Explanatory alphabetical list of Romano-British place-names
Part 3: C
Part 3a: Ca to Ci
[For information as to which names are included in this list and which not, and an explanation of the abbreviations employed, click on Alphabetical List menu provided above]
CACTABACTONION
|
(Celt) |
(Rav) |
(Ptol) |
(AI) |
|
(Mod) |
|
Castaractonion |
Cactabactonion |
Caturactonium |
Cataractoni |
|
Catterick |
|
or |
(136) |
|
(Iter I, V) |
|
(North Yorkshire) |
|
Casataractonion |
|
|
Cataractone |
|
|
|
|
|
|
(Iter II) |
|
|
Note 1:
The traditional view is that this name is derived from Latin cataracta meaning 'waterfall, rapids'. Rivet and Smith prefer a derivation from hypothetical Celtic catu meaning 'battle' and hypothetical Celtic ratis meaning 'rampart', racte sometimes appearing as a variant of ratis. They thus suggest an original Celtic name of the form Catura(c)tonion meaning '(place of the) battle-ramparts'. They then suggest that before the AI was compiled "there occurred an assimilation by Latin speakers in the area to Cataracta via a mistaken association with the rapids of the Swale" (near Richmond).
Cas(a)taractonion is a topographical compound with the transitional element cas(a)t meaning 'steep hill high' and the inversion-type element ract meaning 'hill steep high'. The name presumably refers to the high ground to the west of Catterick Bridge, unless perhaps the Thornbrough promontory can sensibly be said to have steep and high sides. Ravenna's form shows the change of s to c (cf. Lectoseto → Lectoceto). The b in Ravenna's form is presumably the result of some medieval copying error - it is not likely to have been on the first-century map used by the compiler of Ravenna. The forms in Ptolemy and the AI all omit the s and it is this omission which has in the past caused scholars to confuse this name with the Latin cataracta meaning 'waterfall, rapids'. The variety of spelling of the name-ending in the Ptolemy and AI forms is not important. Note that the hill-letters s and r of Cas(a)taractonion correspond to the river-letters b and s, both of which are present in the name of the river, the Swale, flowing past the fort, the b first changing to v and the v later being anglicised to w. The l of Swale is a river-letter applied to minor rivers and tributaries by those who used the hill-letter n, it being noted that the Swale is a tributary of the Ure.
Note 2:
Example of omission of s
Example of s → c
CALAGUM
|
(Celt) |
|
(Ptol) |
|
(Mod) |
|
Calavum |
|
Calagum |
|
Healam Bridge |
|
|
|
|
|
(North Yorkshire) |
Note 1:
Rivet and Smith suggest Calacum as the correct form of this name. Following Jackson they favour a meaning ‘noisy stream’, based on a hypothetical root cal, taken to mean ‘to call’. They identify Burrow-in-Lonsdale as Calacum.
However, at the point where Ptolemy lists Calagum he is following an itinerary down the east side of the Pennines, Calagum being between Caturactonium at Catterick and Isurium at Aldborough. It seems quite clear that Calagum was the Roman fort/settlement at Healam Bridge, on Dere Street. But the name will have been transferred to Healam Bridge from an Iron Age settlement somewhere in that region, most probably East Witton Camp (NGR: SE 120 854) or nearby Castle Steads, Caldbergh (SE 105 853). It is possible that these two enclosures had the same name, one being a replacement for the other. Both enclosures stand on a steep hillside, so the Celtic name was probably Calavum (though with a different ending), where Calav means ‘steep hill slope’. See Home/Chapter 15, Part 1, 2 for an explanation as to why Ptolemy’s Calagum is not likely to have been the same place as Ravenna’s Caluvio or the Galacum of Iter X of the Antonine Itinerary.
Note 2:
Example of v → c → g
[The entry for Calagum was inserted on 01 May 2021]
CALCARIA see TADORITON
CALEBA ARBATIUM
|
(Celt) |
(Rav) |
(Ptol) |
(AI) |
|
(Mod) |
|
Calveva |
Caleba Arbatium (67) |
Caleva |
Galleva Atrebatum |
|
Silchester |
|
or |
|
|
(Iter VII) |
|
(Hampshire) |
|
Calveba |
|
|
Calleva |
|
|
|
|
|
|
(Iter XIII, XIV, XV) |
|
|
Note:
Holder identified a hypothetical stem calli, taken to mean 'wood', and Jackson thought the British name might have been Calleva, meaning '(town in the) woods'.
But it seems unlikely that the name refers to any wood around Silchester. It is much more likely that the name was simply transferred to a new settlement at Silchester from the Iron-Age hillfort at SU 628 631, about one kilometre northwest of Romano-British Calleva. That hillfort stands on a steepish slope and so most probably had a name of the form Calveva or Calveba, where eva/eba is just an ending and the old-style element Calv means ‘steep hill slope’. Calv is similar to Banv, but uses the hill-letter l rather than n and the adjectival c meaning ‘steep’ rather than b meaning ‘high’. And just as Banva changed to Banna so Calve changed to Calle. In the present case there can be no doubt that the name changed from Calveva/Calveba to Calleva in the pre-Roman period since, as Rivet and Smith point out, two coins of Eppillus, king of the Atrebates, have been found, one bearing the inscription REX CALLE and the other CALLEV. In effect the Atrebates did what the Romans were to do on a number of occasions later - they moved the inhabitants of a hillfort (l1-people, perhaps Durotriges) to a new settlement and transferred the name of the old settlement to the new one. Evidence of that new settlement, which in its late phase is believed to have been an oppidum, has been found underneath Romano-British Calleva at Silchester.
