This report extends the work of Corrigan and Gibb (1979) who determined losses of Fe and Na from a typical basaltic liquid during 1-atmosphere experiments using the wire-loop method (utilizing both Pt and Ag40Pd60 wires) by electronmicroprobe analyses of the quenched glasses. They found that Na losses were much more serious than Fe losses, the former being in direct linear proportion to the length of the experiment and in inverse proportion to the mass of the sample, such that the Na loss was 25 % after 120 hours for samples of 60 mg. Even using samples with weights in excess of 200 mg would mean that Na losses would be unlikely to be less than 5 % except in experiments shorter than a few days or on samples with very low initial Na contents.