ABSTRACT

The technique of the graphical dissection of a heterogeneous distribution may be used to show that almost all grain size frequency distributions are bimodal. This consideration together with an investigation of the truncation effects due to sorting leads to the following two hypotheses concerning clastic sediments.

Hypothesis I.--"All clastic sediments are essentially mixtures of three or less fundamental populations of log-normal grain sizes. The fundamental populations are:

(a) "Gravel" with a median of -3.5 to -2 units and a standard deviation of 1.0 to 2.0 units.

(b) "Sand" with a median of 1.5 to 4 units and a standard deviation of 0.4 to 1.0 units.

(c) "Clay" with a median of 7 to 9 units and a standard deviation of 2 to 3 units."

Hypothesis II.--"Sorting in clastic sediments may be recognized only by the degree of truncation of the original fundamental populations or mixtures of these populations."

These two hypotheses lead to the conclusion that most clastic sediments consist of "grains" and "matrix" and the clue to effective porosity development lies in the dispersion of the "matrix."