唐代帝陵

                                             
　　The following study treats a detached Chinese mural painting from the collection of Asian Art at the Mus& Guimet in Paris. The research and restoration of this paint- ing was the subject of a diploma thesis at the IFROA (The French Institute for Conservation of Works of Art), in Paris. Measuring 110 cm by 71, 5 cra, this painting is part of a lot of six paintings, all housed at the Musee Guimet and all of similar dimensions and techniques. This painting represents two female figures, and the artist employed a dry technique on a render made of clay, sand, straw and other vegetal fibers. While this painting has remained in good condition, and is a rare ex- ample of a Chinese mural painting in its original state, it has never before been the object of a study.
　　Historical documentation
　　Little is known of the original provenance of the painting, the date in which it en-tered into French collections, and its iconography. Along with four other paintings from the lot of six, this detached wall painting has been stored flat in drawers dueto its lack of rigid support.
　　The subject represented had not been identified until now. With the help of the po-em by Bai Ju- yi (772-846)①, The song of the everlasting regret, an analysis of the iconography is possible. The text leads us'to discover that the figure on theright is indeed a representation of Yang Guifei, accompanied by a palace lady on the left. Yang Guifei is an historical character of the Tang dynasty (618- 907), the favorite concubine of the emperor Xuanzong who reigned from 713 to 755. Throughout the ages, her story and tragic destiny has much inspired painters,sculptors, poets and dramatists in China.
　　A formal analysis has lead to a hypothesis for dating the painting. Comparing it to Chinese murals from different styles suggests that it may have been executed be-tween the end of the Yuan and the beginning of the Ming dynasties, in the earlyXIVth century.
　　Despite the progress made by this initial study, the identification of the subject and its date remain uncertain.
　　Condition Report
　　In the absence of a rigid support, the painting was manipulated with great difficul-ty. Likewise, transporting the work posed another threat to its fragile support. Its transfer from the Musée Guimet to the fFROA was done by placing the work in acustom- made box with a removable back.
　　During the condition report, no trace of previous restoration was found. In spite of its presumed date, its detachment, and multiple transfers, the painting appeared to be well preserved. The support showed alteration marks, notably on the top andbottom edges which were seriously damaged. During the painting's detachment from the wall, the support was dramatically reduced to only approximately 5 mm of width, rendering the painting fragile. It was thus this new state of fragility thatprovoked considerable damage to the edges.
　　The representation was complete but the coloration dulled by a layer of grayish dust and by the presence of white drippings on the painting's surface. The pictorial layer presented abrasion zones and suffered from a general state of powdering. 0nly a fewlacunae were present on the edges, and the format of the painting leads to believe that the original composition was without doubt of larger dimension. Finally, small fragments of paper peeling away from the back seemed to indicate the presence of apaper mounting on the back of the painting.
　　This condition report revealed two major problems for which an intervention seemed necessary in order to assure the work's conservation: a general powdering of the pictorial layer and the absence of a rigid support.Research for a method of consolidation for a matte pictorial layer, without visual modifications
　　The consolidation of matte and porous pictorial layers often leads to an undesirable modification of the colors and to an undesirable shine. It was thus necessary to es-tablish a method of consolidation that minimized the risk of visual modifications.
　　A certain number of adhesives and application methods, selected among the prod-ucts frequently used for restoring mural paintings, were, therefore, tested on sam-ples reproducing the technique employed for the painting in question. Following
　　these tests, colorimetric measurements allowed us to choose the most appropriate method.
　　Initially, a series of micro- samples were done on the painting in order to determine the technique and the original materials used. Analyzing these samples revealed that the techniques and materials used for this work are in keeping with traditional Chi-
　　nese mural paintings.
　　The support was made of a mixture of clay, rich in smectite- type clay, as well as different vegetal fibers: straw and other fibers generally used for the fabrication of paper. A preparation consisting of barium white was then applied over the entire
　　surface. The contours of the drawing were outlined with the help of ink, then the fields of color made of pigments and vegetal gum were applied. The following pig-ments were identified: atacamite green; a mixture of iron oxides, red lead and ver-milion
　　red; orpiment yellow; a white composed of barium sulfate; carbon black.The blue could not yet be identified, but it is not azurite. Finally, raised oma-ments, consisting of a mixture of calcium carbonate and kaolin were added to the
　　painting's surface before being covered with silver foil.
