Author Archives: merlyngriffiths

‘Priest’ – Part 3: The Conservation Treatment Process

An Introduction

Before reading this post, I must point out that Priest is unlike any other, and is very different to the stained glass panels typically brought to a conservator for treatment. In conservation, every window is assessed individually, and rather than there being a universal ‘one size fits all’ approach, they are treated according to their own specific condition. When Priest arrived in the studio, it was in a very poor condition. It was very unstable, and without treatment, would have been unlikely to last much longer in its existing state. Indeed, pieces of glass had already long-since detached from the plaster support structure, which itself was cracking and had slowly buckled. The instability of Priest meant that as soon as conservation treatment started, it was obvious that this was a panel that would require more than the ideal approach of ‘minimum intervention’, in order to prolong its life into the future. Therefore, the treatment that Priest received is as individual as the panel itself, and is far from the typical approach to the conservation of stained glass. This is something that must be borne in mind whilst reading this post.

Removing the Plate and Cleaning the Glass

The first step in the process of conserving Priest, was to make a cradle that would support the panel, so that it could be laid flat. This cradle was propped up against the inside of the panel, and the whole thing was swiftly – but gently – laid down, so that the exterior faced upwards, in order to access the broken glass backing plate. The nails holding the plate in position were bent so that the two parts of the plate could be removed from the panel.

Detail of the exterior of the panel after the removal of one part of the glass backing plate - Exterior - 'Priest'.

Detail of the exterior of the panel after the removal of one part of the glass backing plate – Exterior Side – ‘Priest’.

Once on the bench, and ensuring that the fragile painted design on the glass was not disturbed, the thick, sticky dust was removed from the surface using a mixture of ethanol and de-ionised water, applied on cotton wool swabs. These were rolled across the surface, ensuring that they did not touch the paint. A gap of a few millimeters was left all around the boundary of the friable paint lines to make sure that it did not come into contact with the cleaning solution. This was all done whilst looking through a microscope, which ensured maximum precision. The edges of the glass along the break in the plate were cleaned thoroughly, again being mindful of the paint on one side, and the two pieces were adhered back together with silicone. This adhesive has a certain degree of flex to it, and so was suitable here due to the scale of the glass sheet and the length of the crack within it – anything more solid – such as an epoxy resin – would have created too much tension in the glass.


Nail bent to facilitate the removal of the glass backing plate – Exterior side – ‘Priest’.

Where some pieces of coloured glass had remained within the plaster matrix, those that had fallen to the bottom of the panel and become wedged between it and the blacking plate were photographed, and then removed into a separate tray. This left the plaster visible, and it was possible to see the many empty gaps where glass should be. The plaster was given a light clean using a soft brush and a low suction vacuum, which had a pair of tights wrapped around the nozzle in order to catch all of the dirt and any loose plaster that came up with the suction. Following this, each of the individual coloured glass pieces was then cleaned with a mixture of ethanol and de-ionised water, on cotton wool swabs. This removed the build up of dust and grease, and left them clean and ready to be re-attached to the plaster. The same was done with the pieces of glass that remained within the plaster matrix.


After the coloured glass pieces had been returned to their original positions – Exterior Side – ‘Priest’.

Detail of the reinstated glass - Exterior Side - 'Priest'.

Detail of the reinstated glass – Exterior Side – ‘Priest’.

Consolidating the Plaster Matrix

Now that the back of the panel had been exposed from underneath the backing plate, it became apparent that the plaster was in worse condition than originally thought. Without the obstruction of the plate, it was possible to see the extent of the cracking, and the relatively large widths that some of these cracks had reached. Before the pieces of coloured glass were reinstated, it was first necessary to fill these cracks with an adhesive. Paraloid B72 was chosen, and was applied to the cracks with a scalpel, which trickled into them in a capillary action. For those areas with wider gaps, the Paraloid was mixed with glass microballoons, and this provided a filler that would ensure that the plaster was bonded together and remained stable.

Consolidating cracks in the plaster - Exterior - 'Priest'.

Consolidating cracks in the plaster – Exterior Side – ‘Priest’.

