Explanatory alphabetical list of Romano-British place-names
Part 1: A
[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]
ABALLABA see AVALAVA
ABISSON
| (Celt) | (Rav) | (Mod) | |||
| Abisson | Abisson (175) | Durham |
Note:
Rivet and Smith see the first element of this name as being derived from hypothetical Celtic ab meaning 'river'. Williams, following Holder, notes that hypothetical isso was a common suffix in Gaul, for example in certain personal names. In place-names, Rivet and Smith observe, its meaning may simply be 'place where', the name as a whole thus perhaps meaning 'river-place, place by a river'.
Abisson, however, is one of about ten names in Roman Britain which have an essa-type ending. The ending came to Britain from the Continent and was originally of the form esca, meaning 'hill steep'. It was regarded as an element in the hill-letter s. By the time the ending reached Britain, however, it appears that it was no longer regarded as an element in the hill-letter s and it appears to have taken on the additional meaning 'overlooking a river'. The ending was thus applied by the people who used the hill-letter s and it was applied by them to places at the top of a steep hill and overlooking a river. In some place-names the essa-ending is added on to the end of a land-name, i.e. a place-name comprising hill-letters, whereas in others it is added on to the end of a river-name. Abisson is a name of the second kind, the b of Abi just being the river-letter b (employed by those who used the hill-letter s). This b became the initial V of Vedra, the Romano-British name of the river Wear. Note that the original Abisson will have been the Iron Age promontory fort now called Maiden Castle, the name simply having been transferred by the Romans to a fort which they built nearby, probably a little downstream at a point controlling a crossing of the river.
[The entry for Abisson was last modified on 03 August 2020]
ABONE
| (Celt) | (Rav) | (AI) | (Mod) | ||
| Abone | Punctuobice (47) | Abone (Traiectus) | Sea Mills | ||
| (a river) | (Iter XIV) | (Bristol) |
Note 1:
Rivet and Smith see a derivation for the AI form from hypothetical British abona meaning 'river'. Richmond and Crawford see Punctuobice as a scribal error for Portu Abone. The argument appears to be that the name had become unintelligible on a map, the first element then being assimilated to Latin punctu meaning 'point' and the second to Latin obice meaning 'bar, bolt; barricade'.
There can be no doubt that the fort at Sea Mills was called Abone, this name having been transferred by the Romans from the river, now the Avon, to the fort. But the name after Abone in Iter XIV is Traiectus and it is clear that this is not a separate place-name - it belongs to Abone and indicates that there was a ferry service at Sea Mills, either across the Avon or across the Severn estuary, perhaps both. It would appear that at some point the army left the site but the vicus lived on under the name Abonetraiectusvicus, and it is this name which leads us to Ravenna's Punctuobice. The form Abonetraiectusvicus was simply shortened by omitting initial A and some internal letters to yield Bonctusvicus. The b then changed to p and the v to b, this yielding Ponctusbicus, and this form, with further minor amendment, yielded Ravenna's Punctuobice.
Note 2:
Example of shortening a name by omitting internal letters
Example of b → p
Example of v → b
AD PONTEM
| (AI) | |||||
| Ad Pontem | |||||
| (Iter VI) |
Note:
Ad Pontem is generally identified as a Roman settlement on the Fosse Way at East Stoke, Thorpe, Nottinghamshire. But that cannot be correct. To understand why one needs to look at the names on both sides of Ad Pontem.
