A Columnar Matter Part III: The Conservation and Installation of a 3rd Millennium BCE Mosaic Column from Al ‘Ubaid

By Marci Jefcoat Burton

As a last installment of the mosaic column from Tell al-Ubaid in Iraq, we are pleased to announce the treatment and remounting of the 4,200(+) year-old shell, pink limestone, and black shale tesserae is complete (Figure 1)! After a number of years in collection storage, the recently conserved column is now in the Middle East Gallery, and ready for visitors when the exhibit opens this Saturday, April 21, 2018. The treatment process was a dynamic collaborative project, involving a team of seven conservators to clean, repair and re-mount hundreds of triangular and diamond shaped tesserae over the span of 18 months.

Figure 1: The previous column support (left), with all tesserae deconstructed from each section. Only mounting materials, such as plaster ground and reconstructions, remain on the supports. (Right) The newly mounted tesserae on Ethafoam column supports.

The tesserae were originally mounted at the archaeological site in 1919 – 1924 on four hollow cylindrical sections made of wire mesh and burlap. The tesserae were imbedded in a thick layer of plaster, and the resulting weight on a somewhat flexible base eventually became a structural problem. In addition, each column section had a large area of plaster reconstruction to continue the mosaic design around the column. For a memory refresher of the previous condition and treatment protocol, visit our blog posts, column blog 1 and column blog 2.

After months of cleaning every tessera and adhering many fragments back together, the tesserae were mounted onto a new support made from solid Ethafoam, a chemically stable and dense foam (Figure 2). Each new column section support was reconstructed to the same measurements as the wire mesh and burlap supports. The previous use of plaster proved too rigid and heavy in the original mounting system, causing many cracks to develop in the plaster and tesserae to loosen over time. For these reasons, plaster was not used again, and the tesserae were mounted to the Ethafoam supports with an acrylic paste (Paraloid® B-72 (ethyl methacrylate (70%) and methyl acrylate (30%) copolymer) in acetone bulked with glass microspheres) toned black with dry pigments. This is lighter in weight compared to plaster, and compatible with the shell and stone materials. The black acrylic paste was made to resemble bitumen, a pine resin material used by the Mesopotamians to originally mount the tesserae to the c. 3rd Millennium BCE column.

Figure 2: Mounting process of column section #2. The cleaned and repaired tesserae are mounted to the Ethafoam cylindrical supports with Paraloid® B-72 (ethyl methacrylate (70%) and methyl acrylate (30%) copolymer) in acetone bulked with glass microspheres and toned black with dry pigments to mimic bitumen. Spatulas were used to spread the acrylic resin on the support and tesserae were imbedded into the mixture.

All four column sections stack and connect with an internal wooden dowel system. To accommodate the large areas that were once plaster restorations, our Photo Studio printed high resolution, archival images of the tesserae, which were custom scaled and fit to each column section (Figure 3). The digital photo fills integrate a complete mosaic design to give the appearance of a fully mosaicked column (Figure 4). This is a great example of how recent technological advancements help conservators with approaches to treatment and display options.

Figure 3: (From left), Conservation Interns Tessa Young, Alyssa Rina and Jennifer Mikes installing digital photo fills to the top and bottom mosaic column sections.

Figure 4: (a): Column section #2, after remounting tesserae. The locations with previous plaster reconstructions are padded with Volara (closed-cell polyethylene foam). (b): Column section #2 after pinning archival photo fill into place over the Volara.

The mosaic column, and over 1,500 objects await your visit in the Middle East Galleries (Figure 5)! To our visitors in the Artifact Lab who witnessed our many, many hours of treating these column sections, we thank you for your brilliant questions, comments and curiosity about the tesserae and our conservation process during our Open Window Sessions. You will find the assembled column has transformed into an impressive and complete piece, and we hope you enjoy it as much as we do!

Figure 5: Bob Thurlow (Special Projects Manager) and Marci Jefcoat Burton (Conservation Curriculum Intern) installing the four mosaic column sections in the Middle Eastern Gallery.

Putting the finishing touches on the shabti box

I have put a lot of work into our troubled shabti box, including investigating and analyzing the varnish (more on the analysis in an upcoming post), doing some pretty cool imaging, and consolidating all flaking and unstable varnish and paint with methylcellulose. After consolidation of the surfaces, the box does not look much different than it did when I started the treatment (and this is a good thing). As a reminder, here is an image of the front of the box before treatment:

shabti box frontAt this point, I could call the treatment done, or take it a step further, by filling in some of the losses of the painted surface, which appear bright white since those losses expose the gesso below. After consulting with Dr. Jen Wegner in the Egyptian Section and with Lynn Grant, the head of our department, I decided to fill in some of the larger losses which really make it difficult to appreciate the object and “read” the designs. I have even heard some visitors refer to the box as “that badly damaged piece of wood”, and that is not what we want people to be thinking when they eventually see this on display. While I know I can never return the box to its original condition, I can reduce the appearance of some of the damage. But how to fill the losses on such a fragile surface, in a way that will be reversible/retreatable?

