Beloved Objects

be Tessa de Alarcon

Like most other Philadelphia residents, the Penn Museum staff are adapting to working from home. As part of this, the Museum staff have recently been posting on the museum’s Instagram feed info on their favorite objects (pennmuseum #VisitFromHome). This got me thinking about the relationship between people and the things we interact with every day. The objects in the museum’s collection, while loved and cared for by the staff, also bear evidence of love and care from before they were in the museum’s collection. One such object recently came across my desk for treatment, E7517A and E7517B, a Nubian wooden box and lid from Karanog. I am not going to talk about the treatment today, so that I can focus on the care it received before it entered the museum.

E7517A and E7517B after treatment.

In the pictures above and the details below you can see that this wooden box has a variety of metal components, including copper alloy straps and a lock plate on the box, and staples on the lid as part of a repair to cracks and breaks through the wood.

The detail of E7517A, the box, on the left shows the lock plate and one of the metal straps, and the detail on the right of E7517B shows the staples on the lid.

Staples like these are a common repair both in antiquity and historically for a variety of materials and are not an unusual feature on objects in the museum’s collections (here are just a few other examples of both types of staples: AF5211, B9220, 2006-15-41, B20014). If you look closely though, you can see that the metal straps and the lock plate go over the inlays on the box. This suggests that these elements were not part of the box originally and were a later addition.

These components are also made from a variety of metals. I tested them using both a magnet to check for iron, and a portable X-ray fluorescence spectrometer (pXRF) and found that they are a range of metals including iron, copper tin alloys (bronze), and copper zinc alloys (brass). Also, parts that appear similar, like the straps are not made up of the same alloying components, some contain lead in addition to the copper and tin, and some have no lead. The staples are also a range of metals including iron, brass, and bronze.

When I started working on the box, I wanted to tease out when these metal components were added as they could have been either ancient or modern. With this type of question, I typically set up an appointment with our archives to look at the original field notes and field photography. However, in this case, much of the data on the excavations at Karanog is online, including pdf’s of the excavation publications. In the museum’s database I found that the box was from a burial: tomb G 445. Going through the publication, I was able to use the context information to find not only a description of where it was found within the tomb, but also a sketch of the burial, a photo of the box, and a detailed description of it in a catalogue of the finds. The box had been found in the burial with two individuals buried one above the other and was found next to their legs.

Drawing of tomb G 445 from Woolley, Leonard, and David Randall-MacIver. Karanòg: The Romano-Nubian Cemetery. Vol. 3. University museum, 1910: 44. The blue highlighting is an addition to point out the location of E7517A and E7517B in the burial.
Plate 22 from Woolley, Leonard, and David Randall-MacIver. Karanòg: The Romano-Nubian Cemetery. Vol. 3. University museum, 1910.

The textual information from the publication includes some important pieces of information: first that, “it had been considerably restored before being deposited in the tomb, brass binding had been added at the corners and the broken lid had been rudely mended with bronze rivets” (Woolley and Randall-MacIver, 44) and that “it remains in the condition in which it was found, no repairs to it having been necessary” (Woolley and Randall-MacIver, 71).  While the language used to describe these metal components seems to me a bit harsh, not only is it described as “rudely mended”, the lock plate is described as “a perfectly useless lock plate”, it does make it clear that these metal components are from when it was in use (Woolley and Randall-MacIver, 44, 71). It should also be noted that the metal identifications given in the publication were not done through analysis, so don’t match with the results I have from pXRF.

Because of the detailed information in the publication, I also know what was in the box when it was excavated: another smaller box (E7510A and E7510B) and two wooden spindle whorls (E7506 and E7507).  These are all shown in the image below.

Objects found inside E7517A: a box and its lid (E7510A and E7510B) and two spindle whorls (E7506 and E7507)

So, all together what does this information tell me about the history of this box? First, the repairs and modifications to this box happened during its use before it was put in the burial of the two individuals in tomb G 445. The fact that the metal components, even similar ones, have different compositions could mean a few things. It could be that it was repaired and modified using scrap metal with the components being made from different scraps, that the repairs occurred at different times and so with different metals, or both. If they were not made using scrap metal, it is possible that some of the straps may have had to be replaced at some point and that may be why some are leaded bronze and some are not. These straps do not appear to have a function and may instead reflect changes in taste. The function of the box may have also changed, and this may be why they needed to add the lock plate. The various metals for the repairs to the lid almost certainly resulted from various treatment campaigns, meaning that it was repaired, used, broken, and repaired again. In any case these modifications and repairs tell a story of care and use and suggest that this box was loved and treasured by the people who owned it. This may also be why it eventually was placed in a burial, perhaps as a particularly prized possession of one or both of the individuals in the burial.

