charters (e. g. Kinloss) is Halech. This form seems to sug- gest that the word is Celtic, especially as a connexion between English hole and healh (Torp, Wortschatz) is not clear. The Breton place-name Lin- halec — salt-lake suggests the connexion. Finchale does not appear in '* the Boldon Book " drawn up in 1 1 83. Simeon of Durham (who died about 1 1 30) gives the name of the place of the Council as Pincanheal and Pincahala, which his Editor (Surtees Soc, Simeon of Durham, i, pp. 22 and 29) has no hesita- tion in changing to Winchanheal and Wincahala respectively. Whichever form we assume, the locality of the Council is yet to seek. Haddan and Stubbs {Councils, iii, p. 444), after pointing out that Smith (Bede, //. E. iii. 27) identifies Pincahala and Finchale with Pgegnalaech (in Anglo-Saxon translation of Bede PeginaleaK), are inclined to the reading Wincanheale, and to connect it with Wincle in Cheshire, but this is Winchul in late twelfth century {B. M. Charters and Rolls) and possibly the Winescol of Domesday Book. For the possible modem form of Psegnalaech see the conjecture, p. 177. The element P*:?^ assumed there seems to occur in other northern place-names, e. g. the Domesday Paghenale, later Panhale, and now Pannel, near Harrogate. As the previous Council of Aclea (782) is not mentioned by Simeon of Durham, Haddan and Stubbs considered it to have been a southern gathering. They add, " In the Durham Ritual published by the Surtees Society there is a note written by a Provost Aldred, at * SuSan Wudigan gaete (= ?Newdigate, Plummer, Anglo-Saxon Chronicle, ii. 78) aet A'clee on West Ssexum', in the tent of Elfsige the Bishop. The note is of the tenth century, but the place is probably the synodical place of encampment. It is not, however, further identified " ; Haddan and Stubbs, Councils, iii. 439 note. The Synod of Aclea or Aclech of 788, Haddan and Stubbs, following Raine {Rich, of Hexham, Surtees Soc, p. 38), would identify with Aycliffe near Darlington, but aclea = oak-wood (original meaning of leah), and there would be no temptation to change it. " Sc) thles cester next THE WARS OF ENGLISH AND NORSEMEN 235 under 790, and also the appointment of his successor ^thelheard [Abbot of the Monastery of Hlud, according to one MS.] in the same year. Baldulf was consecrated (at Hearrahalch, says Simeon) Bishop for Hwitern (Whitehorn in Galloway) in 791, according to the Chronicle, and in the same year, according to Simeon, two sons of ^Ifwald were slain by ^thelred, King of Northumbria, at Wonwaldremer.* Under 792 the Chronicle states that ^thelbryht, King of East Anglia, was beheaded by order of OfFa, King of Mercia, and that Osred, who had suc- ceeded -^Ifwald as King of Northumbria (and had returned, according to Simeon, from exile in the Isle of Man (Eufania), whither he had been driven in 790), was slain (at Aynburg» according to Simeon) and buried at Tynemouth. Simeon of Durham, in recording the death (in 796) of Offa, refers to the the wall " (variants Scydescesire, Cithlescester and Scithles-uastre) has been identified with Chesters, near Chollerton (in 1233 Chelreton, Cal. of Chart. Rolls, i. 171) on the Wall. Both Chollerton and Chollerford seem to contain the name of the Roman station Cilumum {Not. Dign.) and to involve a pos- sible earlyname {CholUrn) of the NorthTyne. If Scythles-cester beCilurnum, it has nothing in common with that name, but it may mean and be a corruption of some other station near the wall, such as " Uxellodama ", which follows, with one interval, Aesica (" Great Chesters") in the Ravenna Geographer. ' Hludense Monasterium \ Haddan and Stubbs {Counc. iii. 468) are inclined, for reasons there given, to place this in Mercia, and to identify it with Louth {Luda before 1166, B. M. Charters, p. 487) or with Lydd or Luddesdown in Kent. Louth does not seem to be recorded in Domesday , and Lydd represents an early Lyde and Lide. Luddes-down appears to involve a personal name. The Domesday Lud-cerce (Lud Church) in Notts, is almost an exact equivalent, and was probably close to the modem Lowdham {Ludham in Domesday). It does not, however, appear in the Taxatio, but it answers the term more fully than any other conjecture. Simeon interprets Hearra-halch as "the Place of Lords". The Anglo- Saxon Hearra certainly meant " Lord", as in the Anglo-Saxon version of Genesis, and in Ccedmon, and halch has been explained above. There appear in the Charters of Melrose {Monastic Annals of Teviotdale, 260) a Harehoch and a Harehouden (now Harden in Koxboroughshire), which may possibly contain the name. Mr. Plunimer suggests Harraton in Durham. Aynburg has not been identified. Wonwaldremer is Winander- mere, now Windermere. 236 BRITISH PLACE-NAMES dyke built by him from sea to sea, that is, from the estuary of the Dee to that of the Severn. The year 793 is noteworthy in the Chronicle for an attack on the Church of Holy Island (Lindisfarena €€) made by the northern heathen pirates who harried Northumberland in the following year, but after despoiling Ecgferthes Minster at Donemuth,! were scattered by a storm, some of their ships being dashed to pieces, and the escaping crews killed by the English at the river mouth. Under 794, the Annals of Ulster (Rolls Series, p. 275) records the burning of Rechra^ and the pillaging of the Isle of Skye i^Ski) by the heathen pirates. The previous year they had devastated all the islands of Britain {ibid.^ p. 275), and in 797 Peel {Inis Pairaicc), in the Isle of Man, and Ireland and Scotland (Alba) were ravaged by them. In 802 and 806 they devastated lona (/ Columbce Cille), and the head of that community removed to Kells in Ireland — the body of St. Columba being taken about the same time to Saul,^ ^ Mr. Plummer says, "Yarrow is called 'Donemouth' as being the junction of the Done with the Tyne" {Anglo-Saxon Chronicle, ii. 64). "In Lindisfarena ^e" means "in the island (J^, a Northumbrian form, presumably, for ige) of the people of Lindisfarne", as we see from Simeon of Durham. It was the custom of the pirates to attack the islands first. It is curious that in a letter to the clergy and nobles of Kent, dated 797, Alcuin warns them of the extreme danger threatening Britain by hitherto unheard-of ravages of pagan pirates, and of the need, therefore, of giving up their internal dissensions. He recommends them to study Gildas the wise, and to learn from him the causes why the Britons lost the rule of their country. 2 Rathlin Island off the coast of Antrim, or Lam-bay Island, both called in early times Rechriu for (?) Rech-eriu = standing before Erin, gen. of Eriu, cf. Welsh Rach-ynis = island off the coast (Strachan's Middle Welsh Grammar), Bec-eriu = little Ireland, was the early name of "Beggery Island" off the coast of Wexford {Martyrology of OettguSj Henry Bradshaw Soc, p. 118). The same name appears as that of a place belonging to Glastonbury in a Charter ascribed to 971 {Cart. Sax. iii. 574), see p. 205. ' The form of this name in The Book of Armagh (a.d. 807) is Saball, and is, as Pedersen has pointed out {Vergl. Gramm., p. 219), the earliest Irish instance of a borrowing from the Latin. The original was Stabulum. THE WARS OF ENGLISH AND NORSEMEN 237 Co. Down, Ireland, where it was when The Book of Armagh was written in 807. The attacks of the Norse pirates became more and more frequent, and they established themselves in forts at Dublin and elsewhere. The Danes seem to have appeared first in these raids, according to The Wars of the Gaedhill with the (?«z7/ (Rolls Series, p. 19), about 850, although Mr. Skene {Celtic Scotland) wrongly calls the early invaders uniformly Danes. ^thelred, King of Northumbria, was slain, according to Simeon of Durham, in 796 [795], at Cobre,^ and was succeeded by Osbald, who, obliged within a month to fly, took refuge with the King of the Picts. Eardwulf, having succeeded, was conse- crated at St. Peter's, York, in 795, but his claims were questioned, and were only decided by a subsequent battle which was fought in 798, according to the Chronicle, at Hweallaege, in Northumberland, and, according to Simeon, at a place called by the English, Billingahoth, close to Walalege.^ Eanbald, Arch- bishop of York, died in 796 (in a monastery called Etclete,' says Simeon), and was buried at York (Eoforwic). In the same year Ceolwulf, King of Mercia, made himself master of Kent and took thence King Eadbryht Prsen prisoner to Mercia. One MS. of the Anglo-Saxon Chronicle relates the death, under 798 [?], at Sudberi (Sudbury, Suffolk) of Alfhun, Bishop of The Celtic word for barn is in Irish Skibber^ in Welsh Ysgubor, and a Breton form Skiber which seems to have found its way to the Continent. See the German forms in Diefenbach's Glossarium, a supplement to Ducange, where Skiibor, Skiber, Sabal, and Sawl are given (under Acervus) as equivalents of Barn. ' Cobre is regarded as a shortened form of Corbrigge = Corbridge, already explained, but it may be for Cover-hAm. 2 HxueallcBg and Simeon's Walalege have been identified with Whalley in Lancashire. The form in Domesday is Wallei, and the earliest docu- ments of the Coucher Book of Whalley (fourteenth century) give Whalleye. Billingahoth does not occur in the latter, but Bilyngton does. ' Etclete is so like the early form of Auckland {Alclet) that, in the absence of any other possible identification, it may be regarded as a mis- reading of that name. 238 BRITISH PLACE-NAMES Dunwich, and of his burial at the latter place [{on) Domuce], and the same MS. records that the body of St. Wihtburg was found at Deorham {Dereham^ Norfolk) in the 5ame year, uncor- rupted fifty-five years after her burial. In the last year of the century (800) the Chronicle records the succession of Ecgbryht to the Kingdom of Wessex, adding that on the same day ^thelmund, an alderman from the Hwiccii, rode over Cynemaeres ford {Kempsford, Glouc.) and was en- countered by the men of Wilts (WilscBte) under their alderman, Weoxtan, and the latter were victorious, both leaders being slain.^ Simeon of Durham records in the same year the conse- cration of Eanbryth at Cettingaham to the bishopric of Hexham. Under 801, Simeon of Durham notes the death of a certain Abbot Edwin, " also called Eda,'' and his burial at Gegen-forda." In the Chronicle from 801 onwards the monotonous record of the succession of kings and bishops is only broken for some years by occasionally a more interesting event, such as, under 813, the harrying of Devon and Cornwall {West Walas) from east to west by Ecgbryht, referred to above, or in 816 the burn- ing of the English School at Rome. In 822 the Chronicle records another Synod held at Clofes-hoo (variant Clo/es-ho), and under the next year we have an account of further fighting with the Welsh (Wala) and Devon men {Defnd) at Gafulford,' and of a battle in which Ecgbryht, King of Wessex, defeated at ^ This is the only mention of the Hwiccii in the Chronicle. Cyne- mseresford was in 1236 Kynemeresford {Cal. Chart. Rolls, i. 222), and in 1 54 1 Kamyseford {B. M. Charters). Cettingham is probably Cottingham, Yorks. (Domesday Cotingeham, compare Catingeuuic). * This is Gainford-on-Tees, see Simeon of Durham, Surtees Soc, i. 39 note ; Gegen, Gegn, or Gean involves the idea of our again, and in composi- tion means ''duplicate". The Genlade of Bede, already explained, contains the same idea. Note how Edwin is shortened to Eda. This process has been previously referred to, and we have, as already stated, many instances of these shortened personal designations in place-names. * Clofes-hoo has been previously dealt with. Gafulford is identified (Plummer's Anglo-Saxon Chronicle, Index, ii) with Galford, Devon, but see p. 141. THE WARS OF ENGLISH AND NORSEMEN 239 Ellendun ^ Beornwulf, King of Mercia, with great slaughter. The same year witnessed the extension of Ecgbryht's power over Kent, Surrey, Sussex, Essex, and East Anglia : in 827 he over- ran all Mercia, together with the country south of the Humber, and led an army to Dore ^ against the Northumbrians, who were obliged to submit to him. He was also, as the Chronicle says, the eighth Bretwalda (variant Bryien Wealda) since the origin of this title. In 828 Ecgbryht led also an army against the North Welsh (North Walas) and reduced them to obedience. The Annales Cambriae record, under 816, an invasion by the Saxons of the district of Snowdon {Monies Ereri) and Roweynauc (Rhyfoniog^ Denbigh), but this, like the invasion of the latter region (here called Rieniich) by Offa, recorded in the same Annales under 795, must be ascribed to the Mercians, and not to Ecgbryht. The record in the same Annales, under 822, of the destruction of the citadel of Deganwy (Arx Deganhut) and of the conquest of Powis {Poywis) by the Saxons possibly refers to the campaign of Ecgbryht in 828. The Scandinavian pirates appeared on the scene again in 832, when they ravaged Sheppey in the Thames (Sceap-tge), and in the following year Ecgbryht fought against thirty-five ships' crews at Char- * The site of Ellendun is supposed (Plummer's Anglo-Saxon Chronicle j Index) to be Wroughton, Wilts. This is fully confirmed by the fact that Wroughton is still known in ecclesiastical registers as Ellingdon. A paper on this subject, written by Mr. T. S. Maskelyne, sent to me by Canon Codrington, D.D., of Chichester, traces the history of the mistake by which the place was identified by J. R. Green (in his Hist, of the Eng. People, p. 45) with Wilton. Sir R. C. Hoare was the first to identify the site with Wroughton {Reg. Wiltun., pp. 54, 55), which name, he says, was added to Elyngdon. ' Dore (Derbyshire), south-west of Sheffield, still retains the name, which it bore also in time of Domesday. It is mentioned also in the Chronicle under 942. Mr. Plummer would make it an English word ( = door) meaning a mountain pass, but it looks more like the cognate Celtic Dor-um, encountered in many place-names with seemingly the same meaning. It may, however, be a river name. The Bdora oi Ravennas seems to be in a sequence with places conjectured to be in Derbyshire, see p. 117. 240 BRITISH PLACE-NAMES mouth {cBt Carrum), but after much slaughter the Danes {Deniscan) held the field. In the following year (835) a large fleet of the pirates, having put into Cornwall, joined forces with the West Welsh (West Walas) and made successful raids into Wessex, of which Ecgbryht having heard, he led an army against them and put to flight both the Cornish {Walas) and Danes {Demscan) at Hengestdun.^ After a glorious reign of thirty-seven years he died in 836. The year 837 witnessed, according to the Chronicle, two fights with the Danes, one at Southampton {Hamtun\ in which the Wessex alderman Wulfheard gained a victory over thirty- three ships' crews of Danes, and the other at Portland {Port),^ where ^thelhelm, the leader, and the Dorset men were defeated by the Danes, and ^thelhelm slain. In the following year (838) the Scandinavian pirates inflicted a defeat on Herebryht, alder- man, in the region of the Marsh-dwellers {Merscware), that is of Romney Marsh, and slew many in Lindsay {Lindesse), East Anglia, and Canterbury. From the middle of the ninth century onwards the attacks of the Scandinavian pirates on the British Isles became more frequent. The year 851 is distinguished by a victory obtained over them at Wicganbeorg by the men of Devon under Ceorl the Alderman, and also by the slaying of a great number of them and the capture of nine of their ships in a sea-fight at Sandwich {SondwicY in Kent, by King ^thelstan — under- * Hengest-dun is Kingston Down, Cornwall. Hengst ( = stallion, horse generally, but possibly also a personal name) enters into several place- names, e. g. Kingston, Cambridge, and Kenstridge, Somerset. 2 Fort is used for what is elsewhere in the Chronicle (982) called Portland. 