The Laboratories

In order to understand the radiocarbon dates, it is necessary to understand the practices of the various laboratories. Canadian archaeological and palaeobiological dates have been obtained from 79 laboratories. I have compiled historical data concerning the operations of most of these laboratories, with special emphasis on pretreatment protocols, attention to fractionation effects, and methods of calculating errors. Such information is unavailable for some laboratories, especially those that have never published formal date lists.
 
 

A (University of Arizona, Tucson). Until May 1958, this laboratory employed the solid-carbon method of sample preparation and then installed carbon-dioxide proportional counters (Shutler and Damon, 1959). Twelve dates from Canadian sites were produced in proportional counters, prior to the installation of liquid scintillation equipment (see Long and Muller, 1981). Three Alaskan samples were analyzed after that change. The one-sigma errors reflect only the counting statistics.

AA (University of Arizona Accelerator, Tucson). This AMS laboratory has dated 76 Alaskan and 11 Canadian samples. I assume that normalized ages are quoted with one-sigma errors.

AECV (Alberta Environmental Centre at Vegreville). This liquid scintillation laboratory was closed in 1995 when the parent facility became the Alberta Research Council. Beginning late in 1986, del C-13 was routinely measured, using the mass spectrometer located in the Department of Chemistry, University of Alberta (L.D. Arnold, p.c. 1998). Most AECV dates have been listed by Beaudoin (1987, 1988, 1991) whose reports present del C-13 measurements for all samples with laboratory numbers above AECV-400. Only the normalized ages are presented by Beaudoin. AECV produced 48 dates from northern Canada archaeological sites, with lab numbers ranging from AECV-536 to AECV-2020. I assume that all of the ages have been normalized on the basis of a measured del C-13 ratio, whether or not the archaeologist reports the ratio itself. Errors are quoted at one sigma.

ANSTO (Australian Nuclear Science and Technology). One AMS date on graphite from a Québec rock art panel was obtained from this laboratory. I assume that it has a one sigma error.

AU (University of Alaska, USA). This liquid scintillation laboratory began operation in Fairbanks in 1969. The one-sigma errors included variations in the sample, background and contemporary standards; no corrections were made for isotopic fractionation (Reeburgh and Young, 1976: 1).

B (Universitat Bern, Switzerland). This gas proportional laboratory originally converted samples to acetylene (Gfeller et al., 1961: 15) and subsequently switched to methane to fill its counters (Oeschger and Riesen, 1965: 1). The one-sigma errors were based on sample activity and the statistical errors of the background and modern standard.

Beta (Beta Analytic, Inc., USA). This laboratory performs liquid scintillation counting at its Miami, Florida, facility and offers direct counting by forwarding samples to several accelerator mass spectrometry (AMS) laboratories. A del C-13 measurement can be obtained for a modest extra fee, and current practice is to report the uncorrected age, del C-13, and the normalized age (“conventional age”) along with the results of a calibration routine. Starting with Beta-70221 in February 1994, Beta routinely supplies a calibration of each date produced, and del C-13 is estimated if it has not been measured (R.E. Hatfield, p.c. 1999). It appears that all AMS dates reported by Beta are corrected for isotopic fractionation, but it is sometimes stated that such a correction has been made without the del C-13 ratio being quoted. The AMS laboratories differ in their methods of measuring del C-13. For example, a Beta Analytic report to E.L. Syms states that “C13/C12 ratios for AMS samples measured at Kiel, Groningen, and ETH are measured in the accelerator itself. The reported ratio represents the natural isotopic ratio plus possible measurement effects. Although the correct adjustment has been applied to derive the Conventional C14 Age, the ratio should not be interpreted as the isotopic ratio of the original” (E.L. Syms, p.c. 1998). At Oxford, del C-13 is measured to within ± 0.5-1.0 ‰, except in a few cases where del C-13 has been estimated (Hedges et al., 1995: 417). The CAMS lab obtains del C-13 measurements from a mass spectrometer independent of its accelerator system (R. McNeely, p.c. 1998). I assume that all AMS dates from Beta have been reported as normalized ages. For the liquid scintillation dates, I have requested the original data sheets from the submitters but have not seen all of them. It is possible that in a few cases I have performed corrections on ages that were already normalized, in which case the normalized age given in this database is meaningless. Errors are quoted at 1-sigma. The Beta web sites is located at www.radiocarbon.com.

BGS (Brock Geological Sciences, Brock University, St. Catherines, Ontario). This liquid scintillation laboratory reports errors representing one sigma.

Birm (University of Birmingham, Birmingham, England). This laboratory uses gas proportional counters, and del C-13 is measured on methane gas samples, the ages usually corrected for fractionation (Williams and Johnson, 1976: 249). Since four of the five Canadian samples consisted of land mammal bone, it is worth noting that the laboratory had adopted the Longin method of collagen extraction by the time of these analyses. The 1-sigma errors reflect the statistical analysis of sample, background, and standard.

BM (British Museum, London). The seven dates from northern Canada were measured in a liquid scintillation counter and were normalized on the basis of measured del C-13 ratios. These dates were analyzed during a period of lab operation that was later found to have produced erroneous results, and the dates were re-evaluated (Bowman et al., 1990). It is clear that collagen was extracted from the bones (Burleigh et al., 1976, 1982), but the method of extraction is not described. The most recent BM date list states that the term “‘collagen’ is used ... to mean the acid insoluble organic fraction” (Ambers et al., 1991: 51). The 1-sigma errors reflect only the counting statistics.