Reattaching the Glass Pieces

After assessing the glass at close proximity, without the backing plate acting as a barrier – it was evident that the glass would need more to secure it than previously intended. The pieces of glass that were still attached to the plaster matrix were generally quite loose, and were likely to fall away from the panel in the near future without a form of barrier to catch them. To solve this problem, we began to look towards the use of fabric as a means of securing the glass, and ensuring that even if it did come loose, it would not fall away from the panel. The stained glass conservators at the Burrell Collection in Glasgow had performed some valuable tests that focused on the use of fabrics as an alternative method of consolidating fractured glass, and so inspiration was taken from here for this project. Information about their investigations can be found on their blog here:

We determined that a fabric called polyester stabiltex would be used in one piece across the entire reverse of Priest, to act as a form of ‘safety net’ for the glass. The weave of this fabric is visible from the reverse of the panel, and if it is viewed at close proximity from the interior, but bends in well enough not to be distracting, especially in areas where there are multiple pieces of glass layered within the plaster matrix. Trials were undertaken of the performance two potential adhesives, in which to attach the fabric. These were Klucel G – a wheat starch paste – and Paraloid B72 – a thermoplastic resin. Though Klucel G adhered well to the glass, and left very minimal visual impact, it did not adhere to the plaster, which was a big problem for this panel. Therefore, though the Paraloid B72 did leave a slight white cast in some areas on the surface of the glass, it adhered well to the plaster (as we have already found out) and from tests carried out it was confirmed that it could be relied upon to remain stable over time. Similarly, another test was undertaken to see how plaster responded to being submerged in the liquids necessary to remove Paraloid B72 and Klucel G. Fortunately, it was acetone – that needed to remove the Paraloid B72 – that caused the least amount of changes in the plaster. Water was the other option – which would remove the Klucel G, however this slightly softened the plaster after an hour. These tests were not exactly representative of the actual fabric removal process, as the panel would not need to be soaked with liquid in order to remove the adhered fabric, rather than lightly brushed with a dampened brush. Regardless of this different, it was informative to see the results, and beneficial to see that the removal of Paraloid B72 would not cause substantial damage to the plaster.

Before applying the polyester stabiltex, the coloured glass pieces of glass were first adhered to the plaster matrix with small dabs of Paraloid B72. Following this, the fabric was laid over the reverse of the panel and the adhesive was applied with a paintbrush.

Applying the Paraloid B72 to the polyester stabiltex - Exterior Side - 'Priest'.

Applying the Paraloid B72 to the polyester stabiltex – Exterior Side – ‘Priest’.

Polyester stabiltex on the exterior side of the panel - Exterior Side - 'Priest'.

Polyester stabiltex on the exterior side of the panel – Exterior Side – ‘Priest’.

Returning the Backing Plate

The final task before the panel could be stood back upright was to reattach the backing plate. It became apparent that the old nails would not provide enough support to ensure that the glass plate would stay in position in the long-term, and so another solution had to be sought. Similarly, the plaster structure itself was not directly fitted into the outer wooden frame, but was actually placed inside a smaller, inner frame, which itself was only attached to the main frame by small sections of plaster, of which some had broken. Therefore, the inherent fragility of the plaster was made more precarious by the instability of the framing system. To solve this dilemma, a more invasive approach that originally desired had to be implemented. This was justified by the fact that without it, the panel’s lifespan would be drastically reduced, and also that despite the necessity for new screw holes having to be drilled into the wooden frame; these holes would be the only direct impact on the original fabric. These holes were also much more desirable than the irretrievable loss from collapse that the panel could face without the stabilisation treatment. The frame was secured by creating a small number of brackets, that were made from manganese bronze, bent into an ‘L’ shape. Each of the two sides had a hole punched through it, and then these were screwed into the wood – one side on the outer frame, and one on the interior frame, so that the inner frame was secured and no longer moved within the outer frame, jeopardising the panel.


Detail of the reverse of ‘Priest’ before the new black card strips were added. Note the brackets holding the inner and outer wooden frames together, and also those holding the backing plate in position – Exterior Side – ‘Priest’.

In order to secure the glass backing plate, the same kind of brackets were created, but this time, only one side was drilled with a hole, and the other was surrounded by plastazote – an inert foam used frequently in conservation as a packing and storage material. One end of the brackets were screwed into the frame, and the plastazote end rested against the glass backing plate – sufficiently enough to support it from loss or movement, but gently enough not to put any pressure on the glass. The brackets were also positioned to avoid contact with any of the fragile paint on the plate.


