Volume 17, Number 1, Jan 1995, p.28

Conference Reviews

Varnishes: Authenticity and Permanence

September 19-20, 1994, Ottawa
Part I

Perhaps the most significant event for painting conservators this year was the colloquium and workshop "Varnishes: Authenticity and Permanence", sponsored by the Canadian Conservation Institute (CCI) and Canadian Heritage. The colloquium was held on Monday and Tuesday (19-20 September 1994) and a workshop with limited enrollment was held Wednesday and Thursday. The colloquium was attended by 160 people from 17 countries from four continents.

The colloquium and workshop were organized by Dr. Leslie Carlyle and James Bourdeau. The original plan was for a small colloquium -- something of a follow-up to the 1990 Brussels IIC conference. The goals of the conference were stated by Leslie Carlyle as being neither an attempt to reach a consensus nor to weigh in for or against any particular varnish, but rather a systematic examination of assumptions made about varnishes and varnishing.

The title of the colloquium "Varnishes: Authenticity and Permanence" was chosen with the following meanings in mind: authenticity connotes the appearance of varnish; and permanence, the stability of a varnish. Appropriately, the first talks were overviews of varnish research and the physical and chemical nature of varnish.

René de la Rie, Head of Scientific Research, National Gallery of Art, Washington, was the first presenter with a most appropriate "An Overview of Research on Picture Varnishes".

René began his presentation with a slide of a 1754 painting showing a restorer in his studio removing a yellow varnish from a painting. René pointed out that of the traditional materials used by artists, natural resin varnishes are the most unstable. Fundamentally, natural resin varnishes are responsible for the varnish cycle.

A difficulty for researchers trying to investigate the varnish problem is that there is no objective measure for a good looking or good handling varnish. The components of that measure are chemical stability and the physical factors influencing the optical and handling properties of a varnish.

There are a number of ways to classify varnishes:(1) natural versus synthetic; (2) chemically: diterpenoid resins (sandarac, rosin); triterpenoid resins (gum mastic, dammar resin); or drying oil.; (3) by size: high molecular verses low molecular weight resins; or (4) by historical usage: the earliest varnishes were oil varnishes; spirit varnishes have been used since the 16th century; dammar began to be used in the 19th century; poly(vinyl acetates) date from the 1930's; poly(alkylmethacrylate)s,the1940's;and poly(cyclohexanone)s (ketone resins), from the late 1940's.

Some problems associated with early synthetic varnishes were the following: crosslinking (some acrylics); instability (ketone resins); and optical problems (polyvinyl acetates and acrylics). Another problem was the varnishing of paintings which were never intended to be varnished.

Next he discussed the functions of a varnish. A varnish should protect the paint surface, although in René's opinion this function has been overemphasized. Varnishes do protect the paint from mild abrasion and grime accumulation, but do not protect against UV and offer only limited protection against oxidation.

The aesthetic function of varnish is to darken colors, increase saturation, and increase gloss. This is caused by the physical leveling of the surface of the painting by the varnish and by the higher refractive index of varnish than that of air. For these reasons it is a mistake to varnish a painting which was not intended to be varnished.

Dr. de la Rie went on to discuss the optical problem of a paint surface. White incident light interacts with the painting surface. An unvarnished paint surface scatters a portion of the white light and the rest enters the paint layer where some colors are absorbed and the rest is reflected back out as colored light. We see both the scattered white light and the colored component. Light interacts differently with a varnished paint surface. Some of the white light is reflected as specular reflection (glare) and the rest enters the paint layer where, again, some colors are absorbed and the rest is reflected out as colored light, however there is no scattered component. The varnished picture appears richer and more saturated.

The air-paint and air-varnish interface was also discussed. An interface between two materials of different refractive index (RI) will cause reflection from the interface. The greater the difference in RI the greater the percentage of light reflected. Everyone has always thought that because low molecular weight varnishes (like dammar or mastic) have higher refractive indecies (closer to that of aged oil) than high molecular weight varnishes (like B-67 etc.) that they looked better on a painting. René doesn't think that refractive index plays as large a role in the difference in optical properties between high and low molecular weight resins as do their viscosity and leveling properties.

Solution viscosity does have a lot to do with the aesthetic function of a varnish. Lower molecular weight resins have lower solution viscosities at a given resin concentration than do high MW resins. This difference in viscosity is important in leveling of the varnish layer. As a varnish dries by solvent evaporation, the resin concentration increases. With drying, sufficient solvent will have evaporated out to cause the varnish to thicken to a point where it will no longer flow. This happens sooner with high molecular weight resins than with low. As long as the varnish flows, it can self-level on the surface of the painting. Due to lower viscosity, low molecular weight resins level more efficiently on the surface of a painting producing a more saturated paint surface.

He went on to discuss the degradation of varnishes - generally photochemically initiated autoxidation caused by oxygen and ultraviolet light. Varnish is most susceptible to degradation because of the high ratio of surface area to volume and because there are no pigments present to absorb UV or inhibit oxidation reactions.

