Revival because of this thread:
link from old RAT
IMO the Doncaster shield is quite tricky. A few thoughts:
1. It´s hard to date. Its taq seems to be ca. 86 CE (phase 2), however it may be from phase 2 since there seems to be no additional stratum between the "bonfire" stratum and the base of the Antonine rampart (?) The specific stratum seems not to have had any accompanying finds (?) which would help in dating.
2. The find situation is such that the sides of the shield board (or, as we may see below, its upper and lower part) were simply not present, which allows all kinds of basic outer shield forms to be deduced from the remains. In regard of the construction of the shield board the most sensible reconstruction would IMO be a long-oval shape which might vary from Kasr-El-Harit to Dura Europos: The shield was not rectangular, other shields or shield fragments found which were constructed in this way had such a form. This goes well along the pictorial evidence from the period in question. Especially in regard of the younger arguments about the form of the shields the Valkenburg shield covers were made for.
3. Shield boss: A blown up pic of the bosses´ cross-section shows that the inner rim of the flange is higher than the outer rim. In my experience this is often the case, it is hardly visible at all, but it´s there if you measure it. Even 0,5 to 1mm make quite a difference of this is in regard of the 80 + cm shield width. E.g. a shield from a circular section of a circle with 60 cm radius with a height difference of 1 cm 10 cm away from the centre would have a depth of ~4 cm after 20 , of ~8 cm after 30 and of ~15 cm after 40 cm. We will all agree upon that the the curvature of a Roman shield was certainly less, and that it may have changed over the course of the whole width, as is the case on the Kasr-El-Harit shield.
addendum: depending on the heat of the fire in which the shield was destroyed the flange may have been deformed (flattened) by the weight of the shield´s dome and / or additional pressure when hot.
4. The shield reinforcing bar is, as usual, useless in determining whether the shield was flat ot not. However, it fits nicely with other such pieces from the Raetian and German limites. BUT this requires to turn the shield to the left by 90°. This would put the iron ferrule into a position similar to those on the Dura Europos shields. The result would be a shield very similar to the Dura shields altogether. The width would fit, anyway (~1m at Dura). Move the shield boss a bit downward from the centre of gravity (as on the Dura shields), and voilá.The preserved curve of the Doncaster board would harmonize with the curvature as well.
Here´s a pic of a soldier with a dished shield with the Dura dimensions and horizontal shield reinforcing bar:
So: The shield bar probably was just in this position, but the shield should be turned.
Christian K.
No reconstruendum => No reconstruction.
Ut desint vires, tamen est laudanda voluntas.