J Archaeol Sci 33 (8)

Forty samples of Roman colourless glass tableware from Binchester, dating from the first to mid-third centuries AD, were analysed using ICP spectrometry and parallels were sought with similar studies of Roman glass from Colchester and Lincoln. Some samples from the Binchester, Colchester and Lincoln (BCL) groups were re-analysed using energy dispersive spectrometry (SEM--EDS) and the results were used to compensate for the differences in reproducibility between the ICP data sets, so that these could be directly compared. The majority of the glass from all three sites was similar but some distinct compositional characteristics were identified that were specific to certain types of ware. There were differences in the concentration of lead in the samples, which appear to be to some extent chronological. The compositional data for the glass from Binchester, Colchester and Lincoln were also compared to literature data for various types of natron glass produced in the first millennium AD. Includes

The prevalence of digital image analysis in studies spanning a wide range of fields makes set-up design decidedly important for repeatability and testing between studies, especially for those interested in the analysis of microscopic properties. This note outlines some guidelines for setting up an image analysis protocol, and provides a case study based on the digital analysis of dental cement. The authors' method allowed them to get data through the conversion of optical luminance into numerical values, and to accurately measure image features through numerical analysis.

Longitudinal and transverse carbon isotope profiles were performed on tooth dentine from five steers (Bos taurus) initially fed C3 and subsequently C4-dominant food. Comparison of different protocols for bioapatite extraction revealed that the use of NaOCl considerably reduced the amplitude of variation of 13C within a tooth. Increasing contribution of C4 food to the carbon isotope composition of bioapatite was found from the tip of the tooth crown to the neck and from the enamel--dentine junction toward the pulp cavity. These findings confirm that the model of dentine growth as a succession of stacked cones applies to bovines. Temporal resolution is estimated to be four months in transverse profiles, significantly better than in longitudinal dentine profiles (eight--nine months) or even in profiles derived from enamel of the same individual (six--seven months). Temporal resolution could be improved by a factor of two by selecting a different sampling zone or refining the sampling protocol. This sampling strategy could also be applied to dentine collagen and has important ecological and archaeological implications including determination of the season of weaning, or the reconstruction of mobility strategies.