There are two degradation pathways which can be mimicked in accelerated aging tests. One is heat aging which causes yellowing. The yellowing produced by heat aging is light catalyzed and can be bleached by light. The other is light aging which is more representative of natural aging. Yellowing is not a good measure of degradation as it is caused by a secondary, non-oxidative, thermal process among photochemically generated autoxidation products.

There are ways to slow down the degradation process in varnish. We can eliminate UV light, minimize visible light, and /or add stabilizers.

There are two broad classes of stabilizers: antioxidants and UV absorbers. Antioxidants work by scavenging free radicals. There are two types: 1) heat stabilizers, which are not of real use to conservators; and 2) hindered amine light stabilizers (HALS). HALS work catalytically and therefore are not consumed as they are used. Tinuvin 292 (manufactured by Ciba Geigy, Additives Division) is the HALS proposed by René for use in conservation. The other class of stabilizers is UV absorbers. Coatings containing UV absorbers only protect the layers under the coating but offer no protection to the layer itself.

The initial results of accelerated aging tests with Tinuvin 292 in dammar (years ago) showed no improvement over dammar without the stabilizer. The accelerated aging tests were repeated with UV filtering Plexiglas between the varnish and the light source in the fadeometer. When UV light was removed, Tinuvin 292 conferred significant stabilization to dammar. The HALS stabilized dammar, when protected from UV radiation, does not yellow or develop UV fluorescence when light aging is followed by heat aging. However, the HALS does not inhibit the formation of oxidation products.

The conclusions from the accelerated aging tests were that dammar is very unstable. Dammar with the addition of 3% Tinuvin 292 (by weight to dry resin) and the exclusion of UV is much more stable. Mastic is less stable and is less responsive to stabilization by Tinuvin 292, although some benefit is seen as long as no UV is present.

René discussed low molecular weight synthetic resins, polymers with molecular weights in the 500-1000 range. We want a prospective varnish to have: optical effects similar to those of natural resins, e.g., high refractive index and good leveling properties; to be soluble in low polarity solvents; and to be stable. Low molecular weight resins, even if they do form occasional cross-links, will not form such large molecules and therefore will not become insoluble like high MW resins.

The following families of low molecular weight synthetic resins were discussed: ketone resins (polycyclo-hexanones) which include Laropal K80, MS2, AW2, and Ketone�N; the reduced ketone resins, MS2A and MS2B; the hydrogenated hydrocarbon resins, Arkon P-90 (Arakawa) and Regalrez 1094 (Hercules, Resins Group); and the experimental aldehyde resins (which are very expensive).

Testing of the LMW synthetic resins has shown that ketone resins are unstable and can only be stabilized with Tinuvin to a degree. Arkon P-90 and Regalrez 1094 with no stabilizer were found to change in artificial aging tests when UV was included, but they remained soluble. Arkon P-90 stabilized with 1% Tinuvin 292 was found to remain stable; but on long term accelerated aging (4,000 hours), it suddenly becomes more polar than the unstabilized product. However, Regalrez 1094 with 1/2% Tinuvin 292 was found to be very stable. Even after extended aging, no change was found in solubility or gel permeation chromatography.

The next topic for discussion was polymeric additives for low molecular weight resin based varnishes. The additives are used to modify the working properties and appearance of varnishes. Kraton G series rubbers (which are styrene-ethylene/butylene-styrene block copolymers) also decrease the brittleness of Arkon and Regalrez resins. Elvacite 2044 can be added to the aldehyde resins to improve durability. Kraton G polymers do not cross-link in a final varnish if present at below 10% of the dry resin, but they shouldn't be used without Tinuvin 292.

The next speaker was Jill Whitten, a Mellon Fellow at The Chicago Art Institute. She has been using Regalrez 1094 and the Kraton G rubbers in various solvents and at various concentrations in treatments of paintings. (See Jill's contribution in Technical Exchange for her suggestions on formulation.) The working properties were chosen to match those of natural resin varnishes.

Solvent selection is a very important aspect of varnish formulation. Jill has chosen to work with the Shell Sol line of petroleum solvents. The solvents can be ranked by their drying times to 90% evaporation. Shell Sol drying times range from 620 to 92,000 seconds. She recommended solvents in the range of 1,600-2,600 seconds as being best for use with Regalrez 1094.

Kraton G rubbers can be added to plasticize the Regalrez 1094, reducing its brittleness. She prefers Kraton G 1657 over the others tested by René. Kraton G works well when added to the Regalrez at between 1.5 -3%, however above 5% the varnish looks 'synthetic'.

Synthetic Resins as Surface Coatings, Alan Phenix, Lecturer, Department of Conservation and Technology, Courtauld Institute of Art, London

Following the 1930 Rome Conference, an international committee of conservators headed by W.G. Constable concluded that no ideal varnish existed. They recommended a coating of dammar or mastic, applied thinly to limit the visual effect of yellowing with age, followed by a coating of wax, to protect against moisture and environmental pollutants.

