Editor Associate Editors

VOLUME 29 / NUMBER 1 / 1987

Published by THE AMERICAN JOURNAL OF SCIENCE

Editor MINZE STUIVER

Associate Editors To serve until January 1, 1989 STEPHEN C PORTER Seattle, Washington

To serve until January 1, 1988 W G MOOK Groningen, The Netherlands HANS OESCHGER Bern, Switzerland

To serve until January 1, 1990 ANDREW MOORE New Haven, Connecticut

To serve until January 1, 1992 CALVIN J HEUSSER Tuxedo, New York

Managing Editor RENEE S KRA

Kline Geology Laboratory Yale University New Haven, Connecticut 06511 ISSN: 0033-8222 NOTICE TO READERS AND CONTRIBUTORS Since its inception, the basic purpose of RADIOCARBON has been the publication of compilations of 14C dates produced by various laboratories. These lists are extremely useful for the dissemination of basic 14C information. In recent years, RADIOCARBON has also been publishing technical and interpretative articles on all aspects of 14C. We would like to encourage this type of publication on a regular basis. In addition, we will be publishing compilations of published and unpublished dates along with interpretative text for these dates on a regional basis. Authors who would like to compose such an article for his/her area of interest should contact the Managing Editor for information. Another section is added to our regular issues, "Notes and Comments." Authors are invited to extend discussions or raise pertinent questions to the results of scientific investigations that have appeared on our pages. The section includes short, technical notes to relay information concerning innovative sample preparation procedures. Laboratories may also seek assistance in technical aspects of radiocarbon dating. Book reviews will also be included for special editions. Manuscripts of radiocarbon papers should follow the recommendations in Suggestions to Authors* and RADIOCARBON Style Guide (R, 1984, v 26, p 152-158). Our deadline schedule for submitting manuscripts is: For Date Vol 29, No. 3, 1987 May 1, 1987 Vol 30, No. 1, 1988 Sept 1, 1987 Vol 30, No. 2, 1988 Jan 1, 1988 Half life of14C. In accordance with the decision of the Fifth Radiocarbon Dating Conference, Cambridge, 1962, all dates published in this volume (as in previous volumes) are based on the Libby value, 5570 ± 30 yr, for the half life. This decision was reaffirmed at the 11th International Radiocarbon Conference in Seattle, Washington, 1982. Because of various uncertainties, when 14C measurements are expressed as dates in years BP the accuracy of the dates is limited, and refinements that take some but not all uncertainties into account may be misleading. The mean of three recent determinations of the half life, 5730 ± 40 yr, (Nature, v 195, no. 4845, p 984, 1962), is regarded as the best value presently available. Published dates in years BP can be converted to this basis by multiplying them by 1.03. AD/BC Dates. In accordance with the decision of the Ninth International Radiocarbon Con- ference, Los Angeles and San Diego, 1976, the designation of AD/BC, obtained by subtracting AD 1950 from conventional BP determinations is discontinued in Radiocarbon. Authors or submitters may include calendar estimates as a comment, and report these estimates as cal AD/BC, citing the specific calibration curve used to obtain the estimate. Calibrated dates will now be reported as "cal BP" or "cal AD/BC" according to the consensus of the Twelfth International Radiocarbon Conference, Trondheim, Norway, 1985. Meaning of b14C. In Volume 3, 1961, we endorsed the notation A (Lamont VIII, 1961) for geochemical measurements of 14C activity, corrected for isotopic fractionation in samples and in the NBS oxalic-acid standard. The value of b14C that entered the calculation of A was defined by reference to Lamont VI, 1959, and was corrected for age. This fact has been lost sight of, by editors as well as by authors, and recent papers have used b14C as the observed deviation from the standard. At the New Zealand Radiocarbon Dating Conference it was recommended to use 614C only for age-corrected samples. Without an age correction, the value should then be reported as percent of modern relative to 0.95 NBS oxalic acid (Pro- ceedings 8th Conference on Radiocarbon Dating, Wellington, New Zealand, 1972). The Ninth International Radiocarbon Conference, Los Angeles and San Diego, 1976, recommended that the reference standard, 0.95 times NBS oxalic acid activity, be normalized to b13C = -19%a. In several fields, however, age corrections are not possible. b19C and A, uncorrected for age, have been used extensively in oceanography, and are an integral part of models and theories. For the present, therefore, we continue the editorial policy of using A notations for samples not corrected for age. *Suggestions to Authors of the Reports of the United States Geological Survey, 6th ed, 1978, Supt of Documents, U S Govt Printing Office, Washington, DC 20402. Vol 29, No. 1 Radiocarbon 1987

CONTENTS

Evaluation and Status of Liquid Scintillation Counting for Radiocarbon Dating Henry A Polach ...... 1 Scintillation Counter Performance at the SMU Radiocarbon Laboratory James M Devine and Herbert Haas ...... 12 Radiocarbon Content of Tropospheric CO2 at China Lake, California 1977-1983 Rainer Berger, TB Jackson, Robert Michael, and HE Suess...... 18 Study of Bone Radiocarbon Dating Accuracy at the University of Arizona NSF Accelerator Facility for Radioisotope Analysis TW Stafford, jr, AJTJull, Klaus Brendel, RC Duhamel, and Douglas Donahue ...... 24 Comparative Study of the Radiocarbon Datirig of Different Bone Collagen Preparations DM Gurfinkel ...... 45

Comparison of Oceanic a14C Data with Those of GEOSECS: Vertical Profiles in 1973 (GEOSECS) and in 1980 at (30°N, 170°E) in the Northwestern Pacific Ocean Toshitaka Gamo, Yoshio Horibe, and Hiromi Kobayashi ...... 53

DATE LISTS ANTW R Vanhoorne and AD Dubois Antwerp University Radiocarbon Dates IV ...... 57

BM Janet Ambers, Richard Burleigh, and Keith Matthews British Museum Natural Radiocarbon Measurements XIX...... 61

HAR AJ Walker, RS Keyzor, and RL Otlet Harwell Radiocarbon Measurements V ...... 78

RT Israel Carmi Rehovot Radiocarbon Measurements III ...... 100

Z Dugan Srdoc, Nada Horuatincic, Bogomil Obelic, Ines Krajcar Bronic, and Adela Sliepcevic Rudjer Boskovic Institute Radiocarbon Measurements IX...... 115

Z Dusan Srdoc, Bogomil Obelic, Adela Sliepcevd, Ines Krajcar Bronic, and Nada Horuatincic Rudjer Boskovic Institute Radiocarbon Measurements X ...... 135

NOTES AND COMMENTS Discussion: Comments on Multiple Dating of a Long Flowstone Profile Rainer Grim and Henry P Schwarez ...... 148

Reply Mebus A Geyh and GJ Hennig ...... 153 Drastic Increase of Background in the Gliwice Radiocarbon Laboratory During Late April, 1986, and its Time Changes Mieczystaw F Pazdur and Andrzej Zastawny ...... 156 [RADIOCARBON, VOL 29, No. 1, 1987, P 1-11] Radiocarbon

1987

EVALUATION AND STATUS OF LIQUID SCINTILLATION COUNTING FOR RADIOCARBON DATING` HENRY A POLACH Radiocarbon Dating Research, Australian National University, Canberra ABSTRACT. An international review of liquid scintillation low-level counting procedures and instrumentation made it possible to assess in detail those elements which lead to high-precision liquid scintillation radiocarbon dating with a Figure of Merit of 32,000. Current research is documented and future possibilities are alluded to.

INTRODUCTION Liquid scintillation (LS) counters are extremely versatile tools which grew, over the last 33 years, out of a need to detect and measure efficiently low energy /3-radioactivity. Most commonly, samples containing 3H and 14C are placed in a sealed counting vial located between two photomultiplier tubes (PMTS) on a common axis. This assembly is surrounded by a minimal lead shield (5cm Pb). Sample count rates are those which are coincident between the two PMTS and which fall within a selected pulse height (energy) window. Generally, there are up to 5 selectable windows spanning the energy of 0-2MeV. Pulse height analyses are enabled as the output of the PMTS is proportional to the energy of the ionizing event. The counts falling within each energy window are recorded. Because of interest in and application of /3-emitting isotopes to biomedical research, LS counters are produced commercially, in great numbers (hence economically) and by many manufacturers, mainly American and European. The detection efficiency for 3H and 14C is high with acceptable backgrounds for most uses (Table 1). For low-level counting of 14C (radiocarbon dating) and 3H (hydrology), at and below natural equilibria abundance levels, the radioisotope signal to background noise ratio of these multi-purpose multi-user LS counters was just acceptable. To improve their performance for low-level counting, attention was being focused progressively and systematically on sample conversion to counting medium, scintillation cocktails, counting vials, shielding and electronics. This paper traces, briefly, the development of 14C dating by LS counting from "give me a better counting liquid" (Arnold, 1958, p 129) to "let's have better commercial counters" (Noakes, 1977, p 189), and leads to the recent development of commercially available LS spectrometers designed specially to cater to ultra low-level, high-precision ,3-counting.

* This paper was presented at the Twelfth International Radiocarbon Conference in Trondheim, Norway, June 24-28, 1985.

1 2 Henry A Polach

TABLE 1 Performance of conventional` LS counters with a <5cm Pb shield surrounding the counting vial and the PMTS Figure of Efficiency Background merit Sample Isotope (E) % (B) cpm E2/B C6H6 ml 3H 67 25 -l80 20 14C 98 20 -480 15 * Commercially available, set up to manufacturer specifications with 20m1, 40K free, glass counting vials

THE LIQUID SCINTILLATION COUNTING SYSTEM The LS counting system consists of a number of discrete components, each one of which needs to be optimized for low-level counting. The components are counting liquid (solvent), scintillation solute (fluorescent molecules), vial, shielding and electronics of counters, data processing, data evaluation, and validation. Sample Preparation Scintillation solvent. From the 3H and 14C ,3-detection point of view, the best scintillation solvent is benzene (C6H6) or a mixture of benzene and 14C toluene (C6H5CH3). For dating, the benzene synthesis has been almost universally adopted as it enables quantitative transfer from sample C to solvent C and the benzene exhibits almost ideal energy transfer properties (eg, Birks, 1974, p 27). The sample carbon is first converted to C02, then to acetylene (C2H2) which in turn is trimerized catalytically to C6H6. Two acetylene syntheses exist: 1) CO2 is precipitated as strontium carbonate which is ignited with magnesium to form strontium carbide and C2H2 by subsequent hydrolysis (Suess, 1954). In the minority of cases, when this technique is now used, consistent yields are obtained by following the procedures established by McDowell and Ryan (1965). 2a) By far the most common and preferred method is the reaction of CO2 at elevated temperatures with lithium (Li) forming lithium carbide which upon hydrolysis yields C2H2 (Barker, 1953). The reaction was well studied and the procedures improved (Tamers, 1965a; Noakes, Kim & Stipp, 1965; Polach & Stipp, 1967; Tamers, 1975; Burleigh, Hewson & Matthews,1977). Routinely, 2 to 20g of elemental carbon can be converted with reproducibly high yields (96 ± 1%) and, using CO2 gas dilution prior to synthesis, the procedure can be applied with equal effectiveness to 0.1 g sample C (Polach, 1969; Polach, Gower & Frazer, 1972). 2b) When samples are large and relatively rich in carbon, a variant of the Barker method is often used. It consists of reacting an organic sample directly with hot Li (Stank, Arslanov & Klener, 1963; Swart, 1964). Benzene is formed from the C2H2 by low temperature catalysis Review of Liquid Scintillation Counting for 14C Dating 3 (Noakes et al, 1963). When CO2 - C6H6 conversion yields are high (>_80%) the isotopic fractionation is minimal; a depletion of b13C in benzene by 1.7 ± l %o was observed (Panarello, Albero & Angiolini, 1983, p 531). Comment: The most often used Li is supplied by Lithium Corporation of America, Bessemer City, North Carolina; the Vanadium V-0701 catalyst by Harshaw Chemical Co, Beechwood, Ohio. A review of various catalysts suitable for benzene synthesis was presented by Kim (1971, p 967). Tritium measurement by benzene was first reported by Tamers and Bibron (1963). The comparative merit of counting aqueous 3H suspensions (TRITON-XTM or DIOXANETM) are reported by Williams and Florkowski (1967). Direct LS counting of CO2 absorbed onto an alkaline scintillator (eg, LUMA- S)RBTM, CARBO-SORBTM) was reported by Eichinger et al (1980). Scintillation solute. The solute acts as an efficient source of photons after accepting energy from the excited solvent molecules due to ionizing events. The most commonly used solute in 14C LS dating is PPO + POPOP dissolved in toluene (1 part) and added to the sample benzene (4 parts), thus forming a scintillation cocktail (Tamers, 1965b). Following a suggestion from Birks, and based on his research (Birks & Poullis, 1972; Birks, 1974), Polach et al (1983a, p 518) made a comparative study of various suitable scintillators. They concluded that for low-level 3H and 14C dry powder butyl-PBD dissolved in sample benzene (15g/1) gave superior and stable performance even under extreme quench conditions.

Counting Vials The function of a vial is to contain the sample solvent and scintillation solute within the space provided between the two PMTS. It is expected that the vial will have a high photon transmission efficiency, a low self-induced radiation component, and will exhibit no memory. The most common vials are made from borosilicate glass (40K and 226Ra free). Special vials are made from fused silica (quartz), polytetrafluoroethylene (PTFE, TeflonTM) or DelrinTM. Commercially available vials are generally cylindrical and have volumes of 7ml or 22m1. The International Electrotechnical Commission (I EC) recommends standard dimensions for LS vials (height = 60.5 ± 2.5mm, outside diameter = 27 ± 1 mm). There are no known specifications relating to wall thickness, uniformity, weight, or purity of materials. Thus, the most common borosilicate glass vials vary both in efficiency and background count rates, not only based on the source of supply but also within a given batch (Garfin- kel et al, 1965; Painter, 1974). 14C Average sample sizes for dating range from 1 to 15ml of sample benzene. To reduce the background of a 20m1 vial when only partially filled with a sample, the glass vial is either reduced in size and suitably mounted to retain the overall dimensions required by LS counters or a full size vial is masked, exposing only the actual sample volume to the PMTs (Tamers, 1965b; McDowell & Ryan, 1965; Polach, 1969). Calf (1969) established that Teflon, which is not affected by aromatic solvents gave the best results for 3H counting and Calf and Polach (1974) investigated Teflon vials, Viton 4 Henry A Polach `0' ring seals and shielding caps for 14C dating. Since then, Teflon vials of improved designs (eg, Noakes, 1977, p 191; Kuc & Rozanski, 1978; Polach et al (1983a, p 517) have become available commercially and gained popu- larity as they are made to specific size requirements with their active volumes placed in the center of the optical axis between PMTs. Quartz vials are used for very small sample counting by Haas (1979) and very large sample counting by Eichinger et al (1980). Delrin was successfully used by Schotterer and Oeschger (1980). In terms of optimal performance, Quartz, Teflon, and Delrin are of equal merit, with Teflon-copper vials rapidly dis- placing the original borosilicate glass vials. Background Reduction Background is the observed count rate when counting a sample not containing the nuclide to be assayed and preferably no other radionuclide. Sources of background in LS counting have been comprehensively assessed and reviewed by Horroks (1974, p 198-207) and Gupta and Polach (1985, p 56-85). Based on the work of these authors, we can now summarize the sources of LS background and, additionally, review the actions taken to remedy them (Table 2). The sources and remedies group themselves into logical units: 1) care in sample preparation, 2) selection of materials, 3) improved shielding (passive and active), 4) electronic optimization, and 5) other physical parame-

TABLE 2 Sources of background in LS counting and recommended remedial actions Source Remedy Natural radioactivity in the materials of Selection of materials scintillation solute, vial and caps, PMTS and shield Sample contamination by radionuclide to be Care in sample preparation; energy discrimi- assayed or other radioactive material nation, pulse shape analysis Environmental gamma and cosmic radiation Passive and active shielding; simultaneous fl- y counting Introduced radioactive sources used for ex- Thicker shield in counters and directional ternal standardization, automatic quench shielding of source correction, etc Radiation-induced X-rays, Cerenkov and Graded shielding; energy discrimination secondary electron emissions, fast and thermal neutrons Chemiluminescence and phosphorescence Selection of solvents and solutes; elimination of UV activation; delayed coincidence monitor or counting of singles (non- coincident mode) Thermionic and secondary electron emis- Shorter coincidence resolution times, cool- sions (afterpulse) from photocathode and ing of PMTs, high voltage reduction; dead dynodes time of 4µs, selection of PMTs PMT crosstalk due to electric discharge and/ Better shielding (passive and active), better or incident ionizing events optical geometry and elimination of light piping. Lesser pulse analysis or pulse height comparison; opaque counting vials Static electricity Temperature and humidity control, ground- ing and positive/negative ionizer Line noise and transients; RF interference `High Isolation Transformers'; RE pickup and anticoincidence gating Review of Liquid Scintillation Counting for'4C Dating 5 tern. Here we shall concern ourselves with progress made in the last three units. Passive shielding. LS counters manufactured in high volume for biomedical research have no need to consider reductions in background beyond those already achieved (Table 1). Good results are obtained with 5cm (or less) of lead surrounding the vials and PMTs; and results comparable to the average gas counters are obtained readily for 14C dating (Table 3, and Polach, 1974, p 167; Polach et al, 1983b, p 422). Physical constraints imposed by compact designs generally preclude additions of lead by low- level users, although, when this was done the background was significantly reduced. Commercially available counters using massive (>10cm Pb) and incorporating graded attenuation concepts were first designed for 3H applications using large volume vials (>25m1) (Noakes, Neary & Spaulding, 1973; Iwakura et al, 1979). An assymetric graded shield, claiming significant reduction in weight without loss of attenuation, was proposed by Kojola et al (1984, p 494). Active shielding. Not withstanding the attenuation of cosmic and environmental radiations by massive graded shielding, some radiation components, such as muons, reach the counting vial. Anderson, Arnold, and Libby (1951) were the first to apply to 14C dating the concept of an anticoincidence counter (guard) located within the shield. This became standard practice in gas proportional counting (eg, review by Mook, 1983). It was used in LS counting for the first time by Pietig and Scharpenseel (1964) and forgotten until it emerged some 12 years later. Then, a number of experimental approaches were tested. Studies involved 1) placing guard counter(s) externally to the shield (Alessio et al, 1976; Noakes, 1977, p 197), 2) the use of NaI(Tl) guard crystals within the shield (Noakes,1977, p

TABLE 3 Factors leading to best performance in 14C dating by LS spectrometry

Function Best performance is achieved by C2H2 synthesis Lithium carbide hydrolysis C6H6 synthesis C2H2 trimerization over Vanadium catalyst CO2 gas dilution prior to synthesis of small samples Fluor Butyl-PBD (15g/1) dissolved in sample C6H6 No UV activation prior to counting Counting vial Teflon, quartz, Delrin, graded in size to suit sample volume, and shielded with copper or lead caps Counting system shielding 1) Passive, assymetric, graded, Pb, Cd, Cu 2) Active, LS guard, assymetric, covering vials as well as PMT assembly Electronics Optimized, high voltage reduction or HI-LO coincidence bias, spectral stabilization, cross talk suppression (lesser pulse or amplitude disparity), RF pickup into A/cos, static noise suppression, cooled counting chamber Environment Shielded room (1.5m concrete), air conditioned and temp 17°C and RH 40% controlled, power line transients isolation, clean `electrical earth' with no other users, ventilated counting chamber Data analysis `Random Access' sample, BKG and STD cycle Multiparameter multichannel `Windowless' software Data acquisition Online dedicated personal computer, floppies and hard disk 6 Henry A Polach

198; Iwakura et al, 1979), and 3) the use of gas, plastic, or LS guards within the shield (Pietig & Scharpenseel, 1964; Punning & Rajamae, 1975; Broda & Radoszewski, 1982; Jiang et al, 1983; Kojola et al, 1984). Systems 2) and 3) achieve the most significant background reductions; the best of these 72-84%, with anticoincidence on. Some of these are commercially available and, because of their very high merit, find favor with newly established laboratories or as replacements of aging gas or liquid counters. Electronic optimization. Electronic optimization, other than the anticoincidence guard, aims to reduce noise (hence background) or increase the long-term stability of LS counters. Involved are high voltage or spectral stabilization (Soini, 1 975a; Berthold,1980), high voltage reduction or variable (HI-LO) coincidence bias (Polach et al, 1983b, p 425-434), chemiluminescence monitor (Soini,1977), lesser pulse analysis (Laney,1971), amplitude disparity discrimination (Soini, 1975b), RF pickup and channeling the amplified signal into the anticoincidence inhibit line. Other physical parameters. Performance of LS counters is also enhanced (increased stability and/or reduction of background) when consideration is given to other physical parameters. Thus, balance point counting, selection of PMTs, considerations of light reflecting properties, and physical shape of sample chamber (Butterfield & Polach, 1983; Erikson, Winn & Hor- rocks, 1983), suppression of line noise (transients) by usage of High Isola- tion Transformers, elimination of static charge by positive and negative ionizers (Kananen et al, 1984), cooling of PMTs, and temperature and humidity control of rooms in which LS counters are placed. Choice of counting vials and scintillation cocktail is already dealt with in preceding chapters. A detailed review of optimization of electronic and physical parameters affecting modifications to old counters and design parameters of new LS counters is given by Gupta and Polach (1985, p 47-98). Data Processing Evaluation and Validation Until the mid-1980s LS and gas counters were just that; they counted the number of events occurring in the region of isotope interest. In gas counting of 3H and 14C, this includes all events occurring within a region (a window) defined by a lower limit discriminator (LLD), set just above electronic noise, and infinity. In LS counting, due to good energy resolution of these isotopes, the window can be restricted to cover a selected percentage of the specified isotope(s) energy range; for 14C dating, this is typically a 60-80% efficiency window. Such procedures resulted in total gross counts, which could be displayed on sealers and output on lister/printers or other data storage devices. Due to excellent energy resolution of LS counters, these, almost from the onset, included some data evaluation parameters such as channel ratio, external standardization, etc. Even so, once a pulse fell within the predefined LS energy window, its energy information was lost as only its presence was recorded as a gross count. The loss of energy information is significant, in terms of data validation, particularly in low- level counting systems, be they based on gas or liquid ,3-detection. Whilst the power of energy resolution in LS counting is inherent to the method, 14C Review of Liquid Scintillation Counting for Dating 7 irrespective of sample size, the need for it, hence development of suitable technology, did not surface until the mid-1980s. Then external multichannel analyzers (MCAs) were interfaced to existing counters with some suc- cess (eg, Gordon et al, 1976). The first interactive and inbuilt MCA-supported LS spectrometer (note shift from word `counter') was reported by Ring, Nguyen and Everett (1980). The MCA-based energy analysis, coupled with microprocessor control and captive online minicomputers (PCs) with data evaluation software for biomedical research, is now available in the upper range models of LS spectrometers, from every major nuclear instru- ment manufacturer. The usefulness of MCA analysis for radiocarbon was first evaluated by Polach et al (1984b). Using an external MCA interfaced to existing LS counters, the authors proposed `Windowless' 14C age determinations and data validation based on software programs they had developed for this purpose. The first spectrometer specially built for low-level counting, incorporating not one but two 2000-channel resolution MCAs (`multiparameter LS spectrometry'), as well as the 14C dating data evaluation software, was built by Polach ei al (1984a). It seems that their counter represents, today, state-of-the-art 14C dating technology.

RESULTS An international survey, initiated in 1984 by Lauri Kaihola, University of Turku, Finland, reviews the field of low-level LS counting and spectrometry especially for 14C dating. In this area the response was almost total. 3H dating was less well covered; the field was not so well known to us and our

TABLE 4 Performance comparison of gas proportional and LS counting systems

Sample Grams C S/B* features

CO2 1.5 19 lab, graded shield, A/cos 1 28 As above but triple A/cos guard 33 U/ground, triple A/cos guard 37 Extrapolated best U/G, triple A/Cos CO2 2.1 40 y-free materials 2 5.8 59 As above CH4 1.5 43 my-free materials 3 CO2 2 to 6 4 to 10 to average systems 4 C6H6 2 to 4 4 to 8 glass, set to manufacturer's specs 5 1 to 5.7 <30 Teflon-copper vials, electronics optimum 4.3 12 EHT low, PMT sel, heavy shield > 10cm 6 4.3 68 As above but deep underground 0.08 5 3m1 T/Cu vial, anti/cos, assymmetric shield 7 0.4 24 3ml T/Cu 0.8 47 3m1 T/Cu 2.4 141 3ml T/Cu 5.7 116 7ml T/Cu 12.2 141 15m1 T/Cu 16.2 141 20ml T/Cu * S/B = signal, 0.95 NBS Ox/background ** 1) (;ulliksen & Nydal, 1979, p 181; 2) Stuiver, Robinson & Yang, 1979, p 213; 3) Oeschger et al, 1979, p 157; 4) Mook,1983; 5) Polach et al, 1983b, p422; 6) Calf & Airey, 1982, p 355; 7) Kojola et al, in press. 8 Henry A Polach inquiry sheets did not necessarily reach the right audience. Hence, the comments that follow are restricted to 14C LS dating. When comparisons to gas counting are being made, as they must be, they are based on published and cited references. Based on the survey it can be stated that 14C LS dating is a fast growing field. Since 1975 the majority (some 94%) of newly established dating laboratories have opted for an LS system. Within these laboratories changes are taking place even today, as the LS a-detection methodology and instrumentation are being continuously improved. Thus, it was not considered relevant to tabulate `who uses what and how' (although this was the initial intention). Such statistical analyses are no longer relevant; instead, a list of steps involved in LS counting, which in the hands of users were most successful, is given in Table 3, and a comparison of results of average and best gas and liquid systems is given in Table 4. Literature reference at this point shall be cited only if they were not already cited in the text above, or if required as proof. The very high S/B of the best of LS systems, in an above ground environment, with a flexibility to count variable size samples from 80mg to 16g is unique in the history of radiocarbon dating. It has been achieved by min- imizing every known factor contributing to background, a concept of total optimization.

PROSPECTS AND CONCLUSIONS A window to see into the future is given but to a few; hence, predic- tions as to future trends will be made on promising research in hand rather than anticipated research development. Even so, it must be stated that the present progress which has achieved a background reduction of 84% in LS counting and raised its Figure of Merit from 500 to 32,000 for 14C can no longer be duplicated. Improvements will follow; they will be based on user and peer group evaluation and interaction with manufacturers and on the commercial viability of the suggestions. Promising are simultaneous a-f3-y LS spectrometry, pulse shape, and rise time analyses to distinguish origin, not only energy, of ionizing events. This may involve NaI(Tl) guards. Already background reductions, using NaI(Tl) guards alone have been achieved equivalent to 5cm of Pb (Noakes, Neary & Spaulding, 1974). Does this lead to small portable, typewriter-size field counters, using CO2 dissolved in a fluor for first order of magnitude field dates? Perhaps more in demand shall be counters using Microchannel Plate PMTs, the small size, inherent low noise, and high gain of which are ideal for smaller, lighter shielding geometry, hence, leading to very cost- effective systems (Spaulding & Noakes, 1983). Certainly, there shall be developments of liquid scintillators with increased sensitivity to y-radiation in the low energy range to act as anticoincidence guards; but then we are nibbling at a few percent of background reduction, worthwhile only for the sake of excellence. Multiparameter multichannel analysis with assymetric shielding leads to high-precision low-level LS counting. This system is with us now. It opens a new and unprecedented dimension in 14C dating. It promises to open new '4C Review of Liquid Scintillation Counting for Dating 9 applications in other fields. I believe that the rules of low-level counting, as applied to '4C dating and related fields, have already been rewritten.

ACKNOWLEDGMENTS The survey, initiated by Lauri Kaihola, University of Turku, made it possible to view the LS low-level counting field in perspective. I thank all those who answered, particularly the radiocarbon dating community, for their complete response. Dilette Polach, Garran ACT, assisted in data evaluation of the survey and tabulated the now superseded information it contained. Her bibliographic files assisted in checking the numerous references. Further, she proofread this text.

REFERENCES Alessio, M, Allegri, L, Bella, F and Improta, S, 1976, Study of background characteristics by means of high efficiency liquid scintillation counter: Nuclear Instruments & Methods, v 137, p 537-543. R, Arnold, J 1958, Archaeology and chemistry, in Bell, C G, Jr and Hayes, F N, eds, Liquid scintillation counting: London, Pergamon Press, p 129. Anderson, E C, Arnold,) R and Libby, W F,1951, Measurement of low level radiocarbon: Rev Sci Instruments, v 22, p 225-230. Barker, H, 1953, Radiocarbon dating: large scale preparation of acetylene from organic material: Nature, v 172, p 631-632. Berthold, F, 1980, A new approach to automatic photornultiplier stabilization for photon and scintillation counters, in Peng, C-T, Horrocks, D L and Alpen, E L, eds, Liquid scintillation counting, recent applications and development: New York, Academic Press, p 273- 280. Birks, J B, 1974, Towards an understanding of the scintillation process in organic molecular systems, in Stanley, P E and Scoggins, B A, eds, liquid scintillation counting, recent developments: London, Academic Press, p 1-38. Birks, J B and Poullis, G G,1972, Liquid scintillators, in Crook, M A, Johnson, P and Scales, B, eds, Liquid scintillation counting: London, Heyden & Sons, v 2, p 1-21. Broda, R and Radoszewski, T, 1982, Scintillation detector with anticoincidence shield for determination of radioactive concentration of standard solutions, in Povinec, P, ed, Low- level counting, Internatl conf on low radioactivities 1980, Proc: Bratislava, VEDA, v 8, 329-333. 14C Burleigh, R, Hewson, A D and Matthews, K J, 1977, Synthesis of benzene for low-level measurement: a review, in Crook, M A and Johnson, P, eds, Liquid scintillation counting: London, Heyden & Sons, v 5, p 205-209. Butterfield, D and Polach, H A, 1983, Effect of vial holder materials and design on low-level 14C scintillation counting, in McQuarrie, S A, Ediss, C and Wiebe, L I, eds, Advances in scintillation counting: Alberta, Univ Alberta Press, p 468-477. Calf, G E, 1969, Teflon vials for liquid scintillation counting of tritium samples: Internatl Jour Appl Radiation Isotopes, v 20, p611-612. Calf, G E and Airey, P L, 1982, Liquid scintillation counting of carbon-14 in a heavy shielded site, in Ambrose W and I)uerden, P, Archaeometry: an Australian perspective: Canberra, ANU Press, p 351-356. Calf, G E and Polach, H A, 1974, Teflon vials for liquid scintillation counting of carbon-14 samples, in Stanley, P E and Scoggins, B A, eds, Liquid scintillation counting, recent developments: London, Academic Press, p 223-234. Eichinger, L, Rauert, W, Salvamoser, J and Wolf, M, 1980, Large-volume liquid scintillation counting of carbon-14, in Stuiver, M and Kra, R S, eds, Internatl 14C conf, 10th, Proc: Radiocarbon, v 22, no. 2, p 417-427. Erikson, K R, Winn, R W and Horrocks, D L, 1983, The optics of LS counting, in McQuarrie, S A, Ediss, C and Wiebe, L I, eds, Advances in scintillation counting: Alberta, Univ Alberta Press, p 90-105. Garfinkel, S B, Manni, W B, Medlock, R W and Yura, 0, 1965, The calibration of the National Bureau of Standards tritiated standard of radioactivity: Internatl Jour Appl Radiation Isotopes, v 16, p 27-33. Cordon, B E, Press, M, Erwin, W and L,emmon, R M, 1976, An interface for routine spectral display from several liquid scintillation counters, in Nouajaim, A A, Ediss, C and Wiebe, I. I, eds, Liquid scintillation counting science and technology: New York, Academic Press, p 173-183. 10 Henry A Polach

Gulliksen, S and Nydal, R, 1979, Further improvement of counter background and shielding, in Berger, R and Suess, H, eds, Radiocarbon dating, Internatl 14C conf, 9th, Proc: Berke- ley/Los Angeles, Univ California Press, p 176-184. Gupta, S K and Polach, H A, 1985, Radiocarbon dating practices at ANU: Garran ACT, Radiocarbon Dating Research. Haas, H, 1979, Specific problems with liquid scintillation counting of small benzene volumes and background countrate estimation, in Berger, R and Suess, H, eds, Radiocarbon dat- 14C ing, Internatl conf, 9th, Proc: Berkeley/Los Angeles, Univ California Press, p 246- 255. Horroks, D S, 1974, Application of liquid scintillation counting: New York, Academic Press. Iwakura, T, Kasida, Y, Inoue, Y and Tokunaga, N, 1979, A low background liquid scintillation counter for measuring low level tritium, in Behaviour of tritium in the environment: Vienna, IAEA, p 163-171. Jiang, H, Lu, S, Fu, S, Zhang, W, Zhang, T, Ye, Y, Li, M, Fu, P, Wang, S, Peng, C and Jiang, P, 1983, Model DYS low-level liquid scintillation counter, in McQuarrie, S A, Ediss, C and Wiebe, L I, eds, Advances in scintillation counting: Alberta, Univ Alberta Press, p 478- 493. Kananen, K, Ala-Uotila, M, Oikari, T and Soini, E, 1984, A study of the effect of humidity and the use of an ioniser on static electricity using an LKB-Wallac 1211 RackBeta liquid scintillation counter: Turku, Wallac Rept, p 3-9. Kim, S M,1971, Low level liquid scintillation counting and evaluation of counting solutions of 14C and 3H, in Horrocks, D L and Peng, C-T, eds, Organic scintillators and liquid scintillation counting: London, Academic Press, p 965-976. Kojola, H, Polach, H, Nurmi, J, Heinonen, A, Oikari, T and Soini, E, in press, Low level liquid scintillation spectrometer for /9-counting, in Nordic conf on application of scientific methods in archaeology, Proc: ISKOS. Kojola, H, Polach, H, Nurmi, J, Oikari, T and Soini, F, 1984, High resolution low level liquid scintillation f3-spectrometer: Internatl Jour Appl Radiation Isotopes, v 35, p 949-952. Kuc, T and Rozanski, K,1978, A small volume Teflon-copper vial for 4C low level liquid scintillation counting: Internatl Jour Appl Radiation Isotopes, v 30, p 452-454. Laney, B H, 1971, Electronic rejection of optical cross talk in twin phototube scintillation counters, in Horrocks, D L and Peng, C-T, Organic scintillators and liquid scintillation counting: New York, Academic Press, p 991-1003. McDowell, L L and Ryan, M F, 1965, USDA Sedimentation Laboratory radiocarbon dates I: Radiocarbon, v 7, p 174-178. Mook, W G,1983, International comparison of proportional gas counters for 14C activity measurements, in Stuiver, M and Kra, R S, eds, Internatl 14C conf, l l th, Proc: Radiocarbon, v 25, no. 2, p 474-484. Noakes, J E, 1977, Considerations for achieving low level radioactivity measurements with liquid scintillation counters, in Crook, M A and Johnson, P, eds, Liquid scintillation counting: London, Heyden & Sons, v 4, p 189-206. Noakes, J E, Isbell, A F, Stipp, J J and Hood, D W, 1963, Benzene synthesis by low temperature catalysis for radiocarbon dating: Geochim et Cosmochim Acta, v 27, p 797-804. Noakes, J E, Kim, S M and Stipp, J J,1965, Chemical and counting advances in liquid scintillation counting, in Chatters, R M and Olson, E A, eds, Internatl 14C conf, 6th, Proc: Clear- inghouse for Fed Sci & Tech Inf, Natl Bur Standards, Washington, DC, p 68-92. Noakes, J E, Neary, M P and Spaulding J D, 1973, Tritium measurements with a new liquid scintillation counter: Nuclear Instruments & Methods, v 109, p 177-187. 1974, A new liquid scintillation counter for measurement of trace amounts of 3H and 14C, in Stanley, P E and Scoggins, B A, eds, Liquid scintillation counting, recent developments: London, Academic Press, p 53-66. Oeschger, H, Lehmann, B, Loosli, H H, Moell, M, Neftel, A, Schotterer, U and Zumbrunn, R, 1979, Recent progress in low level counting and other isotope detection methods, in Berger, R and Suess, H, eds, Radiocarbon dating, Internatl 14C conf, 9th, Proc: Berkeley/ Los Angeles, Univ California Press, p 147-157. Painter, K, 1974, Choice of counting vials for liquid scintillation: a review, in Stanley, P E and Scoggins, B A, eds, Liquid scintillation counting, recent developments: London, Aca- demic Press, p431-45l. Panarello, H 0, Albero, M C and Angiolini, F E, 1983, Stable isotope fractionation during benzene synthesis for radiocarbon dating, in Stuiver, M and Kra, R S, eds, Internatl i4C conf, 11th, Proc: Radiocarbon, v 25, no. 2, p 529-532. Pietig, von, F and Scharpenseel, H W, 1964, Alterbestimmung mit Flussigkeits-Scintillations- Spectrometer. Uber die Wirksamkeit von Abschirmungsabnahme: Atompraxis, v 7, p 1- 3. Polach, H A, 1969, Optimisation of liquid scintillation radiocarbon age determinations and reporting of results: Atomic Energy in Australia, v 12, p 21-28. Review of Liquid Scintillation Counting for 14C Dating 11

Polach, H A, 1974, Application of liquid scintillation spectrometers to radiocarbon dating, in Stanley, P E and Scoggins, B A, eds, Liquid scintillation counting, recent developments: London, Academic Press, p 153-171. Polach, H A, Gower, J and Frazer, I, 1972, Synthesis of high purity benzene for radiocarbon dating by the liquid scintillation method, in Rafter, T A and Taylor, T, eds, Internatl 14C conf, 8th, Proc: Wellington, Royal Soc New Zealand, v 1, p 145-157. Polach, H A, Gower, J, Kojola, H and Heinonen, A, 1983a, An ideal vial and cocktail for low- level scintillation counting, in McQuarrie, S A, Ediss, C and Wiebe, L I, eds, Advances in scintillation counting: Alberta, Univ Alberta Press, p 508-525. Polach, H A, Kojola, H, Nurmi J and Soini, E,1984a, Multiparameter liquid scintillation spectrometry, in Wolfli, W, Polach, H and Andersen, H H, eds, Accelerator mass spectrometry, AMS 1984: Nuclear Instruments & Methods, v 233{B5], p 439-442. Polach, H A, Nurmi, J, Kojola, H and Soini, E, 1983b, Electronic optimisation of scintillation counters for detection of low-level 3H and 14C, in McQuarrie, S A, Ediss, C and Wiebe, L I, eds, Advances in scintillation counting: Alberta, Univ Alberta Press, p 420-441. Polach, H A, Robertson, S, Butterfield, D, Gower, and Soini, E, 1984b, The `Windowless' 14CJ approach to scintillation counting: low-level as an example, in McQuarrie, S A, Ediss, C and Wiebe, L I, eds, Advances in scintillation counting: Alberta, Univ Alberta Press, p 494-507. Polach, H A and Stipp, J J, 1967, Improved synthesis techniques for methane and benzene radiocarbon dating: Internatl Jour Appl Radiation Isotopes, v 18, p 359-364. Punning, J M and Rajamae, R, 1975, Some possibilities for decreasing the background of liquid scintillation beta-ray counters, in Povinec, P and Usacev, S, eds, Low-radioactivity measurements and applications: Bratislava: Slov Pedagog Nakladatelstvo, p 169-171. Ring, J C, Nguyen, D C and Everett, L J,1980, Liquid scintillation counting from gross counts to spectral analysis, in Peng, C-T, Horrocks, D L and Alpen, E L, Liquid scintillation counting, recent applications and development: New York, Academic Press, v 1, p 89- 104. Schotterer, U and Oeschger, H, 1980, Low-level liquid scintillation counting in an underground laboratory, in Stuiver, M and Kra, R S, eds, Internatl 14C conf, 10th, Proc: Radio- carbon, v 22, no. 2, p 505-511. Soini, E, 1975a, Stabilisation of photomultiplier tubes in liquid scintillation counters: Rev Sci Instruments, v 46, p 980-984. 1975b, Rejection of optical cross-talk in photomultiplier tubes in liquid scintillation counters: Turku, Finland, Wallac Rept, p 1-9. 1977, Chemiluminescence monitoring in liquid scintillation counting: Turku, Finland, Wallac Rept, p 3-8. Spaulding, J D and Noakes, J E, 1983, An evaluation of microchannel plate photomultipliers for liquid scintillation counting, in McQuarrie, S A, Ediss, C and Wiebe, L I, eds, Advances in scintillation counting: Alberta, Univ Alberta Press, p 112-122. Stank, I E, Arslanov, Kh A and Klener, I R, 1963, Improved procedure for chemical preparation of samples for radiocarbon dating by the scintillation method: Soviet Radiochemis- try, v 5, 174-180. p 14C Stuiver, M, Robinson, S W and Yang, I C, 1979, dating to 60,000 years BP with proportional counters, in Berger, R and Suess, H, eds, Radiocarbon dating, Internatl ' C conf, 9th, Proc: Berkeley/Los Angeles, Univ California Press, p 202-215. Suess, H E, 1954, Natural radiocarbon measurements by acetylene counting: Science, v 120, p 5-7. Swart, E R, 1964, The direct conversion of wood charcoal to lithium carbide in the production of acetylene for radiocarbon dating: Experientia, v 20, p 47-48. Tamers, M A, 1965a, Chemical yield optimisation of benzene synthesis for radiocarbon dating: Internatl Jour Appl Radiation Isotopes, v 26, p 676-682. 1965b, Routine carbon-14 dating using liquid scintillation techniques, in Chat- ters, R M and Olson, E A, eds, Internatl' C conf, 6th, Proc: Clearinghouse for Fed Sci & Tech Inf, Natl Bur Standards, Washington, DC, USAEC, CONF-650652, p 53-67. 1975, Chemical yield optimisation of benzene synthesis for radiocarbon dating: Internatl Jour Appl Radiation Isotopes, v 26, p 676-682. Tamers, M A and Bibron, R, 1963, Benzene method measures tritium in rain without isotope enrichment: Nucleonics, v 21, p 90-94. Williams, P H and Florkowski, T, 1967, Comparison of Triton-X emulsion systems with diox- ane solutions in liquid scintillation counting of low-level tritium, in Radioactive dating and methods of low-level counting: Vienna, IAEA, p 703-709. [RADIOCARBON, VOL 29, No. 1, 1987, P 12-17] SCINTILLATION COUNTER PERFORMANCE AT THE SMU RADIOCARBON LABORATORY' JAMES M I)EVINE and HERBERT HAAS Radiocarbon Laboratory, Southern Methodist University Dallas, Texas 75275

ABSTRACT. Results are presented of a study of counter performance and vial characteristics for three liquid scintillation counters used at the SMU Radiocarbon Laboratory: the Inter- technique LS20, Packard Tri-Garb 460C, and LKB Wallac Rack Beta 1217. Modifications to photomultiplier tube high voltage, pre-amplifier gain, energy window settings, counting vial design, and sample holder design have resulted in reduced background, higher counting efficiency, and greater long-term stability for the Intertechnique and Packard counters. Square quartz counting vials are used in the Intertechnique and Packard counters with excellent results. Use of Teflon vials in the LKB counter requires careful cleaning procedures and long counting times.

COUNTER AND VIAL CHARACTERISTICS The SMU Radiocarbon Laboratory uses three liquid scintillation counters: an Intertechnique L520, a Packard Tri-Garb 4600, and the low-background "Kangaroo" version of the LKB Rack Beta 1217. The Intertechni- que is a benchtop non-cooled model with the following characteristics: 1) carefully selected photomultiplier (PM) tubes (RCA 4501V3) with low background and high counting efficiency at reduced dynode voltage, 2) a shielded PM tube assembly made from low-activity lead, 3) a modified single-sample manual loading drawer that accommo- dates quartz vials holding 3, 1.5, 0.6, and 0.3m1 counting solutions (Haas, 1979). Temperature sensitivity of counter components requires careful control of laboratory environment. It has been observed that background count rates increase 0.1-0.4cpm during high-temperature excursions, as timing oscillator and energy window settings are temperature dependent. The counter has enjoyed long-term stability when laboratory temperature has been kept stable (± 1°C). The counter has been modified by removal of the'37Cs standard from the cabinet and by replacement of variable potenti- ometers by fixed resistors for energy window adjustments. The Packard counter is a floor-standing model with built-in cooling. Cooling is disabled, however, to avoid moisture condensation within the counter. Sample loading is automated, and 15 programs can be stored for different 10-sample trays. At present, a separate program (tray) is used for each of 4 counting vials (3 at 3m! and 1 at 1.5m1 filling volume). Several modifications were made to this counter: 1) Following a one-year trial period, the vial lift and shutter mechanisms were modified to accept square quartz vials. These vials are inserted in a round black nylon holder with the same outside dimensions as a standard 4 5ml vial. In the cylindrical wall of the holder are two opposed windows through which PM tubes view the vial. Proper alignment of the holder

* This paper is a modified version of the one presented at the Twelfth International Radiocarbon Conference in Trondheim, Norway, June 24-28, 1985.

12 Scintillation Counter Performance at SMU 13 is maintained with a slot along its outside and several guide pins in the lift tube and sample chamber. 2) The original analog amplifier board was replaced by Packard Instru- ments with a design having greater stability and adjustable gain. Gain was boosted to spread the energy spectrum and to separate the background and 14C peaks for energy window adjustment. PM tube high voltage (HV) was decreased to the lower edge of the weakly developed plateau of the PM tube HV vs counting efficiency curve (Fig 1). 3) A focusing and masking insert (similar to the design of Butterfield & Polach, 1983) was placed before each PM tube to reduce crosstalk and reflect scintillation photons to the center of the PM tube front dynode. 4) The 226Ra external quench standard was removed since the channel ratio method, rather than external standardization, is used for quench determination. External standardization is not used due to possible PM tube and counting solution memory effects following standard-generated counts, andsto questions regarding the effectiveness of internal shielding against the Ra standard. 5) Frosted quartz vials replaced clear vials, with resultant background decrease (Table 1) due to crosstalk reduction. Improvements in counter performance with each modification are demonstrated by increased counter efficiency and decreased background (Table 1). Counting solutions comprise benzene with 0.9100 weight percent butyl-PBD scintillant.

>+ 85 cV U w 0 80 01 C C 0 C1 75

16 17 18 19 PM tube high voltage (x100 V) Fig L Efficiency vs PM tube HV for Packard 4600. HV settings for front and rear PM tubes = (0). 14 James M Devine and Herbert Haas

TABLE Counter and vial response characteristics

Vial Background Efficiency Counter (ml) (cpm) (%) E2/B* Intertechnique LS20 3.0 2.6 67.8 RCA 4501 V3 PM tubes 1.5 2.2 68.6 Clear square quartz vials 0.6 2.1 70.1 0.3 1.9 68.6

LKB Rack Beta 1217 3.0 1.8 71.6 Copper/Teflon vials 7.0 3.8 80.8 CBC option Packard Tri-Garb 460C RCA 4501V3 PM tubes 1. Initial testing 15.0 4.4 76.9 Cylindrical glass vials 15.0 masked 4.9 to 3ml vol 2. First upgrade A ch Clear square quartz 0 B ch 2.0 52.9 1400 12.2 vials, flat reflectors 1.5 A ch 5.0 B ch 1.4 50.9

3. Present setup A ch 5.4 82.7 Frosted quartz vials in 3.0 B ch 2.7 black nylon holders, curved reflectors 4.7 ch 2.0 58.0

E2/B Figure of merit = (Counting efficiency2)/Background cpm ** Nox/B Radiocarbon Factor of merit = (0.95 Mod std cpm//Background cpm)

The LKB Rack Beta 1217 counter (low-background "Kangaroo" version), a floor-standing non-cooled model, had acceptable performance characteristics as delivered; however, counting channels were reset from default values to windows allowing maximum' C detection efficiency using Teflon vials. The LKB Teflon/copper 3m1 counting vials required redesign to promote easier sample weighing and reduced sample loss. The original vial design was changed with the addition of a lightweight Teflon screw cap to enable weighing on our analytical balance. The seal between the Teflon cap and vial is maintained by a Teflon disk backed by silicone. Tests have shown that the Teflon vial/cap assembly absorbs on occasion ca 0.5mg over a two-day counting period, presumably due to laboratory humidity fluctua- tions. Observations of vial memory during alternating runs of background and modern standards (0.5cpm background increase following modern standard runs) resulted in our adoption of a cleaning procedure more extensive than that recommended by the manufacturer; vials are rinsed four times in benzene, and vacuum-dried at 60°C for two hours. Vials are then placed in a desiccator until needed for counting, and allowed to equil- ibrate one hour before filling. Filled vials rest in the counter for at least 1.5 days before counting begins because of their high sensitivity to light, and require a much longer time to stabilize at the start of counting than quartz vials. The first five 100-minute counts are routinely omitted to remove effects of initial erratic data. Given these constraints, Teflon vials provide Scintillation Counter Performance at SMU 15 low background and high counting efficiency and figure of merit (Table 1; Gupta & Polach, 1983).

