5 CONCLUSIONS

Limestone samples from a Naga el-Deir stela, which shows a high level of decay due to spalling and delamination in a museum environment, have been found to contain more than 10% by weight of clay minerals that are concentrated along the bedding planes. The mineralogy, morphology, and distribution of the clay were determined with optical microscopy, XRD, and SEM. These data confirm the presence of sepiolite as a major clay constituent, along with minor amounts of palygorskite. Cyclic wetting/drying laboratory tests using ethylene glycol and distilled water, relative humidity cycling, and thermomechanical analyses corroborate a working hypothesis that the swelling of sepiolite and palygorskite is the mechanism responsible for the observed fracture development and spallation damage. When water enters the pore system of the limestone, the clays expand primarily due to osmotic swelling and, to a minor extent, due to crystalline swelling promoted by the presence of sodium ions from halite and soda niter. Due to the presence of salts within the pore system of the limestone, water condensation takes place at RH values below 75%. In a museum environment, RH changes promote water condensation within the pore system of the limestone, resulting in eventual swelling of the clays. As demonstrated by the RH cycling within an environmental chamber, this condensation produces the cracking and the fracture development typically observed in decayed ancient Egyptian limestone sculptures.

To assure the future conservation of similar stone artifacts, a close environmental control of relative humidity and temperature within storage areas is recommended. Salt extraction using water, often a routine procedure, should be discouraged because this treatment promotes damage caused by clay swelling. Conventional treatments using hydrophobic coatings (wax, organic polymers, etc.) have not been found to be effective in most cases. It is proposed that clay stabilization studies involving replacement of existing sodium ions by calcium or magnesium, or by addition of surfactants, be initiated to control swelling. Nevertheless, prior to any conservation intervention, an accurate mineralogical, petrographical, and compositional study of the stone, as well as a complete laboratory testing to identify and evaluate the action of clays, should be developed.

ACKNOWLEDGEMENTS

This work was performed at the Getty Conservation Institute. Carlos Rodriguez-Navarro was supported by a postdoctoral fellowship from the Ministerio de Educaci�n y Ciencia Spain (no. PF94 24232705). Financial support was also provided by the Direccion General de Investigaci�n Cientifica y Tecuica (project no. PB 93/1090). We are indebted to George S. Wheeler for his critical review of the manuscript; Madeleine W. Fang of the Phoebe Hearst Museum of Anthropology, Berkeley, California, who supplied the Naga el-Deir stela samples; Eric Doehne for his help in performing the SEM analyses; and Michael Schilling, who performed the thermomechanical analysis.