Brookhaven(Brookhaven National Laboratory, Upton, New York). This laboratory is not conducting radiocarbon analysis on an ongoing basis, but it filled a particular need by supplying miniature gas proportional counters in which samples down to a level of 10 mg of carbon could be analyzed (Harbottle et al., 1993: 174). This technology was used to date the Smithsonian iron bloom from Kodlunarn Island at a time when AMS dating was in its infancy. The errors represent one sigma.

C (University of Chicago, Chicago, Illinois). This laboratory analyzed the first Canadian radiocarbon sample ever published, a solid-carbon date from the Burley site in Ontario. This database also contains 10 solid-carbon dates on Alaskan samples.

CAMS (Center for Accelerator Mass Spectrometry, Lawrence Livermore Laboratories, California). In addition to AMS dates obtained through Beta Analytic, this laboratory has furnished AMS dates for 328 Canadian and 23 Alaskan samples. Del C-13 ratios have been reported for many of them, and I assume that all of them have been normalized and are given with 1-sigma errors.

CRNL (Chalk River Nuclear Laboratories, Chalk River, Ontario). This tandem accelerator was operated by the Atomic Energy Commission of Canada (Andrews et al., 1980; Brown et al. 1983). It was designed to study a variety of isotopes, and interest in radiocarbon dating was mainly oriented toward dating groundwater in support of a search for nuclear waste disposal sites. Its early experimental analyses employed a variety of samples supplied by archaeologists, including 13 dates from northern Canada. In a few cases del C-13 measurements were performed in a separate mass spectrometer, but in most cases the ages are normalized on the basis of estimated del C-13. Errors represent one sigma. I have invented the lab code “CRNL” and used the numbers assigned by the National Museum of Man when the samples were transmitted to the laboratory.

CWRU (Case Western Reserve University, Cleveland, Ohio). This laboratory issued approximately 100 dates using the CWRU prefix before it was transformed into the Dicar Corporation laboratory with the DIC prefix. Despite an informal agreement that the CWRU lab numbers would remain unchanged (D. Stothers, p.c. 2000), the first DIC date list (Sumodi, 1974) omitted the CWRU label entirely. Some of the dates have been reported with both lab numbers in various publications, and all of them are listed as DIC samples in this database.

DAL (Dalhousie University, Halifax, Nova Scotia). This gas proportional laboratory was built with equipment from the former Ohio Wesleyan University lab (Ogden and Hart, 1976). The counters were filled with methane, and the 1-sigma errors were based on uncertainties of sample, modern and background activity (Ogden and Hart, 1977).

DIC (Radioisotopes Laboratory, Dicar Corporation, Cleveland, Ohio). This liquid scintillation laboratory quotes errors at one sigma. Bone is decomposed in a refrigerated (10°C) acid bath (1.0N), to enhance collagen recovery, and after filtration and washing the residue is treated with cold 1.0N NaOH for two hours prior to combustion (Sumodi, 1974). Some DIC samples have been reported with a CWRU prefix as well as with the DIC laboratory code.

GaK (Gakushuin University, Tokyo). Many of the dates produced by this laboratory were published in date lists (Kigoshi et al., 1969, 1973). These samples are known to have been dated in acetylene-filled proportional counters. Some of the samples were processed in later years, and it is not known whether changes were made to the equipment or procedures. The method of bone pretreatment is described by Kigoshi and Kobayashi (1966: 54): “Bone samples were dated on the organic materials obtained by the following procedure. After washing with distilled water the powdered bone samples were boiled with 10% H2SO4 solution 10 to 40 hours. The extract, a clear solution, was evaporated to almost dryness, and, after concentrated H2SO4 was added, heating was continued until most of the organic compounds became insoluble carbonized or polymerized material. The black residue was washed in water and treated as the usual charcoal sample for dating.” This method represents rather brutal chemistry, and it requires large samples to ensure adequate yield. In my experience, the method produced reliable results, although gentler techniques requiring smaller samples are to be recommended. The 1-sigma errors reflect only the counting statistics.

Gif (Centre National de la Recherche Scientifique, Gif-sur-Yvette [Essonne], France). This proportional counting laboratory reports errors at one sigma. The prefix “Gif” was adopted when the laboratory, previously labelled “Gsy,” was joined together with the Saclay laboratory (Delibrias et al., 1966).

GIN (Geological Institute, Russian Academy of Sciences, Moscow, Russia). This is a liquid-scintillation laboratory.

GrN (University of Groningen, Netherlands). This laboratory began operation with a solid carbon counter, but its operator, Hessel de Vries, installed one of the first carbon dioxide counters (see Willis, 1998, for an amusing account). The only sample in this database was analyzed by proportional counting and has a one-sigma error.

GSC (Geological Survey of Canada). This laboratory employs gas proportional counters of various sizes and operating pressures using CO2 gas. It has published nearly all of its dates in Date Lists that permit researchers to learn many details concerning sample pre-treatment, dilution requirements, and counting times. Its record is also outstanding for the attention given to the identification of sample materials, whether charcoal, wood, bone, or shell. Pre-treatment of bone samples was recognized as problematic early in the history of the laboratory, and experiments were conducted to test varying pre-treatments (Lowdon et al., 1969: 23; Lowdon and Blake, 1970: 48-49). The procedures usually followed are similar to those described by Berger and others (1964), extracting collagen in an insoluble form. However, in recent years the GSC Lab. has sent most bone samples to the Saskatchewan Research Council where soluble collagen extractions are performed. In the late 1960s, GSC began routinely to acquire measurements of del C-13 and to report normalized ages for most samples. The ages of organic materials are normalized to -25 parts per mil, but marine shells were unconventionally corrected to 0 parts per mil. Since 1992, the GSC Lab. reports all of its ages normalized to -25 per mil, but still provides a corrected age for marine shells. GSC practice is also unconventional in reporting errors of ±2 sigma. The age errors include counting errors of sample, background, and standard, and the error in the half-life of C-14. In Date Lists published prior to 1974, the GSC errors also included a term to account for variations in the C-14 concentration of the atmosphere, and the error assigned to an age was always a minimum of ±100 years (Lowdon, et al. 1972; Lowdon and Blake, 1973). In the early 1970s, this term was dropped from the error calculations, and 2-sigma errors of less than ±100 years were routinely reported. When comparing dates from GSC with those from other laboratories, this difference in the error term must always be borne in mind. The GSC web site is located at sts.gsc.nrcan.gc.ca/radiocarbon.

