The Textile Specialty Group Post Prints Papers Delivered at The

American Institute for Conservation of Historic & Artistic Works 41nd Annual Meeting Indianapolis, Indiana May 29-June 1, 2013

Edited by Amanda Holden Rebecca Summerour Emily Schuetz Julia Carlson Glenn Petersen

VOLUME TWENTY-THREE

AIC

AMERICAN INSTITUTE FOR CONSERVATIO N OF HISTORIC AND ARTISTIC WORKS POSTPRINTS of the Textile Specialty Group, Volume Twenty-Th ree, 2013, is published by the Textile Specialty Group of the American Institute for Conservation of Historic & Artistic Works. POSTPRINTS is distributed to members of the Textile Specialty Group. Additional copies may be purchased from the American Institute for Conservation of Historic & Artistic Works.

Papers presented in this publication have been edited for clarity and content but have not undergone a formal process of peer review. Responsibility for methods and/or materials described herein rests solely with the contributors and should not be considered offi cial statements of the Textile Specialty Group or AIC.

Textile Specialty Group Postprints Volume 23, 2013 ii CONTENTS

PREFACE i–vi

PAPERS PRESENTED

A GREEN SOLVENT FOR TEXTILE CONSERVATION?: INVESTIGATING THE USE OF CYCOSILOXANE D5 FOR TEXTILE CONSERVATION CLEANING JULIE BENNER, FRANCIS LENNARD, AND ANITA QUYE 1—16

TREATMENT OF A SUITE OF BAROQUE REVIVAL STYLE SEATING FURNITURE GENEVIEVE BIENIOSEK 17—32

ESTABLISHING A TEXTILE DYE ANALYSIS PROGRAM AT THE INDIANAPOLIS MUSEUM OF ART VICTOR J. CHEN, KATHLEEN KIEFER, NILOO PAYDAR, AND GREGORY D. SMITH 33—42

FINDING THE EASE: APPROACHES TO MOUNTING AND INSTALLATION AT THE ART INSTITUTE OF CHICAGO ISAAC FACIO AND LAUREN CHANG 43—56

AN OLD CASE OF NEW DISPLAY: CONTEMPORARY AND HISTORIC FASHION AT THE VICTORIA AND ALBERT MUSEUM JOANNE HACKETT AND KEIRA MILLER 57—74

EMERGENCE OF “ANTIQUE” SYNTHETIC TEXTILES EBENEZER KOTEI 75—92

MERGING DISCIPLINES: PREPARING A MATISSE SERIGRAPH FOR DISPLAY YADIN LAROCHETTE 93—108

NEW AND CURRENT MATERIALS AND APPROACHES FOR LOCALIZED CLEANING IN TEXTILE CONSERVATION ELIZABETH SHAEFFER AND JOY GARDINER 109—124

Textile Specialty Group Postprints Volume 23, 2013 iii CONTENTS PAPERS PRESENTED

FERROUS ATTRACTIONS: THE SCIENCE BEHIND THE MAGIC GWEN SPICER 125—140

DANCING ON A WIRE: ARTICULATION SOLUTIONS FOR MANNEQUINS IN THE CIRCLE OF DANCE EXHIBITION AT NMAI-NY SHELLY UHLIR 141—160

RENEWING THE PAST: PRESSURE MOUNTING TWO LARGE FRAGMENTED FLAGS JAN VUORI, RENÉE DANCAUSE AND STEFAN MICHALSKI 161—180 POSTER SESSION

RESPONSIBLE STEWARDSHIP: EXPLORING SUSTAINABILITY WITHIN CONSERVATION CHRISTIAN HERNANDEZ 181—194

WHEN LIFE GIVES YOU VELVET … PRESERVATION CONSIDERATIONS IN THE MAKING OF A PERIOD SHADOW BOX FRAME LAUREN ROSS AND MIRANDA HOPE DUNN 195—208

REMOVING MODERN ACCRETIONS: HOT-MELT ADHESIVE, CHEWING GUM, AND PRESSURE SENSITIVE TAPE REBECCA SUMMEROUR, SARAH JANE GRACE OWENS, SHANNON A. BROGDON-GRANTHAM, MARIAN KAMINITZ, AND SUSAN HEALD 209—222

Textile Specialty Group Postprints Volume 23, 2013 iv PREFACE

Th e twenty-third volume of POSTPRINTS contains papers presented at the Textile Specialty Group (TSG) session of the American Institute for Conservation of Historic & Artistic Works (AIC), in Indianapolis, Indiana from May 29–June 1, 2013.

TSG POSTPRINTS is a non-juried publication. Submission of these papers to juried publications, such as the Journal of the American Institute for Conservation, is encouraged. Th e papers chosen from abstracts submitted to the Meeting Chair, Virginia Jarvis Whelan, Textile Specialty Group Vice Chair for 2012–2013, are published as submitted by the authors. Editing of papers was done according to the Journal of the American Institute for Conservation’s “Guidelines for Authors” and the “Best Practices for Online PDF Publication: AIC Specialty Group Annuals & Postprints”, 2014 version. Materials and methods presented within the papers should not be considered offi cial statements of either the Textile Specialty Group, or of the American Institute for Conservation of Historic & Artistic Works.

Th e editors wish to thank the contributors to this publication for their cooperation. Without their enthusiasm and hard work this publication would not have been possible. Th anks are also extended to Aptara Inc. for layout of the papers, as well as Translation Cloud for translating the abstracts into Spanish. Christian Hernandez, Alba Fernandez-Keys, and Elizabeth Shaeff er graciously updated some of the abstracts prior to publication.

Textile Specialty Group Postprints Volume 23, 2013 v Textile Specialty Group Postprints Volume 23, 2013 vi A GREEN SOLVENT FOR TEXTILE CONSERVATION?: INVESTIGATING THE USE OF CYCOSILOXANE D5 FOR TEXTILE CONSERVATION CLEANING

JULIE BENNER, FRANCIS LENNARD, AND ANITA QUYE

ABSTRACT—Concerns about the health and environmental impacts of some solvents used in textile conservation have signaled the need for more environmentally friendly alternatives. At the same time, “green” cleaning solvents have begun to be marketed by the professional dry cleaning industry. One of these alternative solvents, decamethylcyclopentasiloxane (D5), may have potential for use in textile conservation, however no previous studies have shown how it may impact textile artifacts. A series of experiments was performed to test the eff ects of D5 on textiles and to examine its soil removal performance. Analysis of the eff ect of D5 on textile fi bers included Attenuated Total Refl ectance-Fourier Transform Infrared spectroscopy, tensile strength tests, and Scanning Electron Microscopy. Soil removal tests were analyzed using colorimetry and Attenuated Total Refl ectance-Fourier Transform Infrared Spectroscopy. In the results of the analysis, no appreciable diff erence in the condition and composition of treated and untreated samples could be detected. D5 was shown to have measureable eff ect on nonpolar soiling. Assessment of the overall results suggests that there is potential for the use of D5 for textile conservation cleaning.

UN SOLVENTE ECOLÓGICO PARA LA CONSERVACIÓN TEXTIL: RESUMEN—La preocupación por los impactos sobre la salud y el medioambiente de algunos solventes utilizados en la conservación textil ha signado la necesidad de buscar alternativas más ecológicas. Al mismo tiempo, la industria de la limpieza a seco profesional ha comenzado a desarrollar solventes “ecológicos”. Uno de estos solventes alternativos, el decametilciclopentasiloxano (D5) podría ser utilizado en la conservación textil, pero todavía no hay estudios que demuestren su impacto en los tejidos. Se ha realizado una serie de experimentos para probar los efectos del D5 en las telas y examinar su efi cacia en la remoción de la suciedad. Para el análisis del efecto del D5 en fi bras textiles se utilizó la espectroscopia de infrarrojos por transformada de Fourier-refl ectancia total atenuada, pruebas de resistencia a la tensión y microscopio electrónico de barrido. Las pruebas de remoción de suciedad fueron analizadas mediante colorimetría y espectroscopia de infrarrojos por transformada de Fourier-refl ectancia total atenuada. En los resultados de los análisis no se observó ninguna diferencia visible en el estado y la composición de las muestras tratadas y no tratadas. Se ha demostrado que el D5 tiene un efecto mensurable sobre la suciedad no polar. La evaluación de los resultados generales sugiere que el D5 podría ser utilizado como limpiador en la conservación textil.

1. INTRODUCTION

Th e use of solvents can be a powerful tool for the trained conservator in removing soils and stains from historic textiles, particularly soils of an oily or greasy character. Th ough eff ective as soil removers, organic solvents present risks to health and to the environment in their use and disposal. As the ubiquitous trend toward “going green” gains traction in the museum and heritage sector, institutions have begun to develop internal policies calling on greener practice for everything from air conditioning to offi ce supplies (Brophy and Wylie 2008). Conservators have a vested interest in using more environmentally friendly materials and methods: both ICON and the AIC explicitly invoke a conservator’s responsibility to the environment and to health and safety in their guidelines for professional ethics. A GREEN SOLVENT FOR TEXTILE CONSERVATION?: INVESTIGATING THE USE OF CYCOSILOXANE D5 FOR TEXTILE CONSERVATION CLEANING

While there is a need to identify alternatives to solvents used in conservation treatments, the fi eld of green chemistry has begun to off er substitutes for solvents for use in industry, including ionic liquids, supercritical fl uids, and cyclic siloxanes (DeSimone 2002). Th e last few decades of the twentieth century saw far-reaching international legislation restricting chlorinated solvent use in industry, in particular ozone- depleting chlorofl uorocarbons, phasing out their use in both developed countries and the developing world by 2010 (Durkee 2011).

2.1 ALTERNATIVE SOLVENTS IN DRY CLEANING Th e most widely used solvent in the dry cleaning industry, tetrachloroethylene, known to the trade as perchloroethylene or PERC, may be facing similar restriction in the near future. PERC, the dry cleaning industry standard for decades, has been shown to be toxic and carcinogenic, capable of causing cancer and neurological damage (US Environmental Protection Agency 2012). Th e use of PERC is already strictly regulated, but these restrictions may soon be growing more severe. In California, whose environmental laws are generally progressive and oft en predictive of future trends, the California Air Resources Board (CARB) has legislated to have PERC phased out of dry cleaners completely by 2023 (California Air Resources Board 2008). Dry cleaners have begun to embrace new technologies in anticipation of more widespread bans on the use of PERC. Th is is the reason that research into the developments of the dry cleaning industry is thought particularly valuable at this time.

2.2 THE NEED FOR ALTERNATIVE SOLVENTS IN TEXTILE CONSERVATION All organic solvents commonly used in textile conservation are thought be toxic to some degree, and many are thought to have environmental implications beyond immediate exposure. Th e hazardous nature of solvents requires strict health and safety precautions, including proper personal protection, ventilation and disposal. As conservators well know, these practices can be time-consuming and restrictive and must be weighed against the potential benefi ts of reducing oily or greasy soiling.

3. IDENTIFYING ALTERNATIVE SOLVENTS

Investigation into alternative solvents was undertaken at the Centre for Textile Conservation and Technical Art History at the University of Glasgow as a part of a fi nal year Master’s dissertation project. Th e research aimed to survey the developments in products and processes used by dry cleaners in order to identify a solvent as a viable candidate for a series of experimental trials assessing that solvent’s ability to remove soils, as well as its eff ect on textile fi bers. It was reasoned that, despite the vast diff erences between commercial garment cleaning and textile conservation, there may be potential for adapting dry cleaning technology for use in the textile conservation lab.

3.1 DECAMETHYLCYCLOPENTASILOXANE (D5) Th e candidate that was chosen as a result of this initial survey is a cyclic silicone-based polymer with the chemical name decamethylcyclopentasiloxane, known commonly as D5. D5 is used under the tradename GreenEarth® by independent dry cleaners and dry cleaning chains worldwide exclusively through licensing and support by the GreenEarth® company based in Kansas City, Missouri, founded in 1999. According to its website, GreenEarth® is the “world’s largest brand of green dry cleaning” (GreenEarth Cleaning n.d.). At the

Textile Specialty Group Postprints Volume 23, 2013 2 JULIE BENNER, FRANCIS LENNARD, AND ANITA QUYE time of this project, it was found to be the only commercial option for green dry cleaning available in Scotland. President Tim Maxwell of GreenEarth® provided 10 liters of GreenEarth® solvent free of charge for use in this study through UK chemical distributor Alex Reid. At the time of this study, there was no published literature regarding the use of D5 in textile conservation cleaning, though D5 has been tested in the use of microemulsions for the cleaning of painted surfaces (Stavroudis 2012). Th ere has also been recent research into the use of D5, among other non-aqueous solvents, in anti-oxidant and deacidifying treatments to stabilize historic textiles that are iron-tannate dyed (Wilson 2012).

3.1.1 Chemical Structure of D5 D5 is composed of a ring of oxygen and silicon atoms, with methyl groups branching from each silicon atom (fi g. 1). Th ough in structure it is similar to a cyclic hydrocarbon, the presence of the silicon atoms dis- tinguishes it from a classically organic hydrocarbon. Th e chemically symmetrical cyclic structure of D5, with its branching nonpolar methyl groups, indicates that there is no strong dipole moment in the molecule. Th e presence of hydrogen and oxygen throughout the molecule suggest that hydrogen bonding is very likely to occur. It was hypothesized that D5 creates secondary bonds with other molecules primarily through van der Waals and hydrogen forces, with little to no dipole interaction. Using the rule of thumb that ‘like dissolves like’, it can be predicted that D5 may have an eff ect on nonpolar soils. In GreenEarth® dry cleaning, D5 is used in a “charged system”, where water and detergents are added to the processing, presumably to reduce polar soils.

3.1.2 How Green Is D5? Th e EPA has not issued a formal risk assessment on the health eff ects of D5 as of this writing. Th e government of Canada has issued a statement that D5 is not considered to be a high priority for assessment of potential risks to human health (Government of Canada 2012). Th e Environment Agency UK, however, has

Fig.1. Chemical structure of D5, “Decamethylcyclopentasiloxane”. Sigma Aldrich, http://www.sigmaaldrich.com/catalog /product/aldrich/444278?langϭen®ionϭUS (Accessed May 2012)

Textile Specialty Group Postprints Volume 23, 2013 3 A GREEN SOLVENT FOR TEXTILE CONSERVATION?: INVESTIGATING THE USE OF CYCOSILOXANE D5 FOR TEXTILE CONSERVATION CLEANING completed an environmental risk assessment of D5, which is published on its website, stating D5 is both bio- accumulative and persistent in the environment, both in air and water, although the risks to the environment were assessed to be minimal (Brooke et al. 2009). Some studies have indicated that exposure to D5 may cause harm to reproductive and immune functions (California Offi ce of Envrionmental Health Hazard Assessment 2008). Consumers, however, are regularly exposed to low concentrations of D5 in cosmetics and pharmaceuticals. D5 is known as a cyclomethicone in personal products, such as lotion and makeup. Th e term cyclomethicone may indicate a pure cyclosiloxane, including D3, D4, and D6, or mixture of cyclic and linear siloxanes, including polydimethylsiloxane (PDMS). D5 is available from a range of manufacturers and chemical suppliers. GreenEarth® sources D5 from Shin-Etsu, GE® Silicones and Dow Corning® (Dow Corning has recently started marketing their silicones through the name Xiameter®). Th e MSDS for the three manufacturers addresses the safety precaution that D5 be used with adequate ventilation. Following this recommendation, all use of D5 in this study was carried out with the use of fume extraction. Nitrile gloves were worn at all times while handling.

4. EXPERIMENTATION

Th e experimental stage of the research assessed the performance of D5 in the conservation context. Th e testing was designed to show whether D5 would fi t a conservator’s criteria for treatment: providing eff ective soil removal without leaving behind detectable contaminant material or causing appreciable damage to fi bers. Th e experiment was divided into two stages. Th e fi rst stage examined the eff ects of immersion treatment with D5 on textile fi bers alone, without the addition of soils, and the second stage would attempt to measure the ability of D5 to remove soiling from textile substrates. For the fi rst stage, samples of lightweight plain weave cotton and wool were used. Th e diff erent fi ber types were chosen to examine how D5 might aff ect cellulosic and proteinaceous textiles. New fabrics were chosen to eliminate the variability inherent to sampling aged textiles. Th ough the ultimate aim of the project was to gather enough information to assess the possible use of D5 on textile artifacts, an examination of D5’s eff ect on new materials provides a foundation of understanding on the eff ects of D5, paving the way for further study that can take into account the variability of aged and degraded textiles. Half of the samples were treated with D5 and the other half remained as controls. Th e treated samples were placed in 1 L beakers containing 250 mL of D5 solution. Th e immersion treatment was designed to be similar to conservation solvent cleaning in an open bath, allowing the samples to remain immersed for an hour, with gentle agitation by stirring performed every 20 minutes. At the end of the hour, they were removed and dried overnight in the fume hood. D5 is described in the literature as odorless and colorless, and in practice this was observed to be true. With a lower surface tension than water (Table 1), the samples wet out readily. Drying time was not strictly measured in the experiment, but the general impression from the treatment is that D5 takes up to a full twenty-four hours to fully evaporate from surfaces, even under fume extraction. A subset of the treated and untreated samples were then exposed to artifi cial ageing in a 60ЊC lab oven. Due to restraints of time and resources, the ageing period was limited to seven days. During the ageing period, an open beaker of deionized water was placed in the oven and continually topped up in order to raise the overall RH.

Textile Specialty Group Postprints Volume 23, 2013 4 JULIE BENNER, FRANCIS LENNARD, AND ANITA QUYE

Table 1: Properties of decamethylcyclopentasiloxane1

Property Value Melting Point –38º C Boiling Point 211º C Density 0.954 g/cm3 at 25º C Surface Tension2 18.5 mN/m at 25º C Vapor Pressure 33.2 Pa at 25º C Water Solubility 17 μg/l at 23–25º C 1 Adapted from Brooke, D.N., et al. 2 Surface tension of water is 71.97 mN/m at 25º C

4.1 ANALYSIS In order to determine the eff ects of D5 treatment on the sample materials in each variable group, they were analyzed using tensile strength testing, attenuated total refl ectance-Fourier transform infrared spectroscopy (ATR-FTIR), and scanning electron microscope (SEM) analysis.

4.1.1 Tensile Strength Testing To compare tensile strength of treated and untreated samples, specimens of each variable group were cut from the samples and subjected to tensile strain tests. Fiber strength was measured by the amount of physical strain that could be applied before fi ber breakage. Th e amount of elongation exhibited by the fi ber at the maximum strain load was also measured. Specimens measuing 10 cm ϫ 2.5 cm were placed vertically within the grips of an Instron® 5544 tensile testing machine running Bluehill® soft ware to record results. During the test, the Instron® apparatus fi tted with a 1000N load cell stretched the fi bers at a rate of 10 mm per minute until the point of rupture. Data was compiled from the measurements and the variable groups were compared, showing little signifi - cant diff erence in the mean tensile strength (fi g. 2) and elongation values (fi g. 3) between the treated and untreated samples.

4.1.2 ATR-FTIR ATR-FTIR spectroscopy was used to determine if diff erences in molecular composition could be detected between the treated and untreated variable groups. Samples were analyzed using a Perkin Elmer® Spectrum One spectrometer with Universal Sampling Accessory. Th e resulting spectra showed no signifi cant diff erence between the treated and untreated samples. Th ere is relative consistency in frequency, shape and intensity of the spectra for both treated and untreated, aged and unaged samples (fi gs. 4 & 5). Th is result suggests that treatment with D5 caused no detectable change to the molecular composition of the samples.

4.1.3 SEM SEM was used to analyze a small range of of unaged samples to assess whether D5 treatment causes a physical change to the fi ber surfaces. Samples from treated and untreated cotton and wool testing specimens, each under 1 mm in length, were placed on a nickel stage. Th e stage was then placed in a Cressington

Textile Specialty Group Postprints Volume 23, 2013 5 A GREEN SOLVENT FOR TEXTILE CONSERVATION?: INVESTIGATING THE USE OF CYCOSILOXANE D5 FOR TEXTILE CONSERVATION CLEANING

Fig. 2. Comparison of tensile strength and elongation measurements of sample variable groups, cotton.

Fig.3. Comparison of tensile strength and elongation measurements of sample variable groups, wool.

Textile Specialty Group Postprints Volume 23, 2013 6 JULIE BENNER, FRANCIS LENNARD, AND ANITA QUYE

Fig. 4. Comparison of results from ATR-FTIR analysis on sample variable groups, cotton.

Textile Specialty Group Postprints Volume 23, 2013 7 A GREEN SOLVENT FOR TEXTILE CONSERVATION?: INVESTIGATING THE USE OF CYCOSILOXANE D5 FOR TEXTILE CONSERVATION CLEANING

Fig. 5. Comparison of results from ATR-FTIR analysis on sample variable groups, wool.

Textile Specialty Group Postprints Volume 23, 2013 8 JULIE BENNER, FRANCIS LENNARD, AND ANITA QUYE

108/CarbonA carbon coating machine where the surface was sputter-coated with a relatively thin layer of carbon. Aft er a baseline was run using a nickel standard, the stage with the carbon coated samples was placed in the vacuum chamber of the scanning electron microscope (Model: JEOL® JSM-6480LV, running INCA soft ware). Th e samples were analyzed individually, each at low energy (25 kv) with a working distance of 10 nm. Sec- onday Electron Imaging (SEI) was used to take still images, captured at 500ϫ magnifi cation (fi gs. 6–9). Ele- mental analysis was performed with energy dispersive x-ray spectrometry (EDS), providing both graphic spectra and quantitative measurements of atomic composition. No physical or elemental diff erences between the treated and untreated samples were detected by the SEM analysis.

5. SOIL REMOVAL TESTING

In the second stage of the experiment, samples of standardized soiled fabrics were treated with D5 and were analyzed before and aft er treatment using a combination of colorimetry and ATR-FTIR spectroscopy. Standardized soiled fabrics, acquired from Materials Research Products, LTD, were chosen to make up the sample population. Both the cotton and wool were purchased pre-soiled with a combination of olive oil and carbon black. This combination of soils was chosen from a host of soiling options and was selected for economy, availability across fabric types, and for the composition of the soiling itself. It was thought that carbon black and olive oil would allow for testing D5’s effectiveness on both polar and nonpolar soil. Th e obvious disadvantage to the use of standardized soiled fabric is that both the fabric itself and the soiling are new. By using new fabric, the tests would not be able to fully replicate the conditions of age and soiling

Fig. 6. Untreated cotton fi bers at 500ϫ magnifi cation.

Textile Specialty Group Postprints Volume 23, 2013 9 A GREEN SOLVENT FOR TEXTILE CONSERVATION?: INVESTIGATING THE USE OF CYCOSILOXANE D5 FOR TEXTILE CONSERVATION CLEANING

Fig. 7. Cotton fi bers treated with D5 at 500ϫ magnifi cation.

Fig. 8. Untreated wool fi bers at 500ϫ magnifi cation.

Textile Specialty Group Postprints Volume 23, 2013 10 JULIE BENNER, FRANCIS LENNARD, AND ANITA QUYE

Fig. 9. Wool fi bers treated with D5 at 500ϫ magnifi cation. in historic textiles. Using standardized soiling ensured that the composition would be of a known quantity and would be distributed evenly throughout the samples. Th e samples were treated with D5, following the same immersion process performed in the fi rst stage of experimentation.

5.1 COLORIMETRY

Th e standardized soiled cotton and wool fabrics were dark gray in color before treatment, aft er which they appeared visibly lighter. In order to make a quantitative measurement of this change, color value readings of the samples were taken with a Konica-Minolta® C-Series Chromameter. Th ree readings of each sample were taken before and aft er treatment. A Mylar® template was used to ensure constistency in testing locations. Th e results from the colorimetry tests showed a measureable diff erence in the mean color values aft er treatment with D5, particularly between the L, or luminosity values, confi rming the color change that could be visually observed (fi g. 10)

5.2 ATR-FTIR ATR-FTIR analysis was performed before and aft er treatment to assess the eff ect of D5 on the molecular composition of the soiled samples. For both cotton and wool fabrics, there are distinct diff erences in the resulting spectra before and aft er treatment (fi gs. 11 & 12) Th ese diff erences correspond to spectral peaks characteristic of reference spectra for olive oil, at 2900 cm-1 and between 1800-1700 cm-1 (fi g. 13), which suggests that compounds in the olive oil portion of the soiling were soluble in the D5 solution, and were reduced by the immersion treatment.

Textile Specialty Group Postprints Volume 23, 2013 11 A GREEN SOLVENT FOR TEXTILE CONSERVATION?: INVESTIGATING THE USE OF CYCOSILOXANE D5 FOR TEXTILE CONSERVATION CLEANING

Fig. 10. Mean chromameter readings, for cotton and wool samples before and aft er treatment.

Fig. 11. Spectral comparison of pre-soiled cotton samples before and aft er treatment with D5.

Textile Specialty Group Postprints Volume 23, 2013 12 JULIE BENNER, FRANCIS LENNARD, AND ANITA QUYE

Fig. 12. Spectral comparison of pre-soiled wool samples before and aft er treatment with D5.

Fig. 13. Reference spectra for olive oil, from “Spectral Database for Organic Compounds, No.: 2740, Compound Name: Olive Oil, IR: Liquid Film”. http://sdbs.db.aist.go.jp/sdbs (Accessed June 2015).

Textile Specialty Group Postprints Volume 23, 2013 13 A GREEN SOLVENT FOR TEXTILE CONSERVATION?: INVESTIGATING THE USE OF CYCOSILOXANE D5 FOR TEXTILE CONSERVATION CLEANING

6. CONCLUSIONS

Comparative analysis of unsoiled sample fabrics showed no measureable diff erence between controls and those treated with D5. Th is confi rms the expectation that, as a volatile solvent, D5 evaporates completely from fabrics and leaves behind no material residue. When tested on soiling of known character, results indicate that it signifi cantly reduced oily soil on cotton and wool fabrics. Th ese results suggest that D5 may be a suitable choice for reducing non-polar soils on historic textiles, in cases where solvent cleaning is appropriate for the treatment at hand. An area of future study in textile conservation may be to investigate how D5 can be used a charged system, with water or surfactant added for the possible reduction of more polar soils. As a low volatil- ity solvent that is approved for use in personal care products, D5 may present a reduced health and safety risk, however studies are not conclusive on its eff ects from exposure over time. It is still recommended to use PPE and ventilation when handling D5.

ACKNOWLEDGEMENTS

Th e authors would like to acknowledge Karen Th ompson and Sarah Foskett at the Centre for Textile Conser- vation and Technical Art History for vital assistance with the project, as well as Ailsa Boyd, Peter Chang and Liz Tanner at the University of Glasgow. Th e project would not have been possible without the help of Dr. Margaret Smith. Additionally, the authors are very grateful to Joe Barabe of the McCrone Group for graciously facilitating the SEM analysis. Tim Maxwell of GreenEarth® generously provided the D5 used in this study.

REFERENCES

Brooke, D. N., M.J. Crookes, D. Gray and S. Robertson. 2009. Environmental risk assessment report: decamethylcyclopentasiloxane. Environment agency. http://publications.environment-agency.gov.uk/PDF/ SCHO0309BPQX-E-E.pdf (accessed 8/15/12).

Brophy, S. S. and E. Wylie. 2008. Th e green museum. Lanham, Maryland: AltaMira Press.

California Air Resources Board. 2008. Fact sheet: dry cleaning alternative Solvents: health and environmental impacts, March 2008. www.arb.ca.gov/toxics/dryclean/alternativesolvts_e.pdf (accessed 3/20/2012).

California Offi ce of Envrionmental Health Hazard Assessment. 2008. Cyclosiloxanes: materials for the December 4-5, 2008 meeting of the California Environmental Contaminant Biomonitoring Program (CECBP). California offi ce of envrionmental health hazard assessment. December 4-5, 2008. http://oehha. ca.gov/multimedia/biomon/pdf/1208cyclosiloxanes.pdf (accessed 8/18/2012).

DeSimone, J. M. 2002. Practical approaches to green solvents. Science. 297: 799–803.

Durkee, J. B. 2011. Cleaning with solvents. In Developments in surface contamination and cleaning, ed. R. Kohli and K.L. Mittal. Norwich, NewYork: William Andrew.

Government of Canada. 2012. Siloxane D5 (cyclopentasiloxane, decamethyl-). Chemical substances. www. chemicalsubstanceschimiques.gc.ca/challenge-defi /summary-sommaire/batch-lot-2/541-02-6-eng.pho (accessed 07/25/12).

Textile Specialty Group Postprints Volume 23, 2013 14 JULIE BENNER, FRANCIS LENNARD, AND ANITA QUYE

GreenEarth Cleaning. n.d. About us. www.greenearthcleaning.com/?pageϭAboutUs (accessed 04/19/12). Stavroudis, C. 2012. More from CAPS3: Surfactants, silicone-based solvents, and microemulsions.WAAC Newsletter. 34(3): 24–27. US Environmental Protection Agency. 2012. Exisiting chemicals: perchloroethylene fact sheet. http://epa.gov/ oppt/existingchemicals/pubs/perchloroethylene_fact_sheet.html (accessed 08/13/12). Wilson, H. 2012. Investigation of non-aqueous remedial treatment for iron-tannate dyed silk textiles. Ph.D. thesis, University of Manchester. www.escholar.manchester.ac.uk/api/datastream?publicationPidϭuk-ac- man-scw:191168&datastreamIdϭFULL-TEXT.PDF (accessed 07/25/2013).

SOURCES OF MATERIALS Materials Research Products 4 Montpelier Street (#236) London, England SW7 1EX Phone: ϩ44 (0) 20-7823-4146 www.mrpltd.com Alex Reid HQ 9 Ashville Way Whetstone Leicester LE8 6NU Phone: ϩ44 (0) 84-5634-4454 Fax: 0116 275 3838 www.alexreid.com GreenEarth® Dry Cleaning 51 West 135th Street Kansas City, MO 64145-1289 Phone: (877) 926-0895 www.greenearthcleaning.com

AUTHOR BIOGRAPHIES

JULIE BENNER is a 2012 graduate of the University of Glasgow Master’s Program in Textile Conservation. She has served as Assistant Conservator of Textiles at the Chicago History Museum and as an intern at the Field Museum of Natural History and the Art Institute of Chicago. From 2013-2015, she was the Andrew W. Mellon Fellow in Textile Conservation at the Denver Art Museum and currently works in private practice in Denver. Email: [email protected]

FRANCES LENNARD worked for 15 years as a textile conservator in the UK, at the Textile Conservation Centre and in private practice before taking on the role of Convenor of the MA Textile Conservation program at the TCC, University of Southampton. She is currently Senior Lecturer in Textile Conservation at the

Textile Specialty Group Postprints Volume 23, 2013 15 A GREEN SOLVENT FOR TEXTILE CONSERVATION?: INVESTIGATING THE USE OF CYCOSILOXANE D5 FOR TEXTILE CONSERVATION CLEANING

University of Glasgow and Convenor of the MPhil Textile Conservation. She is the co-editor of Tapestry Con- servation: Principles and Practice (with Maria Hayward, 2006) and Textile Conservation: Advances in Practice (with Patricia Ewer, 2010). Address: School of Culture and Creative Arts, University of Glasgow, 8 University Gardens, Glasgow G12 8QH, UK. Email: [email protected]

DR. ANITA QUYE is the Lecturer in Conservation Science at the Centre for Textile Conservation and Technical Art History (CTCTAH), University of Glasgow. Her research specialism is micro-chemical analysis of historical dyes, plastics and organic residues, and her current projects include early synthetic dyes, semi- synthetic fi bers, color preservation, and reconstructions of 19th century textile production. Prior to her lectureship, Anita was the analytical organic chemist at National Museums Scotland. Her extensive publications include books on conservation science for historical tapestries and plastics preservation. Address: School of Culture and Creative Arts, University of Glasgow, 8 University Gardens, Glasgow G12 8QH, UK. Email: [email protected]. Web: http://www.gla.ac.uk/schools/cca/staff /anitaquye

Textile Specialty Group Postprints Volume 23, 2013 16 TREATMENT OF A SUITE OF BAROQUE REVIVAL STYLE SEATING FURNITURE

GENEVIEVE BIENIOSEK

ABSTRACT—A suite of Italian Baroque revival style upholstered seating furniture, composed of 12 armchairs and two settees, are part of the collection at Biltmore, George Washington Vanderbilt’s house and estate in Asheville, North Carolina. Th e pieces are carved in the manner of Italian sculptor Andrea Brustolon, and likely date to the nineteenth century. Part of the suite had been reupholstered in the 1970s using modern materials. Th e remaining pieces retained their original upholstery, including their gauff rage wool plush showcovers. Four chairs and one settee were treated for exhibit using minimally-intrusive upholstery techniques and new, reproduction gauff rage fabric and mocquette braid trim.

TRATAMIENTO DE UNA SUITE DE SILLONES DE ESTILO NEOBARROCO: RESUMEN—Una suite de sillones tapizados de estilo neobarroco italiano compuesta por 12 sillones individuales y dos sofás, forman parte de la colección de Biltmore, de la casa y patrimonio de George Washington Vanderbilt en Asheville, Carolina del Norte. Las piezas están talladas al estilo del escultor italiano Andrea Brustolon, y probablemente daten del siglo 19. Parte de la suite fue retapizada en los años setenta con materiales modernos. El resto de las piezas conserva su tapicería original, incluyendo las fundas de fi eltro gofrado. Para la exhibición, se trataron cuatro sillones y un sofá utilizando técnicas de tapizado mínimamente intrusivas, reproducciones de telas gofradas nuevas y pasamanería mocquette.

