Review of Geophysical
Methods Used in Archaeology

by

Jeffrey C. Wynn

ABSTRACT

Geophysical methods have been used with increasing frequency in archaeology since 1946; aerial photography has been used since 1919. The geophysical methods that are most commonly used at present are electrical resistivity, magnetics, and ground-probing radar. Magnetic detectors, particularly when used in a gradient mode or with a continuously recording base station, are used at almost all sites where any geophysical methods are used. Portable, noncontacting electromagnetic soil-conductivity systems are also being increasingly used because of their very high rate of data acquisition. Less commonly used methods include self-potential (sometimes called spontaneous potential), microgravity, radiometric, thermal infrared imagery, and sonic or seismic techniques. Recent developments in image processing and graphic representation have contributed substantially to the archaeologist’s ability to do "rescue archaeology," that is, to carry out high-speed, nondestructive reconnaissance surveys for ancient human cultural evidence in advance of modern industrial development.

INTRODUCTION

Classical archaeological methods, including trenching with trowels and brushes, require an enormous expenditure of human energy often provided by unpaid student labor. Enthusiasm can provide payment only up to a point, however. On some of the large contract archaeological surveys, such as the Tombigbee Waterway, Dolores Reservoir, and Black Mesa coal mining, archaeological surveyers were paid to scan the surface for evidence of human occupation (John Weymouth, oral communication, 1985). This is extremely inefficient because most buried evidence will be missed, and the effort is not very cost-effective. Several geophysical methods are now available that are capable of mapping as much as a quarter hectare of land (at a 1-m grid spacing) in a day’s time. Some conductivity surveys can cover as much as a hectare in a day under special circumstances. As a result, the enormous expenditure of human hands and-knees effort can now be used much more efficiently. Archaeologists often call ground geophysical methods "archaeological remote sensing," or "archaeogeophysics." They now use these methods with ever-increasing frequency as an important adjunct to several aspects of their work.

Archaeologists began using air photos immediately after World War I. Since the mid 1940s, they have been using other geophysical methods for prospection of archaeological sites. (Prospection is a European term for archaeogeophysics occasionally used by North Americans.) Initially, their efforts focused on resistivity surveys, followed soon after in the early ’50s by magnetic surveys. High-resolution seismics, ground and spaceborne radar, infrared imagery, and self-potential methods were available to the archaeologist by the 1980s.

The reason for this growth in activity is that geophysical methods can provide an extremely rapid, three-diamensional reconnaissance of a site. They can also provide a synoptic view of the potential human cultural resources of a target area. Surface geophysical methods currently in use can usually detect soils disturbed by burials. They can identify hollows and voids in structures such as pyramids and ancient fortifications, and can map buried stone foundations. All this can be done rapidly, without ever disturbing the ground. This ability to explore without damaging a site probably does not seem important to a geologist. An archaeologist, however, knows that excavation opens the way for the inevitable destruction of preserved remains by weather and vandals. Excavation, in fact, destruction. J.W. Weymouth (written communication, 1985) remarks that "archaeology is the one science that destroys its own lab-no repeated experiments ... it better be done right the first time, or not at all."

Modern examples of the use of geophysics include searching for hidden cavities in Chephren’s Pyramid at Giza in Egypt (Moussa and Dolphin, 1977), searching for hidden caves in Victorio Peak, New Mexico (Dolphin et al., 1978), and locating buried Olmec basalt monuments in east-central Mexico (Breiner and Coe, 1972). Lost 200year-old graves in Maryland have been found using geophysical techniques; in one gravesite an iron coffin from the 19th century was identified and dated, and at the same time information was provided on how it was cast (reversed magnetic polarity suggests that it was cast upside down from its final resting configuration; J. Wynn, unpublished data, 1985). Other modern uses of geophysics include mapping of ancient Indian villages in Washington State (Huggins, 1984) and - of submerged paleoshorelines in the search for settlement sites in the Gulf of Mexico (Stright, 1986). Finally, geophysical methods are an integral part of the search for 5000-year-old shaft tombs in Jordan (Frohlich and Lancaster, in press), and the search for Herod’s tomb in Israel (L. Dolphin, oral communication, 1984).

