Te Papa's Blog

The varnish removal of Poedua is progressing slowly and painstakingly and we are now almost halfway through this important part of the treatment.  As we have mentioned before, removal of discoloured varnish can have a dramatic effect on the overall balance, colour and depth of a painting.  Because discoloured varnishes like Poedua’s are usually a yellow-brown colour, they tend to reinforce warm tones but counteract the cooler colours such as some greens, whites and cooler blues.

This image shows Poedua's face and the flowers pre treatment - before surface cleaning and varnish removal. Photograph taken by Michael Hall, 2011 © Te Papa

Every week we are recovering more of the artist’s original paint from beneath the discoloured varnish and overpaints and we have made some nice discoveries.  Most recently, we have revealed the beautiful colours and brush-stroke detail of the flowers in Poedua’s hair. 

In this image after cleaning, you can now see much more detail in the brush-work of the artist and appreciate all the colours of his palette. Photograph by Katherine Campbell, 2011 © Te Papa

Before we cleaned them, the flowers were quite dark and a little cloudy, making them a bit difficult  to interpret.  Now that the surface dirt and varnish has been removed you can see the lovely detail of the brush-work and full range of colours employed by Webber.

This image shows the difference in appearance after cleaning of Poedua's hair, face and the flower. Photograph by Katherine Campbell, 2011 © Te Papa

In the case of Poedua, we want to remove the varnish layer because it is very discoloured and has been applied extremely unevenly, which has become more apparent as the discolouration has progressed.  Similar to the effects of surface dirt, the discoloured varnish present has diminished the tonal ranges of the painting.  

Under UV the uneven application of the varnish becomes really obvious. The bright green fluorescence indicates a natural resin varnish, 2011. Photograph by Mel Carlisle © Te Papa.

The first step is to characterise the varnish and work out how many layers are present.  There are several techniques available to us including, exposing the painting to ultraviolet light and taking cross-sections of the paint and varnish layers.  Looking at the painting under UV lamps causes specific auto-fluorescence in different materials which helps us with identification.  The varnish layer on Poedua fluoresced a bright green colour, indicating a natural resin such as dammar or mastic.  This wasn’t surprising to us since both are common varnishes that we might expect to find on a work such as this. 

Cross-section analysis involves the sampling of a microscopic piece of the painting, usually taken from an area of existing damage.  This sample is set in a polyester resin and ground down with fine sand paper.  The cross-section is then examined using a compound microscope where we can see the building up of paint layers by the artist, surface coatings and later additions.  From the cross-section we took from Poedua we can see that there is only one varnish layer apparent.

Once the cross-section has been prepared, we look at it under magnification and also UV to see the build-up of layers used by the artist, 2011. Photograph by Katherine Campbell © Te Papa.

 

This a cross-section taken from a Don Driver painting in Te Papa's collection. 1980-0013-1; Zither; Don Driver. Under magnification and UV it's possible to see several layers of paint and 2 layers that fluoresce with the appearance of varnish. The cross-section was photographed by Sarah Hillary at the Auckland Art Gallery, 2010 © Te Papa.

Varnish layers become increasingly difficult to remove the older they are.  This is due to alterations in their chemical structure that occur over time.  The solvents that conservators use to remove varnishes are carefully tested in a similar manner to the surface cleaning testing.  We use a variety of solvent mixtures and different preparations and combinations including enzymes, soaps and gels.  Their effectiveness builds further on our knowledge about the varnish layer we are dealing with.  Solvents work by penetrating the varnish layer and swelling the film, bringing the resin into solution and allowing it to be removed from the surface.  After initial testing, we select the most appropriate solvent or solvent mixture and test this in the various areas of the composition.   Constant care and attention is crucial during this process because the varnish layer is usually the last layer before the paint film is encountered.

Initial testing is done under magnification so we can closely monitor the results. As with the procedure for surface cleaning swabs are kept along with notes of the results for comparison, 2011. Photograph by Mel Carlisle © Te Papa.

Once the pale paints have been tested we move onto all the major areas of colour. The two circular areas that appear dry and matt are the patches where the varnish has been removed and the paint layer is exposed. At the end of the treatment a new varnish layer will be applied which will resaturate and intensify the paint surface again, 2011. Photograph by Katherine Campbell © Te Papa.

Poedua is an oil painting on stretched canvas, meaning the composition is painted on a fabric support which is stretched to keep it evenly tensioned, over an expandable wooden stretcher.  Canvas supports began to replace traditional wooden panel supports from about the 16^th century and were the most commonly used type of support by the 18^th century. 

In this photo you can see the back of the canvas support and the wooden stretcher frame which keeps the canvas taut, 2011. Photograph by Michael Hall. MA_I.227733 © Te Papa

The canvas used for this painting is probably linen (untested) and it has a twill weave, which creates a distinct texture which is visible through some areas of the paint layer.  Linen is a common support for paintings of this age, but other fabrics are also used as supports for paintings, including hemp and cotton.  The canvas is attached with small metal tacks around the sides of the stretcher bars.  We think there’s a good chance that this is probably the original attachment because there are no old holes in either the canvas or the stretcher, which can indicate that the canvas has been taken off in the past.

Here you can see a profile view of the painting showing a metal tack attaching the canvas to the stretcher, 2011. Photograph by Michael Hall. MA_I.227731 © Te Papa

The back of the canvas is no longer looking as it would have when Webber first painted it!  The canvas is made of a natural fibre which has darkened and discoloured because of the process of ageing and deterioration.  We have found the canvas to be in quite a good condition, given its age.  There are a few areas we will need to address in our treatment, including some small tears and embrittlement around the edges.  This has occurred because over time the fibres have become inelastic and snap rather than stretch.  There are some areas of damage which we can’t determine the extent of just yet because they have been covered with patches on the back.

Detail of one of the old patches on the back of the canvas. At this stage we are unsure what the patch is covering, 2011. Photograph by Michael Hall. MA_I.227733 © Te Papa

The back of the canvas is also dirty from being exposed to the world for the past 226 years!  There is a layer of dirt and some staining present, which are a combination of age-soiling and various deposits and residues.  We hope to be able to reduce some of this soiling with cleaning, but this action probably won’t dramatically change the way the canvas currently appears.  What we hope to achieve is a reduction of the dirt matter, which if left, can accelerate deterioration.

An interesting discovery for us is that Poedua has never been lined.  Lining is the attachment of a new support to the entire reverse of a painting on canvas.  Its aim is to give additional support and reinforcement to a weakened or damaged canvas.  It was a very common restoration procedure in the past and most paintings of this age have been lined at some point.  The fact that the canvas has remained unlined means that we can see the back of the original canvas and a lovely inscription, presumably by the artist himself. 

Detail of the handwritten inscription on the back of the canvas. The inscription includes information such as the subject's name, her father’s name and the location of their village. Some of the inscription is unclear and we are hopeful that cleaning will reveal it more, 2011. Photograph by Michael Hall. MA_I.227733 © Te Papa