PUVIS DE CHAVANNES'S ALLEGORICAL MURALS IN THE BOSTON PUBLIC LIBRARY: HISTORY, TECHNIQUE, AND CONSERVATION
TERI HENSICK, KATE OLIVIER, & GIANFRANCO POCOBENE
1 APPENDIX 1
1.1 BINDING MEDIA OF THE MURALS
A sample of the ground was hydrolyzed, derivatized with phenylisothiocyanate, and analyzed for amino acids by high-performance liquid chromatography (HPLC) following Waters Chromatography “Pico Tag” method. Analyses were carried out on a Waters instrument that consisted of two 510 pumps, a manual injector, and 991M photo diode array detector. The amino pattern closely matched that of collagen, with the exception of considerably reduced glutamic and aspartic acid levels; these amino acids are typically lost during sample preparation when calcium carbonate is present, as was the case in the ground sample (Halpine 1992). Another portion of the ground was saponified in 10% KOH in methanol, the solution neutralized and then extracted twice with ether. Ether extracts were combined, evaporated to dryness, redissolved in a small amount of methylene chloride, and methylated with 10 μl dimethylformamide dimethyl acetal (Pierce Chemical Co.). Analysis was carried out by gas chromatography—mass spectrometry (GC-MS) on a Hewlett Packard 5890 capillary gas chromatograph equipped with an HP 5971A mass selective detector. The sample was found to contain traces of palmitic and stearic acids, which are typical contaminants in animal glues. It is also possible that these fatty acids came from the overlying oil paint layer.
1.1.2 Paint Binder
Two samples were saponified and analyzed by GC-MS as described above. While traces of hydrocarbons were detected, such as those characteristic of hydrocarbon waxes such as ceresin, the very low levels suggest that they may simply be contaminants in the oil and not indicative of wax intentionally added by the paint manufacturer. Traces of methyl dehydroabietate, an oxidation product commonly found in aged pine resins, were also detected in the two samples. The level was extremely low, so it cannot be concluded that pine resin was an intentional component of the paint. Ratios of the C9 and C8 dicarboxylic acids (diC9/C8) suggest that the oil was not heat-bodied. The palmitic/stearic acid ratios (P/S) are somewhat high for linseed oil; possibly the oil was walnut or a mixture of linseed and poppyseed. Poppyseed was identified in all four samples from the tubes of paint (see next section).
1.1.3 Tube Paints
Four partially dried samples from tubes preserved from Puvis's studio by his heirs were analyzed by GC-MS as described above. The oil in each appears to be poppyseed, based on the high palmitic/stearic acid ratios (P/S). Two samples contained substantial amounts of straight-chain hydrocarbons in the C22-C32 range, maximizing at C26. This hydrocarbon pattern is typical of ceresin wax. The results were as follows (when two analyses were carried out, P/S ratios from both are given):
2 APPENDIX 2
2.1 ANALYSIS OF THE PIGMENTS, MAROUFLAGE ADHESIVE, AND STAINS
Nine paint samples were selected for analysis, including two greens, one red, four blues, and two whites. Cross sections of the paint samples were examined using a Leitz Laborlux biological microscope fitted with visible and ultraviolet light sources. The cross sections were prepared by imbedding the paint samples in cubes of bioplastic polyester resin, which were subsequently cut and polished with micromesh polishing cloths to reveal the structure of the paint layers. The cross sections were analyzed for their elemental composition using a JEOL 6400 scanning electron microscope with a Noran Instruments Z Max 30 Series light x-ray energy dispersive x-ray spectrometer (SEM-EDX). Sample sites were analyzed for 100 counts at an accelerating voltage of 20 kilo electron volts (keV). Standardless quantitative analysis was performed by the calibrated voyager quantitative microanalysis system using ZAF matrix corrections. Fourier transform infrared microspectroscopy (FTIR) was performed using a Spectra-Tech IR-Plan microscope attached to a Nicolet 510M spectrometer with an auxiliary MCT detector. Samples were mounted for analysis on a Spectra-Tech Micro Sample Plan fitted with a diamond window, and data were collected for 200 scans at a spectral resolution of 8 cm−1. The resulting spectra were viewed in absorbance mode between 625 and 4000 wave numbers, and for consistency the CO2 peak was removed and the spectra were baseline corrected.
