Blog: Archeoleg: PAS Cymru

Just a short blog entry today, describing the completed excavation of another area of the soil block, and some of the interesting features that have cropped up.

This section of the block is composed of what appears to be two flat lorica plates, one lying at a 45 degree angle to the other. Note the length of the straight edge of plate; I believe that this plate will be one of the large plates that came across the middle of the cuirass. I have included annotations indicating small areas of potential importance, such as the corroded remains of fittings (see red arrows), which stand proud to the surface of the plates. The gap between the plates, which shows how damaged and broken the edges of the lorica set really are, can be seen in the second picture.

I have found another fragment of plate with a rolled edge (see third photograph), though the roll itself is much narrower in comparison with that exposed within the girth hoop (refer to previous blog entry). The fragment itself is also a little too small to detect any curvature or to easily extrapolate a larger shape, but could this fragment be part of a plate (the breast or backplate) that would have been in contact with the wearer’s neck? All comments and opinions regarding this little hypothesis are welcome.

I have included a macro shot of a small cylindrical item: whilst this may be physically unimpressive, I believe that this could be the iron pin that would have been drawn through a lobate hinge, holding the shoulder plates together.

As mentioned above, obvious fittings that are immediately identifiable still haven’t been found, and careful excavation has only managed to produce vague shapes of what is essentially metal corrosion. I have included in the last photograph a view of an area of probable lorica attachments and fittings, though only a very good quality x-ray will be able to make any sense of these lumpy features.

As a last aside, I thought I should provide a brief explanation for the condition of the buried lorica segmentata. Readers may have noticed how exposed finds lack the thick crusts of rust and voluminous corrosion products typical of a lot of archaeological iron objects: this is most likely because the thin iron plates corroded extremely quickly, with the iron leeching into the soil. Whilst this does mean that I will not have to spend hours removing powdery iron corrosion in order to reach a more certain surface on the iron, it also indicates that the remaining ‘object’ is more of a pseudomorph lying on top of the soil: this is why the ‘plate’ most often does not respond to the pull of a magnet. This level of deterioration will have implications for the eventual conservation treatment of the armour, as I may be unable to extract the iron plates (which have very little physical integrity), from the soil. 

To the right is an image of the lorica segmentata girth hoop, complete with a copper alloy tie ring, after excavation in the conservation laboratory. The general shape of the hoop, which is lying on its side, can be discerned, and it is obvious that the plate is broken in several places. I had hoped to be able to see how the solid tie ring interacted with the girth hoop, but unfortunately too little of the iron plate remains in a stable enough state to remove enough of the obscuring soil currently supporting it. An x-ray of this area of the block will show the complete shape of the tie ring, and hopefully how it is attached to the plate.

To reiterate, I am excavating the artefacts in this soil block in the same way that archaeologists do in the field, and am coming across the same extraneous stones, silty-soil and common artefacts such as ceramic building material and animal bone contained within the burial deposit that are found on the average archaeological site. As the stones, tile, etc, do underlie so much of the fragile iron plate, I will not remove them at this stage, though I am aware they could be covering more artefacts.

Within the hoop further broken plates can be seen, most notably a flat plate with a rolled edge: this is one of the few plates that can at this early stage be attributed to a specific area of the cuirass; the only plates with rolled edges were the lowest girth hoop above the hip, and the breast and back plates. Given the flat nature of this plate, it is most likely the lowest girth hoop.

To give readers an idea of the depth of the archaeological remains in the soil block, I have included photographs of the back of the girth hoop, as well as a photograph showing the interior. Of interest in the third photograph is part of a plate held to the back of the girth hoop by corrosion and burial products, and how thin the lorica plate with a rolled edge is in the fourth.

Not all artefacts excavated in this area of the soil block can be as easily identified as the plate with the rolled edge; some of the iron has suffered greatly, and only vague shapes in the soil can be distinguished. Removing soil from the surfaces of these mineralized objects has been challenging, as I have to be careful not to take away any burial deposit that might contain corrosion that may be part or all that is left of the archaeological object.

I have not recovered as many fittings, such as buckles, rivets, washers and studs as those found on cuirasses from other Roman sites, and it will be these attachments, and their position on the body armour, that will be very important when trying to work out which plates were breastplates, backplates, shoulderguards, etc. As expected, I have not found many of the valuable copper alloy fittings which could have been removed for re-use during and after the occupation of the fortress by military forces, although I have found possible iron components, harder to spot given the corroded nature of the metal. For example, in the last photograph, the red arrows are pointing towards vague rectangular shapes slightly proud of what is an extremely fragmentary lorica plate; these might be the remains of fittings. Again, only high quality x-radiography will provide us with an image which might allow curators to make an identification of what these deteriorated artefacts are.