1. INTRODUCTION AND BACKGROUND

Biltmore house is a gilded age mansion completed in 1895 in the mountains of western North Carolina by George Washington Vanderbilt, the youngest grandson of “the Commodore” Cornelius Vanderbilt. Th e house was designed by Richard Morris Hunt, and the grounds by Frederick Law Olmsted. Biltmore was opened for public tours starting in 1930, and in 1963 it was designated as a National Historic Landmark. Th e house and estate are still owned by descendants of George W. Vanderbilt, and are operated as a privately owned, for- profi t company with around 1800 employees. Biltmore is open 365 days a year, and hosts over one million visitors annually. George W. Vanderbilt traveled extensively throughout his life, and amassed a signifi cant collection of books, art, and furniture from around the world. Th e majority of his collection remains at Biltmore, which currently comprises approximately 23,000 books and 28,000 objects. Many furnishings were purchased or created specifi cally for Biltmore; he also inherited and retained a portion of his father’s, (William Henry Vanderbilt), extensive art collection. Most of the rooms in the house are interpreted for the period of Vanderbilt’s residence (1895–1914), using original, collection objects. One room, the second floor Living Hall, was deinstalled in 2008 and held a temporary exhibit about the process of preservation and conservation at Biltmore. In preparation for its 2013 restoration as a period room, the history and original furnishings were researched. The old- est image found of the second floor Living Hall dates from the early 1950s (fig. 1). By this point, the family was no longer living in the rooms, and the house had been open to tours for many years. This photograph shows the room staged to display items from the collection, such as the vases and a sedan TREATMENT OF A SUITE OF BAROQUE REVIVAL STYLE SEATING FURNITURE

Figure 1: Postcard showing the second fl oor Living Hall (“Upstairs Hall”); image dates to the 1950’s

chair, alongside furniture including the ornately carved, upholstered chairs and settees placed around the room. Th ere are 12 of these matching chairs and the two settees currently in the collection. Unfortunately, no information about them dating to Mr. Vanderbilt’s time has been found. Th e suite was likely made in the 19th century, based on observations of the upholstery and overall condition. Th e style is reminiscent of the work of 17th century Italian sculptor Andrea Brustolon, who infl uenced other furniture made in the mid-to-late 19th century. From the suite, one settee and eight of the chairs were previously reupholstered for display in the fi rst fl oor Music Room. Th e project of fi nishing the never-completed Music Room was undertaken in 1976 by Vanderbilt’s grandson William A. V. Cecil, the current owner of Biltmore. Th ese reupholstered pieces remain an integral part of the room, so the chairs and settee currently on display were not treated as part of the project. One settee and four of the remaining six chairs in storage were selected for display in the second fl oor Living Hall (fi gs. 2-4). Th e settee and three of the chairs had their original upholstery, and one had been reupholstered but was no longer displayed in the Music Room(fi g. 5). One chair with the most intact original upholstery was left untreated in storage for future reference.

Textile Specialty Group Postprints Volume 23, 2013 18 GENEVIEVE BIENIOSEK

Figure 2: Settee before treatment, front view

Figure 3: One of the chairs before treatment, Figure 4: One of the chairs before front view treatment, back view

Textile Specialty Group Postprints Volume 23, 2013 19 TREATMENT OF A SUITE OF BAROQUE REVIVAL STYLE SEATING FURNITURE

Figure 5: One of the chairs reupholstered in the 1970’s, before treatment, front view

2. ANALYSIS OF THE CHAIRS AND SETTEE

2.1 UPHOLSTERY Th e investigation of the upholstery materials and techniques was primarily carried out by senior upholstery conservator Anne Battram. Th e chairs and settee are upholstered with a dead seat and back, and have a fairly severe, only slightly curved profi le. Th e original showcover fabric is gauff rage, which is also described as Utrecht Velvet, or simply “embossed wool plush.” Gauff rage is created by taking velvet or plush fabric and pressing the nap fl at to produce a relief pattern, using hot metal rollers that have a reverse impression of the design. Depending on the profi le of the metal roller, the nap can be pressed to diff erent heights and patterns. Gauff rage mimics the appearance of voided or cut and uncut velvets, but is simpler and faster to produce. Th e gauff ering process may be applied to any type of cloth (with or without nap) to create texture (Bast 1946; Grier 1988; Trench 2000, 186). Th e stamped pattern is sensitive to moisture, so wet cleaning is not an option. Th e showcover fabric on the seat of the settee had almost none of the stamped pattern remaining when it was treated, and a dark tideline around the edges confi rms that water most likely caused the disappearance.

Textile Specialty Group Postprints Volume 23, 2013 20 GENEVIEVE BIENIOSEK

Th ree similar types of plush fabric were used for the original showcover; all have a linen ground with a wool nap. Th e main fabric used for the chairs is wool plush gauff rage with a fairly coarse weave (thread count: warp 38 threads per inch, weft 49 threads per inch), and the selvedges are the same as the ground. Th e settee showcover is wool plush gauff rage that has a much tighter weave (thread count: both warp and weft 72 threads per inch), and the selvedges are white with two blue stripes. Th e stamped pattern appears to be identical for both types of fabric. Th is was a popular pattern that is very common, and was produced by more than one company. Th e exposed showcover was very dirty, and had also faded or undergone a color shift , which was more pronounced on the gauff rage with a coarser weave. During deupholstery, it became apparent that the fabric was once bright gold (fi g. 6). Th e chair showcovers also incorporate smaller pieces of unstamped wool plush used in unobtrusive areas on the sides of the seat, and on the back underneath the arms. Th is unstamped plush is identical to the more tightly-woven gauff rage used on the settee. In a few sections, multiple small pieces of both types of fabric were stitched together to make up the cover; some are less than an inch wide. Most of these pieces are seamed with

Figure 6: Detail of chair during deupholstery, showing the original color of the showcover and trim

Textile Specialty Group Postprints Volume 23, 2013 21 TREATMENT OF A SUITE OF BAROQUE REVIVAL STYLE SEATING FURNITURE butt joins, with narrow strips of fabric added to back longer joins. It is unclear why there are three diff erent types of fabric, and why so many small pieces were patched together on the chair showcovers. All of the work seems to be original, rather than later repairs, because there are no extra tack holes or stitching in the pieced areas. Perhaps the upholsterer ran out of fabric to cover the entire suite and needed to piece together bits trimmed from larger sections. Th e stamped pattern was not positioned quite the same on each chair, which also could be an attempt to save fabric. In some spots the small patches might have been added to correct for a piece accidentally cut too small (an occurrence this upholsterer has experienced). Th e showcover and cotton skimmer were intact except for a tear at the front of the seat on each chair. Th ese tears occurred right where the fabric would rub against the top of the seat rail because there is no edge roll to provide extra padding. Th e tears allowed some access to the underupholstery, which is a simple fl at cake of grass (specifi c type was not identifi ed). Th e cake likely also contains springier horsehair in the center, but this could not be accessed. Th e backs and seats are supported by webbing and a burlap fabric webbing cover (both bast fi ber). Th e original outback fabric was a beige plain weave linen fabric. Most of the original outback on the chairs was lost, having worn through on the unpadded wood back, but fragments remained along the sides and top. Th e settee outback was in much better condition, with only a few tears. Mocquette braid trim (a fl at plush trim, with linen ground and wool nap) was used for both the settee and chairs. Th e trim was stitched to the showcover around the outside edge of the seat and glued to the seat rails, and just stitched along the outside edges of the chair backs. Th e trim was also secured at corners with brass fi nishing nails.

2.2 CONSTRUCTION Th e chairs and settee are designed with removable parts, which greatly simplifi es construction and upholstery (fi g. 7). On the chairs, the decoratively carved crest rail attaches using a metal spring clip which snaps down into a metal slot, similar to a lock plate, on the top of the back. Th ere are smooth-sided, cylindrical metal posts extending up on each side of the top of the back that fi t into holes in the underside of the detachable crest. Th e removable busts are secured atop each arm with buttress-threaded screws that fi t through holes in the stiles. Th e entire back panel of the settee is designed to be removable, and like the chairs, the busts on top of the arms also unscrew (fi g. 8). Th e upholstered back is held in the frame by three turnbuckles. A generous assort- ment of modern nails had also been added over the years. Th e chairs and settee have very similar construction; they use mainly mortise and tenon joints. For the settee, the mortise and tenons joining the x-stretcher to the legs are reinforced by large screws. Th e arm supports attach with pegs through to the top of the arms, pounded fl at and worked into the carved design. On all four chairs, the mortises on the back stiles were cut extra large, and the arm tenons shimmed into place with several bits of wood (fi g. 9). Th is may have been done to ensure the arms would be level, as the elaborate carving at the front arm support couldn’t easily be moved. By cutting the mortises large, the arm position could be adjusted by simply changing the size of the shims. All the pieces have previously been treated for insect infestation, and there are fl ight holes in the wood. Otherwise the woodwork was in good condition, but dirty. Th ere were a few scattered losses on the feet and projecting ornaments, some of which had been previously replaced with carved wood fi lls. Th e settee also had old damage to the proper right back leg, which had been repaired several times.

Textile Specialty Group Postprints Volume 23, 2013 22 GENEVIEVE BIENIOSEK

Figure 7: Th e detached crest rail and busts from a chair

Figure 8: Th e deupholstered settee with back and busts removed; the original skimmer is visible on the seat

Textile Specialty Group Postprints Volume 23, 2013 23 TREATMENT OF A SUITE OF BAROQUE REVIVAL STYLE SEATING FURNITURE

Figure 9: Detail, left side of chair outside back showing the shimmed mortise and tenon joint; also note scribed lines from marking the mortises, and the hole used to secure the detachable bust

3. TREATMENT

Th e treatments aimed to stabilize the furniture and restore a nearly-new appearance, as they would have looked when Vanderbilt lived at Biltmore. Although it is unlikely the furniture will ever be used for seating, the treatment also needed to be fairly robust. Th e pieces will be displayed in the historic house, which is not climate controlled. In warm weather, rooms are ventilated by opening (screened) windows. Due to the high visitation, objects must be cleaned and handled frequently.

Textile Specialty Group Postprints Volume 23, 2013 24 GENEVIEVE BIENIOSEK

Considering the condition of the original gauff rage showcover and its sensitivity to water, and the fact that one chair no longer had any remaining original show fabric, the decision was made to have new reproduction showcover and trim produced to reupholster the pieces. Treatment of the settee and chairs was carried out by Anne Battram, Nancy Rosebrock, and the author. Th e pieces were photographed and vacuum-cleaned. Th e original trim, showcover and outback were removed and placed in storage, along with samples of the removed tacks and nails. Th e skimmer and under upholstery were left intact. Every detail possible was documented during the deupholstery process, to serve as a reference for how to reupholster the pieces in the same manner. Also, as each nail or tack was removed, the hole was lightly marked with chalk. When the all the upholstery layers were removed, there were no unmarked holes visible from any earlier upholstery campaign, indicating the upholstery is original. Aft er each piece had been deupholstered down to the original skimmer, the wood was treated. All the fi nished surfaces were cleaned using water and saliva to remove accumulated dirt and grime. For many years the house was heated by a coal furnace and fi replaces, so similar dirt is found on most of the collection items that come to the lab. Interestingly, the dirt had also selectively fi ltered through the unstamped areas of the gauff rage fabric, leaving a shadow impression of the pattern on the inside back skimmers of the chairs. Major losses in the wood were fi lled using Araldite Epoxy 1253 over a hide glue barrier. eTh damaged settee leg was completely removed to reverse older, inadequate repairs, and then reattached, using Araldite Epoxy 1253 over hide glue barrier. Th e original screw securing the mortise and tenon joint connecting the leg to the x-stretcher was reused by waxing and inserting it into the original hole in the joint before the epoxy had cured. A small plug of Volara foam and a layer of Araldite were used to imitate the original putty fi ll used to disguise the screw head. Insect fl ight holes were filled using pigmented wax. Scratches and areas where the fi nish had worn away were inpainted with Charbonnel Restoration Colors, then all the fi nished surfaces were coated with paste wax and hand buff ed. Aft er treatment of the wood was complete, the pieces were reupholstered, trying to recreate the original upholstery appearance as much as was possible and practical. First a new linen skimmer was added over the existing skimmer. Th is layer acts as an isolating barrier that helps protect the original under upholstery from the new fabric, and also make it less likely dirt will migrate out into the new showcover. Th e linen also functions as sewing base for the attachment of all the upper layers. New, medium weight, bleached linen from Testfabrics was used for the skimmer. Th e linen was attached with new tacks inserted into the existing tack holes, and in a few places with new staples because there was no available hole. It was possible to reuse the tack holes in this situation because the wood was still in very good shape, with holes from only one upholstery campaign. In this situation, tacks will provide a strong, reliable attachment. Th e linen sewing base can also be re-used as the sewing base to attach the next round of reuphol- stery without needing to add more holes into the wood. At the front of the seat where the stuffi ng was worn, polyester batting was added to recreate a more appropriate profi le. A pocket of linen was stitched onto the sewing base and fi lled with polyester batting (fi g. 10). Th e gauff rage showcover was reproduced by Prelle, a 260-year-old company based in Lyon, France. Th ey wove a wool plush fabric with linen ground and had it pressed to match the original gauff rage (fi g. 11). A smaller amount of the plush was left unstamped to be used to duplicate the original appearance of the chairs.

Textile Specialty Group Postprints Volume 23, 2013 25 TREATMENT OF A SUITE OF BAROQUE REVIVAL STYLE SEATING FURNITURE

Figure 10: Detail of chair seat with linen and batting pinned in place

Figure 11: Detail of reproduction gauff rage on the chair; there are actually three heights: the unstamped sections, the fl at background with a subtle striped texture, and an intermediate, partly pressed level that adds extra detail and depth to the pattern

Textile Specialty Group Postprints Volume 23, 2013 26 GENEVIEVE BIENIOSEK

Figure 12: Detail of chair side back showing the unstamped plush pieced to replicate the appearance of the original upholstery

In order to reproduce the original appearance, the new showcover was positioned to match the original pattern placement on the chair as closely as possible, even though this resulted in small sections of waste. The cut fabric pieces were seamed together using a sewing machine to create the seat and inside back panels. These were pinned in place, and then stitched onto the new linen skimmer sewing base using a running backstitch. Areas of unstamped plush in the original showcover were also duplicated (fig. 12).

Textile Specialty Group Postprints Volume 23, 2013 27 TREATMENT OF A SUITE OF BAROQUE REVIVAL STYLE SEATING FURNITURE

Figure 13: Settee aft er treatment

For the trim, reproduction un-cut mocquette braid woven by Heritage Trimmings was used. Th e trim was pinned in place and stitched to the seat fabric along its top edge. Along its bottom edge it was attached to the wood of the seat rail using hot melt glue. For the outback, a plain weave, linen blend fabric woven by Humphries Weaving Company was used. Th e outback for the settee was pinned and stitched in place. At this point, the removable back and busts were replaced on the settee to complete the treatment (fi g. 13). On the chairs, the outback fabric was pinned in place and sewn along the top edge. Th e removable crest rail and busts were reattached (fi g. 14). eTh outback was then sewn down around the sides and bottom, covering the heads of the screws securing the detachable busts (fi g. 15). Finally, the reproduction trim was sewn down the sides of the back, completing the treatment. Th e chairs and settee were installed in the restored second fl oor Living Hall, where they are displayed on custom wood platforms designed to provide extra protection from accidental contact.

CONCLUSIONS

Th e treatments were able to successfully stabilize the original materials and achieve a nearly-new appearance. In this case, the use of historic textile production techniques to create the reproduction gauff rage show fabric resulted in restoration of the upholstery’s original appearance. However, modern reproduction

Textile Specialty Group Postprints Volume 23, 2013 28 GENEVIEVE BIENIOSEK

Figure 14: Chair aft er treatment, front view Figure 15: Chair aft er treatment, back view

products, such as digitally printed fabrics, might be a viable solution for other furniture. As many readers are aware, ordering any type of custom reproducing materials is a time consuming process, including the time needed to produce and approve samples for matching color and overall appearance, and should be factored into treatment plans. It was very fortunate to have such a large suite with multiple examples of original upholstery to study, and to be able to leave an example unrestored. Questions remain about the use of multiple show fabrics and elaborate piecing; perhaps future investigation of the untouched chair and the original upholstery that was removed will shed light on this or other historic upholstery examples.

ACKNOWLEDGEMENTS

Th e author is grateful for the expertise, support and assistance of Nancy Rosebrock, Anne Battram, Darren Poupore, and the entire staff at Biltmore. Great thanks are due to the Cecil family and the Biltmore Company for preserving an amazing group of objects. Generous fi nancial support from Biltmore and the Foundation of the American Institute for Conservation of Historic & Artistic Works (FAIC) George Stout Memorial Fund made presentation of this project possible.

Textile Specialty Group Postprints Volume 23, 2013 29 TREATMENT OF A SUITE OF BAROQUE REVIVAL STYLE SEATING FURNITURE

REFERENCES

Bast, H. 1946. New Essentials of Upholstery. Milwaukee, Wisconsin: Th e Bruce Publishing Company.

Carley, R. D. and R. G. Rennicke 2008. A Guide to Biltmore Estate. Asheville, North Carolina: Th e Biltmore Company.

Grier, K. C. 1988. Culture & Comfort People, Parlors, and Upholstery 1850–1930. Rochester, New York: Th e Strong Museum.

Trench, L., ed. 2000. Materials and Techniques in the Decorative Arts. Chicago, Illinois: Th e University of Chicago Press.

MATERIALS Wool/linen gauff rage is from Prelle 7 Rue Barodet 69004 Lyon France Tel (33) 472 10 11 40 Fax (33) 472 10 11 41 www.prelle.fr/en/

For the new outback, plain woven linen blend fabric (fi ber content 48% viscose, 36% cotton and 16% fl ax) was used from Th e Humphries Weaving Co. Ltd Ashburton Lodge, 64 Cornard Road Sudbury, Suff olk CO10 2XB UK Tel 01787 466670 Fax 01787 466671 www.humphriesweaving.co.uk

New, medium weight, bleached linen used for the new skimmers was purchased from Testfabrics 415 Delaware Avenue PO Box 26 West Pittson, PA 18643 Tel: (570) 603-0432 Fax: (570) 603-0433 www.testfabrics.com

Textile Specialty Group Postprints Volume 23, 2013 30 GENEVIEVE BIENIOSEK

Reproduction trim is 1-3/8” wide un-cut mocquette braid from Heritage Trimmings Ltd Th e Old Mess Room, Colombo Street Derby DE23 8LW UK Tel 01332 343953 Fax 01332 298443 www.heritagetrimmings.co.uk

Other upholstery supplies available at local fabric stores or through B&M Upholstery Supply, LLC 332 Geer Road Lebanon, CT 06249 Tel: (860) 886-1884 Fax: (860) 886-4209 www.bandmupholsterysupply.com

Charbonnel Restoration Colors by Lefranc & Bourgeois are no longer manufactured. Araldite Epoxy 1253 manufactured by Huntsman Advanced Materials Americas Inc. Available from Conservation Resources International, LLC 5532 Port Royal Road Springfi eld, VA 22151 Tel: (800) 634-6932 Fax: (703) 321-0629 www.conservationresources.com

J.E. Moser’s Paste Wax, Premium Dark Brown color (903-244) distributed by Woodworker’s Supply, Inc. 1108 North Glenn Road Casper, Wyoming 82601 Tel: (800) 645-9292 Fax: (800) 853-9663 www.woodworker.com

Volara (fi ne celled polyolefi n foam) manufactured by Voltek Available from Conservation Resources International, LLC 5532 Port Royal Road Springfi eld, VA 22151 Tel: (800) 634-6932 Fax: (703) 321-0629 www.conservationresources.com

Textile Specialty Group Postprints Volume 23, 2013 31 TREATMENT OF A SUITE OF BAROQUE REVIVAL STYLE SEATING FURNITURE

Hide glue (#251 strength prepared with distilled water) available at local hardware stores or through Woodworker’s Supply, Inc. 1108 North Glenn Road Casper, Wyoming 82601 Tel: (800) 645-9292 Fax: (800) 853-9663 www.woodworker.com

AUTHOR BIOGRAPHY

GENEVIEVE BIENIOSEK was a graduate student intern in Furniture and Upholstery Conservation at Biltmore from 2012-2013. She has also held internships at the Cleveland Museum of Art; Shangri-La in Honolulu, Hawaii; and the Smithsonian Museum Conservation Institute in Suitland, Maryland. She received her M.A. in Art Conservation from Buff alo State College in 2013. Contact: [email protected]

Textile Specialty Group Postprints Volume 23, 2013 32 ESTABLISHING A TEXTILE DYE ANALYSIS PROGRAM AT THE INDIANAPOLIS MUSEUM OF ART

VICTOR J. CHEN, KATHLEEN KIEFER, NILOO PAYDAR & GREGORY D. SMITH

ABSTRACT—Knowledge of the dyestuff s on textiles can provide information on the history and origin of the textile as well as the technology employed to create it. Since dyestuff s used on textiles can consist of mixtures of coloring materials of natural or synthetic origin, their identifi cation requires extraction of the colorants from the fi ber samples, resolution of the mixture into individual chemical entities, followed by physical characterization of each dye compound. To accomplish this at the Conservation Science Laboratory at the Indianapolis Museum of Art, we have established the instrumental technique of liquid chromatography coupled to detection by diode array spectroscopy and mass spectrometry. Th is article describes the format of the experiment and some of the challenges encountered in dye extraction, structural characterization, and assignment. Th e results presented include those from studies to characterize reference dyestuff s such as dyer’s broom, Scotch broom and several synthetic dyes, as well as a summary of the results from a detailed analysis of silk fi bers from an Uzbek coat.

CREACIÓN DE UN PROGRAMA PARA ANALIZAR LAS TINTURAS TEXTILES EN EL MUSEO DE ARTE DE INDIANÁPOLIS—El conocimiento de las tinturas de las telas puede aportar mucha información sobre la historia y el origen de la tela así como de la tecnología empleada para su creación. Dado que las tinturas utilizadas en las telas pueden ser mezclas de material colorante de origen natural o sintético, su identificación exige la extracción de los colorantes de las muestras de fibras, la resolución de la mezcla en entidades químicas individuales y la caracterización física de cada compuesto. Para lograr esto en el Laboratorio de Ciencias de la Conservación del Museo de Arte de Indianápolis, hemos estab- lecido la técnica instrumental de cromatografía líquida con detección de matriz de diodos y espectometría de masas. Este artículo describe el formato de nuestros experimentos y algunos de los problemas con los que nos encontramos durante la extracción de la tintura y la caracterización y asignación estructural. Se presentan los resultados de estudios de caracterización de tinturas de referencia, gualda y retama negra, de varias tinturas sintéticas, y un resumen de los resultados de un análisis de fibras de seda de un abrigo uzbeko.

1. INTRODUCTION

Results from scientifi c investigations of objects in cultural heritage can be helpful to conservators in their eff ort to formulate preservation and conservation strategies. Th e same kind of results can also form an objective basis in discussions of art history relating to matters such as authorship, aesthetics, artist’s intent, and chronology (Artioli 2010). A common type of scientifi c investigation requested for textiles is dye analysis aimed at understanding the identity of the dyestuff s used on textile. Dyestuff s derived from both natural and synthetic sources are generally composed of multiple colored and colorless compounds. To identify the dyestuff s used on textiles, it is best to extract these compounds from the textile fi ber, separate them from one another, and then individually identify them, even if this involves the destruction of a small sample taken from the textile. Th e analytical method that is most commonly used to accomplish this is liquid chromatography combined with diode array detection and mass ESTABLISHING A TEXTILE DYE ANALYSIS PROGRAM AT THE INDIANAPOLIS MUSEUM OF ART spectrometry (LC-DAD-MS). Th is technique has been used in the Conservation Science Lab at the India- napolis Museum of Art for dye analysis since it began operation in 2011. Th e goal of this essay is to give the reader a sense of what is involved in this method of dye analysis including a description of the experimental set up along with some of the challenges encountered in sample extraction, characterization of dye molecules, and the issue of specifi cally identifying dyes. Th e work fl ow of dye analysis consists of (1) sample collection, (2) release of the coloring matter into an extract, (3) analysis of the color compounds in the extract by LC-DAD-MS, and (4) interpretation of instrumental data and their assignment to known color compounds that would indicate the use of a particular dyestuff . Th e amount of sample required to make an accurate identifi cation depends on the amount of color compounds present on the dyed textile fi bers. An amount of fi ber in the range of 0.5 to 1.0 mg (about 1 cm or less in most cases) can generally be considered an adequate amount of sample: for LC-DAD-MS. In a favorable situation, wherein the sample is dyed with a high amount of dyestuff comprising a simple mixture of compounds with good detectability, an identifi cation of the dyestuff can be made with 50 ␮g of textile. In some situations adequate availability is unrelated to failure to identify the dyestuff . Factors such as sample degradation, or the data obtained not being easily assignable to any known color compounds, can account for failure.

2. DYE EXTRACTION

Because a textile sample can contain a wide variety of natural and synthetic dyes, we have used a four-step protocol, each making use of a solvent of diff erent solubility parameters to maximize recovery of the colorants. Th e fi rst step involves heating the sample fi ber for 60 min at 80ЊC in 200 ␮L of a solvent composed of methanol:acetone:water in the volumetric ratio 3:3:4 that contains 2 mM oxalic acid, a chelating agent to facilitate disruption of any dye-mordant-substrate interaction. Th is solvent (OAMAW) was reported previously for analysis of natural dyes (Valianou et al. 2009). Th e second step entails heating the recovered fi ber in 200 ␮L of pyridine:water in the volumetric ratio 1:1 (PW), which is a solvent similar to that reported for successful analysis of synthetic dyes (Stefan et al. 2009). Between each extraction, the fi ber sample is washed three times with 200 ␮L of methanol:water in a volumetric ratio1:1 (MW). Since the extraction solvents are not compatible with subsequent instrumental analysis, the extract combined with the associated wash liquids for each step set at 65ЊC is dried by a gentle stream of nitrogen, and the residue re-solubilized in a minimum amount of MW. Since the materials extracted in the two steps likely consist of some overlapping colorants, they are combined to a fi nal volume of 200 ␮L, then fi ltered through a 0.45 ␮m polyvinylidene fl uoride (PVDF) or polytetrafl uoroethylene (PTFE) membrane and are analyzed as a single sample by LC-DAD-MS. Current literature (Mouri et al. 2012) indicates that an alternative to the fi rst two steps would be a solution of PW containing 0.05M oxalic acid (OAPW). Th ese solvents, OAMAW, PW and OAPW are suffi ciently gentle so that esters and O-glycosidic bonds of dyes are unaltered. Th e preservation of these kinds of derivatives of the dyes is important as their presence can be used for identifi cation of the botanical origin of the dyestuff (Mouri and Laursen 2012, Mouri and Laursen 2014). If the sample still remains substantially colored aft er the initial extraction, especially when the sample is originally blue or green and is suspected to contain indigoid dyes, the fi ber is heated in dimethyl sulfoxide to recover any possible indigoids. Additionally, a fi nal step involving digestion of the fi ber with 6N HCl prepared

Textile Specialty Group Postprints Volume 23, 2013 34 VICTOR J. CHEN, KATHLEEN KIEFER, NILOO PAYDAR & GREGORY D. SMITH in methanol:water (HMW) may be incorporated to rule out any remaining tightly bound dyes. Prior to instrumental analysis, the HMW solvent must be removed and the recovered dyes redissolved in MW, fi ltered, and analyzed as described for the fi rst two steps.

3. INSTRUMENTAL ANALYSIS

Conceptually, LC-DAD-MS is a two-stage technique. In the liquid chromatography stage, the extracted color compounds are forced through a small column packed with microparticles (the stationary phase) by a pressurized solvent (the mobile phase) that is composed of an aqueous-organic mixture. The column most commonly used for dye analysis is packed with what is known as ‘reversed phase’ type microparticles that have a nonpolar surface. Differential interactions between dye compounds and the stationary phase cause the dye compounds to travel at an unequal rate through the column, thereby allowing the chemically distinct components of a dyestuff to be separated. The retention time (RT), which is the time required to for a dye to travel through the column, is a characteristic of a dye compound. As the molecules emerge from the column, they are analyzed by a range of detectors. A diode array detector records an electronic absorption spectrum that provides information about the color and the chemical class of the dye. The effluent then enters a mass spectrometer, which scans the mass spectrum (MS) that indicates the molecular mass of the dye and any easily fragmented portions of the dye structure. Additionally, collision of the dye molecules with helium atoms injected into the mass spectrometer creates dye fragments, which when analyzed by a secondary mass spectrum (MS2) gives hints about the functional groups that may be present on the dye molecule. Thus by the technique of LC- DAD-MS, four types of analytical data are obtained for each colorant, such that even when two dyes are found to share more than one parameter, the remaining information can provide discrimination and specification. To identify a dye, the LC-DAD-MS data generated for the sample are compared with the corresponding data from known dye compounds. A match of the four types of analytical data constitutes a successful identifi cation. Unlike infrared (IR) or Raman spectroscopic analysis for which data bases of culturally relevant chemicals exist either online or are provided with the soft ware of the instrument, there are few repositories of retention time, absorption spectra, and mass spectroscopic data for individual organic dyes. In most instances, the assignment of instrumental data to specifi c dye compounds depends either on having prior experience with the dyes and related compounds or on mining published chemical literature for data of known dyes for a match. Once identifi ed, the dyes in the samples should be confi rmed by performing the same instrumental analysis on genuine reference materials, since results from one lab to another may vary to some extent due to diff erences in instrumention or experimental conditions.

4. ANALYSIS OF REFERENCE MATERIALS

To illustrate what the LC-DAD-MS data look like and how they are used, the results from a comparison of two dyestuff s will now be described. Th e dyestuff dyer’s broom (Genista tinctoria) is reported to contain the dyes luteolin, genistein and apigenin (Cardon 2007). A mixture made from pure chemicals of the three compounds obtained from Sigma-Aldrich Co. was run on the LC-DAD-MS instrument.

Textile Specialty Group Postprints Volume 23, 2013 35 ESTABLISHING A TEXTILE DYE ANALYSIS PROGRAM AT THE INDIANAPOLIS MUSEUM OF ART

Figure 1. LC-DAD-MS data obtained from a mixture made from individual pure compound of luteolin, genistein and apigenin. Th e top graph shows the chromatogram of the mixture, followed by their individual electronic spectra in the next row, mass spectra (MS) in the third row and MS2 fragmentation pattern in the lowest row.

Th e data are shown in fi g. 1, wherein the topmost graph is a chromatogram showing the separation of the three compounds based on RT, with the peaks indicating the detection of the presence of luteolin, genistein and apigenin at 13.0, 15.5 and 15.9 min, respectively. Th e three graphs in a row immediately below the chromatogram are the corresponding electronic absorption spectra from the DAD. Th e next three graphs in the third row are MS, followed by fragmentation pattern of MS2 in the last row. It can be seen that luteolin and apigenin cannot be distinguished by their electronic spectra, but genistein stands out among the trio in having a distinct electronic spectrum. On the other hand, genistein and apigenin have same mass, which is diff erent from that of luteolin, so MS can identify luteolin from the three compounds. Th us, retention times, electronic spectra, MS and MS2 data are complementary and provide cross confi rmation of the identity of the three compounds. To gain experience in working with the dyestuff , we purchased a commercial sample of dried plant dyestuff that was labeled as dyer’s broom. An HMW extract of this dyestuff was prepared as described and subjected to analysis by LC-DAD-MS to compare to the reference chromatogram shown again in fi g. 2 at the top. Th e extract gave the chromatogram shown beneath the reference compounds in fi g. 2, wherein only one compound was detected at 8.1 min, and it did not match the elution of reference luteolin, genistein, or apigenin shown in the top chromatogram for the three dyes expected in dyer’s broom.

Textile Specialty Group Postprints Volume 23, 2013 36 VICTOR J. CHEN, KATHLEEN KIEFER, NILOO PAYDAR & GREGORY D. SMITH

Figure 2. Chromatograms of extracts of dyestuff s from various sources. All runs in this fi gure were performed under identical experimental conditions. Top chromatogram shows the elution of pure compound of luteolin, genistein and apiginin with retention times (RT) at 10.4, 11.7 and 12.1 min, respectively. Th ese RTs are not the same as those in fi g. 1 due to slight diff erence in experimental conditions. However, this diff erence in experimental conditions did not aff ect the electronic, MS and MS2 spectra.

Aft er searching data published in the literature, it was found that the electronic spectrum and MS of the mysterious dye in the commercial dyestuff matched the natural dye scoparin (Gattuso et al. 2006), which is found in Scotch broom (Cytisus scoparius), but not in dyer’s broom. While Scotch broom is a rarely used dyestuff (Cardon 2007), it is a popular ornamental plant that is readily available in many plant nurseries. To confi rm our results, HMW extract prepared from aerial parts of a live Scotch broom purchased from the gardening section of Lowe’s Companies Inc. was analyzed. Th e result is shown in the third chromatogram in fi g. 2 and shows that the Lowe’s Scotch broom contained the same dye eluting with the same retention time as the commercial dyestuff . In contrast, an HMW extract of a sample of authentic dyer’s broom acquired from the commercial plant nursery Carolee’s Herb Farm gave luteolin, genistein, and apigenin as described in the literature, but no scoparin. Th us, we concluded that the commercially purchased dried plant dyestuff was Scotch broom rather than dyer’s broom as indicated by the label. We have also analyzed synthetic dyes obtained from chemical supply vendors. Th e results are collected in table 1, where it can be seen that in each case, varying amounts of the component designated by the label on the bottle were detected. For Brilliant Green, Acid Green 16, Rhodamine B, Crystal Violet, and Ethyl Violet,

Textile Specialty Group Postprints Volume 23, 2013 37 ESTABLISHING A TEXTILE DYE ANALYSIS PROGRAM AT THE INDIANAPOLIS MUSEUM OF ART

Table 1: Color Compounds Detected in Commercial Synthetic Dyes Dye Name Supplier Dyes Found (% present)* Brilliant Green Sigma, Cat# B6756 Brilliant Green (96); Desethyl Brilliant Green (4) Acid Green 16 Sigma, Cat# S468770 Acid Green (93); Desethyl Acid Green (7) Rhodamine B Sigma, Cat# R4127 Rhodamine B (98); Desethyl Rhodamine B (2) Crystal Violet Fisher, Cat# S 93213 Crystal Violet (96); Methyl Violet (4) Ethyl Violet MP Biochemicals, Ethyl Violet (98) Cat # 158016 Desethyl Ethyl Violet (1); others (trace) Others (trace) Methyl Violet 2B Fisher, Methyl Violet (44); Crystal Violet (37) Cat # S71970R Desmethyl Methyl Violet isomer 1(1), 2(12); Didesmethyl Methyl Violet isomers 1 (1), 2 (2), 3 (2); Hofmann’s Violet MP Biochemicals, Crystal Violet (11); Methyl Violet (24); Cat # 208546 Desmethyl Methyl Violet isomer 2 (21); Didesmethyl Methyl Violet isomer 3 (43); Th ree other Violet compounds (trace) * Percent of total colored compounds were estimated based on the peak absorbance of the fi rst absorption band on the high wavelength end of the electronic spectrum. the label-specifi ed compound was detected as greater than 90% of the color material present. However, for Methyl Violet 2B and Hofmann’s Violet, many other color substances were present. Th ese additional substances showed spectral properties consistent with derivatives of the major colorants but lacking certain alkyl substituents, and thus they appeared to be degradation products or synthetic variants. Interestingly, these variants are serendipitous reference materials as they are oft en found present in historical artifacts too. From the data discussed so far, it is evident that commercial samples or laboratory prepared reference materials cannot be taken at face value as oft en there are mistakes, adulterations, and unexpected variants present. However, some of the unanticipated results turn out to be a valuable windfall, and the experience gained in characterizing reference materials also provides technical confi dence in the subsequent results obtained from artwork.