[The entry for Caleba was last modified on 29 December 2020]
CALLEVA see CALEBA ARBATIUM
CALUVIO
|
(Celt) |
(Rav) |
|
(AI) |
|
(Mod) |
|
Caluvion |
Caluvio (112) |
|
Galacum |
|
Ingleton |
|
|
|
|
(Iter X) |
|
(North Yorkshire) |
Note 1:
Rivet and Smith equate Ravenna's Caluvio with the Galava of the AI and provide an etymology based on the latter form - they follow Jackson's suggested derivation from hypothetical gala with an ava suffix, the sense being 'vigorous stream'.
Caluvio, however, simply comprises the old-style topographical element caluv meaning 'steep hill slope'. Caluvion will have been the name of the Iron Age settlement known as Yarlsber Camp (NGR: SD 710 726) about one kilometre east of Ingleton. The settlement does indeed stand on the slope of a steep hill, the highest part of the settlement being around the 250 metre contour, the lowest part at around 235 metres. The Romans will simply have transferred the name to some installation which they built at Ingleton. The identification of Ingleton as Caluvio/Galacum is explained in Chapter 15: Roman place-names in the north of England.
The AI form Galacum shows the common interchange of c and g and replaces the uvio of Caluvio by acum, suggesting that someone had mistakenly taken uvio to be just a name ending.
Note 2:
Example of c → g
[The entry for Caluvio was last modified on 26 November 2023]
CAMBODUNO
|
(Celt) |
|
|
(AI) |
|
(Mod) |
|
Cambroduno |
|
|
Camboduno |
|
Eccleshill |
|
|
|
|
(Iter II) |
|
(Bradford) |
Note 1:
Rivet and Smith see this name as being derived from hypothetical Celtic cambo given the meaning 'curved, crooked' and hypothetical dunos, taken to mean 'hill' originally but later taking on the meaning 'fort', the name as a whole thus meaning 'fort at the bend' (of a river).
Camboduno is, however, a straightforward topographical name, but the name is incomplete. The cam element means 'steep hill' and duno means 'summit of hill'. The letter b in topographical names means 'high' and comes before its hill-letter - there is thus a hill-letter missing after the b and it seems most likely that that hill-letter was r. The reason for this is that there is an error in Iter II for the journey from Tadcaster to Manchester. The Iter gives 20 Roman miles as the distance between Tadcaster and Camboduno and then 18 miles for the journey on to Manchester, which is of course much too little. The clue is that the Roman fort at Slack, to the west of Huddersfield, was Ravenna's Cambroianna, originally Cambrolanda, so it would appear that some copyist has been confused by two adjacent names both beginning with Cambro. The 20 Roman miles is the distance from Tadcaster to Cambroduno, but the 18 Roman miles is the distance from Cambroianna to Manchester - the copyist has missed out the journey from Cambroduno to Cambroianna. Cambroduno, referring to a site on the summit of a high, steep hill is a perfect topographical description of Eccleshill in the northern outskirts of Bradford, and Eccleshill is precisely 20 Roman miles from Tadcaster.
Note 2:
Example of missing letter - r
CAMBORICO
|
(Celt) |
|
|
(AI) |
|
(Mod) |
|
Camborico |
|
|
Camborico |
|
Thetford |
|
|
|
|
(Iter V) |
|
(Norfolk) |
Note:
Rivet and Smith prefer the form Camboritum and see this as being derived from hypothetical cambo, taken to mean 'curved, crooked', and hypothetical ritu, taken to mean 'ford', the name as a whole perhaps meaning 'the ford at the bend (of the river)'.
But Camborico is a perfectly acceptable topographical name as it stands. It is a compound of the old-style element cam, meaning 'steep hill', and borico, which may comprise the old-style element bor, meaning 'high hill', with an ico ending, or the transitional element boric, meaning 'high hill steep'. The river-letters corresponding to the hill-letters m and r of Camborico are r and s, which are present in the river-name Isur, the then name of the Ure/Ouse in Yorkshire and perhaps also of the Little Ouse river in Norfolk. The Celtic place called Camborico may then have been adjacent a high, steep hill close to the Little Ouse river. It is thus proposed that Camborico was at Thetford. Presumably the Romans had a post there to control movement across the Thet and Little Ouse rivers. But the name Camborico seems inappropriate for the Celtic hill-fort surrounding the Norman motte and bailey castle in Thetford, unless that hill-fort is on an eminence which does not show up on the topographical maps available to the present writer. It seems much more likely that the original Camborico was the Iceni religious centre which is said to have stood on Gallows hill in the northern outskirts of the modern town. That hill can reasonably be said to be high and steep and it is close to the Little Ouse river. The name may have been shared with the hill-fort down by the Norman castle and then the Romans simply adopted the name for their own post. But note that this identification of Camborico rests upon one or two questionable assumptions:
- some inattentive medieval copyist swapped over the mileages quoted opposite Camborico and Duroliponte in Iter V: the Iter should read
Icinos
Camborico XXV
Duroliponte XXXV
- there was a direct road from Icinos at Caistor St. Edmund over to the Old Buckenham area where the traveller could get on to Margary road 37 heading west for Thetford; and
- there was a direct road from Newmarket to Cambridge.