Pigments were identified by microscopy and SEM-EDX (see table). Analyses showed that lead white was used throughout the painting as highlights and aa matrix for the colors. For example, the highlight from The Inspiring Muses is lead white over a gray layer composed of lead white modified with ultramarine, red oxide, emerald green, and carbon black. The principal green pigments used were emerald green (copper aceto-arsenite) and green earth (hydrous iron, magnesium, and aluminum potassium silicate). Emerald green was found in the grass of Philosophy and in the initial green layer of the foreground of the Muses, together with green earth. The upper layer of the foreground of the Muses contains particles of green earth in a lead white matrix. The principal blues found were cobalt blue (cobalt aluminate) and synthetic ultramarine (sodium aluminum silicate). The water for the Muses was ultramarine mixed with lead white, black, and red in one simple layer. The blue sky in both Chemistry and Aeschylus is cobalt blue mixed with lead white. The blue water of Aeschylus was applied in two layers: ultramarine blue and lead white overlaid by cobalt blue and lead white. Four reds were identified: vermilion (mercuric sulfide), red lake on an aluminum substrate, red ochre, and red lead. History's red robe is built up from a layer of lead white and red ochre followed by a layer containing lead white, calcite, and vermilion. Red lake was found as a minor component of the green grass in the Muses.
The yellowish white ground was identified by SEM-EDX and FTIR as chalk with a proteinaceous binding medium. A small amount of oil was also identified in the ground, which could account for the yellow color of the ground.
2.1.2 Marouflage Adhesive
The marouflage adhesive, which holds the canvas onto the wall, was analyzed by SEM-EDX and FTIR. The sample showed absorbances for lead white, oil, and lead carboxylate (a byproduct of the aging of lead white in the presence of oil).
2.1.3 Drip from Surface of The Inspiring Muses
A drip stain was sampled from the surface of the Muses. FTIR analysis showed major absorbances for protein and lead white. There was a minor absorbance at 1076 wavenumbers that was unidentified.
3 APPENDIX 3
3.1 UV FLUORESCENCE STAINING ANALYSIS OF PAINT SAMPLES FROM THE INSPIRING MUSES AND CLEANING FORMULAS
Samples of The Inspiring Muses were taken in an attempt to characterize the drip and stains and the accumulation of grime on the surface of the paint. Samples were also taken from similar areas before and after grime and accretion removal tests to evaluate the effects of the cleaning solution. These were mounted in Ward's Bioplast and sent to Richard Wolbers for ultraviolet fluorescence analysis and photomicrography. Twelve samples were viewed and photographed in the following sequence: normal light, UV only, UV stained with 4% triphenyl tetrazolium chloride in methanol (TTC), UV stained with .25% rhodamine isothiocyanate in acetone (RITC), and UV stained with .20% rhodamine 123 in acetone (RHO 123).
The results of the examination can be summarized as follows:
3.1.1 Samples of Paint with Surface Bloom but No Deposit from Steam Drips
Boston Public Library (BPL) 1. Green paint. The green paint stained with RHO 123 (for oils) and with TTC (carbohydrates), signaling both materials and therefore potentially water soluble components in the binder. Apart from a single droplet of a noncharacterized material (negative staining with the stains listed above) and a substantial and continuous grime layer, no other surface-accumulated materials were observed.
BPL 3. White paint. The white paint stained only lightly with RHO 123. Little accumulation of grime was noted on the surface, which was slightly autofluorescent, indicating normal aging on the surface.
BPL 5. Blue paint. The blue paint stained RHO 123 positive (oil), but, as in sample BPL 1, there was a slight reaction with TTC (carbohydrate) as well. The surface is obviously aged (autofluorescent) and grime-laden, but no additional deposits were noted on the paint surface.
3.1.2 Samples of Paint with Deposits from Steam Drips
Three samples of the green, white, and blue paint (BPL 2,4, and 6) all exhibited obvious autofluorescent “deposits” along the surface and above the grime layer under UV light. The morphology of the deposited materials was noncrystalline (organic) in nature and may have signaled the extraction and redeposition of a discrete layer of polar organic material. However, this material was negative for all of the applied stains.