5. ANALYSIS OF A MUSEUM OBJECT

A silk embroidered man’s overcoat in traditional Uzbek style with little information about its date of creation or history was recently donated to the Indianapolis Museum of Art. Decorative garments of this kind have been in fashion in Central Asia since historic times. Although no provenance documentation exists, this particular example was thought to originate from the period spanning the late 1800s to early 1900s. To shed light on its creation date and history, a detailed dye analysis using LC-DAD-MS was performed on samples of silk thread taken from the coat. Th e results of this study have now been published (Chen et al. 2016) and will only be sum- marized here. Fourteen color thread samples and two colorless controls were analyzed. From the color threads a total of 58 compounds were identifi ed. Twelve of these color compounds were dyes known to be found in the natural dyestuff s common barberry, madder, larkspur, and cochineal. Th e remaining were synthetic compounds in the class of sulfonated diazo, xanthene, and triarylmethine dyes. Of the synthetic dyes identifi ed, the most recently discovered was Direct Red 23, C.I. 29160, fi rst patented by W.A. Israel and R. Kothe of Bayer and Co in 1900 (Colour Index 1971), thus refi ning the date for the creation of the garment to early 20th century.

Textile Specialty Group Postprints Volume 23, 2013 38 VICTOR J. CHEN, KATHLEEN KIEFER, NILOO PAYDAR & GREGORY D. SMITH

Besides dating of the coat, the data of this investigation also taught us several lessons about dyes. Th e fi rst was that dyes such as carminic acid, the principal component of cochineal dyestuff , and the synthetic dye Rhodamine B observed as contaminants in the embroidery thread samples were likely to have derived from the lining though physical contact. Another lesson learned was that most of the embroidery silk threads were dyed with more than one dye, with the same color on diff erent samples being obtained from blending diff erent dye components, mostly involving only synthetic dyes, and one involving the combinations of natural and synthetic components. It was interesting to note that despite the existence of green synthetic dyes, the green threads were dyed with diff erent combination of blue and yellow. Furthermore, all the synthetic dyes on the garment were found to be heterogeneous with profi les attributable either to crudeness in early synthetic methodology or to aging related degradation. Th is observation highlighted the conceptual relationship between dyes and dyestuff s. Historically, coloring materials referred to as dyestuff s were derived from natural materials like insects or plants (Cardon 2007). Each of these natural dyestuff s is characterized by a set of specifi c colored metabolites, or dyes, in ratios determined by biological, seasonal, and other environmental factors. In most cases, each of the synthetic dyes could be correlated with known compounds assigned a Colour Index (C.I.) number. However, in certain instances, the dyestuff indicated by one C.I. number contains multiple components derived from the particular synthetic route used to prepare them, each of which is also referred to as a dye, and is sometimes given a separate C.I. number. Additional complexity is introduced as a result of chemical degradation during post reaction work up or by the process of aging or by both. Th us, in addition to providing objective data for dating the garment, the in-depth study performed on the coat provided insight into the history of European dye manufacturing and the utilization of these colorants in a Central Asian country in the early 20th century.

6. CONCLUSION

LC-DAD-MS dye analysis has been established at the Conservation Science Lab of the Indianapolis Museum of Art and has been applied successfully to the identifi cation of both natural and synthetic dyes in reference materials and museum artifacts. As interesting data are being generated on articles of cultural heritage from the IMA, continued eff ort is being made to fi ne tune the protocol for increased sensitivity so as to minimize sample requirements. Adjustments are also being made to accommodate a wider variety of fi bers and colorants that can be encountered in artwork.

ACKNOLWEDGEMENTS

We are indebted to Ms. Alba Fernandez-Keys for translating the abstract into Spanish. Th anks are due to Ms. Amanda Holden and Rebecca Summerour, as well as Drs. Michael Columbia and Jay Siegel for helpful discussions throughout this work. Th e authors also acknowledge the generosity of Ms. Carolee Synder of Carolee’s Herb Farm in Hartford, Indiana for the gift of a sample of dryer’s broom.

REFERENCE CITED

Artioli, G. 2010. Scientifi c methods and cultural heritage : an introduction to the application of materials science to archaeometry and conservation science. Oxford; New York: Oxford University Press.

Cardon, D. 2007. Natural dyes: sources, tradition, technology and science. London: Archetype.

Textile Specialty Group Postprints Volume 23, 2013 39 ESTABLISHING A TEXTILE DYE ANALYSIS PROGRAM AT THE INDIANAPOLIS MUSEUM OF ART

Chen, V. J., G. D. Smith, A. Holden, N. Paydar and K. Kiefer. 2016. Chemical analysis of dyes on an Uzbek ceremonial coat: Objective evidence for artifact dating and the chemistry of early synthetic dyes. Dyes and Pigments, 131: 320–332.

Colour Index Volume 4 1971. Bradford, Yorkshire, UK: Society of Dyers and Colourists, American Association of Textile Chemists and Colorists. 4272–4273

Gattuso, G., C. Caristi, C. Gargiulli, E. Bellocco, G. Toscano and U. Leuzzi. 2006. Flavonoid Glycosides in Bergamot Juice (Citrus bergamia Risso). Journal of Agricultural and Food Chemistry, 54: 3929–3935.

Mouri, C. and R. Laursen. 2012. Identifi cation of anthraquinone markers for distinguishing Rubia species in madder-dyed textiles by HPLC. Microchimica Acta, 179: 105–113.

Mouri, C., V. Mozaff arian, X. Zhang and R. Laursen. 2014. Characterization of fl avonols in plants used for textile dyeing and the signifi cance of fl avonol conjugates. Dyes and Pigments, 100: 135–141.

Stefan, A. R., C. R. Dockery, B. M. Baguley, B. C. Vann, A. A. Nieuwland, J. E. Hendrix and S. L. Morgan. 2009. Microextraction, capillary electrophoresis, and mass spectrometry for forensic analysis of azo and methine basic dyes from acrylic fi bers. Analytical and Bioanalytical Chemistry, 394: 2087–2094.

Valianou, L., I. Karapanagiotis and Y. Chryssoulakis. 2009. Comparison of extraction methods for the analysis of natural dyes in historical textiles by high-performance liquid chromatography. Analytical and Bioanalytical Chemistry, 395: 2175–2189.

SOURCES OF MATERIALS

Dyer’s Broom Carolee’s Herb Farm 3305 Country Rd S 100 W Hartford City, IN 47348 Tel: (765) 348-3162 http://www.caroleesherbfarm.com/

Dyer’s Broom (found to be Scotch broom) Kremer Pigments 247 West 29th Street New York, NY 10001 Tel: (212) 219-2394 http://www.kremerpigments.com/

Scotch Broom Lowe’s Companies, Inc. 6002 N Rural St, Indianapolis, IN 46220 Tel: (317) 202-9142 http://http://www.lowes.com/

Textile Specialty Group Postprints Volume 23, 2013 40 VICTOR J. CHEN, KATHLEEN KIEFER, NILOO PAYDAR & GREGORY D. SMITH

Brilliant Green, Acid Green 16 and Rhodamine B Sigma-Aldrich Co., LLC. 3050 Spruce St. St. Louis, MO 63103 Tel: (800) 521-8956 http://www.sigmaaldrich.com/ Crystal Violet and Methyl Violet 2B Fisher Scientifi c Co., LLC. 300 Industry Drive, Pittsburgh, PA 15275 Tel: (877) 885-2081 http://www.fi shersci.com/ Ethyl Violet and Hofmann’s Violet MP Biomedicals, LLC 3 Hutton Center Drive, Suite 100 Santa Ana, CA 92707 Tel: (800) 854-0530 http://www.mpbio.com/

AUTHORS BIOGRAPHIES

VICTOR J. CHEN, a native of Hong Kong, obtained his Ph.D. degree in biochemistry from Iowa State Uni- versity, Ames, Iowa, with specialization in protein chemistry and enzymology. He is a retiree from Eli Lilly and Company, where worked as a research scientist in drug discovery for 25 years. Victor is interested in learning about visual art through studying the science behind the artwork. He joined the Indianapolis Museum of Art Conservation Science Lab in 2011 as a full time volunteer. Email: [email protected]. KATHLEEN KIEFER earned an MS in Art Conservation from the Winterthur/University of Delaware Pro- gram in Art Conservation and a BS in Textiles and Clothing from the University of Texas at Austin. She was an assistant conservator with the Textile Conservation Center at the American Textile History Museum in Lowell, MA, a project conservator for the Peabody Museum of Archaeology and Ethnology, and a textile conservator at Winterthur Museum, Gardens & Library, where she also taught with the graduate conservation training program, and was the senior conservator of textiles for the Indianapolis Museum of Art. She is currently the conservator of textiles for the Art Institute of Chicago, and continues to provide Oddy testing services through her business, Material Culture Conservation, LLC. Email: [email protected]. NILOO PAYDAR received BFA and MFA degrees in textiles from Syracuse University. Since 1986, Niloo has been curator of textile and fashion arts at the Indianapolis Museum of Art, where she has organized more than 35 exhibitions, authored catalogues and books on the museum’s collection. She holds membership in, and has served on advisory boards of, several associations related to textiles and costumes. She was awarded Th e Premier Print Award for the book eTh Fabric of Moroccan Life, Best of Category, 2002, and Th e Joseph V. McMullan Award for Stewardship and Scholarship in Islamic Rugs and Textiles in 2002. Email: [email protected]

Textile Specialty Group Postprints Volume 23, 2013 41 ESTABLISHING A TEXTILE DYE ANALYSIS PROGRAM AT THE INDIANAPOLIS MUSEUM OF ART

GREGORY DALE SMITH holds a Ph.D. in physical/analytical chemistry from Duke University. Dr. Smith was previously the Andrew W. Mellon Assistant Professor of Conservation Science at Buff alo State College, and is currently the Otto N. Frenzel III Senior Conservation Scientist at the Indianapolis Museum of Art. Dr. Smith has extensively published in the fi elds of chemistry and conservation. His research interests include studying condition issues aff ecting modern polymers used in art, pigment degradation processes, preservation environments, and the development and testing of innovative conservation treatments. Email: [email protected]

Textile Specialty Group Postprints Volume 23, 2013 42 FINDING THE EASE: APPROACHES TO MOUNTING AND INSTALLATION AT THE ART INSTITUTE OF CHICAGO

ISAAC FACIO AND LAUREN CHANG

ABSTRACT—Th e conservation staff of the Department of Textiles at the Art Institute of Chicago has recently collaborated with colleagues to develop mounting and installation strategies that will ease the burden of the ever-increasing demand for textiles within the museum without compromising the safety of the art, and which will decrease stress and strain on both the collection and the staff . Th ree examples of adaptations to existing installation methodologies to address specifi c problems are explored. For large African textiles, a three-part modular pin mount was devised which could be easily changed in width without removing it from the wall. For longer lengths of yardage, a “roll-top” with a specialized cradle was redesigned to hold the un-exhibited section of a textile on its storage pipe suspended over the mount surface and without re-rolling. For tapestries, a hollow I-beam of medium density overlay and a coordinating shelf hanging system was designed. A process for installing the tapestry with two pallet stackers and extending end-caps on the I-beam was developed, increasing installation effi ciency and minimizing stress and strain on installation staff . Th e above strategies have transformed some of the more challenging installations into more routine activities.

EN BUSCA DE LA COMODIDAD: MÉTODOS DE MONTAJE E INSTALACIÓN EN EL INSTITUTO DE ARTE DE CHICAGO—El personal de conservación del Departamento de Textiles del Instituto de Arte de Chicago ha colaborado con sus colegas para desarrollar estrategias de montaje e instalación que aliviarán la carga de la creciente demanda de telas dentro del museo sin poner en riesgo la seguridad de las piezas, y que disminuirá la tensión y el esfuerzo tanto para la colección como para el personal. Se están analizando tres ejemplos de adaptaciones de las metodologías de instalación existentes para solu- cionar problemas específicos. Para las grandes telas africanas, se diseñó un montaje modular de tres piezas que puede ensancharse fácilmente sin tener que ser retirado de la pared. Para las telas más largas, se rediseñó un mecanismo enrollable con un soporte especial que mantiene a la parte de la tela no exhibida enrollada en su tubo por encima de la superficie del montaje sin que se desenrolle. Para los tapices, se diseñó una viga hueca en forma de I con revestimiento de densidad media y un sistema colgante de repi- sas. Se desarrolló un proceso para la instalación de tapices con dos apiladores de pallets y topes extensi- bles en la viga, lo que mejora la eficacia de la instalación y minimiza el esfuerzo del personal. Estas estrategias han transformado algunos de los procesos de instalación más difíciles en actividades más rutinarias.

1. INTRODUCTION

Over the last fi ve years the conservation staff of the Department of Textiles has collaborated with colleagues to develop mounting and installation strategies that will ease the burden of the ever-increasing demand for textiles within the museum without compromising the safety of the art. Th ree examples of adaptations to existing installation methodologies to address large African textiles, longer lengths of yardage, and tapestries are explored below. In each situation the resulting methodology builds on the generous expertise of textile conservators, mountmakers and installers past and present. FINDING THE EASE: APPROACHES TO MOUNTING AND INSTALLATION AT THE ART INSTITUTE OF CHICAGO

2. AFRICAN TEXTILE MOUNT

Th e African art galleries were relocated and renovated from 2009 to 2011. An open platform display for pin- mounted textiles was designed within the gallery where textiles were selected in advance for four years of three-month rotations. Twelve to sixteen large textiles of varying sizes from multiple cultures were selected. Building individual mounts for each of these textiles was not an option because storage space is not available. Th e mount system had to remain on the wall during installation due to the lack of space in the gallery because of casework directly in front of this area. Installations typically occur within the 9:00 to 10:30 a.m. time slot before museum public hours, requiring installation, repair, and cleanup to be compact, effi cient, and safe. Th e mount and installation system had to accommodate large textiles of varying cultures and sizes. A three-part modular mount was devised that would accommodate these textiles using the department’s standard pin mounting method: a fabric-covered pin mount that hangs from a cleat system, and is kicked-out from the wall from the bottom edge to a specifi ed angle. Th e three parts were built to specifi c dimensions with short receiving wall cleats distributed across the width of the mount panels specifi cally positioned to receive three diff erent confi gurations of this mount (fi g. 1). Th e sizes and combinations of mount parts were dictated by the various groupings of the textiles. Although a modular mount is not new in concept, it was new to the department at this scale and with these constraints.

Figure 1: Th ree-part modular mount showing the position of the short receiving wall cleats and long mount cleats that are on the reverse side of each section along the top edge. Th e mount height (X) is 306.07 cm (120 ½ in.); the total width (Y) is 453.4 cm (178 ½ in.). Parts a and c are 102.87 cm (40 ½ in.) wide and part b is 247.65 cm (97 ½ in.) wide. Th e platform (d) is a 518.16 cm (204 in.) wide including an adjustable 91.44 cm (36 in.) extension (e).

Textile Specialty Group Postprints Volume 23, 2013 44 ISAAC FACIO AND LAUREN CHANG

Th e mount is 306.07 cm (120 ½ in.) in height with a center panel measuring 247.65 cm (97 ½ in.) in width and two narrow panels of 102.87 cm (40 ½ in.) widths on each side. Th e mount is kicked-out from the wall at a 10Њ angle from six removable C-shaped metal kick-outs distributed and attached across the back of the three panels (fi g. 2). Since the mount is very tall, the top horizontal portion of the C-shape supports the mount about midway to avoid any bowing; the vertical portion rests against the wall and is covered with Volara strips to avoid damage to the exhibition wall; and the lower horizontal portion kicks-out the bottom of the mount.

Figure 2: Side view of the mount showing the C-shaped kick-outs attached to the reverse side of the mount.

Textile Specialty Group Postprints Volume 23, 2013 45 FINDING THE EASE: APPROACHES TO MOUNTING AND INSTALLATION AT THE ART INSTITUTE OF CHICAGO

Th e selected textiles were divided into three groupings, large, medium, and small. Th e largest textile uses the entire 453.39 cm (178 ½ in.) width and the 306.07 cm (120 ½ in.) height of the mount leaving a suffi cient mount border around the object (fi g. 3). In some cases, two smaller textiles will be used to take up this width, placed side by side to fi ll the same space. Medium sized textiles use the 350.52 cm (138 in.) width with two parts of the mount (fi g. 4), and the smaller textiles use the 247.65 cm (97 ½ in.) width of the center panel alone (fi g. 5). Installation and repair of the exhibition mount and space had to be systematic. Two wall cleats support the side panels, and three wall cleats support the center panel. Each cleat is held in place with two 10.16 cm (4 in.) bolts. Th e receiving cleats are spaced out specifi cally so they accept the three combinations of mounts with even support, but not necessarily evenly spaced. A corresponding 10Њ angle on the face provides a can- tilever-rack eff ect of support across the mount cleat. For a two-part display, one of the narrow panels is removed exposing its two receiving cleats. Th e remaining panels are centered within the space by sliding them to the center receiving cleats. Th e exposed cleats on each side of the mount are removed, requiring minimal repair. When displaying the single panel mount, the remaining narrow panel is removed, and the wide panel is transferred to the center receiving cleats, exposing cleats on both sides. Th ey are removed requiring minimal repair to the wall. Th e platform also provides some adjustability to coincide with the size of the mount. To correspond with this single panel confi guration, an extension on the right side is removed. It is centered on the mount providing a 457.2 cm (180 in.) span of protection to the 247.65 cm (97 ½ in.) mount width (fi g. 5).

Figure 3: Th ree-part mount confi guration showing a man’s adinkra wrapper, early 20th c., Asante, Ghana (Th e Art Institute of Chicago). Th e mount height (X) is 306.07 cm (120 ½ in.); the total width (Y) is 453.4 cm (178 ½ in.).

Textile Specialty Group Postprints Volume 23, 2013 46 ISAAC FACIO AND LAUREN CHANG

Figure 4: Two-part mount showing a display cloth (ndop), early/mid-20th c. (before 1937) Bamum, Foumban, Cameroon (Th e Art Institute of Chicago). Th e mount height (X) is 306.07 cm (120 ½ in.); the total width (Y) is 350.52 cm (138 in.).

Figure 5: One-part mount showing a Nupe panel, mid-20th c., Nigeria (Th e Art Institute of Chicago, gift of Richard J. Faletti). Th e mount height (X) is 306.07 cm (120 ½ in.); the total width (Y) is 247.65 cm (97 ½ in.). Th e platform extension is altered to coincide with the smaller scale of the mount.

Textile Specialty Group Postprints Volume 23, 2013 47 FINDING THE EASE: APPROACHES TO MOUNTING AND INSTALLATION AT THE ART INSTITUTE OF CHICAGO

Th e fi nal product allows for a single three-part mount that requires minimal storage space, accommodates a large array of large textiles, and leaves minimal repair of the exhibition wall, for a simple to handle and clean display. However, this system does leave some drawbacks such as the visible seams where the mount panels meet and the mounting of the art has to be done upright, with the mount on the wall, because of the lack of additional foot space in the gallery. With some refi nement to overcome diffi culty in producing a fl ush alignment across the panels, a successful installation has resulted.

3. ROLL TOP

For extra-long lengths of yardage, the “roll-top” was redesigned to hold un-exhibited sections of the textile on the storage pipe suspended in a cradle above its mount. Th e roll-top allows a textile to move from storage to the mount and back to storage all on the same pipe, reducing handling. Th e cradle can accommodate multiple angles to adjust to the needs of the textile or display, while maintaining a smooth transition from pipe to mount. Th e previous method of mounting long lengths of yardage required re-rolling the object from its storage pipe onto a new pipe with the face of the object rolled inward. Th e exhibited length of textile was unrolled onto a wooden stretcher covered with Velcro-compatible cloth, and the remaining textile on the pipe would be wrapped with an additional piece of Velcro-compatible cloth. Th e package was held in place against the top edge of the mount, strapped with hook Velcro straps from the edges of the pipe to the back of the mount. Th is method took a lot of practice to align the cloth cover that gets partially rolled into and around the object. In some cases, because of the thickness of the Velcro-compatible cloth and the amount of textile on the pipe, the package became bulky. One of the goals in preparing objects for exhibition is to avoid over-manipulating the art at any point. Th e above method needed to be re-approached with the goals of reducing handling of the art, and eliminating the Velcro-compatible cloth, which we now know does not have good aging properties. Transferring the art onto a new pipe is a big process. Lengths are typically longer than 244 cm (96 in.) and required re-rolling object panels in the opposite direction from its storage confi guration. It was preferred that the pipe and rolled textile be suspended rather than resting on its own weight against the mount. Curvature or tension on the object needed to be eliminated as the object transitioned from the roll to the mount due to the bulky wrapping material. In the new design, the storage pipe requires approximately 5 cm (2 in.) of exposed pipe on each side of the rolled object that is used to rest on the roll-top cradle. Th e C-shaped wooden cradle mechanism attaches around the top edge and reverse side of the mount. Th e object remains on this storage pipe, is unrolled to the exhibition length onto the mount, and the pipe placed into the cradle (fi g. 6). Th e roll-top consists of two cradles, one on each side of the top edge of the mount (fi g. 7). Each cradle is made up of two 1-inch thick wood blocks: the support block which attaches to the reverse side of the mount, and on top of this the deeper cradle block which is notched to receive the pipe and extends over the top of the mount. Th e support block is attached to the back of the mount with an L-shaped metal plate pointing towards the center of the back of the mount. Th e two blocks are attached together through a dovetail groove, allowing the top block to slide fl ush with the face of the mount, thus providing some adjustability to various mount depths and allowing the object to transition from the pipe it is rolled on to the mount on the same plane. Th e cradles are adjustable for 2.54 cm (1 in.), 3.18 cm (1 ¼ in.), and 3.81 cm (1 ½ in.) depths to coincide with the department’s mount stock. To fi x the top block in fl ush position, a slotted metal plate across the joint in the two blocks is secured with large setscrew on the support block. Th e exposed pipe rests on the cradle and is

Textile Specialty Group Postprints Volume 23, 2013 48 ISAAC FACIO AND LAUREN CHANG

Figure 6: Side view and close up of “Horses,” designed by Saul Steinberg, ca. 1949/52 (Th e Art Institute of Chicago, Elizabeth F. Cheney Fund), rolled and suspended on the roll-top. Th e visible outer side of the roll-top is plain and the adjustment plate faces the center. secured with the twill tape that extends through a hole in the back. Th e twill tape is tacked on the inside of the cradle and pulled through the hole to provide tension to secure the pipe in place. A metal L-plate attached to the bottom is used for hanging the cradle from a J-rail, and also accommodates the depth needed depending on the angle on the mount (fi g. 8). Because it is free hanging and held in place by gravity, the cradle accepts any angle of kick-out from the wall. When a 10Њ angle or above is needed, the upper back corner of the roll- top mechanism pivots towards the wall; to give the space needed, the bottom plate is adjusted by moving it back, pushing the support block away from the wall. Th e fi nal product is successful in that it results in allowing the object to remain on its storage pipe in the same confi guration. Th e system provides that the object is suspended and off its own weight, allows for a smooth transition from pipe to mount, is simple to install, and produces a clean display. It is also adaptable for

Textile Specialty Group Postprints Volume 23, 2013 49 FINDING THE EASE: APPROACHES TO MOUNTING AND INSTALLATION AT THE ART INSTITUTE OF CHICAGO

Figure 7: Roll-top. Left : front view showing the attachment plate that secures to the reverse side of the mount at three points. Center top: the inside side view showing the adjustment plate with three small screws securing the top portion and a large adjusting screw into the bottom portion. Th e 1 in. twill tape is secured to the inside of the cradle and feeds through the block. Center bottom: the L-plate view from below shows the adjusting screws to move the plate back. Right: back view showing the attachment plate that secures to the reverse side of the mount at three points and the dovetail notch that joins the two portions together and allows for smooth, sliding adjustments. Th e total dimensions of the roll-top are 15.24 cm (6 in.) in height (a), 2.54 (1 in.) in width (c) and 10.16 cm (4 in.) in depth (b). horizontal display and has the potential for the dimensions of the blocks and metal plates to be altered to accept other pipe diameters and object sizes. Although this system is very successful, it currently has one drawback in that the roll top, the mechanism that holds the object, also hangs the mount from the wall. Th e authors are in the process of redesigning this element to make it a diff erent unit if necessary.

4. TAPESTRY I-BEAM AND INSTALLATION

In the fall of 2008 the Department of Textiles celebrated the completion of a 15-year research and conservation project of its historic tapestry collection with Th e Divine Art: Four Centuries of European Tapestries, an

Textile Specialty Group Postprints Volume 23, 2013 50 ISAAC FACIO AND LAUREN CHANG

Figure 8: Diagram showing the art unrolled onto the mount on the same plane. Th e mount is kicked out at a 10˚ angle and the L-plate is moved back to avoid touching the wall on the top corner of the roll-top.

Textile Specialty Group Postprints Volume 23, 2013 51 FINDING THE EASE: APPROACHES TO MOUNTING AND INSTALLATION AT THE ART INSTITUTE OF CHICAGO exhibition of 70 tapestries from the permanent collection. Th e installation schedule necessitated raising a minimum of seven tapestries a day. Th e nature of tapestry installation requires a group of individuals working together for a common goal. It is not something that the Department of Textiles staff could install independently. Th erefore, it was crucial to develop a system that was easy to communicate with clarity to a group and in which all team members had confi dence. Neither the Department of Textiles staff , nor the museum installation crew had extensive experience raising or lowering tapestries. Previous tapestry installations involved either pulley or scaff olding systems. Th e in- house team each had specifi c reservations about past installations, including fatigue and safety of the staff . Kirk Vuillemot, assistant conservator for preparation and framing, was concerned about the lack of strength in new two-by-four wood slats, which would need to span 548.64 to 731.52 cm (216 to 288 in.) without bowing. Th e authors were concerned about placing the components of a pulley system over the center of the tapestry without suffi cient clearance on the wall above. Craig Cox, head of art installation, Vuillemot, and the authors worked together to develop a new hanging mechanism and installation method to address those concerns and allow for the installation of 70 tapestries over the course of a 10-day period. Th e end result was a hollow beam of medium density overlay (MDO) that is based upon I-beam construction, but uses two crossbars. Th is hollow I-beam replaced the more traditional two-by-four used by the museum, and the hook side of Velcro was stapled to the I-beam in the same manner as to a two-by-four (fi g. 9). Tapestries can exhibit two kinds of force when hung—the downward force of gravity but also the forward roll of the tapestry being secured on the outer face of a slat, especially a three-dimensional I-beam. Th e I-beam, however, allows those two forces to be addressed separately. To support the downward force, a shelf, also created from MDO, was built as a mate to the I-beam (fi g. 10). Th is shelf is secured to the wall in accordance with the pre-determined hanging height. Once the tapestry is raised, the I-beam rests on this shelf. To counteract the forward roll, the I-beam was fi tted with mending plates at regular intervals. Th ese mending plates secure the I-beam to the wall with bolts once the tapestry is in place on the shelf. As most tapestries extend to the edge of the I-beam, endcaps were fabricated to slide into the hollow ends of each beam. Th e endcaps are 91. 44 cm (36 in.) lengths of two-by-fours that have a square of MDO on one end. Th e MDO square can be painted wall color to give the I-beam a fi nished look and less visually intrusive appearance. Th e endcaps slide out of the hollow up to 61 cm (24 in.) and serve as handles for any movement. Th e endcaps also play a role in the mechanical raising and lowering of the tapestries. Th e three-dimensional I-beam construction works in concert with the mechanical means employed to raise/lower and install/deinstall the tapestries. To fi ght fatigue and maintain the pace of installation, pallet stackers, to which platforms were already secured, were fi tted with custom metal arms designed and produced by Vuillemot. Th ese steel arms extend 61 cm (24 in.) beyond the front edge of the platform and are curved

Figure 9: Full view of the front of the I-beam showing the placement of the mending plates evenly distributed and the shelf spanning about three-quarters of I-beam’s length.

Textile Specialty Group Postprints Volume 23, 2013 52 ISAAC FACIO AND LAUREN CHANG

Figure 10: Th e I-beam is composed of four sides of MDO that create a hollow area. eTh I-beam is 10.8 cm (4 ¼ in.) square. Th e endcap is composed of a 10.8 cm (4 ¼ in.) by 1.9 cm (.75 in.) square piece of MDO secured to a 91.44 cm (36 in.) long two-by-four that is inserted into the hollow I-beam. Th e support shelf is also made of MDO and is 7.62 cm (3 in.) tall with a 7.62 cm (3 in.) depth. upwards at the end. Th e tip of the metal arms have protective felt, as the outer edges can come close to touching the wall on which the tapestry is installed. Installation of a tapestry might take place as follows. Th e shelf on which the I-beam will rest is secured to the wall in accordance with the pre-determined hanging height. Plastic is laid out on the fl oor and the pallet stackers are aligned and ready to drive parallel to the edges of the plastic (fi g. 11). eTh tapestry is rolled out face up on plastic with the top of the tapestry nearest the wall. Th e I-beam is brought between the wall and the top of the tapestry. Th e tapestry is secured with Velcro to the I-beam. Th e pallet stackers are moved forward to be on either side of the tapestry. If space is constricted, the positioning of the pallet-stackers can be swapped and placed in front of a pleated tapestry (fi g. 12). eTh endcaps are extended, and the tapestry is lift ed and placed onto the steel arms of the pallet stackers (fi g. 13). An installation crewmember steps up onto the platform on the pallet stacker to accompany the tapestry. Th e pallet stackers are operated so that they lift the tapestry and an installation crewmember on each platform up and above the shelf in a controlled and synchronized manner. In the same synchronized manner the pallet stackers move closer to the wall and lower the I-beam and tapestry onto the shelf. At this time, the installation crew members can slide the tapestry left or right on the shelf or even slightly raise one side for fi nal placement. Once the tapestry is where is should be, they take turns drilling in the mending plates at either end. Finally, one pallet stacker is driven to secure the remaining mending plates along the center of the I-beam.

Textile Specialty Group Postprints Volume 23, 2013 53 FINDING THE EASE: APPROACHES TO MOUNTING AND INSTALLATION AT THE ART INSTITUTE OF CHICAGO

Figure 11: Diagram showing a tapestry rolled out face up on plastic and secured to the I-beam, viewed from above. Th e pallet stackers are placed on either side of the tapestry with the arm extensions on the inside to receive the I-beam from its endcap. Th e tapestry depicted is Caesar in the Gallic Wars from Th e Story of Caesar and Cleopatra, ca. 1680, aft er a design by Justus van Egmont (1632-1688) (Th e Art Institute of Chicago, gift of Mrs. Chauncey McCormick and Mrs. Richard Ely Danielson).

Figure 12: In more compact spaces, the tapestry is pleated on plastic so that it will unfold as it is raised. Th e pallet stackers are placed in front of it with the arm extensions on the outsides to receive the I-beam from its endcaps.

Textile Specialty Group Postprints Volume 23, 2013 54 ISAAC FACIO AND LAUREN CHANG

Figure 13: Profi le view illustrating the steps of a tapestry installation. 1) Th e shelf is installed to the pre-determined height as shown in the upper left hand of the illustration. 2) Th e tapestry (shown in red) is attached to the I-beam on the ground between the two pallet stackers. 3) Th e tapestry is lift ed by art installation crewmembers from the endcap (shown in gray) to the pallet stacker’s arm extension. 4) Th e tapestry is raised high enough to clear the shelf. 5) It is moved forward and lowered onto the shelf.

Th roughout this process, the installation crewmembers at most bear only a portion of the weight of the tapestry during the initial placement onto the arms of the pallet stackers. Aft er that, either the metal arms or the shelf bears the weight of the tapestry. Each person keeps a hand or applies pressure on the end caps to fi ght the force of the forward roll. Th is system was devised out of the necessity of installing and deinstalling a large number of medium to large tapestries over a short period. During the installation period for the exhibition, the team was able to raise up to 10 tapestries a day. Th e system transfers the majority of the stress and strain of installation to a mechanical means. It also might have applications with other heavy textiles, such as carpets. However, the system does require two components that must be stored, and the I-beams themselves are heavy. Th e authors are in the process of refi ning the I-beam design with the hopes of addressing some of these issues.

Textile Specialty Group Postprints Volume 23, 2013 55 FINDING THE EASE: APPROACHES TO MOUNTING AND INSTALLATION AT THE ART INSTITUTE OF CHICAGO

5. CONCLUSION

As textiles continue to gain popularity and prominence in the museum, there will continue to be a need for innovation in installation. Th e strategies in this paper, which are works-in-progress and an on-going collabor- ative process, have transformed some of the more challenging installations at the Art Institute into more routine activities for both the conservation and installation staff .

ACKNOWLEDGEMENTS

Th e authors would like to thank Kirk Vuillemot and Craig Cox for their contributions to the development of the new installation methods; we would like to acknowledge our many colleagues who have shared their tapestry knowledge and expertise, especially Julia Woodward Dippold and the Department of Textile Conserva- tion at the Metropolitan Museum of Art. We would also like to thank Odile Joassin and Cristina Balloff et Carr for their support and encouragement.

AUTHOR BIOGRAPHIES

ISAAC FACIO part of the conservation staff in the Department of Textiles at the Art Institute of Chicago. He is a graduate of the School of the Art Institute of Chicago and holds a MS in Textile Technologies from the School of Engineering and Physical Sciences at the University of Manchester, UK. Address: Department of Textile, Th e Art Institute of Chicago, 111 S. Michigan Ave. Chicago, IL 60603; [email protected]

LAUREN CHANG is a textile conservator in Chicago. She was the conservator of textiles at the Art Institute of Chicago from 2005 until 2014. From 2003–2005 she was the project conservator at the National Museum of the American Indian for a community-based exhibition working with 11 Native communities from the North Pacifi c Coast. She was awarded Mellon Fellowships at the NMAI from 2001–2002 and the Los Angeles County Museum of Art from 2002–2003. She received an A.B. from Princeton University in Art and Archaeology and an MA from the Textile Conservation Centre, University of Southampton, UK. Address: Chicago, IL; [email protected]

Textile Specialty Group Postprints Volume 23, 2013 56 AN OLD CASE OF NEW DISPLAY: CONTEMPORARY AND HISTORIC FASHION AT THE VICTORIA AND ALBERT MUSEUM

JOANNE HACKETT AND KEIRA MILLER

ABSTRACT—Over the past 10 years, more than 70% of the Victoria and Albert Museum’s public space has been transformed as part of an ambitious plan to turn the museum into a 21st century cultural destination. As part of this process, the fashion display housed in Gallery 40 underwent renovation from January 2011 to May 2012, including new ambient lighting, a new paint scheme, new fl ooring, and new display spaces, but no new cases. Alongside the renovation of the permanent display housed within the same gallery, the V&A launched a temporary exhibition entitled Ballgowns: British Glamour Since 1950. Over 160 outfi ts were conserved and mounted for display for these two projects. Th e permanent display featured many contextual objects along side the garments. Despite the desire to display previously un-exhibited objects, time and budget constraints necessitated selecting mainly objects from previous exhibitions and reusing mounts. Th e open display of the temporary exhibition required particular attention during planning to the prevention of object handling and the mitigation of dust deposition. Unexpectedly high dust levels in the exhibition were monitored and determined to be visitor related. Th e results of the dust study have led to new criteria for future displays planned for this space.