This project was made possible in part by the Institute of Museum and Library Services.

Lessons Learned: You Cannot Treat Dry Wood Like Wet Wood

by Tessa de Alarcon

Documentation of conservation treatments undertaken in the lab is a very important part of what we in the conservation department do at the museum. One of the main reasons we document our treatments is so that conservators in the future don’t have to try to figure out what was done to an object. Instead those future conservators can read our reports and start off knowing the treatment history.  

An array of treatment reports, and log book entries, from the conservation department at the Penn Museum

We are also sometimes those future conservators, looking back at previous treatments. This means that not only can we see an object’s treatment history, we have the opportunity to evaluate and learn from the decisions made by conservators decades ago. Some of these treatments were successful, and some were not. Now we know to avoid those treatments that clearly did not work.

I recently completed a treatment on an object that is a good example of this process. E11151 is a carved wooden figure from Nubia. It was treated before I worked on it by a conservator in the 1970’s. The photo below is how the object looked when it entered the lab at the end of last year.

Before treatment image of E11151 taken in 2019

In the case of this object, it was noted in the 1970’s as having a slightly powdery surface. The conservator conducting the treatment decided to apply a consolidation method that is frequently used on waterlogged wood: immersion in a solution of PEG (polyethylene glycol). PEG treatments are still done today and are very effective at consolidating and stabilizing waterlogged wood before it dries out. But because of this treatment and others done at the museum around this same time period, we have learned that consolidation with PEG is not effective on dry wood, even if it once was waterlogged. E11151 was dry wood and did not come from a waterlogged context.

E11205, E11187, E11203, E11151
E11151 is on the far right.
This publication photo was taken before the object was treated in the 1970’s.

The photo above is a publication photo of the object before it was treated in the 1970’s. While there is no photography with the 1970’s treatment report, there is a sketch which suggests that the object looked similar to this photo when it entered the lab at that time. As you can see there are quite a few differences in the object’s appearance as it entered the lab in 2019 and how it looked before it was treated in the 1970’s.

Before treatment photo of E11151. The red arrow is pointing to an area of thick, white, and hard high molecular weight PEG and the blue arrow is pointing to soft and sticky low molecular weight PEG

Two different molecular weights of PEG were used on this object, a hard high molecular weight one that was used to consolidate the powdery wood and looked white on the surface where it was very thick, and a low molecular weight one that was used to join pieces of the object together and left the surface tacky and sticky. The sticky PEG also trapped a lot of dirt and dust on the surface of the object.

After treatment detail of the top of E11151

The detail above shows the top of the head of the object. The report from the 1970’s states that during treatment the object began to crack and fragment. PEG is typically dissolved in water or ethanol. Both solvents were used in the PEG treatment of this object. These two solvents can be mixed together and during a normal PEG treatment the wet wood starts with PEG in water and then moves to PEG in ethanol: this helps start the drying process. As the waterlogged wood is also already wet, the PEG can penetrate fully into the swollen wood. However, in the case of dry wood, these solvents (ethanol typically has some water in it) introduce moisture into the object, and it starts to swell as wood is very responsive to moisture. It then has to dry out again after the treatment, causing the wood to shrink. This is stressful for the object. This stress is what likely caused the cracking documented in the report and visible in the image above.

E11151, after treatment photo from the treatment completed in 2020

Here you can see what the object looks like now that I have finished my treatment. Consolidation is a very permanent and tricky to reverse treatment, even when adhesives that remain soluble, like PEG, are used. There is currently no way to remove the PEG from this object. All I have done is reduce the PEG on the surface by cleaning it with ethanol. I also used the previous documentation to figure out where small detached fragments went so that it looks more like it did originally. While the treatment that was undertaken in the 1970’s seems to be over-treatment as it caused new problems, some more severe than the problems the object had to begin with, I do also want to recognize that it is because of past experiments like this one that I know not to use PEG on dry wood.

Because wholesale consolidation is a fairly permanent and risky treatment, I think long and hard when I choose a consolidant. I also remain aware that there is a chance that someday, some future conservator, will deem some of my treatments mistakes as well. Hopefully my mistakes will be ones that they can learn from too.

To learn more about PEG treatments for waterlogged organic materials, check out these links:

Ellen Carrlee Conservation blog: High Molecular Weight PEG for basketry

Queen Anne’s Revenge blog: Waterlogged wood – Gotta bulk it up!

Cardiff University Conservation: Newport Ship Workshop

This project was made possible in part by the Institute of Museum and Library Services.