Archaeometry 36 (2)

The recently developed laser microprobe 40argon-39argon technique has been used to give a geological date for a rhyolitic tuff stone axe fragment. The method requires only milligramme sized samples and gives dates of sufficient accuracy to aid in provenancing artefacts to sources, as well as information on the heating history of samples. The axe sample is of Lower Carboniferous date (341±5 Ma) and this limits possible sources to outcrops within the Scottish Midland Valley and small altered exposures in Dartmoor. X-ray fluorescence analysis of the axe suggests the Scottish Midland Valley as the more likely source. The laser argon analysis also shows that the implement had not been heated in antiquity.

Sensitivity, precision and accuracy of quantitative microanalysis (EDS & WDS) of major, minor and trace elements present in ancient glasses is compared. Although there are certain problems with EDS due to peak overlaps and background noise, the precision and minimum detection limits of EDS for most of the important elements in ancient glasses were found to compare favourably with those of WDS. Both EDS and WDS, when quantified by comparison to glass standards, can produce reliable results which are adequate for most technological questions.

A statistical procedure for grouping pottery in provenance studies by chemical data is presented. It is based on the Mahalanobis filter method and P2-statistics, and can be used for both establishing groups and assigning single sherds to already known groups, thus replacing principal components analysis or cluster analysis and avoiding their problems in grouping pottery. The new method is able to consider correlations, uncertainties of measurement and constant shifts of the data in case of dilution. In particular, considering dilution effects results in both a better assignment to and separation of reference groups and is also equivalent to the compositional data approach, if log-transformed data are used. Other distortions of data (eg mixing of clays) can also be considered.

All surface traces of the vallum have been obliterated over a 500m long stretch past the fort of Vindobala at Rudchester, Northumberland. To the west it had previously been delineated by running shear-wave seismic refraction lines to obtain depth profiles of rockhead. More seismic refraction lines have been run, using both shear waves and compressional waves, to the south-west and south, thereby locating the complete course. The seismic refraction technique is successful at this site because the glacial drift cover is thin, so the vallum ditch was cut down into bedrock. During the second phase of seismic work, compressional waves gave better results than shear waves. Previously, test profiles using compressional waves had failed to locate the vallum ditch west of the fort where the surface elevation is lower; this is interpreted as being due to a shallow water table within the drift, which would have eliminated the compressional-wave velocity anomaly. Although the vallum does not pre-date the fort at Vindobala, its asymmetrical course around the fort suggests that it may already have been under construction to the west when the decision to build the fort was taken.