NUEVAS EXHIBICIONES EN LAS VITRINAS DE SIEMPRE: MODA CONTEMPORÁNEA E HISTÓRICA EN EL MUSEO DE VICTORIA Y ALBERTO—En los últimos 10 años, más del 70% del espacio público del Museo de Victoria y Alberto ha sido transformado como parte del ambicioso plan de convertir al museo en un destino cultural del siglo 21. Como parte de este proceso, se renovó la sala de exhibición de moda de la Galería 40. La renovación comenzó en enero de 2011 y fi nalizó en mayo de 2012, e incluyó una nueva ilumi- nación ambiental, un nuevo esquema de pintura, nuevos pisos y nuevos espacios de exhibición, pero sin modifi car las vitrinas. A lo largo de la renovación de la muestra permanente alojada en la misma galería, el V&A lanzó una muestra temporaria llamada: Ballgowns: British Glamour Since 1950. Para estos dos proyectos, se conservaron y montaron más de 160 trajes en exhibición. La muestra permanente incluyó varios objetos contextuales que acompañaron a los trajes. A pesar del deseo de incluir objetos nunca exhibidos, dadas las limitaciones de tiempo y presupuesto, se debieron seleccionar objetos de exhibiciones anteriores y reutilizar los exhibidores. La exhibición abierta de la muestra temporaria requirió de especial atención durante la planifi cación en cuanto a la manipulación de los objetos y la mitigación de la decantación de polvo. Sorprendentemente, se determinó que los elevados niveles de polvo monitoreados durante la muestra estaban relacionados con los visitantes. Los resultados del estudio del polvo dieron lugar a nuevos criterios para las futuras muestras que se harán en este espacio.

1. INTRODUCTION

Beginning in 2001, the Victoria and Albert Museum (V&A) began an ambitious program of renovation. Titled ‘FuturePlan’, this program involves the remodeling and renewal of all of the galleries and public spaces within the museum, while causing as little impact or disruption as possible to the visiting public. So far, over 70% of the museum has been renovated and the galleries have been reorganized into more logical groupings. As part of this rolling renovation, the fashion display housed in Gallery 40 was renovated and remodeled AN OLD CASE OF NEW DISPLAY: CONTEMPORARY AND HISTORIC FASHION AT THE VICTORIA AND ALBERT MUSEUM between January 2011 and May 2012. Th e budget for the renovation was £1 million, the majority of which would be spent on architectural restoration, as the main aims of the project were to increase accessibility and improve the ambient lighting. As FuturePlan continues, the long-term plan for Gallery 40 is to house large scale South East Asian art, not fashion, making this something of an interim refurbishment. As such, there was limited budget to be spent on cases or object lighting as the cases are due to be removed when the gallery is handed over to the Asia department in 2018.

2. A HISTORY OF GALLERY 40 AND THE FASHION DISPLAY

Gallery 40, originally named the Octagon Court, was designed by Aston Webb and was built as part of the 1909 expansion of the museum. Th e space was originally designated as a gallery in which to house a mixed collection of items on loan from other institutions and private lenders (fi g. 1). eTh architectural features of the

Figure 1: Th e Octagon Court, 1920

Textile Specialty Group Postprints Volume 23, 2013 58 JOANNE HACKETT AND KEIRA MILLER gallery included a spacious glass dome, large alcoves, architectural columns, and mosaic fl ooring. From 1933 onwards, the Octagon was given over to the display of historic fashion in small freestanding glass cases under the day-lit dome. In 1934 a large section of a period room with an important mural by Paul Sandby was rescued during the demolition of Drakelow Hall in Derbyshire and acquired by the museum. Th is was erected against one wall of the gallery, where it has remained ever since. Th e gallery remained in this format until 1962 when a large mezzanine was added to make better use of the vast space. Until 2011 the mezzanine housed the V&A’s collection of musical instruments. It was during the 1962 refurbishment that the glass dome was covered with a secondary roof and the windows were covered and insulated. Large glass cases were built below the mezzanine level and around the perimeter of the gallery walls. However, no lighting was added to the gallery, so that the only illumination in the entire space came from within the display cases. Th e gallery became very dark and the dome and architectural features were lost in the gloom. Subsequent minor revamps occurred in 1984 and 2005 during which the mosaic fl oor was carpeted and the cases were clad in dark wood, each adding further to the dark, dull, and shady atmosphere. By 2010 it had become clear that considerable work would be needed to make the gallery into a more inviting and accessible space, bringing it in line with the rest of the FuturePlan redevelopment.

3. PLANNING THE REFURBISHMENT

Planning for the refurbishment began in July 2010 with an assessment of the needs of the collection and the new display. Tom Emerson and John Ross of 6A Architects were appointed to lead the renovation. Preparation work began at the beginning of August with an audit of the collection, followed by 100 days of object decant and the de-installation of the fi nal temporary exhibition in February 2011. Th e empty galleries were handed over to contractors on the 21st of March 2011, by which time 6A Architects had devised a scheme to strip the gallery back in order to reveal as much of the original Aston Webb design as possible by removing the ‘improvements’ made during the 1960s and in succeeding revamps. Th is plan involved opening up the long boarded over original entrances, restoring the mosaic fl oor, and providing both ambient gallery lighting and exhibition lighting for the mezzanine. A lift would be installed to make the mezzanine fully accessible but the space would be left blank for changing displays of costume (fi g. 2). Th e plans provided for minimal improvements to the existing display cases. Ten of the existing display cases on the ground fl oor were to be removed to improve visitor circulation, while a further two cases were to be removed to make way for a fashion-specifi c retail space complete with changing room. Architectural restoration company Hare and Humphries was commissioned to undertake microscopic paint analysis in order to work out the original paint scheme for the gallery and its fi ndings were used to justify a cool, unifi ed, pale grey paint scheme throughout the gallery. As the Victoria and Albert Museum is a Grade I listed building and considered of national importance, the proposed changes to the gallery had to be passed by the conservation offi cers at Kensington and Chelsea Town Hall. Th e reinstallation of the collection was divided between two curatorial teams, with senior fashion curator Claire Wilcox overseeing the permanent display, while curators Sonnet Stanfi ll and Oriole Cullen planned and curated a temporary display, which became part of the museum’s season celebrating British art and design in the years between the London Olympic Games of 1948 and 2012, and was titled Ballgowns: British Glamour Since 1950. However, the conservation and mounting team was to be the same for both projects.

Textile Specialty Group Postprints Volume 23, 2013 59 AN OLD CASE OF NEW DISPLAY: CONTEMPORARY AND HISTORIC FASHION AT THE VICTORIA AND ALBERT MUSEUM

Figure 2: Scheme by Architects 6A for a brighter, airier, and more accessible gallery space

4. THE PERMANENT COLLECTION DISPLAY

Aft er many years of thematic displays of fashion, the decision was made to reinstate a chronological display pattern, starting at 1750. Th e gallery would be divided into four time-period sections: the 18th century, the 19th century, the early 20th century, and the late 20th century to the present. Th e design team allocated a color scheme to each of the sections. Th is would be used for the backdrop of the cases, the mannequin covers, the signage, and the metal work and mounts. Lead curator Claire Wilcox was keen to reintroduce the use of well-fi lled cases with as many contextual additions as possible, including textiles, accessories, magazines, furniture, and dolls in fashionable attire. Alongside these she was eager to display portraits, fashion plates, and photographs to enliven the interpretation of the collection. In total around 100 dressed fi gures would fill the 18 cases on the perimeter of the ground fl oor with an additional 400 contextual objects. Among these contextual objects, it became particularly important to include textiles and lace within the display as the museum’s textile galleries had also been closed and emptied to allow for another major phase of the FuturePlan. Each of the main time-period sections was to be anchored by a ‘star object’, such as the spectacular magenta crinoline used within the 19th century section, which is shown with examples of alternative fabrics, a cage crinoline, shoes, and the jewellery that might have been worn with the dress (fi gs. 3 and 4). While some of the items for display were familiar to the public from V&A publications and exhibitions, a concerted eff ort was made to display items that were long accessioned but never before displayed, such as a 1936 yellow knit bikini accessioned in 1971 (fi g. 5), as well as several new acquisitions. However, due to unforeseen circumstances, this ambitious display had to be planned, conserved, and mounted in only 18 months, and it was soon realized that the backbone of the display would be many garments from previous exhibitions including Th e Golden Age of Couture, Surreal Th ings, and 200 Years of British Fashion. Already conserved and mounted, these objects were ready for display, having recently returned from international tours.

Textile Specialty Group Postprints Volume 23, 2013 60 JOANNE HACKETT AND KEIRA MILLER

Figure 3: ‘Star object’, a crinoline gown of the mid 19th century [T.118-1979]

Textile Specialty Group Postprints Volume 23, 2013 61 AN OLD CASE OF NEW DISPLAY: CONTEMPORARY AND HISTORIC FASHION AT THE VICTORIA AND ALBERT MUSEUM

Figure 4: Plan and elevation for showcase 7, Gallery 40

4.1 CONSERVING AND MOUNTING THE PERMANENT COLLECTION Time and budget constraints also had an impact on conservation and mounting. Unfortunately, many items considered for inclusion in the permanent collection display were unable to be used due to lack of conservation time. Given the short-term nature of the project, and with the gallery only destined to display costume for the next seven to eight years, much of the major conservation work had to be curtailed. Conser- vation assessments became something of an exercise in fi nding objects from within the V&A’s vast collection that required little treatment or that complemented those pieces already conserved and mounted for other exhibitions. For example, the inclusion of the Zemire gown by Christian Dior, which had been displayed in Th e Golden Age of Couture, allowed textile conservator Frances Hartog to wet-clean and conserve the undergarment that had not been previously shown, as it would now be part of the contextual material displayed along with this ‘star object’ (fi gs. 6 and 7).

Textile Specialty Group Postprints Volume 23, 2013 62 JOANNE HACKETT AND KEIRA MILLER

Figure 5: Wool bikini, 1936 [T.294-1971]

Textile Specialty Group Postprints Volume 23, 2013 63 AN OLD CASE OF NEW DISPLAY: CONTEMPORARY AND HISTORIC FASHION AT THE VICTORIA AND ALBERT MUSEUM

Figure 6: Wet-cleaning the under dress for Dior’s Zemire [T.24:2-2009]

Textile Specialty Group Postprints Volume 23, 2013 64 JOANNE HACKETT AND KEIRA MILLER

Figure 7: Zemire, complete with petticoat, alternative bodice, and Dior ephemera

Textile Specialty Group Postprints Volume 23, 2013 65 AN OLD CASE OF NEW DISPLAY: CONTEMPORARY AND HISTORIC FASHION AT THE VICTORIA AND ALBERT MUSEUM

Th e theme of recycling continued with the mounting of the display, as only a small budget was allocated to buy new mannequins. Many old papier-mâché dress-forms from previous displays were cut down and modifi ed with Ethafoam and padding. A bodice from 1869 required a mannequin with a much exaggerated, corseted shape (fi g. 8). With the corresponding skirt fabric accessioned in pieces rather than as a garment, this object had never before been displayed despite having been in the collection since 1964. However, shown alongside a length of skirt fabric displayed on a hanging metal bar with magnets and a cage crinoline mounted on a Perspex frame, it demonstrated technological advances in the construction of garments, fabric, and undergarments in the mid to late 19th century (fi g. 9). While very few new mannequins were used, those that were, were purchased with ease of re-use in mind. As such, the new mannequins within the display were all selected on account of their simple and practical poses.

Figure 8: Ethafoam adaptations to a recycled mannequin

Textile Specialty Group Postprints Volume 23, 2013 66 JOANNE HACKETT AND KEIRA MILLER

Figure 9: Contextual display for 19th century dress [T.10-1964]

Textile Specialty Group Postprints Volume 23, 2013 67 AN OLD CASE OF NEW DISPLAY: CONTEMPORARY AND HISTORIC FASHION AT THE VICTORIA AND ALBERT MUSEUM

5. THE TEMPORARY DISPLAY

Th e inner circle of cases on the ground fl oor and the new exhibition space on the mezzanine are intended for a series of nine-month long temporary exhibitions. Th e fi rst of these was Ballgowns: British Glamour Since 1950. Th is was intended as a companion exhibition to British Design: 1948-2012 Innovation and Design in the Modern Age in the main exhibition space. Th e exhibition featured 72 dresses from noted British designers. Downstairs, the display consisted of historic dresses from the V&A collection arranged in cases to appear as if they were getting ready to attend a ball. Upstairs on the mezzanine a display of contemporary dresses, both accessioned objects and loans, would appear to be dancing along catwalks. As the dresses on the mezzanine would be on open display, preventing the public from handling the garments and keeping dust to a minimum were concerns that were raised early on in the planning phase. Previous exhibitions and dust studies had taught us three valuable lessons about costume on open display: use barriers to keep visitors at least 1 meter away from the costume, raise the costume at least 1 meter from the ground, and provide lowered ceilings where possible (Shah et al. 2011, Adams et al. 2011). Designer Emily Pugh developed a plan to keep the public away from the dresses on the mezzanine with a barrier of gigantic pearls providing the 1 meter separation between the public and the objects, which were placed on 1 meter high plinths. To protect them from dust, the gowns were arranged below domed, fabric- covered canopies, with one or two canopies per plinth.

5.1 CONSERVATION AND MOUNTING OF THE TEMPORARY DISPLAY Most of the dresses selected for this display were in good condition and only 600 hours of conservation was needed for all 72 gowns. Th e dresses were mounted onto a selection of full fi berglass fi gures supplied by both Rootstein and Proportion London Ltd, which were selected by the curators for their fl uid, dance-like poses (fi g. 10). It was necessary to make extreme modifi cations to a number of these fi gures in order to achieve a slim enough body shape for many of the gowns. While these adaptations—which included the removal of mannequin chests, hips, and waists (fi g. 11)—were quite arduous, they allowed us to use fi gures with far more of a sense of movement than those we might have chosen solely for their collections care credentials.

Figure 10: Ballgowns: British Glamour Since 1950

Textile Specialty Group Postprints Volume 23, 2013 68 JOANNE HACKETT AND KEIRA MILLER

Figure 11: Fiberglass mannequin adapted to fi t a particularly slim ball gown

Textile Specialty Group Postprints Volume 23, 2013 69 AN OLD CASE OF NEW DISPLAY: CONTEMPORARY AND HISTORIC FASHION AT THE VICTORIA AND ALBERT MUSEUM

Figure 12: Loss of gloss aft er eight weeks of display, with a small area cleaned for comparison

6. MONITORING AND MAINTAINING THE GALLERY SPACE

Th e real challenge of the ballgown exhibition, as it transpired, was not in mounting or conserving the garments, but in keeping the dresses clean. It became clear that the dust mitigation strategy was not working as well as we would have wished when the mannequins on the mezzanine were vacuumed aft er only eight weeks of open display (fi g. 12). Th e questions that arose: why was the display so dusty, and were the high dust levels a direct result of the recently completed building work? As the space was scheduled to house at least another fi ve temporary costume exhibitions we felt it necessary to investigate the wider implications Dust samples collected in specimen containers and on dusting cloths were sent for analysis by DustScan, who undertook light microscopy to determine the characteristics of the dust. Th eir report revealed that very little building-related debris was present and that the vast majority of the particles were natural and synthetic fi bers, including cotton and fl ax, strongly suggesting that the dust was visitor borne, mostly from clothing (Shah and Miller 2012).

Textile Specialty Group Postprints Volume 23, 2013 70 JOANNE HACKETT AND KEIRA MILLER

Having established this, we resolved to monitor the space to determine the dust deposition rate. A loss of gloss method was used to measure the rate of accumulation. Standard microscope slides were exposed to dust by placing them on horizontal surfaces around the display, both on the plinths and on other surfaces away from the display, to measure the ambient dust level within the space. Th e percentage of loss to surface gloss was then measured every week. A table of results showed that most slides had between 4% and 6% loss of gloss each week (table 1). In real terms, this meant that even within the space of a week dust deposits would alter a garment’s appearance, which would be particularly noticeable on garments with shiny surfaces or surface decoration. When compared to visitor numbers it became clear that the level of dust within the display directly correlated with number of visitors (table 2). Th is confi rmed earlier fi ndings by the National Trust when dust monitoring their buildings at various times of the year, including times when properties were closed for the season (Lloyd et al. 2007). While our dust precautions had made modest limitations to the level of dust on the objects, there was still far too much to ignore, particularly as the conservation cleaning of the garments took on average 6 hours each time it was carried out, which over the duration of the exhibition added up to 18 hours in total.

7. PLANNING FOR FUTURE DISPLAY

Th e next exhibition scheduled for display on the mezzanine level of Gallery 40 was Club to Catwalk: London Fashion in the 1980s, which opened in July 2013, again without cases. With prior knowledge of the space we were able to make informed decisions when assessing the objects for display. While many of the costumes chosen for the mezzanine display were relatively sturdy, any objects that could not be adequately cleaned by either surface cleaning or wet cleaning were rejected, including all aged P.V.C and latex (fi g. 13). As a result of this, designer Th eresa Coburn, who made costumes for band such as Th e Clash and Siouxsie and Th e Banshees, was invited to make a replica of one of her gothic latex creations rather than put extant garments at risk.

Table 1: Percentage loss of gloss across fi ve locations within the gallery % loss of gloss week-1 (60Њ angle) Locations Date glass slide placed in location A B C D E 20 Sep 2012 5.1% 5.2% 4.6% 5.7% 27 Sep 2012 6.7% 6.8% 6.1% 6.7% 04 Oct 2012 7.7% 6.8% 6.9% 5.8% 5.1% 11 Oct 2012 6.8% 5.6% 4.2% 4.3% 3.7% 18 Oct 2012 6.5% 6.5% 6.3% 6.0% 6.0% 25 Oct 2012 6.9% 7.8% 7.7% 6.2% 7.0% 01 Nov 2012 5.8% 6.8% 4.1% 6.0% 5.4% 20 Nov 2012 6.1% 5.9% 4.9% 5.1% 3.4% 27 Nov 2012 7.5% 5.4% 5.4% 3.8% 3.5% 04 Dec 2012 2.9% 3.1% 3.4% 3.3% 3.0% 11 Dec 2012 3.2% 2.8% 2.9% 3.0% 2.6%

Textile Specialty Group Postprints Volume 23, 2013 71 AN OLD CASE OF NEW DISPLAY: CONTEMPORARY AND HISTORIC FASHION AT THE VICTORIA AND ALBERT MUSEUM

Table 2: Comparison: percentage loss of gloss compared to visitor numbers over 11 weeks

However, applying this criterion to the exhibition due to open in spring 2014 titled Wedding Dress: 1775–2014, very few of the garments met the standards required for open display. As a result, for the duration of this exhibition the mezzanine level will be fi tted with a number of brand new display cases, although a small selection of items will remain on open display, protected within closer fi tting individual canopies.

Figure 13: Club to Catwalk: London Fashion in the 1980s. Open display excluding objects too vulnerable to wet-clean or surface clean

Textile Specialty Group Postprints Volume 23, 2013 72 JOANNE HACKETT AND KEIRA MILLER

8. CONCLUSIONS

Th e renovation of Gallery 40 has without a doubt improved the look and feel of the gallery, even with the retention of the old cases and display lighting. Th e 4000 hours spent conserving and mounting the many objects for display, while not enough time to treat everything we might have liked, allowed for the display of several items never before displayed, as well as bringing about a re-evaluation of how we display and select objects. Henceforth, a program of object rotations will allow for the gradual inclusion of many items deemed too time-consuming to conserve in time for the initial opening in 2012. Moving into a radically altered gallery space on the mezzanine and adapting to large-scale open display has necessitated an amount of investigation into how museum visitors impact on the objects they come to observe. While the design and build of Ballgowns: British Glamour Since 1950 was almost entirely successful in preventing the audience from handling any of the objects, the build-up of visitor-borne dust proved to be something of a predicament, which will certainly continue for as long as the space remains dedicated to fashion display. By using this initial exhibition as an opportunity to investigate and understand the implications of visitor-borne dust and the environmental parameters of the exhibition space, we hope, as a department, to be better able to guide the design teams for future displays, ensuring we display our fashion collections in a way that is both safe for the objects and engaging and accessible for our visiting public.

ACKNOWLEDGEMENTS

Th e authors would like to thank Bhavesh Shah of the V&A Science section for coordinating the dust monitoring and for general advice. Th ey would also like to thank designer Emily Pugh, and architects Tom Emerson and John Ross for allowing us to access and reproduce their designs for this paper.

REFERENCES

Adams, A., S. Hunter, and B. Shah. 2011. ‘Diaghilev and the Golden Age of the Ballets Russes, 1909–1929’ exhibition dust monitoring exercise. Unpublished report no. 11/32/BS. V&A Science Section

Lloyd, H., P. Brimblecombe, and K. Lighgow. 2007. Economics of dust. Studies in Conservation 52. 135–146.

Shah, B., S. Hunter, and S. Adams. 2011. Dust to dust. Access to access. V&A Conservation Journal 59. 19–20.

Shar, B., and K. Miller. 2012. ‘Ballgowns’ exhibitions dust monitoring exercise and DustScan report. Unpub- lished report no. 12/98/BS. V&A Science Section

SOURCES OF MATERIALS MANNEQUINS Proportion London Ltd. 9 Dallington St Clerkenwell London EC1V 0LN UK

Textile Specialty Group Postprints Volume 23, 2013 73 AN OLD CASE OF NEW DISPLAY: CONTEMPORARY AND HISTORIC FASHION AT THE VICTORIA AND ALBERT MUSEUM

Tel: ϩ44 (0)20 7251 6943 Fax: ϩ44 (0)20 7250 1798 [email protected]

Adel Rootstein Ltd 9 Beaumont Avenue London W14 9LP UK Tel: ϩ44 (0)207 381 1447 Fax: ϩ44 (0)207 386 9594 [email protected]

AUTHOR BIOGRAPHIES

JOANNE HACKETT is a senior textile conservator at the Victoria and Albert Museum. Before joining the V&A in 2006 she was associate textile conservator at the Indianapolis Museum of Art for two years. Prior to that she was a textile conservator at the Fine Arts Museums of San Francisco for six years. She graduated with an M. S. from the Winterthur/University of Delaware Program in Art Conservation in 1998. Address: Textile Conservation, Victoria and Albert Museum, South Kensington, London, SW7 2RL, UK; [email protected]

KEIRA MILLER graduated from Wimbledon School of Art, gaining a BA (Hons) in Th eatre Design: Costume Interpretation. Since graduation, she worked briefl y for the BBC wardrobe department, before taking up a post in the Textile Conservation Department of the Victoria and Albert Museum, where she specializes in the mounting and packing of textiles and dress objects. Address as for Hackett; [email protected]

Textile Specialty Group Postprints Volume 23, 2013 74 EMERGENCE OF “ANTIQUE” SYNTHETIC TEXTILES

EBENEZER KOTEI

ABSTRACT—Aft er the Du Pont Company began the manufacture of nylon and gave birth to the synthetic polymer fi ber industry the world was presented with an alternative raw material for clothes-making that did not involve the use of natural fi bers or come from natural sources. Hagley Museum and Library in Wilmington, Delaware is the depository for the myriad of sample fi bers, the fi rst spools of fi bers to roll off the mills, the fi rst nylon shirt ever made, a women’s slip made out of parachute grade nylon, and more. Th e collection tells the story of the technological and human challenges that had to be overcome in order to turn petroleum into textile fi bers, and how to romanticize “wrapping ourselves in plastic” as it were. Clothes have been collected and cherished for their beauty, their provenance, and for their unique source of materials. To many people, antique clothing means clothing made of natural fi bers, synthetics simply haven’t been around long enough. Recently clothes made of these latter day materials are also making their way into collectors’ homes and institutions so it is important for textile conservators to pay attention to these man-made materials. So where does synthetic clothing fall in the admixture of textiles worth preserving? What are the conservation issues? Th is paper, with the dates of invention, dates of production, and description of the characteristics of the materials, will hopefully help guide conservators in determining how to approach the conservation of these materials.

SURGIMIENTO DE TELAS SINTÉTICAS “ANTIGUAS”—Después de que la compañía Du Pont lanzara el nylon y creara la industria de la fi bra de polímeros sintéticos, el mundo tuvo una materia prima alternativa para la confección de ropa que no contemplaba el uso de fi bras naturales ni provenía de recursos naturales. El Museo y Biblioteca Hagley de Wilmington, Delaware es el depositario de miles de muestras de fi bras, las primeras bobinas de fi bras que se fabricaron, la primera camisa de nylon, enaguas femeninas hechas con nylon para paracaídas, y mucho más. La colección cuenta la historia de los desafíos tecnológicos y humanos que debieron ser superados para convertir el petróleo en fi bras textiles, y cómo ver el lado positivo de “envolverse en plástico” en aquellas épocas. Los textiles confeccionados con fi bras y telas hechas por el hombre están colmando los depósitos de los museos de todo el mundo. Para la mayoría de las personas, la ropa antigua es sinónimo de ropa hecha con fi bras naturales. Entonces, ¿dónde encuadra la ropa sintética dentro de la gran variedad de textiles que vale la pena preservar? ¿Cuáles son los problemas de conservación relacionados con las fi bras sintéticas? Esperamos que este documento, con las fechas de invención, las fechas de producción y la descripción de los materiales, ayude a los conservadores a saber cómo abordar la conservación de estos materiales.

1. INFLUENCE OF SYNTHETIC TEXTILE FIBERS ON SOCIETY

It began with humankind trying to cover up its nakedness. Leaves, branches, tree barks and animal skin were used. Th is chore went from the mere need to cover up, to exhibiting the glamour of the material, its unique fi t to the human body, and the desire to outshine. For thousands of years designers utilized the natural materials they found in their surroundings; silk, cotton, wool, linen. EMERGENCE OF “ANTIQUE” SYNTHETIC TEXTILES

In 1883 clothing materials began to evolve when scientists in France and England came out with clothing materials that were diff erent from what was known. Th en in 1935, there came a major breakthrough with the discovery of nylon by an American icon company, E. I. Du Pont de Nemours and Company. Th e discovery of nylon, the deeper understanding of polymers that followed, and the myriad of textile fi bers that came in its wake, completely revolutionized the fashion industry. Th ese new materials can be manufactured to look like, feel like, be anything and do anything the imagination desires. Fashion went from big business to mega business. Weather and atmospheric conditions on earth and in space became less of a barrier to humankind.

2. RAYON AND ACETATE

Rayon fi bers are made by dissolving cellulose in caustic soda to form alkaline cellulose. Th is is bathed in carbon disulfi de where the molecular weight of the cellulose is reduced to an appropriate size to form cellulose xanthate. Th e xanthate is dissolved in dilute sodium hydroxide solution to form a very viscous substance called viscose, which is pushed through spinnerets to form rayon fi bers. Rayon is therefore only semi-synthetic and composed of regenerated cellulose. Acetates are made by dissolving cellulose in acetic acid and acetic anhydride to form cellulose acetate. Unlike rayon, cellulose acetate is a chemical compound and its properties are diff erent from those of cellulose or those of rayon. In 1884 a patent for the fi rst commercially successful man-made fi ber was granted to Count Hilaire de Chardonnet in France, for cellulose spun from gun cotton dissolved in ether and alcohol. In 1898 Charles H. Stearn, assisted by Charles Fred Topham, produced the fi rst “artifi cial silk” from viscose and received an English patent (Kaufman, 1993). In 1910 the fi rst “artifi cial silk” produced in the US was by the American Viscose Company formed by Samuel Courtauld and Company, Ltd. at Marcus Hook, Pennsylvania. In 1919 Camille and Henri Dreyfus of England off ered Celanese, the fi rst commercially successful acetate yarn. In 1921 the Du Pont Fibersilk Company, a subsidiary of E. I. Du Pont De Nemours and Company, began production of “artifi cial silk” fi lament yarn at its Buff alo, New York plant. Th ree years later in 1924 the Du Pont Cellophane Co. began production of Celanese acetate yarn in the United States. In the same year, Du Pont coined the name rayon for “artifi cial silk” (Holmes and Hicks, 1983).

2.1 EARLY YEARS OF RAYON Th e early rayon was held in low esteem: • Th ere was only the standard 150 denier, 18 fi lament yarn • Rayon could be highly colored like silk but it was harsh to the touch, weak and lacked uniformity Despite the poor properties rayon found use in women’s hosiery, and some foundation garments (fi g. 1).

2.1.1 Properties Of Rayon • Highly moisture absorbent • Wrinkles easily and becomes weaker when wet • Soft , comfortable feel • Sensitive to mildew, silverfi sh and other clothes insect

Textile Specialty Group Postprints Volume 23, 2013 76 EBENEZER KOTEI

Figure 1: Despite the poor properties, rayon found use in women’s hosiery

2.1.2 Properties Of Acetates • Luxurious feel and appearance • Special dyes developed for acetates allow for a wide range of colors and luster • Soft and excellent drapeability • It dries fast • It is shrink-resistant • Resistant to moths and mildew (Textile Learner, 2012) Table 1: Comparism of the properties between acetates and rayon

Property Acetate Rayon Sensitivity to water Loses strength when wet Loses strength when wet Dyeability Special dyes required Easily dyed with direct, fi ber reactive vat sulfur dyes Biological attack Resistant to mold and mildew Attacked by mold and mildew Shrinkage Shrink-resistant Shrinks and wrinkles easily Chemical stability Weakened by strong alkaline Alkalis have little or no eff ect solutions but attacked by hot dilute acids

Textile Specialty Group Postprints Volume 23, 2013 77 EMERGENCE OF “ANTIQUE” SYNTHETIC TEXTILES

Determined to produce materials that would replace silk, Du Pont and other companies worked hard to improve the material. In 1934 Cordura® high tenacity rayon was developed with a strength of 1.0 gram per denier. Its tenacity was close to 3 grams per denier and was strong enough to be used for strengthening automobile tires and could be used even for military applications. Rayon staple fi ber was also produced and by the 1950’s the quality of rayon was high enough to fi nd uses in raincoats, night dresses, and other fi ne garments. (Holmes and Hicks, 1983).

3. NYLON

In 1935 “Polymer 66”, a polyamide superpolymer was synthesized by Wallace Carothers and his team at the Du Pont Experimental Station in Wilmington, Delaware. Now known as nylon 66, it was produced by the reaction between a diamine monomer called hexamethylendiamine and a diacid monomer called adipic acid. Th e resulting product is polymethylene adipamide, which is a superpolyamide. Hexamethylenediamine and adipic acid both have six carbons in their chain structure, hence the name Polymer 66 or nylon 66 (Corbman, 1983). In 1938 the Du Pont Company formally announced “nylon” as the generic name for all polyamide type superpolymers (Holmes and Hicks, 1983). Th e fi rst public showing of nylon stockings was at the San Francisco World’s Fair in February 1939. Newspaper accounts from the New York Fair in the same year stated “Du Pont has the best leg show at the fair” in response to Miss Chemistry’s modeling of their nylon stockings. Th e following October, nylon hosiery was put on public sale for the fi rst time, only in Wilmington, Delaware—maximum three pairs per customer (fi g. 2). In May 1940 nylon hosiery went on sale in stores throughout the country. In 1941 nylon darning and sewing thread were introduced. Nylon articles became very popular and the public’s yearning for it grew strong, only to be dashed by the outbreak of World War II.

Figure 2: Nylon hosiery that was shaped and stabilized by pre-boarding (a heat-setting process)—Hagley Museum Collection

Textile Specialty Group Postprints Volume 23, 2013 78 EBENEZER KOTEI

Figure 3: Circa 1945 Air Force fl ight suit made of nylon—Hagley Museum Collection

Figure 4: White nylon parachute—Hagley Museum Collection

Textile Specialty Group Postprints Volume 23, 2013 79 EMERGENCE OF “ANTIQUE” SYNTHETIC TEXTILES

Figure 5: Nylon shirt made with fabric designed for fi ltering blood Figure 6: Woman’s slip made with World War plasma—Hagley Museum Collection II surplus nylon—Hagley Museum Collection

3.1 DIVERSION OF NYLON World War II broke out in 1941 and in 1942 the United States War Department ordered all nylon materials to be diverted for the war eff ort. Nylon was used for making parachutes, tow ropes for military trucks and planes, fl ight suits (fi g. 3), tampons for plugging bullet holes in boats, and mosquito nets for soldiers in the tropics. Every nylon fi ber manufactured between February 10, 1942 and VJ-Day in August 1945, was made strictly for World War II use. Yet, some of it was sold for dress-making as reclaimed parachute material, as war excess or it simply found its way onto the Black Market due to high demand (fi gs. 3, 4, 5 & 6). Aft er World War II it took some time to get back to producing nylon for consumption by the general public. However nylon caught on like wild fi re and was soon known in most households. In 1946 nylon was introduced in tricot fabrics commercially. In August 1953 woven nylon bed sheets and pillow cases were introduced by Pepperell Manufacturing Company in Massachusetts and following December, opaque (dull) nylon yarn production began. (Du Pont Company Textile Fiber Department, 1938; Holmes and Hicks, 1983).

3.2 NYLON APPLICATIONS

Apparel: Blouses, dresses (fi g. 7), hosiery, lingerie, underwear, raincoats, ski apparel, windbreakers, swimwear, and cycle wear. Home Uses: Tooth brushes, bedspreads, carpets, curtains, upholstery. Industrial and Other Uses: Tire cord, hoses, conveyer and seat belts, parachutes, racket strings, ropes and nets, sleeping bags, tarpaulins, tents, thread, monofi lament fi shing line, dental fl oss.

