Organic Residues in Egyptian Amphorae

Department of Archaeological Sciences

New approaches to the study of commodities transported to Egypt in Late Bronze age Canaanite amphorae

The project is based at the University of Bradford in conjunction with Dr Carl Heron. Funding is from the Natural Environment Research Council (NERC)

The impetus for this project derives from the research interests of Margaret Serpico (University College, London). A highly successful pilot project was carried out in conjunction with Raymond White, this provided the essential foundation for the development of the research programme outlined here. Janine Bourriau (McDonald Institute for Archaeological Research, Cambridge) has been researching the geographical provenance of Canaanite jars and is also involved with this project.

Specific objectives

The outcome will provide a valuable contribution to the field of trace organic analysis applied to archaeology. Specifically addressed will be the need to establish a more efficient protocol for sample analysis and to clarify compositional variation in ancient organic matter associated with archaeological ceramics. These aspects will be explored through a detailed study of the contents of Late Bronze Age (c. 1500-1050 BC) Canaanite amphorae imported to Egypt. These jars, believed to have been manufactured in Syria/Palestine, were widely exported throughout the Mediterranean and are important indicators of the trade in bulk commodities. In addition, some of the jars transported to Egypt were inscribed, probably on arrival, with the names of the products which they contained. Translation of these texts provide an exceptionally clear indication of the types of commodities in use (higher plant resins, oils, honey). By integrating study of material from Egypt and Israel, this project will help establish links between the sources of the jars, the commodities, the names of those products and ultimately the transport of organic substances in the Mediterranean world.

The proposal has two key scientific objectives, each of which encompasses a number of research themes.

i) To establish a new analytical approach to the investigation and identification of visible and absorbed organic matter associated with archaeological ceramics The aim is to develop and apply a sequential methodology appropriate for the extraction and characterisation of organic matter associated with ancient pottery vessels. In particular, we wish to distinguish the principal contents thought to have been transported in Canaanite amphorae, although the methodology is sufficiently flexible to screen for possible alternatives (e.g. bitumen).
ii) To explore the implications of compositional variation in complex mixtures extracted from archaeological ceramics using solvent extraction and pyrolysis techniques.

We intend to characterize specific compositional patterns in the residues and to correlate these data with the following; (i) the contribution of the burial environment to the composition of organic matter deposited on the surface of, or absorbed into, Canaanite jar fragments in two contrasting burial environments and; (ii) evaluation of vessel re-use by multiple comparisons of surface residue with absorbed organic matter; (iii) variation in resin composition caused by heating; (iv) identification of honey residues and the botanical sources of oil and resin.

Background to the project

Organic chemical analysis has been increasingly recognised as a valuable source of information for the food products and raw commodities used in antiquity. As visible residues rarely survive, previous studies have focused mostly on the investigation of absorbed residues by solvent extraction of pottery sherds. Although considerable progress has been made in this field, certain fundamental issues have remained unresolved. This project, originating from a pilot study of samples from ancient Egypt including a number of Canaanite jar contents, is designed to address some of these problems.

Particularly important is the unprecedented opportunity to explore the extent of compositional variation through study of comparative material; from two archaeological excavations in Egypt (Tell el-Amarna and Memphis) and from a variety of sites in Israel. In total, some 150-200 pottery sherds will be available for study, comprising 75-100 sherds from Tell el-Amarna, 50-75 from Israel and 25-50 from Memphis. While organic preservation at the dry site of Tell el-Amarna is excellent, offering the potential for comparison of visible and absorbed organic matter, no visible residues remain at the wet/dry site of Memphis, or on the material from Israel. This diversity will allow an opportunity to compare visible and absorbed residues and also to assess the effect of differing burial environments on composition.