After some hemming and hawing and some failed tests, I ultimately decided to fill the losses by first placing a small piece of Japanese tissue paper into the loss, then applying a tinted fill mixture over the paper. I did this by doing the following:

1. I took a quick snapshot of the surface I was about to work on. I then downloaded the image and copied it into a Word document. Using the scale in Word, I was able to resize the photograph in order to print it approximately true to size, and then I printed the image in black and white. This took no more than 5 minutes.

2. I placed a piece of Mylar over the B&W print-out and traced the losses I wanted to fill with a black marker.

L-55-23A_template2

B&W image with Mylar template moved off to the right side

3. After trimming the Mylar around one of the tracings, I taped it to a piece of Japanese tissue paper with a small piece of blue tape.

L-55-23A_template34. I cut out the Japanese tissue paper and adhered it into the loss on the shabti box with a small amount of 5% methylcellulose.

5. I then applied a fill mixture over the Japanese tissue paper. The fill mixture is made of 5% methylcellulose, glass microballoons, and powdered pigment.

Fill mixture (in the jar and on the spatula)

Fill mixture (in the jar and on the spatula)

This may sound tedious, but the whole process works very smoothly and relatively quickly. It also minimizes the amount of time I need to spend touching the object and therefore minimizes damage that might be caused by touching the very fragile surface.

I’m not finished, but so far I’m pretty happy with how the front of the box is looking:

L-55-23A_dt01_compressed

Front view, during filling

It’s subtle, but to see the difference that filling makes, here are views before and after, side-by-side:

Picture1

Before treatment (left) and during treatment (right)

The only problem is, I feel like I’ve opened a can of worms. There are so many losses and I am not going to fill them all, but as soon as the larger losses are filled, I start seeing all of the small ones! I think it’s looking better though and I will get some feedback from my colleagues before proceeding further.

 

Treating fragments of a Middle Kingdom painted wooden coffin

If you’ve visited the lab in the last few weeks you may have seen me, head bent at the binocular microscope, working away on fragments of a painted wooden coffin from Abydos. These fragments (7 in total) were excavated in 1901 and have been here at the museum ever since. As I described in a previous post, these boards were severely damaged by termites prior to excavation, and the painted surface, while very well-preserved in some areas, was cracked, flaking, and barely attached in places, not to mention covered with grime.

One of the coffin fragments, which features a portion of a frieze of objects that includes two vessels with spouts and a bolt of clothing.

A before treatment photograph of one of the coffin fragments, which features a portion of a frieze of objects that includes two vessels with spouts and a bolt of clothing.

On the board in the image above, the paint was actually in decent condition. After cleaning the surface with bits of a kneaded rubber eraser, I stabilized the edges around the paint losses with a 2% solution of methyl cellulose in water. With the help of an intern, we sorted through a box of much smaller fragments that presumably had become detached from the 7 larger boards at some point, and we found two small fragments of wood with painted decoration which belonged to this board. These fragments were adhered in place with a 1:1 mixture of 5% methyl cellulose and Jade 403, an ethylene vinyl acetate emulsion.

E12505emends

In this image the red arrows point out the two fragments which were adhered in place after cleaning and consolidation.

Here is a view of that area from the back after mending those fragments:

e12505e_backbeforemendingThe termite damage is evident from the back, and as you can see, the wood is very thin in this area in particular, only about 1mm thick along the join edges between the small fragments and the larger board. The small loss to the right of the upper fragment is an area where the wood and painted surface have been lost completely.

Because the wood is so thin and fragile, I decided to provide some support to this area, by first adhering a piece of Japanese tissue paper over the loss from the back with a 5% solution of methyl cellulose.

e12505e_backmend

A detail of the Japanese tissue paper support adhered over the loss

I then filled the loss and the small gaps along the join edges of that upper fragment from the front, using a fill mixture made from 5% methyl cellulose, glass microballoons, and powdered pigment.

e12505edetail

A detail shot showing the fill from the front

Here is an overall view of the board, after treatment:

E12505Edt02_blogThe fill mixture I used worked nicely, and I’m now using it to stabilize the edges of some of the lifting paint on the other coffin board fragments where the painted surface is in worse condition. I will post photos soon showing what the coffin boards look like before and after treatment.

 

Completing the treatment of Tawahibre’s coffin

Things have been pretty busy around here lately, and I almost forgot to post some updates about several projects. One project in particular is the treatment of Tawahibre’s coffin. We have been working on this 2-part painted wooden coffin in the lab for the last year, and we recently completed its treatment.

