ROMAN DISHED SHIELDS

[jho]

The proceedings of the 16th ROMES in Xanten contain an article by Ratsdorf on evidence for use of dished shields by the Romans.

Ratsdorf has measured some 300 Roman and Germanic shield bosses spanning the complete imperial period. He does not provide all 300 measurements but states that the angle between the boss and the base plate is always larger than 90 degrees (93-125 degrees). It appears that the average angle increases over time.

Ratsdorf concludes that this is evidence of the fact that Romans and Germans (almost) exclusively used dished shields during the Imperial period. He dismisses the idea of a wooden base plate underneath the boss as a reason for this phenomenon due to the fact that, where shield nails were found, they show that the boss had been directly nailed to the shield board.

It would be interesting to know whether RATers with an experience in making shields may come up with an alternative explanation of this phenomenon or whether it may really imply that most or even all shields were dished.

[Medicus matt]

It would be interesting to know whether RATers with an experience in making shields may come up with an alternative explanation of this phenomenon or whether it may really imply that most or even all shields were dished.

You get the same evidence when you look at the bosses from early Anglo-Saxon cemeteries. Having made lots of shields (flat ones, lenticular ones, ones curved in one plane), I've found that the angle of the flange on the boss can't be taken as an indicator for the shape of the board.
Whilst you usually do need some shaping of the flange to ensure a good fit on a lenticular board (altough even this depends on the extent of the dishing at the centre of the board), a shaped flange also ensures a very secure fit on a flat board. Rivetting a shaped boss onto a flat board puts the flange of the boss under compression, meaning that it's much more secure and maybe means that it's able to absorb impact more effectively.

And of course it may be that in many cases making the flange perpendicular to the wall of the boss was deemed to be an unnecessary step in the manufacturing process.

I've not thought to check, but given that a lot of bosses from AS graves are found with their nails in situ, it should be possible to analyse these to see if there's a space between the inside of the boss flange and the wooden board.

[P. Clodius Secundus]

Matt's analysis agrees with my albeit limited experience in this area. When making hand hammered bosses for flat parmas, (I've made about half a dozen) I used to try to get the flange as flat to the surface as possible so it would seat properly. Someone asked me to do a larger Constantine Chi-Rho shield, but they wanted a decorated brass boss that came with a dished edge. I couldn't figure out how to flatten the edge without destroying the raised decoration so I ended up pressing the boss on a dishing stump before riveting it. The edge on that one wound up fitting tighter to the shield face once it was clenched down. It left less of a gap around the rim of the boss where a slashing blade might bite in.

[caiusbeerquitius]

I think Holger Ratsdorf is quite right. In relation to the shield boards we have preserved, this is also quite sensible (Dura, Trier Collection etc.) So far besides the Doncaster shield, which IMO does not provide evidence that it was actually flat, as I have shown here on RAT, the extant examples are lenticular, and if only veryy slightly through being thinned out towards the edges.

[jho]

Thank you everbody for your interesting feedback. So it appears that a sloping boss can be fitted to a flat shield.

If I may, I would like to further inquire whether such a sloping boss would remain sloped even after fitted to a flat shield or whether the pressure of the nails and the shield board would not actually make it flat after prolonged use?

[Medicus matt]

Ah, I'm not debating the evidence presented by extant shield boards, only the premise that the shape of the flanges on the bosses themselves can be taken as conclusive evidence for the shape of the boards, providing an alternative explanation in the process.

[Medicus matt]

Thank you everbody for your interesting feedback. So it appears that a sloping boss can be fitted to a flat shield.

If I may, I would like to further inquire whether such a sloping boss would remain sloped even after fitted to a flat shield or whether the pressure of the nails and the shield board would not actually make it flat after prolonged use?

I've only built one shield using an iron (rather than modern steel) boss. Even though it was softer, it still retained it's shape when fixed to a flat board. I suppose if I'd been really severe when rivetting it might have flattened it but you don't need that much force to peen over iron rivets.

Another thought, a sloped flange would give more handroom inside the boss than a flat flange when fixed to a flat board.

[Robert Vermaat]

Wouldn't we be able to judge this by the length of the nails? I think I recall that most shields have a certain thickness (9 mm or so?), meaning that with a sloping boss the nails/rivets would have to be longer than you'd expect with flat bosses?

[Medicus matt]

Wouldn't we be able to judge this by the length of the nails? I think I recall that most shields have a certain thickness (9 mm or so?), meaning that with a sloping boss the nails/rivets would have to be longer than you'd expect with flat bosses?

Board thickness isn't a constant though, at least not on AS shields. Range is anywhere between 6mm and 12mm. Don't know about Roman.

If you've got the nails, there's a good chance that you've got some evidence of the board remaining. If there's a gap between the head of the nail/rivet and the board that's thicker than the thickness of the flange, that could be evidence that it's an angled flange on a flat board.

Or that the board was faced with something.
Or that it was a badly made shield.
Or something.