The hoop and other items will remain in place until the rest of the block has been excavated, when deconstruction will take place.

At this juncture in the investigation of this block-lift, I am making every effort to outline the relationships (if any) between artefacts. As can be see in the first photograph, plenty of small pieces of iron plate, often with no telling association with larger plates, emerge as soil is scraped away. Aside from photographing their position for future reference, and examining them for signs of the remains of fittings, impressions of textile or leather, there is not much that can be done with these anonymous fragments. Moreover, these fragments often overlie more interesting and coherent features, and so I am generally removing these: I will most likely x-ray these in large batches at a later date. As you can tell by the annotations, I’ve begun to get a good idea of the fragile nature of the fragmentary, corroded copper and iron artefacts mixed in the burial deposit, and have begun to grasp how difficult lifting the larger pieces of lorica will be.

So far I have had limited success at recovering any ‘true edges’ of the iron armour, as most of the vulnerable thin plate has been broken. Finding edges greatly improves our chances of identifying plates, and where two edges have been found, dimensions such as the width of the plate can give us an idea of which part of the lorica cuirass the plate comes from. It also helps us to make educated comparisons with examples of Roman armour found from other sites. For instance, the iron plate recovered in the second photograph has a width of 6.5 cm across, dimensions similar to those recorded for the armour fragments found amongst the Corbridge Hoard, and from the Austrian site, Carnuntum. It also has the very corroded remains of two copper alloy rivets, which improves our understanding of how the cuirass was constructed and held together.

As I work I am keeping the surface of the soil block wet, by spraying it with deionised water. This prevents the soil from drying out too much, separating, and breaking the iron remains as it falls into chunks. As most of the iron is in such a poor condition, consolidation with a removable acrylic adhesive, such as Paraloid B72 (ethyl methacrylate copolymer) is a must (which is why in some photographs the iron surface appears to have a dark sheen to it).  

Whilst excavating an area of the block to the left of the photograph, I came across an exciting, (and sadly, very degraded) find: copper alloy wrapped around a thin iron plate. It can be seen in-situ in the photograph to the right, and after excavation in the photograph below. Sadly, as not much of the object has been recovered, a firm identification of this piece hasn’t been reached yet, though further excavation might yield more clues.

Readers may have noticed that I have begun to clean the outside of what is most likely a girth hoop. The exposed iron plate is 1mm in thickness, and the hoop is broken in several places, that I can see from the surface. When focusing on this feature, I will have to be careful to remove enough soil and other burial debris to reveal the curved plate’s shape, whilst maintaining the earthy support until I am ready to remove the that section of armour from the soil block. The next blog entry will focus on describing the results of excavations in this area, which includes a copper-alloy tie loop, still associated with the iron plate.

After documentation, the next step was to take stock of the overall condition of the block, and to make a preliminary inventory of the types of archaeological materials I could see.

As can be seen in the birds-eye-view photograph, there are some large cracks running across the block: these most likely occurred during lifting and transportation from the site to the museum. Unfortunately, these extend through some of the exposed armour, and areas of thin, mineralized iron plate have broken. This kind of damage, whilst regrettable, couldn’t have been avoided.

I wanted to keep the broken pieces in place for as long as possible, as the positioning of the remains is important to our interpretation of the events taking place in these two rooms of the warehouse. In order to ensure that fragments stayed together (at least for the interim) I used a wax-like substance called Cyclododecane, which I melted and brushed onto the artefacts. The Cyclododecane will eventually sublime altogether, and so I will not have to remove it later.

In addition to the iron plates, there are also a range of other interesting artefacts visible on the surface of the block. There appears to be a scattering of copper alloy scale armour (see photograph for an example), which would also have been worn by a soldier.

At the edge of the block is a large copper alloy stud, its use in antiquity currently unknown: it is not an artefact which has been associated with lorica segmentata. As the photograph shows, the copper has corroded considerably and is very thin.

Scattered amongst these exciting finds are the usual types of artefacts found during archaeological excavations: pieces of red ceramic tile, fragments of pottery, bits of animal bone, and crumbly lumps of white plaster (from the building itself).

Excavating the block will be both time consuming and challenging: I have selected a number of small tools to use. I am unable to place a microscope over the block, and so will be using an optimizer (a visor with magnification lenses) instead- I do not want to miss any small artefacts or details during the excavation.