3 The Scandinavian invaders are generally designated simply as " Kea- then " in the Anglo-Saxon Chronicle. This word in one form or another is common to all the Teutonic dialects. Its assumed connexion with Heath in the signification of" men of the heather" is not at all convincing. Professor Kluge would be inclined to look upon the word {Etym. Diet, sub voce), of which a form appears in the Gothic translation of the Scriptures {Haithnd—'a. feminine noun sing.), as being adopted like '^Church"" THE WARS OF ENGLISH AND NORSEMEN 241 king of Kent — and the Alderman Ealchere. The heathen in- vaders, the Chronicle adds, remained in Thanet this year for the whole winter, a thing they had not done before. The year 851 did not pass away without reverses elsewhere. Three hun- dred and fifty of the invaders' ships having entered the Thames, and ^' Pfaffe" (the Papa previously referred to) in some way from the Greek. It would seem, therefore, that Dr. Vigfusson's suggestion {Icelandic Diet, sub voce) may be correct that it is adapted from the Greek td. Uhtie, the word in the New Testament for gentiles^ see also New Eng. Diet. The Anglo-Saxon Chronicle sometimes specifies them as Danes {Deniscan). In the Irish Annals the Norwegians are called either Fair Gentiles {Find Geintt) or Lochlanns {Lochlanncuh, i. e. men from the lake or fiord-land). The latter term is preserved in the surnames of McLoughlin and McLachlan. The Irish called the Danes Danar or Dub-geinti (Black Gentiles) or Dubgaill (black foreigners), which still exists in the surnames McDougal or McDugald, McDowall = " son of the Dane*' and Doyle. As to Wicganbeorg ( = \Vicga's hill or tumulus), Mr. Stevenson ( Asser's Life of Alfred, p. 175) rightly dismisses the identifica- tion with Wembury or with Weekborough, both in Devon. He suggests Wigborough, Somerset, which Mr. Eyton identified with the Wincheberie of Domesday and Winchin-beria of the Exon. Domesday. But it is not prob- able that the beria termination in Z'^wtfja'ay represents ^^^r^= hill, which appears to be represented apparently therein by berga, representing an oblique case. Beria seems uniformly to represent the dative of Burh = Byrig. The termination bere or bera in the Exon. Domesday seems to represent the Anglo-Saxon beam, dative bearwe, = wood , as in Ac beara, Collabera, Eattebera, Lochesbera, Hundesbere, Laurocabera (now Larkbear). It occurs also alone as in ''Bere Forest". In Exon. Domesday, Beorg, dative Beorge, appears seemingly as berga^ as in Albretes-berga, Lange berga, &c., and takes in modem times the form borough. Burh, dative byrig, is represented in Chentis-beria (now Kentis- bury), Cungres-beria (now Congresbury), &c. In these circumstances Winchin-beria cannot well represent Wicganbeorg, which remains thus unidentified. Sandwich {Sondwic and Sandwic) involves w^V in the sense of "bay", and is, therefore, probably Scandinavian {=Vik), although, as we have seen (p. 210), the term Wicing occurs before the arrival of the Norsemen, and the fVic may be English. It was the chief haven on the south coast for the assembling of the Danish fleets. It is mentioned in the Sagas {Corpus Poeticum Boreale^ ii, 262). Sand-vik (sand-bay) is a common place-name in Iceland and Norway. Q 242 BRITISH PLACE-NAMES the crews took Canterbury (Cantwaraburg) and London (Lundenburg) by storm, and put to flight the King of Mercia (Beorhtwulf) and his army. Passing over the Thames into Surrey, the invaders were encountered at Ockley (Acka) by King ^thelwulf, and his son ^thelbald, together with the Wessex host, who, after slaughtering an unheard-of number of the heathen, gained the day. In one of the MSS. of the Anglo-Saxon Chronicle drawn up at Medeshamstede, that is, Peterborough, we have, under 852, some local details with regard to the endowment of that monas- tery. On condition of his assigning Sleaford [Slioiva/ord) in Lincoln, to Medeshamstede, and paying certain rents in kind, the monks let to Wulfred the land of Sempringham {Sempinga- hani) for his lifetime. The Charter dealing with this arrange- ment is in existence.^ Under 853 the Anglo-Saxon Chronicle records that the Mercians assisted by the King of Wessex attacked the North Welsh {North Walas) and reduced them to obedience. The Chronicle goes on to state that the men of Kent, under Ealhere (probably the Ealchere referred to above), and the men of Surrey, under Huda, fought in the same year with the heathen ^ This Charter is in possession of the Society of Antiquaries, and is printed in Cart. Sax. (vol. ii, p. 57). Sempingham (which lies to the south of Sleaford) seems to be the correct form. The name occurs also in the Domesday Survey of Norfolk in the hundred of Diss {Bice), It is now Shimpling there. Sempinga is seemingly a genitive plural of a patronymic. Slioiva ford contains apparently a river-name. Compounds with involve generally (i) a personal name in the genitive; or (2) the nature of the river at the point (as Stan-ioxdi, Mud-ford, Sand-ford, &c.) ; or (3) the animals, &c., which may cross, as Horse-ford, Shef-ford (Sheep), Oxen-ford^ Wain-ford (Wagon) ; (4) the kind of river, and sometimes its name ; (5) the aids to steady crossing as Staf-ford, that is, with a guiding rail : Barn-ford (that is. Beam = tree laid across) : Trefford (Tree, Treow-ford). *' Sliowa-ford " comes under (4), and the brook is there still known as the *'Slea", for probably Sleow-ea = Tench-river : compare Trout-beck, a fre- quent brook name. THE WARS OF ENGLISH AND NORSEMEN 243 pirates in Thanet {Tenet), but after an initial success were worsted with the loss of both their leaders. In the year 854 the heathen men remained again, the Chronicle states, over winter, taking up their quarters in Sheppey (Sceap-ige, that is, Sheep-island). ^thelbald, who had succeeded his father ^thelwulf as King of Wessex, died in 860, and was buried at Sherborne {cet Sctra- burnan). ^thelbyrht, his brother, now became King of Wessex and Kent, and in his days the heathen men came with a large fleet and stormed Winchester (Win/an Ceas/rey but the Hampshire and Berkshire men put them to flight and held the battlefield. Under 861 in the Anglo-Saxon Chronicle the record occurs of the death of St. Swithun, Bishop (of Winchester), whose name is not mentioned anywhere else in the Chronicle. In 865 the men of Kent tried to buy ofif the heathen army in Thanet, but the latter stole away in the night and ravaged all the east of Kent. Next year a great fleet of pirates arrived, and took up their winter quarters among the East Angles {on East Engluni) providing themselves, says the Chronicle^ with horses. The East Angles made peace with this army. Having crossed the Humber in the following year to York {to Eoforwic Ceastre) the heathen army was attacked there by the Northumbrians,* who stormed the town, but were at last defeated with great slaughter and had to make peace with the enemy. The Welsh Annals {Annates Cambriae under 866) refer * Scira-burna = Clear brook. Wintan-ceastre (Winchester) is based on the ancient Romano-British name Venta-Belgarum, or the Venta of the Belgae, who, according to Caesar, were Continental Celts. The meaning of Venta has been already dealt with, p. 32. ^ The Angles attracted Eburacum as I have already said (p. 37) into a familiar compound Eofor-wic, that is, " Boar-lair or shed." This was, like all the adaptations from the Romano-British nomenclature, a great leap, and it took place at the very outset of the occupation. It may be said, indeed, that all the startling changes in place-names took place shortly after the Teutonic settlement. The forms afterwards were fairly well preserved. Q2 244 BRITISH PLACE-NAMES to this battle at York {Urbs Ebrauc) noting that it was devas- tated by the Danes (Dub gint = black gentiles). The heathen army moved the next year into Mercia, to Nottingham {Snottingaham ^). The Mercians and the men of Wessex tried to dislodge them, but failed, and had to come to terms with them. The heathen went again to York in 869, where they remained a year. In the following year (870), having passed through Mercia to the East Angles, they took up their winter quarters at Thetford {cBt Theodforda'^). Here King Edmund fought against them but was defeated and slain, and the Danes (Deniscan) proceeded to subdue the whole country, and to destroy the minsters. The rich abbey of Medeshamstede was reduced by them " to nothing " on the occasion. Some additional details upon the invaders who arrived in East Anglia in 865 are given in the Chronicle of Melrose ^ a twelfth-century MS. The leaders are there called Iwar {Ivar) ^ Snottinga-ham is said in Asser's Life of Alfred to have been called by the British Tig guocobauc, which Asser, who, as a Welshman of St. David's, knew Welsh, says means " House of Caves" (rather "Cavy-house"), for guocob-auc is an adjective derived from Guocob = ca.ve, modern Guocof. Guocob and Guocov occur in the Book of Landaff, — Wookey and Wookey Hole near Wells, and the early forms are not against the supposition, seem to represent the British word. Compare Vooga (Williams' Cornish Dictionary) which Norden makes Googoo, that is Guocof Cf. Fogo in the Brythonic districts of Scotland. The Tig is the early British form for house, modem Ty. Any one who has seen the caves at Nottingham cut out of the sandstone rock will admit the appropriateness of the name. Snottingham has nothing to do with caves, but is probably, as Mr. Stevenson says, derived from the patronymic Snottinga, descendants or family of a certain Snuot or Snot. Snottar is Anglo-Saxon for Wise, and may be a form of the word. A similar compound exists in Norse names. Another instance of Asser's knowledge of British Place-names is his British equivalent of Thanet = Ruim. This would seem to indicate an acquaintance with Nennius, where this name is also found. See p. 127. 2 Theod-ford made up of Theod^'' people", " public" Sind ford, that is, it was on a public highway. There are several places of this name and similar compounds occur among the Scandinavians, e.g. TheoS-braut = Highroad. THE WARS OF ENGLISH AND NORSEMEN 245 and Hubba, who directed the campaign in Nottingham and York and were in command when King Edmund was slain in East Anglia. In the F. MS. of the Anglo-Saxon Chronicle they are called Ingware and Ubba. The heathen army when it advanced to Reading was, the Melrose Chronicle states, under the leadership of Bagsecg and Halfdene, Ivar having presumably died (Wars of Gcedhill and Gatll, p. 270). In the series of battles fought in this part of Wessex, Baegsecg (Baegscecg in Asser) was killed, and Halfdene then proceeded with part of his host into Northumbria, whence (according to the Anglo-Saxon Chronicle under 875) he harried the Picts and Strath-Clyde Britons. Strath-Clyde had been ravaged by Norsemen (" Loch- lanns" according to Fragments 0/ Irish Annals, p. 192) a few years before 870, and Alt Clyde {All Clut in Welsh Annals = Dumbarton, properly Ail-cluathe^ genitive Ailech Cluathe = the rock of Clyde) taken by siege. The year before these Norse- men (according to Fragments of Irish Annals) had ravaged Fortrenn and made the people tributary.* ^ The various spellings of Ivarr in the Anglo-Saxon and Celtic annals {Imhar in Irish) are doubtless owing to the early form of this Teutonic name, viz. Ingo-tnar. In the Flatey Jarbok, iii. 507, it appears as y'ngvar, but generally in Norse as Ivarr, in which form it has passed into the Scotch personal name Maclver (or McKeever) = Ivarson and Iveson, and the Welsh Ivor. The celebrated Archbishop of Reims, Hincmar, bore the same name. The Ingo is obscure, but Forstemann would connect it with jung (young) and also identify it with the suffix -ing in Teutonic personal names, e.g. Harding, Heming, Helding, &c. Hubba is a name unknown to the Sagas. It is probably a nickname. A similar form Hufa is used as such, meaning " with a helm ", and Hubo, a Teutonic name, is made cognate by Forstemann. Bnegsecg is evidently, from the variants, a corrupt form. It may be an ekename of which the second element is probably, from the form in Asser, Skegg = beard, as in Kolskegg = Black beard. It could hardly represent Berserkr, which was also used {Iceland. Diet, sub voce) as a nickname = "Bear- coat." Half-dene has been already explained. In the Irish Annals it is Alb-dan. Ethelwerds Chronicle states that King Edmund was buried at Beado- rices (/ur/Ae, th&t is, " Beadu-richs farm," the old name of St. Edmundsbury. Beadu-rich ( = battle-king) was a common Saxon name and gave origin both to Battersea and Petersham, which in Doniesday Survey are " Patri- 246 BRITISH PLACE-NAMES In 871 the heathen army came to Reading {to Readmgum), but they were defeated at Engle-field {on Engla-felda ^). Three days after this, King ^thered, and Alfred his brother (the future celebrated King Alfred), led an army to Reading against the heathen men, but after great slaughter on both sides the Danes held their ground. About four days later. King ^thered and Alfred fought another battle with the foe at ^scesdune,'* and cesy" and " Patricesham " respectively. In the original Westminster Charter of 693 {Cart. Sax. i. 116) the former is Batricesege. In 1067 {Brit. Museum Chart, and Rolls : Copy) the former is Batriceseie or Fatriceseia, and the latter is in 1266 {Merton Cart.) Petricheskam. It is clear we have to do in both cases with a Saxon magnate called " Beadu- rich " and not with " St. Peter ", who is wrongly assumed by some to be the Petrich in Petrichesham. Initial P is not an uncommon mutation of B in English. As to the Scandinavian raid upon the Picts and Strath-Clyde Britons, it is clear that the aggressors here were Danes, but the siege of Ail-Cluathe (870 in the Annates Cambriae) was by Lochlanns, who had also in 865 or 869 invaded Fortrenn under Amlaiph (that is, Atnhlaif, pronounced Awlaif = Olaf) and Avisle {Ann. Ult. i, p. 374 = Haisl) aided by the Scandinavians of Ireland and Scotland, and had plundered all Pictland. These were not Danes, as Dr. Todd ( Wars of Gaedhill and Gaill, Ixxix) made them out to be, but, according to Fragments of Annats, p. 159, Norsemen. Fortrenn has been dealt with before. It was the Pictish district proper, and contained Dunkeld and Scone. The similarity of the word to Fortren («= mighty, Felire of Oengus, Jan. 31) probably led to its being considered in the Irish version of Nennius as the eponymus founder of the district. ^ Engla-felda is probably rightly explained by Asser as the " Camp of the English ". Readingum is the locative plural of a patronymic, nomina- tive plural Redingas. The Basengum mentioned lower down is of the same nature. Meretune is probably' Marten, Wilts., which is "Mertone" in Domesday, and "Meretone" (in Himdred of Kinwarestan) in 1227. Winburna is most probably, as was previously stated, a hybrid word involving the British ?F?«« = white and the English buma. 2 ^scesdun is not, as Mr. Stevenson has rightly shown, composed of Ash-\xQe and Down (as Asser states), but involves a personal name ^sc in the genitive case, /^sces, contracted from some such original as /^scwine. The name still exists in Ashdown Park in the manor of Ash-bury (that is, yEscesburh) at the west end of the Berkshire Downs. A Charter ascribed to about 931 {Cart. Sax. ii. 376) gives as a boundary y^scces byries Subgeate, that is, the south entrance to -^scesbury. Mr. Stevenson (Asser's Life of Alfred, p. 238), though tempted to connect the Nachededorne THE WARS OF ENGLISH AND NORSEMEN 247 after great slaughter put them to flight. About a fortnight later, however, ^thered and Alfred were defeated at Basing {cet Basengum) and again after a hard fight at Meretune, where the Danes (Deniscan) possessed the batdefield. After this battle there arrived a great body of pirates {Stimor lida = Summer army), and in the same year King ^-Ethered died, and was buried at Wimborne {cei Winburnan)^ his brother iElfred suc- ceeding to the crown. Alfred became king, as we saw, in 871, and all the manu- scripts of the Anglo-Saxon Chronicle agree in giving 901 as the year of his death. His reign of some thirty years was as event- ful, perhaps, as that of any English Sovereign, and in his strenuous career during this stormy period he well earned the distinction of " The Great " which posterity conferred upon him. The record of iElfred*s deeds is preserved for the most part in the Anglo-Saxon Chronicle^ and in his life by Asser, which supplements the account in the Chronicle by some important particulars. Alfred's biographer was a Welsh Bishop, who says, in his own words (Cap. 79), that "he was brought up, educated, tonsured (as monk), and finally ordained in the holy places of the extreme west of Britain (St. David's) ". His learn- ing seems to have given him a certain fame, and he was invited by Alfred to pay him a visit. Asser relates how he came to the royal vill at Dene ^ in Sussex, where he saw the King for the first time. The King's efforts to secure Asser's services in his literary labours were not at first attended with success. Asser was unwilling to leave his country and his people without consultation with them, but promised to let the King know his decision. On his way back he fell ill of a fever, and lay ill for ( = " barethorne " Hundred, which contained an " Assedone ") of Domesday with the unica spinosa arbor of Asser, will not associate this Assedone with the ^scesdun of the battle. 1 This is doubtless the Dene mentioned in ^Elfired's will, and is repre- sented, as Mr. Stevenson suggests, by east and west Dean, near Eastbourne. It is in the dative or locative of the Anglo-Sax. Denu, a dell. 248 BRITISH PLACE-NAMES over a year at a place he calls Wintonia.