Alan Phenix stated that this recommendation acknowledged the limitations of natural resins as varnishes for paintings, and sought to minimize these limitations through the method of application. It seems that no synthetic materials at the time were considered to have better properties than the natural resins. George Stout stated that poly(vinyl acetate) was "probably the varnish of the future". This was the beginning of a search for synthetic resins that would surpass dammar and mastic in terms of stability and removability, and would successfully reproduce the optical qualities associated with these natural resins.

He described the properties, benefits and limitations of synthetic resin coatings, focusing on PVA, the cyclohexanones, and the acrylics. Varnishes with high refractive indices and good leveling properties produce the best saturation, and strongest color and tonal contrast, most successfully approaching dammar and mastic.

Copal oil produces tough films, which are resistant to solvent. In the 19th century, copal was regarded as valuable for its protective properties. One recommendation was to apply six or more coats of copal until the surface resembles a sheet of glass, safe from the hands of restorers.

For paintings of the late 19th century onward, one needs a broader range of materials to give a larger variety of effects. Some would argue that a matte varnish can give the appearance of an unvarnished painting. There is, however, a different nature of matteness between the inherent matte surface of a painting and one in which the matte elements are introduced in the varnish layer.

There are three criteria for a varnish: (1) having inherent appropriate physical and optical properties; (2) chemical and physical stability, long term maintenance, resistance to photo oxidation; (3) handling properties, to provide the most appropriate finish.

High MW thermoplastic polymers have lower refractive indices and higher viscosity in solution which produces relatively matte varnish surfaces. There are concerns with crosslinking and ease of removability. These polymers, however, have good physical properties: they are flexible, tough, and protect against abrasion.

Low MW resins produce a glossy finish, have good resistance to crosslinking, and though they change in solubility and oxidize, are still removable. The major limitation is that they have poor physical properties, tending towards brittleness and weakness.

Synthetic resins, in the early years, held the promise of increased stability compared to natural resins; now this is a less certain position. Appearance and optical criteria are a matter of debate. Our visual perception is acutely sensitive to surface; however, our ability to describe this is limited. Sara Walden in The Ravished Image states that synthetic varnishes are not compatible with all paintings. There is no ideal varnish. All materials have some failing. Often it can be the way varnish is applied that is problematic and gives inappropriate or undesirable results.

Poly(vinyl acetate) was adopted as a picture varnish in the 1930's. By the 1950's, PVA was one of the most common synthetic varnishes used in North America. Although PVA fulfills some of the criteria for conservation varnishes due to chemical stability and removability, its use in the field has been limited by concerns over physical properties. It has a low refractive index and highly viscous solutions, which tend to produce relatively poor saturation and gloss. It produces a relatively matte surface due to poor leveling and scattering at the paint varnish interface. At room temperature, the resin has a tendency to attract and absorb dirt, which can compromise its inherent color stability.

Acrylic varnishes include the n-butylmethacrylates (e.g., Elvacite 2044, Plexisol P550, Acryloid F10, LeFranc & Bourgeois, Soluvar Gloss (1978) ), the isobutylmethacrylates (e.g., Acryloid B-67, Elvacite 2045, Soluvar Gloss and Matte, Talens Van Gogh varnish), and others including Acryloid B-72.

The methacrylates have the potential for crosslinking, and became a focus for research in the 1950's. In 1957 Thomson found and Feller later confirmed that isoamylmethacrylate (27H) crosslinks and generates insoluble material. Butylmethacrylates also crosslink and become insoluble. Within just ten years, removability decreases significantly. In thin films, there is a higher proportion of insoluble matter.

B72 is the most stable of the methacrylates. It is a higher MW polymer, forming viscous solutions during removal. De Witte and Goessens-Landrie's study of B-72 found poor reversibility even when filtered for UV, however these results are regarded as anomalous.

Poly(vinyl alcohol), synthesized by the hydrolysis of poly(vinyl acetate), was first suggested in 1963 by G. Thomson as a protective coating with low oxygen transmission. Applied over a dammar coating it is effective in reducing yellowing, but there is a bloom problem. There are two concerns with the use of polyvinyl alcohol: the potential to crosslink, and the concern with polar adhesion to a non-polar surface. He gave an example of a painting with an initial varnish of polyvinyl alcohol. The coating separated from the paint, resulting in a whitish, opaque appearance, and required mechanical removal after swelling with MEK.

Polycyclohexanones were first patented by BASF in 1930. Winton Picture varnish, introduced in 1937/38, was formulated with polycyclohexanone and an unknown plasticizer to overcome brittleness. Talens Rembrandt Picture Varnish which was used by Ruhemann was based on AW2 and, later, on other ketone resins.

There are three classes of polycyclohexanone resins: AW2, the earliest, and MS2 a methylcyclohexanone/cyclohexanone oligomer; MS2A and MS2B, reduced forms of MS2 and AW2; and Ketone N and Laropal K-80, cyclohexanone oligomers. Methylcyclo-hexanones are softer and more plastic than cyclohexanone only resins.