ENERGY WINDOW SELECTION Energy window adjustment is a necessary procedure in preparation of a counter for routine operation. The energy distribution of background and 14C pulses differs for each counter and different considerations apply in the choice of lower and upper limits for 14C counting. Spectrum measurements are made easily in the Packard counter, which allows immediate access to any stored channel or group of channels, enabling listing of a complete spectrum. Spectrum measurement in the Intertechnique and LKB counters is time consuming, however, as data must be collected for each channel or group of channels before proceeding to the next set of channels. It has been demonstrated that use of an external multichannel analyzer (Polach et al, 1983b) hastens this procedure appreciably. The Intertechnique counter has a logarithmic amplification stage and presents a nearly symmetric 14C spectrum (Fig 2A). The background curve is nearly flat; thus, window settings are determined wholly by the 14C curve. A wide window records counts in the entire spectrum; setting of a narrow window on the steep high energy flank of the curve acts as a sensitive quench indicator (Baillie, 1960; Gupta & Polach, 1983). A curve shift to lower energy due to quenching causes a proportionally larger drop in measured output in this narrow channel compared to energy loss in the wide window setting. Thus, the two-channel ratio method makes application of an external radioactive source unnecessary. The 14C curve of the LKB counter (Fig 2B) is similar to that of the Intertechnique counter. The background spectrum is flat in the region of interest and rises steeply slightly beyond endpoint. As in the Intertechni- que, two windows are set for quench monitoring. The Wallac amplitude disparity discrimination (CBC) option, based on coincidence timing delay via shaping of PM tube pulses (Soini,1975), accepts a pulse only if pulse height 14() ratio is < 2:1 (for within the coincidence resolution interval (Polach et al, 1983a). Use of CBC halves background count rate and reduces counting efficiency ca 8%. Energy window settings and background cpm are identical for each Teflon vial in use, given strict cleaning and cpm monitoring procedures. The Packard counter presents linearly amplified spectra compared to logarithmic output of the Intertechnique and LKB counters. Initial tests revealed that the sharp background and 14C peaks overlapped completely (Fig 2C); it was not possible to reduce background without substantial 14C count loss. Increase of pre-amplifier gain and reduction of PM tube HV lowered the background peak and separated, partially, background and 14C peaks; introduction of square quartz vials and changes in the optical geometry of the counting chamber spread the 14C spectrum (Fig 2D). The present arrangement sets the lower A-channel window limit within the wide 14C peak (Fig 2E); the background peak lies slightly below the lower window limit. Some background reduction was achieved by frosting the sides of the square quartz vials, which dampens crosstalk. B-channel lower limit is set so 2A. Intertechnique LS 20 2B. LKB Rack Beta 1217 40 40

so 10 iso channel

2C. Packard 460C -initial 2D. Packard 460C -first upgrade 2E. Packard 460C -present A A 40 40 40

a B U B 20

0 40 80 120 0 40 80 120 0 40 80 120 channel channel channel

(---) Fig 2. Modern standard ( ) and background spectra for counters. Window settings = (F-). Scintillation Counter Performance at SMU 17 that the majority of the background peak is distinctly below the lower window limit. Part of the 14C spectrum is excluded, and the upper window limit is positioned near the endpoint of the 14C curve. B-channel counting efficiency is reduced (Table 1), and count rate is sensitive to quenching. A dis- crepancy in sample age calculated from A and B channels (B-channel age > A-channel age) is, thus, a sensitive quench indicator.

SUMMARY Several modifications have optimized our three liquid scintillation counters for routine 14C dating. Reduction of PM tube HV to decrease dark current noise coupled with increase in gain to boost counting efficiency, improvements in optical geometry for the use of square quartz vials, use of low background materials for sample holders (lead, black nylon), and energy window adjustment for quench detection have yielded reduced background and greater stability in counter operation.

ACKNOWLEDGMENTS The support of NSF grant BNS 8211974 is gratefully acknowledged. We thank H Polach and R Kahn for their critical comments.

REFERENCES Baillie, I. A, 1960, Determination of liquid scintillation counting efficiency by pulse height shift: Internatl Jour Appl Radiation Isotopes, v 8, p 1-7. Butterfield, l) and Polach, H,1983, Effects of vial holder materials and design on low-level C14 scintillation counting, in McQuarrie, S A, Ediss, C, and Wiebe, L I, eds, Advances in scintillation counting: Edmonton, Univ Alberta Press, p 468-477. Gupta, S K and Polach, H, Radiocarbon dating practices at ANU: Research School Pacific Studies, ANU, Canberra, 173 p. Haas, H. 1979, Specific problems with liquid scintillation counting of small benzene volumes and background count estimation, in Berger, R and Suess, H E, eds, Radiocarbon dating, Internatl 4C conf, 9th, Proc: Berkeley, Univ California Press, p 246-255. Polach, H, Nurmi, Kojola, H, and Soini, E, 1983a, Electronic optimisation of scintillation J, C14, counters for detection of low-level H3 and in McQuarrie, S A, Ediss, C, and Wiebe, L A, eds, Advances in scintillation counting: Edmonton, Univ Alberta Press, p 420-441. Polach, H, Robertson, S, Butterfield, D and Gower, 983b, The "windowless" approach to C14 J, scintillation counting: low-level as an example, in McQuarrie, S A, Ediss, C, and Wiebe, L 1, eds, Advances in scintillation counting: Edmonton, Univ Alberta Press, p 494-507. Soini, E, 1975, Rejection of optical crosstalk in photomultiplier tubes in liquid scintillation counters: Wallac rept, Turku, Finland, 9 p. [RADIOCARBON, VOL 29, No. 1, 1987, P 18-23]

RADIOCARBON CONTENT OF TROPOSPHERIC CO2 AT CHINA LAKE, CALIFORNIA 1977-1983 RAINER BERGER Institute of Geophysics and Departments of Geography and Anthropology, University of California, Los Angeles, California 90024 T B JACKSON, ROBERT MICHAEL, and H E SUESS Mount Soledad Laboratory University of California, San Diego La Jolla, California 92093 ABSTRACT. The measurements reported here are a continuation of tropospheric radiocarbon measurements in carbon dioxide carried out since 1961 at our China Lake, California collection facility. The data show a continued decrease in radiocarbon activity from ca 3300/oo in 1977 to 215%o in 1983 in agreement with similar analyses in Europe for the same time interval.

INTRODUCTION The measurements reported are part of a continuing study of the I4C content of tropospheric CO2 in the Northern Hemisphere. The program began in 1961.at China Lake, California (35° 32' N, 1170 41' W). The sampling location is exposed to clean maritime air in the summer while in the winter there can be episodes of incursions of continental air via Santa Ana winds. The site was chosen to avoid contaminated air from anthropogenic sources such as the metropolitan Los Angeles area or other major produc- ers of fossil fuel emissions. Results of measurements of prior years were reported previously (Berger & Libby, 1966, 1967, 1968; Berger, Fergus- son & Libby, 1965). In the early 1960s, a pronounced maximum in the 14C content of CO2 in the air occurred approaching a factor of two above the prebomb level. Since then, the 14C content has steadily decreased, due to the uptake of 14C bomb primarily by the world's oceans (Keeling, 1979). This decrease is interesting because it provides information on residence time of CO2 in the atmosphere and allows the dating of relatively young plant material. Tropospheric carbon dioxide undergoes exchange with three main reservoirs on time scales of interest to mankind: the stratosphere, the bio- sphere, and the oceans. Most bomb 14C is injected into the stratosphere (Berger, 1979). Hence, only mixing with stratospheric CO2 can increase 14C in the troposphere. The oceanic reservoir is a complex system containing the bulk of bomb-produced 14C; it is actually a series of interrelated reservoirs with varying time scales for exchange. Most of the upper layers equili- brate with the 14C of the bomb transient on a decadal time scale. The time scales for equilibration with most deeper waters are longer, on the order of centuries (Linick, 1980; Bien, Rakestraw & Suess, 1965).

METHODS Samples were collected in a photographic tray ca 20 x 40 x 6cm containing 2L of 2N carbonate-free sodium hydroxide. This tray is located inside a wire-mesh-protected "bird-house" remote from any contaminating sources of anthropogenic fossil fuel emissions. The height of the collection

18 '4C Content of Tropospheric CO2 at China Lake, California 19 station is 2m above ground level to reduce the amount of dust collected. The unexposed sodium hydroxide solution, as well as the exposed mixture, is shipped in tightly sealed plastic laboratory bottles. Exposure time to atmospheric air is typically one week. In the laboratory, the contents of the plastic bottles are transferred to a round-bottomed flask of 5L capacity. CO2 is gradually liberated by the addition of dilute hydrochloric acid. CO2 produced in the above manner is converted to acetylene by reaction with molten lithium metal to produce Li2C2, followed after cooling, to room temperature by hydrolysis with tritium-free well water. The acetylene is purified for counting by passage through traps of dry ice, traps filled with glass balls coated with P205, and traps filled with activated charcoal at 00. Samples were allowed to age for at least 3 weeks to allow any radon present to decay before radiocarbon analysis. After 14C assay, a small sample of the acetylene was recombusted to CO2 without isotopic fractionation by adding 02 and circulating over Cu0 at 600°C. The S13C of the CO2 was then measured mass spectrometrically (Linick, 1977).

RESULTS AND DISCUSSION Results reported here are based on 95% of the NBS oxalic acid activity normalized to b13C = -19%o (PDB). All sample activities have been normalized to a'3C = -25%o (PDB). Mass spectrometric o'3C measurements were made on all samples and standards using, as previously, CO2 prepared by recombusting an aliquot of the acetylene counting gas sample. Two of the gas proportional counters described by Linick (1977, 1979) were used, ie, 2L quartz counters. The data from our measurements are given in Table 1 and shown in 14C Figure 1. The continuing decrease following the peak produced by the bomb transient is evident from the data. The e-folding time for the excess 14CO2 is found to be 14-15 years. This is very similar to the time scale found from data prior to our sampling period. As noted, the time scale results from a series of complex interactions between the earth's various reservoirs, as well as in situ production. 14C02 From Figure 1, a seasonal structure in tropospheric is seen to be superimposed on the basic decrease. A small increase in 14C appears to occur about the middle of each year, which is barely detectable above the downward trend. Then a major decrease occurs each winter. This structure is due to a combination of atmospheric and oceanic effects. Mixing between the stratosphere and troposphere is not constant, but reaches a maximum around April of each year. This phenomenon is known as the "spring leak." It introduces excess 14002 each year. The effect is com- pounded in the oceans by the warming of surface waters which increases stratification and inhibits mixing of 14C02 into deeper waters resulting in a few per mil increase in the atmosphere. Thus, a combination of these processes can produce a small 14002 maximum in the northern hemispheric summer. During the winter the thermocline deepens, resulting in an increased uptake of atmospheric 14C and a more rapid decrease in the tropospheric level of 14C02. Inasmuch as the sampling location is very iso- 20 Rainer Berger et al

TABLE 1 14C and 13C measurements from China Lake, California La Jolla UCLA no. no. sampled »c 5729 2600 1/14-1/21/77 ±3.5 5728 2601 2/25-3/4/77 3.4 5739 2602 3/28-4/4/77 ±7.0 5730 2603 4/27-5/4/77 ±3.5 5737 2604 6/13-6/20/77 ±3.4 5732 2605 7/31-8/7/77 ±3.3 5731 2606 8/29-9/5/77 ±3.5 5734 2607 9/22-10/4/77 ±3.2 5733 2608 10/28-11/4/77 ±3.3 5735 2609 11/27-12/4/77 ±3.3 5736 2610 12/28-1/4/78 ±3.3 5727 2611 5/23-5/30/78 ±3.5 5726 2612 7/5-7/12/78 ±3.5 5725 2613 7/28-8/4/78 ±3.4 5724 2614 8/28-9/4/78 ±3.3 5738 2615 9/27-10/4/78 ±3.4 5722 2616 10/28-11/4/78 ±3.3 5741 2617 2/12-2/19/79 ±7.0 5740 2618 3/12-3/19/79 ±6.8 5704 2619 5/28-6/4/79 ±6.9 5706 2620 7/4-7/11/79 ±3.3 5719 2621 9/1-9/8/79 3.4 5700 2622 9/28-10/5/79 ±3.6 5721 2623 10/28-11/4/79 3.4 5718 2624 11/28-12/5/79 ±3.3 5723 2625 12/29-1/5/80 ±3.3 5686 2626 4/1-4/8/80 ±3.9 5688 2627 4/27-5/4/80 ±3.4 5690 2628 5/28-6/4/80 ±3.5 '4C Content of "1ropospheric CO2 at China Lake, California 21

TABLE 1 (continued)

La Jolla UCLA no. no. sampled 5692 2629 6/27-7/4/80 ±3.6 5694 2630 7/29-8/5/80 ±3.1 5696 2631 8/28-9/4/80 ±3.6 5698 2632 9/27-10/4/80 ±3.9 5702 2633 10/28-11/4/80 ±3.4 5720 2634 11/27-12/4/80 ±3.6 5711 2635 2/25-3/4/81 ±3.1 5710 2636 3/28-4/4/81 ±3.2 5716 2637 7/28-8/4/81 ±3.2 5717 2638 8/28-9/4/81 ±3.4 5673 2639 2/5-2/12/82 ±7.2 5674 2640 3/1-3/8/82 ±2.9 5675 2641 3/28-4/5/82 ±3.9 5680 2642 4/27-5/4/82 ±3.5 5681 2643 5/28-6/4/82 ±3.8 5676 2644 6/27-7/4/82 ±3.1 5683 2645 8/4-8/11/82 ±3.6 5682 2646 8/31-9/7/82 ±3.6 5677 2647 9/27-10/4/82 ±3.2 5678 2648 l 0/28-11 /4/82 ±3.6 5679 2649 11/27-12/4/82 ±3.2 5684 2650 12/29-1/5/83 ±3.3 5708 2651 5/18-5/25/83 ±3.1 lated, a seasonal varying influence of fossil fuel CO2 must be exceedingly small. In comparison, our radiocarbon results lie well within the date enve- lope of a similar study for central Europe (Levin et al, 1985). Taken together both investigations show that the global troposphere, except for some differences in fine structure, is well mixed between 35° and 48°N lati- tude on an intercontinental scale for the period of 1977-1983. In addition, information about the tropospheric residence time of CO2 can be derived 22 Rainer Berger et al 350

300

250

f T

200 1977 1978 1979 1980 1981 1982 1983 Fig 1. Atmospheric &4C at China Lake, California (35° 32' N, 117° 41' W) through a comparison of the atmospheric data with radiocarbon assays of banded corals (Druffel, ms in preparation).

ACKNOWLEDGMENTS G Plain collected the China Lake samples and deserves our gratitude. This study was funded in part by the State of California Air Resources Board to provide baseline data. This is publication no. 2635 of the Institute of Geophysics and Planetary Physics, UCLA.

REFERENCES Berger, R, 1979, Artificial radiocarbon in the atmosphere, in Berger, R and Suess, H E, eds, Radiocarbon dating, Internatl 14C conf, 9th, Proc: Berkeley/Los Angeles, Univ California Press, p 309-312. Berger, R and Libby, W F, 1966, UCLA radiocarbon dates V: Radiocarbon, v 8, p 467-497. -----1967, UCLA radiocarbon dates VI: Radiocarbon, v 9, p 477-504. -----1968, UCLA radiocarbon dates VIII: Radiocarbon, v 10, no. 2, p 402-416. ----- 1969, UCLA radiocarbon dates IX: Radiocarbon, v 11, no. 1, p 194-209. 14C Content of Tropospheric CO2 at China Lake, California 23 Berger, R, Fergusson, G J and Libby, W F, 1965, UCLA radiocarbon dates IV: Radiocarbon, v 7, p 336-371. Bien, C S, Rakestraw, N W and Suess, H E, 1965, Radiocarbon in the Pacific and Indian Oceans and its relation to deep water movements: Limnol & Oceanog, v 10, Supp R25- R37. D, Keeling, C 1979, The Suess effect: Environment Internatl, v 2, p 229-300. Levin, I, Kromer, B, Schoch-Fischer, H, Bruns, M, Munnich, M, Berdau, D, Vogel, J C and Munnich, K-0, 1985, 25 years of tropospheric 14C observations in central Europe: Radio- carbon, v 27, no. 1, p 1-19. Linick, T W,1977, La Jolla natural radiocarbon measurements VII: Radiocarbon, v 19, no. 1, p 19-48. --- -1979, La Jolla natural radiocarbon measurements VIII: Radiocarbon, v 21, rio. 2, p 186-202. -- -1980, La Jolla natural radiocarbon measurements IX: Radiocarbon, v 22, no. 4, p 1034-1044. Smith, B N and Epstein, 5, 1971, Two categories of'3C/12C ratios in plant tissue: Plant Physiol, v 47, p 380-384. [RADIOCARBON, Vot. 29, No. 1, 1987, P 24-44]

STUDY OF BONE RADIOCARBON DATING ACCURACY AT THE UNIVERSITY OF ARIZONA NSF ACCELERATOR FACILITY FOR RADIOISOTOPE ANALYSIS THOMAS W STAFFORD, JR*, A J T JULL**, KLAUS BRENDELt, RAYMOND C DUHAMELt, and DOUGLAS DONAHUE**

INTRODUCTION Bone would seem to be an ideal material for 14C dating because this calcified tissue contains 20 weight per cent protein. Fossil bone, however, can lose most of its original organic matter and frequently contains contaminants having different 14C ages. Numerous 14C dates on bone have been available to archaeologists and geologists but many age determinations have been inaccurate despite over 30 years of research in the field following the first 14C age determinations on bone (Arnold & Libby, 195k). This situ- ation remained unchanged until simple pretreatments were abandoned and more bone-specific fractions were isolated. The ideal solution is to use accelerator mass spectrometer 14C dating, which facilitates the use of milligram-sized amounts of highly purified compounds-an approach impossi- ble to pursue using conventional 14C decay-counting methods. OBJECTIVES Our principal objective was to determine how bone 14C dates could be made more accurate. Our goal was to improve sample pretreatment chemistry and use TAMS technology to date milligram-sized, highly purified bone constituents. The research was part of a larger study that used stable and 14C isotopes from fossil bones for chronologic, paleoenvironmental, and paleoecologic determinations (Stafford, 1984; Stafford et al, 1985). A secondary objective was to date several bones of unknown age that exhibited a wide range of preservation. The unknown-age fossils provided additional data on the range of ages that would be obtained from various chemical fractions. The accuracy of dates on these bones could be evaluated by knowing whether or not the same fraction dated accurately from the known-age mammoths. PRESENT KNOWLEDGE OF BONE `4C DATING 14C Bone is not usually recommended for dating (Libby, 1955; Olson, 1963) because its 14C ages are either discordant with associated charcoal dates or ages for different fractions from the bone are discordant with each other. Numerous methods have been devised to pretreat fossil bones (Ols- son et al, 1974; El-Daoushy, Olsson & Oro,1978; Taylor, 1982) but all techniques are minor modifications on methods used to extract either inor- * Laboratory of Isotope Geochemistry, Department of Geosciences, University of Ari- zona, Tucson, 85721. Present address: Carnegie Institution of Washington, Geophysical Lab- oratory, 2801 Upton St N W, Washington, D C 20008. ** NSF Accelerator Facility for Radioisotope Analysis, University of Arizona. t Department of Pharmacology, College of Medicine, University of Arizona, Tucson, 85724

24 14C Bone Dating with Arizona TAMS 25 anic carbon (bone apatite carbonate) or organic carbon (bone protein) for 4C dating. 14C The first bone dates were on total carbon from naturally burned (Arnold & Libby, 1951) and unburned bone (de Vries, 1959). 14C dates on organic fractions used the HC1-insoluble residue from artificially pyrolyzed bone (May, 1955) and later that residue treated with NaOH (Vogel & Waterbolk, 1963). Bone protein (approximately collagen) was extracted by HC1 decalcification (Munnich, 1957; Olsson, 1959; Berger, Homey & Libby, 1964; Krueger, 1965; Tamers & Pearson, 1965), with a chelating agent such as EDTA (Berger, Homey & Libby, 1964; Olsson et al, 1974; El-Daoushy, Olsson & Oro, 1978) or rarely with H2SO4 (Sato et al, 1969). Decalcification of bone yields a "weak-acid insoluble residue" that was often contaminated with humates (Vogel & Waterbolk, 1963). Methods to remove humic and fulvic acids from collagen include either NaOH treatment (Berger & Libby, 1966; Haynes, 1 967a) or conversion of the collagen to gelatin (Sinex & Farris, 1959; Longin, 1971). The use of both NaOH- leaching of collagen and gelatin extraction was introduced by Protsch (1975). The most rigorous methods for isolating organic carbon fractions are the chromatographic extraction of total collagen-derived amino acids (Ho, Marcus & Berger, 1969) and the isolation of individual amino acids as hydroxyproline and proline (Wand, 1981; Stafford et al, 1982; Gillespie & Hedges, 1983; Gillespie, Hedges & Wand, 1984). Inorganic carbon from fossil bone has been isolated by either acid hydrolysis of untreated bone (Olsson, 1959) or from bone pretreated with acetic acid (Haynes, 1968) or triammonium acetate (Hassan, Termine & Haynes, 1977), two reagents that are used to remove secondary carbonate contamination. Additional techniques for preparing bone carbonate include sequential HC1 hydrolysis (Haynes, 1968; Hassan, Termine & Haynes, 1977; Sullivan & Krueger, 1981) and differential thermal release of CO2 (Haas & Banewicz, 1980). Many of the inaccurate ages on fossil bone were due to the chemical heterogeneity of the dated fractions. The acid-insoluble residues retain humate contamination that is not removed by any of the described methods. Fortunately, the organic phase is amenable to chemical processing that is specific to the isolation of humates and specific peptides and amino acids. In contrast, inorganic carbon in fossil bones can exchange with environmental carbonate (Hassan, Termine & Haynes, 1977) and it is uncertain whether or not pretreatment methods yield an uncontaminated carbonate phase (Sullivan & Krueger, 1981; Schoeninger & DeNiro, 1982; Haas & Banewicz, 1980). Because no mechanisms are currently known for bone proteins to exchange carbon after burial, we emphasized the dating of the bone's organic phases, which were considered to have the greatest potential for purification and retention of their original 14C integrity. The following experiments evaluate the efficacy of bone dating by accelerator mass spectrometry. We have evaluated both pre-existing and new chemical procedures and make recommendations for testing the accuracy of 14C dates on bone. 26 TW Stafford, fret al METHODS

Experimental Design Fifty-eight 14C dates were determined on fractions from 11 fossil bone specimens. Thirteen dates were made on charcoal, shell, or pedogenic carbonates that were associated with fossil bones. Three known-age mammoth archaeologic sites were initially dated to deter- bones from Clovis-culture 14C mine which of several possible fractions would be most reliable for bone dating. Bone samples of unknown age were chosen for dating because they represented fossils with a range of geologic ages, preservation, and depositional environments. The charcoal and shell samples were dated because they were relevant to interpreting the accuracy of uranium series ages (Bi- 14C schoff & Rosenbauer, 1981) and dates (Bada et al, 1984) on human bone from the Del Mar Man site. The three known-age bones were all mammoth (Mammuthus sp) specimens that were from Clovis Indian mammoth-kill sites that date between 11,000 and 11,500 yr BP (Haynes, 1982). The elephants are independently dated with 14C ages on associated wood or charcoal. The Domebo mammoth was used for the most extensive experimentation and those results were used in determining which fractions would be most suitable for dating from the Dent and Escapule mammoths. Both the Domebo and Dent mammoths had collagen-like amino-acid compositions and contained 0.7%N and 0.8%N, respectively. In contrast, the Escapule mammoth had a non- collagen amino acid composition and contained 0.08%N. Bones of unknown ages and good to extremely good collagen preservation were a Wisconsinan-age whale preserved in permafrost (Beaufort Sea coast whale), human calvaria from an arid cave (Wilsall/Anzick series), and a horse ramus from a hyper-arid cave (Fishbone Cave series). Humid- cave depositional environments were represented by bird and rodent postcranial bones from the Puu Naio Lava Tube series, which included specimens with good to very poor collagen preservation. Human bones that had poor to extremely poor protein preservation and which were from leached and oxidized sediments comprised fossils from the Yuha, La Jolla shores, and Wilson-Leonard series. A list of these dates is presented below.

Chemical Pretreatment of Bone The chemical pretreatment methods used for the bones are summa- rized in Figures 1 and 2. Fossil bones are washed in tap water to remove sediments and broken into l to 3cm fragments that are ultrasonically cleaned in tap water and distilled water. Physically cleaned bone is ground to <63µm or left intact if grinding losses must be avoided. Inorganic carbon is extracted from the OH-apatite phase by hydrolyzing bone powder with 95% H3P04. The bone powder is either untreated or it is extracted for 24hr with 1 M acetic acid under line vacuum. Organic carbon phases are concentrated by decalcifying bone powder in 4°C, 0.6N distilled HCI. The acid insoluble, collagenous residue is sepa- 14C Bone Dating with Arizona TAMS 27

WHOLE BONE

physical cleaning; grinding of bone to <63µm

BONE POWDER

INSOLUBLE RESIDUE HCL SOLUBLE PHASE

H2O washes; lyophylize 0.45µm filtration; reverse phase chromatography of HCI INSOLUBLE COLLAGEN peptides

collagen+H20 at pH 3, N2 purge; PURIFIED PEPTIDES Lyophylize 3h at 110°C, filter, lyophylize chromatographic isolation of individual CRUDE COLLAGEN CRUDE GELATIN amino acids

hydrolysis: hydrolysis: INDIVIDUAL AMINO ACIDS 6N HCI, 110°C, 6N HCI, 110°C DERIVED FROM PEPTIDES 24 hr 24 hr

HYDROLYSED PROTEIN

XAD-2 or charcoal purification XAD-PURIFIED COLLAGEN HYDROLYSATE chromatographic isolation of individual amino acids

INDIVIDUAL COLLAGEN-DERIVED AMINO ACIDS

Fig 1. Flow diagram showing pretreatment methods used for 14C dating fossil bones. Well- to moderately-well-preserved bones require the extraction of gelatin or XAD-treated gelatin, whereas the minimum pretreatment for poorly-preserved bones comprises the XAD- treatment of hydrolyzed protein and preferably the isolation of individual amino acids. RAW BONE

PHYSICAL CLEANING

GRIND TO <63 pm

GELATIN METHOD HYDROLYSIS METHOD ACETIC ACID DIGESTION SOLVENT EXTRACTION CLEAN BONE POWDER Soxhlet 1 M Acetic Acid, 20°C 069 HCI HCI Extraction 24h , Line Vacuum Discard Solution with Acetone H2O Wash Powder H PO 3 4 WEAK ACID ACETONE ACETONE Cu0 Combustion hydrolysis WEAK ACID SOLUBLE WEAK ACID SOLUBLE ACETIC ACID EXTRACTED EXTRACTED SOLUBLE TOTAL TOTAL INORGANIC INSOLUBLE ORGANIC ORGANIC BONE POWDER BONE POWDER CARBON CARBON CARBON RESIDUE CARBON CARBON

10,690 ± 9540 ± 2050 ± 580 8010'450 7300 ± 320 640 10,820 ± 270 480 0.6N HC/ CuD H3PO4 Combustion hydrolysis GELATIN FRACTION HYDROLYSIS METHOD TOTAL BONE CARBONATE ACID ACID SOLUBLE pH 1, H2O, 245 6N HCI, 24h, 110°C CARBON INSOLUBLE PHASE under CO2 l i0oC, under N2 RESIDUE N2

10,760 ± 440 9310 ± 360 9360± 670

SUPERNATANT HOT WATER 66 HCI SUPERNATANT "GELATIN" INSOLUBLE INSOLUBLE RESIDUE RESIDUE SAD-2 COLUMN

10,280 560 ADSORBED FULVIC ACIDS EFFLUENT. ELUTE WITH BASE AMINO ACID OR ORGANIC SOLVENT HYDROLYSATE 6N HCI SUPERNATANT 4910 '320 11,010' 110 INSOLUBLE RESIDUE 11,480. 450 XAD-2 COLUMN Fig 2. Flow dia am with the chemical steps used to iso- ' Nitrosylation late fractions for C dating from the Domebo mammoth. 4:1 12 m HCI & Accelerator 14C ages are given for each fraction and are used ADSORBED FULVIC EFFLUENT 18 m HN03 ACIDS .420 to illustrate the effectiveness of the chemical Purifications and ELUTE WITH BASE 10,810 least reliable. OR ORGANIC SOLVENT which fractions are most or ACETONE ETHER Nitrosylation 5130 ± 290 EXTRACTION 4. 1 12 m HCI, 18 M HNO3

ACETONE/ETHER SOLUBLE, EXTRACTION INSOLUBLE, NON-POLAR POLAR PHASE PHASE IMINO ACIDS "HYDROXY ACIDS" INSOLUBLE, SOLUBLE 786o. 450 11,280. 530 POLAR PHASE NON-POLAR IMINO ACIDS PHASE

0'450' 11,330 - 410 14C Bone Dating with Arizona TAMS 29 rated by centrifugation from the acid soluble phase. The acid-soluble fraction is filtered through 0.45µm teflon Millipore membranes and rotary evaporated. The acid-soluble phase can be further purified by passing it through XAD resin, which is used to remove fulvic acids (Stafford et al, 1982). The acid-insoluble collagen is lyophylized, then hydrolyzed or converted to gelatin. The protein is hydrolyzed by heating ca l Omg of protein per l ml distilled 6N HC1 for 24hr at 110°C. Teflon-sealed tubes purged with nitrogen are used for the hydrolysis. The hydrolyzate solution is filtered before it is passed over XAD. Gelatin is extracted from the weak-acid insoluble residue by heating ca 10mg protein and 1Oml pH 3 water at 90°C for 3 to 4hr. The hydrolysis tubes are purged with nitrogen prior to sealing. The gelatin solution is cen- trifuged and filtered before it is lyophylized. The freeze-dried gelatin is hydrolyzed and purified with XAD. Fulvic acids are removed from the gelatin and collagen hydrolyzates by passing the 6N HC1 through a column of XAD resin. From 10 to 50ml of hydrolyzate is passed through a 1 X40cm glass column of 20-50 mesh XAD-2 resin. To pretreat the resin for use, ca 5OOg Sigma Chemical Com- pany XAD-2 resin are washed exhaustively with acetone, methanol, and water before the resin is extracted 3 alternating times with 3N HC1 and 3M NaOH. The resin is washed finally with 1 N HC1. A bed of resin 20 to 30cm high is poured, capped with glass wool, and equilibrated with 3 bed volumes 6N HC1. The protein hydrolyzate is passed through the resin at 100µl/min or at a flow rate slow enough to adsorb the fulvic acids in the upper 1 /3 of the resin bed. The XAD-purified hydrolyzates are filtered and rotary evaporated. Fulvic acids separated from the hydrolyzed protein can be eluted from the resin by washing the resin with distilled water until the eluaee pH is between 1 and 2. A 1 M NH4OH solution is used to desorb the fulvic acids, which are immediately acidified with HC1 before drying and combustion.

Target Preparation After isolation of the desired organic fraction, the sample was combusted to CO2, which was reduced over magnesium to amorphous carbon. One to 2mg carbon was fused with iron powder to form an iron carbide bead that is mounted into a sample wheel for accelerator dating (Dull, Don- ahue & Zabel, 1983). The purity of CO2 was increased by combusting the sample with 2g CuO powder that contained 1 Owt % Ag powder. The oxi- dant was prepared by mixing Ago powder with CuO powder that had been combusted for lhr at 800°C. The AgO/CuO powder mixture was finally fired at 600°C for l hr. The CO2 gas from one sample, AA-312C, was converted directly to graphitic carbon using the methods ofJull et al (1986) and Vogel et al (1984).

Radiocarbon Measurement 14C/13C The ratios were measured by TAMS at the University of Ari- zona. Sample carbon was converted to Fe-C targets according to Jull, Dona- hue and Zabel (1983). The targets were mounted in a 6-position target wheel, which was placed into the cesium-sputter ion source. Measurements 30 TW Stafford,,Jr et al were made as a series of 4 cycles around the wheel, which contained 4 unknowns, l modern standard, and a background sample. For each cycle, a 14C 13C single measurement comprised 15 samplings of for 50 sec and of for 10 sec. Details of the procedure are given in Donahue et al (1983). Achievable 14C precision using the Fe-C targets was between ±150 and +400 yr for samples <10,000 yr old. The wide variability in precision was 12C- caused by variations in the target's output, which ranged from 0.5 to 3µA (Dull, Donahue & Zabel,1983). Dates reported here use a background 14C. of 2 ± 1% modern This represents our best estimate of the background for these targets; thus, sample ages may differ slightly from previously published values (Stafford et al, 1984; Bada et al, 1984; Taylor et al, 1985). All dates reported here were made between February and June 1984. Since then we have developed a procedure for making graphite targets with a modified Vogel et al (1984) method. When graphite is used we have been able to attain single-target precisions of ± 1%, equivalent to an age precision of ±80 yr or better. Backgrounds for graphite targets are currently ca 0.6% modern 14C. Results for this method of target preparation were reported previously (Dull et al, 1986; Linick et al, 1986). Results The 14C dates are presented in Table 1, which includes only dates for the three known-age mammoths. The elephant 14C dates are listed in order of age for each fraction. The fractions from the Domebo mammoth that were not accurate at 2r were acetone-soluble organic carbon, fulvic acids, OH-apatite, total bone carbon, and acid insoluble collagen. Inaccurate Dent mammoth dates were from acid-insoluble collagen and gelatin, all three fractions of the Escapule mammoth dated <11,000 yr. whereas 14C Epoxy preservative from the Escapule mammoth had an apparent age (3680 ± 210: GX-11261) that was significantly less than the geologic age of the mammoth.

DISCUSSION AND CONCLUSIONS Our results show that the accuracy of bone 14C dates depends on the preservation of the bone protein, upon which fractions are dated, and what contaminants are present. Humic and fulvic acids are the predominant contaminants in fossil bone. The degree to which humates affect a bone's 14C age depends on the weight per cent of humates present and the apparent 14C age of the humate fraction. Permafrost-derived and sometimes arid- cave-derived bones have extremely well-preserved collagen that is amenable to standard biochemical isolation techniques. If humates exist, they are usually in negligible amounts and are restricted to the bone's exterior. Although accurate ages can apparently be obtained often on acid-insoluble collagen and untreated gelatin from well-preserved bones, it is recommended that these fractions be purified with XAD resin, which will stan- dardize procedures for humate removal. Bones that have collagen-like compositions and >0.2% N should be dated using only certain fractions. Recommended fractions are either weak-acid-insoluble collagen or gelatin that is hydrolyzed and purified with XAD resin. Although isolation of individual amino acids may not always be 14C Bone Dating with Arizona TAMS 31

TABLE 1 Target Sample Radiocarbon Lab no. no. description date (yr BP)

Domebo Mammoth AA-822A C-783 Acetone soluble organic carbon extracted from 580 bone powder AA-816 C-650 Fulvic acids from hydrolyzed 0.6N HCl insoluble 760 collagen AA-812 C-561 Fulvic acids from hydrolyzed 0.6N HCl insoluble 320 collagen: NH4OH elution AA-819 C-771 Fulvic acids from hydrolyzed gelatin; acetone elu- 290 tion AA-818 C-662 OH-apatite CO2, untreated bone 320 AA-801 C-474B Total inorganic + organic carbon from untreated 500 bone powder AA-815 C-615 OH-apatite CO2 from acetic acid extracted bone 360 powder AA-822B C-743 0.6N insoluble collagen extracted from acetone- 670 extracted bone powder AA-802A C-477 0.6N HC1 soluble phase from bone powder 480 AA-810 C-556B Imino acids from XAD-2 purified hydrolyzed gela- 450 tin AA-802B C-743 6N HC1 insoluble residue from weak-acid insolu- 560 ble collagen AA-803 C-480 Unpurified gelatin 410 AA-814 C-606 0.6N HCl insoluble collagen 640 AA-804 C-542 Total organic and inorganic carbon after HAc ex- 440 traction of bone powder AA-805 C-543 XAD-purified hydrolyzed gelatin 420 AA-824 C-1002 0.6N HCl insoluble collagen 270 AA-811 C-559 Imino acids from XAD-purified 0.6N HC1 insolu- 450 ble collagen AA-806 C-544B XAD-purified hydrolyzed collagen 710 AA-808 C-555 Acetone/ether soluble a-amino acids from nitro- 530 sylated XAD-purified acid insoluble collagen AA-807 C-551 Acetone/ether soluble a-amino acids from XAD- 470 purified hydrolyzed gelatin AA-825 C-1038 XAD-purified 0.6N HC1 insoluble collagen 450 AA-823 C-978 Elm tree stump associated with mammoth (Leon- 450 hardy & Anderson, 1966)

Dent Mammoth AA-830 C-1220 0.6N HCl insoluble collagen 520 AA-831 C-1221 Unpurified gelatin 350 AA-832 C-1261 XAD-purified collagen hydrolyzate 500 AA-833 C-1267 XAD-purified gelatin hydrolyzate 480 Escapule Mammoth AA-834 C-1259 0.6N HCl insoluble residue from bone powder 470 AA-835 C-1275 Unpurified "gelatin" 270 AA-836 C-1358 XAD-purified, hydrolyzed insoluble residue 280

necessary, their extraction and dating is highly encouraged. HC1 insoluble residues, untreated gelatin, and acid-soluble phases may occasionally yield accurate dates, but there are no known chemical criteria for predicting when dates will be spurious on these fractions. Bones with non-collagen amino acid compositions and <0.2% N do 32 TW Stafford, Jr et al not date as accurately as bones with substantial amounts of collagen. Even XAD treatment may not be effective in yielding accurate ages on bones that are diagenetically altered. Contamination by exogenous amino acids and epoxy residues are the likely causes of the young ages for the Escapule mammoth's fractions. Accurate dates from degraded bone will probably require the exclusive dating of individual amino acids. The worst fractions to use from poorly preserved bone are weak-acid soluble and insoluble phases and any apatite fraction. Non-specific organic fractions should be used only when further pretreatment would lower carbon to sub-milligram levels and only when a minimum-age estimate is acceptable. In summary, the fractions that should not be dated from bones are untreated weak-acid insoluble residues, weak-acid soluble phases, untreated gelatin, apatite carbonate, and humic or fulvic acids. XAD treatment to remove humates should become mandatory for acid insoluble collagen and gelatin from all bones. The isolation of individual amino acids is highly encouraged, especially for bones that have lost >90% of their original organic matter during diagenesis.

BONE RADIOCARBON DATES FROM THE ARIZONA NSF ACCELERATOR FACILITY Beaufort Sea Coast series Gray whale (Eschrichtius sp) rib from marine Flaxman Fm, 20km S of Beaufort Sea coast, Sec 14, Ti 6N, R5W, Teshekpuk C-1 quad, Alaska (70° 44.78' N,153° 06.38' W). Coll Sept 1, 1983 and subm by L I) Carter, USES, Anchorage, Alaska. Date will estimate age of Flaxman transgression. AA-312A. Beaufort Sea coast >26,700 Gray whale rib. Comment: weak-HCI insoluble collagen. Target C-1035. AA-312B. Beaufort Sea coast >27,300 Gray whale rib. Comment: gelatin phase from weak-HC1 insoluble collagen used for AA-312A. Target C-1040. AA-312C. Beaufort Sea coast >38,000 Gray whale rib. Comment: graphite made from gelatin fraction. General Comment: extremely good bone preservation. Bone and its collagen have properties as of modern material. All dates significant at 2r. (LDC): '4C dates on assoc fossils: marine mollusk shells, 42,600 ± 1500 (USGS- 1689), whalebone, 22,530 ± 260 (Beta-6108); seal bone, 19,640 ± 130 (USGS-1515); marine mollusk shells, 20,760 ± 210 (Beta-5869). Whale considered beyond range of 14C dating. Wilsall (Anzick) series Homo sapiens sapiens partial calvaria from Wilsall (Anzick) Paleo-Indian site, 0.6km 5 of Wilsall, Montana (Taylor, 1969; Lahren & Bonnichsen, 14C Bone Dating with Arizona TAMS 33 1974). One skull fragment was coated with hematite, from 3-to-5-yr-old adolescent. Second sample was bleached-white calvarium.

AA-313A. Wilsall (Anzick) 8690 ± 310 Hematite stained, 3-to-5-yr-old adolescent calvaria. Comment: weak- HC1 insoluble collagen dated. Target C-1036.

AA-313B. Wilsall (Anzick) 10,500 ± 400 Hematite stained, 3-to-5-yr-old adolescent calvaria. Comment: gelatin fraction from collagen used for AA-313A. Target C-1037.

AA-313C. Wilsall (Anzick) 8620 ± 340 Bleached calvaria. Comment: weak HCl insoluble collagen. Target C-1039.

AA-313D. Wilsall (Anzick) 8940 ± 370 Bleached calvaria. Comment: gelatin extracted from AA-313G collagen. Target C-1042. General Comment: both calvaria have physical and chemical properties of modern bone.

Cheek Bone Cave series Pocket gopher (Geomys cf bursarius) ramii and unid. larger mammal bone fragments coil Nov 1983 and id by W Klippel; samples from Cheek Bone Cave, 4OMu-261, 13km ESE of Columbia, Maury Go, Tennessee; Stratum 8(2?), l Ocm thick, level 42 of LOIN 99E. Gopher bones were well perserved and angular, fragments 0.5 to 2cm long.

AA-734. Cheek Bend Cave 14,710 ± 490

Unid. mammalian cortical bone, CBC no. 1 colln, Sample A. Comment: bone analysis: 4.22%G, 0.97%H, 0.95%N. Pale yellow-brown to dark brown, moderately hard, chalky surfaced bone.

AA-735. Cheek Bend Cave 6740 ± 280

Unid. large mammal cortical bone, CBC no. 1 colln, Sample B. Very pale yellow, very hard, waxy bone with modern physical and chemical properties. C-1125. General Comment: XAD-2-purified gelatin hydrolyzate dated from all samples. Bones in GBC no. 1 colln range from 0.65% to 1.92%N and are chalky to hard and waxy.

Yuha series Homo sapiens Sapiens post-cranial bone from Yuha cairn burial, W of El Centro, Imperial Co, California. Coll 1971 by M Childers (1974; 1983). Skeletal remains curated in three different collns: Imperial Valley Goll Mus, E1 Centro, California (IVCM), J L Bishoff, USGS, Menlo Park, California (USGS), and R E Taylor, Univ California, Riverside (UCR). 34 TW Stafford, Jr et al

AA-737. Yuha 3930 ± 270 Post-cranial bone no. 1. IVCM colln. Comment: 0.3N HCl-insoluble fraction dated. Target C-674. AA-738. Yuha 1750 ± 230 Post-cranial bone no. 1, IVCM colln. Comment: 0.3N HCl soluble fraction. Target C-673. AA-739. Yuha 2490 ± 300 Post-cranial bone no. 1, IVCM colln. Comment: total inorganic carbon from 95% H3P04 hydrolysis of bone powder. Target C-664. Bone analysis: 2 234U/238U = 1 .24, 0Th/ 34U = 0.05; U/Th date = 5900 + 1000/-800 yr BP (J Bischoff, pers commun, 1983). AA-740. Yuha 2830 ± 260 Caliche, 0.3mm thick coating bone no. 1, IVCM colln. Comment: CO2 evolved by H3P04 hydrolysis. Target C-663. AA-741. Yuha 2460 ± 290 Caliche, 0.3mm thick coating bone no. 1, IVCM colln. Comment: caliche combusted with Cu0 powder. Target C-684. AA-742. Yuha >26,600 Petrocalcic horizon caliche, 3mm thick, frag no. 1, IVCM colln. Com- ment: CO2 evolved by H3P04 hydrolysis. AA-743. Yuha 3030 ± 270 Petrocalcic horizon caliche, 2mm thick, frag no. 2, IVCM colln. Com- ment: CO2 evolved by H3P04 hydrolysis. AA-744. Yuha 2610 ± 200 Post-cranial bone no. 2, UCR colln. Comment: total bone inorganic CO2 evolved by H3P04 hydrolysis. Target C-748. AA-745. Yuha 2840 ± 220 Post-cranial bone no. 3, USGS colln. Comment: total 0.3N HCl soluble organic carbon. Specimen was fragment from bone subm to J Bischoff by M Childers and dated by aspartic acid racemization at 23,600 yr (Bischoff & Childers, 1979). Bone analysis: 8.9ppm U, 234U/238U = 1.21, 230Th/234U = 0.03 (J Bischoff, pers commun, 1983). Target C-759B. AA-746. Yuha 2690 ± 200 Post-cranial bone no. 3, USGS colln. Comment: total inorganic CO2 evolved by H3P04 hydrolysis. Sample as AA-745. Target C-746. General Comment: for Yuha series 14C dates, see Stafford et al (1984). All bones were very poorly preserved. Analyses on unid. cortical fragment from IVCM colln: 7.90%C, 0.32%H, 0.06%N; 8.34%C, 0.26%H, 0.05%N; Bone 14C Dating with Arizona TAMS 35 and 4.57%C, 0.35%H, 0.08%N. Differences between petrocalcic carbonate dates (AA-742, -743) are probably due to mixing of young and old caliche 14C horizons when body was interred. Caliche dated to 22,125 ± 400 230Th (UCLA-2600 "1854") and dated to 19,000 ± 3000 (Bischoff et al, 1976) probably antedates burial. Caliche-coated cobbles and boulders for cairn were taken from older (Pleistocene) deposits. Caliche 3 to 6mm thick and 25mm across that adhered to bone was '4C-dated to 21,500 ± 1000 (GX-2674) (Bischoff et al, 1976). Caliche was probably pre-Holocene carbonates that were later cemented onto bone.

Wilson-Leonard series Homo sapiens sapiens postcranial bone and assoc charcoal from Wilson Leonard site, 41 WM-235, Williamson Go, Texas (3378550N, 617250E) Level 32; bones were in Leanne soil. Coil 1933 by F Weir and subm by FW, State Dept Hwys and Public Transportation.

AA-747. Wilson-Leonard 4650 ± 310 Homo sapiens sapiens bone fragments. Comment: total inorganic CO2 evolved by H3P04 hydrolysis. Target C-1017.

AA-748. Wilson-Leonard 5940 ± 520 Homo sapiens sapiens bone fragments. Comment: total 0.6N HCl soluble organic carbon. Target C-968.

AA-749. Wilson-Leonard 6700 ± 460 Homo sapiens sapiens bone fragments. Comment: total 0.6N HC1 soluble organic carbon. Target C-1117.

AA-751. Wilson-Leonard 5860 ± 270 Second target from AA-750 carbon. Comment: Target C-967.

AA-752. Wilson-Leonard 5440 ± 420 Homo sapiens sapiens bone fragments. Comment: hot-water insoluble organic carbon from 0.6N HCl insoluble residue. Second extraction of this phase. Target C-1116B.

AA-753. Wilson-Leonard 1270 ± 280 Homo sapiens sapiens bone fragments. Comment: gelatin phase dated. Fraction was pale yellow solid resembling inorganic salt. No properties of modern gelatin. Target C-970. General Comment (TWS): bone collns 2 and 3 contained unid. cortical and canceilous fragments that were combined as 22.7g of cleaned, powdered bone; cortical bone analysis: 4.01%C, 0.54%H, 0.09%N; cancellous bone analysis: 3.48%C, 0.56%H, 0.06%N. General Comment (FW): 14C dates on charcoal that is apparently stratigraphically higher than burial: 7470 ± 230 (Tx-4798), on charcoal ca 1.5m above burial and 8820 ± 120 (Tx-4784A), 8860 ± 150 (Tx-4784B) and 36 TW Stafford, J r et al 8940 ± 100 (Tx-4784C) on charcoal ca 1.2m above human skeleton. Leanne soils overlying and enclosing burial had soil 14C dates of 9470 ± 170 (Tx-4787) and 9650 ± 120 (Tx-4793).

La Jolla Shores series Homo sapiens sapiens long bone fragments coil May to July 1926 by M Rogers, San Diego Mus of Man, California. In dune sands (now leveled) on embayment 1.2km N of La Jolla (32° 51' 25" N, 1170 16' 17" W) San Diego Mus site W-2, mus specimen no. SDM-16755. Part of colln of Littoral I culture human limb and rib fragments from same white sand stratum yielding partial human cranium S1)M-16742. White sand stratum is ca 2m below ground level and ca 5.5m asl. 5DM-16755 colln dated by aspartic acid 14C racemization to 28,000 yr BP (Bada, Schroeder & Carter, 1974) and by dating to 1850 ± 200 (UCLA-2368),1930 ± 200 (UCLA-2384) and 1770 + 790 (UCR-1511 D) (Taylor, 1983). Subm 1982 by R Tyson, San Diego Mus of Man.