The Iter VI names Verometo, Margiduno, Ad Pontem and Crococalana are generally identified as Willoughby, Castle Hill, East Stoke and Brough respectively, all on the Fosse Way. But the name Margiduno (referring to a location on the summit of a steep hill) is not at all appropriate for Castle Hill and Crococalana is wholly inappropriate for Brough. Crococalana is an old-style topographical name with three elements all meaning 'steep hill' (the second hill-letter is missing) - and there most certainly are no steep hills at Brough, or anywhere near it on the Fosse Way. It is thus clear that Iter VI does not in fact follow the Fosse Way all the way from Leicester to Lincoln. The Iter VI traveller went up the Fosse Way as far as Six Hills and then turned right on to Margary Road 58a. He followed that road over to the Grantham area, kept going straight on as far as Ermine Street and there turned left for Lincoln. And on this route there are indeed locations which are wholly appropriate for the Iter VI names listed above - and those locations are substantially the correct distances apart. Following the mileage given in Iter VI Verometo, or Vernemeto as it is better known, was on road 58a in the vicinity of Scholes Farm, a little west of Grimston. Note that road 58a is a ridgeway and the area around Scholes Farm is up on top of the ridge, hence the element med in Bernemedo (the original form of Vernemeto), where med is an old-style element meaning 'hill summit'. Twelve Roman miles further on was Margiduno at Croxton Kerrial. At that point on the road there is a northwards extending promontory bounded by steep slopes to the E, N and W - now that is a suitable site for a Margiduno (arg = 'hill steep'; dun = 'summit of hill'). Another seven miles brings us to the crossing of the Witham just south of Grantham, and Ad Pontem will have been at that crossing. It is not clear whether the bridge was at Great Ponton, at Little Ponton or at a point even nearer to Grantham, though the mileage given in Iter VI seems to point to the third possibility - Ad Pontem may thus have been the Roman settlement at Saltersford. A little more than seven Roman miles brings us to Ancaster, and it is Ancaster which must have been Crococalana. The name will originally have been Croconcalana and will have been the name of the Iron Age settlement at SK 987 433, next to Castle Quarry just south of modern Ancaster. The name will have been transferred to the later Iron Age settlement down on the valley floor and it is this which will later have become Romano-British Crococalana. The An of Ancaster is presumably all that is left of the old name. But there is an error in Iter VI as to the distance from Ancaster to Lincoln. That distance is given as xii miles in Iter VI and xiiii miles in Iter VIII, whereas it is actually xvii miles. The v has simply been omitted in Iter VI and miscopied as ii in Iter VIII. The above-mentioned place-names are discussed in more detail in the individual entries.
[The entry for Ad Pontem was last modified on 09 July 2019]
AESICA see ESICA
ALABUM
| (Celt) | (Rav) | (Mod) | |||
| Alabum | Alabum (55) | Llandovery | |||
| (a river) | (Carmarthenshire) |
Note:
Williams appears to favour a derivation from hypothetical alebh or alobh, apparently taken to mean 'ridge, crest; cockscomb'. The name Alabum might then mean 'hill, crest'.
The Llandovery fort stands on raised ground between the river Towy and its tributary the Brân, though closer to the tributary. Alabum appears to be a river-name. It may be a compound in the river-letters l and b, where the river-letter l is generally applied to minor rivers, including tributaries of major rivers. Alabum may thus have been the then name of the Afon Brân and the Romans simply transferred the river-name to their fort. On the other hand Alabum could be a river-name of the kind in which a river-suffix, here ab, including the river-letter b, is attached to a place-name, here Al, the place-name in this case just comprising the hill-letter l. But again the Romans will have transferred the river-name to their fort. Alabum cannot be a topographical land-name in the hill-letter l because in such a name the letter b meaning 'high' would come before the l, not after it as in this name.
ALAVNA
| (Celt) | (Rav) | (Ptol) | |||
| Al Abona | Alavna | Alaunus | |||
| (a river) |
Note 1:
This name, written as Alauna, has caused scholars some difficulty in the past. Watson suggested derivation from a hypothetical British root al or alo meaning 'rock', the name then meaning 'rocky place' when applied to a place or 'rocky one' when applied to a river. Others have suggested that since the name is so widely used it must have an adjectival sense, with Ekwall suggesting 'holy' or 'mighty', Rivet and Smith suggesting 'good' and Pokorny suggesting 'shining, brilliant'.
The name, however, written as Alavna, appears to be a straightforward compound river-name in the river-letters l and b (changed to v). It will have begun as Abona, a simple river-name in the river-letter b. Then the people who used the river-letter l called it the river Abona or, in their language, the Al Abona. The b then changed to v and Alavona was shortened to Alavna. This river-name was adopted by the Romans and given by them to forts which they had built on the banks of the rivers concerned, for example forts at Ilchester, Ilminster (not yet found), Stratford-on-Avon (not yet found), Ravenglass and also apparently at Alnwick (not yet found).
Note 2:
Example of b → v
ALBINUMNO
| (Celt) | (Rav) | (Mod) | |||
| Alvinundo | Albinumno (51) | Welshbury Camp | |||
| (Gloucestershire) |
Note 1:
Rivet and Smith think Albiniano might be the correct form of this name, this comprising a personal name Albinianus, so that the name might mean 'at Albinianus's'.