TERMINOLOGY

Before proceeding further, a digression into essential terminology is necessary. "Archaeogeophysics" and "archaeological prospection" are the terms most commonly applied to the field described in this paper. This field includes geophysical methods used in site prospection but not isotopic provenance or archaeomagnetic dating. Provenance, incidentally, (sometimes spelled "provenience") is the study of the source of ancient artifacts to document ancient trade and communication patterns (Aitken, 1974). The latter technologies are included along with prospection methods in the broader term "archaeophysics." A survey of the broader field of physics applied to archaeology is available in excellent summaries by Aitken (1974), and Wolfman (1984).

Archaeologists frequently use the term rescue archaeology." It refers to emergency evaluation of an area for human cultural resources. This normally is done, under pressure, in advance of industrial development (the bulldozer used to prepare a site for modern human cultural occupation). It is nearly impossible to do this effectively without assistance from geophysical methods. The term "non-destructive archaeology" refers to using remote sensing" methods to provide three-dimensional information about a large tract of land. The key element here is that the evaluation is done without disturbing the land. Archaeologists commonly use the term "remote sensing," incidentally, for more than just photo or LANDSAT image analysis. For them, it includes the whole range of surface geophysical and geochemical methods.

Within the general area of physics used in archaeology, a distinction can be made between prospection and nonprospection methods. Geophysical methods used to prospect for sites not yet found, or to, search for features within sites already known (intra-site mapping) fit within the category of prospection methods. Archaeomagnetism (for example, the dating of kilns and hearths by analysis of the orientation and intensity of their remanent magnetism) is a nonprospecting method. Recent summaries of this specialty are available in Wolfman (1984), and Tarling et al. (1986). Isotopic analysis is another nonprospecting method. It is used both for dating purposes (for example, radiocarbon dating; see Olsson, 1970), as well as for provenance studies. Except for providing the above references as starting points, I will not discuss these other methods further.

GEOPHYSICAL METHODS USED IN ARCHAEOLOGY

Airphotos and Modern Digital Imaging Systems

Probably the first instrument-based prospection method applied to archaeological sites, first used in England just after World War 1, was aerial photography (Beazeley, 1919). Archaeologists have frequently used it since then because it gives a synoptic view, something especially useful before a site is excavated (Aitken, 1974; Binford, 1964). Aerial photogrammetric methods are particularly useful because they help identify soil marks and crop marks. These subtle differences in soil color, moisture content, and texture indicate buried structures such as walls, or past human agricultural activity. A typical application of this method is to photograph a site from a low-flying airplane or a tethered balloon. Fine-scale contour base maps are then constructed from photos taken in stereo pairs. Photos are also used to look for any features not readily visible on the ground that may give the archaeologists some place from which to start their excavation efforts.

Airborne and space-borne digital imagery has been used experimentally since the 1960s (Stringer and Cook, 1974; Lyons and Avery, 1977; Ebert and Lyons, 1978) with only limited success. The resolution of the imagery, especially LANDSAT, is too low to allow effective use in intra-site archaeological prospection. A good summary of the modern state-of the-art is found in Ebert (1984), who points out that such imagery can sometimes identify land forms and vegetative types conducive to human occupation. Berlin et al. (1977) have successfully used Landsat imagery to map areas in northern Arizona that had been cultivated and then abandoned almost 700 years ago.

The U.S. Geological Survey and NASA have experimented with aircraft digital imagery as a tool for mapping mineral resources. This kind of imagery has considerably higher resolution than LANDSAT data, and is limited only partly by the height above ground. It also covers a much wider spectral range than even the space-borne LANDSAT Thematic Mapper system (in one case, 128 spectral channels). To the author’s knowledge, this kind of imagery has not yet been used in an archaeological application, but could prove extremely useful. An extensive bibliography on air- and space-borne remote sensing applications in archaeology is available in Lyons et al. (1980).

Thermal methods, both ground and airborne, have great potential for use in the archaeological sciences. These methods include temperature probes placed in the ground (Benner and Brodkey, 1984), and thermal infrared imagery, used to map subtle temperature differences in exposed surfaces and soils. Thermal-inertia methods, using aircraft thermal imaging systems flown at two different times, are also included in this category; buried stone blocks, for example, lose thermal energy at different rates than the surrounding topsoil, resulting in anomalous day time night time thermal ratios. Specific applications could include mapping voids and other structures in ancient fortifications such as those that exist in Peru, Egypt, and Jerusalem. Perisset and Tabbagh (1981) have demonstrated that digital imagery in the infrared can be used in archaeological applications.