AA-754. La Jolla Shores 3640 ± 360 Homo sapiens sapiens partial radius. Comment: total inorganic CO2 evolved by H3PO4 hydrolysis. Target C-658.

AA-755. La Jolla Shores 5290 ± 270 Homo sapiens sapiens partial radius. Comments: total 0.3N HC1 insoluble organic carbon. Target C-669.

AA-756. La Jolla Shores 6330 ± 280 Homo sapiens sapiens partial radius. Comment: total 0.3N HC1 soluble organic carbon. Target C-675.

A-757. La Jolla Shores 5110 ± 270 Caliche film coating partial radius. Comment: CO2 evolved by hydrolysis with H3PO4. Target C-601.

Fish Bone Cave series Horse (Equus sp) postcranial bone from Fish, Bone Cave, P3e, Winne- muca Lake, Pershing Co, Nevada (40° 12' 08" N,119° 16' 45" W). Coil 1956 by P C Orr (1974) and subm 1984 by R Thompson, Univ Arizona.

AA-759. Fish Bone Cave 12,280 ± 520 Partial Equus sp right ramus, Site Pe/4, Nevada State Mus no. 317. Comment: extremely well-preserved bone. Sample overlay sagebrush (originally id as juniper-bark mat) that was 14C-dated to 11,200 ± 250 (L-245) (Orr,1974; Broecker, Kulp & Tucek, 1956). Ramus has chemical and physical properties of modern bone. HC1 dissolution yields pseudomorph of bone. Bone washed with acetone twice before decalcification. Gelatin is physically identical to modern gelatin. Target C-1276. Bone 14C Dating with Arizona TAMS 37 Puu Naio Lava Tube series Rodent and extinct bird bones from Puu Naio lava tube, Ulupalaku Ranch, Maui, Hawaii (20° 37' N,156° 24' W). Coil Feb 1984 by S Olson and H James, Smithsonian Inst, Washington, D C; subm 1984 by P Martin, Univ Arizona.

AA-760. Puu Naio lava tube 707 ± 350 Rat (Rattus exulans) bones. W12, Unit II. Comment: organic carbon extracted from 450mg combined partial pelvis, 2 femora, tibia, 2 radii and ramus bones. Bone analysis: 7.60%C, 1.27% H, 2.08% N. Bones had physical and chemical properties of modern bone. Target C-1271.

AA-761. Puu Naio lava tube 1850 ± 270 Ibis (Apteribis sp) single, complete femur. E24, 10 to 20cm. Comment: bone analysis: 9.81% C, 1.74% H, 2.75% N. Bones had physical and chemical properties of modern bone. Target C-1272.

AA-762. Puu Naio lava tube 4340 ± 610 Extinct goose (Thambetochen sp) complete femur. W11, cross-section S face, unit III, subunit I, 50 to 60cm. Comment: chalky bone with no spiral breakage possible. Bone analysis: 4.70%C, 0.88%H, 0.3%N. Target C-1273.

AA-763. Puu Naio lava tube 7750 ± 500 Ibis (Apteribis sp) complete tarsomatetarsus. W 11, 90 to l 00cm. Com- ment: chalky bone, readily disaggregated in HCI. Bone analysis: 7.78%C, 1.34%H, 1.64%N. Target C-1274. General Comment: XAD-2-purified collagen hydrolyzates were dated from all bones.

Domebo Mammoth series Postcranial bone from immature mammoth (Mammuthus cf imperator) id by M Mehl (1966). Mammoth was excavated 1962 at Paleo-Indian Dom- ebo site (34Cd-50), ca 4km E of Stecker, Caddo Co, Oklahoma (NE 1 /4, SWI/4, SE1/4, sec 29, T6N, RIOW). For site report, see Leonhardy (1966). Bone was used as known-age fossil for calibrating bone sample preparations. Mammoth dates 11,000 to 11,500 yr BP by assoc with Clovis culture artifacts (Haynes, 1982, 1984).14C dates on assoc wood are 11,045 ± 647 (SM-695) (Leonhardy & Anderson, 1966) and 11,490 ± 450 (AA- 823), which is accelerator relate of SM-695 wood. Wood 7m above bone level dated to 10,123 ± 280 (SM-610).

AA-801. Domebo mammoth 8010 ± 500 Bone. Comment: total inorganic and organic carbon from powdered bone. CuO combustion. Whole bone 14C date. Target C-474B. 38 TW Stafford, Jr et al AA-802A. Domebo mammoth 9540 ± 480 Bone. Comment: weak-HCl-soluble organic carbon from bone powder. Target C-477.

AA-802B. Domebo mammoth 10,280 ± 560 Bone. Comment: 6N HCl insoluble organic carbon from hydrolyzate of 0.6N HC1 insoluble collagen. Target C-743. AA-803. Domebo mammoth 10,350 410 Bone. Comment: unpurified gelatin. Target C-480.

AA-804. Domebo mammoth 10,760 ± 440 Bone. Comment: total organic and inorganic carbon from bone powder after leaching with 1M acetic acid. CuO combustion. Target C-542.

AA-805. Domebo mammoth 10,810 ± 420 Bone. Comment: XAD-2-purified gelatin hydrolyzate. Target C-543.

AA-806. Domebo mammoth 11,010 ± 710 Bone. Comment: XAD-2-purified collagen hydrolyzate. Target C- 544B.

AA-807. Domebo mammoth 11,330 ± 470 Bone. Comment: alpha-hydroxy acids from nitrosylation of XAD-2- purified gelatin hydrolyzate. Target C-551.

AA-808. Domebo mammoth 11,280 ± 530 Bone. Comment: alpha-hydroxy acids from nitrosylation of XAD-2- purified collagen. Target C-555.

AA-810. Domebo mammoth 10,120 ± 450 Bone. Comment: imino acids (hydroxyproline and proline) isolated from XAD-2-purified gelatin hydrolyzate. Target C-556B.

AA-811. Domebo mammoth 10,860 ± 450 Bone. Comment: imino acids (hydroxyproline and proline) from XAD- 2-purified collagen hydrolyzate. Target C-559.

AA-812. Domebo mammoth 4910 ± 320 Bone. Comment: fulvic acids from hydrolyzed collagen. FA eluted from XAD-2 resin with NH4OH (conc). Target C-561.

AA-814. Domebo mammoth 10,690 ± 640 Bone. Comment: 0.6N HCl insoluble collagen. Target C-606.

AA-815. Domebo mammoth 9310 ± 360 Bone. Comment: hydroxy-apatite (dahllite) CO2 from bone powder extracted for 241i with TM acetic acid. H3PO4 hydrolysis. Target C-615. Bone 14C Dating with Arizona TAMS 39 AA-816. Domebo mammoth 4810 ± 760 Bone. Comment: fulvic acids from hydrolzed collagen. Target C-650. AA-818. Domebo mammoth 7300 ± 320 Bone. Comment: hydroxy-apatite CO2 from untreated bone powder. H3P04 hydrolysis. Target C-662. AA-819. Domebo mammoth 5130 ± 290 Bone. Comment: fulvic acids from hydrolyzed gelatin. FA eluted with acetone. Target C-771. AA-822A. Domebo mammoth 2050 ± 580 Bone. Comment: acetone-soluble organic carbon isolated by soxhlet extraction of bone powder. Target C-783. AA-822B. Domebo mammoth 9360 ± 670 Bone. Comment: 0.6N HCl insoluble collagen from acetone-extracted bone powder. Target C-743. AA-823. Domebo mammoth 11,490 ± 450 Stump of elm (cf Ulmus elata assoc with mammoth. Comment: outer 10 rings including bark were dated. Total ring count = 94 ± 1 (id by M Thomp- son). Wood was previously dated to 11,045 ± 647 (SM-695) (Leonhardy & Anderson, 1966, p 24). AA-824. Domebo mammoth 10,820 ± 270 Bone. Comment: 0.6N HCl insoluble collagen from bone powder. Tar- get C-1002. AA-825. Domebo mammoth 11,480 ± 450 Bone. Comment: XAD-2-purified collagen hydrolyzate. Target C-1038. General Comment: fractions of bone previously dated by Leonhardy and Anderson (1966, p 24-25) were untreated tusk: 4952 ± 304 (TBN-311); bone organic carbon soluble in 2N HCl after initial 2% NaOH: 11,220 ± 500 (SI-172) and humic acids extracted after decalcification: 11,200 ± 600 (SI-175). Domebo series represents bone preserved in reduced clay. Analy- ses of cortical bone (micro-Kjedahl): 0.43%N; cortical bone (CHN analyzer): 5.19%C, 0.69%H, 0.69%N; cancellous bone: 4.37%C, 0.59%H, 0.24%N. Uranium analyses by j Bischoff, USGS: 234U/238U = 1.14 ± 0.02, 23sU = 4.57 + 0.09ppm, age = 9512 + 525-400yr. Uranium series age = 11,500 ± 2000 (Szabo,1980).

Dent Mammoth series Postcranial mammoth bone from Clovis culture Dent site, Weld Go, Colorado (40° 19' N,104° 49' W),1.2km SE of Milliken, Colorado (Figgins, 1933; Wormington, 1959; Haynes,1974). Coll Oct 1973 by F Frazier. Site was first unquestionable evidence of assoc of humans and mammoths in 40 TW Stafford, Jr et al North America. Previous 14C date on bone and tusk fragments was 11,200 ± 500 (1-622) (Trautman & Willis, 1966; Haynes, 1967b).

AA-830. Dent mammoth 8250 ± 520 Bone. Comment: weak-HC1-insoluble collagen from bone powder. Tar- get C-1220.

AA-831. Dent mammoth 9240 ± 350 Bone. Comment: unpurified gelatin phase. Target C-1221.

AA-832. Dent mammoth 10,590 ± 500 Bone. Comment: XAD-2-purified hydrolyzed collagen. Target C-1261.

AA-833. Dent mammoth 10,950 ± 480 Bone. Comment: XAD-2-purified gelatin hydrolyzate. Target C-1267.

General Comment: sample was coated with Gelva preservative. Bone powder from cancellous tissue was soxhlet extracted 20hr with ethanol, washed with dirt H2O, extracted I Ohr with acetone and wash in H2O before drying. Extracted powder was used for all subsequent isolations. CHN analysis: 0.83%N (cortical bone); 1.07%N (cancellous bone). Escapule Mammoth series Innominate from adult mammoth (Mammuthus /ParelephasJ columbi) from Escapule site, Clovis culture mammoth kill site (EE:8:28) in Horse Thief Draw, Sec 1, T225, R21 E, Cochise Co, Arizona (Hemmings & Haynes, 1969). Mammoth bones were from erosional contact between Units E and F2. Fossils were overlain by erosional surface dated at 10,900 ± 40 yr BP; occupational surface overlain by organic carbon-rich horizon dated to 10,800 yr BP (Haynes, 1984). Excavated and coil June 1967 (Hem- mings & Haynes, 1969).

AA-834. Escapule mammoth 8500 ± 470 Bone. Comment: 0.6N HC1 insoluble residue from pretreated bone powder. Target C-1259. AA-835. Escapule mammoth 5210 270 Bone. Comment: unpurified gelatin phase. Target C-1275.

AA-836. Escapule mammoth 4610 ± 280 Bone. Comment: XAD-2 purified hydrolyzed weak-HC1-insoluble residue. Target C-1358. General Comment: cancellous tissue from Innominate (6911 /UA3404c) was dated. Sample curated at Univ Arizona. Bone was coated with epoxy preservative (2mm thick) that was physically removed. Underlying cancellous tissue was powdered and washed in ethanol and acetone. Haynes (pers commun) noted that bones had been treated with acetone and methyl ethyl ketone before epoxy was applied. Cancellous tissue: 3.64%C, 0.66%H, !4C Bone Dating with Arizona TAMS 41 0.08%N. Sample was chosen to represent known-age bone with very low organic carbon content and non-collagen amino acid composition; burial was in oxidized clay. Epoxy preservative (2mm thick) was 14C dated to b13C 3680 ± 210 (GX-11261), = -24.2%o (PDB). Epoxy was pretreated (by TWS) with 3 successive 6N HC1 and 1 % NaOH washes.

Del Mar series Charcoal and Chione shell from upper midden (Site W-34) of Del Mar Early Man site (W-34A) on NW point of San Digieto R inlet, Del Mar, San Diego Go, (32° 58' 36" N, 1170 16' 12" W). Samples coil (1974) by R Tyson, during excavation of upper midden (W-34) adjoining loc W-34A that yielded Del Mar Man skull, SDM-16704.

AA-837. Del Mar 3330 220 Charcoal from dm 14. Comment: target C-1284.

AA-838. Del Mar 3520 330 Charcoal from dm 12. Comment: target C-1285.

AA-839. Del Mar 7000 390 Charcoal from dm 11. Comment: target C-1286.

AA-840. Del Mar 4240 300 Charcoal from dm 9. Comment: target C-1294

AA-846. Del Mar 8680 ± 400 Chione shell carbonate, dm 11. Comment: target C-1300.

AA-847. Del Mar 4720 ± 260 Chione shell carbonate, dm 12. Comment: target C-1359.

AA-848. Del Mar 4880 ± 260 Chione shell carbonate, dm 14. Comment: target C-1360.

AA-849. Del Mar 6610 ± 290 Chione shell. Comment: target C-1 361. General Comment: individual charcoal fragments weighed 20 to 10mg; 20 to 406mg charcoal were available from each l0cm level. Charcoal was pretreated 3 times each with 3N HC1(60°C) and 1 % NaOH (60°C) and finally acidified and washed with list H2O. Chione shell carbonate was evolved by using 95% H3PO4. Outer, chalky shell layers were physically removed and remaining hard core etched to half original thickness with 1 N HO. Previ- 14C ous shell carbonate dates from upper midden(W-34) are 4590 ± 60 (LJ- 3175) for dm 3; 5440 ± 70(LJ-3176) for dm 7; 7380 ± 220(LJ-3507) for dm 10 and dm 11, and 9260 ± 100 (LJ-3177) for dm 15; amino acids from '4C Chione shell from dm 10 and 11 of W-34 dated to 12,000 ± 1100 (LJ- 3631) (Masters & Bada, 1977). 42 TW Stafford, Jr et al ACKNOWLEDGMENTS We thank Lisa Warneke and L Toolin for lab assistance. Research was supported by NSF grants EAR-09448 to P Damon and D Donahue, BNS- 82-11864 to A Long, BNS 83-03674 to R Duhamel and grants EAR-82- 16725 and EAR-83-1265 to Vance Haynes, Jr.

REFERENCES Arnold, J R, and Libby, W F,1951, Radiocarbon dates: Science, v 113, p 111-120. Bada, J L, 1985, Aspartic acid racemization ages of California Paleoindian skeletons: Am Anti- quity, v 50, p 645-647. Bada, J L, Schroeder, R A and Carter, G F, 1974, New evidence for the antiquity of man in North America deduced by aspartic acid racemization: Science, v 184, p 791-793. Bada, J L, Gillespie, R, Gowlett, J A Y and Hedges, R E M, 1984, Accelerator mass spectrometry radiocarbon ages of amino acid extracts from California paleoindian skeletons: Nature, v 312, p 442-444. Berger, R and Libby, W F, 1966, UCLA radiocarbon dates V: Radiocarbon, v 8, p 467-497. Berger, R, Homey, A G and Libby, W F, 1964, Radiocarbon dating of bone and shell from their organic components: Science, v 144, p 999-1001. Bischoff, J L and Childers, W M, 1979, Temperature calibration of amino acid racemization: age implications for the Yuha skeleton: Earth Planetary Sci Letters, v 45, p 172-180. Bischoff, J L, Merriam, R, Childers, W M and Protsch R, 1976, Antiquity of man in America indicated by radiocarbon dates on the Yuha burial site: Nature, v 261, p 128-129. Bischoff, J L and Rosenbauer, J, 1981, Uranium series dating of human skeletal remains from the Del Mar and Sunnyvale sites: Science, v 213, p 1003-1005. Broecker, W S, Kulp, L and Tucek, C 5,1956, Lamont natural radiocarbon measurements III: Science, v 124, p 154-165. Childers, W M, 1974, Preliminary report on the Yuha burial, California: Anthropol Jour Canada, v 12, p 2-9. -----1983, The interrelationships of geology, geography and archaeology in the Yuha desert Part III: The Yuha burial: Anthropol Jour Canada, v 21, p 122-127. de Vries, H, 1959, Radiocarbon dating of the Piltdown skull and jaw: Nature, v 184, p 224- 226. Donahue, D J, Jull, A J T, Zabel, T H and Damon, P E, 1983, The use of accelerators for radioisotope dating: Nuclear Instruments & Methods, v 218, p 425-429. El-Daoushy, M F A F, Olsson, I U and Oro, F H, 1978, The EDTA and HCl methods of pre- treating bones: Geol Foren Stockholm Forh, v 100, p 213-219. Figgins, J D, 1933, A further contribution to the antiquity of man in America: Colorado Mus Nat Hist Proc, v 2, no. 2. Gillespie, R and Hedges, R E M, 1983, Sample chemistry for the Oxford high energy mass spectrometer in Stuiver, M and Kra, R S, eds, Internatl 14C conf, 11th, Proc: Radiocar- bon, v 25, no. 2, p 771-774. Gillespie, R, Hedges, RE M and Wand, J 0, 1984, Radiocarbon dating of bone by accelerator mass spectrometry: Jour Archaeol Sci, v 11, p 165-170. Haas, H and Banewicz, J, 1980, Radiocarbon dating of bone apatite using thermal release of CO2 in Stuiver, M and Kra, R S, eds, Internatl 4C conf, 10th, Proc: Radiocarbon, v 22, no. 2, p 537-544. Hassan, A A, Termine, J D and Haynes, C V, Jr, 1977, Mineralogical studies on bone apatite and their implications for radiocarbon dating: Radiocarbon, v 19, no. 3, p 364-374. Haynes, C V, Jr, 1967a, Bone organic matter and radiocarbon dating, in Radiocarbon dating and methods of low level counting: Vienna, IAEA, p 163-168. ----- 1967b, Carbon-14 dates on early man in the New World, in Martin, P S and Wright, H E, eds, Pleistocene extinctions: the search for a cause: New Haven, Yale Univ Press, p 267-286. ----- 1968, Radiocarbon: Analysis of inorganic carbon of fossil bone and enamel: Science, v 161, p 687-688. -----1974, Archaeological geology of some selected paleo-indian suites, in Black, C C, ed, History and prehistory of the Lubbock Lake site: Lubbock, The Mus Jour XV, W Texas Mus Assoc, p 133-139. ----- 1982, Were Clovis progenitors in Beringia?, in Hopkins, D, Mathews, J, Schweiger, C and Young, S eds, Paleoecology of Beringia: New York, Academic Press, p 383-398. -----1984, Stratigraphy and late Pleistocene extinctions in the United States, in Mar- tin, P S and Klein, R G, eds, Quaternary extinctions, a prehistoric revolution: Tucson, Univ Arizona Press, p 345-353. Bone 14C Dating with Arizona TAMS 43

Hemmings, F T and Haynes, C V, 1969, The Escapule mammoth and associated projectile points, San Pedro Valley, Arizona: Jour Ariz Acad Sci, v 5, 184-188. Ho, T Y, Marcus, L F and Berger, R, 1968, Radiocarbon dating of petroleum impregnated bone from tar pits at Rancho La Brea, California: Science, v 164, p 1051-1052. Jull, A J T, Donahue, DJ and Zabel, T H, 1983, Target preparation for radiocarbon dating by tandem accelerator mass spectrometry: Nuclear Instruments & Methods, v 218, 509- 514. Jull, A J T, Donahue, DJ, Hatheway, A L, Linick, T W and Toolin, U, 1986, Production of graphite targets by deposition from CO/H2 for precision accelerator 14C measurements, in Stuiver, M and Kra, R S, eds, Internatl 14C conf, l 2th, Proc: Radiocarbon, v 28, no.2A, p 191-197. Krueger, H W, 1965, The preservation and dating of collagen in ancient bones, in Chatters, R M and Olson, E A, eds, Internatl conf on radiocarbon and tritium dating, 6th, Proc: Clearinghouse for Fed Sci: Techn Inf, Natl Bur Standards, Washington, DC, p 332-327. Lahren, L and Bonnichsen, R, 1974, Bone foreshafts from a Clovis burial in southwestern Montana: Science, v 186, p 147-150. Leonhardy, F D, ed, 1966, Domebo: a Paleo-Indian mammoth kill in the Prairie-plains: Law- son, Oklahoma, Great Plains Hist Assoc, Contr Mus Great Plains No. 1. Leonhardy, F C and Anderson, A D, 1966, Archaeology of the Domebo site, in Leonhardy, F C, ed, Domebo: a Paleo-Indian mammoth kill site in the prairie-plains: Lawson,1)kla- homa, Great Plains Hist Assoc, Contr Mus Great Plains No. 1, p 14-26. Libby, W, 1955, Radiocarbon dating, 2nd ed: Chicago, Univ Chicago Press, 175p. Linick, T W, Jull, A J T, Toolin, L J and Donahue, D J, 1986, Operation of the NSF-Arizona accelerator facility for radioisotope analysis and results from selected collaborative research projects: in Stuiver, M and Kra, R S, eds, Internatl 14C conf, 12th Proc: Radio- carbon, v 28, no. 2A, p 522-533. Longin, R,1971, New method of collagen extraction for radiocarbon dating: Nature, v 230, p 241-242. Masters, P M and Bada, J L, 1977, Racemization of isoleucine in fossil molluscs from Indian middens and interglacial terraces in southern California: Earth Planetary Sci Letters, v 37,p 173-183. C14 May, I, 1955, Isolation of organic carbon from bones for dating: Science, v 121, p 508- 509. Mehl, M G, 1966, The Domebo mammoth: vertebrate paleomortology, in Leonhardy, F C, ed, Domebo: A Paleo-Indian mammoth kill site in the prairie-plains: Lawson, Oklahoma, Great Plains Hist Assoc, Contr Mus Great Plains No. 1, p 27-30. Munnich, K 0,1957, Heidelberg natural radiocarbon measurements I: Science, v 126, p 194- 199. Olson, E A (ms), 1963, The problem of sample contamination in radiocarbon dating: PhD dissert, Columbia Univ. Olsson, I, 1959, Uppsala natural radiocarbon measurements I: Radiocarbon, v 1, p 87-102. Olsson, I U, El-Daoushy, M F A F, Abd-El-Mageed, Al and Klasson, M,1974, A comparison of different methods for pretreatment of bones.I: Geol Foren Stockholm Forh, v 96, p 171- 181. Orr, P C, 1974, Notes on the archaeology of the Winnemucca Caves, 1952-1958: Nevada State Mus Anthropol Papers no. 16, p 46-59. Protsch, R, 1975, The absolute dating of Upper Pleistocene SubSaharan fossil hominids and their place in human evolution: Jour Human Evol, v 4, p 297-322. Sato, J, Sato, T, Otomori, Y and Suzuki, H, 1969, University of Tokyo radiocarbon measurements II: Radiocarbon, v 11, no. 2, p 509-514. Sinex, F M and Fans, B, 1959, Isolation of gelatin from ancient bones: Science, v 129, p 969. Schoeninger, M Y and DeNiro, M J, 1982, Carbon isotope ratios of apatite from fossil bone cannot be used to reconstruct diets of animals: Nature, v 297, p 577-578. Stafford, T W, Jr (ms), 1984, Quaternary stratigraphy, geochronology, and carbon isotope geology of alluvial deposits in the Texas Panhandle: PhD dissert, Tucson, Univ Arizona. Stafford, T W, Jr, Duhamel, R C, Haynes, C and Brendel, K, 1982, Isolation of proline and hydroxyproline from fossil bone: Life Sci, v 31, p 931-938. Stafford, T W, Jr, Jull, A J T, Donahue, D, Brendel, K and Duhamel, R (ms), Studies on the radiocarbon dating of fossil bones by accelerator mass spectrometry and the stable carbon and nitrogen isotopic composition of dated fractions: Paper presented at Internatl 14C conf, 12th, June 24-28, 1985, Trondheim, Norway. Stafford, T W, Jr, Jull, A j T, Zabel, T H, Donahue, DJ, Duhamel, R C, Brendel, K, Haynes, C V, Jr, Bischoff, J L, Payen, LA and Taylor, R E, 1984, Holocene age of the Yuha burial: direct radiocarbon determinations by accelerator mass spectrometry: Nature, v 308, p 446-447. Sullivan, C H and Krueger, H W, 1981, Carbon isotope analysis of separate chemical phases in modern and fossil bone: Nature, v 292, p 333-335. 44 TW Stafford, Jr et al Szabo, B J, 1980, Results and assessment of uranium-series dating of vertebrate fossils from Quaternary alluvium in Colorado: Artic Alpine Research, v 12, p 95-100. Tamers, MA and Pearson, F J, Jr, 1965, Validity of radiocarbon dates on bone: Nature, v 208, p 1053-1055. Taylor, 1), 1969, The Wilsall excavations: an exercise in frustration: Montana Acad Sci Proc, v 29, p 147-150. Taylor, R E, 1982, Problems in the radiocarbon dating of bone, in Currie, L, ed, Nuclear and chemical dating techniques: Washington, D C, Am Chem Soc, p 453-473. ----- 1983, Non-concordance of radiocarbon and amino acid racemization deduced age estimates on human bone, in Stuiver, M and Kra, R S, eds, Internatl 14C conf, 11th, Proc: Radiocarbon, v 25, no. 2, p 647-654. Taylor, R E, Payen, I. A, Gerow, B, Donahue, D, Zabel, T, Jull, T, and Damon, P. 1983, Middle Holocene age of the Sunnyvale human skeleton: Science, v 220, p 1271-1273. Taylor, R E, Payen, I. A, Prior, CA, Slota, P J, Jr, Gillespie, R, Gowlett, J A J, Hedges, R E M, Jull, A J T, Zabel, T H, Donahue, D j and Berger, R, 1985, Major revisions in the Pleisto- cene age assignments for North American human skeletons by C-14 accelerator mass spectrometry: none older than 11,000 C-14 years BP: Am Antiquity, v 50, p 136-140. Trautman, M A and Willis, E H, 1966, Isotopes, Inc. radiocarbon measurements V: Radiocar- bon, v 8, p 161-203. Vogel, J C and Waterbolk, 1963, Groningen radiocarbon dates: Radiocarbon, v 5, p 163- 202. Vogel, j S, Southon, J R, Nelson, I) E and Brown, T A, 1984, Performance of catalytically condensed carbon for use in accelerator mass spectrometry: Nuclear Instruments & Methods, v 223, p 289-293. Wand, J 0 (ms),1981, Microsample preparation for radiocarbon dating: PhD dissert, Oxford Univ. Wehmiller, J F, 1977, Amino acid studies of the Del Mar, California, midden site: apparent rate constants, ground temperature models, and chronological implications: Earth Plane- tary Sci Letters, v 37, p 184-196. Wormington, H M, 1959, Ancient man in North America: Denver, Denver Mus Nat Hist, Pop- ular ser no. 4, 322 p. [RADIOCARBON, VOL 29, No. 1, 1987, P 45-52] COMPARATIVE STUDY OF THE RADIOCARBON DATING OF DIFFERENT BONE COLLAGEN PREPARATIONS

D M GURFINKEL Collegium Archaeometricum, c/o Department of Metallurgy and Materials Science, University of Toronto Toronto, Ontario, Canada M5S 1A4

ABSTRACT. Four different bone collagen preparation procedures were compared and were found useful as a means of assessing the nature of contaminants present in a sample. Weath- ered bone however appeared to contain contaminants that could not be eliminated by any of the procedures studied.

INTRODUCTION In this study the bone samples described were dated by accelerator mass spectrometry (AMS) at the Isotrace Laboratory of the University of Toronto. Four different bone collagen preparation procedures were compared, with the intention of determining the nature of contaminants present and assessing the most reliable age. A comparison was also made between the dating of spongy and compact bone and weathered and unweathered bone.

SAMPLES Three bone samples were used: TO-34: This was a bison metacarpal obtained from the Northern Yukon Research Project of the University of Toronto and found at Old Crow River (locality 11) Yukon Territory. It was from a group of bones several of which had been previously dated by conventional radiocarbon methods although the bison bone obtained here had not been directly dated. TO-30: This was a portion of a whale bone obtained from the Geological Survey of Canada (GSC) which was found at Cape Storm on Ellesmere Island, North West Territory and had been previously dated. TO-33: This was a portion of a whale rib obtained from GSC which was found near Rosse Bay, Ellesmere Island and which had been previously dated. It had been found only partly buried and the exposed portion showed considerable signs of weathering. Because of the small sample size requirement of AMS and the relatively large amount of rib available it was possible to subdivide this sample into three sections: 1) compact bone-good quality (ie, the buried portion of bone) 2) compact bone-weathered (ie, the exposed portion of bone) 3) spongy bone (removed from interior portion of bone, 10cm from one end). This subdivision provided an opportunity to compare both weathered and spongy bone, considered poor dating material, to the better quality compact bone, with the bone sample effectively acting as its own control. A summary of the results of previously obtained ages for these samples is shown in Table 1.

45 46 D M Gurfinkel

TABLE 1 Sample ages obtained by conventional methods

Sample: Age (yr BP) Reference TO-34 Associated samples: Harington (1977) I-7765: 11,910 ± 180 I-3574: 12,460 ± 220 I-7764: 12,275 ± 180 TO-30 Collagen: GSC-1496-A: 9040 ± 120 Blake (1975) GSC-1496-B: 9230 f 140 Lowdon & Blake (1979) GSC-1496-C: 9410 ± 150 GSC-1496-3: 9340 ± 80 GSC-1496-4: 9590 ± 120 Apatite: GSC-1496-2: 8770 ± 240 9040 ± 130 Collagen: U-2510: 9600 ± 120 Olsson & El-Daoushy (1978)

TO-33 GSC-3055: 6920 ± 90 Blake (ms in preparation)

METHODS OF COLLAGEN PREPARATION The basic principle behind any pretreatment is to recover from the sample a fraction that originates from the sample itself, free of carbon-containing environmental contaminants. For bone, collagen is generally considered the preferred dating fraction. Early pretreatment procedures involved the recovery of collagen as the insoluble residue remaining after acid extraction (Tamers & Pearson, 1965; Berger, Homey & Libby, 1964; Sellstedt, Engstrand & Gejvall, 1966). Acid treatment dissolved the bone apatite and decomposed secondary carbonates which constituted a major source of carbon contamination. Humic substances were also recognized as a major source of contamination and as they are largely alkali-soluble (Stevenson, 1982, p 43) NaOH extraction was incorporated as a means of eliminating them from bone (Haynes, 1967). Longin (1971) introduced an additional method for collagen purification which was the conversion of the collagen recovered as an acid-insoluble residue to a soluble gelatin by dissolution in hot water. This was intended to remove hot water-insoluble contaminants such as many humic substances and carbon particulates. Berglund, Hakansson and Lagerlund (1976) incorporated this step after both acid and alkali treatment. Three major steps then, acid extraction, alkali extraction, and gelatin conversion separately or in combination, constitute the general approach taken to the pretreatment of bone. Other pretreatments, of course, have been developed such as EDTA extraction (Olsson et al, 1974) and particularly since the advent of AMS, the isolation of total amino acids and/or specific amino acid fractions (eg, Taylor & Slota, 1979; Donahue, Jull & Zabel, 1984; Gillespie, Hedges & Wand, 1984). '4C Dating of Different Bone Collagen Preparations 47 In order to determine the nature of contaminants present in bone, the major pretreatment steps were incorporated into the scheme shown in Fig- ure T. If secondary carbonates were the major contaminants present, then no difference would be observed between the four collagen preparations as the HCl extraction was common to all. If humic substances were a major and presumably younger-aged contaminant, then methods B, C and D would be expected to yield older ages than method A. If a large proportion of contamination was from highly insoluble and younger-aged material, then methods B and D would yield the oldest ages. Bone samples were prepared by each of the four methods outlined in Figure 1. Only 100mg of starting material were required as the samples, being from Arctic locations, were very well-preserved. Collagen yields are shown in Table 2. A portion of each collagen residue, containing the equivalent of 4 to 6mg carbon, was combusted and acetylene gas synthesized from the resultant carbon dioxide. From each batch of acetylene gas up to four solid carbon targets were prepared by a "cracking" process. The results shown in Figures 2, 3 and 4 are the age determinations obtained from two or three such replicate targets. The combustion, acetylene synthesis and cracking procedures are detailed by Beukens, Gurfinkel, and Lee (1986) and Gurfinkel (ms in preparation).

DISCUSSION OF RESULTS Results obtained for bison bone TO-34 illustrate how the comparison of the four procedures can be used to determine the nature of contami-

Powdered Bone (20 mesh)

Method A Method B Method C Method D 100 mg bone 100 mg bone 100 mg bone 100 mg bone

IN NCI' IN HCI IN HCI IN HCI (5Oml. 3hr, rm (5Om1. 3hr. rm (5Oml. 3hr. rm (50m1. 3hr, rm temp) temp) temp) temp)

filtrate residuerr filtrate residue filtrate residue filtrate residue 'COLLAGEN A' acidified O.1N NaOH O.1N NaOH H20(-pH3) (50m1, 2hr, rm temp) (50m1,2 hr, rm (30m1, lOhr, 90'C) temp)

I 1 filtrateIi residue filtrate residue 'COLLAGEN C' acidified H20(-pH3) residueII filtrate I (30m1.lOhr,90'C) evaporate to dryness + residuefiltrate 'COLLAGEN B' evaporate to (dryness 'COLLAGEN D' * All extractions were performed in glass beakers, with stirring. ** All residues were washed with water to neutralization. Base extract residues were also washed with 1 N HC1 to eliminate any contamination due to atmospheric carbon dioxide. + Filtrate was evaporated under reduced pressure and with gentle heating to a volume of ca 1 ml. This 1 ml volume was transferred to a small test tube and evaporated to dryness under a stream of argon. The test tube was then introduced into the combustion system as described elsewhere (see text). Fig 1. Collagen preparation methods 48 D M Gurfinkel

TABLE 2 Bone collagen yield

Sample % collagen yield* Preparation method A B C D TO-34 10+_1** 9±1 8±1 TO-30 12±1 11±1 TO-33:

Section 1: good quality compact bone 15 ± 1 8 ± 1

Section 2: weathered compact bone 13 ± 1 7 ± 1 Section 3: spongy bone 13 ± 1 12 ± I collagen yields were initially calculated as (wt of collagen residue/initial wt of bone) x 100. However, considerable variability was observed in yields determined this way, attributed to incomplete drying of the collagen residues. For this reason % collagen yields were recalculated based on the amount of carbon dioxide recovered converted to the weight of collagen residue by assuming collagen had a 50% carbon content. These are shown above. The 50% carbon figure was determined by repeated combustion of purified collagen (bovine ten- don, obtained from Sigma Chemical Co) and was consistent with % C as calculated from the known amino acid content of collagen. ** This error estimation was determined by first calculating a single combined error from the errors of each of the directly measured quantities used to calculate yield. This combined error was then doubled as this was considered a reasonable estimate of the additional error in the collagen yield due to small but variable sample losses which occurred during the handling of the residues through the various extraction steps. rants present. It is evident from Figure 2 that younger-aged contaminants were not removed by HC1 extraction alone (Method A) and while the additional gelatin conversion step (Method B) was more effective, alkali extraction was most effective in eliminating the contaminants (Methods C and D). In light of this alkali solubility, these contaminants were most likely humic substances. Because of this, the most reliable age was judged to be the average of Methods C and D: 12,610 ± 70 yr BP, and this was in good agreement with previously obtained dates (Table 1).

11,000 Age (yr 0 BP)

12,000 0

13,000 I A B C 1) Method of collagen preparation

Fig 2. Results for TO-34: bison metacarpal; dotted line indicates average age for Meth- ods C and 1). 14C Dating of Dzferent Bone Collagen Preparations 49

9,000

Age (yr BP)

10,000

A B C I) Method of collagen preparation

Fig 3. Results for TO-30: whale bone; dotted line indicates overall average age.

Although variation in age with chemical treatment was evident, there was little variation in the collagen recovery between the different preparation methods for TO-34 (see Table 2). This was not surprising as only 4% contamination of the collagen residue with modern material having a similar C-content to collagen (or ca 8% contamination with 5000-yr-old material) would be necessary to produce the shift from the maximum to minimum age observed for this sample. Such a level of contamination could not be accurately detected given the error associated with collagen yields in this study (see footnote, Table 2). The ages obtained for TO-30 (Fig 3) as well as the similarity in collagen yields (Table 2) indicate that, for this sample, there was no difference between the preparation procedures. In this case then, the best age esti-

Section 1 Section 2 (good quality (weathered Section 3 compact bone) compact bone) (spongy bone)

6000

Age r (yr f S lse) -4 7000 4 +f .4 S

5000 A B 1) A B C 1) A B C 1) Method of collagen preparation Fig 4. Results for 10-33: whale rib; dotted line indicates average age for each section, bracketed points excluded. See text for explanation. 50 D M Gurfinkel mate was considered to be average of the four preparations, 9350 ± 40 yr BP, which was consistent with other age determinations for this sample (Ta- ble 1). There were three possible reasons for the concordance between the four methods: 1) no major contaminants were present or at least none of an age different from that of the bone. This appeared unlikely as previously obtained apatite dates (see Table 1) were somewhat younger than the collagen dates, suggesting that a younger-aged contaminant was indeed present; 2) the major contaminant present, eg, secondary carbonate, was removed by the HCl extraction which was common to all procedures. This was a likely explanation as the younger-aged contaminant noted above might well have been secondary carbonate; 3) contaminants were present that were not eliminated by any of the procedures used. Unfortunately, the plausibility of this explanation could not be determined from the existing data. The ages obtained for TO-33 (Fig 4), however, did provide an example of the third explanation above, ie, incomplete contaminant removal by all four methods. For this sample three sections were dated and no major differences in age between collagen preparation methods were apparent within each section except Method B in Section 1, which produced an anomalously young date. This result was believed due to an irregular acetylene synthesis process likely caused by a vacuum line leak. The second exception was Section 3, Method C, and an explanation for this follows. The ages obtained for the good quality compact bone, Section 1, which, excluding Method B, averaged 7120 ± 25 yr BP, were in good agreement with the previously obtained age (Table 1) and were assumed to represent the best estimate of the true age of the bone. Compared to these, the weathered bone (Section 2) showed a clear trend toward younger ages, the overall average being 6740 ± 100 yr BP. Both the weathered and good-quality compact bone had similar collagen yields (Table 2). This suggested that the shift in average ages observed was the result of a weathering process which introduced contaminants into the bone but did not cause extensive collagen degradation. Only 2% contamination with modern material (or ca 4% contamination with 5000-yr-old material) would produce the age shift observed here and, as previously noted, with sample TO-34 this was too low a level to be detected from collagen yields. For both bone sections, Method B produced unusually low collagen yields. The reason for this was not clear and may merely represent excessive and coincidental sample losses during preparation. While Method B in Section 1 had an anomalously young age, this anomaly was more likely due to the explanation given above than related to collagen yield. With respect to the spongy bone, all methods except C produced dates consistent with Section 1. While the sections were finely ground to ensure homogeneity, the possibility that the same contaminants present in the weathered bone may have been present in the particular sample aliquot prepared by Method C and were responsible for its younger age cannot be ruled out. The fact that the collagen yield (Table 2) for Method C was some- '4C Dating of DiJJ rent Bone Collagen Preparations 51 what higher than the other three fractions for this section also suggested possible sample heterogeneity.

CONCLUSIONS From the above results the following conclusions can be drawn: 1) the scheme shown in Figure 1 is useful in determining the nature of contaminants present in a sample; 2) the results of TO-34 suggest that alkali-extraction is more effective than Longin's gelatin conversion in removing humic acid contaminants; 3) weathering processes appear to introduce contaminants that are not completely removed by the procedures described here. It should be recognized that the bones used in these experiments were from the Arctic and showed very good collagen preservation. One can only assume that in poorly preserved bone, where contaminants might represent a much higher proportion of total carbon, this problem would be aggravated. Fur- ther investigation in this area is clearly required.

ACKNOWLEDGMENTS The author wishes to acknowledge the financial support of this work by the University of Toronto, the Natural Sciences and Engineering Research Council, Energy, Mines, and Resources of Canada, and Environment Canada. I would like to thank W Blake, Jr, Geological Survey of Canada, and W Irving, University of Toronto, for donating the samples used in this study. Thanks go also to H Polach, Australian National University, for the discussions of the summer of 1983. I wish to extend my appreciation to the able staff of the Isotrace Labo- ratory, especially R Beukens, C Bowen, H Lee, and L Vasserman with whom I worked closely on the radiocarbon dating project and to A E Litherland, Isotrace Director, for encouragement and guidance. Special thanks go to the members of the Collegium Archaeometricum, especially to U M Franklin, whose constant advice, support, and inspiration have been invaluable.

REFERENCES Berger, R, Homey, A G and Libby, W F, 1964, Radiocarbon dating of bone and shells from their organic components: Science, v 144, p 999-1001. Berglund, B E, Hakansson, S and Lagerlund, E, 1976, Radiocarbon-dated mammoth (Mam- muthus primigenius Blumenbach) find in South Sweden: Boreas, v 5, p 177-191. Beukens, R P, Gurfinkel, D M and Lee, H W,1986, Progress at the Isotrace radiocarbon facil- 14C ity, in Stuiver, M and Kra, R S, eds, Internatl conf, 12th, Proc: Radiocarbon, v 28, no. 2, p 229-236. Blake, W, Jr, 1975, Radiocarbon age determinations and post-glacial emergence at Cape Storm, Southern Ellesmere Island, Arctic Canada: Geog Annaler, v 57, ser A, p 1-71. Donahue, R, Jull, A J T and label, T H, 1984, Results of radioisotope measurements at the NSF-University of Arizona tandem accelerator mass spectrometer facility: Nuclear Instruments & Methods Phys Research Sec B, p 162-166. Gillespie, R, Hedges, R E M and Wand, J, 1984, Radiocarbon dating of bone by accelerator mass spectrometry: Jour Archeol Sci, v 11, p 165-170. Harington, C R, (ms) 1977, Pleistocene mammals of the Yukon territory: Ph D dissert, Univ Alberta. Haynes, C V, 1967, Bone organic matter and radiocarbon dating, in Radiocarbon dating and methods of low-level counting: Vienna, Internatl Atomic Energy Agency, p 163-168. 52 I) M Gurfinkel

Longin, R, 1971, New method of collagen extraction for radiocarbon dating: Nature, v 230, no. 5291, p241-242. Lowdon, J A and Blake, W, Jr, 1979, Geological Survey of Canada radiocarbon dates XIX: Geol Survey Canada, Paper 79-7. Olsson, I U and El-Daoushy, M F A F, 1978, Uppsala natural radiocarbon measurements XII: Radiocarbon, v 20, p 469-486. Olsson, I U, El-Daoushy, M, Farid, A F, Abd-El-Mageed, A I and Klasson, M,1974, A comparison of different methods for pretreatment of bones I: Geol Foren Stockholm Forh, v 96, p 171-181. Sellstedt, H, Engstrand, L and Gejvall, N G, 1966, New application of radiocarbon dating to collagen residue in bones: Nature, v 212, p 572-574. Stevenson, F J, 1982, Humus chemistry: Genesis, composition, reactions: New York, Wiley & Sons. Tamers, M A, and Pearson F J, 1965, Validity of radiocarbon dates on bone: Nature, v 208, p 1053-1055. Taylor, R F, and Slota, P, 1979, Fraction studies on marine shell and bone samples for radiocarbon analyses, in Berger, R and Suess, H E, eds, Radiocarbon dating, Internatl 14C conf, 9th, Proc: Berkeley/Los Angeles, Univ California Press, p 422-432. [RADiocARBoN, VOL 29, No. 1, 1987, P 53-56] COMPARISON OF OCEANIC 14C DATA WITH THOSE OF GEOSECS: VERTICAL PROFILES IN 1973 (GEOSECS) AND IN 1980 AT (30° N,170° E) IN THE NORTHWESTERN PACIFIC OCEAN TOSHITAKA GAMO, YOSHIO HORIBE * Ocean Research Institute, University of Tokyo, Minamidai, Nakano-ku Tokyo 164, Japan and HIROMI KOBAYASHI C-14 Dating Laboratory, Faculty of Science, University of Tokyo, Hongo Bunkyo-ku, Tokyo 113, Japan

ABSTRACT. The vertical profile of radiocarbon at (30° N, 170° E) measured in 1980 was compared with the GEOSECS data measured in 1973. 14C was extracted from 200L of sea water, converted to C2H2, and analyzed with a gas proportional counter. Our profile and that of GEOSECS were in good agreement below 700m depth without systematic deviation of &40 values between both measurements. On the other hand, a a14C increase was observed above 700m depth, reflecting the transient addition, in 6.6 years, of bomb 14C to the intermediate layer from the atmosphere.

INTRODUCTION 14C It is well known that is one of the most important geochemical tracers for studying ocean circulation and mixing processes (eg, Broecker & Peng, 1982). The 14C distributions in the Pacific Ocean were obtained in detail during the GEOSECS operation in 1973-1974 (Ostlund & Stuiver, 1980). We re-occupied one of the GEOSECS stations-Station 226 (30°N, 170°E)-during the CYGNUS Expedition (KH-80-2) of R V Hakuho 0140 Maru in 1980 (Horibe,1983), in order to compare our data with those of the GEOSECS group. This paper reports the method for a14C measurement and the result of intercomparison of the &4C profiles.

EXPERIMENTAL Fifteen seawater samples from surface to bottom (as listed in Table 1) were collected by using water-tight PVC 230L water samplers (Horibe, 1981) at Station CYGNUS 11 (30° 32' N,170° 39' E). To measure the precise depth and temperature of each sample, a sonar pinger (Benthos, Model 2216) and reversing thermometers were attached to the sampler. As soon as the sampler was brought on deck, 200L of sea water were transferred into a stainless steel barrel to extract the dissolved total inorganic carbon (CO2). The extraction method was similar to that reported by Ostlund, Dorsey and Rooth (1974). The sea water was acidified with HC1 and stripped of CO2 by circulating C02-free nitrogen carrier gas. The extracted CO2 was absorbed in 800m1 carbonate-free 4N NaOH solution which was prepared by diluting clear saturated solution of NaOH with distilled water just before its use. The flow rate of the carrier gas was

* Present address: Isotope Laboratory, Scripps Institution of Oceanography, University of California, La Jolla 92093 53 54 Toshitaka Gamo et al

TABLE 1 Results of measurements at Station CYGNUS 11 (30° 32' N, 170° 39' E) June 6-7,1980

Sample Depth ASN 14C no. (m) (cpm)

TK-376 10 15.41 0.06 5 TK-375 94 14.90 0.06 5 TK-374 221 14.75 0.05 5 TK-373 467 13.43 0.06 5 TK-372 684 11.95 0.05 4 TK-371 808 10.92 0.05 4 TK-370 1001 10.71 0.05 4 TK-369 1235 10.23 0.05 223 4 TK-368 1559 10.10 0.04 4 TK-367 2022 10.02 0.05 240 4 TK-366 2542 10.02 0.05 0.7 240 4 TK-365 3012 10.27 0.05 0.5 222 4 TK-364 4055 10.31 0.05 0.6 219 4 TK-363 4953 10.33 0.05 0.9 217 4 TK-362 5396 10.47 0.04 206 4

adjusted to ca 6L per minute. By analyzing the remaining CO2 content in the sample sea water with the gas chromatograph (Gamo & Horibe, 1980) at appropriate time intervals during the extraction, it was found that 3 hours were necessary to achieve 95% extraction of CO2. In this study, the extraction was continued for >4 hours for every sample. In the laboratory on land, CO2 was recovered by acidifying the NaOH solution with HC 1, and converted to C2H2 gas for fi counting. Standard gases were prepared from the old NBS oxalic acid by using a wet oxidation method. The ' C measurement was carried out with the gas proportional counter described by Kobayashi et al (1974). A fraction of the C2H2 gas was subjected to quantitative combustion to CO2, and its 813C relative to PDB was measured with the Finnigan MAT 250 mass spectrometer. The Q14C value of the sample was calculated according to the following equation: 014C = (ASN/AABS - 1) x 1000 (%o) 14C where ASN is the net activity of the sample normalized to 313C = - 25%0, and AABS (absolute international standard activity) is 95% of the net oxalic b13C acid activity in AD 1950 which is normalized to = -19%o (Stuiver & Polach,1977). The standard deviation for each z14C data was calculated from statistical error of 3 countings. When the total count is N (counts), its standard deviation is. The counting rate is expressed as N/T (T: counting time) and its standard deviation is N/T2. After the standard deviations for ASN 14C and AABS are obtained as o and o , respectively, that for is calculated by the following equation.