Ketone N and Laropal K80 are brittle and less stable than MS2A and AW2. Laropal K80 is probably the least stable of the ketone resins, being probably no better than dammar or mastic in terms of removability, except that it yellows a bit less.

MS2A is revered and is considered to be the most stable of the ketone resins. Early commercial production yielded inconsistent products. The resin currently produced by Linden Chemicals is carefully controlled ensuring a consistent and reliable product.

"Paraloid (Acryloid) B-72 as a Varnish for Paintings", Stephen Hackney, Senior Conservator of Paintings, Tate Gallery, London

Stephen Hackney's talk outlined the properties of Acryloid B-72 resin (Paraloid in Europe) and a history of its use at the Tate Gallery. He referred to a long list of uses within the conservation literature. It is used commercially for a wide variety of purposes, including flexible ink and aerosol spray media and for coatings on wood. ICI first introduced methylmethacrylate in 1930 under the tradename Perspex. Paraloid B-72 was later produced by Rohm and Haas. The initial formulation consisted of 68% ethylmethacrylate and 32% methylacry-late. In 1978 the formula was changed to 70% ethylmethacrylate and 30% methylacrylate. The arrangement of the monomer units in the polymer is random along the chain.

He identified some of its characteristics. The average molecular weight of B-72 is 11,397. This indicates that B-72 is a much longer molecule than natural resin, and as a consequence is less readily soluble and the solutions formed are more viscous. It dissolves in xylene and is just soluble in warm ethanol. B-72 has a slightly sweet smell, due to the presence of residual monomer. Its glass transition temperature is 40�C; therefore it does not absorb dirt or exhibit cold flow. It is a medium hard resin which makes it appropriate for wood but not for metals.

Dr. Feller identified B-72, along with B-67 and PVA, as class A varnishes. Hackney also noted that B-67 contains unspecified stabilizers which make it almost as stable as B-72.He noted that while there have been reports of the insolubility of B-72, no clear evidence has been presented. Recent studies at CCI found some discoloration in B-72, but added that the discoloration was within an acceptable range. Yellowing of solvents can contribute to a misinterpretation of the cause of yellowed varnishes. The yellowing could also have been due to contamination or extended storage in solution. Hackney reported that B-72 varnishes applied to paintings at the Tate Gallery 29 years ago are still clear and remain soluble in xylene.

Because mastic absorbs UV, forming oxidative products, it offers some protection to the paint layer. B-72 does not absorb UV and therefore does not protect the painting from UV. B-72 does not bloom or blanch with age, unlike dammar or mastic. B-72 films do not form oxidized sites where moisture can accumulate causing bloom.

Hackney compared B-72 specifically and acrylic resins in general, to a variety of criteria for acceptable varnishes. B-72 has a refractive index of 1.483-1.487, lower than that of natural resins, which is 1.539, but close to that of dried oil paint. Recent work by Joyce Townsend shows that the refractive index of oil paint appears to vary throughout the painting and is increased by the addition of mastic or copal, up to 1.55. Laurie's assertion that the refractive index of oil increases with age to 1.57-1.60 has never been confirmed and appears to be incorrect. It might be argued that the refractive index of the varnish should match that of the paint medium.

Some of the acrylic varnishes such as B-72 saturate poorly and do not produce a glossy surface. Rather, B-72 conforms to the substrate, preserving the original texture of the paint surface, important in some applications. Slower evaporating solvents can increase the leveling and therefore the gloss. Hackney observed that wetting can be incomplete on polar surfaces such as oxidized resin layers causing a speckled interface between the two layers, particularly when B-72 is dissolved in xylene.

Dilute solutions of the resin can sink-in in some areas, producing an uneven surface. To achieve an even varnish, Hackney applies the first layer by brush using a 15% solution of B-72 (above 15% is too viscous, below is too thin to cover). For sprayed applications, a 12.5% solution of B-72 is used.

B-72 can be used as an overall surface consolidant or as a minimal coating when the paint is underbound. Hackney recommended about a 5% solution in sprayed applications, but cautioned such treatment is not reversible.

Flexibility is also a factor in the choice of varnish. While B-72 is less flexible than dammar, it is more flexible than MS2A. B-72 does not tend to crack and may therefore offer better protection than MS2A. A common problem with synthetic varnishes in general is their tendency to develop static effects, particularly if sprayed. Synthetic resins do not conduct electricity and as a result the stored charges attract dust particles.

Due to the toxicity of solvents required to dissolve B-72, Hackney suggested that the development of B-72 emulsions could have applications.

The earliest painting to be varnished with B-72 at the Tate Gallery (in 1964) has remained clear and has shown no tendency towards bloom or blanching. The painting has required surface cleanings frequently due to the accumulation of dust and particulate matter.

David Rainford from the FOM Institute for Atomic and Molecular Physics, Amsterdam, spoke on "The Mass Spectrometric Identification of Natural Resins and their Degradation Products in and on Paintings", co-written with Jaap J. Boon.