The order of names in Ravenna appears to indicate that Albinumno was west of the Severn and generally northeast of Lydney. The Celtic name was most probably of the form Alvinundo where old-style alv means ‘hill slope’. Ravenna’s form shows the common change of v to b. The undo element may have changed to unno (compare Ravenna’s Gabaglanda and the later Amboglanna of the Notitia Dignitatum) and medieval m/n confusion may have caused the change of unno to umno. The name Alvinundo is, however, rather odd in having one element referring to a slope and another referring to a summit, the only other example which springs to mind being Celtic Vresmedenaci which became Ravenna’s Bresnetenaci at Ribchester. But the name would be entirely appropriate for the hill-fort now called Welshbury Camp, to the northeast of Cinderford in Gloucestershire, and this hill-fort is in the right area to be Ravenna’s Albinumno. The structure of the hill-fort is unusual in that there is an early enclosure to which a southwards-extending enclosure was added later at its southeastern corner (see, for example, the Pastscape website of Historic England). But it is the location of those enclosures that is important. The later enclosure appears to stand on the summit of the hill at an altitude above the 175 metre contour. The southeast corner of the earlier enclosure is at this same height, but by far the greater part of the earlier enclosure is on the hillside, on the slope. A name such as Alvinon (where inon is just an ending) would thus be appropriate for the early enclosure. And since the later enclosure is on the highest point of the hill an element undo, where nd means ‘hill summit’, would be appropriate for that enclosure. It is thus entirely possible that the early enclosure had a name of the form Alvinon, using the hill-letter l1, and then the hill-fort was taken over by n2-people who applied the element undo because the part of the fort built by them was on the summit of the hill. The name Alvinundo, in which nd meaning ‘hill summit’ is the generic element and is qualified by the earlier name of the hill-fort, taken to be Alvin, would thus be entirely appropriate for the site. Note that the Celtic name may have been Balvinundo, where initial b means ‘high’, or Calvinundo, where initial c means ‘steep’. There are other examples where the Ravenna form of a name appears to have dropped the initial consonant of the Celtic name, eg. Arduaravenatone, Armis, Evidensca, Itucodon and Onno.
It is not clear whether the name as it appears in Ravenna refers to the hill-fort itself or to a Roman fort built somewhere in the vicinity and to which the name of the hill-fort was transferred. Note that there is a Roman road (Margary 614) running north-south at the foot of the western side of the hill on which the hill-fort stands.
Note 2:
Example of v → b
ALICINCA see ALICUNA
ALICUNA
| (Rav) | (Ptol) | (Mod) | |||
| Alicuna (110) | Olicana | Castleshaw | |||
| (Oldham) |
Note:
Richmond and Crawford appear to see the name as being related to Welsh al meaning 'rock', the name perhaps meaning 'rocky hill'. Rivet and Smith appear to doubt that the above form is a correct reading of the name in the Ravenna manuscripts. They seem to prefer a reading Aluna, with a variant form Alicinca. They think the Aluna form might be an abbreviation of Alauna and that the Alicinca variant might have borrowed its last two letters from the beginning of the next-following Ravenna name, Camulodono. They suggest that Alicinca might correspond to Ptolemy's Olicana, a point also made by Richmond and Crawford for their Alicuna.
There is thus some uncertainty as to the form of this name, but as it stands, as given by Richmond and Crawford, it comprises the inversion-type topographical element lic meaning 'hill steep'. The name thus appears entirely appropriate for the fort at Castleshaw, which is surrounded by steep hills. And Rivet and Smith's variant form Alicinca includes two inversion-type elements lic and inc, both meaning 'hill steep'.
ALIONE see ALONE
ALITACENON
| (Celt) | (Rav) | (Mod) | |||
| Alitacenon | Alitacenon (164) | Adel | |||
| (Leeds) |
Note:
Rivet and Smith change this name to Alaunocelum. They see the first part of this form as being the well-known name Alauna (written as Alavna by the present writer since it is a river-compound in the river-letters l and b, changed to v) and the second part as being derived from hypothetical British ocelo, taken to mean 'headland, promontory, spur'.
