Christopher S. Peebles (Glenn A. Blaclc Laboratory of Archaeology, Indiana University, Bloomington)


As archaeologists know all too well, a proper excavation consumes time and money at prodigious rates; it also destroys everything in its path not recovered and recorded by the excavators. In order to minimize these real and destructive costs, archaeologists have adopted a variety of techniques that simultaneously preserve archaeological and financial resources. The most recent additions to these conservative techniques are various geophysical methods: magnetometry, electromagnetic methods, and even ground penetrating "radar." Singly, and especially when used together, these techniques can produce measurements from which one can produce remarkably accurate maps of sub-surface archaeological features.

Pioneering work in the application of geophysical methods to archaeological problems was undertaken by Glenn A. B1ack and Richard Johnston at the Angel Site between 1958 and 1963. Blade, with the collaboration of Judson Mead, an Indiana University geologist, first tried electrical resistivity in the summer of l958. This electromagnetic method measures the degree to which the soil attenuates the transmission of an electric current. Archaeological features, such as buried walls, have a high resistance and retard transmission; other features, such as the contents of pits which retain soil moisture, offer far less resistance. The results of the work in l958 were uneven and very hard to interpret-- largely due to the low sensitivity of the instrument--so resistivity was abandoned at that point as a technique for archaeological prospecting.

At the same time, Black and Mr. Eli Lilly had been following the work of M. J. Aitken and his associates at the Oxford University (UK) Archaeometric Laboratory, specifically their use of proton magnetometers to locate and map buried archaeological features. Magnetometers measure minute differences in the magnetic field of the earth caused by ferrous concentrations (and iron objects) in the soil, thermo-remnant magnetism of fired clays, and differences in the magnetic susceptibility of disturbed soils. In 1959, with the support of the Indiana Historical Society, Black rented a general purpose magnetometer from Varian Associates. It proved a bit better than a "dowsing rod" for locating very large features, but it was not really suited to archaeological work. In l959, again with the support of the Society, Black purchased a magnetometer built specifically for archaeological work from the Oxford Archaeometric Laboratory.

In 1961 and again in 1963 the National Science Foundation awarded funds to Black and to the Indiana Historical Society to evaluate the magnetometer as a tool for archaeological research. At this point Johnston joined the project as a Research Associate to oversee the field and laboratory work. During the Summers of 1961-1963, more than 100,000 square feet of the Angel site were surveyed and more than 7,000 square feet were excavated in order to match anomalous magnetometer readings with the archaeological features that produced those readings. For the most part the strategy was to locate and follow the course of the stockade trench, but other features, such as houses, fire basins, and pits, were sought and located as well. This work, which I believe was the first systematic use of a proton magnetometer for archaeological research in North America, was fully reported by both Johnston and by Black (Black, G. A. and Johnston, R. B., "A Test of Magnetometry as an Aid to Archaeology," American Antiquity, Vol. 28, pp. 199-205, 1962; Black, G. A., Angel Site: An Archaeological Historical, and Ethnological Study, 2 vols., Indiana Historical Society, Indianapolis, 1967; Johnston, R. B., "Proton Magnetometry and its Application to Archaeology: An Evaluation at Angel Site," Indiana Historical Society, Prehistory Research Series, Vol. IV, No. II, 1962).

The magnetometer used by Black and Johnston, although state-of- the-art for 1961, was very slow and not very sensitive. They were able to complete no more than 1 ,5OO readings per day, and the average number of readings per day was far less. Moreover, the precision of their instrument was, at best, 5 nT (nanoTeslas): for comparison, the average magnetic field in southern Indiana is approximately 57,000 nT. Finally, they were not able either to make corrections for the diurnal variation of the magnetic field of the earth or to correct for variation caused by sunspot activity, both of which seriously affect the accuracy of magnetometer measurements.

In the twenty-five years since Black and Johnston worked there have been major changes in geophysical instrumentation, especially in the design and use of magnetometers. First, methods have been developed to control the effects of diurnal and extraordinary magnetic variation due to the rotation of the earth and activity of the sun, respectively. Second, the precision and accuracy of instruments have increased by at least one order of magnitude: precision of 0.1 nT and accuracy of 0.05 nT is routine. Third, the instruments have become sufficiently small and lightweight to be carried in one hand (the instrument used by Black and Johnston, for comparison, filled a Chevrolet Corvair van). Fourth, comparable gains have been made in electromagnetic instruments--resistivity and conductivity meters. Fifth, ground penetrating radar--which works much like conventional radar, except that signals are reflected from below-ground strata and objects--although far more expensive than the other methods, is becoming more widespread as instruments are produced commercially for archaeological applications. Finally, most modem instruments have an internal memory module that stores the value of each measurement and the grid coordinates at which it was taken. As a result, more than ten thousand readings can be taken each day with some instruments--a ten-fold increase from what Johnston could expect on the best of days.

During the Summer of 1989, two instruments were tested at the Angel site: a Geoscan Model 36 Fluxgate Gradiometer (magnetometer) and a Geonics EM38DL Conductivity Meter (a form of resistivity measurement). Both these instruments proved to be quite sensitive, very efficient, and extremely effective in the location of archaeological features (see the abstract by Stephen Ball,89 et al.)

Given the marked improvements in geophysical instruments, plans have been made to conduct a fine-scale geophysical map (points every 1.0 or 2.5 feet) for the whole of the Angel site (ca. 92 acres). [Note: the Angel site was originally surveyed in English units of measure, and these have been retained; comparable metric units would be measurements every 25 cm or every 1.0 m over ca. 37 ha.] This venture will be directed by Professor emeritus of Physics John Weymouth of the University of Nebraska, a leading expert on geophysical applications in archaeology, and Professor Christopher S. Peebles of Indiana University. They will serve as Principal Investigators, and Mr. Stephen Ball, Prehistory Research Fellow of the Glenn A. Black Laboratory of Archaeology will be a Co-Investigator and field director for the project. A proposal for support of this research was submitted to the National Science Foundation in December, 1989.

This project will employ two varieties of magnetometry-- gradiometery and difference magnetometry--and two varieties of electromagnetic measures--resistivity and conductivity. The gradiometry surface win have readings taken every 1.0 feet; readings from the other instruments will be taken every 2.5 feet. Thus the first map will contain measurements from approximately 5,000,000 points, and the other three maps will comprise measurements at each of approximately 1,000,000 points. Data values of this magnitude win require an extremely fast computer (one that is capable of more than 100,000,000 instructions per second: > 100 mips) to manipulate singly and simultaneously these four geophysical "surfaces" and produce inferential maps of archaeological features that lie beneath the surface of the Angel site. It is only within the last few years that such computers have become available for any but projects in cryptography and in physics.

If all goes as planned, this project will consume several person -years of effort and several hundred thousand dollars. In return, it will produce an inventory of the number, variety, and location of archaeological features at the site. Moreover, with a bit of judicious and nondestructive archaeological testing, the nature of these features and their chronological locus in the development of the site can be determined. These data, in turn, should enable us to write the history of the Angel site as a major Mississippian settlement from the tenth to the fifteenth centuries of our era. The result win be the maximum amount of infommation with the absolute minimum of destruction. If, however, excavation of the whole of the Angel site were to be seriously considered, a conservative estimate of costs would be on the order of 300 person years and more than twenty million dollars. The result, if the excavation were done well, would be a large and fine collection, but there would be nothing left on which to apply better archaeological methods in the future.

[return to 1989 abstracts menu] [continue to next]

Created: July 23, 1996
Copyright 1996, Glenn Black Laboratory of Archaeology and The Trustees of Indiana University
Last updated: September 15, 2003