3.1.3 Samples of Paint After Cleaning with Wolbers Diammonium citrate-chelating/ionic strength solution
Three samples of the green, blue, and white paint (BPL 7, 8, and 9) all showed undisrupted surfaces after cleaning with Wolbers's diammonium citrate-chelating/ionic strength solution.
Xylenes/water emulsion formula (for the removal of BEVA 371 adhesive and less tenacious grime layers):
50 ml xylenes, 20 ml Triton X-100, 30 ml water, 2 ml triethanolamine (the aqueous phase was adjusted with dilute HCl acid to a pH of 8.5 prior to emulsification)
Diammonium citrate-chelating/ionic strength solution (for the removal of surface grime and stains):
100 ml water, .5 g deoxycholic acid, 5 ml triethanolamine, 1 g ammonium chloride, 1 g diammonium citrate, .05 g Triton X-100, 1.5 g hydroxypropyl-methyl-cellulose (adjusted with HCl acid to a pH of 8.5)
Baudouin, P.1914. La fresque: Sa technique, ses applications. Paris: Librairie Centrale des Beaux-Arts.
Baudouin, P.1935. Souvenirs sur Puvis de Chavannes. Gazette des Beaux Arts(Paris)13: 295–314.
Béguin, A.1981. Dictionnaire technique de la peinture. Paris: Audré Béguin.
Carpentier, P.1875. Notes sur la peinture à la cire cautérisée ou procédé encaustique, d'âprés les laborieuses recherches de Paillot de Montabert. Paris: Librairie Renouard, Rue de Tournon.
Crowninshield, F.1887. Mural painting. Boston: Ticknor and Co.
d'Argencourt, L.1973. Les peintures murales de Puvis de Chavannes au Musée de Picardie. Thesis, Université de Paris, Sorbonne.
d'Argencourt, L., M.-C.Boucher, D.Druick, and J.Foucart, eds.1977. Puvis de Chavannes. Exhibition catalog. Ottawa: National Gallery of Canada.
Esch, E.1982. The decorative program of the Boston Public Library. Honors thesis, Harvard University.
Forbes, E. W.1940. Letter to Robert Bellows, February 1. Cambridge, Mass.: Harvard University Art Museums.
Galinier, C.1995. La décoration de l'escalier du Musée des Beaux-Arts de Lyon. Unpublished monograph. Service de Restauration des Musées de France, Versailles.
Gautier, T.1861. Abécédaire du salon de 1861. Paris.
Geffroy, G.1897. La vie artistique, vol. 5, ed.H.Floury. Paris.
Gettens, R.J., H.Kuhn, and W. T.Chase. 1967. Lead white. Studies in Conservation12: 125–39.
Halpine, S. M.1992. Amino acid analysis of proteinaceous media from Cosimo Tura'sThe Annunciation with Saint Frances and Saint Louis of Toulouse. Studies in Conservation37: 22–38.
Hamerton, P. G.1882. The graphic arts: A treatise on the varieties of drawing, painting, and engraving in comparison with each other and with nature. Boston: Roberts Brothers.
Lefranc and Bourgeois. 1990. Le livre des couleurs. Le Mans: Lefranc and Bourgeois.
Lefranc and Cie. 1897. Les couleurs et lesvernis de J-G Vibert. Notes et renseignements sur les produits préparés pour la peinture artistique d'après les procédés de J-G Vibert. Paris: Lefranc and Cie.
Mayer, R.1970. The artist's handbook of materials and techniques. New York: Viking Press.
Milner, J.1988. The studios of Paris: The capital of art in the late nineteenth century. New Haven and London: Yale University Press.
Mora, P., Mora, L., and Philippot, P.1994. The conservation of wall paintings. London.
Price, A. B.1977. The decorative aesthetic in the Work of Pierre Puvis de Chavannes. In Pierre Puvis de Chavannes, ed.L.d'Argencourt et al. Ottawa: National Gallery of Canada.
Price, A. B.1994. Pierre Puvis de Chavannes. Amsterdam: Van Gogh Museum.
Sérullaz, A.1995. Delacroix à l'Assemblée Nationale.Paris: Assemblée Nationale.