^ Letters from the King finally induced him to agree to his request, a thing he was the more content to do because, as he says, the King's authority had at this time, and long before, extended to all the south part of Britain : " Himeid, for instance, with all the inhabitants of the Demetic region, had submitted to Alfred. So also had Houil {Higuel, that is modern Howel), son of Ris, King of Gleguising, and Brochmail and Fernmail, sons of Mouric, kings of Guent. Elised, son of Teudubr, King of Brecheniauc, had also submitted to the King. Some of these had been driven, as Asser tells us, by the exactions of Rodri the Great [King of Gweneth = North Wales], and others by the tyranny of the " Aldorman " Eadred (that is Ethelred), and of the Mercians to seek Alfred's protection.^ Asser was honourably received by * It is clear this could not be Winchester, Venta Belgarum, and it is probable, therefore, that it was the Venta Silurum, now Caer Went, which was on the Roman Road to the West, and is styled Guentonia by Welshmen in the tenth century when writing Latin. See Liber LandavensiSy Rh^s- Evans, p. 220. 2 Himeid's death is recorded in the Annates Cambriae under 892. He was King of Dyfed {Demetid). Maes-yvet (Maes Hyveidd of the Myvyr. Arch.), the Welsh name of Old Radnor before, and in the sixteenth century, possibly preserves his name. Rotri {Rodri), King of North Mid Wales and Cardigan, was slain by the vSaxons, according to the Annates Cambriae j in 876, and Howel's death is placed by the same authority under 885 (894 in the Brui). For Gleguissig {Gteguising) see p. 138. A Fernmail, son of Teudubir, is mentioned by Nennius as the King then reigning (785-815, according to Zimmer, Nennius Vindicatus, p. 71) in the two regions Buelt and Guortigirniaun. Buelt is now Buitth, and embraced that part of Powis lying between the Wye and the Severn ; and the lordship of Gwrthre- nion {Gwrthernion : Myrv. Arch., Warthrenion, Geratd. Camb. I tin. c. i) in Radnorshire, preserves seemingly Guor-tigim-iaun, which stands to Guortigern {Vor-tigern — oytxAoxd.) as Cereticiaun (Cardigan) does to Ceretic. The Brochmail and Fernmail of the text are seemingly of later date, but the latter was probably King of the Builth region. Brochmail and Fernmail witness a charter in the time of Nud, Bishop of Llandaff (ninth century), Lib. Land, p. 226. A Teudur, son of the Elised referred to on the pillar at Abbey Crucis (see p. 55), is mentioned under 927-9 as King of Brecknock, Lib. Land., p. 237. THE WARS OF ENGLISH AND NORSEMEN 249 King Alfred at his royal vill of Leonaford,^ and we learn from Alfred's preface to his translation of Gregory's Pastoral Care how he valued Asser's help. The King rewarded him first by granting him the monasteries of Cungresbyri and Banuwille,* and afterwards by the gift of Exanceastre, including the whole diocese {Parochia)^ which embraced also Cornwall {Cornubia). As the See of Exeter was not yet formed, it is clear that this was an earlier ecclesiastical division of which Sherborne was the Bishop's Stool, for the Annals of St, Neots records his death as Bishop of Sherborne in 909. Of Asser's Life of Alfred only one manuscript was in exis- tence in the sixteenth century, and this was unfortunately destroyed in the fire at the Cottonian Library in 1731. Prior to this, however, several copies had luckily been made, upon one or other of which the existing printed editions are based. It seems clear from the opinions of experts'* who had seen the original MS. that it was written in two or more hands about A. D. 1000. The author seems to have completed his biography in 893 (Cap. 91. 4), when -Alfred was forty-five years old. Anything subsequent to this is, therefore, from another hand. Mr. Stevenson's introduction to the latest printed edition of the work (Oxford, 1904) is a marvel of painstaking research, and from this and his numerous valuable notes to the text the reader cannot but arrive at the conviction that Asser*s Life of Alfred is a trustworthy historical document. The Annals of St. Neots, which was formerly associated with Asser's Life of Alfred, and * Mr. Stevenson conjectures this may be Landford, Wilts. 2 Now Congresbury {Thonne restetk S. Congarus on Congreshyrig — see Tha Halgan on Angelcynne, Liebermann, Hanover, 1889) and Banwell (probably from Benawt//e = FTa.yer--we\\). Cf. Bonus and Bedhus pre- viously explained. In a "History of the Bishoprick of Somerset till 1174" {Camden Society, 1840), Congresburia is made the earliest seat of the Bishop, and Bishop Daniel is there said to have removed it to "Cideston" [Tideston], that is, Wells. 3 Among others by Humphrey Wanley, an early Secretary of S.P.C.K., and afterwards Secretary of the Society of Antiquaries. 250 BRITISH PLACE-NAMES was so called from its having been discovered by Leland at St. Neots in Huntingdonshire, is regarded by Mr. Stevenson as a Post-Conquest compilation, made up chiefly from the Anglo- Saxon Chronicle and Asser's Life. Its topographical additions will be noted in their proper place. Within a month of Alfred's accession he had a conflict with the whole army of the Danes at Wilton (Willun), and having with his small force seemingly put them to flight, a ruse on the enemy's part, he lost the battle. Asser adds a topographical detail that Wilton was on a hill on the south side of the river Guilou} The Chronicle and Asser agree in stating that iElfred fought in the same year eight pitched battles with the Danes in the region south of the Thames {Themese), and was day and night on the move owing to incessant attacks of the enemy. A peace was patched up at the end. In 872, the Chronicle tells us, the Danes withdrew from Reading {Readingum) to London (Lunden by rig) and remained the winter there, the Mercians making peace with them, which they also did in the following year when the Danes removed to Northumbria {on Northhymhre) and settled for the winter at Torksey (^Turicesiege) in Lindsey {on Lindesse). Next year (874), the Danes having made their winter settlement at Repton {Hreopedune), drove King Burgrsed over the sea, putting a thane, a creature of their own, in his place. In 875 part of the Danish army (under Healfdan) went from Repton to the River Tyne, and harried the country, extending their attacks to the Picts {Peohlas) and Strath-Clyde Welsh {Sir ceded Walas). Three of their leaders (Godrum, Oscytel, and Anwynd) removed from Repton to Cambridge (Granle hrycgey and settled there for one year. In the same year 1 This is presumably the British form of the river Wiley which gives its name to Wil-tun and fViliumdr, whence IVtl^esa'r, modern Wiltshire. A similar form of river name occurs in Wales, as in the Camguili ( = '* winding-Guili") and in Aber-guilly in Carmarthenshire. Welsh 2 For the various stages of transition between Granta-brycg and Cam- THE WARS OF ENGLISH AND NORSEMEN 251 JEUred, having had a naval engagement with the enemy, took one of their ships and put the others to flight The Danes, having left Cambridge in the following year, proceeded to the Castle of Wareham {Wer/iam), where was a monastery (Asser adds) situated in a most safely protected place between the rivers Frome (Frauu) and Terente [= Trent] in the district (paga) which is called in British Durngueis (Mr. Stevenson reads Durngueir) but in Saxon Thornsceta} Proceeding from hence bridge see Professor Skeat's masterly paper on the subject. The Cam was originally called, it would seem, Granta, which also probably gave its name to Cair Granth (wrongly written Grauth in Nennius's List of Civitates Britanniae) now Grant Chester. The Camborico of the Itinerary ^ although near, has nothing seemingly to do with Cambridge. 1 Mr. H. Bradley, whom Mr. Stevenson follows, sees in Frauu an evolution of a supposed earlier Frdma, of which the river From (given in the Anglo-Saxon Chronicle under 998) preserves the form, a form that would have developed naturally in British lips to Frauv^ and later to Fraw, as in Aberffraw. This would mean that the Saxons had occupied this district before the British change of m to v had become operative (but see Pedersen, V. Gramm. § 99. i, for a possible instance of an aspirated in in the middle syllable in Gaulish). A similar early occupancy is implied, therefore, wherever we find in Britain a river called Fr5me, that is, in Somerset, Gloucester, Hereford, &c. As to the reading Durn- gueir^ which Mr. Stevenson assumes to be that of the original MS., it is considered to be a natural development of the name of Dorchester in Roman times, Dumo-varia, gueir representing varia according to Welsh phonology. But the British dialect in South Britain was not what we know as Welsh, Moreover, Thomsdta (or better DomscEtd) is not given by Asser as the equivalent of Dorchester, but of the district {paga) now Dor-set, an almost exact representative of the Durotriges (cf. Allo-triges and Allo-briges) of Ptolemy (see p. 55). A possible derivation of Dumo- varia raises a further difficulty. Durn-o can be explained here (as it has been by Dr. Whitley Stokes in the personal name Dumcuos) as "fist" (Irish Dom, Welsh Divni), and Varia, as " plays", " theatre". Seeing that there is no evidence of a native stronghold at Dorchester, it was probably founded by the Romans. The amphitheatre they constructed there would furnish to the natives a significant name for the place, and in British would possibly be represented by "Fist-plays" (Dourna in Breton is to fight with the fists). The Cornish word for a theatrical play was Gwarey a seeming direct descendant of Varia. Cf. plain ait Guare in Cornwall. " The Guare Miracle, in [English, a Miracle play .... For representing it they raise an earthen Amphitheatre in some open field," 252 BRITISH PLACE-NAMES to Exeter [Exanceasire —kss&v, Cairuisc)'^ the Danes lost 120 of their ships in a gale off Swanage (cet Swanawic^ a Norse name = Swain's bay). The mounted army proceeded to Exeter. Alfred strove to outride them, but although he failed to prevent their getting into this stronghold, yet he compelled them to give him hostages and make peace. The Danes then withdrew to Mercia. Asser, under the year 878, relates that the pagan army having left Exeter [Eaxeancestre) came to the royal vill of Chippenham {Cippanham), situated on the east bank of the stream "called in British Abon'' (Avony where they wintered, having driven Carew's Survey of Cornwall^ 1769, p. 71. Cf. the Welsh Guare miragl of the lolo MSS., p. 86. The Argento-varia (now ? Horburg) near Argento- ratum (Strassburg) was probably so called from an amphitheatre. The Anglo-Saxon forms of Dorchester, Dorn-wara Ceaster and Dorn-warana C caster y seem to preserve ^wa;'^, and not the genitive plural oi wer^wara, men or warena, ^ Mr. Stevenson regards the uisc in Cair- ( = casird) uisc as indicating a "Welsh development of Isca or {Esca) later than the time of the Saxon occupations of Exeter {Escanceaster), and that Asser heard the name from the natives of Devon or gave it a Welsh form. The latter is probable. Asser seems to show in one or two places an acquaintance with Nennius, in whose list of Civitates occurs Cair Legeion Guar Usic {Uisc\ which the Irish version places at Oper-uisc (Aber-uisc). This is now Caerleon. The form uisc is possibly, therefore, Welsh as well as Irish, and, perhaps, an oblique case. Caer Esca occurs in the Welsh Triads as presumably the early name of Exeter. ^ Cippanham (Chippenham) seems made up of Ciepa (gen. Ciepan) = "trader", and hamm, "meadow." The double/ in Cippa suggests, how- ever, a personal name. "The ham at Cippan-hamme" is a legacy in Alfred's will {Cart. Sax. ii. 178). There seems to have been no Roman road nearer to Chippenham than the Fosse Way by which the pagan army could proceed direct to Cirencester. Asser calls Cirencester Cair Ceri in British. This is another instance of dependency on the Historia of Nennius, where Cair Ceri occurs among the Civitates BritanniaCy most of the names of which seem to belong to Wales and the North. The identification is not supported by its ancient designation Korinnion^ Ptolemy : Cironium Dobunorum, Ravennas. The Commot of Ceri and the Castle oiCruk-ceri{Cru\ieT\) in Radnorshire possibly preserve the Caer Ceri oi Nennius. The form Abon, given by Asser, is almost the same as the Abona of the Ravenna Geographer, about which Sir J. Rhys finds a difficulty. THE WARS OF ENGLISH AND NORSEMEN 253 many of the inhabitants of the region over sea and compelled the rest to submission. But Alfred with a small host betook himself to the protection of woods and moors, where he main- tained his independence with much privation and trouble. In the winter of the same year the brother of Ingwar {Inwar — Asser) and Healfdan sailed from South Wales {Demetic region), where he had wintered, with twenty-three ships and came to Devon {Domnom'a). Here, before the king's stronghold of Cynuit,^ which was well protected on every side but the east, Alfred's men, by an unexpected sally at daybreak, gained a complete victory over the pagans, killing their king and over 800 of his men. In the Eastertide of 879 Alfred and his little host threw up a fortification at ^Ethelingaeg * {^Aelhelinga-etgge in the Chronicle = Athelney at the junction of the Tone and Parret). From this fortress as a centre " iElfred, with the vassals of Somerset ", Asser says (but the Chronicle gives, " of the part that was nighest of the Sumursaete ", Sumurscelna gen. plur.), made frequent successful raids against the pagan army in that region. At length, in the seventh week after Easter, he rode forth to Ecgbryhtes-stan * to the east of Selwood {Seal-ivyda, Chron.: * Cynuit. No word has suffered more than this at the hand of guessers. It has been made into the Domesday Comich^ now Combwich : and even into Cannington {Candetona in Domesday). It has also been identified with an ancient fortification in the parish of Abbotsham near Bideford, for which the name Kenwith Castle has been assumed. Nothing is known of the place, which was clearly in Devon. The efforts made to fix the site in Somerset (and to include the latter in Devon) are based upon the unsupported opinion that the fleet from Demetia came to co-operate with the pagan army at Chippenham. ** Arx Cynuit" would probably be represented in Devon by Dun-cynuit. Domesday gives a Donecheniu {Domnechenif in Exon. D. day) in Cornwall. " iEthelinga-eg means not *' the island of nobles ", but the island of the descendants of some one whose full name was some such form as JEXhtl- beorht, or iEthel-rasd. ' The place-names involved in this enterprise of Alfred have been the subject of many conjectures. As to Ecgbrihtes-stan^ Mr. Stevenson (Asser's Life of Alfred, p. 268) shows clearly that Brixton Deverill is not the place. 254 BRITISH PLACE-NAMES ^^ Seluudu^ that is, Sylva-magna in Latin, Coii maur in British speech/' Asser), where he met [" all ", Asser] the inhabitants of Sumurtun shire [paga], of Wiltun shire, and of Hamtun shire, all those who had not voyaged beyond sea for fear of the pagans. Alfred's host remained one day in camp here, and starting at daybreak came on to ^cglea [Igka and ^glea, in the Chronicle] where they camped for the night. On the All we know is that it was east of Selwood, but the extent of that forest in the ninth century we have no means of ascertaining. In the list of Cities attached to Nennius a Cair Pensa vel Coyt occurs (the reading of the Harleian MS. of the eleventh century : the Vatican MS. of the eleventh century has Pensa vel Coin) which has been identified by Mr. Kerslake with Penselwood (Pen-savel-wood), but no light is forthcoming on Savel. In the fourteenth- century MS. of the Welsh version of Geoffrey of Monmouth {^Red Book of Hergest : Rh^s-Evans, ii, p. 97) a Pen-hwyl-koet occurs, which is there said to be the name of Exeter. In an earlier MS. this identification does not occur. It is a pity, therefore, that Mr. G. Evans should have reprinted this MS. without the readings of the Dingestow and other earlier MSS. Pen-hwyl-koet was identified by Mr. Kerslake with Pen-savel-coit, It seems probable that it was intended to translate Pen selwood, since hwyl and its cognate Irish form Sedl represent Sel (contracted from i>^^/= Anglo-Saxon for the modern "sail"). Initial h is not organic in British speech, but represents an s (cf. Hafren and Sabrin-a), and wy is the representative of the older long e. Thus the British scribe in translating Sel-wood treated it as Sail-wood. Geoffrey has many similar grotesque translations of English names, for instance, his fanciful origins of Walbrook, in the City of London, and of Southampton. If Savel is the correct origin of the Sel or Seal in Selwood it is quite obscure. If the extensive pits in the neighbourhood were meant for granaries, the Irish Sabal derived from the Latin Stabulum (Pedersen, V. G. § 136.) has that meaning, and was possibly borrowed through the British. Pen-Sabal would mean *' Chief Granary". The Pe^is which gave names to Pen-selwood, Pen-domer, &c., seem to be grouped in the Anglo-Saxon Chronicle under the term Aet Peonnum. Domesday selvra, and selva, now Monk -silver, seem to show that Latin Silva was in early use for *'wood", and is probably the origin of this sel as well as the many "silver" streets (i.e. "Forest roads"), &c. Mr. Stevenson adopts the reading Iglea \_Ig-leah, where leah = wood] of the Chronicle, and identifies it with a place called Ilegh in 1439 near Sowley (that is, South-leigh) wood. An " Iley Oak " stood according to tradition near Lord Heytesbury's lodge at Sowley Wood. A Domesday Illega annexed to Knowles St. Giles was in South Petherton Hundred, but that would seemingly shift the locality out of the region of probabilities. THE WARS OF ENGLISH AND NORSEMEN 255 following morning early the standards were moved and the host came to Ethandun.