Alitacenon appears, however, to be a straightforward topographical name comprising the inversion-type element litac meaning 'hill high steep'. Alitacenon was between Maporiton at Bramham and Loxa at Exley Head, Keighley. Adel seems the most likely location - indeed it is possible that the very name Adel is derived from the Alit of Alitacenon. The place-name appears to refer to an early fort and suggests a location close to Black Hill, though there are other slopes to the SE and SW which might also be called high and steep.
ALONE
| (Celt) | (AI) | (ND) | (Mod) | ||
| Alibone | Alone | Alione | Burrow in Lonsdale | ||
| (a river) | (Iter X) | (Lancashire) |
Note:
Rivet and Smith do not see these AI and ND names as referring to the same place, but to two different places both called Alauna. For their derivation of the name Alauna see the entry for Alavna.
This name appears to be a river-name which the Romans transferred to a fort which they built close to the river concerned. The river in this case is the Leck Beck, a tributary of the Lune, and the fort is that called Burrow in Lonsdale, sometimes referred to as Nether Burrow. The name was probably earlier Alibone, a river-compound in the river-letters l and b, where the river-letter l appears to have been applied to minor rivers, including tributaries of major rivers, by those who used the hill-letter n. If then the intervocalic b is dropped one has the Alione form, and if the i is then dropped one has the Alone form. What is a little unsatisfactory here is that the Alione form appears in the ND, which is normally considered to be later than the AI. Regarding the modern river-name it is possible that there was some movement of people within the island after the Romans left and that those who coined inversion-type place-names in the hill-letter l settled in the Lune valley and added their river-letter t to the end of the river-name Alone, this giving a form such as Alonet, which with the omission of intervocalic n and change of t to c yields the form Alec, which may possibly explain the modern river-name Leck.
ALOVERGIUM
| (Celt) | (Rav) | (Mod) | |||
| Alobergium | Alovergium (22) | Ilfracombe | |||
| (Devon) |
Note:
Rivet and Smith prefer the form Alobergium and see this as being derived from hypothetical British alo, taken to mean 'rock', and hypothetical berg, taken to mean 'hill, mountain; heap', the name as a whole then perhaps meaning 'rocky hill'.
Alovergium, however, appears to be a compound in the hill-letters l and r, where the inversion-type element verg means 'slope of hill steep'. It is, however, possible that the verg element was originally the transitional element berg meaning 'high hill steep', the element possibly referring to the hill which shelters Ilfracombe harbour from the open sea, though there are higher hills on the landward side of the harbour. There is, however, a hill-fort on the high, steep hill just E of the harbour, so it is probable that Alobergium was the name of that hill-fort and the Romans simply transferred the name to a fort which they built nearby. The Ilfra of the modern name may be derived from the Alover part of Alovergium, it being noted that the initial Al of Alavna Silva at Ilminster and of Alavna Colonea at Ilchester appears as Il in the modern names, so it is possible that the initial Al of Alovergium became the initial Il of Ilfracombe.
ALUNA see ALICUNA
AMBOGLANNA see GABAGLANDA
ANDERELIONUBA
| (Celt) | (Rav) | (ND) | (Mod) | ||
| (S)anderelionuba | Anderelionuba (68) | Anderidos | Pevensey | ||
| (a river) | (var. Anderitos) | (East Sussex) |
Note 1:
Anderitum is usually taken to be the correct form of this name, and it is traditionally taken to be derived from a hypothetical root ande assigned the meaning 'great, big' and a hypothetical root ritu, taken to mean 'ford'. It is, however, surprising that an adjective as common as 'great, big' should only appear in this one name out of all the known names from Roman Britain. And one needs to be careful with ritu - of the five other Romano-British names said by Rivet and Smith to include this element (Camboritum, Carbantoritum, Duroritum, Maporitum and Tadoritum) three of them (Carbantoritum, Maporitum and Tadoritum) can be shown to have no connection with fords.
Although given as a place-name in Ravenna Anderelionuba is in fact a river-name of the kind comprising a river-element, here the river-letter b, used as a suffix to a place-name, here Anderelion, comprising the old-style element and meaning 'hill summit' and the hill-letters r and l. Anderelion was thus a place on the summit of a hill, or at least on the top of raised ground, adjacent the river concerned. The Romans then transferred the river-name Anderelionuba to a fort which they built close to that river, though of course the fort was not necessarily at or close to the native settlement called Anderelion. Note that the presence of the river-letter b, corresponding to the hill-letter s, in Anderelionuba suggests that the Celtic place-name may actually have been Sanderelion, where the initial S is used in the inversion-type manner.
