THE APPLICATION OF THE PHYSICAL SCIENCES TO ARCHAEOLOGY I. INTRODUCTION Fred H. Stross, Research Associate, Archaeological Research Facility As a discipline, the History of Science has led an unobtrusive exist- ence for some years. By contrast, a more recent undertaking which we may call Science in History, has become a burgeoning field in the past few de- cades. The development of analytical instruments with vastly increased sensitivity, precision, and sample throughput, and of a strikingly powerful computer technology has become a great aid in establishing and testing elements of the fabric of history and, perhaps more significantly, of pre- history. To acknowledge these developments, a symposium entitled The Appli- cation of the Physical Sciences to Archaeology was organized by the writer for the Pacific Division Meeting of the American Association for the Advance- ment of Science,by the California Section of the American Chemical Society, with the co-sponsorship of the San Francisco Society of the Archaeological Institute of America. Its immediate purpose was to note and discuss the most recent advances in the field defined by the title of the symposium. It was held on June 23, 1970, in 1 Le Conte Hall on the Berkeley campus of the University of California, and consisted of a morning and an afternoon session and ended with the showing, at the Lowie Museum of Anthropology, of the color film "Neutron Activation of Pottery," which describes the work of the Lawrence Radiation Laboratory in this area. The symposium was opened by the Vice-Chancellor, Dr. Robert E. Connick, who recalled the role played by this University in the early work on the utilization of carbon 14 especially for dating carbon-containing compounds. The symposium was well attended by an audience predominantly consisting of those seriously interested in the topic. This interest was expressed in the nature and number of questions brought up during the dis- cussion periods. Regrettably, the first lecture scheduled had to be can- celed. It was entitled "The Role of Chemistry in Saving the Nubian Monu- m6nts, with Special Reference to the Abu Simbel Temples," and was to have been given by Dr. Zaki Iskander, Director General for Technical Affairs, Department of Antiquities, Egypt. At the critical time, Dr. Iskander was called to Venice to take part in a session of conservation experts to help develop plans to forestall the subsidence of that city, which has become a substantial threat during the past few years. All papers, with the exception of that of Dr. Iskander, are repro- duced here, in their full text, or, in the case of those of Drs. Curtis and _ l _ 2 Perlman, in the form of abstracts. In the meantime, abstracts of the papers have been published in Science, 171, 831-6, 1971. It is obviously impossible in a single day to provide a technical summarization of the status of the entire field. The following paragraphs are intended to supplement the proceedings at the meeting and at least to indicate some of the areas in which novel and promising work is being done, and what types of instruments are being used in this work. It may be worth stressing that the new technology is serving not only the needs of archaeological research per se, but also related needs, including the conservation of the artifacts that are so often the objectives of our studies. The instruments used in both areas, archaeological research, and conservation, often are the same. It may, for example, be of interest to the archaeologist and the art historian to have at his disposal methods for elemental analysis of very small samples of paint pigments - as in the case of the mysterious Maya Blue - and similarly to the restorer-conservator, who might want to make his restoration as inconspicuously as possible. Conse- quently we may find that an advanced museum laboratory allocates the time available for utilization of expensive analytical instruments to research and day-to-day problems in conservation in a ratio of 2:3, respectively. For convenience, one may still group the activities into inorganic-physical and organic-chemical, although the specialized areas of today do not fall into these classical categories. The most widely used operation in the first group probably is the elemental analysis of the material at hand. Many techniques are available, varying in first cost, cost per analysis, sensitivity, precision, expertise required, and complexity of ancillary facilities. One of the most versatile and promising instruments is the spark source mass spectrometer. With it, essentially all elements can be determined with very high sensitivity. The sample requirements are usually very small. The operation of the instrument can be taught to a skilled non-professional technician, but the supervision and interpretation especially of non-routine samples must be performed by a professional physicist-chemist with the necessary background. The spark- volatilized proportion of the sample that reaches the analyzing section of the instrument is very small, so that even a small sample must be thoroughly mixed to give representative results. Moreover, because of the extreme sen- sitivity of the method, fastidious housekeeping is necessary to avoid contam- ination, which easily causes erratic results. The cost of an installation ranges between $100,000 and $150,000. X-ray fluorescence spectrometry is a very popular technique for elemental analysis because of its relative simplicity and the short time required for analysis in some circumstances. Depending on the accessories available (and the corresponding investment), it is relatively easy to 3 determine, in many cases quite sensitively, most elements present, start- ing with aluminum (atomic number 13). To achieve good precision and to convert the output into absolute units when using different materials, however, careful work and a fair investment in analytical time are nec- essary. The sample size can vary only within relatively narrow limits (roughly of the size range of coins or buttons), except for specially de- signed equipment. The analyzing beam penetrates only on the order of microns, and the surface of the sample to be analyzed must therefore be represent- ative of what is desired in the analysis. That is to say, if the analysis is to reflect the composition of the bulk of the material, the surface must be representative of the bulk. If, on the other hand, one wishes to analyze the surface, as:distinguished from the bulk, of the sample (e.g. to deter- mine surface enrichment), x-ray fluorescence is the method of choice, followed, for instance, by neutron activation analysis to obtain bulk com- position. Equipment ranges from $20,000 up. Many variants of the technique are also available, such as the gamma ray spectrometry developed at the U.C. Lawrence Radiation Laboratory, the electron microprobe, and others. The optical emission spectrometer has been much used for elemental analysis in archaeology, for example by Cann and Renfrew