^ Here JE\hed's host met the whole pagan army, which after great slaughter was put to flight, the escaping fugitives being pursued right up to their stronghold [Geweorc, which may have been Chippenham], where after a siege of fourteen days they were obliged to submit to ^Elfred's con- ditions of peace. The pagans gave hostages, swearing to leave iElfred's kingdom, while their leader, Godrum, promised to receive Baptism, which promise three weeks afterwards he fulfilled, when he and some thirty of the worthiest of his army came to Aller (Aire) over against Athelney. There Alfred received him from the sacred font as an adopted son, and the Chrism-fillet was removed eight days after at the royal vill of Wedmore. The pagan army removed [879] from Chippenham to Cirencester ['* Cirrenceas/re, called Catr ceri in British, which is in the south part of the Huiccii ",^ Asser] and there remained for a year, and in the same year a band of Wikings assembled at Fulham [FuUanhamm] on the Thames and took up their quarters there. Under the year 880, the Chronicle tells us, the heathen army which had been for a year at Cirencester proceeded to East Anglia, and occupied and divided the land.' The force which ^ Ethandtm. A place of this name was left by iElfred to his wife {^Cart. Sax. ii. 178). That this was a royal possession is clear, and as King Edgar granted Edington (Wilts.) to Romsey Abbey in 968, Mr. Stevenson thinks the proof of identity is not seriously questionable [Asser's Life, p. 273]. In a letter to the Athenaum (October 5, 1907) Mr. Stevenson goes anew into the question, and says in regard to Edington (Wilts.), " It cannot by any ordinary process of reasoning be shown to be an impossible site, and its acceptance does not involve any transgressions of the laws of history or philology or logic such as Bishop Clififord and his followers are driven to commit." * See note above on Cair-Ceri and compare Celtic Folk-Lore (Sir John Rh^s, i. 280, where Ceiri is connected with Cawr (Giants = Picts or Scots). ' Evidence from place-names of this and subsequent occupations is clear. Scandinavian elements in British place-names will be dealt with later. 256 BRITISH PLACE-NAMES had occupied Fulham proceeded to Ghent {Gent) in Flanders \^Fr one-land'] and fought with the Franks, proceeding afterwards (882) far up the Meuse (McBse). In the year 882 Alfred had a sea-fight with the Danes, in which he captured four of their ships. The next two years the heathen army were occupied on the Continent, sailing up the Scheldt as far as Conde {CundothY and the Somme to Amiens {E?nbenum). In the year 885 the aforesaid army was split into two divisions, of which one proceeded to the east, and the other began the siege of Rochester {Hro/es ceastre\ throwing up a work against it. yElfred coming to the relief of the besieged city, the enemy made for their ships, having been deprived of their horses. Another sea-fight occurred at Sturemouth (Essex) the same year, in which, returning from a victory over one fleet, Alfred was beaten by another larger one of the Danes {Deniscan), Carl (Charles the Fat), the King of the Franks, is said, by the Chronicle, to have recovered at this time virtually the empire of Charles the Great — except Brittany (Lidwi'ccium, dative plural of Lzdwiccias)? The Chronicle records under 886 that the heathen army pro- ceeded up the Seine {Sigene = Sequana) and settled for the winter. At the same time Alfred occupied London {Lunden- hurg)y where all the English kin, except those who adhered to the Danes, rendered obedience to him.* After having secured the place (burg) he gave it to his son-in-law ^thered to hold. The Anglo-Saxon Chronicle records under 887 that the ^ Condoth is the Anglo-Saxon form of Condate, a common Romano-Celtic word for confluence. Ambiani was the ancient name of the people whose capital was Samaro-briva, that is, the bridge on the Samar-a, now the Somme. 2 See p. 159, where the Letewicion of Nennius is cited as the origin of the Anglo-Saxon form. ^ Asser says that -Alfred, after the burning of cities and slaughter of people, restored London City in a worthy manner, and rendered it habitable. THE WARS OF ENGLISH AND NORSEMEN 257 heathen army proceeded through the bridge at Paris {Part's)^ up the Seine, across to the Marne, and up the latter river to Chezy {Can'ei for Caziei, that is, Asser adds, " royal vill " : it appears in an early form as Casiacum), where, and also ** within " the Yonne (lona), they settled for two winters. It was about this time that Alfred (as Asser tells us) constructed two monas- teries, one at Athelney (approachable only by boats) and the other, a nunnery beside the East gate, at Shaftesbury {Sce/ies- burg)} In 890 the Chronicle records the death of King Godruniy Alfred's godson, who, according to the Annals of Si. NeotSy had established his rule in East Anglia, after King Edmund — the same authority adding that he was buried at Headlega (Hadley, Suffolk). Under 891 we have an account in the Chronicle of the arrival, in a skin-covered coracle without oars, of three Irishmen (Scollas) on the coast of Cornwall {on Cornwallum). They started out with no aim but that of making a pilgrimage, and about the seventh day came to land. Their names are given as Dubslane, Maccbethu, and Maelinmun. They proceeded soon after their arrival to King Alfred. Asser, in dealing with the charitable acts of Alfred, after staling that Alfred helped monasteries and churches in Mercia, Cornwall, Gaul, Armorica, ^ Sceftesburg appears in the Chronicle and Charters as Sceaftesbirig. The first element is the genitive of a personal name Sceaft, shortened from some such form as Scef there {Cart. Sax. i, p. 113) or See/twine (ibid., p. 164) : Sceaft =» spear. Shafton, another name for the place, is clearly from the same Sceaft. In the Latin form of the History of Geoffrey of Monmouth a certain King Hudibras is said to have built Motint Paladur, now Shaftesbury. The Welsh version (in the fourteenth-century MS.) calls the builder Run Falatyr-vras {bras = stout) and the town ** Caer-vynyd Paladur {niynyd = mount) which is now called Kaer Septon '\ Sceptonia is the form given in the Latin version of the Anglo-Saxon Chronicle and hence the later Shafton. The town was walled in Asser's time, as he mentions the East Gate. Bin port, for " Binnan Port " ( = within the Gate), is still one of the streets there. 258 BRITISH PLACE-NAMES and Northumbria, adds, "and even sometimes in Ireland {Hyhernia) also," a fact which may help to explain this voyage. We have in the Chronicle under 893 an account of the arrival at Limenemouth ^ from Boulogne {Bunnan representing Bononia, the Roman name for Gesoriacum\ see p. 63) of 250 ships of the heathen army. This port, adds the Chronicler, is in the eastern part of Kent, at the east end of the great wood that we call Andred^ which is 120 or more miles long and 30 miles broad. They took their ships four miles up the river from its mouth, the Chronicle adds, and there stormed a badly defended stronghold. Then they threw up a strong fort at Apul-dore {Apuldran, Annals of St. Neots, = Apple-tree). In the year 891 (in the Chronicle 893) the Annals 0/ St. Neots records the arrival of Hasteng [Hastengus) with eighty ships at the mouth of the Thames, on one bank of which, at Middel- tuna (Milton), he threw up a fortress, and on the other bank, at Beamfleot (Ben-fleet, Essex), a similar stronghold. The same annals state under 893 that York was captured by the Northmen, the bishop — named Sebur — escaping. In the same year, the same authority adds, " Alfred fought with the Northmen in a place called Fearnhamme (Farnham, Surrey) and put them to flight. Pressed by Alfred they crossed the Thames at the Colne^ river to an island and were there ^ This is the "Ad portum Lemanis" (locative plural) of the Ant. Itin. Lemana is given as the name of the river, which became on Saxon lips Limen-ea {ea — river), now Lympne, Kent. The coast has changed much since Roman times. 2 Colne appears frequently as a river name, and especially in districts where Roman settlements are manifest. Although it cannot indicate that such districts were Coloniae, it may preserve some Anglicized form of that Latin word with its generic meaning of "cultivated place", "farm". Colonus (from Cold) is properly a farmer. The earliest form encountered seems to be Colen as in Colenceaster {Cart. Sax. ii, p. 359). Compare Colen-ea (now Colney, Herts., ibid., i, p. 339). The Cylen in Lindcylne (in Kent, Cart. Sax. i, p. 483) cannot well be the same word ; nor can Cunugl-ae {Cart. Sax. i, p. 240), which Mr. Birch identifies with the Gloucestershire Colne, contain the origin of the word. Cunugl is the THE WARS OF ENGLISH AND NORSEMEN 259 beset by JElhed. Eventually they went into Essex, ^thered, who had been left in charge of London, proceeded with his forces to Beamfleot and captured the heathen army's stronghold there, together with much spoil — including the wife and two sons of Hasteng, whom he brought back to London, and presented to Alfred, but was ordered to restore them at once. Hasteng, however, according to the Annals 0/ Si. Neois^ repaired the fortress at Beamfleot, and threw up another at Shoebury (Sceobyng), the army, which had been at Apuldore, joining his forces there, and also a large number from East Anglia and Northumbria. The Chronicle then records the ascent of the Thames by the heathen army, and their progress westward until they reached the Severn, up which they proceeded. iEthered, having in the meantime collected an army, pursued them to Buttinga-iun on Severn shore {stathe). Here he be- sieged the heathen army and, compelling them by hunger to give battle, won the victory, the remnant of the army seeking safety in flight.* Having reached their head quarters in Essex, and placed their women and ships in safety, the heathen set out, with reinforce- ments, across England, and reached at length a waste Chester in Wirhealum (locative plural of Wirheall, now Wirral), called Lega Ceaster (Chester on the Dee). Hence they proceeded to North Wales. In the following year (895) they proceeded to Essex, to an antecedent of Knoyle in Dorset {Cart. Sax. iii, p. 16). The origin {Cnoll) given in the New English Dictionary for Knoyle is, therefore, not correct. A river Calne occurs in Scotland. ^ This is the raid recorded in the tenth-century MS. of the Annales Cambriae under 895. " Nordmani venerunt et vastaverunt Loycr ( = Loegria, the Welsh name of Mercia probably, but later applied to England) et Bricheniauc (Brecknock) et Guent et Guinn liguiauc." Guinn Liguiatic is in Lib. Land, Gimliviuc and later Gwent loog. Liguiauc seems a personal name, Cymmrodor. Mag. ix. 167. As the Danes are distinctly said to have gone up the Severn, and as the Brut tells us that in 890 (893) they came a second time to Montgomery, the place is evidently Buttington, opposite Welshpool, and not Buttingtontump at the junction of the Severn and Wye. R 2 26o BRITISH PLACE-NAMES island called Meresig^ Mereseg (Mersea, Essex), whence they sailed (896) up the Thames to the Lygan (river Lea), and up this for twenty miles. King Alfred driving them thence they proceeded to Cwat-brycge ^ on Severn, which they fortified. In 897 they went across the Channel to the Seine. In 898 Rollo, the founder of the Dukedom of Normandy, besieged Chartres {Civtlas Carnotiensis), but had at last to raise the siege. The Annals of St, Neots under 900 records the death of Alfred and the succession of his son Eadward, also the defection of the latter's cousin ^thelwold, who withdrew to Northumbria and joined the fortunes of the Northmen there. When the tenth century opens we find Eadward, the son and successor of Alfred, reigning in Wessex, his sister ^thelflaed and her husband ruling in Mercia, where their authority was limited seriously by the Danes, who bore sway also in East Anglia and Northumbria. ^thelwald, Eadward's cousin, before taking 1 Prof. Earle pointed out that there is still a Quat and Quatford near Bridgenorth. Mr. Eyton {^Antiquities of Shropshire, vol. ix, 143-150) derives Quat from Coed (early form Coyt = forest), pointing out that the Wyre, Kinver, and Morf Forests embrace all this district on both sides of the Severn. In an original charter of A.D. 736 \Cart. Sax. i, p. 223) we have the early names of two of these forests, Cynibre (which has become "Kinver", passing through the forms Chenefare [dative, Domesday'] and Kinefar, Testa de Nevill) and Moerheb (which has become **Morfe"). Moerheb is interesting as it is probably the same word as the twelfth-century More/ (now Moray, in Scotland : Murief in Geoffrey of Mon.), representing possibly an earlier *morapa, apa ( = Lat. aqua^ Old High German aha) being, according to Dr. W. Stokes, the Gaulish form of water, giving ep-^ ef-, -aff in the later termination of river- and place-names. Compare also the Morbi-o of the Not. Dig. Moravia in Germany also was in the ninth century Mor-ava and Mor-aha, implying that *Morapa was the original form. Cynibre is also interesting as showing the possible source of some of our place-endings in -far and -ver, and may represent an early *Cunobriga = high-burgh. Bridgenorth does not appear in Domesday y but its early name Bricge occurs in the Anglo-Saxon Chronicle in A.D. 912. The Norse Bryggja, = a landing-place, may be the origin of the bridge in Bridgenorth — the ''north" is obscure if it be not a corruption oiMorf. See p. 61. THE WARS OF ENGLISH AND NORSEMEN 261 refuge in Northumbria, had, as the Chronicle records under 901, seized without the King's leave the residences at Wimborne [cet Winburnan) and at Christchurch {cet Tweoxn-eam, that is, " between the rivers '\ known in the twelfth century as Christe Cerce). King Eadward moved against him and encamped at Badbury Rings {cei Baddan-hyrig^\ ^Ethelwald, who had escaped by night to Northumbria, where he was received as king, came back later on with a fleet to Essex. Under 903 we have the record of the Consecration of the New Minster at Winchester, wherein were placed the relics of St. Judoc,^ and wherein Grimbald the priest, who had come over from the Continent to help -Alfred, and had died in 903, was buried. Under 905 the Chronicle relates how iEthelwald, having induced the Danes of East Anglia to break the peace, ravaged Mercia as far as Cricklade {Crecca-gelade^)^ where, crossing the * Baddan-byrigf Badbury Rings. This seems to be, according to the mileage, the Vindo-Cladia (variant Vindo-gladia) of the ////;. Ant. Vind-o is represented now by the Welsh Gwynn = White ; Irish Finn : cladia (= Kladia?), as Dr. W. Stokes has pointed out {Wortschatz der Keltischen Spracheinheit, p. 82), contains probably the stem Klado (« ditch) pre- served in Welsh Cladd (^Clawdd)^ and Irish Clad. The compound would thus mean " White-ditches ", a natural term for a stronghold with a series of concentric rings cut in the chalk. Baddan is the weak genitive of Badda, a name which occurs frequently, and is probably a shortened form of some such name as Beadn win = " war friend ". For Winbuma, which seems also to contain Vindo, see p. 194. Tweoxn- earn = '' between the rivers " {£a « river), is declined irregularly and ^am is the dative or locative plural. " St. Judoc was a Breton saint of the seventh century. The name is a diminutive from some full form such as Jud-gtial. It has become, both in Brittany and in Cornwall, Just.