From 1993 to the present, Historic Archaeological Research has directed and or supervised a number of archaeological investigations designed around the use of metal detector technology. The objective of each project has been to systematically recover metal artifacts in an effort to enhance the archaeological record. The author has nearly thirty years experience with metal detecting devices; the last eleven have been dedicated to this specific purpose. This is a 2004 revision of the original writing and is intended to provide a general outline for archaeologists considering this approach to survey of historic period sites.

The following is a brief discussion and description of the metal detectors used by HAR; although, there is no intention to go into great detail on this topic. There are perhaps hundreds of makes and models of metal detectors manufactured by several dozen domestic and foreign manufacturers, with a like number of features offered on this wide variety of instruments. In order to reduce variables and control survey parameters, metal detection instruments are selected for complimentary performance characteristics that are fundamental to archaeological reconnaissance. These types of detector have been classified as Type I, II and III. Special purpose instruments may also be useful under certain circumstances and are discussed in further detail below.

Both the Type I and II metal detectors are VLF motion, full range discriminating detectors. Since we are not usually concerned with the instruments ability to discriminate (identify and reject) most metal items, discrimination control really isn't an issue. In fact, detectors intended for use by hobbyists to locate coins and the like are typically designed with features that make them less than desirable for archaeological survey work. Instruments with visual or audio target identification and or additional discrimination enhancements have been observed to skew sampling uniformity. These discrimination features should either be manually defeated by adjustment, or this kind of metal detector should be avoided altogether.

Metal detectors for use in archaeological survey are selected, not for the facility to determine the probable type of material, but rather for inherent sensitivity and a critically quick response to all metal objects, particularly those found in close proximity to one another. The types of detector we have found most useful are considered more basic in design and function yet, when working in tandem, possess the ideal characteristics for thorough artifact recovery.

• Type I VLF: Type I instruments operate near the 5-6 KHz range. These detectors have an affinity for iron artifacts and exhibit exceptional sensitivity to metal items on the high end of the conductivity spectrum i.e. brass, copper and silver.
• Type II VLF: The Type II detectors generally have an operating frequency near 10-15 kHz. These instruments are intrinsically sensitive to metal targets of mid range conductivity such as aluminum, lead, nickel; gold and small irregularly shaped artifacts. The Type II is typically less sensitive to small iron artifacts making this a better choice for working in and around concentrations of ferrous debris.
• Type III Multi-frequency: Recent innovations in metal detector design exploit the response characteristics of both the Type I and Type II instruments by simultaneously utilizing multiple frequencies. This may be accomplished differently by various manufacturers; however, for the purpose of this discussion, the Type III circuit design is considered a combination of the Type I and II detectors.
• Hand-held Magnetometer: This device is designed to locate ferrous metals only and is used to identify the presence of iron and other ferromagnetic materials. The hand-held magnetometer is an extremely sensitive detector capable of identifying artifacts as small as nails near the surface, in addition to larger objects at much greater depths.
• Two-box type Metal Detector: This is a Transmit-Receive (TR) type metal detector capable of locating metal objects that have relatively large surface areas up to extreme depths. The two-box will detect both ferrous and non-ferrous metals; however, it will not indicate on small metal artifacts of any kind, and is typically used in conjunction with or instead of the magnetometer to identify larger deeply buried items. This instrument may also be used to locate voids, and is useful if it is necessary to inductively trace linear features such as utility lines.

Metal detection technicians may work either alone or with the assistance of an excavator. In either case, permitting the technician to take part in the recovery of detected artifacts has been proven to be both expeditious and, perhaps more importantly, provides immediate operator feedback critical to the metal detection process.

Each metal detector, regardless of type classification, is adjusted for response to all metal artifacts unless otherwise instructed by a supervisor. Under some circumstances, it may become necessary to purposely reject concentrations of debris or items too numerous to collect, and a representative sample may be deemed appropriate. Examples of this include modern metal surface waste, disintegrated wire fence material or concentrations of near-surface building debris. Otherwise, all detected metal items are recovered and collected by the operator and or the excavation personnel that have been assigned to assist the technician.

All artifacts of interest are assigned a field specimen number and the location of recovery is marked by a non-conductive pin flag for subsequent mapping. Any unidentified items are considered to be artifacts and are collected for laboratory analysis. The artifact collection bag includes all pertinent information including the date, operator's name, detector type, axis of sweep and the depth of the recovered item. Field specimen bags are subsequently collected and recorded in an FS log. Items recovered but not assigned field specimen identification are collected and counted permitting statistical breakdown of metal artifact density within the survey area.

Wide Area Survey: Large parcels of land may be sampled by deploying metal detector operators to survey multiple parallel transects. The Type I and II instruments, by virtue of different operating frequencies, can be positioned relatively close together without cross-talk and spurious interference. As a rule, Type I and II detector technicians are positioned alternately at five meter intervals. Technicians using Type III instruments may have to be placed in line where minimum interference is obtained. This is a more than adequate technique for quickly sampling larger areas that are not immediately slated for intensive metal artifact recovery. Pin flags placed at recovered artifact locations often indicate distribution patterns and areas of concentration that may merit further more intensive investigation methods.

2-2-90 Method of Intensive Metal Detection: HAR has refined an intensive metal detection survey procedure that we have dubbed the 2-2-90 Method. A grid is established of discretionary size, usually 10-15 meter units with the former being the more desirable. Units larger than 15-meters are difficult for operators to cover completely and should be avoided. Technicians are assigned to grid units and the survey axis is determined by instrument type. Typically, the Type I operators survey units along the N-S axis and technicians using Type II instruments investigate along the E-W axis. The Type III detector may be substituted for either of the other two types.

Each grid square is examined in overlapping sweeps resulting in 100% initial coverage. Following completion of a unit, a second technician with an opposite type detector is assigned to survey the unit perpendicular to the first operator's sweep. Operator skill and detector performance are integrated in this fashion. This technique results in coverage by 2 operators using 2 complimentary detectors, and at a 90 degree angle of approach. Coverage is improved exponentially and either results in an increase of collected artifacts or more thoroughly substantiates the lack of metal remains within a given survey unit.

Metal detection reconnaissance, in conjunction with more traditional archaeological survey, is a pragmatic method that clearly enhances artifact sampling and archaeological research of historic period sites. Some form of these techniques should be considered in all cases where metal artifact distributions are anticipated. For more detailed information, please feel free to contact HAR.