1 = a A 2 x 1000 2 ( AABS 1) 2+ ( SN 2) A ABS Comparison of Oceanic 14C Data 55

TABLE 2 Results of the NBS standard measurements

Aox a13C AAas NBS no. (cpm) (%o) (cpm)

NBS-42 13.02 ± 0.06 0.06 -42 13.08 ± 0.07 0.07 -43 13.14 ± 0.05 0.05

RESULTS AND DISCUSSION Two standard gases (NBS-42 and NBS-43) were prepared for this study. Table 2 shows the results of the standard gas measurements: their b13C, net activity (A0X), and AABS. The standard gases were measured three times at intervals in the course of successive sample measurements. The good reproducibility of the value of AABS (average value: 12.55 ± 0.03 cpm) reflected the stability of the counting system throughout the study. The results of sample measurements and calculated 14C values were 0140 listed in Table 1. Figure 1 is the vertical profile at CYGNUS Station 11 together with that at GEOSECS Station 226 (Ostlund & Stuiver,1980). The former samples were collected June 6-7, 1980, and the latter November 9, 1973. In deep waters below 700m depth, both profiles are in good agreement with each other within the analytical error as indicated by a solid line

,14C ( °i°° ) -200 -100 100 at

III ,o. 4. ,.0.

F T .

3 . f , GEOSECS 226 (NOV 9, 1973 a- CYGNUS 11 (JUN 6-7,1980) 0w r 4I

e r

Fig 1. Comparison of the vertical profiles of &4C between GEOSECS 226 (30° 34' N, 170° 38' E) in 1973 and CYGNUS 11 (30° 32' N, 170° 39' E) in 1980. The solid line was arbi- trarily drawn to show good agreement between both profiles below 700m depth. 56 Toshitaka Gamo et al in Figure 1. Thus, there is no systematic deviation between our &4C measurements and those of the GEOSECS group. As shown in Figure 1, above 700m depth our 1980 014C data at 221 m and 467m depths are significantly greater than the 1973 GEOSECS values interpolated at the same depths. Since it was shown above that there is no analytical bias between both measurements, the increase of &4C is real, possibly reflecting the addition of bomb 14C from the atmosphere. As the o14C increase was limited in the intermediate depth range and no change was observed in the surface water, it is supposed that the bomb 14C was supplied to the intermediate water by transport of sinking surface sea water in the northern region of the Pacific. The similar increase of &4C in the intermediate layer compared with the GEOSECS data was observed also at (25° N,170° E) (GEOSECS 227, CEPHEUS 5) and at (12° 45' N, 173° 14' E) (GEOSECS 229, CEPHEUS 8) during the CEPHEUS Expedition of R V Hakuho Maru in 1982 as presented by Horibe and Gamo (1983). The transient behavior of 14C in the intermediate layer and its oceanographic meanings will be discussed elsewhere in detail together with other results so far obtained in the northwestern Pacific Ocean.

ACKNOWLEDGMENTS The authors thank M Ishiguchi and M Koizumi for 14C measurements. Thanks are also due the officers and crew of R V Hakuho Maru for their help in large-volume water sampling. This work was supported by grants- in-aid (Nos. 57540321, 56117009, 57110008, and 58102009) from the Ministry of Education, Culture and Science to the University of Tokyo.

REFERENCES Broecker, W S and Peng, T -H, 1982, Tracers in the sea: New York, Eldigio Press, 690 p. Gamo, T and Horibe, Y, 1980, Precise determination of dissolved gases in sea water by ship- board gas chromatography: Bull Chem Soc Japan, v 53, p 2839-2842. Horibe, Y, 1981, Preliminary report of the Hakuho Maru Cruise KH-77-3 (Pegasus Expedi- tion): Ocean Research Inst, Univ Tokyo, 55 p. -----1983, Preliminary report of the Hakuho Maru Cruise KH-80-2 (CYGNUS Expe- dition) and the Hakuho Maru Cruise KH-82-1 (CEPHEUS Expedition): Ocean Research Inst, Univ Tokyo, 78 p. Horibe, Y and Gamo, T, 1983, Tracer studies on the transport of carbon dioxide into the ocean in the central North Pacific (abs): IUGG General Assembly, 18th, Hamburg. Kobayashi, H, Hirose, T, Sugino, M and Watanabe, N, 1974, University of Tokyo radiocarbon measurements V: Radiocarbon, v 16, no. 3, p 381-387. Ostlund, H G, Dorsey, H G, and Rooth, C G, 1974, GEOSECS North Atlantic radiocarbon and tritium results: Earth Planetary Sci Letters, v 23, p 69-86. Ostlund, H G and Stuiver, M,1980, GEOSECS Pacific radiocarbon: Radiocarbon, v 22, no. 1, 25-53. p '4C Stuiver, M and Polach, H A, 1977, Discussion: Reporting of data: Radiocarbon, v 19, no. 3, p 355-363. [RADIOCARBON, VOL 29, No. 1, 1987, P 57-60] ANTWERP UNIVERSITY RADIOCARBON DATES IV R VANHOORNE and A D DUBOIS Department of General Botany, University of Antwerp (RUCA) Groenenborgerlaan 171, B - 2020 Antwerp, Belgium

Samples processed since the last list was published (R, 1978, v 20, no. 3, p 192-199) are reported here. The dates were obtained by liquid scintillation counting of benzene, using laboratory procedures outlined in previous articles (R, 1976, v 18, no. 2, p 151-160; R, 1977, v 19, no. 3, p 383- 388).

GEOLOGIC SAMPLES Belgium Booitshoeke series Peat and clay layers in W Belgian coastal plain (51 ° 05' 34" N, 2° 44' 03" E). Coll and subm Nov 1978 by R Vanhoorne and A D Dubois.

ANTW-316. Booitshoeke 1 3210 ± 50 Base of upper peat layer containing numerous reed rhizomes (Phrag- mites australis), 5cm thick, 125 to 130cm depth. Peat is overlain by clay. Comment: date is ca 530 yr younger than IRPA-286 (Baeteman et al, 1979; Baeteman, 1981; Dauchot-Dehon, Van Strydonck & Heylen, 1981a) in same strat position, but containing no reed rhizomes.

ANTW-315. Booitshoeke 2 2950 ± 60 Extracted humic acids from ANTW-316. Humic acids were obtained by boiling peat sample with NaOH, followed by their precipitation by HC 1 2N treatment. Comment: small difference between ANTW-315 and -316 reflects only minor percolation of humic acids through overlying clay.

ANTW-317. Booitshoeke 3 3680 ± 50 Wood (Alnus) in same peat layer as ANTW-316, but l Om NW. Com- ment: date agrees with IRPA-286 (Baeteman et al, 1979; Baeteman, 1981; Dauchot-Dehon, Van Strydonck & Heylen, 1981 a), in same strat position.

ANTW-321. Booitshoeke 4 3130 ± 50 Reed rhizomes from intercalated clay layer, 10cm thick, between upper and lower peat layers, 130 to 140cm depth, immediately underlying ANTW-316. Comment: reed rhizomes are ca 565 yr younger than wood in overlying peat. Preliminary diatom analysis of this clay layer, which was previously observed in borings in that area (Stockmans & Vanhoorne, 1954), indicates that it was deposited under marine tidal flat conditions (L Denys, pers commun, 1985). Strat position and sedimentary environment support opinion of C Baeteman (Baeteman et al, 1979) that this clay layer resulted from Calais IV-B transgression.

57 58 R Vanhoorne and A D Dubois

General Comment: ages of samples containing reed rhizomes (ANTW-316 and -321) reveal that they did not originate from initial colonization after clay deposition, but actually belonged to plants growing several hundreds of years after initiation of upper peat layer. Explanation by deviating carbon isotope fractionation must be rejected, as Phragmites australis belongs to C-3 (Calvin cycle) group of plants, needing little or no correction (Ler- man, 1973). The same explanation, ie, presence of younger reed rhizomes contaminating peat sample, is proposed for unusually young date of IRPA-345 (3200 ± 200 BP) originating from base of upper peat layer (R,1981, v 23, p 346).

ANTW-304. Oostkerke 4180 ± 60 Clayey peat, containing many reed rhizomes, from boring 374 (Belgian Geol Survey) at 505 to 510cm depth (base of peat layer) in W Belgian coastal plain (51 ° 20' 54" N, 2° 44' 54" E). Coll and subm Aug 1978 by C Baete- man and A Tassanasorn. Comment: date is ca 400 yr too young compared with other samples from same layer, ie, ANTW-249, -102, IRPA-282, -288, Hv-8794 (Baeteman, 1981). However, it closely resembles date of IRPA- 292 (R, 1981, v 23, p 33), from same layer, also containing reed rhizomes. This reveals that reed rhizomes did not originate from initial vegetation of peat fm. This conclusion agrees with findings on Booitshoeke series.

Beerse series Peat and extracted humic acids from Beerse (51° 20' 22" N, 40 48' 31" E). Coll and subm Nov 1978 by R Vanhoorne and A D Dubois.

ANTW-310. Beerse 1 9460 ± 80 Peat from 264 to 276cm depth.

ANTW-311. Beerse 2 6080 ± 240 Extracted humic acids from ANTW-310. Humic acids were extracted as above (ANTW-315). General Comment: from strat point of view, peat corresponds to Late Glacial layers described by De Ploey (1961), Vanhoorne (1963), and Riezebos, Slotboom and Vanhoorne (Lithological expression of a palynologically deduced climatic oscillation during the Allerod, ms in preparation) in same clay pit. Obtained date is >1000 yr BP younger than youngest date obtained by Riezebos, Slotboom and Vanhoorne (ms in preparation). This can be explained by presence of younger rootlets within peat. Relatively young date for extracted humic acids reflects their good percolation through sandy soil, overlying peat layer. Salix and Juniperus wood from peat layer is currently being analyzed for its deuterium content (Dubois, The deuterium hydrogen ratio in cellulose as a climatological and a paleoclimatological indicator, ms in preparation). Antwerp University Radiocarbon Dates IV 59 Wuustwezel series Peat and extracted humic acids from Wuustwezel (51 ° 25' 15"N, 4° 35' 43" E). Coil and subm June 1978 by R Vanhoorne and A D Dubois.

ANTW-300. Wuustwezel 1 9930 ±130 Amorphous peat, 104 to 106cm depth. No rootlets visible.

ANTW-308. Wuustwezel 2 9100 ± 30 Amorphous peat, 104 to 106cm depth, adjoining ANTW-300. No rootlets visible. Comment: significantly different age from ANTW-300 reveals contamination probably caused by younger rootlets indistinguish- able from rest of peat.

ANTW-303. Wuustwezel 3 10,370 ± 80 Amorphous peat, 104 to 106cm depth. Visible rootlets removed.

ANTW-313. Wuustwezel 4 8690 ± 90 Extracted humic acids from sample ANTW-303. Humic acids were extracted as above (ANTW-315). General Comment: pollen diagram, characterized by dominance of Pinus and Betula without any thermophilous trees and low NAP percentages reflects Late Glacial forested landscape, which should be set in Allerod. According to generally accepted dates for Allerod in Europe, date was expected to be 3000 yr older. This error is probably caused by occurrence of rootlets in peat, belonging to younger vegetation.

ANTW-305. Wuustwezel 5 7240 ± 80 Slightly clayey amorphous peat, 106 to 111 cm depth. No rootlets visible. Comment: on palynol grounds, open park landscape with a few specimens of Betula and Salix is inferred. However, Salix and Betula may represent dwarf specimens, playing no role in forestation. No thermophilous trees were observed. NAP pollen, in which Artemisia is more important than in Wuustwezel 4, attains almost 70%. According to palynol results, this peat should belong to Older Dryas. Date obtained is 5000 yr too young. There is no doubt that peat was contaminated.

REFERENCES Baeteman, C, (ms) 1981, De Holocene ontwikkeling van de Westelijke Kustvlakte (Belgie): Phi) dissert, Univ Brussels, 297 p. Baeteman, C, Verbruggen, C, with Dauchot-Dehon, M, Heylen, J and Van Strydonck, M, 1979, New approach to the evolution of the so-called surface peat in the western coastal plain of Belgium: Geol Service Belgium, Prof paper, v 11, no. 167. Dauchot-Dehon, M, Van Strydonck, M and Heylen, J, 1981a, Institut Royal du Patrimoine Artistique radiocarbon dates VII: Radiocarbon, v 23, no. 1, p 33-37. ----- 1981b, Institut Royal du Patrimoine Artistique radiocarbon dates VIII: Radio- carbon, v 23, no. 3, p 345-351. De Ploey, J, 1961, Morfologie en kwartair-stratigrafie van de Antwerpse Noorderkempen: Acta Geog Lovaniensia, Leuven, v 1, 130 p. 60 R Vanhoorne and A D Dubois

Lerman, J C,1973, Carbon 14 dating: Origin and correction of isotope fractionation errors in terrestrial living matter, in Rafter, T A and Grant-Taylor, T, eds, Internatl conf on radiocarbon dating, 8th, Proc: Wellington, Royal Soc New Zealand, p HI 6-H28. Stockmans, F and Vanhoorne, R, 1954, Etude botanique du gisement de tourbe de la region de Pervijze (Plaine maritime Belge): Mem Inst Royal Sci Nat Belgique, v 130, 144 p. Vanhoorne, R, 1963, Le niveau d'Allerod de Beerse (Campine Belge): Grana Palynol, v 4, p 449-451. Vanhoorne, R and Van I)ongen, W, 1976, Antwerp University radiocarbon dates I: Radiocar- bon, v 18,00.2, p 151-160. Vanhoorne, R and Van Strydonck, M, 1977, Antwerp University radiocarbon dates II: Radio- carbon, v 19, no. 3, p 383-388. Vanhoorne, R, Van Strydonck, M and Dubois, A 1), 1978, Antwerp University radiocarbon dates III: Radiocarbon, v 20, no. 2, p 192-199. [RADIOCARBON, VOL 29, No. 1, 1987, P 61-77] BRITISH MUSEUM NATURAL RADIOCARBON MEASUREMENTS XIX JANET AMBERS, RICHARD BURLEIGH*, and KEITH MATTHEWS

Research Laboratory, The British Museum, London WC 1 B 3DG, England The following list consists of dates for archaeologic samples mostly measured from June 1984 to June 1985. The dates were obtained by liquid scintillation counting of benzene using the laboratory procedures outlined in previous lists (see, eg, BM-VIII, R,1976, v 18, p 16). Dates are expressed 14C as suggested by Stuiver and Polach (1977), ie, in years relative to AD 1950, based on the Libby half-life for 14C of 5570 yr, and corrected for isotopic fractionation (OC C values are given relative to PDB). No corrections have been made for natural 14C variations. The modern reference standards are the NBS oxalic acids (SRM 4990 and RM 49). Errors quoted are the counting error for the sample, combined with an estimate of the errors contributed by the modern and background samples. This estimate includes both counting and non-counting errors, and is computed from differences in the overall count rates observed among the individual backgrounds and moderns. The overall error is given as ± 1 standard deviation (±1a . Descriptions, comments, and references to publications are based on information supplied by submitters. British isles England Roxby series Charcoal from three similar and adjoining houses (Inman et al, 1.985) from site at Roxby, near Whitby, N Yorkshire (54° 30' N, 0° 50 W, Natl Grid Ref N7 762143). Coll 1982 and subm by D Spratt, York, to establish chronol for site.

1950 ± 150 BM-2207A. Roxby S13C = -23.994o Charcoal from post hole in House no. 3, assoc with Iron Age pottery.

7090 ± 120 BM-2208A. Roxby 3«C = -23.9%o Charcoal from House no. 2 assoc with metallurgic debris. General Comment (DS): BM-2207A agrees with finds from houses (Stanwick- type pottery, beehive quern and glass). BM-2208A is much earlier than expected but comes from area assoc with metal working and probably reflects age of peat or local coal-like material used for fuel. Sample from House no. 1 proved too small for dating.

* Current address: Ethnography Department of the British Museum, Museum of Man- kind, Burlington Gardens, London WIX 2EX, England.

61 62 J Ambers, R Burleigh, and K Matthews Down Farm Pond Barrow series Samples from pond barrow near Down Farm, Woodcutts, Dorset (50° 55' N, 2° 00' W, Nat! Grid Ref SU 002137). Col! 1980 and subm by R Bradley, Dept Archaeol, Univ Reading.

3390 ± 45 BM-2189. Down Farm Pond Barrow b13C = -24.4%o Charcoal, ref F2 L4, (Quercus sp, heartwood >15 yr old) id. by M Rob- inson, Univ Mus, Oxford, from base of pit in cluster of burials (Cluster 1) at edge of pond barrow, assoc with bronze awl and human cremation burial.

3210 ± 45 BM-2190. Down Farm Pond Barrow b13C = -24.1% Charcoal, ref F12 L9, (Quercus, Prunus, Crataegus type, Fraxinus spp, all from young wood 5-25 yr old) id. by M Robinson, from base of stepped pit in cluster of features (Cluster 1) on edge of pond barrow, assoc with Collared Urn containing human cremation, 2 bone awls, and bronze awl.

3670 ± 60 BM-2191. Down Farm Pond Barrow b13C = -24.3%o Charcoal, ref F28 L2, (Quercus sp, >15 yr old) id. by M Robinson, from middle fill of pit in cluster of features (Cluster 2) beside pond barrow, assoc with middle Beaker pottery.

3110 ± 100 BM-2192. Down Farm Pond Barrow b13C - -26.1%o Charcoal, ref PH 10, (Rhamnus type) id. by M Robinson, from fill of 1 of 4 post holes in center of pond barrow.

3190 ± 70 b13C BM-2324. Down Farm Pond Barrow = - 24.1 %o Charcoal, ref F30, (Crataegus type, 5-25 yr old) id. by M Robinson, from fill of stepped pit containing unaccompanied human cremation at edge of barrow (Cluster 2).

3870 ± 50 BM-2325. Down Farm Pond Barrow b13C = -25.2%o Charcoal, ref F31 L2, (Quercus sp, >15 yr old) id. by M Robinson, from middle fill of pit on edge of pond barrow (Cluster 2), assoc with middle Beaker sherds similar to material from upper fill of Dorset Cursus, 40m away.

3570 ± 40 BM-2326. Down Farm Pond Barrow b13C = -23.5%0 Charcoal, ref F3 (Quercus sp, >15 yr old) id. by M Robinson, from fill of pit on edge of pond barrow, assoc with enlarged Food Vessel containing human cremation. British Museum Natural Radiocarbon Measurements XIX 63

3450 ± 50 BM-2327. Down Farm Pond Barrow b13C = -24.3%o Charcoal, ref PH8 (Fraxinus sp, 5-25 yr old) id. by M Robinson, from fill of 1 of 4 post holes at center of pond barrow, assoc with sherds of plain Food Vessel. General Comment (RB): samples date Beaker settlement cut by features of pond barrow (BM-2191, -2325) and features belonging to pond barrow itself. Two Beaker dates (BM-2191, -2325) are consistent with general sequence of Beaker domestic pottery, whilst 6 dates for elements of pond barrow form consistent series in later part of Wessex Early Bronze Age. Site is first pond barrow to be excavated for >30 years, and these are only dates available for this type of monument.

Badshot series Bone and antler samples from primary silts in N ditch (Tr B.IV. Layer 3) of Badshot Long Barrow, at Badshot, Runfold, Surrey (51 ° 10' N, 00 50' W, Natl Grid Ref SU 860480). Coll 1936 to 1937 and subm 1983 by J Cotton, W London Field Unit, Mus London, to date construction of ditch of only known long barrow in Surrey.

4420 ± 90 BM-2272. Badshot b13C = -22.6%o Collagen from bone fragments, ref B.IV.3 (G).

4480 ± 100 BM-2273. Badshot b13C = -22.9%o Collagen from bone fragments, ref B.IV.3 (F).

4600 ± 120 BM-2274. Badshot b13C = -21.4%o Collagen from beam of antler, (red deer) id. by J W Jackson, Univ Man- chester, ref B.IV.3. General Comment (JC): dates constitute useful series for only Neolithic com- munal monument yet discovered in old Surrey. BM-2273 and-2274, from primary chalk silting of N ditch, indicate that monument lies toward latter end of date range for earthen long barrows. BM-2272 from secondary silt of "red brown loam" at E terminal of N ditch gives date for sherds of simple large Mortlake bowl.

Maumbury Rings series Antler picks (red deer) id. by H St George Gray from shafts at Maum- bury Rings (Bradley, 1975), Dorchester, Dorset (50° 40' N, 2° 30' W, Nail Grid Ref SY 691899). Coll 1908 to 1913 by H St George Gray and subm 1983 by R Bradley, Dept Archaeol, Univ Reading, for Dorchester Mus. 64 J Ambers, R Burleigh, and K Matthews 3650 ± 70 BM-2281. Maumbury Rings b'3C = -23.3%o Collagen from red deer antler, ref sample 4 (excavator's no. 160), from uppermost fill of Shaft 3.

3640 ± 70 BM-2282. Maumbury Rings b'3C = -22.3%o Collagen from red deer antler, ref sample 1 (site no. 60), from bottom of Shaft 1. General Comment (RB): two dates confirm excavator's impression that shafts had filled up rapidly, and suggest that site might be contemporary with stone phase at Mount Pleasant (Wainwright, 1979) nearby, rather than primary earthwork on site. Maumbury Rings now appears to be late satellite of massive post circle recently discovered under modern town of Dorchester (P Woodman, pers commun). Wor Barrow series Antler (red deer) id. by A Pitt Rivers from lower part of primary silt in ditch of long barrow at Wor Barrow (Pitt Rivers, 1898), Sixpenny Handley, Dorset (51 ° 00' N, 2° 00' W, Nat! Grid Ref SU 012173). Coll 1893 to 1894 by A Pitt Rivers and subm 1983 by R Bradley for Salisbury Mus. Barrow is one of distinctive oval type found in Cranborne Chase area. Samples give terminus ante quern for barrow construction.

4350 ± 70 b'3C BM-2283. Wor Barrow = -21.9%o Collagen from base of red deer antler, ref sample 1, possibly used as pick, marked "Depth 11 ft. Bottom of SE ditch."

4440 ± 70 b'3C BM-2284. Wor Barrow = -21.D%o Collagen from fork of red deer antler, ref sample 2, marked "Depth 10.5 ft. On bottom of ditch." General Comment (RB): original site report suggests that BM-2284 dates antler on bottom of long barrow ditch and BM-2283 sample within its lowest silts. Dates agree well with other evidence that oval barrows and related enclosures were being built as late as 5th millennium BP (Bradley et al, 1984). Oldbury series Samples from hilifort (Ward Perkins,1944) at Oldbury, near Ightham, Kent (51 ° 10' N, 0° 10' E, Nat! Grid Ref TQ 590580). Coll 1983 and subm by F H Thompson, Soc Antiquaries, London.

2310 ± 50 b'3C BM-2290. Oldbury = -25.6%o Charcoal, (Quercus sp) id. by Janet Ambers, from small hearth or fire in old ground surface beneath main rampart. British Museum Natural Radiocarbon Measurements XIX 65

1840 ± 40 BM-2291. Oldbury b13C = -27.5%o Charcoal, (family Rosaceae, subfamily Pomoideae) id. by Janet Ambers, from small hearth in interior of hillfort near S end.

1910 ± 80 BM-2292. Oldbury b13C = -25.5%o Charcoal, (Corylus sp) id. by Janet Ambers, from hearth in Tr 9. General Comment (FHT): BM-2290 is somewhat earlier than hoped but can only give terminus post quem for rampart construction. Barbed and tanged arrow-head from separate area of old ground surface indicates activity on site earlier than Iron Age. Hearth from which BM-2291 was taken contained no finds; and date, which is later than expected, could conceivably relate to Roman activity. BM-2292 was assoc with trench originally believed to belong to hut, but now known to be part of linear feature containing quantities of late Iron Age pottery with which result agrees.

Springfield series Samples from late Bronze Age enclosure at Springfield Lyons, Spring- field, Essex (51 ° 45' N, 0° 30' E, Nat Grid Ref TL 734082). Coll 1983 and subm by D G Buckley, Essex Co Council.

2780 ± 90 BM-2313. Springfield b13C = -26.8%o Charcoal, ref RC1, (Acer sp) id. by Paula Rudall, Royal Botanic Gar- dens, Kew, from Context 5532, layer within enclosure ditch, assoc with bronze-casting clay molds.

2370 ± 80 BM-2314. Springfield b13C = -23.1%o Charcoal sample, ref RC3, (mixed Quercus sp, Acer sp) id. by Janet Ambers, from Context 5153, secondary fill of enclosure ditch assoc with burned clay and vesiculated pottery. General Comment (DB): BM-2313 provides date for lower fills of enclosure ditch and agrees with postulated date for construction. It also gives date for deposition of bronze casting molds found at bottom of ditch and belonging to Ewart Park traditions. Date is directly comparable with 2 dates from middle silts of outer enclosure ditch at Mucking South Rings (Har-1634, 2770 ± 110 BP and Har-1708, 2810 ± 70 BP ( Jones & Bond, 1980, p 475). BM-2314 provides date for large group of pottery derived from higher ditch silts, comprising late Bronze Age/early Iron Age forms.

Ireland Mount Gabriel series Samples from copper mining site at Mount Gabriel, Co Cork, Eire (51 ° 30' N, 9° 30' W). Measured to confirm age of site of presumed early 66 J Ambers, R Burleigh, and K Matthews Bronze Age mining (Jackson, 1968, 1980, 1984a,b; Briggs, 1983). Com- ment by P T Craddock, Research Lab, British Mus. 3200 ± 110 BM-2271. Mount Gabriel b13C = -24.9%o Peat, ref MG 4, from top of waste heap of mine no. 6. Coil and subm by J Jackson, Dept Geol, Trinity Coil, Dublin. 3130 ± 80 BM-2336. Mount Gabriel b13C = -26.1%o Wood, ref MG-24-0l, from sediment in entrance to mine MG-24, sealed under peat deposit and assoc with stone tools. Coil 1984 and subm by W O'Brien, Dept Archaeol, Univ Coil, Cork. General Comment (PTC): dates confirmed existing date of 3450 ± 120 BP (VRI-66; R, 1970, v 12, p 316). This, together with rest of archaeol and environmental evidence, had been called into question by Briggs (1983) who regarded whole operation as 19th century site. These results suggest Briggs' hypothesis to be untenable, and that mines, together with grooved stone hammers found with them, are correctly assigned to Bronze Age, as elsewhere (Craddock, 1986).

Isle of Man Peel Castle series Human bone samples from cemeteries at Peel Castle, Peel (54° 10' N, 4° 40' W, Natl Grid Ref SU 242846). Coll 1983 and subm by D Freke, St Patrick's Isle Trust. Comment by A F Roberts, Univ Liverpool.

170 ± 50 BM-2303. Peel Castle b13C = -20.3%o Collagen from human long bone, ref PCB 51.

150 ± 40 b'3C BM-2304. Peel Castle = -21.4%o Collagen from human long bone, ref PCB 63.

630 ± 45 BM-2305. Peel Castle b13C = -19.8%o Collagen from human long bone, ref 82-150/A 352.

730 ± 50 BM-2306. Peel Castle b13C = -20.2%o Collagen from human long bone, ref 82-150/A 416. General Comment (AFR): dates are later than expected. BM-2303, -2304 come from area for which there is documentation and oral commun of other use in cal AD 1645 to 1950 (Klein et al, 1982). BM-2305, -2306 come from area believed to be pre-Viking, Celtic monastery of 8th century. BM- 2305 was overlain by hearth archaeomagnetically dated to ca AD 1150. British Museum Natural Radiocarbon Measurements XIX 67 Scotland Strichen series Samples from Recumbent Stone Circle at Strichen, Grampian (57° 35' N, 2° 05' W, Nat! Grid Ref NJ 937545). Coll 1981 and subm by P Abramson, Leeds.

2150 ± 60 BM-2315. Strichen b13C = -27.2%o Charcoal, ref 2b/5T81, (Alnus sp) id. by Janet Ambers, from bottom of shallow gully inside stone circle, sealed by rubble.

3090 ± 60 BM-2316. Strichen b13C = - 25.7%o Charcoal, ref 3/ST 81, (Alnus sp) id. by Janet Ambers, from bottom of disturbed area containing cremation of adult female, assoc with coarse-fab- ric sherds.

2050 ± 80 BM-2317. Strichen b13C = -25.5%o Charcoal, ref 5 ST/81, (Alnus sp) id. by Janet Ambers, from fill of shallow pit regarded as stone hole of prehistoric circle. General Comment (PA): BM-2315 and -2317 are too recent to be regarded as coming from stone circle and must come from later but circle. BM-2316 is from cremation assoc with stone circle and was probably inserted into bank of circle towards end of period of use.

Cyprus Lemba Lakkous series Samples from occupation site at Lemba Lakkous, 4.5km NNW of Paphos, Paphos dist (34° 50' N, 32° 20' E). Coll 1982 and subm by E J Pel- tenburg, Lemba Archaeol Proj, Cyprus, and Dept Archaeol, Univ Edin- burgh, as part of Lemba Archaeol Proj (cf Kissonerga Mosphilia, below; Peltenburg, 1985; and BM-1353, -1354, -1473 to -1476, and -1539 to -1543; R, 1982, v 24, p 238-239).

3930 ± 100 BM-2278. Lemba Lakkous b13C = -24.7%o Charcoal, from pit M34c.3aF9.

5710 ± 100 BM-2280. Lemba Lakkous b13C = - 24.2%o Charcoal, from M33d.7-10. General Comment (EJP): by comparison with other late Neolithic dates from Cyprus, BM-2280, which comes from Period 1 context, is too early and may 68 J Ambers, R Burleigh, and K Matthews derive from land clearance. BM-2278 is only date from Period 2 context and cannot be differentiated from late Chalcolithic Period 3 dates.

4030 ± 110 BM-2279. Kissonerga Mosphilia b13C = -23.00oo Charcoal, ref 82.11-12, from destruction debris of Bldg 3.2, Kisson- erga Mosphilia, Paphos dist (34° 50' N, 32° 25' E). Coll 1982 and subm by E J Peltenburg as part of Lemba Archaeol Proj (cf Lemba Lakkous, above, Peltenburg, 1985; and BM-1353, -1354, -1473 to -1476, and -1539 to 1543; R,1982, v 24, p 238-239). Comment (EJP): date confirms comtempo- raneity of major occupation with that at nearby Lemba.

2090 ± 50 b13C BM-2294. Kyrenia Ship = -24.8%o Pitch, ref Sample KS A, (probably from Aleppo pine, Pinus halepensis), from waterproof lining of late 4th century BC Rhodian amphora from cargo of submerged wreck of Greek merchant ship sunk ca 300 BC 1.6km NE of Kyrenia (35° 20' N, 33° 20' E). Co111969 (sampled 1982) and subm 1984 by M L Katsev, Inst Nautical Archaeol, Am School Classical Studies, Athens. Date of manufacture and filling of amphora expected to be closely contemporaneous with sinking of ship (see BM-1588, -1588A, -1639, R, 1982, v 24, p 239-240). Comment (RB): result agrees statistically with previous dates BM-1588, -1588A (R, 1982, v 24, p 239-240) and P-1621, -1622 (R, 1971, v 13, p 363-364) for wood from ship and almonds from cargo, suggesting pitch sample does not show initial age effect. Calibrated mean date range of 375 BC to AD 25 (Klein et al, 1982) agrees broadly with expected historic date of wreck (ca 310 to 300 BC).

Prance Les Eyzies series Collagen from fragmentary animal bone (probably bison and horse) excavated by R Chappell, Dept Conservation, British Mus, from opposite sides (upper and lower surface) of mass of breccia 30cm thick from floor of Grotte des Eyzies, Les Eyzies, Dordogne (44° 35' N, 1° 0' E), representing late Magdalenian occupation debris. Coll ca 1860 by E Lartet and H Christy and subm 1984 by G de G Sieveking, Dept Prehist and Romano- British Antiquities, British Mus, from specimen in British Mus colln since ca 1865 (Lartet & Christy, 1875). Recently, complete separation of components of breccia has revealed engraved bone pieces, antler biserial points, bone needles, and flint artifacts, and faunal analysis, in particular of rein- deer bone, has provided detailed evidence of hunting and butchery practices. Samples dated to provide chronol basis for this investigation (Am- bers, J & Burleigh, R, Radiocarbon dating of the Magdalenian habitation refuse, in Sieveking, G de G, ed, Les Eyzies mon, ms in preparation).

11,600 ± 380 BM-2285. Les Eyzies b13C = -20.4%o Collagen, ref Sample 1(A). British Museum Natural Radiocarbon Measurements XIX 69

12,590 ± 980 BM-2286. Les Eyzies b13C = 20.2%o Collagen, ref Sample 2(B). General Comment (RB & JA): difference between dates is not statistically significant (large error of BM-2286 arose from small amount of sample available). Mean age of 11,730 ± 350 BP agrees well with other dates for late Magdalenian in Perigord region (cf BM-302, 11,750 ± 300 BP, BM-304, 12,070 ± 180 BP, R,1969, v 11, p 283).

Montgaudier series

Samples from Palaeolithic occupation of cave complex at Montgaudier (Duport, 1976; 1983), Commune Montbron, near Angouleme, Charente (45° 40' N, 0° 30' E). Coll by L Duport, Dept Archaeol, Angouleme and subm by G de G Sieveking, to date assoc works of art, hearths, and habitation levels.

18,090 ± 650 BM-2307. Montgaudier b13C = -19.6%o Collagen from bone fragments, ref Sample 1, from Magdalenian level above Hearth H27/I17 (Duport, 1983). Coll 1980 to 1981.

11,930 ± 190 BM-2308. Montgaudier b13C = -19.8%o Collagen from bone fragments, ref Sample 3, from Loc 12, Layer 2, Sq E'2. Coll 1983.

14,770 ± 270 BM-2309. Montgaudier b13C = -19.8%o Collagen from bone fragments, ref Sample 4, from Loc 12, Layer 2, Sq G2. Coll 1983.

11,690 ± 170 BM-2310. Montgaudier b13C = - 19.4%o Collagen from bone fragments, ref Sample 5, from Loc 12, Layer 3, Sounding. Coll 1983.

20,870 ± 370 BM-2311. Montgaudier b13C = -18.9%o Collagen from bone fragments, ref Sample 6, from Loc 12, Level 4, Sq F'2. Coll 1983. General Comment (GdeGS): BM-2307 gives date for level overlying Hearth 27/127, which is regarded as belonging to same strat level as hearth in Sq J.20, although this is some meters away and deposit is not continuous. Thus, date relates to BM-1913 (18,050 ± 230 BP; R, 1983, v 25, p 47) and confirms unexpectedly early date for Sq J.20 hearth and suggests that this and layer dated by BM-2207 antedate classic Magdalenian III-IV occupa- 70 J Ambers, R Burleigh, and K Matthews tion at Montgaudier. Late Magdalenian finds for these deposits can be regarded as intrusive. BM-2208 to -2311 come from Loc 12, strat deposit on terrace outside cave entrance, close to sounding from which BM-1913 was obtained. BM-2308, -2309 and -2311 are in strat order. BM-2308 dates upper part of Layer 2, BM-2309, lower part and BM-2311, Layer 4, lowest part of strat. Determinations confirm strat sequence and suggest cave was used for long period. BM-2310 was from separate sounding in same area. BM-2308 and -2310 can be regarded as late Magdalenian dates and BM- 2309 as classic Magdalenian. BM-2311 confirms evidence of BM-1913 that lower part of deposit antedated classic Magdalenian III-IV cultural succes- sion in Charente. Site is in hard water area but pretreatment in acid to extract collagen should have removed any possibility of contamination from dissolved carbonate.

Iraq

2310 ± 80 b'3C BM-2293. Khirbet Khatuniyeh = -23.5%o Charcoal, ref KK Al, from Tr Dl, Level 4, destruction level at Khirbet Khatuniyeh, near Eski Mosul (36° 40' N, 42° 50' E). Coll 1984 and subm by J E Curtis, Dept W Asiatic Antiquities, British Mus. Comment (JEC): it is clear from assoc material that destruction level at site should be dated either to end of Late Assyrian period (612 Bc) or later (550 to 500 Bc). Abu Salabikh series Samples from Early Dynastic tell of Abu Salabikh, Qadisiyyah Gover- norate (35° 15' N, 45° 05' E). Coll 1983 and subm by J N Postgate, Fac Oriental Studies, Univ Cambridge, to supplement dating sequence already established for site (R, 1982, v 24, p 163-164).

3700 ± 60 b'3C BM-2328. Abu Salabikh = -25.4%o Charcoal, ref 4J87:108, from burned floor, Phase, Rm 28 (see Post- gate, 1984, p 98-99 for provenience).

4090 ± 60 b'3C BM-2329. Abu Salabikh = -27.5%o Charcoal, ref 6G75:525, from fire installation in W side of Gr 162, Rm 58 (see Postgate, 1984, p 97 for provenience).

4210 ± 70 BM-2330. Abu Salabikh b13C = - 26.6%0 Charcoal, ref 5I88:11, from remains of roof beam from Rm 192, Sq 5188 (see Postgate,1984, p 101-103 for provenience). General Comment (JNP): results conform with chronol order expected on archaeol grounds, also with earlier series: BM-2328 is from context close to BM-1366 (3869 ± 56 BP; R, 1982, p 163), suggesting that beam from which British Museum Natural Radiocarbon Measurements XIX 71 earlier sample was taken had been in use for some time. Both these contexts should be late Early Dynastic III. BM-2329 should be earlier than BM- 1365A to -1365D (R, 1982, p 163) on strat grounds and is therefore satisfactory (ca transition from ED II to ED III). BM-2330 is from ED II house and cal 2650 to 3030 BC (Klein et al, 1982) is acceptable.

Israel Nahal Hemar series Samples from early Neolithic levels in cave deposit (Bar-Yosef,1985, p 8) at Nahal Hemar, ca 10km W of Sodom, Judean Desert (31 ° 10' N, 35° 10' E). Coll 1983 by D Alon and 0 Bar-Yosef and subm by 0 Bar-Yosef, Inst Archaeol, Hebrew Univ, Jerusalem, to provide date for very rare occurrence of site with numerous organic remains (wooden beads, baskets, linen textiles, human skulls, animal remains) of Neolithic period of Levant (cf Jericho), from which sample of flints recovered was insufficient to establish correlation with Jericho sequence.

8250 ± 70 b13C BM-2298. Nahal Hemar = - 20.9%o Charcoal from hearth in Sq F3, Layer 3A.

9110 ± 300 BM-2299. Nahal Hemar b13C = - 23.O%o Strings from basketry, from Sq H4C, Layer 4.

8690 ± 90 BM-2300. Nahal Hemar b13C = - 20.4%0 Strings from Layers 3 and 4. General Comment (OB-Y & RB): results confirm that site is in same range as PPNB of Jericho (9160 to 8450 BP; Burleigh, 1984). BM-2298 agrees with dates of 8100 ± 100 BP (RT-650, unpub) and 8270 ± 80 BP (Pta-3650, unpub) for charcoal from same context (hearth in Layer 3A); BM-2299 agrees with Pta-3625 (unpub), 8850 ± 90 BP for strings from same context (Layer 4).

Italian prehistory series Samples from SE and central Italy subm by R Whitehouse, Univ Lan- caster, to establish 14C chronology for prehist of area, for which few dates are available. Marcianese series Samples from early to middle Neolithic village site (Geniola, 1980, 1982) at Marcianese, comune Lanciano, Chieti prov, S Abruzzo (42° 12' N, 1° 55' E). Coll 1969 by A Geniola, Inst Civilta Preclassiche, Univ Ban. 72 J Ambers, R Burleigh, and K Matthews 6290 ± 60 b13C BM-2250. Marcianese = -24. 7/oo Charcoal, ref sample 1, from lower levels of main but excavated.

6250 ± 90 BM-2251. Marcianese b13C = -24.1%o Charcoal, sample 2, from upper levels of main hut.

6000 ± 110 BM-2252. Marcianese b13C = -19.9%o Collagen from human and animal bone, ref sample 3. General Comment (RW): samples were assoc with Impressed Ware, including some with elaborate decoration but no painted pottery was present. Thus, date was tentatively postulated to be in early 7th millennium BP. Instead, dates form very coherent group falling late in range of dates for Impressed Ware generally. They nonetheless fit into range of dates available for earliest Neolithic sites in central E Italy (Pi-101, 6578 ± 135 BP, from Villaggio Leopardi, Pi-46, 6247 ± 130 BP, from Grotta dei Piccioni, R, 1961, v 3, p 100; R-634a, 6580 ± 75 BP, from Maddalena, and R-598, 6210 ± 75 BP, -598a, 6140 ± 70 BP, -599a, 6260 ± 85 BP, from Ripabianca, R, 1970, v 12, p 603).

Grotta di Cala Scizzo series Samples from small natural cave (Geniola & Tunzi, 1980) used during late to final Neolithic at Grotta di Cala Scizzo, near Punta della Penna, comune Torre a Mare, Bari prov (41 ° 5' N, 4° 30' E), assoc with cult objects. Coll 1977 by A Geniola.

4880 ± 210 BM-2253. Grotta di Cala Scizzo Est b13C = -19.5%o Collagen from animal bone, ref sample 1, from Level I, assoc with late Serra d'Alto and Diana Wares.

4230 ± 100 BM-2254. Grotta di Cala Scizzo b13C = -18.6%o Collagen from animal bone, ref sample 2, from Level II, assoc with Diana and Bellavista wares.

3190 ± 80 BM-2255. Grotta di Cala Scizzo 513C = -20.6%o Collagen from animal bone, ref sample 3, from Level III, assoc with Diana and Bellavista wares. General Comment (RW): of 3 dates from site, only BM-2253 seems acceptable. For further discussion of implications of results, see comment on dates for Grotta 1 di Cala Colombo, BM-2259, -2260, -2301, -2302, below. British Museum Natural Radiocarbon Measurements XIX 73 Santa Barbara series Samples from artificial rock-cut structure, cut in late Neolithic into side of ditch of earlier Neolithic settlement (Geniola, 1979) at Santa Bar- bara, comune Polignano a Mare, Bari prov, C Apulia (41 ° 0' N, 4° 50' E). Coll 1975 & 1978 by A Geniola.

5800 ± 120 b13C BM-2256. Santa Barbara = -18.1b o Collagen from animal bone, ref sample 1, from Level I.

5620 ± 130 BM-2257. Santa Barbara b13C = - 20.9/00 Collagen from animal bone, ref sample 2, from Level II.

5720 ± 120 b13C BM-2258. Santa Barbara = -20.8 oo Collagen from animal bone, ref sample 3, from Level III. General Comment (RW): site is clearly of cult significance and contains deer skulls and number of highly decorated painted vessels of Serra d'Alto Ware, which should, on typologic grounds be assigned to early phase of ware, characterized by cups with bold painted designs on belly (White- house, 1969, p 290-292). Dates form tight cluster which seems early for traditional placing of Serra d'Alto Ware in Neolithic sequence, but is in agreement with earlier suggestion (Whitehouse, 1969, p 290-292; 1978, p 81) that first phase of Serra d'Alto Ware was contemporary with Trichrome Wares. Other dates available for Serra d'Alto Ware are R-284, 5555 ± 75 BP, from Grotta della Madonna (R,1967, v 9, p 355) and Pi-49, 4770 ± 110 BP, from Grotta dei Piccioni (R,1961, v 3, p 100). R-284 has been regarded as anomalous but in view of new dates it now seems acceptable and that Serra d'Alto painted ware tradition began early in 6th millennium BP. For further discussion of dating of Serra d'Alto Ware, see comment on dates for Grotta 1 di Cala Colombo, BM-2259, -2260, -2301, -2302, below.

Grotta 1 di Cala Colombo Samples from small natural cave used for burials and cult purposes in late to final Neolithic in Grotta 1 di Cala Colombo (Geniola, 1976; De Lucia et al, 1977) comune Torre a Mare, Bari prov, C Puglia (41 ° 5' N, 4° 30' E). Coll 1973 by A Geniola.

4070 ± 60 13C BM-2259. Grotta 1 di Cala Colombo b = - 19.7 oo Collagen from animal bone, ref sample 2, from Level I, assoc with 1st use of cave, and with pottery of late Serra d'Alto type and Diana Ware.

4870 ± 90 b13C BM-2260. Grotta 1 di Cala Colombo = -19.6%0 Collagen from animal bone, ref sample 3, from Levels II to IV, period when cave not in use but assoc with some material from Level 1. 74 J Ambers, R Burleigh, and K Matthews 1180 ± 50 5130 BM-2301. Grotta 1 di Cala Colombo = -18.8%o Collagen from animal bone, ref sample 4, from Levels V to VII, assoc with fire in cave and re-use for collective burial of ca 15 individuals with mainly Bellavista type pottery together with some Diana and late Serra d'Alto wares.

4810 ± 180 '3C= BM-2302. Grotta 1 di Cala Colombo -19.7%o Collagen from animal bone, ref sample 5, from Level VII, assoc with collective burial, Bellavista, Diana and Serra d'Alto pottery. General Comment (RW): BM-2301 is clearly anomalous and BM-2259 seems too young. Remaining 2 results, 4870 ± 90 BP (BM-2260) for late Serra d'Alto and Diana Ware and 4810 ± 180 BP (BM-2302) for Bellavista Ware are consistent with other dates for these pottery types. Other dates for Diana Ware come from Grotta della Madonna (R-283, 5110 ± 70 BP; R, 1967, v 9, p 355), Lipari acropolis (R-180, 5000 ± 200 BP; R, 1969, v 11, p 488) and Contrada Diana (R-182, 4885 ± 55 BP; R, 1969, v 11, p 489). Ear- lier dates for Cala Colombo and Cala Scizzo (cf BM-2253, above) are almost identical to date from Contrada Diana. Later date for Bellavista Ware is in accordance with strat and with pottery typology. Most interesting feature of results is indication of long duration of Serra d'Alto Ware, which seems to have lasted from early 6th to early 5th millennium BP, confirming previously doubted evidence from Grotta dei Piccioni (cf comment on dates for Santa Barbara, BM-2256 to -2258, above).

Portugal Segovia series Samples from Iron Age hillfort (Judice Gamito, 1979, 1981, 1982) at Segovia, near Elvas, Alentejo (40° 0' N, 7° 0' W). Coll 1972 and 1982 and subm by Teresa Judice Gamito, Inst Archaeol, Univ Lisbon. 2280 ± 45 b13C BM-2159. Segovia = -21.4 oo Collagen from bone from Layer 5, Sq Al (corresponding to Layers 6-7 of Sq B), close to interface of N defensive wall of main settlement.

2410 ± 50 b'3C BM-2160. Segovia = -19.O%o Collagen from bone from Layer 8, Sq Al, close to interface of N defensive wall of main settlement.

2140 ± 130 BM-2287. Segovia Y3C _ -23. 7%o Charcoal from Area B-111 b, Layer 6, Spit 1 underneath large bldg, assoc with pottery. British Museum Natural Radiocarbon Measurements XIX 75

1220 ± 110 BM-2288. Segovia b13C = -19.9%o Collagen from bone from Sq Al, Layer 4, Spit 1.

890 ± 60 BM-2289. Segovia b13C = -20.5%o Collagen from bone from Sq Al, Layer 5-6, close to N wall. General Comment (TJG): BM-2159 = cal 540-180 BC (Klein et al; R, 1982, v 24, p 103-150) which agrees with expected date of 600-400 BC. BM- 2160 = cal 700-395 BC which overlaps with expected late Bronze Age/early Iron Age date, 750-700 BC by analogy with Medillin, site close to Segovia in Extremadura, (Almagro Gorbea,1977), and with similar layers at E1 Caram- bolo, Cerro Macareno and Colina de los Quemados in Andalucia (Pellicer Catalan, 1976-78). BM-2287 = cal 530 BC-AD 205 and was assoc with pottery of ca 600-500 BC. Date is acceptable but large error term due to small size of sample limits usefulness. BM-2288 came from Iberian horizon with expected date between 200-400 BC. Cal AD 605-1010 does not agree with expectations but is acceptable for known later use of site and probably reflects later intrusion into Iron Age layers. BM-2289, which was expected to give early Iberian date, probably reflects same re-use of site. Spain Ferrandell-Oleza series Samples from Beaker settlement site of Ferrandell-Oleza Old Settle- ment, Valldemosa, Mallorca, Baleares (39° 40' N, 2° 30' E). Coil 1984 and subm by W H Waldren, Donald Baden-Powell Quaternary Research Cen- tre, Pitt Rivers Mus, Univ Oxford, and Dir, Deya Archaeol Mus and Research Center, Deya de Mallorca.

