Skip to Content.

Kent Archaeological Review extract

Some Early Place and River Names in Kent.
by John Evans, FSA, FRGS.

Few pre-English place and river names survive in Kent even in greatly modified forms, and the former are generally attached to Roman sites and their names given in a Latin form. Professor K Jackson has defined Roman-British as "the form of British reported by Roman writers in a latinised spelling." (1) British river names are recognised by the experts under their later modified forms, but some have a literary confirmation.

The principal Romano-British (2) place names were attached to the two towns (civitates or castra) and to the four Saxon Shore forts in the county, and the ancient literary evidence for them is definite. These sources are as follows, with a short reference to each:

Ant. It. The Antonine Itinerary (Itinerarium Antonini Augusti). Considered to date from the end of the third century, it covers the whole Roman empire, giving routes from one place to another with total distances, and the names and distances apart of intermediate stations. The best and classical forms of place names.
Rav. The Anonymous Ravenna Geographer. A compilation which gives a list of the countries, towns and rivers of the known world made by an unknown cleric of Ravenna, and dating from the late seventh century. The names are greatly corrupted.
Peut. The Peutinger Table. A Roman itinerary in the form of a rough map, supposed to date from the fourth century, but derived from earlier sources. Much of the western part of the empire including all but the southern coast of Britain, is missing.
ND. The Notitia Dignitatum. A list of civil and military officers, with the names of their stations, at the beginning of the fifth century.
Ptol. Claudius Ptolemeus, the Geography. AD 150.
Bede. The Venerable Bede, Ecclesiastical History of the English People, circa AD 730.
A S C. The Anglo-Saxon Chronicle. Date of entry given.
BCS. Birch's Cartularium Saxonicum (Saxon Charters). The first figure is the date, the second, the reference.
Nen. The Historia Brittonum of Nennius. Compiled about AD 800 from many and older sources.

Here is a review of thirteen such British and Romano-British place and river names:

Kent Cantium. In Greek Kantion (Strabo 44 BC, Diodorus Siculus 54 BC), Kantion Akron, Ptol. (Kent Foreland), Cantium, Caesar, Cantia, Bede. The Belgic tribe which occupied east, and the northern third, of Kent were known to the Romans as the Cantii. The Old English name was Cent, Caent, and for the people Cantiware, later Cantiwara. Chenth in Domesday. Meaning: Romano-British canto, white, or cantus, border. (Wall. II) (3). From the Cantii (Jackson) ekwall agrees with Wallenberg, but quotes Silvan Evans as offering Welsh caint, plain or open country.
Canterbury Durovernum Cantiacorum. Dorovernon, Ptol. Duroverno, Ant. It., 2, 3, 4. Dorovernum, also, Cantia, Bede. Dorobernia, Asser. Dorwitceastre, A S C, 604. Dorovernia, 605, BCS, 5. Meaning: British duro, fortress, vernum, based on a Celtic word for alders, hence swamp or marsh. "The fort among the alders" (Arch. 93). Wallenberg and Ekwall agree, the latter giving "the swamp by the fort." Cantiacorum, Latin, of the men of Kent, i.e. of the Cantii. The OB name was Cantwarabyrig, the town of the men of Kent.
Rochester Durobrivis, Ant. It., 2, 3, 4. Durobrabis, Rav. Roibis, Peut. Dorubreui, and also civitas Hrofi, Bede. Civitate Hrofibreui, 604, BCS, 3. Civitate Hrobi 842, BCS, 439. Dorobrevi, 844, BCS, 445. The Old English names were variants of Hrofescester. Roves cestre, in Domesday. Meaning: British duro, fortress, British brivis, loc. pl. "at the bridges" (Arch. 93). All experts agree, hence "fortress at the bridges," but Ekwall reads it "the bridges of the stronghold."
Dover, and the River Dour Dubris, loc. pl. Ant. It. 3, 4. Rav., Peut., ND. Dofras, 696, BCS, 716. Meaning: British dubro, dubra, water, river, hence The Waters. All the experts agree.
Richborough Rutupiae. Routoupiai, Ptol. Rutupias, acc. pl. Ammianus Mar cellinus. Rutupis, pl. Rav., ND. Ratupis, Peut. Ritupis, Ant. It., 2. Rutupinas, L. adj. Rutupi Portus, Orosius. Rutubi Portus, Bede, who adds that the English have corrupted the name to Repta-caestir. Later forms are Ratteburg, 1196, Ratborough, 1462, and Richborowe, 1509. Meaning: British ro, great, tupias, unknown meaning (Jackson), but Arch. 93 has British rut, Welsh rhwd once meant filth, cf. Gaul. rutubi, muddy. Hence muddy creek or flats. Possibly from Gaul ritum, Welsh rhyd, with loc. ending meaning "at the ford." Ekwall writes that the exact meaning is unknown. Forster, quoted by Jackson.
Reculver Regulbium, Regulbi, ND. Racuulfe, Bede. Raculf, A S C, 669. supposes a tribal name Rutupones. Raculfcestre, 784, BCS, 173. Ricuulfi, 765, BCS, 199. Roculf, Domesday. Raculvre, 1276. Meaning: British ro, great, and British gulbio, beak; hence, Rogulbium, great headland or the promontory (Jackson). All experts agree.
Lympne, and the River Limene Portus Lemanis, loc. pl. Ant. It., 4. Rav., ND. Lemanio, Peut. Lemannonios bay, Ptol. Meaning: Old Irish lem, Irish leamh, Welsh llwyf, elm tree; hence, river of elms. All experts agree.
Durolevum Roman Station between Canterbury and Rochester., site un known, but probably at the west end of Faversham. Durolevo, Ant. It., 2. Peut. Durolavo, Rav. Meaning: British duro, fortress, British lava, a river name meaning "babbling brook" as in Gaulish labara, Welsh Llafer (Ekwall); hence, "the fort on the Lava."
Noviomagus Roman Station between Vagniacae and London, possibly near Crayford. Noviomagus, Ant. It., 2. Peut. Meaning: British novio, Latin novios, English new, and British magus, Irish mag, Breton, Welsh ma, field, place, and by extension fair or market, hence, Newmarket. There were other places so named in Britain and Gaul.
Vagniacae Roman Station at Springhead. Vagniacis, Ant. It., 2. In reply to my enquiry, Professor Jackson writes that he does not know of any explanation of the element Vagn. Unfortunately, we only have one literary reference to the name, as quoted, and are so unable to give variants which might help.
Darenth river Derquentid, Nen. The British origin of this name is the same for the rivers Derwent, Darwin, Dart, etc. References to the Derwent are Derventio, Ant. It., 1. Rav. ND. Deruuentio, Bede. Meaning: Primitive Welsh Derwn't, Old Welsh Derquentid, Middle Welsh Derwennyd, Anglo-Saxon Deorwente. Meaning: British derva, oak, Welsh derw, oak tree; hence,"the river of oaks."
Medway river Medeuuaege, 765-91, Miodowaege, 880, BCS, 195, Medwaeg, 894, Asser Meaning: Zach. is of opinion that the name is not of Celtic origin, but Ekwall suggests that it derives from Celt medu, Gaul., Old Cornish medu, Welsh medd, meaning mead, with the British river name Wey or Wye; hence, "the mead-coloured Wey." Zach. proposes the root med in Old English meduma, middling, i.e. "river in the middle."
Cray river Wallenberg doubts if the name has a Celtic origin, but Ekwall proposes that the name is derived from Middle Welsh crei, fresh, clean, and thus a British river name. There is a river Crei in Wales.

