Excavations taking place at Richborough, Kent, between 1922 and 1938 unearthed several centuries of Roman occupation on the site. Unfortunately for our purposes, the earlier excavations of the late period (late 3^rd to early 5^th Century) shore-fort in the 1920s were conducted somewhat less meticulously and without the more improved techniques brought to the table a decade later in the 1930s when work commenced on the Claudian – Flavian layers of the site. 

The objects for our first helmet, here tentatively dubbed the Richborough I, and dated thanks to context between AD280 and AD400 onwards) were found in the south side of the fort and originally listed as “an iron plate with thin bronze binding…from a suit of armour (lorica segmentata)”, “iron armour shoulder piece” and other fragments of what the archaeologists believed to be a set of segmented torso armour. We can now see after re-examination that the plate with bronze binding is actually a neck-guard from a later Roman helmet (Figure 1, 1), with other parts forming elements of the bowl and cheek.

Also found was a small box containing six fragments of iron with “bronze strapping” (Figure 1, 2); when put together, they were originally believed to be part of a cheek-piece but is now almost certainly a helmet crest in similar style to the Intercisa IV example we have already looked at. Interestingly the fragments for the crest were found to have remnants of gypsum faced with a red or brown paint, suggesting the crest was a three-dimensional box rather than a flat “fin” as per the Intercisa IV type with a plume of feathers or horsehair secured in the gypsum as we see on some elaborate crest reconstructions from the period and in contemporary frescoes or coinage. It is unclear whether this crest would be permanently attached as the Intercisa IV model or detachable such as the Augst and Budapest examples. 

Two large pieces of iron found in this context form the side of the same helmet the crest fragments belong to; we can see an ear cutaway and a row of punched holes around the lower edge for securing the cheek flaps and inner lining (thus of course providing further suggestion this helmet is an Intercisa-type). X-Ray work shows some decorative pattern on the surface, although likely made of iron as per the rest of the helmet bowl as they are indiscernible on the X-Ray itself beyond a trace outline. The parts identified as the cheek piece itself do indeed bear the hallmarks of belonging to an Intercisa-type helmet; also made of iron, they fit nicely with the bottom of the bowl and ear-hole present on the other fragments. 

Something interesting to note, however, is that the general shape of the helmet bowl seems to differ from that of the traditional late period helmet; as we can see in the diagram, the lower edge rises significantly towards the forehead and over the eyes, a feature we notice on earlier helmets. Lyne identifies the closest contemporary to be an Auxiliary Cavalry Type H from the early 3^rd Century, but with the replaced neck-guard and cheek pieces we see today. Lyne also notes a projecting iron boss above the ear hole, suggesting it originally functioned as a pivot for a separate peak piece in the original Auxiliary form. 

We also see on the diagram an impression of a brow or visor sketched over the forehead area; this is Lyne’s suggestion for how the reconstituted helmet was able to accept Intercisa-style cheek pieces, as they would have been difficult to attach in the helmet’s original form and as such a separate brow would need to be added to the old helmet to make the new version. Whilst this makes sense and seems likely, we must remember this is of course supposition. 

Notably, this helmet also seems to feature no sign of precious metal coating; if it is a repurposed earlier helmet, this would not be surprising, however it is further indication that, despite the stereotype, not all late period Roman helmets were covered in gold. 

The second helmet in the Richborough collection, tentatively named here the Richborough II, is notable for being composed of several pieces of bronze plating. The original finds notebook lists the following: “Bronze plating: one heart-shaped…one strip…and four fragments”. 

The heart-shaped plate (Figure 2, 2, below) is decorated with repousse work and features a central hole, suggested by both Lyne and Miks to be part of a crest holder as some type of anchoring device to prevent the crest coming loose. The long strip (Figure 2, 3, below) also features repousse work, and seemingly fits as a brow-band, mimicking the shape of a visor as it becomes thicker and more bowed towards the centre. Estimates by Lyne suggest that when intact it reached from one ear to the other rather than just forming a central peak. As he also rightly points out, the bowing of the “visor” above the brows mimics the slightly concave profile of the earlier Attic helmets, a decorative feature of votive monuments and high-status sculpture. 

