To journal index
10 August 2011

Conservation of K843 and K682

As a conservation intern with the Staffordshire Hoard, I am in an enviable position this summer. I arrived in Birmingham from the United States at the beginning of June, and I have yet to experience a dull moment. Within the first two weeks alone, I encountered film crews, engaged in public tours and learned about exciting discoveries made by my colleagues.

I have also had the chance to treat several objects. This blog post will detail two of these treatments, which demonstrate the boundless potential of the Staffordshire Hoard to surprise us and present new questions for further research.

Objects Before Treatment

Prior to treatment, the two objects in questions appeared to be constructed similarly (figs. 1 and 2).

K843 Lentoid Plate prior to conservation

Figure 1. K843 Lentoid Plate.

The first, K843, is a stunning lentoid made of gold and garnets, a possible pair to K270, treated by Deborah Magnoler and described by her in a previous blog post.

K682 Cloisonné Strip prior to conservation.

Figure 2. K682 Cloisonné Strip.

The second, K682, is a gold strip. Like the lentoid, it is embellished with cloisonné cells. The cells in both objects were heavily caked with soil prior to cleaning, obscuring the distinctive features of the pieces.

Treatment of K843

The lentoid was my first treatment, and it was a rather striking introduction to the Hoard. I began by examining the object and taking note of its condition and construction. Then I took photographs and wrote a brief report to document the object’s state before treatment. Certain features were apparent before removing the soil. The object was clearly designed to impress, with an estimated 239 cloisonné cells, gold backing foils stamped with two different patterns, and a double border of filigree around the outside edge.

In treating the lentoid, I followed a protocol that is fairly consistent for Hoard objects. Working at magnification levels from 7.5x to 75x, I used a natural thorn held in a pin-vice to carefully break up soil (fig. 3). Then I used a fine, soft brush to lift the particulate from the surface. Where necessary, I used a small amount of water or industrial methylated spirits (95% ethanol + 5% methanol) on a swab to help soften compact soil and remove residues.

Garent being cleaned with a thorn.

Figure 3. Cleaning a garnet with a thorn 25x.

Everything that I removed from the object, including soil, organic material and corrosion products, was photographed in situ and retained in labelled sample vials. In the future, these samples may provide critical information through analysis. Prior to analysis, it is crucial to limit the introduction of modern materials to an object, but consolidation is sometimes necessary when elements are significantly loose or fractured. In treating the lentoid, I consolidated four garnets and one loose backing foil using tiny drops of Paraloid B-72, a reversible adhesive that is often used in conservation.

During treatment discoveries about K843

After treatment, the lentoid’s design can be fully appreciated (fig. 4) and its condition can be accurately assessed.


Lentoid K843 after treatment, garnets can now clearly be seen.

Figure 4. Lentoid K843 after treatment.

Altogether, 211 garnets remain in the object. Some cells are missing garnets or backing foils; and while this interrupts the design, it also reveals the astonishing intricacy of the object’s construction. In the empty cells, possible remnants of a backing paste exist. This would have bulked up the cells beneath the patterned gold foils into which the cut garnets were precisely set (fig. 5).

Garnets, backing foils and empty cells

Figure 5. Garnets, backing foils and empty cells 7.5x.

Several rivet holes and one intact rivet, seemingly in its original location (fig. 6), indicate that the object would have fastened to a support of some kind. For the present, this object remains a mystery.


Figure 6. Extant rivet atop a silver divider 10x.

Treatment of K682

I began treating the cloisonné strip in much the same manner as the lentoid. Even prior to treatment, I made a number of observations about the object. The right end of the strip bends back on itself, and certain features suggest that the two short edges of the strip were originally joined to form an elliptical shape, which possibly would have been used as a hilt collar. One of the most compelling indications for this theory is the presence of excess border extending from the bottom edge on the right of the object and a lost portion of border at the bottom edge of the left side of the object.

Excess border

Figure 7. Excess border measured at 13.44 mm 10x.

I took measurements of these two areas, and they are almost equal in length, about 13.4 mm (figs. 7 and 8).  I also found that the partial cells on the right edge matched up well with the partial cells on the left edge.

Missing border

Figure 8. Missing border measured at 13.49 mm 10x.

During treatment discoveries about K862

Although the contents of the cells were covered in soil, they were, I assumed, likely to contain garnets. I did notice the presence of greenish deposits visible through the soil in a few of the cells, but similar green material was found on the lentoid, and these cells contained garnets (figs. 9 and 10).

Green material on strip

Figure 9. Green material on strip in cell near top edge BT 25x

After the first few cells were excavated, however, it became apparent that this object was very different than the lentoid. None of the cells contain garnets.  Nor do they contain backing foils or any other evidence that they were once set with garnets. Instead, the large majority of cells are partially or completely filled with the compact green material that was only present in traces on the lentoid and assumed to be a corrosion product. After this discovery, I proceeded to remove the loose soil from atop the cells, but I left all of the green deposits in situ. Orange residues were also found in some of the cells, typically on top of the green material.

Green material on lentoid

Figure 10. Green material on lentoid near bottom right edge BT 40x

These findings present more questions than answers. The elaborate cloisonné cells on the object must have been inlaid with something originally. The appearance of the deposits suggests copper corrosion products, possibly from a colorant in glass, but if the deposits are remnants of deteriorated glass, why have glass inlays survived in much better condition on other objects from the Hoard? Another theory is that the cells contained enamel. Enamel is essentially made by fusing powdered glass. The ratio of the components in enamel may be adjusted to change properties such as melting temperature. For example, with a high percentage of alkalis and low lime content, enamel will flow at lower temperatures but may be very prone to decay. Perhaps K682 was originally embellished with a beautiful but short-lived material like this. Further analysis by the British Museum science department will help to solve this mystery and other key questions surrounding the hoard.

Conserving the objects in the Staffordshire Hoard is a crucial step in piecing together their history. As more objects are conserved, we are able to identify candidates for in depth analysis and establish key relationships among individual artefacts. I feel privileged to be part of the team.

During my 9 week placement, I have also recorded video blogs of my experience, which focus on my work in the conservation studio as well as installing the Staffordshire Hoard on the Mercian Trail in Lichfield Cathedral.

Ellen Promise, Conservation intern from Winterthur/University of Delaware Program in Art Conservation.

Cadmium sulphide is used in enamelling and that will decay to a orange/ yellow residue.

Thank you for you comment on the blog. Cadmium Sulphide and other colouring elements for glass and enamel will be one of the areas for analyisis when trying to understand these objects. The analysis is at the early stages and it is hoped that as information is revealed that there will be blogs on the findings.

Regarding the lentoid plaque , K843, it reminds me of the typical lenticular cross section of modern Kukri knife handles as well as some Viking knife handles excavated at Frogel on Gotland, by Professor Dan Carlson of Gotland University. Perhpas it is a pommel plaque from a special Seax. The two rivits indicated could have connecxted it to other hilt elements as a cover for the end of the tang. It appears rather large for such an application ( almost 8 cm long ?), but still it may have been for a special weapon.


Not too far from the shape of the eagles thigh
from Sutton Hoo ?

Excellent work! Precision to the highest standards! Would you have any information on how I could recreate some of the gold and garnet decorated pieces in particular! I know jewellery workshops is the obvious way to go but any info and or where to find it would be great! How can get into any sort of restoration work like this?