Explanatory alphabetical list of Romano-British place-names


Part 6:  H → K



[For information as to which names appear in this list and which not, and an explanation of the abbreviations employed, click on Alphabetical List menu provided above]





This name is known from an inscription (RIB 1225) found at the Roman fort at Risingham, in Northumberland.


Rivet and Smith see this name as being Habitancum and as including the cognomen Habitus. The ancum suffix might be similar, they think, to hypothetical Celto-Latin aco or acum, taken to mean 'property of', so that the name as a whole might indicate that the land on which the Roman fort was built belonged to a family called Habitus.


It is likely, however, that this name was originally somewhat of the form Gamblidandi, comprising the old-style elements Gam, blid and and respectively meaning 'steep hill', 'high hill summit' and 'hill summit'. All that will have happened is that the intial G changed to H, the hill-letters m and l were dropped, d changed to t (a fairly common change in Romano-British place-names) and the second t changed to c (the t/c interchange is also known in Romano-British place-names). The river-letters corresponding to the hill-letters m and l of Gamblidandi are r and t, and these are found, with the t changed to d, in the name Rede of the river flowing past the fort. Note the parallels with the name Gambaglanda (Birdoswald), also an old-style compound in the three hill-letters m, l and n and overlooking a river with a name including the river-letters r and t, namely the river Irthing, where the t has changed to th.


The name Gamblidandi thus refers to a location on the summit of a high, steep hill. Such a name is not appropriate for the actual site of the Roman fort at Risingham. There may have been a Celtic settlement on the summit of a high, steep hill somewhere in the vicinity and the name was then transferred by the Romans to their fort. But note that the name would be entirely appropriate for the Bell Knowe round cairn - the cairn is at the top of a high, steep hill and the Roman fort is at the bottom of that same hill. But if the name of the cairn was transferred by the Romans to their new fort then this surely implies that the local people were aware of the name of the cairn at the date of building of the Roman fort. This might imply that that cairn, believed to date from the Bronze Age, was still in use when the Roman fort was built.







(Celt) (Rav)       (Mod)
Inacisodulno Iaciodulma (95)       Little Brickhill
          (Milton Keynes)


Rivet and Smith, following Richmond and Crawford, see this Ravenna name as a corrupt form of Lactodoro at Towcester. This cannot be correct as the compiler of Ravenna has just written Lectoceto, the correct form of the name of the fort at Towcester, and is hardly likely then to write down a garbled version of Lactodoro, which is itself an erroneous form of the name of the Towcester fort (see Lectoceto). It is more sensible to see Iaciodulma as the name of a fort or posting station on Watling Street somewhere between Towcester and the next but one Ravenna name, Virolanium at St. Albans.


Iaciodulma is indeed a strange name, but if the c, d and l are correct then we are looking for a place at the top of a steep hill. The two most likely locations, both fitting that description, are Little Brickhill and the top of the hill just northwest of Hockliffe. Little Brickhill seems more likely since the river at the foot of the hill is called the Lovat (also known as the Ouzel). Lovat is a Celtic river-name in the river-letters l (applied to minor rivers, including tributaries of major rivers), b changed to v, and t. And the river-letter t is at the end of the river-name, which is where one would expect to find it if the dulma element is the earliest element of the name and the first part of the name comprises inversion-type elements. The first part of the name may have been Inaciso, comprising the hill-letters n and s corresponding to the river-letters l and b, changed to v, in the river-name Lovat. The ma at the end of Iaciodulma probably results from n/m confusion at some stage of medieval copying. The second part of the name was probably originally dulno, where the no ending is fairly common - one sees it, for example, in Duriarno and Duroaverno. Regarding the n/m confusion referred to above, it was apparently fairly common in medieval texts to abbreviate a name by not writing an m or an n, but to represent either by a tilde above the preceding letter in the name. A later copyist, wishing to expand the name, might not have been quite sure whether to write an m or an n. Sometimes he would get it right, sometimes not, so dulno could easily end up as dulma




(Celt) (Rav)       (Mod)
Ibernio Ibernio (37)       Hod Hill
(a river)          (Dorset)


Note 1:

Rivet and Smith suggest hypothetical British Ivernio taken to mean 'place on the Iverno river', though the river-name is also hypothetical. They specifically exclude Hod Hill since they themselves identify that as having been Dunum (Ptolemy's Dunium in the territory of the Durotriges).


