THE ORIGIN,DESTRUCTION AND RESTORATION OF COLOUR IN EGYPTIAN TRAVERTINE*

Travertine was one of ancient Egypt's most popular ornamental stones. Upon prolonged exposure to sunlight, the rock's translucent brownish calcite is bleached to nearly white. Ultraviolet and gamma irradiation experiments were undertaken to investigate the source of the brownish colour and the process by which it fades. It was found that the coloration results from the activation of colour centres by natural radioactivity within the rock, and that these colour centres are deactivated by the ultraviolet component of sunlight. The results also demonstrate that the original brownish colour of sun‐bleached, travertine objects can be restored by artificial gamma irradiation.     

NON‐DESTRUCTIVE PORTABLE GAMMA RAY SPECTROMETRY USED IN PROVENANCING ROMAN GRANITOID COLUMNS FROM LEPTIS MAGNA,NORTH AFRICA*

Portable gamma ray spectrometry (PGRS) provides a non destructive means to analyst quantitatively large artefacts, such as building stones, for the radioelements K. U and Th. Nine Raman granitoid columns at the Leptis Magna Ruins in Windsor Great Park, London, were measured in situ by PGRS. Corrections for the environmental background contribution to the gamma ray flux measured, and for the shape and size of the columns, are described Comparison of the PGRS data with a radioelement data base for Roman granite sources indicates that most of the columns originated in the Troad area of Turkey. Two columns could not be unambiguously provenanced using PGRS alone because there is insufficient difference between radioelement concentrations in certain sources. However, non‐destructive measurements of magnetic susceptibility, used in conjunction with PGRS data, suggest that these two columns originated in the Kozak Daǧ, also in Turkey.     

THE EFFICIENT COMPUTATION OF UNBIASED MAHALANOBIS DISTANCES AND THEIR INTERPRETATION IN ARCHAEOMETRY*

This paper argues for cross-validation to reduce bias in estimating Mahalanobis distances of individuals to groups, a particular problem with small sample sizes. Formulae for the efficient computation of cross-validated distances, which avoid excessive matrix inversion, are given. The use of Mahalanobis distances as measures of consistency, or as indications of outlying values, is described. The little-known technique of gamma plotting is outlined and discussed as an aid to interpreting distances in these terms. Both cross-validation and gamma plotting are illustrated in an example on marble composition.     

OBSIDIAN IN THE MEDITERRANEAN AND THE NEAR EAST: A PROVENANCING SUCCESS STORY

Obsidian provenancing studies comprise one of the most productive and successful research programmes of archaeological science. Obsidian characterization has been successful because workable obsidian is homogeneous on a small scale, analysable by a large number of methods, and is restricted to a small number of mainly readily distinguishable geological sources. Analytical, dating, source, and trade studies within the western Mediterranean, central and eastern Europe, the Aegean, and Anatolia and the Near East during the last 30 years or so are reviewed. Research has shown that distributions are mainly separate in the four regions examined, and that obsidian was traded up to 900km in the prehistoric period. Publications on obsidian in the areas under review reached a peak of frequency in the later 1970s and 1980s, but have now decreased in number. This may reflect changing fashions in archaeometric studies, and a current lack of routine application of the provenancing methods developed.     

The chronology of the Neolithic sequence at Dikili Tash,Macedonia, Greece: TL dating of domestic ovens*

The chronological sequence of a major reference site for Neolithic periods in southeastern Europe and the Aegean, Dikili Tash (Macedonia, Greece), has been reassessed by TL. Baked adobe material samples extracted from domestic ovens, directly associated with specific occupational levels, have been studied. The fine‐grain technique has been used for equivalent dose determination. K, U and Th contents and the equilibrium state of U‐series have been measured by low‐background gamma spectroscopy. Since the archaeological levels were already partly excavated, the environmental dose‐rate was evaluated by a reconstruction technique combining laboratory analysis and on‐site measurements. The TL results (weighted average of multiple age determination, oven 400, middle of the Dikili Tash I (DT‐I) phase, 5500 ± 320 bc ; oven 600, end of DT‐I, 4920 ± 310 bc ; house 3, DT‐II, 4260 ± 280 bc ; house 4, DT‐II, 4510 ± 410 bc ) are in general agreement with the timeframe deduced from previous radiocarbon dates. This convergence strengthens the settlement's chronology, established now by independent dating methods.     

DATING ARCHAEOMETALLURGICAL SLAGS USING THERMOLUMINESCENCE*

Thermoluminescence has a great potential for dating archaeometallurgical slags, because the smelting process leads to a well‐defined resetting of the ‘luminescence clock’. However, the complex compositions of slags have unpleasant consequences for TL measurements if the bulk slag substance is used. To overcome this problem, quartz has been separated out of slag matrices by chemical and physical procedures. The TL measurements were carried out on this defined mineral phase. This method was tested with seven slag samples from different locations and ages. In most cases, the TL ages determined show good agreement with reference data.