2140 ± 80 BM-2297. Ferrande1101eza b13C = -24.2%0 Charcoal from Sec EXW, Quad Q15-16 alongside Water Channel, from level of early Bell Beaker pottery.

3210 ± 80 BM-2312. FerrandellOleza b13C = -19.2/oo Collagen from fragmentary animal bone, most unid., but including small cow and goat, from beneath Water Channel. General Comment (RB): dates are later than expected (>4000 BP); charcoal probably misassoc.

2330 ± 80 b13C BM-2337. Rio Tinto = - 24.3%o Charcoal, ref CL 79 Layer 15, Sample 39, from early mining levels in area T1, at base of Corto Largo sec, Rio Tinto (37° 40' N, 6° 30' W), assoc with evidence of metal extraction. Coil 1979 and subm by B Rothenburg, Inst Archaeo-Metallurg Studies, Univ London. Comment by P T Craddock. 76 J Ambers, R Burleigh, and K Matthews Comment (PTC); sample is from base of massive continuous layers of slag in central area of ancient smelting, thus dating commencement of large-scale operations at Rio Tinto that were to last for next 6 centuries or so.

REFERENCES Almagro Gorbea, M, 1977, Appendix on CM date, in El Bronce Final y el Periodo Orientali- zante en Extremadura: Madrid, Ministeris Culture. Ambers, J and Burleigh, R, in prep, Radiocarbon dating of the Magdalenian habitation refuse, in Sieveking, G de G, Les Eyzies mon, in prep. Bar-Yosef, 0, 1985, A cave in the desert: Nahal Hemar, 9,000-year-old-finds: Jerusalem, Israel Mus. Bradley, R, 1975, Maumbury Rings, Dorchester: the excavations of 1908-13: Archaeologia, v 105, p 1-97. Bradley, R, Cleal, J, Gardiner, J, Green, M and Bowden, M, 1984, The Neolithic sequence in Cranborne Chase, in Bradley, R and Gardiner, J, eds, Neolithic studies: Oxford, Br Archaeol Repts 133, p 87-105. Briggs, C S, 1983, Copper mining at Mount Gabriel, Co Cork: Bronze age bonanza or post- famine fiasco?: Prehist Soc Proc, v 49, p 317-334. Burleigh, R, 1984, Additional radiocarbon dates for Jericho (with an assessment of all the dates obtained), in Kenyon, K M and Holland T A, eds, Excavations at Jericho, v 5: Lon- don, Br School Archaeol Jerusalem, p 760-765. Craddock, P T, 1986, Bronze age metallurgy in Britain: Current Archaeol, v 99, p 106-109. De Lucia, A, Fern, D, Geniola, A, Giove, C, Maggiore, M, Melone, N, Pesce Delfino, V, Pieri, P and Scattarella, V, 1977, La comunita neolitica di cala Colombo presso Torre a Mare (Ban), in Documenti e Monografie 42: Soc Storia Patria Puglia, Ban. Duport, L, 1976, La grotta de Montgaudier in Union Internatl Sci Prehist et Protohist (UISPP) tong, 9th Proc: Livret guide de l'excursion A4 Sud-Ouest (Aquitaine et Char- ente), p 151-158. ----- 1983, Gravures Magdaleniennes de Montgaudier Commune de Montbron (Charente): Poitiers. Geniola, A, 1976, La comunita neolitica di Cala Colombo presso Tome a Mare (Ban): Riv Antropol, v 59, p 189-275. -----1979, Il neolitico delta Puglia settentnionale e centrale, in La Publia dal Paleoli- tico al Tardoromano: Electa, Milano, p 52-93. ----- 1980, Considerazioni conclusive sullo scavo archeologico dell' insediamento neolitici di Marcianese, in Convegno sulla Preistoria-Protostoria della Daunia, 2nd, Proc: San Severo, Civica Amministrazione, p 59-67. -----1982, Marcianese: Itinerari, Lanciano. Geniola, A and Tunzi, A M, 1980, Espressioni cultuali e d'arte nella Grotta di cala Scizzo presso Tome a Mare (Ban): Riv Sci Preistoriche, v 35, pt 1-2, p 125-146. Inman, R, Brown, D R, Goddard, R E and Spratt, D A, 1985, Roxby Iron age settlement and the Iron age in north-east Yorkshire: Prehist Soc Proc, v 51, p 181-214. Jackson, J S, 1968, Bronze age copper mines on Mount Gabriel, W County Cork, Ireland: Archaeol Austriaca, v 43, p 92-103. -----1980, Bronze age copper mining in Counties Cork and Kerry, Ireland, in Crad- dock, P T, ed, Scientific studies in early mining and extractive metallurgy: British Mus Occ paper 20, London, p 9-30. ----- 1984a, Copper mining at Mount Gabriel, Co Cork: Bronze age bonanza or post- famine fiasco? a reply: Prehist Soc Proc, v 50, p 375-377. ----- 1984b, The age of primitive copper mines on Mt Gabriel, West County Cork: Irish Jour Archaeol, v 2, p 41-50. Jones, M V and Bond, D, 1980, Later Bronze age settlement at Mucking, Essex, in Barrett, J and Bradley, R, eds, Settlement and society in the British Later Bronze age: Oxford, Br Archaeol Repts 83. Judice Gamito, T, (ms) 1979, Aspects of settlement, economy and society in southern Portugal from 600 BC till the Roman conquest: M Phil dissent, Umv Cambridge. -----1981, Resistencia a Roma no Sudoeste Peninsular: Lisbon, Historia. -----1982, A Idade do Ferro no sul de Portugal-problemas e prespectivas: Arqueo- logia, v 6, p 3-16. Klein, J, Lerman, J C, Damon, P E and Ralph, E K, 1982, Calibration of radiocarbon dates: Tables based on the consensus data of the Workshop on Calibrating the Radiocarbon Time Scale: Radiocarbon, v 24, no. 2, p 103-150. Lartet, E and Christy, H,1875, Reliquiae Aquitanicae, being contributions to the archaeology British Museum Natural Radiocarbon Measurements XIX " 77 and palaeontology of Perigord and the adjoining provinces of southern France (1865- 1875): London, Williams & Norgate. Pellicer Catalan, M, 1976-78, Problematica general de los inicios de la iberizacion en Andalu- cia Occidental: Barcelona, Inst Prehist Arqueol, Mono XLIX. Peltenburg, E J, 1985, Lemba archaeological project Cyprus 1983: Preliminary report: Levant, v XVII, p 53-64. Pitt Rivers, A, 1898, Excavations in Cranborne Chase, v 4: London, privately printed. Postgate, J N, 1984, Excavations at Abu Salabikh,1983: Iraq, v 46, 95-113. Stuiver, M and Polach, H A, 1977, Discussion: Reporting of 14C data: Radiocarbon, v 19, no. 3, p 355-363. Wainwright, C, 1979, Mount Pleasant, Dorset: Excavations 1970-71: London, Soc Anti- quaries. Ward Perkins, J B, 1944, Excavations on the Iron age hillfort of Oldbury, near Ightham, Kent: Archaeologia, v 49, p 127-176. Whitehouse, R, 1969, The Neolithic pottery sequence in southern Italy: Prehist Soc Proc, v 35, p 267-310. - 1978, Italian prehistory, carbon 14 and the tree-ring calibration, in Blake, H McK, Potter, T W and Whitehouse, I) B, eds, Papers in Italian archaeology I: Oxford, Br Archaeol Repts, S41, p 71-91. [RADIOCARBON, VOL 29, No. 1, 1987, P 78-99] HARWELL RADIOCARBON MEASUREMENTS V A J WALKER, R S KEYZOR*, and R L OTLET Isotope Measurements Laboratory, Nuclear Applications Centre, Atomic Energy Research Establishment, Harwell, Oxfordshire, OXl l ORA, UK

INTRODUCTION The results presented in this list include some recently measured samples (1984) but mostly ones from our earlier years of operation which had not been previously published in RADIOCARBON. It is the first of a number of special lists prepared over the last year so that the backlog of unpublished dates of this laboratory will be cleared. The samples are all archaeologic from the United Kingdom most of which have originated from "rescue"-type excavations. As in previous lists, all samples were measured by liquid scintillation counting (Otlet & Warchal,1978), and the error term quoted is the 10r standard deviation estimate of the full replicate sample reproducibility (Otlet, 1979). The list was produced semi-automatically from the Harwell data base stored on the main frame computer using the procedures described in Otlet and Walker (1983). Calculations are based on the Libby half-life of 5568 years, using NBS oxalic acid standard (x0.95) as "modern," both values treated as constants, with AD 1950 as the reference year. All results are corrected for fractionation according to the quoted b13C (wrt PDB) values measured in this laboratory.

ACKNOWLEDGMENTS We wish to acknowledge the work of our colleagues, G A Bradburn and D G Humphreys, with the laboratory measurements and of E F Westall, S E Hasler, and M Gibson with the preparation of the data in computer readable form. The financial support and cooperation of the staff of the Historic Buildings and Monuments Commission for England (Ancient Monuments Laboratory), through which most of the samples (those additionally referenced with AML numbers) were submitted, is also gratefully acknowledged.

ARCHAEOLOGIC SAMPLES

British Isles England Stonehenge Car Park series Charcoal samples, presumably decayed bark, from Stonehenge Car Park, Stonehenge, Salisbury, Wiltshire (51 ° 10' 49" N, 1° 49' 37" W, Natl Grid Ref SU 121424). Coll Feb 1966 by H L and F J de M Vatcher and subm by H Keeley, Ancient Monuments Lab. For description of excavation see Anonymous (1973).

* also School of Archaeologic Sciences, University of Bradford, UK

78 Harwell Radiocarbon Measurements V 79

9130 ± 180 '3C HAR-455. HK1 b = - 24.2%0 From Hole A, depth 0.76m, halfway between top (natural chalk) and base, on post circumference.

8090 ± 140 HAR-456. HK4 b13C = -25.4%0 AML 733250, from Hole B, depth 0.91m from surface of natural chalk. General Comment (FJdeMV): dates are unexpectedly early; we do not know why this post hole is ca 1000 yr younger than the other which should have been contemporaneous. Excavator thinks that samples, although pertain- ing to original posts, were of poor quality, being fine and mixed with other material.

Marc 3 series Samples from two sites described in Fasham (1980), R6, Burntwood Farm and R5, Bridget's Farm, both in Itchen Valley Parish, Hampshire along planned route of M 3 Motorway. Previous dates from Rte M 3 are reported in Walker and Otlet (1985, p 75-76). All samples coil and subm by P J Fasham, Archaeol Rescue Comm, Winchester.

1980 ± 110 HAR-1021. R6-801 b13C = -19.6%0 Human leg bones, AML 749329, from grave 55 at Burntwood Farm (51 ° 6' 13" N, 1° 16' 11" W, Natl Grid Ref SU 51133411). Coil Oct 1974 and subm Nov 1974. Comment, uncertainties quoted may underestimate actual measurement errors due to low laboratory yields.

3790 ± 70 HAR-1695. R5-816 b13C Charcoal, nut shells, AML 750725, from Bridget's Farm (51 ° 6' 26" N, 1° 15' 58" W, Nail Grid Ref SU 51373453). Coil Dec 1974 and subm May 1976.

2800 ± 80 b13C HAR-2745. R5-817 = -24.2 o Bone, AML 780963, from Bridget's farm (51 ° 6' 26" N, 1° 15' 58" W, Nati Grid Ref SU 51373453). Coil Dec 1974 and subm April 1978. General Comment (PJF): HAR-1021 formed part of study of close dating system for Iron Age and Roman pottery in area; HAR-1695 and -2745 were intended to date series of otherwise undated pits at Bridget's Farm.

Burton Fleming series Human bone, from Iron Age cemetery at Burton Fleming, Yorkshire (54° 6' 41" N, 0° 19' 24" W, Nail Grid Ref TA 096697). Coil July 1974 and 80 AJ Walker, RS Keyzor & RL Otlet subm Jan 1975 by I M Stead, Brit Mus. For description of site see Stead (1977).

2600 ± 70 13C HAR-105 7. FAAM4 = -21.0%0 AML 748109,1 femur/1 fibula.

2520 ± 70 b13C HAR-1058. FAAQ 10 = - 21.1 %o AML 748110, 1 femur/1 fibula.

2050 ± 80 HAR-1129. FACM34 b13C = -20.7%o AML 748112.

2150 ± 150 HAR-1130. FABG7 3130= -21.1%o AML 748111, 1 femur/1 humerus. Comment: small sample size accounts for large error.

Bishophill series Samples from Bishophill I, 58 and 59 Skeldergate, York, North York- shire (53° 57' 20" N, 104k 56" W, Natl Grid Ref SE 60215145). Coll by S Donaghey and subm by H K Kenward, Environmental Archaeol Unit, Univ York. Comments supplied by P V Addyman and R A Hall. Site was excavated by York Archaeol Trust 1973-1975 and yielded evidence of virtually uninterrupted occupation sequence from Roman period to present day. For details of Anglo-Scandinavian phase see Donaghey (1986); for Roman structures dated here see Carver, Donaghey and Sumpter (1978); for biological remains assoc with Roman structures see Hall, Kenward and Wil- liams (1980).

1230 ± 80 HAR-1412. BH155 5130 = -26.00/o0 Charcoal, AML 757497, id by C A Keepax as oak (Quercus sp) and willow (Salix sp) of fairly large timbers from occupation deposit, context 2131. Subm Dec 1975. Comment (RAH): date is compatible with mid-8th century coin from deposit, but pottery in assoc with structure D was 10th century as was beam in slot marking one wall of structure, HAR-1728 (Walker & Otlet, 1985, p 83). Radiocarbon date does not necessarily contradict the later date as it reflects age of timber.

1840 ± 70 13C HAR-1416. BH198 = -27.2 o Wood, AML 757498, id by C A Keepax as oak (Quercus sp), from top part of buried soil, context 2356. Subm Dec 1975. Harwell Radiocarbon Measurements V 81

1820 ± 70 HAR-1417. BH2O1 b13C = -26.7%s Charcoal, AML 757499, id by C A Keepax as oak (Quercus sp) from fragments of fairly large timbers, from rich layer overlying buried soil, context 2360. Subm Dec 1975. General Comment (PVA): HAR-1416 and -1417 confirm clearance of area in early Roman period preceding primary occupation and now allow reconstruction of local ecological conditions from assoc biological remains and soil evidence.

1930 ± 70 b13C HAR-1418. BH202 = -27.9%0 Charcoal, AML 757500, id by C A Keepax as possibly hazel, from fairly large timbers, from layer assoc with Roman destruction levels, context 2361. Subm Dec 1975. Comment (PVA): early date is surprising as deposit overlay 3rd century Roman levels. Charcoal was presumably derived from trees of considerable age.

1750 ± 80 b13C HAR-1729. BH8068 = - 26.8%0 Wood, AML 760237, from timber post in construction pit of Roman well. Subm Feb 1976. Comment (PVA): result accords with estimated date of construction for the Roman well in 2nd to 3rd century AD.

1840 ± 60 '3C HAR-1866. BH220 b = - 2 7.000 Charcoal, AML 766965, from lowest fills of Roman well. Subm Nov 1976.

1830 ± 70 HAR-1927. BH300 b13C = -26.4%0 Oak, AML 766963, from timber lining of Roman well. Subm Nov 1976. Comment (HKK): ca 50 rings in sample.

2150 ± 70 HAR-1928. BH301 b13C = -27.7%0 Oak, AML 766964, from timber lining of Roman well. Subm Nov 1976. Comments (HKK): ca 50 widely-spaced rings in sample; (PVA): early date suggests re-use of old timbers. Winklebury series Continuation of series first reported in Walker and Otlet (1985, p 83).

1930 ± 70 HAR-1765. 13646 b13C = -25.0%10 Charcoal, AML 766704, from next to bottom layer in pit at Winkle- ° bury Camp, Basingstoke, Hants (51 16' 17" N, 1° 7' 14" W, Natl Grid Ref 82 Af Walker, RS Keyzor & RL Otlet SU 61355290). Coil by D Bartlett and subm Sept 1976 by G Wainwright. Comment (GW): date does not agree (Smith, 1979) with implication of pot of 6th to 5th centuries BC found in same pit. Poundbury series Charcoal from post-Roman settlement at Poundbury, Dorchester, Dorset (50° 41' S" N, 2° 26' 47" W, Natl Grid Ref SY 685911). All samples coil and subm by CJ Sparey Green, Dorchester Excavation Comm, Dorset. Continuation of series first reported (Otlet & Walker, 1979, p 365); for discussion see Green (1971, 1974, 1976, 1982). 1590 ± 80 HAR-2281. PC76E276 b13C = -27.5%o AML 767534, from deposit of dark brown soil and chalk flecks in base of pit, E276. Coil Aug 1976 and subm Sept 1976. Comment (CJSG): date is slightly earlier than expected since sample derived from settlement overlying Roman cemetery still in use in late 4th century AD. Archaeologically settlement might be 5th to 8th century AD but close relationship of structures to cemetery could support original occupation soon after cemetery went out of use. Wood from which this charcoal was derived could have been of some age at time of burial. 1490 ± 80 HAR-3080. 0B12 b13C = -24.1%o Additionally referenced as PC71 B, J9, 18, 115, AML 790638, from burned layer in base of corn drying oven B 115. 1530 ± 60 HAR-3081. 0B21 813C = -24.3%o Additionally referenced PC72E, 53, AML 790640, charred grain, from burned layer in base of corn drying oven E 53. General Comment (CJSG): these 3 samples provide date for early post- Roman settlement on site of Late Roman cemetery which, from coin evidence, continued in use to at least AD 380. They imply rapid change in land use at end of Roman period and date for occupation well before arrival of Saxons (ca AD 650) in this area. Saxon settlements have been identified on several sites in S Britain over last decade, but identification of specifically sub-Roman settlement fills major gap in settlement history of S Britain. Aldwark series Wood, from brushwood at base of fortress rampart at site adjacent to 1 to 5 Aldwark, York, North Yorkshire (53° 57' 43" N, 1° 4' 36" W, Natl Grid Ref SE 60585219). Coil by H MacGregor and subm June 1977 by H K Ken- ward, Environmental Archaeol Unit, Univ York. Comments by R A Hall. 940 ± 70 HAR-2300. ALDW/52 313C = -27.8%o AML 776296. Harwell Radiocarbon Measurements V 83

950 ± 80 HAR-2301. ALDW49A b13C = -27.3%0 AML 770035.

910 ± 70 HAR-2302. ALDW49 b13C= -27.3%0 AML 770034. General Comment (RAH): on artifactual evidence rampart is probably 11 th 14C century AD which agrees well with dates. Rampart is ca 1Om SW of Roman NE fortress wall and has either cut away Roman deposits or has been added to Roman rampart as part of new phase of defense work. Bank itself has been cut away by large feature of unknown function and date which was backfilled in 13th century (MacGregor, in press).

Rivenhall series Samples from churchyard and church at Rivenhall, Essex (51 ° 49 42" N, 0039 11" E, Natl Grid Ref TL 828178). Coll Aug 1972 (HAR-2427) and Aug 1973 and subm Aug 1977 by W J Rodwell, Old Vicarage, Chil- compton, Somerset unless otherwise stated. For description of site see Rodwell and Rodwell (1985).

820 ± 60 HAR-2326. GR298 b13C = -20.1%o Bone, AML 766978, from cemetery area 1 in churchyard. Subm Nov 1976 by P J Drury. Comment (WJR): one of series of graves forming part of regularly laid out Saxon cemetery. Stratigraphically earlier than Saxon or Medieval cemetery bldg of unknown function, but later than Grave 326 (HAR-2404) below.

1140 ± 70 HAR-2404. GR326 b13C = -19.6%0 Bone, AML 766980, from cemetery area 1 in churchyard. Subm Nov 1976 by P J Drury. Comment (WJR): one of deep and early, possibly Middle Saxon, graves in area. Part of regularly planned cemetery.

950 ± 60 HAR-2427. R9 b13C = -26.2%0 Wood, AML 776990, id by C A Keepax as oak, from Medieval church. Comment (WJR): sample was sealed in 14th century blocking of late Saxon window and is thought to be part of original window sill.

Hambledon Hill series Charcoal from 1976 excavation season at Neolithic causewayed enclosure, Hambledon Hill, Dorset (50° 51' 34" N, 2° 12' 38" W, Natl Grid Ref ST 852123). All coll Sept/Oct 1976 and Subm Sept 1977 by R Mercer, Dept Archaeol, Univ Edinburgh. For other samples in series see Walker 84 AJ Walker, RS Keyzor & RL Otlet and Otlet (1985, p 85-86). For description of site see Mercer (1980, 1985).

4680 ± 110 b13C HAR-2371. HH7647 = -27.5%o AML 777006, Site L, Outwork Ditch 1, Layer 7. Comments: small sample size accounts for large error. (RM): burned area seemed to have definite N edge abutting Layer 9. Deposit contained bone, flint, and human skeletal material; charcoal fragments were contained within other burned matter.

4610 ± 90 13C HAR-2377. HH7679 = -25.4%o AML 777011, 50% of sample id by C A Keepax as oak from large timbers, from Site G, ditch segment 3, Layer 9A.

4820 ± 120 HAR-2378. HH7662 5'3C = -26.4%o AML 777010, 50% of sample id by C A Keepax as oak and hazel/alder, Site K, outwork ditch 1, Layer 7. Comments: small sample size accounts for large error; (RM): sample came from very ashy area containing charcoal lumps. General Comment (RM): samples represent 2 layers, 7 (HAR-2371, HAR- 2378) and 9 (HAR-2377). Layer 7 overlay primary silt layers 8 and 9 and was composed of vacuous chalk lumps and flint nodules. Layer 9 was characterized by gray charcoal silt in deep bowl-shaped pit.

Billingborough series Charcoal, from Bronze Age settlement, Billingborough, Lincolnshire (52° 53' 9" N, 0° 19' 35" W, Natl Grid Ref TF 126334). Coll Oct 1977 and subm Dec 1977 by P Chowne, South Lincolnshire Archaeol Unit. For description of site see Chowne (1980).

2390 ± 70 5130 HAR-2483. BFE77F98 = -24.8%o AML 777901, id by C A Keepax as oak (large timber), from charred post.

2410 ± 80 HAR-2523. BFE77F43 o13C = - 26.1 /oo AML 78009, id by C A Keepax as mainly hazel/alder with some hawthorn (large timbers), from upper levels of silted-up enclosure ditch.

Custom House series Waterlogged wood samples, id as oak, part of 315-yr tree-ring sequence, from beams forming timber quay of Roman waterfront at Cus- ° tom House, London (51 30' 26" N, 0° 4' 45" W, Natl Grid Ref TQ 33298050). Coll 1973 by Dept Urban Archaeol, London and subm Feb Harwell Radiocarbon Measurements V 85

1978 by j Hillam, Dept Prehist & Archaeol, Univ Sheffield. For description of site see Tatton-Brown (1974).

1970 ± 70 b13C HAR-2532. CUS 1 = -26.6%o AML 780045. Comment (JH): years 91-110, growth allowance 220 years.

1820 ± 70 HAR-2530. CUS 2 b13C = -26.5%o AML 780046. Comment (JH): years 124-133, growth allowance 190 years.

1900 ± 70 HAR-2534. CUS 3 b13C = -25.6%o AML 780047. Comment (JH): years 167-186, growth allowance 145 years. General Comment (JH): with growth allowances shown samples can be correlated with further Roman waterfront structures at New Fresh Wharf/Seal House (assoc dates HAR-1864, -1865, -1867, -1868; Walker & Otlet,1985, p 84). From tree rings Custom House structure appears a few decades earlier than this; for full dendrochronology see Morgan and Shhofield (1978).

Hadstock series Charcoal and bone from church, Hadstock, Essex (52° 4' 43" N, 0° 16' 30" E, Natl Grid Ref TL 55884474). All coil and subm by W J Rod- well, Old Vicarage, Chilcompton, Somerset. See Rodwell (1976).

880 ± 70 HAR-2559. HAD2 b13C = _26.1%o Charcoal, AML 780208, id by C A Keepax as all oak from large timbers, from filling of bell-casting pit inside church. Coll Aug 1974 and subm Feb 1978.

1080 ± 80 b13C HAR-2594. HAD 1 = - 22.0%0 Bone, AML 780207, from colln of various animal bone fragments in Saxon domestic horizon in buried soil below church. Coll Sept 1974 and subm May 1978.

860 ± 70 HAR-2595. HAD4 b13C = -17.7%0 Human bone, AML 780210, from colln in reliquary in floor of S porticus of church. Comments: date checked by replicate measurement, HAR- 2697 below. (WJR): stratigraphically earlier than HAD3, below. Burial 86 AJ Walker, RS Keyzor & RL Otlet would seem to be Norman rather than Saxon deposit contemporary with Saxon church.

600 ± 80 HAR-2606. HAD3 3130 = -19.7%0 Bone, AML 780209, from burial of child in S porticus of church. Coil Aug 1974 and subm May 1978. Comment (WJR): burial sealed filling of reliquary which contains HAD4 (HAR-2697).

780 ± 70 b13C HAR-2697. HAD4 = -19.2 oo Human bone, AML 780210. Comment: replicate check measurement on HAR-2595, above.

Stafford King's Pool series Peat from the King's Pool, Stafford, Staffordshire (52° 48' 28" N, 2° 6' 41" W, Natl Grid Ref SJ 925234). Coil May 1977 by J R A Greig and S M Colledge and subm Feb 1978 by S M Colledge, Dept Plant Biol, Univ Birmingham.

960 ± 80 HAR-2577. KP100 5'3C = -29.6%0 AML 7716223, from 1 m depth within deposit. Subm Oct 1977.

920 ± 60 HAR-2578. KP140 313C = -29.00/o0 AML 7716224, from 1.4m depth within deposit. Subm Oct 1977.

1620 ± 60 HAR-2582. KP230 513C = - 28.8%0 AML 7716225, from 2.3m within deposit. General Comment (SMC): results required to date different levels of pollen diagram. Peat is most recent deposit in deep depression formed by glacial action (Colledge, 1978).

North Ferriby series Wood, from sites assoc with Bronze Age boat finds at North Ferriby, Humberside (53° 42' 50" N, 0° 29' 52" W, Nat! Grid Ref SE 99132524). All samples subm by Nat! Maritime Mus, Greenwich, London.

3150 ± 80 HAR-2759. ARC3083 313C = - 29.2%0 Oak (Quercus sp) from intertidal zone of beach. Coil May 1978 and subm July 1978 by S McGrail. Comment (SMcG): excavation will continue to investigate relationship between possible linear wooden features and site of boat finds. Sample came from same layer as Bronze Age boat dated to ca 1500 BC (Wright,1976). Harwell Radiocarbon Measurements V 87

3500 ± 70 HAR-3682. ASA3215 b13C = -27.4%o Vertical stake, oak (Quercus sp), one of pair located either side of hori- zontal timber of possible trackway, hardstanding or fish weir, see McGrail (1983). Coll Aug 1979 by G T Denford and subm 1980 by G T Denford. Comment (GTD): sample from same layer as three Bronze Age boats dated to mid-2nd millennium BC, also HAR-4204, below.

3420 ± 90 HAR-4204. ARCS3086 b13C = - 28.0/00 Vertical stake, hazel, from possible trackway on foreshore, (site co- ordinates 3355/2199). Coll and subm Dec 1980 by V E Heal. Comment (SMcG): sample 35mm in diam, some bark present. Top of timber found 0.23m and bottom 0.77m below ground surface, penetrates 0.30m into context 8, upper 0.24m in context 5 of site.

960 ± 70 HAR-2835. ARCS2213 b13C = -26.1%o From remains of log boat (dugout canoe with fitted ribs) in Yorkshire Mus, York, found 1838 at Stanley Ferry, Yorkshire (53° 42' 9" N, 1° 27' 38" W, Natl Grid Ref SE 35612305). Coll Aug 1978 and subm Sept 1978 by S McGrail. Comment (SMcG): date substantiates other evidence that some log boat finds are Medieval rather than prehistoric (McGrai1,1978). Somerset Levels series Following 33 dates are for samples coll during excavations 1975-1984 at Somerset Levels site, Somerset. Following general dates, main series is divided into sub-series (Tinney's, Baker Platform, Eclipse site, Meare Vil- lage East and Honeygore). For introduction to project and other results, see Harwell II (Otlet, 1977, p 415-416), Harwell III (Otlet & Walker, 1979, p 360-364) and Harwell IV (Walker & Otlet,1985, p 77-80). Except where noted all samples coil and subm by J M Coles, Dept Archaeol, Cam- bridge.

4590 ± 70 HAR-3078. SLP7901 b13C = - 28.0%0 Wood, from Jones' Track, Ashcott (51 ° 8' 37" N, 2° 46' 45" W, Nati Grid Ref ST 455386). Coil Feb 1979 by C R Sturdy and subm Feb 1979. Comment (JMC): one of several Neolithic structures discovered on Walton Heath (Orme, Caseldine & Morgan, 1982), see HAR-3386, below.

5180 ± 70 HAR-3195. 5LP7902 b13C = -26.8%o Wood, bog oak, from Site 4, Meare Heath Field (51 ° 9' 54" N, 2° 47' 33" W, Nati Grid Ref ST 446410). Coil by R Morgan and subm March 1979. Comment (JMC): date was completely unknown but is very satisfactory as it is roughly contemporary with floating Neolithic tree-ring chronology. For description of site see Coles and Orme (1978). 88 AJ Walker, RS Keyzor & RL Otlet 4420 ± 80 HAR-3386. SLP7904 li13C = -30.9%0 Wood, assoc with HAR-3078, above, at JN79.1, Jones' Track (51 ° 8' 40" N, 2° 46' 45" W, Nat! Grid Ref ST 455387). Col! and subm June 1979. Comment (JMC): same structure as HAR-3078. For description of site see Orme, Caseldine and Morgan (1982).

4340 ± 80 HAR-3387. SLP7905 b13C = - 28.8%0 Wood from near Garvin's Track, Ashcott (51 ° 8' 37" N, 2° 46' 50" W, Nat! Grid Ref ST 454386). Col! June 1979 by B J Orme and subm June 1979. Comment (JMC): closely related to Garvin's Track (HAR-1222, -1219; see Otlet,1979, p 361-362; Coles & Orme,1977; Orme, Caseldine & Mor- gan, 1982).

2810 ± 90 HAR-4477. SLP8103 513C = - 25.0%0 Wood from structure at complex Bronze Age site in raised bog peat at Stileway, Meare (51 ° 9' 52" N, 2° 45' 55" W, Nail Grid Ref ST 465409). Coll June 1981 by KR Campbell and subm June 1981. See Orme et al (1985a) for description of site.

4580 ± 70 b13C HAR-4739. SLP8108 = - 29.1 %o Wood, from structure in peat at Site 81.42, Signal Pole Ground, Ash- cott Heath (51° 9' 53" N, 2° 49' 21" W, Nat! Grid Ref ST 425410). Col! Nov 1981 and subm Dec 1981. See Orme et al (1985b).

3020 ± 60 b13C HAR-4998. SLP8204 = - 29.8 oo Wood, from structure, found in bank of R Gary at Henley Bridge, Sedgemoor (51 ° 5' 19" N, 2° 48' 9" W, Nat! Grid Ref ST 438325). Coll and subm April 1982. Comment (,JMC): date appears to confirm presumed prehistoric date (Coles & Orme, 1983).

3810 ± 70 HAR-5054. SLP8210 b13C = - 28.6%0 Peat from around flint core near line of 2nd millennium BC tracks at Ten Acres, Meare Heath (51 ° 9' 50" N, 2° 47' 44" W, Nat! Grid Ref ST 44374088). Coll and subm June 1982.

3060 ± 70 HAR-5086. SLP8206 b13C = -29.7%o Wood, from newly discovered trackway at Godwin's Track, Sharpham (51 ° 8' 31" N, 2° 45' 48" W, Nail Grid Ref ST 466384). Coll and subm July 1982. Comment (JMC): dates wooden trackway positioned near later Bronze Age complex of Tinney's Ground (Coles et al, 1985a). Harwell Radiocarbon Measurements V 89

3210 ± 80 HAR-5710. SLP8301 b13C = -27.8%o Wood, from dumped brushwood on line of prehistoric route at Stile- way, Meare (51 ° 9' 47" N, 2° 45' 56" W, Nati Grid Ref ST 46474076). Coil and subm July 1983 by B J Orme. Comment (JMC): relates to HAR-1221, (Otlet & Walker, 1979, p 362) but HAR-4477 is more recent (Orme et al, 1985a).

4680 ± 70 HAR-5726. SLP8307 b13C = -30.4%o Wood, from newly discovered hurdle in area of intense Neolithic activity at Walton Heath (51 ° 9' 1" N, 2° 46' 44" W, Nail Grid Ref ST 45533935). Coil and subm Sept 1983. Comment (JMC): one of earliest hurdle-like structures in Somerset Levels. Lower in peat strata than previously known Neolithic hurdles in this area; see Orme et al (1985b).

4660 ± 80 HAR-6264. SLP8401 b13C = - 29.7%o Birch (Betula sp) from wooden structure in low peats at Foster's Wal- ton, Walton Heath (51 ° 8' 56" N, 2° 46' 45" W, Natl Grid Ref ST 455392). Coll Aug 1984 by S D Loxton and subm Aug 1984. Comment (JMC): structure, probably at least 20m long, extended over two peat heads, partly destroyed by peat cutting. It formed part of traditional passageway across moors in 3rd millennium BC, with structures such as Jones', Bisgrove and Garvin's trackways serving as links in network (Orme et al, 1985b).

4790 ± 80 HAR-6265. SLP8403 b13C = -29.4%o Hazel from rods of Neolithic hurdle at Frank's Hurdle Track, Meare Heath (51 ° 9' 51" N, 8° 47' 48" W, Natl Grid Ref ST 443409). Coll Sept 1984 by S D Loxton and subm Sept 1984. Comment (JMC): hurdle is possibly oldest known in Britain.

Tinney's series Wood from Tinney's Ground, Somerset (51 ° 8' 24" N, 2° 45' 33" W, Natl Grid Ref ST 469382). For other dates in sub-series see Walker and Otlet (1985, p 79) and Coles and Orme (1980) for description of site. 2960 ± 70 HAR-2773. SLP7808 b13C = -29.7%o From brushwood trackway intersection with oak planking and wooden artifacts. Colt July 1978 by B J Orme and subm July 1978. Comment (JMC): agrees well with dates for group of Bronze Age structures in same area. 3800 ± 80 HAR-3388. SLP7903 b13C = - 28.7%0 From site 79.7 F2 64.8 m. Coil April 1978 by C Sturdy and subm June 90 AJ Walker, RS Keyzor & RL Otlet 1979. Comment (JMC): indication of structures earlier than main bulk of trackways in same field, now totally destroyed. Baker Platform series Wood from Baker Platform, structure of Neolithic character at West- hay, Somerset (51 ° 10' 36" N, 2° 49' 4" W, Natl Grid Ref ST 42854230). See Coles, Fleming and Orme (1980).

4720 ± 80 13C HAR-2843. SLP7810 = -28.5%o From lowest platform. Coil and subm Sept 1975.

4520 ± 90 HAR-2844. SLP7811 S13C = -29.1%0 From edge of platform. Coil and subm Sept 1978.

4950 ± 80 HAR-2845. SLP7812 o13C = -28.5%o From platform. Coll and subm Sept 1978. Comment (JMC): rather early; most of Baker platform dates are late 3rd millennium BC.

4450 ± 100 13C HAR-2846. SLP7813 = -28.4%o From upper brushwood of platform. Coil and subm Sept 1978.

4540 ± 80 HAR-2919. SLP7814 b13C = - 30.3%0 From lower timbers of Baker platform track extension, discovered Nov 1978. Coil by S C Beckett and subm Nov 1978. Comment (JMC): dates basal structure of Neolithic track at junction with platform.

4520 ± 70 HAR-2920. SLP7815 b13C = -30.2%o From platform track extension, discovered Nov 1978. Coll Nov 1978 by B J Orme and subm Nov 1978, relates to HAR-2919. General Comment (JMC): all dates except HAR-2845 agree with stratigraphy of site and with BM-385 (4450 ± 110 BP), Lu-328 (4280 ± 65 BP) and Q- 987 (4230 ± 60 BP) which date uppermost structures. Eclipse site series Peat samples taken to date local pollen assemblage zone boundaries E4/E5, E3/E4, and E2/E3 at Eclipse Site, Somerset (51 ° 9' 41" N, 2° 47' 17" W, Natl Grid Ref ST 449406). Coll Sept 1980 and subm by A E Caseldine. For description of zones see Beckett and Hibbert (1979) and Orme (1982).

3990 ± 100 '3C= HAR-4542. SLP8107 - 26.7%0 Sphagnum/Calluna peat from 0.42m below present ground surface. Harwell Radiocarbon Measurements V 91 Subm July 1981. Comment (JMC): provides essential date for pollen zone boundary E4/E5 and marks beginning of increased clearance activity in area. Eclipse track was constructed during this phase.

4230 ± 80 b13C HAR-4543. SLP8105 = -28. 1/oo Sphagnum/Calluna/Eriophorum peat from 0.70 to 0.71 m below present ground surface. Subm Aug 1981. Comment (JMC): dates E3/E4 boundary, regional assemblage zone boundary C/D, end of forest regeneration phase.

4640 ± 70 HAR-4544. SLP8106 b13C = -27.6/oo Calluna/Eriophorum peat from 0.94 to 0.95m below present ground surface. Subm Aug 1981. Comment (JMC): dates E2/E3 boundary, regional assemblage zone boundary B/C, end of 1st major clearance in area.

5440 ± 70 HAR-4865. SLP8109 b13C = -27.4/00 From base of peat monolith at E80.212 (Nail Grid Ref ST 449406). Subm Feb 1982. Comment (JMC): dates clay-peat interface. General Comment (JMC): dates are earlier than those obtained for probably comparable boundaries at Meare Heath, Abbot's Way and Sweet Track factory site.

Meare Village East series Charcoal from Meare Village East, Somerset (51 ° 10' 33" N, 2° 47' 28" W, Natl Grid Ref ST 447422). Coil April 1982 by B J Orme and subm April 1982.

2080 ± 60 HAR-5000. SLP8203 b13C = - 26.9/00 Context 82.1119.

1740 ± 60 b'3C HAR-5001. SLP8201 = -27.4/00 Context 82.1073.

2090 ± 70 b13C= HAR-5002. SLP8202 -27.00/o0 Context 82.1094. General Comment (JMC): samples used to date stratigraphic context in Iron Age site. HAR-5001 is from same broad context as HAR-5000; -5002 is from lower layer in same sequence, ie, all three samples are from same mound. For description of site see Orme, Coles and Silvester (1983).

Honeygore series Samples from tracks excavated in Neolithic peat at Honeygore 92 AJ Walker, RS Keyzor & RL Otlet Complex, Somerset Levels (51 ° 10' 51" N, 2° 50' 8" W, Natl Grid Ref ST 416428). For description of site see Coles et al (1985b).

4610 ± 90 HAR-5721. SLP8302 b13C =-28.2%o Wood, from Honeygore track. Coil Aug 1983 by A E Caseldine and subm Aug 1983. Comment (JMC): slightly younger than previous dates which centered on ca 2800 BC (Coles et al, 1985b).

4610 ± 100 HAR-5722. SLP8303 b13C = - 30.2%0 Wood from hurdle A of Honeybee track. Coil and subm Aug 1983. Comment (JMC): cJHAR-5723.

4500 ± 70 HAR-5723. SLP8304 b13C = -28.3%o Wood from hurdle B of Honeybee track. Coil and subm Aug 1983. Comment (JMC): see HAR-5722, above, with which these is good correlation.

4560 ± 80 HAR-5724. SLP8305 b13C = - 28.6%0 Wood from complex of Honeycat track. Coil and subm Aug 1983. Comment (JMC): slightly older than previous dates: HAR-653, -652 (Otlet, 1977, p 415) and Q-427, -429, -320; but see discussion in Coles et al (1985b).

4300 ± 70 b13C= HAR-5727. SLP8306 -27.5 o Peat, from thin peat surface underlying hurdle B of Honeybee track. Coil Aug 1983 and subm Sept 1983. Comment (JMC): younger than wood overlying HAR-5723.

Priory Barn series Wood from Augustinian Priory excavation, Taunton, Somerset (51 ° 1' 4" N, 3° 5'57" W, Nail Grid Ref ST 229249). Coil 1978 by P Leach and subm July 1978 by J Hillam, Dept Prehistory and Archaeol, Univ Shef- field.

840 ± 90 HAR-2804. TAUN40A b13C = -26.4%o AML 781642, 7-year-old ash twig, part of bulk wood sample from Layer 40.

1090 ± 80 HAR-2806. TAUN96 313C = -28.3%o AML 781641, oak from Layer 96. Comment (JH): growth allowance not known but contained at least 20 growth rings. Harwell Radiocarbon Measurements V 93

870 ± 70 HAR-2815. TAUN40B b13C = -27.4%0 AML 781643, willow/poplar twig or branch ca 17 years old, part of bulk wood sample from Layer 40.

Ledston series Samples from Ledston, West Yorkshire (53° 45' 35" N, 10 20' 35" W, Natl Grid Ref SE 433295). Coll Sept 1976 by J Keighley and I Hodder, subm Aug 1978 by R E Yarwood, West Yorkshire Co Council, Wakefield.

2080 ± 100 HAR-2805. LEDSF363 b13C = - 22.0%0 Bone, AML 781940, from burial sealed in bottom of storage pit (F704).

2270 ± 70 HAR-2825. LEDSF418 b13C = -24.6%0 Charcoal, AML 781939, from sample sealed in fill of post hole (F539). General Comment (REY): these are 1st 14C dates for prehistoric site on West Yorkshire limestone. They confirm Iron Age occupation of Ledston and, together with crop mark evidence for this area, suggest extensive prehistoric settlement.

Silver St series Oak samples, taken as part of dendrochronologic study, from Silver St, Glastonbury, Somerset (51° 8' 48" N, 2° 42' 48" W, Nat! Grid Ref ST 501389). Col! 1978 by R H Leech and subm July 1978 by J Hillam. 1340 ± 70 HAR-2812. GL263 b13C = -25.2%0 AML 781638, ca 20 rings from sample containing 38 rings taken close to outer edge of tree.

1450 ± 80 HAR-2813. GL239 b13C = - 26.7%0 AML 781639, from rings 15 to 34 of 79-year-old sample, probably from near outer edge of tree, although no sapwood present.

1470 ± 80 HAR-2814. GL249B b13C = -27.5%0 AML 781640, from sample containing ca 20 rings. Comment (JH): not known if from inner or outer rings. General Comment (JH): dendrochronology suggests that timbers 239 (HAR- 2813) and 263 (HAR-2812) are contemporary but too few rings present to be certain. 94 Af Walker, RS Keyzor & RL Otlet Barton series Wood id by C A Keepax as oak (Quercus sp) from St Peter's Church, Barton-on-Humber (53° 41' 1" N, 0° 25' 59" W, Natl Grid Ref TA 03472195). Coil July 1978 and subm Aug 1978 by W J Rodwell, Old Vic- arage, Chilcompton, Somerset. Site is described in Rodwell and Rodwell (1982).

720 ± 60 HAR-2863. BHO5 5130 = -25.5%o AML 781906, from stump of beam embedded in belfry wall. Comment (WJR): wall supported former 13th to 14th century spire, probably contemporary with HAR-2864, below.

780 ± 80 HAR-2864. BHO6 S13C = -26.2%o AML 781907 from central joist in floor of ringing chamber of tower. Comment (WJR): floor probably contemporary with HAR-2863, above. 570±70 HAR-2865. BHO3 8130 = - 25.0%o AML 781904, from E-W timber of base frame of former 13th to 14th century spire.

960 ± 70 HAR-3106. BHO1 8130 = - 26.0%o AML 781902, from mature timber. Comment (WJR): joist fragment from original Saxon timber supporting floor of upper chamber in Anglo- Saxon W annex of church. Mucking-Central Unit series Charcoal from recut Iron/Bronze Age ring ditch at North Ring, Muck- ing, Essex (51 ° 30' 13" N, 0° 24' 52" E, Natl Grid Ref TQ 67558112). Coil July 1978 by D Bond and subm Oct 1978 by N D Balaam, Ancient Monu- ments Lab. For description of site see Bond (1980).

2630 ± 110 8130 HAR-2893. 23-1658 = -24.2oo AML 7824847, id by C A Keepax as Acer sp (mature timbers) and some Prunus sp, overlying primary fill. 2700 ± 80 HAR-2911. 23.34 8130 = -24.6%o AML 7820949, id by C A Keepax as Acer sp, Quercus sp and Corylus/ Alnus sp, from mature timbers, from primary fill. Warbleton series Charcoal from iron smelting bloomery furnace at Warbleton Parish, Turners Green, Heathfield, East Sussex (50° 57' 5" N, 0° 20' 11" E, Natl Harwell Radiocarbon Measurements V 95 Grid Ref TQ 64091954). Coil Nov 1978 by W R Beswick and 0 Bedwyn and subm Nov 1978 by W R Beswick, Turners House, Heathfield, Sussex. Site is described in Beswick (1978, 1979).

1810 ± 70 HAR-2930. TG2 b13C = -26.3%o AML 785542, id by C A Keepax as oak (Quercus sp) from mature timbers, from reaction zone of furnace still containing some charge.

2040 ± 70 HAR-2932. TG4 bl3C = -25.1%o AML 785543, id by C A Keepax as oak (Quercus sp) from mature timbers, from same context as HAR-2930, above.

1900 ± 70 HAR-3017. TG3 b13C = -24.8%o AML 785541, from smithy area 4.8m from furnace, 0.62m below cultivation surface. General Comment (WRB): HAR-2930, -2932, and -3017 are consistent with dates for two other sites with similar unusual design of bloomery furnace- Levisham, East Yorkshire and Engsbach, Siegerland, West Germany (Rut- ter & Hayes, 1970; Hayes, 1983; Weierhausen, 1939). Dates confirm smithy area was contemporaneous with furnace.

Condicote series Charcoal from early secondary fill of inner ditch of henge monument at Condicote, Gloucestershire (51 ° 57' 13" N, 1° 46' 34" W, Nat! Grid Ref SP 15382841). Coll 1977 and subm 1978 by A Saville, Art Gallery & Mus, Cheltenham. For description of site see Saville (1983).

3720 ± 80 HAR-3064. CH77C69A b13C = -28.00/o0 AML 790631, id by C A Keepax as hawthorn (Rosaceae subfamily Pomoi- deae) and hazel or alder (Corylus sp) from mature timbers, in Layer 9A. Com- ment (AS): sample assoc with prehistoric pottery of Beaker-related type.

3670 ± 100 b13C HAR-3067. CH77C114 = -26.1 /oo AML 790632, id by C A Keepax as oak (Quercus sp) from mature timbers, in Layer 14.

Great Linford series Wood id by C A Keepax as oak (Quercus sp) from mature timbers, from buried supports of Medieval Post Mill (Site 3709) at Great Linford, Buck- inghamshire (52° 4' 12" N, 0° 45' 7" W, Nat! Grid Ref SP 85544208). Coil Sept 1977 and subm Feb 1979 by R J Zeepvat, Bradwel! Abbey Field Cen- tre, Milton Keynes (Zeepvat, 1980). 96 AJ Walker, RS Keyzor &RL Otlet

750 ± 70 HAR-3121. M 1098W 1 B Y3C = -25.5%o AML 790636.

730 ± 80 HAR-3122. M1098W1A Y3C = -25.6 o AML 790635. General Comment (RJZ): beam appears to have measured 0.3m x 0.3m x 5.5m. It was packed with stone in cross-shaped trench. Samples originally from near center of beam.

Rivington-Anglezarke series Peat taken as part of paleo-ecologic studies on Rivington and Angle- zarke Moors, Lancashire. Human exploitation of area dates from Meso- lithic/Neolithic (flint finds and barrow, Bu'lock, 1958) through Historic period to present. Samples coil and subm by M G Bain and B Barnes, Bol- ton Inst Higher Educ, Yorkshire. Dates were required to support strat and palynol evidence and provide chronol framework for record. Similar studies are described in Hibbert, Switzur and West (1971) and Tallis and McGuire (1972).

