ABSTRACT
Recent criticism of sand grain-size analysis as a useful tool in sedimentary petrology is directed at popular techniques of data analysis that do not rest on a correct theoretical basis. Several decades ago, R. A. Bagnold pioneered a new approach that deserves attention before grain-size analysis is discarded for environmental discrimination. Bagnold wrote that transportational sorting and winnowing are independent processes involved in sediment deposition. As a result, grain-size distributions of sand deposits follow log-hyperbolic and not log-normal distributions, and the coarse and fine tails of grain distributions are independent. More recently, D. Love showed that grain-size studies assuming log normality are prone to specific errors, one of which is the false inference of two lognormal populations when a log hyperbolic-population is analyzed. A difficulty, up to now, in the practical application of Bagnold’s approach has been the requirement that sieve-opening diameters be measured accurately. This problem can be eliminated by the use of sedimentation tower techniques. We are using a 2-m tall, 40-cm wide, water-filled sedimentation tower and a sensitive electronic balance coupled to a microprocessor to collect sediment-weight and settling-time data for sand samples. Bagnold’s log differential plots can be plotted directly from such data by the computer and offer the potential to interpret sand grain-size distributions on a sound theoretical basis.