Cartwright, J.1896. Puvis de Chavannes. Art Journal. 190–92.
Farell, E., and A.Snodgrass. 1992–94. Analytical reports on the Boston Public Library murals. Straus Center for Conservation, Cambridge, Mass.
Moore, C.1929. The life and times of Charles Follen McKim. Boston and New York: Houghton Mifflin Co.
Newman, R.1993. Analytical reports on the Boston Public Library murals. Research Laboratory, Museum of Fine Arts, Boston.
Puvisde, Chavannes. 1896. Description of the decorative paintings of Puvis de Chavans. Boston Public Library.
Sharp, W.1898. Puvis de Chavannes: An appreciation. Art Journal. 377–378.
Vachon, M.1895. Puvis de Chavannes. Paris.
Wattenmaker, R.1975. Puvis de Chavannes and the modern tradition. Toronto: Art Gallery of Ontario.
SOURCES OF MATERIALSBEVA 371
Conservator's Products, P.O. Box 411, Chatham, N.J. 07928Acryloid B-72
Rohm and Haas Co., Philadelphia, Pa. Supplied by Conservation Materials, 1165 Marietta Way Sparks, Nev. 89431Deoxycholic acid Triethanolamine Citric acid-diammonium salt Triton X-100 Hydroxypropyl-methyl-cellulose Triphenyl tetrazolium chloride (TTC) Rhodamine isothiocyanate (RITC)
Sigma Chemical, P.O. Box 14508, St. Louis, Mo. 63178Ammonium chloride
Mallinckrodt Specialty Chemicals, Paris, Ky. 40361Hydrochloric acid (reagent A.C.S.)
Fisher Scientific, Springfield, N.J. 07081Mylar (.0005 in.)
Talas, 213 West 35th St., New York, N.Y. 10001Magna Colors
Bocour Artists Colors, New York, N.Y. 10019Willard heated spatula
Willard Developments, Leigh Road, Chichester, West Sussex P019 2T3, U.K.Rhodamine 123 (RHO123)
Kodak Laboratory and Research Products, Rochester, N.Y. 14650
TERI HENSICK has been a paintings conservator at the Straus Center for Conservation, Harvard University Art Museums, since 1980. She holds a B.A. (Phi Beta Kappa) in art history from Wellesley College and trained in paintings conservation in Florence (Universita Internationale dell'Arte), Zurich (Swiss Institute for Art Research), and Nuremberg (Germanisches Nationalmuseum). She interned in paintings conservation at the Harvard University Art Museums in 1976–77 and was assistant and subsequently associate conservator of paintings at the Detroit Institute of Arts from 1977 to 1980. In 1995 she spent two months in France researching Puvis de Chavannes's materials and techniques with the support of a National Endowment for the Arts professional development grant. Address: Straus Center for Conservation, Harvard University Art Museums, 32 Quincy St., Cambridge, Mass. 02138.
KATE OLIVIER received her training at the Courtauld Institute of Art from 1962 to 1965. She worked for six months in Florence following the flood in 1966 and one year in Venice on paintings by Tintoretto. As a private conservator in London she worked regularly for the Department of the Environment, the Royal College of Music, and private conservator Patrick Lindsay. From 1974 to 1976 she was assistant painting conservator at the Winterthur Museum, Delaware. She has been a conservator at the Fogg Art Museum since 1977. Address: Fogg Art Museum, Harvard University, Cambridge, Mass. 02138.
GIANFRANCO POCOBENE received his master of arts in conservation from the Art Conservation Program, Queen's University, in 1984. He was assistant conservator at the Art Conservation Laboratory in Raymond, New Hampshire, from 1984 to 1985. He returned to Queen's to conserve paintings from the university collection from 1985 to 1988. During that time he also worked on paintings from the Alfred Bader Collection and on several mural projects in Canada. From 1988 to 1989 he was a paintings conservation intern at the Center for Conservation, Fogg Art Museum. Upon completion of his internship he became assistant paintings conservator. He is currently associate conservator of paintings at the Straus Center for Conservation, Harvard University Art Museums. Address: Straus Center for Conservation, Harvard University Art Museums, 32 Quincy St., Cambridge, Mass. 02138.