　　Samples, imitating the painting's technique with a powdering pictorial layer and employing the same materials, were fabricated in order to serve as supports for the consolidation tests. One to three layers of adhesives in solution with the following
　　solvents were applied to the samples:
　　Acryloid B372 (r) at 5 % in xylene, toluene and acetone Plexisol P550 at* 5 % in acetone Primal E300* at 5% in ethanol Klucet G* at 2 % in ethanol The solutions were applied with a brush, pulverizing with the help of an aerograph,
　　and on certain samples, according to the methods described by Eric Hansen②, in anatmosphere saturated with solvent vapors. To serve as evidence of the original col-ors, one band on each sample was left untouched. After these applications, a cer-tain number of observations could be made with the naked eye before preceding with the colorimetric measurements using a goniospectrophotometer with optical fibers②,
　　This machine was perfected by the C2RMF (Center of Research and Restoration of the Museums of France). Its use has already been applied to works of art for the
　　recognition of certain pigments, as well as for the characterization of rough surfaces and the measure of the color modifications.
　　Colorimetric measurements of the samples were then taken, both before consolida-tion and of each consolidated band. These measurements allowed us to verify the in-nocuousness of the substances used, and to quantify the difference in color created
　　by the application of each adhesive. The consolidation method by vaporizing Acry-loid B72* at 5 % in toluene seemed to be quite efficient and caused little modifica-tion of the colors. For these reasons, this solution was chosen for the restoration of
　　the painting.
　　Restoring the painting
　　The painting's materials had kept their original physical characteristics, a detail of great interest as much to historians as to conservators. However, the painting's fragility rendered it dangerous to' handle, difficult to conserve, and impossible to
　　exhibit. It was crucial to orient its restoration toward the most minimal intervention possible in order to keep the appearance of the work intact. Thus, we decided to proceed with a type of restoration that Cesare Brandi called "preventive restora-
　　tion"③, which "fully preserves the future appreciation of the work of art, uniquely by keeping it in the state in which it was found."
　　The restoration of the work began with a complete dusting with the help of a small,soft paint brush and a mini- vacuum at various intensities. While controlling with each movement the solidity of the colors, the dust was delicately taken off with the
　　brush, then picked up by the vacuum held at a distance. This operation allowed the painting to regain the intensity of its color and allowed for a better legibility of the image. The white drippings splattered on the pictorial layer were then removed,
　　under binoculars, with a dry technique using a scalpd and a paint brush with close -cropped bristles. Cleaning techniques were then tested on the work in order to determine if it were possible to reduce the grayish layer present on the painting's surface. However, this grayish layer was too closely linked to the fragile pictorial layer, rendering its thickness impossible to reduee.
　　The painting was finally consolidated by vaporization with an aerograph of Acryloid B72* in a solution of 4 % in the toluene in order to permit the application of two
　　layers of adhesive while still minimizing the risks of visual modifications. A piece of the pictorial layer was voluntarily left unconsolidated, m that it can serve as a refer-
　　ence for future analysis.
　　The final intervention was meant to recreate a rigid support for the painting so that it could be handled without risk and eventually exhibited. The pictorial layer was still quite fragile, even after the minimal consolidation. Turning the work over, to
　　gain aecess to the reverse side, had to be done without applying facings. For this,the work had to be prepared. The edges were purified, and put back in place, and
　　the fragments of the support were glued with a solution of 50 % water and 50 % a mixture of two parts Plextol* B500 and one part D 360 . The edges were finally blocked by using modeling clay, then the painting's form was cut in mousse. The
　　painting as well as the reliefs were protected by a layer of Japanese paper and a layer of thin mousse set on the surface. Then, a wood panel was positioned and blocked with elamps to enable the painting to be turned over. After turning it, the back
　　seemed to be in good condition and the paper, indeed covering the entire surface,was also well preserved. It was thus decided that the paper would be conserved and that a method of mounting the painting would be employed. The paper offered vari-
　　ous advantages: a composition, rich in vegetal fibers, similar to the original materi-als used for the support. This composition did not risk hindering the mobility of the render, nor did it constrict the work. Different layers of heavy Japanese paper 
　　were thus applied to the painting, using, for the first, starch glue, and for the following layers, a mixture of Primal E330* at 50% in thickened ethanol with Klueel G*
　　at 10% in ethanol. Finally, to insure that the back was fiat, a paste composed of cellulose pulp and Klucd G* at 5 % in ethanol was applied to the cavities. A layer of 3 mm- thick pressed cork was glued with pure Plextol B500* to the last layer
　　of paper, allowing the reversibility of the mounting. The final support, consisting of aluminum alveotate, was glued with the help of a slow- drying Araldite*. This support was chosen because of its solidity, its lightness, and its imputrescibility.