The advantages of mass spectroscopy (MS) for the identification of varnish and binders in paintings are that only a very small sample is needed and that a large amount of chemical structure information can be obtained. The disadvantages of various means of introducing the sample into the spectrometer were discussed. In-source pyrolysis mass spectrometry (PYMS) and high-temperature gas chromatography mass spectrometry (GCMS) solve the problems differently. In PYMS, the varnish sample is dissolved and coated onto a filament. The filament is heated and sample atoms and decomposition products are generated. The atoms are converted to ions by an ionization source close to the pyrolysis, either chemical ionization or electron impact. In high-temperature GCMS a derivatizing agent is added to the sample and the volatile reaction products are formed on-column, avoiding complex workup procedures.

These MS techniques have been used to identify the media used by a number of artists. The process of cleaning has been examined by analyzing paint both before and after cleaning as well as the residue on the swabs. Modern varnishes and lining adhesives have been found to have penetrated into paint layers.

"Some Examples of the Identification of Natural Resin Varnish Components in Paintings" Raymond White, Principal Scientific Officer, National Gallery, London

Raymond White was not able to attend, and his paper was summarized by Robert Feller.

Analyzing natural resins in the varnish layer of a painting is a demanding undertaking. Varnish additions may be present in the medium of the painting. The paint may also have a complex layer structure. The medium will have undergone centuries of thermodynamic and photolytic chemical change. The problems are accentuated by the increasing complexity of the mixture and by the general increase in the polarity of oxidation products, induced by lengthy periods of aging and exposure to the ambient environment. The inherent difficulties are compounded by the restriction on sample size.

One identifies organic materials within such samples by analyzing the material to establish the presence of a range of possible source-indicator compounds. Often, it is not possible to find indicators of such a specific nature to point to the precise origin of the material. Compounds may merely suggest a particular botanical or zoological genus, or even a family or sub-family. Such compounds are of use to the analyst as indicators, and may be either primary or secondary.

A primary compound is one that is present in the fresh material in reasonably substantial amounts and that possesses a chemical structure that is relatively stable to thermodynamic change and photolytic bombardment in the presence of oxygen. A secondary compound is not an original component of the material, but is formed from a primary indicator present in the fresh product that has degraded or changed into other products during the aging process. Although the primary indicator may no longer be above the detection threshold in a sample, the oxidation/degradation products grow steadily with time and are detectable.

One may have to accept less specific results and have compounds that suggest certain plant compounds. How much of a resin sample needs to be present before one can detect it? How much shows age or very aged? This is a difficult challenge. It is essential to run analytical blanks along with the samples.

Feller noted some problems with these types of analysis. Shellac will not work in a gas chromatograph. One can find very little of primary indicators in copal varnishes because it is cooked during the preparation of the varnish. One must have absolutely clean working methods. If one handles the test tube improperly, it will pick up waxes and oils from your hand, and contaminate the analysis.

He noted that to obtain meaningful results, you must work with the analyst and describe what you're looking for. The conservator has to think through what he wants from the analyst / scientist.

"Yellowness and Removability: How Much Change? How Fast? How Important?" Stefan Michalski, Senior Conservation Scientist, CCI, Ottawa.

Mr. Michalski stated the conclusion that none of the resins under consideration will turn into the crosslinked brown layer that conservators have had to remove from old paintings. These were primarily the result of the addition of drying oils and turpentine.

In a gallery below about 1000 lux, dammar yellows as much as it would in the dark. It remains to be seen whether Tinuvin 292 will help reduce this low light (dark) yellowing as remarkably as it reduces high intensity light yellowing. Irganox 565 has reduced dark yellowing by a rate of 3 - 4 over an 18-year study.

"Dark" yellowing occurs throughout the thickness of the varnish so a thicker layer will look yellow faster. A resin that remains acceptable for 400 years at 10nm thickness may only be acceptable for 100 years at 20nm, and 50 years at 40 nm. The solubility change of dammar with light aging shows reasonable agreement (reciprocity) across all studies from 20 klux to 200 lux. If dammar with Tinuvin 292 under UV filters follows reciprocity, it should remain essentially unchanged for several centuries at gallery light 200 lux.

Because there is little consistent reporting on varnish yellowing, Michalski interpolated and plotted other researcher's results in order to compare the data. From this analysis he was able to make a number of generalizations. Looking at data on removability, he showed that UV filtration will increase the life of a varnish by 10-15 times, adding Tinuvin will increase the stability by another factor of 10.

James Bourdeau, also from CCI and a co-organizer of the colloquium spoke on "Using UV Absorbers in Acrylic Top Coats as a Remedial Treatment for Dammar Varnishes Containing Irganox 565".

In the late 1970's, Raymond Lafontaine's research on stabilized varnishes popularized the use of Irganox 565 as an antioxidant in dammar. Subsequent research by René de la Rie showed that Irganox 565 forms yellow decomposition products on exposure to ultraviolet radiation. The Irganox works well, however, if protected from UV.

Over the years, a number of paintings have been treated at CCI with dammar stabilized with Irganox 565. To reduce the necessity of recleaning the paintings, the Irganox requires UV protection. Jim formulated and tested applying an additional varnish layer which incorporated a UV absorbent as a top coat to the paintings.

