1 INTRODUCTION

Thin-section petrography is polarized light microscopy of rocks and other mineral-containing materials, using samples ground to a standard thickness of 30 μm. The standard thickness gives known colors between crossed polarizers, facilitating comparison of different samples and the use of reference tables of mineral optics for identifying unknowns. Use of a specific, known thickness for all standard petrographic samples differs from the cross sections often used in conservation.

Geologists use thin-section petrography to describe and classify rocks, soils, and sand. Archaeologists and conservation scientists use it to study many inorganic materials used in the production of cultural objects. Purposes of such analyses in cultural object studies include making correct material identifications, grouping similar objects, comparing materials-based groupings to art historical groupings, identifying the geological origin of the object or some of its components, and studying manufacturing technology. For some art materials, structural and mineralogical changes on weathered surfaces in comparison to unaltered interior sections of a sample may provide information concerning the authenticity of a piece. Thin sections have also been used to study the deterioration of inorganic art and architectural materials and to check the effects of conservation treatments on those materials.

One advantage of thin-section petrography is that the necessary equipment (a polarizing microscope) is relatively inexpensive, making the technique potentially available to almost any laboratory as a routine method of analysis. Although a significant investment in training and experience is required to use this technique to full advantage, thin-section analysis is still a very efficient way to obtain crucial information about many inorganic materials. The cost of purchasing and maintaining a polarizing microscope is much less than for other types of equipment used to study inorganic objects (such as a scanning electron microscope, electron beam microprobe, x-ray diffractometer, or elemental analysis instrumentation). All these alternatives, including thin-section petrography, require extensive training and experience in the materials being analyzed and in the theory underlying the instrument.

Although thin-section petrography was developed many decades ago, conservation science still does not utilize it to its full potential. Thin sections are often the most useful starting point for a study of inorganic materials, even if they may sometimes need to be supplemented by other approaches. They provide information not available by other analytical techniques.

To make a thin section, the sample material is mounted on a glass slide using an epoxy resin with a refractive index (1.54–1.55) essentially the same as that of quartz, a ubiquitous mineral found in almost all geological materils. Thin sections are then ground to a uniform thickness of 30 μm. Very porous or friable materials may be impregnated, and techniques for mounting loose grains have also been developed. A thin section may be protected by a coverslip, particularly important for friable specimens and other materials that might easily become damaged. However, if thin sections are left uncovered they can also be used for a variety of additional analytical techniques such as microhardness and microchemical tests, scanning electron microscopy with elemental analysis capabilities, or microprobe analysis (Hutchison 1974; Loretto 1984; Goodhew 1988; Reed 1993). Staining of uncovered sections can aid the identification of minerals that are otherwise difficult to differentiate in thin section. Examples of the use of stains include distinguishing between alkali and plagioclase feldspars and between calcite and dolomite (Allman and Lawrence 1972; Hutchison 1974).

When examined under a polarizing microscope, minerals in mounted thin sections of 30 μm can be identified through a variety of optical properties, usually at magnifications ranging from 16� to 400�. For transmitted plane polarized light, the properties include transparency versus opaqueness, color, pleochroism, refractive index, relief, morphology, and cleavage. Between crossed polarizers, important properties include isotropism versus anisotropism, birefringence, extinction angle, and the presence or absence of other features such as zoning, twinning, undulous extinction, and anomalous polarization colors (Phillips 1971; Kerr 1977; MacKenzie and Guilford 1981; MacKenzie et al. 1982; Williams et al. 1982; Adams et al. 1984; Yardley 1990).

In addition to the identification of specific minerals, thin-section petrography also involves the study of mineral and rock textures, coarseness, and the relative or quantitative percentage of various constitutents. The data collected from thin-section studies can easily be used for statistical analysis by using quantitative measurements, converting binary and categorical information to arithmetic data, or using statistical methods designed for binary and categorical data (Chayes 1956; Keith and Cooper 1974; Stoltman 1989; Agresti 1990; Matthew et al. 1991; Reedy 1991; Reedy and Reedy 1994). One can also decrease the subjective aspects of thin-section analysis by coding sections for “blind” analysis, randomizing the order of analysis, or using semiautomatic image analyzer systems (Middleton et al. 1985; Garrett 1986; Stoltman 1989; Middleton et al. 1991).

To introduce the potential of thin-section petrography for studies of cultural materials, this paper includes brief discussions of a wide variety of relevant materials and research problems. These examples are divided into sections on stone, ceramics, glass and glazes, and other miscellaneous materials.