GX (Geochron Laboratories, Inc., Cambridge, Massachusetts). This laboratory began operation in April 1963 with a methane proportional counter. Bone and shell samples were “treated in a special apparatus which separates and recovers both the carbonate and collagen (or conchiolin) fractions in a manner similar to that of Berger, Horney and Libby (1964)” (Krueger and Weeks, 1965: 47). Approximately 500 dates had been measured by the time the last date list was published (Krueger and Weeks, 1966), and all of the Canadian dates were processed in later years. This laboratory has continued to play an important role in producing pairs of dates on the collagen (or gelatin) versus apatite of bone samples, although experience had soon suggested that a “collagen date is far superior in cases of discordance” (Krueger and Weeks, 1966: 146). Presumably the errors represent one sigma, although this was not explicitly stated in either published date list.

HAR (Atomic Energy Research Establishment, Harwell, Berkshire, England). This liquid scintillation laboratory reports one sigma errors “based on all accountable errors in the measurement process, eg, the allowance for background and calibration uncertainties is based on the observed long term reproducibility; and not merely on the counting statistics of the measured sample. Conventional pretreatment processes have been used throughout. Bone determinations are always based on collagen ages” (Otlet and Slade, 1974: 179).

I (Isotopes - a Teledyne Company, Westwood, New Jersey). Gas proportional counters filled with CO2 were used by this laboratory. Bone samples were pre-treated with a method similar to that of Berger et al. (1964), extracting collagen in an insoluble form (Buckley and Willis, 1970: 87). Subsequent date lists, the last of was I-XII (Buckley and Valdes-Pages, 1981), indicated no change in this practice, although dates analyzed in later years may have been employed different procedures. The error term from Isotopes was based on counting statistics of the sample, background and modern standard, and it included no factor for the half-life of C-14 or for variations in atmospheric concentration of C-14.

IEMAE (Institute of Evolutionary Morphology and Animal Ecology, Russian Academy of Sciences, Moscow, Russia). This is a liquid-scintillation laboratory.

IGAN (Institute of Geography, Russian Academy of Sciences, Moscow, Russia). This is a liquid-scintillation laboratory.

IM (Institute of Permafrost Studies, Siberian Branch of the Russian Academy of Sciences, Yakutsk, Russia). This liquid-scintillation laboratory is no longer active.

K (Danish National Museum, Kopenhagen). This laboratory began operation with a solid carbon counter, but it had switched to a gas proportional counter filled with CO2 by the time the first sample from Canada was analyzed (Tauber, 1960). I have found no information on pre-treatment methods for bone samples in this laboratory. The errors quoted include standard deviations of the count rates for the unknown sample, contemporary standards, and background. “Because possible errors arising from isotopic fractionation, or from fluctuations in the atmospheric C14 activity, are not included, calculated errors smaller than 100 years were increased by rounding to that figure as a minimum” (Tauber, 1973: 86).

Ki (Institute of Radiogeochemistry of the Environment, Ukraine Academy of Science, Kiev, Ukraine). This is a liquid-scintillation laboratory.

KRIL (Institute of Forestry, Siberian Branch of the Russian Academy of Sciences, Krasnoyarsk, Russia). This liquid-scintillation laboratory is no longer active.

L (Lamont Geological Observatory, Columbia University, New York) . This gas proportional laboratory reported one-sigma errors.

LE (Institute of the History of Material Culture, Russian Academy of Sciences, St. Petersburg, Russia). This is a liquid-scintillation laboratory.

LJ (University of California, San Diego, La Jolla, California). This acetylene gas proportional laboratory began operation at the Scripps Institution of Oceanography but was moved to Mt. Soledad in 1969. Its one-sigma errors are based only on the counting statistics of background, standard, and sample activities (Bien and Pandolfi, 1972; Linick, 1977).

LU (Leningrad University [St. Petersburg State University], Russia). This liquid scintillation laboratory has produced one date on a mammoth bone from Yukon, and eight dates on mammoth remains from Wrangel Island. I assume that the quoted errors represent one sigma.

Lv (University of Louvain, Louvain, Belgium). This methane gas proportional laboratory reports one-sigma errors based on the counting variations of the unknown sample, the modern standard and the background (Gilot, 1968).

M (University of Michigan, Ann Arbor). From 1950 through 1952, this laboratory used a solid carbon counter that experienced increasing interference from radioactive fallout due to atmospheric bomb tests (Crane, 1956). M-38, from Alaska, was analyzed during this period. After suspending operations for a while, the lab resumed work with two carbon dioxide-carbon disulfide Geiger counters that were used for all subsequent analyses (see Crane and Griffin, 1958). I have found no statement concerning bone pre-treatment methods, and it is possible that whole bone was employed in the three Canadian dates that are pertinent to this question. M-622 and M-1085 were given “errors at least twice as great as the statistical errors of counting, in order to take account of other errors in the over-all process” (Crane and Griffin, 1966: 256). All other Michigan dates from Canada have errors that reflect only the counting statistics.

