SUMMARY. 1. An experimental study is described, designed to test the hypothesis that the loss of particles from suspension in flowing water follows an exponential decay function, the exponent of which is influenced more by water depth than water velocity. Successive experiments employed suspensions of Lycopodium spores which were introduced into one of the FBA's circulating channels maintained at its Watcrston site, Dorset, under different combinations of water depth and pumping rate.
2. In each experiment, the concentration of introduced spores indeed declined exponentially through time. The bulked, transformed data-set also conformed well to a single regression against a common time scale and which explained over 94% of the accumulated variances.
3. The variance unexplained by the regression was apportioned among components distinguishing between experiments, experimental differences in starting concentration, water depth and pumping rate (velocity). This analysis revealed that, after elimination of different initial concentrations, only water depth produced a significant effect, through its relation to the settling velocity of the Lycopodium. Thus, the hypothesis was not invalidated: water depth and not Row velocity proved to be the main controlling variable determining the rale of sinking loss in these experiments. Flow velocity is, nevertheless, an important component only in the sense that it influences the horizontal distance travelled by the residual spore suspension through the time period required for complete settlement.
4. The consequences of experimental findings are applied to the maintenance of planktonic diatoms in rivers: both the survival of a potential growth inoculum and its downstream dispersion are strongly time-dependent and are enhanced by greater channel depths.