Bespoke bracket holding the glass backing plate in position – Exterior Side – ‘Priest’.


The Frame

The frame was then cleaned with a smoke sponge, which is a form of cleaning apparatus used in conservation – notably that of fire damaged objects. This is made from vulcanized rubber that absorbs dust as it is gently drawn across the surface of an object. The beading on the front of the frame that had come loose was reattached using Paraloid B72, as fixing it back in position with nails would cause too much stress to the panel, due to the requirement of using a hammer to insert them.

The final task for the exterior side of the panel was to remove the black paper around the perimeter. This had acted as a barrier to stop light coming through around the plaster. Over time some of this had been lost, and the remaining paper had weakened and become thin and liable to tearing. The decision was made for this to be removed, and retained for archiving, and a new, sturdier black card reinstated in its place, and in that of the missing card.

Conserving the Interior face of the panel

Due to the panel having been laid flat and then erected back into a vertical position, some pieces of loose plaster inevitably fell away from the interior side of the panel. Most of these areas were reattached using the same adhesive, as before, however for a small number of tiny pieces that fell away, their positions remained unidentified. It was not felt that there was justification to attach them ‘at random’ and so they have been retained for archiving. The repair around the nose was one of the pieces that had detached – but due to the crumbling nature of the repair material, it was not possible to reinstate it to the panel. Again, this has been retained for archiving, and the glass is the nose has been positioned to fit without the later repair addition. Cracks in the plaster were consolidated, and the visible areas of coloured glass cleaned from this side.

Old repair to the nose of the figure, now archived - Interior Side - 'Priest'.

Old repair to the nose of the figure, now archived – Interior Side – ‘Priest’.

Detail of the nose after conservation. Note the original plaster revealed after the archiving of the repair compound - Interior Side - 'Priest'.

Detail of the nose after conservation. Note the original plaster revealed after the archiving of the repair compound – Interior Side – ‘Priest’.

In Conclusion

The conservation of Priest was not without it’s challenges. The method of its original manufacture suggests that it was purely experimental, and was unlikely to have ever been built to last. Certainly, it has become weaker over time, and without treatment, would have fallen into irretrievable disrepair in the not too distant future. Though the conservation treatment carried out was relatively invasive, it has ensured that the panel will remain in a stable condition for a much longer period of time, and that it is safe to display and be enjoyed by visitors to the Stained Glass Museum. This panel is particularly beautiful, and now conserved, its true beauty can be seen for the first time in over half a century, when it was last put on display. This is a brilliant example of an art student’s exploration of materials and his formation of ideas in 3D, and provides an excellent opportunity to see an early piece of work by artist, Geoffrey Clarke. The exhibition in Ely is a fitting way to remember the artist, whilst appreciating his talent and creativity, and it has been a pleasure to preserve one of his most fragile works for you to enjoy.

Before and After Conservation Photographs

'Priest' - Interior Side - Before Conservation

‘Priest’ (1949) – Interior Side – Before Conservation

Priest AC_0005

‘Priest’ (1949) – Interior Side – After Conservation

'Priest' - Exterior Side - Before Conservation

‘Priest’ (1949) – Exterior Side – Before Conservation

Priest AC back_0019

‘Priest’ (1949) – Exterior Side – After Conservation


‘Priest’ – Part 2: The Conservation Treatment Proposal.

Part 1 of this post contained historical information about Priest, an introduction to the materials from which it was made, and an assessment of its current condition. The following post will discuss the proposals for the conservation treatment of this panel.

'Priest' (1949) - Interior view

‘Priest’ (1949) – Interior view

The main concern with the Priest is its instability. The use of two somewhat incompatible materials – glass and plaster – has meant that over time, the plaster support has bowed outwards and the glass has become detached from it, which has compromised both the integral stability of the panel and also its aesthetic appearance.

The priority for the treatment of this panel is to ensure that it will not collapse any further. This will necessitate the removal of the painted backing plate, and whilst this is separate from the panel, there will be the opportunity to reinstate the pieces of glass that have fallen away from the plaster, returning the panel as close as possible to its original intended appearance.