MAG (Quaternary Geology and Geochronology Laboratory, Far Eastern Branch of the Russian Academy of Sciences, Magadan, Russia). Dates from this liquid-scintillation laboratory are reported with one-sigma errors.

Mo (Institute of Geochemistry, Russian Academy of Sciences, Moscow, Russia). This liquid-scintillation laboratory is no longer active.

N (Institute of Physical and Chemical Research [RIKEN], Japan). This proportional counting laboratory uses carbon dioxide counters and quotes errors at one sigma (Yamasaki et al., 1972).

NSRL (Nuclear Structure Research Laboratory, Institute of Arctic and Alpine Research, University of Colorado, Boulder). This AMS laboratory has produced seven dates that appear in this database, and I assume that they are normalized with one-sigma errors.

NUTA (Nagoya University Tandem Accelerator Centre, Nagoya, Japan). This AMS lab has produced four dates that appear in this database, and I assume that they are normalized with one-sigma errors.

OS (National Ocean Sciences AMS Facility, Woods Hole Oceanographic Institution, Woods Hole, Massachusetts).

OxA (Oxford University, England). This AMS laboratory routinely publishes date lists in the journal “Archaeometry,” and all of the Canadian dates have appeared in those lists. “The errors are quoted as one standard deviation, and are our estimate for the total error in the system including the sample chemistry. This estimate includes the statistical precision from the number of 14C nuclei detected (usually the dominant error for dates of less than 5000 years B.P.), the reproducibility of the mass-spectrometric measurements between different targets, and the uncertainty in our estimate of the contamination background” (Gillespie et al., 1985).

P (University of Pennsylvania, Philadelphia). This laboratory began operation in 1959 with a solid carbon counter. Seventeen Alaskan samples, and one from northern Canada (P-62) were analyzed by this method. The laboratory switched to CO2 counters, and all but two of the Canadian dates appear either in date list P-IV (Ralph and Ackerman, 1961) or in date list P-IX (Stuckenrath et al., 1966; see also Rainey and Ralph, 1959). The bone dates listed in P-IV were measured from the combustion products of whole bones, and the normalized ages must be considered minimum ages. Errors quoted reflect the counting statistics of the unknown, the standard, and the background; they do not reflect the half-life error.

PITT (University of Pittsburgh, Pennsylvania). This laboratory was a reincarnation of "Merlin's Lair" operated by the late Robert Stuckenrath (see Adovasio, 1995). I do not know whether the gas proportional counters from moved from the former Smithsonian lab or whether this lab was equipped for liquid scintillation counting. I assume that the quoted errors represent one sigma.

QC (Queens College, City University of New York, Flushing, New York). This laboratory uses liquid scintillation counters for most samples, but a gas proportional counter was added in 1981 to handle the analysis of small samples (Pardi et al., 1984). The 1-sigma errors reflect the standard deviations of the sample, standard, and background activities.

QL (Quaternary Isotope Laboratory, University of Washington, Seattle). This gas proportional laboratory has produced six dates that appear in this database with one-sigma errors.

QU (Centre de Recherches Minérales, Ministère des Richesses Naturelles, Gouvernement du Québec). This laboratory began operation in 1972 with a liquid scintillation counter (Samson et al., 1977), but it is no longer active. In addition to many samples from Québec, 18 samples were dated for the Northern Yukon Research Programme at the University of Toronto, including six samples of bone. It is known that collagen was extracted from the bones, but details of the method employed have not been reported. Quoted errors reflect only the counting statistics at one-sigma.

RIDDL (Radioisotope Direct Detection Laboratory). This AMS lab consisted of an instrumental line operated by Simon Fraser University at the Tandem Accelerator Laboratory of McMaster University in Hamilton, Ontario (Nelson et al., 1986b; Vogel et al., 1987). During its several years of operation it made important contributions to northern Canada archaeology and palaeobiology, especially in the Yukon. Erle Nelson and his colleagues continued to refine methods of extracting collagen, using procedures described elsewhere (Brown et al., 1988; Morlan et al., 1990; Nelson et al., 1986a; Vogel et al., 1991). In a few cases where it was critical to correct for fractionation, del C-13 was measured in a mass spectrometer, but for most of the dates del C-13 was estimated. Errors are quoted at one sigma.

Riga (All-Union Research Institute of Submarine Geology and Geophysics, Riga, Latvia). This liquid-scintillation laboratory is no longer active.

RL (Radiocarbon Ltd.). This is a privately owned laboratory that began operation in New York state in 1969 using two proportional counters filled with CO2 (Tucek, 1971). Conversion to liquid scintillation counting began in 1972, and many samples were counted both as CO2 and as benzene during the transition period (Tucek, 1977). The laboratory was moved to Lampasas, Texas in 1974, where I presume that it continued operation with liquid scintillation counting, but no more date lists were published to confirm this. The lab is no longer active, and its pre-treatment methods have not been described although they are said to be “standard techniques” (Tucek, 1971: 74). Errors are based on counting statistics only, and they are quoted at one sigma.

