Analysis of a Purple Material Found at Akrotiri

About 50 mg of this very light and fine powdered material was analysed non-destructively by XRF using a Cd-1O9 radioactive source and by Proton Induced X-ray Emission (PIXE). The XRF spectrum is shown in Fig. 1 and the elemental composition derived from it is given in Table 1. The large amount of Br indicates a material originating from the sea but the absence of Cl excluded the possibility of sea-water salts and pointed to the presence of an organic dye.

In particular, because of the purple colour the presence of purpura was suspected. Friedlander (1901) characterized the major constituent of the ancient dye Tyrian purple, derived from the mollusc Murex brandaris as the known compound 6,6' - dibromoindigotin, C₁₆H₈N₂O₂Br₂ (Fig. 2). Later the same compound was identified in the purple dyes from Murex trunculus, Purpura aperta, Purpura lapillus (Friedlander 1922) and also from Purpura brandti and Purpura haemastoma and some other molluscs (Baker and Sutherland 1969).

XRD analysis of the bulk sample with a Philips diffractometer using a Co-Ka line allowed the identification of the following phases: aragonite (abundant); calcite (abundant); feldspars (subsidiary); quartz (minor).

Aragonite is the form in which calcium carbonate crystallizes in living seashells, transforming into calcite with a rate which depends on the conditions of burial.

In an attempt to identify a possible organic compound in the sample, a small quantity of the material treated with HCl was dissolved in CHCL₃. The phase of the solvent coloured purple indicating the organic nature of the dye. The examination of this organic compound in an infra-red spectrometer was unsuccessful as no distinct lines of absorption could be detected. If the presence of Br in the sample is due to purpura then the amount of this dye in the bulk sample would have been about 1.5%. Thus the absolute amount of this organic compound in the specimen prepared for IR would be below the detection limit for a positive identification. Unfortunately the attempt to isolate the organic component was unrepeatable due to lack of extra material to be wasted. Although the presence of purpura is not definitely identified, the organic nature of the dye was strongly indicated by its dilution in the CHCl₃.

In spite of this last difficulty, the above-mentioned indications imposed the suggestion that this purple material found at Akrotiri was most probably produced by pulverizing purpura or murex seashells. In order to compare with the ancient material a recent Murex trunculus shell was pulverized and analysed by XRF and XRD. The powder was buff in colour and the Br content was found to be very low (5 ppm) compared to the ancient purple material (1.5%). The main mineral phase present was aragonite, as expected.

Thus it is reasonable to assume that the purple material, due to its high Br content and its intense purple colour, could not be simply a powder of a pulverized shell, but it should probably contain the dried gland of the mollusc in which the accumulation of purpura is high enough to exhibit dye properties. Summarizing, the evidence accumulated so far (i.e. the large amount of Br compared with the usual amount of the element in natural earthy materials, and the large amount of aragonite) indicates the marine origin of the purple material, and more specifically its possible association with the purpura dye derived from Murex seashells. This is further reinforced by the fact that a lot of seashells and especially those of the Murex species have been found fragmented and scattered all over the excavated area. The presence of these molluscs in relatively large quantities suggests a probable local production of the above purple dye (Karali-Yannacopoulou 1988). However, such production requires specific open-air installations, which have not yet been discovered at Akrotiri.

The use of shells around the Mediterranean in purple-dye production is well established. The evidence for this industry is in a tablet of the 13th century BC written in Linear B, found at Knossos (tablet KN X976). It specifically refers to purple-coloured materials.

The form of the material, i.e. a small amount of dye together with a lot of carbonates, excludes its use for textile dyeing. The fact that this pigment has not as yet been identified on wall-paintings points to a use other than for wall-paintings, e.g. a cosmetic use. Such a suggestion is reinforced by the existence of small boxes made out of Pecten jacobaeus Linne seashell at Akrotiri, which were probably used as containers of cosmetics.

The work will continue with the treatment and analysis of living Murex seashells. The use of more sensitive techniques will be necessary in a further effort positively to identify the compound.

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

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