Abstract
Examination of desert varnish samples from Australia by electron microprobe and energy dispersive spectrometry has shown that gross layering of iron‐ and manganese‐enrichments is variable. Surface (upper) varnish may be similar to lower varnish, or be relatively enriched in either iron or manganese. When viewed in sections cut parallel to the varnished surface, microlaminations with high manganese content (up to 30 per cent) were sometimes found in concentrically zoned patterns around grains.Dorn et al. (1986) have proposed that chemical changes occur with increasing varnish age. In this study, the possibility of such changes was examined at the within‐varnish scale by comparing near‐surface and deeper (upper and lower) portions of the same varnishes. Variations in elemental ratios (K + Ca:Ti) at this scale were shown to be largely unsystematic, suggesting either that samples were unrepresentative; that element inputs to varnish have varied over time; that the varnish was too young for differentiation by weathering into an upper and lower layer; that varnish cover formed rapidly and at one time; or that existing varnish cover represents the coalescence of numerous individual patches of varnish, each of which formed over a relatively short period.