However, Ibernio occurs in a group of Ravenna names coming down from Old Sarum, Ansty and Melbury to Ibernio and then going on to Weatherby Castle and probably on to Maiden Castle, so it seems quite clear that Ibernio was the name of the Roman fort built inside the hill-fort on Hod Hill, adjacent the river Iwerne. The name appears to be a river-name in the river-letters b and r corresponding to the hill-letters s and m respectively. It is thus probably safe to conclude that the Celtic name of the hillfort included the hill-letters s and m. The river-letter r corresponding to the hill-letter m is found also in the river-name Stour, this being the river which flows along the western side of the hillfort. There is no sign of the river-letter b in Stour (unless the u was earlier a v, itself a modified river-letter b). But the river-letter t in Stour corresponds to the hill-letter l. This hill-letter was used by the Durotriges and enough Durotrigan material has been found during excavation at Hod Hill to make it safe to conclude that the Durotriges did actually take over the hillfort (from a people who used the hill-letter m). The land-name of the hillfort on Hod Hill thus most probably now included the hill-letters l, s and m in that order. The river-letter s at the beginning of Stour corresponds to the hill-letter r and precisely because the s is at the beginning of Stour we can conclude that it was added to the river-name by people who coined inversion-type place-names in the hill-letter r. It is probably unlikely that any such people seized control of Hod Hill or anywhere downstream from it after Hod Hill fell to the Durotriges, so we can conclude that part at least of the area around the upper reaches of the Stour, upstream from Hod Hill, was occupied by a people who coined inversion-type place-names in the hill-letter r. This should not surprise us as this area is not so very far away from other places which had names including inversion-type elements in the hill-letter r, as for example Omirededertis (Ravenna’s Omiretedertis) at Ham Hill, to the west, and Verlucione at Sandy Lane, to the north. What is not clear is whether those r-people along the upper reaches of the Stour belonged to the Dobunni, who used the hill-letter r, or were a separate group wholly independent of that tribe. But to get back to Roman Ibernio at Hod Hill, at some point the b of Ibernio changed to v and much later the v was anglicized to w, this yielding the modern river-name Iwerne. It is, however, a little surprising that the Romans should have transferred a river-name to the fort which they built inside the old hillfort, rather than just adopt the name of the hillfort itself (unless, of course, the hillfort had been abandoned at an earlier date). Surprising, too, that they should adopt the name of the minor river, the Iwerne, rather than that of the main river, the Stour.


Note 2:

Example of bv


[The entry for Ibernio was last modified on 06 August 2020]




(Celt) (Rav)        
Iberban Iberran (210)        
(a river)          


Note 1:

Rivet and Smith see this name as being certainly corrupt, probably a corrupt form of the name Ibernia, referring to Ireland.


It seems quite clear, however, that this name was simply transferred from the river Iberban (now the Bervie Water) with r b → r r.  Iberran will have been a Flavian fort built fairly close to the Bervie Water, most probably at a location where it could control movement across the Mounth. (A note explaining the structure of the river-name Iberban may be accessed by clicking here).


Note 2:

Example of r b → r r


[The entry for Iberran was last modified on 12 February 2021]




(Celt) (Rav) (Ptol) (AI)   (Mod)
Isca (I)scadumnamorum (16) Isca Isca Dumnoniorum   Exeter
      (Iter XV)    


In the past Isca has been regarded as a river-name, perhaps derived from hypothetical British isca, taken to mean 'water, river'. There has been a great deal of discussion over the years of a variety of hypothetical roots, including esca, eisca and even peisko, the last-named thought to mean 'fish', as well as hypothetical ei, taken to mean 'to go', and hypothetical eis, taken to mean 'to move quickly', but it is fair to say that no generally accepted etymology has emerged. But, to come back to the Isca at Exeter, the river-name is thought to have been transferred to the Roman fortress built close to the river now called the Exe.


