[NB. Detailed information as to the different river-letters and how they were combined to form compound river-names, together with information as to the four categories of Celtic river-names, is given in Chapter 19: the rivers of Roman Britain. Detailed information as to the different hill-letters is given in Chapter 1 and information as to how the hill-letters were combined to form compound place-names is given in Chapter 2].
Ratostabius to Toesobis
[This page has been revised in its entirety. The writer is grateful to Mr. David Forbes for drawing his attention to this particular group of names by expressing his disagreement with some at least of the identifications previously offered here. This has enabled the writer to reconsider the entire group of names, giving rather more weight this time to Ptolemy’s promontories and to his coordinates of latitude and longitude. The results of that reconsideration are presented below].
1 The Deceanglani tribe
This page is primarily concerned with identifying the rivers in the group Ratostabius to Toesobis, but it is convenient to start with discussion of a name which is not a river-name and does not appear in Ravenna or Ptolemy, the tribal name Deceanglani (for reasons which will become clear in paragraph 2.2 below this form of the tribal name is preferred to Rivet and Smith’s shorter form Deceangli). This name (or at least the Deceangl part of it) is said by Rivet and Smith to be inscribed faintly on two lead pigs found at Chester and one found at Hints Common in Staffordshire. The pigs are thought to have come from the lead mines of Flintshire, presumably because of the proximity of Chester to Flintshire, but it is equally possible that they had come from the lead mines in the Vale of Conway. The pigs may merely have been in temporary storage at Chester and Hints Common awaiting onward transport to wherever they were wanted or needed. It is to be noted that the great majority of British Celtic tribal names are based on Celtic topographical place-names, presumably in each case the name of the tribal centre at the time the tribal name was coined (this point is discussed in detail in Home/Chapter 27), and Deceanglani appears very clearly to be a modified topographical place-name. The Dec element appears to be an inversion-type element meaning ‘summit of (hill) steep’, but the hill-letter is missing. The gl element is an old-style element meaning ‘steep hill’. The second element was most probably the old-style element can, also meaning ‘steep hill’. The full topographical name was thus probably somewhat of the form De(x)cecanglanion. At this point the name Canubio of the Roman fort at Caerhun comes to our aid since it appears to be based on a Celtic composite river-name (such names are explained in Home/Chapter 19, 11) of the form Canubena comprising the place-name element can meaning ‘steep hill’ and a river-suffix ubena including the river-letter b, corresponding to the hill-letter s. This is attractive, because we may now assume a full composite river-name for the Conway with a form such as Descecanglubena, where the missing hill-letter in the first element is s, corresponding to the river-letter b in the river-suffix. The Romans may then have reduced the river-name (by deletion of letters) to Canubena and then transferred this name in the form Canubio to the fort which they built at Caerhun, close to the river Conway.
2 The Caeanganorum promontory
Now it is time to bring Ptolemy’s Caeanganorum promontory into the discussion. First the location of the promontory will be discussed and then the meaning of the name.
2.1 Ptolemy tells us that the Caeanganorum promontory was 2 degrees of longitude west of the Mersey estuary, the same longitudinal distance as between the mouths of the rivers Tamar and Exe. That would place the promontory at Great Orme’s Head, by Llandudno. But Ptolemy also indicates that the Caeanganorum promontory was 1 degree of latitude south of the Mersey estuary, and that seems inappropriate for Great Orme’s Head. Moreover, the mouths of the rivers Toesobis, Stuccia and Tuerobis are all said by Ptolemy to be south and east of the Caeanganorum promontory, so, if we accept the coordinates of those river-mouths as being even just approximately correct, then the Caeanganorum promontory cannot have been Great Orme’s Head, since there are no river-mouths southeast of that point. We must therefore ignore Ptolemy’s longitude coordinate for the Caeanganorum promontory and locate the promontory somewhere west and south of Great Orme’s Head. The most likely location would then be one of the promontories at the southern end of the Lleyn peninsula.
2.2 Meaning of Caeanganorum
It seems quite clear that Caeanganorum is merely a modified form of the tribal name Descecanglanorum in what is presumably intended to be the genitive plural form. If we write the tribal name as [Des] ce [c] ang [l] anorum, then delete the letters in brackets, we have Ceanganorum. Ptolemy’s form shows a change in the linking vowel between the first two elements of the name (ae instead of e), but that is of no importance. It is thus evident that we are dealing with one and the same tribal name based on a Celtic topographical place-name. The Caeanganorum promontory will thus be ‘promontory of the Deceanglani’ in the sense of ‘promontory in the territory of the Deceanglani’. The tribal centre, Descecanglanion, will not have been in the vicinity of the promontory – it will have been further east, in the valley of the river Conway. It may have been the large hillfort known as Castell Caer Seion (SH760778) just west of the mouth of the Conway, or possibly the hillfort on Pen-y-Gaer (SH750693), a little southwest of Caerhun. This geographical separation of the tribal centre and the promontory bearing its name will be of some interest in the discussion of Ptolemy’s Octapitarum promontory in paragraph 5 below. One sees the same geographical separation of a tribal centre and a promontory bearing its name in the case of the Novantarum promontory, apparently somewhere on the Rhinns of Galloway, whereas the tribal centre of the Novantae was apparently further east, in the valley of the Water of Luce. Indeed the tribal name Novantae is based on the Celtic composite river-name of the Water of Luce, a name somewhat of the form Abranobandena (Ptolemy’s Abravannus).