Experimental varnishes of Acryloid B-67 and B-72 with UV absorbers Tinuvin 1130 and Tinuvin 327 were tested. The Tinuvin 327 was used at 3% (to weight of resin, not solution) while 1130 was used at 4-4.5% (because of residual polyethylene glycol in the product). The Tinuvin 1130 is slightly yellow and has a sharper UV cut-off (akin to UF3 Plexiglas).The research found that an Acryloid B-72 top coat maintained the gloss of the dammar surface even on aging.

To block sufficient UV to afford protection to the layer below, the top coat of B-72 with Tinuvin 1130 and the HALS Tinuvin 292 should be 2-3 �m thick. This requires 12-15 very thin spray coats. The effectiveness of the coating could be evaluated by examination of the fluorescence of the dammar under UV light. Future research will examine the question of stabilizer migration.

The presentation raised some important questions. The first was about filtering the light source rather than coating the paintings. Jim said that was adequate if one could be assured that the lighting was always free of UV. Stefan Michalski commented that unless UV levels are checked regularly, incandescent bulbs could be replaced with the ubiquitous quartz halogen bulb without the conservator's knowledge. In an answer to a question at the end of the colloquium, Michalski also noted that he saw no need to add a UV absorber if UV is removed at the light source. If a painting is for a private client or the painting may go out on loan, one might add a UV absorber as a component of risk management.

A question was raised about the appearance of inpainting under ultraviolet examination. Jim explained that the UV absorbing varnish blocked a high percentage of the UV, but not all. Under higher intensity UV illumination, the fluorescence and retouch can still be discerned. A photograph with a long exposure shows the fluorescence even better.

The two day conference concluded with a discussion by a panel of the thirteen speakers, moderated by Ian Hodkinson, Professor and Program Director, Master of Art Conservation Program, Queen's University, Kingston.Some of the questions and summaries of the answers pertaining to the above talks follow:

Q: Do you think there is enough now known about the new hydrocarbon resins, for example Regalrez, to use them with confidence now?

There were different opinions but generally it was felt that Regalrez is stable and can now be used with confidence. It was felt that removability is directly related to the light dose, and therefore one should monitor and control the environmental conditions around the work of art. It was recommended that Regalrez be used with Tinuvin, however, some concern was expressed regarding the longevity of Tinuvin 292. Regalrez can be mixed without aromatics. If Kraton rubbers are added, you will need about 1 - 2% aromatic solvent. One response suggested a cautious approach, and a desire to see more workshops showing practical application.

Q: What is the shelflife for Tinuvin and Regalrez?

René de la Rie recommended that solutions be kept only for a short period, a few weeks only. Everything ages more quickly when dissolved. Tests with Tinuvin 292 in solvent solution without resin show that it lasts quite well. It is wise to not keep it in solution too long and to not expose it to light. Regarding the stability of Tinuvin 292, it should last a year, according to Ciba Geigy. Hindered amine light stabilizers by nature start reacting with the free radicals as soon as they are put into a varnish solution.

Q: Mixing resins such as B-72 with dammar, have these been tested? Is there any reason not to mix polymers and resins?

Hackney knows of no one who has tested these. Phenix referred to a study in Poland on Laropal with a range of Plexisols. They observed very different behavior in the tests. Bourdeau referred to Winsor & Newton Conservart, which is a mix of n-butyl methacrylate and cyclohexanone. Steven Prins said the Kecks experimented at the Brooklyn Museum with a mixture of AW2 and n-butyl methacrylate which give a nice well saturated satin finish. Conservart gives the same finish.

Q: Please comment on the anomalous results with the aging of B-72 by DeWitte in approximately 1975-6.

Someone noted that no one else has looked at the relationship between colored substrates and varnish aging. Robert Feller stated that short wave UV will "kill" the acrylics and noted that this would not be present in museums. Michalski stated that no one else has gone beyond the 50-50 acetone-toluene in solubility tests.

Q: Regalrez versus Arkon P90?

René de la Rie stated that at the Metropolitan in NY they were using Arkon on furniture as an inpainting medium, but the polarity of Arkon was so low that it did not mix well with pigments. He does hear that it works, however. Regalrez works. Ketone or Aldehyde may work better because of their higher polarity which improves pigment wetting.

Varnishes: Authenticity and Permanence Workshop

The two day workshop which followed the Varnishes Colloquium was a very practical extension of many of the ideas presented during the colloquium papers. Workshop participants had the opportunity to compare many different types of varnishes, both historic and contemporary, in a practical session.

The historic varnishes provided were based largely upon the research of Dr. Leslie Carlyle of CCI, who has meticulously studied various historic sources including original recipe books, to compile probably the most comprehensive listing of historic varnishes to date. Dr. Carlyle, using several historic recipes, had mixed up a number of varnish solutions as close to the original recipes as was possible in this modern age. She was quick to point out, though, that over the years resin sources have changed, in some cases quite dramatically, thus even her varnishes may not have had the same exact working properties, appearance, coloration, etc. as those originally used.

