I think the replies posted by Andrea and Luis deserve a "public" response and at the risk of alienating all list members I'd like to ask two questions. These questions apply equally to each of the three common methods of treatment (roasting, chloroxing, plasma ashing) and are intended for all people who work with this problem (not just Andrea and Luis). For brevity I'm going to phrase the questions for the roasting method, but you can substitute any of the other methods. Q1. If roasting at 200oC is the right method, how do you view data produced from samples that have been chloroxed or plasma ashed? Are all methods equally valid, or is yours the only true path to enlightenment? Q2. How do you know that your roasted samples yield the true isotopic composition? Are you saying that that you analysed untreated and roasted sub-samples of the same material and obtained identical isotopic values? Or are you saying that untreated and treated sub-samples yielded different isotopic values and that your treated samples yielded isotopic values more consistent with your expectations? If you answer yes to the first question, then treatments are unnecessary. If the answer to the second question is also yes, then how do you know that your expectations are correct? The implication of question 2 is: if you start with a sample of unknown isotopic composition that contains organic matter, structurally-bound water etc. you can either analyse it untreated, in which case its measured composition might be compromised by the presence of the contaminants, or you can treat it, in which case its measured composition might be compromised by the treatment. I don't quite see how this circle can be broken without knowing the true composition to begin with. Since the samples are by definition of unknown composition it seems difficult to adjudicate on the veracity of the resultant data in either instance. I would argue that attempting to break this circle by experiment using artificial mixtures would not be a true representation of real samples, and that different results may be obtained from different samples using different methods. I promise not to post any more messages on this subject!!! Cheers, Steve Department of Earth & Ocean Sciences University of Liverpool 4 Brownlow Street Liverpool L69 3GP UK 0151 794 5163/5164 **************************