Porto Novo / Valkenburg scabbard lockets

(Paul Geddes)

One of the interesting things about our hobby, but one which can also be the most frustrating, is the way our knowledge of the Roman army and its equipment is continually being expanded by new discoveries. Sometimes this allows us to introduce new items which allow for a more interesting interpretation or sometimes a more practical approach to things. However, all too often it tells us that we have we have fallen behind the current state of the evidence and need to make improvements to kit in the light of an enhanced level of knowledge. Regrettably this can end up costing us money we would sometimes rather not spend and this is a common problem experienced by most re-enactment groups of most periods.

In the light of this, it is an unfortunate fact that many of our sword scabbards now fall into this category and need to be upgraded or replaced in the light of what we now know about the evidence they are based on. This includes my own scabbard.

The scabbards I am talking about here are the ones which feature locket plates decorated with two large embossed 'X' shapes. These are based on Micheal Simkins interpretation of pictures of a fragmentary locket plate from Long Windsor (figure 1), now in the Ashmolean Museum. In the mid 1970s, when Simkins proposed the pattern of locket plate commonly seen in the RMRS, far fewer Pompeii type scabbard parts were known than are know to us today. At the time the only known Pompeii type locket plates he knew of were the four from Pompeii itself and one other from Leiden in Holland. The Long Windsor find consisted of a Pompeii type sword and its fragmentary scabbard. Although the remains of the locket plate measured less than half the length of the Pompeii and Leiden examples, Simkins made the quite reasonable assumption that the Long Windsor plate would originally have been of the same dimensions as the other know Pompeii type locket plates, with two decorative fields. His interpretation of the piece was that it had originally had an 'X' shape embossed into it (which had later corroded away) and given that the assumed lower area of the plate was not present, he repeated the pattern on the lower field of his reconstruction which he found on the surviving portion. At the time his interpretation was sensible and reasonable and seemed a good interpretation of the evidence. It became known as the 'Simplified Simkins Pattern' locket.

So far so good then, until new evidence turned up to call Simkins' interpretation of the Long Windsor plate into question. In 1996, Martin White, of the Ermine Street Guard, published an article which convincingly showed, based on a closely comparable locket plate from Valkenburg in Holland(figure 2), that the Simkins' interpretation of the Long Windsor plate had been wrong and that the plate had not been embossed but had instead had the 'X' shape inscribed into it, along with a number of small punched circular shapes. Another comparable locket plate had been discovered some time before at Vindonissa in Switzerland (figure 3) but was not widely known at the time.

Figure 1: The Long Windsor locket plate

Figure 2: The Valkenburg locket plate.

Figure 3: The Vindonissa locket plate.

Figure 4: The Porto Novo locket plate .
Since then five more examples have come to light and they now comprise a distinctive group of a completely different type of locket plate to what had been known before. These other examples are a plate still incorporated in a scabbard from Porto Novofigure 4) an unprovenanced plate currently in a museum in Budapest (figure 5) an unprovenanced find from somewhere in the Balkans, now in a private collection (figure 6), a plate from Lobith in Holland featuring an inscribed warrior in a chariot (figure 7) and a plate still incorporated in the remains of a scabbard from Rajkova-Mogila, currently in the National Museum in Sofia (figure 8). This last example appears to be embossed and features a motif of the wolf and twins.

Figure 5: The Budapest locket plate.

Figure 6
Unprovenanced plate from the Balkans

Figure 7: The Lobith locket plate

Figure 8: The Sofia locket plate.

As will be seen from the pictures, these locket plates are all much shorter that the other type of Pompeii scabbard locket and all feature short legs which extend down to the point where the scabbard would be crossed by the upper cross hanger holding the upper suspension rings (as can still be seen on the Porto Novo and Rajkova-Mogila examples), inscribed horizontal lines at top and bottom and often feature gamma shaped cutouts, further decorative cutouts, radiating inscribed lines, small punched circles, figural designs and scalloped lower edges. All of these plates appear to have featured a retaining band near the top which passed around the rear of the scabbard.

It is worth pointing out at this point that these lockets may come from a transitional type of scabbard. Although the Long Windsor plate was associated with a Pompeii type chape, the Porto Novo scabbard was closer to the dimensions of a Mainz type scabbard but also featured the ‘palmette decoration associated with Pompeii type sheaths.

Figures 9 and 10 show reconstructions of the updated interpretation of the Long Windsor plate and the Vindonissa plate.


Figure 1 N. Griffiths
Figures 2-8 C. Miks
Figures 9-10 Ermine Street Guard


Martin White, 'Pompeii Scabbards' - Exercitus 1996
Christian Miks, Studien Zur Roemischen Schwertbewaffnung In Der Kaiserzeit 2007


A good point that has often been made is that at any one time the Roman army must have had around a quarter of a million sets of equipment in service and with what we have to study comprising less than one percent of this, it is often unwise to get too prescriptive about how something was or was not done. Therefore a certain amount of assumption will always have to be used in addition to the strict evidence which has survived, as long as it starts from the evidence and works outwards from there.

This is generally a safe 'fall back' position. However, in this case we have an instance of the evidence having increased, which in turn has shown that the old interpretation of one part of it had been seriously and verifiably flawed.

Therefore, regrettable as it may be, the time has now come when we need to start phasing the Simkins pattern locket out of the group. This should start with no further examples being introduced, and then continue as individual members either replace their scabbards with more accurate ones or have their scabbards upgraded to fit better with the current state of the evidence, hopefully with minimal cost being involved, until we reach a point in perhaps two or three years' time when the old Simplified Simkins pattern lockets are nothing more than a fond memory to be found in old photos of the group. As long as all of the Simkins plates are replaced with reproductions of designs known to be correct, no similar upgrading of scabbards should be needed again for a very long time (and hopefully never). For those who wished to upgrade, rather than replace, their scabbards, replacement locket plates could easily be produced and fitted.


Figure 9: The Long Windsor reconstruction.

Figure 10: The Vindonissa reconstruction.