As you may remember, when the coffin first came up here, it was covered with a thick layer of dust and grime, the paint was badly flaking in areas, several large pieces of painted gesso were pulling away from the wood support, and there were large cracks throughout.

Before treatment photos (clockwise from left): upper half of coffin showing layer of dust and large cracks and losses; large piece of painted gesso partially detached from top of head; large loss on wig, showing old animal glue adhesive from a previous restoration

Before treatment photos (clockwise from left): upper half of coffin showing layer of dust and large cracks and losses; large piece of painted gesso partially detached from top of head; large loss on wig, showing old, shiny animal glue adhesive from a previous restoration

After cleaning the surface with a brush and vacuum, followed by cosmetic sponges, I consolidated the paint with a methyl cellulose solution, filled in cracks and gaps using Japanese tissue paper and a mixture of methyl cellulose bulked with cellulose powder and glass microballoons, and then toned the fills with acrylic paint. This work is explained in further detail in previous posts, which you can find by clicking on the links included in blue above.

fillingcracks

A detail shot of the wig showing an area with several large open cracks before and after filling with Japanese tissue paper and methyl cellulose/cellulose powder/glass microballoons mixture

Based on a discussion with our Egyptian section curators, I also made some aesthetic fills to mask some large losses, including 2 losses on the wig. We chose not to fill the losses on the nose and chin because filling these losses would require too much guess-work as to the original contours of these features.

Large loss on wig before (left), after application of Japanese tissue paper layer (middle), and after application of fill mixture (right)

Large loss on wig before (left), after application of Japanese tissue paper layer (middle), and during application of fill mixture (right)

Detail of the head and wig before (left) and after (right) conservation treatment, with losses in before treatment photo outlined in red

Detail of the head and wig before (left) and after (right) conservation treatment, with losses on the wig outlined in red. The larger loss on the right is the featured in the previous series of images.

I carried out similar work on the base of the coffin, and now both are complete:

Tawahibre's coffin lid before (left) and after (right) conservation treatment

Tawahibre’s coffin lid before (left) and after (right) conservation treatment

The coffin base before (left) and after (right) conservation treatment.

The coffin base before (left) and after (right) conservation treatment

As you can see, we chose not to fill many of the losses, focusing instead on stabilization.

This work will enable future exhibition of the coffin, and just as importantly, it will make further study of the coffin possible. All along there have been some discrepancies between the name that has always been associated with the coffin (Tawahibre, a woman’s name) and a previous translation in 1946 of the hieroglyphic text on the coffin (which identified the name of a male court official, the son of J-se(t)-N-Ese). There has also been some confusion about the remains once housed in the coffin, which were previously identified as male, but in a 1975 autopsy the remains were confirmed as belonging to a female in her mid-30s. A bit confusing, but hopefully we’re now one step closer to getting this all straightened out!

 

Ungluing, re-gluing and filling the jar.

Statuette of an egyptian potter at work (beginning of  the 2nd mill. B.C).

Statuette of an egyptian potter at work (beginning of the 2nd mill. B.C).

The next step for the Egyptian jar was un-gluing all the fragments …to glue them together again.

We had two different cases: fragments that remained adhered together and fragments that were already separated, bearing remains of an old adhesive on their edges. The old adhesive had to be removed since it had many negative issues. First, it prevented the fragments from being joined back together by creating an unnecessary thickness at their junction. Moreover, when reconstructing the ceramic, the old adhesive prevents the fragments from fitting together well.

This old adhesive had a light brown color and after a few tests, it was found to swell when warm water was applied on it.

Here is what it looked like:

Detail of the break edge of one of the fragments, after applying water steam.

Detail of the break edge of one of the fragments, after applying water steam.

To remove the adhesive from the break edges, we used a Preservation Pencil, a tool looking like a pen and emitting water steam. Once softened, the adhesive was very easily removed with a scalpel or a brush.

And here is the result :

The same fragment edge after the removal of the old adhesive.

The same fragment edge after the removal of the old adhesive.

For the fragments still adhered together, it was a little more difficult since the water had to penetrate inside the jar but not too much because of the water-soluble ink on the surface. Compresses, or poultices, of water were applied on the interior of the ceramic, to cover the breaks. Most of the fragments fell apart quite quickly contrary to areas where the jar was very thick.

Now the building could begin ! … well almost since it was necessary to plan precisely how to proceed and in which order to arrange the fragments. First, we had to find where each of them was going, to estimate the losses. For that purpose every fragment was given a number and they were located on a map so as to keep track of their location.

The map; the numbers were indicated on the fragments with blue scotch tape.

The map; the numbers were indicated on the fragments with blue scotch tape.

Then the gluing really began, using the conservator’s favorite adhesive: Paraloid B72, diluted in acetone.

