In this paper, we explain our routine pretreatment of bone for radiocarbon dating by accelerator mass spectrometry (AMS), based on the specific reaction between amino acids and ninhydrin described by Nelson (1991). The values and uncertainties of the total system background are presented as a function of the carbon sample mass and the reliability of this method is discussed.

We constructed an automated system to transform organic samples to CO2, which included several options such as: combustion in 2 steps with collection of the 2 fractions, volatile fraction combustion, and 13C sampling. The process includes organic matter combustion, CO2 drying, quantification of the mass of carbon, CO2 collection in a glass vial, and eventually 13C sampling. The system is computer-controlled and -monitored. The apparent background age of the automated system reaches 0.191 ± 0.011 pMC (2 σ), equivalent to a 14C age of about 51,700 yr BP, and requires only 30 min of handling, instead of the several days needed when using a manual procedure.

We have analyzed by radiocarbon 27 consecutive single rings, starting from AD 1952, of a preliminarily cross-dated section (DFR 021) of Pinus radiata, which grew in Armidale, northern New South Wales, Australia. The bomb 14C results suggested the possibility of 2 false rings, and, consequently, 2 misidentified rings in the preliminary count for this section. This possibility was supported by a better ring-width correlation between the revised DFR 021 count and other Pinus radiata chronologies in the study region. This indicated that bomb 14C is a useful tool to complement the standard techniques of dendrochronology in tree species where annual rings are not always clearly defined.

Radiocarbon dates of food residue on pottery from northern European inland areas seem to be influenced significantly by the freshwater reservoir effect (“hardwater” effect) stemming from fish and mollusks cooked in the pots. Bones of freshwater fish from Stone Age åmose, Denmark, are demonstrated to be 100 to 500 14C yr older than their archaeological context. Likewise, food residues on cooking pots, seemingly used for the preparation of freshwater fish, are shown to have 14C age excesses of up to 300 yr. It is probable that age excesses of similar or even larger magnitude are involved in food residue dates from other periods and regions. Since this effect cannot, so far, be quantified and corrected for, 14C dating of food residue, which may potentially include material from freshwater ecosystems, should be treated with reserve.

In a core off Pakistan, we obtained 38 14C analyses by accelerator mass spectrometry (AMS) from a 4.4-m-thick, expanded, annually-laminated Last Glacial Maximum (LGM) section, bracketed by bioturbated intervals ascribed to the Heinrich-1 (H1) and Heinrich-2 (H2) equivalent events (52 14C analyses between 24–15 kyr BP). A floating varve age scale, anchored to the oxygen isotope record of the layer-counted GISP2 ice core at the H2/LGM boundary, results in an annually dated record for the LGM from 23,450–17,900 cal BP. The floating varve scale of the LGM provides us with a tentative calibration of local marine AMS 14C age dates to calendar years.

A simple method for preparing pollen concentrates for 14C AMS dating is applied to organic and inorganic deposits from a peat bog in south Westland, New Zealand, from which preliminary AMS dating indicated age inversions and severe younger carbon contamination problems. The AMS ages of the pollen concentrates provided consistently older age estimates for each sample than ages derived from their respective organic residue or combined pollen and organic residue fractions. It is likely that the younger age estimates of the organic residue fractions result from the incorporation of younger plant material into the sample and possible contamination from younger humic acids percolating through the site.

The issue of comparability of measurements (and thus bias, accuracy, and precision of measurement) from diverse laboratories is one which has been the focus of some attention both within the radiocarbon community and the wider user communities. As a result, the 14C community has undertaken a widescale, far-reaching, and evolving program of intercomparisons, to the benefit of laboratories and users alike. The benefit to the users is, however, indirect, since the 14C intercomparisons have not been used to generate “league tables” of performance, but rather to allow individual laboratories to check procedures and modify them as required.

The historical progression of 14C laboratory intercomparisons from the Third (TIRI, completed in 1995, Gulliksen and Scott 1995) and Fourth (FIRI, completed in 2000, Scott 2003; Boaretto et al. 2000; Bryant et al. 2002) suggests that a Fifth (VIRI) should also be expected. We describe the plans for VIRI.

Carmi (2002) is a response to our study published in Radiocarbon 43(1) by Rasmussen et al. (2001). We noted widespread possible exposure to castor oil of the Dead Sea Scrolls (DSS) in the Rockefeller Museum in the 1950s and reported experiments showing that the AAA pretreatment used in the first 2 series of radiocarbon datings of the DSS (Bonani et al. [1992] and Jull et al. [1995]), “cannot be guaranteed to have removed all of the modern carbon in any samples if they had been contaminated with castor oil and hence could have produced some 14C dates that were younger than the texts' true ages.” Carmi, a coauthor of the Bonani et al. (1992) study, criticizes our analysis on 4 grounds:

It is common practice to estimate the age of undated material extracted from a sediment core from radiocarbon or other radiometric dates of samples taken above and below the extracted material. This paper presents a simple expression for the variance of this estimated age. This variance accounts for both 14C dating error and error due to bioturbation.