But even making the above assumptions there is still an error of about one mile in each of the mileages indicated above. The identification of Thetford as Camborico must therefore be provisional.
CAMBROIANNA
|
(Celt) |
(Rav) |
|
|
|
(Mod) |
|
Cambrolanda |
Cambroianna (167) |
|
|
|
Slack |
|
|
|
|
|
|
(W of Huddersfield) |
Note 1:
Rivet and Smith see this name as a duplicate of Gabaglanda, the Ravenna form of Cambaglanda at Birdoswald (though they locate it at Castlesteads).
Cambroianna, however, with two very minor amendments, is a straightforward topographical name - the i should be an l and the second n a d, thus yielding the form Cambrolanda. As for the change from anda to anna compare Ravenna's Gabaglanda with the Amboglanna of the ND. Cambrolanda is a compound in the hill-letters m, r, l and n, where cam means 'steep hill', br means 'high hill' and and means 'hill summit'. The name as a whole thus refers to a location at the top of a high, steep hill, a definition which is entirely appropriate for Slack, to the west of Huddersfield. For an explanation of the identification see Chapter 15: Roman place-names in the north of England.
But note that Cambrolanda will have been the name of an Iron Age hillfort or settlement at the top of a high, steep hill, so if there had been no such hillfort or settlement on the actual site of the Roman fort at Slack then the place-name must have been transferred to Slack from some other location in that region, for example from the hillfort on nearby Old Lindley Moor.
Note 2:
Example of l → i
Example of anda → anna
[The entry for Cambroianna was last modified on 01 February 2024]
CAMULODONO
|
(Celt) |
(Rav) |
(Ptol) |
|
|
(Mod) |
|
Camulodono |
Camulodono (111) |
Camunlodunum |
|
|
Skipton |
|
|
|
|
|
|
(North Yorkshire) |
Note 1:
Camulodunum is generally held to be the correct form of this name. The name is conventionally held to comprise the name of the Celtic war-god Camulos and an element dunum thought to be derived from hypothetical Celtic dunos, believed to have meant 'hill' originally but to have taken on the meaning 'fort' at a later date, the name Camulodunum thus meaning 'fortress of Camulos'.
But Camulodono is in fact a straightforward topographical compound in the hill-letters m, l and n. The cam element means 'steep hill' and don means 'summit of hill'. The reasons for identifying Skipton as Camulodono are explained in full in Chapter 15: Roman place-names in the north of England. Note that the hill-letters m and n correspond to the river-letters r and l (for minor rivers and tributaries), both of which are present in the name Eller of the stream on which Skipton stands. The inversion-type element dono is normally placed at the end of the name, but the people who applied that element to the place-name used their river-letter l in an inversion-type manner, i.e. they placed their river-letter l before the river-letter r corresponding to the hill-letter m. One sees exactly the same development at Cambodunum at Kempten in Bavaria. This name also comprises the hill-letters m and n and Cambodunum stood at the top of a high, steep embankment on the eastern side of the river Iller, which is of course the same name as Eller. In Chapter 15 the writer assumed that the Roman fort had been more or less where Skipton castle now stands, but the name merely refers to a location on the summit of a steep hill, so that the Roman fort may have been almost anywhere along the top of the steep drop down to the Eller Beck at Skipton.
The Ptolemy form has an intrusive n. Alternatively that n may be the surviving ending of the original name cam, meaning 'steep hill', cam being the earliest element in Camunlodunum.
Note 2:
Possible example of intrusive letter - n
Possible example of the survival of the ending of original name
CAMULODUNO
|
(Celt) |
(Rav) |
(Ptol) |
(AI) |
|
(Mod) |
|
Camanulodulo |
Manulodulo Colonia |
Camudolanum |
Camuloduno |
|
Colchester |
|
or |
(99) |
(Iter IX) |
(Essex) |
||
|
Cam(an)uloduno |
|
|
|
|
|
Note:
This name is generally thought to be derived from the name of the Celtic war-god Camulos, with a dunum ending, which is thought to be derived from hypothetical Celtic dunos, taken to have meant 'hill' originally and to have taken on the meaning 'fort' later. The place-name is thus thought to mean 'fortress of Camulos'. It is noteworthy, however, that in the many forms of this name available in the ancient sources - and the name is also given by Pliny, Tacitus and Cassius Dio - not one single form includes the s of Camulos. And in the case of the similar name Camulosessa, which does include an s before the essa-ending, it can be shown that the Celtic name was most probably of the form Lucamosessa (see entry for Camulosessa), so in this case the name had nothing to do with the Celtic war-god Camulos.
