The Roman Map of Britain

Comments on a work by R N Worth 1885, a belated review

    With the advent of Google Books it has become increasingly easy to search the history of the treatment of the Ravenna Cosmography. Recently (January 2007) I came across an article printed in the 1885 edition of Reports and Transactions of the Devonshire Association for the Advancement of Science, Literature, and Art.
    R. N. Worth in his article "Notes on the Ancient Recorded Topography of Devon" makes some surprising observations, associations and conclusions on some place-names. While a great portion of his work is certainly in error there are parts that deserve reading.

p.357
Eltabo and Elconio, bearing in mind that the e would sound a, and that h was a letter never used by the Ravennat, suggested as their first component the Kornu hayle or hel=a river. Assuming a topographical order Eltabo associates itself at once with the Helford river...

I have come to associate El- with the river Hayle rather than the Helford.

p.364
Now the five names which follow Alauna Silva are Omire, Tedertis, Lindinis, Canca, and Dolocindo or Dolotindo. The last of these is clearly Doulting, which lies on the Fosseway at or near the junction of the ancient road traversing the crest of the Mendips. Omire, Tedertis, Lindinis, and Canca in all likelihood are to the south, and I think it highly probable therefore that by Omire our author intends the [p.365] magnificent set of earthworks on Ham (Om) Hill; and by Canca, the equally noteworthy fortified town of Cadbury,* 

[* The name Camel, which occurs several times in the vicinity, is much nearer to Canca, and has led to Cadbury being regarded as Camelot! Tacitus speaks of the Cangi as dwelling in Somerset.] 

    While the association of Dolocindo with Doulting is correct, Doulting is based on the old name of the  river Sheppey. Worth does not assign the name to the Roman town at Shepton Mallet on the Fosse Way.

    There is no Om in the etymology of Ham Hill.

    The faint association of Canca (properly Cantia) with the name Camel (likely an earlier compound of cant as in Quantock Hills) should have been pursued further and would have pointed to Podimore on the Fosse Way.

Lindinis is clearly the "lake fort," and probably an old name of Ilchester, near which we still have the name of Limington. Tedertis may be a form of Petherton. At least these spots occur on or close to the line of the Fosseway in the order given by the Ravennat, and were this the place many arguments might be advanced in support of these suggestions, though I do not lay absolute stress upon them.

    The assignment of Lindinis is correct, though the derivation of the name is not.

    The association of Tedertis (properly Pedretis) with Petherton is correct, but only because Petherton is based on the river-name Parret. The proper assignment belongs to a site on the Parret and the Fosse Way at Ham Hill.

My impression is that the line is continued from Doulting by Clavinio (Clutton) to the mouth of the Avon at Morionio, the latter name appearing to indicate a place by the sea.
And then we appear to make a fresh start from Doulting on the Fosseway. The next name is Bolvelaunio. Now bearing in mind that the Ravennat has no W and substitutes B, there can hardly be a doubt that the place here indicated is Wellow.

    Clavinio as Clutton is wrong, the proper association is Clausentum (Bitterne).

    Recognizing Bolvelaunio as a site on a river Wellow is correct, but it is not the Somerset Wellow but the Hampshire Wellow now known as the Blackwater.

The succeeding group of names comprises Alauna, Coloneas, Aranus or Aramis, Anicetis, Moiezo, and Ibernio. As to the latter, though possibly I stand alone, I have no doubt of its being the modern Iwerne, and of its applying to the great earthworks on Hamildon and Hod Hills near Blandford.

    He is correct in associating Ibernio with Iwerne. Unfortunately, the form Moiezo (properly Meletio in Parthey & Pinder) did not have lead him to the nearby Fontmel.

This gives us a clue that cannot lead us very far wrong. Alauna is probably some place on the Avon between Wellow and Calne, the most likely Coloneas. Sarum or Salisbury does not occur in the Ravennat's list; but assuming that I am right in following this traverse the similarity between Aramis and Harnham the Saliburian suburb is very striking. Yet after all it is only the initial "S" that is wanting for the reproduction of Sarum itself, and that may really be the final "s" of the word preceding, and in any case is within the limits of proven corruption. Have the words been rightly divided in reading the MS.?

    Alauna should be on a road connecting to Bolvelaunio. Alauna does not connect by direct road to Coloneas.

    Worth's note that Coloneas-Aramis/Aranus might be mis-divided pre-dates the most recent editor Schnetz by 55 years! The first name, as Colonea, he associates with Calne, rather than Collingbourne's river.  The second name is properly Sarauus and represents the Roman settlement at Salisbury. Sorviodunum (properly Sarviodunum) is the name of the native hill-fort/settlement recorded in the Antonine Itineraries. An imperfect association, at best, but nevertheless the equation is correct.

p.366
... If the system set forth for the interpretation of the Ravennat is the right one, it will prove itself in the very fact that it is capable of systematic application; and as a system by the results of that application it must stand or fall.

    No citation of this work is listed in Google, merely that it exists as an article.
    To the best of my knowledge none have noted this previously, a shame.