S (Saskatchewan Research Council, Saskatoon). This laboratory began operation at the University of Saskatchewan with support from the Saskatchewan Research Council (SRC) and later the National Research Council of Canada. By 1973, the lab was being operated commercially by the SRC, and the National Museums of Canada soon joined its operation to provide radiocarbon dating service for Canadian archaeologists. The Saskatchewan lab began operation with a screen-wall counter, and it produced 13 dates from solid carbon samples (McCallum, 1955). By the time its second date list appeared (McCallum and Dyck, 1960), the screen-wall counter had been replaced by a gas proportional counter filled with acetylene. This method continued in use until 1989 when the lab was dismantled and reopened off-campus in the new quarters of the SRC. In this new phase of its long history it operated liquid scintillation counters until it was closed in 1999. Some samples were analyzed by both gas proportional and liquid scintillation counting during the transition period, but samples after S-3051 through S-3669 were analyzed by liquid scintillation (J. Zimmer, p.c. 1998). In view of the importance of this laboratory and the large number of bone samples it has processed, I asked Jurgen Wittenberg for a summary of its bone pre-treatment methods. He reported that four methods have been employed (J. Wittenberg, p.c. 1989): S-1~S-50, bone was prepared by hydrolyzation, a difficult method with little yield; S-51~S-399, whole bone was charred; S-400~S-1299, collagen was extracted in an insoluble form; and after S-1300, collagen extractions were solubilized with a method similar to that of Longin (1971). This last method was described in detail (Rutherford and Wittenberg, 1979), and it remained in use thereafter (J. Zimmer, p.c. 1998). The Saskatchewan lab did not measure del C-13 routinely, but such a measurement could be provided on request. In the latter case, the lab reported the measured age and the del C-13 ratio, trusting the client to calculate the normalized age. I presume that the one-sigma errors reflect only the counting statistics, although this was not explicitly stated.

SBAS (see SOAN). This prefix represents a translation of the terms represented by SOAN.

SDAS (see SOAN). This prefix represents a translation of the terms represented by SOAN.

SFU (Simon Fraser University, Burnaby, B.C.) . This liquid scintillation laboratory is no longer in service, but it produced approximately 350 dates, most of them included in this database. Bones were pre-treated according to the method of Longin (1971) or a modification thereof (Nelson and Hobson, 1982). Although no mass spectrometer was locally available, del C-13 ratios were acquired from other laboratories, especially for bone samples and standards, and the results used to calculate normalized ages. The one-sigma errors reflect counting statistics of unknown, standard, and background, as well as an estimate of uncertainty in the relative counting efficiencies.

Shell (Shell Development Company, Houston, Texas). I have no information concerning this laboratory.

SI (Smithsonian Institution, Washington, D.C.). This gas proportional laboratory was established in the basement of the Smithsonian “castle” in 1962 with methane-filled counters (Sigalove and Long, 1964). Samples SI-1 through SI-616 were processed there before the lab was dismantled in 1969 for renovations to the building. The lab re-opened the next year in the Radiation Biology Laboratory at Rockville, Maryland (Stuckenrath and Mielke, 1972). Employing similar equipment but better shielding, samples SI-617 and onward were dated at Rockville until the lab was closed by financial constraints in 1986. Specific methods of bone pre-treatment are not described in the date lists, but the usual method was to extract collagen unless otherwise stated. The lab produced 10 pairs of collagen/apatite comparisons that are published in full herein for the first time. The 1-sigma errors reflect counting statistics but may also be adjusted for small-sample dilution (Stuckenrath and Mielke, 1973: 388).

SM (Socony Mobil Oil Company, Dallas). This methane gas proportional laboratory analyzed one Canadian and 13 Alaskan samples. The one-sigma error reflects counting variations of background, standard, and unknown sample (Bray and Burke, 1960).

SMU (Southern Methodist University, Dallas). This laboratory is operated by the Department of Geological Sciences within the Institute for the Study of Earth and Man. Most of its dates are from liquid scintillation counters, although very small samples can be measured in methane proportional counters acquired from the Mobil Field Research Laboratories (Haynes and Haas, 1974: 368). I assume that the quoted errors represent one sigma.

SOAN (Institute of Geology, Siberian Branch of the Russian Academy of Sciences, Novosibirsk, Russia). This is a liquid-scintillation laboratory.

St (Stockholm, Sweden). This carbon dioxide gas proportional laboratory calculates one-sigma errors from the statistical uncertainties of sample, standard, and background, as well as the error in the half-life of radiocarbon (Ostlund and Engstrand, l960: 186). Most dates younger than 5000 years are corrected for isotopic fractionation by means of del C-13 measurements on the counting gas, and dates lacking such measurements have increased error terms (Ostlund and Engstrand, 1963: 204). Bone is pre-treated with a method modified after Berger et al. (1964) and Sellstadt et al. (1966; Engstrand, 1967).

T (Norwegian Institute of Technology, Trondheim, Norway). This carbon dioxide gas proportional laboratory reports one-sigma errors calculated only from the counting statistics and not including the half-life error. Most samples are measured for isotopic fractionation and normalized (Nydal et al., 1964).

TK (University of Tokyo, Japan). This laboratory counts acetylene in a proportional counter and calculates one-sigma errors based on the sample, standard, and background activities (Sato et al., 1971: 94).

TO (IsoTrace Laboratory, University of Toronto). According to Manley and Jennings (1996), the reporting practices of this AMS laboratory have deviated from the conventional format in some respects. Prior to about TO-1800, IsoTrace would report their dates according to the older GSC format if the submitter was a member of the GSC or if the submitter requested it. However, during this time they would report their dates according to the conventional format if the submitter was not associated with the GSC. Therefore for Isotrace Lab. Nos. less than TO-1800, you will not know from the Lab. number alone which procedure they followed and will need to go to the original reporting form. Since TO-1800, they have followed the conventional format, regardless of the submitter’s affiliation (Manley and Jennings, 1996: 161-162). The “older GSC format” refers to the practice of correcting marine shell dates to 0 parts per mil and dates on other materials to -25 parts per mil. The “conventional format” normalizes all dates to -25 parts per mil. IsoTrace reports normalized ages but seldom if ever reports a del C-13 measurement, and it is not clear whether del C-13 is being measured or estimated by this laboratory.