2940 ± 70 HAR-6206. WH2 b13C = -29.4%o AML 832900, from 1.17 to 1.20m depth, well-marked clearance phase with cultivation and possible relation to metal extraction, at Winter Hill pollen site, Rivington Moor (53° 37 39" N, 2° 31' 7" W, Natl Grid Ref SD 657147). Coll Oct 1979. Comment (BB): date corresponds well with estimates based on peat accumulation rates in other local profiles, and marks beginning of sustained clearance activity in Rivington-Anglezarke uplands.

5660 ± 80 HAR-6207. BB1 b13C = -29.6%o AML 832869, from 2.42 to 2.46m depth, peat initiation and clearance phase, at Black Brook pollen site, Anglezarke Moor (53° 39' 41" N, 2° 33' 25" W, Natl Grid Ref SD 632185). Coll Nov 1980. Comment (BB): date agrees well with initial estimates of age (Mesolithic, ie, ca 5500 BP) based on palynol and strat data. Marks earliest forest clearance phase yet detected in area.

310 ± 80 HAR-6208. PKA1 li13C = -29.6%o AML 832897, id as sphagnum/monocot from 0.35 to 0.37m depth at Pikestones pollen site, Anglezarke Moor (53° 38' 53" N, 2° 33' 41" W, Natl Grid Ref SD 629170). Coll Nov 1980. Comment (BB): date is younger than expected (Norse or Medieval) but accords well with documentary evidence for increased agric activity in 16th and 17th centuries AD on uplands. Date reinforces palynol evidence of truncated profile. Harwell Radiocarbon Measurements V 97

1710 ± 70 b13C HAR-6209. PK1 = - 29.4%o AML 832898, from 0.4 to 0.45m depth at Pikestones pollen site, Anglezarke Moor (53° 38' 59" N, 2° 33' 52" W, Nat! Grid Ref SD 627172). Coll Nov 1980. Comment (BB): dates peat initiation on flanks of moor, with clearance phenomena. Romano-British date is earlier than expected (ca 1000 BP) but there are problems in estimating age based on accumulation rates in profile. Result corresponds with clearances dated from Round Loaf samples (HAR-6211, -6212, below). Palynol evidence from both sites suggests considerable upland exploitation at this time including both arable and pastoral activity.

4740 ± 70 b13C HAR-6210. BB2 = -29.1 %o AML 832899, from 2.20 to 2.24m depth, early clearance phase with evidence of elm decline, at Black Brook pollen site, Anglezarke Moor (53° 36' 59" N, 20° 33' 23" W, Nat! Grid Ref SD 632135). Coll Nov 1980. Comment (BB): agrees well with accumulation rate estimates from Round Loaf site and marks `Elm decline horizon' within published range for NW England.

1550 ± 70 HAR-6211. RL3 b13C = -28.9%o AML 832901, id as Calluna eriophorum, from 0.50 to 0.55m depth, horizon marking end of well-defined clearance phase at Round Loaf pollen site, Anglezarke Moor (53° 39' 25" N, 2° 33' 14" W, Natl Grid Ref SD 634180). Coll Aug 1980. Comment (BB): although earlier than estimated (ca 900 BP), date relates well-marked clearance and agric episode to Romano- British period and accords well with RL4 (HAR-6212).

1710 ± 70 HAR-6212. RL4 b13C = -28.2%o AML 832902, id as Calluna eriophorum, from 0.60 to 0.65m depth, beginning of well-defined clearance phase, at Round Loaf pollen site, Anglezarke Moor (53° 39' 25" N, 2° 33' 14" W, Natl Grid Ref SD 634180). Coll Aug 1980. Comment (BB): date agrees well with RL3 (HAR-6211, above) and marks initiation of well-defined Romano-British upland clearance horizon. This corresponds with clearance evidence of similar date (see HAR-6209, above) from Pikestones pollen site.

REFERENCES Anonymous, 1973, Wiltshire Archaeol Magazine, v 68, p 57. Beckett, S C and Hibbert, F A, 1979, Vegetational change and the influence of prehistoric man in the Somerset Levels: New Phytologist, v 83, p 577-600. Beswick, W R,1978, A note on early ironmaking in Sussex: Sussex Industrial Hist, v 8, p 23. ----- 1979, Ironm ak0ing origins and their early impact on the English Weald: Sussex Industrial Hist, v 9, p Bond, D, 1980, North Ring, Mucking: an enclosure of the Late Bronze Age: East Anglian Archaeol Repts (Essex). 98 AJ Walker, RS Keyzor & RL Otlet Bu'lock, J D, 1958, The Pikestones, a chambered long cairn of Neolithic type on Anglezarke Moors, Lancashire: Trans Lancashire Cheshire Antiquarian Soc, no. 68, p 143-145. Carver, M 0 H, Donaghey, S and Sumpter, A B, 1978, The Colonia: Riverside structures and a well in Skeldergate and buildings in Bishophill: The Archaeol of York, v 4, no. 1, p 25. Chowne, P, 1980, Bronze Age settlement in South Lincolnshire, in Barrett, J C and Bradley, R J, eds, Settlement and society in the Later Bronze Age: British Archaeol Repts, v 83. Coles, J M, Fleming, A M and Orme, B J,1980, The Baker site: a Neolithic platform: Somerset Levels Papers, v 6, p 6-23. Coles, J M and Orme, B J,1977, Garvin's tracks: Somerset Levels Papers, v 3, p 73-81. -----1978, The Meare Heath track: Somerset Levels Papers, v 4, p 11-39. ---1980, Tinney's Ground 1978-1979: Somerset Levels Papers, v 6, p 61-68. -----1983, The Sedgemoor survey 1982: Somerset Levels Papers, v 9, p 6-8. Coles, J M, Orme, B J, Caseldine, A E and Morgan, R A, 1985a, Godwm's track: a Bronze Age structure at Sharpham: Somerset Levels Papers, v 11, p 69-74. ----- 1985b, A Neolithic jigsaw: the Honeygore complex: Somerset Levels Papers, v 11, p 51-61. Colledge, S M, 1978, An interim environmental report, the King's Pool, Stafford: Ancient Monuments Lab Rept, no. 2535. Donaghey, 5, 1986, `58-9 Skeldergate', in Moulden, J and Tweddle, D, eds, Anglo-Scandinav- ian settlement south-west of the Ouse: The Archaeol of York, v 8, no. 1, p 37-48. Fasham, P J, 1980, Excavations on Bridget's and Burntwood Farms, Itchen Valley Parish, Hampshire, 1974: Hants Field Club Archaeol Soc Proc, v 36, p 37-86. Green, CJ S, 1971, Interim report on excavations in the Roman cemetery, Poundbury, Dor- chester, 1970: Dorset Nat Hist Archaeol Soc Proc. ----- 1974, Interim report on excavation in the Roman cemetery, Poundbury, Dor- chester, 1973: Dorset Nat Hist Archaeol Soc Proc, v 95, p 100. ----- 1976, Poundbury, in Dorset archaeol in 1976: Dorset Nat Hist Archaeol Soc Proc, v 98, p 55-56. ----- 1982, The cemetery of a Roman-British Christian community at Poundbury, Dorchester, Dorset, in Pearce, S M, ed, The early church in Western Britain and Ireland: British Archeol Research Repts, v 102. Hall, A R, Kenward, H K and Williams, D, 1980, The past environment of York: Environ- mental evidence from Roman deposits in Skeldergate: The Archaeol of York, v 14, no. 3, p 105. Hayes, R H, 1983, Levisham Moor archaeologic investigation 1957-1978: North Yorks Moors Nat Park Comm/Scarborough Archaeol Hist Soc, p 24. Hibbert, F A, Switzur, V R and West, R G, 1971, Radiocarbon dating of Flandrian pollen zones at Red Moss, Lancashire: Royal Soc [London] Proc B, v 177, p 161-176. MacGregor, H, in press, Structures on the north-east side of Aldwark, in Hall, R A, MacGre- gor, H and Stockwell, M, eds, Medieval tenements in Aldwark and other sites: The Archaeol of York, v 10, no. 2. McGrail, S, 1978, Logboats of England and Wales: Nail Maritime Mus Archaeol ser no. 2, Br Archaeol Repts, v 51, p 275-276. ----- 1981, A medieval logboat from the River Calder at Stanley Ferry: Medieval Archaeol, v 25, p 160-164. -----1983, Interpretation of archaeol evidence for maritime structures, in Annis, P, ed, Sea studies: Natl Maritime Mus, p 33-46. Mercer, R J, 1980, Hambledon Hill, a Neolithic landscape: Edinburgh, Edinburgh Univ Press. ---- 1985, A Neolithic fortress and funeral gate: Scientific Am, v 252, no. 3, p 94- 101. Morgan, R A and Schofield, J, 1978, Tree rings and the archaeol of the Thames waterfront in the City of London, in Fletcher, J M, ed, Dendrochronology in Europe: Council Br Archaeol Repts, Internatl ser, v 51, p 223-238. Orme, B J, 1982, The use of radiocarbon dates from the Somerset Levels: Somerset Levels Papers, v 8, p 9-25. Orme, B J, Caseldine, A E and Morgan, R A, 1982, Recent discoveries on Walton Heath: Gar- vin's, Bisgrove's and Jones' tracks: Somerset Levels Papers, v 8, p 51-64. Orme, B J, Coles, J M and Silvester, R J, 1983, Meare Village East 1982: Somerset Levels Papers, v 9, p 49-74. Orme, B J, Coles, J M, Caseldine, A E and Morgan, R A, 1985a, A later Bronze Age complex at Stileway: Somerset Levels Papers, v 11, p 75-79. -----1985b, Third millennium structures on Walton Heath: Somerset Levels Papers, v 11, p 62-68. Otlet, R L, 1977, Harwell radiocarbon measurements II: Radiocarbon, v 19, no. 3, p 400- 423. Harwell Radiocarbon Measurements V 99

Otlet, R L, 1979, An assessment of errors in liquid scintillation methods of 14C dating, in Berger, R and Suess, H E, eds, Radiocarbon dating, Internatl 14C conf, 9th, Proc: Berke- ley, Univ California Press, p 256-267. Otlet, R Land Walker, A J,1979, Harwell radiocarbon measurements III: Radiocarbon, v 21, no. 3, p 358-383. ----- 1983, The computer writing of radiocarbon reports and further developments in the storage and retrieval of archaeological data, in Mook, W G and Waterbolk, H T, eds,14C and archaeology, Internatl symposium, 1st, Proc: PACT, v 8, p 91-106. Otlet, R L and Warchal, R M, 1978, Liquid scintillation counting of low-level 14C, in Crook, M A and Johnson, P, eds, Liquid scintillation counting, vol 5: London, Heyden, p 210- 218. Rodwell, W J, 1976, The archaeological investigation of Hadstock Church Essex: Antiquaries Jour, v 56, p 55-71. Rodwell, W J and Rodwell, K A, 1982, St Peter's Church, Barton-upon-Humber: Excavation and structural study 1978-1981: AntiquariesJour, v 62, p 283-315. ----- 1985, Rivenhall: investigations on the Roman villa, church and village: Council Br Archaeol Res Rpts, Br ser, v 55. Rutter, J G and Hayes, R H, 1970, A pre-Roman iron-smelting site at Levisham, North Yorks: Bull Hist Metallurgy Soc, v 4, no. 2, p 79. Saville, A, 1983, Excavations at Condicote Henge monument, Gloucestershire 1977: Trans Bristol Gloucester Archaeol Soc, v 101, p 21-47. Smith, K, 1979, The excavation of Winklebury Camp, Basingstoke Hants: Prehist Soc Proc, v 43. Stead, I M, 1977, La Tene burials between Burton Fleming and Rudston, North Humberside: Antiquaries Jour, v 56, pt II, p 224-225. Tallis, J H and McGuire, J,1972, Central Rossendale: the evolution of an upland vegetation. I. The clearance of Woodland: Jour Ecology, v 60, p 721-737. Tatton-Brown, T, 1974, Excavations at the Custom House site, City of London 1973: Trans London Middlesex Archaeol Soc, v 25, p 117-219. Walker, A J and Otlet, R L, 1985, Harwell radiocarbon measurements IV: Radiocarbon, v 27, no. 1, p 74-94. Weierhausen, P, 1939, Vorgeschichtliche eisenhutten im Deutschlands: Leipzig, p 9. Wright, E V, 1976, The North Ferriby boats: Natl Maritime Mus, Mon no. 23. Zeepvat, R J, 1980, Post mills in archaeology: Current Archaeol, v 71, p 375-377. [RADIOCARBON, VOL 29, No. 1, 1987, P 100-] 14] REHOVOT RADIOCARBON MEASUREMENTS III ISRAEL CARMI Department of Isotope Research, The Weizmann Institute of Science, 76100 Rehovot, Israel This list contains results obtained between 1981 and 1985. Since the first description of the laboratory (Carmi, Noter & Schlesinger, 1971) the following changes were made. Two proportional counters are now used: 1) 0.5L volume, 0.865 ± 0.023 cpm background, 12.830 ± .134 cpm NBS oxalic acid standard (old); 2) 0.25L volume, 0.484 ± 0.023 cpm background, 6.185 ± .123 cpm NBS oxalic acid standard (old). The passive shield has been increased by 2cm of mercury next to the counters. For anticoincidence, a modular, hand-made gas counter is used. The laboratory was transferred to the ground floor of a 7-storey building. Data acquisition and processing are done with a scaler/buffer built at the Institute and an IBM PC computer. Samples are filled into the counters and counted for ca l 000min at least twice. The sample preparation method and counter filling pressure have not been changed.

ACKNOWLEDGMENT Thanks are due S Kazes for technical help in the lab.

ARCHAEOLOGIC SAMPLES

Marine Samples Israel Kfar Samir series Prehistoric submerged settlement 2km S of Haifa, 50m offshore (Natl Grid ref 1461-2441).

4800 ± 70 RT-598B. b13C = -26.4%o Oak tree excavated 1 m below sea level (bsl). Coll 1981 by M Evron, Lab Prehist, Univ Haifa.

RT-682A. 6470 ± 130 Wood from construction #5, 4.5m bsl. Coll 1984 by A Raban, Center for Marine Studies, Univ Haifa.

6670 ± 140 RT-682B. b'C = -26.4%o Wood from construction #3.4.5m bsl. Coll 1984 by A Raban. CAHEP series Samples coil during Cesarea Ancient Harbor Excavation Proj by A Raban and N Karmon.

100 Rehovot Radiocarbon Measurements III 101

RT-609. Harbor 1470 ± 50 Coil 1981 (Nat! Grid Ref 1397-2124). Comment (AR): wood from cra- dle, used to lower construction stones during attempt by Emperor Anasta- sius to rebuild harbor.

2210 ± 190 RT-631A. Harbor/ship b13C = -26.4°oo Coll 1983 from 9m bsl (Natl Grid Ref 1398-2124). Comment (AR): beam from entrance to harbor in Herodian period. Sample is from quay or ship. 1970±70 '3C RT-631 B. Harbor b = - 25.8%o Coll 1983 from 6m bs! (Nail Grid Ref 1398-2124). Comment (AR): wood from frames used in construction of harbor.

RT-645. Side plate 1870 ± 60

1990 ± 140 RT-653B. Rib I b13C = -28.0%o

1990 ± 150 RT-653C. Rib II b'3C = -26.6%o

1930 ± 220 RT-680A. Tenon I b13C = -27.1%o

RT-680B. Tenon II 1990 ± 100 Col! 1983 from 2.5m bsl (Nat! Grid Ref 1403-2127). Comment (AR): parts recovered from sunken ship from 1st century AD.

RT-652. Acre 2310 ± 50 Wood from sideplate of ship brought up by dredger during deepening of harbor (Nat! Grid Ref 1569-2583). Coll 1983 and subm by N Karmon.

100 ± 100 b13C RT-684. Dor ship! = -25.1 %o Wood from unid. sunken ship at Dor, 27km S of Haifa (Nail Grid Ref 1424-2237). Coll 1983 by S Wachsmann, Dept Antiquities, Ministry Educ, from 2 to 3m bsl. Comment (SW): possibly from ship that sank in 1664.

1590 ± 110 RT-686A. Dor ship!! b13C = -28.2%o Wood from Byzantine ship at Dor (Nail Grid Ref 1422-2238). Coll 1983 by S Wachsmann, from 2 to 3m bsl (Wachsmann & Raveh, 1984). Comment (SW): ceramics suggest that ship is from 6th-7th century AD. 102 Israel Carmi 990 ± 100 RT-686B. Atlit ship S13C = -28.8%o Wood from sunken ship in Atlit (Nat! Grid Ref 1449-2346). Coil 1982 by S Wachsmann, from 2 to 3m bsl.

1800 ± 100 RT-710. Hahotrim ship b13C = - 25.4%0 Wood from sunken ship in Hahotrin (Nat! Grid Ref 1456-2400). Coil 1984 by S Wachsmann from 2 to 3m bsl.

2100 ± 110 RT-681. Ram 3'3C = -27.6%o Wood from bronze ram. Coil 1980 offshore of Atlit (Nat! Grid Ref 1445-2348). Subm by N Karmon.

8140 ± 130 RT-707. NahalOren Y3C = -26.4%o Charcoal from prehistoric site presently 300m offshore of Nahal Oren. Coil 1984 from 1 m bsl by E Galilee, Center for Marine Studies, Univ Haifa. Italy

RT-705. Oristano 100 ± 100 Wood from sunken ship, 20km N of Oristano, Sardinia, 2m bsl. Coil 1984 by E Galilee.

Terrestrial Samples Israel

RT-611. Olive seeds 320 ± 70 Charred olive (Olea europea) from old agric terrace in Jerusalem. Coil 1982 by G Edelstein, Dept Antiquities, Ministry Educ. Comment (GE): Canaanite pottery was found in terrace but sample is probably of secondary origin.

RT-614. Kaukab 700 ± 70 Olive tree (Oleo europea) from Kaukab in lower Galilee (Nati Grid Ref 1735-2496). Coil 1982 by Y Sela, Jewish Natl Fund. Comment (YS): part of tree exposed by erosion. Uvda Valley series Archaeol excavation in Arava valley 40km N of Eilat. a) Loc 906 (Natl Grid Ref 1468-9297). Charcoal from stone bowl 0.5m below ground surface. Coll 1980 by 0 Yogev, Dept Antiquities, Ministry Educ (Yogev, 1984). Rehovot Radiocarbon Measurements 111 103

6560 ± 90 RT-628A. b13C = -10.9%o

RT-628B. 6400 ± 200 Comment (OY): open sanctuary. Sites with similar stereographic plans have not been previously known before 3rd-4th millennium BC. b) Loc 916 (Natl Grid Ref 1465-9287). Samples from residential site. Coll 1980 by 0 Yogev.

4800 ± 70 RT-640A. b13C= -24.6%o Charcoal from stone cache 1.5m below ground surface.

4400 ± 60 RT-640B. b13C = -23.9%o Charcoal from under secondary wall.

4280 ± 80 RT-640C. 513C = -22.7%o Charcoal from implement 0.7m below ground surface. Comment (OY): residential sites are known in region from 3rd millennium BC.

RT-648A. Shrine 5670 ± 90 Charcoal from massebot shrine (Nail Grid Ref 1495-9255). Coll 1982 by U Avner, Dept Antiquities, Ministry Educ. Comment (UA): evidence for early desert habitation and cult sites (Henry, 1982; Rosen, 1984).

RT-648B. Threshing floor 4250 ± 50 Charcoal from threshing floor (Natl Grid Ref 1495-9255). Coll 1982 by U Avner. Comment (UA): date suggests MBI period but find points to EBII, 400 yr earlier. This date, together with those of samples 714A and 714B point to longer duration of EBII culture in desert compared to more humid regions.

4070 ± 100 RT-714A. Site 9 b13C = -13.7%0 Charcoal from residential site (Nail Grid Ref 9683-1462) 0.6m below ground surface. Coll 1980 by U Avner and 0 Ilan (Dept of Antiquities, Ministry Educ).

3850 ± 100 RT-714B. Site 166 b13C = -18.2%o Charcoal from residential site (Natl Grid Ref 9277-1459) 0.7m below ground surface. Coll 1980 by U Avner.

Dor series Ancient harbor 24km S of Haifa (Natl Grid Ref 1425-2247). 104 Israel Carmi

2830 ± 70 5130 RT-630. Floor 1 -23.4%o Coil 1981 by A Raban.

3640 ± 200 RT-685. Locus 101 513C -25.1%o Coil 1984 by A Raban.

Zalaka series Tumuli tombs field in Wadi Zalaka, E Sinai (Nail Grid Ref 0884-8239) (Avner, 1984). All samples are charcoal, coil 1983 by U Avner. Comment (UA):date supports idea that appearance of tumuli should be moved back to 4th or 5th millennium BC.

5440 ± 80 RT-648E. S13C = - 23.9%0

Har Shani series Charcoal from open shrine 18km NW of Eilat (Natl Grid Ref 1360- 9000). Coil 1981 by U Avner. Comment (UA):evidence suggests that shrine had been in intermittent use between 4th millennium BC and Byzantine times. Date suggests that shrine had been in use by the Nabatean as late as 6th century AD.

1470 ± 60 RT-648F. 5130 = - 22.6%0

RT-648G. 1500 ± 170

Shiqmim series Charcoal from Chalcolithic village near Beer Sheva (Natl Grid Ref 1170-0689). Coil 1982 by T E Levy, Negev Mus, Beer Sheva (Levy, 1983).

5750 ± 180 RT-649B. o13C = - 22.1 %o

6150 ± 180 RT-649D. Locus 412 o13C = -17.7%0

8100 ± 150 b13C RT-650. Nahal Hemar = -23. 7%o Charcoal from cave in Judean Desert (Nail Grid Ref 1675-0645), from Neolithic pre-ceramic B layer which contains intact artifacts. Coil by 0 Bar- Yosef, Inst Archeol, Hebrew Univ. Comment (IC): measurements in other Rehovot Radiocarbon Measurements III 105 labs gave following results: 6230 ± 80 BC (PTA-3650) and 6300 ± 70 BC (BM-2298).

1380 ± 180 RT-656. Kasr El Yahud b13C = -20.2%o Wood from common burial ground at Kasr El Yahud in Lower Jordan R, (Natl Grid Ref 2012-1386). Coll 1983 by J Zias, Dept Antiquities, Minis- try Educ. Yiftahel series Burned bricks from excavation at Yiftahel in lower Galilee (Nail Grid Ref 1710-2405). Coll 1984 by E Brown, Dept Antiquities, Ministry Educ.

RT-702A. 5570 ± 220 Sample from EBI layer.

RT-702B. 7460 ± 210 Sample from Neolithic pre-ceramic layer.

5540 ± 110 RT-718. Silo site b13C = -22.Ooo Triticum diococcum from silo in Chalcolithic site, Golan Heights (Natl Grid Ref 2234-2564). Coll 1981 by C Epstein, Dept Antiquities, Ministry Educ.

Other Countries

RT-612A. Honduras del Oeste 3540 ± 70 Shells (Caracolus excellens) from Santo Domingo. Coll from ancient refuse dump in 1981. Subm by M Vellos Magiolo.

RT-612B. Cacoq 3090 ± 50 Shells (Arca occidentalis) from Ihle a Vache, Haiti. Coll 1982, subm by Clark Moore.

CARBONATE SAMPLES OF BIOGENIC ORIGIN Tiran series Samples from Favel Bay, Straights of Tiran (Natl Grid Ref 1100-0735). Coll 1981 by E Spanier, Center Marine Studies, Univ Haifa.

RT-601A. 1570 ± 80 Chicoreus ramosus (gastropod).

RT-601 B. > 30,000 Fossilized sample of echinoid (sand dollar). 106 Israel Carmi Achziv series Samples from Achziv, 26km N of Haifa (Natl Grid Ref 1596-2718). Coil 1981 by Z Levy and D Neev, Geol Survey Israel, from terrace 7m above sea level. 3240 ± 180 RT-660A. b13C = -4.5%o Cerastoderma glaucum. 6000 ± 170 RT-660B. S13C = -8.9%o Unio sp. 3640 ± 160 '3C= RT-660C. - Q 9/oo Euthria cornea. RT-683A. Acre PMC = 82.0 ± 2.2 Aragonitic shell (Euthria cornea). Coil ca 1935 from beach at Acre, 30km N of Haifa (prebomb sample). Subm 1984 by D Neev. RT-683B. Tel Aviv PMC = 89.0 ± 2.0 Aragonite shell (Euthria cornea) coil ca 1960, from beach at Tel Aviv (prebomb sample). Subm 1984 by D Neev.

Land snail series Land snails were coil by A Karnieli (AK), Desert Research Inst, Sde Boger, and G A Goodfriend (GG), Weizmann Inst. All samples are from Negev Desert except for RT-674 which is from Jamaica. Data is given in Table 1. Results for live samples are given in percent modern corrected for '3C fractionation, in italics. Natl Grid Refs are given where available; for Jamaican sample, international grid is given. b13C values in parenthesis were estimates by GG. Arad snails (sample RT-746A) were excavated by R Ami- ran, Hebrew Univ. H Meinis, Zoology Mus Hebrew Univ (HUZM) provided live-collected prebomb land snail shells. Samples RT-732, -741, -744 are from rodent middens. Comment (GG): most fossil snail material was excavated from loessial sediments. Specimens were thoroughly cleaned of all secondary deposits inside and outside. Ages are reported uncorrected for anomalies to which land snails from carbonate substrates are subject (Goodfriend & Stipp, 1983) which are due to incorporation of carbonate carbon into shell (Goodfriend & Hood, 1983). Reported ages are thus ca 1000-2000 yr too old.

HYDROLOGIC SAMPLES The Arava samples were coll by R Nativ, Desert Research Inst, Sde Boqer. Galilee and Golan Heights samples were coil by M Stiller, Weizmann Inst, and I Carmi (Carmi, Stiller & Kaufman, 1985), except for Lake Kin- TABLE 1 14C in land snails

Sample Colln Nat! Grid Ref Subm BP no. date E N PMC RT-626A 1982 1325-0340 AK seetzeni in loess Sp hinc t hil a zona t a la r -626B 1982 1325-0340 Trochoidea seetzeni 130cm in loess Sphincterochila zonata layer -632A 1983 1310-0299 zonata coil -632B 1983 1310-0299 zonata toll -655A 1983 1310-0299 zonata in loess layer -655B 1983 1310-0299 zonata in loess l -671B 1984 1905-1340 mbriata 170 Is-30 -674 1981 420-566 lucerna Jamaica -675 1984 1325-0340 seetzeni ive toll -679 1984 1325-0340 seetzeni -687 1984 1858-1381 mbriata -693 1942 GG zonata 1.7 live toll -712A 1949 GG caesareans 2.3 Live toll -712B 1952 GG hincterochila mbriata 1.9 Live toll -712C 1949 GG seetzeni 2.6 Live toll -712ll 1955 cc seetzeni 1.7 live toll -712E 1941 GG zonata 1.4 Live toll -721 1985 1400-0418 seetzeni depth Is-168 -722 1985 1400-0418 seetzeni depth Is-168 0 TABLE 1 continued

Sample Colln Nat! Grid Ref BP no. date E N PMC -725 1985 1932-1276 GG Trochoideaseetzeni -4.5 8300 ± 260 Is-84 -726 1984 1556-0742 GG Trochoidea seetzeni -5.3 7250 ± 180 Is-111 -727 1985 1325-0513 GG Trochoidea seetzeni -2.9 8270 ± 180 Is-190 -729A 1985 ] 427-0751 GG Trochoidea seetzeni -5.4 4330 ± 170 Is-202 -7296 1985 1558-0642 GG Trochoideaseetzeni -4.6 6000 ± 180 Is-174 -731 1985 1312-0630 GG Trochoidea seetzeni -4.5 8120 ± 120 Is-215 -732 1985 1311-0636 GG Trochoideaseetzeni -5.3 16160 ± 530 Is-213 -733 1985 1316-0649 GG Trochoidea seetzerei -4.8 7860 ± 260 Is-211 -738 1985 1492-0510 GG Trochoidea seetzeni -2.2 11140 ± 310 Is-263 -739 1985 1275-0441 GG Trochoidea seetzeni (-4.0) 9190 ± 220 Is-258 -740 1985 1316-0641 GG Trochoidea seetzeni (-4.0) 10500 ± 420 Is-280 -741 1985 1441-0664 GG Trochoideaseetzeni (-3.4) 11230 ± 140 Is-291 -742 1985 1395-0688 GG Trochoidea seetzeni (-4.0) 11150 ± 250 Is-281 -743 1985 1297-0616 GG Trochoide¢ seetzeni -4.5 7700 ± 190 Is-277 -744 1985 1316-0642 GG Trochoidea seetzeni -4.0 4690 ± 190 Is-279 -745 1985 1500-0518 GG Trochoidea seetzeni -3.1 5230 ± 120 Is-301 -746A 1970 1620-0765 GG Trochoide¢ seetzeni -4.9 5500 ± 120 Arad-4570, EBII layer 2 -7466 1985 1303-0713 GG Trochoidea seetzeni -6.5 193() ± 180 Is-267 -749 1985 1378-0499 GG Trochoidea seetzeni -3.4 13,200 ± 170 Is-274 -750 1985 1481-0633 GG Trochoideaseetzeni -3.3 10,170 ± 240 Is-29 -751 A 1946 GG Sfihincterochila zonata 0 80.7 ± 1.0 HUZ-SZ-WA, Live co!! -751B 1985 1558-0642 GG Trochoidea seetzeni -4.5 6400 ± 200 Is-318

* AK = Arnon Karnieli; GG = Glenn Goodfriend. Rehovot Radiocarbon Measurements III 109 neret samples which were coil by A Kaufman, Weizmann Inst. Mezar samples were coil by G Shaliv TAHAL, Water Planning for Israel Ltd. Dead Sea flood samples were coil by M Stiller. Lowland, Judean Mts and Judean Desert samples were coil by L Kroiteru, Weizmann Inst. Data is given in Table 2.

TABLE 2 14C in hydrologic samples Sample Nail Grid Ref Colln 8'3C 14C no. Name Type E N date %o PMC Arava Valley RT-600C Barbur 2 Well 1608-0603 5/82 0.1 -615A Zuk Tamrur Well 1746-0748 6/82 0.2 -615B Zuk Tamrur Well 1746-0748 6/82 0.2 -615C Zuk Tamrur Well 1746-0748 6/82 0.2 -615D Zuk Tamrur Well 1746-0748 6/82 2 Well 1608-0603 6/82 0.3 -615F Ein Saharonim Spring 1430-0040 7/82 0.5 -621 Beer Mashchur Well 1430-0080 10/82 0.5 -624A Neot Hakikar Spring 1852-0388 10/82 0.5 -624B Fin Amatzia Spring 1760-0343 10/82 0.4 -624C Ein Ofarim Well 1675-0275 1.1/82 ±0.2 -633B Tamar 11 Well 1800-0450 1/83 0.2 -697A Nevatim Well 1400-0700 3/84 0.3 -697B Nevatim Well 1400-0700 3/84 0.3 Galilee and Golan Heights -643A Dan Spring 2111-2946 2/83 -10.7 1.3 -643B Snir River 2151-2949 2/83 -10.0 ±0.5 -6430 Hermon Spring 2087-2922 2/83 -11.2 0.5 -661A Dan Spring 2111-2946 10/83 -10.1 1.3 -661B Hermon Spring 2087-2922 10/83 -9.5 1.2 -661C Snir River 2151-2949 10/83 -10.0 1.7 -661D Jordan River 2079-2563 10/83 -7.4 1.7 -662 Mezar 2 Well 2156-2355 12/83 -9.8 0.6 -664 Mezar 3 Well 2160-2355 12/83 -15.2 0.7 -729A Kinneret Lake 2035-2350 4/85 -5.5 2.6 -729B Kinneret Lake 2035-2350 4/85 -5.6 1.2 Dead Sea Floods -639A Zohar Flood 1849-0620 11/82 0.4 (inorganic) -639B Zohar Flood 1849-0620 11/82 1.3 (organic) Lowland, Judea Mountains, and Judea Desert -694 Ein Hemed Spring 1620-1337 4/84 -13.8 1.2 -698 Fin Hemed Spring 1620-1337 1/84 -12.8 1.1 -700 Fin Hemed Spring 1620-1337 8/84 78.7 ± 1.9 -701 Ein Sultan Spring 1923-1419 8/84 -13.1 1.9 -703A Fin Farah Spring 1787-1378 9/84 -10.5 0.8 -703B Ein Al Fauar Spring 1832-1356 9/84 -10.2 ±0.9 -703C Ein Qelt Spring 1856-1382 9/84 -11.4 1.0 -703D Fin Sultan Spring 1923-1419 9/84 -11.9 1.0 -703E Lod 4A Well 1408-1533 10/84 -11.4 -703F Lod 26 Well 1415-1591 10/84 -10.7 ±0.6 -703G Rosh Ha'ain 3 Well 1428-1681 10/84 -11.0 0.6 -703H Gimzu Well 1450-1494 10/84 -12.4 1.8 -7031 Kfar Uria 3 Well 1461-1342 10/84 -12.0 -703J Eshtaol 2A Well 1513-1316 10/84 -11.5 0.7 -703K Eshtaol 5 Well 1525-1316 10/84 -12.4 1.9 -703L Modi'in 2 Well 1542-1397 10/84 -12.4 ±0.9 110 Israel Carmi

TABLE 2 (continued) Sample Natl Grid Ref Colln no. Name Type N

-706A Agur 1 Well 0.5 -706B Agur 4 Well 0.4 -706C Hartuv 4 Well 0.7 -706D Eshtaol 7 Well 1.4 -706E Ayalon 3 Well ±0.6 -706G Ein Karem 6 Well 2.0 -706H Ein Karem 1 Well 1.1 -706I Jerusalem 6 Well 0.7 -706J Ein Karem 9 Well 0.9 -708B Jerusalem 4 Well 0.6 -7080 Azariyah Well 0.8 -709D Jericho 5 Well 0.6 -709F Jericho 1 Well -709G Jericho 2 Well 0.6 -713A Ein Farah Spring 1.7 -713B Ein Al Fauar Spring ±0.9 -713C Ein Qelt Spring 1.1

GREENHOUSE SAMPLES Samples measured in experiment to estimate incorporation of added CO2 by greenhouse-grown tomato plants. Coil 1983 by Z Enoch, Dept Agric Meteorol, Agric Research Center, Bet Dagan, Israel. Results are given in percent modern carbon (PMC) (Enoch et al, 1984).

PMC = 115.2 ± 2.7 RT-637AG. 5130 = -25.2%o Tomato plant from unenriched greenhouse.

PMC = 66.7 ± 1.5 RT-637AE. b13C = -37.0%0 Tomato plant from greenhouse enriched with tank CO2. PMC=68.6±0.5 513C RT-637AG. = -37.1 o Tomato plant from greenhouse enriched with CO2 from burned pro- pane-butane.

DEAD SEA WOOD SAMPLES

RT-625. Bottom wood PMC = 116.9 ± 2.5 Piece of wood coated with salt crystals, brought up from bottom of Dead Sea at 100m bsl (Natl Grid Ref 1890-0960) by mud dredger. Coil 1982 by Y Levy, Geol Survey Israel.

320 ± 80 RT-663A. Driftwood 513C = -12.0%0 Driftwood heavily coated with precipitates and held in place by boul- Rehovot Radiocarbon Measurements III 111 ders, exposed when Dead Sea receded to -404.5m below msl (Nat! Grid Ref 1891-1136). Coil 1983 by Z Klein, Hydrol Service Israel.

PMC = 100.0 ± 1.2 RT-683A. Dead Sea Works (30cm) b13C = -22.3%0

RT-683B. Dead Sea Works (80cm) PMC =103.8 ± 1.6 Wood exposed by channel in sediment created by overflow of brine from evaporation ponds of Dead Sea Works (Nat! Grid Ref 1905-0507). Coil 1984 by M Magaritz.

RHIZOFOSSIL SAMPLES Carbonate filling of root-grooves in Judean Desert. Coil 1983-4 by A Danin, Dept Botany, Hebrew Univ, Jerusalem (Danin, Wieder & Magaritz, in press). b13C values in parentheses were estimated by M Magaritz.

30,500 ± 900 RT-646A. Maaleh Adumim b13C = -11.5%o (Nat! Grid Ref 1700-1325) from depth 2m.

29,800 ± 800 RT-646B. Anatot b13C = -10.8%0 (Nat! Grid Ref 1767-1369).

31,400 ± 1200 RT-678A. Maaleh Adumim b13C = -11.0%o (Nat! Grid Ref 1700-1325) from depth 2m.

> 44,000 RT-678C. (b'3C = -11%0) (Nat! Grid Ref 1767-1369) from depth 2 to 3m.

LISAN SAMPLES Lisan series Samples coil near boundaries of late Pleistocene Lisan Lake, precursor of present Dead Sea. Coil 1982 by B Buchbinder, Geol Survey Israel (Buch- binder, 1981).

23,800 ± 400 b13C RT-613A. Hirbet Samra = -0. 72%o Lisan stromatolite from near Jericho (Nat! Grid Ref 1950-1460).

17,600 ± 500 RT-613B. Nahal Mor b13C = - 2.26%0 Lisan stromatolite (Nat! Grid Ref 1850-990). 112 Israel Carmi > 40,000 RT-613C. Zohar 513C = + O.8%o Laminar tufa (Nat! Grid Ref 1843-0630). > 40,000 RT-613D. Zohar li13C = + 0.9%o Postular tufa (Nat! Grid Ref 1845-0630). Nahal Amatzyah series Oolite samples consisting of calcite and aragonite. Ages are given in Druckman, Magaritz & Sneh (in press). PMC = 8.50 ± .25 RT-620A. b13C = - 4.1 Oolite (Nat! Grid Ref 1765-0353). Coil 1982 by M Magaritz. PMC=5.6±.27 RT-620B. i513C = +1. 5%o Oolite (Nat! Grid Ref 1776-0378). Coil 1982 by M Magaritz.

14,600 ± 200 RT-635. b13C= -25.6%o organic matter in clay matrix (Nat! Grid Ref 1768-0367). Coil 1982 by M Magaritz.

HULA CORE SAMPLES Dates from core coil in drilling operation at Hula Basin, N Israel (Natl 13C Grid Ref 1264-0614). Subm 1980 by M Magaritz. In age calculation, = - 25%o was assumed (Kafri, Kaufman & Magaritz, 1983). RT-610A. 20,940 ± 390 Depth 46.5m. RT-610B. 34,000 ± 1700 Depth 55.Om.

CALCITE NODULES SAMPLES Calcite nodules from loess sections, Negev, measured to date environmental changes in upper Pleistocene along desert boundary (Magaritz, in press). Netivot series Section at Netivot, Negev near Beer Sheba (Nail Grid Ref 1110-0930). Coil 1982 by M Magaritz.

7240 ± 90 RT-604C. b13C = -5. 5/oo Depth 80cm. Rehovot Radiocarbon Measurements III

13,630 ± 100 RT-604D. b13C = -3.8%o Depth lm.

27,900 ± 660 RT-619B. b13C = -11.9%o Depth 7m.

35,000 ± 1500 b13C RT-629A. = -9.3%o Depth 5.5m.

24,400 ± 450 RT-629B. b13C = -11.1°/ Depth 7.5m.

Ramat Hovav series Coil 1982 by M Magaritz, from 7 loci along Nahal Sekher, 34km SF of Netivot sec.

10,500 ± 130 RT-604A. b13C = -3.2%o Lacustrine sediment (Natl Grid Ref 131.6-0577) from depth 7m.

11,680 ± 140 b13C RT-604B. = -4. l%oo Lacustrine sediment (Natl Grid Ref 1316-0577) from depth 6m.

25,900 ± 400 RT-606A. b13C = -1.7%o Calcite nodules (Nat! Grid Ref 1308-0575) from depth 4.8m.

30,000 ± 800 b13C RT-606B. = Calcite nodules (Nail Grid Ref 1301-0577) from depth ca lm.

29,000 ± 700 b13C RT-606D. = -2.8%o Calcite nodules (Nat! Grid Ref 1284-0591).

25,900 ± 500 RT-607A. b13C = Calcite nodules (Nail Grid Ref 1286-0591).

21,900 ± 300 RT-607B. b13C = -3.3%,0 Calcite nodules (Nat! Grid Ref 1286-0591). 114 Israel Carmi

16,100 ± 270 b13C RT-607D. = -1 %o

Calcite nodules (Nat! Grid Ref 1308-0575) . > 35,000 RT-608A. 8130 = -1.0%0 Calcite nodules (Nat! Grid Ref 1264-0614). > 35,000 RT-608B. o13C = - 0.700

Calcite nodules (Nat! Grid Ref 1264-0614) .

9300 ± 100 RT-608C. Y3C = -2.4%s Calcite nodules (Nat! Grid Ref 1264-0614).

REFERENCES Avner, U, 1984, Ancient cult sites in the Negev and Sinai deserts: Tel Aviv, v 11, p 115-131. Buchbinder, B, 1981, Morphology, microfabric and origin of stromatolites of the Pleistocene precursor of the Dead Sea, Israel, in Monty, C, ed, Phanerozoic stromatolites: Berlin, Springer Verlag, p 181-195. Carmi, I, Noter, Y and Schlesinger, R, 1971, Rehovot radiocarbon measurements I: Radiocar- bon, v 13, no. 2, 412-419. p 14C Carmi, I, Stiller, M and Kaufman, A,1985, The effect of atmospheric 14C variations on the levels in the Jordan River system: Radiocarbon, v 27, no. 2B, p 305-313. Danin, A, Wieder, M and Magaritz, M, in press, Rhizofossils and root-grooves in the Judean Desert and their paleo-environmental significance: Paleogeog Paleoclimat Paleoecol. Druckman Y, Magaritz, M and Sneh, A, in press, Leaching and cementation of Late Pleisto- cene oolites from Lake Lisan, Dead Sea rift, Israel: Jour Sed Petrol. Enoch, Z H, Carmi, I, Rounick, J S and Magaritz, M, 1984, Use of carbon isotopes to estimate incorporation of added CO2 by greenhouse-grown tomato plants: Plant Physiol, v 76, p 1083-1085. Goodfriend, G A and Stipp, J J, 1983, Limestone and the problem of radiocarbon dating of land-snail shell carbonate: Geology, v 11, p 575-577. Goodfriend, G A and Hood, D G, 1983, Carbon isotope analysis of land snail shells: implications for carbon sources and radiocarbon dating: Radiocarbon, v 25, no. 3, p 810-830. Henry, D, 1982, The prehistory of southern Jordan and relationships with the Levant: Jour Field Archeol, v 9, no. 4, p 417-444. Hohfelder, R, 1981, The ancient harbor of Cesarea Maritima: Archaeology, v 84, p 56-60. Kafri, U, Kaufman, A and Magaritz, M, 1983, The rate of Pleistocene subsidence and sedimentation in the Hula basin as compared with those of other time spans in other Israeli tec- tonic regions: Earth Planetary Sci Letters, v 65, p 126-132. Levy, T E, 1983, The emergence of specialized pastoralism in the southern Levant: World Archaeol, v 15, p 15-36. Magaritz, M, in press, Environmental changes in the upper Pleistocene along the desert boundary, southern Israel: Paleogeog Paleoclimat Paleoecol. Rosen S, 1984, Kvish Harif: preliminary investigations at a late Neolithic site in the Central Negev: Paleorient, v 10, p 111-121. Wachsmann, S and Raveh, K, 1984, A concise history of Dor/Tantura: Internatl Jour Nautical Archaeol Underwater Exploration, v 13, p 223-241. Yogev, 0, 1983, A fifth millennium BCE sanctuary in the `Uvda Valley: Qadmoniot, v 16, no. 4, p 118-122 (in Hebrew). [RADIOCARBON, VOL 29, No. 1, 1987, P 115-134] RUDJER BOSKOVIC INSTITUTE RADIOCARBON MEASUREMENTS IX DUSAN SRDOC, NADA HORVATINCIC, BOGOMIL OBELIC, INES KRAJCAR BRONIC Rudjer Boskovic Institute, PO Box 1016, 41001 Zagreb, Yugoslavia and ADELA SLIEPCEVIC Faculty of Veterinary Medicine, University of Zagreb The following radiocarbon date list contains dates of samples from Plitvice Lakes measured since our previous list dealing with tufa from this region (Srdoc et al, 1982). Tufa measurements from the Knin area in S Croatia and some localities in Bosnia are also listed. Age calculations are based on the Libby half-life (5570 ± 30) yr and reported in years before 1950. Reported ages are based on the initial activity of 85% except for lake sediments where calculations of initial activity have been performed. The modern standard is 0.95 of the NBS oxalic acid activity. Sample pretreatment and counting technique are essentially the same as described in R, 1971, v 13, p 135-140, supplemented by new techniques for groundwater processing (R, 1979, v 21, p 131-137). Statistical processing of data has been computerized (Obelic & Planinic, 1977; Obelic, 1980). The errors quoted correspond to 1 r variation of sample net counting rate and do not include the uncertainty in 14C half-life.

ACKNOWLEDGMENTS This research was made possible by the cooperation and financial support of Plitvice Lakes National Park Authorities and the Community for Scientific Research of Croatia. The following experts took part in various stages of research, field and laboratory work, discussion, and sample collection: M A Geyh and S Merkt, Niedersachsisches Landesamt fur Bodenforschung, Hannover, H Muller, Bundesanstalt fur Geowissenschaften and Rohstoffe, Hannover, and H K S13C Wong, S Kempe and K Emeis, University of Hamburg. measurements were made by J Pezdic, Jozef Stefan Inst, Ljubljana. E Hernaus helped in sample preparation and methane synthesis, A Turkovic in data processing, and P Hojski in technical maintenance.

GEOLOGIC SAMPLES Tufa from Plitvice Lakes area Tufa samples from outcroppings scattered in the Plitvice Lakes area (44° 50' N, 15° 35' E), central Croatia. Sixteen lakes are separated by tufa barriers; 3 major and several minor springs feed lakes and the Korana River. Measurements of 14C activity of dissolved inorganic carbon in water samples as well as that of recent tufa deposited on the surface of polyamide mats and aquatic plants in the Plitvice National Park area performed since 1981 showed a systematic increase of 14C activity from karst springs to the 115 116 Dusan Srdoc et al Korana R mouth. The increase has been attributed to the influx of atmospheric CO2 and to the contribution of terrestrial plants, through detrital decay and root respiration, and a model based on the described process was developed (Srdoc et al, 1986b). Dates help to determine periods of inten- sive tufa formation in the area. Samples coil 1981-1985 by D Srdoc. Crna Rijeka series Crna Rijeka brook feeds Plitvice Lakes joined by Bijela Rijeka brook at Plitvicki Ljeskovac, forming rivulet Matica which discharges into the uppermost Lake Prosce. Crna Rijeka brook flows over solid tufa terraces which belong to earlier phase of tufa deposition. Mud covering riverbed contains large fraction of dolomite (Popovic, Srdoc & Grgic, 1986), thus, its 14C activity does not reflect its age.

69.1 ± 0.6% modern Z-702. Crna Rijeka No. l 1630 + 100 Recent tufa near waterfall, right bank.

47.6 ± 0.5% modern Z-703. Crna Rijeka No. 2 4630 ± 110 Surface layer of tufa upstream from bridge near waterfall.

53.2 ± 0.5% modern Z-751. Crna Rijeka No. 3 3720 ± 100 Tufa, right bank upstream from bridge near waterfall.

48.3 ± 0.5% modern Z-810. Crna Rijeka No. 4 4500 ± 120 Tufa, right bank upstream from waterfall.

8.3 ± 0.3% modern b13C Z-1064. Crna Rijeka No. 5 = + 1.9%o Sandy deposit covering creek bottom.

Bijela Rijeka series Large deposits of tufa above right bank of Bijela Rijeka brook. Tufa rocks, 10 to 12m high above ground level belong to preglacial period (Riss/ Wurm) as determined by 230Th/234U (Srdoc et al, 1986a), whereas river terrace is much younger.

0.0 ± 0.3% modern Z-1050. Bijela Rijeka No. l >37,000 Compact tufa, right bank of Bijela Rijeka, village Plitvicki Ljeskovac.

0.4 ± 0.4% modern Z-1051. Bijela Rijeka No. + 3800 2 34, 700 -3200 Same block as Z-1050. Rudjer Boskovic Institute Radiocarbon Measurements IX 117

73.6 ± 0.5% modern Z-1115. Bijela Rijeka No. 3 1120 ± 140 Coarse calcareous grains, Bijela Rijeka terrace.

3.9 ± 0.3% modern Z-1116. Bijela Rijeka No. 4 24,600 ± 1300 Compact tufa, Plitvicki Ljeskovac.

2.3 ± 0.2% modern Z-1117. Bijela Rijeka No. 5 28,800 ± 2200 Same block as Z-1 H 6. Kavga brook series Right tributary of Matica R near Plitvicki Ljeskovac.

