The accuracy of AMS radiocarbon determinations on very small samples has been tested by measuring a suite of microgram-sized samples of a known-age material. The total measurement precision for the smallest sample (50μg) was found to be ± 3% and the precision improved with larger sample size. The accuracies of the measurements were found to be within the measurement precisions.

The Riga Radiocarbon Laboratory began its serial measurements in 1971. Radiocarbon activity is measured by solution scintillation counters. Benzol, obtained by acetylene trimerization on vanadium pentoxide activated silicagel, is used as 14C activity carrier. The dates refer to 1950 according to the 5568-year half-life of 14C. The statistical error is 1σ (Veksler, Putans & Savvaitov, 1971).

We present data here of samples from coastal regions of the USSR, which reflect their geologic-geomorphologic development (Veksler et al, 1975; Veksler, Putans & Stelle, 1977; Veksler & Prede, 1986).

The occurrence of atmospheric 14C variations complicates calibration, ie, the translation of 14C ages into real calendar ages. The procedure of wiggle matching, however, allows very precise calibration, by matching known 14C variations with wiggles in the floating chronology. In principle, wiggle matching can also be applied to a series of 14C dates from organic (peat) deposits. Where, in general, 14C ages are required at short distances and on small samples, dating by 14C accelerator mass spectrometry (AMS) is required.

Gas chromatographic analysis of the benzene obtained for radiocarbon dating revealed only very low levels (<200 ppm in total) of impurities, most of which were alkylbenzenes. Traces of acetone and phenol were also found. These impurities, at such low levels are thought unlikely to affect counting efficiency.

Radiocarbon ages on handpicked foraminifera from deep-sea cores are revealing that areas of rapid sediment accumulation are in some cases subject to hiatuses, reworking and perhaps secondary calcite deposition. We present here an extreme example of the impacts of such disturbances. The message is that if precise chronologies or meaningful benthic planktic age differences are to be obtained, then it is essential to document the reliability of radiocarbon ages by making both comparisons between coexisting species of planktonic foraminifera and detailed down-core sequences of measurements.

A fast 14C sample preparation technique for organic material in conventional 14C counting is described. The basic difference from conventional preparation is oxidation of the organic substances in a closed system under an oxygen pressure of ca 10 bars. After the explosion-like combustion, the reaction products SO2 and NO2 are oxidized immediately to SO4 2-and NO3 − and precipitated on the wall of the reaction vessel. The residual gas mixture is passed first through a cold trap at −78° to remove water vapor and then through an activated charcoal column at 0° for purification. CO2 is removed from the remaining O2−CO2 gas mixture in a LN2 trap at a pressure of ≃ 100mbar. With this method sample preparation time is reduced from ca 10 hr to ca 1 hr. The efficiency and reproducibility of the procedure is shown with 14C and 13C measurements of a sugar-carbon standard. The results obtained by the new technique agree well with conventionally prepared wood and bone samples.

Our standard procedure of 14C concentration measurements is reported in Riga I. We report here data on samples from the Latvian SSR and Eastern Siberia.

Temporal variations of cosmogenic radionuclide atmospheric concentrations can be caused by such global phenomena as solar activity and geomagnetic field changes as well as atmospheric circulation processes. These causes can be distinguished by the comparison of several isotope records corresponding to the same time period. We discuss a possibility for reconstructing the geomagnetic moment during the last 30,000 years from the comparison of 10Be and 14C concentrations in terrestrial archives. The results agree with conventional paleomagnetic data and promise to enrich our knowledge of geomagnetic field variations and reversals.

Preparation techniques used in dating freshwater peat were evaluated and compared using a homogeneous sample. This paper details various approaches to pretreatment of this material and presents the resultant 14C dates. General trends in the results are discussed.

The radiocarbon dates reported in this list are almost all based on collagen measurements from human archaeologic bone material. This collagen was isolated according to the methods developed by Berger, Horney and Libby (1964), Ho, Marcus and Berger (1969), Longin (1971) and Protsch (1973). All samples were counted as CO2 gas at close to 1 atm in a 7.5L proportional counter with three energy channels. The 14C dates reported are calculated with the Libby half-life of 5568 ± 30 years as required by convention. 13C measurements were used to normalize to −25‰ in the standard manner. The biospheric standard is 95% the count rate of NBS oxalic acid for radiocarbon laboratories. Background is based on CO2 obtained from marble. The accuracy of counter operation is checked against historically dated wood from the funerary boat of Sesostris III, 1872−8 +4 bc (Hayes, 1962). The 14C ages are not tree-ring calibrated.