^ as in St. Just^ see Les Petit s Bollandistes. The new minster was removed to Hyde Abbey in the twelfth century. ^ Crecca-gelade is now Cricklade. Gelade (dative) is " a passage ", and is a collective form of Lad = way — preserved in the numerous "Lodes" on the Severn, e.g. "Saxon's Lode". Crecca is possibly connected with Welsh Craig = ** rock", or with Cruc =» mound. on Bradene. The name Bradcn occurs in Domesday in Bradenc stocky 262 BRITISH PLACE-NAMES Thames, the heathen army raided in and about Bredon forest (on BrcBdene), and then set out homeward. Eadward overran all their country between the dikes {dicum) and the Ouse {Wusan) and as far as the fens {/ennas) northward. On the king's homeward journey the Kentish men {Kentiscan) lagged behind contrary to orders, and were attacked by the Danes, who gained the victory, after much slaughter on each side. This battle seems to have been that fought at Holme (somewhere in East Anglia), which is recorded in one of the MSS. of the Chronicle under an earlier year. In 907 the Reeve of Bath died (cet Badum), and in the same year peace was concluded both with the East Angles and Northumbrians at Yttingaforda. For the next three years we have records of the construction or repairs by the English of strongholds at strategic points in connexion with the move- ments of the Danes. Thus, that at Chester (Ltgcester^ that is, Castra Legionis) was renewed in 907. There was a fight between the Angles and Danes at Tattenhall {Teotan-healh'^), and there is also a Domesday Bradene (dat.) in Somerset. Its modern form of Bredon, a name found both in Wales and in the West Country, suggests that it represents possibly an early Brigo-dun-um, that is, ** hill fortress.'' Between the dikes and the Ouse. " The dikes ", some of which seem to have been in existence in Roman times, were evidently a known term in the tenth century. The Wusa is obscure. It may be a modified form of some Celtic word, or it may be connected with Wdse (mud), compare Wos = wet. Early forms occur in Charters, e.g. Usa {Cod. Dip. 480), Use (ibid., 621). Yttingaforda has not been identified. It may be a mis- reading for Ultingford and be thus represented by Ulting in Essex. ^ Teotan heal contains the weak genitive of a personal name {Teota) contracted from a full form such as Theod-win. Heall is used for a large structure where a chief resided, as well as for ** shed " generally. cet Bremes byrig has been identified by Mr. Plummer with Bromesberrow near Ledbury, Hereford. In favour of the identification is the fact that Bromesgrove represents an earlier Bremesgraf^ but berrow is not the usual modern equivalent of the dative of burh. To Scergeate. Mr. Taylor {Danes in Gloucestershire^ would make this Shrewsbury, but this is linguistically impossible. The place is most probably the Scrget {Codex Diplomat. 1199, where the editor makes it THE WARS OF ENGLISH AND NORSEMEN 263 Staffordshire, in 910, in which the Angles were victorious, and in the same year ^thelflsed built the stronghold {burh) at Bromesberrow [cei Bremes hyrig) near Ledbury, Herefordshire. In 9 1 2 she arrived at Scergeat {to Scergeate) on the eve of the Invention of the Holy Cross, and constructed the burg {burh) there, and in the same year that at Bridgenorth {Bn'cge). In the next year, 913, she came with all the Mercians to Tam- worthy {Jo Taviaweordige ^) and built the burg there, and after- wards that at Stafford {cBt SicEfforda). King Eadward in the same year ordered the northern burg at Hertford (cei Heorot- fordd) between the rivers Maran, Beane, and Lee {betweox Memeran Beneficcan and Lygean) to be built. When he was afterwards at Maldon {^Mceldun) in Essex he constructed the burg at Whitham (Wi/ham), in the same county, whereupon many of the people who had been previously under the Danes Sc[i]rget). It was near Beenham in Berkshire, and close to the Kennet {Cynetd) river. * Tamaweorbige appears in the Charters as Toman-worthtgy involving probably the genitive of a personal name Toma or Tuma^ which occurs. Stafford has been previously explained. The river Memera (variant Mara) is obscure, as is also Bene ficca. Lygean^ now the Lea^ from the variant Ligette, looks as if the n were organic and suggests that the word is connected with the common Celtic river-name, Logana. at Eadesbyrig is identified by Mr. Taylor {Danes in Gloucestershire) ^ probably rightly, with Eddisbury, Cheshire, but the same writer's identifica- tion of z'«/(" silent"), earlier Taguel. 3 Bradan relice. This has been dealt with under A. D. 918. * Faversham was Fefres-ham in 811 {B. M. Charters and Rolls)y and Febresham in 858 (ibid.). In Domesday it was Faversham. The first element is a personal name in the genitive, but there seems to be no THE WARS OF ENGLISH AND NORSEMEN 303 place-names mentioned in this portion, but none of importance. Under 1087 we have mention of Berkeley {Beorclea) * and the Castle of Tonbridge {Tonbrucge), and in 109 1 occurs " Lothian " (into Lothene)^ which is declared to be in England, and into which King Malcolm of Scotland led his troops. The hallowing of the minster at Battle {Baiaille) is recorded under 1094. In 1095 we have an account of an expedition made by William against the Welsh, who had destroyed the Castle of Montgomery (Muntgumni) in Domesday it is Muntgumeri) and slain the garrison. The king's troops proceeded as far as Snowdon {Snawdun : in Welsh Mynydd Eryri), which they reached on All Saints' Day, but, the enemy having retired into their fast- nesses, the king had to return. In 1098 a prodigy is recorded at Finchamstead, Berks., where a fountain of blood was seen, and in the same year Earl Hugh is stated to have been slain by foreign Wikings (Uiwikin- gan) in Anglesey {Angles ege). In mo, the Chronicle tells us, Henry I spent Easter at Marlborough (Mcerle-beorgY In a scrap of another MS. (H), there is an account (under 1114 ?) of a grant by Henry of the Earldom of Northampton Shire (North- ham tun Scire) to David of Scotland, who was the queen's brother, and also grants of Abbacies at Ramsey {Ramesige), Thorney (Thornige\ and Cerne Abbas (Cernet). The same MS. records also the death of the Abbot Nigel at Burton-on-Trent similar Saxon designation. It may possibly be a case of the survival of the Latin Faber, = Smith, in a thoroughly Latinized district such as Kent must have been. * Beorc-lea contains BcorCy ■» birch, and leah^ in its original signification of wood. * Lothenef= Lothian, has already been discussed. Freeman contends {Norman Conquest ^ vol. i, note B) that Lothian was given up to Scotland after, and as a result of, the battle of Carham (see back). The territory covered by the name is vague. It probably extended south of the Tweed, and thus confirms the text. ^ Mcerl'beorgy = Marlborough, shows that beorg ( = hill) becomes, as else- where in the South, borough. The Meerl is obscure. The Domesday form is Marleberg. 304 BRITISH PLACE-NAMES (ByrtuTi)^ and the burning of Chichester {Cicesire) and its Minster. At Pentecost, the MS. continues, King Henry was at St. Albans {St. Albanestowe), from whence he led an expe- dition into Wales, where he built castles, and the Welsh kings submitted to him. Returning to Winchester he bestowed, among dignities to other clerics, the Abbacy of Muchelney {Myclanyge) on the monk Ealdulf. MS. E tells us that in October, 1114, King Henry was at Rowner {Rugenore: in Domesday it is Ruenore) in Hants, and that he kept, in 11 16, Christmas at St. Albans, and Easter at Odeham {Wudiham\ Hants. In 11 21 he spent Christmas at Brampton (Bramtun), Hunts, and in 11 23 the same festival at Dunstable {Dunestaple), and went thence to Woodstock ( Wudestoke), In the latter year he appointed a canon of St. Osyth's (Cicc : in Domesday, Cice\ Essex, archbishop. In 11 24 Ralph Basset and the king's thanes held a meeting at Hundcot {Hundehoge : in Domesday it is Hune- coie), in Leicestershire (Letheccestresctre : cf. the Domesday form Ledecestre). In 11 25 John Bishop of Lothian [Lothene) is men- tioned. In 1 132 the Prior of St. Neots {St. Neod) is named, and in 1 135 is recorded the death of King Henry, and his burial at Reading {Redinge), Stephen of Blois succeeding to the throne. His first act was to take possession of the castles of those noblec> who were disaffected, among these are named the castles of Rock- ingham {Rogi'ngham), Cottingham {Cottngham), Easton {Esiun), Irthlingborough {Hyrtlingheri\ Stanwick {Stanewig), and Ald- wyncle {Aldewingk), all in Northants. The defeat of David, King of Scotland, at the battle of '* the Standard " is recorded under 11 38, and Stephen's death under 1154. The consideration of later records containing place-names is not worth pursuing, as the forms therein presented vary but little from those now in use, and the new terms introduced by the Continental monastic orders, such as BeauUeu, Rievaulx, Jervaulx, Sec, explain themselves. INDEX Ahallaba {Avalana, ?Papcastle), 99, 100, loi, 115, 272. Abbandun (Abingdon), 40, 2*j*j n.^ 379, 292. Abbercurnig (Abercorn), 186. Abbots Anne {Domesday Anne), 144 «. Aben, the river, 219. Aber, meaning of, 186 n, Abercorn. See Abbercurnig. Aberdeen. See Devana, Abergavenny. See Gobannio. Aber-guilly, 250 «. Abingdon. See Abbandun, Abon (?Sea Mills), Abone, Abona (river Avon), 22, 23 w., Ii2,ii3«., 252. Ab-os (or Aboub-os), the river, 98. Abravann-os river-mouth, 93, 95. Ac beara, 241 «. Acemannes Ceastre, 378. See Bathan ceaster. Aclea, Acley (Ockley), 196, 197 «., 242. Aclea (Aycllffe), 234 «. Ac-leag (Oakley), 215. Acviodce (i^ibudes Islands ■= Heb- rides), 90. Ad Ansam (? Stratford), no, in n. Ad Baruae, i8r. Ad Lapidem (Stoneham), 183. Ad Murum (Walton), royal vill called, 178. Ad Pontem (?E. Stoke-on-Trent), 109. Ad-Tanatos (Thanet), 91. See also Tenid. Adamnan's monastery at Hii, 191. Adel, 169 «. Adescancaster. See Exanceastre. "Adgefrin," the royal vill called, 168. Adour, the, 114. Adredes-eya, Andredes-emg, 203, 206. iEcglea. Sec Aclea and Iglea. iEd wines clif (" Eldunum near Mel- rose") now Eildon, Eildon hills), 196. Aedan, King, 162. iEgelbeorht {^gelbiryth)^ a Prank- ish Bishop, 175. /Egelesbyrig, jEgeUsburg (Ayles- bury), 147, 267. -^gelesthrep {^Mgelesforda^ Aylei- ford), 139, 140, 141, 287. ^Idenham (Aldenham), 218. iElfetee {El-vety Elfetu, Elmet), 196. iElfheah (Archbishop), slain, 286; his body taken up, 288. iElfric selected Archbishop of Can- terbury, 281. /Eschongas, 287. See also Hangra. JEsica (Great Chesters), 99, 115. .Esp hangra, 287. i^t-Treocum (Utrecht), 190 «. ALthelingadene^ 283. iEthelingaeg (Athelney), 253. i^thelred crowned at Kingston, 279 ; goes to Normandy, 286; at war with Cnut, 286. -^Ethelric, Bishop of Worcester, buried at Ramsey, 289. iEthelstan's grants of lands, 162; death of king, 273. iEthelwold made Bishop of Win- chester and rebuilds Ely, 277; "father of monks," 280. Aetius, the third Consulate of, 127. Afallen (Avallon), 198. Afene-mutha (Avonmouth), 294. Agelberct (Agelbeorht), Bishop, 182. Ageloco (Littleborough), io8,io9«., 172. See also Segeloci, and Tiouulfingacsestir. " Agricola, a Pelagian," 124. •aha, names ending in, 68. Aibouda (?Uist), 92. Aiboudai, the, 92. 3o6 BRITISH PLACE-NAMES Ail-Cluade. See Nemthur. Ailcluit (Alclut, Dumbarton, Dun- Brettan), 84, 152 «., 153. "Ail-dun," 196 «. "Aircol, son of Trephun," 56. Airer-Gaedhel (Argyll), 84. Akeburg, 171. Ake-mann CcBster (Bath), 41. See Bathan ceaster. Akemann Street, 41. Alauna, Alaun-os (Alnwick, Alne, Alum), 97, 99, 203. Alauna Silva, 96. Alaun-os (? Axe), the river, 96. Alba, extent of, 88. Alban (Scotland), 9, 76, 92. Alban, martyrdom of St., 35, 60, 69, 70. Albans, St. (St. Albanestowe). See Verulamium. Albion (Albiona), 88. See also Alvion. Alcester. See Alne-Chester. Alclet (Hak-let, Auckland), 234 «., 237- Alcluith {Petram Cluithe). See Ail Cluade, Ailcluit, and Alt Clyde. Alcuin, 124 w. Aldborough. See Isourion. Aldewingle (Aldwyncle), 304. Aldhelm, 166. Aler (in Alercumb), 207. Alfred's Description of Europe, 225. Alh-mund, 217. Alione {Alone\ in, 116. AUectus, 62, 64, 65, 66, 67. Aller. See Aire. Allerdale (Alner-dale), 273. Alne. See Alauna. Alne-Chester (Alcester), 203. Alner-dale. See Allerdale. Alney Island. See Olan-ige, Alnwick. See Alauna. Aire (Aller), 255. Alt Clyde, 245. Alum. See Alauna. Alvion (Albion), 92. Ambiani (Amiens), 79. Amboglanis. See Ambo-Glanna. Ambo-Glanna (Burdoswald. ? Gaba Glanda), loi, 115. Ambresbyri (Amesbury), 382. Ambrones, 208. Ambrosius Aurelianus, 122, 149. Ambrosius (the Myrdin of Geoffrey of Monmouth), 136, 137 «. Amesbury. See Ambresbyri. Amiens. See Ambiani and Em- benum. Amwythig (Shrewsbury), 285 «. Anava (Noe), the river, 117. Anavio (? Brough), Anavione, 109, 116, 117. Ancalites, 10. Ancaster. See Causennis. Andeferan (An-dever, Andover, Anton-dover), 97, 144 «., 281. Ande-gav-ia (Angers, Anjou), 109 n. Anderitos (Anderida, Andredes Wald, Andereton, Anderitum), 113. See also Andredes-Cester. Andover. See Andeferan. Andred (Andredesleage), the forest of, 143, 195, 218, 258. Andredes Cester (?Pevensey, Pevt- nisei), Anderida, 143 «. Andredes-eme. See Adredes-eya, Andredes Wald. See Anderitos. Andret, 143 n. Andros (Adros), 90. Aneurin, poems of, 154. Angen-geot-ing, 214. A«gen-lab-es hamm, 214. Angers. See Ande-gav-ia. Angles, 128. Angles-ey (Angles ege), 229 «., 303. See also Mona. Angra and Hanger in place-names, 214, 286 «. Angra (Ongar), 287?/. Anlaf besieged by Eadmund, at Leicester, 274; arrival at London, 281. Annan, the river, 95. Annport, 144 «. Anse, win. Anstige (Anstey), 211 «. Antivestaion. See Belerion. Antona, the river, 22, 117 Note F, 144 «. Antondover. See Andeferan, Anthun, 56. *^-apa^' names ending in, 68, 69. INDEX 307 Apostle's Oak (the) on Abberley Hill, 168. Appenzell, 190. Apuldran (Apul-dore, Appledore), 187 «, 258. Aquamania, 41. Aquis (Buxton), 116. Aqiiis Solis. See Bathan ceaster. -drd, in river-names, 187 «. Arbeia (? South Shields), 114. Arcencale {Ercal), 266 n. Arcunia, 266. Arcuulf, Bishop, 191. Areconio, Ariconium (Deanery of Erchen^€i^^ 107 w., 112, 139 «., 266 «. Aregenua. See Arguenon. Argento varia (Horburg), 151 «., 252 «. Arguenon {Aregenua)^ the river, 155 «. revll. Argyll. See Airer-Gaedhel, Armagh, Book 0/, I28«., I29«. Armorica (Brittany), 9,61, 122 «.; emigration to, 154; state of, be- fore sixth century, 1 54 ; Celtic names in, before and after immi- gration of British, 155; Gaulish vrords in, in fifth century, 156 «. ; British words imported into, 157, 256. Aro Ssetna, 222. Arran, 171 «. Arras, 12. Artbrannan, 50. Arthur, King, place-names in con- nexion with legendary, 149; his victories, 149. Arthuret, 84. See Westmoreland. Artois, 12. Arviragus, 137 n. Arwa (Orwell), the river, 2 87. Arxy the Citadel of Vortigem, be- side the river Teibi, 1 76 n. "ArxCynuit", 253 «. Arx-Decant-orum, Arx Deganhui^ (Deganwy), 94, 239. Ascan-mynster (Axminster), 196. Asclepiodotus, the Praetorian Pre- fect, 62, 64, 66. Ashbourne, 187 «. Ashdown. See Assandun. Ashingdon. See Assandun. Ash-over (Asher) 210 «, Assandun, -^scesdun, ^scesbtirh (Ashingdon, Ashdown), 346 «., 284, 287, 288. Atclete. See Etclete. Ath (Adwy = ford). See Path. Athelney. See iEthelingaeg. AthoU. See Fotla. A[d]trebates, the, 9, 12. AtrebcUioi, the, 103, 104. Atrebatum (Silchester), 9. Attacotts (Atecotts, Atticotts, Atta- cotti), 6i«., 78, 79, 81; harass Britons, 11^ n. Auckland. See Alclet and Etclete. Augustine, St., arrival of, 148; meeting-place with the British Bishops, 164. "Augustine's Oak," 164; ?near Cncklade, 167 ; ada. Auray {fferio-s), the river, 155 n. "Aust Cliff," 303. Aust on Severn, 167. Aust Passage, ii^n. Austlin {Austiriy "Austins rick," Penryn Aws/in^, 202; ?