Painted Backing Plate

The backing plate will need to be removed in order to access the glass and plaster at the back of the panel. This will mean that the panel needs to be laid on its front in a supportive cradle. Once this has been achieved, the nails and wood holding the plate in place will require documentation and then removal, and then the plate can be taken out.

The plate itself is coated with a thin layer of grease and dirt, and so this will need to be removed. It is likely that this will require a ‘wet’ clean – which will involve the use of de-ionised water and ethanol in order to cut through the grease, and this will need to be done on both sides of the glass. The paint used on the outside of the plate will have to be avoided as it has not been fired onto the glass, and so a microscope will need to be used to ensure that no liquid comes into contact with it, otherwise the pigment will dissolve.

The plate is broken into two pieces, and the use of silicone is recommended to adhere these back together. The edges of the glass will requite cleaning with acetone before any adhesive is applied, and this will need to be done very carefully to avoid the paint.

Detail of the break in the glass backing plate (exterior)

Detail of the break in the glass backing plate (exterior)

Glass Pieces

While the backing plate is away from the frame, the pieces of glass that have fallen from the plaster support can be collected, documented, cleaned, and their original positions identified. In the areas where adhesive has previously been used, the outline of the piece of glass formerly attached has left an impression in the adhesive, and this could be used to help identify which piece went where.

Evidence of adhesive on the glass. This can be used to determine where glass pieces have fallen from.

Evidence of adhesive on the glass. This can be used to determine where glass pieces have fallen from.

The glass that is still secure within the plaster can be cleaned with a soft brush to remove any dust. That which is separate from the plaster can be given a wet clean before it is reinstated, to remove any dust, dirt, and adhesive residue.

At this point it is suggested that Paraloid B72 is used to adhere the glass back onto the plaster. This is recommended as it is reversible, and though a ‘wet’ mixture, the acetone within the mixture evaporates at such a rate that it would not ‘sit’ on the surface of the plaster for long enough to cause damage.

Whilst the backing plate is separated from the panel and the back of the panel is exposed, the stability of the glass can be assessed, to ensure that no other pieces are likely to fall from their positions in future. It is not advised that the backing plate be removed more times than absolutely necessary, and so this is a good opportunity to undertake some preventative conservation.


The plaster shall undergo a ‘dry clean’, in order to remove any dust and cobwebs that have formed on the surface during the panel’s years in storage. This will involve using a soft brush, lightly dusted across the surface in order to clean it. It is referred to as a ‘dry’ clean as there is no liquid involved, and due to the nature of plaster, it is preferred that liquid is avoided unless it is necessary.

The cracks in the plaster can be consolidated with Paraloid B72 in order to stabilise the network, as this is integral to holding the glass in place. Though it would be possible to use new plaster to consolidate the cracks in the original plaster, there are numerous problems with this approach. A new batch may not adhere to the older plaster, and there is a possibility that they may separate and fall away at some point in the future. Similarly, in adhering new to old, the moisture in the new mixture would compromise the stability of the older material. As discussed above, the Paraloid B72/ acetone mix will ensure that liquid does not sit on the surface of the plaster for a prolonged period of time.

Cracking in the plaster (exterior)

Cracking in the plaster (interior)



The wooden frame is generally sound, however in areas, parts of the beading has come loose and shifted its position, and some nails are missing. In this instance it would be possible to re-set this by using a new nail in the old nail-hole, and this would ensure that the frame is able to retain its stability and hold the panel in place.

The frame can be cleaned using a soft bush and a smoke sponge. The latter is a vulcanised rubber sponge that is swept gently across a surface and absorbs any dust and grime. This is to be used after a soft brush has removed any harder deposits which may otherwise scratch the wood when a smoke sponge is moved across it.

Cobwebs inside the wooden frame.

Cobwebs inside the wooden frame.


The paint is of a powdery consistency and is not solidly adhered to the glass, which may account for the area of loss. Therefore when being handled in the future, extra care must be taken, ensuring that nothing touches the plate or the paint. Though a ‘ghost’ of the paint is visible, it is not so definite that paint could be applied with confidence, and any moisture on the original cold paint would harm it, and so the painted areas of the backing paint shall be left untreated.

