Page 18 - Petrology of Sedimentary Rocks

As a rule, it is difficult to study surface features in ancient sediments because of
cementation. An attempt may be made by cleaning the grains thoroughly in warm HCI
to remove carbonates and iron stains. They should be studied under highest power of
the binocular microscope, on a black surface and under strong light. After this, the
grain should be mounted in water and examined with petrographic microscope. Electron
microscopy has made great contributions to study of grain surface features (Krinsley),
and intimate details of grain history can be studied.
Graphic and Statistical Analysis of Shape Data. Graphic or statistical analysis is
necessary to show (I) the variation 2 grain morphology with size, which is nearly
always present, and (2) environmental, spatial, or stratigraphic differences between
samples. In comparing a set of samples, one should choose the same size interval
throughout for his analyses, because coarser grains are usually better rounded, and show
different surface features. However on some of the samples, shape study should be
made on all sizes. This is called an H pattern of sampling.

If a significant difference in morphology is found, (for example if beach sands in
an area are more angular than dune sands), you must always consider whether it is due
to (I) an actual difference in the processes going on (that the dunes are rounding more
effectively) or whether (2) it is simply the result of selective sorting, where e.g.
rounded and more spherical grains are left behind by wind currents which selectively
sweep away the more angular and less spherical grains.

To determine if a difference in Form is present between two sets of samples, one
can (I) using a moving circular mask, contour the points on the triangular diagram as is
done in contouring joint or petrofabric diagrams; (2) use the X2 test by counting the
number of particles in each shape “cell” and comparing the results; (3) obtain by
counting a median shape for each set of data; (4) superimpose a sheet of transparent
triangular-ruled graph paper over the data, and assign each point a “percent elongation”
and a “percent platiness” by measuring its distance from the base line of the triangle;
this data may be treated by finding its mean and standard deviation and comparing by
the t test.

Sphericity differences between samples can be evaluated by finding a mean and
standard deviation for each set of particles and comparing by means of the t test.
Roundness is analyzed by counting a large number of particles using comparison
charts; then, use the log transformations of the roundness values (p scale) and compute
means (average roundness) and standard deviations (roundness sorting), then compare by
the t test. The roundness data may also be plotted by means of cumulative curves if
probability paper is used. In doing a set of samples from two or more environments,
formations, or localities, it is well to have an associate mix the samples up so that you
do not know which set you are working on when you count a sample; thus you avoid a
bias and the data is more “honest.”

Surface features may be compared by counting the number of frosted, polished
grains etc. and comparing by the X2 test. Again, an associate should mix the samples
up so you do not know which sets you are counting, in order to avoid bias.

Significance of Grain Morphology

Form and Sphericity are the result of (I) Structure (internal properties, inherited
from the source; (2) Process (work of the depositional environment e.g. glacier, river or

I2