The development from Anderelionuba to Anderitos/Anderidos seems quite clear. Bearing in mind that we are speaking of a river-name, the people who used the hill-letter l2 appear always to have placed their river-letter t at the end of an existing river-name. One sees this in the river-name Derventio and also in Trisantonis (the second t), and it is fairly common in modern river-names such as Tweed (t changed to d), Teviot, Derwent, Dart and Teith (t changed to th). If these people at some time occupied parts of southern England they may have added their river-letter t to the end of Anderelionuba, thus yielding a river-name somewhat of the form Anderelionubita. One should not worry about the i before the t - one sees this i in the modern river-names Nith (t changed to th) and Nidd (t changed to dd) and also in the Romano-British Nido (river-name transferred to a fort, the t being changed to d). This river-name Anderelionubita was then transferred by the Romans to their fort at Pevensey and the name was at some time shortened, understandably enough, to Anderita, or to Anderida in which the t is replaced by d, as often. Presumably there was an early Roman fort called Anderelionuba at Pevensey and the much later Saxon shore fort was called Anderelionubita, shortened to Anderita. The native settlement called Anderelion was presumably upriver from Pevensey, since if it had been at Pevensey the Romans would surely have called their early fort Anderelion rather than have transferred the river-name to the fort. The change of the ending of the name from a in Ravenna to os in the ND is of no relevance to the meaning of the name. And of course if the writer's reasoning is correct, this name has nothing to do with any ford, no matter how great or big.
Note 2:
Example of addition of letter to a name - river-letter t
Example of change of t to d
Example of shortening, simplification of a name by deleting internal letters
[The entry for Anderelionuba was last modified on 08 April 2021]
ANDERIDOS/ANDERITOS see ANDERELIONUBA
ANICETIS
| (Celt) | (Rav) | (Mod) | |||
| Anicetis | Anicetis (35) | Ansty | |||
| or | (Wiltshire) | ||||
| Danicetis |
Note:
Rivet and Smith take the view that this name is probably the designation of an estate, based on the Latin personal name Anicetus. They see Anicetis as being a plural form, perhaps meaning 'the family of Anicetus', or 'the descendants of Anicetus'.
The nicet of Anicetis, however, appears to be a straightforward inversion-type topographical element meaning 'hill steep high'. There is a steep slope in the village of Ansty itself and another just a little south on the other side of the A30. But there is a fairly large hillfort (now called Castle Ditches) about a kilometre to the NNE of the village, which may account for the interest the Romans took in this area - Anicetis appears within a group of Ravenna names which appears to date from a period shortly after the invasion of AD43. It is thus most likely that Anicetis was actually the name of that hillfort - the fort is built around the summit of a steep, high hill - and was later simply transferred to the village now called Ansty. Note that the name may originally have been Danicetis, where danicet means 'summit of hill steep high'.
APAUNARIS
| (Celt) | (Rav) | (Mod) | ||||
| Abravnaris | Apaunaris (20) | Braunton | ||||
| (Devon) |
Note:
Rivet and Smith see Apaunaris as a corrupt form of Aquis Sulis (at Bath). Dillemann sees it as a corruption of A Tamaris.
The order of names in Ravenna appears to suggest that Apaunaris was a Roman fort at Braunton in northwest Devon. But the name was probably transferred by the Romans from the Celtic hill-fort now called The Castle, which lies just north of Braunton at a point near the village of Knowle. That hill-fort does not stand on the summit of a high hill, but on the side of the hill, on the slope. It is probable, then, that the p was originally a b meaning 'high' and the u a v meaning 'slope'. The Celtic name will then have been somewhat of the form Abavnaris, but the hill-letter is missing from the first element of the name. On the assumption that the Braun of modern Braunton is based on the Celtic name the first element may have included the hill-letter r - it may have been of the form Abrav, similar to the Brav of Bravogenium, apparently the original form of Ravenna's Branogenium (at Leintwardine). If this is correct the name of the hill-fort may have been Abravnaris, or possibly Abarvnaris. It is not clear whether the n is intrusive, is the hill-letter n or was the ending of the name before aris was added. The aris ending is presumably the same ending as in Brocara, Valteris and Lavaris.
