Fig. 1.
Fig. 1.

Graph of Hg flux from representative 17 g sand (open diamond) and soil (closed diamond) samples. No difference in flux was noted in the dark; however, an immediate and dramatic increase in flux was seen on exposure to light (time = 0), and flux slowly decreased with continued exposure to light.

 


Fig. 2.
Fig. 2.

Mean Hg flux ± 95% CIs from sand samples (ng h−1 m−2) versus thickness of sample in millimeters. Each point represents the statistical average of Hg emissions collected during a sampling interval of 5.5 d in which the sample was exposed to full-spectrum light; 95% confidence intervals are shown as error bars. Replicate trials were run for most samples. Flux increased proportional to thickness up to ∼2 mm but showed no significantly significant difference (α = 0.05) above this depth.

 


Fig. 3.
Fig. 3.

Mean Hg flux ± 95% CI from soil samples (ng h−1 m−2) versus thickness of sample (mm). Each point represents the statistical average of Hg emissions collected during a sampling interval of 5.5 d in which the sample was exposed to full-spectrum light; 95% CIs are shown as error bars. Replicate trials were run for many samples. The thickest samples have slightly higher average fluxes; however, the R 2 value for the goodness of fit for a trend line with thickness is quite low (0.21).

 


Fig. 4.
Fig. 4.

Emission half-times calculated for sand samples using experimentally determined rate constants.

 


Fig. 5.
Fig. 5.

Mean Hg flux ± 95% CI from sand samples (ng h−1 m−2) during their final 24 h of exposure to light. Mean ± 95% CIs of sand flux from sand in dark shown at bottom.

 


Fig. 6.
Fig. 6.

Emission half-times calculated for soil samples using experimentally determined rate constants.

 


Fig. 7.
Fig. 7.

Mean Hg flux ± 95% CIs from soil samples (ng h−1 m−2) during their final 24 h of exposure to light. Mean ± 95% CIs of soil flux from soil in dark shown at bottom.