63.3 ± 0.7% modern Z-1057. Kavga No. l 2320 ± 150 Tufa block above brook.

60.7 ± 0.6% modern Z-1058. Kavga No. 2 2700 ± 150 Tufa from terrace above brook. Pecina series Spring Pecina near Plitvicki Ljeskovac, left tributary of Bijela Rijeka brook.

77.5 ± 0.7% modern Z-1052. Pecina No. l Modern Recent tufa deposited around spring Pecina.

69.6 ± 0.5% modern Z-1055. Pecina No. 2 1570 ± 140 Hard, porous tufa above spring Pecina.

Plitvicki Ljeskovac series Confluence of Crna Rijeka and Bijela Rijeka brooks is in tufa-covered valley near Plitvicki Ljeskovac. Outcroppings of preglacial tufa and thick deposits of recent tufa are very abundant in valley.

65.3 ± 0.6% modern Z-705. Ex Bio-station 3200 ± 100 Outcroppings of tufa in marshy field, presently flooded area (cf Z-700; R,1982, v 24, p 356). 71.5 ± 0.7% modern Z-1059. Confluence site No.1 1350 ± 140 Recent, porous, coarse-grained tufa covered with moss. 118 Dusan Srdoc et al

1.3 ± 0.4% modern Z-1069. Confluence site No. 2 + 4000 33 000 , - 3500 Compact, inner layer of tufa tube.

0.4% modern + 3300 Z-1070. Confluence site No . 3 31 000 , - 2700 Porous, outer layer of tufa tube. Comment: tubular forms of tufa are frequent, as result of encrustation of wood branches and trunks.

45.8 ± 0.5% modern Z-1135. Matica 4900 ± 160 Fine-grained calcareous deposit, mixed with organic detritus, Matica rivulet mouth. Material transported by water and deposited at river mouth. Matica R flows through preglacial and Holocene tufa deposits. Comment: no formation of recent tufa concretions has been observed in Matica R.

77.5 ± 0.6% modern Z-1031. Prosce, Spiljski vrt Modern Recent tufa under moss, Cave garden. Measurement of initial 14C activity of recently deposited tufa.

Lake Ciginovac series

73.3 ± 0.6% modern Z-1029. Ciginovac No. l Modern Recent tufa under growing moss (Cratoneurum commutatum). Comment: 14C determination of initial activity of sediment in Upper Lakes (cf Z-817: R, 1982, v 24, p 361).

77.9 ± 0.6% modern Z-1416. Ciginovac No. 2 Modern Freshly deposited tufa on surface of artificial substratum (polyamide mat) immersed in water from May to Sept 1984. Microlocation: waterfall connecting lakes Prosce (upper) and Ciginovac (lower). Comment: determination of initial 14C activity.

Veliki Jovinovac series Lake Veliki Jovinovac is in middle of Upper Lakes, Plitvice Natl Park. Entire area covered with thick tufa deposits.

74.4 ± 0.6% modern Z-1001. Veliki Jovinovac No. l Modern Dry, porous tufa, mossy shape, green algae on surface. Rudjer Boskovic Institute Radiocarbon Measurements IX 119

75.4 ± 0.6% modern Z-1002. Veliki jovinovac No. 2 Modern Porous tufa from cave in dry tufa barrier.

74.9 ± 0.5% modern Z-1006. Lake Vir Modern Hard, porous, mossy structured tufa above surface of Lake Vir.

Lake Galovac series Samples of tufa deposits lying above present level of Lake Galovac.

71.2 ± 0.6% modern Z-1003. Galovac No.1 1390± 110 Hard, dry tufa near pathway, right side.

73.9 ± 0.6% modern Z-1004. Galovac No. 2 1080± 110 Porous tufa below pathway.

69.2 ± 0.4% modern Z-1005. Galovac No. 3 1610± 110 Hard tufa barrier.

Lake Gradinsko series Thick deposits of hard and powdered tufa above surface of Lake Gra- dinsko.

51.8 ± 0.4% modern 3950 ± 90 Z-835. Gradinsko No. l b13C = -8.4%o Powdered microcrystalline calcareous deposit 2m above present level of lake (cf Z-832 to Z-834: R, 1984, v 24, p 366).

47.0 ± 0.4% modern Z-836. Gradinsko No. 2 4700 ± 120 Compact tufa 2m above lake surface, near Z-835.

53.8 ± 0.5% modern Z-837. Gradinsko No. 3 3600 ± 120 Porous tufa, mossy structured (Cratoneurum commutatum) ca 2m above lake surface.

50.0 ± 0.4% modern Z-1056. Gradinsko No. 4 4220 ± 150 Tufa 2m above present level of lake. 120 Dusan Srdoc et al Gradina series Isolated hill on peninsula above Lakes Gradinsko and Kozjak, with remnants of prehistoric and medieval ramparts. Hill encircled with deposits of Holocene tufa with outcroppings of preglacial tufa on top. Lake Gra- dinsko is in Upper Triassic well bedded dolomite containing >90% MgCa(C03)2. Stromatolithic fms are very frequent. Systematic measurements of preglacial tufa using fractional dissolution revealed contamination with more recent calcareous material (see Table 1). 230Th/234U dating on samples of preglacial tufa gave age of ca 120,000 yr (Riss/Wurm interglacial) (Srdoc et al, 1986a).

61.2 ± 0.5% modern Z-830. Gradina No. l 2600 ± 110 Powdered, microcrystalline calcareous tufa.

63.4 ± 0.4% modern Z-996. Gradina No. 2 2310 ± 110 Tufa blocks used for building ramparts.

62.4 ± 0.6% modern Z-997. Gradina No. 3 2450 ± 120 Tufa boulders used to build fortification. Comment: only Holocene tufa was used to construct fortifications at Gradina.

2.8 ± 0.3% modern Z-1208. Gradina, hilltop No. l 27,200 ± 1500 Porous tufa, Block C. Upper surface under atmospheric influence. Samples used for testing contamination with recent carbonate.

1.5 ± 0.3% modern Z-1210. Gradina, hilltop No. 2 3200 30,900 + - 2600 Compact tufa, Block S, covered by porous tufa. General Comment: 14C measurements show that each subsequent soluble

TABLE 1 Test of fractionation of sample Z-1211 (Porous tufa, Block S)

Sample Grain size Fraction BP no. (mm) no. (%) modern Z-1268 1-5 I (30) 0.4 400 -1269 1-5 II (30) 0.3 800 -1270 1-5 III (30) 0.3 1100

-1271 <1 I (50) 0.4 400 -1272 <1 II (50) 0.2 1200

-1273 1 I (50) 0.4 1 II (50) 0.3 2400 Rudjer Boskovic Institute Radiocarbon Measurements IX 121 fraction obtained from porous tufa gave successively older age indicating that surface of sample was contaminated by younger carbonates. No consistent effect of grain size on 14C age is observed.

Burget series Freshly deposited tufa on surface of artificial substratum (polyamide mat) immersed in streamwater. Waterfall connecting Lakes Burget (upper) and Kozjak (lower). Comment: determination of initial 14C activity.

85.4 ± 0.7% modern Z-980. Burget No. l Modern Substratum immersed in water from June to Oct 1981.

83.8 ± 0.6% modern Z-1011. Burget No. 2 Modern Substratum immersed in water from Oct 1981 to Apr 1982.

Rjecica series Rjecica brook is a major tributary to Plitvice Lakes. Tufa deposition starts at approx half way between its springs and Lake Kozjak. Upper part of Rjecica bed is cut in old tufa.

84.0 ± 0.6% modern Z-1020. Rjecica No. l Modern Hard, porous, recent tufa from last barrier before brook discharges into Lake Kozjak.

46.3 ± 0.6% modern Z-1038. Rjecica No. 2 4830 ± 170 Hard tufa, riverbed.

56.3 ± 0.4% modern Z-1060. Rjecica No. 3 3260 ± 150 Tufa in form of small pebbles covering riverbed.

70.9 ± 0.6% modern Z-1061. Rjecica No. 4 1420 ± 140 Tufa from river terrace, coarse grains several mm in diameter.

89.0 ± 0.7% modern Z-1068. Rjecica No. 5 Modern Freshly deposited tufa on surface of artificial substratum (polyamide mat) immersed in stream water from July to Oct 1982. General Comment: two distinct 14C activities were found in Rjecica brook: recent activity of fresh tufa is from 84 to 89% modern, whereas tufa riverbed belongs to older, Holocene deposits (Popovic, Srdoc & Grgic, 1986). 122 Dusan Srdoc et al Lake Kozjak series Samples of hard, porous tufa and other calcareous material presently above surface of Lake Kozjak. Tufa outcroppings around lake mark position of dried-up tributaries, and i4C dating helps to reconstruct history of lake.

61.8 ± 0.6% modern Z-676. Kozjak No. l 2500 ± 100 Hard, porous tufa, resembles petrified moss (Cratoneurum commutatum), 7 to 8m above lake surface; sample taken from huge block of tufa emerging from lake.

69.9 ± 0.5% modern Z-844. Kozjak No. 2 1530 ± 100 Hard, porous tufa from NW shore.

53.0 ± 0.4% modern Z-1037. Kozjak No. 3 3750 ± 150 Powdered, microcrystalline calcareous sediment, ca 8 to 1Om above lake surface.

75.2 ± 0.6% modern Z-1082. Kozjak No. 4 950 ± 130 Compact tufa overgrown by moss, left approach to waterfalls connecting Lake Kozjak (upper) and Lake Milanovac (lower).

59.8 ± 0.6% modern Z-1114. Kozjak No. 5 2800 ± 150 Dripstone from cave, 4 to 5m above lake surface. General Comment: all samples belong to warm and humid periods in Holo- cene.

Kozjak barrier series Core samples taken by drilling at barrier connecting Lakes Kozjak and Milanovac. Coll 1983 by S Merkt.

64.1 ± 0.5% modern Z-1463. Kozjak barrier No. l 2230 ± 150 Depth 0.5 to 1.0m.

59.9 ± 0.4% modern Z-1464. Kozjak barrier No. 2 2770 ± 140 Depth 2.2 to 2.6m.

57.2 ± 0.5% modern Z-1457. Kozjak barrier No. 3 2840 ± 120 Depth 3.0 to 3.2m. Rudjer Boskovic Institute Radiocarbon Measurements IX 123 Plitvica series Plitvica brook is major tributary to Korana R, discharging into Korana at "Sastavci" (confluence). Tufa deposition starts at Crkvine, ca 1 km downstream from Plitvica karst spring.

83.0 ± 0.7% modern Z-1066. Plitvica No.1 Modern Freshly deposited tufa on artificial substratum (polyamide mat) immersed in streamwater from May to Oct 1982.

84.6 ± 0.6% modern Z-1067. Plitvica No. 2 Modern Same as Z-1066, except microlocation.

69.9 ± 0.6% modern Z-1118. Plitvica No. 3 1530 ± 140 Hard, porous tufa from riverbed near Hajdukovic Mill.

66.8 ± 0.4% modern Z-1083. Plitvica waterfall No.1 1890 ± 140 Hard, porous tufa from river terrace above 76m high waterfall. Com- ment: recent tufa activity of Rjecica brook (Z-1068, Z-1020, above) and Plit- vica brook are similar.

87.9 ± 0.7% modern 1-1012. Plitvica waterfall No. 2 Modern Freshly deposited tufa on surface of artificial substratum (polyamide mat) immersed in water from Oct 1981 to Apr 1982 under Plitvica brook waterfall. Comment: determination of initial 14C activity.

88.2 ± 0.7% modern Z-1276. Novakovica brod Modern Recent tufa under growing moss (Cratoneurum commutatum). Comment: "C activity increases along river course (Srdoc et al, 1986b).

Hajdukovic pit series Pit, 2 to 3m deep in flat terrace ("polje") near Hajdukovic Mill. Layer of charred decayed leaves entrapped between tufa deposits.

66.3 ± 0.6% modern Z-1119. Hajdukovic pit No.1 3260 ± 130 Charred wood in Hajdukovic pit surrounded by tufa layer.

66.2 ± 0.6% modern Z-1204. Hajdukovic pit No. 2 3300 ± 130 Charred leaves in Hajdukovic pit surrounded by tufa layer. 124 Dusan Srdoc et al

51.5 ± 0.5% modern Z-1205. Hajdukovic pit No. 3 4000 ± 130 Tufa layer above charred leaves.

50.9 ± 0.4% modern Z-1206. Hajdukovic pit No. 4 4100 ± 130 Tufa layer below charred leaves. General Comment: ratio of activity of tufa and that of organic material gives initial activity of tufa at this location equal to 77.3% modern. Smolcic flat series Outcroppings of tufa in small flat above Smolcica pecina cave. Very hard, porous tufa dated by 14C and 230Th/234U method. 14C dating gave inconsistent ages, depending on contamination of samples with more recent carbonates (see Table 2) whereas 230Th/234U dating pointed at warm and humid interglacials (Riss/Wurm, Mindel/Riss) (Srdoc et al, 1986a).

1.5 ± 0.3% modern Z-1213. Smolcic flat No. l 32,000 ± 2900 Hard, compact tufa, Block N.

11.9 ± 0.4% modern Z-1214. Smolcic flat No. 2 15,800 ± 400 Hard, porous tufa. Humus and rootlets in pores and crevices removed mechanically and by washing. Smolcic cave series Dripstones and tufa from Smolcic cave, ca 60m above present level of Korana R. Measurement of tufa and dripstone samples from same site could reveal any difference in degree of contamination with more recent calcareous material. Dripstones are considered less susceptible to contamination, as opposed to porous tufa. Comment: no conclusive remarks can be drawn from this series of measurement.

36.4 ± 0.5% modern 6800 ± 160 Z-932. Smolcic cave No. l b13C = -7.8%o Core of tufa block, entrance to cave.

TABLE 2 Test of fractionation of sample Z-1213 (Compact tufa, Block N)

Sample Grain size Fraction BP no. (mm) no. (%) modern Z-1500 1-5 I (30) 0.3 1100 -1501 1-5 II (30) 0.3 1600 -1502 1-5 III (30) 0.3 >37,000 Rudjer Boskovic Institute Radiocarbon Measurements IX 125 0.2 ± 0.3% modern Z-1008. Smolcic cave No. 2. >37,000 Limestone bedrock.

1.6 ± 0.3% modern Z-1007. Smolcic cave No. 3. 31,600 ± 2800 Core of homogeneous tufa block.

3.4 ± 0.3% modern Z-1144. Smolcic cave No. 4. 25,700 ± 1600 Inner part of tufa covered by flowstone.

5.0 ± 0.3% modern Z-1145. Smolcic cave No. 5 22,600 ± 1100 Outer part of tufa covered by flowstone.

6.0 ± 0.3% modern Z-1146. Smolcic cave No. 6 21,200 ± 900 Partly crystallized tufa.

Siroka Luka series Korana village is above partly cultivated large river terrace Siroka Luka. Terrace consists of calcareous material, mostly tufa, deposited by Korana R and covered with 15 to 20cm layer of soil. Dates help to reconstruct periods of development of terrace.

54.0 ± 0.5% modern Z-1127. Siroka Luka No. l 3610 ± 150 Fine-grained tufa, terrace.

59.2 ± 0.5% modern Z-1132. Siroka Luka No. 2 2870 ± 150 Tufa from cultivated field.

72.5 ± 0.6% modern Z-1133. Siroka Luka No. 3 1230 ± 140 Dry tufa barrier, village Korana, 1 to 2m above river terrace.

83.9 ± 0.6% modern Z-1134. Siroka Luka No. 4 Modern Recent tufa chips, river terrace Siroka Luka under constant influence of Korana R water (DIC activity 90% modern).

Sartuk series Sartuk brook flows in same area of Plitvice National Park where other streams form tufa barriers. No typical tufa has been found in Sartuk. Cal- careous deposits contain much less 14C, consisting of mixtures of weathered 126 Dusan Srdoc et al limestone and dolomite rocks which surround Sartuk and some biogeni- 14C cally deposited calcite. No deposit ages can be deduced from measurements. Z-1026. Sartuk No. l 4.4 ± 0.3% modern Mixture of soil and calcareous deposit.

Z-1027. Sartuk No. 2. 49.9 ± 0.6% modern Calcareous deposit under growing moss.

Z-1028. Sartuk No. 3 11.2 ± 0.3% modern Sandy calcareous deposit. Dolomite with calcite content 5 to 10%, quartz 2 to 5% (Popovic, Srdoc & Grgic,1986). Z-1339. Sartuk No. 4. 43.3 ± 0.5% modern Calcareous deposit, riverbed.

Korana River series Measurement of 14C activity of calcareous deposits, mostly tufa, along Korana R. For details see Srdoc et al (1986b).

93.6 ± 0.5% modern Z-1019. Korana River No. l Modern Recent tufa under growing moss (Cratoneurum commutatum).

90.8 ± 0.6% modern Z-1063. Korana River No. 2 Modern Tufa under moss on waterfall, Tusilovic near Karlovac (44° 20' N, 15° 37'E).

89.5 ± 0.6% modern Z-1065. Korana River No. 3 Modern Freshly deposited tufa on artificial substratum (polyamide mat) immersed in water from Apr to Oct 1982, 0.5km downstream from confluence ("Sastavci"). Comment: determination of initial 14C activity.

Tufa used for building at Plitvice Lakes region

75.5 ± 0.6% modern Z-807. Hajdukovic Mill 900 ± 90 Hard, porous tufa used for construction of mill.

60.8 ± 0.6% modern 2670 ± 100 3130= Z-808. Crkvina -84 o Hard, porous tufa block from foundation of medieval church; origin and denomination of church unknown. Rudjer Boskovic Institute Radiocarbon Measurements IX 127 63.8 ± 0.5% modern Z-916. Old power station 2270 ± 110 Hard, porous tufa from building housing small power station, now out of operation, Lake Burget above Kozjak.

72.4 ± 0.6% modern Z-1018. Old power station 1260 ± 100 Same as Z-916. General Comment: both preglacial (Z-921: R, 1982, v 24, p 355) and Holo- cene tufas were hard enough to be used for construction of buildings. However, it should be pointed out that age of tufa used in construction does not indicate date of erection of building, even though it can be used as terminus post quem non. Pevalek collection series Increased activity of groundwater, tufa, and aquatic plants due to increased activity of atmospheric CO2 in past decades caused by nuclear weapon tests can be obtained by comparison of activity of recent material with that from pre-bomb era, providing that age of latter is known. Samples coll 1919 by late academician, I Pevalek (cf Z-847, -848, -853, -856, -857, -907, -908: R, 1982, v 24, p 365-366). 72.2 ± 0.5% modern Z-1306. Pevalek colln No. l 1280 ± 110 Tufa around wooden branch.

97.3 ± 0.6% modern Z-1307. Pevalek colln No. 2 185 ± 100 Wooden branch encrusted with tufa. General Comment: ratio of activity of tufa and wood gives initial activity of tufa at this location equal to 74.2% modern. Present activity of tufa is 80% modern at this location.

Plitvice Lakes Sediments Lake Kozjak sediment core In autumn 1983 several sediment cores were retrieved from Lakes Kozjak and Prosce, Plitvice Nail Park area (Srdoc et al, 1986c) (Table 3). Lake Kozjak core was retrieved from 24m water depth. First 2m of core gave good stratification with sedimentation rate 0.85mm/a. Next 5m sec gave scattered 14C data in agreement with seismic records. This sec appears to have undergone mass transport and redeposition. Further 5m sec showed good stratification with sedimentation rate 1.1 mm/a. Initial 14C activity of 75% modern was determined by measuring activity of top of sediment layer (cf Z-840 and -841: R, 1982, v 24, p 367) and pre-bomb test tufa (Z-1082: R, 1982, v 24, p 363). Piece of wood (Abies sp) was found in sediment core. Surrounding sediments were carefully collected. Ratio of sediment vs wood activity gave initial 14C activity of 74.4% modern. 128 Dugan Srdoc et al

TABLE 3 Lake Kozjak sediment core

Sample Depth 1 no. (m) % modern BP PDB) Z-1301 0.15 74.8 -1230 0.17 74.9 -1372 0.00-0.20 73.3 150 -1302 0.72 66.3 115 -1303 1.07 64.0 115 -1304 1.45 60.0 120 -1305 1.70 55.6 125 -1232 2.00 55.3 130 -1347 2.40-2.60 52.2 130 -1233 3.12 49.8 145 -1234 3.45 51.5 130 -1235 3.84 52.7 130 -1236 4.30 53.7 130 -1237 4.68 49.3 130 -1373 4.96-5.05 49.8 140 -1374 5.05-5.20 50.5 140 -1375 5.20-5.35 58.4 120 -1376 5.35-5.50 57.4 130 -1172 5.60 54.7 130 -1171 5.70 54.4 150 -1173 5.80 56.9 140 -1240 6.13 49.4 140 -1241 6.50 46.1 140 -1465 6.60-6.80 57.9 120 -1242 6.86 46.1 120 -1243 7.24 58.6 140 -1246 8.06 53.8 145 -1471 8.70-8.90 48.6 140 -1250 9.20 51.0 150 -1253 10.20 47.4 160 -1472 10.20-10.35 43.3 140 -1393 10.55-10.75 41.6 150 -1369 10.90-11.08 38.2 160 -1370 11.08-11.25 38.6 160 -1392 11.75-11.95 35.4 160 -1432 11.95-12.15 39.2 150 -1348 12.15-12.35 43.4 150 Wood (Abies sp) -1168 5.64 120

Lake Kozjak bottom series Sediment from bottom of Lake Kozjak. Coil 1983 by S Merkt, Nied- ersachsisches Landesamt f Bodenforschung, Hannover.

Z-1122. Kozjak No.1 69.0 ± 0.6% modern Bottom of lake. Water depth 43m.

Z-1123. Kozjak No. 2 73.5 ± 0.6% modern Bottom of lake. Water depth 23m. Lake Prosce sediment core Lake Prove core was retrieved from 17.2m water depth and reached clayey residual overlaying bedrock (Table 4). 14C dating of lake sediment revealed uniform sedimentation rate of 1.4mm/a. Pollen analyses of core Rudjer Boskovic Institute Radiocarbon Measurements IX 129 sec revealed major settlement phases in Plitvice Nat! Park area during last 6000 yr (Muller & Obelic, 1986). Piece of wood (Abies sp) was found in sediment core and dated together with surrounding sediment which enabled determination of 14C initial activity of sediment equal to 72% modern. Tufa Deposits in Kninsko Polje Tufa deposits in Kninsko polje (44° 02' N, 16° 11' E), S Croatia consist of Holocene and preglacial deposits. Waters of intermittent Krcic creek and perrenial Krka R are rich in carbonates forming thick tufa and lacustrine sediments along their ancient and recent courses in Kninsko polje val- (cf Z-1189 to -1194; R, 1984, v ley 26, p 455). Tufa samples with apparent C age > ca 20,000 y r are much older, as shown by 230Th/234U analyses (Srdoc et al, 1986a). Krcic series Samples toll 1984 by D Srdoc and B Obelic.

88.8 ± 1.1% modern Z-1323. Krcic No. 1 Modern Recent tufa under moss, Krcic brook near village Krcic. Comment: activity of recent tufa similar to activities measured in Plitvice Nail Park.

TABLE 4 Lake Prosce sediment core Sample Depth 13c no. (m) % modern BP PDB) Z-1441 0.00-0.40 66.2 100 -1398 0.80-1.00 67.6 115 -1658 1.20-1.40 62.6 100 -1399 1.60-1.80 62.4 120 -1659 2.00-2.20 60.7 100 -1661 2.40-2.60 59.0 100 -1407 2.80-3.00 55.7 120 -1662 3.20-3.40 53.8 100 -1422 3.60-3.80 53.2 125 -1663 3.80-4.00 53.7 100 -1424 4.20-4.40 52.0 130 -1664 4.90-5.00 48.8 110 -1430 5.20-5.40 47.5 130 -1665 5.60-5.80 45.1 120 -1431 6.20-6.40 43.8 140 -1666 6.80-7.00 41.9 140 -1438 7.20-7.40 43.8 145 -1667 7.60-7.80 38.7 130 -1436 8.20-8.40 38.4 150 -1668 8.90-9.00 36.6 130 -1437 9.20-9.40 34.5 160 -1669 9.80-10.00 33.7 140 -1433 10.20-10.40 34.5 170 -1670 10.60-10.75 30.3 150 -1671 11.00-11.20 29.6 160 -1435 11.50-11.70 27.8 190 Wood (Abies sp) -1395 11.35-11.50 160 130 Duian Srdoc et al 0.8 ± 0.3% modern Z-1325. Krcic No. 2 > 37,000 Powdered tufa from deposit ca 12m above brook level.

30.7 ± 0.5% modern Z-1321. Krcic No. 3 8000 ± 170 Sample from tufa block lying on brook terrace; first appearance of tufa downstream from Krcic karst spring.

Topoljski buk barrier series Krka R spring in cave under Topolj ski buk barrier. Samples colt and subm 1985 by S Bozicevic, Geol Inst Zagreb.

47.9 ± 0.5% modern Z-1562. Topoljski buk No.1 4570 ± 110 Outer layer of tufa tube, bottom of well, passage under Krcic waterfall. Comment: age similar to Z-1193 (R, 1984, v 26, p 455).

48.4 ± 0.5% modern Z-1564. Topoljski buk No. 2 4480 ± 120 Outer layer of tufa tube, same as Z-1562.

70.6 ± 0.6% modern Z-1563. Topoljski buk No. 3 2770 ± 110 Mud above tufa tube (Z-1562 and -1564) in passage at Krcic waterfall. Comment (SB): expected same age for tufa and mud.

Topolje quarry series Tufa quarry, edge of Kninsko polje. Coll 1984 by D Srdoc and B Obelic.

1.1 ± 0.3% modern 4400 Z-1311 . To p olje No . l 3600 Tufa from lowest layer, Topolje quarry.

1.9 ± 0.3% modern 2400 Z-1313 . Top olje No . 2 2000 Tufa from uppermost layer, Topolje quarry.

4.6 ± 0.3% modern Z-1316. Knin 23,400 ± 900 Powdered tufa from trench 2.5m deep excavated during building of new hospital in Knin. Coll 1984 by D Srdoc and B Obelic. Rudjer Boskovic Institute Radiocarbon Measurements IX 131 Tufa Samples From Bosnia 14C Systematic dating of tufa in karst regions of Yugoslavia (cf Z-1046 to -1049, -1164 to -1167: R, 1984, v 26, p 454-455). 56.3 ± 0.5% modern Z-1351. Jajce 3300 ± 130 Tufa from thick deposit above Pliva R, Jajce (44° 20' N,17° 17' E) Cen- tral Bosnia. Coil 1984 by D Srdoc and B Obelic.

0.3 ± 0.3% modern Z-1354. Janj > 37,000 Tufa from thick deposits above Janj R, Mujdzici near Jajce (44° 14' N, 170 07' E), Central Bosnia. Coil 1984 by D Srdoc and B Obelic.

Z-1552. Banja, Fojnica 9.9 ± 0.4% modern Tufa from thermal spring in Banja near Fojnica (43° 58' N, 170 54' E), alt 670m, Bosnia. Coil and subm 1985 by I Krusic, Geoinzenjering Co, Sarajevo. Kiseljak Slatina series Tufa deposited from thermal springs Kiseljak Slatina near Banja Luka (44° 49' N, 170 18' E), NW Bosnia. Coil 1983 and subm 1985 by D Hrustan- pasic, Geoinzenjering Co, Sarajevo.

Z-1459. Kiseljak Slatina No. l 13.2 ± 0.4% modern Recent porous tufa from hot mineral spring mixed with decayed organic detritus.

1.9 ± 0.3% modern Z-1458. Kiseljak Slatina No. 2 30,600 ± 2500 Porous dry tufa, above water level, partly covered with humus and moss.

Z-1414. Sockovac 5.8 ± 0.3% modern Tufa from borehole OS-2, Sockovac near Gracanica (44° 39' N, 18° 18' E), N Bosnia. Coil and subm 1985 by N Miosic, Geoinzenjering Co, Sara- jevo. General Comment: percent of modern carbon in recent samples indicates ratio of biogenic to inorganic carbon in freshly deposited tufa around hot springs. No age of deposits can be deduced from these data because of insufficient knowledge of aquatic chemistry and isotopic composition of hot spring water.

HYDROGEOLOGIC SAMPLES Plitvice Lakes National Park Surface water 14C activity was measured in 1983 and 1984 to determine 14C distribution patterns along the river course. For detailed discus- 132 Dusan Srdoc et al lion, see Srdoc et al (1986b). Samples were coil by Rudjer Boskovic Inst staff. Crna Rijeka series (f Z-692: R,1982, v 26, p 369. 69.2 ± 0.6% modern b13C Z-1337. Crna Rijeka No. l = -12.6%o River water, coil May 1984.

90.6 ± 0.6% modern 513C Z-1379. Crna Rijeka No. 2 = -13.2%0 Karst spring, coil Sept 1984. 0.5% modern 13C Z-1425. Crna Rijeka No. 3 = -13.2%0 Spring water, coil Dec 1984. 14G General Comment: Crna Rijeka spring water shows large variations of activity of dissolved inorganic carbon (DIG). Mean residence time is 2 yr, calculated by means of exponential model (Krajcar Bronic et a1,1986). Bijela Rijeka series Z-1024. Bijela Rijeka No. l 85.9 ± 0.9% modern Spring water, coll July 1982.

Z-1159. Bijela Rijeka No. 2 83.0 ± 0.9% modern Spring water, coil Oct 1983.

84.7 ± 0.6% modern b13C Z-1281. Bijela Rijeka No. 3. = -12.2 o Spring water, coil Apr 1984.

81.3 ± 0.6% modern Z-1434. Bijela Rijeka No. 4 5130 = -12.6%o River water, coil Dec 1984. General Comment: mean residence time of 4 yr was calculated by using exponential model. 14G activity and o13G of DIG in spring water are fairly constant throughout year. Matica River series 85.3 ± 0.6% modern Z-1280. Matica, mouth No. l 513C = -11.8%0 River water colt Apr 1984, high waters, snow melting. Z-1336. Matica, mouth No. 2 76.9 ± 0.6% modern River water coil May 1984. Rudjer Roskovic Institute Radiocarbon Measurements IX 133

90.1 ± 0.7% modern b'3C Z-1381. Matica, mouth No. 3 = -12.1 %o River water coil Sept 1984.

Plitvica spring series

qZ-708: R, 1982, v 24, p 369.

Z-1025. Plitvica spring No.1 81.9 ± 0.6% modern Spring water, coil July 1982.

Z-1160. Plitvica spring No. 2 83.2 ± 1.1% modern Spring water, coil Oct 1983. General Comment: mean residence time of 3 yr was calculated by means of exponential model.

Korana River series

Z-1279. Korana No.1 95.2 ± 0.7% modern River water coil Apr 1984, Slunj (45° 07' N, 15° 36' E) Croatia.

99.0 ± 0.8% modern b'3C Z-1278. Korana No. 2 = -11.2%o River water toll Apr 1984, village Tusilovic near Karlovac (45° 23' N, 15° 37' E) Croatia.

REFERENCES Krajcar-Bronic, I, Horvatincic, N, Srdoc, D and Obelic, B, 1986, On the initial 14C activity in karst aquifers with short mean residence time in Stuiver, M and Kra, RS, eds, Internatl 14C conf, 12th, Proc: Radiocarbon, v 28, no. 2A, p 436-440. Muller, H and Obelic, B, 1986, Pollen distribution in radiocarbon dated sediment cores from a Plitvice Lake as the indicator of human settlements, in Symposium on sedimentology, 5th, Proc: Brioni/Yugoslavia, June 2-5,1986, p 152-155. Obelic, B,1980, Computer analysis and interpretation of radiocarbon data: Fizika, v 12, suppl 2, p 139-161. Obelic, B and Planinic, J, 1977,.Computer processing of radiocarbon and tritium data, in Povinec, P and Usacev, S, eds, Internatl conf on low-radioactivity measurements and applications, Proc: The High Tatras, Slovenske pedagogicke nakladatelstvo, Bratislava, 117-120. p Popovic, S, Srdoc, l) and Grgic, S, 1986, Investigation of lake sediments and tufa in National Park Plitvice by x-ray diffraction analyses of lake sediment cores, in Symposium on sedimentology, 5th, Proc: Brioni/Yugoslavia, June 2-5, 1986, p 140-142. Srdoc, D, Breyer, B and Sliepcevic, A, 1971, Rudjer Boskovic Institute radiocarbon measurements I: Radiocarbon, v 13, no. 1, p 135-140. Srdoc, D, Horvatincic, N, Obelic, B, Krajcar Bronic, I and O'Malley, P, 1986a, The effects of contamination of calcareous sediments on their radiocarbon 14C age, in Stuiver, M and Kra, R S, eds, Internatl conf, 12th, Proc: Radiocarbon, v 28, no. 2A, p510-514. Srdoc, D, Horvatincic, N, Obelic, B and Sliepcevic, A, 1982, Rudjer Boskovic Institute radiocarbon measurements VII: Radiocarbon, v 24, no. 3, p 352-371. Srdoc, I), Krajcar Bronic, I, Horvatincii, N and Obelic, B, 1986b, Increase of 14C activity of dissolved inorganic carbon along 14C the river course, in Stuiver, M and Kra, R S, eds, Inter- nail conf, 12th, Proc: Radiocarbon, v 28, no. 2A, p515-521. 134 Dusan Srdoc et al Srdoc, D, Obelic, B, Horvatincic, N, Krajcar Bronic, I, Marcenko, E, Merkt, J, Wong, H K and A, 986c, Radiocarbon dating of lake sediments from two karst lakes in Yugo- Sliepcevic, 4 slavia, in Stuiver, M and Kra, RS, eds, Internatl C conf, 12th, Proc: Radiocarbon, v 28, no. 2A, p 495-502. Srdoc, D, Obelic, B, Horvatincic, N, Krajcar Bronic, I and Sliepcevic, A, 1984, Rudjer Boskovic Institute radiocarbon measurements VIII: Radiocarbon, v 26, no. 3, p 449- 460. Srdoc, D, Sliepcevic, A, Obelic, B, Horvatincic, N, 1979, Rudjer Boskovic Institute radiocarbon measurements V: Radiocarbon, v 21, no. 1, p 131-137. [RADIOCARBON, VOL 29, No. 1, 1987, P 135-147] RUDJER BOSKOVIC INSTITUTE RADIOCARBON MEASUREMENTS X DUSAN SRDOC, BOGOMIL,OBELIC, ADELA SLIEPCEVIC*, INES KRAJCAR BRONIC and NADA HORVATINCIC Rudjer Boskovic Institute, PO Box 1016, 41001 Zagreb, Yugoslavia The following radiocarbon date list contains dates of samples measured since our previous list (R, 1984, v 26, no. 3, p 449-460). As before, age calculations are based on the Libby half-life (5570 ± 30) yr and reported in years before 1950. The modern standard is 0.95 of the NBS oxalic acid activity. Sample pretreatment, combustion and counting technique are essentially the same as described in R, 1971, v 13, no. 1, p 135- 140, supplemented by new techniques for groundwater processing (R, 1979, v 21, no. 1, p 131-137). Statistical processing of data has been computerized (Obelic & Plan- inic, 1977; Obelic, 1980). Sample descriptions were prepared with collec- tors and submitters. The errors quoted correspond to Ur variation of sample net counting rate and do not include the uncertainty in 14C half-life. Calculations of age of speleothems and groundwaters are based on the initial activity equal to 0.85 of the NBS oxalic acid activity multiplied by 0.95.

ACKNOWLEDGMENTS We thank E Hernaus for preparation of samples and methane synthesis, A Turkovic for data processing and P Hojski for technical assistance.

ARCHAEOLOGIC SAMPLES

Yugoslavia Croatia Vucedol series Systematic excavations at Eneolithic site Vucedol, "Strein's vineyard," near Danube R, Vukova (45° 21' N, 19° 01' E), 110m alt, E Croatia. Baden, Kostolac, and Vucedol cultural layers are dominant. Samples coll 1984 and 1985; subm by A Durman, Dept Archaeol, Univ Zagreb.

Z-1446. Vucedo1, Quad 45, 1984 4540 ± 130 Charcoal, Pit 47. Comment (AD): expected age: Baden phase.

Z-1617. Vucedo1, Quad 69/99,1985 4500 ± 100 Charcoal, Pit 76, depth 2.9m. Comment (AD): expected age: Baden phase.

* Faculty of Veterinary Medicine, Physics Department, University of Zagreb

135 136 Dusan Srdoc et al Z-1618. Vucedol, Quad 131/161,1984 4300 ± 100 Charcoal, Pit 88, depth 4.2m. Comment (AD): expected age: Baden phase. Z-1619. Vucedol, Quad 86/87,1985 4400 ± 100 Charcoal, Pit 21, depth 2.6m. Comment (AD): expected age: Baden phase. Z-1447. Vucedol, Quad 115/145,1984 4290 ± 120 Charcoal, Pit 14, depth 2.5m. Comment (AD): expected age: Vucedol phase. Z-1449. Vucedol, Quad 107/138,1984 4190 ± 120 Charcoal, Grave 2, depth 2.3m. Comment (AD): expected age: Vucedol phase. Z-1453. Vucedol, Quad 107/138,1984 4290 ± 120 Charcoal, Grave 2, depth 2.9m. Comment (AD): expected age: Vucedol phase. Z-1454. Vucedol, Quad 107/138,1984 4130 ± 120 Charcoal, Grave 2, depth 3.2m. Comment (AD): expected age: Vucedol phase. Z-1621. Vucedol, Quad 62, 1985 4300 ± 100 Charcoal, Pit 2, depth 3.6m. Comment (AD): expected age: Vucedol phase. Z-1622. Vucedol, Quad 36, 1985 4100 ± 100 Charcoal, Pit 6, depth 2.3m. Comment (AD): expected age: Vucedol phase. Z-1637. Vucedol, Quad 36, 1985 4300 ± 100 Charcoal from bottom of grave with 8 skeletons, Pit 6, depth 4.7m. Comment (AD): expected age: Vucedol phase. Z-1624. Vucedol, Quad 33/43,1985 4200 ± 100 Charcoal, Pit 19, depth 2.4m. Comment (AD): expected age: Vucedol phase. General Comment: 14C ages older than classical chronology. Measurements at other sites of Baden culture made at Groningen and Berlin gave 4500 to 3900 yr BP (classical chronology, 4250-3900 yr BP) and of Vucedol culture, 4300 to 4100 yr BP (classical chronology, 4100 to 3700 yr BP) (Dimitrijevic, 1979). Sisak series Dry summer 1985 caused very low waters in Kupa R at Sisak (45° 29' N, 16° 23' E), 99m alt. Parts of wooden posts emerged from river Rudjer Boskovic Institute Radiocarbon Measurements X 137 bed, probably Roman bridge over Kupa R near ancient Siscia. Coil and subm 1985 by M Smalcelj, Dept Archaeol, Univ Zagreb. Comment (MS): expected age: Roman period.

Z-1580. Sisak No, l 1830 100 Tree rings I to 10 of larger sample of wood.

Z-1581. Sisak No. 2 1880 ± 100 Tree rings 1 to 10 of smaller sample of wood. General Comment: excellent agreement with expected age.

Otisic series Charcoal mixed with ash from hearth, 0.6m below ground level in karst depression Otisic near Sinj (43° 42' N, 16° 14' E), alt 434m, S Croatia. Coil and subm 1984 by A Milosevic, Regional Mus, Sinj. Comments (AM, DS): expected age: prehistoric settlement, 1900-1600 BP; dates indicate recent or medieval occupation of site.

Z-1451. Otisic No. l 330 ± 120

Z-1452. Otisic No. 2 530 ± 120 Z-1450. Zadar, Sv.Sime 1830 + 110 Wooden board from wall above N colonnade, St Simeon church in Zadar (44° 07' N, 15° 15' E), S Croatia. Medieval church was probably built on Roman foundation. Coil and subm 1984 by P Vezic, Inst Preservation Cultural Monuments, Zadar. Comment: date confirms Roman occupation of site.

99.2 ± 0.2% modern Z-1566. Cazma Modern Fragments of wooden pipeline, depth 2.35 to 3.20m, Cazma (45° 45' N, 16° 38' E). Coil and subm 1985 by V Strk, Regional Mus Cazma. Comment (VS): expected age: late Middle Age. Comment: previously reported dates (Z-669: R, 1981, v 23, p 410) indicate replacement of destroyed wooden tubings with new ones.

Sv. Ivan, Bol series Fragments of wooden beams and planks from St John's church in Bol, i Brac (43° 16' N, 16° 40' E), S Croatia. Coil and subm 1985 by G Niksic, Regional Inst for Preservation of Cultural Monuments, Split. Medieval church frequently reconstructed. Dates help to determine periods of reconstruction.

Z-1540. Bol No. l 550 ± 100 Wooden beam in SE corner supporting wall. Comment (GN): expected age: 10-1.2 century AD. 138 Dusan Srd& et al Z-1594. Boi No. 2 320 ± 100 Wooden beam in SE corner supporting wall. Repeated measurement of Z-1540. Both dates are younger than expected.

Z-1541. Bol No. 3 490 ± 100 Fragments of wooden shelf, morgue. Comment (GN): expected age: 15 century AD or later.

103.9 ± 0.9% modern Z-1542. Bol No. 4 Modern Fragment of wooden coffin mixed with earth in central grave.

Jazvinacki brig series Charcoal from Podvrsje near Zadar (44° 07' N, 15° 13' E), S Croatia. Coil and subm 1983 by S Batovic, Archaeol Mus, Zadar. Comment (SB): expected age: Bronze Age.

112.4 ± 1.5% modern Z-1265. Jazvinacki brig No. l Modern Charcoal from depth 0 to 0.1 m. Surface covered by grass with pene- trating roots.

Z-1264. Jazvinacki brig No. 2 560 ± 100 Charcoal from depth 0.1 to 0.3m.

Slovenia Z-1466. Divje Babe 27,000 ± 1300 Cave bear bone (U7sus spelaeus), Divje Babe cave near Sebrelje, SW Slovenia. Coll and subm by D Brodar, Slov Acad Sci Arts. Sample from systematic excavations of Paleolithic site (Turk & Dirjec, 1985). Bones buried in mixture of gravel and earth; cryoturbation was evident. Comment (DB): expected age: >20,000 BP. Result agrees with previously dated charcoal (Z- 1032, -1033: R, 1984, v 26, p 450-451). Comment: dated on extracted and pyrolized collagen. Z-1582. Zamedvedca 10,500 ± 200 Fossil wood (Salix sp) in clay, Ljubljansko barje near Ljubljana (46° 00' N, 14° 23' E), 300m alt. Coll and subm 1985 by A Sercelj. Comment (AS): expected age: Pleistocene.

Ajdovska jama cave series Charcoal from Ajdovska jama cave near Nemska Vas, Krsko, E Slove- nia. Coll 1983 and 1985 by T Bregant and subm by A Sercelj. Samples from systematic excavation of Neolithic site. Corresponds to earlier measurements, Z-1042 to -1045 (R,1982, v 26, p 451). Comment (AS): expected age: Neolithic. Rudjer Boskovic Institute Radiocarbon Measurements X 139

Z-1178. Ajdovska jama No. l 5400 150 Charcoal from right entrance to cave. Z-1179. Ajdovska jama No. 3 4700 200 Charcoal from grave in cave.

Z-1554. Ajdovska jama No. 5 4700 120

Charcoal from central cave, depth 1 m.

Z-1602. Ajdovska jama No. 68/85 4850 130 Charcoal from fireplace, depth 1.9m.

Z-1603. Ajdovska jama No. 34/85 2900 ± 120 Charcoal from fireplace in deepest layer with human bones, depth 2.3m.

Moverna vas series Charcoal mixed with earth from hearth, Moverna vas near Crnomelj (45° 39' N,15° 13' E), 160m alt, SE Slovenia. Coll 1984 by M Budja, Fac Sci Arts, Ljubljana, subm by A Sercelj. Dates Neolithic and Eneolithic cultures. Establishment of chronologic model of Neo-eneolithic of W Yugoslavia on basis of vertical stratigraphy and 14C measurements. Comment (MB): expected age: 4000 to 5000 BP.

Z-1474. Moverna vas No.1 5400 140 Charcoal, Layer 7, Quad 2, depth 2.2 to 2.5m.

Z-1475. Moverna vas No. 8 4900 130 Charcoal, Layer 6, Quad 2, depth 1.5 to 1.8m.

Z-1476. Moverna vas No. 9 4050 ± 120 Charcoal, Layer 5, Quad 2, depth 1.5 to 1.8m. Comment (MB): result disagrees with other samples. Subm another sample (Z-1685) from same layer.

Z-1685. Moverna vas No. 11 3900 ± 100 Charcoal, Layer 5, Quad 2, depth 1.5 to 1.8m (cf Z-1476). Comment: both measurements on separate samples (Z-1476 and -1685) gave dates within statistical error.

Donji Lakos series Oakwood (Quercus robur) samples in soil, l to 2m deep. Excavations of Bronze Age site near Donji Lakos (46° 33' N, 16° 26' E), 150m alt. Coll by j Dular, Slov Acad Sci Arts; subm by A Sercelj, 1983. Comment (AS): expected age: 3500 to 4000 yr BP.

Z-1467. Donji Lakos No. 9 1120 ± 100 140 Dusan Srdoc et al Z-1468. Donji Lakos No. 10 1400 ± 100 Z-1469. Donji Lakos No. 11 1020 100 General Comment: results disagree with expected ages. Z-1421. Breg 6630 ± 150 Charcoal from hearth, 0.7 to 0.8m deep in dolomite sand. Mesolithic site near Breg, Skoflijica (45° 59' N, 14° 34' E), 294m alt. Coil and subm 1984.by F Osole, Dept Geol & Palaeontol, Univ Ljubljana. Comment (FO): expected age: 7000-8000 yr BP. Ajdna series Charcoal with recent roots from Ajdna near Bled (46° 25' N, 14° 07' E), W Slovenia. Coil and subm 1985 by A Sercelj. Comment (AS): expected age: Late Classical to Old Slavic times. 109.9 ± 0.7% modern Z.1575. Ajdna No.1 Modern Small pieces of charcoal in soil.

Z-1576. Ajdna No. 2 660 ± 100 Charcoal from hearth near ruins of church.

Neolithic in Serbia Systematic excavations for study of early (Starcevo) and late Neolithic (Vinca) cultures in central Serbia were conducted at Grivac and Divostin as part of joint Yugoslav-American project organized by A McPherron, Dept Anthropol, Univ Pittsburgh, and D Srejovic, Philos Fac, Univ Belgrade. Grivac series Charcoal from Neolithic settlement near Grivac (44° 01' N, 20° 42' E), 18km W of Kragujevac. Coil and subm 1984 by M Bogdanovic, Regional Mus, Kragujevac. 14C measurements made at Berlin gave 7250 yr BP for Starcevo and 6300 to 5930 yr BP for Vinca phase (McPherron & Srejovic, 1971). Z-1507. Grivac No. l 5600 ± 140 Charcoal, Trench A, Layer V. Comment (AMcP): expected age: Vinca phase.

Z-1508. Grivac No. 2 6000 ± 140 Charcoal, Trench A, Layer IX. Comment (AMcP): expected age: Starcevo phase. Z-1505. Divostin D-5/2 6900 ± 150 Charcoal, Layer VII, from systematic excavations near Divostin (44° 02' N, 20° 50' E), 7km W of Kragujevac. Coll and subm 1984 by M Bogdanovic. 14C measurements made at Berlin and British Mus gave 7100 Rudjer Boskovic Institute Radiocarbon Measurements X 141 to 6900 BP for Starcevo phase. Charcoal, 35cm below Level VII (Vinca phase) divided into 3 portions, dated at 3 labs (Bln-898: 5860 BP, BM-574: 5250 BP, Z-336: 6000 BP: R,1977, v 19, p 472). Comment (AMcP): expected age: Starcevo phase. Bosnia

Z-1415. Pustopolje 3260 ± 110 Wooden plank from grave floor, Pustopolje near Kupres (43° 59' N, 17° 17' E), 1150m alt. Plank buried in wet soil, 3.7m deep. Coil and subm 1984 by A Benac, Acad Sci Arts B&H, Sarajevo. Comment (AB): expected age: 3500 BP, based on assoc pottery.

GEOLOGIC SAMPLES

Fossil Wood Samples Z-1317. Jelici >37,000 Fossil wood (unid. sp), Orasnica creek bed near Jelici, 4km NE of Knin (44° 02' N,16° 11' E), Dalmatia, S Croatia. Coil and subm 1984 by A Pavicic, Geol Inst, Zagreb. Comment: unexpectedly well-preserved wood stumps in clayey Neogene layer. Date excludes recent growth of trees on much older soil.

