K-Ar ages of Eocene glauconite pellets from the Isle of Wight are related to quantified amounts of older glauconite pellets and to occurrences of detrital mica/illite particles that might have been added to synsedimentary pellets during reworking processes. Addition of older glauconite did not significantly bias the K-Ar dates, as the results most often provide the expected stratigraphic reference ages or even significantly lower ages. Alternatively, K-Ar dates significantly greater than those suggested by the stratigraphy appear to result from non-glauconized detrital mica in the pellets which is not always removable, even by high-gradient magnetic separation.

Unexpected ‘old’ K-Ar glauconite ages do not result from misapplication of the method, but from inability to systematically identify and remove the `contaminant' particles from glauconite splits. Analysis of highly evolved glauconite separates is definitely appropriate for reliable isotopic age determinations, but it might not be enough for the final selection, as long-lasting diagenesis might have taken over the synsedimentary process. Much if not all depends on the separation and characterization of the separates, but also on the completion of the glauconitization process which effects cannot always be anticipated.