　　Finally, the edges of the painting that were lacunary were reintegrated with the help of a neutral colored mixture of clay, sand, vegetal fibers, and Plextol B500*
　　meant to increase the edge's resistance. This was applied at a distance in reference to the level of the pictorial layer, in order to diminish the bothersome effects caused by the presence of gaps.
　　Following this restoration, the painting could finally be put in a vertical position and
　　could re - appropriate the characteristics of a mural painting, as opposed to being flat as it had been since its deposit. The painting can currently be viewed in its to-tality, which allows for a better appreciation from a removed standpoint. Its picto-
　　rial layer is consolidated, and the image has regained a certain vivacity. Exhibiting this mural painting is now conceivable.
　　Materials
　　Aeryloi'd B 72*, Primal E 330*, Plextol* B 500 and D 360 and Plexisol P 550 manufactured by Rohm and Haas CompanyKlucei G* manufactured by Aqualon Company (Division of Hercules)Araldite * AY 103 manufactured by Ciba-Geigy
　　References
　　BRANDI BRANDI, Cesare, Teoria del restauro, Rome, Edizioni di Sto-ria e Letteratura, 1963.
　　ELIAS ELIAS, Mady, MENU, Michel, "Characterization of Surface States on Patrimonial Works of Art", in preprints of the Sur-face Modification Technologies (SMT) XIV , Paris, Septem-ber 2000.
　　GORDON GORDON, Eric, PHILLIMORE, Elisabeth, "The Treatment of two 13th century Chinese Wall Paintings in the Collection of the Royal Ontario Museum", in ICOM Committee for Conser-uation 7th Triennial Meeting, Copenhagen, September 10-14th 1984, Section 15, mural paintings and mosaics, p 12—14.
　　HANNA HANNA, Seamus, LEE, Nicholas, FOSTER, Geoffrey, "Three bodhisatvas: The Conservation of a 15th century Chinese Wall Painting in the British Museum Collection", in preprints for the UKIC 30th anniversary conference, 1988, London, p 130 - 134.
　　HANSEN HANSEN, Eric, LOWINGER, Rosa, SADOTT, Eileen, "Consolidation of Porous Paint in a Vapor - Saturated Atmo- sphere: A Technique for Minimizing Changes in the Appearance of Powdering, Matte Paint", in Journal of the American Insti- tute for Conservation, vol. 32, nl, Spring 1993, p 1 - 14.
　　LUK LUK, Cynthia, NEUMAN, Ingrid, MARTIN, James, KU- NIEJ BERRY, Cynthia, GREENFIELD, Judy, "The Treat- ment of two Chinese Wall Painting Fragments", in Conserua- tion of Ancient Sites on the Silk Road, International Confer- ence on the 
　　Conservation of Grotto Sites, October 1993, The Getty Conservation Institute, 1997, p 95-104.
　　MORA MORA, P, MORA, L, PHILIPPOT, P, Conservation of Wall Paintings, Butterworths, ICCROM, 1984.
　　PARIS PARIS, Emmanuelle, Etude et restauration d'une peinture murale chinoise conseruee au Music Guimet, Memoire de fin d'etudes IFROA, 2001, consultable ā la bibliotheque de I'IFROA, Paris.
　　PHILLIMORE 　　 PHILLIMORE, Elisabeth, "Initial Report on the Treatment of two 13th century Chinese Wall Paintings in the Royal Ontario Museum, Toronto", in AIC Meeting, Milwaukee,Wisconsin,March 26 - 30th 1982, p 150 - 159.
　　SIMONOT SIMONOT, Lionel, ELIAS, Mady, MENU, Michel, "Pig-ment recognition in works of art thanks to a speetrophotometric database", in preprints of the Colour and Visual Scales, Der- by, U-K, April 2000.
　　YUANZHONG YUANZHONG XU ( translated by), Song of the Immortals, an Anthology of Classical Chinese Poetry, Beijing, New York Press, 1994.
　　① YUANZHONG
　　② HANSEN and LUK
　　③ ELIAS
　　④ BRANDI
　　

唐墓壁画国际学术研讨会论文集/陕西历史博物馆编.--西安：三秦出版社，2006.10