** Au- gustine's Oak," " Austin's Well," *' Aust Cliff," 202, 203. Avalana. See Aballaba. AvanniiSy 95. Avon. See Abon. Avonmouth. See Afene-muiha, Axa (Axe, Exa), the, 96, 199. Axeloduno (Axelodunum, Uxellu- damo, ? EUenborough or Hex- ham), 99, 100, loi, 115, 174W. Axium, the, 96. Axminster. See Ascan-mynster. Aycliffe. See Aclea. Aylesbury. See -^Egelesburg. Aylesford. See /Egelsthrep. Aynburg, 235. Baccan-Celde, 226 «. Badbury. See Baddan-byrtg. Baddan-byrig (? Vindo-gladia, Bad- bury), 16, 113, 261. Badecan - Wiella {Badecanwiellan, Bakewell), 147, 268. Badonic Mount. See Caer Badon. Baenesingtun (Benson), 147. u a 3o8 BRITISH PLACE-NAMES Bagillt, 46. Bakewell. See Badecanwiellan, Bald wines Healh, 217. Balsboro. See Balteresberghe. Balteresberghe (Balsboro, Baltones- boro), 201. Bamborough. See Bebbanburh, Dinguoaroy. Bampton, 175. Ban-Chor (Bangor, Bancornaburg, Bangor-is-coed), 130 «. ; meaning of, 167; "Bangor" applied to Glastonbury, 167 «.; Bangors in Wales, in Brittany, and in Scot- land, 167 n. Bancornaburg. See Ban-Chor. Bangor. See Ban-Chor. Bangor Wydrin in Afallen (Avalon), 198. Banna, 99, 100, loi. Bannagher, 167;?. Bannatia, loi. Bannavem (Bannaventa, Bannavento, Bennavento), 33, 34, 106, 107 w., 109, no, i29«., 130 «. See also Isannavantia. "Bannavem Tabemiae," (Nentre, Nentria or Ventre), 129 «. Banno-vallum, 72 n. Banuwille (Banwell), 249. Bapchild (Baccan-Celde), 226 «. Barbury Camp. See Beranburg. Bardaneu {Bardanege, Bardney), 174. Bards, the, 54. Baregain, 272. See also Burrens. Bare-oak-shire, 11. Barganes lapidum, " Sbeep-fold[s] oiStone[s]," 28. Barham. See Bioraham. Barking. See Bercingon. Barnbougle, 86. Barwick-in-Elmet, 1 70 n. Basincge (Basing - stoke, Basing- werk), 212. Basings {Basengum), 247. Bassas, the river, 150, 152. Bataille (Battle), 303. Batch (Bach), 230. Bat-chet (Bathgate), 86 «., 181 n. Bath. See Bathan ceaster. Batham Gate, 117. Bathan ceaster {Ba^um, Ealdan- byrig, Hydata Thermae Aquis Solis, Bath), 9, 104, 113, 148; hot baths at, i6\n.y 262, 278. Bathgate. See Bat-chet. Battersea involves " Beadu-rich,*' 245 «. Battle. See Bataille. Battle of Stamford Bridge {Stanfora Brycg\ 301. Bdora, 116, 239 «. See also Dote. Beadorices Uurthe (St. Edmunds- bury), 245 «. Beamfleot (Ben-fleet), 208, 258, 259. Beandune, battle of, 175. Beane, the river. See Beneficcan. Beardanig (Beardan-eu, Beardan- ige), 194. Bearruc-scire (Berkshire), 10, 284. Bebban-burh (Bebbanbyrig, Boily- banburhy Bamborough, Dinguo- aroy)^ 146, 15T«., 269, 281. See also Dumono-varia, Argent varia. Beber (iu Beverley), 208. Bech (Batch, Bach), 230. Bedanforda (Bedcanforda, Bedean- ford, Bekeford, Bydanford Bede- forde, Bedford), 147, 266, 285. Beddan-haam, 213. Bede-hus (Bettws), 227 «. Bed-ern, 227 «. Bedford. See Bedanforda. Bedwig, 142. Bedwin. See Biedan Heafde. Beeston. See Bovio. BeferpyttaSf 204. Beferith, 11. Beggares-thorne^ 204. "Beggary-island," 205, 236 «. BekeHa {=parva Ybernta), Bee- eriu, Bickery, Bickworthy, 205. Belerion (Bolerion Antivestaion). See Fenwiht Steort, Belgae, 8, 9, 10, 104. Belgium, 61. Belisama, estuary of the, 93, 95. Bellum Cochboy^ i74«. See also Maserfelth. ''Belly-moor." '^tt Bolgros. *' Benavent " (in Catalonia), 33. *' Benaventa " (S. of Astorga), 33. INDEX 309 Berhichor, 167 «. Beneficcan (Beane), the river, 263. Benesingtim (Benson), 196. Beneventa, 129 n. '* Beneventana Civitas," 34. " Beneventenreut " (in Germany), 33. " Beneventum" (now "Benevento", in S. Italy), 33. Ben-fleet. See Beamfleot. S. Benignus (Beonna), 198. Benson. vSee Benesingtun. Ben well Hill. See Conderco. Beonna. See S. Benignus. Beorclea (Berkeley), 303. Beorg-ford (Berg-ford, Burford), 195, 210. Beorh - ham - stead (Berkhamstead, Berchamstead), 30a. See also Berstead. Beowi, 142. Beowulf, poem of, 130, 131, 143. Beranburg (Beran burh, Barbury Camp), 146; battle of, 160. Bercingon (Barking), 214. Beret, the Northumbrian leader, 185. Bcrctfrid fights Picts, 193. Berct-hun, Abbot, 188. -bere or -bera in place-names, 241 «. Bere Foresi, 241 n. Bere-weg, 214. -berga in place-names, 341 ;/. Bergiona (Iberion), 88. Bergos (Berehu, Burhou), 88 n. Berigon, 88 w. Berkeley. See Beorclea. Berkhamstead. See Beorh-ham- stead, Berkshire. See Bearruc-scire. Beniicii, 26, 28. Bernwood Forest. See Bymewudu. Berricen, 88 w. Berstead, 302. Berwickshire. See Prender^oj/. Betendune. See Vetadi^n. Bettws (Bedhus), 227 w. Beverley. See Beber, and In-dera- uuda. Beverston. See Byferestan. Bibracte (Autun), 11. Bibr-ax (Beuvray), 11. Bibrdci, the, 10. Bickery. See Bekeria. Bickworthy. See Bekeria. Bieda, son of Port, 144. Biedafi Heafdei^. Bed win, Bedewind), 123;/., 217. Billinga byrig, 215. Billingahoth, 237. Billingsley. See BylgesUge. Bilmiga, 222. Binchester. See Vinovia. Bindon, 175. Bioraham (? Barham), 219. Birc hanger , 287//. Birch-over (Bircher), 208, 210 n. Birdoswald, loi. Birinus, 175. Birrens Camps (Burrins Camps), 28, 272 n, Biscopes uuic (Bishop's dwelling), 215. Bishop Felix (Dunwich), 173. Bishopricks in 731, 193. Bishops Waltham. See Wealiham. Bittern , Hants. See Clausentum and "Butter -Bittern". Bitton. See Trajectus. Bituriges (Bourges), Britons received by the Emperor in, 160. Blabo, Blabolg, 106 n. Bladney. See Bledenhithe. Blaecca of Lincoln, baptism of, 171. Blcethgent, brother of Griffin, the Welsh king, 299. Blato-bulgio (Blatobulgium), 105, 106 »., 298 n. Bledenithe (Bled one, Bleadon, Bladney), 200. BUdgint, 299;/. Bledone (Bleadon), 200. Blestio (Monmouth), 1 1 2. Bobing-seata, 219. Bodan (in Bothamley, Ranasbotham, &c.), 207. Boderia, Bodotria Estuary (Firth of B^orth), 44, 45, 97, 152. Bodfari. See Varis. Bodmin manumissions, the, 162. Bodotria. See Boderia. Boece = beech (in Buckhurst), 207. Bogtielt. See Boiillaion. Boisil, Bishop of the Hwiccas, 184. Bolerion. See Belerion. 3IO BRITISH PLACE-NAMES Bolgros ('* Belly-moor"), 298 «. See also hwlch. Bollan-ea, 215. Bomio (? Cowbridge), 112, 129 «. Boniface (Winifrith), educated at Adescancastre, martyred at Doc- cuniy 190 n. Bonny, the river, 152. Bononia. See Gesoriacum. Bon-US (Bede-hus, Bed-ern), 227 w. Bootham Bar. See Galman-hithe. Borcovicioy 115. " Boresti," 47. Boructuari, 189. Bosenhani (Bosan-hamm, Bosham), 182, 291. Bosham. See Bosenham. Bosley Minn, 158 «. -bost^ shortened form of B^stad-r, 229. Boston (Lines.). See Icanho. Bothamley, Ramsbotham. See Bodan. Botulf, St., 176 w., 278. "Bouden Hill" (Badon Mount). See Caer Badon. Boudicca, Queen, 34, 36. Boullaion (? Builth, Boguelt, Buelt), 103. Boulogne. See Gesoriacum. Bourges. See Bituriges. Bovio (Beeston, Buistane), 103, 106. Bowes, 106. Bown-hill {Badon hill). See Caer Badon. Bowne, barony of, 153 n. Bowness (Cumberland), 54. Braboniaco (? KirkbyThore), 105;/., 114. See also Ravonia. Brada7iforda be Afne (Bfadford-on- Avon), 175. Bradan Relice {Steapan Relice^ Flatholm, Echni), 266, 302. Bradene (Bredon), 193, 262 n. Bradford-on-Avon. See Brada7i forda be Afne. Bradwell, 276 «. Brcegentforda (Brentford), 25, 26, 214, 287. Braintree, iii. Bramtun (Brampton), 304. Brancaster, Branchester. See Bra- noduno. Branch Huish {Htwisc), 206. Brancminstre {Branucminstrey i. e. ofS. Branoc, = S. Brandan), 206. Brandon. See Bravonio and Bran- nogenium. Brandon Hill, 294 «. See also p. 206. Brandon, St., memorials of, 206, 294 «. Brannogenion {Bravinnio, ? Bran- don), 102, 103. Branoduno, Branodumim (Brancas- ter, Branchester), 98, 113. '■'■ Branok hyalf hiwisce," 206. Bramic Dtmhead (Downhead), 206. Braunton, Devon, 206. Bravinnio. See Brannogenion. Bravonio (Brandon), 105 w., 112. Bre, in place-names, 105//., 193. Brecenan Mere, 264. Brecknock. See Bricheniauc. Bredon. See Brddene and Briudun, Bregedes-were^ 202. Breguntford (Brentford), See Bra:- gentforda. Bremenio {Bretnenium, Bremenion, Riechester), 53, 99, 105. Bremes byrig (Bromesberrow), 262 «., 263. Bremesgraf {^xoxi\t^^xov%)y 262 n, Bremeta7i7iaco, 105 «., Bremetonaci, Bremetonacum, Bresnetenace Veieranorum (Ribchester), 102, III, 116. *' Brent," 26. Brente (Brent Knoll), 199. " Brente " hill, 26, 199. Brentford. See BrcBgentforda. Bremen-land, 142 n, Breton's lands, the, 162 w. Brettanoii 75. Bretwalda, 239. Bricg Stowe (Bristol), 294. Bricge (Bridgenorth), 260 «., 263. Bricheniauc (Brecknock), 259 w. Bridgenorth. See Bricge. Bridstow, 294?/. Briga {^Brigy Brigs), 28. Brigantes, the, 25, 26 ; State of, 30; 49; 50> 94) 1 01, 102. INDEX 3" '* Brigden," early Brigo-duno-n, 193 «. Brige (Brigae), 113. Brigit and Briganiia, 94. Brigo-duno-n. See " Brigden ". " Briguntford." See Bragentforda, Brisnodes Land-share, 298 «. Bristol. See Bricg St owe. Britain, Provinces of, 71. Br it amity 89. Britannia, Britt6nes,^nVa««/, 88, 89. Britannia Minor, 1 56 n. Britannia Prima (Province of Britain), 71. Britannia Secunda (Province of Britain), 71. Britford, 25 «. British, cities of, 165 n. British troops in Gaul, 158. Britons, 84 ; on the Loire, 160 n. ; under Riothimus cross into Gaul, 159- Brittany. See Armorica and Lid- wiccium. Brittones, 89. Briudun (Bredon, Breodun), monas- tery of, 193. See also Bradene. Briuu (Brue), the river, 199. Brixton Devcrill, 253 n. Brocava, Brocavo (near Brougham Castle), 105;;., 108, 109 «. Brochyl, 216. Brockley Hill. See SuUoniacis and "Tunsteall, the old". Brocoliti. See Procolitia. Broco-mago, 115. Brog in place-names, 105 «. (Brog- 2aos = Breton name for England. Bro-cavo, Bro-vonacis, &c., pos- sibly involve Bro^. Bromesberrow. See Bremes byrig. Bromesgrove. See Bremesgraf. Bromfield. See Bruningafeld. Brom-geheg, 217. Broom, Loch. See Volas bay. Brough. See Anavio, Crococalana and Verteris. Brougham Castle, 108. Brovonacis (Kirkby Thore), 106. Brunanburh, 161, 370, 271, 273. Brue. See Briuu. Bruningafeld (Brunefeld, Brune, Brunanwercj IVeondune, Brom- field), 271, 272, 273. Brut of Layamon, the, 137 «. Brytaen Vechan, 156;/. Bry tan for dan (Britford), 300. Brythons, 9. B ry t /and {\Y ales), 297, 300. Buckfast. See Faestin. Buckhurst. See Boece. Budecakch {Bodecaleie, Butley), aoi, 202. Buelt {Bogitelt, Builth), 29, 103, 248 n. See Boullaion. " Bugga," pet appellation of St. Eadburga, 133 «. Buistane. See Bovio, Bulcan pytte, 217. Bullaeum, 39. Bun-an-Tabema, 129W. Bunchrew, 130W. Bun-doran, Bun-an-dall, 130 «. Bunker (Bunker Hill, Golf bunkers), 330. Burdoswald. See Anibo-Glanna. Burehu. See Bergos. Burford. See Beorg-ford. Burganes Lapidum, 272 «, Burgh and Burgh Castle, 98. See Garianno. "burgs, the five," freed by Ead- mund, 274. Burgundians sent into Britain, 53, Burgundiones, the, 27. -burh in place-names, 241 «. Burhou. See Bergos. Burrens {Burians^ Birrens, Bur- wens, Borwen = Cumulus) ^ 28, 272 w. Burrins Camps. See Birrens Camps. Burrio (? Usk), 112. See also Din- birrion and Wylisce Axa. Burrium, 29. Burton-on-Trent. See Byrtun. Bustadar, 228//., 229;/. Butley. See Budecakch, "Butter" (= Bittern), lai. Buttington, 259 «. Buttingtontump, 259 «. Buxton. See Aquis. bivlch, 298 n. Bydanford. See Bedanforda. Byferestan (Beverston), 293. 3T2 BRITISH PLACE-NAMES Bylgeslege (?Billingsley), 297. Byrgelsun (burying-places), 215. Byrhtmser, Bishop, buried at Lich- field, 289. Byrncwudu (Bern wood Forest), 267. Byrtun (Burton-on-Trent), 304. Cad bury (Camelot), 20, 21. Cadegyr (Categern), 139 «., 140. Coder ^ 16. Cseanganoi, 24. See also " Dece- angli ". Caedualla, king of Gweneddy 173, 183. Caelf-loca-a (Challock), 211 «. Caer Badon {Mons Badottiais, Bown Hill), 1 23, 1 23, 148 «., 149, 150 w., 151, 153. Caer Esca (Exeter), 252 «. Caere. See Hsefe and Caere. Caer Guiragon. See Caerwrangon, Caer-Gworthigirn (? Tintern, Din- thigim, Dindeyrn), 176 n. Caerleon (Mon.), (" Camp of the Legions," Isca), 54, 71, 113. *' Caer LigidW (Lichfield), 289 n. Caerluel (Carlisle). See Luguvallo. C<7«r /%m J (Dumfries), 120, 189. Caer Seiont. See Segontium. Caer Vudeiy 148 w. Caer Went. See Venta Silurum, Caerwrangon {Caer Guiragony Wor- cester), 165 n. See Wigomia. Caesaromago, 108, no, win. Caestri-uuara (Caestniuara), 215, 220 «. Cafortun (Caverton), 211 «. Cairbre Righfada. See Reuda. Cair Ceri. See Cirenceastre. "Cair Glovi." See Gleawan Ce- aster. Cair Granth (Grant Chester), 251 «. Cair Lerion, 265 n. Cair Luit Coyt, 181. Cair Pensa vel Coyt, 254 n. Q^\xn-Ryan. 94. Cairuisc. See Exanceastre. Caistor. See Venta Icinorum. Cait (Caithness^ 76, 77. Calacum {Kalagon\ 101 ; Casterton on Lune, 102 ; ? Overtown on Lune, III. Caladbolg, 106 n. Calat-ria (Kalentyre, Callender), 46. Calat-ros, 46. Calcaria (Tadcaster), 106, 301. Calder, the river, 144 n. Caldour (Caiedo/re), the river, 144 «. " Caledo," 45. Caledofre. See Caldour. Caledouy wood of. See Celidon. Caledonia, 45. Caledonians, 65. Calgacos, 83. Callender. See Calat-ria. Calleva Atrebatum (or Galleva, Sil- chester), 9, 47, 64 w., 103, no, 112, 113. Calne, 277 «., 279. Calne, the river (Scotland), 259 n. Calo-dunum, 277 w. Cal-ston hill, 277 «. CamaldunOy 90. Camboduno. See Campodunum. Camborico, 108. Cambridge. See Grantabrycge, Cambus, 213. Cam-dwr, 150 n. " Camel," river involving, 21, 141. Cameleac, Bishop, 266. Camelot, 20, 21. Camguili, the river, 250;/. Cam-las, 149 w. "Camlet way," 21. Camp-town, 170 w. Campodonum {Donafeld, Slack), 106, 169 «. Campus Lapideus, 88. Camulodunum (Camuloduno, Co- lonia, Colneceastre, Colchester), 13, 19, 20, 21, 35, 36, 102, 103, 108, no, 267. Canary Islands (Makaroi nesoi), 91, 92. Candida Casa (Whitehorn in Gal- loway), 124 w. Candover, 144 «., 281 w. Canegan-Mersces (Canning Marsh, Cannington Hundred, Caningis, Canninga mare), 285. Canning Marsh. See Canegan- Mersces. Canonio (? Kelvedon), no, in n. Canterbury. See Durobemia. INDEX 313 Caniio-n. See Cantium. Cantium (Cantio-n), 9. Cantref Orddwyf (Hundred of the Ordovi), 32, 102. Cantuc-uudu (Quantocks), 199. Cantwarena. See Warena. Canubio (Conway river), 102. Car {Kjarr) in northern place- names, 231. Caratacus, 29, 144 «. Carausius, 59, 60, 62-6, 68, 69, 1 19. Carbantorigon {Carbantiuni), 99. Cari stream, the, 205. Carling-ford, 229 «. Carlisle. See Lugiivallium. Carlisle, Old, 53. Carlued. See i.uguvallo. Carmarthen. See Maridufion, Carnarz'iJW, 10. Carravvburgh. See Procolitia, Carron River, 46. Carron-Water, 186 n. Carrum (Carham), 288. Carriim (Charmouth), 239, 340. Cartismandua iCartimandua), 30, 107 «. Carventana Arx, 33. Carvoran, 115. Cary. See Kari. Cassi, the, 10. Cassiterides, the, 90. Casielluni Menapiorum, 67. Casterton, 102. See also Calaatm. castle in place-names, 296. Castle Hill. See Margiduno. Castle-ford (Yorks.). See Ceaster- forda. Castleford. See Legeolio. Castor, See Castre. Castor on the Nene. See Daro- brivas. Castra Exploratorum, 105. Castre (Castor), 278. Cataracta (Cataractone, Katourak- tonion, Cataractoni, Catterick), loi, 102, 105, 106, 108, 169. Categim (Cadegyr), death of, 139, 140. Cathir, 16. Catraeth. See Galt-reath. Cat-scaul, battle of, 174 «. Catterick. See Cataracta. Catu in Celtic names, 1 39 n. Catn-Vellauni {KatveUaufiot), 103. Cat-vvalader, 139 n. Causennis (? Ancaster), 108, 109 «. Caverton. See Cafortun. Ce, 76. Ceahhyth. See Celchyth. Cealf-loca (Challock), 197 «. Ceangi. See Deceangli. Ceap-stowe (Chepstowe), 208, 294 «. Ceasterforda (Castle-ford, Yorks.), 375 «• Ceaster setna, 2 2o«. Ceastrum ( = York), 196. See also Eburacum. Ceddren (Cheddar), 206. Ceitern (Kernes), 139 w. Celchyth{CealchythjC\it\sQ2i), 157 «., 196, 197, 219. Celidon (Caledon, Celyddon), the wood of, 45, 150, 153, 153. See also Dun-chailden. Celtae, 9. Cemyn^ 300. Cenapa, 1 76 n. Cenepes-moor, 176 «. Cefid, 86 n. Cen-defrion, 281 n. Cen or Can-dover^ 281 n. Cenimagtii {jCeno-manni), 10. Centinces treow, 214. Ceoftefne^ 181 «. Ceolesige (Cholsey), 284. Ceollach (Kelly), Bishop, 178. Ceorlingcburh (Charlbury), 178 n. Ceortes-tge (Chertsey), 183 n. ; canons expelled from, 278. Cerdic. See Cerdices-ora. Cerdices-ford, 144, 145. Cerdices-leaga, 145. Cerdices-ora (Cerdic, Certic), 143 «., 144. Ceretic (Corotic), 144 «. Ceme Abbas. See Cernel. Cernel (Ceme Abbas), 303. Certic, See Cerdices-ora. Cestell Merit, 296 n. • Cet in place-names, 181 «. See LetocettuH. Ceto-bricca, 86 «., 181 n. Cett as tumulus, 181 «. Cett, Bosco de, 181 n. 314 BRITISH PLACE-NAMES Cettingaham, 238. Challock. See Cealf-loca. Char- ford, 144 w. Charing. See Ciomincge. Charing Cross, 219. Charlbury. See Ceorlingcburh. Charmouth. See Carrwn. Charterhouse-on-Mendip, 24. Charters, 213. Chartulary of Red on, names in, 161 «. Charudes, Heretha-land, the country of the, 223 «. Chatti (now Hesse), the, 10 «. Cheddar. See Ceddren. Chelmsford, 108. Chelsea. See Celchyth. Chentis-beria (Kentisbury), 241 n. Chepstow. See Streguil. Cherchehulle. See Cyric-hurh. Cheriton, 90. Chertsey. See Ceortes-tge. Chester. See Deva and Legeceastre. Chesterholm. See Vindolana. Chester-le-Street. See Incunenin- gum, Chesters. See Scythles cester. Chet = Keith in place-names, 86 «., 181 «. Chett, Forest of, 165 «., 181 «. Chetwood, 165 «., 181 n. Chevenage, Cefn^ 158 «. Chichester. See Regno. Chichestra. See Regno. Child-rey. See Cilla-rith. Chiltem, See Ciltinne. Chippenham. See Cippmiham. Chirchind, 52. Q\i\xV{CyriC'bur}i), 263 «. Chiulas. See Cyulas. CJiochilaicus (Hygelac), 131. CholleAoid, 115. See Cilurn. Chollerioxi, 115, 235 «. Cholsey. See Ceolesige. Christchurch. See Tweoxn-eam. Christian-Mal-ford ( = Cristenial = " Christs Cross "), 203. "Christs Cross." See Christian- Mal-ford. Churchill. See Cyric-burh. Churn, the river. See Corinnium. Chute (Chet), Forrest, 165 w., 181 n. Cicc {Cice^ Chich, St.Osyth), i79«., 304. Cicht, 74. "Cideston" (Tideston), Wells, 249 «. Cilla-rith (Child-rey), 171 «. Ciltern Ssetna, 221. Ciltinne (Chiltem), 216, 285. " Cilurn," the river, 115. Cilurnum (Celunno), 115, 235 w. Cinges-tune (Kingston), 279. Ciomincge (Charing), 219. Cippanham (Chippenham), 252. Circinn. See Cirig. Cirenceastre {Cair Ceri, Corinion ( ? Cironium ), Durocornoviu7?iy Cyren-ceastre, Cirencester), 103, 112, T47, 148, 175, 252«., 288. Cirig (Circinn), 52, 76, 77. Cironiufn. See Cirenceastre. Cironiw/i Dobunorum, 252 «. Cissa (son of ^lle), 142 «. See also Cissecester (Chichester), Cissan beorh. Cissan beorh, 142 «. Cissecester (Chichester), 142 «. See also Regno. Cithlescester. See Scythles cester. City of the Legion. See UrbsLegionis. Ciulae. See Cyulas. Clackmannan {Manau), 67, 