Whilst there is little other evidence to confirm this unusual helmet bore a crest, the object indicated in item 4 on the diagram has been proposed as a stiffening piece for an organic crest (either feathers or horsehair), based on the Linz relief (featured in the images below). An analysis of the metal of this “stiffening piece” indicate that a material of some kind (likely leather or fabric) was glued onto it and held in the turned-up end as seen in the diagram. 

Here Lyne cites that the helmet, interestingly, has no indication of any iron corrosion on the fragments at all; he goes on to suggest that this is therefore possibly a leather helmet with metal adornments, used either as a piece of ceremonial or dress headwear, or is a lighter version of a traditional battle helmet (using Vegetius’ example of soldiers complaining about the weight of their equipment and thus shedding them accordingly). Item 5 on the diagram is a second anchoring plate similar to the one already discussed, and 6 is an even more exaggeratedly concave brow-band, suggesting a second of this type of helmet was deposited in the same context and their organic bowls have since disappeared.  It seems more likely, however, that they were instead made of a non-corrosive metal or perhaps the iron bowl was covered with a protective layer of precious metal and the latter was forcibly removed by looters. Regardless of the circumstances, we cannot be sure exactly what form these helmets took – it seems more reasonable to assume the entire panoply was made of a non-corrosive metal, however, than to concede they were a leather cap with metal decorations riveted on. 

It is interesting that this helmet of a seemingly earlier design was found in the same context as a helmet of the 4^th Century, however several explanations can be mustered; the first is simply that the deposition shifted somewhat over time and, coupled with the more lax archaeological practices employed in the 1920s, the correct dating and context of these fragments has been confused. It is also possible that it is another example of helmet re-use, a trend we know is not at all uncommon in the Roman army with at least three generations being noted to have used the same equipment in some cases – if we assume a maximum of approximately 25 years of service on average for a legionary, this gives us the best part of a century’s window in reuse time based on these averages. We could also assume it is possible, albeit unlikely, that were the Richborough II helmets to be made of or be covered in precious metals that their find context was confused upon discovery and dismantlement by looters (although why the bronze decorative plates would be left behind is a notable problem with this argument). 

 Miks dates this helmet to the early 5^th Century AD, although this is seemingly based on his inclusion of a Chi-Rho crest plate (Figure 2, 7) in the same context, which does not seem to have been the case. 

There is a third helmet from Richborough to be discussed; this one in particular is very difficult to analyse, as there is barely any material remaining pertaining to it. Found in Pit 314, a deposit also containing Theodosian-era coinage (11 January 347 – 17 January 395) as well as the skeletons of a one-child family, the only parts confirmed to be from this pit are a 1^st/2^nd Century style crest holder (Figure 3, 1, below) along with iron bowl fragments. Despite the earlier appearance of this fragment, analysis by both Lyne and Russel Robinson note that the earlier examples from Imperial Gallic helmets are usually plain, slotted under a fixed plate on the top of the helmet, and our more elaborate example from Pit 314 does not fit this methodology (seemingly being affixed to the top of the helmet permanently). 

Frustratingly, there is very little else of use surviving from the pit context, except a small fragment that survives from what appears to be a ridge helmet as it features a central “ridge” of some kind running vertically towards the centre (Figure 3, 2, below). It could have belonged to a myriad of late Roman style helmets given the small fragment we have to work with and could be of either the Intercisa or Berkasovo type. As noted by Lyne, a “spokeshave/lathe” (according to the original archaeologists) was discovered in the same pit – his suggestion is that such an object could be a misinterpreted nose guard/brow for a Berkasovo-type helmet, although as the object has since been lost, we cannot be certain whether it was related to the rest of the helmet fragments or not. 

The separate crest “spike”, an archaic device by this time (even if a “modern” version made in the 4^th Century), does make this find context stand out among peers. Lyne’s suggestion is that this uniqueness could imply the owner was the male skeleton found in the pit, and he was the garrison commander. Also found in this pit were found two boxes, one with a highly decorated copper lock plate and the other inlaid with jet; the former was found to contain seven armlets from the fourth century, a finger ring and two earrings. Crucially, the front plate was identical to those on two other boxes thrown into a ditch near where the Richborough I was discovered, along with two more armlet collections, the keyhole of at least one showing signs of being broken open. Fragments of yet another front plate of this type were found on the roadway in the west gate. At the very least, this gives us an impression of a wealthy man (or men, in separate contexts) whose belongings have been looted either upon his death or in the centuries after. 