The Ravenna form of this name has simply lost its initial C, as happened to quite a few other names in Ravenna, e.g. Armis (42), Ardaoneon (43) and Arduaravenatone (9). The place-name thus comprises the inversion-type element dul, meaning 'summit of hill', attached to the old-style name Camanulo, where the an of this form is just the surviving ending of the name when it existed in the Cam form, i.e. before the hill-letter l was added. There are other place-names where an old-style element in one hill-letter is followed by an inversion-type element in that same hill-letter, for example Omirededertis (the original form of Ravenna's Omiretedertis) and Bannatia. The structure of the name Camanulodulo is thus not surprising. The dulo ending must then have changed to the more familiar duno before the Roman invasion, since Rivet and Smith draw our attention to a coin of Cunobelinus, who is thought to have reigned from about 9 AD to about AD 40, that coin including the name CAMULODUNO in its inscription. But it is much more likely that the dulo in Ravenna's form is simply the result of a spelling error and should be duno, the full name Cam(an)uloduno then indicating that the name changed from Cam(an)ulo to Cam(an)uloduno when Colchester was taken over by the Catuvellauni. The initial cam element of the name is an old-style element meaning 'steep hill', so the place-name tells us that the pre-Roman Celtic settlement at Colchester was on the summit of a steep hill. The best location would be up on the ridge later occupied by the Roman fortress and colony, but if archaeologists are quite certain that there had been no Celtic settlement on that site then presumably Celtic Camuloduno was a little to the west, up on the high ground in the Lexden area, from which steep slopes drop down towards the river Colne. A third possible location is Hythe Hill. This is admittedly not so high as the other two suggested locations, but there is nothing in the name Camuloduno to indicate that the hill was high, only that it was steep, and the drop down to the river Colne from Hythe Hill is steep.
The Ptolemy form is presumably just the result of copying error.
[The entry for Camuloduno was last modified on 27 February 2021]
CAMULOSESSA
|
(Celt) |
(Rav) |
|
|
(ND) |
(Mod) |
|
Lucamosessa |
Camulosessa Presidium |
|
|
Praesidio |
Malton |
|
|
(172, 173) |
|
|
|
(North Yorkshire) |
Note 1:
The name Camulosessa is traditionally seen as a compound of Camulo and sessa, where Camulo is taken to be the name of the Celtic war-god Camulos, and sessa is thought to be derived from a hypothetical sessa meaning 'seat', the name as a whole thus meaning 'seat of Camulos'.
The name, however, seems quite clearly to be topographical, though the first element has been slightly rearranged. Camulosessa appears to be a name in the hill-letters m, l and s which correspond to the river-letters r, t and b. These three river-letters are all present in the river-name Derventione (the modern Derwent at Malton), though the t has changed to d and the b to v. Now, the first element in the place-name is the old-style cam (meaning 'steep hill') so the l and s must also be used in the old-style manner. That is to say, Camulos is a purely old-style topographical name where the hill-letters are in the chronological order first m, then l and finally s, and the corresponding river-letters r, t and b must occur in the same chronological order in the river-name. But the order of the river-letters in Derventione is first t (changed to d), then r and finally b (changed to v). Now there can be no doubt as to the spelling of the river-name since it occurs various times in various regions of Britain. It follows that the order of the hill-letters in the place-name must originally have been first l, then m and finally s. In other words the Celtic name of the place may have been Calumosessa. But even this form is not satisfactory since the chronological order of the hill-letters is l, m, s, which is wholly inconsistent with the normal order n1, s, m, r, l1, n2, l2. The first part of the Celtic name must therefore have been somewhat of the form Lucamosessa, where the hill-letters m and l are used in an inversion-type manner, the latest element of the name being Luc, meaning 'hill steep', where the hill-letter l is l1. Note that Lucamos corresponds to the river-name element Derben (precursor of the modern river-name Darwen). The t in Derbentione (earlier form of Derventione) must therefore have been added by people who used the hill-letter l2. These people seem always to have used l2 in an inversion-type manner but to have placed their river-letter t at the end of an existing river-name - this presumably accounts for the second t in some at least of the Trisantonis names and for the final t in modern river-names such as Dart, Teviot, Tweed (t changed to d) and Teith (t changed to th). The people who added the river-letter t (and the ione ending) to the river-name Derben to form the new river-name Derbentione appear to have been the Parisi. Presumably the Romans simply replaced the unfamiliar Lucamo by the more familiar Camulo - after all they had occupied Camuloduno at Colchester for more than 20 years when they built the fort at Malton. We can thus be quite certain that the place-name Camulosessa has no connection with the Celtic war-god Camulos. The development of the essa-ending is discussed fully in Updates 19 October 2015. It was originally esca, an inversion-type element meaning 'hill steep'. It was of course an element in the hill-letter s, and so this hill-letter did not appear earlier in the name. But by the time the ending reached Britain it had changed to essa and it seems it was no longer regarded as an element in the hill-letter s, so the s will be found in the earlier part of the name (where the earlier part is a place-name rather than a river-name), and the ending appears to have taken on the additional meaning of 'overlooking a river'. In other words the ending was employed by people who used the hill-letter s and was applied by them to places at the top of a steep slope and overlooking a river. What this means is that the original fort at Malton will have been at the top of the steep river embankment to the south of the later fort, so traces of it should be sought underneath the vicus of the later fort.
The term Presidium is Latin and is often taken to mean 'small fort, post'. However there were very many Roman forts in Britain, many of them small, but only this particular one includes Presidium in its name, and indeed it is identified simply as Praesidio in the ND. This uniqueness caused the writer to suggest in Chapter 20: Rome's frontiers in northern England that the Malton fort enjoyed some special status, perhaps as a forward command centre at the time when the 9th legion was still based at Lincoln but was active up the eastern side of the Pennines after Vespasian gave the order to bring the Brigantes under Roman control.
