Review A review of factors affecting the composition of early Egyptian glasses and faience: alkali and alkali earth oxides Th. Rehren * University College London, Institute of Archaeology, 31-34 Gordon Square, London WC1H 0PY, United Kingdom Received 6 July 2007; received in revised form 26 September 2007; accepted 29 September 2007 Abstract The composition of Late Bronze Age Egyptian faience and glass is thought to reflect the composition of the raw material used in their pro- duction, particularly plant ash. This paper argues that there are strong and systematic shifts between the plant ash composition and the resulting glass, as a function of inherent technical processes during both the wet preparation and subsequent drying of faience glazes, and the smelting of glass from its raw materials. These factors lead on the one hand to significant differences in glaze composition from identical raw materials due to different glazing techniques being used, and on the other hand to a homogenisation of glass compositions which may obscure more subtle differences in initial raw material composition. The paper aims to explore the various factors at work, briefly summarizing recent publications and current thought on the subject, in the hope to raise awareness of the issues involved, and to stimulate further research. Ó 2007 Elsevier Ltd. All rights reserved. Keywords: Glass; Salt; Egyptian faience; Element partitioning; Differential efflorescence; Analysis 1. Introduction The identification of plant ash as the most likely flux in New Kingdom/Late Bronze Age Egyptian glassmaking by Brill (1970: 111) was a major breakthrough in our understand- ing of the earliest glassmaking tradition. Since then, efforts continue to identify the exact plant species being used (e.g. Santopadre and Verita, 2000; Tite et al., 2006). A key ap- proach is the characterisation of the plant ash component of the vitreous material, using the ratios of soda to potash, lime to magnesia and total alkali oxides to total alkali earth oxides as main descriptors. These are then compared to the same ra- tios obtained from analyses of actual plant ash samples, both ethnographic and experimentally produced (e.g. Lilyquist and Brill, 1993; Tite et al., 2006). Specific questions addressed by this approach include whether different types of plant ash were being used for the production of faience and glass, as seems to be indicated by the systematically different ratios of oxides analysed in the two materials (see e.g. Tite and Shortland, 2003; Tite et al., 2006: 1291). This question is of fundamental importance both for our understanding of the ap- parent sudden origin of glassmaking at around 1600 and 1550 BC in Mesopotamia and Egypt, respectively, and for the orga- nisation of the vitreous industries afterwards. This paper takes the recent increase in quality and quantity of analyses of Egyptian faience and glass (e.g. Shortland and Eremin, 2006; Shortland et al., 2007; Tite et al., 2007) as a starting point to explore whether the oxide ratios as analysed can be taken to represent the original plant ash composition, or whether the inevitable cold and hot processing of raw mate- rials may result in minor and trace element partitioning, which in turn will severely affect the relevant ratios. It is argued that a number of process-specific factors have a strong and system- atic influence on all relevant oxide ratios, and that much of the apparent differences between faience and glass compositions are likely due to the different nature of their production rather than the use of different plant ashes. The same argument can be made for other early vitreous materials, such as those * Tel.: þ44 20 7679 4757; fax: þ44 20 7383 2572. E-mail address: th.rehren@ucl.ac.uk 0305-4403/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.jas.2007.09.005 Journal of Archaeological Science 35 (2008) 1345e1354 http://www.elsevier.com/locate/jas