Seismic Methods

Historically, bosing has proven to be a useful but nonquantitative exploration tool (Aitken, 1974). (Bosing refers to thumping the ground with a heavy rammer- to detect different sounds caused by resonant effects over hollows, structures, and soils of different compaction.) A sonic spectroscope (Carabelli, 1966) was used to test stone and brick walls for voids and variations in thickness. The method uses transmitted sonic energy in the frequency range of about 20-3000 Hz. Resonance effects are detected by means of a wide bandwidth accelerometer or geophone attached to a wall. Typically, the low natural resonant frequency (5-20 Hz) of a solid wall will increase in the vicinity of a void.

Geophysicists have experimented with refraction seismic methods in archaeological applications with relatively little success (Carson, 1962; Dolphin, oral communication, 1985; Aitken, 1974). Refraction methods work best in mapping undisturbed layers that have velocities increasing with depth. The method becomes less useful and interpretation becomes very qualitative and difficult when there are velocity inversions representative of human cultural disturbance, or highly three-dimensional objects such as burial sites or stone foundations.

Seismic-reflection methods (for example, sonar) are known to work well in marine applications. Shipwrecks buried in sediments in the Mediterranean stand out clearly by means of this kind of technique (McGhee et al., 1968; Edgerton, 1972). Seismic-reflection methods have also been used to detect cavities in otherwise homogeneous rock masses or in ancient stone structures. One particular example was the detection of a cavity beneath the ancient mosque/synagogue at Machpelah in modern Israel, where tradition suggests that Abraham, the ancient Hebrew prophet, was buried (Dolphin, oral communication, 1984). Seismic-reflection methods were also used to map faults and cavities in Chephren’s pyramid at Giza in Egypt (Dolphin, 1981).

High-resolution structure mapping technologies are the most recent application of seismic reflection in archaeology. These methods are the same as those used to investigate sites for ocean-drilling platforms. Recently, they were used to map sites of potential human occupation on ancient shorelines beneath the sediments of the Gulf of Mexico (Stright, 1986). These surveys have resolution capabilities sufficient to permit the identification of maddens (trash heaps) and perhaps even dwellings. They hold out the promise of making major contributions to the study of the early human migration across the Bering Straits, where most potential sites are now underwater (Kontrimavichus, 1984; Dikov, 1983; McManus et al. 1983).

Magnetic Methods

Magnetic methods were first used in the 1950s (Belshe, 1957; Aitken et al. 1958), and have since become the backbone of archaeological prospection. They are now used even more frequently than electrical prospection methods. Typically, soils that have had a campfire maintained over them develop an increased magnetic susceptibility resulting from the consequent reducing environment. This reducing environment causes the formation of magnetite if even moderate amounts of iron are present. Magnetometers can easily detect variations of less than 0. 1 percent magnetite content in the soil.

Soils compacted by human occupation or disturbed by a burial will also show a variation from background values of magnetic susceptibility. Burials frequently cause localized oxidation, creating a void in magnetite content. These phenomena have been exploited by geophysicists to produce large-scale maps of aboriginal occupation sites in the Great Plains of North America (Weymouth and Huggins, 1985). They have also been used to map Roman-age occupation sites in Europe (Scollar et al., 1986). Archaeologists now even use susceptibility meters to map soils modified by human activity (Colani and Aitken, 1966).

Field procedures for the magnetic prospection method have been refined considerably in the last two decades. Some archaeologists now use magnetic gradiometers and microcomputer-controlled automatic data gathering systems to make low-noise, high-productivity surveys. It is not unusual for a field crew to acquire as much as a third of a hectare of 1-m-grid data in a single day. Some archaeologists have reintroduced fluxgate magnetometers to speed up the data gathering process even further (Clark, 1986). Experiments have also been carried out to physically model the effects of campfires and human occupation. This results in better interpretation of field acquired data (Weymouth and Huggins, 1985; Weymouth, 1986; Gibson, 1986). Figure 1, provided by John Weymouth, shows the remarkable detail and coverage now possible with magnetic methods.

Archeaomagnetism

Archaeomagnetism is not a prospection method in itself, but is mentioned here for completeness. Archaeologists and physicists have used it to date ancient sites with identifiable hearths and kilns in North America, Britain, and southern Europe (Wolfman, 1984; Tarling et al., 1986). Carefully oriented samples are first acquired from the walls and floors of these kilns and hearths. The age of last firing can then be inferred from the intensity and orientation of the remanent magnetic field of the sample. The process requires the calibration of the declination, inclination, and intensity of the ancient earth field. This process is still being refined by means of dendrochronology (tree-ring analysis) and radiocarbon-14 methods.

