Kalefeld Roman Battlefield

While arguments here rage still (well, perhaps 'rage' is a little strong) over the possible site of Boudicca's last stand, and largely on the basis of topography alone, interesting news from Germany. As if finding the site of the Varusschlacht at Kalkriese wasn't enough, news of a new battlefield, much further into Germany, at Kalefeld. Lots of discussion over at RomanArmyTalk.

This too seems to have begun as a series of metal detector finds (dating back to at least 2008) followed up by targeted excavation. Coin evidence and C14 places the date at or after 235 and a historical context is posited in German campaigns of Maximinus Thrax. At this date the nearest permanent bases were 150 miles away on the Rhine, so this army (however constituted) was a long way from home. The metal finds reported recently are spectacular. The inscribed dolabra places some element at least of legio IIII Flavia at the scene.

Recent Binchester inscription

Some initial comments by RSO Tomlin are reported in the article in Current Archaeology 263. He prefers Sace[llum, shrine, and the text might make more sense as recording the dedication of such. However, I must assume his comments have been simplified for publication - cui praeest does not mean 'dedicated by'. The formulation is surely going to require that the dedication, whether or not of a shrine, is made by the army unit in garrison, cui praeest [insert name] praefectus equitum, 'under the command of [unnamed] prefect

Roman cemetery Gloucester

Looking up the latest issue of the Journal of Roman Archaeology, I found this review by Henry Hurst of the publication of the Gloucester London Road cemetery excavated by Oxford Archaeology.

Quite pertinent to the Twentieth legion (it is the findspot of the tombstone of Lucius Octavius Martialis) with interesting comments on dating. May not be up for long (its a sample review from Vol. 23 and I suppose may be replaced by a sample from Vol. 24 at some point).

Unknown soldiers

The report on the Inversesk inscriptions prompts another thought...


It seems to have become de rigeur in the reporting of a new legionary centurion to attempt to identify him with some previously attested individual (or in this case to bring in individuals whom he cannot be identified with, which demonstrates access to a good database, but seems to add little otherwise [I am, of course, guilty of this myself] ). Little attention appears to be payed to the likelihood of this...

For that we need a few numbers - number of centurions, length of service etc - but data is sketchy, so we will only have broad estimates to play with, but enough to give an order of magnitude.

60 centurions per legion (round numbers); 20 years service; first 2 centuries AD; = 600 per legion
Average (low average) 25 legions = 15,000 centurions serving over this period.

How many do we know of already? Richier 2006 Centuriones ad Rhenum brings together 400-odd for the legions on the Rhine; c. 300 if we stick to the first two centuries AD and an average of 6 legions (8 in the first century; 4 in the second). Summerly 1991 Studies in the legionary centurionate was using Eric Birley's lists and as I recall his numbering got close to 1000 which fits quite well.

So we know the names of about 1/12th-1/15th of the centurions serving over this period. Sound quite good put like that, but 1/12th = 8.33%, ie we don't know the identity of 91.66% of centurions. The odds of a randomly recovered name matching one of those 8.3366% rather than being one of the previously unknown 91.66% are really not very high. Out of every 100 newly discovered centurions (and the rate of discovery is not especially high) we'd only expect 8 to be individuals already known...

...all things being equal. Which, of course they're not. Lots of factors pulling various ways. How many inscribed monuments would a centurion put up? Expensive things to commission. If a centurion was only likely to put up one or two, then once we've found one (or two), we're very unlikely to find another. Or perhaps certain centurions were more wedded to the 'epigraphic habit' and more inclined to leave their name in stone everywhere they went, in which case they would be easier to find. Not easy to know either way, but the main point stands. It's not actually very likely that we will already know of a newly discovered centurion. Certainly worth a comment if we do! Hardly worth noting if we don't.

Inveresk - finally this time

The latest issue of Britannia (Vol. 42 2011) arrived at my door this week and with it more detail of the Inveresk inscriptions. 'To the unconquered god Mithras, C Cas(sius) Fla(vianus) (dedicated this altar)'. That is, Flavus or Flavinus or Flavianus, or indeed possibly Flaccus (Tomlin 2011, 441-4). Still not to be seen in other than an uncleaned state (paint evidently survives and further work will be sensitive). Work towards full publication is ongoing.

Of particular interest from my point of view is the potential link to the Twentieth Legion. Entirely from the form of the letter A, but convincing nonetheless as this is distinctive on the Antonine Wall distance slabs of the legion, So, welcome C Cassius Flavus to the roster of the legion.

I am slightly less convinced by the parallels from Chester (neatly though that might tie things up), at least without seeing them again. RIB461 and RIB497 are referenced, the latter clearly shown with this letter form in RIB I, the former less clearly so.

RIB461 is in Greek in any case, so the choice of letter form may be less relevant. Unfortunately the photographs in the 1955 corpus (Wright and Richmond The Roman Inscribed and Sculptured Stones in the Grosvenor Museum, Chester) are in neither case very clear. 497 is cracked across close to the relevant point (but as drawn would indeed seem to be a clear example). Of the three letter As on 461 the first has no clear crossbar at all, the final one only a medial mark which isn't convincingly part of the letter, leaving only the third which doesn't appear entirely clear either and in the photograph the stone surface appears damaged at the relevant point. A reason for another visit to the Grosvenor Museum, I think.

Binchester Inscription

Reported last week on the blog for the Roman Binchester project:

More learned specialists will doubtless opine shortly, in the meantime a few throughts of my own (since the Binchester blog doesn't appear to take comments).

The final two lines look to be something like

[name of unit] cui prAEST [name of commander] praef eQVITVM

ie ...which is commanded by [name] prefect of cavalry

cf RIB 897 from Papcastle
ala Aug Gordia(na)...cui praest Aemilius Crispinus pref eq

and other examples (RIB 817,1585,1688,1976, 2149).


The upper line would give us ...SACER. This does appear as a cognomen, especially in the celtic provinces, but it's not really in the right place (who's name would it be? the commander's name should come later. if not part of the name of the unit, it should form part of the subject of the dedication).

Much more interesting would be SACER(dos), or some such, which is found in the imperial titulature of Elagabalus.
cf RIB1465 - dedication to the emperor recording rebuilding at Chesters in AD221.

Except the emperor's name - in particular the sacerdos...dei invicti solis elegabali bit - has been erased there, so this is only inferred... and there doesn't seem enough space here to fit all of this, however abbreviated.

Inveresk Again (Again)

A bit more info on this Museum blog
(same photos, though. still none in a cleaner state)

DAEO INVICTO MY C CAS FLA

wouldn't really go with 'to the glory of...'; nor necessarily 'invincible', though this is the usual translation... is 'unconquered' the same as 'unconquerable'?
MY is a fairly extreme abbreviation, given there's clearly no lack of space, but the second inscription is equally terse

SOL C CAS FLA >

That should be C(aius) Cas(sius) Fla(vus), not Fla(vius) which would be the nomen. Sadly the nomenclature is rather undistinctive. I wish them luck in tracing other postings.

The association of Sol and Mithras is much attested, most relevantly perhaps on the dedications at the Temple of Mithras at Housesteads (RIB 1599, 1600)