The site also contains elements that belong to what could be other helmets; in Diagram 3, item 4 demonstrates a small length of copper-alloy, gilded, and designed to sheathe part of a helmet. It bears a marked resemblance to the “fin” shape of an Intercisa IV style helmet crest, although the piece is very badly damaged, and it is difficult to confirm this. We also see a variety of other small finds and fittings in proximity; notably a Chi-Rho monogram (as discussed earlier – Figure 2, 7) which we could expect to find on some designs of Intercisa IV-style crests making it possibly connected to the gilt sheath, and an applique shaped like a clover leaf with a central hole (Figure 2, 8) serving as decoration. 

Finally, we have the remains of one last helmet; in the third diagram, item number 3 (and the subsequent Xray image) show the shape of a cheek piece from another helmet. This piece was found in two fragments, as indicated on the diagram, and we see a copper rivet in the centre of the piece. The helmet was found beneath series of copper foil shapes, which Lyne suggests may have originally been glued to the iron itself or been part of the means by which a chinstrap was secured. The overall form the helmet took, we sadly cannot be sure of.

Some supplementary material found at the site is also worth considering. 

At Richborough, archaeologists also found a series of black glass “jewels” with a pale blue centre, of various ovoid shapes (and sadly unstratified). Seen in diagram III, item number 5, they bear a striking resemblance to the glass gems we see studding the Berkasovo I and Budapest helmets that we covered in our first articles. Indeed, Lyne notes that these gems are nearly identical if not exact matches for the Berkasovo/Budapest ones in size, form and colouration, implying that a helmet of what we currently consider to be the highest quality was either deposited at Richborough or, for whatever reason, spare glass gems for one were left on-site. 

In all, we have quite an interesting picture painted for us of the helmet finds at Richborough; some archaic design, a suggestion of reuse and also an indication of great wealth and status. It is interesting to note that the overall story seems to imply at the very least a spate of site looting following the end of the Roman occupation, but potentially of a more violent end to the fort’s occupants; numerous loose shield bosses were found on the upper levels of the site, and a concentration of human skeletons with spears and arrowheads were discovered in the ditches in front of the West Gate, where the Richborough II helmet was discovered (along with the possible fourth helmet, some other small helmet fragments and a shield boss). Lyne notes as well that at least two of the projectiles from the ditches had bent points where they had rebounded from a hard surface, such as a wall. With most of the helmet fragments coming from these deposits, it gives us an image of a rather sudden end and deposition of the equipment at the turn of the 5^th Century. 

Lyne does note that he believes the antiquated appearance of some helmets found here demonstrates a decline in quality of head protection as used by the very late Roman army and post-Roman British, with any quality items being reserved specifically for senior officers. This may not be the case, however, as the Romans have a habit of revering the antiquated; indeed, the same design of musculata is seen on statuary evidence from the Republican era to beyond the fall of the west, and both mosaic and fresco often display characters with an antiquated look about them even into the 4^th and 5^th Centuries. 

Such a discovery could imply the Romans had a greater admiration of the “classical” era than we first thought, the finds from Richborough aligning well with a lot of their contemporary mosaic and fresco pieces, implying the trend is not simply limited to artwork. On the other hand, it could indeed be a prime example of an empire stretched to the limit in resources and manpower and giving a backwater province such as Britain (despite being heavily militarised for many years) less priority for new equipment than their more desperate cousins along the Danube and Rhine. 

Sadly, it seems to be a situation where, as per usual, we will never know the answer. The Richborough scene is sadly muddled, with both older items and new mingling together. The second helmet is something of a curiosity (particularly coupled with a potential second one) that seems somewhat out of time with its deposit, as does the crest holder from helmet III, both in a 1^st Century or earlier style but found in a IVth – Vth Century context alongside fragments of standard late Roman ridge helmets. The argument of limited resources and wealth stretched to the limit makes sense considering the repurposed helmet I, but helmet II and III appear somewhat ostentatious and the finding of Berkasovo/Budapest style decorative glass gems cannot be ignored. 

In all, quite a jumping off point for discussion and debate on these intriguing finds… 

Images courtesy of Malcom Lyne in “Late Roman Helmets from Richborough”, in the Journal of Roman Military Equipment Studies Issue 5.