3 The river Stuccia
Stuccia was almost certainly Stuccla at an earlier date, this land-name comprising the inversion-type element Stuc, meaning ‘hill high, steep’, qualified by the old-style element cl, meaning ‘steep hill’. The Stuccia does appear to have been the river Ystwyth, as suggested by Rivet and Smith, though the derivation of the modern name given by Rivet and Smith is less than convincing. They follow Holder in deriving the modern river-name Ystwyth from a hypothetical British root stuctio, taken to mean ‘bent, curved’, the modern name Ystwyth then being “a direct and natural descendant of the ancient one”. There are indeed bends in the course of the Ystwyth, though not to a degree in any way noteworthy. In addition there are many, many rivers in Britain which might reasonably be said to be winding, so if there had been a British root stuctio meaning ‘bent, curved’, and if that root had been used in coining river-names, then one might reasonably expect quite a number of rivers to be called Ystwyth or to have names closely similar to Ystwyth. But that appears not to be the case. Alternatively, we may see Stuccla as the surviving part of a composite river-name somewhat of the form Stucclobitena, where the river-letters b and t in the river-suffix correspond to the hill-letters s and l in the place-name element Stuccl. With loss or deletion of the internal letters ucclo and the changes b→v→w and t→th, we obtain the form Stwith. This form is surely much more likely to be the origin of the modern river-name Ystwyth. The place with a name including Stuccl may have been the hillfort on Pendinas Mon, adjacent the mouth of the Ystwyth.
4 The river Toesobis
The mouth of Ptolemy’s Toesobis river was thus somewhere between the Lleyn peninsula and the mouth of the Ystwyth. The most obvious candidate is the river Dovey. All that will have happened over time is that a few internal letters, -eso-, were deleted (a common change in Romano-British names), the initial T changed to D (also a common change) and b changed to v (yet another common change).
5 The Octapitarum promontory
And now a word about Ptolemy’s Octapitarum promontory. Rivet and Smith identify this as St. David’s Head. They may be right, but that is only one possible identification. Again one should note that the great majority of British Celtic tribal names are based on Celtic topographical place-names. Octapitarum looks like a modified form of a topographical place-name. The second element appears to be a modified form of an element such as bild, meaning ‘high hill summit’, where the b has undergone the fairly common change to p, the hill-letter has been deleted and the d has undergone the fairly common change to t. The first element, oct, appears to be the rump of an inversion-type element meaning ‘hill steep, high’. But because the old-style element refers to the summit of a high hill it is quite possible that the initial inversion-type element contained much the same information. We would therefore be looking at an inversion-type element somewhat of the form Demecet, meaning ‘summit of hill steep, high’. The Demecet element uses the hill-letter m because Pembroke (the assumed location of the Octapitarum promontory) was in the territory of the Demetae tribe and they used the hill-letter m. The complete topographical name will thus have been somewhat of the form Demecetabildion, though the ending need not have been ion and the second hill-letter need not have been l; it may have been an r as in Maridunum at Caermarthen. It is quite possible that Demecetabildion was the origin of both the tribal name Demetae and the promontory name Octapitarum. The tribal name would be based on the Demecet part of the name and in the form Demecetae is in fact on record, though it is better known in its abridged form Demetae. Octapitarum is arrived at very simply by deletion of initial Dem, the change of ecet to oct, the common change b→p, deletion of the second hill-letter and the common change d→t (the change from ecet to oct may look rather odd, but then we can’t be sure that the second and third vowels of Demecet are exactly the same as those in the original topographical name). Note that Octapitarum will be a genitive plural form, just like Caeanganorum and Novantarum. Caeanganorum promontory appears to mean ‘promontory of the Deceanglani’ in the sense of ‘promontory in the territory of the Deceanglani’. Likewise Novantarum promontory appears to mean ’promontory of the Novantae’ in the snese of ‘promontory in the territory of the Novantae’ It is thus quite possible that Octapitarum promontory means ‘promontory of the Demetae’ in the sense of ‘promontory in the territory of the Demetae’. And just as the tribal centres of the Novantae and Deceanglani were not in the immediate vicinity of the Novantarum promontory and the Caeanganorum promontory respectively, so it is quite possible that the tribal centre of the Demetae was not on the coast, but was an inland hillfort on the summit of a steep, high hill (the meaning of demecet). The Octapitarum promontory, the ‘promontory of the Demetae,’ the ‘promontory in the territory of the Demetae’ could then be almost any promontory on the rugged coast of southwest Pembroke. St. David’s Head is then just one possible candidate.
6 The river Tuerobis
The mouth of the Tuerobis river was thus somewhere between the mouth of the Ystwyth and the rough, indented coastline of southwest Pembroke. The obvious candidate is the river Teifi, a river of some interest to the Romans as the Roman fort in the vicinity of Lampeter was built close to that river. All that will have happened over time is that the r was dropped or lost and b→v→f, Tueofis then developing further to yield Teifi.
7 The river Tobius
The Tobius, then, will have been a river to the east of southwest Pembroke. The Towy seems the obvious candidate. The b→v→w change would yield Towius and this form may well have developed to Towy.
8 The river Ratostabius
Now, finally, the river Ratostabius, located by Ptolemy somewhere between the Towy and the Severn. This is another composite river-name, the place-name element being Rat, meaning ‘hill high’ and the river suffix being stab, comprising the river-letters s, t and b. Rivet and Smith identify the Taff as the Ratostabius. They may well be right, though Ravenna appears to apply the name Tamion to the same river. Perhaps both Taff and Tam are derived from the tab part of Ratostabius. But we should not lose sight of another possibility. Ravenna gives the name Isca to the Usk, though this will have been the Roman and not the Celtic name of the river (isca is a Celtic land-name meaning ‘hill steep’ transferred by the Romans to a river). Ratostabius may thus possibly have been the Celtic name of the river Usk.
[This page was last modified on 26 September 2024]