Methodology

Conventionally, extraction of soluble lipids is achieved by solvent (e.g., chloroform/methanol, 2/1 v/v) washing of powdered ceramic fragments or visible surface deposits followed by analysis using gas chromatography (GC) and combined gas chromatography/mass spectrometry (GC/MS). Relatively large (e.g., 1-3 grams of ceramic) samples and c. 20 millilitres of solvent are normally required and the procedure (involving ultrasonication, centrifugation, rotary evaporation and so on) is time consuming and can introduce contaminant molecules. The large sample size is also prohibitive for some sherds.

Development of a single step extraction/sample work-up procedure employing the reagent TMTFTH (trimethyl(a, a, a-trifluoro-m-tolyl)ammonium hydroxide), which has not previously been used for the analysis of archaeological material, is underway. Ceramics and visible deposits (where they survive) will be extracted using TMTFTH and analysed for the lipid, sugar, macromolecular and strongly adsorbed composition.

The information on the likely contents supplied from the inscriptions found on some of the jars forms an important control for the project. This is particularly useful for oil and honey as the characterization of these products in ancient samples has consistently proved problematic. Several sources of oil could have been known to the Egyptians, including the drying oil linseed, the semi-drying oils safflower, sesame and castor, and non-drying olive oil. Two other oils, balanos (Balanites aegyptiaca) and moringa (Moringa peregrina) were also available but little is known of their composition. The project will need to establish the composition of these oils by analysis of authentic samples. It is anticipated that unsaturated fatty acids will be depleted or lost in aged oil residues. Therefore relative abundance of these molecules cannot reliably be used to define oil source. However, classification of oil residues according to their drying properties will be investigated by the identification and quantification of free acids (mono- and dicarboxylic) molecules arising from scission of fatty acyl moieties at carbon-carbon double bond positions. Furthermore, characterization of high molecular weight lipid polymer arising from 'drying' reactions will also be investigated. Characterization of honey residues will rely on extracting soluble sugars from the ceramic fabric and implementing a suitable derivatisation procedure. Based on jar inscriptions and the results of the pilot study, it is known that both diterpenoid (pine, cedar and/or fir) and triterpenoid (Pistacia spp.) resins were stored in these jars, as well as the gum-resins frankincense (Boswellia spp.) and myrrh (Commiphora spp.). As most of the visible residues on the sherds for this study appear to be resins, this material will play a key role in the determination of compositional variation between visible and absorbed residues. The principal diterpenoid and triterpenoid composition of these resins is generally well understood for fresh specimens. However, any compositional differences arising from long-term chemical alteration in the burial environment will need to be assessed. Detection of components indicative of heating will be necessary to determine the presence of pitch. The potential of sesquiterpenoids as chemotaxonomic markers for pine, cedar and fir will also be evaluated. The pilot study demonstrated the survival of certain sesquiterpenoids in some visible residues. Variation in their presence between visible and absorbed residues requires further evaluation. Samples will also need to be screened for the possibility of added essential oils mixed with a lipid base or a resin. In addition to the scientific objectives, a number of related issues of particular interest to archaeologists, such as the extent of re-use of jars and the presence of mixtures or of scented and heated products, can also be explored through analysis. From the pilot study, it is clear that these aims can be met most successfully through an integrated scientific and archaeological approach. It is also intended that the results of this project will be integrated into a wider study of Canaanite jars and Mediterranean trade.

In summary, the demands of the research programme include the need to innovate sample extraction, to carry out successfully trace analysis and to interpret comprehensively pattern evident in complex mixtures. The significant challenge lies in using evidence for molecular alteration to understand use (e.g., evidence for heating) of these substances, re-use of jars (comparison of visible and absorbed organic matter from Amarna) and to elucidate post-depositional processes acting on organic matter from at least two contrasting sites. As well as providing key strategies for resolving archaeological problems, this project will establish a new analytical protocol for the small-scale extraction and analysis of many forms of remnant organic matter. The outcome should greatly advance the study of organic residues and will have widespread applicability in archaeological science.

Back to Research in Molecular Archaeology