Z-1456. IKA No. l 24,800 ± 1000 Fragment of wood (unid. sp) from layer of mud 5 to 6m thick inter- spersed with wooden fragments, ca 100m below sea level, at sea water depth 58m. Offshore drilling platform Panon, Adriatic Sea, Oilfield IKA, 60km SW of Pula (46° 08' N,15° 01' E), Istria, W Croatia. Coll and subm 1984 by I Muhovec, Geoexpert Co, Zagreb. Comment (IM): mud layer indicates silting from Po R, Italy.

Z-1162. Petanjci 5790 ± 160 Fragments of wood from 4m depth of Mura R alluvium at Petanjci near Radenci (46° 38' N, 16° 02' E), NE Slovenia. Coil and subm 1983 by J Pezdic, Jozef Stefan Inst, Ljubljana. Comment (JP): dating of young river alluvium.

Loess Samples Loess profiles dated for drafting of geologic map of Yugoslavia.

Z-1455. Novi Sivac No. 2091 23,200 ± 800 Organic component of loess, open profile, depth 8m near Novi Sivac, N Backa (45° 21' N,19° 24' E), Vojvodina. Coll and subm 1984 by S Trifu- novic, Geol Inst, Belgrade.

Z-1444. Mikanovci No. 2194 5860 ± 150 Calcareous concretions (loess kindchen) from silt layer above road near Mikanovci, Slavonia (46° 17' N, 18° 33' E), E Croatia. Cobb and subm 142 Dusan Srdoc et al 1983 by M Brkic, Geol Inst Zagreb. Comment (MB): expected age: Wurm III. Z-1443. Prkovci No. 2329 3850 ± 130 Calcareous concretions from loess, Prkovci, E Slavonia (45° 11' N, 18° 38' E), E Croatia.

Koska series Calcareous concretions (loess kindchen) and snail shells from clayey silt, depth 1.5m, Koska near Nasice (45° 32' N, 18° l 5' E), E Croatia. Coil and subm 1984 by B Korolija, Geol Inst Zagreb. Comments (BK,DS): expected age: younger Pleistocene to Holocene. Calcareous concretions of loess kindchen type show consistently younger age, indicating carbonate- rich freshwater penetration in loess.

Z-1489. Koska, Na-7496 4360 240 Calcareous concretions (loess kindchen).

Z-1490. Koska, Na-7495 17,400 500 Land snail shells.

Z-1481. Stitar No. 3449 2600 ± 400 Bones (unid. sp) from dark clay layer, depth 1.Om below surface, Sava riverbank at Stitar near Zupanja (45° 05' N, 18° 39' E), E Croatia. Coil and subm 1984 by M Brkic. Comment (MB): expected age: Holocene-Atlanti- cum.

Other Samples Z-1161. Sratovci >37,000 Peat from alluvium of Mura R at Sratovci near Radenci (46° 38' N, 16° 02' E), N Slovenia. Coil and subm 1983 by J Pezdic. Ribnisko barje series Peat from peat bog Ribnisko barje (46° 30' N, 15° 17' E), alt 1550, Pohorje Mt, N Slovenia. Coll and subm 1984 by A Sercelj, Slov Acad Sci Arts, Ljubljana. Dates help determine vegetation development periods and peat bog fm.

Z-1365. 90-100cm 3100 ± 120

Z-1366. 190-200cm 3590 ± 120

Z-1477. Cerna 27,300 ± 1400 Clay with black organic detritus, depth 6.5m. Clay deposit in brickyard Cerna, E Slavonia (45° 11' N, 18° 41' E), E Croatia. No pollen grains found Rudjer Boskovic Institute Radiocarbon Measurements X 143 in sample. Coil and subm 1984 by M Brkic. Comment (MB): expected age: Quaternary marshy deposits. Slavonski Samac series Clay with organic detritus and shells (unid. sp), Pleistocene profile, Sava R bank downstream from Slavonski Samac (45° 04' N, 18° 30' E), E Croatia. Coil and subm 1984 by M Brkic. Comment: dates indicate organic material of secondary origin, probably roots.

Z-1479. Slavonski Samac No. 2807 4500 120 Clay with organic detritus.

Z-1442. Slavonski Samac No. 2807 2340 110 Charred wood embedded in clay. >37,000 Z-1511. Radovna b13C River chalk deposit in Radovna R valley in vicinity of Lake Bled. Coil and subm by J Pezdic. Comment (JP): isotopic composition (180 and 13C) reveals detrital sediment deposited during postglacial period. Mineral composition: calcite (60%), dolomite (30%), other minerals (10%). Celarevo series Samples ofv peat, charred leaves, and organic detritus from several cores drilled at Celarevo near Backa Palanka (45° 15' N, 19° 24' E), W Voj- vodina. Geologic survey of area near Danube R. Coil and subm 1985 by N Krstic, Geoinst, Belgrade. Comment (NK): expected age: Holocene. Z-1548. Celarevo IB-36A, No. l 1230 100 Peat from sandy sediment, depth 2.3m.

Z-1549. Celarevo IB-36A, No. 2 530 100 Organic mud from sandy sediment, depth 3.Om. Z-1550. Celarevo IB-36A, No. 3 960 100 Charred leaves, depth 4.3m. Z-1551. Celarevo IB-36A, No. 4 26,200± Diagenetic mud, depth 22m.

Z-1591. Celarevo G-1, No. l 740 ± 100 Peat, depth 2.6 to 2.65m, Borehole G-1. Comment (NK): expected age: younger Holocene (500-1200 yr). 36,300 + 4400 Z1592. Celarevo G-1, No. 2 - 3600 Peat, depth 20.7 to 20.8m, Borehole G-1. 144 Dusan Srdoc et al Z-1597. Celarevo G-2, No. l 370 ± 100 Peat depth 2.7 to 2.8m, Borehole G-2. Comment (NK): expected age: Holocene. Pollen analyses showed corn indicating layer younger than 350 yr. Z-1598. Celarevo G-2, No. 2 8300 ± 1000 Peat, depth 22.5 to 22.6m, Borehole G-2. Z-1599. Celarevo G-3, No.1 480 ± 100 Peat, depth 1.85 to 1.95m. Borehole G-3.

Z-1600. Celarevo G-3, No. 2 560 ± 100 Peat, depth 2.4 to 2.5m, Borehole G-3. Z-1601. Celarevo G-3, No. 3 650 ± 100 Peat, depth 2.9 to 2.95m, Borehole G-3. 39.2 ± 0.5% modern b'3C Z-1510. Lake Bled = -2%o Lake chalk from Lake Bled (44° 22' N,14° 06' E), W Slovenia. Coll and subm by J Pezdic, Jozef Stefan Inst, Ljubljana. Comment (,JP): chemically very pure calcite without admixture of detrital minerals formed by calcite i4C precipitation from water (Dolenc et al, 1984). Comment: activity of sediment indicates ratio of biogenic vs inorganic carbon.

HYDROGEOLOGIC SAMPLES Croatia

3.3 ± 0.3% modern 31,800 + 2700 Z-1503. Samobor No. 2 - 2200 Water from artesian well at 190 to 220m depth, Samobor (45° 48' N,15° 43' E), near Zagreb, W Croatia. Coil and subm May 1985 by N Tipic, INA-Naftaplin Co, Zagreb. Geothermal exploration near Zagreb. Podsused series Thermal ground water from borehole Podsused (44° 49' N, 15° 50' E), W suburb of Zagreb. Coil and subm Aug 1985 by B Munda, INA-Proj Co, Zagreb.

6.4 ± 0.4% modern Z-1558. Podsused PDTE-1 20,700 ± 700 Thermal ground water, depth 600m.

11.1 ± 0.4% modern Z-1559. Podsused PDT-1 16,300 ± 400 Thermal ground water, depth 282m. Rudjer Boskovic Institute Radiocarbon Measurements X 145 Slovenia

3.0 ± 0.3% modern Z-1174. Topolsica 27,000 ± 1800 Water from borehole PC, Topolsica spa near Velenje (46° 24' N, 15° 01' E), N Slovenia. Tritium activity: <0.2 Bq/1. Coil and subm May 1983 by M Veselic, Geol Inst Ljubljana.

48.6 ± 0.8% modern Z-1346. Medijske toplice 4460 ± 530 Water from borehole V-3. Medijske toplice spa near Izlake (43° 42' N,16° 14' E), Central Slovenia. Coil and subm July 1984 by M Veselic.

Kanizarica series Water samples from Kanizarica in Bela Krajina (46° 35' N,15° 03' E), E Slovenia. Coil and subm by P Kralj, Geol Inst, Ljubljana.

7.8 ± 0.4% modern Z-1405. Kanizarica No. l 19,000 ± 600 Water from borehole V-40. Coil Nov 1984. Tritium activity: (0.4 ± 0.3) Bq/1. 13.2 ± 0.4% modern Z-1426. Kanizarica No. 2 14,900 ± 400 Water from borehole V-41. Coll Dec 1984. Tritium activity: < 0.2 Bq/l. 16.0 ± 0.4% modern Z-1497. Kanizarica No. 3 13,400 ± 300 Water from borehole JV-2/82. Coll Apr 1984. Tritium activity: <0.2 Bq/i.

34.1 ± 0.5% modern Z-1498. Kanizarica No. 4 7500 ± 200 Water from borehole V-42/85. Coil Apr 1985. Tritium activity: <0.2 Bq/l. 2.1 ± 0.5% modern 31,900 + 2400 Z-1484. Kanizarica No. 5 - 1600 Water from borehole JV-2/85. Coil Apr 1985.

Bosnia Tuzla series Water from several boreholes from salt mine Tuzla (44° 33' N, 18° 42' E). 146 Dusan Srdoc et al 1.8 ± 0.3% modern 30,800 + 2700 Z-1300. Tuzla No.1 -2100 Water from borehole TD-22, depth 285 to 307m. Coil and subm May 1984 by N Djuric, Salt Mine Co, Tuzla. 5.9 ± 0.3% modern Z-1423. Tuzla No. 2 21,300± 750 Water from borehole TD-23, 375m deep. Coil and subm Nov 1984 by N Djuric. Tritium activity: <0.2 Bq/1. 10.3 ± 0.4% modern Z-1187. Tuzla No. 3 17,000 ±500 Water from borehole TD-17, depth 480m. Coil and subm Nov 1983 by N Miosic. Tritium activity: <0.2 Bq/l. 33.8 ± 0.5% modern Z-1188. Tuzla No. 4 7400 ± 200 Water from spring TD-9 ("Tetima"). Coil and subm Nov 1983 by N Miosic. Tritium activity: (0.6 ± 0.2) Bq/l. 13.0 ± 0.4% modern Z-1183. Gusica haus 15,100 ± 400 Thermal water, Gornji Seher near Banja Luka (44° 45' N, 170 45' E), NW Bosnia. Coil and subm Nov 1983 by N Miosic. Geothermal explora- tions near Banja Luka.

4.4 ± 0.3% modern Z-1186. Laktasi 24,000 ± 1000 Thermal water from Laktasi near Banja Luka (44° 45' N, 170 45' E), NW Bosnia. Coil and subm Nov 1983 by N Miosic. Tritium activity: <0.2 Bq/1. 1.4 ± 0.3% modern 33,000 + 3500 Z-1275. Banja Ilidza - 2800 Water from Ilidza spa near Sarajevo (43° 50' N, 18° 20' E), Bosnia. Coil and subm Feb 1984 by N Miosic. Tritium activity: <0.2 Bq/1. 1.9 ± 0.3% modern 30,500 + 2700 Z-1499. Fojnica - 2300 Thermal water from Fojnica spa, depth 50 to 300m, (43° 58' N,17° 54 E), 670m alt, central Bosnia. Coll and subm Apr 1984 by N Miosic.

REFERENCES Dimitrijevic, S, 1979, Badenska kultura; Vucedolska kultura i Vucedolski kulturni kompleks, in Benac, A, ed, Praistorija jugoslavenskih zemaija, Vol III-Eneolitsko doba: Svjetlost, Sarajevo, p 183-342. Rudjer Boskovic Institute Radiocarbon Measurements X 147 Dolenc, T, Pezdic, J, Misic, J and Ogorelec, B, (ms) 1984, Isotopic and mineralogical characteristics of lake chalk in NW Slovenia: Paper presented at European regional mtg of sedimentology, 5th, Marseilles, France. McPherron, A and Srejovic, D, 1971, Early farming cultures in central Serbia (eastern Yugo- slavia): Prelim rept, Natl Mus Kragujevac, p 1-26. Obelic, B, 1980, Computer analysis and interpretation of radiocarbon data: Fizika, v 12, suppl 2,p 139-161. Obelic, B and Planinic, J, 1977, Computer processing of radiocarbon and tritium data, in Povinec, P and Usacev, S, eds, Internatl conf on low-radioactivity measurement and applications, Proc: The High Tatras, Slovenske pedagogicke nakladatelstvo, Bratislava, p 117- 120. Srdoc, D, Breyer, B and Sliepcevic, A, 1971, Rudjer Boskovic Institute radiocarbon measurements I: Radiocarbon, v 13, no. 1, p 135-140. Srdoc, D, Horvatincic, N, Obelic, B and Sliepcevic, A, 1982, Rudjer Boskovic Institute radiocarbon measurements VII: Radiocarbon, v 24, no. 3, p 352-371. Srdoc, D, Obelic, B, Horvatincic, N, Krajcar Bronic, I and Sliepcevic, A, 1984, Rudjer Boskovic Institute radiocarbon measurements VIII: Radiocarbon, v 26, no. 3, p 449- 460. Srdoc, D, Sliepcevic, A, Obelic, B and Horvatincic, N, 1977, Rudjer Boskovic Institute radiocarbon measurements IV: Radiocarbon, v 19, no. 3, p 465-475. ----- 1979, Ruder Boskovic Institute radiocarbon measurements V: Radiocarbon, v 21,no.1,p 131-137. -----1981, Rudjer Boskovic Institute radiocarbon measurements VI: Radiocarbon, v 23, no. 3, p410-421. Turk, I and Dirjec, J, 1985, Reka, Divje Babe I, Varstvo spomenikov: Ljubljana, v 27, p 189- 191. [RADIOCARBON, VOL 29, No. 1, 1987, P 148-152] NOTES AND COMMENTS

DISCUSSION COMMENTS ON MULTIPLE DATING OF A LONG FLOWSTONE PROFILE RAINER GRUN and HENRY P SCHWARCZ Department of Geology, McMaster University, Hamilton Ontario, Canada We strongly welcome the investigation by Geyh and Hennig (1986) and agree with their conclusion that the boundaries of the interglacial periods cannot be determined exactly using the methods applied so far in the Heg- gen cave. However, since we have been engaged for several years in the application of these methods (except 14C), and since the paper gives a very different interpretation of the ESR results from that presented by one of us (Grun, 1985), we feel the need to make some comments on this paper. Samples were collected in the Heggen cave during three campaigns.

First, samples were taken for paleomagnetic study (see Fig 1: Al -E3l ; cir- cles), which were additionally investigated by U-series (Peters, 1981) and ESR. During a second visit to the site, the profiles HA and HR were cut and later studied by ESR only. Subsequently, a third profile was cut near the sample locations Al -B9 and studied by Geyh and Hennig. As can be seen in Figure 1, the numbering of the samples in series A-E does not coincide with the numerical sequence of series HA or HB. Although Geyh and Hen- nig propose to use the paleomagnetic sequence as an indication of the rate of deposition, the different numbering in the respective sample profiles might have led to some confusion. The Brunhes/Matuyama boundary was, in fact, observed between samples C18 and C19 at a depth of ca 60cm and not between samples HAl 7 and HAl 8 at a depth of 70cm, as cited by Geyh and Hennig. Unfortunately, the scale in Figure 1 of their paper does not agree with the depths quoted in Table 1. Therefore, it is rather difficult to determine from where the samples for 14C and U-series dating were taken.

ESR DATING The naive reader might get the impression from the discussion by Geyh and Hennig that ESR is anything but a dating method. We would like to clarify the presentation of the data and to suggest an alternate explanation. The whole profile can be subdivided into 13 units. The correlation of the different sample sites is shown in Figure 1 (Fig 1 of Geyh & Hennig shows an area to the right of profile HB). The upper parts of profiles HA and HR were connected to the ceiling with calcite drapes. The ESR results for these samples (HA1-HA3a and HB 1-HB4a1) therefore scatter quite randomly and are not included in the following discussion. We shall limit our further discussion to the upper part of the deposit (Units 1-5). The problems of ESR dating in the lower part (recrystallization, U-mobilization, alpha-paleodose, etc) are discussed elsewhere in detail (Grun, 1986).

148 Comments on Multiple Dating of a Long Flowstone Profile 149

632 HA 43 431 813 34 X75 165 110 1399 174 A 413 5 \ 1.4331 (111 111 142 -A 63 299 150 2 [38] 154 139 A3.11,38 198 2 /1 61 5 101 \ l39 \ A 4.135 20 171 129 161 B6225 (2161 303 239 382 31Q _ B7263116lt1 425 386 10 264 618 4 79 B8S368 308 397 B9472 (3021 312 10 343 349 184 171 ??8 B 1134312371 C15 181 188 172 _ B 34 5 12651 C 16.2161>2551 - - - ` 12 231 B134051'4001 C172371>2541 ,------_,; B14509 ['.1.001 15 266 -C18279[>237] B 15?3413161 131 C 1 C20 274 12851 300 D2219812801 519 _ _

552 713 12531 395 522 D254 3512351 10 864 556 D26S4109 985 1030 [240] 1124 750 0285081>234]- 597 524 25 534

1254] D30 269_ 520 560 , D_31.264(>2671 828 032 26 490 34 06 1>328] 708 99 783 753 659 514 30 627 E31.1142 1044

1941 E33 652 301

Fig 1. (Fig 76 in Grun, 1985). Schematic presentation of the sampling sites with U series (Peters, 1981; in brackets) and ESR age results (Grun, 1985). = samples for NRM: --- = correlation of the samples according to the speleothem layers. 150 Rainer Grim and Henry P Schwarcz We can roughly correlate between the sequence of ESR samples and Geyh and Hennig's sequence as follows: ESR sequence Geyh and Hennig Unit 1: Al-A2 Layer 1-2 Unit 2: A3-A4, HB4a2-HB6B Layer 3-4 Unit 3: B6-B7, HA3B-HA5, HB7-HB9 Layer 5-7 Unit 4: B8-B9, HA8-HA10, HB10-HB12 Layer 8-9 Unit 5: B 11-B 13, HA8-HA 10 Layer 10-12 Geyh and Hennig present, in their Figure 2, the "scattering of AD with depth." It must be noted, however, that the layers given on the x-axis do not correspond to the layers as presented in their Figure 1 and Table 1, but are taken from the HA profile. Also, we cannot reproduce their correlation of the plots, eg, the AD of Sample B9 is plotted together with those for HA4 and HB9, which is obviously wrong (this might be due to the confusion in labeling noted previously). Since the layers of their Figure 2 do not agree with the layers as used for 14C and U series dating, it is not surprising that, as the authors state: "the apparent stepwise increase of the 14C AD. . . does not coincide with the boundaries reflected by the and U/ Th data." More important is the question of the possible significance of the plot of AD vs depth presented by the authors. The accumulated dose is only one of many parameters used to calculate an ESR age. Therefore, a plot of AD alone is not very meaningful. For example, the samples with the highest ages (above 1 Ma) have ADs varying between 11.8 to 44 krad and the highest AD (95 krad) yielded an age of only 769 ka. For these reasons, no significant conclusions about the validity of ESR dating can be drawn from their Figure 2. In our Figure 2 we show the plot of ESR ages vs depth for the Units 1-5. The ESR age estimates of ca 40 ka for the upper samples (Holocene

Unit 0 100 200 300 400 500 AgelkaJ avAgefkal

asp0 1 0 49.7±14.4 ---- o 0-6------No 2 -----A-°`b------00N 139.5 ± 16.9 0 0 0 0 0 0 226 ±42 3 0 0 ------0 a. ----a------0o 0 It 0 303 ± 93

0 0 5 0 401±82 0

Fig 2. ESR age results for Units 1-5 (see Fig 1). Comments on Multiple Dating of a Long Flowstone Profile 151 according to 14C and U series) are attributed by Geyh and Hennig to the large interlaboratory variation in AD estimates (25 - >_ 100%) as shown by the E.SR comparison project (Hennig, Geyh & Grun, 1985). If this were an explanation of the four-fold overestimation in age, this systematic error should occur throughout the profile (the laboratory reproducibility for our ESR data was in the range of 10-30%; see Fig 2), but this is obviously not the case. Even though the age data show a scatter of up to 30% (Unit 4), the ESR results of the various units can easily be correlated with the warm stages of the oxygen isotope record. The scatter can be accounted by the failure of one or more of the assumptions made for ESR age calculations: constant alpha-efficiency and 234U/ 238U ratio; closed system behavior (lack of recrystallization or U mobilization). Up to now, we have no reasonable explanation for the high dates of Unit 1 and the slightly high ages for Unit 2 (which should be ca 125 ka). Although we do not insist that Unit 1 is not Holocene, we also do not know whether Geyh and Hennig in fact used the same samples for their measurements. It must be mentioned that the U series results of Peters (1981) on the samples used for NRM yielded 11 ka for Al and 38 ka for A2; the latter obviously agrees with the ESR results. Geyh and Hennig further assert: "... most of the samples deeper than Layer 2 yielded ESR ages with a range of 200 to 400 ka without an obvious trend towards an increase with depth." It is apparent from Figure 2 that there is indeed a trend of increasing ESR ages with depth (with a correlation coefficient of 0.85). Possible explanations for the scatter are given above. The ESR dates for Units 3 and 4 agree fairly well with the Geyh and Hennig U series results of 216 ± 77 and 300 ± 76 ka, respectively (the age of Unit 5 is beyond the Th/U dating range).

U SERIES DATA Geyh and Hennig state that the brownish layers might contain clay minerals being pushed from the deeper part into this layer. This should be documented by the presence of a 232Th peak of the alpha spectra. Unfortu- nately, the authors do not present the isotopic results of the Th/U analyses. For this reason it is also difficult to estimate the reliability of ages for the lower layers, since the 230Th/232Th ratio might be so low as to contribute significantly to the error in the age.

STABLE ISOTOPE ANALYSES As noted by Geyh and Hennig, oxygen isotopic data for calcite deposited in caves can only be used to infer paleoclimatic conditions if the calcite was precipitated in isotopic equilibrium with drip waters. It is, in fact, possible to determine whether this is the case, as pointed out long ago by Hendy (1971). Deposits that satisfy these criteria have been shown to record paleotemperatures in other caves, as shown by several studies in this laboratory (eg, Gascoyne, Ford & Schwarcz, 1981). Essentially, one must test whether the oxygen isotopic composition of a single growth layer is constant over a considerable area of the layer. In cases where this criterion is not satisfied, there is no simple relation between paleoclimate and the isotopic composition of the speleothem. The calcite deposited on such a 152 Rainer Grun and Henry P Schwarcz speleothem is isotopically out of equilibrium with the cave drip water, and an isotopic profile such as shown by Geyh and Hennig (Fig 3) would depend on such factors as the humidity in the cave, the degree of supersaturation of calcite in the cave, and the position on the speleothem where the profile was taken. Even under conditions of equilibrium deposition, we do not yet know how to interpret profiles of carbon isotopic composition (although pronounced variations such as shown by Geyh and Hennig have been seen in other speleothem records and presumably are somehow related to environmental changes above the cave).

RATE OF SPELEOTHEM GROWTH The authors also make some claims about the relationship between growth rates of speleothems and climate. We would like to note here that, contrary to these authors, speleothems can grow during glacial periods, although normally only in caves located at low latitudes (but may grow even under active glaciers (Gascoyne et al, 1983)). Further, the rate of accumulation of calcite on a speleothem is probably not a simple function of climate. It depends principally on the degree of supersaturation of the drip water with calcite, plus the rate of release of CO2 from the cave atmosphere. These variables can be controlled by such parameters as the depth of soil above the cave, the size (or existence) of an opening connecting the cave atmosphere to the external atmosphere, etc. Temperature alone is not a significant control, except that when the temperature falls below zero, deposition stops. We have studied caves in which deposition continued at an approximately uniform rate until it was apparently abruptly stopped by freezing in the cave (Gascoyne, Ford & Schwarcz,1981).

REFERENCES Gascoyne, M, Ford, D C and Schwarcz, H P, 1981, Late Pleistocene chronology and paleoclimate of Vancouver Island determined from cave deposits: Can Jour Earth Sci, v 18, p 1643-1652. Gascoyne, M, Latham, A G, Harmon, R S and Ford, D C, 1983. The antiquity of Castleguard Cave, Columbia icefields, Alberta, Canada: Jour Arctic & Alpine Research, v 15, p 463- 470. Geyh, M A and Hennig, G 1986, Multiple dating of a long flowstone profile, in Stuiver, M J, 14C and Kra, R S, eds, Internatl conf, 12th, Proc: Radiocarbon, v 28, no 2A, p 503-509. Grun, R, 1985, Beitrage zur ESR-datierung: Sonderveroff Geol Inst Univ Koln, v 59, p 1- 157. ----- 1986, ESR-dating speleothem records: Limits of the method, in Ikeya, M and Miki, T, eds, ESR dating and dosimetry: Tokyo, Tonics Pub Go, Ltd, p 61-72. Hendy, C H, 1971, The isotopic geochemistry of speleothems. Part 1. The calculation of the effects of different modes of formation on the isotopic composition of speleothems and their applicability as paleoclimatic indicators: Geochim et Cosmochim Acta, v 35, p 801- 824. Hennig, G J, Geyh, M A and Grun, R, 1985, The interlaboratory comparison project of ESR dating-Phase II: Nuclear Tracks, v 10, p 945-952. Peters, R J,1981, Neutronenaktivierungsanalytische Bestimmungen von Spurenelementvaria- tionen in Hohlensinter-Records and deren Absolutdatierung uber die Th-230/U-234- Isotopenanalyse: Unverffentliche Diplom-Arbeit, Univ Koln, 124 p. [RADIOCARBON, Voi. 29, No. 1, 1987, P 153-155] REPLY

MEBUS A GEYH AND G J HENNIG

STRATIGRAPHY Grun and Schwarcz claim that the numbering of the flowstone profile conflicts with that published by Grun (1985). Due to the complex sequence of non-continuous sedimentologic and crystallographic strata first brought to attention in our study, all profiles from different sampling projects have somewhat differing numbers. However, our numbers for Layers 1-11 (Fig 1, above) are identical to the labels (black dots in Fig 1, above) of the first paleomagnetically analyzed profile. We did not assign the Bruhnes/Matuyama boundary between Layers HA 17 and HA 18 but noted that "from layer 18 downwards, the magnetic orientation is reversed." The depth scale in Figure 1 conflicts with the mean thickness of the layers in Table 1.

ESR DATING The purpose of our Figure 2 (data from Grun, 1985) was to demon- strate that three stratigraphically related vertical profiles in the same flowstone being only a few decimeters apart from each other, yielded completely different AD/depth relationships. This is independent of using either the original depth data (Grun,1985), or our normalized ones. In any case, only one of the three profiles (shown as a bold line, Fig 2, our paper) shows a linear trend with depth (restricted to 40cm below the top). It is well-known that additional parameters have to be taken into account for the transformation of AD values to ESR ages. However, the main parameter is the uranium content of the speleothem, which is rather uniformly distributed (250 ± 60ppm) in the upper part of the vertical flowstone profile. But, as seen from Figure 2, above, the ESR ages for the same layers scatter by up to a factor of three. Therefore, the question should be raised if deletion of the other results (without age/depth trends) is permit- ted without additional arguments. We are convinced that ESR analyses can yield very reliable dates if the suitability of the samples is proven or disproven, eg, by independent crystallographic, sedimentologic, mineralogic, or trace element analysis. Ob- viously, in the case of very slowly growing flowstone, zones containing suitable and unsuitable material follow so closely that samples a few centi- meters thick (as used for ESR by Grun) most probably contain some unsuitable material. Moreover, Grun had only broken samples, which made strati- fied sampling difficult. These and other reasons already mentioned may be responsible for the Holocene samples yielding ESR ages of ca 40,000 BP + 30% (Grun & Schwarcz).

U/TH DATING A complete data list could not be published due to the restriction in the length of the paper. Grun and Schwarcz stated "we do not know whether Geyh and Hennig in fact used the same samples for their measure- 153 154 Reply ments" of 14C and U/Th. But this was clearly mentioned in the section on 234U/230Th data. The data of Peters (1981) cannot be compared to ours as he used thick layers of up to 10cm for U/Th dating. Indeed, the 232Th content of the samples is increased in samples with obviously too large U/Th ages. However, there is no radiometric method to ascertain whether 23 Th was diagenetically moved by recrystallization or remained in situ.

STABLE ISOTOPE ANALYSES The problems and presuppositions inherent in the transformation of stable isotope data in paleotemperature were first discussed by Fanditis and Ehhalt (1970). Hendy worked as a postgraduate student under Ehhalt and published his identical results one year later. Hence, our citation is correct. As we proved analytically, conditions of isotope equilibrium were not ful- filled for the investigated flowstone (there is a strong correlation between b180). Y3C and Therefore, b values were applied only stratigraphically, correlated to sedimentologic findings and the raw time scale, and any paleoclimatic interpretation was avoided based only on 518Q values.

RATE OF SPELEOTHEM GROWTH Speleothem growth is at least one order of magnitude larger during interglacial periods than during glacial periods (Geyh & Franke, 1970). Geyh, Franke and Dreybroth (1982) discussed this problem in more detail and initiated a theoretical study (Dreybroth,1980). He confirmed quantita- tively the theoretical approach by Franke (1971) who evaluated the parameters determining the growth rate and the shape of stalagmites. The qualitative statements by Grun and Schwarcz were already quantified (Dreybroth, 1980, 1982). In flowstones that have grown during glacial periods, the sedimentation rate (if not zero) seems to be smaller than the erosion rate. This may be different for stalagmites which have sedimentation rates one order of magnitude higher.

CONCLUSION 14C, U/Th, and paleomagnetic data are in fairly good agreement for a long flowstone profile. However, the age resolution achieved is smaller than methodologically expected. Diagenetic processes modify the specific activities of 14C, U, and Th isotopes, as well as the number of trapped elec- trons used for ESR dating. Hence, an exact fixation of the sediment boundaries between glacial and interglacial periods is not possible.

REFERENCES Dreybroth, W, 1981, The kinetics of calcite precipitation from thin films of calcareous solutions and the growth of speleothems, revisited: Chem Geol, v 32, p 237-245. -----1982, A possible mechanism for growth of calcite speleothems without participa- tion of biogenic dioxide: Earth Planetary Sci Letters, v 58, p 293-299. Fanditis, J and Ehhalt, D H, 1970, Variations of the carbon and oxygen isotopic compositions in stalagmites and stalactites: evidence for non-equilibrium isotopic fractionation: Earth Planetary Sci Letters, v 10, p 136-144. Mebus A Geyh and G J Hennig 155 Franke, H W, 1971, Morphologie and Stratigraphie des Tropfsteines-Ruckschlusse auf GroJ3en des Palaoklimas: Geol Jb, v 89, p 473-501. Geyh, M A and Franke, H W, 1970, Zur Wachstumsgeschwindigkeit von Stalagmiten: Atom- praxis, v 16, p 1-3. Geyh, M A, Franke, H W and Dreybroth, W, 1982, Anomal grofle 5°C-Werte von Hochge- birgssinter-Vergeblicher Versuch einer palaoklimatischen Deutung: Holloch, v 5, p 49- 61, Langnau am Albis. R, Grun, 1985, Beitrage zur ESR-Datierung. Sonderveroff: Geol Inst, Univ Koln, v 59, p 1- 157. Peters, R J,1981, Neutronenaktivierungsanalytische Bestimmungen von Spurenelementvaria- tionen in Hohlensinter-Records and deren Absolutdatierung uber die Th-230/U-234- Isotopenanalyse: Unveroffentlichte Diplom-Arbeit, Univ Koln, 124 p. [RADIOCARBON, Voi. 29, No. 1, 1987, P 156-158] DRASTIC INCREASE OF BACKGROUND IN THE GLIWICE RADIOCARBON LABORATORY DURING LATE APRIL, 1986, AND ITS TIME CHANGES* MIECZYST AW F PAZDUR and ANDRZEJ ZASTAWNY Radiocarbon Laboratory, Institute of Physics, Silesian Technical University, Krzywoustego 2, PL-44-100 Gliwice, Poland ABSTRACT. Preliminary observations of changes of the background of three proportional counters during May and June,1986, are presented. Some isotopes contributing to the rise of background were identified on the basis of their half-life determinations. Starting from April 29, 1986, a significant increase of the background of proportional counters in the Radiocarbon Laboratory in Gliwice, caused by damage to a nuclear power plant in Chernobyl near Kiev, USSR, was noted. An analysis of this increase during the first three months is presented for different proportional counters. Three proportional counters are used for natural radiocarbon measurements in our laboratory. Each counter forms a separate unit with its own anticoincidence and material shielding. All units are located in the same semi-underground room. Basic parameters of the units are given in Table 1 (see also Pazdur et al, 1982). An increase of the background counting rate was noted on counter No. 1 (L1) late morning, April 30. This counter was filled a day before with inactive CO2 for routine background counting rate measurement. The same gas was left in this counter till the end of May. Two others were filled with inactive CO2 a few days later. The resulting record of background counting rate changes is shown in Figure 1. A rapid increase of background of counter No. l on April 29 and 30 should be noted with the rate exceeding 1.5 cpm per hour. The maximum of this peak was not detected, but it occurred perhaps on May 1. Mean counting rate recorded from April 30 to May 2 was equal to ca 42cpm, ie, ca 6 times greater than before contamination. Ca 10% of partial results were rejected as outliers by a statistical microcomputer program gathering the data (Wal- anus, 1986). Thus, we conclude that the peak value was greater, but not more than by some 10-20% of the recorded mean value. Maximum increase of background of counters No. 2 and 3 was estimated at ca 6cpm and 1.2cpm, ie, ca 100% and 60% of the pre-contamination levels, respectively. Possible explanation of these differences is in the fact that counters No. 2 and 3 have internal shielding made of a layer of old lead (counter No.

TABLE I Parameters of the proportional counters Total Counter volume pressure no. (dm3) (mm Hg)

1 2.9 1700 0.050 2 4.5 1700 0.050 3 1.5 760 0.030

*This study was sponsored by Central Research Project CPBP 01.06 from Warsaw Uni- versity. 156 Increase of Background in Gliwice, April 1986 157 Ai A2 c m cpm 6 counter L1 Bo =7,0 cpm

o L2 Bo = 6D cpm x !1 L3 Bo = 1,9 cpm

A3 cpm

1,2

0,8

0,4

5 10 15 20 25 30 4 9 14 19 24 29 MAY JUNE Fig I. Increase of background above the pre-contamination level

In Ai,2 lnA3

3,0 0

2,0 -1

w 2

1 2 3 4 5 6 t (104 mm) Fig 2. Semilogarithmic plot of post-contamination changes of background counting rate in different counters 158 Mieczysaw F Pazdur and Andrzej Zastawny 2) and of purified mercury (counter No. 3), and moreover, both units are tightly covered with plastic foil. Post-contamination decrease of background counting rate of all units is shown in Figure 2 in a semilogarithmic plot. Three different time constants can be clearly distinguished in the plot of counter No. 1, corresponding to the half-life values equal to 4.7d, 8.7d, and 37.Od. Decrease of background of units No. 2 and 3 in May can be described by a single decay law with half-life values equal respectively to 7.4d and 7.8d. The contamination of units No. 2 and 3 and the medium-term contamination of unit No. 1 should thus be ascribed to 13J with a half-life equal to 8.1 d, produced in fission of 235U with yield r = 2.9% (Kimel & Mashkovich, 1966). The short- term component in unit No. 1 wasprobably caused by 133Xe with a half-life equal to 5.3d and r 1.9% and 132`Te (half-life 3.2d, r = 4.4%). There are four products of induced fission of 235U with half-lives ranging from 1 to 2 months, which could be responsible for the long-term component in unit 89Sr, 91Y, 95Zr, 1o3Ru, No. 1. These nuclides include and their total abundance in products of 235U fission is close to 20%. Systematic measurements of the background counting rate of counter No. 3, after careful cleaning of the whole unit, indicate that it was stable during June and July, at the level of 2.10 ± 0.05cpm. Preliminary measurements of the total gamma radioactivity of dust from an electrostatic dust collector installed near the counters have shown high radioactivity of dust particles. Most likely the large increase of background in counter No. 1 was caused by dust particles trapped at the HV connections at the end of the proportional counter.

REFERENCES Kimel, L R and Mashkovich, V P, 1966, Zashchita of ionizirujushchih izluchenii: Moskva, Atomizdat, p 47-49. Pazdur, A, Awsiuk, R, Bluszcz, A, Pazdur, M F, Walanus, A and Zastawny, A, 1982, Gliwice radiocarbon dates VII: Radiocarbon, v 24, no. 2, p 171-181. Walanus, A, 1986, A 14C electronic measurement system with a micro-computer, in Stuiver, M 14C and Kra, R S, eds, Internatl conf, 12th, Proc: Radiocarbon, v 28, no. 2A, p 569- 570. Radiocarbon Proceedings of the 12th International Radiocarbon Conference Trondheim, Norway, June 24-28, 1985

Radiocarbon, Volume 28, Number 2A & B, 1986 @ $60.00 Please order from: Radiocarbon Kline Geology Laboratory Yale University PO Box 6666 New Haven, CT 06511

Number 2A contains papers presented at the conference. Topics include Accelera- tor Mass Spectrometric (AMS) Techniques, Natural Variations, The carbon cycle, Methods and Applications, Conventional Techniques, Anthropogenic Variations, Applications in Archaeology, and Reporting and Management of Data.

Number 2B is the special Calibration Issue which contains the latest calibration curves and tables derived from dendrochronological methods. The issue presents, for the first time, all of the most precise calibrations to date covering a time interval of up to 10,000 years. U S Bristlecone Pine chronologies are matched with those of Irish and German Oak and a calibration curve for marine samples is also presented. This issue is proving extremely important to archaeologists in calculating precise radiocarbon ages.

Both issues are available separately at $30.00 each. AVAILABLE PUBLICATIONS

Proceedings of the 12th International Radiocarbon Conference $60.00 (Vol 28, Nos. 2A & 2B, 1986) Proceedings of the 11th International Radiocarbon Conference 50.00 (Vol 25, No. 2, 1983) Proceedings of the 10th International Radiocarbon Conference 60.00 (Vol 22, Nos. 2 & 3, 1980)

Vol 29, Nos. 1-3, 1987 75.00/vol Inst 50.00/vol Ind

Vol 28, Nos. 1, 2 A & B, 3, 1986 80.00/vol Inst 60.00/vol Ind

Vol 27, Nos. 1, 2 A & B, 3, 1985 75.00/vol Inst (includes Proceedings) 50.00/vol Ind Vol 26, Nos. 1-3, 1984 75.00/vol Inst 50.00/vol Ind

Vol 25, Nos. 1-3, 1983 75.00/vol Inst (includes Proceedings) 50.00/vol Ind Vo124, Nos. 1-3, 1982 60.00/vol Inst 40.00/vol Ind Vol 23, Nos. 1-3, 1981 60.00/vol Inst 40.00/vol Ind Vol 22, Nos. 1-4,1980 80.00/vol Inst (includes Proceedings) 60.00/vol Ind Vols 11-21, 1969-1979 60.00/vol Inst 40.00/vol Ind Vols 1-9,1959-1967 20.00/vol 20.00

Comprehensive Index (1950-1965) 20.00 Out of print: Vol 10, Nos. 1 & 2, 1968 Vol 12, No. 2, 1970 Vol 13, No. 1, 1971

Postage outside USA on all vols 10.00/vol

Prepayment requested on all issues New Books on Archaeology Archaeological Results from Accelerator Dating edited by J. A. J. Gowlett and R. E. M. Hedges Twenty-two papers from a conference at the Oxford Research Laboratory for Archaeology in November 1985 at which archaeologists and scientists put the latest accelerator dates in context. The papers fall into five groups: Origins of Agriculture, Plant and Animal Domestication Dating early moves towards agriculture in the Middle East (D R Harris); Seeds of discontent: accelerator dates on some charred plant remains from the Natufian Culture (A J Legge) ; New radiocarbon dates for early sheep at Tell Abu Hureyra, Syria (A J Legge and P Rowley-Conwy). Early Human Remains New radiocarbon dates for two Mesolithic burials in Denmark (S H Anderson, T S Constandse-Westermann, R R Newell, R Gillespie, J A J Gowlett and R E M Hedges), Direct dates for the fossil hominid record (C Stringer); Oxford radiocarbon dates for early man in America (J A J Gowlett). The Upper Palaeolithic Lessons of context and contamination in dating the Upper Palaeolithic (J A J Gowlett and R E M Hedges); Radiocarabon accelerator dating in the earlier Upper Palaeolithic (P A Mellars and H M Bricker); A.M.S. results from Cheddar Gorge - trodden and untrodden `lifeways' (R M Jacobi); Dating late Devensian - early Flandrian barbed points (J Cook and R N E Barton); Radiocarbon dates from the Pitstone soil at Pitstone, Bucks. (J G Evans); Complementarity of conventional and accelerator dating: examples in Pleistocene extinctions (R. Burleigh); Dating results from Palaeolithic sites and palaeoenvironments in Epirus (North-west Greece) (G N Bailey, CS Gamble, H P Fliggs, C Roubet, D P Webley, J A J Gowlett, D A Sturdy and C Turner); Radiocarbon accelerator dates for Upper Palaeolithic sites in European U.S.S.R (0 Soffer). Later Prehistory Prospects for dating Neolithic sites and monuments in the Cotswolds and adjacent areas (T C Darvill); Radiocarbon dating for the Giants' Hills 2 Long Barrow, Skendleby, Lincs. (J G Evans and D D A Simpson); Radiocarbon dating of Hazleton north chambered tomb: a preliminary statement (A Saville); Radiocarbon and the cursus problem (R Bradley); Radiocarabon: a means to understanding the role of Bronze Age metalwork (S Needham); Methodological issues in the study of Bronze Age chronology (R Bradley). Radiocarbon Methodology and Technology The terminology of time (R Gillespie and J A J Gowlett); The future prospects of accelerator dating (R E M Hedges). The volume ends with an index to the four accelerator datelists published in `Archaeometry'. (Oxford University Committe for Archaeology, Monograph 11, 1986) Pb, 176pp with text illustrations. ISBN 0 947816 11 9. Post free price £14.00, US $24.00 Claim 20% discount on this book for pre paid orders before April 1987

The Radiocarbon User's Handbook by Richard Gillespie A practical guide for field workers and radiocarbon users explaining how to collect samples, how they are treated in the laboratory and how to interpret and use the results. `An invaluable aid to all actively engaged in dating past events' - J. Desmond Clark. (Oxford University Committee for Archaeology, Monograph 3) A5, 42pp. Revised edition 1986. ISBN 0 947816 03 8. Post free price £2.50, US $4.75 Archaeology, Dendrochronology and the Radiocarbon Calibration Curve Ten papers given at an Edinburgh University `Workshop' in October 1982, ed. B. S. Ottaway (Edinburgh University, Dept. of Archaeology, Occasional Paper 9, 1983) A4, 99pp. ISSN 0144-3317 (9). Post free price £8.50, US $14.50.

Available direct from Oxbow Books. Prices include worldwide mailing costs (by air to USA). Send cheques in Sterling or US dollars; or charge to Mastercard, Amex, Access or Eurocard; or pay to our Giro Account 25 606 4008. Oxbow Book News - our quarterly guide to new books on all aspects of Archaeology - is sent free to all those on our mailing list. Are you?

Oxbow Books, 10 St. Cross Road, Oxford 0X1 3TU, England

RADIOCARBON

Editor: MINZE STUIVER Managing Editor: RENEE S KRA Published by THE AMERICAN JOURNAL OF SCIENCE Editors: JOHN RODGERS, JOHN H OSTROM, ROBERT A BERNER, DANNY M RYE Managing Editor: MARIE C CASEY

Published three times a year, in Winter, Spring, and Summer, at Yale University, New Haven, Connecticut 06511. Subscription rate $75.00 (for institutions), $50.00 (for individuals), available only in whole volumes. The Proceedings of the Twelfth International Radiocarbon Conference, Vol 28, Nos. 2A and 2B, 1986 are $60.00. No. 2B, the Special Calibration Issue, is available separately for $30.00. The full subscription for 1986 which includes the Proceedings is $80.00 (institutions) and $60.00 (individuals). The Proceedings of the Eleventh International Radio- carbon Conference, Vol 25, No. 2, 1983 are $50.00, and the Proceedings of the Tenth International Radiocarbon Conference, Vol 22, Nos. 2 and 3, 1980 are $60.00. Back issues and price lists may be obtained from the office of RADIOCARBON. All correspondence and manuscripts should be addressed to the Managing Editor, RADIOCARBON, Kline Geology Laboratory, Yale University, 210 Whitney Ave, PO Box 6666, New Haven, Connecticut 06511. Reprints. The minimum reprint order for each article will be 50 copies without cover. No reprints will be furnished free of charge unless page charges are paid. The cost of additional copies will, of course, be greater if the article is accompanied by plates involving unusual expense. Copies will be furnished with a printed cover giving the title, author, volume, page, and year, when specially ordered. Page charges. Each institution sponsoring research reported in a technical paper or a date list, will be asked to pay a charge of $80.00 per printed page. Institutions or authors paying such charges will be entitled to 100 free reprints without covers. No charges will be made if the author indicates that his institution is unable to pay them, and payment of page charges on an article will not in any case be a condition for its acceptance. Missing issues will be replaced without charge only if claim is made within three months (six months for India and Australia) after the publication date. Claire for missing issues will not be honored if absence results from failure by the subscriber to notify the Journal of an address change. Illustrations should include explanation of symbols used. Copy that cannot be reproduced cannot be accepted; it should be capable of reduction to not more than 10 by 17.5, all lettering being at least 1/6 inch high after reduction. When necessary, one large map or table can be accepted, if it will not exceed 17.5 inches in width after reduction. Line drawings should be in black India ink on white drawing board, tracing cloth, or coordinate paper printed in blue and should be accompanied by clear ozalids or reduced photographs for use by the reviewers. Photographs should be positive prints. Photostatic and typewritten material cannot be accepted as copy for illustrations. Plates (photographs) and figures (line drawings) should each be numbered consecutively through each article, using arabic numerals. If two photographs form one plate, they are figures A and B of that plate. All measurements should be given in SI (metric units). Citations. A number of radiocarbon dates appear in publications without laboratory citation or refererence to published date lists. We ask that laboratories remind submitters and users of radiocarbon dates to include proper citation (laboratory number and date-list citation) in all publications in which radiocarbon dates appear. Radiocarbon Measurements: Comprehensive Index, 1950-1965. This index covers all published 14C measurements through Volume 7 of RADIOCARBON, and incorporates revisions made by all laboratories. It is available to all subscribers to RADIOCARBON at $20.00 US per copy. List of laboratories. The comprehensive list of laboratories at the end of each volume appears in the third number of each volume. Accelerator laboratories have been added to the list. Changes in names or addresses, additions, or deletions should be reported to the Man- aging Editor by May 1. Annual Index. All dates appear in index form at the end of the third number of each volume. Authors of date lists are asked to supply indexed material of archaeologic samples only with their date lists. Vol 29, No. 1 Radiocarbon 1987 CONTENTS

Evaluation and Status of Liquid Scintillation Counting for Radiocarbon Dating Henry A Polach ...... 1

Scintillation Counter Performance at the SMU Radiocarbon Laboratory James M Devine and Herbert Haas ...... 12

Radiocarbon Content of Tropospheric CO2 at China Lake, California 1977-1983 Rainer Berger, TB Jackson, Robert Michael, and HE Suess 18

Study of Bone Radiocarbon Dating Accuracy at the University of Arizona NSF Accelerator Facility for Radioisotope Analysis TW Stafford, Jr, AJT full, Klaus Brendel, RC Duhamel, and Douglas Donahue ...... 24

Comparative Study of the Radiocarbon Dating of Different Bone Collagen Preparations DM Gurfinkel ...... 45

Comparison of Oceanic 014C Data with Those of GEOSECS: Vertical Profiles in 1973 (GEOSECS) and in 1980 at (30°N, 170°E) in the Northwestern Pacific Ocean Toshitaka Gamo, Yoshio Horibe, and Hiromi Kobayashi..... 53

DATE LISTS ...... 57 NOTES AND COMMENTS ...... 148