But in fact isca is an inversion-type topographical element meaning 'hill steep', the hill in question at Exeter being the steep river embankment on the eastern side of the Exe. The element is also present in Esica, the Trajanic fortlet at Haltwhistle Burn (this name later being transferred to Great Chesters). The element also occurs in a number of names on the Continent, usually as the final element of a compound name, as in Matisco (Macon, north of Lyon), Petinesca (Studen, northern Switzerland) and Vindisca (Venasque in southern France). All of these places were located at the top of a steep slope. Isca at Exeter is simply a place-name in the hill-letter s and the name was transferred by the Romans to the river now called the Exe. It follows that Isca was not the Celtic name of the river - this name may have been Ravenna's Maviasarna (written as two separate names, Mavia and Sarna, in Ravenna), a river-name of the kind where a river-element, here sarna, is used as a suffix to a place-name, here Mavia. Note that Mavia may be an abbreviation of Mestevia, apparently the name of the Roman fort at Tiverton, on the river Exe.


As to Ravenna's form of the name, the order of names in Ravenna is Melamoni, then Scadumnamorum, but this should clearly read Melamon and Iscadumnamorum. The second m in Ravenna's form may just be a copying error for ni. And of course the dumnamorum/dumnoniorum part of the name refers to the local tribe, the Dumnoni




(Celt)   (Ptol)      
Iscalis (a place), or   Iscalis      
Iscalis (a river)          



Ptolemy places Iscalis in the territory of the Belgae.


Rivet and Smith see the isca part of the name as a 'water' name, Iscalis perhaps meaning 'place on the Isca river'.


Names of this kind cause some difficulty because the letter l was used both as a hill-letter and as a river-letter (though obviously not by the same group of Celts), so that it is sometimes not clear whether a name is a place-name in the hill-letters s and l, or a river-name in which the river-letter l is used as a suffix (or prefix) to a place-name. Added to that is the inconvenient custom of the Romans of transferring river-names to places and place-names to rivers. Thus, in the present case, Iscalis could be a place-name in the hill-letters s and l, where the inversion-type element sc means 'hill steep', the name then being transferred by the Romans to the river on which Iscalis stood, or it could be a river-name comprising the river-letter l added as a suffix to the place-name Isca, where Isca was a Celtic settlement on the river concerned. But by whatever means, Iscalis came to be applied to the river, now the Axe in Somerset. It is likely that later English settlers saw the l of the name as the letter they knew 'the' Celts had used simply to signify 'river', so those settlers saw Isca as the proper name of the river, this Isca later changing to Axe. It is alternatively possible that Iscalis, like a number of other Romano-British names, was simply reversed at some stage, this yielding the form Lacsis. The later English settlers would then have seen Acs as the proper name of the river, this being written down as Axe. But we can conclude that there was a place called Iscalis or Isca on or close to the river Axe and that that place was adjacent a steep hill. Of course if Iscalis was originally the name of the river, then the Romans will have transferred the name to a fort or settlement which they built close to the river Axe, though not necessarily adjacent a steep hill.




(Celt)   (Ptol) (AI)   (Mod)
Isur   Isurium Isurium   Aldborough
(a river)     (Iter I)   (North Yorkshire)
      (Iter II)    
      (Iter V)    


Jackson referred to a hypothetical British river-name Isura, though pointed out that some scholars doubted whether such a name is really the origin of the modern river-name Ure. But whatever the original form of the river-name the Romans are assumed to have transferred the river-name to the new capital of the Brigantes, Isurium, built at Aldborough, just outside modern Boroughbridge in Yorkshire.


There is apparently no problem here. The name Isurium will be derived from a river-name Isur, this being a river-compound in the river-letters s and r, these corresponding to the hill-letters r and m, both of which are present in the name of the Roman fort further upriver at Bainbridge, namely Smetriadum, as indeed is the hill-letter s corresponding to the river-letter b in Bain, the name of the tributary which joins the Ure just north of Bainbridge. But note that it is probable that Smetriadum was Celtic Smedriladum orSmedriadunum and that the river Ure was called Urus, meaning the 'Ur river', prior to about 130BC. Then, after the changeover to inversion-type names the name changed to Isur, meaning the 'river Ur'. For a more detailed discussion see the entry for Smetriadum.


[The entry for Isurium Brigantium was last modified on 17 May 2021]




(Celt) (Rav)       (Mod)
Litucodon Itucodon (179)       Thockrington

Note 1:

Rivet and Smith change the spelling to Itunodunum, and appear to see this form as being derived from hypothetical pituna, taken to mean 'water, river' and hypothetical dunos, taken to have meant 'hill' originally and to have taken on the meaning 'fort' later.