[The entry for Apaunaris was last modified on 29 March 2022]
ARAMIS see ARANUS
ARANUS
| (Celt) | (Rav) | (Mod) | |||
| Saranus | Aranus (34) | Old Sarum | |||
| (Wiltshire) |
Note:
Rivet and Smith prefer the reading Aramis and see this name as a duplicate of Ravenna's Armis (42), but the order of names in Ravenna makes it clear that the two names refer to two quite different places. Rivet and Smith appear to see Armis as being a river-name ultimately derived from hypothetical Indo-European ora, taken to mean 'to set in motion'. Alternatively, if the name is adjusted to hypothetical Arnus, they see it as being a Celtic river-name Arno. As a third possibility, if the form Aramis be taken as correct, they see this as a British equivalent of hypothetical Gaulish aramon, taken to be related to the modern English word 'arm', the meaning of the name then being perhaps 'arm of the sea'.
Ravenna includes three adjacent names Alavna Coloneas Aranus. It seems quite clear, however, that the Ravenna text has been wrongly divided at this point and that the names should be Alavna Colonea Saranus. Saranus is simply a compound in the hill-letters s and r. Note that the river-letter corresponding to the hill-letter s is b and this survives, changed to v, in the name of the river Avon, which flows past the foot of the hill crowned by the Old Sarum fort.
ARBEIA
| (Celt) | (ND) | (Mod) | |||
| Verbeia | Arbeia | Newton Kyme | |||
| (a river) | (North Yorkshire) |
Note 1:
Rivet and Smith offer no derivation of their own and do not indicate that they are aware of a derivation proposed by anyone else. They identify South Shields as Arbeia on the ground that "this is the only suitable place for a Numerus Barcariorum", this being the unit stated in the ND to be stationed at Arbeia.
However, if one respects the order of the names in the ND it is clear that Arbeia was the Roman fort at Newton Kyme. An altar found at Ilkley bears an inscription (RIB 635) which reads, translated, 'sacred to Verbeia'. As Rivet and Smith point out Verbeia will in fact have been the then name of the river Wharfe. The name is a river-name of the kind comprising a river-element, here the river-letter b, used as a suffix to a place-name, here the element ver meaning 'slope of hill'. The place with the element ver in its name will of course have been a native stronghold or settlement, not a Roman fort, though if it was a stronghold then ver was probably ber originally, this element meaning 'high hill'. The Romans simply transferred the name of the river to the fort which they built on the banks of the Wharfe at Newton Kyme, though the initial V has been dropped or lost in the ND's Arbeia. The name Arbeia is discussed in detail in Chapter 7 of the Home menu.
Note 2:
Example of missing initial letter - V
[The entry for Arbeia was last modified on 03 August 2020]
ARDAONEON
| (Celt) | (Rav) | (Mod) | |||
| Cardadonecon | Ardaoneon (43) | Harting Beacon | |||
| (West Sussex) |
Note 1:
Rivet and Smith think the correct form might be Portus Ardaoni or Ardaonium. They appear to follow Richmond and Crawford in taking the view that the first part of the name might be derived from hypothetical British ard, taken to mean 'height'. They tentatively identfy the Roman fort at Portchester as Portus Ardaonium.
The writer’s reasons for associating Ravenna’s Ardaoneon with Harting Beacon are given in Chapter 11 (Roman place-names in central southern England) of the Home menu. The hill on which the hillfort stands has very steep sides and so it is probable that the Celtic name had a C at the beginning, the element Card meaning ‘steep hill summit’. For some reason Ravenna drops the initial letter of quite a number of names, for example (C)armis, (Lu)coganges¸ (L)itucodon. The second part of Ardaoneon does of course look very odd, but it seems most likely that it is a duno-type element conveying the same information as the first part of the name. The element will then be donec meaning ‘summit of hill steep’. One sees the same element as denac in Vresmedenaci, apparently the Celtic form of Bresnetenaci at Ribchester. All that will have happened is that intervocalic d and c of Cardadonecon have been dropped, this being a common change in Romano-British place-names. The name Cardadonecon indicates that the hillfort had been occupied by an r-people, but was taken over by a people using the hill-letter n2, i.e. by the Atrebates, where the term Atrebates is used broadly here so as to embrace also the groups later referred to by Ptolemy as the Belgae and the Regni. All three groups, if indeed they were separate groups at that early date, appear to have used the hill-letter n2 (see 'Ptolemy’s Celtic tribes in Britain').