Textile Specialty Group Postprints Volume 23, 2013 80 EBENEZER KOTEI

NYLON is not a trade-mark name. Since Du Pont coined the name to apply to any of various synthetic polyamide fi bers formed by the condensation reaction between an amino group of one molecule and a carboxylic acid group of another, the following materials also qualify to be called nylon: Table 2: Other types of important nylon textiles

Date Material Properties Some applications 1963 Nomex® Fiber HT-1 High temperature resistance Fire-resistant uniforms and garment 1968 Qiana® Superior-to-silk properties. Silky High end fashion for women. appearance, fi ne supple hand and Also used in men’s ties and feel. Resistant to water spotting, shirt. resistant to chemicals, heat and light, wrinkle resistant and high neatness retention. 1973 Kevlar® - High strength industrial fi ber Bullet-proof materials Aromatic polyamide

3.2.1 Qiana Qiana® (fi g. 8), was particularly important because it boasted the superior-to-silk qualities listed above. But it faced an uphill battle from the start because silk had reigned for four centuries and could not be easily replaced. Th ere were manufacturing problems with handling; the process required new and complex technologies, and its limited specialty market did not justify its high manufacturing cost. Th ese problems were mostly resolved by the 1970s, but other problems emerged beyond the control of the Du Pont Company: • A weak textile economy emerged at the start of the 1970’s • Casual wear was growing rapidly as elegant dress wear declined in popularity • Th ere was a dramatic increase in fi ber producers and fi ne denier polyester became available at cheaper prices (Holmes and Hicks, 1983). Th ough Qiana was made into many high end women and men clothing it went out of production without making a high impact on the industry.

3.2.2 Properties of Nylon Fabric • Nylon is a thermoplastic and will melt rather than burn • It is strong and durable • Resistant to abrasion • Retains its shape very well • Resistant to moths and mildew, but sensitive to ants and roaches • Absorbs and holds body oils • Collects static electricity • Yellows when exposed to light, and may pill

Textile Specialty Group Postprints Volume 23, 2013 81 EMERGENCE OF “ANTIQUE” SYNTHETIC TEXTILES

Figure 7: McMullen dress of fi rst 100% spun nylon—Hagley Museum Collection

Textile Specialty Group Postprints Volume 23, 2013 82 EBENEZER KOTEI

Figure 8: Ladies’ designer evening gown made of white Qiana®—Hagley Museum Collection

Textile Specialty Group Postprints Volume 23, 2013 83 EMERGENCE OF “ANTIQUE” SYNTHETIC TEXTILES

4. ACRYLIC AND MODACRYLIC

Acrylic fi bers are made from acrylonitrile polymer. Du Pont scientists wanted to make useable fi bers out of simple vinyl type chemicals; polyacrylonitrile was cheap and plentiful. Typical co-monomers are vinyl acetate or methyl acrylate. To be called acrylic fi ber in the U.S, the polymer must contain at least 85% acrylonitrile monomer. Modacrylic is a manufactured fi ber in which the fi ber-forming substance is any long chain synthetic polymer composed of less than 85% but at least 35% by weight of acrylonitrile units (-CH2CH[CN]-)X (Kadolph and Langford, 2002). Th e fi ber was made by dissolving polyacrylonitrile in dimethyl formamide. In August 1948 Du Pont announced that the name for its new acrylic fi ber would be Orlon®. Th e fi rst continuous yarn of Orlon® acrylic was produced in July 1950.

4.1 PROPERTIES OF ACRYLICS Acrylic fi bers are: • Smooth • Moth and mildew proof • Resistance to acid attack • Resistant to deterioration by sunlight • Fairly resistant to abrasion and slightly weaker than nylon and polyester • Highly wrinkle resistant

4.2 ACRYLIC APPLICATIONS Apparel: Sweaters, socks, fl eece wear, circular knit apparel, sportswear and children’s wear Home Furnishings: Blankets, area rugs, upholstery pile, luggage, awnings, outdoor furniture Industrial Uses: Asbestos replacement, concrete and stucco reinforcement. Its smoothness, sunlight resistance, and wrinkle resistance made it perfect for awnings, drapes and win- dow curtains and its resistance to acids made it perfect for chemical fi lters, automobile tops and other uses that exposed it to the elements. Fabrics made from the staple form of Orlon® became the most sought aft er materials for knit goods. Th e continuous yarn was not so successful. Further innovation led to improvements in the process, material and mechanical knowledge, leading to the development of fi bers with special properties. One such development was bi-component fi bers in which two polymers of diff erent properties are pulled together to form a single fi ber. Th e two polymers behave diff erently under diff erent moisture and temperature conditions, thereby imparting special characteristics to the fi ber as a whole. In 1960 Orlon® Sayelle® was introduced. Th e two components in “Sayelle” produced a reversible crimp, expanding on wetting and contracting on drying, resulting in a permanent spiral crimp. It felt remarkably wool-like. It was immediately accepted by the public and developed into prestige knitwear (fi gs. 9 & 10) (Holmes and Hicks, 1983).

4.3 WASH AND WEAR In 1959 a machine-washable-with-no-pressing seersucker suit (60% Orlon®/40% cotton) was introduced by Haspel Brothers of New Orleans. Seeing the suit being washed in a machine, Collins Th ompson initiated the term “wash and wear.”

Textile Specialty Group Postprints Volume 23, 2013 84 EBENEZER KOTEI

Figure 9: Man’s Orlon® Sweater—Hagley Museum Collection

Figure 10: Acrylic staple is best suited for knits—Hagley Museum Collection

Textile Specialty Group Postprints Volume 23, 2013 85 EMERGENCE OF “ANTIQUE” SYNTHETIC TEXTILES

5. POLYESTER

Th e history of the development of polyester fi bers is worth a mention. Before Du Pont invented nylon, they invented polyester polymer which they called Polymer V. Polymer V was made out of aliphatic molecules. Th ough it had high molecular weight the ester linkages were too weak and unstable. It was a low melting point material that was prone to hydrolysis. It could not be made into textile fi bers. Th e team shelved the material and started working on polyamides instead, leading to the discovery of nylon. Two British scientists, Winfi eld and Dickson picked up from where the Carothers team had left it and replaced the aliphatic molecules with aromatic ones. Th e resulting polyethylene terephthalate (PET) polyester was more resistant to hydrolysis and had a higher melting point. Aft er the two had secretly licensed the material in Britain, the British chemical company, Imperial Chemical Industries (ICI) acquired the rights and became the fi rst to actually make polyester fi bers (Hounshell and Smith, 1988). Du Pont meanwhile went back to do more research on the material they had shelved, also using terephthalic acid, and came up with PET. Aft er discovering that ICI had also produced PET, the two companies compared notes and found that Du Pont’s yarn had superior properties to that of ICI. Du Pont had a better catalyst for the ester exchange and had more routes for preparing the polymer. Aft er patent and licensing issues were settled, the fi rst commercial yarn of polyester fi ber was spun by Du Pont at Seaford, Delaware in April 1949. Full production followed in September 1950 (Holmes and Hicks, 1983).

5.1 PROPERITES OF POLYESTER FIBERS Polyester fi bers are: • Insensitive to water • Resistant to common solvents • Tear resistance very close to that of wool, even when wet. Th ey are strong materials that hold their strength through laundering and resist stretching through wash • Resistant to moth, mildew and silver fi sh. In the early days it was common for polyester fabrics to be made out of 100% polyester. Several clothing items made out of 100% polyester came with all the accompanying properties of the pure material; dyeing was almost impossible because the commonly known dyes would not work, developed heat in sewing machines, had static. Dyeing techniques were gradually improved, using dispersed colors. All the colors required the use of a very expensive benzoic acid carrier, making polyester too expensive for most consumers. Th e characteristics of the original material were not very good at fi rst and the manufacturing processes were complex and unreliable. Further innovations, improvements and ingenuity led to refi nements. In February 1950, Fiber V was introduced in men’s tropical-weight tailored suits by Witty Brothers. In August 1950 Fiber V staple yarn was produced at Seaford and made available commercially (fi g. 11). In 1951 the Du Pont Company announced Dacron® as the generic name for its polyester fi ber. In 1952 home sewing thread of Dacron® was introduced, made by Brooks Brothers. In January 1953, 60/40 Dacron®/cotton blend concept was introduced by Brooks Brothers in Brooksweave shirts. In March the same year, Dacron® polyester staple production was started by Du Pont. In January 1955 Dacron® polyester fi berfi ll was announced. In 1960 the fi rst woven 65/35 Dacron®/cotton blended bed sheets were introduced by Wamsutta.

Textile Specialty Group Postprints Volume 23, 2013 86 EBENEZER KOTEI

Figure 11: Suit made from fi rst Dacron® tow made on a commercial machine—Hagley Museum Collection

Textile Specialty Group Postprints Volume 23, 2013 87 EMERGENCE OF “ANTIQUE” SYNTHETIC TEXTILES

In 1961 Type 88 Dacron®, the fi rst commercial spiral crimp polyester fi berfi ll was introduced. In 1973 Hollofi l, Dacron® polyester hollow fi lling fi ber with greater bulk than similar weight conventional fi lling fi ber was introduced (Holmes and Hicks, 1983).

5.1.1 Polyester Applications • Th e continuous fi lament yarn is used in curtains, dress fabrics, men’s shirts. It is also used in bed sheets, pants, and suits. • Th e staple fi ber is used mostly as a blend in wool for making men and women’s clothing, knits and other washables.

6. SPANDEX (LYCRA)

Th e need to create an elastic material for figure control in the garment industry was tough to master. Designers stretched several ribbons of rubber, which they held taut and wrapped together with cotton threads or yarns. Th ese provided elasticity for some garments, but had the disadvantage of being heavy, stiff and bulky. In the mid-1950’s Du Pont began the manufacture of an elastomeric yarn based on segmented polyurethane. Th e fi ber could be stretched to many times its length but would spring back to its original length. Th ey called it Fiber T-80. In 1958, aft er further refi nement, it was designated Fiber K and was sold for such applications as foundation garments, surgical hose and swim suits. It had high restraining power accom- panied by good stretch, so only small amounts were needed in fashion design. Fiber K came out of Type 124 segmented polyurethane fi ber and could be used “as-is”. It provided the elasticity needed for fi gure control, tightening, or stretching. On October 28, 1959 Fiber K was announced and trademarked as Lycra®. Commercial production started in March 1962.

6.1 PROPERTIES OF LYCRA SPANDEX Th e original Type 124 fi ber: • Was very good as spandex but had poor aging properties • Was poor in mechanical quality and fi nishing • Discolored on exposure to sunlight • Discolored on exposure to chlorine bleaches and atmospheric aldehydes. • Discolored in the nitrogen oxides of gas-fi red clothes dryers Aft er a series of tests and attempts to fi x the problem failed, the structure of the polymer was changed and a new fi ber Type 125 was produced. Type 125 had: • Superior whiteness retention • Reduced discoloration in sunlight • Less restraining power than 124 Seeking further improvements, Type 126 came out in 1967, also with diff erent characteristics. In the long run all three types remained on the market, and customers used diff erent types for diff erent applications.

Textile Specialty Group Postprints Volume 23, 2013 88 EBENEZER KOTEI

Despite the complex nature and manufacturing diffi culties of spandex, several other companies made it: U. S. Rubber Company had “Vyrene”, Chemstrad had “Blue C”, American Cynamid had “NUMA”. Lycra from DuPont had the best properties and controlled the market. Towards the end of the 1960’s Spandex sales began to wane as consumption of foundation garments, in the sense of powerful corsets and girdles, began to decrease (Holmes and Hicks, 1983).

7. BLENDS

Blending is a concept that leverages the individual properties of two or more fi bers to create a fabric with diff erent, desirable properties (fi gs. 12a &12b). Blending may be done before or during weaving of the fi bers into fabrics. Some of the advantages of blending are: • Increase in abrasion resistance • Decorative eff ects with diff erential dye pick-up • Increase in pilling resistance • Increase in bulk without an accompanying increase in weight Unless blending is done carefully and thoughtfully certain problems may be encountered in the fi nished product that may aff ect its application. Diff erences in elastic modulus of the two materials may lead to reduction in tensile strength. Under load, the higher modulus material such as cotton may break before the lower modulus material such as nylon.

Figure 12a: Baseball uniform worn by New York Yankee’s pitcher Bobby Shantz was A wool/nylon blend—Hagley Museum Collection

Textile Specialty Group Postprints Volume 23, 2013 89 EMERGENCE OF “ANTIQUE” SYNTHETIC TEXTILES

Figure 12b: Baseball uniform worn by New York Yankee’s pitcher Bobby Shantz was A wool/nylon blend—Hagley Museum Collection

8. STORAGE OF SYNTHETIC TEXTILES

Synthetic textiles are ideally stored in environmentally controlled storage, following typical museum protocol such as reducing light exposure and folds in garments, and using conservation-grade materials. At the Hagley Museum the storage room is kept at constant temperatures of 62 F and relative humidity of 45%. Rayon and acetate objects benefi t from cold storage as well as the use of oxygen scavengers. Acid fume absorbers such as zeolite, are used to scavenge degradation products in closed boxes.

9. CONCLUSIONS

Synthetic textiles have been around for over eight decades and now appear in almost every fi ne fashion in every country. Clothing made of synthetic fi bers is fi lling storage rooms and galleries in some of the biggest and richest museums around the globe. Antique synthetic textiles have emerged. Yet synthetic textiles have not enjoyed the same care and protection as textiles made of natural materials. Th ough environmentalists will tell us that plastics are non-biodegradable, for the purposes of textile conservation professionals, synthetic textiles do degrade enough to warrant the careful care, handling and storage that natural fi ber textiles have enjoyed over the decades. Colors fade, dyes migrate, some materials shrink in water and other wash solutions, materials melt in excessive heat, materials wrinkle with poor storage or display, and the semi-synthetics such as rayon are attacked by biological agents of decay. Conservation of clothing made of synthetic fi bers should become a priority for all conservators.

Textile Specialty Group Postprints Volume 23, 2013 90 EBENEZER KOTEI

REFERENCES

Hounshell, D. A; J. K. Smith, 1988. Science and Corporate Strategy. Cambridge University Press, Cambridge, New York.

Holmes, D. F; E. M. Hicks: A History of the Du Pont Company’s Textile Fibers Department. E. I. Du Pont de Nemours and Co, Wilmington, Delaware, 1983.

Du Pont Company Textile Fiber Department, 1938. Du Pont at Two World Fairs. Du Pont Company Magazine. September 6.

Bowen, H. J. 1938. A Son is born to the Royal family of Rayon. Du Pont Company Magazine. June 11.

Corbman, Bernard P. Textiles: Fiber to Fabric, 6th ed. McGraw-Hill, 1983

Rutledge, C. H. 1965. “Tongue-in-cheek” Report on First 25 Years of Nylon. Proceedings of the 61st Annual Convention of the National Association of Hosiery Manufacturers. Atlanta, Georgia.

Online: http://cdm15017.contentdm.oclc.org/cdm/ref/collection/p16038coll8/id/6 (accessed 05/08/16).

Textile Learner; January 2012. http://textilelearner.blogspot.com/2012/01/acetate-fi ber-characteristics-of.html

Kauff man, G. B. - Rayon: Th e First Semi-Synthetic Fiber Product; Journal of Chemical Education, 1993, vol. 70, number 11, pp. 887-893; American Chemical Society, Division of Chemical Education, Madison, WI, ETATS-UNIS (1924) (Revue)

Kadolph, S.J; A.L. Langford, (2002). Textiles (9th Edition). New Jersey: Pearson Education, Inc.

AUTHOR BIOGRAPHY

Since September 1988, EBENEZER KOTEI has been the Objects Conservator at the Hagley Museum and Library in Wilmington, Delaware. He is also an Adjunct Assistant Professor with Wintethur/University of Delaware Program in Art Conservation. Kotei graduated in Materials Science and Conservation of Cultural Property from the University of London, Institute of Archaeology (University College, London) in 1986. He worked at the Passmore Edwards Museum in East Ham, London, and from 1987 to 1988, served a one year, Andrew W. Mellon post-graduate fellowship with the Detroit Institute of Art, Detroit, Michigan. Address: Hagley Museum and Library, 298 Buck Road East, Wilmington, Delaware 19807. Tel. (302) 658 2400. E-mail: [email protected]

Textile Specialty Group Postprints Volume 23, 2013 91

MERGING DISCIPLINES: PREPARING A MATISSE SERIGRAPH FOR DISPLAY

YADIN LAROCHETTE

ABSTRACT—In the summer of 1946, Henri Matisse (1869–1954) lay bedridden as he directed his nurse to pin a series of white paper cut-outs inspired by memories of French Polynesia on to the walls of his Paris bedroom. Later, upon being approached by textile printer Zika Ascher, he agreed to turn the wall compositions into two large serigraphs, titled Océanie, la mer and Océanie, le ciel. Made with oil bound pigments on dyed, un-primed linen, each measures approximately 1.5 meters (5 feet) high by 3.7 meters (twelve feet) wide. Th e author undertook the conservation treatment and preparation for display of one of the prints of Océanie, le ciel. Unlike many of its counterparts, this particular serigraph had never been mounted before and was in good condition. Th e treatment included surface cleaning, paint consolidation, and crease reduction. Since stitching the body of the work to a support would have permanently altered the surface and been visually obtrusive, a system involving stitching the fabric only at the edges was developed to make the serigraph appear taut, while at the same time honoring the diff erent surface characteristics of the paint and dyed linen.

FUSIÓN DE DISCIPLINAS: PREPARACIÓN DE UNA SERIGRAFÝA DE MATISSE PARA SU EXHIBICIÓN—En el verano de 1946, Henri Matisse (1869–1954), postrado en su cama, pidió a su enfermera que colocara en la pared de su habitación de París una serie de recortes inspirados en recuerdos de la Polinesia Francesa. Más tarde, ante el pedido del impresor textil Zika Ascher, Matisse aceptó convertir las composiciones de la pared en dos grandes serigrafías tituladas Océanie, la mer y Océanie, le ciel. Hechas con pigmentos al óleo sobre lienzo teñido sin imprimar, cada serigrafía mide aproximadamente 1,50mt (5 pies) de alto por 3,70mts (12 pies) de ancho. El autor realizó el tratamiento de conservación y preparación para la exhibición de una de las impresiones de Océanie, le ciel. A diferencia de muchas de sus copias, esta serigrafía en particular nunca había sido mon- tada y se encontraba en buen estado. El tratamiento consistió en la limpieza de la superfi cie, la consolidación de la pintura y la reducción de los pliegues. Dado que coser el cuerpo de la obra a un soporte habría provo- cado alteraciones permanentes en la superfi cie y habría sido visualmente chocante, se desarrolló un sistema para coser la tela solo por los bordes para que la serigrafía quede tensa y se puedan resaltar las diferentes car- acterísticas de la superfi cie de la pintura y el lienzo teñido.

1. INTRODUCTION

1.1 THE ARTIST AND HIS NEW MEDIUM Henri Matisse (1869–1954), already a well-established painter, faced several challenges in the late 1930s and early 1940s. As World War II exploded around him, he and his wife of 41 years separated (Spurling 2005). During this diffi cult period, in 1941 at the age of 72, he was diagnosed with cancer and underwent a colostomy (Russell 1999). Th e surgery left him bed ridden or in a wheelchair most of the time, but this did not stop his creative drive. One of the means he found to express himself in his newly limited physical condition was to make paper cut-outs and collages. He called it “painting with scissors”. He created and MERGING DISCIPLINES: PREPARING A MATISSE SERIGRAPH FOR DISPLAY surrounded himself with images that inspired and lift ed his spirits, including memories of past exotic travels. “Only what I created aft er the illness constitutes my real self: free, liberated”, he is quoted as saying (Henri-matisse.net 2011).

1.2 THE ARTWORK As Matisse lay in bed in his Paris apartment in 1946, he asked his night nurse to pin cut-outs he had made out of white paper on to two adjacent walls in his bedroom (Berggruen and Hollein 2006, Spurling 2005). Th e shapes were reminiscent of a trip to Tahiti he had taken many years before. “It’s as though my memory had suddenly taken the place of the outside world. Sixteen years aft er my trip to Tahiti, my memories are fi nally coming back to me. Th ere, swimming every day in the lagoon, I took such intense pleasure in contemplating the submarine world.” (National Gallery of Australia 2013). Leaves, birds, algae, fi sh, and other wildlife lledfi the room, distilled as stylized white shapes on the beige walls, which were buff ed with age and city living. Th e wallpaper, which was most likely originally cream-colored, had not been changed in over 20 years and had darkened over time. One newspaper article said he initially directed the location of the shapes to cover stains (August 1998). Some time aft er decorating the walls, he was approached by two diff erent parties to design textiles. One was the Manufacture nationale de Gobelins, the government-run French tapestry workshop located in Paris. When Matisse off ered to turn the bedroom wall compositions into tapestry designs (or cartoons), the workshop demurred, concerned that the buff beige color of the wall would be too hard to replicate in dyed wool weft . Th ey suggested that a blue ground would work better, so he prepared two diff erent cartoons in blue. Th e results were Polynésie, le ciel (Polynesia, the sky) and Polynésie, la mer (Polynesia, the sea), each measuring approximately 2 meters (6 ½ feet) by 3.1 meters (10 ¼ feet). Zikmund “Zika” George Ascher, on the other hand, was starting a new textile printing enterprise and was up for the challenge. Born into a textile printing family in Czechoslovakia, he and his wife Lida settled in London during World War II. As Europe recovered, they continued the family tradition by opening a textile design and printing business, but with a new twist. Zika Ascher approached preeminent artists of the time, including Pablo Picasso, Henry Moore, Jean Cocteau, and Henri Matisse, among others, to create designs to print onto silk scarves. Th e scarves ended up being a success, and led to designing fabrics used by various haute couture designers such as Christian Dior, Guy Laroche, and Roberto Capucci. Matisse and Zika Ascher worked closely to achieve the most accurate reproduction of the wall compositions as possible. Matisse was reluctant to have a tracing of each wall made, concerned that there would be some inevitable re-interpretation by the person doing the tracing. Eff orts were made to photograph each wall and print it at actual size, but that proved ineff ective. In the end, each wall was eventually traced. Th e Gobelins representatives were correct in assessing the diffi culty of achieving an accurate match for the wall color. Several dyed linen swatches were sent back and forth between Ascher’s dyers and Matisse before a paintings restorer was brought in to the bedroom to color match by painting out color samples. Th ese colors still did not appeal, and Matisse ended up choosing a shade between two of the restorer’s samples (Christie’s 2008). Th e dyers, incredibly, under the circumstances, achieved the desired color. Ascher requested the assistance of the Belfast Silk and Rayon Company in Ireland to develop the white paint to replicate the cut-outs and to help with the silk-screening on the dyed linen fabric under his supervi- sion. FTIR spectroscopy of the paint used on one of the panels was conducted by the Canadian Conservation

Textile Specialty Group Postprints Volume 23, 2013 94 YADIN LAROCHETTE

Institute in 1998 in preparation for the conservation treatment of one of the Océanie, la mer serigraphs owned by the National Gallery of Canada. Th e results indicate that the paint was composed of titanium white, calcium carbonate, and kaolin in a drying oil medium (Vuori et. al. 1998). Th e two wall compositions of cut-out paper were now transformed to oil-bound paint on dyed, medium weight linen fabric: Océanie, le ciel (Oceania, the sky) and Océanie, la mer (Oceania, the sea). Th irty serigraphs of each design were made, measuring approximately 1.5 meters (5 feet) high by 3.7 meters (12 feet) wide. Henri Matisse was very pleased with the results, and signed them in ink with the edition number. Half of each edition was placed on the market and he kept the other half for himself. In 2010, one of the serigraphs of Océanie, le ciel found its way to Larochette Textile Conservation LLC to be prepared for display in a home (fi g. 1). Th e owners requested that the edition number not be mentioned and that they remain anonymous. At the time, Larochette Textile Conservation LLC was operating out of a studio space owned by and shared with Sharon Shore, proprietor of Caring for Textiles. All images included in this paper were taken at Shore’s studio.

Figure 1: Th e serigraph as it came into the studio

Textile Specialty Group Postprints Volume 23, 2013 95 MERGING DISCIPLINES: PREPARING A MATISSE SERIGRAPH FOR DISPLAY

2. CONDITION

Th e serigraph arrived to the conservation studio in pristine condition save sharp creases, which were most likely incurred during rolling. Th ere were other soft er undulations in the dyed fabric surrounding painted areas. Th ese latter distortions to the plane appeared to have been made during the printing process as the fabric reacted to the take-up of paint. Some paint abrasion was evident at creases within painted areas. Th e artwork arrived stretched on a wood strainer covered with non-archival foam core board. Extant sel- vages at the top and lower end, and cut edges at the sides wrapped to the back of the strainer and were pinned in place approximately every six inches with metal dressmaker pins. Th is mount was intended to be temporary, according to the information given. Tension was uneven, with undulations throughout the fabric, and surface particulate was noted on both the front and back of the textile once removed from the strainer. Th e paint seeped to the back in several areas during printing, suggesting that the fabric was saturated in these discrete areas. Also worth noting are the transfers of oil that apparently took place from one of the prints of Océanie, la mer to this serigraph. It appears as if the oil binding seeped from one work to the other (most likely they were stored stacked fl at at some point). A “ghost” of the images in Océanie, la mer appears as slightly darker shades of beige throughout unpainted areas of this print of Océanie, le ciel.

3. THE CONSERVATION TREATMENT

Th e conservation treatment proposal included removing the textile from the current strainer, surface cleaning the front and back, consolidating abraded areas of paint, reducing the sharp creases, and preparing the textile for display on a new strainer. A team was assembled to review treatment protocol, which included the author, textile conservator Cara Varnell and paintings conservator Joe Fronek. Sharon Shore was consulted throughout the project, and conservation assistant Kimberly Brunner helped with the treatment.

3.1 SURFACE CLEANING Th e front and back of the serigraph were surface cleaned with an adjustable suction Rainbow vacuum with a water trap, using a micro brush attachment for the non-painted areas (fi g. 2). Painted areas were cleaned with a soft sable brush, holding the vacuum nozzle at a close enough distance to suction air-borne debris. Large particulates were removed with tweezers. Th e textile was placed fl at on a series of tables, and a large plank or bridge, owned by Caring for Textiles, allowed access to the entire textile.

3.2 PAINT CONSOLIDATION Joe Fronek conducted the consolidation of the abraded paint using a combination of isinglass and gelatin with low heat. Th e aim was to prevent further fl aking. In-painting specks of loss was considered, but discussions with the owners led to a minimal approach.

3.3 CREASE REDUCTION Sharp, curvilinear creases present throughout the serigraph were reduced with humidifi cation via weighted Gore-Tex membrane and damp blotting paper. Templates were made marking the shape and location of each crease, and blotting paper was cut to fi t into each shape (fi gs. 3–5). Th is allowed for

Textile Specialty Group Postprints Volume 23, 2013 96 YADIN LAROCHETTE

Figure 2: Surface cleaning the back of the serigraph

maximum control for the discrete application of the custom-cut damp blotters. Th e blotters were left in place under weights for 25 to 45 minutes, depending on degree of planar deformation. Severely creased areas required two to three applications (fi gs. 6–7). Undulations located around painted areas which are apparently inherent to the work and made during the printing process, were left untreated.

4. PREPARATION FOR DISPLAY

During the design of the mount for the serigraph, the owners requested that the serigraph look “crisp”, like a painting. Th ey did not want the edges to be overly rounded or that the mount system read visually like a padded surface. Th is was indeed challenging. Deformations to the painted surface were apparent on the version of Océanie, la mer conserved by the Canadian Conservation Institute prior to treatment (Vuori et. al., 1998). From the images included in the report, it appeared as if stretching the fabric had expanded the paint layer. Upon releasing the tension on the fabric, the paint layer had nowhere to go but ripple.

Textile Specialty Group Postprints Volume 23, 2013 97 MERGING DISCIPLINES: PREPARING A MATISSE SERIGRAPH FOR DISPLAY

Figure 3: Polyester fi lm template exposing creases only

Figure 4: Blotting paper cut to fi t within template

Textile Specialty Group Postprints Volume 23, 2013 98 YADIN LAROCHETTE

Figure 5: Crease reduction set-up with Gore-Tex membrane, blotting paper, and polyester fi lm

Figure 6: Detail before crease reduction, under raking light

Textile Specialty Group Postprints Volume 23, 2013 99 MERGING DISCIPLINES: PREPARING A MATISSE SERIGRAPH FOR DISPLAY

Figure 7: Detail aft er crease reduction, under raking light

Th e question was: how to create a crisp look with out jeopardizing the pristine surface of the serigraph and honoring the characteristics of the un-primed, dyed linen and the white paint layers?

4.1 MOCK-UPS: FACED STRAINER VS. BEVELED STRAINER Two mock-ups were made in order to determine the most eff ect mounting system. Th e fi rst mock-up involved a faced strainer using Coroplast. Th e Coroplast was covered with polyester needle-punch batting to soft en the edges and provide a bit of “give” to the support (fi gs. 8–9). Th e second mock up included a wood strainer with beveled bars and recessed crossbars (fi g. 10), covered in fabric, to reduce potential marking of the textile. Th e outside edges of the bars were slightly rounded to prevent sharp creases. Th e clients chose this latter system. Paintings conservators use the term “loose lining” for this type of approach, since other types of linings on paintings are usually adhered to the artwork. Th e term “support fabric” will be used here, since it is a term textile conservators are familiar with; its mission, aft er all, is to off er overall support to the artwork.

4.2 THE SUPPORT FABRIC With the beveled strainer system established, the conservators needed a fabric that had a bit of a nap to it to help provide purchase and support the artwork overall. Stitching through the fl at, pristine surface of the artwork in the manner textiles are traditionally supported was out of the question, since it would leave permanent marks. Polyester suede was considered, but it was too bulky to provide clean corners, and most importantly, it was not commercially available in the width needed. Samples of two pieces of polyester suede

Textile Specialty Group Postprints Volume 23, 2013 100 YADIN LAROCHETTE

Figure 8: Mock-up of faced strainer before adhering Coroplast

Figure 9: Mock-up of faced strainer aft er adhering Coroplast and covering with polyester needle-punch batting

Textile Specialty Group Postprints Volume 23, 2013 101 MERGING DISCIPLINES: PREPARING A MATISSE SERIGRAPH FOR DISPLAY

Figure 10: Beveled strainer for second mock-up stitched together suggested that joins would show through the serigraph (fi gs. 11–12). Adhesives and fabric tapes were considered briefl y, but the team did not want to risk adhesives having direct contact with the back of the artwork. Heavy weight cotton “scenery” muslin, advertised as being used in theater backdrops, comes in the desired width and off ers enough body to help support the serigraph. Tests using various brushes to raise the nap on the muslin were made, and results suggested that a plastic bristle brush produced the appropriate texture to off er purchase for the serigraph. Th e fabric was washed, ironed, and brushed prior to being attached to the new strainer. Th e profi le of the fabric-covered mock-up is shown in fi gure 13.

4.3 THE STRAINER AND ATTACHING THE ART WORK Th e new strainer, built to spec by fi ne woodworker Robert Espinoza, was made of molding grade pine with slightly rounded edges to avoid creating sharp creases in the artwork when mounted. Th e wood was sealed with three coats of a clear Varathane (dipropylene glycol monomethyl ether, silicon dioxide, and methyl-n 2 pyrrolidone) manufactured by Rustoleum. Th e prepared support fabric was wrapped to the back of the strainer and stapled with Arrow Monel rust- proof staples. Th e conservators used a technique learned from Sharon Shore to prevent the various folds of fabric from sticking out at the corners of the strainer. Prior to attaching the support fabric, a triangular recess was fi led into each of the four strainer corners. Th e depth of the recess matched the depth of the folds of the two fabrics: that of the cotton support fabric and that of the linen serigraph. Each fi led corner was re-coated with sealant prior to stapling the support fabric.

Textile Specialty Group Postprints Volume 23, 2013 102 YADIN LAROCHETTE

Figure 11: Mock-up of beveled strainer with polyester suede; partially applied

Figure 12: Mock-up of beveled strainer with polyester suede fully applied and visible seams

Textile Specialty Group Postprints Volume 23, 2013 103 MERGING DISCIPLINES: PREPARING A MATISSE SERIGRAPH FOR DISPLAY

Figure 13: Mock-up of beveled strainer covered in fabric (corner shows soft ened edges)

Th e conservators also wanted to prevent direct contact between the back of the artwork and the staples. Th is was easy enough to do along the sides, where the support fabric was attached near the inner edges of the bars, allowing plenty of space between the staples and the edges of the serigraph linen. To prevent contact at the corners, the support fabric was hand-stitched in place rather than stapled, using Gütermann polyester thread (dtex 300(2)). Since the artwork had never been mounted for display before and there was no evidence of tacking margins or turn over edge, the responsibility was enormous. In order for the owners to determine the desired margin, other examples were examined via photos and several images were taken showing various margin widths. Th e serigraph was then pinned in place along the perimeter with entomology pins. Slight tension was applied, and measurements were taken during the process to ensure even distribution. Th e edges of the artwork were wrapped around to the back, and hand-stitched to the cotton support lining with Mettler no. 60 fi ne cotton machine embroidery thread. Protective strips of sheer polyester fabric were sub- sequently hand sewn over the cut edges to prevent further unraveling. Staples were covered with Lineco

Textile Specialty Group Postprints Volume 23, 2013 104 YADIN LAROCHETTE

Figure 14: Serigraph aft er treatment, showing profi le

frame sealing tape containing an aluminum barrier and a pH neutral acrylic adhesive. See fi gure 14 for a view of the mounted work.

5. CONCLUSION

Th e mounting technique proved eff ective. Th e brushed cotton fabric off ers suffi cient purchase to support the serigraph, with enough tension to off er a pristine, painterly look without jeopardizing the planarity of the paint layers and/or the surrounding un-primed fabric. Marking of the turned-over edges were kept to a minimum with the combination of slightly rounded strainer edges and hand stitched attachment. Before this paper was proposed for presentation, the author checked in with the owners to make sure that the system still looked good aft er three years. All is well and they are more than pleased. Fluctuations in temperature and humidity within the home do not appear to have aff ected the linen serigraph.

ACKNOWLEDGEMENTS

Th is project would not have been possible without the insight and talents of many, including Sharon Shore, textile conservator and proprietor of Caring for Textiles and Cara Varnell, textile conservator and proprietor of Textile Arts Conservation. Joe Fronek, head of paintings conservation at the Los Angeles County Museum of Art, was very generous with his time and expertise, and having both his mental and physical input was imperative. Anne Ruggles, former conservator of paintings, National Gallery of Canada (NGC), was equally generous with information regarding the condition of a version of Océanie, la mer owned by the NGC, and provided the author with the treatment report written and conducted at the Canadian Conservation Institute under the direction of Jan Vuori, senior textile conservator. Stephen Gritt, director of conservation at the NGC was very helpful in providing images for the oral version of this presentation. Laura Rivers, associate paintings conservator, J. Paul Getty Museum, was very patient and off ered great insight during our hikes as the author prepared the talk. Robert Espinoza, fi ne wood worker and builder of the strainer, has prepared displays for various museums in the Los Angeles area and brought new perspectives to the table. And last but not least, thanks to Kimberly Brunner for helping with the treatment and being such a good sport for working days on end on “the bridge”.