Additional Materials

The removal of the exterior backing plate is necessary to reinstate the missing glass, which would include removing the nails, wood and plaster holding it in place, and manoeuvring it around the black paper attached to the wooden frame. Anything taken away in order to remove this plate would be catalogued, and reinstated in the same position when the plate is put back into place.

There are some areas where the black paper used to block out the light between the panel and the frame is missing. In those areas, new black paper shall be added to the frame to maintain this function.

There is already substantial cracking of the plaster that was designed to hold the plate in place, and in some areas bits have become loose and fallen away. Therefore when these areas are put back, and after the plate is replaced, their reinforcement with Paraloid B72 will not only hold them back in their original positions but also ensure their strength into future.


Broken segments of plaster used to hold the backing plate in position.

To Be Continued…

Part three of this blog post will concentrate on the conservation process itself, and will guide through the decision making processes and chosen treatment methods used to conserve Priest.

 Merlyn Griffiths



‘Priest’ – Part 1: An evaluation of the materials used in the construction of a Geoffrey Clarke panel, and an analysis of their condition.

History and Description

'Priest' (1949) - Interior view

‘Priest’ (1949) – Interior view

Of all the panels, Priest is certainly the most fragile. Like St. Sebastian, this panel was created in 1949 during Clarke’s time as a student at the Royal College of Art in London. It is very experimental by nature, and utilises a number of different materials in its construction, including wood, plaster and glass.

The panel shows the bust of a figure against a darker background. It may be assumed from the name of the piece that this figure is a Priest, and his hands look as though they are about to join in a praying gesture. His hair is blonde and chin length, and his eyes, which face directly towards the viewer, are a piercing blue, and appear to follow you around the room. The panel is set into a wooden frame, and rather than being a stained glass window in the traditional sense, in this panel the lead has been substituted for plaster, and the intervening spaces filled with mosaic pieces of glass.

'Priest' (1949) - External view

‘Priest’ (1949) – External view

The process in which this panel was made began with the painting of a rough design on a piece of clear glass with black paint. This was then turned over, and plaster was applied onto the surface of the glass, following the lines of the painted design. Pieces of coloured glass were then set into this plaster while it was still wet, and more plaster was built up around it, a process that was repeated in areas so numerous layers of glass were embedded in one place. Once dry, the an additional layer of mortar was applied from the front, and was painted a dark grey, so there was more contrast between the glass and the ‘lead’ substitute. This process has created a form of mosaic, similar to dalle de verre windows, but on a much smaller scale. Dalle de verre is the technique of setting large blocks of coloured glass into a concrete surround, a process which was in use around the time Clarke was making Priest, though he apparently never did make large scale versions of this method using concrete himself.

The Condition of the Panel

The interior of the panel viewed from an angle, note the 3D plaster

The interior of the panel viewed from an angle, note the 3D plaster

There are a number of problems with this panel and it is incredibly unstable, due to the nature of the materials used and the manner in which they have been applied.

In order to formulate a conservation proposal, it is necessary to first assess the condition of the panel, and to identify where the issues lie. The rest of this blog post will be divided into sections that describe each of the materials used, and discuss their condition. A condition report can also help identify if there have been any previous attempts to repair or restore a panel, which can be valuable when there is no written documentation about this available.


There are two types of glass used within Priest. The first is the backing plate, measuring 545mm (h) x 390mm (w) x 2mm (d). This is made of horticultural glass, the kind you may recognise from greenhouses. This is even in texture, and clear with a slight tint of green. This is why you may hear clear glass referred to as ‘tint’ glass – if you look at it from the side, you will see that there is a green hue to the material, though it may appear perfectly clear from the front.


  • Dusty inside and out
  • A large semi-circular crack is in the lower left hand corner (when viewed from the reverse of the panel).
Detail of the break in the glass backing plate (exterior)

Detail of the break in the glass backing plate (exterior of panel)

The second type of glass used in this panel is the various coloured pieces set into the plaster. The glass is of a variety of different colours, with purple, yellow, blue, red, turquoise and rose among the most prominent. These pieces of glass are relatively thin, and appear to be machine-made, due to their very flat surfaces and even thickness. There are a few pieces which are what we would call ‘antique’ glass. Later in his career, Clarke was fond of using this traditionally made ‘mouth blown’ glass from the Hartley Wood and Co. glasshouse in Sunderland. Mouth blown, or antique glass is still made today, and is often favoured during the conservation work of medieval glass, as it can be made to match the original very well. This can be identified as having more character than machine-made glass, with more irregularities in its thickness and discrepancies in its texture.