Note 2:
Example of replacing an unfamiliar name by a similar familiar name
CAMUNLODUNUM see CAMULODONO
CANTIVENTI
|
(Celt) |
(Rav) |
|
(AI) |
(ND) |
(Mod) |
|
Candiventena |
Cantiventi |
|
Clanoventa |
Glannibanta |
Watercrook |
|
(a river) |
(115) |
|
(Iter X) |
|
(Cumbria) |
Note 1:
Rivet and Smith take Glannoventa to be the correct form, this being a hybrid of the AI and ND forms. They take the first element to be derived from hypothetical glanno given the meaning 'bank, shore'. The second element has caused scholars considerable difficulty in the past. Williams suggested that venta might mean 'field', with a secondary meaning 'market place'. Following that suggestion Glannoventa might perhaps mean 'shore field' or 'shore market'. Rostaing, on the other hand, suggested a hypothetical pre-Indo-European vin meaning 'hill'. For a full discussion of the problem see Rivet and Smith 1979, 262.
But Cantiventi, which is the earliest form of the name available from our sources, appears to be a river-name of the kind comprising a river-suffix attached to a place-name with one or more hill-letters. The river-suffix would be vent (seen also in Derventio and Aventio) comprising the river-letters b (changed to v) and t corresponding to the hill-letters s and l respectively. The place-name might have been the transitional element cant (meaning 'steep hill high') but is perhaps more likely to have been the old-style element Cand (meaning ‘steep hill summit’). Candion will have been the name of the hillfort now known as ‘Castlesteads, The Helm’ (NGR: SD 531 887) overlooking the river Kent from the summit of a steep hill a little south of Oxenholme. The river-name will have been somewhat of the form Candiventena and this form, with the common d→t change and modification of the ending, yielded the Cantiventi of Ravenna. Note that the hill-letter n of Cand may be n1, suggesting that the hillfort may date back to the 3rd or even 4th century BC. The Roman fort at Watercrook lies in a loop of the river Kent, so the Romans will simply have transferred the name of the river to their fort. And of course the Cantiventi, i.e. the ‘Cant river’, is now the river Kent. Presumably C [antiv] ent [i] → Cent → Kent.
The forms in the AI and ND both have an intrusive l and the ND form shows the change of v to b. The vowel changes are of no consequence.
Note 2:
Example of intrusive letter - l
Example of v → b
[The entry for Cantiventi was last modified on 15 April 2021]
CANUBIO
|
(Celt) |
(Rav) |
|
(AI) |
|
(Mod) |
|
Canubio |
Canubio (83) |
|
Conovio |
|
Caerhun |
|
(a river) |
|
|
(Iter XI) |
|
(Conwy) |
Note 1:
Rivet and Smith see Canovium as being the correct form of this name and Jackson believes this to be derived from hypothetical cano meaning 'reed'.
Ravenna's Canubio, however, appears to be based on a composite river-name Canubena comprising the river-letter b used as a suffix to the old-style place-name Can, meaning 'steep hill'. Can was not a free-standing place-name but one element of the compound place-name Descecanglanion, this being the name of the tribal centre of the Deceanglani tribe. Descecanglanion was most probably the hillfort known as Castell Caer Seion just west of the mouth of the river Conwy in North Wales. Descecanglubena was the composite river-name of the Conwy, comprising the place-name element Descecangl and the river-suffix ubena. The Romans shortened Descecanglubena to Canubena and then transferred the shortened river-name in the form Canubio to the fort which they built at Caerhun.
The AI form shows the change from b to v. The change had evidently already taken place by the time the inscription RIB 2265 was produced (which Rivet and Smith date to AD 120-121), since this gives the spelling Kanovio. Again the changes in the vowels are of no consequence.
Note 2:
Example of b → v
[The entry for Canubio was last modified on 21 September 2024]
CANZA
|
(Celt) |
(Rav) |
|
|
|
(Mod) |
|
Cantia |
Canza (27) |
|
|
|
Henton |
|
|
|
|
|
|
(Somerset) |
Note:
There appears to be no generally accepted etymology for this name. Jackson suggested three possible roots: firstly, a hypothetical canto, taken to mean 'rim, border, circumference, circle'; secondly, Middle Welsh cant, taken to mean 'host, party'; and thirdly, hypothetical canto, taken to mean 'a hundred'. Quentel, citing hypothetical Gaulish cant, identified in many place-names the senses '100', 'circle' and 'edge'. Richmond and Crawford concluded that the name might mean 'corner stream', and Rivet and Smith thought 'river at the edge' might also be possible, perhaps in the sense of a river at the edge of a tribal area.
But cant appears to be a straightforward transitional topographical element meaning 'steep hill high'. One sees the element in Cunetione (Mildenhall), in Gabrocentio (Hard Knott) and in Concanata (Bowness-on-Solway, Concanata apparently being the original form of the ND's Congavata). At the point where Canza appears in Ravenna the compiler appears to be listing place-names on alignments out of Martock/Ham Hill (in Somerset). The name before Canza is Lindinis, which appears to refer to Dundon Hill. Now if one draws a line on a map from Martock/Ham Hill to Dundon Hill and then keeps straight on, one comes to Henton in Somerset, to the west of Wells. Henton lies on a thread of high land striding E-W over what was presumably marshland in the Romano-British period and, just as important for present purposes, it lies at the foot of a steep, high hill, so that the name Cantia would be appropriate for the location. Moreover, it is possible that the Hen of Henton is derived from the Can of Canza.