Textile Specialty Group Postprints Volume 23, 2013 105 MERGING DISCIPLINES: PREPARING A MATISSE SERIGRAPH FOR DISPLAY

REFERENCES

August, M. 1998. Th e artists: Henri Matisse. www.apnewsarchive.com/1998/Th e-Artists-Henri-Matisse/id- 8f992aa039416d2c623724ce665a190f (accessed 10/13/13). Berggruen, O. and M. Hollein. 2006. Henri Matisse: Drawing with scissors: Masterpieces from the late years. New York: Prestel. Christie’s. 2008. Sale 2045, lot 75. Henri Matisse, Océanie, la mer. http://www.christies.com/lotfi nder/drawings- watercolors/henri-matisse-oceanie-la-mer-5145347-details.aspx (accessed 10/14/13). Henri-matisse.net. 2011. Paper cut outs (gouaches découpés). www.henri-matisse.net/cut_outs.html (accessed 10/13/13). National Gallery of Australia. 2013. Henri Matisse, Océanie, le ciel. Online Catalog. http://nga.gov.au/international/Catalogue/Detail.cfm?IRNϭ148088 (accessed 10/13/13). Russell, J. 1999. Matisse: Father and son. New York: Henry N. Abrams Inc. Spurling, H. 2005. Matisse the master: A life of Henri Matisse: Th e conquest of colour 1909–1954. New York: Alfred A. Knopf. Vuori, J., D.D. Hartin, E. Moff att, and J. Sirois. 1998. Examination, analysis and stain removal testing of Oceanie, la mer for National Gallery of Canada, Ottowa, Ontario, Canada. Unpublished report. Th e Canadian Conservation Institute.

SOURCES OF MATERIALS “Scenery” 100% Cotton Muslin Dharma Trading Co. 1604 -4th St. San Rafael, CA 94901 (800) 542-5227 www.dharmatrading.com

100% Polyester fabric F&S Fabrics 10629 W. Pico Blvd. Los Angeles, CA 90064 (310) 475-1632 www.fandsfabrics.com

Mettler no. 60 100% Cotton Th read Sewing Arts Center 3330 Pico Blvd. Santa Monica, CA 90405 (310) 450-4300 sewingartscenter.com

Textile Specialty Group Postprints Volume 23, 2013 106 YADIN LAROCHETTE

Gütermann 100% Polyester Th read F&S Fabrics 10629 W. Pico Blvd. Los Angeles, CA 90064 (310) 475-1632 www.fandsfabrics.com

Varathane (made by Rustoleum) Home Depot 12975 W. Jeff erson Blvd. Los Angeles, CA 90066 (310) 822-3330 www.homedepot.com

AUTHOR BIOGRAPHY

YADIN LAROCHETTE received a B.A. in Art History from the University of California at Berkeley with honors in 1994 and an M.S. degree in Art Conservation from the Winterthur/University of Delaware Program in 2004, majoring in textiles with a concentration in preventive conservation. She is a fi ft h generation French tapestry weaver, and wove professionally prior to entering the conservation fi eld. She founded Larochette Textile Conservation LLC, a private practice, in the Los Angeles area in 2005 where she worked on projects involving various private textile collections and numerous institutions. She closed her practice in 2015 upon accepting a position with Tru Vue Inc., as their Museum and Conservation Liaison. Email: [email protected].

Textile Specialty Group Postprints Volume 23, 2013 107

NEW AND CURRENT MATERIALS AND APPROACHES FOR LOCALIZED CLEANING IN TEXTILE CONSERVATION

ELIZABETH SHAEFFER AND JOY GARDINER

ABSTRACT—For this paper the localized cleaning of textiles with poultices and gels was reviewed in the existing textile conservation literature as well as in related literature in the fi elds of paper, objects, and painting conservation. Th ese current and developing approaches to aqueous cleaning enable stains and residues to be addressed individually through the precise application of specialized cleaning solutions. Cellulose pulp, methyl cellulose, xanthan gum, Laponite® RD, agarose, and gellan gum were investigated. Th e chemical and physical properties of each material are discussed, and their compatibilities with additives including enzymes, chelating agents, solvents, and bleaches are compared. Situations in which the use of one material might be preferable to another are suggested. Recipes for an agarose gel testing kit are provided and opportunities for future research are proposed.

MATERIALES Y MÉTODOS NUEVOS Y ACTUALES PARA LA LIMPIEZA LOCALIZADA EN LA CONSERVACIÓN TEXTIL—En este artículo, se analizó el tema de la limpieza localizada de textiles con emplastos y geles abordado en la literatura existente de conservación textil y en literatura relacionada en las áreas de conservación del papel, objetos y pinturas. Estos métodos de limpieza acuosa actuales y en desarrollo permiten el tratamiento individual de manchas y residuos por la aplicación precisa de soluciones de limpieza especializadas. Se investigó sobre la pulpa de celulosa, la celulosa metílica, la goma xantana, Laponite® RD, la agarosa y la goma gellan. Se discute sobre las propiedades químicas y físicas de cada material y se comparan sus compatibilidades con aditivos, incluyendo las enzimas, los agentes quelantes, los solventes y los blanqueadores. Se sugieren situaciones en las que el uso de un material podría ser preferible a otro. Se publican recetas para armar un kit de prueba del gel de agarosa y se proponen oportunidades para futuras investigaciones.

1. INTRODUCTION

Textiles may require local cleaning for a variety of reasons: the presence of water-sensitive dyes or fi nishes, the need to reduce the quantity of expensive additives such as enzymes, or the overall fragility of an object. To address these issues, textile conservators have traditionally relied on suction or absorbent materials such as blotter paper in order to control the spread of moisture. Drawing on cleaning techniques employed in related conservation specialties including paper, objects, and paintings, textile conservators have begun to experiment with a variety of poulticing and gelling materials to introduce aqueous solutions or solvents to a textile. In addition to permitting the localization of treatment to a small area by controlling the spread of a solution into the substrate, gels and poultices can increase the contact time between the cleaning solution and the object. Th ey can also slow the evaporation rate of solvents incorporated into the solution, minimize mechanical action and abrasion, retain and distribute heat for enzymatic preparations, act as soil suspension aids, and reduce the volume of solution required for treatment (Lemiski 1998; Wolbers and Stavroudis 2012). Designing a poultice system requires the identifi cation of a solution appropriate for removing a particular stain and the selection of a compatible poulticing agent. Th e concentration of the system must be carefully controlled to prevent the formation of tidelines, while still delivering enough solution to permit the soil to be NEW AND CURRENT MATERIALS AND APPROACHES FOR LOCALIZED CLEANING IN TEXTILE CONSERVATION solubilized. Th e same poultice may have widely varying results on diff erent fabrics, depending on the weave density, fi ber type, and sizing. Poultices can generally be divided into three types: cellulose-based poultices, viscous gels, and rigid gels. While the materials within each class show certain similarities in the ways in which they are prepared and applied, their working properties diff er. Th is is particularly evident in terms of their compatibility with additives including solvents, chelators, enzymes, and bleaches. Depending on the requirements of a particular treatment, one poultice may be more suitable than another. Cellulosic poultices such as blotter paper, cotton sheeting, and purifi ed cellulose pulp draw solubilized soils out of a textile by capillary action as the poultice dries and its pore size shrinks. While they are simple to prepare and leave no residues, they are generally better suited to absorbing stains than to delivering specialized cleaning solutions. Viscous gels such as methyl cellulose, xanthan gum, and Laponite® RD (Southern Clay Products, Inc.) thicken solutions, enabling them to be applied to discrete locations. Th eir thixotropic or sheer-thinning nature allows them to suspend soils with a small degree of agitation. However since they are amorphous, these gels can become trapped in the interstices of a fabric making them diffi cult to remove. Rigid gels such as agarose and gellan gum, on the other hand, form interlinked polymer networks with pores. Th e cleaning solution fl ows through the pores. As the gel dries, the pore size shrinks, extracting solubilized soils from the substrate by capillary action. Although the rigidity of these gels makes them easy to apply and remove without leaving residues, controlling the lateral spread of the solution presents a challenge. As with any localized treatment, the introduction of moisture to an isolated area of a textile via poultice can result in the concentration of acids at the wet-dry interface and the formation of tidelines. Although these may not be visible at fi rst, it is possible that they will become apparent with age, and therefore the risks of the treatment must be considered carefully before proceeding (Dupont 1996).

2. CELLULOSIC POULTICES

Absorbent cellulosic material, such as blotter paper or cotton fabric, is frequently used in textile conservation to reduce stains. For instance, a drying cloth may be placed over a textile following wet-cleaning in order to draw solubilized soils to the point of evaporation in the drying cloth, preventing their re-deposition in the textile and the formation of tidelines, or even controlling dye bleed (Francis 1992). Contact cleaning follows a similar logic: clean cloth or blotter is placed against a damp and dirty textile to draw out soils. Th e cloth may be allowed to come to complete dryness (Johnson-Dibb 1995), or can be stroked across the surface of the textile (Bede 2002). Blotter paper has also been used to deliver enzymatic solutions, which were kept warm with a heating pad (Segal 1994). While these techniques can certainly be eff ective, another poulticing material, purifi ed cellulose pulp, is able to achieve closer surface contact due to its short fi ber length and malleable nature, oft en resulting in superior extraction of stains. At Winterthur Museum, the procedure for using this material on textiles begins by moistening a local area by brush and poulticing with blotter until no more discoloration is drawn up. Th e process is then repeated with cellulose pulp, removing signifi cant amounts of additional soiling. Another successful technique utilizes damp Arbocel® BC 1000 cellulose powder (J. Rettenmaier & Söhne GmbH ϩ Co KG) mixed 1 g per 2 mL water. It is applied to a stain, surrounded by a ring of dry powder to prevent the formation of tidelines and allowed to dry. An additional layer of dry powder can be applied on top to improve wicking capacity (Lemiski 1998).

Textile Specialty Group Postprints Volume 23, 2013 110 ELIZABETH SHAEFFER AND JOY GARDINER

Cellulose pulp has also been used for cleaning ceramics (Pouliot et al. 2013), wall paintings (Dei et al. 1998), and stone (Grissom 1988; Vergès-Belmin et al. 2011).

3. VISCOUS GELS

3.1 METHYL CELLULOSE In the conservation fi eld, methyl cellulose and other cellulose ethers have been in use since the 1940s, and have been applied as soil suspension aids, consolidants, adhesives, paper sizes, and as poultices on varied substrates including textiles, paper, paint (Eisenberg and Irgang 2000), stone (Lauff enburger 1992; Garland and Rogers 1995), encaustic (Rothe 1982), glass (Mijangos et al. 1995), and metals (Lupu and Balta 1998). Methyl cellulose is produced by treating alkaline cellulose with methyl chloride. In the chemical reaction that follows, hydroxyl groups (-OH) on the cellulose are substituted with methoxides (-OCH3), altering the solubility of the polymer (fi g. 1). At low concentrations around 0.5% w/v, it is used as a paper size (Baker 1982; Areal Guerra et al. 1995), with lower molecular weight formulations off ering superior penetration due to the ability of the short polymer chains to slip between the paper fi bers (Banik 1990). Solutions of high molecular weight methyl cellulose such as Dow Methocel™ A4M (4000 cps) at 2-3% w/v have been suggested for use as poultices for paper and textiles due to their high viscosity, which inhibits lateral moisture spread (Baker 1982; Smith 1987; Larochette 2003). Some of the benefi ts of methyl cellulose are its neutral pH, its compatibility with cellulosic textiles, and its ability to deliver chelating (Bedynski 1995; Lupu and Balta 1998; Wolbers and Stavroudis 2012; Sahmel et al. 2012) and enzymatic solutions (Hatton 1977; Smith 1987; Segal 1994; Mijangos et al. 1995; Blüher and Banik 1996; Stockman 2007). It has also been used successfully in combination with the bleaching agents sodium borohydride (Larochette 2003) and hydrogen peroxide (McFarland 1997). Its primary drawbacks are that it precipitates out of solution in the presence of large amounts of salts (Feller and Wilt 1990) and that as noted above, it can be diffi cult to fully remove from a textile, leaving shiny residues that stiff en the fabric. Th e issue of rinsibility has been a problem with methyl cellulose since its introduction as a poulticing material, and it has been addressed in a variety of ways. Past publications have advocated using lower concentrations (2% vs. 4% w/v) with short application times (5 minutes vs. 10 minutes) to reduce the deposition of residues

Figure 1. Methyl cellulose monomer unit.

Textile Specialty Group Postprints Volume 23, 2013 111 NEW AND CURRENT MATERIALS AND APPROACHES FOR LOCALIZED CLEANING IN TEXTILE CONSERVATION

(Smith 1987), or adding sodium chloride to the rinse solution to disrupt hydrogen bonding between the methyl cellulose and the textile (Larochette 2003). Another method that has been tested in recent years uses very high concentrations of methyl cellulose, around 50% w/v, to produce a malleable putty that can be shaped before being applied to a textile (Naunton 2010, Sahmel et al. 2012). While the severely restricted volume of water in the poultice prevents the formation of visible tidelines, it also greatly increases treatment time and may still leave residues (Sahmel et al. 2012), making it impractical for many situations. Th e problem of clearing residues has also led to interest in using a barrier layer between the methyl cellulose and the object. On paper, gampi tissue barriers have been used to eliminate the need for mechanical reduction of residues, however they have also been found to reduce the effi cacy of the cleaning solution (Stockman 2007). In experiments undertaken at Winterthur Museum, 2.5% w/v Dow Methocel™ A4M was applied to linen and cotton substrates over a gampi barrier. Th e gampi not only prevented the transfer of residues, but also the transfer of the solution, moistening the fabrics only slightly even aft er an hour of contact. Th is may suggest that this technique is ineff ective for use with textiles, however additional testing with other types of fabric and concentrations of methyl cellulose is required.

3.1.1 Albertina Kompresse To specifi cally address the issue of releasing backings adhered with non-swellable starch pastes, a pre-packaged system of methyl cellulose with the enzyme α-amylase has been developed. Th e Albertina Kompresse is a poultice consisting of three layers: an interleaving sheet, a non-woven polyester fabric impregnated with α-amylase, and moistening material. To use the poultice, each layer is wetted by brush or spray, stacked on the backing, and then covered with Mylar and weight (fi g. 2). Th e starch paste may be released in as little as 20 minutes (Schwarz 2000). Two studies by Haldane and McClean (2003) and by Whapp (2007) have independently reported that the Albertina Kompresse allowed backings adhered to textiles with starch paste to be successfully removed, however both also cite the need for additional rinsing to reduce the remaining adhesive residues. While this poulticing system is not diffi cult or time consuming to prepare, delivering the appropriate amount of wetness requires a certain amount of practice and experience. A signifi cant drawback to this product is its expense. A single 8 ϫ 12” pack of the three-layered system costs $67 in 2013 (Talas). 3.2 XANTHAN GUM Xanthan gum, a polysaccharide derived from the Xanthomonas campestris bacterium, is commonly used for thickening foods and stabilizing emulsions in cosmetic products. Th e backbone of the xanthan gum

Figure 2. Albertina Kompresse application. 1) Weight, 2) Polyester fi lm, 3) Moistening material, 4) Amylase poultice, 5) Paper interleaf, 6) Textiles or paper adhered together with starch paste (adapted from Talas product literature 2013).

Textile Specialty Group Postprints Volume 23, 2013 112 ELIZABETH SHAEFFER AND JOY GARDINER

Figure 3. Xanthan gum monomer unit, with negatively charged side chains.

polymer is identical to that of cellulose, with the unique character of the gum being derived from the tri-saccharide side chains on alternate sugar units (fi g. 3). Th e anionic nature of these chains aff ects the conformation of the molecule when in solution. At alkaline pH or low salt concentrations, the side chains are negatively charged and repel each other (fi g. 4). In this conformation, gels are formed when the polymers become physically entangled. In solutions at lower pH or with added electrolytes, the charge on the side chains is neutralized, reducing their electrostatic repulsion and allowing them to wrap around and hydrogen bond to the backbone. Hydrogen bonding between these helical rods and physical entanglement of their uncoiled tails form a gel network interspersed with microscopic “pockets”. When force is applied to the gel, the network exhibits shear thinning, disaggre- gating into individual polymers and reducing viscosity. When the force is removed the network reforms and viscosity increases. (Vanderbilt Minerals, LLC 2013). While the conservation literature on this material is limited (Wolbers and Stavroudis 2012), it is of particular interest due to its compatibility with chelating agents and high concentrations of acids and salts. Unique among the materials considered in this article, xanthan gum can also be used to stabilize oil-in-water emulsions due to the hydrophobicity of its “pockets”, allowing non-water miscible solvents such as benzyl alcohol or toluene to be incorporated into a gel up to approximately 20% v/v. It may be less compatible with enzymes, however, since the gel’s anionic charge could interact with positively charged amino acids, inhibiting their functionality (Wolbers 2013). As with other viscous gels, xanthan gum must be rinsed with large volumes of water. Pharmaceutical grade xanthan gums that form clear gels such as Vanzan® NF-C (Vanderbilt Minerals, LLC) are preferred for conservation treatments. It is also important to use the gel within a few days aft er preparing it or to add a wide-spectrum biocide such as Germaben II (International Specialty Products, Inc.) to prevent mold growth.

Textile Specialty Group Postprints Volume 23, 2013 113 NEW AND CURRENT MATERIALS AND APPROACHES FOR LOCALIZED CLEANING IN TEXTILE CONSERVATION

Figure 4. Xanthan gum in solution a) at alkaline pH or without electrolytes, b) at lower pH or with electrolytes, and c) forming a gel network (adapted from Vanderbilt Minerals, LLC 2013).

3.3 LAPONITE® RD Laponite® RD is a synthetic silicate used commercially to alter the rheology of paints, personal care products, and cleaning sprays. For conservation applications, it has been used as a poulticing material on textiles (Potter 2003), paper (Stockman 2007; Warda et al. 2007), and ceramics (Pouliot et al. 2013). In its dry form, Laponite has a layered structure of disc-shaped silicate crystals (fi g. 5). When water is added, “electrical double layers” form, in which sodium on the surface of the crystals is hydrated and then held electrostatically between them. With the addition of polar compounds such as salts or soluble impurities, the laponite crystals may re-arrange into a “house of cards” structure stabilized by electrostatic forces. Like xanthan gum, it exhibits shear-thinning properties. It is most stable in the range pH 6 to pH 13 (Southern Clay Products, Inc. 2013).

Figure 5. Laponite RD structure: a) dry, b) with the introduction of water, forming “electrical double layers”, and c) in a “house of cards” arrangement with additional salts (adapted from Southern Clay Products, Inc. 2013).

Textile Specialty Group Postprints Volume 23, 2013 114 ELIZABETH SHAEFFER AND JOY GARDINER

Figure 6. Agarose monomer unit.

Because it is inorganic, Laponite is more compatible with bleaches than the other gelling materials reviewed here. While it can be combined with enzymes, its viscosity may be aff ected by chelating agents since it is stabilized by inorganic salts (Wolbers 2013). It also leaves residues that may be abrasive. With artifi cial ageing, Laponite residues have been found to contribute to the discoloration of a paper substrate (Warda et al. 2007). In the same study, gampi tissue barriers were shown to prevent the deposition of residues. Resi- dues have also been reduced by preparing the Laponite at 12.5% w/v, which forms a very solid gel that can be gently lift ed from a textile surface with a microspatula (Potter 2003).

4. RIGID GELS

4.1 AGAROSE Agarose is a polysaccharide that forms rigid gels at low concentrations. It is separated from agar, a red algae exudate, the other component of which is the non-gelling polysaccharide agaropectin. Agarose is a disaccha- ride of 1,3-linked β-D-galactopyranose and 1,4-linked 3,6-anhydro-α-L-galactopyranose (fi g. 6) (Araki 1956). In a conservation context, agarose and agar have been used on painted surfaces (Campani et al. 2007), textiles (Ellis 2009; Sahmel et al. 2012), stained glass (Mijangos et al. 2005), polychrome wood (Guerra and Abeni 2008), ceramics (Pouliot et al. 2013), and paper (Van Dyke 2004; Stockman 2007; Warda et al. 2007). Agarose is insoluble at room temperature, but when heated to over 85ЊC in water, it dissolves into a ran- dom coil conformation. As the solution cools, the agarose twines together into double helices. Th e uncoiled ends of each helix tangle together, producing a rigid three-dimensional polymer network interspersed with pores (fi g. 7) (Arnott et al. 1974). Th e size of these pores is dependent on the concentration of agarose. Lower concentrations of gel have larger pores and release large volumes of solution whereas higher concentrations yield gels that release very little moisture. Th e useful concentration range for textile conservation is between 1 and 4% w/v, and should be tailored for each treatment. Agarose has many advantages as a gelling material. Not only can it be combined with chelating agents (Pouliot et al. 2013; Sahmel et al. 2012) and enzymes (Mijangos et al. 2005; Ellis 2009), it can also be made into a gel for delivering water-miscible solvents such as ethanol and acetone by soaking it in solvent and allowing the solvent to replace the water by diff usion (Pouliot et al. 2013). It has also been suggested that it can be paired with bleaches (Burgess 1988; Pouliot et al. 2013), however with prolonged exposure these may degrade the polymer. Its most obvious advantages are that it is easy to manipulate and position on a textile, that no mechanical action is required during treatment, and that the gel itself leaves no residues (Warda et al. 2007). Treated areas that cannot be fl ushed with water to remove cleaning solutions, either due to the textile’s fragility or the presence of sensitive media, could potentially be rinsed by repeatedly applying fresh agarose gels prepared with only distilled water in order to dilute the solution. Although it is more expensive than

Textile Specialty Group Postprints Volume 23, 2013 115 NEW AND CURRENT MATERIALS AND APPROACHES FOR LOCALIZED CLEANING IN TEXTILE CONSERVATION

Figure 7. Agarose gel formation (adapted from Arnott et al. 1974; Låås 1975). other gelling agents, agarose could potentially be made more economical by rinsing gels in distilled water and reusing them. It is also thermo-reversible, so scraps can be reheated and cast again to form continuous sheets. Th e greatest challenge when working with agarose is to determine what concentration will deliver enough solution to achieve the desired eff ects without delivering so much that it spreads laterally through the textile, causing tidelines or interacting with sensitive media. Cyclododecane has been used previously in agarose gel treatments as a hydrophobic barrier to prevent the spread of a cleaning solution and to protect fugitive dyes (Sahmel et al. 2012).

4.1.1 Agarose Test Kit In the Winterthur/University of Delaware Program in Art Conservation, an agarose-based test kit developed by Richard Wolbers is employed to measure the pH and conductivity of textiles, as well as to test how eff ectively diff erent chelating agents remove stains. Th e information gathered from this test kit can be used to formulate gels for localized cleaning as well as to decide on appropriate parameters for wet-cleaning baths. Th e gels are prepared at 5% w/v to limit the formation of tidelines, and can be dried further by dabbing them with lens tissue. Germaben II is added to inhibit biological growth. Gels are cast to a depth of approximately 0.25 cm in small molds and are kept sealed in plastic bags to prevent moisture loss (fi gs. 8 and 9). Aft er many months at room temperature they maintain their functionality. When performing tests, small pieces of gel are cut with a 4 mm dermal punch for convenience, but can also be cut with a scalpel. When testing conductivity and pH, the pieces of gel remain in contact with the object for 20 minutes. Measurements are made with a conductivity meter (e.g. Horiba Twin Cond B-173) and a pH meter (e.g. Horiba Twin pH B-212) with reservoirs. Th e gels are positioned in the reservoirs between the electrodes and distilled water is added to connect the electrodes and complete the circuit. Although these measurements do not give “absolute” values for pH and conductivity, this information is particularly informative when measured and compared before and aft er cleaning. Th e chelating kit consists of gels with: a weak chelator, citric acid; a strong chelator, ethylenediaminetet- raacetic acid (EDTA); and an iron chelator, N,N-bis(2-hydroxybenzyl)ethylenediamine-N,N-diacetic acid (HBED). Each chelating agent is prepared at both pH 6.5 and pH 8, since the effi cacy of the solution may be pH dependent. Small pieces of gel are allowed to dry fully on the textile in order to maximize contact time

Textile Specialty Group Postprints Volume 23, 2013 116 ELIZABETH SHAEFFER AND JOY GARDINER

Figure 8. Agarose test kit composed of a dermal punch and gels sealed in plastic bags.

pH/CONDUCTIVITY GELS • 100 mL distilled water • 2 drops Germaben II

CHELATING GELS pH 6.5 pH 8 • 100 mL distilled water • 100 mL distilled water • 0.5 g chelator (EDTA, HBED, • 0.5 g chelator (EDTA, HBED, citric acid) citric acid) • 0.5 g citric acid (buff er, added to • 0.5 g tetrasodium borate deca- EDTA and HBED only) hydrate (buff er) • NaOH/HCl (to adjust pH) • NaOH/HCl (to adjust pH) • 2 drops Germaben II • 2 drops Germaben II

Figure 9. Recipes for an agarose gel test kit. Th e solutions are prepared to the appropriate pH on a stirring plate before adding 5 g agarose (5% w/v) and heating the solution in short pulses in the microwave or in a bain marie until the agarose is dissolv ed. Note that chelators can also be added as salts, rather than as free acids as described here. Th e amount of chelator required does not change. Sodium or potassium cations are preferred to ammonium, which over time converts to ammonia gas.

Textile Specialty Group Postprints Volume 23, 2013 117 NEW AND CURRENT MATERIALS AND APPROACHES FOR LOCALIZED CLEANING IN TEXTILE CONSERVATION

Figure 10. High-acyl gellan gum monomer unit. and capillary action. If the dried gel sticks to the textile, it can be loosened with a moistened brush. Th e color of the gels and the areas of the textile beneath them are compared qualitatively to determine which combination of chelator and pH is most eff ective.

5. GELLAN GUM

Gellan gum is a rigid gel that has recently been introduced to the conservation community as an alternative to agarose, one of its primary appeals being that it is less expensive. While de-acetylated gellan gum has been found useful for paper objects (Iannuccelli and Sotgiu 2010), the high-acyl form (fi g. 10) may be preferred for use with textiles due to its greater elasticity and ability to conform to weave interstices, resulting in better surface contact (Peranteau 2013). Gellan gum diff ers from agarose in that its polymer network is anionic rather than neutral, and can be stabilized by divalent metal ions such as calcium in addition to physical intertwining and hydrogen bonding (fi g. 11). Th e high-acyl form relies less on cation stabilization than the low-acyl form

Figure 11. Gellan gum gel formation, b) in the presence of cations, represented as black dots, and c) without cations (adapted from Morris et al. 2012).

Textile Specialty Group Postprints Volume 23, 2013 118 ELIZABETH SHAEFFER AND JOY GARDINER

Poultice Chelators Enzymes Bleach Solvents Leaves Residues RIGID Agarose yes yes limited water-miscible no GELS Gellan Gum limited limited limited water-miscible no Methyl Cellulose yes yes limited water-miscible yes VISCOUS Xanthan Gum yes limited limited water-miscible yes GELS and hydrophobic Laponite RD no no yes water-miscible yes CELLULOSE Cellulose Pulp yes yes limited water-miscible no POULTICE

Figure 12. Performance comparison of gelling and poulticing materials. due to its higher degree of polymerization. It forms gels that are easy to work with at a concentration of 1% w/v, without the addition of other salts (Peranteau 2013). While the low price of gellan gum makes it appropriate for large-scale treatments, due to its anionic character it may be less compatible with additives such as chelators and enzymes. Rather than incorporating these materials into the gel solution before it is cast, they can be applied to the gel’s surface once it has set (Iannucelli and Sotgiu 2010). Further testing is required to determine how these additives aff ect the gel structure.

6. CONCLUSIONS

Th is literature review provides an assessment of poulticing materials employed across conservation sub-specialties. Th e appropriateness of each material with textiles was considered, in terms of their physical properties as well as their compatibility with diff erent additives (fi g. 12). For localized stain reduction, cellulose pulp demonstrates superior results over sheets of paper or cloth due to its ability to conform more closely to the surface of a textile. Viscous gels are diffi cult to remove fully and may change the hand and sheen of a fabric, while rigid gels are easily lift ed away without leaving residues. Although it is easier to control the lateral spread of water through a textile with viscous gels than rigid gels, the volume of water that is required for rinsing them away may negate some of the benefi ts of this control. While all of the materials explored can be combined with water-miscible solvents, only xanthan gum forms emulsions with hydrophobic solvents. Laponite® RD is the most compatible with bleaches, and agarose can be combined with the widest range of cleaning solutions without leaving residues. Th ere remains considerable opportunity for future research on poulticing materials, including application methods, ageing characteristics, and the chemical eff ects of diff erent additives on the gels. Other research questions might compare the residues left by methyl cellulose and xanthan gum, both quantita- tively and in terms of their ageing properties. Additional investigation is required to confi rm speculations about potential applications of agarose: its reusability, eff ectiveness as a tool for the controlled dilution of other cleaning solutions, and a side-by-side comparison with gellan gum in combination with diff erent additives, to name a few.

Textile Specialty Group Postprints Volume 23, 2013 119 NEW AND CURRENT MATERIALS AND APPROACHES FOR LOCALIZED CLEANING IN TEXTILE CONSERVATION

ACKNOWLEDGMENTS

Th e authors wish to thank Richard Wolbers, associate professor, Art Conservation Department, University of Delaware, for generously sharing his expertise and advice, without whom this study would not have been possible. Th anks to Laura Mina and Katherine Sahmel, Mellon Fellows present and former at the Philadelphia Museum of Art for allowing the use of their research, which was not yet published at the time that this study was conducted. Additional thanks to Bruno Pouliot, Lauren Fair, Chela Metzger, and Joan Irving, adjunct assistant professors in the Winterthur/University of Delaware Program in Art Conservation in objects, objects, library, and paper conservation respectively, for their insight into how these materials are used in their fi elds. Financial support to Elizabeth Shaeff er while at the University of Delaware was provided by the National Endowment for the Humanities, the Andrew W. Mellon Foundation, and the Society of Winterthur Fellows.

REFERENCES

Araki, C. 1956. Structure of the agarose constituent of agar-agar. Bulletin of the Chemical Society of Japan 29(4): 543–544.

Areal Guerra, R., J. M. Gibert Vives, J. M. Dagà Monmany, and J. Fernández Garrido. 1995. Procedure for simultaneous deacidifi cation and sizing of paper. Restaurator 16: 175–193.

Arnott, S., A. Fulmer, W. E. Scott, I. C. Dea, R. Moorhouse, and D. A. Rees. 1974. Th e agarose double helix and its function in agarose gel structure. Journal of Molecular Biology 90: 269–284.

Baker, C. 1982. Methylcellulose and sodium carboxymethylcellulose: uses in paper conservation. AIC Book and Paper Group postprints 1. Washington DC: AIC-BPG. 16-19.

Banik, G. and W. K. Sobotka. 1990. Deacidifi cation and strengthening of bound newspapers through aqueous immersion. In Paper preservation: current issues and recent developments, ed. P. Luner. Atlanta: TAPPI. 137–143.

Bede, D. 2002. Conservation of the Wright Flyer III: serendipity and substantiation. Strengthening the bond: science and textiles. North American Textile Conservation Conference 2002 preprints. Philadelphia, PA. 15–24.

Bedynski, M. 1995. Th e use of enzymes in the conservation treatment of a seventeenth-century wall map of America. Journal of the IIC-Canadian Group 20: 11–18.

Blüher, A. and G. Banik. 1996. Th e application of enzyme-containing methylcellulose gels for the removal of starch-based adhesives in albums. ICOM Committee for Conservation preprints. 11th triennial meeting, Edinburgh. London: James & James. 2: 494–499.

Burgess, H. D. 1988. Practical considerations for conservation bleaching. Journal of the IIC-Canadian Group 11: 11–26.

Campani, E., A. Casoli, P. Cremonesi, I. Saccani, and E. Signorini. 2007. L’uso di agarosio e agar per la prepara- zione di “gel rigidi”, trans. D. Kunzelman. Saonara: Il Prato.

Dei, L., A. Ahle, P. Baglioni, D. Dini, and E. Ferroni. 1998. Green degradation products of azurite in wall paintings: identifi cation and conservation treatment. Studies in Conservation 42(2): 80–88.

Textile Specialty Group Postprints Volume 23, 2013 120 ELIZABETH SHAEFFER AND JOY GARDINER

Dupont, A. L. 1996. Degradation of cellulose at the wet/dry interface I: the eff ect of some conservation treatments on brown lines. Restaurator 17: 1–21.

Eisenberg, J. and H. Irgang. 2000. Shade of faith: the conservation of a painted sukkah interior. Conservation combinations. North American Textile Conservation Conference 2000 preprints. Ashville, NC. 69–78.

Ellis, S. 2009. A passage in the life of a palampore: conservation. Journal of the Canadian Association for Conservation 34: 21–28.

Feller, R. L. and M. Wilt. 1990. Evaluation of cellulose ethers for conservation. Vol. 3. Research in conservation. Los Angeles: Getty Conservation Institute.

Francis, K. 1992. Observations on the drying behavior of textiles. Textile Specialty Group postprints 2. AIC 20th Annual Meeting. Buff alo, NY. 1–6.

Garland, K. M. and J. C. Rogers. 1995. Th e disassembly and reassembly of an Egyptian limestone sculpture. Studies in conservation 40(1): 1–9.

Grissom, C. A. 1988. Methyl cellulose poultice cleaning of a large marble sculpture. Proceedings of the VIth International Congress on deterioration and conservation of stone. Torún, Poland. 551–562.

Guerra, P. and M. Abeni. 2008. Un esempio di applicazione di gel rigidi di agar. Progetto restauro 47: 15–19.

Haldane, E. and L. McClean. 2003. Adapting techniques: cleaning a covenanting banner. In Dust, sweat and tears: recent advances in cleaning techniques. London: United Kingdom Institute for Conservation of Historic and Artistic Works. 15–27.

Hatton, M. 1977. Enzymes in a viscous medium. Th e Paper Conservator 2: 9.

Iannuccelli, S. and S. Sotgiu. 2010. Wet treatments of works of art on paper with rigid gellan gels. Th e Book and Paper Group annual 29. Washington DC: AIC-BPG. 25–39.