Interior of the panel, areas where the glass has fallen away from the plaster are illuminated by the light shining through from behind

Areas where the glass has fallen away from the plaster are illuminated by the light shining through from behind (interior of panel)


  • Some pieces of glass have become detached from the plaster and have fallen down and become lodged between the plaster and the backing plate, leaving holes in the design.
  • Some more pieces of glass have become loose, and can be seen hanging from their plaster surround.
  • Dusty inside and out.
A piece of glass barely attached to the plaster (exterior)

A piece of glass barely attached to the plaster (exterior of panel)



The plaster used in this panel is a light cream colour when viewed from the reverse. From the front, this has been ‘painted’ with a viscous concrete or plaster mixture to darken it. There are some areas where a lighter grey and a more yellow-toned substance have been used to stabilise the plaster network, which shows that there have been at least two separate attempts to repair the plasterwork in the past, judging by the different approaches to the choice of repair material.

Cracking in the plaster (exterior)

Cracking in the plaster (exterior of panel)


  • The plaster is cracked, especially around the nose of the figure.
  • On the left hand side of the face, some of the plaster has fallen away.
  • Dusty inside and out.
  • The yellow-toned material added around the nose area has lost its adhesion to the original plaster, and is visibly loose.
  • From the reverse, where pieces of glass have fallen away, there are remnants of an adhesive, which have been applied to hold the glass in place, but have failed. This may be evidence of a repair made after the piece was constructed.
  • Around the backing plate, sections of plaster have crumbled and become loose.
  • The whole of the plaster structure is bowing out towards the front – away from the backing plate that it was originally stuck to – due to its weight.
Evidence of earlier repair to the plaster (interior)

The areas surrounding the nose is evidence of an earlier repair to the plaster, and has become loose (interior of panel)

Remains of silicone on the glass and plaster (exterior of panel) - also note the dust on the surface of the glass

Remains of adhesive on the glass and plaster (exterior of panel) – also note the dust on the surface of the glass


The frame surrounding the panel is made from wood painted a dark grey shade, and comprises of a thicker surround with a thinner wooden beading inside – which is visible from the front. Only visible from the reverse of the panel is a smaller frame that the backing plate is attached to with plaster.


  • Generally in good condition.
  • Inner beading has become loose in areas.
  • Cobwebs and dust over reverse of the frame.
  • When stood on lower edge, the panel does rock slightly.
  • Some of the frame is scuffed along the front.


The black paint lines on the exterior of the backing plate are applied in relatively thick strokes, but in a somewhat ‘sketchy’ manner. This paint is referred to as a ’cold paint’ as it has not been fired to the glass, as common in more traditional painting methods used in stained glass.

'Priest’ (exterior of panel) from an angle; note the area where the paint has been lost, in contrast to the more defined black paint

‘Priest’ (exterior of panel) from an angle; note the central area where the paint has been lost, in contrast to the more defined black paint surrounding it


The consistency of the paint is powdery and therefore is unstable.

There is an area where a proportion of the paint has been lost, though a ‘ghost’ or trace of the paint is still visible in certain light.

Additional Materials

The backing plate is held in position with nails, packed underneath with small off-cuts of an engineered wood. On the reverse of the panel, thin black card or sugar paper was originally attached in strips to each side of the frame with nails. This was done to block out the light when viewed from the front. An eyelet is screwed into each side of the frame and a length of picture framing string is attached.


  • The string has weakened with age and so could no longer be relied upon to hang the heavy panel.
  • All of the nails are very rusty.
  • Some of the black paper is missing from the reverse of the panel and so light is let through, which is disruptive.
  • The pieces of wood used to hold the backing plate in place are dusty and loose.
Priest 11

Wood and nails used to help secure the glass backing plate in position (exterior of panel)

To Be Continued…

Part two of this blog post will discuss the proposal made for the conservation of this panel. It will take each of the concerns highlighted in this condition report and propose a solution. The main aim of this proposal will be to stabilise the panel, as its fragility is currently a concern, and will limit it’s longevity, options for display and for its storage if left in its current state.

 Merlyn Griffiths