It is of course not clear from the name itself whether Ravenna refers to the hillfort, to a Roman post possibly built inside the hillfort or to a Roman post built somewhere else in the vicinity and to which the name of the hillfort was transferred. It may, however, be possible that the Hart of Harting is just an anglicised version of the Card of Cardadonecon.
Note 2:
Example of missing initial letter - C
[The entry for Ardaoneon was last modified on 11 May 2019]
ARDOTALIA see ZERDOTALIA
ARDUARAVENATONE
| (Celt) | (Rav) | (Mod) | |||
| Arduaravenatone | Arduaravenatone (9) | Okehampton | |||
| or | (Devon) | ||||
| Carduaravenatone | |||||
| (a river) |
Note 1:
Rivet and Smith appear to regard the duaravenatone part of the name as a garbled version of Derventio. If the name refers to a place, they argue, then Derventio might be preceded by A or Ad, as if the name had been read from an itinerary. If the name is that of a river then initial Ar might be a mistake for the abbreviation Fl of Latin Flumen, meaning ‘river’. Richmond and Crawford see a possible derivation from ardu, taken to mean ‘height’, and aravenatone, related perhaps to Welsh araf meaning ‘gentle’ or arafhan meaning ‘to quieten’.
Arduaravenatone is a river-name transferred by the Romans to the fort which they built at Okehampton. In other words Arduaravenatone was the then name of the river now called the Okement. It is a river-name of the kind comprising a river-suffix, in this case venatone, attached to a place-name, in this case Arduara, where Ard is an old-style element meaning ‘hill-summit’. But because the name includes an old-style element in the hill-letter r and the hill-summit letter d Arduara will have been a Celtic hill-fort/settlement on top of a hill adjacent the river. The only such hill-fort/settlement on the OS maps appears to be that on Castle Hill, just west of Woolleigh Barton, though the river is today called the Torridge at that point. The hill has very steep sides to the north and south of the hill-fort/settlement, so the Celtic name was probably Carduara, where card means ‘steep hill summit’. This means that the Celtic river-name Arduaravenatone/Carduaravenatone was applied to the river Okement and that part of the modern Torridge from the confluence with the Okement down to Bideford, the upper reaches of the modern Torridge thus having had some other Celtic name. The above argument is borne out by the modern name Bideford. Bideford is given as Bedeford in the Domesday Book and Bede is just a shortened form of venatone with the v changed to b and the t to d, both common changes in Romano-British place- and river-names. One may deduce from the above that the English settlement called Bedeford was founded, and named, before the river-name changed to Torridge.
Note 2:
Example of missing initial letter - C
[The entry for Arduaravenatone was last modified on 03 August 2020]
ARGISTILLUM
| (Celt) | (Rav) | (Mod) | |||
| Argistilum | Argistillum (63) | Stretton Grandison? | |||
| (Herefordshire) |
Note:
Williams seemed to see this name as being derived from a hypothetical prefix are or ar, taken to mean 'in front of', and a British word which gave rise to Welsh gwystl meaning 'hostage', the place-name then meaning 'at the hostage'. Rivet and Smith seem prepared to accept the view of Williams, but think it also possible that a hypothetical Celtic arganto, taken to mean 'silver', might be involved.
On the assumption that there should be only one l, the name comprises the inversion-type elements arg and st respectively meaning 'hill steep' and 'hill high', together with the hill-letter l. The hill in question might be one of those to the east and south of the site of the fort on the south side of the river Frome, or that on the Stretton Grandison side of the river. However, the identification of Argistillum with the fort at Stretton Grandison can only be provisional - it is clear from Ravenna that Argistillum was between Kenchester and Worcester, and Stretton Grandison is currently the only known fort along that alignment. But Stretton Grandison is some considerable distance from Worcester, so there may have been another and as yet undiscovered fort between the two.