Johnson-Dibb, R. 1995. Contact cleaning of textiles. Th e Textile Specialty Group postprints 5. AIC 23rd Annual Meeting. St. Paul, MN. 35–37.

Låås, T. 1975. Agar derivatives for chromatography, electrophoresis and gel-bound enzymes. Ph.D. diss., Uppsala Universitet.

Larochette, Y. 2003. Determining the effi cacy of cyclododecane as a barrier for a reduction bleaching treatment of a silk embroidered linen napkin. 29th annual ANAGPIC student conference papers. New York, NY. 52–61.

Lauff enburger, J. A., C. A. Grisson and A. E. Charola. 1992. Changes in gloss of marble surfaces as a result of methylcellulose poulticing. Studies in conservation 37(3): 155–164.

Lemiski, S. 1998. An investigation of poultice materials for textile conservation. Textile conservation newsletter. 1–15.

Lupu, M. I. A. and Z. I. Balta. 1998. A possible method based on CMC for cleaning metalwork decorative art objects. Metal 98: proceedings of the international conference on metals conservation. Draguignan-Figanières, France. 173–176.

Textile Specialty Group Postprints Volume 23, 2013 121 NEW AND CURRENT MATERIALS AND APPROACHES FOR LOCALIZED CLEANING IN TEXTILE CONSERVATION

McFarland, M. R. 1997. Th e whitening eff ects of peroxide gels on darkened lead white paint. Th e Book and Paper Group annual 16. Washington DC: AIC-BPG. 55–65.

Mijangos, C. N. Valentín, and A. Sánchez. 2005. Geles poliméricos en la conservación y limpieza del patrimonio histórico. Revista de plásticos modernos 89(583): 67–72.

Morris, E. R., K. Nishinari, and M. Rinaudo. 2012. Gelation of gellan - a review. Food hydrocolloids 28: 373–411.

Naunton, M. 2010. Use of gel poultices for adhesive removal. Textile Specialty Group Postprints 19. AIC 37th Annual Meeting. Los Angeles, CA. 184–186.

Peranteau, A. 2013. Gellan gum as a material for local stain reduction. Conserving modernity: the articulation of innovation. North American Textile Conservation Conference 2013 preprints. San Francisco, CA. 72–85.

Potter, J. 2003. Experiments in rust removal on a painted Indian chintz. In Dust, sweat and tears: recent advances in cleaning techniques. London: United Kingdom Institute for Conservation of Historic and Artistic Works. 39–57.

Pouliot, B., L. Fair, and R. Wolbers. 2013. Rethinking the approach: techniques explored at Winterthur for the stain reduction of ceramics. Recent advances in glass, stained-glass, and ceramics conservation 2013. ICOM Committee for Conservation Glass and Ceramics Working Group interim meeting and forum of the International Scientifi c Committee for the Conservation of Stained Glass (Corpus Vitrearum-ICOMOS) preprints. Amsterdam, the Netherlands. 211–223.

Rothe, A. 1982. Conservation report on the mummy portrait of Isidora. Th e J. Paul Getty Museum journal 10: 195–196.

Sahmel, K., L. Mina, K. Sutherland, and N. Shibayama. 2012. Removing dye bleed from a sampler: new methods for an old problem. Textile Specialty Group Postprints 22. AIC 40th annual meeting. Albuquerque, NM. 78–90.

Schwarz, I. 2000. A pre-packaged a-amylase poulticing system: Albertina-Kompresse. Th e Book and Paper Group annual 15. Washington DC: AIC-BPG. 97–104.

Segal, J. 1994. New techniques for the application of enzymes. Conservation of historic and artistic works on paper: proceedings of a conference. Ottawa, Canada. 205–208.

Smith, L. M. 1987. Use of enzymes in textile cleaning. AIC preprints. American Institute for Conservation 15th Annual Meeting, Vancouver. Washington, DC: AIC. 124–131.

Southern Clay Products, Inc. 2013. Laponite product information. Gonzales, TX.

Stockman, D. 2007. Treatment options for oil stains on paper. Th e Book and Paper Group annual 26. Washington DC: AIC. 115–126.

Vanderbilt Minerals, LLC. 2013. Vanzan® xanthan gum product information. Norwalk, CT.

Van Dyke, Y. 2004. Practical applications of protease enzymes in paper conservation. Th e Book and Paper Group annual 23. Washington, DC: AIC-BPG. 93–107.

Textile Specialty Group Postprints Volume 23, 2013 122 ELIZABETH SHAEFFER AND JOY GARDINER

Vergès-Belmin, V., A. Heritage, and A. Bourgès. 2011. Powdered cellulose poultices in stone and wall painting conservation: myths and realities. Studies in conservation 56(4): 281–297. Warda, J., I. Brückle, A. Bezúr, and D. Kushel. 2007. Analysis of agarose, carbopol, and laponite gel poultices in paper conservation. Journal of the American Institute for Conservation 46(3): 263–279. Whapp, F. 2007. Th e treatment of two Coptic tapestry fragments. Victoria and Albert conservation journal 55: 11–13. Wolbers, R. 2013. Personal communication. Winterthur/University of Delaware Program in Art Conservation. Wolbers, R. and C. Stavroudis. 2012. Aqueous methods for the cleaning of paintings. In Conservation of easel paintings, ed. J. Hill Stoner and R. Rushfi eld. New York: Routledge. 500–523.

SOURCES OF MATERIALS Agarose LE (Benchmark Scientifi c, Inc.) Universal Medical Inc. PO Box 467 Norwood, MA 02062 Tel: (800) 423-2767 Fax: (800) 535-6229 www.universalmedicalinc.com Arbocel BC 1000 cellulose fi ber (J. Rettenmaier & Söhne GmbH ϩ Co KG) Kremer Pigments Inc. 247 West 29th Street New York, NY 10001 Tel: (212) 219-2394 or (800) 995-5501 Fax: (212) 219-2395 www.kremerpigments.com Germaben II (Ashland, Inc.) Xanthan Gum NF-C (Vanderbilt Minerals, LLC) Th e Personal Formulator 97 South Red Willow Rd. Evanston, WY 82930 Tel: (307) 789-7288 www.personalformulator.com Laponite RD (Southern Clay Products, Inc.) Methocel A4M methyl cellulose (Dow Chemical Company) Talas 330 Morgan Ave Brooklyn, NY 11211 Tel: (212) 219-0770 Fax: (212) 219-0735 www.talasonline.com

Textile Specialty Group Postprints Volume 23, 2013 123 NEW AND CURRENT MATERIALS AND APPROACHES FOR LOCALIZED CLEANING IN TEXTILE CONSERVATION

Kelcogel CG-HA and Kelcogel CG-LA C.P. Kelco U.S., Inc. 3100 Cumberland Blvd. Suite 600 Atlanta, GA 30339 Tel: (800) 535-2687 Fax: (678) 247-2797 www.cpkelco.com

AUTHORS’ BIOGRAPHIES

ELIZABETH SHAEFFER is a Fellow in the Winterthur/University of Delaware Program in Art Conservation majoring in Textiles. She has interned at the Art Institute of Chicago, the Philadelphia Museum of Art, the Field Museum, and the Museo Histórico Nacional in Santiago, Chile. She received her BA in Anthropology from the University of Chicago. elizabeth.a.shaeff [email protected]

JOY GARDINER is a Winterthur Assistant Professor in Art Conservation for the Winterthur/University of Delaware Art Conservation Program and a Textile Conservator and the Assistant Director of Conservation at the Winterthur Museum, Garden & Library. She received her MS in conservation from the Winterthur/UD program and her BFA from Moore College of Art in Philadelphia majoring in Textile Arts. [email protected] Winterthur Museum, Garden & Library 5105 Kennett Pike Wilmington, DE 19735

Textile Specialty Group Postprints Volume 23, 2013 124 FERROUS ATTRACTIONS: THE SCIENCE BEHIND THE MAGIC

GWEN SPICER

ABSTRACT—Th e question of how to fasten or secure an artifact has long been a focus of art conservators in all specialties. We have stitched and mounted items for decades. With each method, the attempt has always been to keep the conservation treatment as reversible as possible. Th e relatively recent development of strong permanent rare earth magnets off ers the possibility of a new type of reversible fastener. Neodymium rare earth magnets are far stronger than earlier permanent magnets and have only truly entered the market since 1990. Th ey have great potential as a new tool for conservators. Before these new magnets become part of our future, a fuller understanding of how they work is needed. Specifi cally, categorization of magnetic systems will aid conservators in determining which attributes a magnet should have for each specifi c project. Th is paper describes a workshop held at the 2013 Annual Meeting of the American Institute for Conser- vation that explored magnetic systems. Participants used “jigs” with various combinations of magnets, metal components, and weights to demonstrate magnetic systems and their parts. Diff erent methods of implemen- tation and the strengths of commonly available magnets were explored. Additional topics included: what makes magnets “permanent”, when magnets were developed, and how magnets diff er from one another.

ATRACCIONES FÉRREAS, LA CIENCIA DETRÁS DE LA MAGIA—La sujeción de las piezas de arte ha sido un tema central para los conservadores de arte de todas las especialidades. Hemos cosido y montado piezas por décadas. Con cada método, siempre se ha procurado que el tratamiento de conservación sea lo más reversible posible. El desarrollo relativamente reciente de los imanes de tierras raras ofrece la posibilidad de tener un nuevo método de sujeción reversible. Los imanes de neodimio son mucho más fuertes que los imanes permanentes anteriores y recién ingresaron al mercado en 1990. Pueden llegar a ser una gran herra- mienta para los conservadores. Antes de que estos imanes formen parte de nuestro futuro, debemos saber mejor cómo funcionan. Específi camente, la categorización de los sistemas magnéticos ayudará a los conservadores a determinar qué atributos debe tener un imán para cada proyecto específi co. Este documento describe un taller de exploración de sistemas magnéticos realizado en la Asamblea Anual del Instituto Americano de Conservación 2013. Los participantes utilizaron “guías” con diferentes combina- ciones de imanes, componentes metálicos y pesas para demostrar los sistemas magnéticos y sus partes. Se exploraron diferentes métodos de implementación y las fortalezas de los imanes comunes. Otros de los temas abordados fueron: qué convierte a los imanes en imanes “permanentes”, cuándo se desarrollaron los imanes y en qué se diferencia un imán de otro.

1. INTRODUCTION

At AIC’s 2013 annual meeting a hands-on session on the use of magnets in conservation was presented. Art conservators have been using magnets for many years, but mostly in a very limited way (Dignard 1992; Spicer 2016b). Perhaps a system has not been fully developed or described in literature, it is not part of our training, or it is a practice that is too new to be embraced. Th is session’s purpose was to change that and give conservators hands-on experience with magnets. FERROUS ATTRACTIONS: THE SCIENCE BEHIND THE MAGIC

Figure 1: Participants at the hand-on session held at 41st AIC Annual meeting in Indianapolis, IN.

Th e focus of the session was for participants to learn and understand the three main parts of a magnet system: magnet, gap, and receiving material, so that they may use this knowledge in their own practice. Each part of the magnet system works in tandem to achieve the best combination for the artifact. Th ese three parts, in various combinations, were experimented with during the session (fi g. 1). Th e goal of the session was for conservators to become acquainted with the diverse variables of magnet systems. A range of Neodymium and ferrite fl exible magnets were selected as the magnets. Mylar, fabrics, and other materials were included as gap materials. Finally, the ferromagnetic receiving materials included were steel plates and preparations of iron powder in a range of several thicknesses. Th e session was divided into parts. First, the types and properties of permanent magnets were described, along with their diff erences, and the parts to any magnet system developed. Th is was followed by the hands- on activity and a discussion of observations.

2. PERMANENT MAGNETS Table 1: Types of Permanent Magnets Alnico Ferrite Samarium Neodymium

Chemical Al-Ni-Fe-Co Fe2O3 SmCo2 Nd2Fe14B structure Date 1935 1951 1965 1985 Method of Cast or sintered Bonded Sintered Sintered or Bonded, manufactured

Textile Specialty Group Postprints Volume 23, 2013 126 GWEN SPICER

Table 1: Types of Permanent Magnets—Continued Alnico Ferrite Samarium Neodymium Structure Face-centered Hexagonal crystal Multi-phase structure, structure tetragonal crystal structure Direction Isotopic and Isotopic and anisotropic anisotropic Demagnetizing Can be easily Keep away from Can be demagnetized Tough to demagnetized. When Rare earth magnets by NdFeB magnets. demagnetize. Th is also repetitively placed (Samarium and But they do not weaker means that they can north pole to north Neodymium). others. easily demagnetize pole ends together, it other classes of quickly weakens itself. magnets like SmCo or Alnico or Ferrite. Heat Tolerance Maximum working Maximum working Maximum working Maximum working temperature is 540 ЊC temperature is 300 ЊC temperature is 300 ЊC temperature is only (1004 ЊF). Th e Curie (572 ЊF). (572 ЊF). Th e Curie 150 ЊC (302 ЊF). Th e Temperature for alnico Temperature for SmCo Curie Temperature for magnets is a blistering magnets is 750 ЊC NdFeB magnets is 860 ЊC (1580 ЊF). (1382 ЊF). Very 310 ЊC (590 ЊF). respectable for a sintered magnet. Moisture/ Resistant to corrosion Resistant to corrosion Relative resistant to Corrodes easily and Oxidation corrosion. requires a coating. Mechanical Shock Very resistant Brittle and chip or Brittle and chip Brittle and chip crack easily or crack easily. Best or crack easily. Best to separate with a to separate with a cushioning material. cushioning material. Common Use First man-made Electronic inductors, Hard drives, printers Green energy, hybrid permanent magnet. transformers, and and other computer cars, wind turbines, ear —Generators electromagnets. Ferrite components phones, cell phones —Engines powders are used to coat magnetic recording taps. Trade Comments Cobalt from Zaire Cobalt Rare earths from China Br(T) 0.6–1.4 0.2–0.4 0.8–1.1 1.0–1.4 Br (gauss) 12,500 3,900 10,500 12,800 Hci 275 100–300 600–2000 750–2000 BHmax 10–88 10–40 120–200 200–440 Tc 700–860 450 720 310–400 Tmax (C) – max 540 300 300 150 temperature of use

Textile Specialty Group Postprints Volume 23, 2013 127 FERROUS ATTRACTIONS: THE SCIENCE BEHIND THE MAGIC

3. CREATING A MAGNETIC SYSTEM

When using and selecting magnets of any type, three key components are in play. 1. Th e strength of the magnet itself. Magnetic strength is measured and described in Gauss. 2. Th e receiving component. Th is is the material that is magnetized in this system. Magnetized regions of this material impact the magnet’s ability to be magnetized 3. Th e magnetic fi eld distance. Th is is the space between the magnet and the magnetized metal. It is also called “the gap”, as it is created by the layers between the magnet and the receiving ferromagnetic material. Each of these components is signifi cant in how the magnet behaves and is able to perform the task (Feymann 1964; Livingston 1996; Magnet Story 1998; Spicer 2016a). Th e balancing of these three parts determines a successful system. No one method appears to be prescribed. Instead each component is adjusted for any particular situation. Th is is further complicated by the wide variety of needs and requirements of each artifact. It is only by knowing the parts and their interactions that a system can be created for a specifi c task. Th e developed system needs to be strong enough to support the artifact, while not being so strong that it creates damage. Each variable can be slightly altered to reach the desired outcome. Each component is described below along with known alternatives.

3.1 STRENGTH OF THE MAGNET Magnets are purchased with a set polar direction. Th e most common magnet has north and south faces located on the largest surfaces of the magnet. Th ese magnets are axially oriented, so the fl at surfaces of these magnets have the strongest pull force present. Th is is because all of the magnetic fi elds are coming or going from this center spot. Th e polar direction can also be oriented side-to-side, making diaxially- oriented magnets. Th e pull force of a magnet is measured in Gauss both from its center and from its outer surface. Th is is the amount of force necessary to pull the magnet straight from the surface of a steel plate. Th e grade of a Neodymium magnet greatly alters its properties such as, strength, brittleness, Curie temperature, etc (Spicer 2014). Neodymium grades that are commonly used by conservators, are Grade N35, N42 or N52. Th e grade of a Neodymium magnet can be thought of as the properties of the magnetic material itself and how the behavior is aff ected. Th e Neodymium rare earth magnet grades are represented with both letters and numbers. A few suppliers use their own systems. Th e number represents the strength of a magnet, and generally speaking, the higher the number, the stronger the magnet. An example is N52, which compared to a N42 of the same size is about 20% stronger, and has a higher pull force of its surface fi eld. Also the higher the number, the more brittle the magnet becomes. Breakage can occur easily especially as the magnet becomes thinner. As an example, a N52 magnet that is quite thin will easily break and should be supported if frequently handled. Th e numbers used by most suppliers correspond to the Maximum Energy Product (MGOe) designation. Th erefore the N42 is 40-42MGOe and the N52 is 49.5-52MGOe. Th e letter represents both their manufacturing method, as well as their formulations. Sintered magnets are represented as N, M, H grades and bonded magnets as BDM grade. Bonded magnets should be considered if the potential of high humidity conditions exist. Additional alloys in mixture with Neodymium, like Terbium and Dysprosium, are added to maintain a magnet’s magnetic properties at higher temperatures (Brown 2004, Jones 2011) and are represented by other letters.

Textile Specialty Group Postprints Volume 23, 2013 128 GWEN SPICER

3.2 RECEIVING COMPONENT (THE MAGNETIZED MATERIAL) Th e receiving component is also an important factor in the strength of a magnet’s pull force (Spicer 2015). Metals are divided into three groups; ferromagnetics are very attractive, paramagnetics are weakly attractive, and diamagnetics are opposed to magnetic fi elds. Th e system will not function fully if the receiving component is not properly considered, as the full strength of a magnet will only be achieved with suffi cient ferromagnetic material. Ferromagnetic metals that are most attractive to magnets include nickel, cobalt, and iron. Within the structure of these materials are small regions or domains that are aligned by permanent magnets, as shown in this illustration (fi g. 2). Th e amount of alignment within the domains or saturation enables the strength of the magnet to be optimized. Th is is how the receiving substrate becomes a temporary or “soft ” magnet. For a given sized magnet, there is a corresponding thickness at which the steel is saturated. If you use a steel plate that is thicker, you should not see any real increase in the pull force. However, when you attach a magnet to thinner steel sheets you will see diminished pull strength and the magnet will behave as a lower strength magnet. Th is occurs because the ferromagnetic material will not become magnetically saturated. Th is means that the receiving material can’t hold all the magnet’s fl ux (the amount of magnetic fi eld passing through a given

Figure 2: Domain regions within a ferromagnetic material.

Textile Specialty Group Postprints Volume 23, 2013 129 FERROUS ATTRACTIONS: THE SCIENCE BEHIND THE MAGIC surface). To utilize 100% of the magnet’s pull force, you would with a thicker plate. When this is the case, some of the magnetic fi eld will extend behind the steel, because the steel isn’t thick enough to shield it all. If another ferromagnetic material is placed behind it, this too will be attracted and become a soft magnet. In this way the force fi eld can travel to several neighboring layers of ferromagnetic materials, increasing the magnetic force as needed. However, if the ferromagnetic material is thicker than the magnetic fi eld’s strength, then the reverse side of the metal shows no magnetic attraction. When using rare earth magnets the lowest and most minimum gauge steel plate to use is 24-gauge. A 22-gauge or thicker would be more optimal (note: the lower the number, the thicker the steel). Th e ferromagnetic metal is an important, but oft en over-looked component of a magnetic system. It was only a few years ago that the gauge of a steel sheet used was fi rst mentioned in conservation literature (Halbrow and Taira 2011; Hovey 2012). It is only through control of all the variables: the magnet, the ferromagnetic material, and the layers between, that a system can be reproduced and adapted to any situation.

3.3 THE MAGNETIC FIELD DISTANCE OR THE GAP Th e magnetic fi eld distance, also called the gap, is composed of the artifact along with various materials used as padding, and barriers. When the layers between the magnet and the receiving metal is widened, the magnetic force is dissipated. Th e strength of the magnetic field falls off inversely with the cube of the distance from the magnet’s center. Th is can also be calculated from the magnet’s surface area. Determining the possible gap of any particular type of magnet is based on its strength, size, and shape. Th e size and grade of the magnet contributes to its pull force as stated earlier, which is measured by surface of coverage. If the receiving side is outside the magnetic fi eld of the magnet, it results in very weak, or zero, attraction between the two surfaces. In essence, as the space between the magnet and the receiving side increases, the magnetic fi eld decreases (Feymann 1964; Livingston 1996; Spicer 2010; Spicer 2016a). Organic materials that are commonly used to soft en the hard surface of the mount and even the magnet itself, have a variety of density, loft , compactness, and friction. Th e materials used in the gap can add or subtract to the performance of the system. Not all gap materials behave the same way. Th us a system that is designed for one material might not work the same way for another.

4. THE HANDS-ON SESSION

Following the introduction to magnets and magnetic systems, participants were divided into groups. Each group received one of fi ve diff erent magnetic system variations. All groups received the same gap materials: cotton fabric, paper, Mylar, two thicknesses of polyester batting, and an ultra suede. Each of the materials was chosen to represent the various materials of an artifact, mount, or barrier layer. Th e Mylar was included because conservators frequently incorporate it as a barrier. Each group was given a “jig” made with ¾” PVC pipe for the legs and an upper horizontal aluminum “L” piece (fi g. 3). Wooden blocks with secured magnets or ferromagnetic materials were made to act in two ways: to rest on the upper edge of the aluminum bar and to independently support a small weighted bucket. In this way, participants were instructed to add weight to their given system and record the weight at which their particular confi guration of magnets, gap material, and receiver failed. Each group was given a set of pre-weighed sand bags, starting with an eighth of a pound. Th ese jigs tested the sheer strength of the magnetic systems (fi g. 4).

Textile Specialty Group Postprints Volume 23, 2013 130 GWEN SPICER

Figure 3: Th e jig is made of 3/4” PVC pipe for the legs, and an upper horizontal aluminum “L” piece. Wooden blocks with secured magnets, or ferromagnetic materials, are made to rest on the upper edge of the aluminum bar and support a weighting bucket (Woods, 2013; Spicer, 2014).

Figure 4: Wooden blocks with secured magnets or ferromagnetic materials made to rest on the upper edge of the aluminum bar and support a weighting bucket.

Textile Specialty Group Postprints Volume 23, 2013 131 FERROUS ATTRACTIONS: THE SCIENCE BEHIND THE MAGIC

Several of the systems were based on confi gurations previously described in conservation literature (Potje 1988; Ritschel 2011; Heer 2012; Migdail 2012; Spicer and Owens 2013; Stein 2013; Spicer and Dunphy 2015). Th ey were included for participants to see not only how they worked, but more importantly, how they could be altered for diff erent situations. Since more has begun to be written about magnets, much of the tests focused on the variations of ferromagnetic material. Th e magnet selection was more constant. Below, each test is described. Th e Neodymium magnet for all cases were disc, grade N42, 1/8” thick, and in a range of diameters. Th e poles were all axially oriented.

4.1 TESTS Th e fi ve tests were performed to test the sheer strength of the magnets. Each of the tests was performed by four diff erent groups (see table 2). Some of the materials used in the tests needed additional information and is included below.

4.1.2 Metallic Cups Several suppliers sell metallic cups in which rare earth magnets fi t. Th e cups are designed to be embedded into wood, in order to have a fl ush relief, and are secured with a screw. Th e cups are made of steel with a nickel-plated coating. While the type of metal used enhances the magnetic power as previously described, it is the presence of the additional vertical sides of the cups that increases and focuses their magnetic fi elds, creating added strength.

4.1.3 Flexible magnets Th e fl exible magnet is a type of ceramic magnet where the magnetic material is dispersed in a binder, such as vinyl or rubber, when the magnet is formed. Th e pull force is quite weak, but in order to increase the strength, the polar directions are arranged as a Halbach Array (fi g. 5). It is this alternating polarity that creates a modest attraction. Th ickness of the fl exible material is in direct relation to the pull force of the magnet. Flexible magnet specifi cations are simple: Th e thicker the sheet, the stronger the magnet. Th e weak strength of this material was confi rmed by all test groups.

4.1.4 Iron Powder As an experimental receiving agent, iron powder, commonly known as “Magnetic Paint”, was prepared in three diff erent ways. First, it was mixed with an acrylic paint according to the manufacturers directions. Second, it was bulked with epoxy and applied directly to a wooden block. Th ird, it was bulked with epoxy but forced into needle-punched polyester batting. Th e varying amounts of the iron powder were added to modify the variables in the kits.

Figure 5: Halback array-alternating polar direction to increase pull force.

Textile Specialty Group Postprints Volume 23, 2013 132 GWEN SPICER magnet is exible magnets are in a steel cup. cup. in a steel magnets are mbinations of magnet, steel magnet, steel of mbinations very are e larger magnets in this test Keep the one rare earth rare the one magnet separate Keep Th exible magnets. exible 0.125 fl of magnets. A second strip exible the iron has three thicknessesmethods is test applying and of Neodymium three and bar steel thicknesses has two of is test thicknesses three and bar steel thickness of has three is test three and washers thicknesses steel has two is test of the other and a block on one bars, steel 24-gauge has two is test provided. In the literature, Mylar was used as a separating bar- was used as a separating Mylar the literature, In provided. bars. steel and magnets exible fl of combinations various Test rier. recording test Return amounts. weight observations and Record presenter. sheet to the fl from Th has increas- magnets. Each option Neodymium two and powder magnet is in a One ½” diameter powder. iron of amounts ing powder iron magnets and of combinations various Test cup. steel test Return amounts. weight observations and Record samples. presenter. sheet to recording Th ½” diameter magnets (grade 42). Two Record bars. steel magnets and of combinations various Test sheet to recording test Return amounts. weight observations and presenter. Th of fl Th magnets are ½” diameter magnets (grade N42). Two Neodymium washers. magnets and of combinations various Test cup. in a steel recording test Return amounts. weight observations and Record presenter. sheet to having Avoid powerful. care in handling is necessary. Great shock can demagnetize abruptly, magnets hit one-another them. Th sleeve. webbing a prepared which slides into and powder-coated and bar aluminum an to both attached magnets are Neodymium various test to is provided webbing loose. A second prepared co various Test magnet placements. amounts. weight observations and sleeves. Record webbing and presenter. sheet to recording test Return # 3 4 4 # 1 2 4 # 5 2 4 # 4 2 4 # 1 4 2 4 (lbs) 1 ½ 1/4 (lbs) 1 ½ 1/4 (lbs) 1 ½ 1/4 (lbs) ½ oz 1/8 1/4 (lbs) 5 1 ½ 1/4 Table 2: Overview All Tests of Table with pockets for for pockets with ½” magnets (24 gauge) steel sleeves Removable Side Removable sleeve webbing • • coated powder- webbing 2 • ¼” ½” ½” in cup magnet: Flexible (0.03) (0.06) (0.125) ½” dia. ½” ½” in cup 1” 1” in cup ½” ½” in cup 1” 1” in cup ” Side magnet attached to to magnet attached aluminum 5/8” batting magnet (0.06) exible Receiving Side Magnetic side Weights Notes “ Slat • (24 gauge) bar Steel diameter 3/4” • • 3/8” & cups, Empty .001 .025 (24 gauge) ½” in cup (.001) tape foil (0.01) steel (.025), (24 gauge) steel fl fender washer fender washer steel thicker metallic ½” and cups, 1” in cups powder Iron “Magnetic” paint • surface on painted • epoxy with mixed into embedded • Halbrow and and Halbrow Potje 1988; Potje Velcro Alternative Alternative Velcro

Yellow : Idea Publication: 2012 Wood Orange: Gauge Steel Publication: 2012 2011; Hovey Taira Red: Magnets Flexible Publication: 2013 2012; Migdail Heer Local Spots Local Publication: Team Green: et al 2007; Spicer Keynan 2011; Hovey 2009; Ritschel Owens and 2012; Spicer 2013 Blue: Powder Iron Publication: 2013 2008; Stein Sheesley

Textile Specialty Group Postprints Volume 23, 2013 133 FERROUS ATTRACTIONS: THE SCIENCE BEHIND THE MAGIC

5. OBSERVATIONS

Th e fi nal section of the session was when, as a group, we discussed our observations of the various trials. Each group was allowed to speak while a volunteer recorded the comments. Th e recorded comments are collated in table 3. Th e observations are divided into comments about the test and the gap materials. Expressed observations were mainly related to the gap materials used by participants.

5.1 DISCUSSIONS OF THE TESTS Any discussion relating to weighted tests, as in this hands-on session, requires a mention of static and dynamic forces. Static can be described as a load that moves slowly, as opposed to one having acceleration. Th e method of weight placement in the bucket by any one group can greatly aff ect the results. Th erefore a weight that is gently and slowly placed will have a higher weight result than a dynamic test where the weights are dropped.

Table 3: Comments recorded during session

Test Test Comments Gap Comments Green One group recommended, “buy the cup!” Th e felt/batting diminished the strength of the While, they also mentioned that it left a mark or magnet’s strength. Th is observation represents impression on the paper. the whole ideas of the thicker the gap, the weaker the pull strength. Blue Th e powdered iron embedded into the batting Mylar on the outside was better than when created the best results. Groups clearly saw that placed on the inside. Th is was noticed by other the increase in the concentration of iron pow- groups too. Nap-to-nap surface was better. der held better. Th e 1” disc magnet in a cup did Alluding to the fact that friction is playing a role not hold more weight than the ½” disc in a cup in the system. on average. Th is was seen on all tests. Orange Th e thin foil (.001) steel did not even hold the When the Mylar was next to the steel, it failed at bucket. (Th e average weight was 40 grams) 24 ½ lb. where as, when the fabric was placed next gauge steel held the cup. to the steel, it stayed at ½ lbs. Other groups also noticed this. Best results were when the suede was between and in the gap. Red Th e overall concession was that Flexible mag- All felt that the strongest was with the suede in nets do not hold much weight. One group was the gap. able to hold as much as 1-½ pounds using the 0.125 thick magnets. Yellow Not a lot of sheer strength None Magnet needs to be smooth when using the cup Mock-up is essential Discussion of how to adjust the lower lip of the “L” slat.

Textile Specialty Group Postprints Volume 23, 2013 134 GWEN SPICER

Table 4: N 42 1/2” disc. ϫ 1/8” thick with all of the ferromagnetic options

.001 steel .01 steel .025 steel Fender Th ick Painted Epoxy mix Embedded washer batting Less than ½ lbs 1 1/8 lbs ¾ lbs 1 lb Less than ~40 grams 1/8 lbs 40 grams 40 grams

Participants quickly found that the amount and thickness of the ferromagnetic materials greatly aff ected the strength of the magnets. Th is was seen regardless of what form of ferromagnetic material was used: washers, steel sheet, or powdered iron. Neither the foil tape (0.001), nor the powdered iron in the paint medium was found to be strong enough to hold the bucket with any of the magnets. Large diff erences in magnet size did not aff ect the pull strength (1” to ½” was the same) (table 4). In all groups, the same ratio of decrease in pull force of any test was seen. In essence, as more layers were placed between, and the thickness of material was increased, the pull force decreased. Besides the distance that aff ects the force fi eld, surface quality and area of the materials are also at play. Th e physical properties (smoothness, friction, or static) all contribute to a small degree (table 5). Participants noticed diff erences with the suede and Mylar layers. Th ese contributing forces need more investigation. Friction, as with other mounting systems, works the same; the smoother the surface of the gap material, the lower the hold. When developing a magnet system, the types of materials selected to be surrounding the artifact (i.e. the materials in the gap) can play a role in the success of the system and when in close proximity, they can have the ability of added attraction. With more grab between the layers, less pull force is necessary, and hence the chance for damaging of the artifact due to compression can be reduced. Th e presence of the pull force applies the necessary pressure to lock the fi bers together. Slippage is hence reduced (SmallCorp 2012). It is common for conservators to use Mylar as a barrier, but this might need to be reconsidered when it comes to magnetic systems. Mylar is oft en used as a barrier in order to prevent unwanted materials to transfer to the artifact (Heer 2011; Migdail 2012). As a barrier, Mylar does quite well, however, as its smooth surface works counter to the holding powers of the magnetic pull force. Static charge that is builds up with Mylar does not aff ect the magnet itself or the ferromagnetic material. But the static can have a role with the other materials and the artifact. During this hands-on training session, participants noted a marked diff erence depending on where the Mylar layer was located. One might want to rough up the surface in the location of the magnet.

Table 5: Gap material characteristics

Static Friction Surface Cotton Acrylic Polyester Th ickness Paper s x 0.0036 Mylar x s x 0.003 Fabric x r/s x 0.011 Batting x r x 0.095 Suede x r 0.025 Webbing x x 0.02 KEY: s-smooth, r-rough

Textile Specialty Group Postprints Volume 23, 2013 135 FERROUS ATTRACTIONS: THE SCIENCE BEHIND THE MAGIC

6. CONCLUSIONS

Th e activity was designed as a learning experience while also serving as a fun introduction to magnetic systems. It appears that both were achieved. Participants were able to deal with many of the issues in creating and altering magnetic systems. Tests were quickly performed. Th is could have altered the results, as the weights were added more quickly or heavy-handedly than might have been done if more time had been provided. Th is gives a less precise result than in a more “controlled” experiment. However, none of the group observations mentioned this type of phenome- non. Most signifi cantly, the importance of the ferromagnetic materials and the gap materials in a magnetic system was demonstrated. Th is allowed for a fuller understanding of the all the parts of a magnetic system.

ACKNOWLEDGMENTS

A special thanks is extended to SmallCorp, Inc who generously made all of the jigs and components for the session. Th is hands-on session would never have gotten off the ground without several individuals, Van Wood, who created the jigs from drawings; Bob Hunnes who created the initial drawing of the jigs, without which there would be no activity. My very deep gratitude goes to Virginia Whelan, the organizer of the TSG meeting this year; and Joy Gardiner who made a fateful phone call late last summer. Special thanks also go to Kathleen Kiefer who made the many pre-weighted sand bags and Robin Hanson who created the beautiful bags outfi t- ted with unique buttons. I also want to thank Ruth Seyler, Eric Pourchot, Steven Charles and everyone else at the AIC offi ce, who graciously helped at the annual meeting, ensuring that all of the components were where they needed to be. Finally, extensive thanks go to those who helped with the preparations before the session and this paper, including Nicolette Cook, Denise Migdail and Barbara Owens.

REFERENCES

Brown, D.N., B. Smith, B.M. Ma and P. Campbell. 2004. Th e Dependence of magnetic properties and hot work- ability of rare earth-iron-boride magnets upon composition. IEEE Transactions of Magnetics. 40 (4) July 2004.