[The entry for Canza was last modified on 23 March 2021]
CARBANTIUM
|
(Celt) |
(Rav) |
|
|
|
(Mod) |
|
Carbandium |
Carbantium (161) |
|
|
|
Harrogate |
|
|
|
|
|
|
(North Yorkshire) |
Note:
Rivet and Smith, following Richmond and Crawford, see this name as a corruption of Ptolemy's Carbantorigum. Unfortunately it is in fact Carbantorigum which is the corrupt form - it is a conflation of the two adjacent Ravenna names Carbantium and Tadoriton in their original forms, i.e. Carband(ium + Cartad)oriton, with the subsequent changes d → t and t → g.
Carbantium is a topographical compound in the hill-letters r and n. The element Car is an old-style element meaning 'steep hill'. The element bant is impossible as the hill-letter n is qualified by b and t, both of which mean 'high' in topographical names. The second element of the name must therefore originally have been old-style band meaning 'high hill summit'. Carbandium was thus an Iron Age settlement at the top of a steep, high hill, the name then being transferred to a Roman fort, the order of names in Ravenna indicating that that fort was at Harrogate. The Harr of the modern name is presumably just an anglicized version of the Carb of Carbantium, the C changing to H and the rb to rr.
CARBANTORIGUM
|
(Celt) |
|
(Ptol) |
|
|
|
|
Carband(ium) + (Cartad)oriton |
|
Carbantorigum |
|
|
|
Note 1:
The correct form is normally assumed to be Carbantoritum, assumed to be derived from hypothetical carbanto, taken to mean 'wagon, chariot', and hypothetical ritu, given the meaning 'ford', the name thus meaning 'wagon-ford, chariot-ford'.
However, Ptolemy's Carbantorigum, Uxellum and Corda appear to be not in the territory of the Selgovae, as stated in the Geography, but in the north of England. And Carbantorigum appears to be not a place-name, but a conflation of the two adjacent Ravenna names Carbandium (original form of Carbantium) and Cartadoriton (original form of Tadoriton). Some copyist was apparently copying Carbandium - he wrote down Carband and then when he turned to his primary text again his eyes fell on the Car of Cartadoriton rather than the Car of Carbandium and then he wrote down what he saw after the d, namely oriton. He had thus produced the false name Carbandoriton, and it is this name, with the minor change of d → t and t → g, which has come down to us as Ptolemy's Carbantorigum.
Note 2:
Example of confusion of two names
CATARACTONE/CATARACTONI/CATURACTONIUM see CACTABACTONION
CAUSENNIS
|
(Celt) |
|
|
(AI) |
|
(Mod) |
|
Arcusendis |
|
|
Causennis |
|
Grantham |
|
or |
|
|
(Iter V) |
|
(Lincolnshire) |
|
Racusendis |
|
|
|
|
|
Note:
Rivet and Smith point out that the element caus is known in Gaulish toponymy, but that no meaning can be assigned to it. They also point out that Holder listed a suffix enna, apparently used in such names as Arduenna, Ravenna and Vienna, but suggest no meaning for this either.
Causennis appears, however, to be a topographical name, but one which has been modified. The au and nn letter combinations look suspect. There was probably a consonant between the a and the u in the original Celtic name and the nn will have been some other letter combination originally, probably nd (cf. Cambaglanda → Amboglanno of the ND). It is here proposed that the Celtic name had been of the form Carusendis and that Carusendis was located on a promontory overlooking Grantham from the east, at a point to the west of Cold Harbour on Ermine Steet. The name itself is an old-style compound with the elements car, meaning 'steep hill', the hill-letter s and end meaning 'hill summit'. Several points speak in favour of the above form of the name and the identification proposed:
- the location is substantially 30 Roman miles from Water Newton and 26 Roman miles from Lincoln, these being the figures quoted in Iter V for the journeys from Durobrivas to Causennis and from Causennis to Lindo;
- the promontory stands some 70 metres above the town of Grantham and has fairly steep sides - the name Carusendis is thus appropriate for the location;
- the hill-letters r, s and n of Carusendis correspond to the river-letters s, b and m (for major rivers), and the b (changed to v and then the v anglicised to w) and m are seen in the name Witham, the river at Grantham; and
- it is possible that with the common changes of c→ g, omission of intervocalic s, and d → t, the Carusend of Carusendis became the Grant of Grantham.
But the form Carusendis is not entirely satisfactory. This is because, contrary to the normal chronological order of the hill-letters, n2 came before l1 in Lincolnshire, this being evidenced by the names Croconcalana (original form of Crococalana) and Banvobalum (original form of Bannovalum). The initial L of Lindum is thus the hill-letter l1 used in an inversion-type manner. What this means is that the river-letter t, changed to th, was added to the river-name after the river-letter m. It follows that the river letter b, changed to v and the latter then anglicised to w, was added after the river-letter t. What this means is that the hill-letters s and r must have been added to the place-name which became Causennis in an inversion-type manner. We are therefore looking for a Celtic place-name somewhat of the form Arcusendis or Racusendis. We would therefore have to assume that the Romans dropped the r, changed nd to nn and rearranged Acusennis to give Causennis. This is credible, but whether Arcusend/Racusend developed to form the Grant of modern Grantham is not clear.