U (Fysiska institutionen, Uppsala universitet, Uppsala, Sweden). This gas proportional laboratory uses carbon dioxide counters. It has devoted special attention to bone pre-treatment methods, often dating a single bone five or six times with different pretreatments and reporting each of them. The 1-sigma error term includes counting statistics, uncertainty in the del C-13 measurements, and several minor technical aspects (Olsson and El-Daoushy, 1978).

Ua (Uppsala Tandem Accelerator Laboratory, Sweden). This AMS laboratory reports one-sigma errors, and I assume that all of its dates are normalized.

UCLA (University of California, Los Angeles). Carbon dioxide gas is counter in a proportional counter at this laboratory. The quoted error is "always at least a one-sigma statistical counting error," and del C-13 is measured in "critical cases" to adjust for fractionation (Berger and Libby, 1969: 194).

UCR (University of California, Riverside). Carbon dioxide proportional counters include one capable of accommodating samples containing as little as 250 mg of carbon. The one-sigma counting errors reflect sample, standard, and background activities (Taylor, 1982: 54).

UGa (University of Georgia, Athens). This liquid scintillation laboratory quotes one-sigma errors that reflect the counting statistics of the unknown, standard, and background (Noakes and Herz, 1983).

UL (University of Laval, Québec). This is a liquid scintillation laboratory that quotes one-sigma errors, but no other information is available.

UQ (University of Québec at Montréal). This liquid scintillation laboratory quotes one-sigma errors based only on the counting statistics of sample, standards, and background. No corrections are made for isotopic fractionation or reservoir effects (Dubois et al., 1988).

USGS (U.S. Geological Survey, Menlo Park). Carbon dioxide proportional counters are located 9.5 m below ground to minimize background. "The reported results closely follow the guidelines of Stuiver and Polach (1977), although the standard error for analyses earlier than USGS-500 are based solely upon counting statistics and do not include uncertainty in voltage, pressure, temperature, and del C-13" (Robinson and Trimble, 1983: 143).

UW (University of Washington, Seattle). This methane gas proportional laboratory quotes one-sigma errors and makes no corrections for isotopic fractionation (Fairhall et al., 1966).

W (U.S. Geological Survey, Reston, Virginia). This is a gas proportional laboratory with acetylene counters. It began operation in Washington, D.C., and was moved to Reston in the mid-1970s. The meaning of its 1-sigma error term seems to have varied, sometimes reflecting only the counting statistics (Kelley et al., 1979: 306), sometimes taking into account fractionation both in the laboratory and in nature (Kelley et al., 1978: 283), and sometimes excluding natural fractionation (Spiker et al., 1977). It appears that the error terms on the dates from northern Canada are larger than the statistical counting errors.

WAT (University of Waterloo, Waterloo, Ontario). This liquid scintillation laboratory “was built primarily for dating groundwater and a combustion line for organic samples was added in 1975” (Berry and Drimmie, 1982: 68). The latter has produced several dozen dates on Canadian archaeological sites. Samples analyzed after WAT-318 received routine del C-13 measurements, and the dates are normalized to -25 parts per mil. This laboratory is unconventional in reporting 2-sigma errors that reflect counting statistics only.

WK (University of Waikato, New Zealand). This liquid scintillation laboratory quotes errors at one sigma. Its web site is located at www.radiocarbondating.com.

WSU (Washington State University, Pullman). This is a gas proportional laboratory with methane counters. Its dates are quoted with 1-sigma errors that reflect only the counting statistics (Chatters, 1968; Sheppard and Chatters, 1976).

Y (Yale University, New Haven). This gas proportional laboratory quoted one-sigma errors “based only on the standard deviations in counting rates of sample and standards” (Stuiver, 1969: 545).

References:

Ambers, J., Matthews, K., and Bowman, S.
1991 British Museum natural radiocarbon measurements XXII. Radiocarbon 33(1): 51-68.

Andrews, H.R., Ball, G.C., Brown, R.M., Davies, W.G., Imahori, Y., and Milton, J.C.D.
1980 Progress in radiocarbon dating with the Chalk River MP Tandem Accelerator. Radiocarbon 22(3): 822-829.

Beaudoin, A.B.
1987 Alberta radiocarbon dates 1982-1986. In Archaeology in Alberta 1986, edited by M. Magne. Edmonton: Archaeological Survey of Alberta, Occasional Paper 31: 197-213.

1988 Alberta radiocarbon dates 1986-1987. In Archaeology in Alberta 1987, edited by M. Magne. Edmonton: Archaeological Survey of Alberta, Occasional Paper No. 32: 159-167.

Beaudoin, A.B.
1991 Alberta radiocarbon dates 1988-1989. In Archaeology in Alberta 1988 and 1989, edited by M. Magne. Edmonton: Archaeological Survey of Alberta, Occasional Paper No. 33: 239-253.

Berger, R., Horney, A.G., and Libby, W.F.
1964 Radiocarbon dating of bone and shell from their organic components. Science 144: 999-1001.

Berry, J.C., and Drimmie, R.J.
1982 University of Waterloo radiocarbon dates I. Radiocarbon 24(1): 68-82.

Bien, G.S. and Pandolfi, L.J.
1972 La Jolla natural radiocarbon measurements VI. Radiocarbon 14(2): 368-379.

Bowman, S.G.E., Ambers, J.C., and Leese, M.N.
1990 Re-evaluation of British Museum radiocarbon dates issued between 1980 and 1984. Radiocarbon 32(1): 59-70.

Bray, E.E. and Burke, W.H., Jr.
1960 Socony Mobil radiocarbon dates I. American Journal of Science Radiocarbon Supplement 2: 97-111.

Brown, R.M., Andrews, H.R., Ball, G.C., Burn, N., Davies, W.G., Imahori, Y., Milton, J.C.D., and Workman, W. 1983 Recent 14C measurements with the Chalk River tandem accelerator. Radiocarbon 25(2): 701-710.

