Journal of A rchaeological S cience (1995) 22, 239–250 The Role of the Environment in Uranium Uptake by Buried Bone Andrew R. Millard and Robert E. M. Hedges Research Laboratory for Archaeology and the History of Art, University of Oxford, 6 Keble Road, Oxford OX1 3QJ, U.K. (Received 21 April 1994, revised manuscript accepted 27 September 1994) Previous studies by a number of workers have shown that uranium is inhomogeneously distributed in excavated bones. It has been suggested that the higher concentrations of uranium found towards the outside of some bones may indicate that the uranium has been taken up by a diffusion process. This paper briefly outlines a quantitative model of uranium uptake by diffusion and chemical reaction, and the measurement of uranium distributions in a number of samples. The relationship of these distributions to environmental factors and to other diagenetic changes is explored, and comparisons made between them and the model. Finally the implications of these results for the uranium series dating of bones are considered, and the way forward in modelling uranium uptake suggested. Keywords: ARCHAEOLOGICAL BONE DIAGENESIS, URANIUM SERIES DATING, HISTOLOGY, HYDROLOGY, FISSION TRACK MAPPING, DELAYED NEUTRON COUNTING. Introduction T he mechanism of uranium uptake by buried bone has direct implications for dating bone by the uranium-series method, but it also gives general insight into diagenetic processes because it is a clear example of change resulting from interaction of the bone with its environment. In the past a number of attempts have been made to date bone by the uranium- series method, starting with Cherdyntsev et al. (1963, 1965). For the purposes of this technique, bone is problematic because it contravenes the basic assump- tion of uranium-series dating; namely that of a closed system with uranium present at the time of deposition. In living bone there is no uranium and therefore that which is found in excavated bone was taken up after burial. Various quantitative models of uranium uptake by bone have been proposed, including the Szabo– Rosholt model (Szabo & Rosholt, 1969), the early uptake and linear uptake models used in Electron Spin Resonance (ESR) dating (Grün & Schwarcz, 1987) and the model of Chen & Yuan (1988). There are also qualitative models suggesting early uptake through reduction of uranyl ions by organic matter (e.g. Szabo, 1980; Rae & Ivanovich, 1987). However, a survey of published uranium-series dates on bone, shows that only 60% agree with comparative dates at the 2 range (Millard, 1993). This level of agreement suggests that there is some hope for using the method, but that it cannot currently be reliably applied. Various workers have examined uranium distri- butions in bone, and a proportion of the results are suggestive of a ‘‘diffusion’’ process for uranium uptake (e.g. Badone & Farquhar, 1982). We have developed a diffusion–adsorption model of uranium uptake by bone, which, if proved realistic, would allow better estimates of dates from uranium series measurements (Millard & Hedges, in prep.). This model also predicts the spatial distribution of uranium in bone (and tooth), in terms only of age, site hydrology and groundwater uranium concentration. Here we report a first attempt to compare the predictions of the model with measure- ments on uranium distributions on a set of bone samples from a variety of sites. This is intended to be a test of our proposed model for uranium uptake. It shows some of the strengths and limitations of the model, and indicates the extent of the dataset required for a more decisive testing. Uranium Uptake Model Details of our uranium uptake model have and will appear elsewhere (Millard & Hedges, 1992, in prep.), so it will only be very briefly outlined here. Uranium uptake by buried bone is viewed as a process of diffusion of uranyl ions from soil/groundwater into water contained in the pore structure of the bone, followed by chemisorption of the uranyl onto the surface of the hydroxyapatite of the bone mineral. This process is su fficient to account for the magnitude and time scale of uranium uptake for bone reported in the literature. We are well aware that other processes may be taking place, such as reduction of uranyl to uranium (IV), and incorporation of uranium into the crystal 239 0305–4403/95/020239+ 12 $08.00/0  1995 Academic Press Limited