[The entry for Causennis was last modified on 04 December 2020]
CELOVION
|
(Celt) |
(Rav) |
|
|
(ND) |
(Mod) |
|
Celovion |
Celovion, Celunno |
|
|
Cilurno |
Chesters on the |
|
|
(178)(147) |
|
|
|
North Tyne |
|
|
|
|
|
|
(Northumberland) |
Note 1:
Rivet and Smith follow Richmond and Crawford in seeing Gelovium as being perhaps the correct form of Celovion. Richmond and Crawford suggested a derivation from hypothetical gelov taken to mean 'sword', the name then perhaps meaning 'place on sword river'. Dillemann saw Celovion as being perhaps a corruption of Cilurnum. It is now clear that Dillemann was right to equate Celovion with Cilurno, though Celovion is not a corruption of Cilurno but an earlier version of the same name. Earlier writers have indeed equated Celunno with Cilurno, Williams drawing attention to Irish cilorn, Old Welsh cilurnn and Welsh celwrn meaning 'bucket, pail'. Rivet and Smith suggest hypothetical cilurno, perhaps referring to a natural feature such as a pool in the North Tyne.
In Chapter 20: Rome’s frontiers in northern England, the writer explained that Celovion was a Flavian fort at Chesters on the North Tyne, Celunno a Trajanic fort and Cilurno the Hadrian’s Wall fort, though the change from Celunno to Cilurno may have taken place some considerable time after the Hadrianic fort was built. It seems clear that Celovion was originally the name of the Iron Age settlement standing on the steep slope (Celov means ‘steep hill slope’) immediately east of the village of Wall, on the eastern side of the North Tyne. The Romans will simply have transferred the name of that settlement to their Flavian fort, built somewhere in the vicinity, though not necessarily on exactly the same site as the later Hadrianic fort. The name was later transferred, slightly modified to Celunno, to a nearby Trajanic frontier fort, presumably located on the Stanegate and possibly close to the point where the road crossed the North Tyne. Some 20 years later the name will have been transferred to the Hadrian’s Wall fort a little upstream at Chesters, the name at some point changing to the Cilurno of the ND. Note that all the way from Bowness to Carrawburgh Hadrian’s Wall was built a little forward of the Trajanic frontier forts, in some places by as little as 200 metres, though at Carrawburgh by more than 3 kilometres. It seems clear that this forward spacing of Hadrian’s Wall was also observed in the Chesters area. Presumably the Trajanic forts remained in operation throughout the period when the Wall was being built, most of them being dismantled after the troops moved into their new forts built on the Wall itself.
Note 2:
Example of change of ending from ovion to unno
Example of nn → r n
[The entry for Celovion was last modified on 18 March 2021]
CELUNNO see CELOVION
CERMA
|
(Celt) |
(Rav) |
|
|
|
(Mod) |
|
Cerma |
Cerma (205) |
|
|
|
Dalginross |
|
|
|
|
|
|
(Perth and Kinross) |
Note:
Rivet and Smith see Cerma as a corrupt form of Ptolemy's tribal name Caereni, taking the view that the compiler of Ravenna misread a tribal name on his map as a habitation name. But that argument is unconvincing since the Caereni are believed to have occupied an area in Sutherland, in the far north of Scotland, whereas Ravenna is at this point listing names in central Scotland, so that such a misreading of the map is highly unlikely.
Cerma appears to be a straightforward topographical name in which the hill-letter m is qualified by the old-style element cer meaning 'steep hill'. The steep hill in question may be that just south of the Dalginross fort or that on the other side of the river Earn, immediately north of the village of Comrie. It is likely, indeed, that the very name Comrie is derived from Romano-British Cerma. But note that the name Cerma may originally have been that of a Celtic hill-fort and was simply transferred by the Romans to their own fort at Dalginross. There are two hill-forts in the vicinity, one further up the river Earn at Dundurn and the other a little downstream at Thornhill Lodge.
CERMIUM
|
(Celt) |
(Rav) |
|
|
|
(Mod) |
|
Cermium |
Cermium (223) |
|
|
|
Forteviot? |
|
|
|
|
|
|
(Perth and Kinross) |
Note:
Rivet and Smith see this name as being a duplicate of Cerma (205), but it is clear from the order of names in Ravenna that the two names refer to two quite different places. Likewise, for reasons given in the entry for Cerma, there is no justification for following Rivet and Smith in seeing Cermium as a corrupt form of the tribal name Caereni.
Cermium appears to be a topographical name in which the hill-letter m is qualified by the old-style element cer meaning 'steep hill'. On spacing grounds the Roman fort called Cermium should have been on or close to the Water of May in Strathearn. There have been finds of Bronze Age material at Forteviot, on the Water of May, and indeed of Roman pottery, though this is no proof of Roman occupation. The village of Forteviot stands on a terrace above the Water of May and that terrace appears to have steepish sides on its W and NE sides, though whether this accounts for the cer element of the name Cermium is not clear. That element might equally refer to the steep slope on the northern bank of the river Earn at that point or to the steep northern flanks of the hills to the south of Forteviot. But note that Cermium may actually have been the name of one of the local hill-forts and was simply adopted by the Romans for the fort which they built down in the valley. There are two hill-forts in the vicinity, one on the Law of Dumbuils, a little southeast of the village of Forgandenny, and the other a little further south, this fort now being known as Castle Law fort, Forgandenny.
CIBRA
|
(Celt) |
(Rav) |
|
|
|
(Mod) |
|
Cibra |
Cibra (199) |
|
|
|
Bearsden |
|
|
|
|
|
|
(East Dunbartonshire) |
Note:
Rivet and Smith see Cibra as a corrupt form of Ptolemy's Coria of the Damnoni, and they provide a derivation on the basis of the Coria form.
