N. F. Hughes has shown how the processes of stratigraphic correlation (using miospores) can be broken down into discrete components- biorecords, comparison records, events, and correlation brackets. Critical attention has been directed almost exclusively at the first of these. This paper concentrates on the latter three in the context of a miospore-based correlation study of Late Triassic strata in Svalbard. No biorecords are described, but traditional identification is avoided in favour of graded comparisons with published records. 'Events' can be qualitatively correlated or, preferably, seriated to give acomposite sequence to taxon range limits. Bracket correlations become unwieldy for inter-regional correlation, and comparison with a probabilistically derived global composite is more easily depicted by the graphic correlation method. It is believed that Hughes's challenging standards of scientific integrity can be maintained through selective use of existing published data if nomenclatural considerations are set aside, at least for the purpose of correlation.