For contemporary varnishes, James Bourdeau, a paintings conservator at CCI, had solicited current materials in use from a number of conservators throughout the U.S., Canada, Britain, and other parts of Europe. He compiled these results into several recipes currently in use by a wide spectrum of practicing conservators, as well as various varnish practices which are utilized, including various ways of layering natural resins, synthetic resins and combinations of both. A number of these contemporary varnish solutions were also prepared for use by the workshop participants.

A Workshop Handbook which resulted from Dr. Carlyle's research and Mr. Bourdeau's findings, was published in conjunction with the workshop, and is available for approximately $20.00 US (check with CCI for the exact cost) from:

Conservators should find it an invaluable resource for various varnish recipes and practices.

The practical part of the workshop centered around the participants applying all of the various varnish solutions to 14 prepared canvas boards. The canvas boards had been painted with oil paints or acrylic emulsion paints, depending on the proposed varnishes to be applied. Each board was designated with a specific range of varnishes to be applied including: traditional varnishes 1750-1900; varnishes to monitor yellowing; mastic vs. dammar vs. Regalrez 1094; proprietary products; gloss and saturation in natural resins vs. synthetics; layering over different media; and traditional resins vs. synthetic resins over different media. Extra boards were provided for participants to use as they wished. I chose to apply various synthetic resins, with an eye on Regalrez 1094, over acrylic emulsion paints to see if there was any possibility of removal in the future. One of the potential discoveries from these boards is to see how they will age in the various climates and environments where they end up along with their "parent" conservators.

In between running from varnish station to varnish station, participants had the opportunity to listen to colleagues during demonstrations including: Nancy Binnie, CCI Conservation Scientist/Researcher, on color and gloss readings of varnishes; Stephen Hackney and Alan Phenix on techniques for applying B-72; Rica Jones substituting for Marion Barclay on brush varnishing with natural resins; and Jill Whitten on techniques for applying Regalrez 1094. The latter demonstration was particularly interesting as it focused in a very practical session on the research of René de la Rie, and the use of Regalrez 1094 both with and without added Kraton-G series rubbers to alter handling properties. Regalrez 1094 appears to be especially promising regarding appearance and saturation similar to natural resins, but with the benefit of long-term stability and removability.

It was especially interesting working with Dr. Carlyle and the various historic varnish solutions which she had prepared. She noted that dammar resin varnishes were not recorded in available literature until the mid-nineteenth century, thus probably were not widely used as an artist's material until some time later. Instead, historically, the most common varnishes utilized were mastic and copal-oil varnishes. Accordingly, dammar varnish, as far as appearance goes, may not be appropriate on paintings prior to the mid-nineteenth century. It was also very interesting to see the varnishes prepared using historic recipes. Many of these varnishes upon preparation were significantly colored, some definitely yellowed or golden, others actually brown. Seeing this, it made one rethink using clear colorless varnishes on certain historic paintings. Certainly artists had to be aware of the dramatic visual change that occurred once these coatings were applied, thus compensation using the original palette for the subsequent varnish may have had more of an effect than we realize today. This is an area that will require much further study, but perhaps the idea of the "old master glow" should not be so readily dismissed.

Since this was the first attempt at a workshop of this magnitude, there were some aspects that simply could not be addressed. There was no experimentation with spray application of varnishes, all varnishes were applied by brushing, which greatly limited the visual effects that could be achieved. Also, although there was considerable discussion about handling properties, it was virtually impossible to get a feel for the varying handling properties of different varnishes when they were being applied in one or two inch strips. Still, in all, the workshop proved to be an incredible learning experience for everyone involved, and hopefully will spawn future workshops where participants may continue to experiment and learn.

Painted Wood: History and Conservation

November 12-14, 1994
Williamsburg, Virginia

A wide variety of disciplines and aesthetics were represented at the Painted Wood symposium held at Colonial Williams-burg on November 12-14. The keynote address by Wendy Cooper united the variety of subjects by stressing the need for collaboration between those with various backgrounds and the need to balance ethical issues and practical concerns in designing a well-planned conservation treatment. Wendy Samet, Chris Augerson and Mark Harpainter submitted papers that discussed aesthetics and treatment issues in objects that bridge painting conservation and furniture conservation. Augerson and Harpainter presented the painted wooden art of Arthur and Lucia Mathews from early 20th century California, illustrating the need for literacy in both fine and decorative arts. Samet's paper compared and contrasted the inpainting needs of traditional paintings and painted furniture.

Three of the case studies presented were the result of collaborative efforts between curators and conservators. Merri Ferrell and Marc Williams discussed how document research and sensitive conservation resulted in preservation of an omnibus and identification of the decorative artist. Lynne Dakin Hastings and Deborah Bigelow discussed how their collaboration on a set of painted furniture resulted in a treatment that harmonized the various pieces and their setting. Peter Sixbey and Rick Parker discussed their collaboration in the treatment decisions regarding carousel animals. Sixbey's observation that carousel animals were shop produced and not the vision of one person reinforced Samet's thesis that easel paintings and painted furniture need to be approached with the different artistic intents in mind.