First steps of the gluing.

First steps of the gluing.

The more the jar grew, the more it needed a support, first on the outside, since its bottom is rounded….

A good support was provided by this bucket filled with glass balloons, heavy enough to stabilize the jar.

A good support was provided by this bucket filled with glass balloons, heavy enough to stabilize the jar.

…then on the inside to prevent it from collapsing on itself because of some particularly heavy fragments.

The jar was growing and needed internal support; the white material inside is a plastic bag filled with polyethylene fiber.

The jar was growing and needed internal support; the white material inside is a plastic bag filled with polyethylene fiber.

Losses in the ceramic had to be filled at the same time as the gluing to provide structural support to the jar and prevent it from collapsing.  Moreover those areas to fill would have been difficult to reach once the gluing was complete.

There was one large loss that definitely needed to be filled since one of the surrounding fragments was holding by only a few millimeters to another one.

Filling this area was a bit tricky. The fill material needed a support to be applied on the jar. Japanese tissue paper was glued inside of it and strengthened by applying several layers of Paraloid B72. It also needed to be shaped according to the curve of the jar.

On the left: The area to be filled.                        On the right: Japanese tissue paper used as a support to hold the fill material.

On the left: The area to be filled.  On the right: Japanese tissue paper used as a  support to hold the fill material.

On the left: The inside of the jar with the "tricky fragment" held in place by the japanese tissue paper.   On the right: Applying the fill material.

On the left: The inside of the jar with the “tricky fragment” held in place by the japanese tissue paper. On the right: Applying the fill material.

The fill material used is a mixture of Paraloid B72 and glass micro-balloons, looking like a very light white powder; plaster is also traditionally used to fill losses, but glass micro-balloons are lighter and don’t bring any salts to the ceramic. That kind of fill is also reversible and completely neutral towards the ceramic.

Here’s the fill once finished and polished with a heat spatula, ready to be painted.

The fill almost finished: the building can go on.

The fill almost finished: the building can go on.

More fills and building to come in a next post !

Slowly, but surely

Sometimes when working on a large, complex project, it can be hard to see progress – once certain areas are addressed/stabilized I just start focusing on all of the other problems. In these cases, I find it really helpful to write about the work, to go through the photos I’ve taken so far, and to reflect on how far we’ve come. One of the more complex treatments we’re working on in the Artifact Lab is Tawahibre’s coffin.

The last time you saw Tawahibre on the blog, she was all tied up, Lilliputian-style.

Tawahibre capturedSince that last post, we actually have made quite a bit of progress, and have started realigning and filling areas where the gesso and smaller wood components have cracked and separated from the wood ground below.

One very precarious area has been a large section on the lower proper left side of the coffin – when the coffin came into the lab for treatment, this section was only just barely attached along the top, with the help of two wooden dowels as well. In addition to being just about ready to detach, this section was also very distorted and misaligned, with areas of the painted surface overlapping and abrading each other.

tawahibre PL detail BT with arrows

Before treatment detail of this large partially detached section. It was just barely attached along the top (indicated by red arrows) and by 2 wooden dowels (circled in green).

Here is a view of this section, before treatment, from above (the red arrows are just pointing out the area that I’m talking about, for clarity).

tawahibre PL detail overhead BT with arrowsAfter working to humidify and realign this area as much as possible, I prepared it for filling and stabilizing by lining the wood support below and the inside surface of the detached section as possible with Japanese tissue paper, adhered with methyl cellulose adhesive. The Japanese tissue paper will serve to make these fills more easily reversible in the future.

Tawahibre PL detail DT with arrows

Preparing this section for stabilization and filling. The red arrows are indicating the Japanese tissue paper used to line the inner surfaces of the coffin before filling.

To secure this section to the rest of the coffin, I applied a fill mixture between the large partially detached section and the wood support below. The fill mixture was made using 5% methyl cellulose adhesive in 1:1 water/ethanol bulked with a 1:1 ratio of alpha cellulose and 3M glass microballoons. The alpha cellulose and microballoons were chosen to create a lightweight, relatively dry, and easily moldable fill – they also make this mixture a bright white color. After applying the fill material, this section was again bound with the twill tape and ethafoam blocks to hold everything together while the fill dried.

Detail of this section after filling. Note-no straps are needed to hold it in place!!!

Detail of this section after filling.

And here is a detail showing this section from above – I think it makes a nice comparison with the before treatment shot from a similar angle, above.

Tawahibre PL detail above DT2So far this has been a successful course of treatment and we have filled several areas on the coffin. Our current goal is to get the lid stabilized enough so that we can separate it from the base, so that we can continue to work on both sections with better access to some of the very unstable, fragile areas.

Special thanks to my conservation colleagues for their help with brainstorming, problem-solving, and carrying out this treatment!