Dignard, C. 1992. Tear repair of skins with minimal access to their backs: the treatment of a kayak. Leather Conservation News 7(2): 1–8.

Feynman, R., R. Leighton and M. Sands. 1964. Magnetic materials. Th e Feynman Lectures of Physics. California Institute of Technology, Addison-Wesley Publishing Company Inc.: Reading, MA.

Heer, S, S. Freeman, M. Harnly, L. Kaneshiro, E. Mack and R. Stroud. 2012. Th e Sacrifi ce by James Nachtwey: gallery installation of a 32 Foot photograph with fl exible magnets. Poster at the 40th Annual Meeting of the American Institute for Conservation, Albuquerque. NM: J. Paul Getty Museum.

Holbrow, K., and K. Taira. 2011. http://www.conservation-wiki.com/index.php?titleϭMagnet_Mounts (accessed 10/27/13).

Hovey, D. 2012. Short communication: receiver systems for object mounts: A design concept for safe installation and display fl exibility. Journal of the American Institute for Conservation 51(1): 51–58. Jones, N. 2011. Th e Pull of stronger magnets. Nature. 472. April 7, 2011.

Textile Specialty Group Postprints Volume 23, 2013 136 GWEN SPICER

Keynan, D., J. Barten, and E. Estabrook. 2007. Installation methods for Robert Rymen’s wall-mounted work. Th e Paper Conservator 31:7–16. Livingston, J. D. 1996. Driving force, the natural magic of magnets. Harvard University Press, Cambridge, MA. Migdail, D. K. 2012. Personal communication, Asian Art Museum, San Francsico, CA. Potje, K. 1988. A Traveling exhibition of oversized drawings. Th e Book and Paper Group 7, Washington D.C.: AIC. 7: 52–55. Ritschel, C. and K. Douglas. 2011. “Lace in fashion” On the Body Fourth AICCM Textile Symposium, National Gallery of Australia, Melbourne, Australia. Sheesley, S. 2008. Vortices and Reveries: an innovative treatment of oversized three-dimensional paper objects. Th e Book and Paper Annual, 27:79–88. Spicer, G. 2009. Th e Re-tuft ing a Hunzinger arm-chair using magnets. North American Textile Conservation Conference. Quebec City, Canada. . 2010. Defying gravity with magnetism. AIC News (Nov. 2010) 35(6): 1, 3–5. . 2013. An alternative to Velcro? Upper edge hanging methods using rare earth magnets. Western Association of Art Conservation 35(3): 20–24. . 2013. Conservation treatment of a Hunzinger cantilevered armchair including the use of magnets to create tuft ing. Journal of the American Institute for Conservation vol. 52(2): 107–122. and Owens. 2013. Magnetic mounts. Inside the Conservator’s Studio. January 11, 2013. . 2013. A Ferrous attractions. Inside the Conservator’s Studio. August 14, 2013. Spicer, G. 2014. What is your base knowledge about Rare Earth or Neodymium magnets? Inside the Conserva- tor’s Studio. December 31. http://insidetheconservatorsstudio.blogspot.com/2014/12/what-is-your-base- knowledge-about-rare.html (accessed 3 June 2016). . 2015. A magnet is only as strong as . . . Inside the Conservator’s Studio. May 28. http://insidetheconservatorsstudio.blogspot.com/2015/05/a-magnet-is-only-as-strong-as.html (accessed 3 June 2016). . 2016a. Ferrous Attractions: Th e science behind the conservation of using rare-earth magnets. Journal of the American Institute for Conservation. 55 (2): Publication pending. . 2016b. Magnet use trends in art conservation, revealed from a survey and literature search. E-Conservation. www.e-conservation.org. Publication pending. Spicer, G and M. Dunphy. 2015. Magnets as an alternative to Velcro. Poster presented at Th e American Insti- tute for Conservation’s 43th Annual Meeting, Miami, FL. http://www.conservation-us.org/docs/default- source/annualmeeting/2015am_poster_90.pdf?sfvrsn=2 (accessed 3 June 2016). Stein, R. 2013. Personal communication. Chief Conservator, Michael C. Carlos Museum, Atlanta, GA. Th e Magnet Story. 1998. Electronic Materials Manufactures Association of Japan (EMAJ) Woods, V. 2012 and 2013. Personal communications. SmallCorp, Inc. Greenfi eld, MA.

Textile Specialty Group Postprints Volume 23, 2013 137 FERROUS ATTRACTIONS: THE SCIENCE BEHIND THE MAGIC

SOURCES OF MATERIALS

Disc Neodymium with nickel-plating, ¼”, ½”, and 1”, Grade 42; Flexible magnets: Ferrite bonded strips with synthetic rubber, style 0.03, 0.06, and 0.125; Cups for magnets: Steel, ½” and 1”; Steel washer 1 1/8”

McMaster-Carr P.O. Box 5370 Princeton, NJ 08543-5370 (609) 689-3000 (609) 259-3575 www.mcmaster.com

Washers: 1” Fender Local hardware

Steel: .001, .01, and .025 (24 gauge) Iron powder Magically Magnetic PO Box 219 Saxonburg, PA 16056 (724) 352-3747 www.lyt.com, [email protected],

1/16” and 4.5 oz., Buff alo Felt Products Corp. 14 Ransier Drive Buff alo, NY 14224 (716) 674-7990, ext. 207 www.buff alofelt.com

Mylar: 2 mil uncoated polyester fi lm (TFM001010), Talas 330 Morgan Ave Brooklyn, NY 11211 (212) 219-0770 (212) 219-0735 http://www.talasonline.com

Novasuede: nylon fi ber Majilite 1530 Broadway Road Dracut, MA 01826 (978) 441 6800 (978) 441 0835 www.majilite.com

Textile Specialty Group Postprints Volume 23, 2013 138 GWEN SPICER

Paper: 100% cotton rag, unbuff ered, Photo-Tex tissue, pH 6.8-7.2 (7-1185), Archivart 40 Eisenhower Drive Paramus, NJ 07652 (888) 846-6847 (870) 268-0581 www.archivart.com

100% cotton, 1-1/2” wide (Tape 4); unbleached, desized, cotton print cloth, 108 g/m2 (400U) Testfabrics 415 Delaware Ave PO Box 26 West Pittson, PA 18643 (570) 603-0432 (570) 603-0433 www.testfabrics.com

GWEN SPICER is a textile, upholstery and objects conservator in private practice. She earned her MA in Art Conservation from Buff alo State College, and has since taught and lectured around the world. In her private practice, she assists many individuals and organizations of all sizes with storage, collection care, and exhibitions, and has become known for her innovative conservation treatments. She is a Fellow of AIC. Contact: 305 Clipp Rd., Delmar, NY 12054. Tel: 518-765-2142. [email protected]. www.spicerart.com.

Textile Specialty Group Postprints Volume 23, 2013 139

DANCING ON A WIRE: ARTICULATION SOLUTIONS FOR MANNEQUINS IN THE CIRCLE OF DANCE EXHIBITION AT NMAI-NY

SHELLY UHLIR

ABSTRACT—Th e creation and fabrication of ten articulated custom mannequins for garments and associated objects in the Circle of Dance exhibition at the National Museum of the American Indian—New York is discussed in this paper. Th e exhibition, which opened in September 2012, showcases dances of ten tribes from diff erent geographical locations in the Western Hemisphere. Each of the mannequins is articulated in a gesture specifi c to an important dance from each tribe, so creating the illusion of specifi c movement as well as providing distinctive facial features and hand gestures was imperative. A variety of arm and leg connection techniques that utilized mechanical and magnetized attachments were employed. Th e faces, hands and feet were created from Fosshape, a felt-like material that hardens somewhat when steamed. Topics discussed are: choosing poses for maximum illusion of motion; creating gesture through arm, leg and head connection techniques; and extension of the articulation to the expression of the faces and hands through the use of Fosshape.

BAILANDO EN LA CUERDA FLOJA: SOLUCIONES DE ARTICULACIÓN PARA LOS MANIQUÍES DE LA EXHIBICIÓN “CIRCLE OF DANCE” DEL NMAI-NY—Este documento habla sobre la creación y fabricación de diez maniquíes articulados para lucir la vestimenta y objetos asociados de la exhibición “Circle of Dance” del Museo Nacional Indo-Americano de Nueva York. La exhibición, que se inauguró en septiembre de 2012, presenta danzas de diez tribus de diferentes lugares geográfi cos del hemisferio occidental. La pose de cada maniquí representa un gesto específi co de una danza importante de cada tribu, por lo que fue necesario crear la ilusión de un movimiento específi co y mostrar los rasgos faciales y gestos corporales distintivos de cada una. Se utilizó una variedad de técnicas de conexión de brazos y piernas con sistemas mecánicos y magnéticos. Los rostros, manos y pies fueron hechos en Fosshape, un material similar al fi eltro que se endurece con vapor. Los temas de discusión son: la elección de las poses para generar una mayor sensación de movimiento; la creación de gestos a través de las técnicas de conexión de los brazos, piernas y cabeza; y la extensión de la articulación a la expresión de los rostros y manos con el uso del Fosshape.

1. EXHIBITION BACKGROUND

Th e Circle of Dance exhibition was installed in September 2012 in the Diker Pavilion, the fi rst fl oor gallery space at the National Museum of the American Indian’s (NMAI) New York branch, located in the Alexander Hamilton Customs House in lower Manhattan. Th e room is a large, oval space generally dedicated to educational and other programming events. Ten exhibition cases are built into the walls around the room. Each of the ten cases for this exhibition displays an outfi t representing a tribe from one area of the western hemisphere through a depiction of a culturally specifi c dance. Th e exhibition is expected to be on display for at least fi ve years (NMAI 2012). Th e primary goal of this exhibition was to illuminate the signifi cance of each dance and to highlight the unique characteristics of its movements and music. Th erefore, creating the illusion of specifi c movement in the mannequins was imperative. With one mannequin in each case, an additional challenge was to avoid creating a ‘people trapped behind glass’ look. While appropriate gesture was DANCING ON A WIRE: ARTICULATION SOLUTIONS FOR MANNEQUINS IN THE CIRCLE OF DANCE EXHIBITION AT NMAI-NY important, the forms were also to be clearly seen as exhibition mannequins, not as realistic-looking people. A fi lm loop showing dances at the front of the exhibition and historical photographs on text panels fi ll in the visual blanks on what the real dancers look like. Th e mountmaking challenge was to fi nd a way to fulfi ll these goals respectfully and accurately while staying in budget and keeping the collections safe over the long-term.

2. MANNEQUIN CONCEPTS

Custom-carved Ethafoam mannequins similar to those made for previous NMAI exhibitions were created for the basic garment supports (Uhlir 2012). Initially, the designers of Circle of Dance were inspired by the dramatic use of mannequins in the Alexander McQueen Savage Beauty exhibition at the Metropolitan Museum of Art (Metropolitan Museum of Art 2011). Th ey hoped to fi nd a similarly exciting style of pre-made mannequin that would be appropriate for this show. However, the garments in NMAI’s collection are rarely made to a specifi c dress size and the styles of fashion mannequins available did not suit the traditional Native American dance poses and facial features in the exhibition, so standard off -the-shelf mannequins were not a practical option. Methods of fabricating mannequins from Ethafoam have been extensively published, so only the general process used at NMAI will be discussed here: once the team had chosen an appropriate pose and a mountmaking plan was sketched out, paper templates were created using tracings and measurements from the garments. Template shapes were transferred to Ethafoam then cut to fi t and heat-welded together. Th ese blocks were carved and fi t to the clothing, then covered with padding and fi nishing material. For the visual elements of faces, hands and feet, a variety of styles and materials were investigated, from wire Calder-like profi le structures to carved Ethafoam heads coated with FoamCoat, but for this exhibition, the most success was achieved with Fosshape 300, the lighter weight version of the material. Use of Fosshape is discussed later in this paper. A wide range of gesture and articulation was achieved in these mannequins (fi g. 1). Th ere were three general areas of focus that helped to bring these mannequins to life. Th ey were: overall pose articulation, joint articulation, and extremity articulation.

3. OVERALL POSE ARTICULATION: Implying Kinetics through Pose Choices

To narrow down the range of movement that would be indicative of each dance and to fi nd dynamic yet appropriate poses, curators, tribal members, education specialists, as well as images and videos were consulted. Conversations with the designer helped reduce pose choices further to those that would fi t within the special restrictions of the exhibition cases. Each deck was 24 inches deep, but the glass doors slanted back to reduce the depth at the top of the case to 19.75 inches (fi g. 2). Concurrently, conservators advised on which poses would be least stressful to the garments. To translate this input into a functional mannequin pose that still implied movement, fi ve general details were considered: age appropriate posture, weight distribution, natural shoulders, torso twist and head tilt. First of all, the age and social position of the person was considered. Th e three mannequins shown in fi gure 3a-c demonstrate these diff erences. Th e man on the left would have been a powerful middle-aged

Textile Specialty Group Postprints Volume 23, 2013 142 SHELLY UHLIR

Fig. 1: Circle of Dance mannequins awaiting dressing. National Museum of the American Indian Smithsonian Institution. Photo by Conservation Staff .

Tlingit leader, with a strong, straight back, leaning forward as he stomps. Th e woman on the right implied a more elderly Mapuche woman who is in a more arched and shuffl ing posture. In the center, a young Yakama girl, around seven years old, was represented in an open, playful pose. Th e Hopi Butterfl y Dancer’s movements (fi g. 4a-b) needed to be subtle and controlled, but even within its subtlety it demonstrates two additional motion-creating details. Her pose was frozen in mid-step to add a feeling of imbalance. Th is created an illusion of movement when the viewer followed the step to its natural landing point. Several of the other mannequins incorporated this contrapposto as well. Additionally, her right shoulder was lowered slightly in order to add another subtle hint of action. Creating the slope and natural asymmetry of the shoulders was a small thing that seemed to make a big diff erence in all the mannequins. Symmetrical shoulders tend to look stiff and perfect but therefore unnatural in a state of dance. Th e Yupik Quyana (Th ank You) Song Dancer shown in fi gure 5a-b has a slightly more expansive pose. Th e pose gets its apparent movement from several elements. As with the Hopi dancer, she is slightly off -balance with one foot engaged and one about to step. Her right shoulder lowers as her upper torso leans back. Additionally, her torso twists naturally toward her raised left hand. All the mannequin torsos were made in two pieces, split at the waist, to allow a twist of the upper torso as needed (fi g. 5b). Th ese two-piece torsos also aided in dressing. Additionally, the face has been adjusted to look at her left hand. All the hollow mannequin faces were pinned around more solid internal structures, such as the neck. Th is allowed for an adjustment of the head tilt or eye glance to best capture the gesture of the dance. Where an object needed to be supported by the head, a separate padded brass or foam mount was made, allowing the face to remain somewhat mobile for slight adjustment case side.

Textile Specialty Group Postprints Volume 23, 2013 143 DANCING ON A WIRE: ARTICULATION SOLUTIONS FOR MANNEQUINS IN THE CIRCLE OF DANCE EXHIBITION AT NMAI-NY

Fig. 2: Case drawing for the Circle of Dance cases. National Museum of the American Indian, Smithsonian Institution. Drawing by Gerald Breen.

Textile Specialty Group Postprints Volume 23, 2013 144 SHELLY UHLIR

Fig. 3: Tlingit Ku.eex’ Entrance dancer mannequin (a). National Museum of the American Indian, Smithsonian Institution (21/6806 shirt, 08/1888 headdress, 21/1650 blanket, 03/3913 dance apron, 16/2771 leggings, 08/3429 moccasins, 05/4185 rattle). Photo by Ernest Amoroso; Yakama Girl’s Fancy Shawl dancer mannequin (b). National Museum of the American Indian, Smithsonian Institution (26/8788). Photo by Ernest Amoroso; Mapuche Mutrum Purun dancer mannequin (c). National Museum of the American Indian, Smithsonian Institution (EP0953). Photo by Ernest Amoroso.

Fig. 4: Th e Hopi Butterfl y Dancer mannequin before dressing (a). National Museum of the American Indian, Smithsonian Institution. Photo by Conservation Staff ; and the Hopi mannequin ensemble dressed (b). National Museum of the American Indian, Smithsonian Institution (26/8785 kopatsoki, 16/6629 earrings, 24/7776 manta, 25/5435 sash, 09/0566 anklets, 25/6226 bracelets, 25/6329 bracelet). Photo by Ernest Amoroso.

Textile Specialty Group Postprints Volume 23, 2013 145 DANCING ON A WIRE: ARTICULATION SOLUTIONS FOR MANNEQUINS IN THE CIRCLE OF DANCE EXHIBITION AT NMAI-NY

Fig. 5: Th e Yup’ik Quyana (Th ank You) Dancer mannequin ensemble (a). National Museum of the American Indian, Smithsonian Institution (25/8687 dance fans). Collection of Chuna McIntyre (headdress, parka, belt, mukluks, leggings, necklace) Photo by Ernest Amoroso; and the Yup’ik mannequin before dressing (b). National Museum of the American Indian, Smithsonian Institution. Photo by Conservation Staff .

4. JOINT ARTICULATION: Gesture and Movement through the Connections at the Joints

Th e next aspect of articulation came at the arm and leg connections. Although each mannequin required a slightly diff erent set of solutions, simplicity was the ultimate goal. Keeping the needs of each piece of the ensemble in mind, arm and leg connections were designed to require as little manipulation of the garment as possible.

4.1 THE SUBTLE APPROACH In the case of the Hopi dancer, a simple plug connection was implemented at the shoulder because the arms were not holding substantial weight (fi g. 6a-b). Th e square Ethafoam plug kept the lightweight arms from rotating. Th e plug was cut slightly larger than the receiving hole to allow some friction and keep the arm in

Textile Specialty Group Postprints Volume 23, 2013 146 SHELLY UHLIR

Fig. 6: Details of the Hopi mannequin arm, before insertion (a); and inserted (b). National Museum of the American Indian, Smithsonian Institution. Photo by Conservation Staff . place, but it was not so tight that the arm would be hard to insert. Th e twill tape and T-pin connection provided added security to keep the arm from releasing as well as some adjustability in the angle of the arm as needed. To avoid potential collapse over time, the lift ed left leg was permanently attached to the torso by an internal aluminum structure similar to that shown in fi gure 7a. Th e right leg was free fl oating and cut to fi t around the steel support pole. A twill tape wrap connects with Velcro at the ankle and holds the leg in place around the pole so that the ankle band does not need to take any expansive pressure (fi g. 7b). Conservators stitched a twill-tape band and false ties to the ankle bands so the actual ties would not need to be used. Th ere is a magnet embedded at the top of the leg which connects to the steel support “rake” structure that supports the torso (fi g. 7c). Twill tape and t-pins added extra security to hold the leg in position.

4.2 THE DYNAMIC APPROACH Th e Quechua Scissor Dancer mannequin in fi gure 8a-b demonstrates a very diff erent set of arm and leg connections. Where the Hopi dance is modest and subdued, the Scissor Dance is extremely athletic. Th is dance required a pose that would indicate the energy of the dance, but still fi t within the confi nes of the exhibit case. Th e ensemble was not an accessioned outfi t but was on loan from the Museum’s education department and had

Textile Specialty Group Postprints Volume 23, 2013 147 DANCING ON A WIRE: ARTICULATION SOLUTIONS FOR MANNEQUINS IN THE CIRCLE OF DANCE EXHIBITION AT NMAI-NY

Fig. 7: Th is internal hip structure, using aluminum tubing, is similar to that used in the Hopi mannequin (a); Hopi mannequin detail showing the right leg connection along the support pole at the ankle (b); and a detail showing a magnet connection to steel support rake for the torso (c). National Museum of the American Indian, Smithsonian Institution. Photo by Conservation Staff .

Fig. 8: Side view of the Quechua mannequin in process, showing the steel wall bracket (a); and the same mannequin installed (b). National Museum of the American Indian, Smithsonian Institution. (EP0954) Photo by Conservation Staff .

Textile Specialty Group Postprints Volume 23, 2013 148 SHELLY UHLIR been worn during a performance at the Museum. Th e fabric and seams were strong, so bends in the knees and arms were easy to attain. Even so, folds and creases in the fabric were minimized. For the Scissor Dancer mannequin, two diff erent arm connections were implemented. Th e left arm was fi xed because the hand would only be carrying a light scarf and the jacket had a functioning zipper that allowed the front to open fully to slide over the extended arm. However, because the right arm of the form would carry the weight of heavy iron “scissors”, an aluminum keyed-arm connection that had been designed and fabricated by Robert Patterson, NMAI Exhibits Specialist, for a previous exhibition (Uhlir 2012) was used (fi gs. 9a-c). Th e keyed connection was not ideal since it required a fair degree of shoulder manipulation to engage, but since the open jacket allowed good visibility and accessibility during dressing and the garment was strong, it was decided that this hardware would be acceptable to use in this case. Once the key was engaged, the arm stayed strongly fi xed in its position.

Fig. 9: Th e Quechua mannequin torso in process, exposing the keyed-arm connection (a); a detail of a keyed-arm connection before insertion (b); and a close-up of the connection engaged (c). National Museum of the American Indian, Smithsonian Institution. Photo by Conservation Staff .

Textile Specialty Group Postprints Volume 23, 2013 149 DANCING ON A WIRE: ARTICULATION SOLUTIONS FOR MANNEQUINS IN THE CIRCLE OF DANCE EXHIBITION AT NMAI-NY

Th e legs of this dancer were also very animated. To allow for some onsite adjustment of the knees, aluminum and plastic Variloc joints made by Adjustable Locking Technologies were used (Williams 2012). Squeez- ing the joint releases the teeth of the joint to allow incremental movement, then lock back into position when released (fi gs. 10a-c). Th is joint was useful in this application but would not be appropriate for a more delicate garment if the access to the joint could only be external. It also worked here because the joints only needed to support the weight of the legs and shoes, not the entire form. Th e joints were not strong enough to solidly support the full weight of an Ethafoam mannequin without some supplementation.

Fig. 10: Th e Quechua mannequin before dressing (a); Here, the Variloc joint is exposed during the process of making the legs (b); To release the Variloc joint for adjustment, a squeezing action is necessary (c). National Museum of the American Indian, Smithsonian Institution (EP0954). Photo by Conservation Staff .

Textile Specialty Group Postprints Volume 23, 2013 150 SHELLY UHLIR

Th e right leg was fi xed to the torso with a supplemental aluminum tube embedded in the foam. Th e left leg is connected to the hip with two magnets to keep it from twisting and twill tape with t-pins to secure. To increase the illusion of the leap, a steel wall armature that connected the two halves of the torso together at the waist and allowed for adjustability in the torso twist on site was fabricated (fi g. 8a).

5. EXTREMITY ARTICULATION: Expression and Gesture in the Heads and Hands

To complete the illusion of movement in these mannequins, expressive faces and hands were essential. Foss- hape, a felt-like polyester-based material that hardens somewhat when steamed was critical to our fi nal process (Amnéus and Miles 2012). Its ease of use, light weight characteristics, and relatively low cost made it a perfect material for this project.

5.1 Heads In an eff ort to save time, ready-made Ethafoam heads from Dorfmann were purchased, which gave a good proportional starting point and were easy to adapt as needed. To model the features, a pulpy cellulose material called Sculptamold, a non-toxic, kid-friendly product was used. It comes as a loose, dry powder but when mixed with water becomes a sloppy cross between clay and plaster. It was used to add brows, noses, and cheekbones to the Ethafoam heads (fi g. 11a) that would then be used as a mold for the Fosshape. Th e specifi cs of each face were based on photographs of dancers chosen by the curator or a representative of the tribal community. Since the Fosshape tends to shrink across the high points, more time was spent on the developing the details of those areas. For our purposes, the Sculptamold areas were left rough, but it can be wet smoothed or dry sanded if needed. Once the mold was made and dried, Fosshape 300 was loosely pinned inside-out over the form using a general pattern that placed seams in the less noticeable areas above the hairline and along the jaw line (fi g. 11b). Th e lighter weight Fosshape was used because it took details slightly better than the 600 weight. Also, the strength of the heavier weight was not required for these items and the lower cost was more practical. Even though faces were not entirely symmetrical, they were close enough to allow the material to be pinned inside out then removed from the mold for stitching. Th e Fosshape was then machine stitched as far as possible, trimmed, then re-positioned over the mold right-side out and hand stitched the rest of the way to close around the entire head mold (fi g. 11c). Once the head was covered, it was steamed with a standard Jiff y steamer. Heat- ing the Fosshape stretched the Fosshape around the face form. To capture more detail in the facial features, the freshly steamed material was pushed into the concavities of the mold until the material cooled and hardened. Aft er at least fi ft een minutes of cooling time to allow the material to fully set up, the stiff ened Fosshape was cut off the mold in a non-visible area in the back of the head with a rotary cutter then peeled off the form. Th e Fosshape snapped back into position easily aft er being mildly distorted. In cases where the head would need to receive mounts to hold earrings or headdresses, or when a tight wig might collapse the Fosshape, the inside of the hollow head would be supplemented with Ethafoam (fi g. 11d). Additionally, when weight on the head would be too much for the foam, a supplemental brass tube was inserted into the neck and head areas. Th e Fosshape was wrapped around this support and pinned to the foam to secure it in place. Most of the faces were fi nished with a neutral grey matte low-VOC latex paint. Th e color was chosen to be a similar value to the dark blue of the case fabric yet to avoid looking too much like a particular skin color. In the case of the Hopi mannequin, representation of face and body paint was also required. Th is was added with acrylic paints mixed with ultra-matte medium (fi g. 11e). Th e painted Fosshape was too abrasive to be in contact with the objects, so wherever an object overlapped, a barrier such as acrylic felt or Tyvek was applied (fi g. 12).

Textile Specialty Group Postprints Volume 23, 2013 151 DANCING ON A WIRE: ARTICULATION SOLUTIONS FOR MANNEQUINS IN THE CIRCLE OF DANCE EXHIBITION AT NMAI-NY

Fig. 11: Making heads from Fosshape. Ethafoam heads are modeled with Sculptamold to create specifi c defi nition (a); When the Sculptamold is dry, Fosshape is loosely pinned inside out to the head form (b); then remove, trimmed and refi tted right-side out before stitching and steaming (c); Th e steaming process hardens the material. Here, the back of Hopi mannequin head shows its cut line, internal foam support and pinning connections to the neck (d); All visible areas were painted with low-VOC latex. Th e Hopi mannequin head required additional face painting detail, which was done with acrylic paint and ultra-matte medium (e); and fi nally, a detail of the Hopi mannequin head fully dressed with necklace, earrings and Kopatsoki. National Museum of the American Indian, Smithsonian Institution. (26/8785 Kopatsoki; 16/6629 earrings; 26/8833 necklace) Photo by Conservation Staff .

Nine of the ten mannequins also had synthetic wigs chosen in styles and shades appropriate to the dancer. Any additional objects attached to the heads, such as earrings, were supported with custom made padded brass mounts secured into the Ethafoam substructure (fi g. 11f).

5.2 Hands and Feet Th e process of making the hands and feet was similar to that of the heads, but due to the specifi c poses required, the molds for the Fosshape were cast instead of carved from Ethafoam. Ten separate hand poses and two diff erent foot poses were needed for the eight mannequins with visible hands or feet. Aft er a technical demonstration by former Andrew W. Mellon Fellow, Lauren Horelick, volunteers with similar sized hands worked in teams to make one hand each for a mannequin (fi g. 13a). Each volunteer was given a photo of a dancer in a specifi c pose to inspire them. Since the Fosshape retains some fl exibility even aft er steaming, the

Textile Specialty Group Postprints Volume 23, 2013 152 SHELLY UHLIR

Fig. 12: Mellon Fellow, Sarah Grace Owens, applying acrylic felt to Yup’ik Quyana Dancer mannequin head. National Museum of the American Indian, Smithsonian Institution. (Collection of Chuna McIntyre). Photo by Conservation Staff .

Fig. 13: Mellon Fellows Peter McElhinney and Lauren Horelick preparing to make a hand cast (a); and hand casts and foam arms for the Hopi mannequin ready for Fosshape application (b). National Museum of the American Indian, Smithsonian Institution. Photo by Conservation Staff .

Textile Specialty Group Postprints Volume 23, 2013 153 DANCING ON A WIRE: ARTICULATION SOLUTIONS FOR MANNEQUINS IN THE CIRCLE OF DANCE EXHIBITION AT NMAI-NY hand position or size could be adjusted slightly in its attachment to the arm or by trimming and re-stitching which allowed a certain latitude in the fi nal cast poses. Aft er the casts and forms were ready, all visible areas were wrapped in Fosshape 300, keeping stitched areas and seams to the unseen side (fi g. 14a). Because the Fosshape stretched to a more mitten-like shape if the fi ngers were not isolated during the steaming process, stitching between the fi ngers helped to keep them defi ned (fi g. 14b). For the sake of consistency, other visible areas of the mannequin’s skin were also covered with Fosshape (fi g. 13b).

Fig. 14: Steps to making a mannequin hand from Fosshape. First of all, the hand cast is loosely covered in Fosshape 300 and stitched to close, creating a seam in a less visible area (a); Th e material is stitched between the fi ngers before steaming to help maintain defi nition (b); Aft er steaming and cutting, the Fosshape peels off the form then pops back into shape (c); and a detail of the Hopi mannequin hand aft er painting and barrier layers are applied (d). National Museum of the American Indian, Smithsonian Institution. Photo by Conservation Staff .

Textile Specialty Group Postprints Volume 23, 2013 154 SHELLY UHLIR

Th e wrapped casts were steamed and cooled then cut with a rotary cutter on the unseen side of the hand (fi g. 14c). Th e hand was then peeled from the cast and quickly popped back into shape. Th e Fosshape had enough fl exibility to wrap around whatever wrist size was needed, so in many cases, the hands could just be t-pinned into Ethafoam in the mannequin’s wrist or forearm. Finally, like the heads, the hands were painted with low-VOC latex then acrylic paints. In fi gure 14d, the silver cloth and Tyvek used as barrier material between the painted Fosshape and the silver and yarn bracelets on the Hopi mannequin is visible. Th e Lakota Northern Traditional dancer seen in fi gure 15a needed to give the illusion of holding a feather fan and a beaded stick in its hands. Th is required separate mounts to support the objects while the Fosshape hands would cover and hide the mounted areas. Th e brass support mounts were padded with acrylic felt, both

Fig. 15: Lakota mannequin installed (a); Installation of the fan into the mount (b); Painted Fosshape hand fl exing to cover the fan and the brass mount (c); and a detail of the fan installed (d). National Museum of the American Indian, Smithsonian Institution. (26/7485 roach, 01/3799 fan, EP0956 arm bands, EP 0958 ankle bells, 26/7485 roach, hair ties, vest, shirt, choker, breastplate, bandolier, belt, apron, trailer, knee bands, anklets, moccasins, wrist cuff s, dance stick) Photo by Conservation Staff .

Textile Specialty Group Postprints Volume 23, 2013 155 DANCING ON A WIRE: ARTICULATION SOLUTIONS FOR MANNEQUINS IN THE CIRCLE OF DANCE EXHIBITION AT NMAI-NY inside the mount and over the edges, to avoid any scratching or cutting. Th e handle of the fan had been wrapped in a prop handkerchief for curatorial reasons, but here it also acted as an additional level of protective padding for the fan as well. Each wrist was outfi tted with a brass receiver plate (fi g. 15b). Spikes in one end of the plate inserted into the foam at the wrist and a tube in the other received the mount. Th e mount could be telescoped as needed to fi t the hand then set by tightening a 4-40 set screw. To install, the hand was fl exed around the object (fi g. 15c). eTh Fosshape was cut with scissors to give the impression of being pressed against the object while not really doing so. Once adjusted, the hand was then pinned at the wrist for security. A felt barrier covered the pinned areas and protected the beaded cuff that covered them (fi g. 15d).

6. CONCLUSIONS

Creating the illusion of movement in museum mannequins is a challenging and time-consuming task, best saved for exhibitions where the pose is an important part of the story of the garment. Not all clothing is strong enough to take the manipulation required to fi t and dress a custom mannequin. Not all exhibition schedules can allow the extra time and research it takes to safely and accurately depict the correct pose. Th e Circle of Dance exhibition, by its very name, was to highlight specifi c dances, so this challenge had to be embraced. Naturalistic motion and articulation needed to be achieved and adjusted to fi t within the vision of the curator, the physical limits of the display case, the condition of the garments themselves and the exhibition budget. Although fabrication of the mannequins for this exhibition took twice the normal time, this project off ered excellent opportunities to experiment with new methods and materials. Th e outcome was overall positive but with room for improvement and experimentation in the future.

ACKNOWLEDGEMENTS

Th anks to Cynthia Amnéus for the Fosshape inspiration and for her help with the steamer details. To the Circle of Dance Team—Cécile Ganteaume, Peter Brill, Jen Tozer, Gerry Breen, Susan Heald and Sarah Owens— thanks for pushing the boundaries beyond our comfort level and having the confi dence and patience to work through the experimentation phase. I am ever grateful to Robert Patterson for his knowledge of all metals but specifi cally for the design and fabrication of the keyed-arm system and the steelwork for the Scissor Dancer’s armature. I would like to thank Natalie Gallelli for generously sharing her expertise on casting processes, Lauren Horelick for showing us how it is done and all the hand models at our casting party: Shannon Brogdon-Grantham, Tom Evans, Susan Heald, Pete McElhinney, Kelly McHugh, Sarah Owens, Jon Pressler, and Rebecca Summerour for donating their extremities for a day.

REFERENCES

Amnéus, C. & M. Miles. 2012. A Method for Invisibly Mounting Costume Using Fosshape. Journal of the American Institute for Conservation. 51(1): 3–14.

Metropolitan Museum of Art. 2011. Alexander McQueen: Savage Beauty: May 4-August 7, 2011, http://blog. metmuseum.org/alexandermcqueen (accessed 7/23/13).

National Museum of the American Indian (NMAI). 2012. Circle of Dance. http:/nmai.si.edu/exhibitions/ circleofdance (accessed 7/8/13).

Textile Specialty Group Postprints Volume 23, 2013 156 SHELLY UHLIR

Uhlir, S. 2012. Mounting Multiple-Piece Ensembles for an Exhibition of Plains, Plateau, and Great Basin Attire at the National Museum of the American Indian. Journal of the American Institute for Conservation. 51(1): 99–116.

Williams, J. 2012. Dancing Pants! Articulated Mannequins (Legs only) for Katharine Hepburn Costumes. Poster presented at the 3rd International Mountmaking Forum, Field Museum, Chicago, IL. www.kent.edu/ museum/news/upload/dancing-pants-handouts.pdf (accessed 10/27/2013).