Page 9 - Petrology of Sedimentary Rocks

PROPERTIES OF SEDIMENTARY ROCKS

Grain Size

Quantitative measurement of such grain size parameters as size and sorting is
required for precise work. To measure grain size, one must first choose a scale.
Inasmuch as nature apparently favors ratio scales over arithmetic scales, the grain size
scale, devised by Udden, is based on a constant ratio of 2 between successive classes;
names for the class intervals were modified by Wentworth. Krumbein devised the phi
(41) scale as a logarithmic transformation of the Wentworth scale, and modern data is
nearly always stated in Q terms because mathematical computations are much simpli-
fied.

Grain size of particles larger than several centimeters is us,Jally determined by
direct measurement with calipers or meter sticks; particles down to about 4$
(0.062 mm) are analyzed by screening; and silts and clays (fine than 44) are analyzed by
pipette or hydrometer, utilizing differential settling rates in water. Sands can also be
measured by petrographic microscope or by various settling devices.

Results of grain-size analysis may be plotted as histograms, cumulative curves or
frequency curves. Curve data is summarized by means of mathematical parameters
allowing ready comparison between samples. As measures of average size the median,
mode, and mean are frequently used; as a measure of sorting the standard deviation is
best. Skewness and kurtosis are also useful parameters for some work.

Geological significance of the measures is not fully known. Many generalizations
have been made on far too little evidence, but significant studies are now going on. As
of our present state of knowledge, we can make a few educated guesses about the
meaning of grain-size statistics.
The significance of mean grain size is not yet well enough known to make any
positive statements; volumes of data on recent sediments must be collected before we
can say anything really meaingful. To be sure there is a certain correlation of grain
size with environments (see page l07)--e.g. you usually do not find conglomerates in
swamps or silts on beaches-- but there is a great deal of overlapping. These questions
can only be solved by integration of size analysis results with shape study, detailed field
mapping, study of sedimentary structures, fossils, thickness pattern changes, etc. It is
extremely risky to postulate anything based on size analysis from one sample or one
thin section, unless that sample is known to be representative of a large thickness of
section. One thing that is an especially common error is the idea that if a sediment is
fine it must be far from the source, while if it is coarse it must be near to a source.
Many studies of individual environments show sediments getting finer away from the
source but these changes are so varied that they can be deciphered only by extensive
field and laboratory work. For example, strong rivers may carry large pebbles a
thousand or so miles away from their source, while on the other hand the finest clays
and silts are found in playa lakes a matter of a few miles from encircling rugged
mountains. Grain size depends largely on the current strength of the local environment
(together with size of available particles), not on distance. Flood plainlays may lie
immediately adjacent to course fluvial gravels, both being the very same distance from
their source. One must work up a large number of samples before anything much can be

3