Large numbers of 14C dates of the base and top of Holocene peat layers may be plotted in 14C histograms in order to establish statistically a chronology of periods of essentially clastic sedimentation and peat formation. Due to the non-linearity of the 14C time scale in terms of calendar years, clustering of 14C dates on random peat growth may occur. This seriously hampers the interpretation of histograms. A quantitative method and computer program were developed to correct the histograms for this effect. The correction factor that has to be applied depends on the calibration curve and the interval width of the correction parameter dy. For peat samples, an interval width of 100 14C yr and a calibration curve based on a 100-yr moving average seems to be a reasonable choice.

The Menlo Park Radiocarbon Laboratory utilizes gas proportional counting of carbon dioxide in counters installed 9.5m below the ground surface for background reduction (Robinson, 1977, 1979; Robinson & Trimble, 1981, 1983). The reported results of geologic samples closely follow the guidelines of Stuiver and Polach (1977). The standard errors for analyses earlier than USGS-500 are based solely upon counting statistics, while later analyses include estimates of uncertainty in voltage, pressure, temperature, and δ13C. Unless otherwise noted, organic samples are pretreated in NaOH and HCl, and carbonate samples in HCl.

Origins of carbon in potsherds are studied through field experiments which involve firing pots containing selected components and laboratory analyses. As temper seems to be the main source of carbon, radiocarbon dates may be too old if calcium carbonate is used. We can avoid the carbon from the clay by low-temperature burning of samples.

Twenty-eight 14C analyses are reported for carbonized roots and other plant material collected from beneath 15 prehistoric lava flows erupted from the northeast rift zone (NERZ) of Mauna Loa Volcano (ML) utilizing the recovery techniques of Lockwood and Lipman (1980). Most samples were collected from the Hilo 7 1/2’ quadrangle during field work for a geologic map of that quadrangle (Buchanan-Banks, unpub data); a few sample sites are located in adjacent quadrangles: Piihonua to the west and Mountain View to the south. Altitudes are given in English units as well as metric to facilitate locating sites on USGS topographic maps.

The standard procedure for measuring 14C at the Gif-sur-Yvette Tandetron AMS facility is described. A new sample manipulator and automated measurements are being used and have been operational for six months. Evidence of isotopic fractionation in the sputter ion source is provided. We take this into account by measuring the 13C/12C ratio of the sample in the accelerator.

This is a list of radiocarbon analyses determined since our previous list (R, 1987, v 29, no. 1, p 135–147). As before, with the exception of speleothems and groundwaters, the calculations follow conventional radiocarbon protocol (Stuiver & Polach, 1977). Sample pretreatment, combustion and counting techniques 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). Speleothem and groundwater model ages (given in the comment section of each listing) are based on an initial activity of 85 ± 10 percent modern carbon (pMC), as recommended in Srdoč et al (1986), without adjustment for δ13C. These model ages can be converted from the 5570-yr half-life to the chronometrically more correct half-life of 5730 yr by multiplying by the factor 1.029. The “± figures” on these model ages are statistical combinations of the 1σ counting uncertainty and the ± 10% uncertainty in the assumed 85 pMC. Ages and “± figures” of all other samples are adjusted for stable isotope fractionation according to the recommendations in Stuiver and Polach (1977). Data processing is done by computer (Obelić & Planinić, 1977; Obelić, 1980). Sample descriptions are prepared in collaboration with collectors and submitters.

This list contains the results of 14C determinations obtained at the laboratory in 1986 and 1987. δ13C was measured by the British Museum Research Laboratory or the Free University of Brussels: *δ13C is estimated (Stuiver & Polach, 1977).

Determination of the more common light elements such as nitrogen and chlorine at trace levels is difficult because of their high abundance on sample surfaces, in materials used to build analysis instruments, and in the residual gas of the instrument vacuum. We present here a new approach to analysis of these elements: accelerator mass spectrometry (AMS) combined with neutron activation. The problem of contamination is overcome by using neutron activation to produce long-lived radioisotopes which generally have low concentrations in the environment. For measurement of 14N and 35Cl, AMS can provide sensitive background-free measurements of their neutron activation products 14C and 36Cl and, in addition, can provide depth profiles. These are the first results of this new method: depth profiles of nitrogen and chlorine implanted in semiconductor grade silicon.

Since our publication of the high-precision calibration curve for the larger part of the 4th millennium BC (de Jong, Becker & Mook, 1986), we found that the dendrochronologic scale (cal bc) needs a correction of 26 years. Instead of using the zero-point of the Niederwill chronology (4039 bc) which was floating at the time, our dendrochronologic scale was erroneously based on the zero-point of the Hohenheim master chronology, which, in its 1986 stage, extended to 4065 bc.