Certain points at once become apparent from a consideration of these placenames. Although they are recorded in Latin forms, yet they are entirely British in origin, for there are no Latin elements as in Caesarornagus, nor are any named for a god or person, as Londinium, "Londinos's Town," or Eburacum, "the estate of Eburos." On the contrary, both river and place-names are descriptive of associated natural features and circumstances, the exception being Durobrivae. Thus, the places are on the Lava, by the swamp or on the muddy creek, or by the ford, the! fort and port on the river of elms, the fort on the headland, the fort by the, waters while the rivers are clear, babbling, brown or where oaks abound. As to the county name, it would appear to be a question as to which came first, the land or the people. If the name Cantium (Kantion) predated the Belgic invasion or there was a tribe in Gaul called the Cantii, then the question could be resolved, but we do not know. It is perhaps not necessary to, assume that there were Belgic camps at Durobrivae and Durolevum, although there certainly was one such close to the site of Durovernum, for the native name for a fort could have been bestowed upon their own works by the Romans, who were fully acquainted with the Belgic speech and customs. The Romano-British name of Rochester raises the important question as to the antiquity of the bridge there. Were the Belgae capable of building a timber bridge across a possibly narrower but swifter Medway? There were three Roman stations with the suffix briva in their names, but there were several others whose names embodied the Latin pontem, and Professor Jackson suggests that perhaps the native term was used for rough timber structures while the Latin word described superior Roman works. He has also mentioned to me the possibility that Rochester "bridges" were bridges of boats, and this may well have been the case. But the whole question is bound up with that of the degree of Romanisation in the southeast of Britain which occurred between the near century, which lays between the incursions of Caesar and the Claudian conquest.

The evolution of these names subsequent to the Roman period is of great interest but the writer is not competent, neither is there space to develop the subject properly here. The reader is referred to the works consulted, especially that of Professor Jackson. But the following points can be made:

  1. The official Roman names of these places are best recorded in the Antonine Itinery, but we know nothing of the "popular" names, which have developed during the long period of occupation. For, in the mouths of even educated Romano-Britons, long names would tend to be clipped, and it may be that the names which we regard as "corrupt" were actually "popular" names.
  2. Civil administration of a Roman type did not break down with the departure of the Roman military establishment in AD 410; it was still operating in the south-east until near the middle of the century.
  3. The once-held opinion that the sea-invaders exterminated the Romano-British population is now known to be exaggerated. Large groups of Romano-Britons remained, and some of these were probably bi-lingual, speaking both British and Latin. From them, the invaders must have learned of the names of towns and forts which they, in their turn, must have tried to pronounce in the phonetics of their own language. Names transmitted would have been those of everyday speech.
  4. But the Roman official place-names did survive in a literary form for they were well known to Bede and to the scribes who drew up some of the early Kentish Saxon charters. As we shall see, these writers sometimes used the old names, and at other times the Old English names derived from them. The close connections of the early Kings of Kent with the Continent, and the establishment of the Roman Christian church with its Latin culture were probably factors which helped in the preservation of the old official names.
  5. The evidence for the continuity of occupation of Roman civil and military centres in Kent is mixed; for while Noviomagus, Vagniacae and Durolevum sank out of sight and out of mind, others, such as Rochester and Canterbury, survived and became the seats of the two earliest Roman Christian bishops in the first years of the seventh century, which was in contradiction to the usual early English customs whereby bishoprics were attached either to royal villas or to towns of English foundation. Continuity at Rochester seems to be indicated by the number of early pagan Saxon cemeteries in and around the town. It is not necessary to assume that the three lost castra noted above were destroyed during the conquest, for urban life declined during the last century of Roman rule, and many towns were partly depopulated or actually abandoned before the end of the Roman province. The forts of the Saxon Shore were probably evacuated in AD 410, but Dover, Richborough and Reculver were eventually re-occupied, although Lemanis, which has no post-Roman history, seems to have been permanently abandoned. But the A S C story of the sack of Anderita should give us pause before we adopt theories of abandonment.
  6. It should be noted that Romano-British towns, when walled and protected, were called castra (s. castrum), and this latter word was rendered in Old Welsh as cair, later caer, and in Old English as caestir, ceaster, later chester. Rochester retains the name, and at times Canterbury, Richborough and Reculver were all recorded as chesters. The bury and borough of Canterbury and Richborough derive from Old English burg, burh, town; the dative form of burh is byrig.
  7. Although the full story of the sound substitution which occurred when Romano-British names were adapted to Anglo-Saxon cannot be told here, yet a few points can be made. Dubris to Dofras, Dover illustrates the softening of b to f or v, and it should be noted that both the Romano-British and Anglo-Saxon names were in the plural, which leads Professor Jackson to suggest that the transformation of Roman-British place-names into Anglo-Saxon was made, in part, by the Britons during the bilingual period when they were speaking both their own and their conquerors' languages. Regulbium to Raculf, Reculver; there was no intervocal g in Anglo-Saxon, so c or gg was substituted, while b was softened to f or v and the Latin case-ending was discarded. Durobrivis to Hrofecaester, Rochester; according to Ekwall, the first syllable had always been weak, hence Robrivis, which reminds us of the corrupt Roibis of the Ravenna; the loss of the Latin case-ending and the v gave Robri, softening of b to f Rofri, and dissimulation of r finally produced Rofi. In spite of these changes, the old official name was kept in memory, for Bede uses it (Dorubreui) and, strangely enough, a Saxon charter of 844 has the form Dorobrevi. Yet more singular is the hybrid form Hrofibrevi used in Ethelbert's charter of 604, which the late Canon G. M. Livett regarded possibly as a sign of the genuineness of this charter (4). We have here an extraordinary overlapping of corrupt Romano-British forms with the developing Anglo-Saxon names. Bede writes of "Dorubrevis which the Englishmen called Hrofecaester," and explains the English name by a mythical Saxon chief Hrof, whose castle it was supposed to be. A more strange case is that of Canterbury, which lost its Romano-British name entirely although, as usual, Bede uses it constanly, and as late as the ninth century Asser refers to it as Dorubernia. But Bede also sometimes used Cantia for kingdom, city and see, and it may well have been the popular name for the nine syllable Romano-British official designation. We have some evidence for its early use, for Nennius preserves a list of the 28 cities of Britain, and in one of these, Cair Ceint, we can recognise Canterbury. Now a much earlier writer, Gildas, in the middle of the sixth century, also mentions this list, but does not stay to mention the cities. If we assume that the names were the same as those of the Nennian list, and there is no reason why we should not, then this Old Welsh list of names goes back at least to the beginning of the sixth century, which suggests that the name Cantia for Durovernum was very early in use. The evolution of the name Rutupiae is as controversial as the meaning of the Romano-British name, and the experts disagree on both these matters. As will have been noticed, the first syllable has used almost all the possible combinations as Ra, Ri, Ru and Re; the Antonine Itinerary names are generally accepted as the most accurate, which in this case is Ritupis. Bede writes of "Rutubi portus now corruptly called Reptacaestir," a change which illustrates metathesis, the transportation of t-p to pt. Since then, the name has undergone many changes from Repta to Raette, and then Rattes-, Rets-, Raborough and finally, in 1509, to Richborowe. There is no agreement among the experts as to how these changes came about, so the amateur must keep silent.
The remarks made in this paragraph give some examples of the probable changes in the evolution of the names; they make no pretence to be complete, nor are they in their correct sequences.
  1. Jackson, K., Language and History in Early Britain, 1953-5.
  2. Languages and grammar contractions used:
    C, Celtic;
    B, British;
    RB, Romano-British,
    Gaul., Gaulish;
    Bret., Breton;
    W, Welsh;
    Pr. W, Primitive Welsh;
    OW, Old Welsh;
    MW, Middle Welsh.,
    Old Ir, Old Irish;
    Ir, Irish;
    O. Corn., Old Cornish;
    L. Latin.,
    AS, Anglo-Saxon;
    OE, Old English;
    pl., plural;
    d. pl. dative plural;
    loc. pl. locative plural;
    Ice. locative;
    ace. pl. accusative plural;
    adj. adjective.
  3. Principal authors consulted, with contractions used:
    • Jackson.—Jackson, K., Language and History in Early Britain, 1953-5.
    • Arch. 93.—Richmond, I.A. and Crawford, O.G.S., The British Section of the Ravenna Cosmography, in Archaeologia 93, 1-50.
    • Jackson.—Jackson, K., On some Romano-British Place Names, in JRS, XXXVIII (1948).
    • Wall. I—Wallenberg, J. W., Kentish Place Names, 1931.
    • Wall. II—Wallenberg, J. W., The Place Names of Kent, 1934.
    • Ekwall I—Ekwall, E., Dictionary of English Place Names, 1947.
    • Ekwall II—Ekwall, E., English River Names.
    • Zach.—Zachrissen, R. E., Romans, Kelts and Saxons in Early Britain, Uppsala 1927.
  4. Livett, G. M. in Arch. Cant., XXI, 19.
Close Window
Accessed this page via search engine or bookmark? Full K A R Index here.