Electrical Methods

Electrical methods were first used in an archaeological application in the 1940s (Atkinson, 1952; Aitken, 1974). They have been used extensively ever since, especially in Europe. These methods are divided into the non-contacting electromagnetic (sometimes called EM, or induction) methods, and the soil contacting (sometimes called galvanic or resistivity) methods.

Some of the initial work with EM methods was carried out by Scollar (1962), Foster (1968), Tite and Mullins (1969, 1970), and Tabbagh (1974). Modern shallow-penetration, high-resolution EM systems are now routinely used in archaeology. They are ideally suited for site mapping because of their resolution and speed. Frolich and Lancaster (in press) have shown that data can be acquired as fast as a person can walk. The method proved invaluable in the mapping of 5000year-old shaft tombs in Jordan and similar sites in Saudi Arabia, Bahrain, Kuwait, and elsewhere (B. Frohlich, oral communication, 1984). Tabbagh (1986) has given a good summary of EM work up to the 1980s, and introduces the concept of using careful parameter control to derive magnetic susceptibility information at the same time that conductivity information is acquired. EM methods were also used in archaeological prospection in the United States with considerable success (Bevan, 1983; Wynn and Sherwood, 1984).

Galvanic, or soil-conduction electrical methods have been used since the 1950s (Atkinson, 1952; Aitken, 1974). The best-known method is resistivity profiling, and typically a Wenner or pole-pole array is used. Sumner (1976) has provided a good description of methods and arrays in common use along with their relative advantages. Induced polarization and self-potential methods are also soil-contacting (Wynn and Sherwood, 1984).

Resistivity methods are especially helpful in detecting gross porosity changes caused by buried stone structures. These changes are almost always invisible to present-day surficial examination. Resistivity methods were extensively used for archaeology in England (Aitken, 1974; Clark, 1986) and in Italy (Carabelli, 1967; Linington, 1970a). They are particularly useful where high water tables prevent mapping of Roman ruins in England or Europe (Aitken, .1974; Pattantyus-A, 1986). Automatic data-acquisition systems using string-and-pulley field coordinate input methods are now used to produce field maps such as Figure 2 (from Wynn and Sherwood, 1984).

The induced-polarization (IP) method has been used with moderate success since the 1960s (Aspinall and Lynam, 1968,1970). IP is useful because it can provide information on the presence of disturbed clay- or pyrite-rich horizons in an area where there has been human occupation. Limited field experience suggests that the IP method provides information of greater clarity than resistivity methods (Aitken, 1974: 191). The requirement of nonpolarizing electrodes slows down the field work considerably, however. The method is only rarely used now because of this time constraint and the cost of necessary sophisticated electronic equipment.

The self-potential (SP) method has proven to be the least expensive geophysical method used in archaeological applications. The equipment consists of only a digital volt meter, some wire, and several low noise, nonpolarizing electrodes. Among other things, the method was used in a marine application to map the presence of an ironclad ship sunk in 1864 during the U.S. Civil War (Corwin, 1973). Wynn and Sherwood (1984) have shown that the method often gives anomalous responses over archaeological targets in areas where one or more other geophysical methods have failed to indicate anything unusual.

The cause of the SP response is thought to be the streaming potential, or selective stripping of ions from percolating rain water by the soil. This effect is different for areas where the soil homogeneity has been disturbed by burials or soil compaction. Electrode-drift noise levels usually encountered in mineral resource SP surveys are about 10 - 30 millivolts. The principle difficulty with the SP method is that archaeological responses are in the range of this mineral survey noise threshold. Stringent field procedures and some filtering can lower this threshold to manageable levels, however. Figure 3 (modified from Wynn and Sherwood, 1984) shows the typical response found over archaeological targets.

A new method for archaeological prospection became available in the mid-1970s. This was ground-probing radar, originally developed for rapid engineering reconnaissance of building sites (Moffat, 1974; Morey, 1974; Cook, 1974; Vickers and Dolphin, 1975; Dolphin et al., 1978; Ulriksen, 1982). The method uses continuous (sometimes pulsed) radar transmission from one or two towed antennas. These typically operate in the 150-500 megahertz range, and give reflections from conductivity contrasts caused by metallic objects or disturbed soil horizons. Sites of houses long since lost have been identified by mapping reflections over soil horizons excavated for a basement. Burials and ancient man-made structures often give similar soil-layer-interruption signatures. Vaughan (1986) mapped burials and whale vertebrae of a Basque whaling station in Labrador using groundprobing radar. Figure 4 (Bevan, 1983), shows a typical record obtained by means of a ground-probing radar system.