ARICONIO
| (Celt) | (AI) | (Mod) | |||
| Ariconio | Ariconio | Weston-under-Penyard | |||
| or | (Iter XIII) | (Herefordshire) | |||
| Dariconio |
Note:
Rivet and Smith see this name as being derived from hypothetical British are or ari, taken to mean 'in front of', and hypothetical conio of unknown meaning.
The Romano-British town called Ariconio is described as having been at Bury Hill, a little to the east of Weston-under-Penyard. The name appears to be a straightforward topographical name where the inversion-type element ric means 'hill steep'. The steep hill in question is presumably that immediately southwest of Weston-under-Penyard itself. It is probable, however, that Ariconio had been the name of the hill-fort in Chase Wood, a little further to the SW of Weston-under-Penyard and that the Romans simply transferred the name to the new settlement at Bury Hill. The hill on which that hill-fort stands has very steep sides. Since the hill-fort is at the summit of a steep hill it is possible that the Celtic name had in fact been Dariconio, where daric means ‘summit of hill steep’. There are other examples where the Romano-British form of a name appears to have dropped the initial consonant of the Celtic name, eg. Arduaravenatone, Armis, Evidensca, Itucodon and Onno.
ARMIS
| (Celt) | (Rav) | (Mod) | |||
| Carnis | Armis (42) | Butser Hill | |||
| (Hampshire) |
Note 1:
Rivet and Smith see this name as a duplicate of their Aramis, which is the same name as Richmond and Crawford's Aranus (Old Sarum), though it is clear from the order of names in Ravenna that Armis and Aramis were two quite different places. Rivet and Smith take the view that Armis might be a river-name, ultimately derived perhaps from hypothetical Indo-European ora, taken to mean 'to set in motion'. And if Armis is adjusted to hypothetical Arnus, then it is the common hypothetical Celtic river-name Arno.
The writer’s reasons for associating Ravenna’s Armis with the hillfort on Butser Hill are explained in Chapter 11 (Roman place-names in central, southern England) of the Home menu. This is not a hillfort in the conventional sense, one with a complete circuit of bank and ditch. Rather it is a hilltop having various spurs branching off it, each spur being interrupted by at least one ditch. Presumably at other points the very steep slopes were thought to give adequate protection. But precisely because of those very steep slopes it is probable that the Celtic name had an initial C, the element Car meaning ‘steep hill’. The hill-letter m is normally chronologically earlier than the hill-letter r, so it is probable that the m of Armis is just the result of m/n confusion and that the name should really be Carnis, this indicating that the hillfort was at some time taken over by people who used the hill-letter n2, i.e. by the Atrebates, where the term Atrebates is used broadly here so as to embrace also the groups later referred to by Ptolemy as the Belgae and the Regni. All three groups, if indeed they were separate groups at that early date, appear to have used the hill-letter n2 (see 'Ptolemy’s Celtic tribes in Britain'). This occupation by an n2 -people is seen also in the names of other hill-forts in that region – Claducendum (Clausentum) at Old Winchester Hill, Cardadonecon (Ardaoneon) at Harting Beacon and (Navi)magno (Navimago) at The Trundle – so it is likely also to have occurred at Butser Hill.
It is of course not clear whether with Armis Ravenna is referring to the hillfort itself, to a Roman post possibly built inside the hillfort or to a Roman post built somewhere else in the vicinity and to which the name of the hillfort was transferred.
Note 2:
Example of missing initial letter - C
[The entry for Armis was last modified on 03 August 2020]
AVALAVA
| (Celt) | (Rav) | (ND) | (Mod) | ||
| Avalava | Avalava (153) | Aballaba | Burgh-by-Sands | ||
| (Cumbria) |
Note 1:
Conventionally the form Aballava is assumed to be correct and to be derived from hypothetical Celtic aballa meaning 'apple'. Rivet and Smith suggest the name might allude to a sacred apple tree or to an apple orchard. Such a meaning, however, seems somewhat incongruous for a military fort.
The val of Ravenna's form seems to be a straightforward topographical element meaning 'side of hill' or 'slope' and Burgh-by-Sands III did indeed stand on a slope. This fort appears to have been the earliest of the three forts at Burgh-by-Sands, so Avalava was presumably the original Celtic name. The first v changed to b very early since Aballava is the form appearing on the Rudge cup and Amiens patera, and both cup and patera appear to have been produced during the Trajanic period.
Note 2:
Example of v changed to b
AXELODUNO see OLERICA