in their studies of Mediterraean and Near Eastern obsidian. Getting accurate results is a laborious process. When using this technique, the sample is destroyed, while x-ray fluorescence and neutron activation in principle are non-destruct- ive, although special objectives may make destruction of the sample necessary. The electron scanning microscope also shows promising new develop- ments. This technique is particularly attractive in that it can be combined with non-dispersive x-ray spectrometry and then permits correlation of shape with composition, and presents a three-dimensional view of the object - but on a minute scale. A typical field of scan is on the order of 1000 Angstrom units, which may be both an advantage and a disadvantage. The price at present ranges upward of $50,000, but there are designs for production at a substantially reduced cost. Neutron activation analysis is another attractive technique for elemental analysis. Even more than in the case of the other techniques, we pay for what we get. Small instruments useful for the determination of only a few elements can be installed for as littl-eas $50,000. The large reactors that have been used for pioneering and most informative analyses of ancient glass, pottery, coins, and other metal objects, etc., have mostly been govern- ment installations, essentially accessible only to the personnel permanently associated with it. Occasionally outsiders may obtain analyses through personal acquaintance or on a commercial basis. The excellent results obtain- able by means of such installations are recorded in the work of Sayre, Perl- man and Asaro, Gordus, and others in the U.S., and of their European counter- parts. To take advantage of the capabilities of neutron activation analysis, 4 such standard but expensive instruments as multichannel analyzers, and sophis- ticated computer services should be available. The installations can deter- mine upwards of 30 (as high as 50) elements with varying, but often very high, sensitivities and precision. The reactors can handle very large samples, if required, and the analytical results are good averages of the concentration of each of the elements over the whole sample. Other techniques are also available, such as atomic spectroscopy, emission spectroscopy, and the classical wet chemistry. Under special con- ditions these may still be effective, and in the hands of skilled analysts can give excellent results. For laboratories planning new installations, however, the techniques indicated further above will probably be more attractive. Among the many problems for which elemental analysis has been used during the past years, the determination of minor- and trace-element patterns of composition of obsidian has been of significance. Assuming relative homo- geneity of individual lava flows and diversity between the flows, one can, by sensitive analysis of source samples and artifacts, in many cases establish correlation between an archaeological site at which the obsidian was found, and its volcanic source. Such studies have been made for Mediterranean regions and the Middle East, and for substantial sections of North and Central America. Thus it has been concluded that some desirable varieties of obsidian in Mexico and Central America were traded as far as about one thousand miles as early as in Middle-, and possibly Early Preclassic times. In the inorganic-physical category there are 4lso interesting develop- ments in the thermoluminescent method for establishing the date of the last firing of ceramic materials. Continued work on the method during the past few years, reported in this symposium, have made this method quite reliable, if the "archaeological" and the maximum "spurious" age are not too close to- gether. It appears now that recent work in the Conservation Center of the Los Angeles County Museum promises to make the determinations both simpler and more dependable. In the organic-chemical field there also has been considerable pro- gress. Asphalts, resins, varnishes, solvents, are only a few of the multi- tude of substances that have been subjected to such experimentation. Amber has been studied by infrared analysis quite extensively during the past few years. Gaschromatography, and pyrolysis-gas chromatography are used, but probably more in the service of conservation than in archaeological research. Such complex installations as gas chromatography - high resolution mass spectro- meters (price in the $100,000 range), which can not only separate organic mixtures into their component groups, but also identify these components, are justifiable only where there are continued demands for this kind of analysis. Less elaborate techniques, such as thin-layer, and paper chromatography, for which equipment is available in the few-hundred to few-thousand dollar range, are very serviceable in the hands of a skilled chromatographer, where rapid 5 mass production of analyses is not a prime requirement. One must bear in mind, of course, that gas chromatography is concerned with volatilizable sub- stances, while the liquid chromatographic methods indicated are used to analyze liquid and soluble materials. There is a large overlap, but not complete identity in these substances, and to a certain extent these tech- niques can be considered complementary rather than interchangeable. After the presentation of the papers, Dr. Albert B. Elsasser conduct- ed the participants of the symposium on a tour of an exhibition entitled Science and Archaeology, organized by himself and the writer, and shown at the R. H. Lowie Museum of Anthropology at the University of California at Berkeley from November 1969 on, which was held over especially for the sym- posium that is the subject of this publication. The topics of the individual exhibits were: Authentication of Antiquities; Cosmic Rays through the Pyramids; The Magnetometer; Rapid Chemical Analysis by X-Ray Fluorescence (including a irorking instrument operating at the exhibition); Element Analysis of Obsidian; Prehistoric Human Coprolites; Reconstruction of an Ancient Plastic Art; Obsidian Hydration Dating; Radiocarbon Dating; Potassium-Argon Dating; Thermoluminescent Measurement; Nuclear Fingerprinting of Ancient Pottery*). The exhibition, arranged with a particular eye toward stimulating interest in interdisciplinary studies, was much frequented by academic study groups, including many secondary school classes. After the Exhibition closed in Berkeley, it was shown on the university campuses at Riverside and Scripps College at Claremont, California. * Contributors were L. W. Alvarez, F. Asaro, C. R. Berger, H. R. Bowman, G. H. Curtis, A. B. Elsasser, R. D. Giauque, R. F. Heizer, W. F. Libby, H. F. Morrison, L. K. Napton, I. Perlman, R. J. Rodden, F. H. Stross, R. E. Taylor, from several campuses of the University of California; J. V. Noble, from the Metropolitan Museum of Art, New York, D. P. Stevenson and J. R. Weaver (retired), from Shell Development Company, Emeryville, California.