The principal weakness of the groundprobing radar method is that it cannot normally penetrate below a clay horizon (though it will detect gaps in one readily enough). The equipment is also very expensive, and interpretation of the data can be complex and difficult. Often, the use of an incorrect antenna means that important features are obscured or missed entirely. This happens because of poor resolution (using an antenna of too low a frequency) or excessive attenuation (using an antenna of too high a frequency) for the feature of interest. Efforts are only just beginning to apply moveout corrections and signal processing technologies to radar data (Gary Olhoeft, U.S. Geological Survey, oral communication, 1985). These methods, of course, have been routinely used with seismic data in the search for hydrocarbons for more than two decades.

Space-borne applications of radar to archaeological prospecting have suddenly appeared in the literature, almost seemingly by accident. Recently, dendritic features were observed in Shuttle Imaging Radar (SIR-A) data acquired in southern Egypt and northern Sudan. Closer examination in the field has shown that these features were caused by ancient drainage systems beneath the sand cover. Related Stone Age occupation sites buried beneath at least five m of this unusually dry sand in the eastern Sahara were also found (McCauley et al., 1982). Under conditions found in most of the rest of the world, radar energy cannot penetrate more than a few centimeters because of the presence of water in the soil. Penetration can be improved only by coupling the antenna directly to the ground surface.

Less Commonly Used Methods

Radiometric data have been experimented with at least since 1967 (Peschel, 1967). Their use stems from the fact that soils normally have greater uranium, thorium, and potassium content than underlying sedimentary rock. Consequently one would suspect that soil-filled pits would give anomalously high radioactivity. Human remains and middens also have high levels of phosphate in them along with attendant radiogenic isotopes (Eldt, 1977). However, soils only a few centimeters thick can mask any radioactivity generated by sources buried beneath them. Measurement times with a small portable crystal detector are also on the order of a minute per reading. These limitations will probably prevent the method from ever being widely used.

Neutron scattering can detect moisture content in anthropogenic soils (Alldred and Shepherd, 1963). It is theoretically possible to detect stone structures beneath a porous soil, or voids behind a rock or brick wall. Penetration limits of 10-20 cm, however, make it hardly worth serious consideration. Aitken (1974) did state that a void behind one brick thickness of a wall would be detectable.

Gravity methods were tried in archaeology several times since 1965 (Linington, 1966; Kolendo et al., 1973; Fajklewicz et al., 1982). Anthropogenic gravity signatures are typically extremely small, however, and usually close to the instrumental noise threshold.

Centimeter-precision accuracy in elevation control is absolutely necessary when applying the gravity method in archaeology. This accuracy requirement, and similar precision requirements in calculating terrain corrections (with the enormous computational labor this requires), have significantly dampened the interest of geophysicists who might have considered applying this method.

DATA ACQUISITION, PROCESSING, AND IMAGE ENHANCEMENT

In the last two decades, geophysicists working in archaeological applications have introduced many innovations to speed up data acquisition and the presentation/interpretation of the data. Many of these innovations have simply been applied to archaeology as the technology became available in other fields. The most important developments in the field equipment include automated microprocessor controlled field data and coordinate acquisition systems (Weymouth and Huggins, 1985). Other innovations, such as streamlined field procedures and customized instrumentation, demonstrate the blossoming of human genius when faced with a tedious and time-consuming task.

Linington (1970b) reported initial efforts to develop filtering procedures for archaeogeophysical data. The desire to extract every useful piece of information from field data has preoccupied geophysicists working in archaeological applications ever since. One of the more significant breakthroughs in the last decade has been the introduction of image processing display and filtering technology. These technologies derive mainly from electrical engineering and remote sensing (geophysicists’ meaning of the word) research. Skilled interpreters often miss subtle information existing in the data when it is presented in the form of contoured maps. Filtering methods and image-enhancing algorithms of remarkable sophistication are now commonly applied to archaeological geophysical data, which are then displayed in the form of an image rather than a contour map. This usually enhances subtle linear features otherwise lost in the presence of strong anomalies of geologic origin (Scollar et al., 1986).