Cibra, however, appears to be a straightforward old-style topographical name in which the element cibr means 'steep high hill'. As explained in Chapter 22 (The Antonine Wall) of the Home menu, Cibra was an early Antonine fortlet located a little to the east of the Antonine Wall fort at Bearsden and built as part of the preparations for the construction of the Wall itself. In all probability the name was transferred to the Wall fort at Bearsden when this was built.
[The entry for Cibra was last modified on 26 September 2020]
CICUTIO
|
|
(Rav) |
|
|
|
(Mod) |
|
|
Cicutio (56) |
|
|
|
Pumpsaint |
|
|
|
|
|
|
(Carmarthenshire) |
Note:
Rivet and Smith prefer the form Cicucium and point out that Williams had identified a root cic, thought to be related to Irish cich, meaning 'pap, breast', and Welsh cig, meaning 'meat, flesh'. Richmond and Crawford described the hill on which the Y Gaer fort stood as being "a rounded spur which is strikingly like a female breast in contour and outline". So, perhaps with Irish cich in mind, they suggested that Cicutio might have been the fort at Y Gaer, west of Brecon.
Cicutio appears, however, to be a topographical name which has lost its hill-letters (cf. the cocc of Coccio or Coccuveda). The name may originally have had a form such as Circultio, comprising the old-style element cir, meaning 'steep hill', and the transitional element cult meaning 'steep hill high'. Circultio will have been the name of a Roman fort and the Romans simply transferred the name to the river on which the fort stood, presumably the river Cothi in South Wales. Circultio is suggested purely by way of example - the original Celtic name may have had a different combination of hill-letters. But note that no matter which hill-letters were present in the name, the name as a whole refers to a steep, high hill and is thus eminently suitable for Pumpsaint on the river Cothi - there are steep, high hills immediately west, north and east of the fort site.
CILURNO see CELOVION
CINDOCELLUM
|
(Celt) |
(Rav) |
|
|
|
(Mod) |
|
Cindocelum |
Cindocellum (204) |
|
|
|
Dumbuck |
|
or |
|
|
|
|
or |
|
Cindoceldum |
|
|
|
|
Dumbarton |
|
|
|
|
|
|
(West Dunbartonshire) |
Note 1:
The second element of this name is generally assumed to be derived from hypothetical ocelo, taken to mean 'headland, promontory, spur'. The cindo element has caused some difficulty in the past. Rostaing suggested derivation from a pre-Indo-European root kend or kind meaning 'mountain'. Richmond and Crawford thought Cindo might be a slight corruption of Cinto, this being taken to be derived from hypothetical cinto, thought to mean 'first'. The name as a whole could thus mean 'chief headland' or 'principal spur'.
But Cindocelum, assuming there should be only one l, is a straightforward Celtic topographical compound in which the old-style element cel meaning 'steep hill' is qualified by the earlier old-style element cind meaning 'steep hill summit'. Note that the name actually indicates that there was a Bronze Age or Iron Age structure called Cindo on the summit of a steep hill and that the name later passed to the hill itself, this becoming known as 'Cindo steep hill', i.e. as Cindocelum. The Roman fort called Cindocelum then took its name from the hill. Thus if the Roman fort was at Dumbarton, perhaps to protect a harbour used by the Roman navy, then there must have been a Bronze Age or Iron Age structure on the summit of Dumbarton Rock. If the Roman fort was in the Dumbuck area, perhaps to control what is believed to have been the lowest fording point on the Clyde, then Cindocelum may have been the Iron Age fort on Sheep Hill, this being said to stand on top of a Bronze Age dun. Note that the cel element, with the river-letter t added as a suffix, yields a river-name somewhat of the form celota, which is of course Ptolemy's Clota, the modern Clyde. Alternatively, the ll of Ravenna's Cindocellum may have been ld in the Celtic name, the latter then comprising two elements cind and celd both meaning 'steep hill summit'. It is even possible that Cindo was the name of the Bronze Age dun on Sheep Hill and Cindoceldum the name of the Iron Age hill-fort.
The above might be thought to favour Dumbuck as the site of Ravenna's Cindocellum, but there is an argument in favour of Dumbarton, too. If Ravenna's Cindocellum had earlier been Cindosellum (cf. Ravenna's Lectoceto which was originally Lectoseto) then we would have in the place-name Cindosellum the hill-letters n and s corresponding to the river-letters l and b, and both of these river-letters appear, with the b changed to v, in the river-name Leven, the Leven being the tributary which joins the Clyde at Dumbarton, immediately adjacent Dumbarton Rock.
Note 2:
Example of doubled consonant - l
[The entry for Cindocellum was last modified on 08 April 2021]
CIRONIUM DOBUNORUM
|
(Celt) |
(Rav) |
(Ptol) |
|
|
(Mod) |
|
Cironium |
Cironium Dobunorum (66) |
Corinium |
|
|
Cirencester |
|
|
|
|
|
|
(Gloucestershire) |
Note:
There appears to be no generally accepted etymology for this name.
The first element of this name, no matter whether spellt Cir or Cor, is an old-style topographical element meaning 'steep hill'. There are a few steepish slopes not too far from Cirencester, but the name seems much more appropriate for Bagendon, about three miles to the north. It seems likely, then, that Cironium was the name of the Celtic settlement at Bagendon and was simply transferred by the Romans to a fort which they built at Cirencester, the vicus of this fort developing later into the civitas capital of the Dobunni.