Brown, T.A., Nelson, D.E., Vogel, J.S., and Southon, J.R.
1988 Improved collagen extraction by modified Longin method. Radiocarbon 30(2): 171-177.

Buckley, J.D. and Valdes-Pages, C.
1981 Teledyne Isotopes radiocarbon measurements XII. Radiocarbon 23(3): 329-344.

Buckley, J.D. and Willis, E.H.
1970 Isotopes' radiocarbon measurements VIII. Radiocarbon 12(1): 87-129.

Burleigh, R., Ambers, J., and Matthews, K.
1982 British Museum natural radiocarbon measurements XV. Radiocarbon 24(3): 262-290.

Burleigh, R., Hewson, A., and Meeks, N.
1976 British Museum natural radiocarbon measurements VIII. Radiocarbon 18(1): 16-42.

Chatters, R.M.
1968 Washington State University natural radiocarbon measurements I. Radiocarbon 10(2): 479-498.

Crane, H.R.
1956 University of Michigan radiocarbon dates I. Science 124: 664-672.

Crane, H.R. and Griffin, J.B.
1958 University of Michigan radiocarbon dates III. Science 128: 1117-1123.

Crane, H.R., and Griffin, J.B.
1966 University of Michigan radiocarbon dates XI. Radiocarbon 8: 265-285.

Delibrias, G., Guillier, M.T., and Labeyrie, J.
1966 Gif natural radiocarbon measurements II. Radiocarbon 8: 74-95.

Dubois, J.M.M., Occhietti, S., Pichet, P., Pagé, P., Jacob, C., and Bigras, P.
1988 Université du Québec à Montréal GEOTOP radiocarbon dates I. Radiocarbon 30(3): 355-365.

Engstrand, L.G.
1967 Stockholm natural radiocarbon measurements VII. Radiocarbon 9: 387-438.

Fairhall, A.W., Schell, W.R., and Young, J.A.
1966 Radiocarbon dating at the University of Washington III. Radiocarbon 8: 498-506.

Gillespie, R., Gowlett, J.A.J., Hall, E.T., Hedges, R.E.M., and Perry, C.
1985 Radiocarbon dates from the Oxford AMS system: datelist 2. Archaeometry 27(2): 237-246

Gilot, E.
1968 Louvain natural radiocarbon measurements VI. Radiocarbon 10(1): 55-60.

Harbottle, G., Creswell, R.G., and Stoenner, R.W.
1993 Carbon-14 dating of iron blooms from Kodlunarn Island. In Archaeology of the Frobisher voyages, edited by W.W. Fitzhugh and J.S. Olin. Washington: Smithsonian Institution Press, pp. 173-180.

Hedges, R.E.M., Houseley, R.A., Ramsey, C.B., and Van KLinken, G.J.
1995 Radiocarbon dates from the Oxford AMS system: Archaeometry datelist 20. Archaeometry 37(2): 417-430.

Kelley, L., Spiker, E., and Rubin, M.
1978 US Geological Survey, Reston, Virginia, radiocarbon dates XIV. Radiocarbon 20(2): 283-312.

Kelley, M.L., Spiker, E.C., Lipman, P.W., Lockwood, J.P., Holcomb, R.T., and Rubin, M.
1979 US Geological Survey, Reston, Virginia, radiocarbon dates XV: Mauna Loa and Kilauea volcanoes, Hawaii. Radiocarbon 21(2): 306-320.

Kigoshi, K., Aizawa, H., and Suzuki, N.
1969 Gakushuin natural radiocarbon measurements VII. Radiocarbon 11(2): 295-326.

Kigoshi, K. and Kobayashi, H.
1966 Gakushuin natural radiocarbon measurements V. Radiocarbon 8: 54-73.

Kigoshi, K., Suzuki, N., and Fukatsu, H.
1973 Gakushuin natural radiocarbon measurements VIII. Radiocarbon 15(1): 42-67

Krueger, H.W. and Weeks, C.F.
1965 Geochron Laboratories, Inc. radiocarbon measurements I. Radiocarbon 7: 47-53.

1966 Geochron Laboratories, Inc. radiocarbon measurements II. Radiocarbon 8: 142-160.

Linick, T.W.
1977 La Jolla natural radiocarbon measurements VII. Radiocarbon 19(1): 19-48.

Long, A., and Muller, A.B.
1981 Arizona radiocarbon dates X. Radiocarbon 23(2): 191-217.

Longin, R.
1971 New method of collagen extraction for radiocarbon dating. Nature 230: 241-242.

Lowdon, J.A. and Blake, W., Jr.
1970 Geological Survey of Canada Radiocarbon Dates IX. Radiocarbon 12(1): 46-86.

Lowdon, J.A. and Blake, W., Jr.
1973 Geological Survey of Canada radiocarbon dates XIII. Geological Survey of Canada, Paper 73-7: 61 pp.

Lowdon, J.A., Wilmeth, R., and Blake, W., Jr.
1969 Geological Survey of Canada radiocarbon dates VIII. Radiocarbon 11(1): 22-42.

Lowdon, J.A., Wilmeth, R., and Blake, W., Jr.
1972 Geological Survey of Canada radiocarbon dates XII. Geological Survey of Canada, Paper 72-7: 26 pp.

Manley, W.F. and Jennings, A.E.
1996 Radiocarbon date list VIII: Eastern Canadian Arctic, Labrador, northern Quebec, east Greenland shelf, Iceland shelf, and Antarctica. Boulder: Institute of Arctic and Alpine Research, Occasional Paper No. 50.

McCallum, K.J. and Dyck, W.
1960 University of Saskatchewan radiocarbon dates II. Radiocarbon 2: 73-81.