In the Federal Republic of Germany and elsewhere in Europe, image enhancement methods are routinely applied by archaeologists to generate land utilization maps (Scollar et al., 1986). These maps show potential human occupation sites that are ubiquitous on the continent. The sites are not usually apparent to an archaeologist from surface examination of a site, however. Maps of this sort are now used to decide whether or not a new land development will be allowed to proceed. There are political consequences when large development projects are held up by the process of archaeological certification, of course. This has led to significant governmental financial support for the development of fast, efficient methods of data acquisition and image presentation (Scollar et al., 1986).

LIMITATIONS OF GEOPHYSICAL METHODS

Geophysical methods used in archaeology are not an unqualified panacea for the archaeologist. In fact, there are several reasons why geophysical methods do not work, or are not cost-effective, in archaeological applications. A primary reason is that they are for the most part instrumentation, computer, and interpretation intensive. Use of a geophysical consultant can be prohibitively expensive, and some ground-probing radar systems cost upwards of $50,000. Data processing and image enhancement methods are also expensive, usually requiring custom application to each data set. Archaeologists do not usually deal with the relatively large costs that geoscientists are much more accustomed to.

Nonanthropogenic sources for geophysical anomalies are also a major problem with geophysical measurements over archaeological sites. Often, anomalies caused by ancient human cultural activity lie beneath the noise threshold of the surrounding geologic environment. A frustrating example is that the normal variation of magnetite content in soils and rocks underlying a site frequently exceeds the anthropogenic anomalies by an order of magnitude or more. This problem is not insurmountable, however. The higher frequency content of shallow archaeological sources means that filtering can remove most of the unwanted noise.

Sometimes, ancient anthropogenic anomalies are unobservable due to the large variations caused by modern cultural interference (powerlines, roads, and so on). This means that there will be areas, especially where population density is large, where geophysical prospecting for archaeological sites can be carried out with only extreme difficulty.

It is also not unusual for physical properties to vary little with the human disturbance of a solid horizon. This problem forces the archaeologist or geophysicist to search for more than one physical property in which contrasts are sufficient to be useful in mapping. Often the problem can be mitigated with careful field procedures that increase the signal-to-noise ratio in the data. This is especially true of the self-potential method.

Resolution and depth limitations are important restraints on the use of geophysical methods in most applications. Because most anthropogenic features are near the surface, this does not usually cause problems for an archaeologist working on the earth’s surface. A notable exception is work conducted by Ain Shams University and Stanford Research Institute at the pyramid of Chephren. Here radar echo and seismic-reflection searches for hidden chambers were frustrated by moisture and clay content, and the presence of many joints in the limestone blocks. Signals had to traverse distances as much as 100 m because of the size of the pyramid (Moussa and Dolphin, 1977).

FUTURE DEVELOPMENTS ON THE HORIZON

The trend in geophysical data gathering in archaeological prospecting is toward acquiring and using very large data bases. This is being done with microprocessor-controlled instruments and innovative developments in field procedures. Together these permit gathering of large numbers of data points in a relatively short amount of time. Large databases mean that the archaeologist can cover larger areas with higher resolution than ever before. The larger amounts of data and higher resolution also mean we can expect increased performance from digital image-enhancement algorithms; there is more data to work with. Consequently, ground geophysical methods can now provide the kind of synoptic site examination formerly available only with aerial photography.

High resolution airborne and spaceborne imagery systems, such as the Large Format Camera on the U.S. Space Shuttle (Doyle, 1985), are now coming online. This means that archaeologists can examine large areas with a resolution good enough for archaeological mapping. Visible and near-infrared, as well as thermal inertia systems also hold considerable promise for the archaeologist. Analytical combinations of several types of ground data and imagery will probably become the norm in the next several years, as it is in the geophysics profession in general. Techniques now available to register and overlay different gridded data sets will probably lead to major breakthroughs in site identifications during the remainder of the 20th century.

There has been extensive use of laboratory measurements in provenance studies, and elsewhere in archaeology and anthropology (Aitken, 1974). Little work has been done, however, in characterization and identification of geophysical signatures of anthropogenic materials. The complexity of compaction and weathering processes, and variations in environmental conditions (including moisture and magnetite content) certainly make this a daunting project. Until this work is undertaken, geophysicists will be unable to carry out detailed quantitative interpretations of archaeogeophysical data. Generic studies may turn out not to be very useful; we won’t know until they have been tried, however.