The definitions of folk art, decorative arts, and fine arts and their treatment implications were a theme throughout most of the talks. Valerie Reich Hunt and Robert Shaw discussed their work with folk art in the Shelburne Museum. Hunt noted the many ethical treatment issues involved with objects that were used, maintained and modified during their functional lives. Stephen Ray and Julie Reilly further developed this topic, discussing their standard methodology for decision-making in folk art treatment.

Questions regarding the purpose and artistic intent of the object were addressed with different philosophies. Some chose to return "maintained and repainted" objects to the original surface appearance, such as Rick Parker's treatment of the carousel horse. As a result, Parker discovered evidence of shop practices that were not known through written documentation. Other conservators returned objects to the appearance when last used, such as Marc Williams' and Jennifer Baker's recovery of the last paint scheme of a horse-drawn vehicle. Other objects, such as the cornice of Olana, were returned to an appearance suitable to the interpretive time period of the site. Hunt and Shaw mentioned the treatment philosophy of the Shelburne, that of returning objects to the 1950's appearance when the folk art left private ownership and entered the museum. The difficulties in making these decisions underscore the need for a multidisciplinary and collaborative approach to conservation.

Many of the speakers felt that one of the best vehicles for clarifying some of these treatment issues was through increased connoisseurship and technical studies involving the review of period documentation, gathering empirical evidence through analysis, and duplication of period techniques. Margaret Ballardie summarized materials and techniques documented in period japanning handbooks and those found by empirical evidence on japanned furniture. Her presentation included a demonstration of japanning techniques. Her demonstration was well complimented by Ina Brosseau Marx's demonstration of period graining techniques.

There were also several systematic technical studies including Nancie Ravenel's research on four painted 16th century Italian picture frames. Frances Gruber Safford presented her thorough research on the designs, materials and techniques used on American Mannerist "Saybrook" chests. Based on her systematic study, Safford was able to classify the stylistic changes into three distinct genres.

Chris Shelton presented a technical study of three techniques for decorating gold and white furniture, illustrated by objects made in 18th century Philadelphia. Robert Mussey presented a systematic study of verte antique decoration. By researching the four different traditions of grainers, gilders, decorative painters and lacquerers, Mussey was able to demonstrate how their familiarity with different materials resulted in four different methods of imitating corroded bronze. Susan Buck presented research on Shaker furniture and dwellings. She discussed the changing appearance of Shaker interiors as the community regularly stripped and repainted their furniture in accordance with their philosophical principles and the availability of commercially prepared materials.

All the presentations at the symposium utilized the skills of the historian, the conservator, the scientist and the artist to better understand the various painted wood objects that span fine arts and decorative arts. The symposium stressed that there are many different ways of considering objects and that all of these views must be considered in treatment decisions. The success of these presentations ensures that the future publication of the papers of this symposium will become a primer for those who work with painted wood.

And From An Ethnographic Conservator's Point of View

For conservators in general, Richard Wolbers gave a paper about a new adhesive that he has been researching for a number of years: Poly (2-ethyl-2-oxazoline). This adhesive has a very wide solubility range: from water to many organic solvents. It will wet many different surfaces, can be mixed with other polymers, can be heat sealed, and will adhere to a wide range of surfaces. Initial testing also indicates that the material is very stable.

Jonathon Thornton presented a thoroughly researched and fascinating paper about The Use of Dyes and Colored Varnishes in Wood Polychromy. It got many people in the audience thinking about how many "yellowed" varnish layers they had removed over the years that may have been intentionally applied.

One of the most interesting overriding themes of the conference from an ethnographic conservator's point of view was the obvious change in ethical approaches to inpainting, cleaning and reintegration that has slowly been occurring in all conservation specialties. These ethical changes seem to have clearly been influenced by ethnographic ethics. Everyone is now doing less interventive treatment than before, and questioning what they are doing to an object and why, in a way that I have previously only heard in ethnographic conservation discussions.

Lori van Handel and David Arnold presented case studies using the Ultrasonic mister, developed at the CCI by Stefan Michalski and Carole Dignard for the consolidation of flaking paint, which were particularly useful to those conservators who are unfamiliar with the merits of this homemade piece of equipment. The mister allows for the successful consolidation of very powdery and porous paints.

Eric Hansen and Mitchell Bishop's paper Factors Affecting the Retreatment of Previously Consolidated Painted Wooden Objects is one that holds out hope for conservators faced with retreatment problems. Hansen discussed redistributing the old consolidant, pre consolidation, and removal of the old consolidant as treatment options.

Nancy Odegaard's paper An Investigation of the Nature of Paint on Wood in the Indigenous Southwest of North America made a valuable contribution to understanding the function, meaning and cultural considerations of paint on objects from the Southwest. Paint is the most important form of decoration of wooden objects; each color has meaning. Pigments can complete, or make an object sacred.