Morlan, R.E., Nelson, D.E., Brown, T.A., Vogel, J.S., and Southon, J.R.
1990 Accelerator mass spectrometry dates on bones from Old Crow Basin, northern Yukon Territory. Canadian Journal of Archaeology 14: 75-92.

Nelson, D.E. and Hobson, K.
1982 Simon Fraser University Radiocarbon Dates I. Radiocarbon 24(3): 344-351.

Nelson, D.E., Morlan, R.E., Vogel, J.S., Southon, J.R., and Harington, C.R.
1986 a New dates on northern Yukon artifacts: Holocene not Upper Pleistocene. Science 232: 749-751.

Nelson, D.E., Vogel, J.S., Southon, J.R., and Brown, T.A.
1986 b Accelerator radiocarbon dating at Simon Fraser University. Radiocarbon 28(2A): 215-222.

Noakes, J.E. and Herz, N.
1983 University of Georgia radiocarbon dates VII. Radiocarbon 25(3): 919-929.

Nydal, R., Lövseth, K., Skullerud, K.E., and Holm, M.
1964 Trondheim natural radiocarbon measurements IV. Radiocarbon 6: 280-290.

Ogden, J.G. III, and Hart, W.C.
1976 Dalhousie University natural radiocarbon measurements I. Radiocarbon 18(1): 43-49.

Ogden, J.G. III, and Hart, W.C.
1977 Dalhousie University natural radiocarbon measurements II. Radiocarbon 19(3): 392-399.

Olsson, I.U., and El-Daoushy, M.F.A.F.
1978 Uppsala natural radiocarbon measurements XII. Radiocarbon 20(3): 469-486.

Ostlund, H.G. and Engstrand, L.G.
1960 Stockholm natural radiocarbon measurements III. American Journal of Science Radiocarbon Supplement 2: 186-196.

Ostlund, H.G. and Engstrand, L.G.
1963 Stockholm natural radiocarbon measurements V. Radiocarbon 5: 203-227.

Otlet, R.L. and Slade, B.S.
1974 Harwell radiocarbon measurements I. Radiocarbon 16(2): 178-191.

Pardi, R.R., Tomecek, L., and Newman, W.S.
1984 Queens Collage radiocarbon measurements IV. Radiocarbon 26(3): 412-430.

Rainey, F. and Ralph, E.
1959 Radiocarbon dating in the Arctic. American Antiquity 24(4): 365-374.

Ralph, E.K. and Ackerman, R.E.
1961 University of Pennsylvania radiocarbon dates IV. Radiocarbon 3: 4-14.

Rutherford, A.A. and Wittenberg, J.
1979 Evidence in support of soluble collagen extraction for radiocarbon bone dating. Saskatoon: Saskatchewan Research Council Report No. C79-22, 8 pp.

Samson, C., Barrette, L., Lasalle, P., and Fortier, J.
1977 Québec radiocarbon measurements I. Radiocarbon 19(1): 96-100.

Sellstedt, H., Engstrand, L.G., and Gejvall, N.G.
1966 New application of radiocarbon dating of collagen residues in bone. Nature 212(5062): 572-574.

Sheppard, J.C. and Chatters, R.M.
1976 Washington State University natural radiocarbon measurements II. Radiocarbon 18(1): 140-149.

Sigalove, J.J. and Long, A.
1964 Smithsonian Institution radiocarbon measurements I. Radiocarbon 6: 182-188.

Spiker, E., Kelley, L., Oman, C., and Rubin, M.
1977 US Geological Survey radiocarbon dates XII. Radiocarbon 19(2): 332-353.

Stuckenrath, R. and Mielke, J.E.
1972 Smithsonian Institution radiocarbon measurements VII. Radiocarbon 14(2): 401-412.

Stuckenrath, R. and Mielke, J.E.
1973 Smithsonian Institution radiocarbon measurements VIII. Radiocarbon 15(2): 388-424.

Stuckenrath, R.J., Coe, W.R., and Ralph, E.K.
1966 University of Pennsylvania radiocarbon dates IX. Radiocarbon 8: 348-385

Stuiver, M.
1969 Yale natural radiocarbon measurements IX. Radiocarbon 11(2): 545-658.

Sumodi, A.J.
1974 Radioisotopes Laboratory radiocarbon date list I. Radiocarbon 16(2): 166-177.

Tauber, H.
1960 Copenhagen radiocarbon dates IV. Radiocarbon Supplement 2: 12-25.

1973 Copenhagen radiocarbon dates X. Radiocarbon 15(1): 86-112.

Tucek, C.S.
1971 Radiocarbon, Ltd. natural radiocarbon measurements I. Radiocarbon 13(1): 74-77.

1977 Radiocarbon, Ltd. natural radiocarbon measurements II. Radiocarbon 19(2): 245-262.

Vogel, J.S., Brown, T.A., Southon, J.R., and Nelson, D.E.
1991 Accelerator radiocarbon dates from the NOGAP archaeology project. In NOGAP archaeology project: an integrated archaeological research and management approach, edited by J. Cinq-Mars and J-L. Pilon. Canadian Archaeological Association Occasional Paper 1: 143-147.

Vogel, J.S., Nelson, D.E., and Southon, J.R.
1987 14C background levels in an accelerator mass spectrometry system. Radiocarbon 29(3): 323-333.

Williams, R.E.G. and Johnson, A.S.
1976 Birmingham University radiocarbon dates X. Radiocarbon 18(3): 249-267.

Willis, E.H.
1998 Hessel de Vries and the Groningen laboratory: recollections of pioneers in radiocarbon dating. Radiocarbon 40(3): vi-viii.