Analysis of Pigments from Thera

A large number of fresco fragments from Thera have been studied. All the samples were taken from the same destruction level of the Akrotiri excavations and can therefore be dated to the period between 1600 B.C. and 1500 B.C.

The study of the pigments which were analyzed by a combination of X-Ray fluorescence, X-Ray diffraction and mineralogical techniques. The X-Ray analysis revealed three types of blue pigments used in Thera. Mainly Egyptian Blue, glaucophane and a mixture of the two. Several combinations of pigments giving rise to subtle colour variations were observed. Black pigments were rather puzzling in that some of them are carbon as expected and some of them are manganese compounds. The rest of the pigments red, pink, yellow, orange and brown are very similar to those previously examined from Knossos.

• INTRODUCTION

The need for scientific study of frescoes from the Minoan and Mycenaean periods was first recognized by Durm in 1907 (Shaw 1973) and then by Evans and his scientific assistant, Noel Heaton. The latter analyzed in 1910 and 1911 numerous plaster and pigment specimens of frescoes from Knossos and he went on to discuss the nature of the painting technique used by the Minoans (Heaton 1911). Since then several analytical studies have been made on the pigments and plasters from many sites which have yielded important frescoes. Until recently all the analyses were done by standard wet chemical and optical microscopy methods. In the case of the plasters most of the analysis included determinations of the percentage composition of the elements associated with the clay impurity in the plasters. These analyses have demonstrated the apparent radical change in plaster composition between the early Minoan 2600 - 2200 B.C. samples and the late Minoan and Mycenaean 1500 - 1400 B.C. samples as Heaton first noted (1911, 698). On the other hand, the analysis of the pigments of painted samples has been essential in identifying the materials used for different colours.

Thus, for example, the blue colour on a sarcophagus from the Aghia Triada excavation was identified as lapis lazuli (Levi 1956). Heaton also discovered malachite (copper carbonate) as the source of green at Tyrins (Heaton 1912).

• EXPERIMENTAL TECHNIQUES

In the present work we have undertaken the analysis by various experimental techniques of plasters and various colours from Thera which were obtained at our laboratory. Several experimental techniques were used giving complementary information as e.g. x-ray fluorescence (XRF), x-ray diffraction and mineralogical examination. The analysis of fresco samples by x-ray fluorescence is non-destructive but other techniques required some scratching of small amounts of material (up to 10 mg) from the pigment surface and the body of the plasters.

All samples were examined under a stereomagnifier (x 10) before commencement of any analytical work in order to determine the nature of the colour and the number of paint layers, the depths of penetration of the pigment and the presence of slip straw remain in the plaster as string and brush marks on the surface. X-ray fluorescence was used to identify the elements present in the pigments and their relative proportions were determined as major, minor or trace constituents after subtraction of the background of intensity. The detection limit was estimated to 0.01 %. Elements with an atomic number less than 19 (such as carbon, magnesium and aluminum) could not be detected. X-ray diffraction was used to identify the crystalline compounds in the pigments of selected samples. The sample was collected by scratching the surface with a razor blade. In the cases in which there was more than one pigment layer, or the nature of the layer altered with thickness, each layer or "phase" was scratched off successively. All the samples were examined mineralogically in order to supplement the information gained from the Debey-Scherrer technique.

A polarizing light Ortholux microscope was used for that purpose. A series of liquids having a range of refractive indices was used to determine the identity of each individual crystalline phase in the samples. Also tests were made to determine the solubility of the pigment layer of all samples in various solvents: hot and cold water, acetone and dilute sodium hydroxide.

• SAMPLES

A relatively large number of pigment samples from Thera have been examined. The painted plaster fragments vary in size from 1 - 2.5 cm² while the plaster which is one layer in all varies in thickness from 0.4 - 0.8 cm.

Most of the samples were kindly provided by the Centre of Conservation of Antiquities in Athens.

All the samples were taken from the same distribution level of the Akrotiri excavations and can therefore be dated to the period between 1600 B.C. and 1500 B.C. Representative results from the analyses of pigments and plasters are given in Table 1.

The most important aspects of these investigations and their implications for archaeological problems can be summarized as follows:

Results:

The results of x-ray analysis of the Thera pigments indicate that these pigments are very similar to those previously examined from Knossos (Profi et al. 1976) especially when one considers the blue pigments used.

The XRF data show that all samples can be divided into three groups:

(a)     Group I. The prominent element is copper. All other elements present are in trace quantities with the exception of iron which amounts to a few percent relative to copper.

(b)   Group II. The prominent element is iron. This group is characterized by a complete absence of copper.

(c)   Group III. Iron is the prominent element with varying amounts of copper.

From all the experimental evidence one may say that the pigments belonging to group I are pure Egyptian Blue, while pigments from group II are identified as pure glaucophane. Samples from group III contain Egyptian Blue and some glaucophane. The black pigments are rather puzzling in that some of them are carbon as expected, but some are Mn-compounds. Mn-minerals are in general not very easy to identify from their diffraction patterns or optical images, still all the indications we have been able to obtain from the experimental methods employed in the present study suggest that these black pigments are most probably Mn-oxides.

The blue pigments are again of great interest - at Knossos, glaucophane has been identified as a blue pigment in many samples. There are however also some samples exhibiting pure Egyptian Blue. What is impressive is the mixture of the two blue pigments. This may be due either merely to colour shade restrictions, or to the fact that Egyptian Blue was a pigment imported from Egypt and thus expensive and available only in limited quantities. Glaucophane however leaves a lot to be desired as a pigment when compared to Egyptian Blue, hence its absence from later wall painting (e.g. Mycenae) (Profi et al. 1974) when commercial relations with the Nile Valley became closer. The presence of tin was noticed once more in all Egyptian Blue samples, in support to our previous suggestion (Profi et al. 1974; 1976) associating the source of Cu in Egyptian Blue with bronze. The rest of the pigments red, pink, yellow, orange and brown are natural earth ochres, i.e. hematite, and/or iron hydroxides, either crystalline (goethite) or amorphous (limonite). What should be noted here is that the iron hydroxides can vary in colour from yellow to red and brown, and that all these varieties occur in mixtures in the pigment sample in varying quantities, frequently surrounding or attached to the clay mineral particles.

-------------------------------------

-------------------------------