100. Clceighangra (Clayhanger), 286, 287 «. Clanoventa. See Glannibanta, Claudian the poet quoted, 80. Clausentum (? Bittern), 53, 54, 64^., no. Clayhanger. See Claighangra. Cledemuthan (Clede-mouth), 267. See also Dou Clediv. Clevo. See Gleawan Ceaster. Clewer. See CI if wara. Clif wara (Clewer), 217. Cliffe at Hoo, 194. Clist-ton. See Glis-tun. Cloister words from Greek, 172 n. Closeworth. See Cloveshou. Clota et Bodotria. See Boderia. Cloveshou (Cloveshdh, Clofeshos), 194, 216, 238. Cloves-wurthe (Cloves-uude, Close- worth), 194 //. INDEX 3^5 Cliid (rock), 212. Clyde, the. See Klota. Clyde-Forth wall, 49,51. Clysts, the, 283 «. Cnobhere, 1 76 n. Cnobheresburg. See Garianno. Cnofwealh (Cnobwalch), 176 «. Cnut, attack on London, 286; Cnut and Eadmund divide kingdom between them, 288 ; Cnut sole monarch, 288 ; he receives sub- mission of King Malcolm, 289 ; dies at Shaftesbury and is buried at Winchester, 289. Cobre, 237. (? Cover-h&m.) Coccio (? Wigan), 1 1 1 . Cocidius ( = Mars), 116. Cockermouth. See Glatinibanta. Cock shot (Cock shade), 213. Cocuneda (Coquet), 99, 100. " Coel, King," 63. Cofa, 210. Cofan-treo (Coventry), 297. Cogidubn-i, lion. Coit maur. See Seal-ivyda, Colanica. See Kolanika. Colchester. See Camulodunum. Coldingham, 179 w. Colen-ea (Colne), 218. Coleshill. See Counsylht. Coligny, inscriptions at, 87. Colling (Cooling), 215. Col man, Bishop of Lindisfame, 180. Colne, the river, 218, 258. Colneceastre, See Camulodunum. Colonia. See Camulodunum. Cohidi urbs {Cohtdes-huTh) , 179 n. Columba, 49. Columcelli, 190. Combretonio, no, in. Combroges. See Cymry. Comich (Combwich), 253;/. Contpendio (Compiegne), 182. Concangios. See Deceangli. Concha-ceastre, I79«. See also Incuneningum. Conchenn, (Conchend, Cyngen, Doghead, Ogam Cuna-Cena), 55, 68. Conchess, name of St. Patrick's mother, 73. Condate (Northwich), 102, 103, 106 «., 1 1 1. See also Kind Street. Conde. See Cundoth. Conder, the river, 144 «. Conderco (Condecor = Benwell Hill) , 115- Congangios. See Deceangli. Congavata. See Incuneningum. Congresbury. See Cungres-beria. Conovio (Conway), in. Conway river. See Canubio. Cooling Castle, 217. See Culinga. Coonore, 143 «• Coquet, the river, 100, 116. Corbridge. See Corstopitum. Corda, 99. CoTfe, 298 n. Corfes Geat (Corfe Castle), 376 «., 278. Corin (Chum), the river, aSS «. Corinium, 288 n. Corio ( = Irish Cuire^ host), 155 «. See also Tri-corii, Petni corii. Coriosolites, the, 155 «. Corio-sopites f 155 w. Corio-vallum (Heer-le), 155 «. Com in " Abercom ", 186 n. Cornish compared with Breton place- names, 155-62. Comovii, the, 1 1 2 «. Corotic. See Ceretic. Corsept, 155 w. Corstopitum (Corbridge), 100, 105, 155 ». Coscaer, 158 n. Cotmgham (Cottingham), 304. Council of Aries, British Bishops at the, 70. Coimsylht (Coleshill), 24, 46. Count of the Saxon Shore, 79. Coventry. See Cqfan-treo. Cowbridge. See Bomio. Crailing. See Traver-er-lin. Crann, 86 «. Cray, the river. See Crecga. Cray ford. See Crecganford. Crecca-gelade (Crick lade), 261. Crecga, the (Cray), 141. Crecgafiford {Cx2iyiox^)y 141. Crediton (? Cridian-tun), removal of Bishop Leofric's See from, 131 w., 277 «. 3i6 BRITISH PLACE-NAMES Credy, the river, 377. Creech^ 39. Creech-hill. See Crich hulk. Creech Michael. See Crycbeorh, Criccyl, 265 «. Crich htdle (Creech-hill), 199. Crick-howel, 39 n. Cricklade. See Crecca-gelade. CHdian-tun (Crediton), 277 «. Cristemal. See Christian-Mal-ford. Crococalana (?Brough), 109, no. Croes-oswald (Oswestre). See Maserfelth. Croyland. See Cruland. Cr«<: ( = tumulus), 204. See Pn eel- ancyrca. Cructan. See Crycbeorh, Cruimther, 121 n, Cruk-ceri (Crukeri), 252 «. Cruland (Croyland), 178 «. Crummock brook, 273. Cruthenech, See Crtithni. Cfuthni (or Cruthenech), 74, 75 ; sons of, 76, 83. (Cf. Ogam in- scription, Luguqrittj where qritt = cruth.) Crycbeorh (Cructan, Creech Mich- ael), 39. Cuichelm, King of Wessex, 175. Cuit, the syllable, 85. Cul-chet (Culquith), 86 n. Culinga (Cooling Castle), 2 1 7. Culquith. See Cul-chet. Cumber, meaning of, 1 86 71. Cumberland (Cumerland), in; an English possession, 273; ravaged by Eadmund, 2 74, and by ^thel- red, 283. Cumbroges (Combroges, Cyrary), 26, 136 «., 299 «. Cutnenesora, 142 n. Cumrew, in. Cuncha-Chester. See Inamenin- gum. Cundoth (Conde), 256. *'Cunedda and his sons" arrive in Wales from Strathclyde, 136?/. Cu-neit, 82. Cunet-io(Cunetione,Kennet),ii,ii3. Cungres-beria (Cungersbury, Cun- gresbyri, Congresbury), 241 n., 249. Cuno — in Celtic personal names, 55. Cunobellinos, 12, 19. Cunoglas, 68. Cf. Glasi-Cunas, C«««^/ (Cunugl-ae, Knoyle), 258 «. Cununga-Chester. See Incunenin- gum. Cursitter, 227 «. Cuthbert of Lindisfarne, his conse- cration, 184 ; at Cloveshou Synod, 194, 195. Cu-Ulaid, 82. Cwat-brycge, 260. See also Qttat. Cweorn Cleofu, 204. Cwicelmes-hlsew, 284. Cyddivfr, 150 «. Cymbeline (Shakespeare's), 66. Cymedes halh, 216. Cymen (son of iElle), 142. See also Cymenesora, Cynienes-denu, Citmenesora. Cymenes-denUy 142;;,, 217. Cymenes-ora, 142 «., 143. Cymry. See Ctimbroges. " Cynebellinga-gemsere," 12. Cynegils, King of Wessex, 175. Cynemseres ford {Kempsford), 238. Cynetan (Kennet), 284. Cyngen, 55. See also Conchenn. Cynibre (Kinver), 216, 260 «. Cynlas (a Welsh brook), 150 «. Cynomannia (Cenomani = Maine) , 299 n. Cynuit, 253. Cyren-ceastre. See Cirenceastre. Cyric-burh (? Cherchehulle, Chur- chill), 263 «., 264. Cyulas {Cmlae, Chiulas~\ong ships), 122, 127 «. Dacore (Dacre, Bacara), 187, 269 «. Dacre, 269. See Ea?notu7n. Dsegstan. See Egesan-stan. dcel ( = valley), 23. Dael (in Doverdale), 208. -dsel ( = portion), 23. DcBrettta Mutha. See Darent. Dagenham. See Deccan-haam. Dal-chet (Dalkeith), 86 w. Dalreudini, 84. Dalriada, 84. Damnonioi (Dumnonii), the, 31, 99, 104. INDEX 317 Damnonion (Devonian) , promon- tory, the, Okrion, 94, 96. Danai, the, 189. Danes, 224, bought off for ;^io,ooo, 281 ; destroy Bamborongh, 281 ; fleet ravages country about Sand- wich, 284; ;^36,ooo tribute paid to, 285. See also Northmen, coming of, 223 et seq. Danihel, Bishop in the Isle of Wight, 166. Dano (Doncaster), 108, 109;/., no, 114. Dante, a resemblance of the trilogy of, 188 «, Dante and bolgia^ 298 n, Daoulas, 159 n. Darent {Darenta Mutha, Der\vent), 141, 291. See also Den 150^ i5i> i53> I59«->I74«' Dubris. See Dofra-n. Duecaledonian Sea, the, 92. Duin Necthan (Dunnichen, Dun- nechtyn, Dunnichen Moss = Nech- tanesmere), 185 n. See also Linn Gar an. Du-las. See Dubglas. Dumbarton. See Ailcluit. Dumfries. See Caer Pheris. Dumna. See Dotimna. Dumnonii, the. See Damnonioi. Dun-bulg, 106 «. Dun-Brettan. See Ailcluit, Duncan. See Duno-Catus. Duncansby Head. See VirvMrum promontory. Dunchad, Abbot, 192. See also Duno-Catus. Dun-Chailden (Dunkeld), 45, 153. Dunestaple (Dunstable), 304. Dunfoeder {1 Fothurdun), 269. Dun-fries. See Caer Pheris, Dungeness. See Ncesse. Dun-gleddy. See Dou Clediv. Duningcland, 219. INDEX 319 Dunion {^ Muridununiy Ridunto^ ?Sidmouth or Seaton), 104. Dunipace, 152. Dunkeld. See Dun-Chailden. Dunmay, 52. Dunmyat (Dalmyot), 52. Dunnet Head. See Tarvedum, the promontory of. Dunnichen, Dunnechtyn. See Duin Necthati. Dunnichen Moss. See Duin Nee- than. Duno-Catus (Duncan), * the fortress warrior': derivative Mac Dhun- cadh = MacConkey or Maco- nochie), 139 «., 192 «. Dunpeleder (Dmmpellier), 86 n. Dun-rechet(?Dunragit),i24w. See also Reget. Dunstable. See Durocobritis. Dunstan, Abbot of Gloestingaberig, 274; banished by Eadwig, 277; made Bishop of Worcester by Edgar, and Archbishop, 277 ; his death, 280. Dun-uualhi Pincerni, 215. Dunuuallan, 215. Dunwal-ing daenn, 288. Dunwich. See Donunoc. Durcinate, 175 «. Durham, Cuthbert's body removed to, 186. Durlstone Head, 157 «. Durngueis (Durngueir, Thornsaia, Dornsceta ) , 1 3 8 «. , 251. Durnono-varia^ Dumo-varia, Dorft- warn Ceaster (Dorchester), 113, 138 »., 151 «., 251 71., 252. Dumo-varia, involves varia = theatre, 251 n. Dur-o {Duron) y 187 «. Duro-bemia {Duro-vemiim, Dar- vernoUy Duroavef-us^ Cantwara- burg, Dorwit-ceastre,Canterbury), 38, 103, 106, 107 «., 187/1., 242. Durobrivas (? Castor on the Nene), 108. Duro-brivis (Rochester), Hrofibrevi^ Hrofesbretay Hrofes-ccesiir), 37, 38, 106, 107, 256, 280. Durocobrivis (Dunstable), 106, 107 w., 109, no. Durocomovio. See Cirenceaster. Dtirocornovitim. See Corinion. Durolevo (?near Sittingbourne), 106, 107 «., 280 w. Duroli (Duroliponte, ? Godman- chester), 108 n. Durolito (? Writola-buma), Writtle, no. III, 214. Durotriges, the {por-sata)^ 104, 138 «. Durovernum (Canterbury). See Duro-bemia. Dux Britanniarum, 114. Dwf[n]las (Dawlish), 150//., 159//. Dyfed. See Demetai. Dyrham. See Deorham. ea = river, in river and place-names, 68, i79«. Eadburga, St., I33«. Eadesbyrig (Eddisbury), 263 «., 264. Eadmund ^Etheling, succeeds i^thelstan in 940, 273 ; over- runs Northumberland, 274; be- sieges Anlaf at Leicester, 274; ravages Cumberland, and leases it to King of Scots, 274 ; stabbed at Pncklechurch in 946, 274. Eadmund iEtheling (known as "Ironside"), chosen king in 1016, 286 ; fights at Pen and Sher- ston, 286 ; attacks enemy in Kent, 287 ; is defeated at Ashingdon, 287 ; divides kingdom with Cnut, 288 ; dies, and is buried at Glas- tonbury, 288. Eadolfesnasse {Eadulfes vase, the Naze), 291, 296. Eadred, succeeds Eadmund i^thel- ing in 946, 274; reduces North- umbrians to obedience, 274; ravages Northumberland, 274; forces slaughtered at Ceasterforda, 275 ; imprisons Archbishop Wul- stan, 275 ; slaughter at Theotford, 276 ; dies at Frome, and is buried at Ealdminster^ 277. Eadward, succeeds Cnut in 1042, 289; his favour to Normans, 292; comes to Westminster and has the Minster consecrated (1065), 301 ; 320 BRITISH PLACE-NAMES his death, and burial in the Min- ster, 301. Eadward, succeeds Eadgar in 975, 278; he is slain at Corfe Castle and buried first at Wareham, 278. Eadwig, succeeds Eadred in 957 and banishes Dunstan, 277. Eadwin, King of Northumbrians, baptism of, 168. Ea-landj 68. Ealdminster (Old minster), 277. Eamotum (? Emmet or Dacre), 269, 273. Earldom of Northampton Shire granted to David of Scotland, 303. Earnes-Scaga (Herneshaw), 188. East Anglia, ravaged by Danes, 285. East Saxons (Essex), 129. Easton. See Estun. East-ryge (Eastrgena, Eastr^ge, Eastry), 178^., 179«., 219. Ea-ton, 68. Eatumberg (? Emborough, Etene- berga), 206. Ebbsfleet, 133. See Eopwines Jledt. Ebchester. See Vindomora. Ebriacus (now Yvre-le-Polin), 71 n. Eburacum (Eburaco, Eborakon, Eo- forwic Ceastre, York), 37, 71, loi, 102, 106, 108, no, 112, 237, 243, 344. Eburiacus (now E vry near Melun) ,71. Eburo-Caslum, 71 «., 99 «. Eburodunum (now Yverdon), 71 n. Eb(ur)oudouron (nowlvorne), 71 n. Eburomagus, 71 n. £bur6s ( = Yew), 71 n. Ecbryhtes-stan, 253. Eceni. See Venta-Icenorum. Ecgberct, Father, 185, 189, 193. Echni^ island of, 266. Eddesbury. See Eadesbyng. Eden, the river, 95, i2o«. Edinburgh. See Mount Agned. Edington. See Ethandim. Edisc (in Hides-Edisc), 208. Edisc-ueard, 208. Effe-cestre^ 276 «. Egelis Brich ( = Varia Capelld)^ 208. Egesan-stan (Daegstan), Daegsa-stan, 162, 163. Egisauuda (Fearful wood), 309 n. Egonesham (Eynsham), 147. Eildon. See Eldunum. 137 ^-j 138 «. See also Westymar. Guened ( Venedotiaj now Gwynedd), 135. 136 «• "Guent,"32. Guent, 259 n. Guentonia, 248 «. Guem in Guem-s-ey, 1 7 1 «. Guilou (Wiley), the river, 221 «., 250. See also Wiltun. Guinn liguiauc (Gwent loog), 259 n. Cf. Irish Imliuch (Emly), and Amlwch. Guinnion, the Castle of, 150, 152. Guital (?Vitalis), 56. Gulb(=beak), 189 «. Gumeninga hergae (? Harrow-on- the-Hill\ 209, 216. Gunnis, Guunnesse. See Gewis. guocob, in place-names, 244 n. Guorthigem, 127 «., 128, 134-7. See Vortigem, the Irish form of the name used by Gildas and Bede. Guorthemir, son of Guorthigirn, 138, 139- Guortigimiaun, 248 n. Guotodin. See Gododin, Guthlac, St., 194. Gwas, the use of, 8 2 «. Gweek ( = Wick), 166 «. Gwenedd (Gweneth, N. Wales), 32, 173 «., 248. Gweneth. See Gwenedd. Gwent loog. See Guinn liguiauc. Gwern-og. See Vern-um. Gwerthyr, 84. Gwic-Wern, 165 «. Gwig, 165 n. Gwynedd. See Gwenedd, Gwyr y Gogledd, 152. Gwysmeuric {Gwys Veuruc), 137 «. Gypeswic (Iipsv/ich), 280 «., 281. Gyrvii, the, 178. Gyrwa, North and Suth, 220. '' Gyrwa river," 178 «. Gyrwe, a Teutonic tribe (?), I'jSn. " Gyrwe-fenn," 178 w. Gytha, Harold's mother, retires to Flatholm, 302. Hadley. See Headlega. Hadrian's wall, 49. Hsefe and Caere, between (Avon and Carron, the rivers), 192. Haerg. See Gumeninga hergae. Haesel-hyrst gate, 218. HcBstinga Ceastre (Hastings), 292, 295» HcBstinga Forty 301. Haethfelth (? Hatfield chase), battle of, 173. See Meicen^ battle of. Hafing seota. See Hrempinguuic. Hafren (Severn), 24, Initial S be- comes H in British c. 300 A. D. : Pedersen § 303. hage (hedge, wood, enclosure), 212 «. Hagustaldesham (Hexham), 99, 115; meaning of, I74«., 184 «. See also Axeloduno, Hale (Healh), 180 w. Hal^-let, 234;/. Halech, 234. Halha, 234 n. Halidene. See Hefenfelth. Halignesse beorge, 214. "Hallelujah victory," 119, 126, 127. Hallington. See Hefenfelth. Halugh-ton (Haughton, Halech), 234 n. Ham (home), 213. See also Hamm. Hamble. See Homel-ea. HajHfti (land drained by dykes), 213. See Ricinga-haam, &c. Hamtun {Hamtune — Southampton, Northampton), 240, 264, 266, 285. Hanc-hemstede, 214. Hanger, 214. Hangra in place-names, 286 «. Hanratty. See St. Indracht. INDEX 325 Harbour, 155 «. Harda-Cnut succeeds Harold ; dies at Lambeth, 289. Hardwick, 166 w. Harelioch, 235 «. Harehouden (Harden), 235 w. Harold succeeds Cnut as King in 1034, 289; dies at Oxford (1040) and is buried at Westminster, 289. Harold at Bylgeslege^ 297 ; invades South Wales in 1062, 299 ; suc- ceeds King Eadwardin 1066, 301. Harraton, 235 n. Harrow. See Gumeninga hergae. Hartboume. See Heorat buma. Hartleford. See loratla-forda. Hartle/^^/. See Herut-eu. Harudes, 223;/. Hasel-over (Hasler), 208, 210;/. Hashhangra^ 287 «. Hasler. See Hasel-over. Hastings. See HcBstinga Ceastre, Hatfield Chase. See Haethfelth. Haughton. See Halugh-ton. Haug-r (Donners haugk), 231. Havvrds. See HorSar. Hayes. See Linga IIcBse. Headlega (Hadley), 257. Heahhaam (Higham), 215. Hearrahalch, 235. "Heathen," meaning of, 240 »., 241 n. Hehelmes Landschere^ 298 n. Hebrides. See Acmodct. Hedene (Hide-clothed), 212. '^ Ileer-W {Cono-vallum), 1^6 fi. Hefenfelth (Halidene, Hallington\ i74«. Hegestaldes setly 204. Hei-weg, 214. Helena (wife of Constantius), 62. Hemgisl, Abbot, 198. Hendrica, 221, 222. Ilengest (stallion), and Hengist, 132. Hengestdun (Hingston Down), 240. Hengestes-ige (Hinksey), 132. Hengestes-ricg (Henstridge), 132. Hengist arrives in Britain, 127 «., I28«. ; name mentioned in Beo- wulf, 131, 132 «., 133, 139. Henry I, spends Easter at Marl- borough in mo, 303. Henstridge. See Hengestes-ricg. Heofeshamme. See Eofeshamrae. Heorat buma (Hartboume), 218. Heorotforda (Hertford), 263. Herberct, Presbyter, of St. Herbert's Isle, Derwent water, 186. Herctilisy promontory of, 94. Herefinna, 221. Hereford Port (Hereford), 297. Herepath, io«., 155 «. Heretha-land (HorSar, Hawrds, Charudes, Harudes), 223 «. Herio-s. See Auray. Heriri, montes*(Snowdon), 136, 239, 303. Herneshavv. See Earnes-Scaga. Hertford. See Heorotforda. Henit-eu (Herot-ea, Hartle/V?^;/), 179 »., 180. Hesket (Hesketh), 86 «. Hesse, 10. Heuuald (Black and White) Pres- byters, 190. Hew Goose {Uch coed), 158 n. Hexham. See Hagustaldesham and Axeloduno. Hibaldestowe, 294 n. " Hiberni," forms of, 72. Hicca, 221. Hides- Ed isc. See Edisc. High Cross. See Venonio. High Riechester. See Bremenio. High Tor, 212. Higham. See Heahhaam. Hii (Hy,Iona), 146, 178, 185 «., 190, 191, 193, 225, 236. Hii Columcelli (/ Columba Cille), 190, 236. Himeid, King of Dyfed {Demetid), 248. See also Maes-yvet. Hincmar, Archbishop, 245 n. Hingston Down. See Hengestdun. Hinksey. See Hengestes-ige. Hin-Locrinus {Locrinus, Lloegr), 198. Historia Britonum, 65. Hjallt-land (Shetland), 225 «. Hjalpands-ey (Shapinsay), 225 «. Hlid (Lydd), 216, 235 «. Hlidaford (Lydford), 282. 