Graphical methods for the determination of standard deviations of grain-size distributions have been proposed by Otto (1939) and Inman (1952). A mathematical test based on Students' "t" distribution is suggested for comparing computed standard deviation or other parameters of a grain-size distribution with those determined by graphical short-cuts. Differences between computed standard deviations and those determined by lnman's graphic method show a probability of less than 0.001 of occurring in a normal population. A slight modification of Inman's formula raises the probability to less than 0.05 but greater than 0.02.

You do not currently have access to this article.