Itucodon appears, however, to be a straightforward topographical name, though it has lost its initial letter. The don element of the name is an inversion-type topographical element meaning 'summit of hill' and the Ituc part of the name is an inversion-type element meaning 'hill high steep', but the hill-letter is missing. It seems probable that the initial hill-letter was an l - one sees this initial l in a number of names up the eastern side of the country, for example in Olcaclavis at Elginhaugh, Lecidanisca (original form of Evidensca) at Inveresk and Lincovigla at Lanchester. The Celtic name will thus have been Litucodon, the lituc element being the same as the litac element of Alitacenon (164).


Note 2:

Example of missing initial letter - L




(Celt) (Rav)     (ND) (Mod)
Isulimocenon Iuliocenon (116)     Tunnocelo Ambleside
or         (Cumbria) 


Note 1:

Rivet and Smith accept Horsley's proposal that the correct form is Itunocelum, the first element perhaps just meaning 'river' and being based perhaps on hypothetical pituna, perhaps simply meaning 'water'. The second element is thought to be derived from hypothetical British ocelo given the meaning 'headland, promontory, spur'.


But Iuliocenon is much earlier than Tunnocelo and does not include any element resembling ocelo, so the above derivation of the second element can be dismissed. Indeed it seems quite clear that Iuliocenon was a Flavian fort at Ambleside and Tunnocelo a later fort. The development of the earlier name to the later seems quite clear - the initial i changed to t and the letters l and n swapped places to yield Tuniocelon. This last form then changed to the Tunnocelo of the ND. But Iuliocenon will not have been the original Celtic name. The initial iu is unusual and suspect and there may have been a consonant between the i and the o. The names of the rivers at Ambleside come to our aid here - these are the Brathay and the Rothay. These are normally considered to be Old Norse names, but they are just as likely to be Celtic names using the river-letters b, r and t (changed to th). The corresponding hill-letters are s, m and l. The l is already present in Iuliocenon, so perhaps we have to fit in an s and an m somehow. This would seem to require that the Celtic name had been somewhat of the form Isulimocenon, this being an inversion-type name with the hill-letters in the chronological order first m, then l and finally s. The earliest river-letter is r, corresponding to the hill-letter m. The hill-letter l will be l2 and, as usual, the corresponding river-letter t, here changed to th, is placed at the end of the existing river-name, in this case after the river-letter r. This accounts for the river-name Rothay and the rathay of Brathay. But then the people who added the hill-letter s in an inversion-type manner to the place-name also added their river-letter b in an inversion-type manner, i.e. generic term before the qualifier, and it is this which yields the river-name Brathay. But of course the first element of the place-name might have the meaning 'hill steep' just as the moc element means 'hill steep'. In this case the Celtic name will have been somewhat of the form Sigulimocenon. Note that there are steep hills to the east and northwest of the Ambleside fort.


Note 2: 

Example of initial i → t

Example of ni → nn

Example of rearranging the letters of a name




(Celt) (Rav)       (Mod)
Luba Iupania (49)       Chepstow
(a river)         (Monmouthshire) 


Note 1:

Richmond and Crawford thought Lupania the correct form of this name. Rivet and Smith thought it possible that this form might be based on a hypothetical river-name Lupa, but thought it also possible that Iupania is a corrupt form of some other name.


One of the surviving Ravenna texts gives the form Iupavia, but leaving that on one side Ravenna gives two adjacent names as Iupania Metambala. It seems likely that the original names were Luba Nemedonbala, and this text was at some point wrongly divided to yield Lubane Medonbala, Lubane then changing to Lubania and at some later point to Iupania. The original name Luba is a river-name of the kind comprising a place-name, here just the hill-letter l, with a river-suffix, here uba, the same suffix as is seen in the river-name Anderelionuba (transferred by the Romans to a fort and listed as such in Ravenna). At some stage during the Romano-British period Luba appears to have changed to Luva and then presumably Anglo-Saxon settlers took the l to be the letter they knew 'the' Celts had used simply to mean 'river' (cf. river Alavna at Stratford-on-Avon, proper name taken to be Avna, now Avon). For those Anglo-Saxon settlers the proper name of the river was thus Uva, which was then anglicized to Uwa, this presumably being the origin of the modern river-name Wye. The Romans simply transferred the original river-name Luba to a fort which they built close to the river Wye, presumably to control a crossing of the river and most probably at Chepstow.


Note 2:

Example of b → v



[The entry for Iupania was last modified on 17 May 2021]