326 BRITISH PLACE-NAMES hlinc (slope), 214. Hlud, Monastery of, 235. Hocneratun (Hokenarton, Hook Norton), 265 n. Hdh (promontory), 194 «. Hokenarton. See Hocneratun. Holegn (in Holenhyrst), 207. Holenhyrst. See Holegn. Holni^ 231. Holm-ryge, 1*^8 n. Holy Island. See Fame Islands. Holy Places, a work on the, pre- sented to King Aldfrid by Adam- nan, 191. Holy well. See Tegingel. Homel-ea (Hamble),*the river, 21, 145 «., 172;?. Honeffe, 69. " Honey Child " in Romney Marsh, 226 «. Honorius, Archbishop of Canter- bury, 178. Hook Norton. See Hocneratun, Hope ( = a bay and river valley), 231. Horburg. See Argento Varia. Horca or Orca (Okrion-point), the Lizard, 89, 96. Hornblawerton (Homblotton), 207. Homcastle, 102. HortJar (Hawrds, Charudes, Ha- rudes), 223;?. Horsa arrives in Britain, 127 w., i2Sn., 131, 132 n., 133; his death, 139, 140. Horsa/a/d {Hor&f all) f 132. Horsaleah (Horsley), 132. Horsfall. See Horsafald. Horsham, 132. Horsley. See Horsaleah. Horstead, 140. Howth Hill, 229W. Hiemping-uuic (Hafing seota), 219. Hreod-ham, 217. Hreopa-dun (Hreopedune, Repton), 196 «., 250. " Hreth " Goths, 223. Hreutford (Hreodford— Redford or Redbridge), 172??., 183. Hris (Rhys), slain, 296. Hrither (= cattle), 208 «., 216. Hroching (Rocking), 218. Hrofes-castir, HrojFes-breta, Hrofi- brevi (Rochester). See Duro- brevis. ZTf-other, 183 «. Hudan-fedt, 132, 133;?., 215. Hugabeorg, 214. Hughes, Professor M'^Kenny, and the landing-place of Augustine, 135- Huic, 165 «. See also Hwicna. Huicceundu, 165 «. Huiccii (Huich), 164, 216. Huiccii, territory of, 166. Humbran muthe (H umber estuary), 281. Humir, the, 98. Hunapa, 69. Hundehoge {Hunecote, Hundcot), 304. Hiinengraber, 82 «. Hunger-\\\\\, 287 «. Hunni, 189. Hunno {Onno), 115. Hiinoz, 82 «. Huntandun (Huntingdon), 267. Husmerae, province of, 216. Hussa, son of Ida, founder of the kingdom of Bernicia, 163. Hwcerwella-n (Wherwell), 294. Hweallsege (Whalley), 237. Hwiccas, 184. See aho ffmccii. Hwicna biscop, 164 «. Hwicna Gemcere, 16^ n. Hwinca, 221. Hivitan Wylles geat (Whitewell's gate), 274, 275 «. Hwitciricean (Whitchurch), T03, 283. Hwitem, 235. Hwyte Celdan, 226 «. Hy (Ion a). See Hii. Hybernia, 90. Hydata Thertna. See Bathan- Ceastre. Hygelac. See Chochilaicus. Hyppeles Fleote (near Sandwich), 134- Hyrtlingberi (Irthlingborough), 304- Hywel Dda, 267 «. lam (or Iron) language, 75 n. Iberran, loi. INDEX 327 Iberion, 88. Ican^ \*l(i n. See also Iceni. Icanho (Boston), 176. Iceland discovered, 225. Icenan, the river, 108 «. See also Itchington. Icenhtldeweg, ion. Icenhylt (Icknield)^ 12, 13. Iceni, the, 12, 20, 176 w. Icht, see of. See Muir-n-Icht, Icinos (?Ick worth), 108. Ickham. See loccham. Icknield. See Icenhylt. Ickworth. See Icinos. / Columbcs Cille. See Hii. Iction. See Ition. Ictis, 89. Ida, founder of the kingdom of Ber- nicia, 163. Idla. See Yddil. lena, the (Wigton-bay), 93, 95. lerne (Ivema), 89. t]^'™ island, 179 «. Iglea (-^glea, MgUd). See Aclea. Ikenoi, the, 103. //a, the river, 97. Ilchester. See Ivel- Chester. Ilegh (South-leigh\ 254/1. "IleyOak," 254 w. //^^/5/5a// (Ulf-ketels-heall), 284 «. Ilkley. See OHcana, Imperial Succession in Welsh Ge- nealogies, 58. Inber Domnann, 99. Incnneningum {Congavafa, Kunka- Cester, Concha-ceastre, Cunanga- Chester, Chester-le-Street), 115, 116, i79«., 188. In-dera-uuda (Beverley), 188. Indracht, St. (Hanratty), 206. Ine, King of Wessex, 194. /«awy/rm (Glastonbury), 200. See also Glastenic. 7n Feppingtim, 178. Ingetlingum (Gilling), 178. Ingo-mar. See Ivarr. In gyruiim (Yarrow), 178 «., 191. See also Donemouth. Inhrypum (Ripon), 191. See also Ripar. Inisboufind, 180. Inis Metgoi'i. See Lindisfam. Inis Patraicc (Peel), 236. inis Prydain (Welsh name for Great Britain), 74. Inlet (Yenlade), 121, 189. Inscribed stone at Deerhurst, 298//. Inscriptions at Coligny, 87. Inscriptions at Rom, 87. Inscriptions at Todi, 87. "Insi-Gall," 27 «. In Undalum (Oundle), 191. InvcTy meaning of, i86«. loccham (Ickham), 218. lona. See Hii. loratla-forda (Hartleford), a 18. Ipswich. See Gypeswic. Ircingafelda (Erchenfield). Sec Areconio. Ireland. See ** luuerna ". Ireland, Ecgfrid invades, 184, 185. Irin. See luuerna. Iris {Irin), See luuerna. Irish sources in Nennius, 128 «.; hermits in Iceland, 225 ; among Norse settlers in Iceland, 225 «. ; pilgrims land in Cornwall, 257. Irruaith, 223 w. Irthlingborough. See Hyrtlingberi, Isannantia. See Isannavantia. Isannavantia (Isannantia), 109, See also Bannaventa and Bannarento. Isam^ 75 n. Isbister, 227//. Isca. See Caerleon. Isca Dumnuniorum. See Exan- ceastre. Isenan awelm, 215. Iska (Exeter). See Exanceastrt. /f>^a-mouth. See Exanmuthan. Iskalis, 9«., 104. Isotirion {Isubrigantum, Isurium, [01, 102, 'rj 106, Aldborough), 108. Isubrigantum. See Iscurion. Isurium. See Isourion. Itchington, 108 n. lihtige, 1 1 1 «. Ition (ikion, Iction), 74, 90. Itios river-mouth, the (?the Sound of Sleat), 92, 93. Ituna, the river, 93, 95, 120 n. IiidoH'byrig {luthan bin'g), 179;/.; 328 BRITISH PLACE-NAMES Wulstan imprisoned in, 275, 276 n. See Geoda, Ythanceaster. "luuerna" (Hiberni,Iverna, Ivernia, Evernili Patria, Eriu, Erinn, Iwerddon, Ireland), 72 «., 73, 89. Ivarr (Ingo-mar), 245 n. Ivel, 141, 221 «. Ivel-Chester (Ilchester), 141, 221 «. Iverna, Ivernia. See luuerna. Ivorne. See Eb(ur)odouron. Iwerddon. See luuerna. Iwys. See Gewis. Jsickmenis, 158 «. Jarls, flight of the, 224 n. Jedburgh. See Ged-burgh. Jemeppe, 69. Joseph of Arimathea, 197. Judoc, St. {/ud-gual,/usf), 261 n. Jud-gual. See St. Judoc. Jugantes (Evigantes) , State of the, 30. Jutes, the, 128, 131 «., 140. futhan-hirig. See ludanbyrig. Kaer Septon. See Sceftesburg. Kaer vudei (? Woodchester), 123 «. Kaianganoi^ promontory of the, 93. Kaianganoi {Ganganoi), 95. Kaili-os, the mouth of the (? Port Gordon), 97. Kainos Limen (? Portsmouth), 97. See also Fo7'tesmutIia. Kairlium, 153. See Chester and Dumbarton. Kalagon. See Calacum. Kaledofr. See Cen-defrion. Kaledonian forest, the, 100. See also Caledonia. Kaledonioi, the, 100, loi. See also Caledonia. Kalentyre. See Calat-ria. Ka^noulodounon. See Camulodu- num. Kantioi (Kent), 103, 104. Kantion, Noukafttion, the promon- tory (Kent),? Beachy Head, 89, 97. Kari (Gary), 201. KarnSnakai. See Kerones. KarnSnes. See Kerones. Karreg Tylhvaen, 157 n. Kassiterides (tin-islands), 89. Kassiteros, See Kattiteros. Katouraktonion. See Cataracta. Kattiteros {Kassiteros), 89. Katvellaunoif the. See Catu-Vel- launi. ''Keiths," in Scotland, 86«., 181 «. Kelda ( = a well), 2 26 n. Kelly. See Ceollach. Kelvedon. See Canonio. Kemesegia (Kempsey), 289. Kempsey. See Kentesegia. Kempsford. See Cynemseres ford. Ken, the prefix, 86 n. Kenchester (Hereford), 54. See also Magnis. Kenetbury (Kintbury), 11, 284. Keneth, King of Scotland, 206. Kenion, the (? Fowey), 96. Kennet. See Cunetio. Kenstey, Bishop of Dinnurrin, 193 n. Kent, evidence of Jutish place-names in, 226 «. ; peace bought in East Kent for ^3,000, 285. Kent. See Kantioi. Kentisbury. See Chentis-beria. Kenver, 216. Ker ( = horn), the root, 183 n. " Kernes and Gallowglasses," (Shakespeare), 139 «. Kernwy, 203. Kerones, the (Kreones, Karndnes, Karnonakai), 100. Kerriemuir, 270. Kesteven. See Ket-stefena. Ket-stefena (Kesteven), 181 «. Keynor, 143 «. Kidderminster, 150 n. (of. also Welsh Cyt-tir = a common). Kiddermore Green, 1 50 w. Kit-, in place-names, 190. Kilda, St., 226 «. Kilpatrick, 190. Kimble, 12. Kimbolton, 12. Kind-street, 102, 106 «., 107 «. King Lear (Shakespeare), 66. Kingston. See Cinges-tune. Kinnaird's Head, 97. Kinross. See *'Fif and Fothrif". Kintbury. See Kenetbury. " Kinver." See Cynibre. Kirkby Thore. See Braboniaco. KirkmtuUoch, 52. INDEX 329 "Kirkja," use of word by Norse- men, 226. Kirknewton, carved sarcophagus at, 168. Kirtlington. See Kyrtlingtun. Kits Cotty-house, 140. Kjarr, 231. Kleit-r, 16. Klota (Clyde), 92. Knoyle. See Cunugl. Kolanika {Colanica), 99. Korio, 99. Koritavo {Koritanot), the, 103. Kornavioi, the, 100, loi, 103. Kreones. See Kerones. " Kunka-Cester." See Congavata. Kyrtlingtun (Kirtlington), 279. Laam (in Lamhythe, Lampyttas), 208. Lackford, 108 «. Lactodoro (Lacto-duro, Lactodrodo, Lado-donim, Tofeceaster, Tow- cester), 106, 107 «., 109, 266, 267. Laech. See ledh. Lagecio. See Legeolio. Lagny. See Latiniacuvi. "Lakehead" (Penne locos), 109 it. Lamb-ay, 229 n. Lambeth. See Lambhythe. Lambhythe, Lamhytha (Lambeth), 289. See also Latn seat has and Laam. Lampyttas (loam-pits), 208. Lamseathas (Loampits), 290 ;/. See also Lambhythe. "Lanchers" at Worth Maltravers, 298 «. Lanchester (Durham), 53. Lanchester north of Binchester, 114. Lanchester on the Lune, 114, Lancing y 142 «., 143 w. See also Wlencing. Landford. See Leonaford. Landndma B6c, 225 n. LMndschar die, 298 n. Landscor-ford, 298 «. Land'scoru, 298 «. Land's End. See Penwiht Steort, Penbryn Penwaeth. Langatreo (Longtree H.), 293 n. Lange herga, 241 «. Zflw/iJ^fl/ (Lantocai), 198, 201. Larkbear. See Laurocabera, Lasket in Cumberland, 181 «. Lasket grove, 181 «. Lastingaeu (Lastingham), 179, 180. Lastingham. See Lastingaeu. Latch, 231. Latin words borrowed from, n8. Latiniacuni (Lagny), 177- Laurocabera (Larkbear), 241 «. Lausiac of Palladius quoted, 172 n. Lavatres, Lavatris, Levatris, io6, 108, 109 «., 114. Lavingtunes dices, 214. Layer Marney. See Legra marany. Lea, the river. See Lygan. Leach, 231. lead, 82 «. iedh in place-names, 145 «., 177 n. Lech, 231. Le(c)to-cet-um. See [L]Eto-cet- um. Ledecestre. See Ligera Ceastre. Leden-ioxd (Leadon), 301, 282 «. Leeds. See Loidis. Leek, 231. Legeceastre (Lega Ceaster, Leg- ceastre, Ligcester, Urbe Legionis, Carlegion, Chester), 150, 151 ; battle of Chester (616), and slaughtering of monks, 1 68 ; West Chester, 173 «., 259, 262, 277, 283, 298. See also Deva. Legecio. See Legeolio. Legeolio (Lagecio, Castleford), 108, 109 n., no. Ivegions in Britain, 117. Legra {Legra marany, Layer Marney), 265 n. See also Ratae. Leicester. See Ligera Ceastre. Leighton Buzzard. See Lygtun. Lein-ster, 229 «. Leith (Lode), 120. Leixlip, 2 29«. Lemanis (Lemannis, Limen - ea, Lympne, Liminge, Limen wero), 90, 108, 113, 214, 215, 218, 219. Lemannonion bay (? Loch Fine or Loch Linhe), 92, 100. Len-borough. See Lygeanburg. Leofric, Bishop of Worcester, death of, 289. 33© BRITISH PLACE-NAMES Leomynstre, Leof minster (Leomin- ster), 390, 295. Leonaford (?Landfoid), 249. -let and gel&t (a junction of roads), place-names ending in, 189. Letch, in place-names, 231. Letewic (Lledewic), 159 «. Letewicion, 159 «. LetheccEstrescire . See L igera Ceastre. [Z] Eto - cet •• ufn {Le{c)to - cet - icm, Luit-coet, Litcidfeld, Lyccidfelth, Llwyd coed, Licetfelda, Lichfield), 38, 106, 181 «., 194, 296. Leucaro (? Loghor), 1 1 2, Levatris. See Lavatres. Liber Vitae, the, 200. Licetfelda. See \^L\Eto-cet'Utn. Lichfield. See [L']Eto-cet-um. Licitit. See Caer Ligidit. Lid, the stream, 282 n. Lidwiccium (Brittany), 256. See also Armorica. Ligera Ceastre {Ligora, Ligara- ceaster, Letheccestrescire^ Leicester), 264, 274, 304. See also Ratae. Ligualid. See Lugu-vallium. Ligures, the, 88. Limen-ea. See Lemanis. Limenemouth, 258. Limen wero. See Lemanis. Limfin (lime heap), 212. Liminge. See I^emanis. "Limit," the (N. Wall), 105, 106 n. Limnus, See Lemanis. Lincoln. See Lindocolina, Lind {Llynn), 16, 170 «. Lind-Cylne in Kent, 171 «., 275 n. Lindes farona, 220.' Lindesige {Lindessef Lindsay), 240, 281. Lindis, the river, 171 w. Lindisfari distinct from Lindisfarne, 193 «. Lindisfam {Medcaut, Inis Metgoit, Lindisfarena €€) , meaning of name, 170 «., 171 «. ; assigned to Aedan by Oswald, 174, 178; Cuthbert's body removed to, 186, 196 ; attack on the Church, 236. Lindissi (Lindissa, "Lindis-faro- rum "), 170 w. Lindocolina (Lindon, Lindo, Lin- dum, Lind-Cylene, Lincoln), 16, 99, 103, 108, 109, no, 171, 274, 275 «. Lindsay. See Lindesige. Linga Hme (Hayes), 219. Linhalec, 1 59 n. Linhe, Loch. See Lemannonion bay. Also Longos, the river. Linn. See Lind. Linn Garan, 185 «. Linnuis, 150. Litau, I59«. Litchfield in Hants. See Lydes- chulv. Litherland, 209 «. Littleborough, See Ageloco. Litus Fresicum(¥ris\2LVi Shore), 120. Livermere, 209 «. Liverpool, 209 w. Livius Gallus, 67. Liwellid. See Lugu-vallium. Lizard, the. See Horca. Llanaledh (St. Germains), 279 «. Llancarvan, 34, 129 «. [In Life of S. Finnian ** Garbayn, alio nomine Nont " is mentioned along with a civitas Melboc. Carvan is the same name as the Ogam Carbagni.l Llandaff, Book of, names m, 16 1 n. Lledewic. See Letewic. Llewelyn, 12 «. See Lugu-Vallium. Lloegr. See Loegria. Llwyd coed. See Luit Coyt. Llyn. See Lind. loca (in Pynloke, &c.), 197 «. Lochie, Loch, 92. Lochlannach (men from fiords), 241 n. Locrinus, 198. ** Loddera-wyllon," 204. ^^ lodes'" on the Severn, 120, 189. Lodonia. See Lothene. Lceck-r, 231. Loegria (Loycr), 198, 259 «. Loghor. See Lucaro. Loidis (Leeds), royal vill in region of, 170. Loingseach (Lynch), 122 n. Londinium (London, Londinio,Lon- donion, Lunden byrg), 35, 36, 78, 79, 103, 106, 107, 108, 109, no, 141, 242. INDEX 331 Longae Naves, 122 n. Longeas, 122 n. Longmi'nd, 158 w. Longo-vicio (Longo-vico), 114. Longos, the river, 92. Lossiemouth. See Lojca, the. Lothene (Lothian), i-jo ft., 303, 304. Lothian. See Lothene. Loucopibia, 99. Louenlitton. See Luetensium. Lougoi, the, 100, 10 1. Loiiih, 170, 235 «. Lowdham, 335. Lowther, the, 109;/. Loxa, the (Lossiemouth), 97. Loycr. See Loegria. Loyt Coyt. See {^L\etocetuvu Luce-hoy and river, 94. Lucius and Eleutherus, fable of, 197. Lucotion, 99. Lud-ccrce, 235 n. Luddesdown, 235 «. Luel, 187 M. Luetensium {Luentinum, LuentU non), 103, 288. Ltigu-dun-um, i^n. Lugitvalliuin, Lugu-vallum (Caer- | Ligualid, Liwellid, Carlued, I Luguvallo, Lugubalia, Carlisle), 12, 13, 71, 106, 108, 137 «., 186, 187 «. Luit-coet. See [L'jEto-cet-um. Lunden byrg. See Londinium. Lune, the, 102, iii «., 115. Lusebyrge, 218. Lyccidfelth. See [L']EtO'Cet-um. Lydd. SeeHlid. Lydeschulv (Litchfield in Hants), 181 n. Lydford. See Hlidaford. Lydney, 20. Lygan (Lea), the river, 260, 263. Lygeanburg (Len-borough), 147. Lygtun (Leighton Buzzard), 265. Lympne. See Lemanis. Lynch. See Loingseach. Lynn. See Lind. Ma and Maes, 166 n. MacAlpin, King Kenneth, 84. Macbeth, King, 297. MacConkey or Maconochie, 192 n. Mac Dhuncadh (MacConkey or Ma- conochie), 192 n. MacNachten. See Nechtan. Maeatae. See Maiatai. Msedham (Medeham), 215. MjEgla, son of Port, 144. Mae I, the use of, 82 «. Mseldun {Afeldufta, Maldon), 263, 281 ». " Maelmin " (Melmin, Melmun) , the royal vill, 168. Mael-ros (Melrose), 168 «., 190. Maen, 112 w. il/<7«z]f«(IsleofMan). SeeMonapia. Maercrjedes-burn, 143. Marle-beorg (Marlborough), 303. Maes-Aleth, 136. Mats Garmon, 127. Maes-yvet (Maes Hyveidd, Old Radnor), 248. Magen. See Magnis. Magenasaete. See Magonsete. Mageo. See Mayo. Magesatas (Magon-Setum), 287 w., 288. Magesitania, 166. Magh-girginn. See Cirig. Magio-vinto (Magiovento, Magio- vinio, Fenny Stratford), 106, 107 ;/., 109, no. Mag-ith, 1 1 1 «. Maglone, 114. Magnis {Magen, Maen, ? Ken- chester, or Carvoran), 112, 115. Cf. Magos. Magnum Monasterium, I24«. See Whitehom. Magonsete (Magenasrete), 166. Magos, 112 n. Maia, 100, loi. Maiatai {^Maeatae), the, 52, loo, loi, 163. Maine. See Cynomannia. Mailros, 184. Maio, 1 01. Mais, loi. Malahide Bay, 99. il/a/a/^j(?Mull),92. Maldon. See Mreldun. ** Maleventum," 33, 34. 332 BRITISH PLACE-NAMES Mamble, Momela, Mamele, 268 n. Mameceaster. See Mancunio. Mam-ilet, Matn-heilad^ Mam-hilad^ 268 n. Mamucio. See Mancunio. Man, Isle of. See Monapia. Manand (Manann), in Cafnpo, 60, 67, 192 n, Manapia (? on the Wexford Coast), 68. See also Monapia. manau Guodoton {Manaw of the Gododin or Votadinoi), 60, 67. Mancetter. See Manduesedo. Manchester. See Mancunio. Mancunio (Mancunium, Mame- ceaster, Manchester), 26, 102, 106, III, 268. Mandu in personal names, 107 n. Manduesedo (Mancetter), 106, 107 «. Manna (Isle of Man). See Monapia . Map, Walter, 65. Maple-durham, 187 n. Maran, the river. See Memeran. Mara-scalc (now Marshall), 83 n. Margiduno (Castle Hill), 109, no. Maridu7ion (Maridunum, Muriduno, Carmarthen), 29, 103, 104, 112, 113. Marlborough, 113. Marshall. See Mara-scalc. Marten. See Meretune. Martenesie, 206. Martin, St., episcopal See and Church of, 124 n.\ oratory of, 215. Mascusius, " High Admiral," 206. Maserfelth (? Oswestry, Croes-os- wald, Bellum Cochboy, " Wane- loc," ?Wenloc), i74«. Mathuedoi, Count of Poher, 160. Maxim Guletec, 56. Maxima Caesariensis (Province of Britain), 71. " Mayo of the Saxons," 180. Meanuari {Meaniuard)^ East and West Meon and Meonstoke, 182. Mearcfleot, 217, Meare. See Ferramere. Meave, Queen (Mab, Medb., Medvii in Ogam). See Pit Meave. Medcaut. See Lindisfarn. Medeshamsted {Medeshamstede, Me- des-wael, Peterboro), i*j6n., 177, 244, 277; relics transferred to, 278. Medewsegan (Med way), 285 «. Mediolano {Mediolanion, ? Whit- church), 102, 103, 106, 107 «., III. meends, in place-names, 158 «. Megas Limen (? Poole Harbour), 96. Meguuines paeth, 214. Meicen, battle of, i73«. Melandra Castle^ 116. Meldu7ia. See Mseldun. Melrose. See Mael-ros. Melvas, King, 153. Memeran (Maran), the river, 263. Menapia, Menapioi (in the Low Countries), 60, 67, 90. See also Monapia. Mendip. See Munidop. Menevia (St. Davids), 67. See also Monapia. vienez, mene, menedh^ 158 «. Meoles at Shrewsbury, 231. Meols {Mel-r), 231. Meosgelegeo, 216. Merantun, 195. Meresig {Mereseg, Mersea), 260. Meretune (Marten), 246 n. Merlin. See Myrdin. Merscuuare, 215; (Merscware) of Romney (Romenal), the, 215, 240. Mertai, the. See Smertai. Messappia, early name of Milan, 69. Metaris estuary, the (" the Wash "), 98. Meuric (Marius), 137 «. See also Gwysmeuric. Mevania, name given by Bede to Man and Anglesea, 67. See also Manapia. Miathi, battle of the. See Maiatai. MicheW